From 772a896a56e0e3ef9424a025cec9176f9d8f4552 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Sat, 11 Apr 2026 14:06:00 +0200 Subject: [PATCH 001/525] scsi: target: configfs: Bound snprintf() return in tg_pt_gp_members_show() target_tg_pt_gp_members_show() formats LUN paths with snprintf() into a 256-byte stack buffer, then will memcpy() cur_len bytes from that buffer. snprintf() returns the length the output would have had, which can exceed the buffer size when the fabric WWN is long because iSCSI IQN names can be up to 223 bytes. The check at the memcpy() site only guards the destination page write, not the source read, so memcpy() will read past the stack buffer and copy adjacent stack contents to the sysfs reader, which when CONFIG_FORTIFY_SOURCE is enabled, fortify_panic() will be triggered. Commit 27e06650a5ea ("scsi: target: target_core_configfs: Add length check to avoid buffer overflow") added the same bound to the target_lu_gp_members_show() but the tg_pt_gp variant was missed so resolve that here. Cc: Martin K. Petersen Fixes: c66ac9db8d4a ("[SCSI] target: Add LIO target core v4.0.0-rc6") Assisted-by: gregkh_clanker_t1000 Signed-off-by: Greg Kroah-Hartman Link: https://patch.msgid.link/2026041159-garter-theft-3be0@gregkh Signed-off-by: Martin K. Petersen --- drivers/target/target_core_configfs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c index d93773b3227c..2b19a956007b 100644 --- a/drivers/target/target_core_configfs.c +++ b/drivers/target/target_core_configfs.c @@ -3249,7 +3249,7 @@ static ssize_t target_tg_pt_gp_members_show(struct config_item *item, config_item_name(&lun->lun_group.cg_item)); cur_len++; /* Extra byte for NULL terminator */ - if ((cur_len + len) > PAGE_SIZE) { + if (cur_len > TG_PT_GROUP_NAME_BUF || (cur_len + len) > PAGE_SIZE) { pr_warn("Ran out of lu_gp_show_attr" "_members buffer\n"); break; From bc0fcb9823cd0894934cf968b525c575833d7078 Mon Sep 17 00:00:00 2001 From: Yilin Zhu Date: Sun, 12 Apr 2026 13:07:54 +0800 Subject: [PATCH 002/525] ipv6: xfrm6: release dst on error in xfrm6_rcv_encap() xfrm6_rcv_encap() performs an IPv6 route lookup when the skb does not already have a dst attached. ip6_route_input_lookup() returns a referenced dst entry even when the lookup resolves to an error route. If dst->error is set, xfrm6_rcv_encap() drops the skb without attaching the dst to the skb and without releasing the reference returned by the lookup. Repeated packets hitting this path therefore leak dst entries. Release the dst before jumping to the drop path. Fixes: 0146dca70b87 ("xfrm: add support for UDPv6 encapsulation of ESP") Cc: stable@kernel.org Reported-by: Yifan Wu Reported-by: Juefei Pu Co-developed-by: Yuan Tan Signed-off-by: Yuan Tan Suggested-by: Xin Liu Tested-by: Ruide Cao Signed-off-by: Yilin Zhu Signed-off-by: Ren Wei Reviewed-by: Simon Horman Signed-off-by: Steffen Klassert --- net/ipv6/xfrm6_protocol.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/ipv6/xfrm6_protocol.c b/net/ipv6/xfrm6_protocol.c index ea2f805d3b01..9b586fcec485 100644 --- a/net/ipv6/xfrm6_protocol.c +++ b/net/ipv6/xfrm6_protocol.c @@ -88,8 +88,10 @@ int xfrm6_rcv_encap(struct sk_buff *skb, int nexthdr, __be32 spi, dst = ip6_route_input_lookup(dev_net(skb->dev), skb->dev, &fl6, skb, flags); - if (dst->error) + if (dst->error) { + dst_release(dst); goto drop; + } skb_dst_set(skb, dst); } From ec54093e6a8f87e800bb6aa15eb7fc1e33faa524 Mon Sep 17 00:00:00 2001 From: Michael Bommarito Date: Sun, 19 Apr 2026 18:35:42 -0400 Subject: [PATCH 003/525] xfrm: ah: account for ESN high bits in async callbacks AH allocates its temporary auth/ICV layout differently when ESN is enabled: the async ahash setup appends a 4-byte seqhi slot before the ICV or auth_data area, but the async completion callbacks still reconstruct the temporary layout as if seqhi were absent. With an async AH implementation selected, that makes AH copy or compare the wrong bytes on both the IPv4 and IPv6 paths. In UML repro on IPv4 AH with ESN and forced async hmac(sha1), ping fails with 100% packet loss, and the callback logs show the pre-fix drift: ah4 output_done: esn=1 err=0 icv_off=20 expected_off=24 ah4 input_done: esn=1 auth_off=20 expected_auth_off=24 icv_off=32 expected_icv_off=36 Reconstruct the callback-side layout the same way the setup path built it by skipping the ESN seqhi slot before locating the saved auth_data or ICV. Per RFC 4302, the ESN high-order 32 bits participate in the AH ICV computation, so the async callbacks must account for the seqhi slot. Post-fix, the same IPv4 AH+ESN+forced-async-hmac(sha1) UML repro shows the corrected offset (ah4 output_done: esn=1 err=0 icv_off=24 expected_off=24) and ping succeeds; net/ipv4/ah4.o and net/ipv6/ah6.o build clean at W=1. IPv6 AH+ESN was not exercised at runtime, and the change has not been tested against a real async hardware AH engine. Fixes: d4d573d0334d ("{IPv4,xfrm} Add ESN support for AH egress part") Fixes: d8b2a8600b0e ("{IPv4,xfrm} Add ESN support for AH ingress part") Fixes: 26dd70c3fad3 ("{IPv6,xfrm} Add ESN support for AH egress part") Fixes: 8d6da6f32557 ("{IPv6,xfrm} Add ESN support for AH ingress part") Cc: stable@vger.kernel.org Assisted-by: Codex:gpt-5-4 Assisted-by: Claude:claude-opus-4-7 Signed-off-by: Michael Bommarito Signed-off-by: Steffen Klassert --- net/ipv4/ah4.c | 14 ++++++++++++-- net/ipv6/ah6.c | 14 ++++++++++++-- 2 files changed, 24 insertions(+), 4 deletions(-) diff --git a/net/ipv4/ah4.c b/net/ipv4/ah4.c index 5fb812443a08..4366cbac3f06 100644 --- a/net/ipv4/ah4.c +++ b/net/ipv4/ah4.c @@ -124,9 +124,14 @@ static void ah_output_done(void *data, int err) struct iphdr *top_iph = ip_hdr(skb); struct ip_auth_hdr *ah = ip_auth_hdr(skb); int ihl = ip_hdrlen(skb); + int seqhi_len = 0; + __be32 *seqhi; + if (x->props.flags & XFRM_STATE_ESN) + seqhi_len = sizeof(*seqhi); iph = AH_SKB_CB(skb)->tmp; - icv = ah_tmp_icv(iph, ihl); + seqhi = (__be32 *)((char *)iph + ihl); + icv = ah_tmp_icv(seqhi, seqhi_len); memcpy(ah->auth_data, icv, ahp->icv_trunc_len); top_iph->tos = iph->tos; @@ -270,12 +275,17 @@ static void ah_input_done(void *data, int err) struct ip_auth_hdr *ah = ip_auth_hdr(skb); int ihl = ip_hdrlen(skb); int ah_hlen = (ah->hdrlen + 2) << 2; + int seqhi_len = 0; + __be32 *seqhi; if (err) goto out; + if (x->props.flags & XFRM_STATE_ESN) + seqhi_len = sizeof(*seqhi); work_iph = AH_SKB_CB(skb)->tmp; - auth_data = ah_tmp_auth(work_iph, ihl); + seqhi = (__be32 *)((char *)work_iph + ihl); + auth_data = ah_tmp_auth(seqhi, seqhi_len); icv = ah_tmp_icv(auth_data, ahp->icv_trunc_len); err = crypto_memneq(icv, auth_data, ahp->icv_trunc_len) ? -EBADMSG : 0; diff --git a/net/ipv6/ah6.c b/net/ipv6/ah6.c index cb26beea4398..de1e68199a01 100644 --- a/net/ipv6/ah6.c +++ b/net/ipv6/ah6.c @@ -317,14 +317,19 @@ static void ah6_output_done(void *data, int err) struct ipv6hdr *top_iph = ipv6_hdr(skb); struct ip_auth_hdr *ah = ip_auth_hdr(skb); struct tmp_ext *iph_ext; + int seqhi_len = 0; + __be32 *seqhi; extlen = skb_network_header_len(skb) - sizeof(struct ipv6hdr); if (extlen) extlen += sizeof(*iph_ext); + if (x->props.flags & XFRM_STATE_ESN) + seqhi_len = sizeof(*seqhi); iph_base = AH_SKB_CB(skb)->tmp; iph_ext = ah_tmp_ext(iph_base); - icv = ah_tmp_icv(iph_ext, extlen); + seqhi = (__be32 *)((char *)iph_ext + extlen); + icv = ah_tmp_icv(seqhi, seqhi_len); memcpy(ah->auth_data, icv, ahp->icv_trunc_len); memcpy(top_iph, iph_base, IPV6HDR_BASELEN); @@ -471,13 +476,18 @@ static void ah6_input_done(void *data, int err) struct ip_auth_hdr *ah = ip_auth_hdr(skb); int hdr_len = skb_network_header_len(skb); int ah_hlen = ipv6_authlen(ah); + int seqhi_len = 0; + __be32 *seqhi; if (err) goto out; + if (x->props.flags & XFRM_STATE_ESN) + seqhi_len = sizeof(*seqhi); work_iph = AH_SKB_CB(skb)->tmp; auth_data = ah_tmp_auth(work_iph, hdr_len); - icv = ah_tmp_icv(auth_data, ahp->icv_trunc_len); + seqhi = (__be32 *)(auth_data + ahp->icv_trunc_len); + icv = ah_tmp_icv(seqhi, seqhi_len); err = crypto_memneq(icv, auth_data, ahp->icv_trunc_len) ? -EBADMSG : 0; if (err) From 7b03c93d2beb91c6abae322a1f25447b5b3bb9e6 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 15 Apr 2026 08:08:06 +0200 Subject: [PATCH 004/525] scsi: sg: Don't use GFP_ATOMIC in sg_start_req() sg_start_req() is called from normal user context and can sleep when waiting for memory. Switch it to use GFP_KERNEL, which fixes allocation failures seen with the bio_alloc rework. Fixes: b520c4eef83d ("block: split bio_alloc_bioset more clearly into a fast and slowpath") Reported-by: Shin'ichiro Kawasaki Signed-off-by: Christoph Hellwig Tested-by: Shin'ichiro Kawasaki Reviewed-by: John Garry Reviewed-by: Hannes Reinecke Reviewed-by: Bart Van Assche Link: https://patch.msgid.link/20260415060813.807659-2-hch@lst.de Signed-off-by: Martin K. Petersen --- drivers/scsi/sg.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index 2b4b2a1a8e44..74cd4e8a61c2 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -1801,7 +1801,7 @@ sg_start_req(Sg_request *srp, unsigned char *cmd) } res = blk_rq_map_user_io(rq, md, hp->dxferp, hp->dxfer_len, - GFP_ATOMIC, iov_count, iov_count, 1, rw); + GFP_KERNEL, iov_count, iov_count, 1, rw); if (!res) { srp->bio = rq->bio; From 04631f55afc543d5431a2bdee7f6cc0f2c0debe7 Mon Sep 17 00:00:00 2001 From: Ranjan Kumar Date: Tue, 14 Apr 2026 16:38:11 +0530 Subject: [PATCH 005/525] scsi: mpt3sas: Limit NVMe request size to 2 MiB The HBA firmware reports NVMe MDTS values based on the underlying drive capability. However, because the driver allocates a fixed 4K buffer for the PRP list, accommodating at most 512 entries, the driver supports a maximum I/O transfer size of 2 MiB. Limit max_hw_sectors to the smaller of the reported MDTS and the 2 MiB driver limit to prevent issuing oversized I/O that may lead to a kernel oops. Cc: stable@vger.kernel.org Fixes: 9b8b84879d4a ("block: Increase BLK_DEF_MAX_SECTORS_CAP") Reported-by: Mira Limbeck Closes: https://lore.kernel.org/r/291f78bf-4b4a-40dd-867d-053b36c564b3@proxmox.com Link: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=9b8b84879d4a Suggested-by: Keith Busch Signed-off-by: Ranjan Kumar Tested-by: Mira Limbeck Link: https://patch.msgid.link/20260414110811.85156-1-ranjan.kumar@broadcom.com Signed-off-by: Martin K. Petersen --- drivers/scsi/mpt3sas/mpt3sas_scsih.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/mpt3sas/mpt3sas_scsih.c b/drivers/scsi/mpt3sas/mpt3sas_scsih.c index 6ff788557294..12caffeed3a0 100644 --- a/drivers/scsi/mpt3sas/mpt3sas_scsih.c +++ b/drivers/scsi/mpt3sas/mpt3sas_scsih.c @@ -2738,8 +2738,20 @@ scsih_sdev_configure(struct scsi_device *sdev, struct queue_limits *lim) pcie_device->enclosure_level, pcie_device->connector_name); + /* + * The HBA firmware passes the NVMe drive's MDTS + * (Maximum Data Transfer Size) up to the driver. However, + * the driver hardcodes a 4K buffer size for the PRP list, + * accommodating at most 512 entries. This strictly limits + * the maximum supported NVMe I/O transfer to 2 MiB. + * + * Cap max_hw_sectors to the smaller of the drive's reported + * MDTS or the 2 MiB driver limit to prevent kernel oopses. + */ + lim->max_hw_sectors = SZ_2M >> SECTOR_SHIFT; if (pcie_device->nvme_mdts) - lim->max_hw_sectors = pcie_device->nvme_mdts / 512; + lim->max_hw_sectors = min(lim->max_hw_sectors, + pcie_device->nvme_mdts >> SECTOR_SHIFT); pcie_device_put(pcie_device); spin_unlock_irqrestore(&ioc->pcie_device_lock, flags); From d65efdf467ff935e35dfe6aa9a7ab93f17ac07ee Mon Sep 17 00:00:00 2001 From: Tomas Henzl Date: Tue, 14 Apr 2026 14:41:18 +0200 Subject: [PATCH 006/525] scsi: smartpqi: Silence a recursive lock warning On systems with multiple controllers debug kernel shows WARNING: possible recursive locking detected during shutdown. Each controller does have its own ctrl_info (and mutex) and that isn't correctly recognized by debug kernel. Suppress the warning by releasing the mutex at the end of pqi_shutdown(). Signed-off-by: Tomas Henzl Acked-by: Don Brace Link: https://patch.msgid.link/20260414124118.23661-1-thenzl@redhat.com Signed-off-by: Martin K. Petersen --- drivers/scsi/smartpqi/smartpqi_init.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/scsi/smartpqi/smartpqi_init.c b/drivers/scsi/smartpqi/smartpqi_init.c index b4ed991976d0..2026ac645d6a 100644 --- a/drivers/scsi/smartpqi/smartpqi_init.c +++ b/drivers/scsi/smartpqi/smartpqi_init.c @@ -9427,6 +9427,7 @@ static void pqi_shutdown(struct pci_dev *pci_dev) pqi_crash_if_pending_command(ctrl_info); pqi_reset(ctrl_info); + pqi_ctrl_unblock_device_reset(ctrl_info); } static void pqi_process_lockup_action_param(void) From 68c3a65a5a8e85643745fdde02cb63904e165620 Mon Sep 17 00:00:00 2001 From: Brian Bunker Date: Thu, 16 Apr 2026 09:55:12 -0700 Subject: [PATCH 007/525] scsi: scsi_dh_alua: Increase default ALUA timeout to maximum spec value The ALUA handler maps a 0 value (no implicit transition timeout provided by the target) to the ALUA_FAILOVER_TIMEOUT constant, currently 60 seconds. This means the kernel already does not accept an infinite transition time. However, 60 seconds is insufficient for some arrays that may take longer to complete ALUA transitions. Since the highest value allowed by the SCSI specification for the implicit transition timeout is a single byte (255 seconds), change the default to 255. This way, when a target does not provide an explicit transition timeout, we default to the maximum value the spec allows rather than an arbitrary 60 second limit. Co-developed-by: Krishna Kant Signed-off-by: Krishna Kant Co-developed-by: Riya Savla Signed-off-by: Riya Savla Signed-off-by: Brian Bunker Reviewed-by: Hannes Reinecke Link: https://patch.msgid.link/20260416165512.26497-2-brian@purestorage.com Signed-off-by: Martin K. Petersen --- drivers/scsi/device_handler/scsi_dh_alua.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/device_handler/scsi_dh_alua.c b/drivers/scsi/device_handler/scsi_dh_alua.c index efb08b9b145a..80ab0ff921d4 100644 --- a/drivers/scsi/device_handler/scsi_dh_alua.c +++ b/drivers/scsi/device_handler/scsi_dh_alua.c @@ -37,7 +37,7 @@ #define TPGS_MODE_EXPLICIT 0x2 #define ALUA_RTPG_SIZE 128 -#define ALUA_FAILOVER_TIMEOUT 60 +#define ALUA_FAILOVER_TIMEOUT 255 /* max 255 (8-bit value) */ #define ALUA_FAILOVER_RETRIES 5 #define ALUA_RTPG_DELAY_MSECS 5 #define ALUA_RTPG_RETRY_DELAY 2 From 1dc39ed655750d6c679d3ada4adf4a937f2a63fc Mon Sep 17 00:00:00 2001 From: Hugo Villeneuve Date: Fri, 17 Apr 2026 16:07:31 -0400 Subject: [PATCH 008/525] scsi: pmcraid: Fix typo in comments Fix typo in structure comment. Signed-off-by: Hugo Villeneuve Link: https://patch.msgid.link/20260417200738.3920001-1-hugo@hugovil.com Signed-off-by: Martin K. Petersen --- drivers/scsi/pmcraid.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/pmcraid.h b/drivers/scsi/pmcraid.h index 9f59930e8b4f..cd059b7599b4 100644 --- a/drivers/scsi/pmcraid.h +++ b/drivers/scsi/pmcraid.h @@ -657,7 +657,7 @@ struct pmcraid_hostrcb { */ struct pmcraid_instance { /* Array of allowed-to-be-exposed resources, initialized from - * Configutation Table, later updated with CCNs + * Configuration Table, later updated with CCNs */ struct pmcraid_resource_entry *res_entries; From 47e66bec3edaebd7c52d8ee981065a4c83b3072f Mon Sep 17 00:00:00 2001 From: Yihang Li Date: Mon, 20 Apr 2026 10:10:44 +0800 Subject: [PATCH 009/525] scsi: hisi_sas: Fix sparse warnings in prep_ata_v3_hw() In prep_ata_v3_hw(), add cpu_to_le32() to fix warning: drivers/scsi/hisi_sas/hisi_sas_v3_hw.c:1448:26: sparse: sparse: invalid assignment: |= drivers/scsi/hisi_sas/hisi_sas_v3_hw.c:1448:26: sparse: left side has type restricted __le32 drivers/scsi/hisi_sas/hisi_sas_v3_hw.c:1448:26: sparse: right side has type unsigned int Fixes: 8aa580cd9284 ("scsi: hisi_sas: Enable force phy when SATA disk directly connected") Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202604191850.IVYPTaML-lkp@intel.com/ Signed-off-by: Yihang Li Link: https://patch.msgid.link/20260420021044.3339459-1-liyihang9@huawei.com Signed-off-by: Martin K. Petersen --- drivers/scsi/hisi_sas/hisi_sas_v3_hw.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c index fda07b193137..14d563e82d20 100644 --- a/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c +++ b/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c @@ -1491,7 +1491,7 @@ static void prep_ata_v3_hw(struct hisi_hba *hisi_hba, phy_id = device->phy->identify.phy_identifier; hdr->dw0 |= cpu_to_le32((1U << phy_id) << CMD_HDR_PHY_ID_OFF); - hdr->dw0 |= CMD_HDR_FORCE_PHY_MSK; + hdr->dw0 |= cpu_to_le32(CMD_HDR_FORCE_PHY_MSK); hdr->dw0 |= cpu_to_le32(4U << CMD_HDR_CMD_OFF); } From 36920f30e78e69df01f9691c470b6f3ba8aebf98 Mon Sep 17 00:00:00 2001 From: Corey Minyard Date: Mon, 20 Apr 2026 12:50:09 -0500 Subject: [PATCH 010/525] ipmi: Check event message buffer response for bad data The event message buffer response data size got checked later when processing, but check it right after the response comes back. It appears some BMCs may return an empty message instead of an error when fetching events. There are apparently some new BMCs that make this error, so we need to compensate. Reported-by: Matt Fleming Closes: https://lore.kernel.org/lkml/20260415115930.3428942-1-matt@readmodwrite.com/ Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: Signed-off-by: Corey Minyard --- drivers/char/ipmi/ipmi_si_intf.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 4a9e9de4d684..08c208cc64c5 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -630,7 +630,13 @@ static void handle_transaction_done(struct smi_info *smi_info) */ msg = smi_info->curr_msg; smi_info->curr_msg = NULL; - if (msg->rsp[2] != 0) { + /* + * It appears some BMCs, with no event data, return no + * data in the message and not a 0x80 error as the + * spec says they should. Shut down processing if + * the data is not the right length. + */ + if (msg->rsp[2] != 0 || msg->rsp_size != 19) { /* Error getting event, probably done. */ msg->done(msg); From b06cf63d83d3b3744d3aefdd2f3ced25e99d7ec1 Mon Sep 17 00:00:00 2001 From: Wang Shuaiwei Date: Tue, 14 Apr 2026 11:37:18 +0800 Subject: [PATCH 011/525] scsi: ufs: core: Fix bRefClkFreq write failure in HS-LSS mode According to the UFS spec, the bRefClkFreq attribute can only be written when both sub-links are in LS-MODE. However, in HS LSS mode with resetmode = HS_MODE, if the UFS device's default bRefClkFreq value differs from the host controller's dev_ref_clk_freq setting, the write operation will fail. To fix this issue, introduce ufshcd_get_op_mode() function to detect the current link operational mode. Call ufshcd_set_dev_ref_clk() only when both sub-links are in LS-MODE to ensure the attribute can be written successfully. Signed-off-by: Wang Shuaiwei Link: https://patch.msgid.link/20260414033718.1459540-1-wangshuaiwei1@xiaomi.com Reviewed-by: Peter Wang Signed-off-by: Martin K. Petersen --- drivers/ufs/core/ufshcd.c | 30 ++++++++++++++++++++++++++++-- include/ufs/unipro.h | 5 +++++ 2 files changed, 33 insertions(+), 2 deletions(-) diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c index 4805e40ed4d7..c3f08957d179 100644 --- a/drivers/ufs/core/ufshcd.c +++ b/drivers/ufs/core/ufshcd.c @@ -9259,6 +9259,30 @@ static void ufshcd_config_mcq(struct ufs_hba *hba) hba->nutrs); } +/** + * ufshcd_get_op_mode - get UFS operating mode. + * @hba: per-adapter instance + * + * Use the PA_PWRMODE value to represent the operating mode of UFS. + * + */ +static enum ufs_op_mode ufshcd_get_op_mode(struct ufs_hba *hba) +{ + u32 mode; + u8 rx_mode; + u8 tx_mode; + + ufshcd_dme_get(hba, UIC_ARG_MIB(PA_PWRMODE), &mode); + rx_mode = (mode >> PWRMODE_RX_OFFSET) & PWRMODE_MASK; + tx_mode = mode & PWRMODE_MASK; + + if ((rx_mode == SLOW_MODE || rx_mode == SLOWAUTO_MODE) && + (tx_mode == SLOW_MODE || tx_mode == SLOWAUTO_MODE)) + return LS_MODE; + + return HS_MODE; +} + static int ufshcd_post_device_init(struct ufs_hba *hba) { int ret; @@ -9281,11 +9305,13 @@ static int ufshcd_post_device_init(struct ufs_hba *hba) return 0; /* - * Set the right value to bRefClkFreq before attempting to + * Set the right value to bRefClkFreq in LS_MODE before attempting to * switch to HS gears. */ - if (hba->dev_ref_clk_freq != REF_CLK_FREQ_INVAL) + if (ufshcd_get_op_mode(hba) == LS_MODE && + hba->dev_ref_clk_freq != REF_CLK_FREQ_INVAL) ufshcd_set_dev_ref_clk(hba); + /* Gear up to HS gear. */ ret = ufshcd_config_pwr_mode(hba, &hba->max_pwr_info.info, UFSHCD_PMC_POLICY_DONT_FORCE); diff --git a/include/ufs/unipro.h b/include/ufs/unipro.h index f849a2a101ae..9c168703b104 100644 --- a/include/ufs/unipro.h +++ b/include/ufs/unipro.h @@ -333,6 +333,11 @@ enum ufs_eom_eye_mask { #define DME_LocalTC0ReplayTimeOutVal 0xD042 #define DME_LocalAFC0ReqTimeOutVal 0xD043 +enum ufs_op_mode { + LS_MODE = 1, + HS_MODE = 2, +}; + /* PA power modes */ enum ufs_pa_pwr_mode { FAST_MODE = 1, From 2f3835771dff512750205aa5f5f61aec0f2b8cb7 Mon Sep 17 00:00:00 2001 From: Carlos Bilbao Date: Tue, 14 Apr 2026 21:07:28 -0700 Subject: [PATCH 012/525] scsi: target: iscsi: reject invalid size Extended CDB AHS If ecdb_ahdr->ahslength is zero, two bugs follow: kmalloc(be16_to_cpu(ecdb_ahdr->ahslength) + 15, ...) allocates 15 bytes, but the immediately following memcpy writes ISCSI_CDB_SIZE (16) bytes into it, a one-byte heap overflow. Also: memcpy(cdb + ISCSI_CDB_SIZE, ecdb_ahdr->ecdb, be16_to_cpu(ecdb_ahdr->ahslength) - 1); (u16)0 - 1 promotes to (int)-1 which converts to SIZE_MAX as size_t, causing a massive out-of-bounds write. Reject ahslength == 0 with ISCSI_REASON_PROTOCOL_ERROR before the kmalloc. Also reject ahslength values that exceed the actual AHS buffer advertised. Fixes: 8f1f7d297bce ("scsi: target: iscsi: Add support for extended CDB AHS") Signed-off-by: Carlos Bilbao Reviewed-by: Dmitry Bogdanov Link: https://patch.msgid.link/20260415040728.187680-1-carlos.bilbao@kernel.org Signed-off-by: Martin K. Petersen --- drivers/target/iscsi/iscsi_target.c | 22 ++++++++++++++++++---- 1 file changed, 18 insertions(+), 4 deletions(-) diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c index e80449f6ce15..cb832fd523af 100644 --- a/drivers/target/iscsi/iscsi_target.c +++ b/drivers/target/iscsi/iscsi_target.c @@ -995,6 +995,7 @@ int iscsit_setup_scsi_cmd(struct iscsit_conn *conn, struct iscsit_cmd *cmd, int data_direction, payload_length; struct iscsi_ecdb_ahdr *ecdb_ahdr; struct iscsi_scsi_req *hdr; + u16 ahslength, cdb_length; int iscsi_task_attr; unsigned char *cdb; int sam_task_attr; @@ -1108,14 +1109,27 @@ int iscsit_setup_scsi_cmd(struct iscsit_conn *conn, struct iscsit_cmd *cmd, ISCSI_REASON_CMD_NOT_SUPPORTED, buf); } - cdb = kmalloc(be16_to_cpu(ecdb_ahdr->ahslength) + 15, - GFP_KERNEL); + ahslength = be16_to_cpu(ecdb_ahdr->ahslength); + if (!ahslength) { + pr_err("Extended CDB AHS with zero length, protocol error.\n"); + return iscsit_add_reject_cmd(cmd, + ISCSI_REASON_PROTOCOL_ERROR, buf); + } + if (ahslength > (hdr->hlength * 4) - 3) { + pr_err("Extended CDB AHS length %u exceeds available PDU buffer.\n", + ahslength); + return iscsit_add_reject_cmd(cmd, + ISCSI_REASON_PROTOCOL_ERROR, buf); + } + + cdb_length = ahslength - 1 + ISCSI_CDB_SIZE; + + cdb = kmalloc(cdb_length, GFP_KERNEL); if (cdb == NULL) return iscsit_add_reject_cmd(cmd, ISCSI_REASON_BOOKMARK_NO_RESOURCES, buf); memcpy(cdb, hdr->cdb, ISCSI_CDB_SIZE); - memcpy(cdb + ISCSI_CDB_SIZE, ecdb_ahdr->ecdb, - be16_to_cpu(ecdb_ahdr->ahslength) - 1); + memcpy(cdb + ISCSI_CDB_SIZE, ecdb_ahdr->ecdb, cdb_length - ISCSI_CDB_SIZE); } data_direction = (hdr->flags & ISCSI_FLAG_CMD_WRITE) ? DMA_TO_DEVICE : From 846c76ecc02973b05ae909dd4248c11bfa277fc1 Mon Sep 17 00:00:00 2001 From: KaFai Wan Date: Tue, 21 Apr 2026 23:58:01 +0800 Subject: [PATCH 013/525] bpf: Reject TCP_NODELAY in TCP header option callbacks A BPF_SOCK_OPS program can enable BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG and then call bpf_setsockopt(TCP_NODELAY) from BPF_SOCK_OPS_HDR_OPT_LEN_CB or BPF_SOCK_OPS_WRITE_HDR_OPT_CB. In these callbacks, bpf_setsockopt(TCP_NODELAY) can reach __tcp_sock_set_nodelay(), which can call tcp_push_pending_frames(). >From BPF_SOCK_OPS_HDR_OPT_LEN_CB, tcp_push_pending_frames() can call tcp_current_mss(), which calls tcp_established_options() and re-enters bpf_skops_hdr_opt_len(). BPF_SOCK_OPS_HDR_OPT_LEN_CB -> bpf_setsockopt(TCP_NODELAY) -> tcp_push_pending_frames() -> tcp_current_mss() -> tcp_established_options() -> bpf_skops_hdr_opt_len() -> BPF_SOCK_OPS_HDR_OPT_LEN_CB >From BPF_SOCK_OPS_WRITE_HDR_OPT_CB, tcp_push_pending_frames() can call tcp_write_xmit(), which calls tcp_transmit_skb(). That path recomputes header option length through tcp_established_options() and bpf_skops_hdr_opt_len() before re-entering bpf_skops_write_hdr_opt(). BPF_SOCK_OPS_WRITE_HDR_OPT_CB -> bpf_setsockopt(TCP_NODELAY) -> tcp_push_pending_frames() -> tcp_write_xmit() -> tcp_transmit_skb() -> tcp_established_options() -> bpf_skops_hdr_opt_len() -> bpf_skops_write_hdr_opt() -> BPF_SOCK_OPS_WRITE_HDR_OPT_CB This leads to unbounded recursion and can overflow the kernel stack. Reject TCP_NODELAY with -EOPNOTSUPP in bpf_sock_ops_setsockopt() when bpf_setsockopt() is called from BPF_SOCK_OPS_HDR_OPT_LEN_CB or BPF_SOCK_OPS_WRITE_HDR_OPT_CB. Fixes: 7e41df5dbba2 ("bpf: Add a few optnames to bpf_setsockopt") Closes: https://lore.kernel.org/bpf/d1d523c9-6901-4454-a183-94462b8f3e4e@std.uestc.edu.cn/ Reported-by: Quan Sun <2022090917019@std.uestc.edu.cn> Reported-by: Yinhao Hu Reported-by: Kaiyan Mei Signed-off-by: KaFai Wan Signed-off-by: Martin KaFai Lau Reviewed-by: Jiayuan Chen Link: https://patch.msgid.link/20260421155804.135786-2-kafai.wan@linux.dev --- net/core/filter.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/core/filter.c b/net/core/filter.c index 5fa9189eb772..96849f4c1fbc 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -5833,6 +5833,12 @@ BPF_CALL_5(bpf_sock_ops_setsockopt, struct bpf_sock_ops_kern *, bpf_sock, if (!is_locked_tcp_sock_ops(bpf_sock)) return -EOPNOTSUPP; + /* TCP_NODELAY triggers tcp_push_pending_frames() and re-enters these callbacks. */ + if ((bpf_sock->op == BPF_SOCK_OPS_HDR_OPT_LEN_CB || + bpf_sock->op == BPF_SOCK_OPS_WRITE_HDR_OPT_CB) && + level == SOL_TCP && optname == TCP_NODELAY) + return -EOPNOTSUPP; + return _bpf_setsockopt(bpf_sock->sk, level, optname, optval, optlen); } From 54377fcab51f6f1f8807827d3751be42279e1a6a Mon Sep 17 00:00:00 2001 From: KaFai Wan Date: Tue, 21 Apr 2026 23:58:02 +0800 Subject: [PATCH 014/525] bpf: Reject TCP_NODELAY in bpf-tcp-cc A BPF TCP congestion control program can call bpf_setsockopt() from its callbacks. In current kernels, if it calls bpf_setsockopt(TCP_NODELAY) from cwnd_event_tx_start(), the call can re-enter the TCP transmit path before the outer tcp_transmit_skb() has completed and advanced the send head. This can re-trigger CA_EVENT_TX_START and lead to unbounded recursion: tcp_transmit_skb() -> tcp_event_data_sent() -> tcp_ca_event(sk, CA_EVENT_TX_START) -> cwnd_event_tx_start() -> bpf_setsockopt(TCP_NODELAY) -> tcp_push_pending_frames() -> tcp_write_xmit() -> tcp_transmit_skb() This leads to unbounded recursion and can overflow the kernel stack. Reject TCP_NODELAY with -EOPNOTSUPP for bpf-tcp-cc by introducing a dedicated setsockopt proto for BPF_PROG_TYPE_STRUCT_OPS TCP congestion control programs. To keep it simple, all tcp-cc ops is rejected for TCP_NODELAY. Fixes: 7e41df5dbba2 ("bpf: Add a few optnames to bpf_setsockopt") Suggested-by: Martin KaFai Lau Signed-off-by: KaFai Wan Signed-off-by: Martin KaFai Lau Reviewed-by: Jiayuan Chen Link: https://patch.msgid.link/20260421155804.135786-3-kafai.wan@linux.dev --- include/linux/bpf.h | 1 + net/core/filter.c | 24 ++++++++++++++++++++++++ net/ipv4/bpf_tcp_ca.c | 2 +- 3 files changed, 26 insertions(+), 1 deletion(-) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index b4b703c90ca9..01e203964892 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -3725,6 +3725,7 @@ extern const struct bpf_func_proto bpf_for_each_map_elem_proto; extern const struct bpf_func_proto bpf_btf_find_by_name_kind_proto; extern const struct bpf_func_proto bpf_sk_setsockopt_proto; extern const struct bpf_func_proto bpf_sk_getsockopt_proto; +extern const struct bpf_func_proto bpf_sk_setsockopt_nodelay_proto; extern const struct bpf_func_proto bpf_unlocked_sk_setsockopt_proto; extern const struct bpf_func_proto bpf_unlocked_sk_getsockopt_proto; extern const struct bpf_func_proto bpf_find_vma_proto; diff --git a/net/core/filter.c b/net/core/filter.c index 96849f4c1fbc..2914f5330310 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -5688,6 +5688,30 @@ const struct bpf_func_proto bpf_sk_getsockopt_proto = { .arg5_type = ARG_CONST_SIZE, }; +BPF_CALL_5(bpf_sk_setsockopt_nodelay, struct sock *, sk, int, level, + int, optname, char *, optval, int, optlen) +{ + /* + * TCP_NODELAY triggers tcp_push_pending_frames() and re-enters + * CA_EVENT_TX_START in bpf_tcp_cc. + */ + if (level == SOL_TCP && optname == TCP_NODELAY) + return -EOPNOTSUPP; + + return _bpf_setsockopt(sk, level, optname, optval, optlen); +} + +const struct bpf_func_proto bpf_sk_setsockopt_nodelay_proto = { + .func = bpf_sk_setsockopt_nodelay, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_BTF_ID_SOCK_COMMON, + .arg2_type = ARG_ANYTHING, + .arg3_type = ARG_ANYTHING, + .arg4_type = ARG_PTR_TO_MEM | MEM_RDONLY, + .arg5_type = ARG_CONST_SIZE, +}; + BPF_CALL_5(bpf_unlocked_sk_setsockopt, struct sock *, sk, int, level, int, optname, char *, optval, int, optlen) { diff --git a/net/ipv4/bpf_tcp_ca.c b/net/ipv4/bpf_tcp_ca.c index 008edc7f6688..791e15063237 100644 --- a/net/ipv4/bpf_tcp_ca.c +++ b/net/ipv4/bpf_tcp_ca.c @@ -168,7 +168,7 @@ bpf_tcp_ca_get_func_proto(enum bpf_func_id func_id, */ if (prog_ops_moff(prog) != offsetof(struct tcp_congestion_ops, release)) - return &bpf_sk_setsockopt_proto; + return &bpf_sk_setsockopt_nodelay_proto; return NULL; case BPF_FUNC_getsockopt: /* Since get/setsockopt is usually expected to From 52b6b5334924d8f083a2abe8edeface9206e13ee Mon Sep 17 00:00:00 2001 From: KaFai Wan Date: Tue, 21 Apr 2026 23:58:03 +0800 Subject: [PATCH 015/525] selftests/bpf: Test TCP_NODELAY in TCP hdr opt callbacks Add a sockops selftest for the TCP_NODELAY restriction in BPF_SOCK_OPS_HDR_OPT_LEN_CB and BPF_SOCK_OPS_WRITE_HDR_OPT_CB. With BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG enabled, bpf_setsockopt(TCP_NODELAY) returns -EOPNOTSUPP from BPF_SOCK_OPS_HDR_OPT_LEN_CB and BPF_SOCK_OPS_WRITE_HDR_OPT_CB, avoiding unbounded recursion and kernel stack overflow. Other cases continue to work as before, including BPF_SOCK_OPS_PASSIVE_ESTABLISHED_CB. Signed-off-by: KaFai Wan Signed-off-by: Martin KaFai Lau Reviewed-by: Jiayuan Chen Link: https://patch.msgid.link/20260421155804.135786-4-kafai.wan@linux.dev --- .../selftests/bpf/prog_tests/tcp_hdr_options.c | 4 ++++ .../bpf/progs/test_misc_tcp_hdr_options.c | 15 ++++++++++++++- 2 files changed, 18 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/prog_tests/tcp_hdr_options.c b/tools/testing/selftests/bpf/prog_tests/tcp_hdr_options.c index 56685fc03c7e..80e6315da2a5 100644 --- a/tools/testing/selftests/bpf/prog_tests/tcp_hdr_options.c +++ b/tools/testing/selftests/bpf/prog_tests/tcp_hdr_options.c @@ -507,6 +507,10 @@ static void misc(void) ASSERT_EQ(misc_skel->bss->nr_hwtstamp, 0, "nr_hwtstamp"); + ASSERT_TRUE(misc_skel->bss->nodelay_est_ok, "nodelay_est_ok"); + ASSERT_TRUE(misc_skel->bss->nodelay_hdr_len_reject, "nodelay_hdr_len_reject"); + ASSERT_TRUE(misc_skel->bss->nodelay_write_hdr_reject, "nodelay_write_hdr_reject"); + check_linum: ASSERT_FALSE(check_error_linum(&sk_fds), "check_error_linum"); sk_fds_close(&sk_fds); diff --git a/tools/testing/selftests/bpf/progs/test_misc_tcp_hdr_options.c b/tools/testing/selftests/bpf/progs/test_misc_tcp_hdr_options.c index d487153a839d..ed5a0011b863 100644 --- a/tools/testing/selftests/bpf/progs/test_misc_tcp_hdr_options.c +++ b/tools/testing/selftests/bpf/progs/test_misc_tcp_hdr_options.c @@ -29,6 +29,10 @@ unsigned int nr_syn = 0; unsigned int nr_fin = 0; unsigned int nr_hwtstamp = 0; +bool nodelay_est_ok = false; +bool nodelay_hdr_len_reject = false; +bool nodelay_write_hdr_reject = false; + /* Check the header received from the active side */ static int __check_active_hdr_in(struct bpf_sock_ops *skops, bool check_syn) { @@ -300,7 +304,7 @@ static int handle_passive_estab(struct bpf_sock_ops *skops) SEC("sockops") int misc_estab(struct bpf_sock_ops *skops) { - int true_val = 1; + int true_val = 1, false_val = 0, ret; switch (skops->op) { case BPF_SOCK_OPS_TCP_LISTEN_CB: @@ -316,10 +320,19 @@ int misc_estab(struct bpf_sock_ops *skops) case BPF_SOCK_OPS_PARSE_HDR_OPT_CB: return handle_parse_hdr(skops); case BPF_SOCK_OPS_HDR_OPT_LEN_CB: + ret = bpf_setsockopt(skops, SOL_TCP, TCP_NODELAY, &true_val, sizeof(true_val)); + if (ret == -EOPNOTSUPP) + nodelay_hdr_len_reject = true; return handle_hdr_opt_len(skops); case BPF_SOCK_OPS_WRITE_HDR_OPT_CB: + ret = bpf_setsockopt(skops, SOL_TCP, TCP_NODELAY, &true_val, sizeof(true_val)); + if (ret == -EOPNOTSUPP) + nodelay_write_hdr_reject = true; return handle_write_hdr_opt(skops); case BPF_SOCK_OPS_PASSIVE_ESTABLISHED_CB: + ret = bpf_setsockopt(skops, SOL_TCP, TCP_NODELAY, &false_val, sizeof(false_val)); + if (!ret) + nodelay_est_ok = true; return handle_passive_estab(skops); } From 2c7e33f1fc2e75fcfb4aa5d840bcd2e8b53c1847 Mon Sep 17 00:00:00 2001 From: KaFai Wan Date: Tue, 21 Apr 2026 23:58:04 +0800 Subject: [PATCH 016/525] selftests/bpf: Verify bpf-tcp-cc rejects TCP_NODELAY Add a bpf_tcp_ca selftest for the TCP_NODELAY restriction in bpf-tcp-cc. Update bpf_cubic to exercise init() and cwnd_event_tx_start(), and check that both callbacks reject bpf_setsockopt(TCP_NODELAY) with -EOPNOTSUPP. Signed-off-by: KaFai Wan Signed-off-by: Martin KaFai Lau Link: https://patch.msgid.link/20260421155804.135786-5-kafai.wan@linux.dev --- .../testing/selftests/bpf/prog_tests/bpf_tcp_ca.c | 4 ++++ tools/testing/selftests/bpf/progs/bpf_cubic.c | 14 ++++++++++++++ 2 files changed, 18 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/bpf_tcp_ca.c b/tools/testing/selftests/bpf/prog_tests/bpf_tcp_ca.c index f829b6f09bc9..fe30181e6336 100644 --- a/tools/testing/selftests/bpf/prog_tests/bpf_tcp_ca.c +++ b/tools/testing/selftests/bpf/prog_tests/bpf_tcp_ca.c @@ -112,6 +112,10 @@ static void test_cubic(void) ASSERT_EQ(cubic_skel->bss->bpf_cubic_acked_called, 1, "pkts_acked called"); + ASSERT_TRUE(cubic_skel->bss->nodelay_init_reject, "init reject nodelay option"); + ASSERT_TRUE(cubic_skel->bss->nodelay_cwnd_event_tx_start_reject, + "cwnd_event_tx_start reject nodelay option"); + bpf_link__destroy(link); bpf_cubic__destroy(cubic_skel); } diff --git a/tools/testing/selftests/bpf/progs/bpf_cubic.c b/tools/testing/selftests/bpf/progs/bpf_cubic.c index ce18a4db813f..ebd5a1e69f56 100644 --- a/tools/testing/selftests/bpf/progs/bpf_cubic.c +++ b/tools/testing/selftests/bpf/progs/bpf_cubic.c @@ -16,6 +16,7 @@ #include "bpf_tracing_net.h" #include +#include char _license[] SEC("license") = "GPL"; @@ -170,10 +171,18 @@ static void bictcp_hystart_reset(struct sock *sk) ca->sample_cnt = 0; } +bool nodelay_init_reject = false; +bool nodelay_cwnd_event_tx_start_reject = false; + SEC("struct_ops") void BPF_PROG(bpf_cubic_init, struct sock *sk) { struct bpf_bictcp *ca = inet_csk_ca(sk); + int true_val = 1, ret; + + ret = bpf_setsockopt(sk, SOL_TCP, TCP_NODELAY, &true_val, sizeof(true_val)); + if (ret == -EOPNOTSUPP) + nodelay_init_reject = true; bictcp_reset(ca); @@ -189,8 +198,13 @@ void BPF_PROG(bpf_cubic_cwnd_event_tx_start, struct sock *sk) { struct bpf_bictcp *ca = inet_csk_ca(sk); __u32 now = tcp_jiffies32; + int true_val = 1, ret; __s32 delta; + ret = bpf_setsockopt(sk, SOL_TCP, TCP_NODELAY, &true_val, sizeof(true_val)); + if (ret == -EOPNOTSUPP) + nodelay_cwnd_event_tx_start_reject = true; + delta = now - tcp_sk(sk)->lsndtime; /* We were application limited (idle) for a while. From 4a1b534177395627579c1fb9e7f9100ee88955dd Mon Sep 17 00:00:00 2001 From: Baochen Qiang Date: Tue, 10 Feb 2026 11:07:31 +0800 Subject: [PATCH 017/525] wifi: ath12k: prepare REO update element only for primary link Commit [1] introduces dp->reo_cmd_update_rx_queue_list for the purpose of tracking all pending REO queue flush commands. The helper ath12k_dp_prepare_reo_update_elem() allocates an element and populates it with REO queue information, then add it to the list. The element would be helpful during clean up stage to finally unmap/free the corresponding REO queue buffer. In MLO scenarios with more than one links, for non dp_primary_link_only chips like WCN7850, that helper is called for each link peer. This results in multiple elements added to the list but all of them pointing to the same REO queue buffer. Consequently the same buffer gets unmap/freed multiple times: BUG kmalloc-2k (Tainted: G B W O ): Object already free ----------------------------------------------------------------------------- Allocated in ath12k_wifi7_dp_rx_assign_reoq+0xce/0x280 [ath12k_wifi7] age=7436 cpu=10 pid=16130 __kmalloc_noprof ath12k_wifi7_dp_rx_assign_reoq ath12k_dp_rx_peer_tid_setup ath12k_dp_peer_setup ath12k_mac_station_add ath12k_mac_op_sta_state [...] Freed in ath12k_dp_rx_tid_cleanup.part.0+0x25/0x40 [ath12k] age=1 cpu=27 pid=16137 kfree ath12k_dp_rx_tid_cleanup.part.0 ath12k_dp_rx_reo_cmd_list_cleanup ath12k_dp_cmn_device_deinit ath12k_core_stop ath12k_core_hw_group_cleanup ath12k_pci_remove Fix this by allowing list addition for primary link only. Note dp_primary_link_only chips like QCN9274 are not affected by this change, because that's what they were doing in the first place. Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.1.c5-00302-QCAHMTSWPL_V1.0_V2.0_SILICONZ-1.115823.3 Fixes: 3bf2e57e7d6c ("wifi: ath12k: Add Retry Mechanism for REO RX Queue Update Failures") # [1] Closes: https://bugzilla.kernel.org/show_bug.cgi?id=221011 Signed-off-by: Baochen Qiang Reviewed-by: Vasanthakumar Thiagarajan Link: https://patch.msgid.link/20260210-ath12k-rxtid-double-free-v1-1-8b523fb2886d@oss.qualcomm.com Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath12k/dp_rx.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/wireless/ath/ath12k/dp_rx.c b/drivers/net/wireless/ath/ath12k/dp_rx.c index 250459facff3..25557dea5826 100644 --- a/drivers/net/wireless/ath/ath12k/dp_rx.c +++ b/drivers/net/wireless/ath/ath12k/dp_rx.c @@ -565,6 +565,9 @@ static int ath12k_dp_prepare_reo_update_elem(struct ath12k_dp *dp, lockdep_assert_held(&dp->dp_lock); + if (!peer->primary_link) + return 0; + elem = kzalloc_obj(*elem, GFP_ATOMIC); if (!elem) return -ENOMEM; From f3ba9e05cc7b65f41f58bb4808f6c3a8f7894bb1 Mon Sep 17 00:00:00 2001 From: Aaradhana Sahu Date: Fri, 10 Apr 2026 12:43:00 +0530 Subject: [PATCH 018/525] wifi: ath12k: fix OF node refcount imbalance in WSI graph traversal ath12k_core_get_wsi_info() traverses the WSI (Wired Serial Interface) device graph starting from dev->of_node. The current code uses dev->of_node directly as the local traversal pointer and calls of_node_put() on error. Since the driver does not own a reference to dev->of_node, dropping it during traversal results in the following OF refcount underflow: OF: ERROR: of_node_release() detected bad of_node_put() on /soc@0/wifi@c000000 CPU: 1 UID: 0 PID: 210 Comm: insmod Not tainted 6.19.0-rc4-next-20260109-00023-g797dd36dc178 #26 PREEMPT Hardware name: Qualcomm Technologies, Inc. IPQ5332 MI01.2 (DT) Call trace: show_stack+0x18/0x24 (C) dump_stack_lvl+0x60/0x80 dump_stack+0x18/0x24 of_node_release+0x164/0x1a0 kobject_put+0xb4/0x278 of_node_put+0x18/0x28 ath12k_core_init+0x29c/0x5d4 [ath12k] ath12k_ahb_probe+0x950/0xc14 [ath12k] platform_probe+0x5c/0xa4 really_probe+0xc0/0x3ec __driver_probe_device+0x80/0x170 driver_probe_device+0x3c/0x120 __driver_attach+0xc4/0x218 OF: ERROR: next of_node_put() on this node will result in a kobject warning 'refcount_t: underflow; use-after-free.' Fix this by explicitly acquiring a reference to the starting node using of_node_get() and attaching automatic cleanup via __free(device_node). Each discovered WSI node is stored in ag->wsi_node[] with its own of_node_get() reference. These references are later released in ath12k_core_free_wsi_info() during driver teardown. Also remove unnecessary memset() of wsi_node array since cleanup now explicitly sets pointers to NULL. Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.6-01243-QCAHKSWPL_SILICONZ-1 Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.6-01275-QCAHKSWPL_SILICONZ-1 Fixes: 908c10c860e0 ("wifi: ath12k: parse multiple device information from Device Tree") Signed-off-by: Aaradhana Sahu Reviewed-by: Rameshkumar Sundaram Reviewed-by: Baochen Qiang Link: https://patch.msgid.link/20260410071300.2323603-1-aaradhana.sahu@oss.qualcomm.com Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath12k/core.c | 77 ++++++++++++++++---------- 1 file changed, 48 insertions(+), 29 deletions(-) diff --git a/drivers/net/wireless/ath/ath12k/core.c b/drivers/net/wireless/ath/ath12k/core.c index 2519e2400d58..980a12fb2c6e 100644 --- a/drivers/net/wireless/ath/ath12k/core.c +++ b/drivers/net/wireless/ath/ath12k/core.c @@ -1838,10 +1838,22 @@ static struct ath12k_hw_group *ath12k_core_hw_group_alloc(struct ath12k_base *ab return ag; } +static void ath12k_core_free_wsi_info(struct ath12k_hw_group *ag) +{ + int i; + + for (i = 0; i < ag->num_devices; i++) { + of_node_put(ag->wsi_node[i]); + ag->wsi_node[i] = NULL; + } + ag->num_devices = 0; +} + static void ath12k_core_hw_group_free(struct ath12k_hw_group *ag) { mutex_lock(&ath12k_hw_group_mutex); + ath12k_core_free_wsi_info(ag); list_del(&ag->list); kfree(ag); @@ -1867,52 +1879,59 @@ static struct ath12k_hw_group *ath12k_core_hw_group_find_by_dt(struct ath12k_bas static int ath12k_core_get_wsi_info(struct ath12k_hw_group *ag, struct ath12k_base *ab) { - struct device_node *wsi_dev = ab->dev->of_node, *next_wsi_dev; - struct device_node *tx_endpoint, *next_rx_endpoint; - int device_count = 0; + struct device_node *next_wsi_dev; + int device_count = 0, ret = 0; + struct device_node *wsi_dev; - next_wsi_dev = wsi_dev; - - if (!next_wsi_dev) + wsi_dev = of_node_get(ab->dev->of_node); + if (!wsi_dev) return -ENODEV; do { - ag->wsi_node[device_count] = next_wsi_dev; + if (device_count >= ATH12K_MAX_DEVICES) { + ath12k_warn(ab, "device count in DT %d is more than limit %d\n", + device_count, ATH12K_MAX_DEVICES); + ret = -EINVAL; + break; + } - tx_endpoint = of_graph_get_endpoint_by_regs(next_wsi_dev, 0, -1); + ag->wsi_node[device_count++] = of_node_get(wsi_dev); + + struct device_node *tx_endpoint __free(device_node) = + of_graph_get_endpoint_by_regs(wsi_dev, 0, -1); if (!tx_endpoint) { - of_node_put(next_wsi_dev); - return -ENODEV; + ret = -ENODEV; + break; } - next_rx_endpoint = of_graph_get_remote_endpoint(tx_endpoint); + struct device_node *next_rx_endpoint __free(device_node) = + of_graph_get_remote_endpoint(tx_endpoint); if (!next_rx_endpoint) { - of_node_put(next_wsi_dev); - of_node_put(tx_endpoint); - return -ENODEV; + ret = -ENODEV; + break; } - of_node_put(tx_endpoint); - of_node_put(next_wsi_dev); - next_wsi_dev = of_graph_get_port_parent(next_rx_endpoint); if (!next_wsi_dev) { - of_node_put(next_rx_endpoint); - return -ENODEV; + ret = -ENODEV; + break; } - of_node_put(next_rx_endpoint); + of_node_put(wsi_dev); + wsi_dev = next_wsi_dev; + } while (ab->dev->of_node != wsi_dev); - device_count++; - if (device_count > ATH12K_MAX_DEVICES) { - ath12k_warn(ab, "device count in DT %d is more than limit %d\n", - device_count, ATH12K_MAX_DEVICES); - of_node_put(next_wsi_dev); - return -EINVAL; + if (ret) { + while (--device_count >= 0) { + of_node_put(ag->wsi_node[device_count]); + ag->wsi_node[device_count] = NULL; } - } while (wsi_dev != next_wsi_dev); - of_node_put(next_wsi_dev); + of_node_put(wsi_dev); + return ret; + } + + of_node_put(wsi_dev); ag->num_devices = device_count; return 0; @@ -1983,9 +2002,9 @@ static struct ath12k_hw_group *ath12k_core_hw_group_assign(struct ath12k_base *a ath12k_core_get_wsi_index(ag, ab)) { ath12k_dbg(ab, ATH12K_DBG_BOOT, "unable to get wsi info from dt, grouping single device"); + ath12k_core_free_wsi_info(ag); ag->id = ATH12K_INVALID_GROUP_ID; ag->num_devices = 1; - memset(ag->wsi_node, 0, sizeof(ag->wsi_node)); wsi->index = 0; } From c4b6ad0e14f5df942eed5ebadaff84b468bd2496 Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Sat, 18 Apr 2026 22:37:00 +0300 Subject: [PATCH 019/525] wifi: ath10k: snoc: select POWER_SEQUENCING The commit afcf3ec615c9 ("wifi: ath10k: snoc: support powering on the device via pwrseq") made ath10k SNOC driver use devm_pwrseq_get(). Select the corresponding Kconfig symbol to make sure that API call is always available and doesn't return an error per se. Fixes: afcf3ec615c9 ("wifi: ath10k: snoc: support powering on the device via pwrseq") Reported-by: Luca Weiss Closes: https://lore.kernel.org/r/DHUHU7UIT487.139L3KIVRVREU@fairphone.com Signed-off-by: Dmitry Baryshkov Reviewed-by: Rameshkumar Sundaram Link: https://patch.msgid.link/20260418-ath10k-snoc-pwrseq-v1-1-832594ba3294@oss.qualcomm.com Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath10k/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/wireless/ath/ath10k/Kconfig b/drivers/net/wireless/ath/ath10k/Kconfig index 876aed765833..efb9f022d8c6 100644 --- a/drivers/net/wireless/ath/ath10k/Kconfig +++ b/drivers/net/wireless/ath/ath10k/Kconfig @@ -46,6 +46,7 @@ config ATH10K_SNOC depends on ARCH_QCOM || COMPILE_TEST depends on QCOM_SMEM depends on QCOM_RPROC_COMMON || QCOM_RPROC_COMMON=n + select POWER_SEQUENCING select QCOM_SCM select QCOM_QMI_HELPERS help From 4498664e2d5888efabb96428196a926acdaa25ed Mon Sep 17 00:00:00 2001 From: Yu-Hsiang Tseng Date: Thu, 23 Apr 2026 02:08:14 +0800 Subject: [PATCH 020/525] wifi: ath12k: use lockdep_assert_in_rcu_read_lock() for RCU assertions Two functions in ath12k assert that the caller holds an RCU read lock: ath12k_mac_get_arvif() and ath12k_p2p_noa_update_vdev_iter(). Both use: WARN_ON(!rcu_read_lock_any_held()); On kernels using preemptible RCU (CONFIG_PREEMPT=y or CONFIG_PREEMPT_RT=y) without CONFIG_DEBUG_LOCK_ALLOC, this produces a false positive splat whenever these functions are invoked from paths that do hold the RCU read lock (e.g. firmware stats processing or mac80211 interface iteration). Root cause: - Without CONFIG_DEBUG_LOCK_ALLOC, rcu_read_lock_any_held() is a static inline that returns !preemptible() as a proxy for "in an RCU read section". - With preemptible RCU, rcu_read_lock() does not disable preemption. A task can therefore be preemptible while legitimately holding an RCU read lock, making the proxy unreliable. - Callers such as ath12k_wmi_tlv_rssi_chain_parse() (via guard(rcu)()) and ieee80211_iterate_active_interfaces_atomic() do hold the RCU read lock, so these warnings are incorrect. Typical splat seen on a WCN7850 station with periodic fw stats processing: WARNING: drivers/net/wireless/ath/ath12k/mac.c:791 at ath12k_mac_get_arvif+0x9e/0xd0 [ath12k] Tainted: G W O 6.19.13-rt #1 PREEMPT_RT Call Trace: ath12k_wmi_tlv_rssi_chain_parse+0x69/0x170 [ath12k] ath12k_wmi_tlv_iter+0x7f/0x120 [ath12k] ath12k_wmi_tlv_fw_stats_parse+0x342/0x6b0 [ath12k] ath12k_wmi_op_rx+0xe9e/0x3150 [ath12k] ath12k_htc_rx_completion_handler+0x3df/0x5b0 [ath12k] ath12k_ce_per_engine_service+0x325/0x3e0 [ath12k] ath12k_pci_ce_workqueue+0x20/0x40 [ath12k] Replace WARN_ON(!rcu_read_lock_any_held()) with lockdep_assert_in_rcu_read_lock(), which is gated on CONFIG_PROVE_RCU and therefore compiles out entirely when PROVE_RCU is disabled. PROVE_RCU kernels continue to get the full lockdep-based check, and the new helper precisely checks for rcu_read_lock() rather than any RCU variant, which better matches the callers' expectations. Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.1.c5-00302-QCAHMTSWPL_V1.0_V2.0_SILICONZ-1.115823.3 Fixes: 3dd2c68f206e ("wifi: ath12k: prepare vif data structure for MLO handling") Suggested-by: Baochen Qiang Suggested-by: Sebastian Andrzej Siewior Reviewed-by: Baochen Qiang Reviewed-by: Rameshkumar Sundaram Signed-off-by: Yu-Hsiang Tseng Reviewed-by: Sebastian Andrzej Siewior Link: https://patch.msgid.link/20260422180814.1938317-1-asas1asas200@gmail.com Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath12k/mac.c | 2 +- drivers/net/wireless/ath/ath12k/p2p.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath12k/mac.c b/drivers/net/wireless/ath/ath12k/mac.c index fbdfe6424fd7..df2334f3bad6 100644 --- a/drivers/net/wireless/ath/ath12k/mac.c +++ b/drivers/net/wireless/ath/ath12k/mac.c @@ -788,7 +788,7 @@ struct ath12k_link_vif *ath12k_mac_get_arvif(struct ath12k *ar, u32 vdev_id) /* To use the arvif returned, caller must have held rcu read lock. */ - WARN_ON(!rcu_read_lock_any_held()); + lockdep_assert_in_rcu_read_lock(); arvif_iter.vdev_id = vdev_id; arvif_iter.ar = ar; diff --git a/drivers/net/wireless/ath/ath12k/p2p.c b/drivers/net/wireless/ath/ath12k/p2p.c index 59589748f1a8..19ebcd1d8eb2 100644 --- a/drivers/net/wireless/ath/ath12k/p2p.c +++ b/drivers/net/wireless/ath/ath12k/p2p.c @@ -123,7 +123,7 @@ static void ath12k_p2p_noa_update_vdev_iter(void *data, u8 *mac, struct ath12k_p2p_noa_arg *arg = data; struct ath12k_link_vif *arvif; - WARN_ON(!rcu_read_lock_any_held()); + lockdep_assert_in_rcu_read_lock(); arvif = &ahvif->deflink; if (!arvif->is_created || arvif->ar != arg->ar || arvif->vdev_id != arg->vdev_id) return; From 375e4e33c18dfa05c5dfd5f3dfffeb29343dd4c7 Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Tue, 21 Apr 2026 23:54:12 -0700 Subject: [PATCH 021/525] bpf: Fix NULL pointer dereference in bpf_sk_storage_clone and diag paths bpf_selem_unlink_nofail() sets SDATA(selem)->smap to NULL before removing the selem from the storage hlist. A concurrent RCU reader in bpf_sk_storage_clone() can observe the selem still on the list with smap already NULL, causing a NULL pointer dereference. general protection fault, probably for non-canonical address 0xdffffc000000000a: KASAN: null-ptr-deref in range [0x0000000000000050-0x0000000000000057] RIP: 0010:bpf_sk_storage_clone+0x1cd/0xaa0 net/core/bpf_sk_storage.c:174 Call Trace: sk_clone+0xfed/0x1980 net/core/sock.c:2591 inet_csk_clone_lock+0x30/0x760 net/ipv4/inet_connection_sock.c:1222 tcp_create_openreq_child+0x35/0x2680 net/ipv4/tcp_minisocks.c:571 tcp_v4_syn_recv_sock+0x123/0xf90 net/ipv4/tcp_ipv4.c:1729 tcp_check_req+0x8e1/0x2580 include/net/tcp.h:855 tcp_v4_rcv+0x1845/0x3b80 net/ipv4/tcp_ipv4.c:2347 Add a NULL check for smap in bpf_sk_storage_clone(). bpf_sk_storage_diag_put_all() has the same issue. Add a NULL check and pass the validated smap directly to diag_get(), which is refactored to take smap as a parameter instead of reading it internally. bpf_sk_storage_diag_put() uses diag->maps[i] which is always valid under its refcount, so diag->maps[i] is passed directly to diag_get(). Fixes: 5d800f87d0a5 ("bpf: Support lockless unlink when freeing map or local storage") Reported-by: Xiang Mei Acked-by: Amery Hung Signed-off-by: Weiming Shi Signed-off-by: Martin KaFai Lau Link: https://patch.msgid.link/20260422065411.1007737-2-bestswngs@gmail.com --- net/core/bpf_sk_storage.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/net/core/bpf_sk_storage.c b/net/core/bpf_sk_storage.c index 14eb7812bda4..dc3e8fce8809 100644 --- a/net/core/bpf_sk_storage.c +++ b/net/core/bpf_sk_storage.c @@ -172,7 +172,7 @@ int bpf_sk_storage_clone(const struct sock *sk, struct sock *newsk) struct bpf_map *map; smap = rcu_dereference(SDATA(selem)->smap); - if (!(smap->map.map_flags & BPF_F_CLONE)) + if (!smap || !(smap->map.map_flags & BPF_F_CLONE)) continue; /* Note that for lockless listeners adding new element @@ -531,10 +531,10 @@ bpf_sk_storage_diag_alloc(const struct nlattr *nla_stgs) } EXPORT_SYMBOL_GPL(bpf_sk_storage_diag_alloc); -static int diag_get(struct bpf_local_storage_data *sdata, struct sk_buff *skb) +static int diag_get(struct bpf_local_storage_map *smap, + struct bpf_local_storage_data *sdata, struct sk_buff *skb) { struct nlattr *nla_stg, *nla_value; - struct bpf_local_storage_map *smap; /* It cannot exceed max nlattr's payload */ BUILD_BUG_ON(U16_MAX - NLA_HDRLEN < BPF_LOCAL_STORAGE_MAX_VALUE_SIZE); @@ -543,7 +543,6 @@ static int diag_get(struct bpf_local_storage_data *sdata, struct sk_buff *skb) if (!nla_stg) return -EMSGSIZE; - smap = rcu_dereference(sdata->smap); if (nla_put_u32(skb, SK_DIAG_BPF_STORAGE_MAP_ID, smap->map.id)) goto errout; @@ -596,9 +595,11 @@ static int bpf_sk_storage_diag_put_all(struct sock *sk, struct sk_buff *skb, saved_len = skb->len; hlist_for_each_entry_rcu(selem, &sk_storage->list, snode) { smap = rcu_dereference(SDATA(selem)->smap); + if (!smap) + continue; diag_size += nla_value_size(smap->map.value_size); - if (nla_stgs && diag_get(SDATA(selem), skb)) + if (nla_stgs && diag_get(smap, SDATA(selem), skb)) /* Continue to learn diag_size */ err = -EMSGSIZE; } @@ -665,7 +666,7 @@ int bpf_sk_storage_diag_put(struct bpf_sk_storage_diag *diag, diag_size += nla_value_size(diag->maps[i]->value_size); - if (nla_stgs && diag_get(sdata, skb)) + if (nla_stgs && diag_get((struct bpf_local_storage_map *)diag->maps[i], sdata, skb)) /* Continue to learn diag_size */ err = -EMSGSIZE; } From 6451d58a355642b612f2bf948ad39108c998ac2a Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Mon, 20 Apr 2026 19:48:41 +0000 Subject: [PATCH 022/525] sockmap: Fix sk_psock_drop() race vs sock_map_{unhash,close,destroy}(). syzbot reported a splat in sock_map_destroy() [0], where psock was NULL even though sk->sk_prot still pointed to tcp_bpf_prots[][]. The stack trace shows how badly the path was excercised, see inet_release() calls tcp_close(), not sock_map_close() yet, but finally reaching sock_map_destroy(). The root cause is a lack of synchronisation. Even if sk_psock_get() fails to bump psock->refcnt, it does not guarantee that sk_psock_drop() has finished, and thus sk->sk_prot might not have been restored to the original one. Commit 4b4647add7d3 ("sock_map: avoid race between sock_map_close and sk_psock_put") attempted to address this, but it was insufficient for two reasons. It did not cover sock_map_unhash() and sock_map_destroy(), and it missed the corner case where sk_psock() is NULL. On non-x86 platforms, sk_psock_restore_proto(sk, psock) and rcu_assign_sk_user_data(sk, NULL) can be reordered because there is no address dependency between sk->sk_prot and sk->sk_user_data. sk_psock_get() returning NULL implies nothing about sk->sk_prot. Let's simply retry sk_psock_get() in the unlikely case. Note that we cannot avoid loop even if we added memory barrier in sk_psock_drop() and sock_map_psock_get_checked(). Also note that sock_map_destroy() cannot be called from softirq while sock_map_close() has also been running. It is because sock_map_destroy() requires SOCK_DEAD, so sock_map_destroy() cannot happen until sock_map_close() has finished the saved_close() (which is tcp_close()). [0]: WARNING: CPU: 1 PID: 8459 at net/core/sock_map.c:1667 sock_map_destroy+0x28b/0x2b0 net/core/sock_map.c:1667 Modules linked in: CPU: 1 UID: 0 PID: 8459 Comm: syz.0.1109 Not tainted syzkaller #0 PREEMPT_{RT,(full)} Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 07/12/2025 RIP: 0010:sock_map_destroy+0x28b/0x2b0 net/core/sock_map.c:1667 Code: 8b 36 49 83 c6 38 4c 89 f0 48 c1 e8 03 42 80 3c 38 00 74 08 4c 89 f7 e8 93 62 22 f9 4d 8b 3e e9 79 ff ff ff e8 a6 2b c3 f8 90 <0f> 0b 90 eb 9c e8 9b 2b c3 f8 4c 89 e7 be 03 00 00 00 e8 0e 4e bc RSP: 0018:ffffc9000d067be8 EFLAGS: 00010293 RAX: ffffffff88fb30aa RBX: ffff888024832000 RCX: ffff888024283b80 RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000 RBP: 0000000000000001 R08: 0000000000000000 R09: 0000000000000000 R10: dffffc0000000000 R11: ffffed100862e946 R12: dffffc0000000000 R13: ffff888024832000 R14: ffffffff995b2208 R15: ffffffff88fb2e20 FS: 0000555579a7d500(0000) GS:ffff8881269c2000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00002000000048c0 CR3: 000000003713a000 CR4: 00000000003526f0 Call Trace: inet_csk_destroy_sock+0x166/0x3a0 net/ipv4/inet_connection_sock.c:1294 __tcp_close+0xcc1/0xfd0 net/ipv4/tcp.c:3262 tcp_close+0x28/0x110 net/ipv4/tcp.c:3274 inet_release+0x144/0x190 net/ipv4/af_inet.c:435 __sock_release net/socket.c:649 [inline] sock_close+0xc0/0x240 net/socket.c:1439 __fput+0x45b/0xa80 fs/file_table.c:468 task_work_run+0x1d4/0x260 kernel/task_work.c:227 resume_user_mode_work include/linux/resume_user_mode.h:50 [inline] exit_to_user_mode_loop+0xec/0x110 kernel/entry/common.c:43 exit_to_user_mode_prepare include/linux/irq-entry-common.h:225 [inline] syscall_exit_to_user_mode_work include/linux/entry-common.h:175 [inline] syscall_exit_to_user_mode include/linux/entry-common.h:210 [inline] do_syscall_64+0x2bd/0x3b0 arch/x86/entry/syscall_64.c:100 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x7f265847ebe9 Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 a8 ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007ffd158dfbd8 EFLAGS: 00000246 ORIG_RAX: 00000000000001b4 RAX: 0000000000000000 RBX: 000000000002ddb0 RCX: 00007f265847ebe9 RDX: 0000000000000000 RSI: 000000000000001e RDI: 0000000000000003 RBP: 00007f26586a7da0 R08: 0000000000000001 R09: 0000000e158dfecf R10: 0000001b30a20000 R11: 0000000000000246 R12: 00007f26586a5fac R13: 00007f26586a5fa0 R14: ffffffffffffffff R15: 00007ffd158dfcf0 Fixes: 1aa12bdf1bfb ("bpf: sockmap, add sock close() hook to remove socks") Fixes: b05545e15e1f ("bpf: sockmap, fix transition through disconnect without close") Fixes: d8616ee2affc ("bpf, sockmap: Fix sk->sk_forward_alloc warn_on in sk_stream_kill_queues") Reported-by: syzbot+b0842d38af58376d1fdc@syzkaller.appspotmail.com Closes: https://lore.kernel.org/bpf/69cec5ef.050a0220.2dbe29.0009.GAE@google.com/ Signed-off-by: Kuniyuki Iwashima Signed-off-by: Martin KaFai Lau Reviewed-by: Jiayuan Chen Link: https://patch.msgid.link/20260420194846.1089595-1-kuniyu@google.com --- net/core/sock_map.c | 39 +++++++++++++++++++++++++-------------- 1 file changed, 25 insertions(+), 14 deletions(-) diff --git a/net/core/sock_map.c b/net/core/sock_map.c index 02a68be3002a..99e3789492a0 100644 --- a/net/core/sock_map.c +++ b/net/core/sock_map.c @@ -1630,18 +1630,23 @@ void sock_map_unhash(struct sock *sk) void (*saved_unhash)(struct sock *sk); struct sk_psock *psock; +retry: rcu_read_lock(); psock = sk_psock(sk); if (unlikely(!psock)) { rcu_read_unlock(); saved_unhash = READ_ONCE(sk->sk_prot)->unhash; + if (unlikely(saved_unhash == sock_map_unhash)) + goto retry; } else { saved_unhash = psock->saved_unhash; sock_map_remove_links(sk, psock); rcu_read_unlock(); + + if (WARN_ON_ONCE(saved_unhash == sock_map_unhash)) + return; } - if (WARN_ON_ONCE(saved_unhash == sock_map_unhash)) - return; + if (saved_unhash) saved_unhash(sk); } @@ -1652,20 +1657,25 @@ void sock_map_destroy(struct sock *sk) void (*saved_destroy)(struct sock *sk); struct sk_psock *psock; +retry: rcu_read_lock(); psock = sk_psock_get(sk); if (unlikely(!psock)) { rcu_read_unlock(); saved_destroy = READ_ONCE(sk->sk_prot)->destroy; + if (unlikely(saved_destroy == sock_map_destroy)) + goto retry; } else { saved_destroy = psock->saved_destroy; sock_map_remove_links(sk, psock); rcu_read_unlock(); sk_psock_stop(psock); sk_psock_put(sk, psock); + + if (WARN_ON_ONCE(saved_destroy == sock_map_destroy)) + return; } - if (WARN_ON_ONCE(saved_destroy == sock_map_destroy)) - return; + if (saved_destroy) saved_destroy(sk); } @@ -1676,32 +1686,33 @@ void sock_map_close(struct sock *sk, long timeout) void (*saved_close)(struct sock *sk, long timeout); struct sk_psock *psock; +retry: lock_sock(sk); rcu_read_lock(); - psock = sk_psock(sk); + psock = sk_psock_get(sk); if (likely(psock)) { saved_close = psock->saved_close; sock_map_remove_links(sk, psock); - psock = sk_psock_get(sk); - if (unlikely(!psock)) - goto no_psock; rcu_read_unlock(); sk_psock_stop(psock); release_sock(sk); cancel_delayed_work_sync(&psock->work); sk_psock_put(sk, psock); + + /* Make sure we do not recurse. This is a bug. + * Leak the socket instead of crashing on a stack overflow. + */ + if (WARN_ON_ONCE(saved_close == sock_map_close)) + return; } else { saved_close = READ_ONCE(sk->sk_prot)->close; -no_psock: rcu_read_unlock(); release_sock(sk); + + if (unlikely(saved_close == sock_map_close)) + goto retry; } - /* Make sure we do not recurse. This is a bug. - * Leak the socket instead of crashing on a stack overflow. - */ - if (WARN_ON_ONCE(saved_close == sock_map_close)) - return; saved_close(sk, timeout); } EXPORT_SYMBOL_GPL(sock_map_close); From 1081de1accb2b224516cca7071122c59532d0b22 Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Thu, 23 Apr 2026 11:38:32 -0700 Subject: [PATCH 023/525] bpf: Fix NULL pointer dereference in bpf_skb_fib_lookup() When tot_len is not provided by the user, bpf_skb_fib_lookup() resolves the FIB result's output device via dev_get_by_index_rcu() to check skb forwardability and fill in mtu_result. The returned pointer is dereferenced without a NULL check. If the device is concurrently unregistered, dev_get_by_index_rcu() returns NULL and is_skb_forwardable() crashes at dev->flags: KASAN: null-ptr-deref in range [0x00000000000000b0-0x00000000000000b7] Call Trace: is_skb_forwardable (include/linux/netdevice.h:4365) bpf_skb_fib_lookup (net/core/filter.c:6446) bpf_prog_test_run_skb (net/bpf/test_run.c) __sys_bpf (kernel/bpf/syscall.c) Add the missing NULL check, returning -ENODEV to be consistent with how bpf_ipv4_fib_lookup() and bpf_ipv6_fib_lookup() handle the same condition. Fixes: 4f74fede40df ("bpf: Add mtu checking to FIB forwarding helper") Reported-by: Xiang Mei Signed-off-by: Weiming Shi Signed-off-by: Martin KaFai Lau Acked-by: Paul Chaignon Link: https://patch.msgid.link/20260423183831.1325480-2-bestswngs@gmail.com --- net/core/filter.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/core/filter.c b/net/core/filter.c index 2914f5330310..bc96c18df4e0 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -6473,6 +6473,8 @@ BPF_CALL_4(bpf_skb_fib_lookup, struct sk_buff *, skb, * against MTU of FIB lookup resulting net_device */ dev = dev_get_by_index_rcu(net, params->ifindex); + if (unlikely(!dev)) + return -ENODEV; if (!is_skb_forwardable(dev, skb)) rc = BPF_FIB_LKUP_RET_FRAG_NEEDED; From b5c111f4967ba4fdecdd318923ec7b081e9ef95f Mon Sep 17 00:00:00 2001 From: Amery Hung Date: Thu, 23 Apr 2026 15:23:55 -0700 Subject: [PATCH 024/525] bpf: Fix sk_local_storage diag dumping uninitialized special fields Call check_and_init_map_value() after the copy_map_value() to zero out special field regions. diag_get() copies sk_local_storage map values into a netlink message using copy_map_value{_locked}(), which intentionally skip special fields. However, the destination buffer from nla_reserve_64bit() is not zeroed and the skipped regions contain uninitialized skb data can be sent to userspace. Fixes: 1ed4d92458a9 ("bpf: INET_DIAG support in bpf_sk_storage") Signed-off-by: Amery Hung Signed-off-by: Martin KaFai Lau Link: https://patch.msgid.link/20260423222356.155387-1-ameryhung@gmail.com --- net/core/bpf_sk_storage.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/core/bpf_sk_storage.c b/net/core/bpf_sk_storage.c index dc3e8fce8809..ecd659f79fd4 100644 --- a/net/core/bpf_sk_storage.c +++ b/net/core/bpf_sk_storage.c @@ -557,6 +557,7 @@ static int diag_get(struct bpf_local_storage_map *smap, sdata->data, true); else copy_map_value(&smap->map, nla_data(nla_value), sdata->data); + check_and_init_map_value(&smap->map, nla_data(nla_value)); nla_nest_end(skb, nla_stg); return 0; From 5f69165b7e4215f02247b0c64052c71b2f66d73a Mon Sep 17 00:00:00 2001 From: "Mukesh Kumar Chaurasiya (IBM)" Date: Sun, 26 Apr 2026 15:17:25 +0530 Subject: [PATCH 025/525] rust/drm: import ARef from sync crate ARef is defined in sync and is getting used from types causing the build to fail. Fix this by using ARef from sync module. Fixes: 80df573af9ef ("rust: drm: gem: shmem: Add DRM shmem helper abstraction") Signed-off-by: Mukesh Kumar Chaurasiya (IBM) Link: https://patch.msgid.link/20260426094725.2188668-2-mkchauras@gmail.com [ Add missing Fixes: tag. - Danilo ] Signed-off-by: Danilo Krummrich --- rust/kernel/drm/gem/shmem.rs | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/rust/kernel/drm/gem/shmem.rs b/rust/kernel/drm/gem/shmem.rs index d025fb035195..e1b648920d2f 100644 --- a/rust/kernel/drm/gem/shmem.rs +++ b/rust/kernel/drm/gem/shmem.rs @@ -19,10 +19,8 @@ }, error::to_result, prelude::*, - types::{ - ARef, - Opaque, // - }, // + sync::aref::ARef, + types::Opaque, // }; use core::{ ops::{ From 15e8bae5d930c91b8739a87d75db0a6efca3cb32 Mon Sep 17 00:00:00 2001 From: Danilo Krummrich Date: Thu, 23 Apr 2026 14:46:35 +0200 Subject: [PATCH 026/525] MAINTAINERS: nova: update mailing list The nouveau mailing list has some issues (e.g. with stripping Cc entries from replies when using notmuch + b4 based workflows). Besides that, having a separate mailing list for nova also helps to better distinguish nova from nouveau and makes it easier to track nova-specific discussions. Replace the nouveau mailing list with the new nova-gpu@lists.linux.dev mailing list for both nova-core and nova-drm, and remove the patchwork entries, since those are bound to the nouveau mailing list and not used by nova anyway. Link: https://lore.kernel.org/all/bc2517c2-6772-4cbd-8fd7-6dbdcdd13eab@nvidia.com/ Reviewed-by: Joel Fernandes Reviewed-by: John Hubbard Acked-by: Alexandre Courbot Link: https://patch.msgid.link/20260423124649.38793-1-dakr@kernel.org Signed-off-by: Danilo Krummrich --- MAINTAINERS | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index 2fb1c75afd16..5c9272622033 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -8193,10 +8193,9 @@ F: include/uapi/drm/nouveau_drm.h CORE DRIVER FOR NVIDIA GPUS [RUST] M: Danilo Krummrich M: Alexandre Courbot -L: nouveau@lists.freedesktop.org +L: nova-gpu@lists.linux.dev S: Supported W: https://rust-for-linux.com/nova-gpu-driver -Q: https://patchwork.freedesktop.org/project/nouveau/ B: https://gitlab.freedesktop.org/drm/nova/-/issues C: irc://irc.oftc.net/nouveau T: git https://gitlab.freedesktop.org/drm/rust/kernel.git drm-rust-next @@ -8205,10 +8204,9 @@ F: drivers/gpu/nova-core/ DRM DRIVER FOR NVIDIA GPUS [RUST] M: Danilo Krummrich -L: nouveau@lists.freedesktop.org +L: nova-gpu@lists.linux.dev S: Supported W: https://rust-for-linux.com/nova-gpu-driver -Q: https://patchwork.freedesktop.org/project/nouveau/ B: https://gitlab.freedesktop.org/drm/nova/-/issues C: irc://irc.oftc.net/nouveau T: git https://gitlab.freedesktop.org/drm/rust/kernel.git drm-rust-next From aa23c94cc433b145d1ce93820ecdfe16d8940e28 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 Mar 2026 12:08:21 +0100 Subject: [PATCH 027/525] media: venus: fix QCOM_MDT_LOADER dependency When build-testined with CONFIG_QCOM_MDT_LOADER=m and VIDEO_QCOM_VENUS=y, the kernel fails to link: x86_64-linux-ld: drivers/media/platform/qcom/venus/firmware.o: in function `venus_boot': firmware.c:(.text+0x1e3): undefined reference to `qcom_mdt_get_size' firmware.c:(.text+0x25a): undefined reference to `qcom_mdt_load' firmware.c:(.text+0x272): undefined reference to `qcom_mdt_load_no_init' The problem is the conditional 'select' statement. Change this to make the driver built-in here regardless of CONFIG_ARCH_QCOM, same as for the similar IRIS driver. Signed-off-by: Arnd Bergmann Reviewed-by: Konrad Dybcio Reviewed-by: Dikshita Agarwal Fixes: 0399b696f7f4 ("media: venus: fix compile-test build on non-qcom ARM platform") Cc: stable@vger.kernel.org Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/venus/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/media/platform/qcom/venus/Kconfig b/drivers/media/platform/qcom/venus/Kconfig index ffb731ecd48c..63ee8c78dc6d 100644 --- a/drivers/media/platform/qcom/venus/Kconfig +++ b/drivers/media/platform/qcom/venus/Kconfig @@ -4,7 +4,7 @@ config VIDEO_QCOM_VENUS depends on VIDEO_DEV && QCOM_SMEM depends on (ARCH_QCOM && ARM64 && IOMMU_API) || COMPILE_TEST select OF_DYNAMIC if ARCH_QCOM - select QCOM_MDT_LOADER if ARCH_QCOM + select QCOM_MDT_LOADER select QCOM_SCM select VIDEOBUF2_DMA_CONTIG select V4L2_MEM2MEM_DEV From a297c5165f91366cbc3490e630aabd1c0f70efb8 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Thu, 5 Feb 2026 15:56:19 +0100 Subject: [PATCH 028/525] media: iris: fix QCOM_MDT_LOADER dependency When build-testined with CONFIG_QCOM_MDT_LOADER=m and VIDEO_QCOM_IRIS=y, the kernel fails to link: x86_64-linux-ld: drivers/media/platform/qcom/iris/iris_firmware.o: in function `iris_fw_load': iris_firmware.c:(.text+0xb0): undefined reference to `qcom_mdt_get_size' iris_firmware.c:(.text+0xfd): undefined reference to `qcom_mdt_load' The problem is the conditional 'select' statement. Change this to make the driver built-in here regardless of CONFIG_ARCH_QCOM. Signed-off-by: Arnd Bergmann Reviewed-by: Konrad Dybcio Reviewed-by: Dikshita Agarwal Reviewed-by: Bryan O'Donoghue Fixes: d19b163356b8 ("media: iris: implement video firmware load/unload") Cc: stable@vger.kernel.org Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/iris/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/media/platform/qcom/iris/Kconfig b/drivers/media/platform/qcom/iris/Kconfig index 3c803a05305a..5498f48362d1 100644 --- a/drivers/media/platform/qcom/iris/Kconfig +++ b/drivers/media/platform/qcom/iris/Kconfig @@ -3,7 +3,7 @@ config VIDEO_QCOM_IRIS depends on VIDEO_DEV depends on ARCH_QCOM || COMPILE_TEST select V4L2_MEM2MEM_DEV - select QCOM_MDT_LOADER if ARCH_QCOM + select QCOM_MDT_LOADER select QCOM_SCM select VIDEOBUF2_DMA_CONTIG help From f27cfdcfc916bb59297825805f4c3499f89f9e76 Mon Sep 17 00:00:00 2001 From: Dikshita Agarwal Date: Mon, 16 Feb 2026 12:37:42 +0530 Subject: [PATCH 029/525] media: iris: Fix use-after-free in iris_release_internal_buffers() The recent change in commit 1dabf00ee206 ("media: iris: gen1: Destroy internal buffers after FW releases") introduced a regression where session_release_buf() may free the buffer. The caller, iris_release_internal_buffers(), continued to access `buffer` after the call, leading to a potential use-after-free. Fix this by setting BUF_ATTR_PENDING_RELEASE before calling session_release_buf(), and reverting the flag if the call fails. This ensures no dereference occurs after potential freeing. Reported-by: Dan Carpenter Closes: https://lore.kernel.org/lkml/aYXvKAX3Pg3sL37P@stanley.mountain/#r Signed-off-by: Dikshita Agarwal Reviewed-by: Vikash Garodia Fixes: 1dabf00ee206 ("media: iris: gen1: Destroy internal buffers after FW releases") Cc: stable@vger.kernel.org Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/iris/iris_buffer.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/media/platform/qcom/iris/iris_buffer.c b/drivers/media/platform/qcom/iris/iris_buffer.c index 9151f43bc6b9..1d53c7414b75 100644 --- a/drivers/media/platform/qcom/iris/iris_buffer.c +++ b/drivers/media/platform/qcom/iris/iris_buffer.c @@ -582,10 +582,12 @@ static int iris_release_internal_buffers(struct iris_inst *inst, continue; if (!(buffer->attr & BUF_ATTR_QUEUED)) continue; - ret = hfi_ops->session_release_buf(inst, buffer); - if (ret) - return ret; buffer->attr |= BUF_ATTR_PENDING_RELEASE; + ret = hfi_ops->session_release_buf(inst, buffer); + if (ret) { + buffer->attr &= ~BUF_ATTR_PENDING_RELEASE; + return ret; + } } return 0; From 4a49ae56b0e4268d48fd96babe0cc68596bc301a Mon Sep 17 00:00:00 2001 From: Thomas Fourier Date: Fri, 13 Feb 2026 10:13:27 +0100 Subject: [PATCH 030/525] media: iris: Fix dma_free_attrs() size in iris_hfi_queues_init() The core->iface_q_table_vaddr buffer is alloc'd with size queue_size but freed with sizeof(*q_tbl_hdr) which is different. Change the dma_free_attrs() size. Signed-off-by: Thomas Fourier Reviewed-by: Dikshita Agarwal Fixes: d7378f84e94e ("media: iris: introduce iris core state management with shared queues") Cc: stable@vger.kernel.org Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/iris/iris_hfi_queue.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/media/platform/qcom/iris/iris_hfi_queue.c b/drivers/media/platform/qcom/iris/iris_hfi_queue.c index b3ed06297953..bf6db23b53e2 100644 --- a/drivers/media/platform/qcom/iris/iris_hfi_queue.c +++ b/drivers/media/platform/qcom/iris/iris_hfi_queue.c @@ -263,7 +263,7 @@ int iris_hfi_queues_init(struct iris_core *core) GFP_KERNEL, DMA_ATTR_WRITE_COMBINE); if (!core->sfr_vaddr) { dev_err(core->dev, "sfr alloc and map failed\n"); - dma_free_attrs(core->dev, sizeof(*q_tbl_hdr), core->iface_q_table_vaddr, + dma_free_attrs(core->dev, queue_size, core->iface_q_table_vaddr, core->iface_q_table_daddr, DMA_ATTR_WRITE_COMBINE); return -ENOMEM; } From 95a337f92f0a602d4f935315bfbc8bf07f475e65 Mon Sep 17 00:00:00 2001 From: Vikash Garodia Date: Fri, 13 Mar 2026 18:49:36 +0530 Subject: [PATCH 031/525] media: iris: switch to hardware mode after firmware boot Currently the driver switches the vcodec GDSC to hardware (HW) mode before firmware load and boot sequence. GDSC can be powered off, keeping in hw mode, thereby the vcodec registers programmed in TrustZone (TZ) carry default (reset) values. Move the transition to HW mode after firmware load and boot sequence. The bug was exposed with driver configuring different stream ids to different devices via iommu-map. With registers carrying reset values, VPU would not generate desired stream-id, thereby leading to SMMU fault. For vpu4, when GDSC is switched to HW mode, there is a need to perform the reset operation. Without reset, there are occasional issues of register corruption observed. Hence the vpu GDSC switch also involves the reset. Co-developed-by: Vishnu Reddy Signed-off-by: Vishnu Reddy Signed-off-by: Vikash Garodia Reviewed-by: Dikshita Agarwal Reviewed-by: Dmitry Baryshkov [bod: occassional => occasional] Fixes: dde659d37036 ("media: iris: Introduce vpu ops for vpu4 with necessary hooks") Cc: stable@vger.kernel.org Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/iris/iris_core.c | 4 ++++ .../platform/qcom/iris/iris_hfi_common.c | 4 ++++ drivers/media/platform/qcom/iris/iris_vpu2.c | 1 + drivers/media/platform/qcom/iris/iris_vpu3x.c | 9 +++---- drivers/media/platform/qcom/iris/iris_vpu4x.c | 24 ++++++++++--------- .../platform/qcom/iris/iris_vpu_common.c | 16 ++++++++----- .../platform/qcom/iris/iris_vpu_common.h | 3 +++ 7 files changed, 38 insertions(+), 23 deletions(-) diff --git a/drivers/media/platform/qcom/iris/iris_core.c b/drivers/media/platform/qcom/iris/iris_core.c index 8406c48d635b..dbaac01eb15a 100644 --- a/drivers/media/platform/qcom/iris/iris_core.c +++ b/drivers/media/platform/qcom/iris/iris_core.c @@ -75,6 +75,10 @@ int iris_core_init(struct iris_core *core) if (ret) goto error_unload_fw; + ret = iris_vpu_switch_to_hwmode(core); + if (ret) + goto error_unload_fw; + ret = iris_hfi_core_init(core); if (ret) goto error_unload_fw; diff --git a/drivers/media/platform/qcom/iris/iris_hfi_common.c b/drivers/media/platform/qcom/iris/iris_hfi_common.c index 92112eb16c11..621c66593d88 100644 --- a/drivers/media/platform/qcom/iris/iris_hfi_common.c +++ b/drivers/media/platform/qcom/iris/iris_hfi_common.c @@ -159,6 +159,10 @@ int iris_hfi_pm_resume(struct iris_core *core) if (ret) goto err_suspend_hw; + ret = iris_vpu_switch_to_hwmode(core); + if (ret) + goto err_suspend_hw; + ret = ops->sys_interframe_powercollapse(core); if (ret) goto err_suspend_hw; diff --git a/drivers/media/platform/qcom/iris/iris_vpu2.c b/drivers/media/platform/qcom/iris/iris_vpu2.c index 9c103a2e4e4e..01ef40f38957 100644 --- a/drivers/media/platform/qcom/iris/iris_vpu2.c +++ b/drivers/media/platform/qcom/iris/iris_vpu2.c @@ -44,4 +44,5 @@ const struct vpu_ops iris_vpu2_ops = { .power_off_controller = iris_vpu_power_off_controller, .power_on_controller = iris_vpu_power_on_controller, .calc_freq = iris_vpu2_calc_freq, + .set_hwmode = iris_vpu_set_hwmode, }; diff --git a/drivers/media/platform/qcom/iris/iris_vpu3x.c b/drivers/media/platform/qcom/iris/iris_vpu3x.c index fe4423b951b1..3dad47be78b5 100644 --- a/drivers/media/platform/qcom/iris/iris_vpu3x.c +++ b/drivers/media/platform/qcom/iris/iris_vpu3x.c @@ -234,14 +234,8 @@ static int iris_vpu35_power_on_hw(struct iris_core *core) if (ret) goto err_disable_hw_free_clk; - ret = dev_pm_genpd_set_hwmode(core->pmdomain_tbl->pd_devs[IRIS_HW_POWER_DOMAIN], true); - if (ret) - goto err_disable_hw_clk; - return 0; -err_disable_hw_clk: - iris_disable_unprepare_clock(core, IRIS_HW_CLK); err_disable_hw_free_clk: iris_disable_unprepare_clock(core, IRIS_HW_FREERUN_CLK); err_disable_axi_clk: @@ -266,6 +260,7 @@ const struct vpu_ops iris_vpu3_ops = { .power_off_controller = iris_vpu_power_off_controller, .power_on_controller = iris_vpu_power_on_controller, .calc_freq = iris_vpu3x_vpu4x_calculate_frequency, + .set_hwmode = iris_vpu_set_hwmode, }; const struct vpu_ops iris_vpu33_ops = { @@ -274,6 +269,7 @@ const struct vpu_ops iris_vpu33_ops = { .power_off_controller = iris_vpu33_power_off_controller, .power_on_controller = iris_vpu_power_on_controller, .calc_freq = iris_vpu3x_vpu4x_calculate_frequency, + .set_hwmode = iris_vpu_set_hwmode, }; const struct vpu_ops iris_vpu35_ops = { @@ -283,4 +279,5 @@ const struct vpu_ops iris_vpu35_ops = { .power_on_controller = iris_vpu35_vpu4x_power_on_controller, .program_bootup_registers = iris_vpu35_vpu4x_program_bootup_registers, .calc_freq = iris_vpu3x_vpu4x_calculate_frequency, + .set_hwmode = iris_vpu_set_hwmode, }; diff --git a/drivers/media/platform/qcom/iris/iris_vpu4x.c b/drivers/media/platform/qcom/iris/iris_vpu4x.c index a8db02ce5c5e..02e100a4045f 100644 --- a/drivers/media/platform/qcom/iris/iris_vpu4x.c +++ b/drivers/media/platform/qcom/iris/iris_vpu4x.c @@ -252,21 +252,10 @@ static int iris_vpu4x_power_on_hardware(struct iris_core *core) ret = iris_vpu4x_power_on_apv(core); if (ret) goto disable_hw_clocks; - - iris_vpu4x_ahb_sync_reset_apv(core); } - iris_vpu4x_ahb_sync_reset_hardware(core); - - ret = iris_vpu4x_genpd_set_hwmode(core, true, efuse_value); - if (ret) - goto disable_apv_power_domain; - return 0; -disable_apv_power_domain: - if (!(efuse_value & DISABLE_VIDEO_APV_BIT)) - iris_vpu4x_power_off_apv(core); disable_hw_clocks: iris_vpu4x_disable_hardware_clocks(core, efuse_value); disable_vpp1_power_domain: @@ -359,6 +348,18 @@ static void iris_vpu4x_power_off_hardware(struct iris_core *core) iris_disable_power_domains(core, core->pmdomain_tbl->pd_devs[IRIS_HW_POWER_DOMAIN]); } +static int iris_vpu4x_set_hwmode(struct iris_core *core) +{ + u32 efuse_value = readl(core->reg_base + WRAPPER_EFUSE_MONITOR); + + if (!(efuse_value & DISABLE_VIDEO_APV_BIT)) + iris_vpu4x_ahb_sync_reset_apv(core); + + iris_vpu4x_ahb_sync_reset_hardware(core); + + return iris_vpu4x_genpd_set_hwmode(core, true, efuse_value); +} + const struct vpu_ops iris_vpu4x_ops = { .power_off_hw = iris_vpu4x_power_off_hardware, .power_on_hw = iris_vpu4x_power_on_hardware, @@ -366,4 +367,5 @@ const struct vpu_ops iris_vpu4x_ops = { .power_on_controller = iris_vpu35_vpu4x_power_on_controller, .program_bootup_registers = iris_vpu35_vpu4x_program_bootup_registers, .calc_freq = iris_vpu3x_vpu4x_calculate_frequency, + .set_hwmode = iris_vpu4x_set_hwmode, }; diff --git a/drivers/media/platform/qcom/iris/iris_vpu_common.c b/drivers/media/platform/qcom/iris/iris_vpu_common.c index 548e5f1727fd..69e6126dc4d9 100644 --- a/drivers/media/platform/qcom/iris/iris_vpu_common.c +++ b/drivers/media/platform/qcom/iris/iris_vpu_common.c @@ -292,14 +292,8 @@ int iris_vpu_power_on_hw(struct iris_core *core) if (ret && ret != -ENOENT) goto err_disable_hw_clock; - ret = dev_pm_genpd_set_hwmode(core->pmdomain_tbl->pd_devs[IRIS_HW_POWER_DOMAIN], true); - if (ret) - goto err_disable_hw_ahb_clock; - return 0; -err_disable_hw_ahb_clock: - iris_disable_unprepare_clock(core, IRIS_HW_AHB_CLK); err_disable_hw_clock: iris_disable_unprepare_clock(core, IRIS_HW_CLK); err_disable_power: @@ -308,6 +302,16 @@ int iris_vpu_power_on_hw(struct iris_core *core) return ret; } +int iris_vpu_set_hwmode(struct iris_core *core) +{ + return dev_pm_genpd_set_hwmode(core->pmdomain_tbl->pd_devs[IRIS_HW_POWER_DOMAIN], true); +} + +int iris_vpu_switch_to_hwmode(struct iris_core *core) +{ + return core->iris_platform_data->vpu_ops->set_hwmode(core); +} + int iris_vpu35_vpu4x_power_off_controller(struct iris_core *core) { u32 clk_rst_tbl_size = core->iris_platform_data->clk_rst_tbl_size; diff --git a/drivers/media/platform/qcom/iris/iris_vpu_common.h b/drivers/media/platform/qcom/iris/iris_vpu_common.h index f6dffc613b82..dee3b1349c5e 100644 --- a/drivers/media/platform/qcom/iris/iris_vpu_common.h +++ b/drivers/media/platform/qcom/iris/iris_vpu_common.h @@ -21,6 +21,7 @@ struct vpu_ops { int (*power_on_controller)(struct iris_core *core); void (*program_bootup_registers)(struct iris_core *core); u64 (*calc_freq)(struct iris_inst *inst, size_t data_size); + int (*set_hwmode)(struct iris_core *core); }; int iris_vpu_boot_firmware(struct iris_core *core); @@ -30,6 +31,8 @@ int iris_vpu_watchdog(struct iris_core *core, u32 intr_status); int iris_vpu_prepare_pc(struct iris_core *core); int iris_vpu_power_on_controller(struct iris_core *core); int iris_vpu_power_on_hw(struct iris_core *core); +int iris_vpu_set_hwmode(struct iris_core *core); +int iris_vpu_switch_to_hwmode(struct iris_core *core); int iris_vpu_power_on(struct iris_core *core); int iris_vpu_power_off_controller(struct iris_core *core); void iris_vpu_power_off_hw(struct iris_core *core); From 3d9593ad1a58c5acc3e5fa2a48222bb7632e6812 Mon Sep 17 00:00:00 2001 From: Vishnu Reddy Date: Thu, 5 Mar 2026 18:58:31 +0530 Subject: [PATCH 032/525] media: iris: fix use-after-free of fmt_src during MBPF check During concurrency testing, multiple instances can run in parallel, and each instance uses its own inst->lock while the core->lock protects the list of active instances. The race happens because these locks cover different scopes, inst->lock protects only the internals of a single instance, while the Macro Blocks Per Frame (MBPF) checker walks the core list under core->lock and reads fields like fmt_src->width and fmt_src->height. At the same time, iris_close() may free fmt_src and fmt_dst under inst->lock while the instance is still present in the core list. This allows a situation where the MBPF checker, still iterating through the core list, reaches an instance whose fmt_src was already freed by another thread and ends up dereferencing a dangling pointer, resulting in a use-after-free. This happens because the MBPF checker assumes that any instance in the core list is fully valid, but the freeing of fmt_src and fmt_dst without removing the instance from the core list is not correct. The correct ordering is to defer freeing fmt_src and fmt_dst until after the instance has been removed from the core list and all teardown under the core lock has completed, ensuring that no dangling pointers are ever exposed during MBPF checks. Reviewed-by: Vikash Garodia Signed-off-by: Vishnu Reddy Reviewed-by: Dikshita Agarwal Fixes: 5ad964ad5656 ("media: iris: Initialize and deinitialize encoder instance structure") Cc: stable@vger.kernel.org Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/iris/iris_vdec.c | 6 ------ drivers/media/platform/qcom/iris/iris_vdec.h | 1 - drivers/media/platform/qcom/iris/iris_venc.c | 6 ------ drivers/media/platform/qcom/iris/iris_venc.h | 1 - drivers/media/platform/qcom/iris/iris_vidc.c | 6 ++---- 5 files changed, 2 insertions(+), 18 deletions(-) diff --git a/drivers/media/platform/qcom/iris/iris_vdec.c b/drivers/media/platform/qcom/iris/iris_vdec.c index 719217399a30..99d544e2af4f 100644 --- a/drivers/media/platform/qcom/iris/iris_vdec.c +++ b/drivers/media/platform/qcom/iris/iris_vdec.c @@ -61,12 +61,6 @@ int iris_vdec_inst_init(struct iris_inst *inst) return iris_ctrls_init(inst); } -void iris_vdec_inst_deinit(struct iris_inst *inst) -{ - kfree(inst->fmt_dst); - kfree(inst->fmt_src); -} - static const struct iris_fmt iris_vdec_formats_cap[] = { [IRIS_FMT_NV12] = { .pixfmt = V4L2_PIX_FMT_NV12, diff --git a/drivers/media/platform/qcom/iris/iris_vdec.h b/drivers/media/platform/qcom/iris/iris_vdec.h index ec1ce55d1375..5123d2a340e1 100644 --- a/drivers/media/platform/qcom/iris/iris_vdec.h +++ b/drivers/media/platform/qcom/iris/iris_vdec.h @@ -9,7 +9,6 @@ struct iris_inst; int iris_vdec_inst_init(struct iris_inst *inst); -void iris_vdec_inst_deinit(struct iris_inst *inst); int iris_vdec_enum_fmt(struct iris_inst *inst, struct v4l2_fmtdesc *f); int iris_vdec_try_fmt(struct iris_inst *inst, struct v4l2_format *f); int iris_vdec_s_fmt(struct iris_inst *inst, struct v4l2_format *f); diff --git a/drivers/media/platform/qcom/iris/iris_venc.c b/drivers/media/platform/qcom/iris/iris_venc.c index aa27b22704eb..4d886769d958 100644 --- a/drivers/media/platform/qcom/iris/iris_venc.c +++ b/drivers/media/platform/qcom/iris/iris_venc.c @@ -79,12 +79,6 @@ int iris_venc_inst_init(struct iris_inst *inst) return iris_ctrls_init(inst); } -void iris_venc_inst_deinit(struct iris_inst *inst) -{ - kfree(inst->fmt_dst); - kfree(inst->fmt_src); -} - static const struct iris_fmt iris_venc_formats_cap[] = { [IRIS_FMT_H264] = { .pixfmt = V4L2_PIX_FMT_H264, diff --git a/drivers/media/platform/qcom/iris/iris_venc.h b/drivers/media/platform/qcom/iris/iris_venc.h index c4db7433da53..00c1716b2747 100644 --- a/drivers/media/platform/qcom/iris/iris_venc.h +++ b/drivers/media/platform/qcom/iris/iris_venc.h @@ -9,7 +9,6 @@ struct iris_inst; int iris_venc_inst_init(struct iris_inst *inst); -void iris_venc_inst_deinit(struct iris_inst *inst); int iris_venc_enum_fmt(struct iris_inst *inst, struct v4l2_fmtdesc *f); int iris_venc_try_fmt(struct iris_inst *inst, struct v4l2_format *f); int iris_venc_s_fmt(struct iris_inst *inst, struct v4l2_format *f); diff --git a/drivers/media/platform/qcom/iris/iris_vidc.c b/drivers/media/platform/qcom/iris/iris_vidc.c index bd38d84c9cc7..5eb1786b0737 100644 --- a/drivers/media/platform/qcom/iris/iris_vidc.c +++ b/drivers/media/platform/qcom/iris/iris_vidc.c @@ -289,10 +289,6 @@ int iris_close(struct file *filp) v4l2_m2m_ctx_release(inst->m2m_ctx); v4l2_m2m_release(inst->m2m_dev); mutex_lock(&inst->lock); - if (inst->domain == DECODER) - iris_vdec_inst_deinit(inst); - else if (inst->domain == ENCODER) - iris_venc_inst_deinit(inst); iris_session_close(inst); iris_inst_change_state(inst, IRIS_INST_DEINIT); iris_v4l2_fh_deinit(inst, filp); @@ -304,6 +300,8 @@ int iris_close(struct file *filp) mutex_unlock(&inst->lock); mutex_destroy(&inst->ctx_q_lock); mutex_destroy(&inst->lock); + kfree(inst->fmt_src); + kfree(inst->fmt_dst); kfree(inst); return 0; From 3e0b2053751657ed2924adfe3ff25b1450231e33 Mon Sep 17 00:00:00 2001 From: Dmitry Baryshkov Date: Fri, 27 Mar 2026 22:19:55 +0200 Subject: [PATCH 033/525] media: qcom: iris: increase H265D_MAX_SLICE to fix H.265 decoding on SC7280 Follow the commit bfe1326573ff ("venus: Fix for H265 decoding failure.") and increase H265D_MAX_SLICE following firmware requirements on that platform. Otherwise decoding of the H.265 streams fails with the "insufficient scratch_1 buffer size" from the firmware. Signed-off-by: Dmitry Baryshkov Reviewed-by: Dikshita Agarwal Reviewed-by: Vikash Garodia Reviewed-by: Konrad Dybcio [bod: Fixed commit log withthe => with the] Fixes: e1f5d32608ec ("media: iris: Add internal buffer calculation for HEVC and VP9 decoders") Cc: stable@vger.kernel.org Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/iris/iris_vpu_buffer.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/media/platform/qcom/iris/iris_vpu_buffer.h b/drivers/media/platform/qcom/iris/iris_vpu_buffer.h index 12640eb5ed8c..8c0d6b7b5de8 100644 --- a/drivers/media/platform/qcom/iris/iris_vpu_buffer.h +++ b/drivers/media/platform/qcom/iris/iris_vpu_buffer.h @@ -67,7 +67,7 @@ struct iris_inst; #define SIZE_DOLBY_RPU_METADATA (41 * 1024) #define H264_CABAC_HDR_RATIO_HD_TOT 1 #define H264_CABAC_RES_RATIO_HD_TOT 3 -#define H265D_MAX_SLICE 1200 +#define H265D_MAX_SLICE 3600 #define SIZE_H265D_HW_PIC_T SIZE_H264D_HW_PIC_T #define H265_CABAC_HDR_RATIO_HD_TOT 2 #define H265_CABAC_RES_RATIO_HD_TOT 2 From dd1b373941079cc102cc18bc68884e18245f5912 Mon Sep 17 00:00:00 2001 From: Wenmeng Liu Date: Fri, 13 Mar 2026 18:13:02 +0800 Subject: [PATCH 034/525] media: qcom: camss: Fix csid IRQ offset for sa8775p Fix BUF_DONE_IRQ_STATUS_RDI_OFFSET calculation for csid lite on sa8775p platform. The offset should be 0 for csid lite on sa8775p, Signed-off-by: Wenmeng Liu Reviewed-by: Bryan O'Donoghue Fixes: ed03e99de0fa ("media: qcom: camss: Add support for CSID 690") Cc: stable@vger.kernel.org Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/camss/camss-csid-gen3.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/media/platform/qcom/camss/camss-csid-gen3.c b/drivers/media/platform/qcom/camss/camss-csid-gen3.c index 664245cf6eb0..bd059243790e 100644 --- a/drivers/media/platform/qcom/camss/camss-csid-gen3.c +++ b/drivers/media/platform/qcom/camss/camss-csid-gen3.c @@ -48,9 +48,9 @@ #define IS_CSID_690(csid) ((csid->camss->res->version == CAMSS_8775P) \ || (csid->camss->res->version == CAMSS_8300)) #define CSID_BUF_DONE_IRQ_STATUS 0x8C -#define BUF_DONE_IRQ_STATUS_RDI_OFFSET (csid_is_lite(csid) ?\ - 1 : (IS_CSID_690(csid) ?\ - 13 : 14)) +#define BUF_DONE_IRQ_STATUS_RDI_OFFSET (csid_is_lite(csid) ? \ + ((IS_CSID_690(csid) ? 0 : 1)) : \ + ((IS_CSID_690(csid) ? 13 : 14))) #define CSID_BUF_DONE_IRQ_MASK 0x90 #define CSID_BUF_DONE_IRQ_CLEAR 0x94 #define CSID_BUF_DONE_IRQ_SET 0x98 From fe56c674118aa46da1a3e65aa22ca709ebd7d812 Mon Sep 17 00:00:00 2001 From: Wenmeng Liu Date: Fri, 13 Mar 2026 18:13:03 +0800 Subject: [PATCH 035/525] media: qcom: camss: Fix csid clock configuration for sa8775p Fix the mismatch between clock list and clock rate table for CSID lite instances. The current implementation has 5 clocks defined but only 2 are actually needed (vfe_lite_csid and vfe_lite_cphy_rx), while the clock rate table doesn't match this configuration. Update both clock list and rate table to maintain consistency: - Remove unused clocks: cpas_vfe_lite, vfe_lite_ahb, vfe_lite - Update clock rate table to match the remaining two clocks Signed-off-by: Wenmeng Liu Reviewed-by: Bryan O'Donoghue Fixes: ed03e99de0fa ("media: qcom: camss: Add support for CSID 690") Cc: stable@vger.kernel.org Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/camss/camss.c | 40 +++++++++-------------- 1 file changed, 15 insertions(+), 25 deletions(-) diff --git a/drivers/media/platform/qcom/camss/camss.c b/drivers/media/platform/qcom/camss/camss.c index 00b87fd9afbd..cb0134718985 100644 --- a/drivers/media/platform/qcom/camss/camss.c +++ b/drivers/media/platform/qcom/camss/camss.c @@ -3598,12 +3598,10 @@ static const struct camss_subdev_resources csid_res_8775p[] = { /* CSID2 (lite) */ { .regulators = {}, - .clock = { "cpas_vfe_lite", "vfe_lite_ahb", - "vfe_lite_csid", "vfe_lite_cphy_rx", - "vfe_lite"}, + .clock = { "vfe_lite_csid", "vfe_lite_cphy_rx" }, .clock_rate = { - { 0, 0, 400000000, 400000000, 0}, - { 0, 0, 400000000, 480000000, 0} + { 400000000, 480000000 }, + { 400000000, 480000000 } }, .reg = { "csid_lite0" }, .interrupt = { "csid_lite0" }, @@ -3617,12 +3615,10 @@ static const struct camss_subdev_resources csid_res_8775p[] = { /* CSID3 (lite) */ { .regulators = {}, - .clock = { "cpas_vfe_lite", "vfe_lite_ahb", - "vfe_lite_csid", "vfe_lite_cphy_rx", - "vfe_lite"}, + .clock = { "vfe_lite_csid", "vfe_lite_cphy_rx" }, .clock_rate = { - { 0, 0, 400000000, 400000000, 0}, - { 0, 0, 400000000, 480000000, 0} + { 400000000, 480000000 }, + { 400000000, 480000000 } }, .reg = { "csid_lite1" }, .interrupt = { "csid_lite1" }, @@ -3636,12 +3632,10 @@ static const struct camss_subdev_resources csid_res_8775p[] = { /* CSID4 (lite) */ { .regulators = {}, - .clock = { "cpas_vfe_lite", "vfe_lite_ahb", - "vfe_lite_csid", "vfe_lite_cphy_rx", - "vfe_lite"}, + .clock = { "vfe_lite_csid", "vfe_lite_cphy_rx" }, .clock_rate = { - { 0, 0, 400000000, 400000000, 0}, - { 0, 0, 400000000, 480000000, 0} + { 400000000, 480000000 }, + { 400000000, 480000000 } }, .reg = { "csid_lite2" }, .interrupt = { "csid_lite2" }, @@ -3655,12 +3649,10 @@ static const struct camss_subdev_resources csid_res_8775p[] = { /* CSID5 (lite) */ { .regulators = {}, - .clock = { "cpas_vfe_lite", "vfe_lite_ahb", - "vfe_lite_csid", "vfe_lite_cphy_rx", - "vfe_lite"}, + .clock = { "vfe_lite_csid", "vfe_lite_cphy_rx" }, .clock_rate = { - { 0, 0, 400000000, 400000000, 0}, - { 0, 0, 400000000, 480000000, 0} + { 400000000, 480000000 }, + { 400000000, 480000000 } }, .reg = { "csid_lite3" }, .interrupt = { "csid_lite3" }, @@ -3674,12 +3666,10 @@ static const struct camss_subdev_resources csid_res_8775p[] = { /* CSID6 (lite) */ { .regulators = {}, - .clock = { "cpas_vfe_lite", "vfe_lite_ahb", - "vfe_lite_csid", "vfe_lite_cphy_rx", - "vfe_lite"}, + .clock = { "vfe_lite_csid", "vfe_lite_cphy_rx" }, .clock_rate = { - { 0, 0, 400000000, 400000000, 0}, - { 0, 0, 400000000, 480000000, 0} + { 400000000, 480000000 }, + { 400000000, 480000000 } }, .reg = { "csid_lite4" }, .interrupt = { "csid_lite4" }, From d31fac47b39f5e1ed85a587688ca70b793e421b4 Mon Sep 17 00:00:00 2001 From: Wenmeng Liu Date: Fri, 13 Mar 2026 18:13:04 +0800 Subject: [PATCH 036/525] media: qcom: camss: Add missing clocks for VFE lite on sa8775p Add missing required clocks (cpas_ahb and camnoc_axi) for VFE lite instances on sa8775p platform. These clocks are necessary for proper VFE lite operation: Reviewed-by: Bryan O'Donoghue Signed-off-by: Wenmeng Liu Fixes: e7b59e1d06fb ("media: qcom: camss: Add support for VFE 690") Cc: stable@vger.kernel.org Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/camss/camss.c | 40 ++++++++++++++--------- 1 file changed, 25 insertions(+), 15 deletions(-) diff --git a/drivers/media/platform/qcom/camss/camss.c b/drivers/media/platform/qcom/camss/camss.c index cb0134718985..9335636d7c4d 100644 --- a/drivers/media/platform/qcom/camss/camss.c +++ b/drivers/media/platform/qcom/camss/camss.c @@ -3742,15 +3742,17 @@ static const struct camss_subdev_resources vfe_res_8775p[] = { /* VFE2 (lite) */ { .regulators = {}, - .clock = { "cpas_vfe_lite", "vfe_lite_ahb", + .clock = { "cpas_ahb", "cpas_vfe_lite", "vfe_lite_ahb", "vfe_lite_csid", "vfe_lite_cphy_rx", - "vfe_lite"}, + "vfe_lite", "camnoc_axi"}, .clock_rate = { - { 0, 0, 0, 0 }, + { 0 }, + { 0 }, { 300000000, 400000000, 400000000, 400000000 }, { 400000000, 400000000, 400000000, 400000000 }, { 400000000, 400000000, 400000000, 400000000 }, { 480000000, 600000000, 600000000, 600000000 }, + { 400000000 }, }, .reg = { "vfe_lite0" }, .interrupt = { "vfe_lite0" }, @@ -3765,15 +3767,17 @@ static const struct camss_subdev_resources vfe_res_8775p[] = { /* VFE3 (lite) */ { .regulators = {}, - .clock = { "cpas_vfe_lite", "vfe_lite_ahb", + .clock = { "cpas_ahb", "cpas_vfe_lite", "vfe_lite_ahb", "vfe_lite_csid", "vfe_lite_cphy_rx", - "vfe_lite"}, + "vfe_lite", "camnoc_axi"}, .clock_rate = { - { 0, 0, 0, 0 }, + { 0 }, + { 0 }, { 300000000, 400000000, 400000000, 400000000 }, { 400000000, 400000000, 400000000, 400000000 }, { 400000000, 400000000, 400000000, 400000000 }, { 480000000, 600000000, 600000000, 600000000 }, + { 400000000 }, }, .reg = { "vfe_lite1" }, .interrupt = { "vfe_lite1" }, @@ -3788,15 +3792,17 @@ static const struct camss_subdev_resources vfe_res_8775p[] = { /* VFE4 (lite) */ { .regulators = {}, - .clock = { "cpas_vfe_lite", "vfe_lite_ahb", + .clock = { "cpas_ahb", "cpas_vfe_lite", "vfe_lite_ahb", "vfe_lite_csid", "vfe_lite_cphy_rx", - "vfe_lite"}, + "vfe_lite", "camnoc_axi"}, .clock_rate = { - { 0, 0, 0, 0 }, + { 0 }, + { 0 }, { 300000000, 400000000, 400000000, 400000000 }, { 400000000, 400000000, 400000000, 400000000 }, { 400000000, 400000000, 400000000, 400000000 }, { 480000000, 600000000, 600000000, 600000000 }, + { 400000000 }, }, .reg = { "vfe_lite2" }, .interrupt = { "vfe_lite2" }, @@ -3811,15 +3817,17 @@ static const struct camss_subdev_resources vfe_res_8775p[] = { /* VFE5 (lite) */ { .regulators = {}, - .clock = { "cpas_vfe_lite", "vfe_lite_ahb", + .clock = { "cpas_ahb", "cpas_vfe_lite", "vfe_lite_ahb", "vfe_lite_csid", "vfe_lite_cphy_rx", - "vfe_lite"}, + "vfe_lite", "camnoc_axi"}, .clock_rate = { - { 0, 0, 0, 0 }, + { 0 }, + { 0 }, { 300000000, 400000000, 400000000, 400000000 }, { 400000000, 400000000, 400000000, 400000000 }, { 400000000, 400000000, 400000000, 400000000 }, { 480000000, 600000000, 600000000, 600000000 }, + { 400000000 }, }, .reg = { "vfe_lite3" }, .interrupt = { "vfe_lite3" }, @@ -3834,15 +3842,17 @@ static const struct camss_subdev_resources vfe_res_8775p[] = { /* VFE6 (lite) */ { .regulators = {}, - .clock = { "cpas_vfe_lite", "vfe_lite_ahb", + .clock = { "cpas_ahb", "cpas_vfe_lite", "vfe_lite_ahb", "vfe_lite_csid", "vfe_lite_cphy_rx", - "vfe_lite"}, + "vfe_lite", "camnoc_axi"}, .clock_rate = { - { 0, 0, 0, 0 }, + { 0 }, + { 0 }, { 300000000, 400000000, 400000000, 400000000 }, { 400000000, 400000000, 400000000, 400000000 }, { 400000000, 400000000, 400000000, 400000000 }, { 480000000, 600000000, 600000000, 600000000 }, + { 400000000 }, }, .reg = { "vfe_lite4" }, .interrupt = { "vfe_lite4" }, From 23c39cb598977f10909a2387c5e5f34afc1d6933 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Fri, 20 Mar 2026 16:18:24 +0100 Subject: [PATCH 037/525] media: qcom: camss: avoid format string warning clang-22 warns about csiphy_match_clock_name() taking a variable format string that is not checked against the 'int index' argument: drivers/media/platform/qcom/camss/camss-csiphy.c:566:44: error: diagnostic behavior may be improved by adding the 'format(printf, 2, 3)' attribute to the declaration of 'csiphy_match_clock_name' [-Werror,-Wmissing-format-attribute] 561 | static bool csiphy_match_clock_name(const char *clock_name, const char *format, | __attribute__((format(printf, 2, 3))) 562 | int index) 563 | { 564 | char name[16]; /* csiphyXXX_timer\0 */ 565 | 566 | snprintf(name, sizeof(name), format, index); | ^ drivers/media/platform/qcom/camss/camss-csiphy.c:561:13: note: 'csiphy_match_clock_name' declared here 561 | static bool csiphy_match_clock_name(const char *clock_name, const char *format, | ^ Change the function to use a snprintf() style format string that allows this to be checked at the call site. Signed-off-by: Arnd Bergmann Reviewed-by: Bryan O'Donoghue Signed-off-by: Bryan O'Donoghue Signed-off-by: Hans Verkuil --- drivers/media/platform/qcom/camss/camss-csiphy.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/media/platform/qcom/camss/camss-csiphy.c b/drivers/media/platform/qcom/camss/camss-csiphy.c index 62623393f414..78a1b568dbae 100644 --- a/drivers/media/platform/qcom/camss/camss-csiphy.c +++ b/drivers/media/platform/qcom/camss/camss-csiphy.c @@ -558,12 +558,16 @@ static int csiphy_init_formats(struct v4l2_subdev *sd, return csiphy_set_format(sd, fh ? fh->state : NULL, &format); } -static bool csiphy_match_clock_name(const char *clock_name, const char *format, - int index) +static bool __printf(2, 3) +csiphy_match_clock_name(const char *clock_name, const char *format, ...) { char name[16]; /* csiphyXXX_timer\0 */ + va_list args; + + va_start(args, format); + vsnprintf(name, sizeof(name), format, args); + va_end(args); - snprintf(name, sizeof(name), format, index); return !strcmp(clock_name, name); } From 711a9c018ad252b2807f85d44e1267b595644f9b Mon Sep 17 00:00:00 2001 From: Rio Liu Date: Wed, 15 Apr 2026 16:57:13 +0000 Subject: [PATCH 038/525] wifi: mac80211: skip ieee80211_verify_sta_ht_mcs_support check in non-strict mode Some Xfinity XB8 firmware advertises >1 spatial stream MCS indexes in their basic HT-MCS set. On cards with lower spatial streams, the check would fail, and we'd be stuck with no HT when in fact work fine with its own supported rate. This change makes it so the check is only performed in strict mode. Fixes: 574faa0e936d ("wifi: mac80211: add HT and VHT basic set verification") Signed-off-by: Rio Liu Link: https://patch.msgid.link/99Mv9QEceyPrQhSP52MtAVmz0_kWJmzqotJjD9YW6LGLqk-AZloAueUyHCURilFkuqOh6Ecv8i2KKdSE1ujP3AnbU5QEouVisT1w_V3xdfc=@r26.me Signed-off-by: Johannes Berg --- net/mac80211/mlme.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 160ae65a5c64..298ebff6bbf8 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -437,6 +437,15 @@ ieee80211_verify_sta_ht_mcs_support(struct ieee80211_sub_if_data *sdata, memcpy(&sta_ht_cap, &sband->ht_cap, sizeof(sta_ht_cap)); ieee80211_apply_htcap_overrides(sdata, &sta_ht_cap); + /* + * Some Xfinity XB8 firmware advertises >1 spatial stream MCS indexes in + * their basic HT-MCS set. On cards with lower spatial streams, the check + * would fail, and we'd be stuck with no HT when it in fact work fine with + * its own supported rate. So check it only in strict mode. + */ + if (!ieee80211_hw_check(&sdata->local->hw, STRICT)) + return true; + /* * P802.11REVme/D7.0 - 6.5.4.2.4 * ... From c623b63580880cc742255eaed3d79804c1b91143 Mon Sep 17 00:00:00 2001 From: Marek Szyprowski Date: Thu, 16 Apr 2026 11:33:39 +0200 Subject: [PATCH 039/525] wifi: brcmfmac: Fix potential use-after-free issue when stopping watchdog task Watchdog task might end between send_sig() and kthread_stop() calls, what results in the use-after-free issue. Fix this by increasing watchdog task reference count before calling send_sig() and dropping it by switching to kthread_stop_put(). Cc: stable@vger.kernel.org Fixes: 373c83a801f1 ("brcmfmac: stop watchdog before detach and free everything") Fixes: a9ffda88be74 ("brcm80211: fmac: abstract bus_stop interface function pointer") Signed-off-by: Marek Szyprowski Acked-by: Arend van Spriel Link: https://patch.msgid.link/20260416093339.2066829-1-m.szyprowski@samsung.com Signed-off-by: Johannes Berg --- drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c index 30f6fcb68632..8fb595733b9c 100644 --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c @@ -2476,8 +2476,9 @@ static void brcmf_sdio_bus_stop(struct device *dev) brcmf_dbg(TRACE, "Enter\n"); if (bus->watchdog_tsk) { + get_task_struct(bus->watchdog_tsk); send_sig(SIGTERM, bus->watchdog_tsk, 1); - kthread_stop(bus->watchdog_tsk); + kthread_stop_put(bus->watchdog_tsk); bus->watchdog_tsk = NULL; } @@ -4567,8 +4568,9 @@ void brcmf_sdio_remove(struct brcmf_sdio *bus) if (bus) { /* Stop watchdog task */ if (bus->watchdog_tsk) { + get_task_struct(bus->watchdog_tsk); send_sig(SIGTERM, bus->watchdog_tsk, 1); - kthread_stop(bus->watchdog_tsk); + kthread_stop_put(bus->watchdog_tsk); bus->watchdog_tsk = NULL; } From 1f4f78bf8549e6ac4f04fba4176854f3a6e0c332 Mon Sep 17 00:00:00 2001 From: Tristan Madani Date: Fri, 17 Apr 2026 11:11:44 +0000 Subject: [PATCH 040/525] wifi: b43: enforce bounds check on firmware key index in b43_rx() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The firmware-controlled key index in b43_rx() can exceed the dev->key[] array size (58 entries). The existing B43_WARN_ON is non-enforcing in production builds, allowing an out-of-bounds read. Make the B43_WARN_ON check enforcing by dropping the frame when the firmware returns an invalid key index. Suggested-by: Jonas Gorski Acked-by: Michael Büsch Fixes: e4d6b7951812 ("[B43]: add mac80211-based driver for modern BCM43xx devices") Cc: stable@vger.kernel.org Signed-off-by: Tristan Madani Link: https://patch.msgid.link/20260417111145.2694196-1-tristmd@gmail.com Signed-off-by: Johannes Berg --- drivers/net/wireless/broadcom/b43/xmit.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/broadcom/b43/xmit.c b/drivers/net/wireless/broadcom/b43/xmit.c index 7651b1bdb592..f0b082596637 100644 --- a/drivers/net/wireless/broadcom/b43/xmit.c +++ b/drivers/net/wireless/broadcom/b43/xmit.c @@ -702,7 +702,8 @@ void b43_rx(struct b43_wldev *dev, struct sk_buff *skb, const void *_rxhdr) * key index, but the ucode passed it slightly different. */ keyidx = b43_kidx_to_raw(dev, keyidx); - B43_WARN_ON(keyidx >= ARRAY_SIZE(dev->key)); + if (B43_WARN_ON(keyidx >= ARRAY_SIZE(dev->key))) + goto drop; if (dev->key[keyidx].algorithm != B43_SEC_ALGO_NONE) { wlhdr_len = ieee80211_hdrlen(fctl); From a035766f970bde2d4298346a31a80685be5c0205 Mon Sep 17 00:00:00 2001 From: Tristan Madani Date: Fri, 17 Apr 2026 11:11:45 +0000 Subject: [PATCH 041/525] wifi: b43legacy: enforce bounds check on firmware key index in RX path Same fix as b43: the firmware-controlled key index in b43legacy_rx() can exceed dev->max_nr_keys. The existing B43legacy_WARN_ON is non-enforcing in production builds, allowing an out-of-bounds read of dev->key[]. Make the check enforcing by dropping the frame for invalid indices. Fixes: 75388acd0cd8 ("[B43LEGACY]: add mac80211-based driver for legacy BCM43xx devices") Cc: stable@vger.kernel.org Signed-off-by: Tristan Madani Link: https://patch.msgid.link/20260417111145.2694196-2-tristmd@gmail.com Signed-off-by: Johannes Berg --- drivers/net/wireless/broadcom/b43legacy/xmit.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/broadcom/b43legacy/xmit.c b/drivers/net/wireless/broadcom/b43legacy/xmit.c index efd63f4ce74f..ee199d4eaf03 100644 --- a/drivers/net/wireless/broadcom/b43legacy/xmit.c +++ b/drivers/net/wireless/broadcom/b43legacy/xmit.c @@ -476,7 +476,8 @@ void b43legacy_rx(struct b43legacy_wldev *dev, * key index, but the ucode passed it slightly different. */ keyidx = b43legacy_kidx_to_raw(dev, keyidx); - B43legacy_WARN_ON(keyidx >= dev->max_nr_keys); + if (B43legacy_WARN_ON(keyidx >= dev->max_nr_keys)) + goto drop; if (dev->key[keyidx].algorithm != B43legacy_SEC_ALGO_NONE) { /* Remove PROTECTED flag to mark it as decrypted. */ From 3994b4afd521d60e47e012fe2ed7b606aaec370b Mon Sep 17 00:00:00 2001 From: Amir Mohammad Jahangirzad Date: Sat, 18 Apr 2026 04:12:47 +0330 Subject: [PATCH 042/525] wifi: libertas: fix integer underflow in process_cmdrequest() The existing validation only checks if recvlength exceeds LBS_CMD_BUFFER_SIZE, but doesn't check the lower bound. When a USB device sends a response shorter than MESSAGE_HEADER_LEN, the subtraction (recvlength - MESSAGE_HEADER_LEN) wraps to a huge value, causing memcpy to corrupt the heap. Add the same lower bound check that libertas_tf already has. Signed-off-by: Amir Mohammad Jahangirzad Link: https://patch.msgid.link/20260418004247.368944-1-a.jahangirzad@gmail.com Signed-off-by: Johannes Berg --- drivers/net/wireless/marvell/libertas/if_usb.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/marvell/libertas/if_usb.c b/drivers/net/wireless/marvell/libertas/if_usb.c index 4fae0e335136..a00d53350fa9 100644 --- a/drivers/net/wireless/marvell/libertas/if_usb.c +++ b/drivers/net/wireless/marvell/libertas/if_usb.c @@ -633,9 +633,10 @@ static inline void process_cmdrequest(int recvlength, uint8_t *recvbuff, unsigned long flags; u8 i; - if (recvlength > LBS_CMD_BUFFER_SIZE) { + if (recvlength < MESSAGE_HEADER_LEN || + recvlength > LBS_CMD_BUFFER_SIZE) { lbs_deb_usbd(&cardp->udev->dev, - "The receive buffer is too large\n"); + "The receive buffer is invalid: %d\n", recvlength); kfree_skb(skb); return; } From 381cd547bc6e35a610c5dfebe554d891eea40f03 Mon Sep 17 00:00:00 2001 From: Michael Bommarito Date: Tue, 21 Apr 2026 18:45:52 -0400 Subject: [PATCH 043/525] wifi: nl80211: require admin perm on SET_PMK / DEL_PMK NL80211_CMD_SET_PMK and NL80211_CMD_DEL_PMK manage the offloaded 4-way-handshake PMK state used by drivers advertising NL80211_EXT_FEATURE_4WAY_HANDSHAKE_STA_1X. The only in-tree driver that wires up both ->set_pmk / ->del_pmk and advertises the feature today is brcmfmac, so the practical reach of this patch is narrow. Both ops were introduced without a .flags gate, so the generic netlink layer dispatches them to an unprivileged caller instead of rejecting with -EPERM at the permission check. Every other connection-state op in the adjacent block (CONNECT, ASSOCIATE, AUTHENTICATE, SET_KEY, ...) carries GENL_UNS_ADMIN_PERM; SET_PMK / DEL_PMK were introduced without the flag in 2017 and left unchanged by later refactors. Johannes checked the original Intel submission history and confirmed there is no admin check in any prior revision either, so this seems likely to be a simple oversight rather than an intentional carve-out. Require GENL_UNS_ADMIN_PERM so the genl layer performs the same capable(CAP_NET_ADMIN) check as its siblings. wpa_supplicant already needs CAP_NET_ADMIN for every other nl80211 op it issues, so supplicant operation is unaffected. The worst case the missing gate enables today is an unprivileged local process on a multi-user system invalidating the offloaded PMK state of another user's 4-way-handshake session, forcing a full EAP re-auth on the next reconnect. Verified in UML: an unprivileged probe (uid=1000) sees SET_MULTICAST_TO_UNICAST (sibling op with GENL_UNS_ADMIN_PERM) return -EPERM on both pre- and post-fix kernels, while SET_PMK / DEL_PMK return -ENODEV from nl80211_pre_doit()'s wdev lookup pre- fix (proving dispatch crossed the genl permission check) and -EPERM post-fix (rejected at the genl layer as intended). Suggested-by: Johannes Berg Fixes: 3a00df5707b6 ("cfg80211: support 4-way handshake offloading for 802.1X") Assisted-by: Claude:claude-opus-4-7 Signed-off-by: Michael Bommarito Acked-by: Arend van Spriel Link: https://patch.msgid.link/20260421224552.4044147-1-michael.bommarito@gmail.com Signed-off-by: Johannes Berg --- net/wireless/nl80211.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index f334cdef8958..67088804dcc7 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -19828,6 +19828,7 @@ static const struct genl_small_ops nl80211_small_ops[] = { .cmd = NL80211_CMD_SET_PMK, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, .doit = nl80211_set_pmk, + .flags = GENL_UNS_ADMIN_PERM, .internal_flags = IFLAGS(NL80211_FLAG_NEED_NETDEV_UP | NL80211_FLAG_CLEAR_SKB), }, @@ -19835,6 +19836,7 @@ static const struct genl_small_ops nl80211_small_ops[] = { .cmd = NL80211_CMD_DEL_PMK, .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, .doit = nl80211_del_pmk, + .flags = GENL_UNS_ADMIN_PERM, .internal_flags = IFLAGS(NL80211_FLAG_NEED_NETDEV_UP), }, { From 9b55d5c1f5e481e391957f9096d798ca331c461b Mon Sep 17 00:00:00 2001 From: Michael Bommarito Date: Tue, 21 Apr 2026 20:06:51 -0400 Subject: [PATCH 044/525] wifi: mac80211: check ieee80211_rx_data_set_link return in pubsta MLO path __ieee80211_rx_handle_packet() resolves the link via ieee80211_rx_data_set_link() on the pubsta->mlo path but ignores the helper's return value. Inside the helper, rx->link = rcu_dereference(rx->sdata->link[link_id]); can leave rx->link NULL if link_id references a slot already cleared by ieee80211_vif_set_links() during station-initiated ML reconfiguration (see mlme.c's ieee80211_ml_reconfiguration(), which invalidates sdata->link[] before the matching ieee80211_sta_remove_link() loop walks the link-sta hash). RX dispatch still resolves a link_sta from the hash and then drops into ieee80211_prepare_and_rx_handle(), which dereferences link->conf->addr. Every other user site of ieee80211_rx_data_set_link() checks the return and bails on failure; only this branch did not. Mirror the safe pattern. Fixes: e66b7920aa5a ("wifi: mac80211: fix initialization of rx->link and rx->link_sta") Assisted-by: Claude:claude-opus-4-7 Signed-off-by: Michael Bommarito Link: https://patch.msgid.link/20260422000651.4184602-1-michael.bommarito@gmail.com Signed-off-by: Johannes Berg --- net/mac80211/rx.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 3e5d1c47a5b0..5a92413a911f 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -5380,7 +5380,9 @@ static void __ieee80211_rx_handle_packet(struct ieee80211_hw *hw, if (!link_sta) goto out; - ieee80211_rx_data_set_link(&rx, link_sta->link_id); + if (!ieee80211_rx_data_set_link(&rx, + link_sta->link_id)) + goto out; } if (ieee80211_prepare_and_rx_handle(&rx, skb, true)) From 7a5b81e0c87a075afd572f659d8eb68c9c4cd2ba Mon Sep 17 00:00:00 2001 From: Catherine Date: Fri, 24 Apr 2026 21:14:36 +0800 Subject: [PATCH 045/525] wifi: mac80211: drop stray 'static' from fast-RX rx_result ieee80211_invoke_fast_rx() is documented as safe for parallel RX, but its per-invocation rx_result is declared static. Concurrent callers then share one instance and can overwrite each other's result between ieee80211_rx_mesh_data() and the switch on res. That can make a packet that was queued or consumed by ieee80211_rx_mesh_data() fall through into ieee80211_rx_8023(), or make a packet that should continue return as queued. Make res an automatic variable so each invocation keeps its own result. Fixes: 3468e1e0c639 ("wifi: mac80211: add mesh fast-rx support") Cc: stable@vger.kernel.org Signed-off-by: Catherine Link: https://patch.msgid.link/20260424131435.83212-2-enderaoelyther@gmail.com Signed-off-by: Johannes Berg --- net/mac80211/rx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 5a92413a911f..d18e962126ce 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -4971,7 +4971,7 @@ static bool ieee80211_invoke_fast_rx(struct ieee80211_rx_data *rx, struct sk_buff *skb = rx->skb; struct ieee80211_hdr *hdr = (void *)skb->data; struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); - static ieee80211_rx_result res; + ieee80211_rx_result res; int orig_len = skb->len; int hdrlen = ieee80211_hdrlen(hdr->frame_control); int snap_offs = hdrlen; From 58c0ac6125d89bf6ec65a521eaeb52a0e8e20a9f Mon Sep 17 00:00:00 2001 From: Vasant Hegde Date: Mon, 20 Apr 2026 08:42:03 +0000 Subject: [PATCH 046/525] iommu/amd: Use maximum Event log buffer size when SNP is enabled on Family 0x19 Due to CVE-2023-20585, the Event log buffer must use the maximum supported size (512K) on Milan/Genoa (Family 0x19) systems when SNP is enabled, to mitigate a potential security vulnerability. All other systems continue to use the default Event log buffer size (8K). Apply the errata fix by making the following changes: * Introduce new global variable (amd_iommu_evtlog_size) to have event log buffer size. Adjust variable size for family 0x19. * Since 'iommu_snp_enable()' must be called after the core IOMMU subsystem is initialized, it cannot be moved to the early init stage. The SNP errata must also be applied after the 'iommu_snp_enable()' check. Therefore, 'alloc_event_buffer()' and 'iommu_enable_event_buffer()' are now called in the IOMMU_ENABLED state, after the errata is applied. * Adjust alloc_event_buffer() and iommu_enable_event_buffer() to handle all IOMMU instances. * Also rename EVT_* macros to make it more readable. Link: https://www.amd.com/en/resources/product-security/bulletin/amd-sb-3016.html Cc: Borislav Petkov Cc: Suravee Suthikulpanit Cc: Joerg Roedel Signed-off-by: Vasant Hegde Tested-by: Dheeraj Kumar Srivastava Signed-off-by: Joerg Roedel --- drivers/iommu/amd/amd_iommu.h | 2 + drivers/iommu/amd/amd_iommu_types.h | 10 ++- drivers/iommu/amd/init.c | 120 +++++++++++++++++++--------- drivers/iommu/amd/iommu.c | 2 +- 4 files changed, 91 insertions(+), 43 deletions(-) diff --git a/drivers/iommu/amd/amd_iommu.h b/drivers/iommu/amd/amd_iommu.h index 1342e764a548..f1c486dcf0f3 100644 --- a/drivers/iommu/amd/amd_iommu.h +++ b/drivers/iommu/amd/amd_iommu.h @@ -11,6 +11,8 @@ #include "amd_iommu_types.h" +extern int amd_iommu_evtlog_size; + irqreturn_t amd_iommu_int_thread(int irq, void *data); irqreturn_t amd_iommu_int_thread_evtlog(int irq, void *data); irqreturn_t amd_iommu_int_thread_pprlog(int irq, void *data); diff --git a/drivers/iommu/amd/amd_iommu_types.h b/drivers/iommu/amd/amd_iommu_types.h index c685d3771436..c3430c09bc5c 100644 --- a/drivers/iommu/amd/amd_iommu_types.h +++ b/drivers/iommu/amd/amd_iommu_types.h @@ -15,6 +15,7 @@ #include #include #include +#include #include #include #include @@ -141,7 +142,6 @@ #define MMIO_STATUS_GALOG_INT_MASK BIT(10) /* event logging constants */ -#define EVENT_ENTRY_SIZE 0x10 #define EVENT_TYPE_SHIFT 28 #define EVENT_TYPE_MASK 0xf #define EVENT_TYPE_ILL_DEV 0x1 @@ -259,8 +259,12 @@ #define MMIO_CMD_BUFFER_TAIL(x) FIELD_GET(MMIO_CMD_TAIL_MASK, (x)) /* constants for event buffer handling */ -#define EVT_BUFFER_SIZE 8192 /* 512 entries */ -#define EVT_LEN_MASK (0x9ULL << 56) +#define EVTLOG_ENTRY_SIZE 0x10 +#define EVTLOG_SIZE_SHIFT 56 +#define EVTLOG_SIZE_DEF SZ_8K /* 512 entries */ +#define EVTLOG_LEN_MASK_DEF (0x9ULL << EVTLOG_SIZE_SHIFT) +#define EVTLOG_SIZE_MAX SZ_512K /* 32K entries */ +#define EVTLOG_LEN_MASK_MAX (0xFULL << EVTLOG_SIZE_SHIFT) /* Constants for PPR Log handling */ #define PPR_LOG_ENTRIES 512 diff --git a/drivers/iommu/amd/init.c b/drivers/iommu/amd/init.c index 56ad020df494..d8dc5c6db29d 100644 --- a/drivers/iommu/amd/init.c +++ b/drivers/iommu/amd/init.c @@ -132,6 +132,8 @@ struct ivhd_entry { u8 uid; } __attribute__((packed)); +int amd_iommu_evtlog_size = EVTLOG_SIZE_DEF; + /* * An AMD IOMMU memory definition structure. It defines things like exclusion * ranges for devices and regions that should be unity mapped. @@ -865,35 +867,47 @@ void *__init iommu_alloc_4k_pages(struct amd_iommu *iommu, gfp_t gfp, } /* allocates the memory where the IOMMU will log its events to */ -static int __init alloc_event_buffer(struct amd_iommu *iommu) +static int __init alloc_event_buffer(void) { - iommu->evt_buf = iommu_alloc_4k_pages(iommu, GFP_KERNEL, - EVT_BUFFER_SIZE); + struct amd_iommu *iommu; - return iommu->evt_buf ? 0 : -ENOMEM; -} - -static void iommu_enable_event_buffer(struct amd_iommu *iommu) -{ - u64 entry; - - BUG_ON(iommu->evt_buf == NULL); - - if (!is_kdump_kernel()) { - /* - * Event buffer is re-used for kdump kernel and setting - * of MMIO register is not required. - */ - entry = iommu_virt_to_phys(iommu->evt_buf) | EVT_LEN_MASK; - memcpy_toio(iommu->mmio_base + MMIO_EVT_BUF_OFFSET, - &entry, sizeof(entry)); + for_each_iommu(iommu) { + iommu->evt_buf = iommu_alloc_4k_pages(iommu, GFP_KERNEL, + amd_iommu_evtlog_size); + if (!iommu->evt_buf) + return -ENOMEM; } - /* set head and tail to zero manually */ - writel(0x00, iommu->mmio_base + MMIO_EVT_HEAD_OFFSET); - writel(0x00, iommu->mmio_base + MMIO_EVT_TAIL_OFFSET); + return 0; +} - iommu_feature_enable(iommu, CONTROL_EVT_LOG_EN); +static void iommu_enable_event_buffer(void) +{ + struct amd_iommu *iommu; + u64 entry; + + for_each_iommu(iommu) { + BUG_ON(iommu->evt_buf == NULL); + + if (!is_kdump_kernel()) { + /* + * Event buffer is re-used for kdump kernel and setting + * of MMIO register is not required. + */ + entry = iommu_virt_to_phys(iommu->evt_buf); + entry |= (amd_iommu_evtlog_size == EVTLOG_SIZE_DEF) ? + EVTLOG_LEN_MASK_DEF : EVTLOG_LEN_MASK_MAX; + + memcpy_toio(iommu->mmio_base + MMIO_EVT_BUF_OFFSET, + &entry, sizeof(entry)); + } + + /* set head and tail to zero manually */ + writel(0x00, iommu->mmio_base + MMIO_EVT_HEAD_OFFSET); + writel(0x00, iommu->mmio_base + MMIO_EVT_TAIL_OFFSET); + + iommu_feature_enable(iommu, CONTROL_EVT_LOG_EN); + } } /* @@ -984,15 +998,20 @@ static int __init alloc_cwwb_sem(struct amd_iommu *iommu) return 0; } -static int __init remap_event_buffer(struct amd_iommu *iommu) +static int __init remap_event_buffer(void) { + struct amd_iommu *iommu; u64 paddr; pr_info_once("Re-using event buffer from the previous kernel\n"); - paddr = readq(iommu->mmio_base + MMIO_EVT_BUF_OFFSET) & PM_ADDR_MASK; - iommu->evt_buf = iommu_memremap(paddr, EVT_BUFFER_SIZE); + for_each_iommu(iommu) { + paddr = readq(iommu->mmio_base + MMIO_EVT_BUF_OFFSET) & PM_ADDR_MASK; + iommu->evt_buf = iommu_memremap(paddr, amd_iommu_evtlog_size); + if (!iommu->evt_buf) + return -ENOMEM; + } - return iommu->evt_buf ? 0 : -ENOMEM; + return 0; } static int __init remap_command_buffer(struct amd_iommu *iommu) @@ -1044,10 +1063,6 @@ static int __init alloc_iommu_buffers(struct amd_iommu *iommu) ret = remap_command_buffer(iommu); if (ret) return ret; - - ret = remap_event_buffer(iommu); - if (ret) - return ret; } else { ret = alloc_cwwb_sem(iommu); if (ret) @@ -1056,10 +1071,6 @@ static int __init alloc_iommu_buffers(struct amd_iommu *iommu) ret = alloc_command_buffer(iommu); if (ret) return ret; - - ret = alloc_event_buffer(iommu); - if (ret) - return ret; } return 0; @@ -2893,7 +2904,6 @@ static void early_enable_iommu(struct amd_iommu *iommu) iommu_init_flags(iommu); iommu_set_device_table(iommu); iommu_enable_command_buffer(iommu); - iommu_enable_event_buffer(iommu); iommu_set_exclusion_range(iommu); iommu_enable_gt(iommu); iommu_enable_ga(iommu); @@ -2957,7 +2967,6 @@ static void early_enable_iommus(void) iommu_disable_event_buffer(iommu); iommu_disable_irtcachedis(iommu); iommu_enable_command_buffer(iommu); - iommu_enable_event_buffer(iommu); iommu_enable_ga(iommu); iommu_enable_xt(iommu); iommu_enable_irtcachedis(iommu); @@ -3070,6 +3079,7 @@ static void amd_iommu_resume(void *data) for_each_iommu(iommu) early_enable_iommu(iommu); + iommu_enable_event_buffer(); amd_iommu_enable_interrupts(); } @@ -3399,6 +3409,23 @@ static __init void iommu_snp_enable(void) #endif } +static void amd_iommu_apply_erratum_snp(void) +{ +#ifdef CONFIG_KVM_AMD_SEV + if (!amd_iommu_snp_en) + return; + + /* Errata fix for Family 0x19 */ + if (boot_cpu_data.x86 != 0x19) + return; + + /* Set event log buffer size to max */ + amd_iommu_evtlog_size = EVTLOG_SIZE_MAX; + pr_info("Applying erratum: Increase Event log size to 0x%x\n", + amd_iommu_evtlog_size); +#endif +} + /**************************************************************************** * * AMD IOMMU Initialization State Machine @@ -3435,6 +3462,21 @@ static int __init state_next(void) case IOMMU_ENABLED: register_syscore(&amd_iommu_syscore); iommu_snp_enable(); + + amd_iommu_apply_erratum_snp(); + + /* Allocate/enable event log buffer */ + if (is_kdump_kernel()) + ret = remap_event_buffer(); + else + ret = alloc_event_buffer(); + + if (ret) { + init_state = IOMMU_INIT_ERROR; + break; + } + iommu_enable_event_buffer(); + ret = amd_iommu_init_pci(); init_state = ret ? IOMMU_INIT_ERROR : IOMMU_PCI_INIT; break; @@ -4037,7 +4079,7 @@ int amd_iommu_snp_disable(void) return 0; for_each_iommu(iommu) { - ret = iommu_make_shared(iommu->evt_buf, EVT_BUFFER_SIZE); + ret = iommu_make_shared(iommu->evt_buf, amd_iommu_evtlog_size); if (ret) return ret; diff --git a/drivers/iommu/amd/iommu.c b/drivers/iommu/amd/iommu.c index 01171361f9bc..157505d96fdd 100644 --- a/drivers/iommu/amd/iommu.c +++ b/drivers/iommu/amd/iommu.c @@ -1010,7 +1010,7 @@ static void iommu_poll_events(struct amd_iommu *iommu) iommu_print_event(iommu, iommu->evt_buf + head); /* Update head pointer of hardware ring-buffer */ - head = (head + EVENT_ENTRY_SIZE) % EVT_BUFFER_SIZE; + head = (head + EVTLOG_ENTRY_SIZE) % amd_iommu_evtlog_size; writel(head, iommu->mmio_base + MMIO_EVT_HEAD_OFFSET); } From 1f44aab79bac31f459422dfb213e907bb386509c Mon Sep 17 00:00:00 2001 From: Vasant Hegde Date: Mon, 20 Apr 2026 08:42:04 +0000 Subject: [PATCH 047/525] iommu/amd: Use maximum PPR log buffer size when SNP is enabled on Family 0x19 Due to CVE-2023-20585, the PPR log buffer must use the maximum supported size (512K) on Genoa (Family 0x19, model >= 0x10) systems when SNP is enabled, to mitigate a potential security vulnerability. Note that Family 0x19 models below 0x10 (Milan) do not support PPR when SNP is enabled. Hence the PPR log size increase is only applied for model >= 0x10. All other systems continue to use the default PPR log buffer size (8K). Apply the errata fix by making the following changes: - Introduce global new variable (amd_iommu_pprlog_size) to have PPR log buffer size. Adjust variable size for Genoa family. - Extend 'amd_iommu_apply_erratum_snp()' to also set the PPR log buffer size to maximum for Family 0x19 model >= 0x10 when SNP is enabled. - Rename PPR_* macros to make it more readable. Link: https://www.amd.com/en/resources/product-security/bulletin/amd-sb-3016.html Cc: Borislav Petkov Cc: Suravee Suthikulpanit Cc: Joerg Roedel Signed-off-by: Vasant Hegde Tested-by: Dheeraj Kumar Srivastava Signed-off-by: Joerg Roedel --- drivers/iommu/amd/amd_iommu.h | 1 + drivers/iommu/amd/amd_iommu_types.h | 11 ++++++----- drivers/iommu/amd/init.c | 13 ++++++++++++- drivers/iommu/amd/ppr.c | 8 +++++--- 4 files changed, 24 insertions(+), 9 deletions(-) diff --git a/drivers/iommu/amd/amd_iommu.h b/drivers/iommu/amd/amd_iommu.h index f1c486dcf0f3..834d8fabfba3 100644 --- a/drivers/iommu/amd/amd_iommu.h +++ b/drivers/iommu/amd/amd_iommu.h @@ -12,6 +12,7 @@ #include "amd_iommu_types.h" extern int amd_iommu_evtlog_size; +extern int amd_iommu_pprlog_size; irqreturn_t amd_iommu_int_thread(int irq, void *data); irqreturn_t amd_iommu_int_thread_evtlog(int irq, void *data); diff --git a/drivers/iommu/amd/amd_iommu_types.h b/drivers/iommu/amd/amd_iommu_types.h index c3430c09bc5c..f9f718087893 100644 --- a/drivers/iommu/amd/amd_iommu_types.h +++ b/drivers/iommu/amd/amd_iommu_types.h @@ -267,11 +267,12 @@ #define EVTLOG_LEN_MASK_MAX (0xFULL << EVTLOG_SIZE_SHIFT) /* Constants for PPR Log handling */ -#define PPR_LOG_ENTRIES 512 -#define PPR_LOG_SIZE_SHIFT 56 -#define PPR_LOG_SIZE_512 (0x9ULL << PPR_LOG_SIZE_SHIFT) -#define PPR_ENTRY_SIZE 16 -#define PPR_LOG_SIZE (PPR_ENTRY_SIZE * PPR_LOG_ENTRIES) +#define PPRLOG_ENTRY_SIZE 0x10 +#define PPRLOG_SIZE_SHIFT 56 +#define PPRLOG_SIZE_DEF SZ_8K /* 512 entries */ +#define PPRLOG_LEN_MASK_DEF (0x9ULL << PPRLOG_SIZE_SHIFT) +#define PPRLOG_SIZE_MAX SZ_512K /* 32K entries */ +#define PPRLOG_LEN_MASK_MAX (0xFULL << PPRLOG_SIZE_SHIFT) /* PAGE_SERVICE_REQUEST PPR Log Buffer Entry flags */ #define PPR_FLAG_EXEC 0x002 /* Execute permission requested */ diff --git a/drivers/iommu/amd/init.c b/drivers/iommu/amd/init.c index d8dc5c6db29d..3bdb380d23e9 100644 --- a/drivers/iommu/amd/init.c +++ b/drivers/iommu/amd/init.c @@ -133,6 +133,7 @@ struct ivhd_entry { } __attribute__((packed)); int amd_iommu_evtlog_size = EVTLOG_SIZE_DEF; +int amd_iommu_pprlog_size = PPRLOG_SIZE_DEF; /* * An AMD IOMMU memory definition structure. It defines things like exclusion @@ -3423,6 +3424,16 @@ static void amd_iommu_apply_erratum_snp(void) amd_iommu_evtlog_size = EVTLOG_SIZE_MAX; pr_info("Applying erratum: Increase Event log size to 0x%x\n", amd_iommu_evtlog_size); + + /* + * Set PPR log buffer size to max. + * (Family 0x19, model < 0x10 doesn't support PPR when SNP is enabled). + */ + if (boot_cpu_data.x86_model >= 0x10) { + amd_iommu_pprlog_size = PPRLOG_SIZE_MAX; + pr_info("Applying erratum: Increase PPR log size to 0x%x\n", + amd_iommu_pprlog_size); + } #endif } @@ -4083,7 +4094,7 @@ int amd_iommu_snp_disable(void) if (ret) return ret; - ret = iommu_make_shared(iommu->ppr_log, PPR_LOG_SIZE); + ret = iommu_make_shared(iommu->ppr_log, amd_iommu_pprlog_size); if (ret) return ret; diff --git a/drivers/iommu/amd/ppr.c b/drivers/iommu/amd/ppr.c index e6767c057d01..1f8d2823bea4 100644 --- a/drivers/iommu/amd/ppr.c +++ b/drivers/iommu/amd/ppr.c @@ -20,7 +20,7 @@ int __init amd_iommu_alloc_ppr_log(struct amd_iommu *iommu) { iommu->ppr_log = iommu_alloc_4k_pages(iommu, GFP_KERNEL | __GFP_ZERO, - PPR_LOG_SIZE); + amd_iommu_pprlog_size); return iommu->ppr_log ? 0 : -ENOMEM; } @@ -33,7 +33,9 @@ void amd_iommu_enable_ppr_log(struct amd_iommu *iommu) iommu_feature_enable(iommu, CONTROL_PPR_EN); - entry = iommu_virt_to_phys(iommu->ppr_log) | PPR_LOG_SIZE_512; + entry = iommu_virt_to_phys(iommu->ppr_log); + entry |= (amd_iommu_pprlog_size == PPRLOG_SIZE_DEF) ? + PPRLOG_LEN_MASK_DEF : PPRLOG_LEN_MASK_MAX; memcpy_toio(iommu->mmio_base + MMIO_PPR_LOG_OFFSET, &entry, sizeof(entry)); @@ -201,7 +203,7 @@ void amd_iommu_poll_ppr_log(struct amd_iommu *iommu) raw[0] = raw[1] = 0UL; /* Update head pointer of hardware ring-buffer */ - head = (head + PPR_ENTRY_SIZE) % PPR_LOG_SIZE; + head = (head + PPRLOG_ENTRY_SIZE) % amd_iommu_pprlog_size; writel(head, iommu->mmio_base + MMIO_PPR_HEAD_OFFSET); /* Handle PPR entry */ From f9471dc1a7177f594acf8106cc4a6ab33bf0bcb6 Mon Sep 17 00:00:00 2001 From: Mostafa Saleh Date: Fri, 24 Apr 2026 11:50:51 +0000 Subject: [PATCH 048/525] iommu/pages: Fix iommu_pages_flush_incoherent() for non-x86 The dma_sync_single_for_device() function expects a dma_addr_t, but iommu_pages_flush_incoherent() was incorrectly passing a virtual address. Since iommu_pages_start_incoherent() enforces a 1:1 mapping between DMA addresses and physical addresses (checked via WARN_ON), we can convert the virtual address to a physical address before passing it to the DMA API. This also matches the behaviour of the other non-x86 in iommu_pages_free_incoherent(), which uses virt_to_phys(virt); Fixes: 36ae67b13976 ("iommu/pages: Add support for incoherent IOMMU page table walkers") Signed-off-by: Mostafa Saleh Reviewed-by: Lu Baolu Reviewed-by: Pranjal Shrivastava Reviewed-by: Jason Gunthorpe Signed-off-by: Joerg Roedel --- drivers/iommu/iommu-pages.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/iommu/iommu-pages.h b/drivers/iommu/iommu-pages.h index ae9da4f571f6..e9e605b5fa3a 100644 --- a/drivers/iommu/iommu-pages.h +++ b/drivers/iommu/iommu-pages.h @@ -137,7 +137,7 @@ static inline void iommu_pages_flush_incoherent(struct device *dma_dev, void *virt, size_t offset, size_t len) { - dma_sync_single_for_device(dma_dev, (uintptr_t)virt + offset, len, + dma_sync_single_for_device(dma_dev, virt_to_phys(virt) + offset, len, DMA_TO_DEVICE); } void iommu_pages_stop_incoherent_list(struct iommu_pages_list *list, From 26735dfdd8930d9ef1fa92e590a9bf77726efdf6 Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Fri, 17 Apr 2026 13:13:31 +0200 Subject: [PATCH 049/525] pmdomain: core: Fix detach procedure for virtual devices in genpd If a device is attached to a PM domain through genpd_dev_pm_attach_by_id(), genpd calls pm_runtime_enable() for the corresponding virtual device that it registers. While this avoids boilerplate code in drivers, there is no corresponding call to pm_runtime_disable() in genpd_dev_pm_detach(). This means these virtual devices are typically detached from its genpd, while runtime PM remains enabled for them, which is not how things are designed to work. In worst cases it may lead to critical errors, like a NULL pointer dereference bug in genpd_runtime_suspend(), which was recently reported. For another case, we may end up keeping an unnecessary vote for a performance state for the device. To fix these problems, let's add this missing call to pm_runtime_disable() in genpd_dev_pm_detach(). Reported-by: Geert Uytterhoeven Closes: https://lore.kernel.org/all/CAMuHMdWapT40hV3c+CSBqFOW05aWcV1a6v_NiJYgoYi0i9_PDQ@mail.gmail.com/ Fixes: 3c095f32a92b ("PM / Domains: Add support for multi PM domains per device to genpd") Cc: stable@vger.kernel.org Tested-by: Geert Uytterhoeven Signed-off-by: Ulf Hansson --- drivers/pmdomain/core.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/pmdomain/core.c b/drivers/pmdomain/core.c index 4d32fc676aaf..71e930e80178 100644 --- a/drivers/pmdomain/core.c +++ b/drivers/pmdomain/core.c @@ -3089,6 +3089,7 @@ static const struct bus_type genpd_bus_type = { static void genpd_dev_pm_detach(struct device *dev, bool power_off) { struct generic_pm_domain *pd; + bool is_virt_dev; unsigned int i; int ret = 0; @@ -3098,6 +3099,13 @@ static void genpd_dev_pm_detach(struct device *dev, bool power_off) dev_dbg(dev, "removing from PM domain %s\n", pd->name); + /* Check if the device was created by genpd at attach. */ + is_virt_dev = dev->bus == &genpd_bus_type; + + /* Disable runtime PM if we enabled it at attach. */ + if (is_virt_dev) + pm_runtime_disable(dev); + /* Drop the default performance state */ if (dev_gpd_data(dev)->default_pstate) { dev_pm_genpd_set_performance_state(dev, 0); @@ -3123,7 +3131,7 @@ static void genpd_dev_pm_detach(struct device *dev, bool power_off) genpd_queue_power_off_work(pd); /* Unregister the device if it was created by genpd. */ - if (dev->bus == &genpd_bus_type) + if (is_virt_dev) device_unregister(dev); } From ec1fcddb3117d9452210e838fd37389ee61e10e8 Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 8 Apr 2026 14:11:21 +0000 Subject: [PATCH 050/525] pmdomain: mediatek: fix use-after-free in scpsys_get_bus_protection_legacy() In scpsys_get_bus_protection_legacy(), of_find_node_with_property() returns a device node with its reference count incremented. The function then calls of_node_put(node) before checking whether syscon_regmap_lookup_by_phandle() returns an error. If an error occurs, dev_err_probe() dereferences the node pointer to print diagnostic information, but the node memory may have already been freed due to the earlier of_node_put(), leading to a use-after-free vulnerability. Fix this by moving the of_node_put() call after the error check, ensuring the node is still valid when accessed in the error path. Fixes: c29345fa5f66 ("pmdomain: mediatek: Refactor bus protection regmaps retrieval") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Signed-off-by: Ulf Hansson --- drivers/pmdomain/mediatek/mtk-pm-domains.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/pmdomain/mediatek/mtk-pm-domains.c b/drivers/pmdomain/mediatek/mtk-pm-domains.c index d2b8d0332951..e1cfd4223473 100644 --- a/drivers/pmdomain/mediatek/mtk-pm-domains.c +++ b/drivers/pmdomain/mediatek/mtk-pm-domains.c @@ -1015,6 +1015,7 @@ static int scpsys_get_bus_protection_legacy(struct device *dev, struct scpsys *s struct device_node *node, *smi_np; int num_regmaps = 0, i, j; struct regmap *regmap[3]; + int ret = 0; /* * Legacy code retrieves a maximum of three bus protection handles: @@ -1065,11 +1066,14 @@ static int scpsys_get_bus_protection_legacy(struct device *dev, struct scpsys *s if (node) { regmap[2] = syscon_regmap_lookup_by_phandle(node, "mediatek,infracfg-nao"); num_regmaps++; - of_node_put(node); - if (IS_ERR(regmap[2])) - return dev_err_probe(dev, PTR_ERR(regmap[2]), + if (IS_ERR(regmap[2])) { + ret = dev_err_probe(dev, PTR_ERR(regmap[2]), "%pOF: failed to get infracfg regmap\n", node); + of_node_put(node); + return ret; + } + of_node_put(node); } else { regmap[2] = NULL; } From 3f91484f6c13c434bd573ca6b6779c26adb0ddab Mon Sep 17 00:00:00 2001 From: Aaro Koskinen Date: Mon, 13 Apr 2026 21:49:12 +0300 Subject: [PATCH 051/525] USB: omap_udc: DMA: Don't enable burst 4 mode Commit 65111084c63d7 ("USB: more omap_udc updates (dma and omap1710)") added setting for DMA burst 4 mode. But I think this should be undone for two reasons: - It breaks DMA on 15xx boards - transfers just silently stall. - On newer OMAP1 boards, like Nokia 770 (omap1710), there is no measurable performance impact when testing TCP throughput with g_ether with large 15000 byte MTU size. It's also worth noting that when the original change was made, the OMAP_DMA_DATA_BURST_4 handling in arch/arm/plat-omap/dma.c was broken, and actually resulted in the same as the OMAP_DMA_DATA_BURST_DIS i.e. burst disabled. This was fixed not until a couple kernel releases later in an unrelated commit 1a8bfa1eb998a ("[ARM] 3142/1: OMAP 2/5: Update files common to omap1 and omap2"). So based on this it seems there was never really a very good reason to enable this burst mode in omap_udc, so remove it now to allow 15xx DMA to work again (it provides 2x throughput compared to PIO mode). Fixes: 65111084c63d ("[PATCH] USB: more omap_udc updates (dma and omap1710)") Cc: stable Signed-off-by: Aaro Koskinen Link: https://patch.msgid.link/ad06qHLclWHeSGnV@darkstar.musicnaut.iki.fi Signed-off-by: Greg Kroah-Hartman --- drivers/usb/gadget/udc/omap_udc.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/drivers/usb/gadget/udc/omap_udc.c b/drivers/usb/gadget/udc/omap_udc.c index 91139ae668f4..f3ca79cece1b 100644 --- a/drivers/usb/gadget/udc/omap_udc.c +++ b/drivers/usb/gadget/udc/omap_udc.c @@ -733,8 +733,6 @@ static void dma_channel_claim(struct omap_ep *ep, unsigned channel) if (status == 0) { omap_writew(reg, UDC_TXDMA_CFG); /* EMIFF or SDRC */ - omap_set_dma_src_burst_mode(ep->lch, - OMAP_DMA_DATA_BURST_4); omap_set_dma_src_data_pack(ep->lch, 1); /* TIPB */ omap_set_dma_dest_params(ep->lch, @@ -756,8 +754,6 @@ static void dma_channel_claim(struct omap_ep *ep, unsigned channel) UDC_DATA_DMA, 0, 0); /* EMIFF or SDRC */ - omap_set_dma_dest_burst_mode(ep->lch, - OMAP_DMA_DATA_BURST_4); omap_set_dma_dest_data_pack(ep->lch, 1); } } From 0b9fcab1b8608d429e5f239afb197de928d4de7d Mon Sep 17 00:00:00 2001 From: Felix Gu Date: Tue, 7 Apr 2026 21:21:22 +0800 Subject: [PATCH 052/525] usb: ulpi: fix memory leak on ulpi_register() error paths Commit 01af542392b5 ("usb: ulpi: fix double free in ulpi_register_interface() error path") removed kfree(ulpi) from ulpi_register_interface() to fix a double-free when device_register() fails. But when ulpi_of_register() or ulpi_read_id() fail before device_register() is called, the ulpi allocation is leaked. Add kfree(ulpi) on both error paths to properly clean up the allocation. Fixes: 01af542392b5 ("usb: ulpi: fix double free in ulpi_register_interface() error path") Cc: stable Signed-off-by: Felix Gu Reviewed-by: Heikki Krogerus Link: https://patch.msgid.link/20260407-ulpi-v1-1-f3fafe53f7b2@gmail.com Signed-off-by: Greg Kroah-Hartman --- drivers/usb/common/ulpi.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/usb/common/ulpi.c b/drivers/usb/common/ulpi.c index b34fb65813c4..9b69148128e5 100644 --- a/drivers/usb/common/ulpi.c +++ b/drivers/usb/common/ulpi.c @@ -286,12 +286,15 @@ static int ulpi_register(struct device *dev, struct ulpi *ulpi) ACPI_COMPANION_SET(&ulpi->dev, ACPI_COMPANION(dev)); ret = ulpi_of_register(ulpi); - if (ret) + if (ret) { + kfree(ulpi); return ret; + } ret = ulpi_read_id(ulpi); if (ret) { of_node_put(ulpi->dev.of_node); + kfree(ulpi); return ret; } From 2909f0d4994fb4306bf116df5ccee797791fce2c Mon Sep 17 00:00:00 2001 From: Amit Sunil Dhamne Date: Tue, 14 Apr 2026 00:58:32 +0000 Subject: [PATCH 053/525] usb: typec: tcpm: reset internal port states on soft reset AMS Reset internal port states (such as vdm_sm_running and explicit_contract) on soft reset AMS as the port needs to negotiate a new contract. The consequence of leaving the states in as-is cond are as follows: * port is in SRC power role and an explicit contract is negotiated with the port partner (in sink role) * port partner sends a Soft Reset AMS while VDM State Machine is running * port accepts the Soft Reset request and the port advertises src caps * port partner sends a Request message but since the explicit_contract and vdm_sm_running are true from previous negotiation, the port ends up sending Soft Reset instead of Accept msg. Stub Log: [ 203.653942] AMS DISCOVER_IDENTITY start [ 203.653947] PD TX, header: 0x176f [ 203.655901] PD TX complete, status: 0 [ 203.657470] PD RX, header: 0x124f [1] [ 203.657477] Rx VDM cmd 0xff008081 type 2 cmd 1 len 1 [ 203.657482] AMS DISCOVER_IDENTITY finished [ 203.657484] cc:=4 [ 204.155698] PD RX, header: 0x144f [1] [ 204.155718] Rx VDM cmd 0xeeee8001 type 0 cmd 1 len 1 [ 204.155741] PD TX, header: 0x196f [ 204.157622] PD TX complete, status: 0 [ 204.160060] PD RX, header: 0x4d [1] [ 204.160066] state change SRC_READY -> SOFT_RESET [rev2 SOFT_RESET_AMS] [ 204.160076] PD TX, header: 0x163 [ 204.162486] PD TX complete, status: 0 [ 204.162832] AMS SOFT_RESET_AMS finished [ 204.162840] cc:=4 [ 204.162891] AMS POWER_NEGOTIATION start [ 204.162896] state change SOFT_RESET -> AMS_START [rev2 POWER_NEGOTIATION] [ 204.162908] state change AMS_START -> SRC_SEND_CAPABILITIES [rev2 POWER_NEGOTIATION] [ 204.162913] PD TX, header: 0x1361 [ 204.165529] PD TX complete, status: 0 [ 204.165571] pending state change SRC_SEND_CAPABILITIES -> SRC_SEND_CAPABILITIES_TIMEOUT @ 60 ms [rev2 POWER_NEGOTIATION] [ 204.166996] PD RX, header: 0x1242 [1] [ 204.167009] state change SRC_SEND_CAPABILITIES -> SRC_SOFT_RESET_WAIT_SNK_TX [rev2 POWER_NEGOTIATION] [ 204.167019] AMS POWER_NEGOTIATION finished [ 204.167020] cc:=4 [ 204.167083] AMS SOFT_RESET_AMS start [ 204.167086] state change SRC_SOFT_RESET_WAIT_SNK_TX -> SOFT_RESET_SEND [rev2 SOFT_RESET_AMS] [ 204.167092] PD TX, header: 0x16d [ 204.168824] PD TX complete, status: 0 [ 204.168854] pending state change SOFT_RESET_SEND -> HARD_RESET_SEND @ 60 ms [rev2 SOFT_RESET_AMS] [ 204.171876] PD RX, header: 0x43 [1] [ 204.171879] AMS SOFT_RESET_AMS finished This causes COMMON.PROC.PD.11.2 check failure for TEST.PD.VDM.SRC.2_Rev2Src test on the PD compliance tester. Signed-off-by: Amit Sunil Dhamne Fixes: 8d3a0578ad1a ("usb: typec: tcpm: Respond Wait if VDM state machine is running") Fixes: f0690a25a140 ("staging: typec: USB Type-C Port Manager (tcpm)") Cc: stable Reviewed-by: Badhri Jagan Sridharan Acked-by: Heikki Krogerus Link: https://patch.msgid.link/20260414-fix-soft-reset-v1-1-01d7cb9764e2@google.com Signed-off-by: Greg Kroah-Hartman --- drivers/usb/typec/tcpm/tcpm.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/usb/typec/tcpm/tcpm.c b/drivers/usb/typec/tcpm/tcpm.c index dfbb94ddc98a..cd5560d6f19e 100644 --- a/drivers/usb/typec/tcpm/tcpm.c +++ b/drivers/usb/typec/tcpm/tcpm.c @@ -5935,6 +5935,8 @@ static void run_state_machine(struct tcpm_port *port) /* remove existing capabilities */ tcpm_partner_source_caps_reset(port); tcpm_pd_send_control(port, PD_CTRL_ACCEPT, TCPC_TX_SOP); + port->vdm_sm_running = false; + port->explicit_contract = false; tcpm_ams_finish(port); if (port->pwr_role == TYPEC_SOURCE) { port->upcoming_state = SRC_SEND_CAPABILITIES; From f6ec9bb4acc7182b25a793ad094a764e1cb819a7 Mon Sep 17 00:00:00 2001 From: Xu Yang Date: Fri, 24 Apr 2026 15:40:09 +0800 Subject: [PATCH 054/525] usb: typec: tcpm: fix debug accessory mode detection for sink ports The port in debug accessory mode can be either a source or sink. The previous tcpm_port_is_debug() function only checked for source port. Commit 8db73e6a42b6 ("usb: typec: tcpm: allow sink (ufp) to toggle into accessory mode debug") changed the detection logic to support both roles, but left some logic in _tcpm_cc_change() unchanged, This causes the state machine to transition to an incorrect state when operating as a sink in debug accessory mode. Log as below: [ 978.637541] CC1: 0 -> 5, CC2: 0 -> 5 [state TOGGLING, polarity 0, connected] [ 978.637567] state change TOGGLING -> SRC_ATTACH_WAIT [rev1 NONE_AMS] [ 978.637596] pending state change SRC_ATTACH_WAIT -> DEBUG_ACC_ATTACHED @ 180 ms [rev1 NONE_AMS] [ 978.647098] CC1: 5 -> 0, CC2: 5 -> 5 [state SRC_ATTACH_WAIT, polarity 0, connected] [ 978.647115] state change SRC_ATTACH_WAIT -> SRC_ATTACH_WAIT [rev1 NONE_AMS] It should go to SNK_ATTACH_WAIT instead of SRC_ATTACH_WAIT state. To fix this, add tcpm_port_is_debug_source() and tcpm_port_is_debug_sink() helper to explicitly identify the power mode in debug accessory mode. Update the state transition logic in _tcpm_cc_change() to ensure the state machine transitions comply with Type-C specification. Also update the logic in run_state_machine() to keep consistency. Fixes: 8db73e6a42b6 ("usb: typec: tcpm: allow sink (ufp) to toggle into accessory mode debug") Cc: stable Signed-off-by: Xu Yang Acked-by: Heikki Krogerus Reviewed-by: Amit Sunil Dhamne Link: https://patch.msgid.link/20260424074009.2979266-1-xu.yang_2@nxp.com Signed-off-by: Greg Kroah-Hartman --- drivers/usb/typec/tcpm/tcpm.c | 25 ++++++++++++++++--------- 1 file changed, 16 insertions(+), 9 deletions(-) diff --git a/drivers/usb/typec/tcpm/tcpm.c b/drivers/usb/typec/tcpm/tcpm.c index cd5560d6f19e..55fee96d3342 100644 --- a/drivers/usb/typec/tcpm/tcpm.c +++ b/drivers/usb/typec/tcpm/tcpm.c @@ -732,9 +732,14 @@ static const char * const pd_rev[] = { (tcpm_cc_is_source((port)->cc2) && \ !tcpm_cc_is_source((port)->cc1))) +#define tcpm_port_is_debug_source(port) \ + (tcpm_cc_is_source((port)->cc1) && tcpm_cc_is_source((port)->cc2)) + +#define tcpm_port_is_debug_sink(port) \ + (tcpm_cc_is_sink((port)->cc1) && tcpm_cc_is_sink((port)->cc2)) + #define tcpm_port_is_debug(port) \ - ((tcpm_cc_is_source((port)->cc1) && tcpm_cc_is_source((port)->cc2)) || \ - (tcpm_cc_is_sink((port)->cc1) && tcpm_cc_is_sink((port)->cc2))) + (tcpm_port_is_debug_source(port) || tcpm_port_is_debug_sink(port)) #define tcpm_port_is_audio(port) \ (tcpm_cc_is_audio((port)->cc1) && tcpm_cc_is_audio((port)->cc2)) @@ -5176,7 +5181,7 @@ static void run_state_machine(struct tcpm_port *port) tcpm_set_state(port, SNK_UNATTACHED, PD_T_DRP_SNK); break; case SRC_ATTACH_WAIT: - if (tcpm_port_is_debug(port)) + if (tcpm_port_is_debug_source(port)) tcpm_set_state(port, DEBUG_ACC_ATTACHED, port->timings.cc_debounce_time); else if (tcpm_port_is_audio(port)) @@ -5434,7 +5439,7 @@ static void run_state_machine(struct tcpm_port *port) tcpm_set_state(port, SRC_UNATTACHED, PD_T_DRP_SRC); break; case SNK_ATTACH_WAIT: - if (tcpm_port_is_debug(port)) + if (tcpm_port_is_debug_sink(port)) tcpm_set_state(port, DEBUG_ACC_ATTACHED, PD_T_CC_DEBOUNCE); else if (tcpm_port_is_audio(port)) @@ -5454,7 +5459,7 @@ static void run_state_machine(struct tcpm_port *port) if (tcpm_port_is_disconnected(port)) tcpm_set_state(port, SNK_UNATTACHED, PD_T_PD_DEBOUNCE); - else if (tcpm_port_is_debug(port)) + else if (tcpm_port_is_debug_sink(port)) tcpm_set_state(port, DEBUG_ACC_ATTACHED, PD_T_CC_DEBOUNCE); else if (tcpm_port_is_audio(port)) @@ -6362,10 +6367,10 @@ static void _tcpm_cc_change(struct tcpm_port *port, enum typec_cc_status cc1, switch (port->state) { case TOGGLING: - if (tcpm_port_is_debug(port) || tcpm_port_is_audio(port) || + if (tcpm_port_is_debug_source(port) || tcpm_port_is_audio(port) || tcpm_port_is_source(port)) tcpm_set_state(port, SRC_ATTACH_WAIT, 0); - else if (tcpm_port_is_sink(port)) + else if (tcpm_port_is_debug_sink(port) || tcpm_port_is_sink(port)) tcpm_set_state(port, SNK_ATTACH_WAIT, 0); break; case CHECK_CONTAMINANT: @@ -6373,9 +6378,11 @@ static void _tcpm_cc_change(struct tcpm_port *port, enum typec_cc_status cc1, break; case SRC_UNATTACHED: case ACC_UNATTACHED: - if (tcpm_port_is_debug(port) || tcpm_port_is_audio(port) || + if (tcpm_port_is_debug_source(port) || tcpm_port_is_audio(port) || tcpm_port_is_source(port)) tcpm_set_state(port, SRC_ATTACH_WAIT, 0); + else if (tcpm_port_is_debug_sink(port)) + tcpm_set_state(port, SNK_ATTACH_WAIT, 0); break; case SRC_ATTACH_WAIT: if (tcpm_port_is_disconnected(port) || @@ -6397,7 +6404,7 @@ static void _tcpm_cc_change(struct tcpm_port *port, enum typec_cc_status cc1, } break; case SNK_UNATTACHED: - if (tcpm_port_is_debug(port) || tcpm_port_is_audio(port) || + if (tcpm_port_is_debug_sink(port) || tcpm_port_is_audio(port) || tcpm_port_is_sink(port)) tcpm_set_state(port, SNK_ATTACH_WAIT, 0); break; From aad35f9c926ec220b0742af1ada45666ae667956 Mon Sep 17 00:00:00 2001 From: Selvarasu Ganesan Date: Fri, 17 Apr 2026 12:03:11 +0530 Subject: [PATCH 055/525] usb: dwc3: Move GUID programming after PHY initialization The Linux Version Code is currently written to the GUID register before PHY initialization. Certain PHY implementations (such as Synopsys eUSB PHY performing link_sw_reset) clear the GUID register to its default value during initialization, causing the kernel version information to be lost. Move the GUID register programming to occur after PHY initialization completes to ensure the Linux version information persists. Fixes: fa0ea13e9f1c ("usb: dwc3: core: write LINUX_VERSION_CODE to our GUID register") Cc: stable Reported-by: Pritam Manohar Sutar Signed-off-by: Selvarasu Ganesan Acked-by: Thinh Nguyen Link: https://patch.msgid.link/20260417063314.2359-1-selvarasu.g@samsung.com Signed-off-by: Greg Kroah-Hartman --- drivers/usb/dwc3/core.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/usb/dwc3/core.c b/drivers/usb/dwc3/core.c index 58899b1fa96d..65213896de99 100644 --- a/drivers/usb/dwc3/core.c +++ b/drivers/usb/dwc3/core.c @@ -1359,12 +1359,6 @@ int dwc3_core_init(struct dwc3 *dwc) hw_mode = DWC3_GHWPARAMS0_MODE(dwc->hwparams.hwparams0); - /* - * Write Linux Version Code to our GUID register so it's easy to figure - * out which kernel version a bug was found. - */ - dwc3_writel(dwc, DWC3_GUID, LINUX_VERSION_CODE); - ret = dwc3_phy_setup(dwc); if (ret) return ret; @@ -1398,6 +1392,12 @@ int dwc3_core_init(struct dwc3 *dwc) if (ret) goto err_exit_phy; + /* + * Write Linux Version Code to our GUID register so it's easy to figure + * out which kernel version a bug was found. + */ + dwc3_writel(dwc, DWC3_GUID, LINUX_VERSION_CODE); + dwc3_core_setup_global_control(dwc); dwc3_core_num_eps(dwc); From 7a400c6fe3617e31e690e3f7ca37bb335e0498f3 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Mon, 20 Apr 2026 18:11:03 +0200 Subject: [PATCH 056/525] usb: usblp: fix heap leak in IEEE 1284 device ID via short response usblp_ctrl_msg() collapses the usb_control_msg() return value to 0/-errno, discarding the actual number of bytes transferred. A broken printer can complete the GET_DEVICE_ID control transfer short and the driver has no way to know. usblp_cache_device_id_string() reads the 2-byte big-endian length prefix from the response and trusts it (clamped only to the buffer bounds). The buffer is kmalloc(1024) at probe time. A device that sends exactly two bytes (e.g. 0x03 0xFF, claiming a 1023-byte ID) leaves device_id_string[2..1022] holding stale kmalloc heap. That stale data is then exposed: - via the ieee1284_id sysfs attribute (sprintf("%s", buf+2), truncated at the first NUL in the stale heap), and - via the IOCNR_GET_DEVICE_ID ioctl, which copy_to_user()s the full claimed length regardless of NULs, up to 1021 bytes of uninitialized heap, with the leak size chosen by the device. Fix this up by just zapping the buffer with zeros before each request sent to the device. Cc: Pete Zaitcev Assisted-by: gkh_clanker_t1000 Cc: stable Link: https://patch.msgid.link/2026042002-unicorn-greedily-3c63@gregkh Signed-off-by: Greg Kroah-Hartman --- drivers/usb/class/usblp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/usb/class/usblp.c b/drivers/usb/class/usblp.c index 669b9e6879bf..e9b848622a3a 100644 --- a/drivers/usb/class/usblp.c +++ b/drivers/usb/class/usblp.c @@ -1377,6 +1377,7 @@ static int usblp_cache_device_id_string(struct usblp *usblp) { int err, length; + memset(usblp->device_id_string, 0, USBLP_DEVICE_ID_SIZE); err = usblp_get_id(usblp, 0, usblp->device_id_string, USBLP_DEVICE_ID_SIZE - 1); if (err < 0) { dev_dbg(&usblp->intf->dev, From b38e53cbfb9d84732e5984fbd73e128d592415c5 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Mon, 20 Apr 2026 18:11:04 +0200 Subject: [PATCH 057/525] usb: usblp: fix uninitialized heap leak via LPGETSTATUS ioctl Just like in a previous problem in this driver, usblp_ctrl_msg() will collapse the usb_control_msg() return value to 0/-errno, discarding the actual number of bytes transferred. Ideally that short command should be detected and error out, but many printers are known to send "incorrect" responses back so we can't just do that. statusbuf is kmalloc(8) at probe time and never filled before the first LPGETSTATUS ioctl. usblp_read_status() requests 1 byte. If a malicious printer responds with zero bytes, *statusbuf is one byte of stale kmalloc heap, sign-extended into the local int status, which the LPGETSTATUS path then copy_to_user()s directly to the ioctl caller. Fix this all by just zapping out the memory buffer when allocated at probe time. If a later call does a short read, the data will be identical to what the device sent it the last time, so there is no "leak" of information happening. Cc: Pete Zaitcev Assisted-by: gkh_clanker_t1000 Cc: stable Link: https://patch.msgid.link/2026042011-shredder-savage-48c6@gregkh Signed-off-by: Greg Kroah-Hartman --- drivers/usb/class/usblp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/usb/class/usblp.c b/drivers/usb/class/usblp.c index e9b848622a3a..746414763da5 100644 --- a/drivers/usb/class/usblp.c +++ b/drivers/usb/class/usblp.c @@ -1178,7 +1178,7 @@ static int usblp_probe(struct usb_interface *intf, } /* Allocate buffer for printer status */ - usblp->statusbuf = kmalloc(STATUS_BUF_SIZE, GFP_KERNEL); + usblp->statusbuf = kzalloc(STATUS_BUF_SIZE, GFP_KERNEL); if (!usblp->statusbuf) { retval = -ENOMEM; goto abort; From 74f192205c48333de054620a79d7ce9f4515fb0b Mon Sep 17 00:00:00 2001 From: Sarthak Sharma Date: Mon, 27 Apr 2026 16:54:47 +0530 Subject: [PATCH 058/525] selftests: kselftest: fix wrong test number in ksft_exit_skip ksft_exit_skip() increments ksft_xskip before printing the KTAP result. As a result, ksft_test_num() already includes the skipped test. Adding 1 to ksft_test_num() increments the printed test number again, producing an incorrect test number and wrong KTAP output. Drop the extra increment and print ksft_test_num() directly. Link: https://lore.kernel.org/r/20260427112447.147985-1-sarthak.sharma@arm.com Fixes: b85d387c9b09 ("kselftest: fix TAP output for skipped tests") Signed-off-by: Sarthak Sharma Signed-off-by: Shuah Khan --- tools/testing/selftests/kselftest.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/kselftest.h b/tools/testing/selftests/kselftest.h index 6d809f08ab7b..60838b61a2da 100644 --- a/tools/testing/selftests/kselftest.h +++ b/tools/testing/selftests/kselftest.h @@ -450,7 +450,7 @@ static inline __noreturn __printf(1, 2) void ksft_exit_skip(const char *msg, ... */ if (ksft_plan || ksft_test_num()) { ksft_cnt.ksft_xskip++; - printf("ok %u # SKIP ", 1 + ksft_test_num()); + printf("ok %u # SKIP ", ksft_test_num()); } else { printf("1..0 # SKIP "); } From 465b05bae5ac553c13315681c1490dc565337771 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Thomas=20Wei=C3=9Fschuh?= Date: Wed, 22 Apr 2026 14:32:33 +0200 Subject: [PATCH 059/525] selftests: harness: Restore order of test functions MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The recent addition of explicit constructor orders for fixture tests broke the ordering of those relative to non-fixture tests and the reverse-constructor-order detection. Restore the ordering of the test functions relative to each other by using the same explicit test order for all test registrations and __constructor_order_first(). Rename the constant, as it is not specific to TEST_F() anymore. Link: https://lore.kernel.org/r/20260422-kselftests-harness-order-v2-1-93ea980ea3ac@linutronix.de Fixes: 6be268151426 ("selftests/harness: order TEST_F and XFAIL_ADD constructors") Signed-off-by: Thomas Weißschuh Reviewed-by: Kees Cook Signed-off-by: Shuah Khan --- tools/testing/selftests/kselftest_harness.h | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/tools/testing/selftests/kselftest_harness.h b/tools/testing/selftests/kselftest_harness.h index 75fb016cd190..cfdce9cd252e 100644 --- a/tools/testing/selftests/kselftest_harness.h +++ b/tools/testing/selftests/kselftest_harness.h @@ -76,7 +76,7 @@ static inline void __kselftest_memset_safe(void *s, int c, size_t n) memset(s, c, n); } -#define KSELFTEST_PRIO_TEST_F 20000 +#define KSELFTEST_PRIO_TEST 20000 #define KSELFTEST_PRIO_XFAIL 20001 #define TEST_TIMEOUT_DEFAULT 30 @@ -194,7 +194,7 @@ static inline void __kselftest_memset_safe(void *s, int c, size_t n) .fixture = &_fixture_global, \ .termsig = _signal, \ .timeout = TEST_TIMEOUT_DEFAULT, }; \ - static void __attribute__((constructor)) _register_##test_name(void) \ + static void __attribute__((constructor(KSELFTEST_PRIO_TEST))) _register_##test_name(void) \ { \ __register_test(&_##test_name##_object); \ } \ @@ -238,7 +238,7 @@ static inline void __kselftest_memset_safe(void *s, int c, size_t n) FIXTURE_VARIANT(fixture_name); \ static struct __fixture_metadata _##fixture_name##_fixture_object = \ { .name = #fixture_name, }; \ - static void __attribute__((constructor)) \ + static void __attribute__((constructor(KSELFTEST_PRIO_TEST))) \ _register_##fixture_name##_data(void) \ { \ __register_fixture(&_##fixture_name##_fixture_object); \ @@ -364,7 +364,7 @@ static inline void __kselftest_memset_safe(void *s, int c, size_t n) _##fixture_name##_##variant_name##_object = \ { .name = #variant_name, \ .data = &_##fixture_name##_##variant_name##_variant}; \ - static void __attribute__((constructor)) \ + static void __attribute__((constructor(KSELFTEST_PRIO_TEST))) \ _register_##fixture_name##_##variant_name(void) \ { \ __register_fixture_variant(&_##fixture_name##_fixture_object, \ @@ -468,7 +468,7 @@ static inline void __kselftest_memset_safe(void *s, int c, size_t n) fixture_name##_teardown(_metadata, self, variant); \ } \ static struct __test_metadata *_##fixture_name##_##test_name##_object; \ - static void __attribute__((constructor(KSELFTEST_PRIO_TEST_F))) \ + static void __attribute__((constructor(KSELFTEST_PRIO_TEST))) \ _register_##fixture_name##_##test_name(void) \ { \ struct __test_metadata *object = mmap(NULL, sizeof(*object), \ @@ -1323,7 +1323,7 @@ static int test_harness_run(int argc, char **argv) return KSFT_FAIL; } -static void __attribute__((constructor)) __constructor_order_first(void) +static void __attribute__((constructor(KSELFTEST_PRIO_TEST))) __constructor_order_first(void) { __constructor_order_forward = true; } From cfcbfe5cb11650d53f7cafd7adfd556690b77114 Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Tue, 21 Apr 2026 08:06:44 -0700 Subject: [PATCH 060/525] PCI: Don't fallback to bus reset after failed slot reset If a bus has hotplug slots that implement the slot's reset_slot callback, it is not safe to do the non-slot specific bus reset, so don't fallback to it. If a slot reset does fail, the subsequent bus reset will attempt a 2nd link reset on top of previous and fail to handle the hotplug events. Fixes: 8238cb69c01fe ("PCI: Make reset_subordinate hotplug safe") Signed-off-by: Keith Busch Signed-off-by: Bjorn Helgaas Link: https://patch.msgid.link/20260421150644.3543733-1-kbusch@meta.com --- drivers/pci/pci.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 8f7cfcc00090..d34266651ad0 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -5607,13 +5607,14 @@ static int pci_try_reset_bus(struct pci_bus *bus) * reset for affected devices * * This function will first try to reset the slots on this bus if the method is - * available. If slot reset fails or is not available, this will fall back to a + * available. If slot reset is not available, this will fall back to a * secondary bus reset. */ static int pci_reset_bridge(struct pci_dev *bridge, bool restore) { struct pci_bus *bus = bridge->subordinate; struct pci_slot *slot; + int ret = 0; if (!bus) return -ENOTTY; @@ -5627,19 +5628,17 @@ static int pci_reset_bridge(struct pci_dev *bridge, bool restore) goto bus_reset; list_for_each_entry(slot, &bus->slots, list) { - int ret; - if (restore) ret = pci_try_reset_slot(slot); else ret = pci_slot_reset(slot, PCI_RESET_DO_RESET); if (ret) - goto bus_reset; + break; } mutex_unlock(&pci_slot_mutex); - return 0; + return ret; bus_reset: mutex_unlock(&pci_slot_mutex); From ada95e5e603bc6e353ee029f2ba7a7d9a42ad018 Mon Sep 17 00:00:00 2001 From: Cosmin Ratiu Date: Wed, 22 Apr 2026 17:06:46 +0300 Subject: [PATCH 061/525] tools/selftests: Use a sensible timeout value for iperf3 client The default timeout of cmd() is 5 seconds and Iperf3Runner requests the iperf3 client to run for 10 seconds, which clearly doesn't work since commit [1] enforced the timeout parameter. Use a value derived from duration as timeout (+5 seconds for startup/teardown/various other overhead). [1] commit f0bd19316663 ("selftests: net: fix timeout passed as positional argument to communicate()") Signed-off-by: Cosmin Ratiu Signed-off-by: Steffen Klassert --- tools/testing/selftests/drivers/net/lib/py/load.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/drivers/net/lib/py/load.py b/tools/testing/selftests/drivers/net/lib/py/load.py index f181fa2d38fc..e24660e5c27f 100644 --- a/tools/testing/selftests/drivers/net/lib/py/load.py +++ b/tools/testing/selftests/drivers/net/lib/py/load.py @@ -48,7 +48,10 @@ class Iperf3Runner: Starts the iperf3 client with the configured options. """ cmdline = self._build_client(streams, duration, reverse) - return cmd(cmdline, background=background, host=self.env.remote) + kwargs = {"background": background, "host": self.env.remote} + if not background: + kwargs["timeout"] = duration + 5 + return cmd(cmdline, **kwargs) def measure_bandwidth(self, reverse=False): """ From e64e03b478e2da7093564819e903932fca2ddfa1 Mon Sep 17 00:00:00 2001 From: Cosmin Ratiu Date: Wed, 22 Apr 2026 17:06:47 +0300 Subject: [PATCH 062/525] tools/selftests: Add a VXLAN+IPsec traffic test There are VXLAN tests and IPsec tests, but there is no test that combines the two protocols and exercises the tunnel-over-ipsec code paths. Fix that by adding a traffic test with VXLAN and IPsec using crypto offload. This is runnable on HW which supports ESP offload (so no nsim unfortunately). Traffic is done with iperf3 and the test validates that there are no packet drops and iperf3 can get to at least 100 Mbps (a very conservative value on today's crypto offload HW, as it can typically reach multi-Gbps rates). Ran right now, the test fails due to a recently exposed bug in xfrm, which will be fixed in the next patch: # ./tools/testing/selftests/drivers/net/hw/ipsec_vxlan.py TAP version 13 1..4 # Check| At ./tools/testing/selftests/drivers/net/hw/ipsec_vxlan.py, # line 161, in test_vxlan_ipsec_crypto_offload: # Check| ksft_eq(drops_after - drops_before, 0, # Check failed 189 != 0 TX drops during VXLAN+IPsec # Check| At ./tools/testing/selftests/drivers/net/hw/ipsec_vxlan.py, # line 163, in test_vxlan_ipsec_crypto_offload: # Check| ksft_ge(bw_gbps, 0.1, # Check failed 0.0015058278404812596 < 0.1 Minimum 100Mbps over # VXLAN+IPsec not ok 1 ipsec_vxlan.test_vxlan_ipsec_crypto_offload.outer_v4_inner_v4 ... Signed-off-by: Cosmin Ratiu Signed-off-by: Steffen Klassert --- .../testing/selftests/drivers/net/hw/Makefile | 1 + tools/testing/selftests/drivers/net/hw/config | 5 + .../selftests/drivers/net/hw/ipsec_vxlan.py | 204 ++++++++++++++++++ 3 files changed, 210 insertions(+) create mode 100755 tools/testing/selftests/drivers/net/hw/ipsec_vxlan.py diff --git a/tools/testing/selftests/drivers/net/hw/Makefile b/tools/testing/selftests/drivers/net/hw/Makefile index 85ca4d1ecf9e..3b6ff4708005 100644 --- a/tools/testing/selftests/drivers/net/hw/Makefile +++ b/tools/testing/selftests/drivers/net/hw/Makefile @@ -30,6 +30,7 @@ TEST_PROGS = \ gro_hw.py \ hw_stats_l3.sh \ hw_stats_l3_gre.sh \ + ipsec_vxlan.py \ iou-zcrx.py \ irq.py \ loopback.sh \ diff --git a/tools/testing/selftests/drivers/net/hw/config b/tools/testing/selftests/drivers/net/hw/config index dd50cb8a7911..ae0168c2bbe6 100644 --- a/tools/testing/selftests/drivers/net/hw/config +++ b/tools/testing/selftests/drivers/net/hw/config @@ -12,5 +12,10 @@ CONFIG_NET_IPGRE=y CONFIG_NET_IPGRE_DEMUX=y CONFIG_NETKIT=y CONFIG_NET_SCH_INGRESS=y +CONFIG_INET6_ESP=y +CONFIG_INET6_ESP_OFFLOAD=y +CONFIG_INET_ESP=y +CONFIG_INET_ESP_OFFLOAD=y CONFIG_UDMABUF=y CONFIG_VXLAN=y +CONFIG_XFRM_USER=y diff --git a/tools/testing/selftests/drivers/net/hw/ipsec_vxlan.py b/tools/testing/selftests/drivers/net/hw/ipsec_vxlan.py new file mode 100755 index 000000000000..0740a4d85240 --- /dev/null +++ b/tools/testing/selftests/drivers/net/hw/ipsec_vxlan.py @@ -0,0 +1,204 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0 +"""Traffic test for VXLAN + IPsec crypto-offload.""" + +import os + +from lib.py import ksft_run, ksft_exit, ksft_eq, ksft_ge +from lib.py import ksft_variants, KsftNamedVariant, KsftSkipEx +from lib.py import CmdExitFailure, NetDrvEpEnv, cmd, defer, ethtool, ip +from lib.py import Iperf3Runner + +# Inner tunnel addresses - TEST-NET-2 (RFC 5737) / doc prefix (RFC 3849) +INNER_V4_LOCAL = "198.51.100.1" +INNER_V4_REMOTE = "198.51.100.2" +INNER_V6_LOCAL = "2001:db8:100::1" +INNER_V6_REMOTE = "2001:db8:100::2" + +# ESP parameters +SPI_OUT = "0x1000" +SPI_IN = "0x1001" +# 128-bit key + 32-bit salt = 20 bytes hex, 128-bit ICV +ESP_AEAD = "aead 'rfc4106(gcm(aes))' 0x" + "01" * 20 + " 128" + + +def xfrm(args, host=None): + """Runs 'ip xfrm' via shell to preserve parentheses in algo names.""" + cmd(f"ip xfrm {args}", shell=True, host=host) + + +def check_xfrm_offload_support(): + """Skips if iproute2 lacks xfrm offload support.""" + out = cmd("ip xfrm state help", fail=False) + if "offload" not in out.stdout + out.stderr: + raise KsftSkipEx("iproute2 too old, missing xfrm offload") + + +def check_esp_hw_offload(cfg): + """Skips if device lacks esp-hw-offload support.""" + check_xfrm_offload_support() + try: + feat = ethtool(f"-k {cfg.ifname}", json=True)[0] + except (CmdExitFailure, IndexError) as e: + raise KsftSkipEx(f"can't query features: {e}") from e + if not feat.get("esp-hw-offload", {}).get("active"): + raise KsftSkipEx("Device does not support esp-hw-offload") + + +def get_tx_drops(cfg): + """Returns TX dropped counter from the physical device.""" + stats = ip("-s -s link show dev " + cfg.ifname, json=True)[0] + return stats["stats64"]["tx"]["dropped"] + + +def setup_vxlan_ipsec(cfg, outer_ipver, inner_ipver): + """Sets up VXLAN tunnel with IPsec transport-mode crypto-offload.""" + vxlan_name = f"vx{os.getpid()}" + local_addr = cfg.addr_v[outer_ipver] + remote_addr = cfg.remote_addr_v[outer_ipver] + + if inner_ipver == "4": + inner_local = f"{INNER_V4_LOCAL}/24" + inner_remote = f"{INNER_V4_REMOTE}/24" + addr_extra = "" + else: + inner_local = f"{INNER_V6_LOCAL}/64" + inner_remote = f"{INNER_V6_REMOTE}/64" + addr_extra = " nodad" + + if outer_ipver == "6": + vxlan_opts = "udp6zerocsumtx udp6zerocsumrx" + else: + vxlan_opts = "noudpcsum" + + # VXLAN tunnel - local side + ip(f"link add {vxlan_name} type vxlan id 100 dstport 4789 {vxlan_opts} " + f"local {local_addr} remote {remote_addr} dev {cfg.ifname}") + defer(ip, f"link del {vxlan_name}") + ip(f"addr add {inner_local} dev {vxlan_name}{addr_extra}") + ip(f"link set {vxlan_name} up") + + # VXLAN tunnel - remote side + ip(f"link add {vxlan_name} type vxlan id 100 dstport 4789 {vxlan_opts} " + f"local {remote_addr} remote {local_addr} dev {cfg.remote_ifname}", + host=cfg.remote) + defer(ip, f"link del {vxlan_name}", host=cfg.remote) + ip(f"addr add {inner_remote} dev {vxlan_name}{addr_extra}", + host=cfg.remote) + ip(f"link set {vxlan_name} up", host=cfg.remote) + + # xfrm state - local outbound SA + xfrm(f"state add src {local_addr} dst {remote_addr} " + f"proto esp spi {SPI_OUT} " + f"{ESP_AEAD} " + f"mode transport offload crypto dev {cfg.ifname} dir out") + defer(xfrm, f"state del src {local_addr} dst {remote_addr} " + f"proto esp spi {SPI_OUT}") + + # xfrm state - local inbound SA + xfrm(f"state add src {remote_addr} dst {local_addr} " + f"proto esp spi {SPI_IN} " + f"{ESP_AEAD} " + f"mode transport offload crypto dev {cfg.ifname} dir in") + defer(xfrm, f"state del src {remote_addr} dst {local_addr} " + f"proto esp spi {SPI_IN}") + + # xfrm state - remote outbound SA (mirror, software crypto) + xfrm(f"state add src {remote_addr} dst {local_addr} " + f"proto esp spi {SPI_IN} " + f"{ESP_AEAD} " + f"mode transport", + host=cfg.remote) + defer(xfrm, f"state del src {remote_addr} dst {local_addr} " + f"proto esp spi {SPI_IN}", host=cfg.remote) + + # xfrm state - remote inbound SA (mirror, software crypto) + xfrm(f"state add src {local_addr} dst {remote_addr} " + f"proto esp spi {SPI_OUT} " + f"{ESP_AEAD} " + f"mode transport", + host=cfg.remote) + defer(xfrm, f"state del src {local_addr} dst {remote_addr} " + f"proto esp spi {SPI_OUT}", host=cfg.remote) + + # xfrm policy - local out + xfrm(f"policy add src {local_addr} dst {remote_addr} " + f"proto udp dport 4789 dir out " + f"tmpl src {local_addr} dst {remote_addr} proto esp mode transport") + defer(xfrm, f"policy del src {local_addr} dst {remote_addr} " + f"proto udp dport 4789 dir out") + + # xfrm policy - local in + xfrm(f"policy add src {remote_addr} dst {local_addr} " + f"proto udp dport 4789 dir in " + f"tmpl src {remote_addr} dst {local_addr} proto esp mode transport") + defer(xfrm, f"policy del src {remote_addr} dst {local_addr} " + f"proto udp dport 4789 dir in") + + # xfrm policy - remote out + xfrm(f"policy add src {remote_addr} dst {local_addr} " + f"proto udp dport 4789 dir out " + f"tmpl src {remote_addr} dst {local_addr} proto esp mode transport", + host=cfg.remote) + defer(xfrm, f"policy del src {remote_addr} dst {local_addr} " + f"proto udp dport 4789 dir out", host=cfg.remote) + + # xfrm policy - remote in + xfrm(f"policy add src {local_addr} dst {remote_addr} " + f"proto udp dport 4789 dir in " + f"tmpl src {local_addr} dst {remote_addr} proto esp mode transport", + host=cfg.remote) + defer(xfrm, f"policy del src {local_addr} dst {remote_addr} " + f"proto udp dport 4789 dir in", host=cfg.remote) + + +def _vxlan_ipsec_variants(): + """Generates outer/inner IP version variants.""" + for outer in ["4", "6"]: + for inner in ["4", "6"]: + yield KsftNamedVariant(f"outer_v{outer}_inner_v{inner}", outer, inner) + + +@ksft_variants(_vxlan_ipsec_variants()) +def test_vxlan_ipsec_crypto_offload(cfg, outer_ipver, inner_ipver): + """Tests VXLAN+IPsec crypto-offload has no TX drops.""" + cfg.require_ipver(outer_ipver) + check_esp_hw_offload(cfg) + + setup_vxlan_ipsec(cfg, outer_ipver, inner_ipver) + + if inner_ipver == "4": + inner_local = INNER_V4_LOCAL + inner_remote = INNER_V4_REMOTE + ping = "ping" + else: + inner_local = INNER_V6_LOCAL + inner_remote = INNER_V6_REMOTE + ping = "ping -6" + + cmd(f"{ping} -c 1 -W 2 {inner_remote}") + + drops_before = get_tx_drops(cfg) + + runner = Iperf3Runner(cfg, server_ip=inner_local, + client_ip=inner_remote) + bw_gbps = runner.measure_bandwidth(reverse=True) + + cfg.wait_hw_stats_settle() + drops_after = get_tx_drops(cfg) + + ksft_eq(drops_after - drops_before, 0, + comment="TX drops during VXLAN+IPsec") + ksft_ge(bw_gbps, 0.1, + comment="Minimum 100Mbps over VXLAN+IPsec") + + +def main(): + """Runs VXLAN+IPsec crypto-offload GSO selftest.""" + with NetDrvEpEnv(__file__, nsim_test=False) as cfg: + ksft_run([test_vxlan_ipsec_crypto_offload], args=(cfg,)) + ksft_exit() + + +if __name__ == "__main__": + main() From fa90a3145c0340c3f624206a81637c542254ea1d Mon Sep 17 00:00:00 2001 From: Cosmin Ratiu Date: Wed, 22 Apr 2026 17:06:48 +0300 Subject: [PATCH 063/525] xfrm: Don't clobber inner headers when already set On VXLAN over IPsec egress, xfrm{4,6}_transport_output() blindly overwrite inner_transport_header (== the inner TCP header saved in VXLAN iptunnel_handle_offloads() -> skb_reset_inner_headers()) with the current transport_header (== the VXLAN outer UDP header set by udp_tunnel_xmit_skb()). This was a latent bug, harmless until commit [1] added a doff validation check in qdisc_pkt_len_segs_init() for encapsulated GSO packets. With the wrong inner_transport_header set by xfrm, qdisc_pkt_len_segs_init() interprets inner_transport_header as a TCP header, reads doff=0 from the upper byte of the VNI and drops the packet with DROP_REASON_SKB_BAD_GSO. Besides the use in GSO to determine the header size of segmented packets, inner_transport_header might be used by drivers to set up inner checksum offloading by pointing the HW to the inner transport header. A quick browse through available drivers shows that mlx5 uses skb->csum_start specifically for this scenario, while others either don't support VXLAN over IPsec crypto offload (ixgbe) or the HW is capable of parsing the packets itself (nfp, Chelsio). But in all cases, it is more correct to let the inner_transport_header point to the innermost header instead of overwriting it in xfrm. So fix this by guarding all four inner header save sites in xfrm_output.c (xfrm{4,6}_transport_output, xfrm{4,6}_tunnel_encap_add) with a check for skb->inner_protocol. When inner_protocol is set, a tunnel layer (VXLAN, Geneve, GRE, etc.) has already saved the correct inner header offsets and they must not be overwritten. When inner_protocol is zero, no prior tunnel encapsulation exists and xfrm must save the inner headers itself. The tunnel mode checks are only added for completion, since they aren't strictly required, as xfrm_output() forces software GSO in tunnel mode before encap. This makes the previously added test pass: # ./tools/testing/selftests/drivers/net/hw/ipsec_vxlan.py TAP version 13 1..4 ok 1 ipsec_vxlan.test_vxlan_ipsec_crypto_offload.outer_v4_inner_v4 ok 2 ipsec_vxlan.test_vxlan_ipsec_crypto_offload.outer_v4_inner_v6 ok 3 ipsec_vxlan.test_vxlan_ipsec_crypto_offload.outer_v6_inner_v4 ok 4 ipsec_vxlan.test_vxlan_ipsec_crypto_offload.outer_v6_inner_v6 # Totals: pass:4 fail:0 xfail:0 xpass:0 skip:0 error:0 [1] commit 7fb4c1967011 ("net: pull headers in qdisc_pkt_len_segs_init()") Fixes: f1bd7d659ef0 ("xfrm: Add encapsulation header offsets while SKB is not encrypted") Signed-off-by: Cosmin Ratiu Signed-off-by: Steffen Klassert --- net/xfrm/xfrm_output.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/net/xfrm/xfrm_output.c b/net/xfrm/xfrm_output.c index a9652b422f51..cc35c2fcbbe0 100644 --- a/net/xfrm/xfrm_output.c +++ b/net/xfrm/xfrm_output.c @@ -66,7 +66,9 @@ static int xfrm4_transport_output(struct xfrm_state *x, struct sk_buff *skb) struct iphdr *iph = ip_hdr(skb); int ihl = iph->ihl * 4; - skb_set_inner_transport_header(skb, skb_transport_offset(skb)); + if (!skb->inner_protocol) + skb_set_inner_transport_header(skb, + skb_transport_offset(skb)); skb_set_network_header(skb, -x->props.header_len); skb->mac_header = skb->network_header + @@ -167,7 +169,9 @@ static int xfrm6_transport_output(struct xfrm_state *x, struct sk_buff *skb) int hdr_len; iph = ipv6_hdr(skb); - skb_set_inner_transport_header(skb, skb_transport_offset(skb)); + if (!skb->inner_protocol) + skb_set_inner_transport_header(skb, + skb_transport_offset(skb)); hdr_len = xfrm6_hdr_offset(x, skb, &prevhdr); if (hdr_len < 0) @@ -276,8 +280,10 @@ static int xfrm4_tunnel_encap_add(struct xfrm_state *x, struct sk_buff *skb) struct iphdr *top_iph; int flags; - skb_set_inner_network_header(skb, skb_network_offset(skb)); - skb_set_inner_transport_header(skb, skb_transport_offset(skb)); + if (!skb->inner_protocol) { + skb_set_inner_network_header(skb, skb_network_offset(skb)); + skb_set_inner_transport_header(skb, skb_transport_offset(skb)); + } skb_set_network_header(skb, -x->props.header_len); skb->mac_header = skb->network_header + @@ -321,8 +327,10 @@ static int xfrm6_tunnel_encap_add(struct xfrm_state *x, struct sk_buff *skb) struct ipv6hdr *top_iph; int dsfield; - skb_set_inner_network_header(skb, skb_network_offset(skb)); - skb_set_inner_transport_header(skb, skb_transport_offset(skb)); + if (!skb->inner_protocol) { + skb_set_inner_network_header(skb, skb_network_offset(skb)); + skb_set_inner_transport_header(skb, skb_transport_offset(skb)); + } skb_set_network_header(skb, -x->props.header_len); skb->mac_header = skb->network_header + From db57a1aa54ff68669781976e4edb045e09e2b65b Mon Sep 17 00:00:00 2001 From: Jeongjun Park Date: Thu, 23 Apr 2026 02:38:46 +0900 Subject: [PATCH 064/525] wifi: rsi: fix kthread lifetime race between self-exit and external-stop RSI driver use both self-exit(kthread_complete_and_exit) and external-stop (kthread_stop) when killing a kthread. Generally, kthread_stop() is called first, and in this case, no particular issues occur. However, in rare instances where kthread_complete_and_exit() is called first and then kthread_stop() is called, a UAF occurs because the kthread object, which has already exited and been freed, is accessed again. Therefore, to prevent this with minimal modification, you must remove kthread_stop() and change the code to wait until the self-exit operation is completed. Cc: Reported-by: syzbot+5de83f57cd8531f55596@syzkaller.appspotmail.com Closes: https://lore.kernel.org/all/69e5d03b.a00a0220.1bd0ca.0064.GAE@google.com/ Fixes: 4c62764d0fc2 ("rsi: improve kernel thread handling to fix kernel panic") Signed-off-by: Jeongjun Park Link: https://patch.msgid.link/20260422173846.37640-1-aha310510@gmail.com Signed-off-by: Johannes Berg --- drivers/net/wireless/rsi/rsi_common.h | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/net/wireless/rsi/rsi_common.h b/drivers/net/wireless/rsi/rsi_common.h index 591602beeec6..3cdf9ded876d 100644 --- a/drivers/net/wireless/rsi/rsi_common.h +++ b/drivers/net/wireless/rsi/rsi_common.h @@ -70,12 +70,11 @@ static inline int rsi_create_kthread(struct rsi_common *common, return 0; } -static inline int rsi_kill_thread(struct rsi_thread *handle) +static inline void rsi_kill_thread(struct rsi_thread *handle) { atomic_inc(&handle->thread_done); rsi_set_event(&handle->event); - - return kthread_stop(handle->task); + wait_for_completion(&handle->completion); } void rsi_mac80211_detach(struct rsi_hw *hw); From a9e8765fd206388d5672db229784982bf559f097 Mon Sep 17 00:00:00 2001 From: Thorsten Blum Date: Wed, 22 Apr 2026 14:25:27 +0200 Subject: [PATCH 065/525] efivarfs: use QSTR() in efivarfs_alloc_dentry Use QSTR() and drop strlen() in efivarfs_alloc_dentry(). Signed-off-by: Thorsten Blum Signed-off-by: Ard Biesheuvel --- fs/efivarfs/super.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/fs/efivarfs/super.c b/fs/efivarfs/super.c index 1c5224cf183e..733c19571f1c 100644 --- a/fs/efivarfs/super.c +++ b/fs/efivarfs/super.c @@ -191,13 +191,10 @@ static const struct dentry_operations efivarfs_d_ops = { static struct dentry *efivarfs_alloc_dentry(struct dentry *parent, char *name) { + struct qstr q = QSTR(name); struct dentry *d; - struct qstr q; int err; - q.name = name; - q.len = strlen(name); - err = efivarfs_d_hash(parent, &q); if (err) return ERR_PTR(err); From b336e40c62fbdc4b8a1f09a4ada31f4a90c69eb1 Mon Sep 17 00:00:00 2001 From: Thomas Huth Date: Mon, 27 Apr 2026 17:56:30 +0200 Subject: [PATCH 066/525] efi: pstore: Drop efivar lock when efi_pstore_open() returns with an error If kzalloc fails, the function returns -ENOMEM without calling efivar_unlock(). Since open() returned an error, the calling site in pstore_get_backend_records() won't call the close() function, so the lock is never released. Thus drop the lock in case of errors here. Fixes: 859748255b434 ("efi: pstore: Omit efivars caching EFI varstore access layer") Assisted-by: Claude:claude-opus-4-6 Signed-off-by: Thomas Huth Signed-off-by: Ard Biesheuvel --- drivers/firmware/efi/efi-pstore.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/firmware/efi/efi-pstore.c b/drivers/firmware/efi/efi-pstore.c index a253b6144945..a5db3534f0a6 100644 --- a/drivers/firmware/efi/efi-pstore.c +++ b/drivers/firmware/efi/efi-pstore.c @@ -60,8 +60,10 @@ static int efi_pstore_open(struct pstore_info *psi) return err; psi->data = kzalloc(record_size, GFP_KERNEL); - if (!psi->data) + if (!psi->data) { + efivar_unlock(); return -ENOMEM; + } return 0; } From f1fb23a0a0fcbdb66672da51d7d63a259f6396ca Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Mon, 27 Apr 2026 11:32:36 -0700 Subject: [PATCH 067/525] fbdev: ipu-v3: clean up kernel-doc warnings Correct all kernel-doc warnings: - fix a typedef kernel-doc comment - mark a list_head as private - use Returns: for function return values Warning: include/video/imx-ipu-image-convert.h:31 struct member 'list' not described in 'ipu_image_convert_run' Warning: include/video/imx-ipu-image-convert.h:40 function parameter 'ipu_image_convert_cb_t' not described in 'void' Warning: include/video/imx-ipu-image-convert.h:40 expecting prototype for ipu_image_convert_cb_t(). Prototype was for void() instead Warning: include/video/imx-ipu-image-convert.h:66 No description found for return value of 'ipu_image_convert_verify' Warning: include/video/imx-ipu-image-convert.h:90 No description found for return value of 'ipu_image_convert_prepare' Warning: include/video/imx-ipu-image-convert.h:125 No description found for return value of 'ipu_image_convert_queue' Warning: include/video/imx-ipu-image-convert.h:163 No description found for return value of 'ipu_image_convert' Signed-off-by: Randy Dunlap Reviewed-by: Philipp Zabel Signed-off-by: Helge Deller --- include/video/imx-ipu-image-convert.h | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) diff --git a/include/video/imx-ipu-image-convert.h b/include/video/imx-ipu-image-convert.h index 003b3927ede5..6b77968a6a15 100644 --- a/include/video/imx-ipu-image-convert.h +++ b/include/video/imx-ipu-image-convert.h @@ -27,12 +27,13 @@ struct ipu_image_convert_run { int status; + /* private: */ /* internal to image converter, callers don't touch */ struct list_head list; }; /** - * ipu_image_convert_cb_t - conversion callback function prototype + * typedef ipu_image_convert_cb_t - conversion callback function prototype * * @run: the completed conversion run pointer * @ctx: a private context pointer for the callback @@ -60,7 +61,7 @@ void ipu_image_convert_adjust(struct ipu_image *in, struct ipu_image *out, * @out: output image format * @rot_mode: rotation mode * - * Returns 0 if the formats and rotation mode meet IPU restrictions, + * Returns: 0 if the formats and rotation mode meet IPU restrictions, * -EINVAL otherwise. */ int ipu_image_convert_verify(struct ipu_image *in, struct ipu_image *out, @@ -77,11 +78,11 @@ int ipu_image_convert_verify(struct ipu_image *in, struct ipu_image *out, * @complete: run completion callback * @complete_context: a context pointer for the completion callback * - * Returns an opaque conversion context pointer on success, error pointer + * In V4L2, drivers should call ipu_image_convert_prepare() at streamon. + * + * Returns: an opaque conversion context pointer on success, error pointer * on failure. The input/output formats and rotation mode must already meet * IPU retrictions. - * - * In V4L2, drivers should call ipu_image_convert_prepare() at streamon. */ struct ipu_image_convert_ctx * ipu_image_convert_prepare(struct ipu_soc *ipu, enum ipu_ic_task ic_task, @@ -122,6 +123,8 @@ void ipu_image_convert_unprepare(struct ipu_image_convert_ctx *ctx); * In V4L2, drivers should call ipu_image_convert_queue() while * streaming to queue the conversion of a received input buffer. * For example mem2mem devices this would be called in .device_run. + * + * Returns: 0 on success or -errno on error. */ int ipu_image_convert_queue(struct ipu_image_convert_run *run); @@ -155,6 +158,9 @@ void ipu_image_convert_abort(struct ipu_image_convert_ctx *ctx); * On successful return the caller can queue more run requests if needed, using * the prepared context in run->ctx. The caller is responsible for unpreparing * the context when no more conversion requests are needed. + * + * Returns: pointer to the created &struct ipu_image_convert_run that has + * been queued on success; an ERR_PTR(errno) on error. */ struct ipu_image_convert_run * ipu_image_convert(struct ipu_soc *ipu, enum ipu_ic_task ic_task, From 0b996ae54d876b41c52dd7cfc512eb008a47d781 Mon Sep 17 00:00:00 2001 From: Chen Ni Date: Tue, 28 Apr 2026 11:17:25 +0800 Subject: [PATCH 068/525] fbdev: defio: Remove duplicate include of linux/module.h Remove duplicate inclusion of linux/module.h in fb_defio.c to clean up redundant code. Signed-off-by: Chen Ni Signed-off-by: Helge Deller --- drivers/video/fbdev/core/fb_defio.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/video/fbdev/core/fb_defio.c b/drivers/video/fbdev/core/fb_defio.c index a12dd25ab697..fd00b86e1ae6 100644 --- a/drivers/video/fbdev/core/fb_defio.c +++ b/drivers/video/fbdev/core/fb_defio.c @@ -14,7 +14,6 @@ #include #include #include -#include #include #include #include From d237f719b2726c0e6d62bfa1543f53b624471929 Mon Sep 17 00:00:00 2001 From: Thomas Zimmermann Date: Tue, 28 Apr 2026 10:28:43 +0200 Subject: [PATCH 069/525] lib/fonts: Fix bit position when rotating by 180 degrees MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix the horizontal bit position when rotating a glyph by 180°. The original code in rotate_ud() rounded the value in width up to a multiple of 8, aka the bit pitch, and calculated the rotated pixel from that value. The new code stores the glyph's pitch in bit_pitch, but fails to update the rotated pixel's output accordingly. Simply replacing the variable does this. The bug can be reproduced by setting a font with an unaligned width, such as sun12x22, like this: setfont sun12x22 echo 2 > /sys/class/graphics/fbcon/rotate Without the fix, the font looks distorted. Fixes: a30e9e6b018f ("lib/fonts: Refactor glyph-rotation helpers") Closes: https://lore.gitlab.freedesktop.org/drm-ai-reviews/review-patch7-20260407092555.58816-8-tzimmermann@suse.de/ Cc: dri-devel@lists.freedesktop.org Cc: linux-fbdev@vger.kernel.org Signed-off-by: Thomas Zimmermann Signed-off-by: Helge Deller --- lib/fonts/font_rotate.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/lib/fonts/font_rotate.c b/lib/fonts/font_rotate.c index 065e0fc0667b..275406008823 100644 --- a/lib/fonts/font_rotate.c +++ b/lib/fonts/font_rotate.c @@ -106,7 +106,7 @@ static void __font_glyph_rotate_180(const unsigned char *glyph, for (y = 0; y < height; y++) { for (x = 0; x < width; x++) { if (font_glyph_test_bit(glyph, x, y, bit_pitch)) { - font_glyph_set_bit(out, width - (1 + x + shift), height - (1 + y), + font_glyph_set_bit(out, bit_pitch - 1 - x - shift, height - 1 - y, bit_pitch); } } From 76b48a70b16b4036814964b039cde413e0164416 Mon Sep 17 00:00:00 2001 From: Li RongQing Date: Fri, 6 Feb 2026 00:08:36 -0500 Subject: [PATCH 070/525] IB/hfi1: Fix potential use-after-free in PIO and SDMA map teardown The current teardown logic for dd->pio_map and dd->sdma_map frees the structures while they might still be accessed by RCU readers. Although the pointer is nulled under a spinlock, the memory is reclaimed before waiting for the grace period to end. This patch fixes the sequence by: 1. Extracting the pointer under the lock. 2. Clearing the RCU-protected pointer. 3. Waiting for readers to finish with synchronize_rcu(). 4. Finally freeing the memory. Fixes: 7724105686e7 ("IB/hfi1: add driver files") Link: https://patch.msgid.link/r/20260206050836.5890-1-lirongqing@baidu.com Signed-off-by: Li RongQing Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/hfi1/pio.c | 5 ++++- drivers/infiniband/hw/hfi1/sdma.c | 4 +++- 2 files changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/pio.c b/drivers/infiniband/hw/hfi1/pio.c index 51afaac88c72..9121d83bf88a 100644 --- a/drivers/infiniband/hw/hfi1/pio.c +++ b/drivers/infiniband/hw/hfi1/pio.c @@ -1942,13 +1942,16 @@ int pio_map_init(struct hfi1_devdata *dd, u8 port, u8 num_vls, u8 *vl_scontexts) void free_pio_map(struct hfi1_devdata *dd) { + struct pio_vl_map *map; + /* Free PIO map if allocated */ if (rcu_access_pointer(dd->pio_map)) { spin_lock_irq(&dd->pio_map_lock); - pio_map_free(rcu_access_pointer(dd->pio_map)); + map = rcu_access_pointer(dd->pio_map); RCU_INIT_POINTER(dd->pio_map, NULL); spin_unlock_irq(&dd->pio_map_lock); synchronize_rcu(); + pio_map_free(map); } kfree(dd->kernel_send_context); dd->kernel_send_context = NULL; diff --git a/drivers/infiniband/hw/hfi1/sdma.c b/drivers/infiniband/hw/hfi1/sdma.c index e5f442938177..cfd9dd0f7e81 100644 --- a/drivers/infiniband/hw/hfi1/sdma.c +++ b/drivers/infiniband/hw/hfi1/sdma.c @@ -1255,6 +1255,7 @@ void sdma_clean(struct hfi1_devdata *dd, size_t num_engines) { size_t i; struct sdma_engine *sde; + struct sdma_vl_map *map; if (dd->sdma_pad_dma) { dma_free_coherent(&dd->pcidev->dev, SDMA_PAD, @@ -1291,10 +1292,11 @@ void sdma_clean(struct hfi1_devdata *dd, size_t num_engines) } if (rcu_access_pointer(dd->sdma_map)) { spin_lock_irq(&dd->sde_map_lock); - sdma_map_free(rcu_access_pointer(dd->sdma_map)); + map = rcu_access_pointer(dd->sdma_map); RCU_INIT_POINTER(dd->sdma_map, NULL); spin_unlock_irq(&dd->sde_map_lock); synchronize_rcu(); + sdma_map_free(map); } kfree(dd->per_sdma); dd->per_sdma = NULL; From 4c6f86d85d03cdb33addce86aa69aa795ca6c47a Mon Sep 17 00:00:00 2001 From: Michael Bommarito Date: Tue, 14 Apr 2026 07:15:55 -0400 Subject: [PATCH 071/525] RDMA/rxe: Reject unknown opcodes before ICRC processing Even after applying commit 7244491dab34 ("RDMA/rxe: Validate pad and ICRC before payload_size() in rxe_rcv"), a single unauthenticated UDP packet can still trigger panic. That patch handled payload_size() underflow only for valid opcodes with short packets, not for packets carrying an unknown opcode. The unknown-opcode OOB read described below predates that commit and reaches back to the initial Soft RoCE driver. The check added there reads pkt->paylen < header_size(pkt) + bth_pad(pkt) + RXE_ICRC_SIZE where header_size(pkt) expands to rxe_opcode[pkt->opcode].length. The rxe_opcode[] array has 256 entries but is only populated for defined IB opcodes; any other entry (for example opcode 0xff) is zero-initialized, so length == 0 and the check degenerates to pkt->paylen < 0 + bth_pad(pkt) + RXE_ICRC_SIZE which does not constrain pkt->paylen enough. rxe_icrc_hdr() then computes rxe_opcode[pkt->opcode].length - RXE_BTH_BYTES which underflows when length == 0 and passes a huge value to rxe_crc32(), causing an out-of-bounds read of the skb payload. Reproduced on v7.0-rc7 with that fix applied, QEMU/KVM with CONFIG_RDMA_RXE=y and CONFIG_KASAN=y, after rdma link add rxe0 type rxe netdev eth0 A single 48-byte UDP packet to port 4791 with BTH opcode=0xff and QPN=IB_MULTICAST_QPN triggers: BUG: KASAN: slab-out-of-bounds in crc32_le+0x115/0x170 Read of size 1 at addr ... The buggy address is located 0 bytes to the right of allocated 704-byte region Call Trace: crc32_le+0x115/0x170 rxe_icrc_hdr.isra.0+0x226/0x300 rxe_icrc_check+0x13f/0x3a0 rxe_rcv+0x6e1/0x16e0 rxe_udp_encap_recv+0x20a/0x320 udp_queue_rcv_one_skb+0x7ed/0x12c0 Subsequent packets with the same shape fault on unmapped memory and panic the kernel. The trigger requires only module load and "rdma link add"; no QP, no connection, and no authentication. Fix this by rejecting packets whose opcode has no rxe_opcode[] entry, detected via the zero mask or zero length, before any length arithmetic runs. Cc: stable@vger.kernel.org Fixes: 8700e3e7c485 ("Soft RoCE driver") Link: https://patch.msgid.link/r/20260414111555.3386793-1-michael.bommarito@gmail.com Assisted-by: Claude:claude-opus-4-6 Signed-off-by: Michael Bommarito Reviewed-by: Zhu Yanjun Signed-off-by: Jason Gunthorpe --- drivers/infiniband/sw/rxe/rxe_recv.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/infiniband/sw/rxe/rxe_recv.c b/drivers/infiniband/sw/rxe/rxe_recv.c index f79214738c2b..2d5e701ff961 100644 --- a/drivers/infiniband/sw/rxe/rxe_recv.c +++ b/drivers/infiniband/sw/rxe/rxe_recv.c @@ -330,6 +330,17 @@ void rxe_rcv(struct sk_buff *skb) pkt->qp = NULL; pkt->mask |= rxe_opcode[pkt->opcode].mask; + /* + * Unknown opcodes have a zero-initialized rxe_opcode[] entry, so + * both mask and length are 0. Reject them before any length math: + * rxe_icrc_hdr() would otherwise compute length - RXE_BTH_BYTES + * and pass the underflowed value to rxe_crc32(), producing an + * out-of-bounds read. + */ + if (unlikely(!rxe_opcode[pkt->opcode].mask || + !rxe_opcode[pkt->opcode].length)) + goto drop; + if (unlikely(pkt->paylen < header_size(pkt) + bth_pad(pkt) + RXE_ICRC_SIZE)) goto drop; From 1114c87aa6f195cf07da55a27b2122ae26557b26 Mon Sep 17 00:00:00 2001 From: Michael Bommarito Date: Sat, 18 Apr 2026 12:21:41 -0400 Subject: [PATCH 072/525] RDMA/rxe: Reject non-8-byte ATOMIC_WRITE payloads atomic_write_reply() at drivers/infiniband/sw/rxe/rxe_resp.c unconditionally dereferences 8 bytes at payload_addr(pkt): value = *(u64 *)payload_addr(pkt); check_rkey() previously accepted an ATOMIC_WRITE request with pktlen == resid == 0 because the length validation only compared pktlen against resid. A remote initiator that sets the RETH length to 0 therefore reaches atomic_write_reply() with a zero-byte logical payload, and the responder reads sizeof(u64) bytes from past the logical end of the packet into skb->head tailroom, then writes those 8 bytes into the attacker's MR via rxe_mr_do_atomic_write(). That is a remote disclosure of 4 bytes of kernel tailroom per probe (the other 4 bytes are the packet's own trailing ICRC). IBA oA19-28 defines ATOMIC_WRITE as exactly 8 bytes. Anything else is protocol-invalid. Hoist a strict length check into check_rkey() so the responder never reaches the unchecked dereference, and keep the existing WRITE-family length logic for the normal RDMA WRITE path. Reproduced on mainline with an unmodified rxe driver: a sustained zero-length ATOMIC_WRITE probe repeatedly leaks adjacent skb head-buffer bytes into the attacker's MR, including recognisable kernel strings and partial kernel-direct-map pointer words. With this patch applied the responder rejects the PDU and the MR stays all-zero. Cc: stable@vger.kernel.org Fixes: 034e285f8b99 ("RDMA/rxe: Make responder support atomic write on RC service") Link: https://patch.msgid.link/r/20260418162141.3610201-1-michael.bommarito@gmail.com Assisted-by: Claude:claude-opus-4-7 Signed-off-by: Michael Bommarito Reviewed-by: Zhu Yanjun Signed-off-by: Jason Gunthorpe --- drivers/infiniband/sw/rxe/rxe_resp.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index 9faf8c09aa8e..9cb2f6fbf2dd 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -540,7 +540,19 @@ static enum resp_states check_rkey(struct rxe_qp *qp, } skip_check_range: - if (pkt->mask & (RXE_WRITE_MASK | RXE_ATOMIC_WRITE_MASK)) { + if (pkt->mask & RXE_ATOMIC_WRITE_MASK) { + /* IBA oA19-28: ATOMIC_WRITE payload is exactly 8 bytes. + * Reject any other length before the responder reads + * sizeof(u64) bytes from payload_addr(pkt); a shorter + * payload would read past the logical end of the packet + * into skb->head tailroom. + */ + if (resid != sizeof(u64) || pktlen != sizeof(u64) || + bth_pad(pkt)) { + state = RESPST_ERR_LENGTH; + goto err; + } + } else if (pkt->mask & RXE_WRITE_MASK) { if (resid > mtu) { if (pktlen != mtu || bth_pad(pkt)) { state = RESPST_ERR_LENGTH; From c488df06bd552bb8b6e14fa0cfd5ad986c6e9525 Mon Sep 17 00:00:00 2001 From: Junrui Luo Date: Fri, 24 Apr 2026 13:51:02 +0800 Subject: [PATCH 073/525] RDMA/mlx5: Fix error path fall-through in mlx5_ib_dev_res_srq_init() mlx5_ib_dev_res_srq_init() allocates two SRQs, s0 and s1. When ib_create_srq() fails for s1, the error branch destroys s0 but falls through and unconditionally assigns the freed s0 and the ERR_PTR s1 to devr->s0 and devr->s1. This leads to several problems: the lock-free fast path checks "if (devr->s1) return 0;" and treats the ERR_PTR as already initialised; users in mlx5_ib_create_qp() dereference the freed SRQ or ERR_PTR via to_msrq(devr->s0)->msrq.srqn; and mlx5_ib_dev_res_cleanup() dereferences the ERR_PTR and double-frees s0 on teardown. Fix by adding the same `goto unlock` in the s1 failure path. Cc: stable@vger.kernel.org Fixes: 5895e70f2e6e ("IB/mlx5: Allocate resources just before first QP/SRQ is created") Link: https://patch.msgid.link/r/SYBPR01MB7881E1E0970268BD69C0BA75AF2B2@SYBPR01MB7881.ausprd01.prod.outlook.com Reported-by: Yuhao Jiang Signed-off-by: Junrui Luo Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mlx5/main.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index 109661c2ac12..8115ae869ef2 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -3392,6 +3392,7 @@ int mlx5_ib_dev_res_srq_init(struct mlx5_ib_dev *dev) "Couldn't create SRQ 1 for res init, err=%pe\n", s1); ib_destroy_srq(s0); + goto unlock; } devr->s0 = s0; From 8135b89a376ab2c20db3921c9c38cf83994b1f1c Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Thu, 23 Apr 2026 09:10:52 +0200 Subject: [PATCH 074/525] Revert "parisc: led: fix reference leak on failed device registration" This reverts commit 707610bcccbd0327530938e33f3f33211a640a4e. platform_device_register() is going to be fixed instead. Signed-off-by: Helge Deller --- drivers/parisc/led.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/parisc/led.c b/drivers/parisc/led.c index b299fcc48b08..016c9d5a60a8 100644 --- a/drivers/parisc/led.c +++ b/drivers/parisc/led.c @@ -543,10 +543,8 @@ static void __init register_led_regions(void) static int __init startup_leds(void) { - if (platform_device_register(&platform_leds)) { - pr_info("LED: failed to register LEDs\n"); - platform_device_put(&platform_leds); - } + if (platform_device_register(&platform_leds)) + printk(KERN_INFO "LED: failed to register LEDs\n"); register_led_regions(); return 0; } From b8425ceefee5df2b8490fa70a07e4e402c752492 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Fri, 24 Apr 2026 12:32:11 +0200 Subject: [PATCH 075/525] parisc: drivers: switch to dynamic root device Driver core expects devices to be dynamically allocated and will, for example, complain loudly if a device that lacks a release function is ever freed. Use root_device_register() to allocate and register the root device instead of open coding using a static device. While at it, drop the redundant additional reference taken at init. Signed-off-by: Johan Hovold Signed-off-by: Helge Deller --- arch/parisc/kernel/drivers.c | 21 ++++++++++----------- 1 file changed, 10 insertions(+), 11 deletions(-) diff --git a/arch/parisc/kernel/drivers.c b/arch/parisc/kernel/drivers.c index bc47bbe3026e..b52ad704ec8a 100644 --- a/arch/parisc/kernel/drivers.c +++ b/arch/parisc/kernel/drivers.c @@ -41,9 +41,7 @@ const struct dma_map_ops *hppa_dma_ops __ro_after_init; EXPORT_SYMBOL(hppa_dma_ops); -static struct device root = { - .init_name = "parisc", -}; +static struct device *root; static inline int check_dev(struct device *dev) { @@ -89,7 +87,7 @@ static int for_each_padev(int (*fn)(struct device *, void *), void * data) .obj = data, .fn = fn, }; - return device_for_each_child(&root, &recurse_data, descend_children); + return device_for_each_child(root, &recurse_data, descend_children); } /** @@ -290,7 +288,7 @@ const struct parisc_device * find_pa_parent_type(const struct parisc_device *padev, int type) { const struct device *dev = &padev->dev; - while (dev != &root) { + while (dev != root) { struct parisc_device *candidate = to_parisc_device(dev); if (candidate->id.hw_type == type) return candidate; @@ -319,7 +317,7 @@ static void get_node_path(struct device *dev, struct hardware_path *path) dev = dev->parent; } - while (dev != &root) { + while (dev != root) { if (dev_is_pci(dev)) { unsigned int devfn = to_pci_dev(dev)->devfn; path->bc[i--] = PCI_SLOT(devfn) | (PCI_FUNC(devfn)<< 5); @@ -482,7 +480,7 @@ static struct parisc_device * __init alloc_tree_node( static struct parisc_device *create_parisc_device(struct hardware_path *modpath) { int i; - struct device *parent = &root; + struct device *parent = root; for (i = 0; i < 6; i++) { if (modpath->bc[i] == -1) continue; @@ -755,7 +753,7 @@ parse_tree_node(struct device *parent, int index, struct hardware_path *modpath) struct device *hwpath_to_device(struct hardware_path *modpath) { int i; - struct device *parent = &root; + struct device *parent = root; for (i = 0; i < 6; i++) { if (modpath->bc[i] == -1) continue; @@ -880,7 +878,7 @@ void __init walk_central_bus(void) { walk_native_bus(CENTRAL_BUS_ADDR, CENTRAL_BUS_ADDR + (MAX_NATIVE_DEVICES * NATIVE_DEVICE_OFFSET), - &root); + root); } static __init void print_parisc_device(struct parisc_device *dev) @@ -907,9 +905,10 @@ void __init init_parisc_bus(void) { if (bus_register(&parisc_bus_type)) panic("Could not register PA-RISC bus type\n"); - if (device_register(&root)) + + root = root_device_register("parisc"); + if (IS_ERR(root)) panic("Could not register PA-RISC root device\n"); - get_device(&root); } static __init void qemu_header(void) From 0de4cb473aed57ee4ba7e0551ad27bddc19fc519 Mon Sep 17 00:00:00 2001 From: Breno Leitao Date: Tue, 28 Apr 2026 08:10:43 -0700 Subject: [PATCH 076/525] workqueue: fix devm_alloc_workqueue() va_list misuse devm_alloc_workqueue() built a va_list and passed it as a single positional argument to the variadic alloc_workqueue() macro: va_start(args, max_active); wq = alloc_workqueue(fmt, flags, max_active, args); va_end(args); C does not allow forwarding a va_list through a ... parameter. alloc_workqueue() expands to alloc_workqueue_noprof(), which runs its own va_start() over its ... params, so the inner vsnprintf(wq->name, sizeof(wq->name), fmt, args) in __alloc_workqueue() received the outer va_list object as the first variadic slot rather than the caller's actual format arguments. Add a new static helper alloc_workqueue_va() that wraps __alloc_workqueue() and runs wq_init_lockdep() on success, and fold both alloc_workqueue_noprof() and devm_alloc_workqueue_noprof() onto it as suggested by Tejun. The wq_init_lockdep() step is required on the devm path too, otherwise __flush_workqueue()'s on-stack COMPLETION_INITIALIZER_ONSTACK_MAP would NULL-deref wq->lockdep_map. No caller changes are required. devm_alloc_ordered_workqueue() is a macro forwarding to devm_alloc_workqueue() and inherits the fix. Two in-tree callers actively trigger the broken path on every probe: drivers/power/supply/mt6370-charger.c:889 drivers/power/supply/max77705_charger.c:649 both of which use devm_alloc_ordered_workqueue(dev, "%s", 0, dev_name(dev)). A standalone reproducer module is available at[1]. Link: https://github.com/leitao/debug/blob/main/workqueue/valist/wq_va_test.c [1] Fixes: 1dfc9d60a69e ("workqueue: devres: Add device-managed allocate workqueue") Signed-off-by: Breno Leitao Signed-off-by: Tejun Heo --- include/linux/workqueue.h | 6 ++++-- kernel/workqueue.c | 28 +++++++++++++++++++--------- 2 files changed, 23 insertions(+), 11 deletions(-) diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h index ab6cb70ca1a5..6177624539b3 100644 --- a/include/linux/workqueue.h +++ b/include/linux/workqueue.h @@ -534,8 +534,10 @@ alloc_workqueue_noprof(const char *fmt, unsigned int flags, int max_active, ...) * Pointer to the allocated workqueue on success, %NULL on failure. */ __printf(2, 5) struct workqueue_struct * -devm_alloc_workqueue(struct device *dev, const char *fmt, unsigned int flags, - int max_active, ...); +devm_alloc_workqueue_noprof(struct device *dev, const char *fmt, + unsigned int flags, int max_active, ...); +#define devm_alloc_workqueue(...) \ + alloc_hooks(devm_alloc_workqueue_noprof(__VA_ARGS__)) #ifdef CONFIG_LOCKDEP /** diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 5f747f241a5f..24d0265191d4 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -5906,6 +5906,20 @@ static struct workqueue_struct *__alloc_workqueue(const char *fmt, return NULL; } +static struct workqueue_struct *alloc_workqueue_va(const char *fmt, + unsigned int flags, + int max_active, + va_list args) +{ + struct workqueue_struct *wq; + + wq = __alloc_workqueue(fmt, flags, max_active, args); + if (wq) + wq_init_lockdep(wq); + + return wq; +} + __printf(1, 4) struct workqueue_struct *alloc_workqueue_noprof(const char *fmt, unsigned int flags, @@ -5915,12 +5929,8 @@ struct workqueue_struct *alloc_workqueue_noprof(const char *fmt, va_list args; va_start(args, max_active); - wq = __alloc_workqueue(fmt, flags, max_active, args); + wq = alloc_workqueue_va(fmt, flags, max_active, args); va_end(args); - if (!wq) - return NULL; - - wq_init_lockdep(wq); return wq; } @@ -5932,15 +5942,15 @@ static void devm_workqueue_release(void *res) } __printf(2, 5) struct workqueue_struct * -devm_alloc_workqueue(struct device *dev, const char *fmt, unsigned int flags, - int max_active, ...) +devm_alloc_workqueue_noprof(struct device *dev, const char *fmt, + unsigned int flags, int max_active, ...) { struct workqueue_struct *wq; va_list args; int ret; va_start(args, max_active); - wq = alloc_workqueue(fmt, flags, max_active, args); + wq = alloc_workqueue_va(fmt, flags, max_active, args); va_end(args); if (!wq) return NULL; @@ -5951,7 +5961,7 @@ devm_alloc_workqueue(struct device *dev, const char *fmt, unsigned int flags, return wq; } -EXPORT_SYMBOL_GPL(devm_alloc_workqueue); +EXPORT_SYMBOL_GPL(devm_alloc_workqueue_noprof); #ifdef CONFIG_LOCKDEP __printf(1, 5) From c4cca236968683eb0d59abfb12d5c7e4d8514227 Mon Sep 17 00:00:00 2001 From: Corey Minyard Date: Mon, 20 Apr 2026 12:39:51 -0500 Subject: [PATCH 077/525] ipmi: Add limits to event and receive message requests The driver would just fetch events and receive messages until the BMC said it was done. To avoid issues with BMCs that never say they are done, add a limit of 10 fetches at a time. In addition, an si interface has an attn state it can return from the hardware which is supposed to cause a flag fetch to see if the driver needs to fetch events or message or a few other things. If the attn bit gets stuck, it's a similar problem. So allow messages in between flag fetches so the driver itself doesn't get stuck. This is a more general fix than the previous fix for the specific bad BMC, but should fix the more general issue of a BMC that won't stop saying it has data. This has been there from the beginning of the driver. It's not a bug per-se, but it is accounting for bugs in BMCs. Reported-by: Matt Fleming Closes: https://lore.kernel.org/lkml/20260415115930.3428942-1-matt@readmodwrite.com/ Fixes: <1da177e4c3f4> ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Signed-off-by: Corey Minyard --- drivers/char/ipmi/ipmi_si_intf.c | 54 +++++++++++++++++++++++++------- drivers/char/ipmi/ipmi_ssif.c | 23 ++++++++++++-- 2 files changed, 64 insertions(+), 13 deletions(-) diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 08c208cc64c5..7c3c463e08da 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -168,6 +168,10 @@ struct smi_info { OEM2_DATA_AVAIL) unsigned char msg_flags; + /* When requesting events and messages, don't do it forever. */ + unsigned int num_requests_in_a_row; + bool last_was_flag_fetch; + /* Does the BMC have an event buffer? */ bool has_event_buffer; @@ -410,7 +414,10 @@ static void start_getting_msg_queue(struct smi_info *smi_info) start_new_msg(smi_info, smi_info->curr_msg->data, smi_info->curr_msg->data_size); - smi_info->si_state = SI_GETTING_MESSAGES; + if (smi_info->si_state != SI_GETTING_MESSAGES) { + smi_info->num_requests_in_a_row = 0; + smi_info->si_state = SI_GETTING_MESSAGES; + } } static void start_getting_events(struct smi_info *smi_info) @@ -421,7 +428,10 @@ static void start_getting_events(struct smi_info *smi_info) start_new_msg(smi_info, smi_info->curr_msg->data, smi_info->curr_msg->data_size); - smi_info->si_state = SI_GETTING_EVENTS; + if (smi_info->si_state != SI_GETTING_EVENTS) { + smi_info->num_requests_in_a_row = 0; + smi_info->si_state = SI_GETTING_EVENTS; + } } /* @@ -595,6 +605,7 @@ static void handle_transaction_done(struct smi_info *smi_info) smi_info->si_state = SI_NORMAL; } else { smi_info->msg_flags = msg[3]; + smi_info->last_was_flag_fetch = true; handle_flags(smi_info); } break; @@ -646,6 +657,11 @@ static void handle_transaction_done(struct smi_info *smi_info) } else { smi_inc_stat(smi_info, events); + smi_info->num_requests_in_a_row++; + if (smi_info->num_requests_in_a_row > 10) + /* Stop if we do this too many times. */ + smi_info->msg_flags &= ~EVENT_MSG_BUFFER_FULL; + /* * Do this before we deliver the message * because delivering the message releases the @@ -684,6 +700,11 @@ static void handle_transaction_done(struct smi_info *smi_info) } else { smi_inc_stat(smi_info, incoming_messages); + smi_info->num_requests_in_a_row++; + if (smi_info->num_requests_in_a_row > 10) + /* Stop if we do this too many times. */ + smi_info->msg_flags &= ~RECEIVE_MSG_AVAIL; + /* * Do this before we deliver the message * because delivering the message releases the @@ -825,6 +846,26 @@ static enum si_sm_result smi_event_handler(struct smi_info *smi_info, goto out; } + /* + * If we are currently idle, or if the last thing that was + * done was a flag fetch and there is a message pending, try + * to start the next message. + * + * We do the waiting message check to avoid a stuck flag + * completely wedging the driver. Let a message through + * in between flag operations if that happens. + */ + if (si_sm_result == SI_SM_IDLE || + (si_sm_result == SI_SM_ATTN && smi_info->waiting_msg && + smi_info->last_was_flag_fetch)) { + smi_info->last_was_flag_fetch = false; + smi_inc_stat(smi_info, idles); + + si_sm_result = start_next_msg(smi_info); + if (si_sm_result != SI_SM_IDLE) + goto restart; + } + /* * We prefer handling attn over new messages. But don't do * this if there is not yet an upper layer to handle anything. @@ -852,15 +893,6 @@ static enum si_sm_result smi_event_handler(struct smi_info *smi_info, } } - /* If we are currently idle, try to start the next message. */ - if (si_sm_result == SI_SM_IDLE) { - smi_inc_stat(smi_info, idles); - - si_sm_result = start_next_msg(smi_info); - if (si_sm_result != SI_SM_IDLE) - goto restart; - } - if ((si_sm_result == SI_SM_IDLE) && (atomic_read(&smi_info->req_events))) { /* diff --git a/drivers/char/ipmi/ipmi_ssif.c b/drivers/char/ipmi/ipmi_ssif.c index b49500a1bd36..f3798f4e6a63 100644 --- a/drivers/char/ipmi/ipmi_ssif.c +++ b/drivers/char/ipmi/ipmi_ssif.c @@ -225,6 +225,9 @@ struct ssif_info { bool has_event_buffer; bool supports_alert; + /* When requesting events and messages, don't do it forever. */ + unsigned int num_requests_in_a_row; + /* * Used to tell what we should do with alerts. If we are * waiting on a response, read the data immediately. @@ -413,7 +416,10 @@ static void start_event_fetch(struct ssif_info *ssif_info, unsigned long *flags) } ssif_info->curr_msg = msg; - ssif_info->ssif_state = SSIF_GETTING_EVENTS; + if (ssif_info->ssif_state != SSIF_GETTING_EVENTS) { + ssif_info->num_requests_in_a_row = 0; + ssif_info->ssif_state = SSIF_GETTING_EVENTS; + } ipmi_ssif_unlock_cond(ssif_info, flags); msg->data[0] = (IPMI_NETFN_APP_REQUEST << 2); @@ -436,7 +442,10 @@ static void start_recv_msg_fetch(struct ssif_info *ssif_info, } ssif_info->curr_msg = msg; - ssif_info->ssif_state = SSIF_GETTING_MESSAGES; + if (ssif_info->ssif_state != SSIF_GETTING_MESSAGES) { + ssif_info->num_requests_in_a_row = 0; + ssif_info->ssif_state = SSIF_GETTING_MESSAGES; + } ipmi_ssif_unlock_cond(ssif_info, flags); msg->data[0] = (IPMI_NETFN_APP_REQUEST << 2); @@ -843,6 +852,11 @@ static void msg_done_handler(struct ssif_info *ssif_info, int result, ssif_info->msg_flags &= ~EVENT_MSG_BUFFER_FULL; handle_flags(ssif_info, flags); } else { + ssif_info->num_requests_in_a_row++; + if (ssif_info->num_requests_in_a_row > 10) + /* Stop if we do this too many times. */ + ssif_info->msg_flags &= ~EVENT_MSG_BUFFER_FULL; + handle_flags(ssif_info, flags); ssif_inc_stat(ssif_info, events); deliver_recv_msg(ssif_info, msg); @@ -876,6 +890,11 @@ static void msg_done_handler(struct ssif_info *ssif_info, int result, ssif_info->msg_flags &= ~RECEIVE_MSG_AVAIL; handle_flags(ssif_info, flags); } else { + ssif_info->num_requests_in_a_row++; + if (ssif_info->num_requests_in_a_row > 10) + /* Stop if we do this too many times. */ + ssif_info->msg_flags &= ~RECEIVE_MSG_AVAIL; + ssif_inc_stat(ssif_info, incoming_messages); handle_flags(ssif_info, flags); deliver_recv_msg(ssif_info, msg); From 09dd798270ff582d7309f285d4aaf5dbebae01cb Mon Sep 17 00:00:00 2001 From: Corey Minyard Date: Mon, 20 Apr 2026 13:02:18 -0500 Subject: [PATCH 078/525] ipmi:si: Return state to normal if message allocation fails There were places where nothing would get started if a message allocation failed, so the driver needs to return to normal state. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: Signed-off-by: Corey Minyard --- drivers/char/ipmi/ipmi_si_intf.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 7c3c463e08da..9a9d12be9bf7 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -497,15 +497,19 @@ static void handle_flags(struct smi_info *smi_info) } else if (smi_info->msg_flags & RECEIVE_MSG_AVAIL) { /* Messages available. */ smi_info->curr_msg = alloc_msg_handle_irq(smi_info); - if (!smi_info->curr_msg) + if (!smi_info->curr_msg) { + smi_info->si_state = SI_NORMAL; return; + } start_getting_msg_queue(smi_info); } else if (smi_info->msg_flags & EVENT_MSG_BUFFER_FULL) { /* Events available. */ smi_info->curr_msg = alloc_msg_handle_irq(smi_info); - if (!smi_info->curr_msg) + if (!smi_info->curr_msg) { + smi_info->si_state = SI_NORMAL; return; + } start_getting_events(smi_info); } else if (smi_info->msg_flags & OEM_DATA_AVAIL && From a8aebe93a4938c0ca1941eeaae821738f869be3d Mon Sep 17 00:00:00 2001 From: Corey Minyard Date: Tue, 21 Apr 2026 06:50:22 -0500 Subject: [PATCH 079/525] ipmi:ssif: NULL thread on error Cleanup code was checking the thread for NULL, but it was possibly a PTR_ERR() in one spot. Spotted with static analysis. Link: https://sourceforge.net/p/openipmi/mailman/message/59324676/ Fixes: 75c486cb1bca ("ipmi:ssif: Clean up kthread on errors") Cc: # 91eb7ec72612: ipmi:ssif: Remove unnecessary indention Cc: stable@vger.kernel.org Signed-off-by: Corey Minyard --- drivers/char/ipmi/ipmi_ssif.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/char/ipmi/ipmi_ssif.c b/drivers/char/ipmi/ipmi_ssif.c index f3798f4e6a63..f419b46bf002 100644 --- a/drivers/char/ipmi/ipmi_ssif.c +++ b/drivers/char/ipmi/ipmi_ssif.c @@ -1905,6 +1905,7 @@ static int ssif_probe(struct i2c_client *client) "kssif%4.4x", thread_num); if (IS_ERR(ssif_info->thread)) { rv = PTR_ERR(ssif_info->thread); + ssif_info->thread = NULL; dev_notice(&ssif_info->client->dev, "Could not start kernel thread: error %d\n", rv); From 3e75021f615ceee8562e6455c335936b39929ffb Mon Sep 17 00:00:00 2001 From: Troy Mitchell Date: Fri, 24 Apr 2026 16:20:32 +0800 Subject: [PATCH 080/525] clk: spacemit: k3: mark top_dclk as CLK_IS_CRITICAL top_dclk is the DDR bus clock. If it is gated by clk_disable_unused, all memory-mapped bus transactions cease to function, causing DMA engines to hang and general system instability. Mark it CLK_IS_CRITICAL so the CCF never gates it during the unused clock sweep. Fixes: e371a77255b8 ("clk: spacemit: k3: add the clock tree") Reviewed-by: Brian Masney Signed-off-by: Troy Mitchell Signed-off-by: Stephen Boyd --- drivers/clk/spacemit/ccu-k3.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/clk/spacemit/ccu-k3.c b/drivers/clk/spacemit/ccu-k3.c index e98afd59f05c..bb8b75bdbdb3 100644 --- a/drivers/clk/spacemit/ccu-k3.c +++ b/drivers/clk/spacemit/ccu-k3.c @@ -846,7 +846,7 @@ static const struct clk_parent_data top_parents[] = { CCU_PARENT_HW(pll6_d3), }; CCU_MUX_DIV_GATE_FC_DEFINE(top_dclk, top_parents, APMU_TOP_DCLK_CTRL, 5, 3, - BIT(8), 2, 3, BIT(1), 0); + BIT(8), 2, 3, BIT(1), CLK_IS_CRITICAL); static const struct clk_parent_data ucie_parents[] = { CCU_PARENT_HW(pll1_d8_307p2), From 79a1886be1564a009cd2a003bada15ed6153f819 Mon Sep 17 00:00:00 2001 From: Jerome Brunet Date: Wed, 25 Feb 2026 17:55:05 +0100 Subject: [PATCH 081/525] clk: eyeq: use the auxiliary device creation helper MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The auxiliary device creation of this driver is simple enough to use the available auxiliary device creation helper. Use it and remove some boilerplate code. Tested-by: Théo Lebrun # On Mobileye EyeQ5 Signed-off-by: Jerome Brunet Reviewed-by: Luca Ceresoli Signed-off-by: Théo Lebrun Signed-off-by: Stephen Boyd --- drivers/clk/clk-eyeq.c | 57 +++++++++--------------------------------- 1 file changed, 12 insertions(+), 45 deletions(-) diff --git a/drivers/clk/clk-eyeq.c b/drivers/clk/clk-eyeq.c index c1dccedf8d5b..ae1aaf274336 100644 --- a/drivers/clk/clk-eyeq.c +++ b/drivers/clk/clk-eyeq.c @@ -321,38 +321,18 @@ static void eqc_probe_init_fixed_factors(struct device *dev, } } -static void eqc_auxdev_release(struct device *dev) -{ - struct auxiliary_device *adev = to_auxiliary_dev(dev); - - kfree(adev); -} - -static int eqc_auxdev_create(struct device *dev, void __iomem *base, - const char *name, u32 id) +static void eqc_auxdev_create_optional(struct device *dev, void __iomem *base, + const char *name) { struct auxiliary_device *adev; - int ret; - adev = kzalloc_obj(*adev); - if (!adev) - return -ENOMEM; - - adev->name = name; - adev->dev.parent = dev; - adev->dev.platform_data = (void __force *)base; - adev->dev.release = eqc_auxdev_release; - adev->id = id; - - ret = auxiliary_device_init(adev); - if (ret) - return ret; - - ret = auxiliary_device_add(adev); - if (ret) - auxiliary_device_uninit(adev); - - return ret; + if (name) { + adev = devm_auxiliary_device_create(dev, name, + (void __force *)base); + if (!adev) + dev_warn(dev, "failed creating auxiliary device %s.%s\n", + KBUILD_MODNAME, name); + } } static int eqc_probe(struct platform_device *pdev) @@ -364,7 +344,6 @@ static int eqc_probe(struct platform_device *pdev) unsigned int i, clk_count; struct resource *res; void __iomem *base; - int ret; data = device_get_match_data(dev); if (!data) @@ -378,21 +357,9 @@ static int eqc_probe(struct platform_device *pdev) if (!base) return -ENOMEM; - /* Init optional reset auxiliary device. */ - if (data->reset_auxdev_name) { - ret = eqc_auxdev_create(dev, base, data->reset_auxdev_name, 0); - if (ret) - dev_warn(dev, "failed creating auxiliary device %s.%s: %d\n", - KBUILD_MODNAME, data->reset_auxdev_name, ret); - } - - /* Init optional pinctrl auxiliary device. */ - if (data->pinctrl_auxdev_name) { - ret = eqc_auxdev_create(dev, base, data->pinctrl_auxdev_name, 0); - if (ret) - dev_warn(dev, "failed creating auxiliary device %s.%s: %d\n", - KBUILD_MODNAME, data->pinctrl_auxdev_name, ret); - } + /* Init optional auxiliary devices. */ + eqc_auxdev_create_optional(dev, base, data->reset_auxdev_name); + eqc_auxdev_create_optional(dev, base, data->pinctrl_auxdev_name); if (data->pll_count + data->div_count + data->fixed_factor_count == 0) return 0; /* Zero clocks, we are done. */ From a25ab518f355e1f0dcbea24ee26418dfcd6944b5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Th=C3=A9o=20Lebrun?= Date: Wed, 25 Feb 2026 17:55:06 +0100 Subject: [PATCH 082/525] clk: eyeq: add EyeQ5 children auxiliary device for generic PHYs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Grow our clk-eyeq family; it knows how to spawn reset provider and pin controller children. Expand with a generic PHY driver on EyeQ5. Reviewed-by: Luca Ceresoli Signed-off-by: Théo Lebrun Signed-off-by: Stephen Boyd --- drivers/clk/clk-eyeq.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/clk/clk-eyeq.c b/drivers/clk/clk-eyeq.c index ae1aaf274336..d9303c2c7aa5 100644 --- a/drivers/clk/clk-eyeq.c +++ b/drivers/clk/clk-eyeq.c @@ -110,6 +110,7 @@ struct eqc_match_data { const char *reset_auxdev_name; const char *pinctrl_auxdev_name; + const char *eth_phy_auxdev_name; unsigned int early_clk_count; }; @@ -360,6 +361,7 @@ static int eqc_probe(struct platform_device *pdev) /* Init optional auxiliary devices. */ eqc_auxdev_create_optional(dev, base, data->reset_auxdev_name); eqc_auxdev_create_optional(dev, base, data->pinctrl_auxdev_name); + eqc_auxdev_create_optional(dev, base, data->eth_phy_auxdev_name); if (data->pll_count + data->div_count + data->fixed_factor_count == 0) return 0; /* Zero clocks, we are done. */ @@ -520,6 +522,7 @@ static const struct eqc_match_data eqc_eyeq5_match_data = { .reset_auxdev_name = "reset", .pinctrl_auxdev_name = "pinctrl", + .eth_phy_auxdev_name = "phy", .early_clk_count = ARRAY_SIZE(eqc_eyeq5_early_plls) + ARRAY_SIZE(eqc_eyeq5_early_fixed_factors), From 4ac170432cf74b753cf59bcd0d449dced48585da Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Th=C3=A9o=20Lebrun?= Date: Wed, 25 Feb 2026 17:55:07 +0100 Subject: [PATCH 083/525] reset: eyeq: drop device_set_of_node_from_dev() done by parent MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Our parent driver (clk-eyeq) now does the device_set_of_node_from_dev(dev, dev->parent) call through the newly introduced devm_auxiliary_device_create() helper. Doing it again in the reset-eyeq probe would be redundant. Drop both the WARN_ON() and the device_set_of_node_from_dev() call. Also fix the following comment that talks about "our newfound OF node". Signed-off-by: Jerome Brunet Reviewed-by: Philipp Zabel Acked-by: Philipp Zabel Signed-off-by: Théo Lebrun Signed-off-by: Stephen Boyd --- drivers/reset/reset-eyeq.c | 24 ++---------------------- 1 file changed, 2 insertions(+), 22 deletions(-) diff --git a/drivers/reset/reset-eyeq.c b/drivers/reset/reset-eyeq.c index 791b7283111e..1a3857983897 100644 --- a/drivers/reset/reset-eyeq.c +++ b/drivers/reset/reset-eyeq.c @@ -422,13 +422,6 @@ static int eqr_of_xlate_twocells(struct reset_controller_dev *rcdev, return eqr_of_xlate_internal(rcdev, reset_spec->args[0], reset_spec->args[1]); } -static void eqr_of_node_put(void *_dev) -{ - struct device *dev = _dev; - - of_node_put(dev->of_node); -} - static int eqr_probe(struct auxiliary_device *adev, const struct auxiliary_device_id *id) { @@ -439,21 +432,8 @@ static int eqr_probe(struct auxiliary_device *adev, int ret; /* - * We are an auxiliary device of clk-eyeq. We do not have an OF node by - * default; let's reuse our parent's OF node. - */ - WARN_ON(dev->of_node); - device_set_of_node_from_dev(dev, dev->parent); - if (!dev->of_node) - return -ENODEV; - - ret = devm_add_action_or_reset(dev, eqr_of_node_put, dev); - if (ret) - return ret; - - /* - * Using our newfound OF node, we can get match data. We cannot use - * device_get_match_data() because it does not match reused OF nodes. + * Get match data. We cannot use device_get_match_data() because it does + * not accept reused OF nodes; see device_set_of_node_from_dev(). */ match = of_match_node(dev->driver->of_match_table, dev->of_node); if (!match || !match->data) From 69056753231f483cc1e40db52228aac42fd4c93d Mon Sep 17 00:00:00 2001 From: Brian Masney Date: Wed, 15 Apr 2026 16:30:49 -0400 Subject: [PATCH 084/525] MAINTAINERS: add myself as a reviewer for the clk subsystem I've reviewed a lot clk patches for parts of the subsystem that typically doesn't get much review. Add myself as a reviewer so that I don't miss anything. Link: https://lore.kernel.org/linux-clk/?q=f%3Abmasney%40redhat.com Signed-off-by: Brian Masney Signed-off-by: Stephen Boyd --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index 2fb1c75afd16..384ce5ee7323 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -6358,6 +6358,7 @@ F: include/uapi/linux/comedi.h COMMON CLK FRAMEWORK M: Michael Turquette M: Stephen Boyd +R: Brian Masney L: linux-clk@vger.kernel.org S: Maintained Q: http://patchwork.kernel.org/project/linux-clk/list/ From de019f203b0d472c98ead4081ad4f05d92c9b826 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 7 Apr 2026 11:50:27 +0200 Subject: [PATCH 085/525] clk: rk808: fix OF node reference imbalance The driver reuses the OF node of the parent multi-function device but fails to take another reference to balance the one dropped by the platform bus code when unbinding the MFD and deregistering the child devices. Fix this by using the intended helper for reusing OF nodes. Fixes: 2dc51ca822e4 ("clk: RK808: Reduce 'struct rk808' usage") Cc: stable@vger.kernel.org # 6.5 Cc: Sebastian Reichel Signed-off-by: Johan Hovold Reviewed-by: Sebastian Reichel Reviewed-by: Brian Masney Reviewed-by: Heiko Stuebner Signed-off-by: Stephen Boyd --- drivers/clk/clk-rk808.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/clk/clk-rk808.c b/drivers/clk/clk-rk808.c index f7412b137e5e..5a75b5c91555 100644 --- a/drivers/clk/clk-rk808.c +++ b/drivers/clk/clk-rk808.c @@ -153,7 +153,7 @@ static int rk808_clkout_probe(struct platform_device *pdev) struct rk808_clkout *rk808_clkout; int ret; - dev->of_node = pdev->dev.parent->of_node; + device_set_of_node_from_dev(dev, dev->parent); rk808_clkout = devm_kzalloc(dev, sizeof(*rk808_clkout), GFP_KERNEL); From 883a32793c86091ea37bb84f88cc697d019e7a5d Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Tue, 28 Apr 2026 12:38:47 +0200 Subject: [PATCH 086/525] efi/libstub: Move efi_relocate_kernel() into its only remaining user LoongArch is the only arch that still uses efi_relocate_kernel(), so before making changes to it that LoongArch needs, turn it into a private function. Move efi_low_alloc_above() into mem.c while at it, and drop the relocate.c source file altogether. Tested-by: WANG Rui Reviewed-by: Thomas Huth Signed-off-by: Ard Biesheuvel --- drivers/firmware/efi/libstub/Makefile | 2 +- drivers/firmware/efi/libstub/efistub.h | 7 - drivers/firmware/efi/libstub/loongarch-stub.c | 79 +++++++++ drivers/firmware/efi/libstub/mem.c | 82 +++++++++ drivers/firmware/efi/libstub/relocate.c | 166 ------------------ 5 files changed, 162 insertions(+), 174 deletions(-) delete mode 100644 drivers/firmware/efi/libstub/relocate.c diff --git a/drivers/firmware/efi/libstub/Makefile b/drivers/firmware/efi/libstub/Makefile index 983a438e35f3..cfedb3025c26 100644 --- a/drivers/firmware/efi/libstub/Makefile +++ b/drivers/firmware/efi/libstub/Makefile @@ -66,7 +66,7 @@ KBUILD_AFLAGS := $(KBUILD_CFLAGS) -D__ASSEMBLY__ lib-y := efi-stub-helper.o gop.o secureboot.o tpm.o \ file.o mem.o random.o randomalloc.o pci.o \ skip_spaces.o lib-cmdline.o lib-ctype.o \ - alignedmem.o relocate.o printk.o vsprintf.o + alignedmem.o printk.o vsprintf.o # include the stub's libfdt dependencies from lib/ when needed libfdt-deps := fdt_rw.c fdt_ro.c fdt_wip.c fdt.c \ diff --git a/drivers/firmware/efi/libstub/efistub.h b/drivers/firmware/efi/libstub/efistub.h index 979a21818cc1..fd91fc15ec81 100644 --- a/drivers/firmware/efi/libstub/efistub.h +++ b/drivers/firmware/efi/libstub/efistub.h @@ -1104,13 +1104,6 @@ efi_status_t efi_allocate_pages_aligned(unsigned long size, unsigned long *addr, efi_status_t efi_low_alloc_above(unsigned long size, unsigned long align, unsigned long *addr, unsigned long min); -efi_status_t efi_relocate_kernel(unsigned long *image_addr, - unsigned long image_size, - unsigned long alloc_size, - unsigned long preferred_addr, - unsigned long alignment, - unsigned long min_addr); - efi_status_t efi_parse_options(char const *cmdline); void efi_parse_option_graphics(char *option); diff --git a/drivers/firmware/efi/libstub/loongarch-stub.c b/drivers/firmware/efi/libstub/loongarch-stub.c index 736b6aae323d..8c538a5243d9 100644 --- a/drivers/firmware/efi/libstub/loongarch-stub.c +++ b/drivers/firmware/efi/libstub/loongarch-stub.c @@ -14,6 +14,85 @@ extern int kernel_asize; extern int kernel_fsize; extern int kernel_entry; +/** + * efi_relocate_kernel() - copy memory area + * @image_addr: pointer to address of memory area to copy + * @image_size: size of memory area to copy + * @alloc_size: minimum size of memory to allocate, must be greater or + * equal to image_size + * @preferred_addr: preferred target address + * @alignment: minimum alignment of the allocated memory area. It + * should be a power of two. + * @min_addr: minimum target address + * + * Copy a memory area to a newly allocated memory area aligned according + * to @alignment but at least EFI_ALLOC_ALIGN. If the preferred address + * is not available, the allocated address will not be below @min_addr. + * On exit, @image_addr is updated to the target copy address that was used. + * + * This function is used to copy the Linux kernel verbatim. It does not apply + * any relocation changes. + * + * Return: status code + */ +static +efi_status_t efi_relocate_kernel(unsigned long *image_addr, + unsigned long image_size, + unsigned long alloc_size, + unsigned long preferred_addr, + unsigned long alignment, + unsigned long min_addr) +{ + unsigned long cur_image_addr; + unsigned long new_addr = 0; + efi_status_t status; + unsigned long nr_pages; + efi_physical_addr_t efi_addr = preferred_addr; + + if (!image_addr || !image_size || !alloc_size) + return EFI_INVALID_PARAMETER; + if (alloc_size < image_size) + return EFI_INVALID_PARAMETER; + + cur_image_addr = *image_addr; + + /* + * The EFI firmware loader could have placed the kernel image + * anywhere in memory, but the kernel has restrictions on the + * max physical address it can run at. Some architectures + * also have a preferred address, so first try to relocate + * to the preferred address. If that fails, allocate as low + * as possible while respecting the required alignment. + */ + nr_pages = round_up(alloc_size, EFI_ALLOC_ALIGN) / EFI_PAGE_SIZE; + status = efi_bs_call(allocate_pages, EFI_ALLOCATE_ADDRESS, + EFI_LOADER_DATA, nr_pages, &efi_addr); + new_addr = efi_addr; + /* + * If preferred address allocation failed allocate as low as + * possible. + */ + if (status != EFI_SUCCESS) { + status = efi_low_alloc_above(alloc_size, alignment, &new_addr, + min_addr); + } + if (status != EFI_SUCCESS) { + efi_err("Failed to allocate usable memory for kernel.\n"); + return status; + } + + /* + * We know source/dest won't overlap since both memory ranges + * have been allocated by UEFI, so we can safely use memcpy. + */ + memcpy((void *)new_addr, (void *)cur_image_addr, image_size); + + /* Return the new address of the relocated image. */ + *image_addr = new_addr; + + return status; +} + efi_status_t handle_kernel_image(unsigned long *image_addr, unsigned long *image_size, unsigned long *reserve_addr, diff --git a/drivers/firmware/efi/libstub/mem.c b/drivers/firmware/efi/libstub/mem.c index 9c82259eea81..59f3f83de50c 100644 --- a/drivers/firmware/efi/libstub/mem.c +++ b/drivers/firmware/efi/libstub/mem.c @@ -124,3 +124,85 @@ void efi_free(unsigned long size, unsigned long addr) nr_pages = round_up(size, EFI_ALLOC_ALIGN) / EFI_PAGE_SIZE; efi_bs_call(free_pages, addr, nr_pages); } + +/** + * efi_low_alloc_above() - allocate pages at or above given address + * @size: size of the memory area to allocate + * @align: minimum alignment of the allocated memory area. It should + * a power of two. + * @addr: on exit the address of the allocated memory + * @min: minimum address to used for the memory allocation + * + * Allocate at the lowest possible address that is not below @min as + * EFI_LOADER_DATA. The allocated pages are aligned according to @align but at + * least EFI_ALLOC_ALIGN. The first allocated page will not below the address + * given by @min. + * + * Return: status code + */ +efi_status_t efi_low_alloc_above(unsigned long size, unsigned long align, + unsigned long *addr, unsigned long min) +{ + struct efi_boot_memmap *map __free(efi_pool) = NULL; + efi_status_t status; + unsigned long nr_pages; + int i; + + status = efi_get_memory_map(&map, false); + if (status != EFI_SUCCESS) + return status; + + /* + * Enforce minimum alignment that EFI or Linux requires when + * requesting a specific address. We are doing page-based (or + * larger) allocations, and both the address and size must meet + * alignment constraints. + */ + if (align < EFI_ALLOC_ALIGN) + align = EFI_ALLOC_ALIGN; + + size = round_up(size, EFI_ALLOC_ALIGN); + nr_pages = size / EFI_PAGE_SIZE; + for (i = 0; i < map->map_size / map->desc_size; i++) { + efi_memory_desc_t *desc; + unsigned long m = (unsigned long)map->map; + u64 start, end; + + desc = efi_memdesc_ptr(m, map->desc_size, i); + + if (desc->type != EFI_CONVENTIONAL_MEMORY) + continue; + + if (desc->attribute & EFI_MEMORY_HOT_PLUGGABLE) + continue; + + if (efi_soft_reserve_enabled() && + (desc->attribute & EFI_MEMORY_SP)) + continue; + + if (desc->num_pages < nr_pages) + continue; + + start = desc->phys_addr; + end = start + desc->num_pages * EFI_PAGE_SIZE; + + if (start < min) + start = min; + + start = round_up(start, align); + if ((start + size) > end) + continue; + + status = efi_bs_call(allocate_pages, EFI_ALLOCATE_ADDRESS, + EFI_LOADER_DATA, nr_pages, &start); + if (status == EFI_SUCCESS) { + *addr = start; + break; + } + } + + if (i == map->map_size / map->desc_size) + return EFI_NOT_FOUND; + + return EFI_SUCCESS; +} diff --git a/drivers/firmware/efi/libstub/relocate.c b/drivers/firmware/efi/libstub/relocate.c deleted file mode 100644 index d4264bfb6dc1..000000000000 --- a/drivers/firmware/efi/libstub/relocate.c +++ /dev/null @@ -1,166 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 - -#include -#include - -#include "efistub.h" - -/** - * efi_low_alloc_above() - allocate pages at or above given address - * @size: size of the memory area to allocate - * @align: minimum alignment of the allocated memory area. It should - * a power of two. - * @addr: on exit the address of the allocated memory - * @min: minimum address to used for the memory allocation - * - * Allocate at the lowest possible address that is not below @min as - * EFI_LOADER_DATA. The allocated pages are aligned according to @align but at - * least EFI_ALLOC_ALIGN. The first allocated page will not below the address - * given by @min. - * - * Return: status code - */ -efi_status_t efi_low_alloc_above(unsigned long size, unsigned long align, - unsigned long *addr, unsigned long min) -{ - struct efi_boot_memmap *map __free(efi_pool) = NULL; - efi_status_t status; - unsigned long nr_pages; - int i; - - status = efi_get_memory_map(&map, false); - if (status != EFI_SUCCESS) - return status; - - /* - * Enforce minimum alignment that EFI or Linux requires when - * requesting a specific address. We are doing page-based (or - * larger) allocations, and both the address and size must meet - * alignment constraints. - */ - if (align < EFI_ALLOC_ALIGN) - align = EFI_ALLOC_ALIGN; - - size = round_up(size, EFI_ALLOC_ALIGN); - nr_pages = size / EFI_PAGE_SIZE; - for (i = 0; i < map->map_size / map->desc_size; i++) { - efi_memory_desc_t *desc; - unsigned long m = (unsigned long)map->map; - u64 start, end; - - desc = efi_memdesc_ptr(m, map->desc_size, i); - - if (desc->type != EFI_CONVENTIONAL_MEMORY) - continue; - - if (desc->attribute & EFI_MEMORY_HOT_PLUGGABLE) - continue; - - if (efi_soft_reserve_enabled() && - (desc->attribute & EFI_MEMORY_SP)) - continue; - - if (desc->num_pages < nr_pages) - continue; - - start = desc->phys_addr; - end = start + desc->num_pages * EFI_PAGE_SIZE; - - if (start < min) - start = min; - - start = round_up(start, align); - if ((start + size) > end) - continue; - - status = efi_bs_call(allocate_pages, EFI_ALLOCATE_ADDRESS, - EFI_LOADER_DATA, nr_pages, &start); - if (status == EFI_SUCCESS) { - *addr = start; - break; - } - } - - if (i == map->map_size / map->desc_size) - return EFI_NOT_FOUND; - - return EFI_SUCCESS; -} - -/** - * efi_relocate_kernel() - copy memory area - * @image_addr: pointer to address of memory area to copy - * @image_size: size of memory area to copy - * @alloc_size: minimum size of memory to allocate, must be greater or - * equal to image_size - * @preferred_addr: preferred target address - * @alignment: minimum alignment of the allocated memory area. It - * should be a power of two. - * @min_addr: minimum target address - * - * Copy a memory area to a newly allocated memory area aligned according - * to @alignment but at least EFI_ALLOC_ALIGN. If the preferred address - * is not available, the allocated address will not be below @min_addr. - * On exit, @image_addr is updated to the target copy address that was used. - * - * This function is used to copy the Linux kernel verbatim. It does not apply - * any relocation changes. - * - * Return: status code - */ -efi_status_t efi_relocate_kernel(unsigned long *image_addr, - unsigned long image_size, - unsigned long alloc_size, - unsigned long preferred_addr, - unsigned long alignment, - unsigned long min_addr) -{ - unsigned long cur_image_addr; - unsigned long new_addr = 0; - efi_status_t status; - unsigned long nr_pages; - efi_physical_addr_t efi_addr = preferred_addr; - - if (!image_addr || !image_size || !alloc_size) - return EFI_INVALID_PARAMETER; - if (alloc_size < image_size) - return EFI_INVALID_PARAMETER; - - cur_image_addr = *image_addr; - - /* - * The EFI firmware loader could have placed the kernel image - * anywhere in memory, but the kernel has restrictions on the - * max physical address it can run at. Some architectures - * also have a preferred address, so first try to relocate - * to the preferred address. If that fails, allocate as low - * as possible while respecting the required alignment. - */ - nr_pages = round_up(alloc_size, EFI_ALLOC_ALIGN) / EFI_PAGE_SIZE; - status = efi_bs_call(allocate_pages, EFI_ALLOCATE_ADDRESS, - EFI_LOADER_DATA, nr_pages, &efi_addr); - new_addr = efi_addr; - /* - * If preferred address allocation failed allocate as low as - * possible. - */ - if (status != EFI_SUCCESS) { - status = efi_low_alloc_above(alloc_size, alignment, &new_addr, - min_addr); - } - if (status != EFI_SUCCESS) { - efi_err("Failed to allocate usable memory for kernel.\n"); - return status; - } - - /* - * We know source/dest won't overlap since both memory ranges - * have been allocated by UEFI, so we can safely use memcpy. - */ - memcpy((void *)new_addr, (void *)cur_image_addr, image_size); - - /* Return the new address of the relocated image. */ - *image_addr = new_addr; - - return status; -} From ad6f4f3ea72f866176f9dd6031c8778da088c686 Mon Sep 17 00:00:00 2001 From: WANG Rui Date: Mon, 27 Apr 2026 16:47:20 +0800 Subject: [PATCH 087/525] efi/loongarch: Implement efi_cache_sync_image() Provide a LoongArch implementation of efi_cache_sync_image() to ensure instruction cache coherency after the kernel image is relocated. Signed-off-by: WANG Rui Reviewed-by: Huacai Chen Signed-off-by: Ard Biesheuvel --- drivers/firmware/efi/libstub/loongarch.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/firmware/efi/libstub/loongarch.c b/drivers/firmware/efi/libstub/loongarch.c index 9825f5218137..f7938d5c196a 100644 --- a/drivers/firmware/efi/libstub/loongarch.c +++ b/drivers/firmware/efi/libstub/loongarch.c @@ -18,6 +18,11 @@ efi_status_t check_platform_features(void) return EFI_SUCCESS; } +void efi_cache_sync_image(unsigned long image_base, unsigned long alloc_size) +{ + asm volatile ("ibar 0" ::: "memory"); +} + struct exit_boot_struct { efi_memory_desc_t *runtime_map; int runtime_entry_count; From cda92ac47c024d84f6b8294e462d6272039a10ac Mon Sep 17 00:00:00 2001 From: WANG Rui Date: Mon, 27 Apr 2026 16:47:21 +0800 Subject: [PATCH 088/525] efi/libstub: Synchronize instruction cache after kernel relocation The relocated kernel image is copied to its new location using memcpy(). On architectures with separate instruction and data caches, the copied instructions may remain stale in the instruction cache, leading to the execution of outdated contents. Call efi_cache_sync_image() after the relocation copy to ensure the instruction cache is synchronized with the updated memory contents before control is transferred to the relocated kernel. Signed-off-by: WANG Rui Reviewed-by: Huacai Chen Signed-off-by: Ard Biesheuvel --- drivers/firmware/efi/libstub/loongarch-stub.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/firmware/efi/libstub/loongarch-stub.c b/drivers/firmware/efi/libstub/loongarch-stub.c index 8c538a5243d9..c87ac7025107 100644 --- a/drivers/firmware/efi/libstub/loongarch-stub.c +++ b/drivers/firmware/efi/libstub/loongarch-stub.c @@ -86,6 +86,7 @@ efi_status_t efi_relocate_kernel(unsigned long *image_addr, * have been allocated by UEFI, so we can safely use memcpy. */ memcpy((void *)new_addr, (void *)cur_image_addr, image_size); + efi_cache_sync_image(new_addr, image_size); /* Return the new address of the relocated image. */ *image_addr = new_addr; From 28465227c80fe417b4013c432be1f3737cb9f9a3 Mon Sep 17 00:00:00 2001 From: Ruijie Li Date: Wed, 29 Apr 2026 00:41:43 +0800 Subject: [PATCH 089/525] xfrm: provide message size for XFRM_MSG_MAPPING The compat 64=>32 translation path handles XFRM_MSG_MAPPING, but xfrm_msg_min[] does not provide the native payload size for this message type. Add the missing XFRM_MSG_MAPPING entry so compat translation can size and translate mapping notifications correctly. Fixes: 5461fc0c8d9f ("xfrm/compat: Add 64=>32-bit messages translator") Cc: stable@kernel.org Reported-by: Yuan Tan Reported-by: Yifan Wu Reported-by: Juefei Pu Reported-by: Xin Liu Signed-off-by: Ruijie Li Signed-off-by: Ren Wei Signed-off-by: Steffen Klassert --- net/xfrm/xfrm_user.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index d56450f61669..38a90e5ee3d9 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -3323,6 +3323,7 @@ const int xfrm_msg_min[XFRM_NR_MSGTYPES] = { [XFRM_MSG_GETSADINFO - XFRM_MSG_BASE] = sizeof(u32), [XFRM_MSG_NEWSPDINFO - XFRM_MSG_BASE] = sizeof(u32), [XFRM_MSG_GETSPDINFO - XFRM_MSG_BASE] = sizeof(u32), + [XFRM_MSG_MAPPING - XFRM_MSG_BASE] = XMSGSIZE(xfrm_user_mapping), [XFRM_MSG_SETDEFAULT - XFRM_MSG_BASE] = XMSGSIZE(xfrm_userpolicy_default), [XFRM_MSG_GETDEFAULT - XFRM_MSG_BASE] = XMSGSIZE(xfrm_userpolicy_default), }; From 14acf9652e5690de3c7486c6db5fb8dafd0a32a3 Mon Sep 17 00:00:00 2001 From: Michal Kosiorek Date: Wed, 29 Apr 2026 10:54:51 +0200 Subject: [PATCH 090/525] xfrm: defensively unhash xfrm_state lists in __xfrm_state_delete KASAN reproduces a slab-use-after-free in __xfrm_state_delete()'s hlist_del_rcu calls under syzkaller load on linux-6.12.y stable (reproduced on 6.12.47, also reachable via the same code path on torvalds/master and on the ipsec tree). Nine unique signatures cluster in the xfrm_state lifecycle, the load-bearing one being: BUG: KASAN: slab-use-after-free in __hlist_del include/linux/list.h:990 [inline] BUG: KASAN: slab-use-after-free in hlist_del_rcu include/linux/rculist.h:516 [inline] BUG: KASAN: slab-use-after-free in __xfrm_state_delete net/xfrm/xfrm_state.c Write of size 8 at addr ffff8881198bcb70 by task kworker/u8:9/435 Workqueue: netns cleanup_net Call Trace: __hlist_del / hlist_del_rcu __xfrm_state_delete xfrm_state_delete xfrm_state_flush xfrm_state_fini ops_exit_list cleanup_net The other observed signatures hit the same slab object from __xfrm_state_lookup, xfrm_alloc_spi, __xfrm_state_insert and an OOB write variant of __xfrm_state_delete, all on the byseq/byspi hash chains. __xfrm_state_delete() guards its byseq and byspi unhashes with value-based predicates: if (x->km.seq) hlist_del_rcu(&x->byseq); if (x->id.spi) hlist_del_rcu(&x->byspi); while everywhere else in the file (e.g. state_cache, state_cache_input) the safer hlist_unhashed() check is used. xfrm_alloc_spi() sets x->id.spi = newspi inside xfrm_state_lock and then immediately inserts into byspi, but a path that observes x->id.spi != 0 outside of xfrm_state_lock can still skip-or-hit the byspi unhash inconsistently with whether x is actually on the list. The same holds for x->km.seq versus byseq, and the bydst/bysrc unhashes have no predicate at all, so a second __xfrm_state_delete() on the same object writes through LIST_POISON pprev. The defensive change here: - Use hlist_del_init_rcu() instead of hlist_del_rcu() on bydst, bysrc, byseq and byspi so a second deletion is a no-op rather than a write through LIST_POISON pprev. The byseq/byspi nodes are already initialised in xfrm_state_alloc(). - Test hlist_unhashed() rather than the value predicate for byseq/byspi, so the unhash decision tracks list state rather than mutable scalar fields. Empirical verification: applied this patch on top of v6.12.47, rebuilt, and re-ran the same syzkaller harness for 1h16m on a previously-crashy configuration that produced ~100 hits each of slab-use-after-free Read in xfrm_alloc_spi / Read in __xfrm_state_lookup / Write in __xfrm_state_delete. After the patch, 7.1M execs across 32 VMs at ~1550 exec/sec produced zero xfrm_state UAF/OOB hits. /proc/slabinfo confirms the xfrm_state slab is actively allocated and freed during the run (~143 KiB resident), so the fuzzer is still exercising those code paths -- they just no longer crash. Reproduction: - Linux 6.12.47 x86_64 + KASAN_GENERIC + KASAN_INLINE + KCOV - syzkaller @ 746545b8b1e4c3a128db8652b340d3df90ce61db - 32 QEMU/KVM VMs x 2 vCPU on AWS c5.metal bare metal - 9 unique signatures collected in ~9h, all within xfrm_state lifecycle Fixes: fe9f1d8779cb ("xfrm: add state hashtable keyed by seq") Fixes: 7b4dc3600e48 ("[XFRM]: Do not add a state whose SPI is zero to the SPI hash.") Reported-by: Michal Kosiorek Tested-by: Michal Kosiorek Cc: stable@vger.kernel.org Signed-off-by: Michal Kosiorek Signed-off-by: Steffen Klassert --- net/xfrm/xfrm_state.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c index 1748d374abca..686014d39429 100644 --- a/net/xfrm/xfrm_state.c +++ b/net/xfrm/xfrm_state.c @@ -818,17 +818,17 @@ int __xfrm_state_delete(struct xfrm_state *x) spin_lock(&net->xfrm.xfrm_state_lock); list_del(&x->km.all); - hlist_del_rcu(&x->bydst); - hlist_del_rcu(&x->bysrc); - if (x->km.seq) - hlist_del_rcu(&x->byseq); + hlist_del_init_rcu(&x->bydst); + hlist_del_init_rcu(&x->bysrc); + if (!hlist_unhashed(&x->byseq)) + hlist_del_init_rcu(&x->byseq); if (!hlist_unhashed(&x->state_cache)) hlist_del_rcu(&x->state_cache); if (!hlist_unhashed(&x->state_cache_input)) hlist_del_rcu(&x->state_cache_input); - if (x->id.spi) - hlist_del_rcu(&x->byspi); + if (!hlist_unhashed(&x->byspi)) + hlist_del_init_rcu(&x->byspi); net->xfrm.state_num--; xfrm_nat_keepalive_state_updated(x); spin_unlock(&net->xfrm.xfrm_state_lock); From 38694f4639c45599161860e828dc4ac77abf8cea Mon Sep 17 00:00:00 2001 From: Edward Srouji Date: Mon, 27 Apr 2026 14:02:32 +0300 Subject: [PATCH 091/525] RDMA/mlx5: Fix UAF in SRQ destroy due to race with create A race condition exists between mlx5_cmd_destroy_srq() and mlx5_cmd_create_srq() that can lead to a use-after-free (UAF) [1]. After destroy_srq_split() releases the SRQ to firmware, the SRQN can be immediately reallocated for a new SRQ being created concurrently. If the create path stores the new SRQ in the xarray before the destroy path erases it, the destroy will incorrectly delete the new SRQ's entry. Later accesses then hit freed memory. Fix by replacing the unconditional xa_erase_irq() with xa_cmpxchg_irq() that only erases the entry if it hasn't already been replaced (still contains XA_ZERO_ENTRY), preserving any newly created SRQ. [1] RIP: 0010:mlx5_cmd_destroy_srq+0xd8/0x110 [mlx5_ib] Code: 89 e1 ba 06 04 00 00 4c 89 f6 48 89 ef e8 80 19 70 e1 c6 83 a0 0f 00 00 00 fb 5b 44 89 e8 5d 41 5c 41 5d 41 5e c3 cc cc cc cc <0f> 0b 48 89 c2 83 e2 03 48 83 fa 02 75 08 48 3d 05 c0 ff ff 77 08 RSP: 0018:ff110001037b7d08 EFLAGS: 00010286 RAX: 0000000000000000 RBX: ff1100010bb9c000 RCX: 0000000000000000 RDX: 0000000000000000 RSI: 0000000000000000 RDI: ff110001037b7c90 RBP: ff1100010bb9cfa0 R08: 0000000000000000 R09: 0000000000000000 R10: ff110001037b7da0 R11: ff11000104f29580 R12: ff1100010e2ac090 R13: 000000000000000d R14: 0000000000000001 R15: ff11000105336300 FS: 00007fa24787c740(0000) GS:ff1100046eb8d000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fa247984e90 CR3: 0000000109d59005 CR4: 0000000000373eb0 Call Trace: mlx5_ib_destroy_srq+0x25/0xa0 [mlx5_ib] ib_destroy_srq_user+0x21/0x90 [ib_core] uverbs_free_srq+0x1b/0x50 [ib_uverbs] destroy_hw_idr_uobject+0x1e/0x50 [ib_uverbs] uverbs_destroy_uobject+0x35/0x180 [ib_uverbs] __uverbs_cleanup_ufile+0xdd/0x140 [ib_uverbs] uverbs_destroy_ufile_hw+0x38/0xf0 [ib_uverbs] ib_uverbs_close+0x17/0xa0 [ib_uverbs] __fput+0xe0/0x2a0 __x64_sys_close+0x3a/0x80 do_syscall_64+0x55/0xac0 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7fa247984ea4 Code: 00 f7 d8 64 89 01 48 83 c8 ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa 80 3d a5 51 0e 00 00 74 13 b8 03 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 3c c3 0f 1f 00 55 48 89 e5 48 83 ec 10 89 7d RSP: 002b:00007ffecfa79498 EFLAGS: 00000202 ORIG_RAX: 0000000000000003 RAX: ffffffffffffffda RBX: 0000200000000080 RCX: 00007fa247984ea4 RDX: 0000000000000040 RSI: 0000200000000200 RDI: 0000000000000003 RBP: 00007ffecfa794e0 R08: 00007ffecfa794e0 R09: 00007ffecfa794e0 R10: 0000000000000000 R11: 0000000000000202 R12: 0000000000000001 R13: 0000000000000000 R14: 0000200000000000 R15: 0000200000000009 ---[ end trace 0000000000000000 ]--- Fixes: fd89099d635e ("RDMA/mlx5: Issue FW command to destroy SRQ on reentry") Link: https://patch.msgid.link/r/20260427-security-bug-fixes-v3-1-4621fa52de0e@nvidia.com Signed-off-by: Edward Srouji Reviewed-by: Michael Guralnik Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mlx5/srq_cmd.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mlx5/srq_cmd.c b/drivers/infiniband/hw/mlx5/srq_cmd.c index 8b3385396599..c1a088120915 100644 --- a/drivers/infiniband/hw/mlx5/srq_cmd.c +++ b/drivers/infiniband/hw/mlx5/srq_cmd.c @@ -683,7 +683,14 @@ int mlx5_cmd_destroy_srq(struct mlx5_ib_dev *dev, struct mlx5_core_srq *srq) xa_cmpxchg_irq(&table->array, srq->srqn, XA_ZERO_ENTRY, srq, 0); return err; } - xa_erase_irq(&table->array, srq->srqn); + + /* + * A race can occur where a concurrent create gets the same srqn + * (after hardware released it) and overwrites XA_ZERO_ENTRY with + * its new SRQ before we reach here. In that case, we must not erase + * the entry as it now belongs to the new SRQ. + */ + xa_cmpxchg_irq(&table->array, srq->srqn, XA_ZERO_ENTRY, NULL, 0); mlx5_core_res_put(&srq->common); wait_for_completion(&srq->common.free); From 9bee81cc5e8811c8bbe67fbf5214a7998457324b Mon Sep 17 00:00:00 2001 From: Edward Srouji Date: Mon, 27 Apr 2026 14:02:33 +0300 Subject: [PATCH 092/525] RDMA/mlx5: Fix UAF in DCT destroy due to race with create A potential race condition exists between mlx5_core_destroy_dct() and mlx5_core_create_dct() that can lead to a use-after-free. After _mlx5_core_destroy_dct() releases the DCT to firmware, the DCTN can be immediately reallocated for a new DCT being created concurrently. If the create path stores the new DCT in the xarray before the destroy path erases it, the destroy will incorrectly delete the new DCT's entry. Later accesses then hit freed memory. Fix by replacing the unconditional xa_erase_irq() with xa_cmpxchg_irq() that only erases the entry if it hasn't already been replaced (still contains XA_ZERO_ENTRY), preserving any newly created DCT. Fixes: afff24899846 ("RDMA/mlx5: Handle DCT QP logic separately from low level QP interface") Link: https://patch.msgid.link/r/20260427-security-bug-fixes-v3-2-4621fa52de0e@nvidia.com Signed-off-by: Edward Srouji Reviewed-by: Michael Guralnik Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mlx5/qpc.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mlx5/qpc.c b/drivers/infiniband/hw/mlx5/qpc.c index 146d03ae40bd..a7a4f9420271 100644 --- a/drivers/infiniband/hw/mlx5/qpc.c +++ b/drivers/infiniband/hw/mlx5/qpc.c @@ -314,7 +314,14 @@ int mlx5_core_destroy_dct(struct mlx5_ib_dev *dev, xa_cmpxchg_irq(&table->dct_xa, dct->mqp.qpn, XA_ZERO_ENTRY, dct, 0); return err; } - xa_erase_irq(&table->dct_xa, dct->mqp.qpn); + + /* + * A race can occur where a concurrent create gets the same dctn + * (after hardware released it) and overwrites XA_ZERO_ENTRY with + * its new DCT before we reach here. In that case, we must not erase + * the entry as it now belongs to the new DCT. + */ + xa_cmpxchg_irq(&table->dct_xa, dct->mqp.qpn, XA_ZERO_ENTRY, NULL, 0); return 0; } From 610771c62e2ac5bca851fc5a6f8af1cdd83f189a Mon Sep 17 00:00:00 2001 From: Maher Sanalla Date: Mon, 27 Apr 2026 14:02:34 +0300 Subject: [PATCH 093/525] IB/core: Fix IPv6 netlink message size in ib_nl_ip_send_msg() When resolving an RDMA-CM IPv6 address, ib_nl_ip_send_msg() sends a netlink request to the userspace daemon to perform IP-to-GID resolution in certain cases. The function allocates the netlink message buffer using nla_total_size(sizeof(size)), which passes 8 bytes (the size of size_t) instead of 16 bytes (the size of an IPv6 address). This results in an 8-byte under-allocation. This is currently masked by nlmsg_new() over-allocation of the skb in its internal logic. However, the code remains incorrect. Fix the issue by supplying the proper IPv6 address length to nla_total_size(). Fixes: ae43f8286730 ("IB/core: Add IP to GID netlink offload") Link: https://patch.msgid.link/r/20260427-security-bug-fixes-v3-3-4621fa52de0e@nvidia.com Signed-off-by: Maher Sanalla Reviewed-by: Patrisious Haddad Signed-off-by: Edward Srouji Signed-off-by: Jason Gunthorpe --- drivers/infiniband/core/addr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c index a40a765f0307..27992c38ad90 100644 --- a/drivers/infiniband/core/addr.c +++ b/drivers/infiniband/core/addr.c @@ -149,7 +149,7 @@ static int ib_nl_ip_send_msg(struct rdma_dev_addr *dev_addr, attrtype = RDMA_NLA_F_MANDATORY | LS_NLA_TYPE_IPV6; } - len = nla_total_size(sizeof(size)); + len = nla_total_size(size); len += NLMSG_ALIGN(sizeof(*header)); skb = nlmsg_new(len, GFP_KERNEL); From 1f3b337af2231b1e83c9052f771b201f5cbb9997 Mon Sep 17 00:00:00 2001 From: Michael Guralnik Date: Mon, 27 Apr 2026 14:02:35 +0300 Subject: [PATCH 094/525] RDMA/core: Fix rereg_mr use-after-free race MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When a driver creates a new MR during rereg_user_mr, a race window exists between rdma_alloc_commit_uobject() for the new MR and the point where the code reads that MR to populate the response keys. A concurrent rereg_mr or destroy_mr could destroy the MR in this window and cause UAF in the first thread. Racing flow between two rereg_mr calls: CPU0 CPU1 ---- ---- rereg_user_mr(mr_handle) uobj_get_write(mr_handle) -> mr0 mr1 = driver→rereg() rdma_alloc_commit_uobject(mr1) // mr1 replaced mr0 and is unlocked uobj_put_destroy(mr0) rereg_user_mr(mr_handle) uobj_get_write(mr_handle) -> mr1 mr2 = driver→rereg() rdma_alloc_commit_uobject(mr2) // mr2 replaced mr1 and is unlocked uobj_put_destroy(mr1) // Destroys mr1! resp.lkey = mr1->lkey; // UAF - mr1 was freed! resp.rkey = mr1->rkey; // UAF - mr1 was freed! Fix by storing lkey/rkey in local variables before the new MR is unlocked and using the local variables to set the user response. Fixes: 6e0954b11c05 ("RDMA/uverbs: Allow drivers to create a new HW object during rereg_mr") Link: https://patch.msgid.link/r/20260427-security-bug-fixes-v3-4-4621fa52de0e@nvidia.com Signed-off-by: Michael Guralnik Reviewed-by: Maher Sanalla Signed-off-by: Edward Srouji Signed-off-by: Jason Gunthorpe --- drivers/infiniband/core/uverbs_cmd.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index a768436ba468..91a62d2ade4d 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -778,6 +778,7 @@ static int ib_uverbs_rereg_mr(struct uverbs_attr_bundle *attrs) struct ib_pd *orig_pd; struct ib_pd *new_pd; struct ib_mr *new_mr; + u32 lkey, rkey; ret = uverbs_request(attrs, &cmd, sizeof(cmd)); if (ret) @@ -846,6 +847,8 @@ static int ib_uverbs_rereg_mr(struct uverbs_attr_bundle *attrs) new_mr->uobject = uobj; atomic_inc(&new_pd->usecnt); new_uobj->object = new_mr; + lkey = new_mr->lkey; + rkey = new_mr->rkey; rdma_restrack_new(&new_mr->res, RDMA_RESTRACK_MR); rdma_restrack_set_name(&new_mr->res, NULL); @@ -871,11 +874,13 @@ static int ib_uverbs_rereg_mr(struct uverbs_attr_bundle *attrs) mr->iova = cmd.hca_va; mr->length = cmd.length; } + lkey = mr->lkey; + rkey = mr->rkey; } memset(&resp, 0, sizeof(resp)); - resp.lkey = mr->lkey; - resp.rkey = mr->rkey; + resp.lkey = lkey; + resp.rkey = rkey; ret = uverbs_response(attrs, &resp, sizeof(resp)); From 6009cca96fcb01182cede725ad61e9e3810f3932 Mon Sep 17 00:00:00 2001 From: Michael Guralnik Date: Mon, 27 Apr 2026 14:02:36 +0300 Subject: [PATCH 095/525] RDMA/mlx5: Fix null-ptr-deref in Raw Packet QP creation Raw Packet QPs are unique in that they support separate send and receive queues, using 2 different user-provided buffers. They can also be created with one of the queues having size 0, allowing a send-only or receive-only QP. The Raw Packet RQ umem is created in the common user QP creation path, which allows zero-length queues. Add a later validation of the RQ umem in Raw Packet QP creation path when an RQ was requested. This prevents possible null-ptr dereference crashes, as seen in the below trace: Oops: general protection fault, probably for non-canonical address 0xdffffc0000000006: 0000 [#1] SMP KASAN KASAN: null-ptr-deref in range [0x0000000000000030-0x0000000000000037] CPU: 6 UID: 0 PID: 3539 Comm: raw_packet_umem Not tainted 6.19.0-rc1+ #166 NONE Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014 RIP: 0010:__mlx5_umem_find_best_quantized_pgoff+0x37/0x280 [mlx5_ib] Code: ff df 41 57 49 89 ff 41 56 41 55 41 89 d5 41 54 4d 89 cc 4c 8d 4f 30 55 4c 89 ca 48 89 f5 53 48 c1 ea 03 48 89 cb 48 83 ec 18 <80> 3c 02 00 44 89 04 24 0f 85 01 02 00 00 48 ba 00 00 00 00 00 fc RSP: 0018:ff1100013966f4e0 EFLAGS: 00010282 RAX: dffffc0000000000 RBX: 00000000ffffffc0 RCX: 00000000ffffffc0 RDX: 0000000000000006 RSI: 00000ffffffff000 RDI: 0000000000000000 RBP: 00000ffffffff000 R08: 0000000000000040 R09: 0000000000000030 R10: 0000000000000000 R11: 0000000000000000 R12: ff1100013966f648 R13: 0000000000000005 R14: ff1100013966f980 R15: 0000000000000000 FS: 00007fae6c82f740(0000) GS:ff11000898ba1000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000200000000000 CR3: 000000010f96c005 CR4: 0000000000373eb0 Call Trace: create_qp+0x747d/0xc740 [mlx5_ib] ? is_module_address+0x18/0x110 ? _create_user_qp.constprop.0+0x18e0/0x18e0 [mlx5_ib] ? __module_address+0x49/0x210 ? is_module_address+0x68/0x110 ? static_obj+0x67/0x90 ? lockdep_init_map_type+0x58/0x200 mlx5_ib_create_qp+0xc85/0x2620 [mlx5_ib] ? find_held_lock+0x2b/0x80 ? create_qp+0xc740/0xc740 [mlx5_ib] ? lock_release+0xcb/0x260 ? lockdep_init_map_type+0x58/0x200 ? __init_swait_queue_head+0xcb/0x150 create_qp.part.0+0x558/0x7c0 [ib_core] ib_create_qp_user+0xa0/0x4f0 [ib_core] ? rdma_lookup_get_uobject+0x1e4/0x400 [ib_uverbs] create_qp+0xe4f/0x1d10 [ib_uverbs] ? ib_uverbs_rereg_mr+0xd40/0xd40 [ib_uverbs] ? ib_uverbs_cq_event_handler+0x120/0x120 [ib_uverbs] ? __might_fault+0x81/0x100 ? lock_release+0xcb/0x260 ? _copy_from_user+0x3e/0x90 ib_uverbs_create_qp+0x10a/0x150 [ib_uverbs] ? ib_uverbs_ex_create_qp+0xe0/0xe0 [ib_uverbs] ? __might_fault+0x81/0x100 ? lock_release+0xcb/0x260 ib_uverbs_write+0x7e5/0xc90 [ib_uverbs] ? uverbs_devnode+0xc0/0xc0 [ib_uverbs] ? lock_acquire+0xfa/0x2b0 ? find_held_lock+0x2b/0x80 ? finish_task_switch.isra.0+0x189/0x6c0 vfs_write+0x1c0/0xf70 ? lockdep_hardirqs_on_prepare+0xde/0x170 ? kernel_write+0x5a0/0x5a0 ? __switch_to+0x527/0xe60 ? __schedule+0x10a3/0x3950 ? io_schedule_timeout+0x110/0x110 ksys_write+0x170/0x1c0 ? __x64_sys_read+0xb0/0xb0 ? trace_hardirqs_off.part.0+0x4e/0xe0 do_syscall_64+0x70/0x1360 entry_SYSCALL_64_after_hwframe+0x4b/0x53 RIP: 0033:0x7fae6ca3118d Code: ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 5b cc 0c 00 f7 d8 64 89 01 48 RSP: 002b:00007ffe678ca308 EFLAGS: 00000213 ORIG_RAX: 0000000000000001 RAX: ffffffffffffffda RBX: 00007ffe678ca448 RCX: 00007fae6ca3118d RDX: 0000000000000070 RSI: 0000200000000280 RDI: 0000000000000003 RBP: 00007ffe678ca320 R08: 00000000ffffffff R09: 00007fae6c8ec5b8 R10: 0000000000000064 R11: 0000000000000213 R12: 0000000000000001 R13: 0000000000000000 R14: 00007fae6cb71000 R15: 0000000000404df0 Modules linked in: mlx5_ib mlx5_fwctl mlx5_core bonding ip6_gre ip6_tunnel tunnel6 ip_gre gre rdma_ucm ib_uverbs rdma_cm iw_cm ib_ipoib ib_cm ib_umad ib_core rpcsec_gss_krb5 auth_rpcgss oid_registry overlay nfnetlink zram zsmalloc fuse scsi_transport_iscsi [last unloaded: mlx5_core] ---[ end trace 0000000000000000 ]--- RIP: 0010:__mlx5_umem_find_best_quantized_pgoff+0x37/0x280 [mlx5_ib] Fixes: 0fb2ed66a14c ("IB/mlx5: Add create and destroy functionality for Raw Packet QP") Link: https://patch.msgid.link/r/20260427-security-bug-fixes-v3-5-4621fa52de0e@nvidia.com Signed-off-by: Michael Guralnik Reviewed-by: Maher Sanalla Signed-off-by: Edward Srouji Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mlx5/qp.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index 8f50e7342a76..1611a704c1b3 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -1603,6 +1603,11 @@ static int create_raw_packet_qp(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, } if (qp->rq.wqe_cnt) { + if (!rq->base.ubuffer.umem) { + err = -EINVAL; + goto err_destroy_sq; + } + rq->base.container_mibqp = qp; if (qp->flags & IB_QP_CREATE_CVLAN_STRIPPING) From 20e81c64c905bd765e69ef07920d2b1130dc79b6 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 29 Apr 2026 09:44:16 -1000 Subject: [PATCH 096/525] workqueue: Annotate alloc_workqueue_va() with __printf(1, 0) alloc_workqueue_va() forwards its va_list to __alloc_workqueue() which ultimately feeds vsnprintf(). __alloc_workqueue() already carries __printf(1, 0); the new wrapper needs the same annotation so format string checking propagates through the forwarding. Fixes: 0de4cb473aed ("workqueue: fix devm_alloc_workqueue() va_list misuse") Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202604300347.2LgXyteh-lkp@intel.com/ Signed-off-by: Tejun Heo --- kernel/workqueue.c | 1 + 1 file changed, 1 insertion(+) diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 24d0265191d4..3d2e3b2ec528 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -5906,6 +5906,7 @@ static struct workqueue_struct *__alloc_workqueue(const char *fmt, return NULL; } +__printf(1, 0) static struct workqueue_struct *alloc_workqueue_va(const char *fmt, unsigned int flags, int max_active, From 1049970d7583194eedc30e45a3c898b2cb1c30ba Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Mon, 27 Apr 2026 14:34:45 +0200 Subject: [PATCH 097/525] netfilter: replace skb_try_make_writable() by skb_ensure_writable() skb_try_make_writable() only works on clones and uncloned packets might have their network header in paged fragments. nft_fwd needs to work for the ingress and egress hooks, but the egress hook where skb->data points to the mac header, use skb_network_offset() to include the mac header. The flowtable is fine since it already uses the transport offset. Fixes: d32de98ea70f ("netfilter: nft_fwd_netdev: allow to forward packets via neighbour layer") Fixes: 7d2086871762 ("netfilter: nf_flow_table: move ipv4 offload hook code to nf_flow_table") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_flow_table_ip.c | 4 ++-- net/netfilter/nft_fwd_netdev.c | 5 +++-- 2 files changed, 5 insertions(+), 4 deletions(-) diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c index fd56d663cb5b..dbd7644fdbeb 100644 --- a/net/netfilter/nf_flow_table_ip.c +++ b/net/netfilter/nf_flow_table_ip.c @@ -524,7 +524,7 @@ static int nf_flow_offload_forward(struct nf_flowtable_ctx *ctx, return 0; } - if (skb_try_make_writable(skb, thoff + ctx->hdrsize)) + if (skb_ensure_writable(skb, thoff + ctx->hdrsize)) return -1; flow_offload_refresh(flow_table, flow, false); @@ -1037,7 +1037,7 @@ static int nf_flow_offload_ipv6_forward(struct nf_flowtable_ctx *ctx, return 0; } - if (skb_try_make_writable(skb, thoff + ctx->hdrsize)) + if (skb_ensure_writable(skb, thoff + ctx->hdrsize)) return -1; flow_offload_refresh(flow_table, flow, false); diff --git a/net/netfilter/nft_fwd_netdev.c b/net/netfilter/nft_fwd_netdev.c index 4bce36c3a6a0..2cc809303ce8 100644 --- a/net/netfilter/nft_fwd_netdev.c +++ b/net/netfilter/nft_fwd_netdev.c @@ -100,6 +100,7 @@ static void nft_fwd_neigh_eval(const struct nft_expr *expr, int oif = regs->data[priv->sreg_dev]; unsigned int verdict = NF_STOLEN; struct sk_buff *skb = pkt->skb; + int nhoff = skb_network_offset(skb); struct net_device *dev; int neigh_table; @@ -111,7 +112,7 @@ static void nft_fwd_neigh_eval(const struct nft_expr *expr, verdict = NFT_BREAK; goto out; } - if (skb_try_make_writable(skb, sizeof(*iph))) { + if (skb_ensure_writable(skb, nhoff + sizeof(*iph))) { verdict = NF_DROP; goto out; } @@ -132,7 +133,7 @@ static void nft_fwd_neigh_eval(const struct nft_expr *expr, verdict = NFT_BREAK; goto out; } - if (skb_try_make_writable(skb, sizeof(*ip6h))) { + if (skb_ensure_writable(skb, nhoff + sizeof(*ip6h))) { verdict = NF_DROP; goto out; } From 0a0b35f0bf10b4c2be607465f5c9c12c8681305b Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Mon, 27 Apr 2026 14:34:48 +0200 Subject: [PATCH 098/525] netfilter: nft_fwd_netdev: add device and headroom validate with neigh forwarding The ttl field has been decremented already and evaluation of this rule would proceed, just drop this packet instead if there is no destination device to forwards this packet. This is exactly what nf_dup already does in this case. Moreover, check for headroom and call skb_expand_head() like in the IP output path to ensure there is sufficient headroom when forwarding this via neigh_xmit(). Fixes: d32de98ea70f ("netfilter: nft_fwd_netdev: allow to forward packets via neighbour layer") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nft_fwd_netdev.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nft_fwd_netdev.c b/net/netfilter/nft_fwd_netdev.c index 2cc809303ce8..605b1d42abce 100644 --- a/net/netfilter/nft_fwd_netdev.c +++ b/net/netfilter/nft_fwd_netdev.c @@ -102,6 +102,7 @@ static void nft_fwd_neigh_eval(const struct nft_expr *expr, struct sk_buff *skb = pkt->skb; int nhoff = skb_network_offset(skb); struct net_device *dev; + unsigned int hh_len; int neigh_table; switch (priv->nfproto) { @@ -153,8 +154,19 @@ static void nft_fwd_neigh_eval(const struct nft_expr *expr, } dev = dev_get_by_index_rcu(nft_net(pkt), oif); - if (dev == NULL) - return; + if (dev == NULL) { + verdict = NF_DROP; + goto out; + } + + hh_len = LL_RESERVED_SPACE(dev); + if (unlikely(skb_headroom(skb) < hh_len && dev->header_ops)) { + skb = skb_expand_head(skb, hh_len); + if (!skb) { + verdict = NF_STOLEN; + goto out; + } + } skb->dev = dev; skb_clear_tstamp(skb); From 1d47b55b36d2ec73fe6901212c8b28a593c3b27c Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Mon, 27 Apr 2026 14:34:50 +0200 Subject: [PATCH 099/525] netfilter: nft_fwd_netdev: use recursion counter in neigh egress path nft_fwd_neigh can be used in egress chains (NF_NETDEV_EGRESS). When the forwarding rule targets the same device or two devices forward to each other, neigh_xmit() triggers dev_queue_xmit() which re-enters nf_hook_egress(), causing infinite recursion and stack overflow. Move the nf_get_nf_dup_skb_recursion() accessor and NF_RECURSION_LIMIT to the shared header nf_dup_netdev.h as a static inline, so that nft_fwd_netdev can use the recursion counter directly without exported function call overhead. Guard neigh_xmit() with the same recursion limit already used in nf_do_netdev_egress(). [ Updated to cache the nf_get_nf_dup_skb_recursion pointer. --pablo ] Fixes: f87b9464d152 ("netfilter: nft_fwd_netdev: Support egress hook") Reported-by: Xiang Mei Signed-off-by: Weiming Shi Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_dup_netdev.h | 13 +++++++++++++ net/netfilter/nf_dup_netdev.c | 16 ---------------- net/netfilter/nft_fwd_netdev.c | 8 ++++++++ 3 files changed, 21 insertions(+), 16 deletions(-) diff --git a/include/net/netfilter/nf_dup_netdev.h b/include/net/netfilter/nf_dup_netdev.h index b175d271aec9..609bcf422a9b 100644 --- a/include/net/netfilter/nf_dup_netdev.h +++ b/include/net/netfilter/nf_dup_netdev.h @@ -3,10 +3,23 @@ #define _NF_DUP_NETDEV_H_ #include +#include +#include void nf_dup_netdev_egress(const struct nft_pktinfo *pkt, int oif); void nf_fwd_netdev_egress(const struct nft_pktinfo *pkt, int oif); +#define NF_RECURSION_LIMIT 2 + +static inline u8 *nf_get_nf_dup_skb_recursion(void) +{ +#ifndef CONFIG_PREEMPT_RT + return this_cpu_ptr(&softnet_data.xmit.nf_dup_skb_recursion); +#else + return ¤t->net_xmit.nf_dup_skb_recursion; +#endif +} + struct nft_offload_ctx; struct nft_flow_rule; diff --git a/net/netfilter/nf_dup_netdev.c b/net/netfilter/nf_dup_netdev.c index e348fb90b8dc..3b0a70e154cd 100644 --- a/net/netfilter/nf_dup_netdev.c +++ b/net/netfilter/nf_dup_netdev.c @@ -13,22 +13,6 @@ #include #include -#define NF_RECURSION_LIMIT 2 - -#ifndef CONFIG_PREEMPT_RT -static u8 *nf_get_nf_dup_skb_recursion(void) -{ - return this_cpu_ptr(&softnet_data.xmit.nf_dup_skb_recursion); -} -#else - -static u8 *nf_get_nf_dup_skb_recursion(void) -{ - return ¤t->net_xmit.nf_dup_skb_recursion; -} - -#endif - static void nf_do_netdev_egress(struct sk_buff *skb, struct net_device *dev, enum nf_dev_hooks hook) { diff --git a/net/netfilter/nft_fwd_netdev.c b/net/netfilter/nft_fwd_netdev.c index 605b1d42abce..b9e88d7cf308 100644 --- a/net/netfilter/nft_fwd_netdev.c +++ b/net/netfilter/nft_fwd_netdev.c @@ -95,6 +95,7 @@ static void nft_fwd_neigh_eval(const struct nft_expr *expr, struct nft_regs *regs, const struct nft_pktinfo *pkt) { + u8 *nf_dup_skb_recursion = nf_get_nf_dup_skb_recursion(); struct nft_fwd_neigh *priv = nft_expr_priv(expr); void *addr = ®s->data[priv->sreg_addr]; int oif = regs->data[priv->sreg_dev]; @@ -153,6 +154,11 @@ static void nft_fwd_neigh_eval(const struct nft_expr *expr, goto out; } + if (*nf_dup_skb_recursion > NF_RECURSION_LIMIT) { + verdict = NF_DROP; + goto out; + } + dev = dev_get_by_index_rcu(nft_net(pkt), oif); if (dev == NULL) { verdict = NF_DROP; @@ -170,7 +176,9 @@ static void nft_fwd_neigh_eval(const struct nft_expr *expr, skb->dev = dev; skb_clear_tstamp(skb); + (*nf_dup_skb_recursion)++; neigh_xmit(neigh_table, dev, addr, skb); + (*nf_dup_skb_recursion)--; out: regs->verdict.code = verdict; } From 6813985ca456d1f5677ad9554f55805cbf27e16f Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Tue, 28 Apr 2026 17:35:18 +0200 Subject: [PATCH 100/525] netfilter: x_tables: add .check_hooks to matches and targets Add a new .check_hooks interface for checking if the match/target is used from the validate hook according to its configuration. Move existing conditional hook check based on the match/target configuration from .checkentry to .check_hooks for the following matches/targets: - addrtype - devgroup - physdev - policy - set - TCPMSS - SET This is a preparation patch to fix nft_compat, not functional changes are intended. Based on patch from Florian Westphal. Signed-off-by: Pablo Neira Ayuso --- include/linux/netfilter/x_tables.h | 8 +++ net/netfilter/x_tables.c | 79 +++++++++++++++++++++++++++--- net/netfilter/xt_TCPMSS.c | 33 +++++++------ net/netfilter/xt_addrtype.c | 25 +++++++--- net/netfilter/xt_devgroup.c | 18 +++++-- net/netfilter/xt_physdev.c | 24 ++++++--- net/netfilter/xt_policy.c | 24 +++++++-- net/netfilter/xt_set.c | 39 +++++++++------ 8 files changed, 189 insertions(+), 61 deletions(-) diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h index 77c778d84d4c..a81b46af5118 100644 --- a/include/linux/netfilter/x_tables.h +++ b/include/linux/netfilter/x_tables.h @@ -146,6 +146,9 @@ struct xt_match { /* Called when user tries to insert an entry of this type. */ int (*checkentry)(const struct xt_mtchk_param *); + /* Called to validate hooks based on the match configuration. */ + int (*check_hooks)(const struct xt_mtchk_param *); + /* Called when entry of this type deleted. */ void (*destroy)(const struct xt_mtdtor_param *); #ifdef CONFIG_NETFILTER_XTABLES_COMPAT @@ -187,6 +190,9 @@ struct xt_target { /* Should return 0 on success or an error code otherwise (-Exxxx). */ int (*checkentry)(const struct xt_tgchk_param *); + /* Called to validate hooks based on the target configuration. */ + int (*check_hooks)(const struct xt_tgchk_param *); + /* Called when entry of this type deleted. */ void (*destroy)(const struct xt_tgdtor_param *); #ifdef CONFIG_NETFILTER_XTABLES_COMPAT @@ -279,8 +285,10 @@ bool xt_find_jump_offset(const unsigned int *offsets, int xt_check_proc_name(const char *name, unsigned int size); +int xt_check_hooks_match(struct xt_mtchk_param *par); int xt_check_match(struct xt_mtchk_param *, unsigned int size, u16 proto, bool inv_proto); +int xt_check_hooks_target(struct xt_tgchk_param *par); int xt_check_target(struct xt_tgchk_param *, unsigned int size, u16 proto, bool inv_proto); diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c index 9f837fb5ceb4..2c67c2e6b132 100644 --- a/net/netfilter/x_tables.c +++ b/net/netfilter/x_tables.c @@ -477,11 +477,9 @@ int xt_check_proc_name(const char *name, unsigned int size) } EXPORT_SYMBOL(xt_check_proc_name); -int xt_check_match(struct xt_mtchk_param *par, - unsigned int size, u16 proto, bool inv_proto) +static int xt_check_match_common(struct xt_mtchk_param *par, + unsigned int size, u16 proto, bool inv_proto) { - int ret; - if (XT_ALIGN(par->match->matchsize) != size && par->match->matchsize != -1) { /* @@ -530,6 +528,14 @@ int xt_check_match(struct xt_mtchk_param *par, par->match->proto); return -EINVAL; } + + return 0; +} + +static int xt_checkentry_match(struct xt_mtchk_param *par) +{ + int ret; + if (par->match->checkentry != NULL) { ret = par->match->checkentry(par); if (ret < 0) @@ -538,8 +544,34 @@ int xt_check_match(struct xt_mtchk_param *par, /* Flag up potential errors. */ return -EIO; } + return 0; } + +int xt_check_hooks_match(struct xt_mtchk_param *par) +{ + if (par->match->check_hooks != NULL) + return par->match->check_hooks(par); + + return 0; +} +EXPORT_SYMBOL_GPL(xt_check_hooks_match); + +int xt_check_match(struct xt_mtchk_param *par, + unsigned int size, u16 proto, bool inv_proto) +{ + int ret; + + ret = xt_check_match_common(par, size, proto, inv_proto); + if (ret < 0) + return ret; + + ret = xt_check_hooks_match(par); + if (ret < 0) + return ret; + + return xt_checkentry_match(par); +} EXPORT_SYMBOL_GPL(xt_check_match); /** xt_check_entry_match - check that matches end before start of target @@ -1012,11 +1044,9 @@ bool xt_find_jump_offset(const unsigned int *offsets, } EXPORT_SYMBOL(xt_find_jump_offset); -int xt_check_target(struct xt_tgchk_param *par, - unsigned int size, u16 proto, bool inv_proto) +static int xt_check_target_common(struct xt_tgchk_param *par, + unsigned int size, u16 proto, bool inv_proto) { - int ret; - if (XT_ALIGN(par->target->targetsize) != size) { pr_err_ratelimited("%s_tables: %s.%u target: invalid size %u (kernel) != (user) %u\n", xt_prefix[par->family], par->target->name, @@ -1061,6 +1091,23 @@ int xt_check_target(struct xt_tgchk_param *par, par->target->proto); return -EINVAL; } + + return 0; +} + +int xt_check_hooks_target(struct xt_tgchk_param *par) +{ + if (par->target->check_hooks != NULL) + return par->target->check_hooks(par); + + return 0; +} +EXPORT_SYMBOL_GPL(xt_check_hooks_target); + +static int xt_checkentry_target(struct xt_tgchk_param *par) +{ + int ret; + if (par->target->checkentry != NULL) { ret = par->target->checkentry(par); if (ret < 0) @@ -1071,6 +1118,22 @@ int xt_check_target(struct xt_tgchk_param *par, } return 0; } + +int xt_check_target(struct xt_tgchk_param *par, + unsigned int size, u16 proto, bool inv_proto) +{ + int ret; + + ret = xt_check_target_common(par, size, proto, inv_proto); + if (ret < 0) + return ret; + + ret = xt_check_hooks_target(par); + if (ret < 0) + return ret; + + return xt_checkentry_target(par); +} EXPORT_SYMBOL_GPL(xt_check_target); /** diff --git a/net/netfilter/xt_TCPMSS.c b/net/netfilter/xt_TCPMSS.c index 116a885adb3c..80e1634bc51f 100644 --- a/net/netfilter/xt_TCPMSS.c +++ b/net/netfilter/xt_TCPMSS.c @@ -247,6 +247,21 @@ tcpmss_tg6(struct sk_buff *skb, const struct xt_action_param *par) } #endif +static int tcpmss_tg4_check_hooks(const struct xt_tgchk_param *par) +{ + const struct xt_tcpmss_info *info = par->targinfo; + + if (info->mss == XT_TCPMSS_CLAMP_PMTU && + (par->hook_mask & ~((1 << NF_INET_FORWARD) | + (1 << NF_INET_LOCAL_OUT) | + (1 << NF_INET_POST_ROUTING))) != 0) { + pr_info_ratelimited("path-MTU clamping only supported in FORWARD, OUTPUT and POSTROUTING hooks\n"); + return -EINVAL; + } + + return 0; +} + /* Must specify -p tcp --syn */ static inline bool find_syn_match(const struct xt_entry_match *m) { @@ -262,17 +277,9 @@ static inline bool find_syn_match(const struct xt_entry_match *m) static int tcpmss_tg4_check(const struct xt_tgchk_param *par) { - const struct xt_tcpmss_info *info = par->targinfo; const struct ipt_entry *e = par->entryinfo; const struct xt_entry_match *ematch; - if (info->mss == XT_TCPMSS_CLAMP_PMTU && - (par->hook_mask & ~((1 << NF_INET_FORWARD) | - (1 << NF_INET_LOCAL_OUT) | - (1 << NF_INET_POST_ROUTING))) != 0) { - pr_info_ratelimited("path-MTU clamping only supported in FORWARD, OUTPUT and POSTROUTING hooks\n"); - return -EINVAL; - } if (par->nft_compat) return 0; @@ -286,17 +293,9 @@ static int tcpmss_tg4_check(const struct xt_tgchk_param *par) #if IS_ENABLED(CONFIG_IP6_NF_IPTABLES) static int tcpmss_tg6_check(const struct xt_tgchk_param *par) { - const struct xt_tcpmss_info *info = par->targinfo; const struct ip6t_entry *e = par->entryinfo; const struct xt_entry_match *ematch; - if (info->mss == XT_TCPMSS_CLAMP_PMTU && - (par->hook_mask & ~((1 << NF_INET_FORWARD) | - (1 << NF_INET_LOCAL_OUT) | - (1 << NF_INET_POST_ROUTING))) != 0) { - pr_info_ratelimited("path-MTU clamping only supported in FORWARD, OUTPUT and POSTROUTING hooks\n"); - return -EINVAL; - } if (par->nft_compat) return 0; @@ -312,6 +311,7 @@ static struct xt_target tcpmss_tg_reg[] __read_mostly = { { .family = NFPROTO_IPV4, .name = "TCPMSS", + .check_hooks = tcpmss_tg4_check_hooks, .checkentry = tcpmss_tg4_check, .target = tcpmss_tg4, .targetsize = sizeof(struct xt_tcpmss_info), @@ -322,6 +322,7 @@ static struct xt_target tcpmss_tg_reg[] __read_mostly = { { .family = NFPROTO_IPV6, .name = "TCPMSS", + .check_hooks = tcpmss_tg4_check_hooks, .checkentry = tcpmss_tg6_check, .target = tcpmss_tg6, .targetsize = sizeof(struct xt_tcpmss_info), diff --git a/net/netfilter/xt_addrtype.c b/net/netfilter/xt_addrtype.c index a77088943107..913dbe3aa5e2 100644 --- a/net/netfilter/xt_addrtype.c +++ b/net/netfilter/xt_addrtype.c @@ -153,14 +153,10 @@ addrtype_mt_v1(const struct sk_buff *skb, struct xt_action_param *par) return ret; } -static int addrtype_mt_checkentry_v1(const struct xt_mtchk_param *par) +static int addrtype_mt_check_hooks(const struct xt_mtchk_param *par) { - const char *errmsg = "both incoming and outgoing interface limitation cannot be selected"; struct xt_addrtype_info_v1 *info = par->matchinfo; - - if (info->flags & XT_ADDRTYPE_LIMIT_IFACE_IN && - info->flags & XT_ADDRTYPE_LIMIT_IFACE_OUT) - goto err; + const char *errmsg; if (par->hook_mask & ((1 << NF_INET_PRE_ROUTING) | (1 << NF_INET_LOCAL_IN)) && @@ -176,6 +172,21 @@ static int addrtype_mt_checkentry_v1(const struct xt_mtchk_param *par) goto err; } + return 0; +err: + pr_info_ratelimited("%s\n", errmsg); + return -EINVAL; +} + +static int addrtype_mt_checkentry_v1(const struct xt_mtchk_param *par) +{ + const char *errmsg = "both incoming and outgoing interface limitation cannot be selected"; + struct xt_addrtype_info_v1 *info = par->matchinfo; + + if (info->flags & XT_ADDRTYPE_LIMIT_IFACE_IN && + info->flags & XT_ADDRTYPE_LIMIT_IFACE_OUT) + goto err; + #if IS_ENABLED(CONFIG_IP6_NF_IPTABLES) if (par->family == NFPROTO_IPV6) { if ((info->source | info->dest) & XT_ADDRTYPE_BLACKHOLE) { @@ -211,6 +222,7 @@ static struct xt_match addrtype_mt_reg[] __read_mostly = { .family = NFPROTO_IPV4, .revision = 1, .match = addrtype_mt_v1, + .check_hooks = addrtype_mt_check_hooks, .checkentry = addrtype_mt_checkentry_v1, .matchsize = sizeof(struct xt_addrtype_info_v1), .me = THIS_MODULE @@ -221,6 +233,7 @@ static struct xt_match addrtype_mt_reg[] __read_mostly = { .family = NFPROTO_IPV6, .revision = 1, .match = addrtype_mt_v1, + .check_hooks = addrtype_mt_check_hooks, .checkentry = addrtype_mt_checkentry_v1, .matchsize = sizeof(struct xt_addrtype_info_v1), .me = THIS_MODULE diff --git a/net/netfilter/xt_devgroup.c b/net/netfilter/xt_devgroup.c index 9520dd00070b..6d1a44ab5eee 100644 --- a/net/netfilter/xt_devgroup.c +++ b/net/netfilter/xt_devgroup.c @@ -33,14 +33,10 @@ static bool devgroup_mt(const struct sk_buff *skb, struct xt_action_param *par) return true; } -static int devgroup_mt_checkentry(const struct xt_mtchk_param *par) +static int devgroup_mt_check_hooks(const struct xt_mtchk_param *par) { const struct xt_devgroup_info *info = par->matchinfo; - if (info->flags & ~(XT_DEVGROUP_MATCH_SRC | XT_DEVGROUP_INVERT_SRC | - XT_DEVGROUP_MATCH_DST | XT_DEVGROUP_INVERT_DST)) - return -EINVAL; - if (info->flags & XT_DEVGROUP_MATCH_SRC && par->hook_mask & ~((1 << NF_INET_PRE_ROUTING) | (1 << NF_INET_LOCAL_IN) | @@ -56,9 +52,21 @@ static int devgroup_mt_checkentry(const struct xt_mtchk_param *par) return 0; } +static int devgroup_mt_checkentry(const struct xt_mtchk_param *par) +{ + const struct xt_devgroup_info *info = par->matchinfo; + + if (info->flags & ~(XT_DEVGROUP_MATCH_SRC | XT_DEVGROUP_INVERT_SRC | + XT_DEVGROUP_MATCH_DST | XT_DEVGROUP_INVERT_DST)) + return -EINVAL; + + return 0; +} + static struct xt_match devgroup_mt_reg __read_mostly = { .name = "devgroup", .match = devgroup_mt, + .check_hooks = devgroup_mt_check_hooks, .checkentry = devgroup_mt_checkentry, .matchsize = sizeof(struct xt_devgroup_info), .family = NFPROTO_UNSPEC, diff --git a/net/netfilter/xt_physdev.c b/net/netfilter/xt_physdev.c index d2b0b52434fa..dd98f758176c 100644 --- a/net/netfilter/xt_physdev.c +++ b/net/netfilter/xt_physdev.c @@ -91,6 +91,21 @@ physdev_mt(const struct sk_buff *skb, struct xt_action_param *par) return (!!ret ^ !(info->invert & XT_PHYSDEV_OP_OUT)); } +static int physdev_mt_check_hooks(const struct xt_mtchk_param *par) +{ + const struct xt_physdev_info *info = par->matchinfo; + + if (info->bitmask & (XT_PHYSDEV_OP_OUT | XT_PHYSDEV_OP_ISOUT) && + (!(info->bitmask & XT_PHYSDEV_OP_BRIDGED) || + info->invert & XT_PHYSDEV_OP_BRIDGED) && + par->hook_mask & (1 << NF_INET_LOCAL_OUT)) { + pr_info_ratelimited("--physdev-out and --physdev-is-out only supported in the FORWARD and POSTROUTING chains with bridged traffic\n"); + return -EINVAL; + } + + return 0; +} + static int physdev_mt_check(const struct xt_mtchk_param *par) { const struct xt_physdev_info *info = par->matchinfo; @@ -99,13 +114,6 @@ static int physdev_mt_check(const struct xt_mtchk_param *par) if (!(info->bitmask & XT_PHYSDEV_OP_MASK) || info->bitmask & ~XT_PHYSDEV_OP_MASK) return -EINVAL; - if (info->bitmask & (XT_PHYSDEV_OP_OUT | XT_PHYSDEV_OP_ISOUT) && - (!(info->bitmask & XT_PHYSDEV_OP_BRIDGED) || - info->invert & XT_PHYSDEV_OP_BRIDGED) && - par->hook_mask & (1 << NF_INET_LOCAL_OUT)) { - pr_info_ratelimited("--physdev-out and --physdev-is-out only supported in the FORWARD and POSTROUTING chains with bridged traffic\n"); - return -EINVAL; - } #define X(memb) strnlen(info->memb, sizeof(info->memb)) >= sizeof(info->memb) if (info->bitmask & XT_PHYSDEV_OP_IN) { @@ -141,6 +149,7 @@ static struct xt_match physdev_mt_reg[] __read_mostly = { { .name = "physdev", .family = NFPROTO_IPV4, + .check_hooks = physdev_mt_check_hooks, .checkentry = physdev_mt_check, .match = physdev_mt, .matchsize = sizeof(struct xt_physdev_info), @@ -149,6 +158,7 @@ static struct xt_match physdev_mt_reg[] __read_mostly = { { .name = "physdev", .family = NFPROTO_IPV6, + .check_hooks = physdev_mt_check_hooks, .checkentry = physdev_mt_check, .match = physdev_mt, .matchsize = sizeof(struct xt_physdev_info), diff --git a/net/netfilter/xt_policy.c b/net/netfilter/xt_policy.c index b5fa65558318..ff54e3a8581e 100644 --- a/net/netfilter/xt_policy.c +++ b/net/netfilter/xt_policy.c @@ -126,13 +126,10 @@ policy_mt(const struct sk_buff *skb, struct xt_action_param *par) return ret; } -static int policy_mt_check(const struct xt_mtchk_param *par) +static int policy_mt_check_hooks(const struct xt_mtchk_param *par) { const struct xt_policy_info *info = par->matchinfo; - const char *errmsg = "neither incoming nor outgoing policy selected"; - - if (!(info->flags & (XT_POLICY_MATCH_IN|XT_POLICY_MATCH_OUT))) - goto err; + const char *errmsg; if (par->hook_mask & ((1 << NF_INET_PRE_ROUTING) | (1 << NF_INET_LOCAL_IN)) && info->flags & XT_POLICY_MATCH_OUT) { @@ -144,6 +141,21 @@ static int policy_mt_check(const struct xt_mtchk_param *par) errmsg = "input policy not valid in POSTROUTING and OUTPUT"; goto err; } + + return 0; +err: + pr_info_ratelimited("%s\n", errmsg); + return -EINVAL; +} + +static int policy_mt_check(const struct xt_mtchk_param *par) +{ + const struct xt_policy_info *info = par->matchinfo; + const char *errmsg = "neither incoming nor outgoing policy selected"; + + if (!(info->flags & (XT_POLICY_MATCH_IN|XT_POLICY_MATCH_OUT))) + goto err; + if (info->len > XT_POLICY_MAX_ELEM) { errmsg = "too many policy elements"; goto err; @@ -158,6 +170,7 @@ static struct xt_match policy_mt_reg[] __read_mostly = { { .name = "policy", .family = NFPROTO_IPV4, + .check_hooks = policy_mt_check_hooks, .checkentry = policy_mt_check, .match = policy_mt, .matchsize = sizeof(struct xt_policy_info), @@ -166,6 +179,7 @@ static struct xt_match policy_mt_reg[] __read_mostly = { { .name = "policy", .family = NFPROTO_IPV6, + .check_hooks = policy_mt_check_hooks, .checkentry = policy_mt_check, .match = policy_mt, .matchsize = sizeof(struct xt_policy_info), diff --git a/net/netfilter/xt_set.c b/net/netfilter/xt_set.c index 731bc2cafae4..4ae04bba9358 100644 --- a/net/netfilter/xt_set.c +++ b/net/netfilter/xt_set.c @@ -430,6 +430,29 @@ set_target_v3(struct sk_buff *skb, const struct xt_action_param *par) return XT_CONTINUE; } +static int +set_target_v3_check_hooks(const struct xt_tgchk_param *par) +{ + const struct xt_set_info_target_v3 *info = par->targinfo; + + if (info->map_set.index != IPSET_INVALID_ID) { + if (strncmp(par->table, "mangle", 7)) { + pr_info_ratelimited("--map-set only usable from mangle table\n"); + return -EINVAL; + } + if (((info->flags & IPSET_FLAG_MAP_SKBPRIO) | + (info->flags & IPSET_FLAG_MAP_SKBQUEUE)) && + (par->hook_mask & ~(1 << NF_INET_FORWARD | + 1 << NF_INET_LOCAL_OUT | + 1 << NF_INET_POST_ROUTING))) { + pr_info_ratelimited("mapping of prio or/and queue is allowed only from OUTPUT/FORWARD/POSTROUTING chains\n"); + return -EINVAL; + } + } + + return 0; +} + static int set_target_v3_checkentry(const struct xt_tgchk_param *par) { @@ -459,20 +482,6 @@ set_target_v3_checkentry(const struct xt_tgchk_param *par) } if (info->map_set.index != IPSET_INVALID_ID) { - if (strncmp(par->table, "mangle", 7)) { - pr_info_ratelimited("--map-set only usable from mangle table\n"); - ret = -EINVAL; - goto cleanup_del; - } - if (((info->flags & IPSET_FLAG_MAP_SKBPRIO) | - (info->flags & IPSET_FLAG_MAP_SKBQUEUE)) && - (par->hook_mask & ~(1 << NF_INET_FORWARD | - 1 << NF_INET_LOCAL_OUT | - 1 << NF_INET_POST_ROUTING))) { - pr_info_ratelimited("mapping of prio or/and queue is allowed only from OUTPUT/FORWARD/POSTROUTING chains\n"); - ret = -EINVAL; - goto cleanup_del; - } index = ip_set_nfnl_get_byindex(par->net, info->map_set.index); if (index == IPSET_INVALID_ID) { @@ -672,6 +681,7 @@ static struct xt_target set_targets[] __read_mostly = { .family = NFPROTO_IPV4, .target = set_target_v3, .targetsize = sizeof(struct xt_set_info_target_v3), + .check_hooks = set_target_v3_check_hooks, .checkentry = set_target_v3_checkentry, .destroy = set_target_v3_destroy, .me = THIS_MODULE @@ -682,6 +692,7 @@ static struct xt_target set_targets[] __read_mostly = { .family = NFPROTO_IPV6, .target = set_target_v3, .targetsize = sizeof(struct xt_set_info_target_v3), + .check_hooks = set_target_v3_check_hooks, .checkentry = set_target_v3_checkentry, .destroy = set_target_v3_destroy, .me = THIS_MODULE From 2f768d638d977eff824f64dcc9639e3fea32da8f Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Tue, 28 Apr 2026 19:04:07 +0200 Subject: [PATCH 101/525] netfilter: nft_compat: run xt_check_hooks_{match,target}() from .validate Several matches and one target check that the hook is correct from checkentry(), however, the basechain is only available from nft_table_validate(). This patch uses xt_check_hooks_{match,target}() from the nft_compat expression .validate path. This patch sets the table in the nft_ctx struct in nft_table_validate() which is required by this patch. Based on patch from Florian Westphal. Fixes: 0ca743a55991 ("netfilter: nf_tables: add compatibility layer for x_tables") Reported-by: Xiang Mei Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 1 + net/netfilter/nft_compat.c | 45 +++++++++++++++++++++++++++-------- 2 files changed, 36 insertions(+), 10 deletions(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index d20ce5c36d31..38e33c66c618 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -4205,6 +4205,7 @@ static int nft_table_validate(struct net *net, const struct nft_table *table) struct nft_chain *chain; struct nft_ctx ctx = { .net = net, + .table = (struct nft_table *)table, .family = table->family, }; int err = 0; diff --git a/net/netfilter/nft_compat.c b/net/netfilter/nft_compat.c index decc725a33c2..0caa9304d2d0 100644 --- a/net/netfilter/nft_compat.c +++ b/net/netfilter/nft_compat.c @@ -261,10 +261,10 @@ nft_target_init(const struct nft_ctx *ctx, const struct nft_expr *expr, return ret; } - nft_target_set_tgchk_param(&par, ctx, target, info, &e, proto, inv); - nft_compat_wait_for_destructors(ctx->net); + nft_target_set_tgchk_param(&par, ctx, target, info, &e, proto, inv); + ret = xt_check_target(&par, size, proto, inv); if (ret < 0) { if (ret == -ENOENT) { @@ -353,8 +353,6 @@ static int nft_target_dump(struct sk_buff *skb, static int nft_target_validate(const struct nft_ctx *ctx, const struct nft_expr *expr) { - struct xt_target *target = expr->ops->data; - unsigned int hook_mask = 0; int ret; if (ctx->family != NFPROTO_IPV4 && @@ -377,11 +375,21 @@ static int nft_target_validate(const struct nft_ctx *ctx, const struct nft_base_chain *basechain = nft_base_chain(ctx->chain); const struct nf_hook_ops *ops = &basechain->ops; + unsigned int hook_mask = 1 << ops->hooknum; + struct xt_target *target = expr->ops->data; + void *info = nft_expr_priv(expr); + struct xt_tgchk_param par; + union nft_entry e = {}; - hook_mask = 1 << ops->hooknum; if (target->hooks && !(hook_mask & target->hooks)) return -EINVAL; + nft_target_set_tgchk_param(&par, ctx, target, info, &e, 0, false); + + ret = xt_check_hooks_target(&par); + if (ret < 0) + return ret; + ret = nft_compat_chain_validate_dependency(ctx, target->table); if (ret < 0) return ret; @@ -515,10 +523,10 @@ __nft_match_init(const struct nft_ctx *ctx, const struct nft_expr *expr, return ret; } - nft_match_set_mtchk_param(&par, ctx, match, info, &e, proto, inv); - nft_compat_wait_for_destructors(ctx->net); + nft_match_set_mtchk_param(&par, ctx, match, info, &e, proto, inv); + return xt_check_match(&par, size, proto, inv); } @@ -614,8 +622,6 @@ static int nft_match_large_dump(struct sk_buff *skb, static int nft_match_validate(const struct nft_ctx *ctx, const struct nft_expr *expr) { - struct xt_match *match = expr->ops->data; - unsigned int hook_mask = 0; int ret; if (ctx->family != NFPROTO_IPV4 && @@ -638,11 +644,30 @@ static int nft_match_validate(const struct nft_ctx *ctx, const struct nft_base_chain *basechain = nft_base_chain(ctx->chain); const struct nf_hook_ops *ops = &basechain->ops; + unsigned int hook_mask = 1 << ops->hooknum; + struct xt_match *match = expr->ops->data; + size_t size = XT_ALIGN(match->matchsize); + struct xt_mtchk_param par; + union nft_entry e = {}; + void *info; - hook_mask = 1 << ops->hooknum; if (match->hooks && !(hook_mask & match->hooks)) return -EINVAL; + if (NFT_EXPR_SIZE(size) > NFT_MATCH_LARGE_THRESH) { + struct nft_xt_match_priv *priv = nft_expr_priv(expr); + + info = priv->info; + } else { + info = nft_expr_priv(expr); + } + + nft_match_set_mtchk_param(&par, ctx, match, info, &e, 0, false); + + ret = xt_check_hooks_match(&par); + if (ret < 0) + return ret; + ret = nft_compat_chain_validate_dependency(ctx, match->table); if (ret < 0) return ret; From 8bedb6c46945752a688d9b0cf2021e0e68b1876c Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Tue, 28 Apr 2026 19:37:57 +0200 Subject: [PATCH 102/525] netfilter: xt_CT: fix usersize for v1 and v2 revision While resurrecting the conntrack-tool test cases I found following bug: In: iptables -I OUTPUT -t raw -p 13 -j CT --timeout test-generic Out: [0:0] -A OUTPUT -p 13 -j CT --timeout test Data after first four bytes of the timeout policy name is never copied to userspace because its treated as kernel-only. Fixes: ec2318904965 ("xtables: extend matches and targets with .usersize") Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- net/netfilter/xt_CT.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/net/netfilter/xt_CT.c b/net/netfilter/xt_CT.c index 498f5871c84a..d2aeacf94230 100644 --- a/net/netfilter/xt_CT.c +++ b/net/netfilter/xt_CT.c @@ -354,7 +354,7 @@ static struct xt_target xt_ct_tg_reg[] __read_mostly = { .family = NFPROTO_IPV4, .revision = 1, .targetsize = sizeof(struct xt_ct_target_info_v1), - .usersize = offsetof(struct xt_ct_target_info, ct), + .usersize = offsetof(struct xt_ct_target_info_v1, ct), .checkentry = xt_ct_tg_check_v1, .destroy = xt_ct_tg_destroy_v1, .target = xt_ct_target_v1, @@ -366,7 +366,7 @@ static struct xt_target xt_ct_tg_reg[] __read_mostly = { .family = NFPROTO_IPV4, .revision = 2, .targetsize = sizeof(struct xt_ct_target_info_v1), - .usersize = offsetof(struct xt_ct_target_info, ct), + .usersize = offsetof(struct xt_ct_target_info_v1, ct), .checkentry = xt_ct_tg_check_v2, .destroy = xt_ct_tg_destroy_v1, .target = xt_ct_target_v1, @@ -398,7 +398,7 @@ static struct xt_target xt_ct_tg_reg[] __read_mostly = { .family = NFPROTO_IPV6, .revision = 1, .targetsize = sizeof(struct xt_ct_target_info_v1), - .usersize = offsetof(struct xt_ct_target_info, ct), + .usersize = offsetof(struct xt_ct_target_info_v1, ct), .checkentry = xt_ct_tg_check_v1, .destroy = xt_ct_tg_destroy_v1, .target = xt_ct_target_v1, @@ -410,7 +410,7 @@ static struct xt_target xt_ct_tg_reg[] __read_mostly = { .family = NFPROTO_IPV6, .revision = 2, .targetsize = sizeof(struct xt_ct_target_info_v1), - .usersize = offsetof(struct xt_ct_target_info, ct), + .usersize = offsetof(struct xt_ct_target_info_v1, ct), .checkentry = xt_ct_tg_check_v2, .destroy = xt_ct_tg_destroy_v1, .target = xt_ct_target_v1, From 63bac027860308d1344f761cb47aabb3b30973fd Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Wed, 29 Apr 2026 08:21:35 +0200 Subject: [PATCH 103/525] netfilter: nf_tables: fix netdev hook allocation memleak with dormant tables sashiko says: could the related code in __nf_tables_abort() leak the struct nft_hook objects when the table is dormant? In __nf_tables_abort(), when rolling back a NEWCHAIN transaction that updates hooks, the code conditionally unregisters and frees the hooks only if the table is not dormant [..] if (!(table->flags & NFT_TABLE_F_DORMANT)) { nft_netdev_unregister_hooks(net, &nft_trans_chain_hooks(trans), true); } ... nft_trans_destroy(trans); Unfortunately netdev family mixes hook registration and allocation. Push table struct down and only check for the flag to unregister. Fixes: 216e7bf7402c ("netfilter: nf_tables: skip netdev hook unregistration if table is dormant") Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 34 ++++++++++++++++++++-------------- 1 file changed, 20 insertions(+), 14 deletions(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 38e33c66c618..87387adbca65 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -407,6 +407,7 @@ static void nft_netdev_unregister_trans_hook(struct net *net, } static void nft_netdev_unregister_hooks(struct net *net, + const struct nft_table *table, struct list_head *hook_list, bool release_netdev) { @@ -414,8 +415,10 @@ static void nft_netdev_unregister_hooks(struct net *net, struct nf_hook_ops *ops; list_for_each_entry_safe(hook, next, hook_list, list) { - list_for_each_entry(ops, &hook->ops_list, list) - nf_unregister_net_hook(net, ops); + if (!(table->flags & NFT_TABLE_F_DORMANT)) { + list_for_each_entry(ops, &hook->ops_list, list) + nf_unregister_net_hook(net, ops); + } if (release_netdev) nft_netdev_hook_unlink_free_rcu(hook); } @@ -452,20 +455,25 @@ static void __nf_tables_unregister_hook(struct net *net, struct nft_base_chain *basechain; const struct nf_hook_ops *ops; - if (table->flags & NFT_TABLE_F_DORMANT || - !nft_is_base_chain(chain)) + if (!nft_is_base_chain(chain)) return; basechain = nft_base_chain(chain); ops = &basechain->ops; + /* must also be called for dormant tables */ + if (nft_base_chain_netdev(table->family, basechain->ops.hooknum)) { + nft_netdev_unregister_hooks(net, table, &basechain->hook_list, + release_netdev); + return; + } + + if (table->flags & NFT_TABLE_F_DORMANT) + return; + if (basechain->type->ops_unregister) return basechain->type->ops_unregister(net, ops); - if (nft_base_chain_netdev(table->family, basechain->ops.hooknum)) - nft_netdev_unregister_hooks(net, &basechain->hook_list, - release_netdev); - else - nf_unregister_net_hook(net, &basechain->ops); + nf_unregister_net_hook(net, &basechain->ops); } static void nf_tables_unregister_hook(struct net *net, @@ -11282,11 +11290,9 @@ static int __nf_tables_abort(struct net *net, enum nfnl_abort_action action) break; case NFT_MSG_NEWCHAIN: if (nft_trans_chain_update(trans)) { - if (!(table->flags & NFT_TABLE_F_DORMANT)) { - nft_netdev_unregister_hooks(net, - &nft_trans_chain_hooks(trans), - true); - } + nft_netdev_unregister_hooks(net, table, + &nft_trans_chain_hooks(trans), + true); free_percpu(nft_trans_chain_stats(trans)); kfree(nft_trans_chain_name(trans)); nft_trans_destroy(trans); From e6a650acbd991bba279f2580853aed9a8d166e6f Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Wed, 29 Apr 2026 21:18:25 +0200 Subject: [PATCH 104/525] parisc: Fix build failure for 32-bit kernel with PA2.0 instruction set The CONFIG_PA11 option can not be used as a reliable check if we build a 32-bit kernel which needs the 32-bit VDSO. Instead depend on CONFIG_64BIT and CONFIG_COMPAT only. Reported-by: Christoph Biedl Tested-by: Christoph Biedl Signed-off-by: Helge Deller --- arch/parisc/Makefile | 16 +++++++++++----- arch/parisc/kernel/Makefile | 7 +++++-- 2 files changed, 16 insertions(+), 7 deletions(-) diff --git a/arch/parisc/Makefile b/arch/parisc/Makefile index edab2a948352..4391783521bd 100644 --- a/arch/parisc/Makefile +++ b/arch/parisc/Makefile @@ -174,15 +174,21 @@ ifeq ($(KBUILD_EXTMOD),) # this hack. prepare: vdso_prepare vdso_prepare: prepare0 - $(if $(CONFIG_64BIT),$(Q)$(MAKE) \ - $(build)=arch/parisc/kernel/vdso64 include/generated/vdso64-offsets.h) - $(if $(CONFIG_PA11)$(CONFIG_COMPAT),$(Q)$(MAKE) \ +ifdef CONFIG_64BIT + $(Q)$(MAKE) $(build)=arch/parisc/kernel/vdso64 include/generated/vdso64-offsets.h + $(if $(CONFIG_COMPAT),$(Q)$(MAKE) \ $(build)=arch/parisc/kernel/vdso32 include/generated/vdso32-offsets.h) +else + $(Q)$(MAKE) $(build)=arch/parisc/kernel/vdso32 include/generated/vdso32-offsets.h +endif endif -vdso-install-$(CONFIG_PA11) += arch/parisc/kernel/vdso32/vdso32.so +ifdef CONFIG_64BIT +vdso-install-y += arch/parisc/kernel/vdso64/vdso64.so vdso-install-$(CONFIG_COMPAT) += arch/parisc/kernel/vdso32/vdso32.so -vdso-install-$(CONFIG_64BIT) += arch/parisc/kernel/vdso64/vdso64.so +else +vdso-install-y += arch/parisc/kernel/vdso32/vdso32.so +endif install: KBUILD_IMAGE := vmlinux zinstall: KBUILD_IMAGE := vmlinuz diff --git a/arch/parisc/kernel/Makefile b/arch/parisc/kernel/Makefile index 2f3441769ac5..49f937c2abbe 100644 --- a/arch/parisc/kernel/Makefile +++ b/arch/parisc/kernel/Makefile @@ -46,6 +46,9 @@ obj-$(CONFIG_KEXEC_FILE) += kexec_file.o # vdso obj-y += vdso.o -obj-$(CONFIG_64BIT) += vdso64/ -obj-$(CONFIG_PA11) += vdso32/ +ifdef CONFIG_64BIT +obj-y += vdso64/ obj-$(CONFIG_COMPAT) += vdso32/ +else +obj-y += vdso32/ +endif From 100201d349edd226ca3470c894c92dccc67ee7a8 Mon Sep 17 00:00:00 2001 From: Fabio Porcedda Date: Mon, 27 Apr 2026 11:17:46 +0200 Subject: [PATCH 105/525] USB: serial: option: add Telit Cinterion LE910Cx compositions Add the following Telit Cinterion LE910Cx compositions: 0x1251: RNDIS + tty (AT/NMEA) + tty (AT) + tty (AT) + tty (SAP) T: Bus=01 Lev=01 Prnt=21 Port=06 Cnt=01 Dev#=108 Spd=480 MxCh= 0 D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs= 1 P: Vendor=1bc7 ProdID=1251 Rev=03.18 S: Manufacturer=Android S: Product=LE910C1-EU S: SerialNumber=0123456789ABCDEF C: #Ifs= 6 Cfg#= 1 Atr=a0 MxPwr=500mA I: If#= 0 Alt= 0 #EPs= 1 Cls=02(commc) Sub=02 Prot=ff Driver=rndis_host E: Ad=82(I) Atr=03(Int.) MxPS= 8 Ivl=32ms I: If#= 1 Alt= 0 #EPs= 2 Cls=0a(data ) Sub=00 Prot=00 Driver=rndis_host E: Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms I: If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=84(I) Atr=03(Int.) MxPS= 10 Ivl=32ms I: If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=85(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=86(I) Atr=03(Int.) MxPS= 10 Ivl=32ms I: If#= 4 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=87(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=88(I) Atr=03(Int.) MxPS= 10 Ivl=32ms I: If#= 5 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=05(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=89(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=8a(I) Atr=03(Int.) MxPS= 10 Ivl=32ms 0x1253: ECM + tty (AT/NMEA) + tty (AT) + tty (AT) + tty (SAP) T: Bus=01 Lev=01 Prnt=21 Port=06 Cnt=01 Dev#=121 Spd=480 MxCh= 0 D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs= 1 P: Vendor=1bc7 ProdID=1253 Rev=03.18 S: Manufacturer=Android S: Product=LE910C1-EU S: SerialNumber=0123456789ABCDEF C: #Ifs= 6 Cfg#= 1 Atr=a0 MxPwr=500mA I: If#= 0 Alt= 0 #EPs= 1 Cls=02(commc) Sub=06 Prot=00 Driver=cdc_ether E: Ad=82(I) Atr=03(Int.) MxPS= 16 Ivl=32ms I: If#= 1 Alt= 1 #EPs= 2 Cls=0a(data ) Sub=00 Prot=00 Driver=cdc_ether E: Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms I: If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=84(I) Atr=03(Int.) MxPS= 10 Ivl=32ms I: If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=85(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=86(I) Atr=03(Int.) MxPS= 10 Ivl=32ms I: If#= 4 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=87(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=88(I) Atr=03(Int.) MxPS= 10 Ivl=32ms I: If#= 5 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=05(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=89(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=8a(I) Atr=03(Int.) MxPS= 10 Ivl=32ms 0x1254: tty (AT) + tty (AT) T: Bus=01 Lev=01 Prnt=21 Port=06 Cnt=01 Dev#=122 Spd=480 MxCh= 0 D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs= 1 P: Vendor=1bc7 ProdID=1254 Rev=03.18 S: Manufacturer=Android S: Product=LE910C1-EU S: SerialNumber=0123456789ABCDEF C: #Ifs= 2 Cfg#= 1 Atr=a0 MxPwr=500mA I: If#= 0 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=82(I) Atr=03(Int.) MxPS= 10 Ivl=32ms I: If#= 1 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=84(I) Atr=03(Int.) MxPS= 10 Ivl=32ms 0x1255: tty (AT/NMEA) + tty (AT) + tty (AT) + tty (SAP) T: Bus=01 Lev=01 Prnt=21 Port=06 Cnt=01 Dev#=123 Spd=480 MxCh= 0 D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs= 1 P: Vendor=1bc7 ProdID=1255 Rev=03.18 S: Manufacturer=Android S: Product=LE910C1-EU S: SerialNumber=0123456789ABCDEF C: #Ifs= 4 Cfg#= 1 Atr=a0 MxPwr=500mA I: If#= 0 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=82(I) Atr=03(Int.) MxPS= 10 Ivl=32ms I: If#= 1 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=84(I) Atr=03(Int.) MxPS= 10 Ivl=32ms I: If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=85(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=86(I) Atr=03(Int.) MxPS= 10 Ivl=32ms I: If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=option E: Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=87(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=88(I) Atr=03(Int.) MxPS= 10 Ivl=32ms Cc: stable@vger.kernel.org Signed-off-by: Fabio Porcedda Signed-off-by: Johan Hovold --- drivers/usb/serial/option.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c index c71461893d20..42e4cecd28ac 100644 --- a/drivers/usb/serial/option.c +++ b/drivers/usb/serial/option.c @@ -1513,7 +1513,11 @@ static const struct usb_device_id option_ids[] = { { USB_DEVICE_INTERFACE_CLASS(TELIT_VENDOR_ID, 0x1231, 0xff), /* Telit LE910Cx (RNDIS) */ .driver_info = NCTRL(2) | RSVD(3) }, { USB_DEVICE_AND_INTERFACE_INFO(TELIT_VENDOR_ID, 0x1250, 0xff, 0x00, 0x00) }, /* Telit LE910Cx (rmnet) */ + { USB_DEVICE_INTERFACE_CLASS(TELIT_VENDOR_ID, 0x1251, 0xff) }, /* Telit LE910Cx (RNDIS) */ { USB_DEVICE_INTERFACE_CLASS(TELIT_VENDOR_ID, 0x1252, 0xff) }, /* Telit LE910Cx (MBIM) */ + { USB_DEVICE_INTERFACE_CLASS(TELIT_VENDOR_ID, 0x1253, 0xff) }, /* Telit LE910Cx (ECM) */ + { USB_DEVICE_INTERFACE_CLASS(TELIT_VENDOR_ID, 0x1254, 0xff) }, /* Telit LE910Cx */ + { USB_DEVICE_INTERFACE_CLASS(TELIT_VENDOR_ID, 0x1255, 0xff) }, /* Telit LE910Cx */ { USB_DEVICE(TELIT_VENDOR_ID, 0x1260), .driver_info = NCTRL(0) | RSVD(1) | RSVD(2) }, { USB_DEVICE(TELIT_VENDOR_ID, 0x1261), From 7dd57d7a6350770dfc283287125c409e995200e0 Mon Sep 17 00:00:00 2001 From: Karol Wachowski Date: Thu, 30 Apr 2026 11:56:44 +0200 Subject: [PATCH 106/525] accel/ivpu: Disallow re-exporting imported GEM objects Prevent re-exporting of imported GEM buffers by adding a custom prime_handle_to_fd callback that checks if the object is imported and returns -EOPNOTSUPP if so. Re-exporting imported GEM buffers causes loss of buffer flags settings, leading to incorrect device access and data corruption. Reported-by: Yametsu Fixes: 57557964b582 ("accel/ivpu: Add support for userptr buffer objects") Reviewed-by: Andrzej Kacprowski Signed-off-by: Karol Wachowski Cc: # v6.19+ --- drivers/accel/ivpu/ivpu_drv.c | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/drivers/accel/ivpu/ivpu_drv.c b/drivers/accel/ivpu/ivpu_drv.c index 2801378e3e19..3b7b008bccfe 100644 --- a/drivers/accel/ivpu/ivpu_drv.c +++ b/drivers/accel/ivpu/ivpu_drv.c @@ -537,6 +537,26 @@ static const struct file_operations ivpu_fops = { #endif }; +static int ivpu_gem_prime_handle_to_fd(struct drm_device *dev, struct drm_file *file_priv, + u32 handle, u32 flags, int *prime_fd) +{ + struct drm_gem_object *obj; + + obj = drm_gem_object_lookup(file_priv, handle); + if (!obj) + return -ENOENT; + + if (drm_gem_is_imported(obj)) { + /* Do not allow re-exporting */ + drm_gem_object_put(obj); + return -EOPNOTSUPP; + } + + drm_gem_object_put(obj); + + return drm_gem_prime_handle_to_fd(dev, file_priv, handle, flags, prime_fd); +} + static const struct drm_driver driver = { .driver_features = DRIVER_GEM | DRIVER_COMPUTE_ACCEL, @@ -545,6 +565,7 @@ static const struct drm_driver driver = { .gem_create_object = ivpu_gem_create_object, .gem_prime_import = ivpu_gem_prime_import, + .prime_handle_to_fd = ivpu_gem_prime_handle_to_fd, .ioctls = ivpu_drm_ioctls, .num_ioctls = ARRAY_SIZE(ivpu_drm_ioctls), From d3b7a868f1aeb6a43de880a3709ef3a0341f2c2a Mon Sep 17 00:00:00 2001 From: Kurt Borja Date: Wed, 29 Apr 2026 08:20:56 -0500 Subject: [PATCH 107/525] platform/wmi: Fix unchecked min_size in wmidev_invoke_method() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit After calling wmidev_evaluate_method(), if the ACPI core does not return an out object, then wmidev_invoke_method() bypasses the min_size check and returns 0. Add a check for min_size if there is not an out object. Fixes: 1aeded2f55f0 ("platform/wmi: Extend wmidev_query_block() to reject undersized data") Closes: https://sashiko.dev/#/patchset/20260406203237.2970-1-W_Armin%40gmx.de Signed-off-by: Kurt Borja Reviewed-by: Armin Wolf Link: https://patch.msgid.link/20260429-invoke-fix-v1-1-ce938eb80cd3@gmail.com Reviewed-by: Ilpo Järvinen Signed-off-by: Ilpo Järvinen --- drivers/platform/wmi/core.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/platform/wmi/core.c b/drivers/platform/wmi/core.c index 7aa40dab6145..5a2ffcbab6af 100644 --- a/drivers/platform/wmi/core.c +++ b/drivers/platform/wmi/core.c @@ -411,6 +411,9 @@ int wmidev_invoke_method(struct wmi_device *wdev, u8 instance, u32 method_id, obj = aout.pointer; if (!obj) { + if (min_size != 0) + return -ENOMSG; + out->length = 0; out->data = ZERO_SIZE_PTR; From c2d4b76458c9ebf81eae2916970320f1f61ad002 Mon Sep 17 00:00:00 2001 From: Krishna Chomal Date: Wed, 29 Apr 2026 23:39:53 +0530 Subject: [PATCH 108/525] platform/x86: hp-wmi: silence unknown board warning for 8D41 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The HP Omen Max 16-ah0xxx, DMI board ID 8D41 is currently marked with victus_s_thermal_params in the victus_s_thermal_profile_boards[] list. This disables thermal profile readback and adds a dmesg warning during driver init for "unknown board". After testing we know that (similar to another HP Omen Max 16 device, board ID 8D87), the embedded controller on this board does not expose thermal profile which means we have to intentionally disable EC readback. Changing its driver_data to omen_v1_no_ec_thermal_params is sufficient to silence the warning. Tested-by: Benjamin Y Signed-off-by: Krishna Chomal Link: https://patch.msgid.link/20260429180953.129885-1-krishna.chomal108@gmail.com Reviewed-by: Ilpo Järvinen Signed-off-by: Ilpo Järvinen --- drivers/platform/x86/hp/hp-wmi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/platform/x86/hp/hp-wmi.c b/drivers/platform/x86/hp/hp-wmi.c index d1cc6e7d176c..24c151289dd3 100644 --- a/drivers/platform/x86/hp/hp-wmi.c +++ b/drivers/platform/x86/hp/hp-wmi.c @@ -243,7 +243,7 @@ static const struct dmi_system_id victus_s_thermal_profile_boards[] __initconst }, { .matches = { DMI_MATCH(DMI_BOARD_NAME, "8D41") }, - .driver_data = (void *)&victus_s_thermal_params, + .driver_data = (void *)&omen_v1_no_ec_thermal_params, }, { .matches = { DMI_MATCH(DMI_BOARD_NAME, "8D87") }, From 863810d4985ad214f70c1623f24384ccc850f2a2 Mon Sep 17 00:00:00 2001 From: Yufei CHENG Date: Mon, 27 Apr 2026 00:50:34 +0800 Subject: [PATCH 109/525] platform/x86: lenovo: wmi-other: Fix uninitialized variable in lwmi_om_hwmon_write() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When the flag relax_fan_constraint is set, local variable 'raw' is never assigned, and lwmi_om_hwmon_write() will pass uninitialized value to lwmi_om_fan_get_set() resulting in undefined behavior. This flag allows user to bypass minimum fan RPM divisor rounding, but assignment to 'raw' only happens in the non-relaxed path. Fix by defaulting 'raw' to user provided 'val' in the else branch. Fixes: 51ed34282f63 ("platform/x86: lenovo-wmi-other: Add HWMON for fan reporting/tuning") Reviewed-by: Rong Zhang Signed-off-by: Yufei CHENG Link: https://patch.msgid.link/20260426165034.9073-1-cd345al@gmail.com Reviewed-by: Ilpo Järvinen Signed-off-by: Ilpo Järvinen --- drivers/platform/x86/lenovo/wmi-other.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/platform/x86/lenovo/wmi-other.c b/drivers/platform/x86/lenovo/wmi-other.c index 6040f45aa2b0..6c2febe1a595 100644 --- a/drivers/platform/x86/lenovo/wmi-other.c +++ b/drivers/platform/x86/lenovo/wmi-other.c @@ -349,6 +349,8 @@ static int lwmi_om_hwmon_write(struct device *dev, enum hwmon_sensor_types type, */ if (!relax_fan_constraint) raw = val / LWMI_FAN_DIV * LWMI_FAN_DIV; + else + raw = val; err = lwmi_om_fan_get_set(priv, channel, &raw, true); if (err) From 92a8b5e2eff6920bf815cd6a80b088ec3fdf01a3 Mon Sep 17 00:00:00 2001 From: Yuriy Padlyak Date: Thu, 30 Apr 2026 01:09:03 +0300 Subject: [PATCH 110/525] ALSA: hda/realtek: Fix speaker silence after S3 resume on Xiaomi Mi Laptop Pro 15 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The Xiaomi Mi Laptop Pro 15 (TM1905, subsystem 1d72:1905) ships with the Realtek ALC256 codec on Intel Comet Lake PCH-LP. After S3 resume the codec sets coefficient register 0x10 to 0x0220 instead of 0x0020 — bit 9 is erroneously set, which silences the internal speaker. Bluetooth and HDMI audio are unaffected because they use different paths. This is the same mechanism fixed for Clevo NJ51CU by commit edca7cc4b0ac ("ALSA: hda/realtek: Fix quirk for Clevo NJ51CU"), but the existing ALC256_FIXUP_MIC_NO_PRESENCE_AND_RESUME also reconfigures pin 0x19 as a front mic, which is wrong for this Xiaomi where pin 0x19 default is 0x411111f0 (disabled). Add a minimal fixup that only clears the stuck coef bit, and add the Xiaomi SSID to the quirk table. Verified by reading coef 0x10 with hda-verb after resume (returns 0x0220), writing 0x0020, and confirming the internal speaker resumes output. With this fixup applied the bit is cleared on every codec init, including post-resume. Signed-off-by: Yuriy Padlyak Cc: Tested-by: Yuriy Padlyak Link: https://patch.msgid.link/20260429220903.14918-1-yuriypadlyak@gmail.com Signed-off-by: Takashi Iwai --- sound/hda/codecs/realtek/alc269.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/sound/hda/codecs/realtek/alc269.c b/sound/hda/codecs/realtek/alc269.c index a9cd03bb73c4..9a5747c2e359 100644 --- a/sound/hda/codecs/realtek/alc269.c +++ b/sound/hda/codecs/realtek/alc269.c @@ -3390,6 +3390,19 @@ static void alc256_fixup_mic_no_presence_and_resume(struct hda_codec *codec, } } +static void alc256_fixup_xiaomi_pro15_resume(struct hda_codec *codec, + const struct hda_fixup *fix, + int action) +{ + /* + * On the Xiaomi Mi Laptop Pro 15 (TM1905, SSID 1d72:1905) the ALC256 + * codec sets coefficient 0x10 bit 9 to 1 after S3 resume, silencing + * the internal speaker. Bluetooth and HDMI audio are unaffected. + * Clear the bit so the speaker keeps working across suspend cycles. + */ + alc_update_coef_idx(codec, 0x10, 1<<9, 0); +} + static void alc256_decrease_headphone_amp_val(struct hda_codec *codec, const struct hda_fixup *fix, int action) { @@ -4052,6 +4065,7 @@ enum { ALC256_FIXUP_SYSTEM76_MIC_NO_PRESENCE, ALC233_FIXUP_NO_AUDIO_JACK, ALC256_FIXUP_MIC_NO_PRESENCE_AND_RESUME, + ALC256_FIXUP_XIAOMI_PRO15_RESUME, ALC285_FIXUP_LEGION_Y9000X_SPEAKERS, ALC285_FIXUP_LEGION_Y9000X_AUTOMUTE, ALC287_FIXUP_LEGION_16ACHG6, @@ -6240,6 +6254,10 @@ static const struct hda_fixup alc269_fixups[] = { .chained = true, .chain_id = ALC269_FIXUP_HEADSET_MODE_NO_HP_MIC }, + [ALC256_FIXUP_XIAOMI_PRO15_RESUME] = { + .type = HDA_FIXUP_FUNC, + .v.func = alc256_fixup_xiaomi_pro15_resume, + }, [ALC287_FIXUP_LEGION_16ACHG6] = { .type = HDA_FIXUP_FUNC, .v.func = alc287_fixup_legion_16achg6_speakers, @@ -7774,6 +7792,7 @@ static const struct hda_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x1d72, 0x1602, "RedmiBook", ALC255_FIXUP_XIAOMI_HEADSET_MIC), SND_PCI_QUIRK(0x1d72, 0x1701, "XiaomiNotebook Pro", ALC298_FIXUP_DELL1_MIC_NO_PRESENCE), SND_PCI_QUIRK(0x1d72, 0x1901, "RedmiBook 14", ALC256_FIXUP_ASUS_HEADSET_MIC), + SND_PCI_QUIRK(0x1d72, 0x1905, "Xiaomi Mi Laptop Pro 15", ALC256_FIXUP_XIAOMI_PRO15_RESUME), SND_PCI_QUIRK(0x1d72, 0x1945, "Redmi G", ALC256_FIXUP_ASUS_HEADSET_MIC), SND_PCI_QUIRK(0x1d72, 0x1947, "RedmiBook Air", ALC255_FIXUP_XIAOMI_HEADSET_MIC), SND_PCI_QUIRK(0x1e39, 0xca14, "MEDION NM14LNL", ALC233_FIXUP_MEDION_MTL_SPK), From 0bf00859d7a5ab685901c36f29df063b825cfaaa Mon Sep 17 00:00:00 2001 From: Fernando Fernandez Mancera Date: Tue, 28 Apr 2026 12:25:46 +0200 Subject: [PATCH 111/525] netfilter: nf_socket: skip socket lookup for non-first fragments Both nft_socket and xt_socket relies on L4 headers to perform socket lookup in the slow path. For fragmented packets, while the IP protocol remains constant across all fragments, only the first fragment contains the actual L4 header. As the expression/match could be attached to a chain with a priority lower than -400, it could bypass defragmentation. Add a check for fragmentation in the lookup functions directly so the problem is handled for both nft_socket and xt_socket at the same time. In addition, future users of the functions would not need to care about this. Fixes: 902d6a4c2a4f ("netfilter: nf_defrag: Skip defrag if NOTRACK is set") Fixes: 554ced0a6e29 ("netfilter: nf_tables: add support for native socket matching") Signed-off-by: Fernando Fernandez Mancera Signed-off-by: Pablo Neira Ayuso --- net/ipv4/netfilter/nf_socket_ipv4.c | 3 +++ net/ipv6/netfilter/nf_socket_ipv6.c | 5 +++-- 2 files changed, 6 insertions(+), 2 deletions(-) diff --git a/net/ipv4/netfilter/nf_socket_ipv4.c b/net/ipv4/netfilter/nf_socket_ipv4.c index 5080fa5fbf6a..f9c6755f5ec5 100644 --- a/net/ipv4/netfilter/nf_socket_ipv4.c +++ b/net/ipv4/netfilter/nf_socket_ipv4.c @@ -94,6 +94,9 @@ struct sock *nf_sk_lookup_slow_v4(struct net *net, const struct sk_buff *skb, #endif int doff = 0; + if (ntohs(iph->frag_off) & IP_OFFSET) + return NULL; + if (iph->protocol == IPPROTO_UDP || iph->protocol == IPPROTO_TCP) { struct tcphdr _hdr; struct udphdr *hp; diff --git a/net/ipv6/netfilter/nf_socket_ipv6.c b/net/ipv6/netfilter/nf_socket_ipv6.c index ced8bd44828e..893f2aeb4711 100644 --- a/net/ipv6/netfilter/nf_socket_ipv6.c +++ b/net/ipv6/netfilter/nf_socket_ipv6.c @@ -100,6 +100,7 @@ struct sock *nf_sk_lookup_slow_v6(struct net *net, const struct sk_buff *skb, const struct in6_addr *daddr = NULL, *saddr = NULL; struct ipv6hdr *iph = ipv6_hdr(skb), ipv6_var; struct sk_buff *data_skb = NULL; + unsigned short fragoff = 0; int doff = 0; int thoff = 0, tproto; #if IS_ENABLED(CONFIG_NF_CONNTRACK) @@ -107,8 +108,8 @@ struct sock *nf_sk_lookup_slow_v6(struct net *net, const struct sk_buff *skb, struct nf_conn const *ct; #endif - tproto = ipv6_find_hdr(skb, &thoff, -1, NULL, NULL); - if (tproto < 0) { + tproto = ipv6_find_hdr(skb, &thoff, -1, &fragoff, NULL); + if (tproto < 0 || fragoff) { pr_debug("unable to find transport header in IPv6 packet, dropping\n"); return NULL; } From 009d203e56dbe8db2589455b9e3644955f30313a Mon Sep 17 00:00:00 2001 From: Fernando Fernandez Mancera Date: Tue, 28 Apr 2026 12:25:47 +0200 Subject: [PATCH 112/525] netfilter: nf_tables: skip L4 header parsing for non-first fragments The tproxy, osf and exthdr (SCTP) expressions rely on the presence of transport layer headers to perform socket lookups, fingerprint matching, or chunk extraction. For fragmented packets, while the IP protocol remains constant across all fragments, only the first fragment contains the actual L4 header. The expressions could be attached to a chain with a priority lower than -400, bypassing defragmentation. Or could be used in stateless environments where defragmentation is not happening at all. This could result in garbage data being used for the matching. Add a check for pkt->fragoff so only unfragmented packets or the first fragment is processed. Fixes: 133dc203d77d ("netfilter: nft_exthdr: Support SCTP chunks") Fixes: 4ed8eb6570a4 ("netfilter: nf_tables: Add native tproxy support") Fixes: b96af92d6eaf ("netfilter: nf_tables: implement Passive OS fingerprint module in nft_osf") Signed-off-by: Fernando Fernandez Mancera Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_core.c | 2 +- net/netfilter/nft_exthdr.c | 2 +- net/netfilter/nft_osf.c | 2 +- net/netfilter/nft_tproxy.c | 8 ++++---- 4 files changed, 7 insertions(+), 7 deletions(-) diff --git a/net/netfilter/nf_tables_core.c b/net/netfilter/nf_tables_core.c index 5ddd5b6e135f..8ab186f86dd4 100644 --- a/net/netfilter/nf_tables_core.c +++ b/net/netfilter/nf_tables_core.c @@ -153,7 +153,7 @@ static bool nft_payload_fast_eval(const struct nft_expr *expr, if (priv->base == NFT_PAYLOAD_NETWORK_HEADER) ptr = skb_network_header(skb) + pkt->nhoff; else { - if (!(pkt->flags & NFT_PKTINFO_L4PROTO)) + if (!(pkt->flags & NFT_PKTINFO_L4PROTO) || pkt->fragoff) return false; ptr = skb->data + nft_thoff(pkt); } diff --git a/net/netfilter/nft_exthdr.c b/net/netfilter/nft_exthdr.c index 0407d6f708ae..e6a07c0df207 100644 --- a/net/netfilter/nft_exthdr.c +++ b/net/netfilter/nft_exthdr.c @@ -376,7 +376,7 @@ static void nft_exthdr_sctp_eval(const struct nft_expr *expr, const struct sctp_chunkhdr *sch; struct sctp_chunkhdr _sch; - if (pkt->tprot != IPPROTO_SCTP) + if (pkt->tprot != IPPROTO_SCTP || pkt->fragoff) goto err; do { diff --git a/net/netfilter/nft_osf.c b/net/netfilter/nft_osf.c index c02d5cb52143..45fe56da5044 100644 --- a/net/netfilter/nft_osf.c +++ b/net/netfilter/nft_osf.c @@ -33,7 +33,7 @@ static void nft_osf_eval(const struct nft_expr *expr, struct nft_regs *regs, return; } - if (pkt->tprot != IPPROTO_TCP) { + if (pkt->tprot != IPPROTO_TCP || pkt->fragoff) { regs->verdict.code = NFT_BREAK; return; } diff --git a/net/netfilter/nft_tproxy.c b/net/netfilter/nft_tproxy.c index f2101af8c867..89be443734f6 100644 --- a/net/netfilter/nft_tproxy.c +++ b/net/netfilter/nft_tproxy.c @@ -30,8 +30,8 @@ static void nft_tproxy_eval_v4(const struct nft_expr *expr, __be16 tport = 0; struct sock *sk; - if (pkt->tprot != IPPROTO_TCP && - pkt->tprot != IPPROTO_UDP) { + if ((pkt->tprot != IPPROTO_TCP && + pkt->tprot != IPPROTO_UDP) || pkt->fragoff) { regs->verdict.code = NFT_BREAK; return; } @@ -97,8 +97,8 @@ static void nft_tproxy_eval_v6(const struct nft_expr *expr, memset(&taddr, 0, sizeof(taddr)); - if (pkt->tprot != IPPROTO_TCP && - pkt->tprot != IPPROTO_UDP) { + if ((pkt->tprot != IPPROTO_TCP && + pkt->tprot != IPPROTO_UDP) || pkt->fragoff) { regs->verdict.code = NFT_BREAK; return; } From 952e121c96137c73bd3e59bb20a93ef659376947 Mon Sep 17 00:00:00 2001 From: Fernando Fernandez Mancera Date: Tue, 28 Apr 2026 12:25:48 +0200 Subject: [PATCH 113/525] netfilter: xtables: fix L4 header parsing for non-first fragments Multiple targets and matches relies on L4 header to operate. For fragmented packets, every fragment carries the transport protocol identifier, but only the first fragment contains the L4 header. As the 'raw' table can be configured to run at priority -450 (before defragmentation at -400), the target/match can be reached before reassembly. In this case, non-first fragments have their payload incorrectly parsed as a TCP/UDP header. This would be of course a misconfiguration scenario. In most of the cases this just lead to a unreliable behavior for fragmented traffic. Add a fragment check to ensure target/match only evaluates unfragmented packets or the first fragment in the stream. Fixes: 902d6a4c2a4f ("netfilter: nf_defrag: Skip defrag if NOTRACK is set") Signed-off-by: Fernando Fernandez Mancera Signed-off-by: Pablo Neira Ayuso --- net/netfilter/xt_TPROXY.c | 11 +++++++++-- net/netfilter/xt_ecn.c | 4 ++++ net/netfilter/xt_hashlimit.c | 4 +++- net/netfilter/xt_osf.c | 3 +++ net/netfilter/xt_tcpmss.c | 4 ++++ 5 files changed, 23 insertions(+), 3 deletions(-) diff --git a/net/netfilter/xt_TPROXY.c b/net/netfilter/xt_TPROXY.c index e4bea1d346cf..5f60e7298a1e 100644 --- a/net/netfilter/xt_TPROXY.c +++ b/net/netfilter/xt_TPROXY.c @@ -86,6 +86,9 @@ tproxy_tg4_v0(struct sk_buff *skb, const struct xt_action_param *par) { const struct xt_tproxy_target_info *tgi = par->targinfo; + if (par->fragoff) + return NF_DROP; + return tproxy_tg4(xt_net(par), skb, tgi->laddr, tgi->lport, tgi->mark_mask, tgi->mark_value); } @@ -95,6 +98,9 @@ tproxy_tg4_v1(struct sk_buff *skb, const struct xt_action_param *par) { const struct xt_tproxy_target_info_v1 *tgi = par->targinfo; + if (par->fragoff) + return NF_DROP; + return tproxy_tg4(xt_net(par), skb, tgi->laddr.ip, tgi->lport, tgi->mark_mask, tgi->mark_value); } @@ -106,6 +112,7 @@ tproxy_tg6_v1(struct sk_buff *skb, const struct xt_action_param *par) { const struct ipv6hdr *iph = ipv6_hdr(skb); const struct xt_tproxy_target_info_v1 *tgi = par->targinfo; + unsigned short fragoff = 0; struct udphdr _hdr, *hp; struct sock *sk; const struct in6_addr *laddr; @@ -113,8 +120,8 @@ tproxy_tg6_v1(struct sk_buff *skb, const struct xt_action_param *par) int thoff = 0; int tproto; - tproto = ipv6_find_hdr(skb, &thoff, -1, NULL, NULL); - if (tproto < 0) + tproto = ipv6_find_hdr(skb, &thoff, -1, &fragoff, NULL); + if (tproto < 0 || fragoff) return NF_DROP; hp = skb_header_pointer(skb, thoff, sizeof(_hdr), &_hdr); diff --git a/net/netfilter/xt_ecn.c b/net/netfilter/xt_ecn.c index b96e8203ac54..a8503f5d26bf 100644 --- a/net/netfilter/xt_ecn.c +++ b/net/netfilter/xt_ecn.c @@ -30,6 +30,10 @@ static bool match_tcp(const struct sk_buff *skb, struct xt_action_param *par) struct tcphdr _tcph; const struct tcphdr *th; + /* this is fine for IPv6 as ecn_mt_check6() enforces -p tcp */ + if (par->fragoff) + return false; + /* In practice, TCP match does this, so can't fail. But let's * be good citizens. */ diff --git a/net/netfilter/xt_hashlimit.c b/net/netfilter/xt_hashlimit.c index 3bd127bfc114..2704b4b60d1e 100644 --- a/net/netfilter/xt_hashlimit.c +++ b/net/netfilter/xt_hashlimit.c @@ -658,6 +658,8 @@ hashlimit_init_dst(const struct xt_hashlimit_htable *hinfo, if (!(hinfo->cfg.mode & (XT_HASHLIMIT_HASH_DPT | XT_HASHLIMIT_HASH_SPT))) return 0; + if (ntohs(ip_hdr(skb)->frag_off) & IP_OFFSET) + return -1; nexthdr = ip_hdr(skb)->protocol; break; #if IS_ENABLED(CONFIG_IP6_NF_IPTABLES) @@ -681,7 +683,7 @@ hashlimit_init_dst(const struct xt_hashlimit_htable *hinfo, return 0; nexthdr = ipv6_hdr(skb)->nexthdr; protoff = ipv6_skip_exthdr(skb, sizeof(struct ipv6hdr), &nexthdr, &frag_off); - if ((int)protoff < 0) + if ((int)protoff < 0 || ntohs(frag_off) & IP6_OFFSET) return -1; break; } diff --git a/net/netfilter/xt_osf.c b/net/netfilter/xt_osf.c index dc9485854002..e8807caede68 100644 --- a/net/netfilter/xt_osf.c +++ b/net/netfilter/xt_osf.c @@ -27,6 +27,9 @@ static bool xt_osf_match_packet(const struct sk_buff *skb, struct xt_action_param *p) { + if (p->fragoff) + return false; + return nf_osf_match(skb, xt_family(p), xt_hooknum(p), xt_in(p), xt_out(p), p->matchinfo, xt_net(p), nf_osf_fingers); } diff --git a/net/netfilter/xt_tcpmss.c b/net/netfilter/xt_tcpmss.c index 0d32d4841cb3..b9da8269161d 100644 --- a/net/netfilter/xt_tcpmss.c +++ b/net/netfilter/xt_tcpmss.c @@ -32,6 +32,10 @@ tcpmss_mt(const struct sk_buff *skb, struct xt_action_param *par) u8 _opt[15 * 4 - sizeof(_tcph)]; unsigned int i, optlen; + /* this is fine for IPv6 as xt_tcpmss enforces -p tcp */ + if (par->fragoff) + return false; + /* If we don't have the whole header, drop packet. */ th = skb_header_pointer(skb, par->thoff, sizeof(_tcph), &_tcph); if (th == NULL) From ef4f741e8627512cb8c82f59a1fc7aacd854aadf Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Thu, 30 Apr 2026 16:49:48 +0200 Subject: [PATCH 114/525] netfilter: flowtable: ensure sufficient headroom in xmit path Check for headroom and call skb_expand_head() like in the IP output path to ensure there is sufficient headroom for the mac header when forwarding this packet as suggested by sashiko. Fixes: b5964aac51e0 ("netfilter: flowtable: consolidate xmit path") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_flow_table_ip.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c index dbd7644fdbeb..8d5fb7e940a1 100644 --- a/net/netfilter/nf_flow_table_ip.c +++ b/net/netfilter/nf_flow_table_ip.c @@ -471,8 +471,17 @@ struct nf_flow_xmit { static unsigned int nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, struct nf_flow_xmit *xmit) { - skb->dev = xmit->outdev; - dev_hard_header(skb, skb->dev, ntohs(skb->protocol), + struct net_device *dev = xmit->outdev; + unsigned int hh_len = LL_RESERVED_SPACE(dev); + + if (unlikely(skb_headroom(skb) < hh_len && dev->header_ops)) { + skb = skb_expand_head(skb, hh_len); + if (!skb) + return NF_STOLEN; + } + + skb->dev = dev; + dev_hard_header(skb, dev, ntohs(skb->protocol), xmit->dest, xmit->source, skb->len); dev_queue_xmit(skb); From d6cc7c99bf1f73eda7d565d224d791d16239bb41 Mon Sep 17 00:00:00 2001 From: Sanman Pradhan Date: Thu, 16 Apr 2026 21:59:30 +0000 Subject: [PATCH 115/525] hwmon: (ltc2992) Clamp threshold writes to hardware range ltc2992_set_voltage(), ltc2992_set_current(), and ltc2992_set_power() do not validate the user-supplied value before converting it to a register value. This can result in: 1. Negative input values wrapping to large positive register values. For power, the negative long is implicitly cast to u64 in mul_u64_u32_div(), producing an incorrect value. For voltage and current, the negative converted value wraps when passed to ltc2992_write_reg() as a u32. 2. Intermediate arithmetic exceeding the range representable in u64 on 64-bit platforms. In ltc2992_set_voltage(), (u64)val * 1000 can exceed U64_MAX when val is a large positive long. In ltc2992_set_current(), (u64)val * r_sense_uohm can overflow similarly. In ltc2992_set_power(), the computed value may not fit in u64. 3. Register values exceeding the hardware field width. Voltage and current threshold registers are 12-bit (stored left-justified in 16 bits), and power threshold registers are 24-bit. Without clamping, bits above the field width are truncated in ltc2992_write_reg(). Fix by clamping negative values to zero, clamping positive values to the rounded hardware-representable maximum (the value returned by the read path for a full-scale register) to prevent intermediate overflow, and clamping the converted register value to the hardware field width before writing. The existing conversion formula and rounding behavior are preserved. In the power write path, cancel the factor of 1000 from both the numerator (r_sense_uohm * 1000) and the denominator (VADC_UV_LSB * IADC_NANOV_LSB) to also eliminate a u32 overflow of r_sense_uohm * 1000 when r_sense_uohm exceeds about 4.29 ohms. Fixes: b0bd407e94b03 ("hwmon: (ltc2992) Add support") Cc: stable@vger.kernel.org Signed-off-by: Sanman Pradhan Link: https://lore.kernel.org/r/20260416215904.101969-2-sanman.pradhan@hpe.com Signed-off-by: Guenter Roeck --- drivers/hwmon/ltc2992.c | 35 ++++++++++++++++++++++++++++------- 1 file changed, 28 insertions(+), 7 deletions(-) diff --git a/drivers/hwmon/ltc2992.c b/drivers/hwmon/ltc2992.c index 1fcd320d6161..106973619676 100644 --- a/drivers/hwmon/ltc2992.c +++ b/drivers/hwmon/ltc2992.c @@ -431,10 +431,16 @@ static int ltc2992_get_voltage(struct ltc2992_state *st, u32 reg, u32 scale, lon static int ltc2992_set_voltage(struct ltc2992_state *st, u32 reg, u32 scale, long val) { - val = DIV_ROUND_CLOSEST(val * 1000, scale); - val = val << 4; + u32 reg_val; + long vmax; - return ltc2992_write_reg(st, reg, 2, val); + vmax = DIV_ROUND_CLOSEST_ULL(0xFFFULL * scale, 1000); + val = max(val, 0L); + val = min(val, vmax); + reg_val = min(DIV_ROUND_CLOSEST_ULL((u64)val * 1000, scale), + 0xFFFULL) << 4; + + return ltc2992_write_reg(st, reg, 2, reg_val); } static int ltc2992_read_gpio_alarm(struct ltc2992_state *st, int nr_gpio, u32 attr, long *val) @@ -559,9 +565,15 @@ static int ltc2992_get_current(struct ltc2992_state *st, u32 reg, u32 channel, l static int ltc2992_set_current(struct ltc2992_state *st, u32 reg, u32 channel, long val) { u32 reg_val; + long cmax; - reg_val = DIV_ROUND_CLOSEST(val * st->r_sense_uohm[channel], LTC2992_IADC_NANOV_LSB); - reg_val = reg_val << 4; + cmax = DIV_ROUND_CLOSEST_ULL(0xFFFULL * LTC2992_IADC_NANOV_LSB, + st->r_sense_uohm[channel]); + val = max(val, 0L); + val = min(val, cmax); + reg_val = min(DIV_ROUND_CLOSEST_ULL((u64)val * st->r_sense_uohm[channel], + LTC2992_IADC_NANOV_LSB), + 0xFFFULL) << 4; return ltc2992_write_reg(st, reg, 2, reg_val); } @@ -634,9 +646,18 @@ static int ltc2992_get_power(struct ltc2992_state *st, u32 reg, u32 channel, lon static int ltc2992_set_power(struct ltc2992_state *st, u32 reg, u32 channel, long val) { u32 reg_val; + u64 pmax, uval; - reg_val = mul_u64_u32_div(val, st->r_sense_uohm[channel] * 1000, - LTC2992_VADC_UV_LSB * LTC2992_IADC_NANOV_LSB); + uval = max(val, 0L); + pmax = mul_u64_u32_div(0xFFFFFFULL, + LTC2992_VADC_UV_LSB / 1000 * + LTC2992_IADC_NANOV_LSB, + st->r_sense_uohm[channel]); + uval = min(uval, pmax); + reg_val = min(mul_u64_u32_div(uval, st->r_sense_uohm[channel], + LTC2992_VADC_UV_LSB / 1000 * + LTC2992_IADC_NANOV_LSB), + 0xFFFFFFULL); return ltc2992_write_reg(st, reg, 3, reg_val); } From 2da0c1fd01dbd6b22844e8676585153dfc660cbe Mon Sep 17 00:00:00 2001 From: Sanman Pradhan Date: Thu, 16 Apr 2026 21:59:40 +0000 Subject: [PATCH 116/525] hwmon: (ltc2992) Fix u32 overflow in power read path ltc2992_get_power() computes the divisor for mul_u64_u32_div() as r_sense_uohm * 1000. This multiplication overflows u32 when r_sense_uohm exceeds about 4.29 ohms (4294967 micro-ohms), producing a truncated divisor and an incorrect power reading. Cancel the factor of 1000 from both the numerator (VADC_UV_LSB * IADC_NANOV_LSB = 312500000) and the divisor (r_sense_uohm * 1000), giving (VADC_UV_LSB / 1000) * IADC_NANOV_LSB = 312500 as the numerator and plain r_sense_uohm as the divisor. The cancellation is exact because LTC2992_VADC_UV_LSB (25000) is divisible by 1000. This is the read-path counterpart of the write-path fix applied in the preceding patch. Fixes: b0bd407e94b03 ("hwmon: (ltc2992) Add support") Cc: stable@vger.kernel.org Signed-off-by: Sanman Pradhan Link: https://lore.kernel.org/r/20260416215904.101969-3-sanman.pradhan@hpe.com Signed-off-by: Guenter Roeck --- drivers/hwmon/ltc2992.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/hwmon/ltc2992.c b/drivers/hwmon/ltc2992.c index 106973619676..2617c4538af9 100644 --- a/drivers/hwmon/ltc2992.c +++ b/drivers/hwmon/ltc2992.c @@ -637,8 +637,10 @@ static int ltc2992_get_power(struct ltc2992_state *st, u32 reg, u32 channel, lon if (reg_val < 0) return reg_val; - *val = mul_u64_u32_div(reg_val, LTC2992_VADC_UV_LSB * LTC2992_IADC_NANOV_LSB, - st->r_sense_uohm[channel] * 1000); + *val = mul_u64_u32_div(reg_val, + LTC2992_VADC_UV_LSB / 1000 * + LTC2992_IADC_NANOV_LSB, + st->r_sense_uohm[channel]); return 0; } From 5ed26ffe57ffca054b7a141ff0c1a07bf3a80f6c Mon Sep 17 00:00:00 2001 From: Ninad Naik Date: Sat, 18 Apr 2026 00:44:11 +0530 Subject: [PATCH 117/525] Documentation: hwmon: fix link to ideapad-laptop.c file The ideapad-laptop.c file now exists inside drivers/platform/x86/lenovo/ directory. Updating the GitHub link to the correct path. Signed-off-by: Ninad Naik Link: https://lore.kernel.org/r/20260417191411.713958-1-ninadnaik07@gmail.com Signed-off-by: Guenter Roeck --- Documentation/hwmon/yogafan.rst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/hwmon/yogafan.rst b/Documentation/hwmon/yogafan.rst index c553a381f772..68761947a1a8 100644 --- a/Documentation/hwmon/yogafan.rst +++ b/Documentation/hwmon/yogafan.rst @@ -135,4 +135,4 @@ References 4. **Lenovo IdeaPad Laptop Driver:** Reference for DMI-based hardware feature gating in Lenovo laptops. - https://github.com/torvalds/linux/blob/master/drivers/platform/x86/ideapad-laptop.c + https://github.com/torvalds/linux/blob/master/drivers/platform/x86/lenovo/ideapad-laptop.c From 1a1414c675ee1b637bbe3840241555a49c61b123 Mon Sep 17 00:00:00 2001 From: Sasha Levin Date: Sat, 25 Apr 2026 20:03:19 -0400 Subject: [PATCH 118/525] hwmon: Remove stale CONFIG_SENSORS_SBRMI Makefile reference MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit kconfiglint reports: X001: CONFIG_SENSORS_SBRMI referenced in Makefile but not defined in any Kconfig The SB-RMI hardware monitoring driver was originally introduced in commit 5a0f50d110b3 ("hwmon: Add support for SB-RMI power module") with both a Kconfig entry (CONFIG_SENSORS_SBRMI) and a Makefile line (obj-$(CONFIG_SENSORS_SBRMI) += sbrmi.o) in drivers/hwmon/. Commit e156586764050 ("hwmon/misc: amd-sbi: Move core sbrmi from hwmon to misc") moved the driver to drivers/misc/amd-sbi/ to support additional functionality beyond hardware monitoring. That commit correctly removed the Kconfig entry from drivers/hwmon/Kconfig, moved the source file drivers/hwmon/sbrmi.c to drivers/misc/amd-sbi/sbrmi.c, and created new Kconfig/Makefile entries in drivers/misc/amd-sbi/ with a renamed symbol (CONFIG_AMD_SBRMI_I2C). However, the Makefile line in drivers/hwmon/Makefile was not removed in that commit. The orphaned line references a CONFIG symbol that no longer exists and a source file that is no longer present, so it has no effect on the build — but it is dead code that should be cleaned up. Remove the stale Makefile reference. Assisted-by: Claude:claude-opus-4-6 kconfiglint Signed-off-by: Sasha Levin Link: https://lore.kernel.org/r/20260426000319.55908-1-sashal@kernel.org Signed-off-by: Guenter Roeck --- drivers/hwmon/Makefile | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile index 4788996aa137..982ee2c6f9de 100644 --- a/drivers/hwmon/Makefile +++ b/drivers/hwmon/Makefile @@ -201,7 +201,6 @@ obj-$(CONFIG_SENSORS_PWM_FAN) += pwm-fan.o obj-$(CONFIG_SENSORS_QNAP_MCU_HWMON) += qnap-mcu-hwmon.o obj-$(CONFIG_SENSORS_RASPBERRYPI_HWMON) += raspberrypi-hwmon.o obj-$(CONFIG_SENSORS_SBTSI) += sbtsi_temp.o -obj-$(CONFIG_SENSORS_SBRMI) += sbrmi.o obj-$(CONFIG_SENSORS_SCH56XX_COMMON)+= sch56xx-common.o obj-$(CONFIG_SENSORS_SCH5627) += sch5627.o obj-$(CONFIG_SENSORS_SCH5636) += sch5636.o From 174606451fbb17db506ebaacdd5e203e57773d5f Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Fri, 24 Apr 2026 22:50:51 +0900 Subject: [PATCH 119/525] hwmon: (corsair-psu) Close HID device on probe errors corsairpsu_probe() opens the HID device before sending the device init and firmware-info commands. If either command fails, the error path jumps directly to fail_and_stop and skips hid_hw_close(). Use the existing fail_and_close label for those post-open failures so the open count and low-level close callback are balanced before hid_hw_stop(). Fixes: d115b51e0e56 ("hwmon: add Corsair PSU HID controller driver") Cc: stable@vger.kernel.org Signed-off-by: Myeonghun Pak Reviewed-by: Wilken Gottwalt Link: https://lore.kernel.org/r/20260424135107.13720-1-mhun512@gmail.com Signed-off-by: Guenter Roeck --- drivers/hwmon/corsair-psu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/hwmon/corsair-psu.c b/drivers/hwmon/corsair-psu.c index dddbd2463f8d..76f3e1da68d0 100644 --- a/drivers/hwmon/corsair-psu.c +++ b/drivers/hwmon/corsair-psu.c @@ -796,13 +796,13 @@ static int corsairpsu_probe(struct hid_device *hdev, const struct hid_device_id ret = corsairpsu_init(priv); if (ret < 0) { dev_err(&hdev->dev, "unable to initialize device (%d)\n", ret); - goto fail_and_stop; + goto fail_and_close; } ret = corsairpsu_fwinfo(priv); if (ret < 0) { dev_err(&hdev->dev, "unable to query firmware (%d)\n", ret); - goto fail_and_stop; + goto fail_and_close; } corsairpsu_get_criticals(priv); From 18ed60e33e6c77d62409c1343dec1c61bae3d2e7 Mon Sep 17 00:00:00 2001 From: Jeremy Kerr Date: Wed, 29 Apr 2026 16:21:41 +0800 Subject: [PATCH 120/525] net: mctp: test: use a zeroed struct sockaddr_mctp Invalid sockaddr padding will cause bind() to fail; ensure we have a zeroed address in the testcase. Fixes: 0d8647bc74cb ("net: mctp: don't require a route for null-EID ingress") Signed-off-by: Jeremy Kerr Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260429-dev-mctp-test-fixes-v1-1-1127b7425809@codeconstruct.com.au Signed-off-by: Jakub Kicinski --- net/mctp/test/route-test.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mctp/test/route-test.c b/net/mctp/test/route-test.c index e1033643fab0..e4b230ef6099 100644 --- a/net/mctp/test/route-test.c +++ b/net/mctp/test/route-test.c @@ -920,9 +920,9 @@ static void mctp_test_route_input_cloned_frag(struct kunit *test) static void mctp_test_route_input_null_eid(struct kunit *test) { struct mctp_hdr hdr = RX_HDR(1, 10, 0, FL_S | FL_E | FL_TO); + struct sockaddr_mctp addr = { 0 }; struct sk_buff *skb_pkt, *skb_sk; struct mctp_test_dev *dev; - struct sockaddr_mctp addr; struct socket *sock; u8 type = 0; int rc; From 76872971064133474d9b891da05db8f7586fcc11 Mon Sep 17 00:00:00 2001 From: Jeremy Kerr Date: Wed, 29 Apr 2026 16:21:42 +0800 Subject: [PATCH 121/525] net: mctp: test: Use dev_direct_xmit for TX to our test device In our test cases, we typically feed a packet sequence into the routing code, then inspect the device's TXed skbs to assert specific behaviours. Using dev_queue_xmit() for our TX path introduces a fair bit of complexity between the test packet sequence and the test device's ndo_start_xmit callback; which may mean that the skbs have not hit the device at the point we're inspecting the TXed skb list. Use dev_direct_xmit instead, as we want a direct a path as possible here, and the test dev does not need any queueing, scheduling or flow control. Fixes: 6ab578739a4c ("net: mctp: test: move TX packetqueue from dst to dev") Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-lkp/202604281320.525eee17-lkp@intel.com Signed-off-by: Jeremy Kerr Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260429-dev-mctp-test-fixes-v1-2-1127b7425809@codeconstruct.com.au Signed-off-by: Jakub Kicinski --- net/mctp/test/utils.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mctp/test/utils.c b/net/mctp/test/utils.c index c3987d5ade7a..6eef8d485c25 100644 --- a/net/mctp/test/utils.c +++ b/net/mctp/test/utils.c @@ -116,7 +116,7 @@ void mctp_test_destroy_dev(struct mctp_test_dev *dev) static int mctp_test_dst_output(struct mctp_dst *dst, struct sk_buff *skb) { skb->dev = dst->dev->dev; - dev_queue_xmit(skb); + dev_direct_xmit(skb, 0); return 0; } From ba6b328588f7cb7cf4aca33bae565c2914d74786 Mon Sep 17 00:00:00 2001 From: David Gow Date: Sat, 25 Apr 2026 11:41:23 +0800 Subject: [PATCH 122/525] rust: arch: um: Fix building 32-bit UML with GCC 32-bit UML builds can be configured either by setting CONFIG_64BIT=n or with SUBARCH=i386. Both work with Rust-for-Linux when clang is the compiler, but when SUBARCH=i386, we don't set a bindgen target correctly if gcc is the compiler. Add the appropriate bindgen target configuration for i386, as is done in Makefile.clang. [ For reference, the errors look like: BINDGEN rust/bindings/bindings_generated.rs error: unsupported option '-mno-sse' for target '' ... error: unknown target triple 'unknown' panicked at .../bindgen-0.72.1/ir/context.rs:562:15: libclang error; possible causes include: ... - Miguel ] Fixes: ab0f4cedc355 ("arch: um: rust: Add i386 support for Rust") Signed-off-by: David Gow Link: https://patch.msgid.link/20260425034125.53866-1-david@davidgow.net [ Added space in title. - Miguel ] Signed-off-by: Miguel Ojeda --- rust/Makefile | 2 ++ 1 file changed, 2 insertions(+) diff --git a/rust/Makefile b/rust/Makefile index b361bfedfdf0..b9e9f512cec3 100644 --- a/rust/Makefile +++ b/rust/Makefile @@ -403,6 +403,8 @@ BINDGEN_TARGET_x86 := x86_64-linux-gnu BINDGEN_TARGET_arm64 := aarch64-linux-gnu BINDGEN_TARGET_arm := arm-linux-gnueabi BINDGEN_TARGET_loongarch := loongarch64-linux-gnusf +# This is only for i386 UM builds, which need the 32-bit target not -m32 +BINDGEN_TARGET_i386 := i386-linux-gnu BINDGEN_TARGET_um := $(BINDGEN_TARGET_$(SUBARCH)) BINDGEN_TARGET := $(BINDGEN_TARGET_$(SRCARCH)) From 83ac2870310b694775ab7e8f0244fdd94fc21926 Mon Sep 17 00:00:00 2001 From: Gary Guo Date: Mon, 27 Apr 2026 16:43:00 +0100 Subject: [PATCH 123/525] rust: pin-init: internal: move alignment check to `make_field_check` Instead of having the reference creation serving dual-purpose as both for let bindings and alignment check, detangle them so that the alignment check is done explicitly in `make_field_check`. This is more robust against refactors that may change the way let bindings are created. Cc: stable@vger.kernel.org Reviewed-by: Alice Ryhl Signed-off-by: Gary Guo Link: https://patch.msgid.link/20260427-pin-init-fix-v3-1-496a699674dd@garyguo.net [ Reworded for typo. - Miguel ] Signed-off-by: Miguel Ojeda --- rust/pin-init/internal/src/init.rs | 78 ++++++++++++++---------------- 1 file changed, 37 insertions(+), 41 deletions(-) diff --git a/rust/pin-init/internal/src/init.rs b/rust/pin-init/internal/src/init.rs index daa3f1c6466e..0a6600e8156c 100644 --- a/rust/pin-init/internal/src/init.rs +++ b/rust/pin-init/internal/src/init.rs @@ -249,10 +249,6 @@ fn init_fields( }); // Again span for better diagnostics let write = quote_spanned!(ident.span()=> ::core::ptr::write); - // NOTE: the field accessor ensures that the initialized field is properly aligned. - // Unaligned fields will cause the compiler to emit E0793. We do not support - // unaligned fields since `Init::__init` requires an aligned pointer; the call to - // `ptr::write` below has the same requirement. let accessor = if pinned { let project_ident = format_ident!("__project_{ident}"); quote! { @@ -367,49 +363,49 @@ fn init_fields( } } -/// Generate the check for ensuring that every field has been initialized. +/// Generate the check for ensuring that every field has been initialized and aligned. fn make_field_check( fields: &Punctuated, init_kind: InitKind, path: &Path, ) -> TokenStream { - let field_attrs = fields + let field_attrs: Vec<_> = fields .iter() - .filter_map(|f| f.kind.ident().map(|_| &f.attrs)); - let field_name = fields.iter().filter_map(|f| f.kind.ident()); - match init_kind { - InitKind::Normal => quote! { - // We use unreachable code to ensure that all fields have been mentioned exactly once, - // this struct initializer will still be type-checked and complain with a very natural - // error message if a field is forgotten/mentioned more than once. - #[allow(unreachable_code, clippy::diverging_sub_expression)] - // SAFETY: this code is never executed. - let _ = || unsafe { - ::core::ptr::write(slot, #path { - #( - #(#field_attrs)* - #field_name: ::core::panic!(), - )* - }) - }; - }, - InitKind::Zeroing => quote! { - // We use unreachable code to ensure that all fields have been mentioned at most once. - // Since the user specified `..Zeroable::zeroed()` at the end, all missing fields will - // be zeroed. This struct initializer will still be type-checked and complain with a - // very natural error message if a field is mentioned more than once, or doesn't exist. - #[allow(unreachable_code, clippy::diverging_sub_expression, unused_assignments)] - // SAFETY: this code is never executed. - let _ = || unsafe { - ::core::ptr::write(slot, #path { - #( - #(#field_attrs)* - #field_name: ::core::panic!(), - )* - ..::core::mem::zeroed() - }) - }; - }, + .filter_map(|f| f.kind.ident().map(|_| &f.attrs)) + .collect(); + let field_name: Vec<_> = fields.iter().filter_map(|f| f.kind.ident()).collect(); + let zeroing_trailer = match init_kind { + InitKind::Normal => None, + InitKind::Zeroing => Some(quote! { + ..::core::mem::zeroed() + }), + }; + quote! { + #[allow(unreachable_code, clippy::diverging_sub_expression)] + // We use unreachable code to perform field checks. They're still checked by the compiler. + // SAFETY: this code is never executed. + let _ = || unsafe { + // Create references to ensure that the initialized field is properly aligned. + // Unaligned fields will cause the compiler to emit E0793. We do not support + // unaligned fields since `Init::__init` requires an aligned pointer; the call to + // `ptr::write` for value-initialization case has the same requirement. + #( + #(#field_attrs)* + let _ = &(*slot).#field_name; + )* + + // If the zeroing trailer is not present, this checks that all fields have been + // mentioned exactly once. If the zeroing trailer is present, all missing fields will be + // zeroed, so this checks that all fields have been mentioned at most once. The use of + // struct initializer will still generate very natural error messages for any misuse. + ::core::ptr::write(slot, #path { + #( + #(#field_attrs)* + #field_name: ::core::panic!(), + )* + #zeroing_trailer + }) + }; } } From 68bf102226cf2199dc609b67c1e847cad4de4b57 Mon Sep 17 00:00:00 2001 From: Gary Guo Date: Mon, 27 Apr 2026 16:43:01 +0100 Subject: [PATCH 124/525] rust: pin-init: fix incorrect accessor reference lifetime When a field has been initialized, `init!`/`pin_init!` create a reference or pinned reference to the field so it can be accessed later during the initialization of other fields. However, the reference it created is incorrectly `&'static` rather than just the scope of the initializer. This means that you can do init!(Foo { a: 1, _: { let b: &'static u32 = a; } }) which is unsound. This is caused by `&mut (*#slot).#ident`, which actually allows arbitrary lifetime, so this is effectively `'static`. Somewhat ironically, the safety justification of creating the accessor is.. "SAFETY: TODO". Fix it by adding `let_binding` method on `DropGuard` to shorten lifetime. This results in exactly what we want for these accessors. The safety and invariant comments of `DropGuard` have been reworked; instead of reasoning about what caller can do with the guard, express it in a way that the ownership is transferred to the guard and `forget` takes it back, so the unsafe operations within the `DropGuard` can be more easily justified. Fixes: 42415d163e5d ("rust: pin-init: add references to previously initialized fields") Cc: stable@vger.kernel.org Signed-off-by: Gary Guo Link: https://patch.msgid.link/20260427-pin-init-fix-v3-2-496a699674dd@garyguo.net [ Reworded for missing word. - Miguel ] Signed-off-by: Miguel Ojeda --- rust/pin-init/internal/src/init.rs | 106 +++++++++++++---------------- rust/pin-init/src/__internal.rs | 28 +++++--- 2 files changed, 66 insertions(+), 68 deletions(-) diff --git a/rust/pin-init/internal/src/init.rs b/rust/pin-init/internal/src/init.rs index 0a6600e8156c..487ee0013faf 100644 --- a/rust/pin-init/internal/src/init.rs +++ b/rust/pin-init/internal/src/init.rs @@ -249,18 +249,6 @@ fn init_fields( }); // Again span for better diagnostics let write = quote_spanned!(ident.span()=> ::core::ptr::write); - let accessor = if pinned { - let project_ident = format_ident!("__project_{ident}"); - quote! { - // SAFETY: TODO - unsafe { #data.#project_ident(&mut (*#slot).#ident) } - } - } else { - quote! { - // SAFETY: TODO - unsafe { &mut (*#slot).#ident } - } - }; quote! { #(#attrs)* { @@ -268,51 +256,31 @@ fn init_fields( // SAFETY: TODO unsafe { #write(&raw mut (*#slot).#ident, #value_ident) }; } - #(#cfgs)* - #[allow(unused_variables)] - let #ident = #accessor; } } InitializerKind::Init { ident, value, .. } => { // Again span for better diagnostics let init = format_ident!("init", span = value.span()); - // NOTE: the field accessor ensures that the initialized field is properly aligned. - // Unaligned fields will cause the compiler to emit E0793. We do not support - // unaligned fields since `Init::__init` requires an aligned pointer; the call to - // `ptr::write` below has the same requirement. - let (value_init, accessor) = if pinned { - let project_ident = format_ident!("__project_{ident}"); - ( - quote! { - // SAFETY: - // - `slot` is valid, because we are inside of an initializer closure, we - // return when an error/panic occurs. - // - We also use `#data` to require the correct trait (`Init` or `PinInit`) - // for `#ident`. - unsafe { #data.#ident(&raw mut (*#slot).#ident, #init)? }; - }, - quote! { - // SAFETY: TODO - unsafe { #data.#project_ident(&mut (*#slot).#ident) } - }, - ) + let value_init = if pinned { + quote! { + // SAFETY: + // - `slot` is valid, because we are inside of an initializer closure, we + // return when an error/panic occurs. + // - We also use `#data` to require the correct trait (`Init` or `PinInit`) + // for `#ident`. + unsafe { #data.#ident(&raw mut (*#slot).#ident, #init)? }; + } } else { - ( - quote! { - // SAFETY: `slot` is valid, because we are inside of an initializer - // closure, we return when an error/panic occurs. - unsafe { - ::pin_init::Init::__init( - #init, - &raw mut (*#slot).#ident, - )? - }; - }, - quote! { - // SAFETY: TODO - unsafe { &mut (*#slot).#ident } - }, - ) + quote! { + // SAFETY: `slot` is valid, because we are inside of an initializer + // closure, we return when an error/panic occurs. + unsafe { + ::pin_init::Init::__init( + #init, + &raw mut (*#slot).#ident, + )? + }; + } }; quote! { #(#attrs)* @@ -320,9 +288,6 @@ fn init_fields( let #init = #value; #value_init } - #(#cfgs)* - #[allow(unused_variables)] - let #ident = #accessor; } } InitializerKind::Code { block: value, .. } => quote! { @@ -335,18 +300,41 @@ fn init_fields( if let Some(ident) = kind.ident() { // `mixed_site` ensures that the guard is not accessible to the user-controlled code. let guard = format_ident!("__{ident}_guard", span = Span::mixed_site()); + + // NOTE: The reference is derived from the guard so that it only lives as long as the + // guard does and cannot escape the scope. If it's created via `&mut (*#slot).#ident` + // like the unaligned field guard, it will become effectively `'static`. + let accessor = if pinned { + let project_ident = format_ident!("__project_{ident}"); + quote! { + // SAFETY: the initialization is pinned. + unsafe { #data.#project_ident(#guard.let_binding()) } + } + } else { + quote! { + #guard.let_binding() + } + }; + res.extend(quote! { #(#cfgs)* - // Create the drop guard: + // Create the drop guard. // - // We rely on macro hygiene to make it impossible for users to access this local - // variable. - // SAFETY: We forget the guard later when initialization has succeeded. - let #guard = unsafe { + // SAFETY: + // - `&raw mut (*slot).#ident` is valid. + // - `make_field_check` checks that `&raw mut (*slot).#ident` is properly aligned. + // - `(*slot).#ident` has been initialized above. + // - We only need the ownership to the pointee back when initialization has + // succeeded, where we `forget` the guard. + let mut #guard = unsafe { ::pin_init::__internal::DropGuard::new( &raw mut (*slot).#ident ) }; + + #(#cfgs)* + #[allow(unused_variables)] + let #ident = #accessor; }); guards.push(guard); guard_attrs.push(cfgs); diff --git a/rust/pin-init/src/__internal.rs b/rust/pin-init/src/__internal.rs index 90adbdc1893b..5720a621aed7 100644 --- a/rust/pin-init/src/__internal.rs +++ b/rust/pin-init/src/__internal.rs @@ -238,32 +238,42 @@ struct Foo { /// When a value of this type is dropped, it drops a `T`. /// /// Can be forgotten to prevent the drop. +/// +/// # Invariants +/// +/// - `ptr` is valid and properly aligned. +/// - `*ptr` is initialized and owned by this guard. pub struct DropGuard { ptr: *mut T, } impl DropGuard { - /// Creates a new [`DropGuard`]. It will [`ptr::drop_in_place`] `ptr` when it gets dropped. + /// Creates a drop guard and transfer the ownership of the pointer content. + /// + /// The ownership is only relinguished if the guard is forgotten via [`core::mem::forget`]. /// /// # Safety /// - /// `ptr` must be a valid pointer. - /// - /// It is the callers responsibility that `self` will only get dropped if the pointee of `ptr`: - /// - has not been dropped, - /// - is not accessible by any other means, - /// - will not be dropped by any other means. + /// - `ptr` is valid and properly aligned. + /// - `*ptr` is initialized, and the ownership is transferred to this guard. #[inline] pub unsafe fn new(ptr: *mut T) -> Self { + // INVARIANT: By safety requirement. Self { ptr } } + + /// Create a let binding for accessor use. + #[inline] + pub fn let_binding(&mut self) -> &mut T { + // SAFETY: Per type invariant. + unsafe { &mut *self.ptr } + } } impl Drop for DropGuard { #[inline] fn drop(&mut self) { - // SAFETY: A `DropGuard` can only be constructed using the unsafe `new` function - // ensuring that this operation is safe. + // SAFETY: `self.ptr` is valid, properly aligned and `*self.ptr` is owned by this guard. unsafe { ptr::drop_in_place(self.ptr) } } } From 838d852da8503372f3a1779bfbd1ccb93153ab4e Mon Sep 17 00:00:00 2001 From: Miguel Ojeda Date: Sun, 26 Apr 2026 16:42:00 +0200 Subject: [PATCH 125/525] rust: allow `clippy::collapsible_match` globally The `clippy::collapsible_match` lint [1] can make code harder to read in certain cases [2], e.g. CLIPPY P rust/libmacros.so - due to command line change warning: this `if` can be collapsed into the outer `match` --> rust/pin-init/internal/src/helpers.rs:91:17 | 91 | / if nesting == 1 { 92 | | impl_generics.push(tt.clone()); 93 | | impl_generics.push(tt); 94 | | skip_until_comma = false; 95 | | } | |_________________^ | = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#collapsible_match = note: `-W clippy::collapsible-match` implied by `-W clippy::all` = help: to override `-W clippy::all` add `#[allow(clippy::collapsible_match)]` help: collapse nested if block | 90 ~ TokenTree::Punct(p) if skip_until_comma && p.as_char() == ',' 91 ~ && nesting == 1 => { 92 | impl_generics.push(tt.clone()); 93 | impl_generics.push(tt); 94 | skip_until_comma = false; 95 ~ } | The lint does not have much upside -- when the suggestion may be a good one, it would still read fine when nested anyway. And it is the kind of lint that may easily bias people to just apply the suggestion instead of allowing it. [ In addition, as Gary points out [3], the suggestion is also wrong [4] and in the process of being fixed [5], possibly for Rust 1.97.0: Link: https://lore.kernel.org/rust-for-linux/DI3YV94TH9I3.1SOHW51552497@garyguo.net/ [3] Link: https://github.com/rust-lang/rust-clippy/issues/16875 [4] Link: https://github.com/rust-lang/rust-clippy/pull/16878 [5] - Miguel ] Thus just let developers decide on their own. Cc: stable@vger.kernel.org # Needed in 6.12.y and later (Rust is pinned in older LTSs). Link: https://rust-lang.github.io/rust-clippy/master/index.html#collapsible_match [1] Link: https://lore.kernel.org/rust-for-linux/CANiq72nWYJna_hdFxjQCQZK6yJBrr1Mb86iKavivV0U0BgufeA@mail.gmail.com/ [2] Reviewed-by: Gary Guo Link: https://patch.msgid.link/20260426144201.227108-1-ojeda@kernel.org Signed-off-by: Miguel Ojeda --- Makefile | 1 + 1 file changed, 1 insertion(+) diff --git a/Makefile b/Makefile index e27c91ea56fc..621d84aa4700 100644 --- a/Makefile +++ b/Makefile @@ -486,6 +486,7 @@ export rust_common_flags := --edition=2021 \ -Wclippy::as_ptr_cast_mut \ -Wclippy::as_underscore \ -Wclippy::cast_lossless \ + -Aclippy::collapsible_match \ -Wclippy::ignored_unit_patterns \ -Aclippy::incompatible_msrv \ -Wclippy::mut_mut \ From 2adc8664018c1cc595c7c0c98474a33c7fe32a85 Mon Sep 17 00:00:00 2001 From: Miguel Ojeda Date: Sun, 26 Apr 2026 16:42:01 +0200 Subject: [PATCH 126/525] rust: allow `clippy::collapsible_if` globally Similar to `clippy::collapsible_match` (globally allowed in the previous commit), the `clippy::collapsible_if` lint [1] can make code harder to read in certain cases. Thus just let developers decide on their own. In addition, remove the existing `expect` we had. Cc: stable@vger.kernel.org # Needed in 6.12.y and later (Rust is pinned in older LTSs). Suggested-by: Gary Guo Link: https://lore.kernel.org/rust-for-linux/DGROP5CHU1QZ.1OKJRAUZXE9WC@garyguo.net/ Link: https://rust-lang.github.io/rust-clippy/master/index.html#collapsible_if [1] Reviewed-by: Gary Guo Link: https://patch.msgid.link/20260426144201.227108-2-ojeda@kernel.org Signed-off-by: Miguel Ojeda --- Makefile | 1 + drivers/android/binder/range_alloc/array.rs | 1 - 2 files changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index 621d84aa4700..28f4ae452441 100644 --- a/Makefile +++ b/Makefile @@ -486,6 +486,7 @@ export rust_common_flags := --edition=2021 \ -Wclippy::as_ptr_cast_mut \ -Wclippy::as_underscore \ -Wclippy::cast_lossless \ + -Aclippy::collapsible_if \ -Aclippy::collapsible_match \ -Wclippy::ignored_unit_patterns \ -Aclippy::incompatible_msrv \ diff --git a/drivers/android/binder/range_alloc/array.rs b/drivers/android/binder/range_alloc/array.rs index ada1d1b4302e..081d19b09d4b 100644 --- a/drivers/android/binder/range_alloc/array.rs +++ b/drivers/android/binder/range_alloc/array.rs @@ -204,7 +204,6 @@ pub(crate) fn reservation_abort(&mut self, offset: usize) -> Result // caller will mark them as unused, which means that they can be freed if the system comes // under memory pressure. let mut freed_range = FreedRange::interior_pages(offset, size); - #[expect(clippy::collapsible_if)] // reads better like this if offset % PAGE_SIZE != 0 { if i == 0 || self.ranges[i - 1].endpoint() <= (offset & PAGE_MASK) { freed_range.start_page_idx -= 1; From dff205550e714f1cb65f27409586e2612905fa88 Mon Sep 17 00:00:00 2001 From: Gui-Dong Han Date: Thu, 16 Apr 2026 21:57:03 +0800 Subject: [PATCH 127/525] hwmon: (lm63) Add locking to avoid TOCTOU The functions show_fan(), show_pwm1(), show_temp11(), temp2_crit_hyst_show(), and show_lut_temp_hyst() access shared cached data without holding the update lock. This can cause TOCTOU races if the cached values change between the checks and the later calculations. Those cached values are updated in lm63_update_device(). In the general case, the affected functions combine multiple cached values without locking and can therefore observe a mixed old/new snapshot. In addition, show_fan() reads data->fan[nr] locklessly while lm63_update_device() updates data->fan[0] in two steps, which can expose an intermediate torn value and potentially trigger a divide-by-zero error. This means that converting the macro to a function is not sufficient to fix show_fan(). Hold the update lock across the whole read and calculation sequence so that the values remain stable. Check the other functions in the driver as well. Keep them unchanged because they either do not access shared cached values multiple times or already do so under lock. Link: https://lore.kernel.org/linux-hwmon/CALbr=LYJ_ehtp53HXEVkSpYoub+XYSTU8Rg=o1xxMJ8=5z8B-g@mail.gmail.com/ Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Fixes: e872c91e726e ("hwmon: (lm63) Add support for unsigned upper temperature limits") Fixes: d216f6809eb6 ("hwmon: (lm63) Expose automatic fan speed control lookup table") Signed-off-by: Gui-Dong Han Link: https://lore.kernel.org/r/20260416135703.53262-1-hanguidong02@gmail.com [groeck: Use lm63_update_device() to get driver data in temp2_crit_hyst_store] Signed-off-by: Guenter Roeck --- drivers/hwmon/lm63.c | 39 ++++++++++++++++++++++++++++++--------- 1 file changed, 30 insertions(+), 9 deletions(-) diff --git a/drivers/hwmon/lm63.c b/drivers/hwmon/lm63.c index 035176a98ce9..30500b4d2221 100644 --- a/drivers/hwmon/lm63.c +++ b/drivers/hwmon/lm63.c @@ -333,7 +333,13 @@ static ssize_t show_fan(struct device *dev, struct device_attribute *devattr, { struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr); struct lm63_data *data = lm63_update_device(dev); - return sprintf(buf, "%d\n", FAN_FROM_REG(data->fan[attr->index])); + int fan; + + mutex_lock(&data->update_lock); + fan = FAN_FROM_REG(data->fan[attr->index]); + mutex_unlock(&data->update_lock); + + return sprintf(buf, "%d\n", fan); } static ssize_t set_fan(struct device *dev, struct device_attribute *dummy, @@ -366,12 +372,14 @@ static ssize_t show_pwm1(struct device *dev, struct device_attribute *devattr, int nr = attr->index; int pwm; + mutex_lock(&data->update_lock); if (data->pwm_highres) pwm = data->pwm1[nr]; else pwm = data->pwm1[nr] >= 2 * data->pwm1_freq ? 255 : (data->pwm1[nr] * 255 + data->pwm1_freq) / (2 * data->pwm1_freq); + mutex_unlock(&data->update_lock); return sprintf(buf, "%d\n", pwm); } @@ -529,6 +537,7 @@ static ssize_t show_temp11(struct device *dev, struct device_attribute *devattr, int nr = attr->index; int temp; + mutex_lock(&data->update_lock); if (!nr) { /* * Use unsigned temperature unless its value is zero. @@ -544,7 +553,10 @@ static ssize_t show_temp11(struct device *dev, struct device_attribute *devattr, else temp = TEMP11_FROM_REG(data->temp11[nr]); } - return sprintf(buf, "%d\n", temp + data->temp2_offset); + temp += data->temp2_offset; + mutex_unlock(&data->update_lock); + + return sprintf(buf, "%d\n", temp); } static ssize_t set_temp11(struct device *dev, struct device_attribute *devattr, @@ -592,9 +604,14 @@ static ssize_t temp2_crit_hyst_show(struct device *dev, struct device_attribute *dummy, char *buf) { struct lm63_data *data = lm63_update_device(dev); - return sprintf(buf, "%d\n", temp8_from_reg(data, 2) - + data->temp2_offset - - TEMP8_FROM_REG(data->temp2_crit_hyst)); + int temp; + + mutex_lock(&data->update_lock); + temp = temp8_from_reg(data, 2) + data->temp2_offset + - TEMP8_FROM_REG(data->temp2_crit_hyst); + mutex_unlock(&data->update_lock); + + return sprintf(buf, "%d\n", temp); } static ssize_t show_lut_temp_hyst(struct device *dev, @@ -602,10 +619,14 @@ static ssize_t show_lut_temp_hyst(struct device *dev, { struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr); struct lm63_data *data = lm63_update_device(dev); + int temp; - return sprintf(buf, "%d\n", lut_temp_from_reg(data, attr->index) - + data->temp2_offset - - TEMP8_FROM_REG(data->lut_temp_hyst)); + mutex_lock(&data->update_lock); + temp = lut_temp_from_reg(data, attr->index) + data->temp2_offset + - TEMP8_FROM_REG(data->lut_temp_hyst); + mutex_unlock(&data->update_lock); + + return sprintf(buf, "%d\n", temp); } /* @@ -616,7 +637,7 @@ static ssize_t temp2_crit_hyst_store(struct device *dev, struct device_attribute *dummy, const char *buf, size_t count) { - struct lm63_data *data = dev_get_drvdata(dev); + struct lm63_data *data = lm63_update_device(dev); struct i2c_client *client = data->client; long val; int err; From c9e3878ae2f57fd6786279cf5d9dc6e6e1b52f5a Mon Sep 17 00:00:00 2001 From: Timur Tabi Date: Thu, 30 Apr 2026 17:38:29 -0500 Subject: [PATCH 128/525] Revert "drm/nouveau/gsp: add support for GA100" This reverts commit 20e0c197802c545db220157fafd567a10f2b7672. Despite claiming to add GA100 support, that commit actually has quite a few problems. It falsely claims that there is no VBIOS. GA100 does have a VBIOS, but it has no display engine, so it cannot use the PRAMIN method the read VBIOS and must fall back to using PROM. For whatever reason, the VBIOS on GA100 has an "Init-from-ROM" (IFR) header where the PCI Expansion ROM would normally be found. So to find that ROM, Nouveau needs to parse the IFR header. The commit also falsely claimed that there is no graphics (GR) engine. So rather than try to fix that commit, just revert it and start over from scratch. Signed-off-by: Timur Tabi Link: https://patch.msgid.link/20260430223838.2530778-2-ttabi@nvidia.com Signed-off-by: Danilo Krummrich --- .../gpu/drm/nouveau/nvkm/engine/device/base.c | 11 +++++++++-- .../gpu/drm/nouveau/nvkm/subdev/gsp/ga100.c | 4 ++++ .../gpu/drm/nouveau/nvkm/subdev/gsp/tu102.c | 18 +++++------------- 3 files changed, 18 insertions(+), 15 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c index 72848ed80df7..b101e14f841e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c @@ -2513,6 +2513,7 @@ static const struct nvkm_device_chip nv170_chipset = { .name = "GA100", .bar = { 0x00000001, tu102_bar_new }, + .bios = { 0x00000001, nvkm_bios_new }, .devinit = { 0x00000001, ga100_devinit_new }, .fault = { 0x00000001, tu102_fault_new }, .fb = { 0x00000001, ga100_fb_new }, @@ -2529,7 +2530,6 @@ nv170_chipset = { .vfn = { 0x00000001, ga100_vfn_new }, .ce = { 0x000003ff, ga100_ce_new }, .fifo = { 0x00000001, ga100_fifo_new }, - .sec2 = { 0x00000001, tu102_sec2_new }, }; static const struct nvkm_device_chip @@ -3341,7 +3341,6 @@ nvkm_device_ctor(const struct nvkm_device_func *func, case 0x166: device->chip = &nv166_chipset; break; case 0x167: device->chip = &nv167_chipset; break; case 0x168: device->chip = &nv168_chipset; break; - case 0x170: device->chip = &nv170_chipset; break; case 0x172: device->chip = &nv172_chipset; break; case 0x173: device->chip = &nv173_chipset; break; case 0x174: device->chip = &nv174_chipset; break; @@ -3361,6 +3360,14 @@ nvkm_device_ctor(const struct nvkm_device_func *func, case 0x1b6: device->chip = &nv1b6_chipset; break; case 0x1b7: device->chip = &nv1b7_chipset; break; default: + if (nvkm_boolopt(device->cfgopt, "NvEnableUnsupportedChipsets", false)) { + switch (device->chipset) { + case 0x170: device->chip = &nv170_chipset; break; + default: + break; + } + } + if (!device->chip) { nvdev_error(device, "unknown chipset (%08x)\n", boot0); ret = -ENODEV; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/ga100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/ga100.c index fdd820eeef81..27a13aeccd3c 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/ga100.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/ga100.c @@ -41,11 +41,15 @@ ga100_gsp_flcn = { static const struct nvkm_gsp_func ga100_gsp = { .flcn = &ga100_gsp_flcn, + .fwsec = &tu102_gsp_fwsec, .sig_section = ".fwsignature_ga100", .booter.ctor = tu102_gsp_booter_ctor, + .fwsec_sb.ctor = tu102_gsp_fwsec_sb_ctor, + .fwsec_sb.dtor = tu102_gsp_fwsec_sb_dtor, + .dtor = r535_gsp_dtor, .oneinit = tu102_gsp_oneinit, .init = tu102_gsp_init, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/tu102.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/tu102.c index dd82c76b8b9a..19cb269e7a26 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/tu102.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/tu102.c @@ -318,13 +318,8 @@ tu102_gsp_oneinit(struct nvkm_gsp *gsp) if (ret) return ret; - /* - * Calculate FB layout. FRTS is a memory region created by the FWSEC-FRTS firmware. - * FWSEC comes from VBIOS. So on systems with no VBIOS (e.g. GA100), the FRTS does - * not exist. Therefore, use the existence of VBIOS to determine whether to reserve - * an FRTS region. - */ - gsp->fb.wpr2.frts.size = device->bios ? 0x100000 : 0; + /* Calculate FB layout. */ + gsp->fb.wpr2.frts.size = 0x100000; gsp->fb.wpr2.frts.addr = ALIGN_DOWN(gsp->fb.bios.addr, 0x20000) - gsp->fb.wpr2.frts.size; gsp->fb.wpr2.boot.size = gsp->boot.fw.size; @@ -348,12 +343,9 @@ tu102_gsp_oneinit(struct nvkm_gsp *gsp) if (ret) return ret; - /* Only boot FWSEC-FRTS if it actually exists */ - if (gsp->fb.wpr2.frts.size) { - ret = nvkm_gsp_fwsec_frts(gsp); - if (WARN_ON(ret)) - return ret; - } + ret = nvkm_gsp_fwsec_frts(gsp); + if (WARN_ON(ret)) + return ret; /* Reset GSP into RISC-V mode. */ ret = gsp->func->reset(gsp); From a177ae30f78688f75ef9c6277a152c5d6979b10e Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Thu, 30 Apr 2026 16:49:51 +0200 Subject: [PATCH 129/525] netfilter: flowtable: fix inline vlan encapsulation in xmit path Several issues in the inline vlan support: - The layer 2 encapsulation representation in the tuple takes encap[0] as the outer header and encap[1] as the inner header as seen from the ingress path. Reverse the encap loop to push first the inner then the outer vlan header. - Postpone pushing the layer 2 header once destination device is known. This allows to calculate the needed hearoom via LL_RESERVED_SPACE to accommodate the layer 2 headers. - Add and use nf_flow_vlan_push() as suggested by Eric Woudstra, this is a simplified version of skb_vlan_push() for egress path only. Fixes: c653d5a78f34 ("netfilter: flowtable: inline vlan encapsulation in xmit path") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_flow_table_ip.c | 110 ++++++++++++++++++++----------- 1 file changed, 73 insertions(+), 37 deletions(-) diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c index 8d5fb7e940a1..0ce3c209050c 100644 --- a/net/netfilter/nf_flow_table_ip.c +++ b/net/netfilter/nf_flow_table_ip.c @@ -462,32 +462,6 @@ static void nf_flow_encap_pop(struct nf_flowtable_ctx *ctx, nf_flow_ip_tunnel_pop(ctx, skb); } -struct nf_flow_xmit { - const void *dest; - const void *source; - struct net_device *outdev; -}; - -static unsigned int nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, - struct nf_flow_xmit *xmit) -{ - struct net_device *dev = xmit->outdev; - unsigned int hh_len = LL_RESERVED_SPACE(dev); - - if (unlikely(skb_headroom(skb) < hh_len && dev->header_ops)) { - skb = skb_expand_head(skb, hh_len); - if (!skb) - return NF_STOLEN; - } - - skb->dev = dev; - dev_hard_header(skb, dev, ntohs(skb->protocol), - xmit->dest, xmit->source, skb->len); - dev_queue_xmit(skb); - - return NF_STOLEN; -} - static struct flow_offload_tuple_rhash * nf_flow_offload_lookup(struct nf_flowtable_ctx *ctx, struct nf_flowtable *flow_table, struct sk_buff *skb) @@ -553,6 +527,32 @@ static int nf_flow_offload_forward(struct nf_flowtable_ctx *ctx, return 1; } +/* Similar to skb_vlan_push. */ +static int nf_flow_vlan_push(struct sk_buff *skb, __be16 proto, u16 id, + u32 needed_headroom) +{ + if (skb_vlan_tag_present(skb)) { + struct vlan_hdr *vhdr; + + if (skb_cow_head(skb, needed_headroom + VLAN_HLEN)) + return -1; + + __skb_push(skb, VLAN_HLEN); + if (skb_mac_header_was_set(skb)) + skb->mac_header -= VLAN_HLEN; + + vhdr = (struct vlan_hdr *)skb->data; + skb->network_header -= VLAN_HLEN; + vhdr->h_vlan_TCI = htons(skb_vlan_tag_get(skb)); + vhdr->h_vlan_encapsulated_proto = skb->protocol; + skb->protocol = skb->vlan_proto; + skb_postpush_rcsum(skb, skb->data, VLAN_HLEN); + } + __vlan_hwaccel_put_tag(skb, proto, id); + + return 0; +} + static int nf_flow_pppoe_push(struct sk_buff *skb, u16 id) { int data_len = skb->len + sizeof(__be16); @@ -739,17 +739,19 @@ static int nf_flow_tunnel_v6_push(struct net *net, struct sk_buff *skb, } static int nf_flow_encap_push(struct sk_buff *skb, - struct flow_offload_tuple *tuple) + struct flow_offload_tuple *tuple, + struct net_device *outdev) { + u32 needed_headroom = LL_RESERVED_SPACE(outdev); int i; - for (i = 0; i < tuple->encap_num; i++) { + for (i = tuple->encap_num - 1; i >= 0; i--) { switch (tuple->encap[i].proto) { case htons(ETH_P_8021Q): case htons(ETH_P_8021AD): - skb_reset_mac_header(skb); - if (skb_vlan_push(skb, tuple->encap[i].proto, - tuple->encap[i].id) < 0) + if (nf_flow_vlan_push(skb, tuple->encap[i].proto, + tuple->encap[i].id, + needed_headroom) < 0) return -1; break; case htons(ETH_P_PPP_SES): @@ -762,6 +764,44 @@ static int nf_flow_encap_push(struct sk_buff *skb, return 0; } +struct nf_flow_xmit { + const void *dest; + const void *source; + struct net_device *outdev; + struct flow_offload_tuple *tuple; +}; + +static void __nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, + struct nf_flow_xmit *xmit) +{ + struct net_device *dev = xmit->outdev; + unsigned int hh_len = LL_RESERVED_SPACE(dev); + + if (unlikely(skb_headroom(skb) < hh_len && dev->header_ops)) { + skb = skb_expand_head(skb, hh_len); + if (!skb) + return; + } + + skb->dev = dev; + dev_hard_header(skb, dev, ntohs(skb->protocol), + xmit->dest, xmit->source, skb->len); + dev_queue_xmit(skb); +} + +static unsigned int nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, + struct nf_flow_xmit *xmit) +{ + if (xmit->tuple->encap_num) { + if (nf_flow_encap_push(skb, xmit->tuple, xmit->outdev) < 0) + return NF_DROP; + } + + __nf_flow_queue_xmit(net, skb, xmit); + + return NF_STOLEN; +} + unsigned int nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, const struct nf_hook_state *state) @@ -806,9 +846,6 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, if (nf_flow_tunnel_v4_push(state->net, skb, other_tuple, &ip_daddr) < 0) return NF_DROP; - if (nf_flow_encap_push(skb, other_tuple) < 0) - return NF_DROP; - switch (tuplehash->tuple.xmit_type) { case FLOW_OFFLOAD_XMIT_NEIGH: rt = dst_rtable(tuplehash->tuple.dst_cache); @@ -838,6 +875,7 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, WARN_ON_ONCE(1); return NF_DROP; } + xmit.tuple = other_tuple; return nf_flow_queue_xmit(state->net, skb, &xmit); } @@ -1128,9 +1166,6 @@ nf_flow_offload_ipv6_hook(void *priv, struct sk_buff *skb, &ip6_daddr, encap_limit) < 0) return NF_DROP; - if (nf_flow_encap_push(skb, other_tuple) < 0) - return NF_DROP; - switch (tuplehash->tuple.xmit_type) { case FLOW_OFFLOAD_XMIT_NEIGH: rt = dst_rt6_info(tuplehash->tuple.dst_cache); @@ -1160,6 +1195,7 @@ nf_flow_offload_ipv6_hook(void *priv, struct sk_buff *skb, WARN_ON_ONCE(1); return NF_DROP; } + xmit.tuple = other_tuple; return nf_flow_queue_xmit(state->net, skb, &xmit); } From 69c54f80f4a7072b51b5b5939185ca5e572be982 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Thu, 30 Apr 2026 16:49:53 +0200 Subject: [PATCH 130/525] netfilter: flowtable: fix inline pppoe encapsulation in xmit path Address two issues in the inline pppoe encapsulation: - Add needs_gso_segment flag to segment PPPoE packets in software given that there is no GSO support for this. - Use FLOW_OFFLOAD_XMIT_DIRECT since neighbour cache is not available in point-to-point device, use the hardware address that is obtained via flowtable path discovery (ie. fill_forward_path). Fixes: 18d27bed0880 ("netfilter: flowtable: inline pppoe encapsulation in xmit path") Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_flow_table.h | 4 ++- net/netfilter/nf_flow_table_core.c | 1 + net/netfilter/nf_flow_table_ip.c | 42 +++++++++++++++++++++++++-- net/netfilter/nf_flow_table_path.c | 7 ++++- 4 files changed, 49 insertions(+), 5 deletions(-) diff --git a/include/net/netfilter/nf_flow_table.h b/include/net/netfilter/nf_flow_table.h index b09c11c048d5..7b23b245a5a8 100644 --- a/include/net/netfilter/nf_flow_table.h +++ b/include/net/netfilter/nf_flow_table.h @@ -148,9 +148,10 @@ struct flow_offload_tuple { /* All members above are keys for lookups, see flow_offload_hash(). */ struct { } __hash; - u8 dir:2, + u16 dir:2, xmit_type:3, encap_num:2, + needs_gso_segment:1, tun_num:2, in_vlan_ingress:2; u16 mtu; @@ -232,6 +233,7 @@ struct nf_flow_route { u32 hw_ifindex; u8 h_source[ETH_ALEN]; u8 h_dest[ETH_ALEN]; + u8 needs_gso_segment:1; } out; enum flow_offload_xmit_type xmit_type; } tuple[FLOW_OFFLOAD_DIR_MAX]; diff --git a/net/netfilter/nf_flow_table_core.c b/net/netfilter/nf_flow_table_core.c index 2c4140e6f53c..785d8c244a77 100644 --- a/net/netfilter/nf_flow_table_core.c +++ b/net/netfilter/nf_flow_table_core.c @@ -122,6 +122,7 @@ static int flow_offload_fill_route(struct flow_offload *flow, flow_tuple->tun = route->tuple[dir].in.tun; flow_tuple->encap_num = route->tuple[dir].in.num_encaps; + flow_tuple->needs_gso_segment = route->tuple[dir].out.needs_gso_segment; flow_tuple->tun_num = route->tuple[dir].in.num_tuns; switch (route->tuple[dir].xmit_type) { diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c index 0ce3c209050c..2eba64eb393a 100644 --- a/net/netfilter/nf_flow_table_ip.c +++ b/net/netfilter/nf_flow_table_ip.c @@ -553,7 +553,8 @@ static int nf_flow_vlan_push(struct sk_buff *skb, __be16 proto, u16 id, return 0; } -static int nf_flow_pppoe_push(struct sk_buff *skb, u16 id) +static int nf_flow_pppoe_push(struct sk_buff *skb, u16 id, + u32 needed_headroom) { int data_len = skb->len + sizeof(__be16); struct ppp_hdr { @@ -562,7 +563,7 @@ static int nf_flow_pppoe_push(struct sk_buff *skb, u16 id) } *ph; __be16 proto; - if (skb_cow_head(skb, PPPOE_SES_HLEN)) + if (skb_cow_head(skb, needed_headroom + PPPOE_SES_HLEN)) return -1; switch (skb->protocol) { @@ -755,7 +756,8 @@ static int nf_flow_encap_push(struct sk_buff *skb, return -1; break; case htons(ETH_P_PPP_SES): - if (nf_flow_pppoe_push(skb, tuple->encap[i].id) < 0) + if (nf_flow_pppoe_push(skb, tuple->encap[i].id, + needed_headroom) < 0) return -1; break; } @@ -769,6 +771,7 @@ struct nf_flow_xmit { const void *source; struct net_device *outdev; struct flow_offload_tuple *tuple; + bool needs_gso_segment; }; static void __nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, @@ -789,10 +792,41 @@ static void __nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, dev_queue_xmit(skb); } +static unsigned int nf_flow_encap_gso_xmit(struct net *net, struct sk_buff *skb, + struct nf_flow_xmit *xmit) +{ + struct sk_buff *segs, *nskb; + + segs = skb_gso_segment(skb, 0); + if (IS_ERR(segs)) + return NF_DROP; + + if (segs) + consume_skb(skb); + else + segs = skb; + + skb_list_walk_safe(segs, segs, nskb) { + skb_mark_not_on_list(segs); + + if (nf_flow_encap_push(segs, xmit->tuple, xmit->outdev) < 0) { + kfree_skb(segs); + kfree_skb_list(nskb); + return NF_STOLEN; + } + __nf_flow_queue_xmit(net, segs, xmit); + } + + return NF_STOLEN; +} + static unsigned int nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, struct nf_flow_xmit *xmit) { if (xmit->tuple->encap_num) { + if (skb_is_gso(skb) && xmit->needs_gso_segment) + return nf_flow_encap_gso_xmit(net, skb, xmit); + if (nf_flow_encap_push(skb, xmit->tuple, xmit->outdev) < 0) return NF_DROP; } @@ -876,6 +910,7 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, return NF_DROP; } xmit.tuple = other_tuple; + xmit.needs_gso_segment = tuplehash->tuple.needs_gso_segment; return nf_flow_queue_xmit(state->net, skb, &xmit); } @@ -1196,6 +1231,7 @@ nf_flow_offload_ipv6_hook(void *priv, struct sk_buff *skb, return NF_DROP; } xmit.tuple = other_tuple; + xmit.needs_gso_segment = tuplehash->tuple.needs_gso_segment; return nf_flow_queue_xmit(state->net, skb, &xmit); } diff --git a/net/netfilter/nf_flow_table_path.c b/net/netfilter/nf_flow_table_path.c index 6bb9579dcc2a..9e88ea6a2eef 100644 --- a/net/netfilter/nf_flow_table_path.c +++ b/net/netfilter/nf_flow_table_path.c @@ -86,6 +86,7 @@ struct nft_forward_info { u8 ingress_vlans; u8 h_source[ETH_ALEN]; u8 h_dest[ETH_ALEN]; + bool needs_gso_segment; enum flow_offload_xmit_type xmit_type; }; @@ -138,8 +139,11 @@ static void nft_dev_path_info(const struct net_device_path_stack *stack, path->encap.proto; info->num_encaps++; } - if (path->type == DEV_PATH_PPPOE) + if (path->type == DEV_PATH_PPPOE) { memcpy(info->h_dest, path->encap.h_dest, ETH_ALEN); + info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT; + info->needs_gso_segment = 1; + } break; case DEV_PATH_BRIDGE: if (is_zero_ether_addr(info->h_source)) @@ -279,6 +283,7 @@ static void nft_dev_forward_path(const struct nft_pktinfo *pkt, memcpy(route->tuple[dir].out.h_dest, info.h_dest, ETH_ALEN); route->tuple[dir].xmit_type = info.xmit_type; } + route->tuple[dir].out.needs_gso_segment = info.needs_gso_segment; } int nft_flow_route(const struct nft_pktinfo *pkt, const struct nf_conn *ct, From e027c218c482c6a0ae1948129ccda3b0a2033368 Mon Sep 17 00:00:00 2001 From: Robert Marko Date: Tue, 28 Apr 2026 15:41:01 +0200 Subject: [PATCH 131/525] net: phy: micrel: fix LAN8814 QSGMII soft reset LAN8814 QSGMII soft reset was moved into the probe function to avoid triggering it for each of 4 PHY-s in the package. However, that broke QSGMII link between the MAC and PHY on most LAN8814 PHY-s, specificaly for us on the Microchip LAN969x switch. Reading the QSGMII status registers it was visible that lanes were only partially synced. It looks like the reset timing is crucial, so lets move the reset back into the .config_init function but guard it with phy_package_init_once() to avoid it being triggered on each of 4 PHY-s in the package. Change the probe function to use phy_package_probe_once() for coma and PtP setup. Fixes: 96a9178a29a6 ("net: phy: micrel: lan8814 fix reset of the QSGMII interface") Signed-off-by: Robert Marko Link: https://patch.msgid.link/20260428134138.1741253-1-robert.marko@sartura.hr Signed-off-by: Jakub Kicinski --- drivers/net/phy/micrel.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/drivers/net/phy/micrel.c b/drivers/net/phy/micrel.c index 2aa1dedd21b8..e211a523c258 100644 --- a/drivers/net/phy/micrel.c +++ b/drivers/net/phy/micrel.c @@ -4548,6 +4548,13 @@ static int lan8814_config_init(struct phy_device *phydev) struct kszphy_priv *lan8814 = phydev->priv; int ret; + if (phy_package_init_once(phydev)) + /* Reset the PHY */ + lanphy_modify_page_reg(phydev, LAN8814_PAGE_COMMON_REGS, + LAN8814_QSGMII_SOFT_RESET, + LAN8814_QSGMII_SOFT_RESET_BIT, + LAN8814_QSGMII_SOFT_RESET_BIT); + /* Based on the interface type select how the advertise ability is * encoded, to set as SGMII or as USGMII. */ @@ -4655,13 +4662,7 @@ static int lan8814_probe(struct phy_device *phydev) priv->is_ptp_available = err == LAN8814_REV_LAN8814 || err == LAN8814_REV_LAN8818; - if (phy_package_init_once(phydev)) { - /* Reset the PHY */ - lanphy_modify_page_reg(phydev, LAN8814_PAGE_COMMON_REGS, - LAN8814_QSGMII_SOFT_RESET, - LAN8814_QSGMII_SOFT_RESET_BIT, - LAN8814_QSGMII_SOFT_RESET_BIT); - + if (phy_package_probe_once(phydev)) { err = lan8814_release_coma_mode(phydev); if (err) return err; From 3744b0964d5267c0b651bcd8f8c25db6bf4ccbac Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 29 Apr 2026 17:46:48 +0200 Subject: [PATCH 132/525] ipv6: Implement limits on extension header parsing ipv6_{skip_exthdr,find_hdr}() and ip6_{tnl_parse_tlv_enc_lim, protocol_deliver_rcu}() iterate over IPv6 extension headers until they find a non-extension-header protocol or run out of packet data. The loops have no iteration counter, relying solely on the packet length to bound them. For a crafted packet with 8-byte extension headers filling a 64KB jumbogram, this means a worst case of up to ~8k iterations with a skb_header_pointer call each. ipv6_skip_exthdr(), for example, is used where it parses the inner quoted packet inside an incoming ICMPv6 error: - icmpv6_rcv - checksum validation - case ICMPV6_DEST_UNREACH - icmpv6_notify - pskb_may_pull() <- pull inner IPv6 header - ipv6_skip_exthdr() <- iterates here - pskb_may_pull() - ipprot->err_handler() <- sk lookup The per-iteration cost of ipv6_skip_exthdr itself is generally light, but skb_header_pointer becomes more costly on reassembled packets: the first ~1232 bytes of the inner packet are in the skb's linear area, but the remaining ~63KB are in the frag_list where skb_copy_bits is needed to read data. Initially, the idea was to add a configurable limit via a new sysctl knob with default 8, in line with knobs from commit 47d3d7ac656a ("ipv6: Implement limits on Hop-by-Hop and Destination options"), but two reasons eventually argued against it: - It adds to UAPI that needs to be maintained forever, and upcoming work is restricting extension header ordering anyway, leaving little reason for another sysctl knob - exthdrs_core.c is always built-in even when CONFIG_IPV6=n, where struct net has no .ipv6 member, so the read site would need an ifdef'd fallback to a constant anyway Therefore, just use a constant (IP6_MAX_EXT_HDRS_CNT). All four extension header walking functions are now bound by this limit. Note that the check in ip6_protocol_deliver_rcu() happens right before the goto resubmit, such that we don't have to have a test for ipv6_ext_hdr() in the fast-path. There's an ongoing IETF draft-iurman-6man-eh-occurrences to enforce IPv6 extension headers ordering and occurrence. The latter also discusses security implications. As per RFC8200 section 4.1, the occurrence rules for extension headers provide a practical upper bound which is 8. In order to be conservative, let's define IP6_MAX_EXT_HDRS_CNT as 12 to leave enough room for quirky setups. In the unlikely event that this is still not enough, then we might need to reconsider a sysctl. Signed-off-by: Daniel Borkmann Reviewed-by: Ido Schimmel Reviewed-by: Eric Dumazet Reviewed-by: Justin Iurman Link: https://patch.msgid.link/20260429154648.809751-1-daniel@iogearbox.net Signed-off-by: Jakub Kicinski --- include/net/dropreason-core.h | 6 ++++++ include/net/ipv6.h | 3 +++ net/ipv6/exthdrs_core.c | 7 +++++++ net/ipv6/ip6_input.c | 5 +++++ net/ipv6/ip6_tunnel.c | 4 ++++ 5 files changed, 25 insertions(+) diff --git a/include/net/dropreason-core.h b/include/net/dropreason-core.h index e0ca3904ff8e..2f312d1f67d6 100644 --- a/include/net/dropreason-core.h +++ b/include/net/dropreason-core.h @@ -99,6 +99,7 @@ FN(FRAG_TOO_FAR) \ FN(TCP_MINTTL) \ FN(IPV6_BAD_EXTHDR) \ + FN(IPV6_TOO_MANY_EXTHDRS) \ FN(IPV6_NDISC_FRAG) \ FN(IPV6_NDISC_HOP_LIMIT) \ FN(IPV6_NDISC_BAD_CODE) \ @@ -494,6 +495,11 @@ enum skb_drop_reason { SKB_DROP_REASON_TCP_MINTTL, /** @SKB_DROP_REASON_IPV6_BAD_EXTHDR: Bad IPv6 extension header. */ SKB_DROP_REASON_IPV6_BAD_EXTHDR, + /** + * @SKB_DROP_REASON_IPV6_TOO_MANY_EXTHDRS: Number of IPv6 extension + * headers in the packet exceeds IP6_MAX_EXT_HDRS_CNT. + */ + SKB_DROP_REASON_IPV6_TOO_MANY_EXTHDRS, /** @SKB_DROP_REASON_IPV6_NDISC_FRAG: invalid frag (suppress_frag_ndisc). */ SKB_DROP_REASON_IPV6_NDISC_FRAG, /** @SKB_DROP_REASON_IPV6_NDISC_HOP_LIMIT: invalid hop limit. */ diff --git a/include/net/ipv6.h b/include/net/ipv6.h index d042afe7a245..1dec81faff28 100644 --- a/include/net/ipv6.h +++ b/include/net/ipv6.h @@ -90,6 +90,9 @@ struct ip_tunnel_info; #define IP6_DEFAULT_MAX_DST_OPTS_LEN INT_MAX /* No limit */ #define IP6_DEFAULT_MAX_HBH_OPTS_LEN INT_MAX /* No limit */ +/* Hard limit on traversed IPv6 extension headers */ +#define IP6_MAX_EXT_HDRS_CNT 12 + /* * Addr type * diff --git a/net/ipv6/exthdrs_core.c b/net/ipv6/exthdrs_core.c index 49e31e4ae7b7..9d06d487e8b1 100644 --- a/net/ipv6/exthdrs_core.c +++ b/net/ipv6/exthdrs_core.c @@ -73,6 +73,7 @@ int ipv6_skip_exthdr(const struct sk_buff *skb, int start, u8 *nexthdrp, __be16 *frag_offp) { u8 nexthdr = *nexthdrp; + int exthdr_cnt = 0; *frag_offp = 0; @@ -82,6 +83,8 @@ int ipv6_skip_exthdr(const struct sk_buff *skb, int start, u8 *nexthdrp, if (nexthdr == NEXTHDR_NONE) return -1; + if (unlikely(exthdr_cnt++ >= IP6_MAX_EXT_HDRS_CNT)) + return -1; hp = skb_header_pointer(skb, start, sizeof(_hdr), &_hdr); if (!hp) return -1; @@ -190,6 +193,7 @@ int ipv6_find_hdr(const struct sk_buff *skb, unsigned int *offset, { unsigned int start = skb_network_offset(skb) + sizeof(struct ipv6hdr); u8 nexthdr = ipv6_hdr(skb)->nexthdr; + int exthdr_cnt = 0; bool found; if (fragoff) @@ -216,6 +220,9 @@ int ipv6_find_hdr(const struct sk_buff *skb, unsigned int *offset, return -ENOENT; } + if (unlikely(exthdr_cnt++ >= IP6_MAX_EXT_HDRS_CNT)) + return -EBADMSG; + hp = skb_header_pointer(skb, start, sizeof(_hdr), &_hdr); if (!hp) return -EBADMSG; diff --git a/net/ipv6/ip6_input.c b/net/ipv6/ip6_input.c index 967b07aeb683..8972863c93ee 100644 --- a/net/ipv6/ip6_input.c +++ b/net/ipv6/ip6_input.c @@ -403,6 +403,7 @@ INDIRECT_CALLABLE_DECLARE(int tcp_v6_rcv(struct sk_buff *)); void ip6_protocol_deliver_rcu(struct net *net, struct sk_buff *skb, int nexthdr, bool have_final) { + int exthdr_cnt = IP6CB(skb)->flags & IP6SKB_HOPBYHOP ? 1 : 0; const struct inet6_protocol *ipprot; struct inet6_dev *idev; unsigned int nhoff; @@ -487,6 +488,10 @@ void ip6_protocol_deliver_rcu(struct net *net, struct sk_buff *skb, int nexthdr, nexthdr = ret; goto resubmit_final; } else { + if (unlikely(exthdr_cnt++ >= IP6_MAX_EXT_HDRS_CNT)) { + SKB_DR_SET(reason, IPV6_TOO_MANY_EXTHDRS); + goto discard; + } goto resubmit; } } else if (ret == 0) { diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index c468c83af0f2..9d1037ac082f 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -399,11 +399,15 @@ __u16 ip6_tnl_parse_tlv_enc_lim(struct sk_buff *skb, __u8 *raw) unsigned int nhoff = raw - skb->data; unsigned int off = nhoff + sizeof(*ipv6h); u8 nexthdr = ipv6h->nexthdr; + int exthdr_cnt = 0; while (ipv6_ext_hdr(nexthdr) && nexthdr != NEXTHDR_NONE) { struct ipv6_opt_hdr *hdr; u16 optlen; + if (unlikely(exthdr_cnt++ >= IP6_MAX_EXT_HDRS_CNT)) + break; + if (!pskb_may_pull(skb, off + sizeof(*hdr))) break; From 26ebd12e67bfc3543d77ce586c33ef29fcafab20 Mon Sep 17 00:00:00 2001 From: Wei Fang Date: Wed, 29 Apr 2026 16:19:30 +0800 Subject: [PATCH 133/525] net: enetc: fix VSI mailbox timeout handling and DMA lifecycle In the current VSI mailbox implementation, the VSI allocates a DMA buffer to store the message sent to the PSI. When the PSI receives the message request from the VSI, the hardware copies the message data from this DMA buffer to PSI's DMA buffer for processing. When enetc_msg_vsi_send() times out, two scenarios can occur: 1) Use-after-free: If the hardware hasn't completed message copying when the VSI frees the buffer, the hardware may subsequently copy the data from freed memory to PSI's DMA buffer. 2) Message race: If PSI hasn't processed the previous message when the next message is sent, the VSI may receive the previous message's reply, leading to incorrect handling. To address these issues, implement the following changes: - Check the mailbox busy status before sending a new message. If the mailbox is in busy state, it indicates the previous message is still being processed, so return an error immediately. - Add the 'msg' field to struct enetc_si to preserve the DMA buffer information. The caller of enetc_msg_vsi_send() no longer frees the DMA buffer. Instead, defer freeing until it is safe to do so (when mailbox is not busy on next send). - Add cleanup in enetc_vf_remove() to free the last message buffer. This ensures the DMA buffer remains valid during message copying and prevents message reply mismatches. Fixes: beb74ac878c8 ("enetc: Add vf to pf messaging support") Signed-off-by: Wei Fang Link: https://patch.msgid.link/20260429081930.3259824-1-wei.fang@nxp.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/freescale/enetc/enetc.h | 1 + .../net/ethernet/freescale/enetc/enetc_vf.c | 42 +++++++++++++++---- 2 files changed, 35 insertions(+), 8 deletions(-) diff --git a/drivers/net/ethernet/freescale/enetc/enetc.h b/drivers/net/ethernet/freescale/enetc/enetc.h index e663bb5e614e..e691144e8756 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc.h +++ b/drivers/net/ethernet/freescale/enetc/enetc.h @@ -330,6 +330,7 @@ struct enetc_si { struct workqueue_struct *workqueue; struct work_struct rx_mode_task; struct dentry *debugfs_root; + struct enetc_msg_swbd msg; /* Only valid for VSI */ }; #define ENETC_SI_ALIGN 32 diff --git a/drivers/net/ethernet/freescale/enetc/enetc_vf.c b/drivers/net/ethernet/freescale/enetc/enetc_vf.c index 6c4b374bcb0e..df8e95cc47d0 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_vf.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_vf.c @@ -17,11 +17,36 @@ static void enetc_msg_vsi_write_msg(struct enetc_hw *hw, enetc_wr(hw, ENETC_VSIMSGSNDAR0, val); } +static void enetc_msg_dma_free(struct device *dev, struct enetc_msg_swbd *msg) +{ + if (msg->vaddr) { + dma_free_coherent(dev, msg->size, msg->vaddr, msg->dma); + msg->vaddr = NULL; + } +} + static int enetc_msg_vsi_send(struct enetc_si *si, struct enetc_msg_swbd *msg) { + struct device *dev = &si->pdev->dev; int timeout = 100; u32 vsimsgsr; + /* The VSI mailbox may be busy if last message was not yet processed + * by PSI. So need to check the mailbox status before sending. + */ + vsimsgsr = enetc_rd(&si->hw, ENETC_VSIMSGSR); + if (vsimsgsr & ENETC_VSIMSGSR_MB) { + /* It is safe to free the DMA buffer here, the caller does + * not access the DMA buffer if enetc_msg_vsi_send() fails. + */ + enetc_msg_dma_free(dev, msg); + dev_err(dev, "VSI mailbox is busy\n"); + return -EIO; + } + + /* Free the DMA buffer of the last message */ + enetc_msg_dma_free(dev, &si->msg); + si->msg = *msg; enetc_msg_vsi_write_msg(&si->hw, msg); do { @@ -32,12 +57,15 @@ static int enetc_msg_vsi_send(struct enetc_si *si, struct enetc_msg_swbd *msg) usleep_range(1000, 2000); } while (--timeout); - if (!timeout) + if (!timeout) { + dev_err(dev, "VSI mailbox timeout\n"); + return -ETIMEDOUT; + } /* check for message delivery error */ if (vsimsgsr & ENETC_VSIMSGSR_MS) { - dev_err(&si->pdev->dev, "VSI command execute error: %d\n", + dev_err(dev, "VSI command execute error: %d\n", ENETC_SIMSGSR_GET_MC(vsimsgsr)); return -EIO; } @@ -50,7 +78,6 @@ static int enetc_msg_vsi_set_primary_mac_addr(struct enetc_ndev_priv *priv, { struct enetc_msg_cmd_set_primary_mac *cmd; struct enetc_msg_swbd msg; - int err; msg.size = ALIGN(sizeof(struct enetc_msg_cmd_set_primary_mac), 64); msg.vaddr = dma_alloc_coherent(priv->dev, msg.size, &msg.dma, @@ -67,11 +94,7 @@ static int enetc_msg_vsi_set_primary_mac_addr(struct enetc_ndev_priv *priv, memcpy(&cmd->mac, saddr, sizeof(struct sockaddr)); /* send the command and wait */ - err = enetc_msg_vsi_send(priv->si, &msg); - - dma_free_coherent(priv->dev, msg.size, msg.vaddr, msg.dma); - - return err; + return enetc_msg_vsi_send(priv->si, &msg); } static int enetc_vf_set_mac_addr(struct net_device *ndev, void *addr) @@ -259,6 +282,7 @@ static void enetc_vf_remove(struct pci_dev *pdev) { struct enetc_si *si = pci_get_drvdata(pdev); struct enetc_ndev_priv *priv; + struct enetc_msg_swbd msg; priv = netdev_priv(si->ndev); unregister_netdev(si->ndev); @@ -270,7 +294,9 @@ static void enetc_vf_remove(struct pci_dev *pdev) free_netdev(si->ndev); + msg = si->msg; enetc_pci_remove(pdev); + enetc_msg_dma_free(&pdev->dev, &msg); } static const struct pci_device_id enetc_vf_id_table[] = { From 694de316f607fe2473d52ca0707e3918e72c1562 Mon Sep 17 00:00:00 2001 From: Jiawen Wu Date: Wed, 29 Apr 2026 16:37:42 +0800 Subject: [PATCH 134/525] net: libwx: fix VF illegal register access Register WX_CFG_PORT_ST is a PF restricted register. When a VF is initialized, attempting to read this register triggers an illegal register access, which lead to a system hang. When the device is VF, the bus function ID can be obtained directly from the PCI_FUNC(pdev->devfn). Fixes: a04ea57aae37 ("net: libwx: fix device bus LAN ID") Cc: stable@vger.kernel.org Signed-off-by: Jiawen Wu Link: https://patch.msgid.link/4D1F4452D21DE107+20260429083743.88961-1-jiawenwu@trustnetic.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/wangxun/libwx/wx_hw.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/wangxun/libwx/wx_hw.c b/drivers/net/ethernet/wangxun/libwx/wx_hw.c index d3772d01e00b..2451f6b20b11 100644 --- a/drivers/net/ethernet/wangxun/libwx/wx_hw.c +++ b/drivers/net/ethernet/wangxun/libwx/wx_hw.c @@ -2480,8 +2480,11 @@ int wx_sw_init(struct wx *wx) wx->oem_svid = pdev->subsystem_vendor; wx->oem_ssid = pdev->subsystem_device; wx->bus.device = PCI_SLOT(pdev->devfn); - wx->bus.func = FIELD_GET(WX_CFG_PORT_ST_LANID, - rd32(wx, WX_CFG_PORT_ST)); + if (pdev->is_virtfn) + wx->bus.func = PCI_FUNC(pdev->devfn); + else + wx->bus.func = FIELD_GET(WX_CFG_PORT_ST_LANID, + rd32(wx, WX_CFG_PORT_ST)); if (wx->oem_svid == PCI_VENDOR_ID_WANGXUN || pdev->is_virtfn) { From 7a33345153eeeda195c55f15be27074e4c3b5109 Mon Sep 17 00:00:00 2001 From: Jiawen Wu Date: Wed, 29 Apr 2026 16:37:43 +0800 Subject: [PATCH 135/525] net: libwx: use request_irq for VF misc interrupt Currently, request_threaded_irq() is used with a primary handler but a NULL threaded handler, while also setting the IRQF_ONESHOT flag. This specific combination triggers a WARNING since the commit aef30c8d569c ("genirq: Warn about using IRQF_ONESHOT without a threaded handler"). WARNING: kernel/irq/manage.c:1502 at __setup_irq+0x4fa/0x760 Fix the issue by switching to request_irq(), which is the appropriate interface or a non-threaded interrupt handler, and removing the unnecessary IRQF_ONESHOT flag. Fixes: eb4898fde1de ("net: libwx: add wangxun vf common api") Cc: stable@vger.kernel.org Signed-off-by: Jiawen Wu Link: https://patch.msgid.link/786DDC7D5CCA6D0A+20260429083743.88961-2-jiawenwu@trustnetic.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/wangxun/libwx/wx_vf_common.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/wangxun/libwx/wx_vf_common.c b/drivers/net/ethernet/wangxun/libwx/wx_vf_common.c index 29cdbed2e5ec..94ff8f5f0b4c 100644 --- a/drivers/net/ethernet/wangxun/libwx/wx_vf_common.c +++ b/drivers/net/ethernet/wangxun/libwx/wx_vf_common.c @@ -99,8 +99,8 @@ int wx_request_msix_irqs_vf(struct wx *wx) } } - err = request_threaded_irq(wx->msix_entry->vector, wx_msix_misc_vf, - NULL, IRQF_ONESHOT, netdev->name, wx); + err = request_irq(wx->msix_entry->vector, wx_msix_misc_vf, + 0, netdev->name, wx); if (err) { wx_err(wx, "request_irq for msix_other failed: %d\n", err); goto free_queue_irqs; From 75df490c9e8457990c8b227650f6491218ce018b Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Wed, 29 Apr 2026 14:02:31 +0200 Subject: [PATCH 136/525] net: airoha: Move entries to queue head in case of DMA mapping failure in airoha_dev_xmit() In order to respect the original descriptor order and avoid any potential IOMMU fault or memory corruption, move pending queue entries to the head of hw queue tx_list if the DMA mapping of current inflight packet fails in airoha_dev_xmit routine. Fixes: 3f47e67dff1f7 ("net: airoha: Add the capability to consume out-of-order DMA tx descriptors") Signed-off-by: Lorenzo Bianconi Link: https://patch.msgid.link/20260429-airoha-xmit-unmap-error-path-v2-1-32e43b7c6d25@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/airoha/airoha_eth.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/airoha/airoha_eth.c b/drivers/net/ethernet/airoha/airoha_eth.c index f8b3d53bccad..d0c0c0ec8a80 100644 --- a/drivers/net/ethernet/airoha/airoha_eth.c +++ b/drivers/net/ethernet/airoha/airoha_eth.c @@ -2120,14 +2120,12 @@ static netdev_tx_t airoha_dev_xmit(struct sk_buff *skb, return NETDEV_TX_OK; error_unmap: - while (!list_empty(&tx_list)) { - e = list_first_entry(&tx_list, struct airoha_queue_entry, - list); + list_for_each_entry(e, &tx_list, list) { dma_unmap_single(dev->dev.parent, e->dma_addr, e->dma_len, DMA_TO_DEVICE); e->dma_addr = 0; - list_move_tail(&e->list, &q->tx_list); } + list_splice(&tx_list, &q->tx_list); spin_unlock_bh(&q->lock); error: From aaadccde312f1f6c752461e015adcaa25d463cbc Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 29 Apr 2026 07:57:13 +0530 Subject: [PATCH 137/525] octeontx2-af: npc: cn20k: Propagate MCAM key-type errors on cn20k npc_mcam_idx_2_key_type() can fail; callers used to ignore it and still used kw_type when enabling, configuring, copying, and reading MCAM entries. That could program or decode hardware with an undefined key type. Return -EINVAL when key-type lookup fails. Return -EINVAL from npc_cn20k_copy_mcam_entry() when src and dest key types differ instead of failing silently. Change npc_cn20k_{enable,config,copy,read}_mcam_entry() to return int on success or error. Thread those errors through the cn20k MCAM write and read mbox handlers, the cn20k baseline steer read path, NPC defrag move (disable/copy/enable with dev_err and -EFAULT), and the DMAC update path in rvu_npc_fs.c. Make npc_copy_mcam_entry() return int so the cn20k branch can return npc_cn20k_copy_mcam_entry() without a void/int mismatch, and fail NPC_MCAM_SHIFT_ENTRY when copy fails. Cc: Suman Ghosh Cc: Dan Carpenter Fixes: 6d1e70282f76 ("octeontx2-af: npc: cn20k: Use common APIs") Link: https://lore.kernel.org/netdev/adiQJvuKlEhq2ILx@stanley.mountain/ Signed-off-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260429022722.1110289-2-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../ethernet/marvell/octeontx2/af/cn20k/npc.c | 124 +++++++++++++----- .../ethernet/marvell/octeontx2/af/cn20k/npc.h | 20 +-- .../ethernet/marvell/octeontx2/af/rvu_npc.c | 18 ++- .../marvell/octeontx2/af/rvu_npc_fs.c | 20 ++- 4 files changed, 125 insertions(+), 57 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c index 7291fdb89b03..7170dcf26200 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c @@ -798,7 +798,7 @@ void npc_cn20k_load_mkex_profile(struct rvu *rvu, int blkaddr, iounmap(mkex_prfl_addr); } -void +int npc_cn20k_enable_mcam_entry(struct rvu *rvu, int blkaddr, int index, bool enable) { @@ -808,7 +808,9 @@ npc_cn20k_enable_mcam_entry(struct rvu *rvu, int blkaddr, u64 cfg, hw_prio; u8 kw_type; - npc_mcam_idx_2_key_type(rvu, index, &kw_type); + if (npc_mcam_idx_2_key_type(rvu, index, &kw_type)) + return -EINVAL; + if (kw_type == NPC_MCAM_KEY_X2) { cfg = rvu_read64(rvu, blkaddr, NPC_AF_CN20K_MCAMEX_BANKX_CFG_EXT(mcam_idx, @@ -819,7 +821,7 @@ npc_cn20k_enable_mcam_entry(struct rvu *rvu, int blkaddr, rvu_write64(rvu, blkaddr, NPC_AF_CN20K_MCAMEX_BANKX_CFG_EXT(mcam_idx, bank), cfg); - return; + return 0; } /* For NPC_CN20K_MCAM_KEY_X4 keys, both the banks @@ -836,6 +838,8 @@ npc_cn20k_enable_mcam_entry(struct rvu *rvu, int blkaddr, NPC_AF_CN20K_MCAMEX_BANKX_CFG_EXT(mcam_idx, bank), cfg); } + + return 0; } void @@ -1042,9 +1046,9 @@ npc_cn20k_set_mcam_bank_cfg(struct rvu *rvu, int blkaddr, int mcam_idx, } } -void npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, - u8 intf, struct cn20k_mcam_entry *entry, - bool enable, u8 hw_prio, u8 req_kw_type) +int npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, + u8 intf, struct cn20k_mcam_entry *entry, + bool enable, u8 hw_prio, u8 req_kw_type) { struct npc_mcam *mcam = &rvu->hw->mcam; int mcam_idx = index % mcam->banksize; @@ -1052,10 +1056,13 @@ void npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, int kw = 0; u8 kw_type; - /* Disable before mcam entry update */ - npc_cn20k_enable_mcam_entry(rvu, blkaddr, index, false); + if (npc_mcam_idx_2_key_type(rvu, index, &kw_type)) + return -EINVAL; + + /* Disable before mcam entry update */ + if (npc_cn20k_enable_mcam_entry(rvu, blkaddr, index, false)) + return -EINVAL; - npc_mcam_idx_2_key_type(rvu, index, &kw_type); /* CAM1 takes the comparison value and * CAM0 specifies match for a bit in key being '0' or '1' or 'dontcare'. * CAM1 = 0 & CAM0 = 1 => match if key = 0 @@ -1120,9 +1127,11 @@ void npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, /* PF installing VF rule */ npc_cn20k_set_mcam_bank_cfg(rvu, blkaddr, mcam_idx, bank, kw_type, enable, hw_prio); + + return 0; } -void npc_cn20k_copy_mcam_entry(struct rvu *rvu, int blkaddr, u16 src, u16 dest) +int npc_cn20k_copy_mcam_entry(struct rvu *rvu, int blkaddr, u16 src, u16 dest) { struct npc_mcam *mcam = &rvu->hw->mcam; u64 cfg, sreg, dreg, soff, doff; @@ -1132,10 +1141,15 @@ void npc_cn20k_copy_mcam_entry(struct rvu *rvu, int blkaddr, u16 src, u16 dest) dbank = npc_get_bank(mcam, dest); sbank = npc_get_bank(mcam, src); - npc_mcam_idx_2_key_type(rvu, src, &src_kwtype); - npc_mcam_idx_2_key_type(rvu, dest, &dest_kwtype); + + if (npc_mcam_idx_2_key_type(rvu, src, &src_kwtype)) + return -EINVAL; + + if (npc_mcam_idx_2_key_type(rvu, dest, &dest_kwtype)) + return -EINVAL; + if (src_kwtype != dest_kwtype) - return; + return -EINVAL; src &= (mcam->banksize - 1); dest &= (mcam->banksize - 1); @@ -1170,6 +1184,8 @@ void npc_cn20k_copy_mcam_entry(struct rvu *rvu, int blkaddr, u16 src, u16 dest) if (src_kwtype == NPC_MCAM_KEY_X2) break; } + + return 0; } static void npc_cn20k_fill_entryword(struct cn20k_mcam_entry *entry, int idx, @@ -1179,16 +1195,17 @@ static void npc_cn20k_fill_entryword(struct cn20k_mcam_entry *entry, int idx, entry->kw_mask[idx] = cam1 ^ cam0; } -void npc_cn20k_read_mcam_entry(struct rvu *rvu, int blkaddr, u16 index, - struct cn20k_mcam_entry *entry, - u8 *intf, u8 *ena, u8 *hw_prio) +int npc_cn20k_read_mcam_entry(struct rvu *rvu, int blkaddr, u16 index, + struct cn20k_mcam_entry *entry, + u8 *intf, u8 *ena, u8 *hw_prio) { struct npc_mcam *mcam = &rvu->hw->mcam; u64 cam0, cam1, bank_cfg, cfg; int kw = 0, bank; u8 kw_type; - npc_mcam_idx_2_key_type(rvu, index, &kw_type); + if (npc_mcam_idx_2_key_type(rvu, index, &kw_type)) + return -EINVAL; bank = npc_get_bank(mcam, index); index &= (mcam->banksize - 1); @@ -1298,6 +1315,8 @@ void npc_cn20k_read_mcam_entry(struct rvu *rvu, int blkaddr, u16 index, cfg = rvu_read64(rvu, blkaddr, NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(index, 0, 1)); entry->vtag_action = cfg; + + return 0; } int rvu_mbox_handler_npc_cn20k_mcam_write_entry(struct rvu *rvu, @@ -1335,11 +1354,10 @@ int rvu_mbox_handler_npc_cn20k_mcam_write_entry(struct rvu *rvu, if (is_pffunc_af(req->hdr.pcifunc)) nix_intf = req->intf; - npc_cn20k_config_mcam_entry(rvu, blkaddr, req->entry, nix_intf, - &req->entry_data, req->enable_entry, - req->hw_prio, req->req_kw_type); + rc = npc_cn20k_config_mcam_entry(rvu, blkaddr, req->entry, nix_intf, + &req->entry_data, req->enable_entry, + req->hw_prio, req->req_kw_type); - rc = 0; exit: mutex_unlock(&mcam->lock); return rc; @@ -1361,11 +1379,13 @@ int rvu_mbox_handler_npc_cn20k_mcam_read_entry(struct rvu *rvu, mutex_lock(&mcam->lock); rc = npc_mcam_verify_entry(mcam, pcifunc, req->entry); - if (!rc) - npc_cn20k_read_mcam_entry(rvu, blkaddr, req->entry, - &rsp->entry_data, &rsp->intf, - &rsp->enable, &rsp->hw_prio); + if (rc) + goto fail; + rc = npc_cn20k_read_mcam_entry(rvu, blkaddr, req->entry, + &rsp->entry_data, &rsp->intf, + &rsp->enable, &rsp->hw_prio); +fail: mutex_unlock(&mcam->lock); return rc; } @@ -1375,11 +1395,13 @@ int rvu_mbox_handler_npc_cn20k_mcam_alloc_and_write_entry(struct rvu *rvu, struct npc_mcam_alloc_and_write_entry_rsp *rsp) { struct rvu_pfvf *pfvf = rvu_get_pfvf(rvu, req->hdr.pcifunc); + struct npc_mcam_free_entry_req free_req = { 0 }; struct npc_mcam_alloc_entry_req entry_req; struct npc_mcam_alloc_entry_rsp entry_rsp; struct npc_mcam *mcam = &rvu->hw->mcam; u16 entry = NPC_MCAM_ENTRY_INVALID; - int blkaddr, rc; + struct msg_rsp free_rsp; + int blkaddr, rc, err; u8 nix_intf; blkaddr = rvu_get_blkaddr(rvu, BLKTYPE_NPC, 0); @@ -1415,12 +1437,23 @@ int rvu_mbox_handler_npc_cn20k_mcam_alloc_and_write_entry(struct rvu *rvu, else nix_intf = pfvf->nix_rx_intf; - npc_cn20k_config_mcam_entry(rvu, blkaddr, entry, nix_intf, - &req->entry_data, req->enable_entry, - req->hw_prio, req->req_kw_type); + rc = npc_cn20k_config_mcam_entry(rvu, blkaddr, entry, nix_intf, + &req->entry_data, req->enable_entry, + req->hw_prio, req->req_kw_type); mutex_unlock(&mcam->lock); + if (rc) { + free_req.hdr.pcifunc = req->hdr.pcifunc; + free_req.entry = entry_rsp.entry; + err = rvu_mbox_handler_npc_mcam_free_entry(rvu, &free_req, &free_rsp); + if (err) + dev_err(rvu->dev, + "%s: Error to free mcam idx %u\n", + __func__, entry_rsp.entry); + return rc; + } + rsp->entry = entry_rsp.entry; return 0; } @@ -1480,9 +1513,9 @@ int rvu_mbox_handler_npc_cn20k_read_base_steer_rule(struct rvu *rvu, read_entry: /* Read the mcam entry */ - npc_cn20k_read_mcam_entry(rvu, blkaddr, index, - &rsp->entry, &intf, - &enable, &hw_prio); + rc = npc_cn20k_read_mcam_entry(rvu, blkaddr, index, + &rsp->entry, &intf, + &enable, &hw_prio); mutex_unlock(&mcam->lock); out: return rc; @@ -3607,9 +3640,30 @@ int npc_defrag_move_vdx_to_free(struct rvu *rvu, NPC_AF_CN20K_MCAMEX_BANKX_STAT_EXT(midx, bank)); - npc_cn20k_enable_mcam_entry(rvu, blkaddr, old_midx, false); - npc_cn20k_copy_mcam_entry(rvu, blkaddr, old_midx, new_midx); - npc_cn20k_enable_mcam_entry(rvu, blkaddr, new_midx, true); + /* If bug happened during copy/enable mcam, then there is a bug in allocation + * algorithm itself. There is no point in rewinding and returning, as it + * will face further issue. Return error after printing error + */ + if (npc_cn20k_enable_mcam_entry(rvu, blkaddr, old_midx, false)) { + dev_err(rvu->dev, + "%s: Error happened while disabling old_mid=%u\n", + __func__, old_midx); + return -EFAULT; + } + + if (npc_cn20k_copy_mcam_entry(rvu, blkaddr, old_midx, new_midx)) { + dev_err(rvu->dev, + "%s: Error happened while copying old_midx=%u new_midx=%u\n", + __func__, old_midx, new_midx); + return -EFAULT; + } + + if (npc_cn20k_enable_mcam_entry(rvu, blkaddr, new_midx, true)) { + dev_err(rvu->dev, + "%s: Error happened while enabling new_mid=%u\n", + __func__, new_midx); + return -EFAULT; + } midx = new_midx % mcam->banksize; bank = new_midx / mcam->banksize; diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h index 815d0b257a7e..8f3eea9cfb1d 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h @@ -320,16 +320,16 @@ void npc_cn20k_dft_rules_free(struct rvu *rvu, u16 pcifunc); int npc_cn20k_dft_rules_idx_get(struct rvu *rvu, u16 pcifunc, u16 *bcast, u16 *mcast, u16 *promisc, u16 *ucast); -void npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, - u8 intf, struct cn20k_mcam_entry *entry, - bool enable, u8 hw_prio, u8 req_kw_type); -void npc_cn20k_enable_mcam_entry(struct rvu *rvu, int blkaddr, - int index, bool enable); -void npc_cn20k_copy_mcam_entry(struct rvu *rvu, int blkaddr, - u16 src, u16 dest); -void npc_cn20k_read_mcam_entry(struct rvu *rvu, int blkaddr, u16 index, - struct cn20k_mcam_entry *entry, u8 *intf, - u8 *ena, u8 *hw_prio); +int npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, + u8 intf, struct cn20k_mcam_entry *entry, + bool enable, u8 hw_prio, u8 req_kw_type); +int npc_cn20k_enable_mcam_entry(struct rvu *rvu, int blkaddr, + int index, bool enable); +int npc_cn20k_copy_mcam_entry(struct rvu *rvu, int blkaddr, + u16 src, u16 dest); +int npc_cn20k_read_mcam_entry(struct rvu *rvu, int blkaddr, u16 index, + struct cn20k_mcam_entry *entry, u8 *intf, + u8 *ena, u8 *hw_prio); void npc_cn20k_clear_mcam_entry(struct rvu *rvu, int blkaddr, int bank, int index); int npc_mcam_idx_2_key_type(struct rvu *rvu, u16 mcam_idx, u8 *key_type); diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c index c2ca5ed1d028..ecaf0946b852 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c @@ -241,7 +241,10 @@ void npc_enable_mcam_entry(struct rvu *rvu, struct npc_mcam *mcam, if (index < 0 || index >= mcam->banksize * mcam->banks) return; - return npc_cn20k_enable_mcam_entry(rvu, blkaddr, index, enable); + if (npc_cn20k_enable_mcam_entry(rvu, blkaddr, index, enable)) + dev_err(rvu->dev, "Error to %s mcam %u entry\n", + enable ? "enable" : "disable", index); + return; } index &= (mcam->banksize - 1); @@ -589,8 +592,8 @@ void npc_read_mcam_entry(struct rvu *rvu, struct npc_mcam *mcam, NPC_AF_MCAMEX_BANKX_CFG(src, sbank)) & 1; } -static void npc_copy_mcam_entry(struct rvu *rvu, struct npc_mcam *mcam, - int blkaddr, u16 src, u16 dest) +static int npc_copy_mcam_entry(struct rvu *rvu, struct npc_mcam *mcam, + int blkaddr, u16 src, u16 dest) { int dbank = npc_get_bank(mcam, dest); int sbank = npc_get_bank(mcam, src); @@ -630,6 +633,7 @@ static void npc_copy_mcam_entry(struct rvu *rvu, struct npc_mcam *mcam, NPC_AF_MCAMEX_BANKX_CFG(src, sbank)); rvu_write64(rvu, blkaddr, NPC_AF_MCAMEX_BANKX_CFG(dest, dbank), cfg); + return 0; } u64 npc_get_mcam_action(struct rvu *rvu, struct npc_mcam *mcam, @@ -3266,7 +3270,10 @@ int rvu_mbox_handler_npc_mcam_shift_entry(struct rvu *rvu, npc_enable_mcam_entry(rvu, mcam, blkaddr, new_entry, false); /* Copy rule from old entry to new entry */ - npc_copy_mcam_entry(rvu, mcam, blkaddr, old_entry, new_entry); + if (npc_copy_mcam_entry(rvu, mcam, blkaddr, old_entry, new_entry)) { + rc = NPC_MCAM_INVALID_REQ; + break; + } /* Copy counter mapping, if any */ cntr = mcam->entry2cntr_map[old_entry]; @@ -3284,7 +3291,8 @@ int rvu_mbox_handler_npc_mcam_shift_entry(struct rvu *rvu, /* If shift has failed then report the failed index */ if (index != req->shift_count) { - rc = NPC_MCAM_PERM_DENIED; + if (!rc) + rc = NPC_MCAM_PERM_DENIED; rsp->failed_entry_idx = index; } diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc_fs.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc_fs.c index b45798d9fdab..fe10554b1f0e 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc_fs.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc_fs.c @@ -1980,13 +1980,15 @@ static int npc_update_dmac_value(struct rvu *rvu, int npcblkaddr, ether_addr_copy(rule->packet.dmac, pfvf->mac_addr); - if (is_cn20k(rvu->pdev)) - npc_cn20k_read_mcam_entry(rvu, npcblkaddr, rule->entry, - cn20k_entry, &intf, - &enable, &hw_prio); - else + if (is_cn20k(rvu->pdev)) { + if (npc_cn20k_read_mcam_entry(rvu, npcblkaddr, rule->entry, + cn20k_entry, &intf, + &enable, &hw_prio)) + return -EINVAL; + } else { npc_read_mcam_entry(rvu, mcam, npcblkaddr, rule->entry, entry, &intf, &enable); + } npc_update_entry(rvu, NPC_DMAC, &mdata, ether_addr_to_u64(pfvf->mac_addr), 0, @@ -2038,8 +2040,12 @@ void npc_mcam_enable_flows(struct rvu *rvu, u16 target) continue; } - if (rule->vfvlan_cfg) - npc_update_dmac_value(rvu, blkaddr, rule, pfvf); + if (rule->vfvlan_cfg) { + if (npc_update_dmac_value(rvu, blkaddr, rule, pfvf)) + dev_err(rvu->dev, + "Update dmac failed for %u, target=%#x\n", + rule->entry, target); + } if (rule->rx_action.op == NIX_RX_ACTION_DEFAULT) { if (!def_ucast_rule) From 1100af13fd14b523f1b0634c14be497b41c78958 Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 29 Apr 2026 07:57:14 +0530 Subject: [PATCH 138/525] octeontx2-af: npc: cn20k: Drop debugfs_create_file() error checks in init debugfs is not intended to be checked for allocation failures the way other kernel APIs are: callers should not fail probe or subsystem init because a debugfs node could not be created, including when debugfs is disabled in Kconfig. Replacing NULL checks with IS_ERR() checks is similarly wrong for optional debugfs. Remove dentry checks and -EFAULT returns from npc_cn20k_debugfs_init(). See: https://staticthinking.wordpress.com/2023/07/24/ debugfs-functions-are-not-supposed-to-be-checked/ Cc: Dan Carpenter Fixes: 528530dff56b ("octeontx2-af: npc: cn20k: add debugfs support") Link: https://lore.kernel.org/netdev/adjNGPWKMOk3KgWL@stanley.mountain/ Reviewed-by: Simon Horman Signed-off-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260429022722.1110289-3-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../marvell/octeontx2/af/cn20k/debugfs.c | 33 ++++++------------- 1 file changed, 10 insertions(+), 23 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/debugfs.c b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/debugfs.c index 3debf2fae1a4..6f13296303cb 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/debugfs.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/debugfs.c @@ -249,34 +249,21 @@ DEFINE_SHOW_ATTRIBUTE(npc_defrag); int npc_cn20k_debugfs_init(struct rvu *rvu) { struct npc_priv_t *npc_priv = npc_priv_get(); - struct dentry *npc_dentry; - npc_dentry = debugfs_create_file("mcam_layout", 0444, rvu->rvu_dbg.npc, - npc_priv, &npc_mcam_layout_fops); + debugfs_create_file("mcam_layout", 0444, rvu->rvu_dbg.npc, + npc_priv, &npc_mcam_layout_fops); - if (!npc_dentry) - return -EFAULT; + debugfs_create_file("mcam_default", 0444, rvu->rvu_dbg.npc, + rvu, &npc_mcam_default_fops); - npc_dentry = debugfs_create_file("mcam_default", 0444, rvu->rvu_dbg.npc, - rvu, &npc_mcam_default_fops); + debugfs_create_file("vidx2idx", 0444, rvu->rvu_dbg.npc, + npc_priv, &npc_vidx2idx_map_fops); - if (!npc_dentry) - return -EFAULT; + debugfs_create_file("idx2vidx", 0444, rvu->rvu_dbg.npc, + npc_priv, &npc_idx2vidx_map_fops); - npc_dentry = debugfs_create_file("vidx2idx", 0444, rvu->rvu_dbg.npc, - npc_priv, &npc_vidx2idx_map_fops); - if (!npc_dentry) - return -EFAULT; - - npc_dentry = debugfs_create_file("idx2vidx", 0444, rvu->rvu_dbg.npc, - npc_priv, &npc_idx2vidx_map_fops); - if (!npc_dentry) - return -EFAULT; - - npc_dentry = debugfs_create_file("defrag", 0444, rvu->rvu_dbg.npc, - npc_priv, &npc_defrag_fops); - if (!npc_dentry) - return -EFAULT; + debugfs_create_file("defrag", 0444, rvu->rvu_dbg.npc, + npc_priv, &npc_defrag_fops); return 0; } From adb5ff41efbc0a9d86fabf880076973379db6e49 Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 29 Apr 2026 07:57:15 +0530 Subject: [PATCH 139/525] octeontx2-af: npc: cn20k: Propagate errors in defrag MCAM alloc rollback npc_defrag_alloc_free_slots() allocates MCAM indexes in up to two passes on bank0 then bank1. On failure it rolls back by freeing entries already placed in save[]. __npc_subbank_alloc() can return a negative errno while only part of the indexes are valid. The rollback loop used rc for npc_mcam_idx_2_subbank_idx() as well, so a successful lookup stored zero in rc and a later __npc_subbank_free() failure could still end with return 0 when the allocation path had also left rc at zero (for example shortfall after zero return values from the alloc helpers). Jump to the rollback path immediately when either __npc_subbank_alloc() call fails, preserving its errno. If both calls succeed but the total allocated count is still less than cnt, set rc to -ENOSPC before rollback. Use a separate err variable for npc_mcam_idx_2_subbank_idx() so a successful lookup no longer clears a non-zero rc from the allocation phase. Cc: Dan Carpenter Fixes: 645c6e3c1999 ("octeontx2-af: npc: cn20k: virtual index support") Link: https://lore.kernel.org/netdev/adjNJEpILRZATB2N@stanley.mountain/ Reviewed-by: Simon Horman Signed-off-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260429022722.1110289-4-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/marvell/octeontx2/af/cn20k/npc.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c index 7170dcf26200..87da43088b67 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c @@ -2338,6 +2338,7 @@ static int __npc_subbank_alloc(struct rvu *rvu, struct npc_subbank *sb, __npc_subbank_mark_free(rvu, sb); err1: kfree(save); + *alloc_cnt = 0; return rc; } @@ -3515,7 +3516,7 @@ static int npc_defrag_alloc_free_slots(struct rvu *rvu, { int alloc_cnt1, alloc_cnt2; struct npc_subbank *sb; - int rc, sb_off, i; + int rc, sb_off, i, err; bool deleted; sb = &npc_priv.sb[f->idx]; @@ -3529,6 +3530,7 @@ static int npc_defrag_alloc_free_slots(struct rvu *rvu, NPC_MCAM_LOWER_PRIO, false, cnt, save, cnt, true, &alloc_cnt1); + if (alloc_cnt1 < cnt) { rc = __npc_subbank_alloc(rvu, sb, NPC_MCAM_KEY_X2, sb->b1b, @@ -3544,15 +3546,17 @@ static int npc_defrag_alloc_free_slots(struct rvu *rvu, dev_err(rvu->dev, "%s: Failed to alloc cnt=%u alloc_cnt1=%u alloc_cnt2=%u\n", __func__, cnt, alloc_cnt1, alloc_cnt2); + rc = -ENOSPC; goto fail_free_alloc; } + return 0; fail_free_alloc: for (i = 0; i < alloc_cnt1 + alloc_cnt2; i++) { - rc = npc_mcam_idx_2_subbank_idx(rvu, save[i], - &sb, &sb_off); - if (rc) { + err = npc_mcam_idx_2_subbank_idx(rvu, save[i], + &sb, &sb_off); + if (err) { dev_err(rvu->dev, "%s: Error to find subbank for mcam idx=%u\n", __func__, save[i]); From d7e5940c4c508df73b15d9bc29628a83b3674fff Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 29 Apr 2026 07:57:16 +0530 Subject: [PATCH 140/525] octeontx2-af: npc: cn20k: Fix target map and rule npc_defrag_move_vdx_to_free() disables, copies, and enables the MCAM entry at a new index but previously left entry2target_pffunc[] and the mcam_rules list still keyed to the old index. Copy the target PF association to the new slot, clear the old one, and retarget the rule entry so software state matches the relocated hardware context. Fixes: 645c6e3c1999 ("octeontx2-af: npc: cn20k: virtual index support") Signed-off-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260429022722.1110289-5-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../ethernet/marvell/octeontx2/af/cn20k/npc.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c index 87da43088b67..70ce3f49adc1 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c @@ -3602,9 +3602,10 @@ int npc_defrag_move_vdx_to_free(struct rvu *rvu, struct npc_defrag_node *v, int cnt, u16 *save) { + u16 new_midx, old_midx, vidx, target_pf; struct npc_mcam *mcam = &rvu->hw->mcam; + struct rvu_npc_mcam_rule *rule, *tmp; int i, vidx_cnt, rc, sb_off; - u16 new_midx, old_midx, vidx; struct npc_subbank *sb; bool deleted; u16 pcifunc; @@ -3723,8 +3724,21 @@ int npc_defrag_move_vdx_to_free(struct rvu *rvu, mcam->entry2pfvf_map[new_midx] = pcifunc; /* Counter is not preserved */ mcam->entry2cntr_map[new_midx] = new_midx; + target_pf = mcam->entry2target_pffunc[old_midx]; + mcam->entry2target_pffunc[new_midx] = target_pf; + mcam->entry2target_pffunc[old_midx] = NPC_MCAM_INVALID_MAP; + npc_mcam_set_bit(mcam, new_midx); + /* Note: list order is not functionally required for mcam_rules */ + list_for_each_entry_safe(rule, tmp, &mcam->mcam_rules, list) { + if (rule->entry != old_midx) + continue; + + rule->entry = new_midx; + break; + } + /* Mark as invalid */ v->vidx[vidx_cnt - i - 1] = -1; save[cnt - i - 1] = -1; From d2dabf09632c84b7acdc0fb2eeb6b6fe9c0f9106 Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 29 Apr 2026 07:57:17 +0530 Subject: [PATCH 141/525] octeontx2-af: npc: cn20k: Clear MCAM entries by index and key width Replace the old four-argument CN20K MCAM clear with a per-bank static helper and npc_cn20k_clear_mcam_entry() that takes a logical MCAM index, resolves the key width via npc_mcam_idx_2_key_type(), and clears either one bank (X2) or every bank (X4). Call it from npc_clear_mcam_entry() on cn20k and log when key-type lookup fails. Use the per-bank helper from npc_cn20k_config_mcam_entry() for pre-program clears. For loopback VFs, use the promisc MCAM index as ucast_idx when copying RSS action for promisc, matching cn20k default-rule layout. Cc: Suman Ghosh Fixes: 6d1e70282f76 ("octeontx2-af: npc: cn20k: Use common APIs") Signed-off-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260429022722.1110289-6-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../ethernet/marvell/octeontx2/af/cn20k/npc.c | 37 ++++++++++++++++--- .../ethernet/marvell/octeontx2/af/cn20k/npc.h | 3 +- .../ethernet/marvell/octeontx2/af/rvu_npc.c | 17 ++++++++- 3 files changed, 48 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c index 70ce3f49adc1..112c37c190b1 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c @@ -842,8 +842,8 @@ npc_cn20k_enable_mcam_entry(struct rvu *rvu, int blkaddr, return 0; } -void -npc_cn20k_clear_mcam_entry(struct rvu *rvu, int blkaddr, int bank, int index) +static void +npc_clear_x2_entry(struct rvu *rvu, int blkaddr, int bank, int index) { rvu_write64(rvu, blkaddr, NPC_AF_CN20K_MCAMEX_BANKX_CAMX_INTF_EXT(index, bank, 1), @@ -877,6 +877,33 @@ npc_cn20k_clear_mcam_entry(struct rvu *rvu, int blkaddr, int bank, int index) NPC_AF_CN20K_MCAMEX_BANKX_STAT_EXT(index, bank), 0); } +int +npc_cn20k_clear_mcam_entry(struct rvu *rvu, int blkaddr, int mcam_idx) +{ + struct npc_mcam *mcam = &rvu->hw->mcam; + int bank = npc_get_bank(mcam, mcam_idx); + u8 kw_type; + int index; + + if (npc_mcam_idx_2_key_type(rvu, mcam_idx, &kw_type)) + return -EINVAL; + + index = mcam_idx & (mcam->banksize - 1); + + if (kw_type == NPC_MCAM_KEY_X2) { + npc_clear_x2_entry(rvu, blkaddr, bank, index); + return 0; + } + + /* For NPC_MCAM_KEY_X4 keys, both the banks + * need to be programmed with the same value. + */ + for (bank = 0; bank < mcam->banks_per_entry; bank++) + npc_clear_x2_entry(rvu, blkaddr, bank, index); + + return 0; +} + static void npc_cn20k_get_keyword(struct cn20k_mcam_entry *entry, int idx, u64 *cam0, u64 *cam1) { @@ -1071,7 +1098,7 @@ int npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, */ if (kw_type == NPC_MCAM_KEY_X2) { /* Clear mcam entry to avoid writes being suppressed by NPC */ - npc_cn20k_clear_mcam_entry(rvu, blkaddr, bank, mcam_idx); + npc_clear_x2_entry(rvu, blkaddr, bank, mcam_idx); npc_cn20k_config_kw_x2(rvu, mcam, blkaddr, mcam_idx, intf, entry, bank, kw_type, kw, req_kw_type); @@ -1096,8 +1123,8 @@ int npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, } /* Clear mcam entry to avoid writes being suppressed by NPC */ - npc_cn20k_clear_mcam_entry(rvu, blkaddr, 0, mcam_idx); - npc_cn20k_clear_mcam_entry(rvu, blkaddr, 1, mcam_idx); + npc_clear_x2_entry(rvu, blkaddr, 0, mcam_idx); + npc_clear_x2_entry(rvu, blkaddr, 1, mcam_idx); npc_cn20k_config_kw_x4(rvu, mcam, blkaddr, mcam_idx, intf, entry, diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h index 8f3eea9cfb1d..2f761b97f91b 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h @@ -330,8 +330,7 @@ int npc_cn20k_copy_mcam_entry(struct rvu *rvu, int blkaddr, int npc_cn20k_read_mcam_entry(struct rvu *rvu, int blkaddr, u16 index, struct cn20k_mcam_entry *entry, u8 *intf, u8 *ena, u8 *hw_prio); -void npc_cn20k_clear_mcam_entry(struct rvu *rvu, int blkaddr, - int bank, int index); +int npc_cn20k_clear_mcam_entry(struct rvu *rvu, int blkaddr, int index); int npc_mcam_idx_2_key_type(struct rvu *rvu, u16 mcam_idx, u8 *key_type); u16 npc_cn20k_vidx2idx(u16 index); u16 npc_cn20k_idx2vidx(u16 idx); diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c index ecaf0946b852..44ca65efc80f 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c @@ -261,6 +261,13 @@ static void npc_clear_mcam_entry(struct rvu *rvu, struct npc_mcam *mcam, int bank = npc_get_bank(mcam, index); int actbank = bank; + if (is_cn20k(rvu->pdev)) { + if (npc_cn20k_clear_mcam_entry(rvu, blkaddr, index)) + dev_err(rvu->dev, "%s Failed to clear mcam %u\n", + __func__, index); + return; + } + index &= (mcam->banksize - 1); for (; bank < (actbank + mcam->banks_per_entry); bank++) { rvu_write64(rvu, blkaddr, @@ -755,9 +762,15 @@ void rvu_npc_install_promisc_entry(struct rvu *rvu, u16 pcifunc, /* If the corresponding PF's ucast action is RSS, * use the same action for promisc also + * Please note that for lbk(s) "index" and "ucast_idx" + * will be same. */ - ucast_idx = npc_get_nixlf_mcam_index(mcam, pcifunc, - nixlf, NIXLF_UCAST_ENTRY); + if (is_lbk_vf(rvu, pcifunc)) + ucast_idx = index; + else + ucast_idx = npc_get_nixlf_mcam_index(mcam, pcifunc, + nixlf, NIXLF_UCAST_ENTRY); + if (is_mcam_entry_enabled(rvu, mcam, blkaddr, ucast_idx)) *(u64 *)&action = npc_get_mcam_action(rvu, mcam, blkaddr, ucast_idx); From 2b6d6bb7282c34dd8c04ee782393231acf5a26e2 Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 29 Apr 2026 07:57:18 +0530 Subject: [PATCH 142/525] octeontx2-af: npc: cn20k: Fix bank value For X4 keys its loop reused the bank parameter as the loop counter, so bank no longer reflected the caller's bank after the loop and the control flow was hard to follow. Program NPC_AF_CN20K_MCAMEX_BANKX_CFG_EXT directly in npc_cn20k_config_mcam_entry(): one CFG write for X2 using the computed bank, and one CFG write per bank inside the X4 action loop. Enable the entry at the end with npc_cn20k_enable_mcam_entry(..., true) instead of embedding the enable bit in bank_cfg via the removed helper. Cc: Suman Ghosh Fixes: 4e527f1e5c15 ("octeontx2-af: npc: cn20k: Add new mailboxes for CN20K silicon") Signed-off-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260429022722.1110289-7-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../ethernet/marvell/octeontx2/af/cn20k/npc.c | 100 +++++++----------- 1 file changed, 41 insertions(+), 59 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c index 112c37c190b1..4773277fd409 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c @@ -1045,34 +1045,6 @@ static void npc_cn20k_config_kw_x4(struct rvu *rvu, struct npc_mcam *mcam, kw, req_kw_type); } -static void -npc_cn20k_set_mcam_bank_cfg(struct rvu *rvu, int blkaddr, int mcam_idx, - int bank, u8 kw_type, bool enable, u8 hw_prio) -{ - struct npc_mcam *mcam = &rvu->hw->mcam; - u64 bank_cfg; - - bank_cfg = (u64)hw_prio << 24; - if (enable) - bank_cfg |= 0x1; - - if (kw_type == NPC_MCAM_KEY_X2) { - rvu_write64(rvu, blkaddr, - NPC_AF_CN20K_MCAMEX_BANKX_CFG_EXT(mcam_idx, bank), - bank_cfg); - return; - } - - /* For NPC_MCAM_KEY_X4 keys, both the banks - * need to be programmed with the same value. - */ - for (bank = 0; bank < mcam->banks_per_entry; bank++) { - rvu_write64(rvu, blkaddr, - NPC_AF_CN20K_MCAMEX_BANKX_CFG_EXT(mcam_idx, bank), - bank_cfg); - } -} - int npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, u8 intf, struct cn20k_mcam_entry *entry, bool enable, u8 hw_prio, u8 req_kw_type) @@ -1080,6 +1052,7 @@ int npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, struct npc_mcam *mcam = &rvu->hw->mcam; int mcam_idx = index % mcam->banksize; int bank = index / mcam->banksize; + u64 bank_cfg = (u64)hw_prio << 24; int kw = 0; u8 kw_type; @@ -1119,41 +1092,50 @@ int npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(mcam_idx, bank, 1), entry->vtag_action); - goto set_cfg; + + /* Set HW priority */ + rvu_write64(rvu, blkaddr, + NPC_AF_CN20K_MCAMEX_BANKX_CFG_EXT(mcam_idx, bank), + bank_cfg); + + } else { + /* Clear mcam entry to avoid writes being suppressed by NPC */ + npc_clear_x2_entry(rvu, blkaddr, 0, mcam_idx); + npc_clear_x2_entry(rvu, blkaddr, 1, mcam_idx); + + npc_cn20k_config_kw_x4(rvu, mcam, blkaddr, + mcam_idx, intf, entry, + kw_type, req_kw_type); + for (bank = 0; bank < mcam->banks_per_entry; bank++) { + /* Set 'action' */ + rvu_write64(rvu, blkaddr, + NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(mcam_idx, + bank, 0), + entry->action); + + /* Set TAG 'action' */ + rvu_write64(rvu, blkaddr, + NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(mcam_idx, + bank, 1), + entry->vtag_action); + + /* Set 'action2' for inline receive */ + rvu_write64(rvu, blkaddr, + NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(mcam_idx, + bank, 2), + entry->action2); + + /* Set HW priority */ + rvu_write64(rvu, blkaddr, + NPC_AF_CN20K_MCAMEX_BANKX_CFG_EXT(mcam_idx, bank), + bank_cfg); + } } - /* Clear mcam entry to avoid writes being suppressed by NPC */ - npc_clear_x2_entry(rvu, blkaddr, 0, mcam_idx); - npc_clear_x2_entry(rvu, blkaddr, 1, mcam_idx); - - npc_cn20k_config_kw_x4(rvu, mcam, blkaddr, - mcam_idx, intf, entry, - kw_type, req_kw_type); - for (bank = 0; bank < mcam->banks_per_entry; bank++) { - /* Set 'action' */ - rvu_write64(rvu, blkaddr, - NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(mcam_idx, - bank, 0), - entry->action); - - /* Set TAG 'action' */ - rvu_write64(rvu, blkaddr, - NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(mcam_idx, - bank, 1), - entry->vtag_action); - - /* Set 'action2' for inline receive */ - rvu_write64(rvu, blkaddr, - NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(mcam_idx, - bank, 2), - entry->action2); - } - -set_cfg: /* TODO: */ /* PF installing VF rule */ - npc_cn20k_set_mcam_bank_cfg(rvu, blkaddr, mcam_idx, bank, - kw_type, enable, hw_prio); + if (npc_cn20k_enable_mcam_entry(rvu, blkaddr, index, enable)) + return -EINVAL; return 0; } From f6803eb070bfb9a5114d16ae15053106bc7842ae Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 29 Apr 2026 07:57:19 +0530 Subject: [PATCH 143/525] octeontx2-af: npc: cn20k: Fix MCAM actions read npc_cn20k_read_mcam_entry() always reloaded action and vtag_action from bank 0 after programming the CAM words. Use the bank returned by npc_get_bank() for the ACTION reads as well, and read those registers once up front so both X2 and X4 paths share the same metadata. Return directly from the X2 keyword path now that the action fields are already populated. Cc: Suman Ghosh Fixes: 6d1e70282f76 ("octeontx2-af: npc: cn20k: Use common APIs") Signed-off-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260429022722.1110289-8-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../ethernet/marvell/octeontx2/af/cn20k/npc.c | 26 +++++++++---------- 1 file changed, 13 insertions(+), 13 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c index 4773277fd409..bb0a9ac7aab3 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c @@ -1219,6 +1219,18 @@ int npc_cn20k_read_mcam_entry(struct rvu *rvu, int blkaddr, u16 index, bank = npc_get_bank(mcam, index); index &= (mcam->banksize - 1); + cfg = rvu_read64(rvu, blkaddr, + NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(index, bank, 0)); + entry->action = cfg; + + cfg = rvu_read64(rvu, blkaddr, + NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(index, bank, 1)); + entry->vtag_action = cfg; + + cfg = rvu_read64(rvu, blkaddr, + NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(index, bank, 2)); + entry->action2 = cfg; + cfg = rvu_read64(rvu, blkaddr, NPC_AF_CN20K_MCAMEX_BANKX_CAMX_INTF_EXT(index, bank, 1)) & 3; @@ -1268,7 +1280,7 @@ int npc_cn20k_read_mcam_entry(struct rvu *rvu, int blkaddr, u16 index, bank, 0)); npc_cn20k_fill_entryword(entry, kw + 3, cam0, cam1); - goto read_action; + return 0; } for (bank = 0; bank < mcam->banks_per_entry; bank++, kw = kw + 4) { @@ -1313,18 +1325,6 @@ int npc_cn20k_read_mcam_entry(struct rvu *rvu, int blkaddr, u16 index, npc_cn20k_fill_entryword(entry, kw + 3, cam0, cam1); } -read_action: - /* 'action' is set to same value for both bank '0' and '1'. - * Hence, reading bank '0' should be enough. - */ - cfg = rvu_read64(rvu, blkaddr, - NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(index, 0, 0)); - entry->action = cfg; - - cfg = rvu_read64(rvu, blkaddr, - NPC_AF_CN20K_MCAMEX_BANKX_ACTIONX_EXT(index, 0, 1)); - entry->vtag_action = cfg; - return 0; } From afb474bd4ffc314de766afc295ac64b42856f48e Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 29 Apr 2026 07:57:20 +0530 Subject: [PATCH 144/525] octeontx2-af: npc: cn20k: Initialize default-rule index outputs up front npc_cn20k_dft_rules_idx_get() wrote USHRT_MAX into individual outputs only on some error paths (lbk promisc lookup, VF ucast lookup, and the PF rule walk), which could leave other caller slots stale across retries. Set every non-NULL bcast/mcast/promisc/ucast pointer to USHRT_MAX once at entry, then drop the duplicate assignments on failure. Successful lookups still overwrite the relevant slot before returning. Fixes: 09d3b7a1403f ("octeontx2-af: npc: cn20k: Allocate default MCAM indexes") Signed-off-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260429022722.1110289-9-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c index bb0a9ac7aab3..b3f34b84c114 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c @@ -4016,6 +4016,13 @@ int npc_cn20k_dft_rules_idx_get(struct rvu *rvu, u16 pcifunc, u16 *bcast, void *val; int i, j; + for (i = 0; i < ARRAY_SIZE(ptr); i++) { + if (!ptr[i]) + continue; + + *ptr[i] = USHRT_MAX; + } + if (!npc_priv.init_done) return 0; @@ -4031,7 +4038,6 @@ int npc_cn20k_dft_rules_idx_get(struct rvu *rvu, u16 pcifunc, u16 *bcast, npc_dft_rule_name[NPC_DFT_RULE_PROMISC_ID], pcifunc); - *ptr[0] = USHRT_MAX; return -ESRCH; } @@ -4051,7 +4057,6 @@ int npc_cn20k_dft_rules_idx_get(struct rvu *rvu, u16 pcifunc, u16 *bcast, npc_dft_rule_name[NPC_DFT_RULE_UCAST_ID], pcifunc); - *ptr[3] = USHRT_MAX; return -ESRCH; } @@ -4071,7 +4076,6 @@ int npc_cn20k_dft_rules_idx_get(struct rvu *rvu, u16 pcifunc, u16 *bcast, __func__, npc_dft_rule_name[i], pcifunc); - *ptr[j] = USHRT_MAX; continue; } From 013717353c03b65f5b00a5cefa1515b6b45777b7 Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 29 Apr 2026 07:57:21 +0530 Subject: [PATCH 145/525] octeontx2-af: npc: cn20k: Tear down default MCAM rules explicitly on free npc_cn20k_dft_rules_free() used the NPC MCAM mbox "free all" path, which does not match how cn20k tracks default-rule MCAM slots indexes. Resolve the default-rule indices, then for each valid slot clear the bitmap entry, drop the PF/VF map, disable the MCAM line, clear the target function, and npc_cn20k_idx_free(). Remove any matching software mcam_rules nodes. On hard failure from idx_free, WARN and stop so the box stays up for analysis. In npc_mcam_free_all_entries(), prefetch the same default-rule indices and, on cn20k, skip bitmap clear and idx_free when the scanned entry is one of those reserved defaults (they are released by npc_cn20k_dft_rules_free). Fixes: 09d3b7a1403f ("octeontx2-af: npc: cn20k: Allocate default MCAM indexes") Signed-off-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260429022722.1110289-10-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../ethernet/marvell/octeontx2/af/cn20k/npc.c | 51 ++++++++++++---- .../ethernet/marvell/octeontx2/af/rvu_npc.c | 59 +++++++++++++------ 2 files changed, 82 insertions(+), 28 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c index b3f34b84c114..1129565a01bd 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c @@ -4178,11 +4178,11 @@ static bool npc_is_cgx_or_lbk(struct rvu *rvu, u16 pcifunc) void npc_cn20k_dft_rules_free(struct rvu *rvu, u16 pcifunc) { - struct npc_mcam_free_entry_req free_req = { 0 }; + struct npc_mcam *mcam = &rvu->hw->mcam; + u16 ptr[4] = {[0 ... 3] = USHRT_MAX}; + struct rvu_npc_mcam_rule *rule, *tmp; unsigned long index; - struct msg_rsp rsp; - u16 ptr[4]; - int rc, i; + int blkaddr, rc, i; void *map; if (!npc_priv.init_done) @@ -4240,14 +4240,43 @@ void npc_cn20k_dft_rules_free(struct rvu *rvu, u16 pcifunc) } free_rules: + blkaddr = rvu_get_blkaddr(rvu, BLKTYPE_NPC, 0); + if (blkaddr < 0) + return; + for (int i = 0; i < 4; i++) { + if (ptr[i] == USHRT_MAX) + continue; - free_req.hdr.pcifunc = pcifunc; - free_req.all = 1; - rc = rvu_mbox_handler_npc_mcam_free_entry(rvu, &free_req, &rsp); - if (rc) - dev_err(rvu->dev, - "%s: Error deleting default entries (pcifunc=%#x\n", - __func__, pcifunc); + mutex_lock(&mcam->lock); + npc_mcam_clear_bit(mcam, ptr[i]); + mcam->entry2pfvf_map[ptr[i]] = NPC_MCAM_INVALID_MAP; + npc_cn20k_enable_mcam_entry(rvu, blkaddr, ptr[i], false); + mcam->entry2target_pffunc[ptr[i]] = 0x0; + mutex_unlock(&mcam->lock); + + rc = npc_cn20k_idx_free(rvu, &ptr[i], 1); + if (rc) { + /* Non recoverable error. Let us WARN and return. Keep system alive to + * enable debugging + */ + WARN(1, "%s Error deleting default entries (pcifunc=%#x) mcam_idx=%u\n", + __func__, pcifunc, ptr[i]); + return; + } + } + + mutex_lock(&mcam->lock); + list_for_each_entry_safe(rule, tmp, &mcam->mcam_rules, list) { + for (int i = 0; i < 4; i++) { + if (ptr[i] != rule->entry) + continue; + + list_del(&rule->list); + kfree(rule); + break; + } + } + mutex_unlock(&mcam->lock); } int npc_cn20k_dft_rules_alloc(struct rvu *rvu, u16 pcifunc) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c index 44ca65efc80f..5d349d131fdb 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c @@ -2521,33 +2521,58 @@ void npc_mcam_clear_bit(struct npc_mcam *mcam, u16 index) static void npc_mcam_free_all_entries(struct rvu *rvu, struct npc_mcam *mcam, int blkaddr, u16 pcifunc) { + u16 dft_idxs[NPC_DFT_RULE_MAX_ID] = {[0 ... NPC_DFT_RULE_MAX_ID - 1] = USHRT_MAX}; + bool cn20k_dft_rl; u16 index, cntr; int rc; + npc_cn20k_dft_rules_idx_get(rvu, pcifunc, + &dft_idxs[NPC_DFT_RULE_BCAST_ID], + &dft_idxs[NPC_DFT_RULE_MCAST_ID], + &dft_idxs[NPC_DFT_RULE_PROMISC_ID], + &dft_idxs[NPC_DFT_RULE_UCAST_ID]); + /* Scan all MCAM entries and free the ones mapped to 'pcifunc' */ for (index = 0; index < mcam->bmap_entries; index++) { - if (mcam->entry2pfvf_map[index] == pcifunc) { + if (mcam->entry2pfvf_map[index] != pcifunc) + continue; + + cn20k_dft_rl = false; + + if (is_cn20k(rvu->pdev)) { + if (dft_idxs[NPC_DFT_RULE_BCAST_ID] == index || + dft_idxs[NPC_DFT_RULE_MCAST_ID] == index || + dft_idxs[NPC_DFT_RULE_PROMISC_ID] == index || + dft_idxs[NPC_DFT_RULE_UCAST_ID] == index) { + cn20k_dft_rl = true; + } + } + + /* Disable the entry */ + npc_enable_mcam_entry(rvu, mcam, blkaddr, index, false); + + if (!cn20k_dft_rl) { mcam->entry2pfvf_map[index] = NPC_MCAM_INVALID_MAP; /* Free the entry in bitmap */ npc_mcam_clear_bit(mcam, index); - /* Disable the entry */ - npc_enable_mcam_entry(rvu, mcam, blkaddr, index, false); - - /* Update entry2counter mapping */ - cntr = mcam->entry2cntr_map[index]; - if (cntr != NPC_MCAM_INVALID_MAP) - npc_unmap_mcam_entry_and_cntr(rvu, mcam, - blkaddr, index, - cntr); mcam->entry2target_pffunc[index] = 0x0; - if (is_cn20k(rvu->pdev)) { - rc = npc_cn20k_idx_free(rvu, &index, 1); - if (rc) - dev_err(rvu->dev, - "Failed to free mcam idx=%u pcifunc=%#x\n", - index, pcifunc); - } } + + /* Update entry2counter mapping */ + cntr = mcam->entry2cntr_map[index]; + if (cntr != NPC_MCAM_INVALID_MAP) + npc_unmap_mcam_entry_and_cntr(rvu, mcam, + blkaddr, index, + cntr); + + if (!is_cn20k(rvu->pdev) || cn20k_dft_rl) + continue; + + rc = npc_cn20k_idx_free(rvu, &index, 1); + if (rc) + dev_err(rvu->dev, + "Failed to free mcam idx=%u pcifunc=%#x\n", + index, pcifunc); } } From bc968f61bf0ad4f085559e5e3d168105fdf88204 Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 29 Apr 2026 07:57:22 +0530 Subject: [PATCH 146/525] octeontx2-af: npc: cn20k: Reject missing default-rule MCAM indices When cn20k default L2 rules are not installed, npc_cn20k_dft_rules_idx_get() leaves broadcast, multicast, promiscuous, and unicast slots at USHRT_MAX. npc_get_nixlf_mcam_index() previously returned that sentinel as a valid MCAM index, so callers could program hardware with an invalid index. Return -EINVAL from the cn20k branches of npc_get_nixlf_mcam_index() when the requested slot is still USHRT_MAX. Harden cn20k NPC MCAM entry helpers to reject out-of-range indices before touching hardware. Drop the early bounds check in npc_enable_mcam_entry() for cn20k so invalid indices are validated inside npc_cn20k_enable_mcam_entry() instead of being silently ignored. In rvu_npc_update_flowkey_alg_idx(), treat negative MCAM indices like out-of-range values, and only update RSS actions for promiscuous and all-multi paths when the resolved index is non-negative. Cc: Suman Ghosh Fixes: 6d1e70282f76 ("octeontx2-af: npc: cn20k: Use common APIs") Signed-off-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260429022722.1110289-11-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../ethernet/marvell/octeontx2/af/cn20k/npc.c | 14 +- .../ethernet/marvell/octeontx2/af/cn20k/npc.h | 1 + .../ethernet/marvell/octeontx2/af/rvu_nix.c | 3 + .../ethernet/marvell/octeontx2/af/rvu_npc.c | 137 +++++++++++++++++- .../marvell/octeontx2/af/rvu_npc_fs.c | 10 +- 5 files changed, 155 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c index 1129565a01bd..6b3f453fd500 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.c @@ -808,6 +808,9 @@ npc_cn20k_enable_mcam_entry(struct rvu *rvu, int blkaddr, u64 cfg, hw_prio; u8 kw_type; + if (index < 0 || index >= mcam->total_entries) + return -EINVAL; + if (npc_mcam_idx_2_key_type(rvu, index, &kw_type)) return -EINVAL; @@ -1056,6 +1059,9 @@ int npc_cn20k_config_mcam_entry(struct rvu *rvu, int blkaddr, int index, int kw = 0; u8 kw_type; + if (index < 0 || index >= mcam->total_entries) + return -EINVAL; + if (npc_mcam_idx_2_key_type(rvu, index, &kw_type)) return -EINVAL; @@ -1148,6 +1154,9 @@ int npc_cn20k_copy_mcam_entry(struct rvu *rvu, int blkaddr, u16 src, u16 dest) int bank, i, sb, db; int dbank, sbank; + if (src >= mcam->total_entries || dest >= mcam->total_entries) + return -EINVAL; + dbank = npc_get_bank(mcam, dest); sbank = npc_get_bank(mcam, src); @@ -1213,6 +1222,9 @@ int npc_cn20k_read_mcam_entry(struct rvu *rvu, int blkaddr, u16 index, int kw = 0, bank; u8 kw_type; + if (index >= mcam->total_entries) + return -EINVAL; + if (npc_mcam_idx_2_key_type(rvu, index, &kw_type)) return -EINVAL; @@ -4170,7 +4182,7 @@ int rvu_mbox_handler_npc_get_dft_rl_idxs(struct rvu *rvu, struct msg_req *req, return 0; } -static bool npc_is_cgx_or_lbk(struct rvu *rvu, u16 pcifunc) +bool npc_is_cgx_or_lbk(struct rvu *rvu, u16 pcifunc) { return is_pf_cgxmapped(rvu, rvu_get_pf(rvu->pdev, pcifunc)) || is_lbk_vf(rvu, pcifunc); diff --git a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h index 2f761b97f91b..3d5eb952cc07 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h +++ b/drivers/net/ethernet/marvell/octeontx2/af/cn20k/npc.h @@ -335,5 +335,6 @@ int npc_mcam_idx_2_key_type(struct rvu *rvu, u16 mcam_idx, u8 *key_type); u16 npc_cn20k_vidx2idx(u16 index); u16 npc_cn20k_idx2vidx(u16 idx); int npc_cn20k_defrag(struct rvu *rvu); +bool npc_is_cgx_or_lbk(struct rvu *rvu, u16 pcifunc); #endif /* NPC_CN20K_H */ diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c index ef5b081162eb..f977734ae712 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c @@ -3577,6 +3577,9 @@ static int nix_update_mce_rule(struct rvu *rvu, u16 pcifunc, mcam_index = npc_get_nixlf_mcam_index(mcam, pcifunc & ~RVU_PFVF_FUNC_MASK, nixlf, type); + if (mcam_index < 0) + return -EINVAL; + err = nix_update_mce_list(rvu, pcifunc, mce_list, mce_idx, mcam_index, add); return err; diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c index 5d349d131fdb..3c814d157ab9 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c @@ -163,14 +163,35 @@ int npc_get_nixlf_mcam_index(struct npc_mcam *mcam, if (rc) return -EFAULT; + if (is_lbk_vf(rvu, pcifunc)) { + if (promisc == USHRT_MAX) + return -EINVAL; + return promisc; + } + + if (is_cgx_vf(rvu, pcifunc)) { + if (ucast == USHRT_MAX) + return -EINVAL; + + return ucast; + } + switch (type) { case NIXLF_BCAST_ENTRY: + if (bcast == USHRT_MAX) + return -EINVAL; return bcast; case NIXLF_ALLMULTI_ENTRY: + if (mcast == USHRT_MAX) + return -EINVAL; return mcast; case NIXLF_PROMISC_ENTRY: + if (promisc == USHRT_MAX) + return -EINVAL; return promisc; case NIXLF_UCAST_ENTRY: + if (ucast == USHRT_MAX) + return -EINVAL; return ucast; default: return -EINVAL; @@ -238,9 +259,6 @@ void npc_enable_mcam_entry(struct rvu *rvu, struct npc_mcam *mcam, int actbank = bank; if (is_cn20k(rvu->pdev)) { - if (index < 0 || index >= mcam->banksize * mcam->banks) - return; - if (npc_cn20k_enable_mcam_entry(rvu, blkaddr, index, enable)) dev_err(rvu->dev, "Error to %s mcam %u entry\n", enable ? "enable" : "disable", index); @@ -434,6 +452,15 @@ static u64 npc_get_default_entry_action(struct rvu *rvu, struct npc_mcam *mcam, index = npc_get_nixlf_mcam_index(mcam, pf_func, nixlf, NIXLF_UCAST_ENTRY); + + if (index < 0) { + dev_err(rvu->dev, + "%s: failed to get ucast entry pcifunc:0x%x\n", + __func__, pf_func); + /* Action 0 is drop */ + return 0; + } + bank = npc_get_bank(mcam, index); index &= (mcam->banksize - 1); @@ -700,6 +727,12 @@ void rvu_npc_install_ucast_entry(struct rvu *rvu, u16 pcifunc, index = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_UCAST_ENTRY); + if (index < 0) { + dev_err(rvu->dev, + "%s: Error to get ucast entry for pcifunc=%#x\n", + __func__, pcifunc); + return; + } /* Don't change the action if entry is already enabled * Otherwise RSS action may get overwritten. @@ -755,11 +788,21 @@ void rvu_npc_install_promisc_entry(struct rvu *rvu, u16 pcifunc, index = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_PROMISC_ENTRY); + /* In cn20k, default indexes are installed only for CGX mapped + * and lbk interfaces + */ if (is_cgx_vf(rvu, pcifunc)) index = npc_get_nixlf_mcam_index(mcam, pcifunc & ~RVU_PFVF_FUNC_MASK, nixlf, NIXLF_PROMISC_ENTRY); + if (index < 0) { + dev_err(rvu->dev, + "%s: Error to get promisc entry for pcifunc=%#x\n", + __func__, pcifunc); + return; + } + /* If the corresponding PF's ucast action is RSS, * use the same action for promisc also * Please note that for lbk(s) "index" and "ucast_idx" @@ -770,6 +813,12 @@ void rvu_npc_install_promisc_entry(struct rvu *rvu, u16 pcifunc, else ucast_idx = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_UCAST_ENTRY); + if (ucast_idx < 0) { + dev_err(rvu->dev, + "%s: Error to get ucast/promisc entry for pcifunc=%#x\n", + __func__, pcifunc); + return; + } if (is_mcam_entry_enabled(rvu, mcam, blkaddr, ucast_idx)) *(u64 *)&action = npc_get_mcam_action(rvu, mcam, @@ -844,6 +893,14 @@ void rvu_npc_enable_promisc_entry(struct rvu *rvu, u16 pcifunc, index = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_PROMISC_ENTRY); + + if (index < 0) { + dev_err(rvu->dev, + "%s: Error to get promisc entry for pcifunc=%#x\n", + __func__, pcifunc); + return; + } + npc_enable_mcam_entry(rvu, mcam, blkaddr, index, enable); } @@ -884,6 +941,12 @@ void rvu_npc_install_bcast_match_entry(struct rvu *rvu, u16 pcifunc, index = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_BCAST_ENTRY); + if (index < 0) { + dev_err(rvu->dev, + "%s: Error to get bcast entry for pcifunc=%#x\n", + __func__, pcifunc); + return; + } if (!hw->cap.nix_rx_multicast) { /* Early silicon doesn't support pkt replication, @@ -948,12 +1011,25 @@ void rvu_npc_install_allmulti_entry(struct rvu *rvu, u16 pcifunc, int nixlf, index = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_ALLMULTI_ENTRY); + if (index < 0) { + dev_err(rvu->dev, + "%s: Error to get mcast entry for pcifunc=%#x\n", + __func__, pcifunc); + return; + } /* If the corresponding PF's ucast action is RSS, * use the same action for multicast entry also */ ucast_idx = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_UCAST_ENTRY); + if (ucast_idx < 0) { + dev_err(rvu->dev, + "%s: Error to get ucast entry for pcifunc=%#x\n", + __func__, pcifunc); + return; + } + if (is_mcam_entry_enabled(rvu, mcam, blkaddr, ucast_idx)) *(u64 *)&action = npc_get_mcam_action(rvu, mcam, blkaddr, ucast_idx); @@ -1018,6 +1094,13 @@ void rvu_npc_enable_allmulti_entry(struct rvu *rvu, u16 pcifunc, int nixlf, index = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_ALLMULTI_ENTRY); + if (index < 0) { + dev_err(rvu->dev, + "%s: Error to get mcast entry for pcifunc=%#x\n", + __func__, pcifunc); + return; + } + npc_enable_mcam_entry(rvu, mcam, blkaddr, index, enable); } @@ -1130,8 +1213,12 @@ void rvu_npc_update_flowkey_alg_idx(struct rvu *rvu, u16 pcifunc, int nixlf, index = mcam_index; } - if (index >= mcam->total_entries) + if (index < 0 || index >= mcam->total_entries) { + dev_err(rvu->dev, + "%s: Invalid mcam index, pcifunc=%#x\n", + __func__, pcifunc); return; + } bank = npc_get_bank(mcam, index); index &= (mcam->banksize - 1); @@ -1175,16 +1262,18 @@ void rvu_npc_update_flowkey_alg_idx(struct rvu *rvu, u16 pcifunc, int nixlf, /* If PF's promiscuous entry is enabled, * Set RSS action for that entry as well */ - npc_update_rx_action_with_alg_idx(rvu, action, pfvf, index, - blkaddr, alg_idx); + if (index >= 0) + npc_update_rx_action_with_alg_idx(rvu, action, pfvf, index, + blkaddr, alg_idx); index = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_ALLMULTI_ENTRY); /* If PF's allmulti entry is enabled, * Set RSS action for that entry as well */ - npc_update_rx_action_with_alg_idx(rvu, action, pfvf, index, - blkaddr, alg_idx); + if (index >= 0) + npc_update_rx_action_with_alg_idx(rvu, action, pfvf, index, + blkaddr, alg_idx); } } @@ -1197,12 +1286,22 @@ void npc_enadis_default_mce_entry(struct rvu *rvu, u16 pcifunc, int index, blkaddr, mce_idx; struct rvu_pfvf *pfvf; + /* multicast pkt replication is not enabled for AF's VFs & SDP links */ + if (is_lbk_vf(rvu, pcifunc) || is_sdp_pfvf(rvu, pcifunc)) + return; + blkaddr = rvu_get_blkaddr(rvu, BLKTYPE_NPC, 0); if (blkaddr < 0) return; index = npc_get_nixlf_mcam_index(mcam, pcifunc & ~RVU_PFVF_FUNC_MASK, nixlf, type); + if (index < 0) { + dev_err(rvu->dev, + "%s: Error to get entry for pcifunc=%#x, type=%u\n", + __func__, pcifunc, type); + return; + } /* disable MCAM entry when packet replication is not supported by hw */ if (!hw->cap.nix_rx_multicast && !is_vf(pcifunc)) { @@ -1231,6 +1330,10 @@ static void npc_enadis_default_entries(struct rvu *rvu, u16 pcifunc, struct npc_mcam *mcam = &rvu->hw->mcam; int index, blkaddr; + /* only CGX or LBK interfaces have default entries */ + if (is_cn20k(rvu->pdev) && !npc_is_cgx_or_lbk(rvu, pcifunc)) + return; + blkaddr = rvu_get_blkaddr(rvu, BLKTYPE_NPC, 0); if (blkaddr < 0) return; @@ -1240,6 +1343,12 @@ static void npc_enadis_default_entries(struct rvu *rvu, u16 pcifunc, pfvf->nix_rx_intf)) { index = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_UCAST_ENTRY); + if (index < 0) { + dev_err(rvu->dev, + "%s: Error to get ucast entry for pcifunc=%#x\n", + __func__, pcifunc); + return; + } npc_enable_mcam_entry(rvu, mcam, blkaddr, index, enable); } @@ -3897,6 +4006,12 @@ int rvu_mbox_handler_npc_read_base_steer_rule(struct rvu *rvu, /* Read the default ucast entry if there is no pkt steering rule */ index = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_UCAST_ENTRY); + if (index < 0) { + mutex_unlock(&mcam->lock); + rc = NIX_AF_ERR_AF_LF_INVALID; + goto out; + } + read_entry: /* Read the mcam entry */ npc_read_mcam_entry(rvu, mcam, blkaddr, index, &rsp->entry, &intf, @@ -3970,6 +4085,12 @@ void rvu_npc_clear_ucast_entry(struct rvu *rvu, int pcifunc, int nixlf) ucast_idx = npc_get_nixlf_mcam_index(mcam, pcifunc, nixlf, NIXLF_UCAST_ENTRY); + if (ucast_idx < 0) { + dev_err(rvu->dev, + "%s: Error to get ucast entry for pcifunc=%#x\n", + __func__, pcifunc); + return; + } npc_enable_mcam_entry(rvu, mcam, blkaddr, ucast_idx, false); diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc_fs.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc_fs.c index fe10554b1f0e..6ae9cdcb608b 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc_fs.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc_fs.c @@ -1444,7 +1444,7 @@ static int npc_install_flow(struct rvu *rvu, int blkaddr, u16 target, struct msg_rsp write_rsp; struct mcam_entry *entry; bool new = false; - u16 entry_index; + int entry_index; int err; installed_features = req->features; @@ -1477,6 +1477,14 @@ static int npc_install_flow(struct rvu *rvu, int blkaddr, u16 target, if (req->default_rule) { entry_index = npc_get_nixlf_mcam_index(mcam, target, nixlf, NIXLF_UCAST_ENTRY); + + if (entry_index < 0) { + dev_err(rvu->dev, + "%s: Error to get ucast entry for target=%#x\n", + __func__, target); + return -EINVAL; + } + enable = is_mcam_entry_enabled(rvu, mcam, blkaddr, entry_index); } From baa3c65435fb3f450b262672bc06db887a92d397 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Thu, 30 Apr 2026 21:55:01 +0200 Subject: [PATCH 147/525] netfilter: flowtable: use skb_pull_rcsum() to pop vlan/pppoe header This adjusts the checksum, if required, after pulling the layer 2 header, either the pppoe header or the inner vlan header in the double-tagged vlan packets. Fixes: 4cd91f7c290f ("netfilter: flowtable: add vlan support") Fixes: 72efd585f714 ("netfilter: flowtable: add pppoe support") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_flow_table_ip.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c index 2eba64eb393a..9c05a50d6013 100644 --- a/net/netfilter/nf_flow_table_ip.c +++ b/net/netfilter/nf_flow_table_ip.c @@ -445,13 +445,13 @@ static void nf_flow_encap_pop(struct nf_flowtable_ctx *ctx, switch (skb->protocol) { case htons(ETH_P_8021Q): vlan_hdr = (struct vlan_hdr *)skb->data; - __skb_pull(skb, VLAN_HLEN); + skb_pull_rcsum(skb, VLAN_HLEN); vlan_set_encap_proto(skb, vlan_hdr); skb_reset_network_header(skb); break; case htons(ETH_P_PPP_SES): skb->protocol = __nf_flow_pppoe_proto(skb); - skb_pull(skb, PPPOE_SES_HLEN); + skb_pull_rcsum(skb, PPPOE_SES_HLEN); skb_reset_network_header(skb); break; } From 845db023a8aeba8b14315a846dcfba31ee727fb1 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Fri, 1 May 2026 19:23:12 +0800 Subject: [PATCH 148/525] ublk: don't issue uring_cmd from fallback task work When ublk_ch_uring_cmd_cb() runs as fallback task work (e.g., because the submitting task is exiting), the command should not be issued as current is a kworker, not the daemon task. This can cause io->task to capture the wrong task in __ublk_fetch(), leading to a task mismatch warning in ublk_uring_cmd_cancel_fn(). Check tw.cancel and return -ECANCELED instead of issuing the command from fallback context. Fixes: 3421c7f68bba ("ublk: make sure io cmd handled in submitter task context") Signed-off-by: Ming Lei Link: https://patch.msgid.link/20260501112312.947327-1-tom.leiming@gmail.com Signed-off-by: Jens Axboe --- drivers/block/ublk_drv.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c index 8e5f3738c203..d10460d29e4a 100644 --- a/drivers/block/ublk_drv.c +++ b/drivers/block/ublk_drv.c @@ -3496,8 +3496,10 @@ static void ublk_ch_uring_cmd_cb(struct io_tw_req tw_req, io_tw_token_t tw) { unsigned int issue_flags = IO_URING_CMD_TASK_WORK_ISSUE_FLAGS; struct io_uring_cmd *cmd = io_uring_cmd_from_tw(tw_req); - int ret = ublk_ch_uring_cmd_local(cmd, issue_flags); + int ret = -ECANCELED; + if (!tw.cancel) + ret = ublk_ch_uring_cmd_local(cmd, issue_flags); if (ret != -EIOCBQUEUED) io_uring_cmd_done(cmd, ret, issue_flags); } From 56722cfbb78d7eb41756cd78dc5192d08bd14f3d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?R=C3=A1mon=20van=20Raaij?= Date: Thu, 30 Apr 2026 21:12:24 +0200 Subject: [PATCH 149/525] ALSA: hda/realtek: Add codec SSID quirk for Lenovo Yoga Pro 9 16IMH9 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The Yoga Pro 9 16IMH9 (codec SSID 17aa:38d6) shares PCI audio device subsystem ID 17aa:3811 with the Legion S7 15IMH05. The existing SND_PCI_QUIRK entry for the Legion routes both machines to ALC287_FIXUP_LEGION_15IMHG05_SPEAKERS, which does not bind the TAS2781 smart amplifiers, resulting in near-silent built-in speakers. Add an HDA_CODEC_QUIRK entry immediately before the conflicting PCI quirk that matches the Yoga Pro 9's unique codec SSID and routes it to ALC287_FIXUP_TAS2781_I2C. Codec quirks are evaluated after PCI quirks and take precedence, leaving the Legion S7 15IMH05 entry unaffected. This follows the same pattern used to disambiguate PCI SSID 17aa:3847 (shared between Yoga Pro 7 14IMH9 and Legion 7 16ACHG6), where a HDA_CODEC_QUIRK for codec SSID 17aa:38cf resolves the conflict. Signed-off-by: Rámon van Raaij Link: https://patch.msgid.link/20260430191224.patch1-ramon@vanraaij.eu Signed-off-by: Takashi Iwai --- sound/hda/codecs/realtek/alc269.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/sound/hda/codecs/realtek/alc269.c b/sound/hda/codecs/realtek/alc269.c index 9a5747c2e359..e061673d3c2e 100644 --- a/sound/hda/codecs/realtek/alc269.c +++ b/sound/hda/codecs/realtek/alc269.c @@ -7647,6 +7647,10 @@ static const struct hda_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x17aa, 0x3801, "Lenovo Yoga9 14IAP7", ALC287_FIXUP_YOGA9_14IAP7_BASS_SPK_PIN), HDA_CODEC_QUIRK(0x17aa, 0x3802, "DuetITL 2021", ALC287_FIXUP_YOGA7_14ITL_SPEAKERS), SND_PCI_QUIRK(0x17aa, 0x3802, "Lenovo Yoga Pro 9 14IRP8", ALC287_FIXUP_TAS2781_I2C), + /* Yoga Pro 9 16IMH9 shares PCI SSID 17aa:3811 with Legion S7 15IMH05; + * use codec SSID to distinguish them + */ + HDA_CODEC_QUIRK(0x17aa, 0x38d6, "Lenovo Yoga Pro 9 16IMH9", ALC287_FIXUP_TAS2781_I2C), SND_PCI_QUIRK(0x17aa, 0x3811, "Legion S7 15IMH05", ALC287_FIXUP_LEGION_15IMHG05_SPEAKERS), SND_PCI_QUIRK(0x17aa, 0x3813, "Legion 7i 15IMHG05", ALC287_FIXUP_LEGION_15IMHG05_SPEAKERS), SND_PCI_QUIRK(0x17aa, 0x3818, "Lenovo C940 / Yoga Duet 7", ALC298_FIXUP_LENOVO_C940_DUET7), From ad39a189bfebb3de580f390bc000f9e121c6aca3 Mon Sep 17 00:00:00 2001 From: Pauli Virtanen Date: Thu, 30 Apr 2026 23:49:50 +0300 Subject: [PATCH 150/525] ALSA: usb-audio: add min_mute quirk for Razer Nommo V2 X ID 1532:055e Razer USA, Ltd Razer Nommo V2 X is tested to have muted min playback volume. Apply quirk for that. Link: https://gitlab.freedesktop.org/pipewire/pipewire/-/work_items/5235 Signed-off-by: Pauli Virtanen Link: https://patch.msgid.link/94449577332d14d7974864903825f27e5824ddbc.1777579951.git.pav@iki.fi Signed-off-by: Takashi Iwai --- sound/usb/quirks.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/sound/usb/quirks.c b/sound/usb/quirks.c index 0b4ecc2c6bcc..f77ff05dff0b 100644 --- a/sound/usb/quirks.c +++ b/sound/usb/quirks.c @@ -2366,6 +2366,8 @@ static const struct usb_audio_quirk_flags_table quirk_flags_table[] = { QUIRK_FLAG_IGNORE_CTL_ERROR), DEVICE_FLG(0x152a, 0x880a, /* NeuralDSP Quad Cortex */ 0), /* Doesn't have the vendor quirk which would otherwise apply */ + DEVICE_FLG(0x1532, 0x055e, /* Razer Nommo V2 X */ + QUIRK_FLAG_MIXER_PLAYBACK_MIN_MUTE), DEVICE_FLG(0x154e, 0x1002, /* Denon DCD-1500RE */ QUIRK_FLAG_ITF_USB_DSD_DAC | QUIRK_FLAG_CTL_MSG_DELAY), DEVICE_FLG(0x154e, 0x1003, /* Denon DA-300USB */ From 2e42a17b8f6bc3c0cd69d7556b588011d3ec2394 Mon Sep 17 00:00:00 2001 From: Eliot Courtney Date: Thu, 23 Apr 2026 21:36:52 +0900 Subject: [PATCH 151/525] rust: drm: gem: clean up GEM state in init failure case MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Currently, if `drm_gem_object_init` fails, the object is freed without any cleanup. Perform the cleanup in that case. Cc: stable@vger.kernel.org Fixes: c284d3e42338 ("rust: drm: gem: Add GEM object abstraction") Signed-off-by: Eliot Courtney Reviewed-by: Alice Ryhl Reviewed-by: Onur Özkan Link: https://patch.msgid.link/20260423-fix-gem-1-v1-1-e12e35f7bba9@nvidia.com [ Move safety comment closer to unsafe block to avoid a clippy warning. - Danilo ] Signed-off-by: Danilo Krummrich --- rust/kernel/drm/gem/mod.rs | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/rust/kernel/drm/gem/mod.rs b/rust/kernel/drm/gem/mod.rs index 75acda7ba500..01b5bd47a333 100644 --- a/rust/kernel/drm/gem/mod.rs +++ b/rust/kernel/drm/gem/mod.rs @@ -277,8 +277,17 @@ pub fn new(dev: &drm::Device, size: usize, args: T::Args) -> Result` is always treated as pinned. let ptr = KBox::into_raw(unsafe { Pin::into_inner_unchecked(obj) }); From 41ca998fbe30755191342b58e4f642cf3052ef2b Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 1 May 2026 19:07:19 +0200 Subject: [PATCH 152/525] parisc: Fix 64-bit kernel build when CONFIG_COMPAT=n VDSO32_SYMBOL() is used in signal.c, defining the value to zero avoids liker issues when CONFIG_COMPAT=n. Signed-off-by: Helge Deller --- arch/parisc/include/asm/vdso.h | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/arch/parisc/include/asm/vdso.h b/arch/parisc/include/asm/vdso.h index 5501560f5ffe..e5cca3c9c8e7 100644 --- a/arch/parisc/include/asm/vdso.h +++ b/arch/parisc/include/asm/vdso.h @@ -6,13 +6,14 @@ #ifdef CONFIG_64BIT #include +#define VDSO64_SYMBOL(tsk, name) ((tsk)->mm->context.vdso_base + (vdso64_offset_##name)) #endif #if !defined(CONFIG_64BIT) || defined(CONFIG_COMPAT) #include -#endif - -#define VDSO64_SYMBOL(tsk, name) ((tsk)->mm->context.vdso_base + (vdso64_offset_##name)) #define VDSO32_SYMBOL(tsk, name) ((tsk)->mm->context.vdso_base + (vdso32_offset_##name)) +#else +#define VDSO32_SYMBOL(tsk, name) 0UL +#endif #endif /* __ASSEMBLER__ */ From cb48828f06afa232cc330f0f4d6be101067810b3 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Thu, 23 Apr 2026 20:17:45 +0100 Subject: [PATCH 153/525] selftests/rseq: Don't run tests with runner scripts outside of the scripts The rseq selftests include two runner scripts run_param_test.sh and run_syscall_errors_test.sh which set up the environment for test binaries and run them with various parameters. Currently we list these test binaries in TEST_GEN_PROGS but this results in the kselftest framework running them directly as well as via the runners, resulting in duplication and spurious failures when the environment is not correctly set up (eg, if glibc tries to use rseq). Move the binaries the runners invoke to TEST_GEN_PROGS_EXTENDED, binaries listed there are built but not run by the framework. The param_test benchmarks are not moved since they are not run by run_param_test.sh. Fixes: 830969e7821a ("selftests/rseq: Implement time slice extension test") Signed-off-by: Mark Brown Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Link: https://patch.msgid.link/20260423-selftests-rseq-use-runner-v1-1-e13a133754c1@kernel.org Cc: stable@vger.kernel.org --- tools/testing/selftests/rseq/Makefile | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/tools/testing/selftests/rseq/Makefile b/tools/testing/selftests/rseq/Makefile index 4ef90823b652..0d1947c0d623 100644 --- a/tools/testing/selftests/rseq/Makefile +++ b/tools/testing/selftests/rseq/Makefile @@ -14,12 +14,15 @@ LDLIBS += -lpthread -ldl # still track changes to header files and depend on shared object. OVERRIDE_TARGETS = 1 -TEST_GEN_PROGS = basic_test basic_percpu_ops_test basic_percpu_ops_mm_cid_test param_test \ - param_test_benchmark param_test_compare_twice param_test_mm_cid \ - param_test_mm_cid_benchmark param_test_mm_cid_compare_twice \ - syscall_errors_test slice_test +TEST_GEN_PROGS = basic_test basic_percpu_ops_test basic_percpu_ops_mm_cid_test \ + param_test_benchmark param_test_mm_cid_benchmark slice_test -TEST_GEN_PROGS_EXTENDED = librseq.so +TEST_GEN_PROGS_EXTENDED = librseq.so \ + param_test \ + param_test_compare_twice \ + param_test_mm_cid \ + param_test_mm_cid_compare_twice \ + syscall_errors_test TEST_PROGS = run_param_test.sh run_syscall_errors_test.sh From 2cb68e45120dfc66404c7547d95b8ac6ff0b25ce Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 28 Apr 2026 10:10:19 +0200 Subject: [PATCH 154/525] rseq: Set rseq::cpu_id_start to 0 on unregistration The RSEQ rework changed that to RSEQ_CPU_UNINITILIZED, which is obviously incompatible. Revert back to the original behavior. Fixes: 0f085b41880e ("rseq: Provide and use rseq_set_ids()") Reported-by: Dmitry Vyukov Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Link: https://patch.msgid.link/20260428224427.271566313%40kernel.org Cc: stable@vger.kernel.org --- kernel/rseq.c | 20 +++++++++----------- 1 file changed, 9 insertions(+), 11 deletions(-) diff --git a/kernel/rseq.c b/kernel/rseq.c index 38d3ef540760..b9f11931ef78 100644 --- a/kernel/rseq.c +++ b/kernel/rseq.c @@ -236,11 +236,6 @@ static int __init rseq_debugfs_init(void) } __initcall(rseq_debugfs_init); -static bool rseq_set_ids(struct task_struct *t, struct rseq_ids *ids, u32 node_id) -{ - return rseq_set_ids_get_csaddr(t, ids, node_id, NULL); -} - static bool rseq_handle_cs(struct task_struct *t, struct pt_regs *regs) { struct rseq __user *urseq = t->rseq.usrptr; @@ -384,19 +379,22 @@ void rseq_syscall(struct pt_regs *regs) static bool rseq_reset_ids(void) { - struct rseq_ids ids = { - .cpu_id = RSEQ_CPU_ID_UNINITIALIZED, - .mm_cid = 0, - }; + struct rseq __user *rseq = current->rseq.usrptr; /* * If this fails, terminate it because this leaves the kernel in * stupid state as exit to user space will try to fixup the ids * again. */ - if (rseq_set_ids(current, &ids, 0)) - return true; + scoped_user_rw_access(rseq, efault) { + unsafe_put_user(0, &rseq->cpu_id_start, efault); + unsafe_put_user(RSEQ_CPU_ID_UNINITIALIZED, &rseq->cpu_id, efault); + unsafe_put_user(0, &rseq->node_id, efault); + unsafe_put_user(0, &rseq->mm_cid, efault); + } + return true; +efault: force_sig(SIGSEGV); return false; } From e9766e6f7d330dce7530918d8c6e3ec96d6c6e24 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 28 Apr 2026 10:14:41 +0200 Subject: [PATCH 155/525] rseq: Protect rseq_reset() against interrupts rseq_reset() uses memset() to clear the tasks rseq data. That's racy against membarrier() and preemption. Guard it with irqsave to cure this. Fixes: faba9d250eae ("rseq: Introduce struct rseq_data") Reported-by: Dmitry Vyukov Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Link: https://patch.msgid.link/20260428224427.353887714%40kernel.org Cc: stable@vger.kernel.org --- include/linux/rseq.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/include/linux/rseq.h b/include/linux/rseq.h index b9d62fc2140d..f446909551df 100644 --- a/include/linux/rseq.h +++ b/include/linux/rseq.h @@ -119,6 +119,8 @@ static inline void rseq_virt_userspace_exit(void) static inline void rseq_reset(struct task_struct *t) { + /* Protect against preemption and membarrier IPI */ + guard(irqsave)(); memset(&t->rseq, 0, sizeof(t->rseq)); t->rseq.ids.cpu_id = RSEQ_CPU_ID_UNINITIALIZED; } From 010b7723c0a3b9ad58f50b715dbe2e7781d29400 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 28 Apr 2026 09:34:45 +0200 Subject: [PATCH 156/525] rseq: Don't advertise time slice extensions if disabled If time slice extensions have been disabled on the kernel command line, then advertising them in RSEQ flags is wrong. Adjust the conditionals to reflect reality, fixup the misleading comments about the gap of these flags and the rseq::flags field. Fixes: d6200245c75e ("rseq: Allow registering RSEQ with slice extension") Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Link: https://patch.msgid.link/20260428224427.437059375%40kernel.org Cc: stable@vger.kernel.org --- include/uapi/linux/rseq.h | 5 ++++- kernel/rseq.c | 9 +++++---- 2 files changed, 9 insertions(+), 5 deletions(-) diff --git a/include/uapi/linux/rseq.h b/include/uapi/linux/rseq.h index f69344fe6c08..ca6fe1f9d05e 100644 --- a/include/uapi/linux/rseq.h +++ b/include/uapi/linux/rseq.h @@ -28,7 +28,7 @@ enum rseq_cs_flags_bit { RSEQ_CS_FLAG_NO_RESTART_ON_PREEMPT_BIT = 0, RSEQ_CS_FLAG_NO_RESTART_ON_SIGNAL_BIT = 1, RSEQ_CS_FLAG_NO_RESTART_ON_MIGRATE_BIT = 2, - /* (3) Intentional gap to put new bits into a separate byte */ + /* (3) Intentional gap to keep new bits separate */ /* User read only feature flags */ RSEQ_CS_FLAG_SLICE_EXT_AVAILABLE_BIT = 4, @@ -161,6 +161,9 @@ struct rseq { * - RSEQ_CS_FLAG_NO_RESTART_ON_PREEMPT * - RSEQ_CS_FLAG_NO_RESTART_ON_SIGNAL * - RSEQ_CS_FLAG_NO_RESTART_ON_MIGRATE + * + * It is now used for feature status advertisement by the kernel. + * See: enum rseq_cs_flags_bit for further information. */ __u32 flags; diff --git a/kernel/rseq.c b/kernel/rseq.c index b9f11931ef78..586f58f652c6 100644 --- a/kernel/rseq.c +++ b/kernel/rseq.c @@ -462,10 +462,11 @@ SYSCALL_DEFINE4(rseq, struct rseq __user *, rseq, u32, rseq_len, int, flags, u32 return -EFAULT; if (IS_ENABLED(CONFIG_RSEQ_SLICE_EXTENSION)) { - rseqfl |= RSEQ_CS_FLAG_SLICE_EXT_AVAILABLE; - if (rseq_slice_extension_enabled() && - (flags & RSEQ_FLAG_SLICE_EXT_DEFAULT_ON)) - rseqfl |= RSEQ_CS_FLAG_SLICE_EXT_ENABLED; + if (rseq_slice_extension_enabled()) { + rseqfl |= RSEQ_CS_FLAG_SLICE_EXT_AVAILABLE; + if (flags & RSEQ_FLAG_SLICE_EXT_DEFAULT_ON) + rseqfl |= RSEQ_CS_FLAG_SLICE_EXT_ENABLED; + } } scoped_user_write_access(rseq, efault) { From 01eb80b767430ae868c48ad106c60eb61a508c85 Mon Sep 17 00:00:00 2001 From: Alok Tiwari Date: Fri, 10 Apr 2026 04:20:12 -0700 Subject: [PATCH 157/525] accel/qaic: fix incorrect counter check in RAS message decode The UE and UE_NF cases check ce_count against UINT_MAX before incrementing their respective counters. This is logically incorrect and prevents ue_count and ue_nf_count from incrementing when ce_count reaches UINT_MAX. Fixes: c11a50b170e7 ("accel/qaic: Add Reliability, Accessibility, Serviceability (RAS)") Signed-off-by: Alok Tiwari Reviewed-by: Jeff Hugo Signed-off-by: Jeff Hugo Link: https://patch.msgid.link/20260410112015.592546-1-alok.a.tiwari@oracle.com --- drivers/accel/qaic/qaic_ras.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/accel/qaic/qaic_ras.c b/drivers/accel/qaic/qaic_ras.c index cc0b75461e1a..6791af366cba 100644 --- a/drivers/accel/qaic/qaic_ras.c +++ b/drivers/accel/qaic/qaic_ras.c @@ -497,11 +497,11 @@ static void decode_ras_msg(struct qaic_device *qdev, struct ras_data *msg) qdev->ce_count++; break; case UE: - if (qdev->ce_count != UINT_MAX) + if (qdev->ue_count != UINT_MAX) qdev->ue_count++; break; case UE_NF: - if (qdev->ce_count != UINT_MAX) + if (qdev->ue_nf_count != UINT_MAX) qdev->ue_nf_count++; break; default: From 5234094c0150338c35280a75cc7842015f76b725 Mon Sep 17 00:00:00 2001 From: Stefan Metzmacher Date: Wed, 29 Apr 2026 15:43:35 +0200 Subject: [PATCH 158/525] smb: smbdirect: make use of DEFAULT_SYMBOL_NAMESPACE and EXPORT_SYMBOL_GPL This is a better solution than EXPORT_SYMBOL_FOR_MODULES(__sym, "cifs,ksmbd") as it makes it possible to rebuild smbdirect.ko against a running kernel and then load the existing cifs.ko and ksmbd.ko from the running kernel. Suggested-by: Christoph Hellwig Link: https://lore.kernel.org/linux-cifs/aehrPuY60VMcYGU8@infradead.org/ Cc: Steve French Cc: Tom Talpey Cc: Long Li Cc: Namjae Jeon Cc: Christoph Hellwig Cc: linux-cifs@vger.kernel.org Cc: samba-technical@lists.samba.org Signed-off-by: Stefan Metzmacher Signed-off-by: Steve French --- fs/smb/client/smbdirect.c | 2 ++ fs/smb/server/transport_rdma.c | 2 ++ fs/smb/smbdirect/accept.c | 2 +- fs/smb/smbdirect/connect.c | 4 ++-- fs/smb/smbdirect/connection.c | 16 ++++++++-------- fs/smb/smbdirect/debug.c | 2 +- fs/smb/smbdirect/devices.c | 2 +- fs/smb/smbdirect/internal.h | 1 + fs/smb/smbdirect/listen.c | 2 +- fs/smb/smbdirect/mr.c | 6 +++--- fs/smb/smbdirect/public.h | 2 -- fs/smb/smbdirect/rw.c | 2 +- fs/smb/smbdirect/socket.c | 20 ++++++++++---------- 13 files changed, 33 insertions(+), 30 deletions(-) diff --git a/fs/smb/client/smbdirect.c b/fs/smb/client/smbdirect.c index 75f9f91a7ec9..b9826185de18 100644 --- a/fs/smb/client/smbdirect.c +++ b/fs/smb/client/smbdirect.c @@ -558,3 +558,5 @@ void smbd_debug_proc_show(struct TCP_Server_Info *server, struct seq_file *m) server->rdma_readwrite_threshold, m); } + +MODULE_IMPORT_NS("SMBDIRECT"); diff --git a/fs/smb/server/transport_rdma.c b/fs/smb/server/transport_rdma.c index a8242c00096f..346c051e31f5 100644 --- a/fs/smb/server/transport_rdma.c +++ b/fs/smb/server/transport_rdma.c @@ -540,3 +540,5 @@ static const struct ksmbd_transport_ops ksmbd_smb_direct_transport_ops = { .rdma_write = smb_direct_rdma_write, .free_transport = smb_direct_free_transport, }; + +MODULE_IMPORT_NS("SMBDIRECT"); diff --git a/fs/smb/smbdirect/accept.c b/fs/smb/smbdirect/accept.c index 704b271af3a8..529740005838 100644 --- a/fs/smb/smbdirect/accept.c +++ b/fs/smb/smbdirect/accept.c @@ -854,4 +854,4 @@ struct smbdirect_socket *smbdirect_socket_accept(struct smbdirect_socket *lsc, return nsc; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_accept); +EXPORT_SYMBOL_GPL(smbdirect_socket_accept); diff --git a/fs/smb/smbdirect/connect.c b/fs/smb/smbdirect/connect.c index 8addee43a381..cd726b399afe 100644 --- a/fs/smb/smbdirect/connect.c +++ b/fs/smb/smbdirect/connect.c @@ -60,7 +60,7 @@ int smbdirect_connect(struct smbdirect_socket *sc, const struct sockaddr *dst) */ return 0; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connect); +EXPORT_SYMBOL_GPL(smbdirect_connect); static int smbdirect_connect_setup_connection(struct smbdirect_socket *sc) { @@ -922,4 +922,4 @@ int smbdirect_connect_sync(struct smbdirect_socket *sc, return 0; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connect_sync); +EXPORT_SYMBOL_GPL(smbdirect_connect_sync); diff --git a/fs/smb/smbdirect/connection.c b/fs/smb/smbdirect/connection.c index 822366718d45..fe9912e53da6 100644 --- a/fs/smb/smbdirect/connection.c +++ b/fs/smb/smbdirect/connection.c @@ -706,7 +706,7 @@ bool smbdirect_connection_is_connected(struct smbdirect_socket *sc) return false; return true; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_is_connected); +EXPORT_SYMBOL_GPL(smbdirect_connection_is_connected); int smbdirect_connection_wait_for_connected(struct smbdirect_socket *sc) { @@ -779,7 +779,7 @@ int smbdirect_connection_wait_for_connected(struct smbdirect_socket *sc) return 0; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_wait_for_connected); +EXPORT_SYMBOL_GPL(smbdirect_connection_wait_for_connected); void smbdirect_connection_idle_timer_work(struct work_struct *work) { @@ -958,7 +958,7 @@ int smbdirect_connection_send_batch_flush(struct smbdirect_socket *sc, return ret; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_send_batch_flush); +EXPORT_SYMBOL_GPL(smbdirect_connection_send_batch_flush); struct smbdirect_send_batch * smbdirect_init_send_batch_storage(struct smbdirect_send_batch_storage *storage, @@ -976,7 +976,7 @@ smbdirect_init_send_batch_storage(struct smbdirect_send_batch_storage *storage, return batch; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_init_send_batch_storage); +EXPORT_SYMBOL_GPL(smbdirect_init_send_batch_storage); static int smbdirect_connection_wait_for_send_bcredit(struct smbdirect_socket *sc, struct smbdirect_send_batch *batch) @@ -1263,7 +1263,7 @@ int smbdirect_connection_send_single_iter(struct smbdirect_socket *sc, bcredit_failed: return ret; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_send_single_iter); +EXPORT_SYMBOL_GPL(smbdirect_connection_send_single_iter); int smbdirect_connection_send_wait_zero_pending(struct smbdirect_socket *sc) { @@ -1288,7 +1288,7 @@ int smbdirect_connection_send_wait_zero_pending(struct smbdirect_socket *sc) return 0; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_send_wait_zero_pending); +EXPORT_SYMBOL_GPL(smbdirect_connection_send_wait_zero_pending); int smbdirect_connection_send_iter(struct smbdirect_socket *sc, struct iov_iter *iter, @@ -1373,7 +1373,7 @@ int smbdirect_connection_send_iter(struct smbdirect_socket *sc, return total_count; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_send_iter); +EXPORT_SYMBOL_GPL(smbdirect_connection_send_iter); static void smbdirect_connection_send_io_done(struct ib_cq *cq, struct ib_wc *wc) { @@ -1937,7 +1937,7 @@ int smbdirect_connection_recvmsg(struct smbdirect_socket *sc, goto again; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_recvmsg); +EXPORT_SYMBOL_GPL(smbdirect_connection_recvmsg); static bool smbdirect_map_sges_single_page(struct smbdirect_map_sges *state, struct page *page, size_t off, size_t len) diff --git a/fs/smb/smbdirect/debug.c b/fs/smb/smbdirect/debug.c index a66a19d4a463..05ba7c8d165e 100644 --- a/fs/smb/smbdirect/debug.c +++ b/fs/smb/smbdirect/debug.c @@ -85,4 +85,4 @@ void smbdirect_connection_legacy_debug_proc_show(struct smbdirect_socket *sc, atomic_read(&sc->mr_io.ready.count), atomic_read(&sc->mr_io.used.count)); } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_legacy_debug_proc_show); +EXPORT_SYMBOL_GPL(smbdirect_connection_legacy_debug_proc_show); diff --git a/fs/smb/smbdirect/devices.c b/fs/smb/smbdirect/devices.c index 44962f221c35..7adacbdfe12e 100644 --- a/fs/smb/smbdirect/devices.c +++ b/fs/smb/smbdirect/devices.c @@ -238,7 +238,7 @@ u8 smbdirect_netdev_rdma_capable_node_type(struct net_device *netdev) return RDMA_NODE_UNSPECIFIED; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_netdev_rdma_capable_node_type); +EXPORT_SYMBOL_GPL(smbdirect_netdev_rdma_capable_node_type); __init int smbdirect_devices_init(void) { diff --git a/fs/smb/smbdirect/internal.h b/fs/smb/smbdirect/internal.h index 2d5acf2c21bc..82529b41708b 100644 --- a/fs/smb/smbdirect/internal.h +++ b/fs/smb/smbdirect/internal.h @@ -6,6 +6,7 @@ #ifndef __FS_SMB_COMMON_SMBDIRECT_INTERNAL_H__ #define __FS_SMB_COMMON_SMBDIRECT_INTERNAL_H__ +#define DEFAULT_SYMBOL_NAMESPACE "SMBDIRECT" #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt #include "smbdirect.h" diff --git a/fs/smb/smbdirect/listen.c b/fs/smb/smbdirect/listen.c index 143a7618d95f..2f78bcaedbf8 100644 --- a/fs/smb/smbdirect/listen.c +++ b/fs/smb/smbdirect/listen.c @@ -90,7 +90,7 @@ int smbdirect_socket_listen(struct smbdirect_socket *sc, int backlog) */ return 0; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_listen); +EXPORT_SYMBOL_GPL(smbdirect_socket_listen); static int smbdirect_new_rdma_event_handler(struct rdma_cm_id *new_id, struct rdma_cm_event *event) diff --git a/fs/smb/smbdirect/mr.c b/fs/smb/smbdirect/mr.c index 5228e699cd5d..9e6ac9d8717a 100644 --- a/fs/smb/smbdirect/mr.c +++ b/fs/smb/smbdirect/mr.c @@ -380,7 +380,7 @@ smbdirect_connection_register_mr_io(struct smbdirect_socket *sc, mutex_unlock(&mr->mutex); return NULL; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_register_mr_io); +EXPORT_SYMBOL_GPL(smbdirect_connection_register_mr_io); void smbdirect_mr_io_fill_buffer_descriptor(struct smbdirect_mr_io *mr, struct smbdirect_buffer_descriptor_v1 *v1) @@ -397,7 +397,7 @@ void smbdirect_mr_io_fill_buffer_descriptor(struct smbdirect_mr_io *mr, } mutex_unlock(&mr->mutex); } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_mr_io_fill_buffer_descriptor); +EXPORT_SYMBOL_GPL(smbdirect_mr_io_fill_buffer_descriptor); /* * Deregister a MR after I/O is done @@ -490,4 +490,4 @@ void smbdirect_connection_deregister_mr_io(struct smbdirect_mr_io *mr) if (!kref_put(&mr->kref, smbdirect_mr_io_free_locked)) mutex_unlock(&mr->mutex); } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_deregister_mr_io); +EXPORT_SYMBOL_GPL(smbdirect_connection_deregister_mr_io); diff --git a/fs/smb/smbdirect/public.h b/fs/smb/smbdirect/public.h index 50088155e7c3..d4fb36e65254 100644 --- a/fs/smb/smbdirect/public.h +++ b/fs/smb/smbdirect/public.h @@ -13,8 +13,6 @@ struct smbdirect_socket; struct smbdirect_send_batch; struct smbdirect_mr_io; -#define __SMBDIRECT_EXPORT_SYMBOL__(__sym) EXPORT_SYMBOL_FOR_MODULES(__sym, "cifs,ksmbd") - #include u8 smbdirect_netdev_rdma_capable_node_type(struct net_device *netdev); diff --git a/fs/smb/smbdirect/rw.c b/fs/smb/smbdirect/rw.c index c2f46b17731e..6fe38042cfb9 100644 --- a/fs/smb/smbdirect/rw.c +++ b/fs/smb/smbdirect/rw.c @@ -252,4 +252,4 @@ int smbdirect_connection_rdma_xmit(struct smbdirect_socket *sc, kfree(msg); goto out; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_connection_rdma_xmit); +EXPORT_SYMBOL_GPL(smbdirect_connection_rdma_xmit); diff --git a/fs/smb/smbdirect/socket.c b/fs/smb/smbdirect/socket.c index 1b4ab01b745e..39cca7219c4d 100644 --- a/fs/smb/smbdirect/socket.c +++ b/fs/smb/smbdirect/socket.c @@ -20,7 +20,7 @@ bool smbdirect_frwr_is_supported(const struct ib_device_attr *attrs) return false; return true; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_frwr_is_supported); +EXPORT_SYMBOL_GPL(smbdirect_frwr_is_supported); static void smbdirect_socket_cleanup_work(struct work_struct *work); @@ -107,7 +107,7 @@ int smbdirect_socket_create_kern(struct net *net, struct smbdirect_socket **_sc) alloc_failed: return ret; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_create_kern); +EXPORT_SYMBOL_GPL(smbdirect_socket_create_kern); int smbdirect_socket_init_accepting(struct rdma_cm_id *id, struct smbdirect_socket *sc) { @@ -148,7 +148,7 @@ int smbdirect_socket_create_accepting(struct rdma_cm_id *id, struct smbdirect_so alloc_failed: return ret; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_create_accepting); +EXPORT_SYMBOL_GPL(smbdirect_socket_create_accepting); int smbdirect_socket_set_initial_parameters(struct smbdirect_socket *sc, const struct smbdirect_socket_parameters *sp) @@ -189,14 +189,14 @@ int smbdirect_socket_set_initial_parameters(struct smbdirect_socket *sc, return 0; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_set_initial_parameters); +EXPORT_SYMBOL_GPL(smbdirect_socket_set_initial_parameters); const struct smbdirect_socket_parameters * smbdirect_socket_get_current_parameters(struct smbdirect_socket *sc) { return &sc->parameters; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_get_current_parameters); +EXPORT_SYMBOL_GPL(smbdirect_socket_get_current_parameters); int smbdirect_socket_set_kernel_settings(struct smbdirect_socket *sc, enum ib_poll_context poll_ctx, @@ -220,7 +220,7 @@ int smbdirect_socket_set_kernel_settings(struct smbdirect_socket *sc, return 0; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_set_kernel_settings); +EXPORT_SYMBOL_GPL(smbdirect_socket_set_kernel_settings); void smbdirect_socket_set_logging(struct smbdirect_socket *sc, void *private_ptr, @@ -240,7 +240,7 @@ void smbdirect_socket_set_logging(struct smbdirect_socket *sc, sc->logging.needed = needed; sc->logging.vaprintf = vaprintf; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_set_logging); +EXPORT_SYMBOL_GPL(smbdirect_socket_set_logging); static void smbdirect_socket_wake_up_all(struct smbdirect_socket *sc) { @@ -663,13 +663,13 @@ int smbdirect_socket_bind(struct smbdirect_socket *sc, struct sockaddr *addr) return 0; } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_bind); +EXPORT_SYMBOL_GPL(smbdirect_socket_bind); void smbdirect_socket_shutdown(struct smbdirect_socket *sc) { smbdirect_socket_schedule_cleanup_lvl(sc, SMBDIRECT_LOG_INFO, -ESHUTDOWN); } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_shutdown); +EXPORT_SYMBOL_GPL(smbdirect_socket_shutdown); static void smbdirect_socket_release_disconnect(struct kref *kref) { @@ -712,7 +712,7 @@ void smbdirect_socket_release(struct smbdirect_socket *sc) */ kref_put(&sc->refs.destroy, smbdirect_socket_release_destroy); } -__SMBDIRECT_EXPORT_SYMBOL__(smbdirect_socket_release); +EXPORT_SYMBOL_GPL(smbdirect_socket_release); int smbdirect_socket_wait_for_credits(struct smbdirect_socket *sc, enum smbdirect_socket_status expected_status, From e768103cfbac30a49860aca08a7710d39dbdd470 Mon Sep 17 00:00:00 2001 From: Stefan Metzmacher Date: Wed, 29 Apr 2026 15:43:36 +0200 Subject: [PATCH 159/525] smb: smbdirect: introduce and use include/linux/smbdirect.h This makes it easier to rebuild cifs.ko and ksmbd.ko against a running kernel. Suggested-by: Christoph Hellwig Link: https://lore.kernel.org/linux-cifs/aehrPuY60VMcYGU8@infradead.org/ Cc: Steve French Cc: Tom Talpey Cc: Long Li Cc: Namjae Jeon Cc: Christoph Hellwig Cc: linux-cifs@vger.kernel.org Cc: samba-technical@lists.samba.org Signed-off-by: Stefan Metzmacher Signed-off-by: Steve French --- MAINTAINERS | 1 + fs/smb/client/smbdirect.c | 1 - fs/smb/client/smbdirect.h | 2 +- fs/smb/server/transport_rdma.c | 1 - fs/smb/server/transport_rdma.h | 2 +- fs/smb/smbdirect/internal.h | 3 +- fs/smb/smbdirect/smbdirect.h | 52 ------------------- .../public.h => include/linux/smbdirect.h | 50 ++++++++++++++++-- 8 files changed, 49 insertions(+), 63 deletions(-) delete mode 100644 fs/smb/smbdirect/smbdirect.h rename fs/smb/smbdirect/public.h => include/linux/smbdirect.h (76%) diff --git a/MAINTAINERS b/MAINTAINERS index 2c67ee25ffe6..060dca38dbf7 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -24650,6 +24650,7 @@ S: Maintained F: fs/smb/client/smbdirect.* F: fs/smb/smbdirect/ F: fs/smb/server/transport_rdma.* +F: include/linux/smbdirect.h SMC91x ETHERNET DRIVER M: Nicolas Pitre diff --git a/fs/smb/client/smbdirect.c b/fs/smb/client/smbdirect.c index b9826185de18..563ef488a225 100644 --- a/fs/smb/client/smbdirect.c +++ b/fs/smb/client/smbdirect.c @@ -9,7 +9,6 @@ #include "cifs_debug.h" #include "cifsproto.h" #include "smb2proto.h" -#include "../smbdirect/public.h" /* Port numbers for SMBD transport */ #define SMB_PORT 445 diff --git a/fs/smb/client/smbdirect.h b/fs/smb/client/smbdirect.h index 287ac849213d..be205ec02077 100644 --- a/fs/smb/client/smbdirect.h +++ b/fs/smb/client/smbdirect.h @@ -12,7 +12,7 @@ #include "cifsglob.h" -#include "../smbdirect/smbdirect.h" +#include extern int rdma_readwrite_threshold; extern int smbd_max_frmr_depth; diff --git a/fs/smb/server/transport_rdma.c b/fs/smb/server/transport_rdma.c index 346c051e31f5..b6d63ff8a8a3 100644 --- a/fs/smb/server/transport_rdma.c +++ b/fs/smb/server/transport_rdma.c @@ -18,7 +18,6 @@ #include "smb_common.h" #include "../common/smb2status.h" #include "transport_rdma.h" -#include "../smbdirect/public.h" #define SMB_DIRECT_PORT_IWARP 5445 diff --git a/fs/smb/server/transport_rdma.h b/fs/smb/server/transport_rdma.h index bde3d88aecc7..8b78917a1795 100644 --- a/fs/smb/server/transport_rdma.h +++ b/fs/smb/server/transport_rdma.h @@ -25,6 +25,6 @@ static inline void init_smbd_max_io_size(unsigned int sz) { } static inline unsigned int get_smbd_max_read_write_size(struct ksmbd_transport *kt) { return 0; } #endif -#include "../smbdirect/smbdirect.h" +#include #endif /* __KSMBD_TRANSPORT_RDMA_H__ */ diff --git a/fs/smb/smbdirect/internal.h b/fs/smb/smbdirect/internal.h index 82529b41708b..e9959e6dc13a 100644 --- a/fs/smb/smbdirect/internal.h +++ b/fs/smb/smbdirect/internal.h @@ -9,9 +9,8 @@ #define DEFAULT_SYMBOL_NAMESPACE "SMBDIRECT" #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt -#include "smbdirect.h" +#include #include "pdu.h" -#include "public.h" #include diff --git a/fs/smb/smbdirect/smbdirect.h b/fs/smb/smbdirect/smbdirect.h deleted file mode 100644 index bbab5f7f7cc9..000000000000 --- a/fs/smb/smbdirect/smbdirect.h +++ /dev/null @@ -1,52 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-or-later */ -/* - * Copyright (C) 2025 Stefan Metzmacher - */ - -#ifndef __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_H__ -#define __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_H__ - -#include - -/* SMB-DIRECT buffer descriptor V1 structure [MS-SMBD] 2.2.3.1 */ -struct smbdirect_buffer_descriptor_v1 { - __le64 offset; - __le32 token; - __le32 length; -} __packed; - -/* - * Connection parameters mostly from [MS-SMBD] 3.1.1.1 - * - * These are setup and negotiated at the beginning of a - * connection and remain constant unless explicitly changed. - * - * Some values are important for the upper layer. - */ -struct smbdirect_socket_parameters { - __u64 flags; -#define SMBDIRECT_FLAG_PORT_RANGE_ONLY_IB ((__u64)0x1) -#define SMBDIRECT_FLAG_PORT_RANGE_ONLY_IW ((__u64)0x2) - __u32 resolve_addr_timeout_msec; - __u32 resolve_route_timeout_msec; - __u32 rdma_connect_timeout_msec; - __u32 negotiate_timeout_msec; - __u16 initiator_depth; /* limited to U8_MAX */ - __u16 responder_resources; /* limited to U8_MAX */ - __u16 recv_credit_max; - __u16 send_credit_target; - __u32 max_send_size; - __u32 max_fragmented_send_size; - __u32 max_recv_size; - __u32 max_fragmented_recv_size; - __u32 max_read_write_size; - __u32 max_frmr_depth; - __u32 keepalive_interval_msec; - __u32 keepalive_timeout_msec; -} __packed; - -#define SMBDIRECT_FLAG_PORT_RANGE_MASK ( \ - SMBDIRECT_FLAG_PORT_RANGE_ONLY_IB | \ - SMBDIRECT_FLAG_PORT_RANGE_ONLY_IW) - -#endif /* __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_H__ */ diff --git a/fs/smb/smbdirect/public.h b/include/linux/smbdirect.h similarity index 76% rename from fs/smb/smbdirect/public.h rename to include/linux/smbdirect.h index d4fb36e65254..97f5ba730fa7 100644 --- a/fs/smb/smbdirect/public.h +++ b/include/linux/smbdirect.h @@ -3,11 +3,51 @@ * Copyright (C) 2025, Stefan Metzmacher */ -#ifndef __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_PUBLIC_H__ -#define __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_PUBLIC_H__ +#ifndef __LINUX_SMBDIRECT_H__ +#define __LINUX_SMBDIRECT_H__ -struct smbdirect_buffer_descriptor_v1; -struct smbdirect_socket_parameters; +#include + +/* SMB-DIRECT buffer descriptor V1 structure [MS-SMBD] 2.2.3.1 */ +struct smbdirect_buffer_descriptor_v1 { + __le64 offset; + __le32 token; + __le32 length; +} __packed; + +/* + * Connection parameters mostly from [MS-SMBD] 3.1.1.1 + * + * These are setup and negotiated at the beginning of a + * connection and remain constant unless explicitly changed. + * + * Some values are important for the upper layer. + */ +struct smbdirect_socket_parameters { + __u64 flags; +#define SMBDIRECT_FLAG_PORT_RANGE_ONLY_IB ((__u64)0x1) +#define SMBDIRECT_FLAG_PORT_RANGE_ONLY_IW ((__u64)0x2) + __u32 resolve_addr_timeout_msec; + __u32 resolve_route_timeout_msec; + __u32 rdma_connect_timeout_msec; + __u32 negotiate_timeout_msec; + __u16 initiator_depth; /* limited to U8_MAX */ + __u16 responder_resources; /* limited to U8_MAX */ + __u16 recv_credit_max; + __u16 send_credit_target; + __u32 max_send_size; + __u32 max_fragmented_send_size; + __u32 max_recv_size; + __u32 max_fragmented_recv_size; + __u32 max_read_write_size; + __u32 max_frmr_depth; + __u32 keepalive_interval_msec; + __u32 keepalive_timeout_msec; +} __packed; + +#define SMBDIRECT_FLAG_PORT_RANGE_MASK ( \ + SMBDIRECT_FLAG_PORT_RANGE_ONLY_IB | \ + SMBDIRECT_FLAG_PORT_RANGE_ONLY_IW) struct smbdirect_socket; struct smbdirect_send_batch; @@ -143,4 +183,4 @@ void smbdirect_connection_legacy_debug_proc_show(struct smbdirect_socket *sc, unsigned int rdma_readwrite_threshold, struct seq_file *m); -#endif /* __FS_SMB_COMMON_SMBDIRECT_SMBDIRECT_PUBLIC_H__ */ +#endif /* __LINUX_SMBDIRECT_H__ */ From 9900b9fee5a0e0f72d7c744b37c7c851d5785ac6 Mon Sep 17 00:00:00 2001 From: Yi Kuo Date: Wed, 29 Apr 2026 18:00:11 +0800 Subject: [PATCH 160/525] smb: smbdirect: fix MR registration for coalesced SG lists ib_dma_map_sg() modifies the provided scatterlist and returns the number of mapped entries, which can be fewer than the requested mr->sgt.nents if the DMA controller coalesces contiguous memory segments. Passing the original, uncoalesced count to ib_map_mr_sg() causes memory registration failures if coalescing actually occurs. Capture the actual mapped count returned by ib_dma_map_sg() and pass it to ib_map_mr_sg() to ensure correct MR registration. Also update the ib_dma_map_sg() error logging to drop the error pointer formatting, since the return value is an integer count rather than an error code. Ensure a proper error code (-EIO) is assigned when DMA mapping or MR registration fails. Fixes: de5ef8ec3c46 ("smb: smbdirect: introduce smbdirect_mr.c with client mr code") Closes: https://bugzilla.kernel.org/show_bug.cgi?id=221408 Reviewed-by: Stefan Metzmacher Acked-by: Namjae Jeon Signed-off-by: Yi Kuo Signed-off-by: Steve French --- fs/smb/smbdirect/mr.c | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-) diff --git a/fs/smb/smbdirect/mr.c b/fs/smb/smbdirect/mr.c index 9e6ac9d8717a..15c6363a2f97 100644 --- a/fs/smb/smbdirect/mr.c +++ b/fs/smb/smbdirect/mr.c @@ -269,7 +269,7 @@ smbdirect_connection_register_mr_io(struct smbdirect_socket *sc, { const struct smbdirect_socket_parameters *sp = &sc->parameters; struct smbdirect_mr_io *mr; - int ret, num_pages; + int ret, num_pages, num_mapped; struct ib_reg_wr *reg_wr; num_pages = iov_iter_npages(iter, sp->max_frmr_depth + 1); @@ -300,19 +300,22 @@ smbdirect_connection_register_mr_io(struct smbdirect_socket *sc, num_pages, iov_iter_count(iter), sp->max_frmr_depth); smbdirect_iter_to_sgt(iter, &mr->sgt, sp->max_frmr_depth); - ret = ib_dma_map_sg(sc->ib.dev, mr->sgt.sgl, mr->sgt.nents, mr->dir); - if (!ret) { + num_mapped = ib_dma_map_sg(sc->ib.dev, mr->sgt.sgl, mr->sgt.nents, mr->dir); + if (!num_mapped) { smbdirect_log_rdma_mr(sc, SMBDIRECT_LOG_ERR, - "ib_dma_map_sg num_pages=%u dir=%x ret=%d (%1pe)\n", - num_pages, mr->dir, ret, SMBDIRECT_DEBUG_ERR_PTR(ret)); + "ib_dma_map_sg num_pages=%u dir=%x num_mapped=%d\n", + num_pages, mr->dir, num_mapped); + ret = -EIO; goto dma_map_error; } - ret = ib_map_mr_sg(mr->mr, mr->sgt.sgl, mr->sgt.nents, NULL, PAGE_SIZE); - if (ret != mr->sgt.nents) { + ret = ib_map_mr_sg(mr->mr, mr->sgt.sgl, num_mapped, NULL, PAGE_SIZE); + if (ret != num_mapped) { smbdirect_log_rdma_mr(sc, SMBDIRECT_LOG_ERR, - "ib_map_mr_sg failed ret = %d nents = %u\n", - ret, mr->sgt.nents); + "ib_map_mr_sg failed ret = %d num_mapped = %u\n", + ret, num_mapped); + if (ret >= 0) + ret = -EIO; goto map_mr_error; } From f93836b236773862e9ee268a82e3614caf77ea01 Mon Sep 17 00:00:00 2001 From: Aleksander Jan Bajkowski Date: Thu, 30 Apr 2026 23:34:33 +0200 Subject: [PATCH 161/525] net: usb: r8152: add TRENDnet TUC-ET2G v2.0 The TRENDnet TUC-ET2G V2.0 is an RTL8156B based 2.5G Ethernet controller. Add the vendor and product ID values to the driver. This makes Ethernet work with the adapter. Signed-off-by: Aleksander Jan Bajkowski Reviewed-by: Andrew Lunn Reviewed-by: Birger Koblitz Link: https://patch.msgid.link/20260430213435.21821-1-olek2@wp.pl Signed-off-by: Jakub Kicinski --- drivers/net/usb/r8152.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/usb/r8152.c b/drivers/net/usb/r8152.c index 7337bf1b7d6a..1ace1d2398c9 100644 --- a/drivers/net/usb/r8152.c +++ b/drivers/net/usb/r8152.c @@ -10138,6 +10138,7 @@ static const struct usb_device_id rtl8152_table[] = { { USB_DEVICE(VENDOR_ID_DELL, 0xb097) }, { USB_DEVICE(VENDOR_ID_ASUS, 0x1976) }, { USB_DEVICE(VENDOR_ID_TRENDNET, 0xe02b) }, + { USB_DEVICE(VENDOR_ID_TRENDNET, 0xe02c) }, {} }; From 4f34002e2e37639133693c13a2a9977fab86880d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 30 Apr 2026 07:06:11 +0000 Subject: [PATCH 162/525] ipmr: prevent info-leak in pmr_cache_report() Yiming Qian reported: ipmr_cache_report()` allocates a report skb with `alloc_skb(128, GFP_ATOMIC)` and appends a `struct igmphdr` using `skb_put()`. In the non-`IGMPMSG_WHOLEPKT` path it initializes only: - `igmp->type` - `igmp->code` but does not initialize: - `igmp->csum` - `igmp->group` Later, `igmpmsg_netlink_event()` copies the bytes after `sizeof(struct igmpmsg)` into the `IPMRA_CREPORT_PKT` netlink attribute and emits `RTM_NEWCACHEREPORT` on `RTNLGRP_IPV4_MROUTE_R`. As a result, 6 bytes of stale heap data from the skb head are disclosed to userspace. Let's use skb_put_zero() instead of skb_put() to fix this bug. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: Yiming Qian Signed-off-by: Eric Dumazet Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260430070611.4004529-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/ipv4/ipmr.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c index 2058ca860294..05fb6eefe0be 100644 --- a/net/ipv4/ipmr.c +++ b/net/ipv4/ipmr.c @@ -1112,11 +1112,12 @@ static int ipmr_cache_report(const struct mr_table *mrt, msg->im_vif_hi = vifi >> 8; ipv4_pktinfo_prepare(mroute_sk, pkt, false); memcpy(skb->cb, pkt->cb, sizeof(skb->cb)); - /* Add our header */ - igmp = skb_put(skb, sizeof(struct igmphdr)); + /* Add our header. + * Note that code, csum and group fields are cleared. + */ + igmp = skb_put_zero(skb, sizeof(struct igmphdr)); igmp->type = assert; msg->im_msgtype = assert; - igmp->code = 0; ip_hdr(skb)->tot_len = htons(skb->len); /* Fix the length */ skb->transport_header = skb->network_header; } From 4b9e327991815e128ad3af75c3a04630a63ce3e0 Mon Sep 17 00:00:00 2001 From: Kai Zen Date: Thu, 30 Apr 2026 18:26:48 +0300 Subject: [PATCH 163/525] net: rtnetlink: zero ifla_vf_broadcast to avoid stack infoleak in rtnl_fill_vfinfo rtnl_fill_vfinfo() declares struct ifla_vf_broadcast on the stack without initialisation: struct ifla_vf_broadcast vf_broadcast; The struct contains a single fixed 32-byte field: /* include/uapi/linux/if_link.h */ struct ifla_vf_broadcast { __u8 broadcast[32]; }; The function then copies dev->broadcast into it using dev->addr_len as the length: memcpy(vf_broadcast.broadcast, dev->broadcast, dev->addr_len); On Ethernet devices (the overwhelming majority of SR-IOV NICs) dev->addr_len is 6, so only the first 6 bytes of broadcast[] are written. The remaining 26 bytes retain whatever was previously on the kernel stack. The full struct is then handed to userspace via: nla_put(skb, IFLA_VF_BROADCAST, sizeof(vf_broadcast), &vf_broadcast) leaking up to 26 bytes of uninitialised kernel stack per VF per RTM_GETLINK request, repeatable. The other vf_* structs in the same function are explicitly zeroed for exactly this reason - see the memset() calls for ivi, vf_vlan_info, node_guid and port_guid a few lines above. vf_broadcast was simply missed when it was added. Reachability: any unprivileged local process can open AF_NETLINK / NETLINK_ROUTE without capabilities and send RTM_GETLINK with an IFLA_EXT_MASK attribute carrying RTEXT_FILTER_VF. The kernel walks each VF and emits IFLA_VF_BROADCAST, leaking 26 bytes of stack per VF per request. Stack residue at this call site can include return addresses and transient sensitive data; KASAN with stack instrumentation, or KMSAN, will flag the nla_put() when reproduced. Zero the on-stack struct before the partial memcpy, matching the existing pattern used for the other vf_* structs in the same function. Fixes: 75345f888f70 ("ipoib: show VF broadcast address") Cc: stable@vger.kernel.org Signed-off-by: Kai Zen Link: https://patch.msgid.link/3c506e8f936e52b57620269b55c348af05d413a2.1777557228.git.kai.aizen.dev@gmail.com Signed-off-by: Jakub Kicinski --- net/core/rtnetlink.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index b613bb6e07df..df042da422ef 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -1572,6 +1572,7 @@ static noinline_for_stack int rtnl_fill_vfinfo(struct sk_buff *skb, port_guid.vf = ivi.vf; memcpy(vf_mac.mac, ivi.mac, sizeof(ivi.mac)); + memset(&vf_broadcast, 0, sizeof(vf_broadcast)); memcpy(vf_broadcast.broadcast, dev->broadcast, dev->addr_len); vf_vlan.vlan = ivi.vlan; vf_vlan.qos = ivi.qos; From c6bebaa744f7579eb72800a262fbfeb93e40db04 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 30 Apr 2026 16:48:36 +0000 Subject: [PATCH 164/525] ipv4: igmp: annotate data-races in igmp_heard_query() Multiple cpus can run igmp_heard_query() concurrently. Add missing READ_ONCE()/WRITE_ONCE() over following in_dev fields. - mr_qrv - mr_qi - mr_qri - mr_v1_seen - mr_v2_seen Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot+ae9a171f239b14485310@syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/69f38675.050a0220.3cbe47.0002.GAE@google.com Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260430164836.872079-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/ipv4/igmp.c | 58 ++++++++++++++++++++++++++++++------------------- 1 file changed, 36 insertions(+), 22 deletions(-) diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index a674fb44ec25..a9ad39064f3b 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -122,16 +122,29 @@ * contradict to specs provided this delay is small enough. */ -#define IGMP_V1_SEEN(in_dev) \ - (IPV4_DEVCONF_ALL_RO(dev_net(in_dev->dev), FORCE_IGMP_VERSION) == 1 || \ - IN_DEV_CONF_GET((in_dev), FORCE_IGMP_VERSION) == 1 || \ - ((in_dev)->mr_v1_seen && \ - time_before(jiffies, (in_dev)->mr_v1_seen))) -#define IGMP_V2_SEEN(in_dev) \ - (IPV4_DEVCONF_ALL_RO(dev_net(in_dev->dev), FORCE_IGMP_VERSION) == 2 || \ - IN_DEV_CONF_GET((in_dev), FORCE_IGMP_VERSION) == 2 || \ - ((in_dev)->mr_v2_seen && \ - time_before(jiffies, (in_dev)->mr_v2_seen))) +static bool IGMP_V1_SEEN(const struct in_device *in_dev) +{ + unsigned long seen; + + if (IPV4_DEVCONF_ALL_RO(dev_net(in_dev->dev), FORCE_IGMP_VERSION) == 1) + return true; + if (IN_DEV_CONF_GET((in_dev), FORCE_IGMP_VERSION) == 1) + return true; + seen = READ_ONCE(in_dev->mr_v1_seen); + return seen && time_before(jiffies, seen); +} + +static bool IGMP_V2_SEEN(const struct in_device *in_dev) +{ + unsigned long seen; + + if (IPV4_DEVCONF_ALL_RO(dev_net(in_dev->dev), FORCE_IGMP_VERSION) == 2) + return true; + if (IN_DEV_CONF_GET((in_dev), FORCE_IGMP_VERSION) == 2) + return true; + seen = READ_ONCE(in_dev->mr_v2_seen); + return seen && time_before(jiffies, seen); +} static int unsolicited_report_interval(struct in_device *in_dev) { @@ -954,23 +967,21 @@ static bool igmp_heard_query(struct in_device *in_dev, struct sk_buff *skb, int max_delay; int mark = 0; struct net *net = dev_net(in_dev->dev); - + unsigned long seen; if (len == 8) { + seen = jiffies + READ_ONCE(in_dev->mr_qrv) * READ_ONCE(in_dev->mr_qi) + + READ_ONCE(in_dev->mr_qri); if (ih->code == 0) { /* Alas, old v1 router presents here. */ max_delay = IGMP_QUERY_RESPONSE_INTERVAL; - in_dev->mr_v1_seen = jiffies + - (in_dev->mr_qrv * in_dev->mr_qi) + - in_dev->mr_qri; + WRITE_ONCE(in_dev->mr_v1_seen, seen); group = 0; } else { /* v2 router present */ max_delay = ih->code*(HZ/IGMP_TIMER_SCALE); - in_dev->mr_v2_seen = jiffies + - (in_dev->mr_qrv * in_dev->mr_qi) + - in_dev->mr_qri; + WRITE_ONCE(in_dev->mr_v2_seen, seen); } /* cancel the interface change timer */ WRITE_ONCE(in_dev->mr_ifc_count, 0); @@ -995,6 +1006,8 @@ static bool igmp_heard_query(struct in_device *in_dev, struct sk_buff *skb, if (!max_delay) max_delay = 1; /* can't mod w/ 0 */ } else { /* v3 */ + unsigned long mr_qi; + if (!pskb_may_pull(skb, sizeof(struct igmpv3_query))) return true; @@ -1015,15 +1028,16 @@ static bool igmp_heard_query(struct in_device *in_dev, struct sk_buff *skb, * received value was zero, use the default or statically * configured value. */ - in_dev->mr_qrv = ih3->qrv ?: READ_ONCE(net->ipv4.sysctl_igmp_qrv); - in_dev->mr_qi = IGMPV3_QQIC(ih3->qqic)*HZ ?: IGMP_QUERY_INTERVAL; - + WRITE_ONCE(in_dev->mr_qrv, + ih3->qrv ?: READ_ONCE(net->ipv4.sysctl_igmp_qrv)); + mr_qi = IGMPV3_QQIC(ih3->qqic)*HZ ?: IGMP_QUERY_INTERVAL; + WRITE_ONCE(in_dev->mr_qi, mr_qi); /* RFC3376, 8.3. Query Response Interval: * The number of seconds represented by the [Query Response * Interval] must be less than the [Query Interval]. */ - if (in_dev->mr_qri >= in_dev->mr_qi) - in_dev->mr_qri = (in_dev->mr_qi/HZ - 1)*HZ; + if (READ_ONCE(in_dev->mr_qri) >= mr_qi) + WRITE_ONCE(in_dev->mr_qri, (mr_qi/HZ - 1) * HZ); if (!group) { /* general query */ if (ih3->nsrcs) From a5148bc2fa27092862ac4b9e7b5c8340d60cff34 Mon Sep 17 00:00:00 2001 From: Alex Cheema Date: Wed, 29 Apr 2026 18:57:39 +0100 Subject: [PATCH 165/525] net: usb: cdc_ncm: add Apple Mac USB-C direct networking quirk Apple Silicon Macs expose two CDC NCM "private" data interfaces over USB-C with VID:PID 0x05ac:0x1905 and product string "Mac". This is the same protocol Apple already ships on iPhone (0x05ac:0x12a8) and iPad (0x05ac:0x12ab) for RemoteXPC since iOS 17 -- both data interfaces lack an interrupt status endpoint, so they rely on the FLAG_LINK_INTR- conditional bind path introduced in commit 3ec8d7572a69 ("CDC-NCM: add support for Apple's private interface"). The id_table currently has entries for iPhone and iPad but not for the Mac. Without a match, cdc_ncm falls through to the generic CDC NCM class-match entry, which uses the FLAG_LINK_INTR-having cdc_ncm_info struct, so bind_common() fails on the missing status endpoint and no netdev appears. Add id_table entries for both interface numbers (0 and 2) of the Mac, bound to the existing apple_private_interface_info driver_info. Verified empirically on a Mac Studio M3 Ultra running macOS 26.5: when a Mac is connected via USB-C, ioreg shows VID 0x05ac, PID 0x1905, product string "Mac", with two NCM data interfaces at numbers 0 and 2. The same PID is presented by all current Apple Silicon Mac models (MacBook Pro/Air, Mac mini, Mac Studio across the M-series), mirroring Apple's single-PID-per-family pattern from iPhone/iPad. After this patch, plugging a Mac into a Linux host running the patched kernel produces two enx... interfaces (one per data interface), "ip -br link" lists them as UP, and standard userspace networking (DHCP, NetworkManager shared mode, etc.) works without any modprobe overrides or out-of-tree modules. Signed-off-by: Alex Cheema Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260429175739.34426-1-alex@exolabs.net Signed-off-by: Jakub Kicinski --- drivers/net/usb/cdc_ncm.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/net/usb/cdc_ncm.c b/drivers/net/usb/cdc_ncm.c index bb9929727eb9..0223a172851e 100644 --- a/drivers/net/usb/cdc_ncm.c +++ b/drivers/net/usb/cdc_ncm.c @@ -2012,6 +2012,14 @@ static const struct usb_device_id cdc_devs[] = { .driver_info = (unsigned long)&apple_private_interface_info, }, + /* Mac */ + { USB_DEVICE_INTERFACE_NUMBER(0x05ac, 0x1905, 0), + .driver_info = (unsigned long)&apple_private_interface_info, + }, + { USB_DEVICE_INTERFACE_NUMBER(0x05ac, 0x1905, 2), + .driver_info = (unsigned long)&apple_private_interface_info, + }, + /* Ericsson MBM devices like F5521gw */ { .match_flags = USB_DEVICE_ID_MATCH_INT_INFO | USB_DEVICE_ID_MATCH_VENDOR, From 6d4106e8df94c0c52cf3ca6a6a0d01567fb3844e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 30 Apr 2026 08:00:56 +0000 Subject: [PATCH 166/525] net/sched: sch_pie: annotate more data-races in pie_dump_stats() My prior patch missed few READ_ONCE()/WRITE_ONCE() annotations. Fixes: 5154561d9b11 ("net/sched: sch_pie: annotate data-races in pie_dump_stats()") Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260430080056.35104-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/sched/sch_pie.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/net/sched/sch_pie.c b/net/sched/sch_pie.c index fb53fbf0e328..b41f2def2e2c 100644 --- a/net/sched/sch_pie.c +++ b/net/sched/sch_pie.c @@ -219,16 +219,14 @@ void pie_process_dequeue(struct sk_buff *skb, struct pie_params *params, * packet timestamp. */ if (!params->dq_rate_estimator) { - vars->qdelay = now - pie_get_enqueue_time(skb); + WRITE_ONCE(vars->qdelay, + backlog ? now - pie_get_enqueue_time(skb) : 0); if (vars->dq_tstamp != DTIME_INVALID) dtime = now - vars->dq_tstamp; vars->dq_tstamp = now; - if (backlog == 0) - vars->qdelay = 0; - if (dtime == 0) return; @@ -376,7 +374,7 @@ void pie_calculate_probability(struct pie_params *params, struct pie_vars *vars, if (qdelay > (PSCHED_NS2TICKS(250 * NSEC_PER_MSEC))) delta += MAX_PROB / (100 / 2); - vars->prob += delta; + WRITE_ONCE(vars->prob, vars->prob + delta); if (delta > 0) { /* prevent overflow */ @@ -401,7 +399,7 @@ void pie_calculate_probability(struct pie_params *params, struct pie_vars *vars, if (qdelay == 0 && qdelay_old == 0 && update_prob) /* Reduce drop probability to 98.4% */ - vars->prob -= vars->prob / 64; + WRITE_ONCE(vars->prob, vars->prob - vars->prob / 64); WRITE_ONCE(vars->qdelay, qdelay); vars->backlog_old = backlog; @@ -501,7 +499,7 @@ static int pie_dump_stats(struct Qdisc *sch, struct gnet_dump *d) { struct pie_sched_data *q = qdisc_priv(sch); struct tc_pie_xstats st = { - .prob = q->vars.prob << BITS_PER_BYTE, + .prob = READ_ONCE(q->vars.prob) << BITS_PER_BYTE, .delay = ((u32)PSCHED_TICKS2NS(READ_ONCE(q->vars.qdelay))) / NSEC_PER_USEC, .packets_in = READ_ONCE(q->stats.packets_in), @@ -512,7 +510,7 @@ static int pie_dump_stats(struct Qdisc *sch, struct gnet_dump *d) }; /* avg_dq_rate is only valid if dq_rate_estimator is enabled */ - st.dq_rate_estimating = q->params.dq_rate_estimator; + st.dq_rate_estimating = READ_ONCE(q->params.dq_rate_estimator); /* unscale and return dq_rate in bytes per sec */ if (st.dq_rate_estimating) From 4bc852006b62eae8ea77e797192d089367e854ff Mon Sep 17 00:00:00 2001 From: Sagarika Sharma Date: Thu, 30 Apr 2026 20:09:00 +0000 Subject: [PATCH 167/525] ipv6: update route serial number on NETDEV_CHANGE When using IPv6 ECMP routes, if a netdev listed as a nexthop experiences a carrier change event (e.g., a bond device generating a NETDEV_CHANGE event after its slaves go linkdown), established connections utilizing that nexthop fail to fail over to other available nexthops. Instead, these connections stall or drop. This happens because the IPv6 FIB code does not invalidate the socket's cached destination when a NETDEV_CHANGE event occurs. While fib6_ifdown() correctly marks the nexthop with RTNH_F_LINKDOWN, it leaves the route's serial number unchanged. As a result, sockets with a previously cached dst do not realize the route is no longer viable and continue to try using the non-functional nexthop. This behavior contrasts with IPv4, which actively flushes cached destinations on a NETDEV_CHANGE event (see fib_netdev_event() in net/ipv4/fib_frontend.c). Fix this by updating the route serial number in fib6_ifdown() when setting RTNH_F_LINKDOWN. This invalidates stale cached destinations, forcing sockets to perform a new route lookup and fail over to a functioning nexthop. Fixes: 51ebd3181572 ("ipv6: add support of equal cost multipath (ECMP)") Signed-off-by: Sagarika Sharma Reviewed-by: Kuniyuki Iwashima Reviewed-by: Ido Schimmel Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20260430200909.527827-2-sharmasagarika@google.com Signed-off-by: Jakub Kicinski --- net/ipv6/route.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 19eb6b702227..0dc0316530ca 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -4995,6 +4995,7 @@ static int fib6_ifdown(struct fib6_info *rt, void *p_arg) rt->fib6_flags & (RTF_LOCAL | RTF_ANYCAST)) break; rt->fib6_nh->fib_nh_flags |= RTNH_F_LINKDOWN; + fib6_update_sernum(net, rt); rt6_multipath_rebalance(rt); break; } From d1ae37dc6881a6a9113c8545cdbba731393d8dcd Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Thu, 30 Apr 2026 20:09:01 +0000 Subject: [PATCH 168/525] selftest: net: Add test for TCP flow failover with ECMP routes. Without the previous commit, TCP failed to switch to alternative IPv6 routes immediately upon carrier loss. It would persist with the dead route until reaching the threshold net.ipv4.tcp_retries1, leading to unnecessary delays in failover. Let's add a selftest for this scenario to ensure TCP fails over immediately upon a carrier loss event. Before: TEST: TCP IPv4 failover [ OK ] TEST: TCP IPv6 failover [FAIL] After: TEST: TCP IPv4 failover [ OK ] TEST: TCP IPv6 failover [ OK ] Signed-off-by: Kuniyuki Iwashima Signed-off-by: Sagarika Sharma Link: https://patch.msgid.link/20260430200909.527827-3-sharmasagarika@google.com Signed-off-by: Jakub Kicinski --- tools/testing/selftests/net/Makefile | 1 + .../selftests/net/tcp_ecmp_failover.sh | 216 ++++++++++++++++++ 2 files changed, 217 insertions(+) create mode 100755 tools/testing/selftests/net/tcp_ecmp_failover.sh diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile index a275ed584026..f3da38c54d27 100644 --- a/tools/testing/selftests/net/Makefile +++ b/tools/testing/selftests/net/Makefile @@ -96,6 +96,7 @@ TEST_PROGS := \ srv6_hl2encap_red_l2vpn_test.sh \ srv6_iptunnel_cache.sh \ stress_reuseport_listen.sh \ + tcp_ecmp_failover.sh \ tcp_fastopen_backup_key.sh \ test_bpf.sh \ test_bridge_backup_port.sh \ diff --git a/tools/testing/selftests/net/tcp_ecmp_failover.sh b/tools/testing/selftests/net/tcp_ecmp_failover.sh new file mode 100755 index 000000000000..5768aa8bff6a --- /dev/null +++ b/tools/testing/selftests/net/tcp_ecmp_failover.sh @@ -0,0 +1,216 @@ +#!/bin/bash +# SPDX-License-Identifier: GPL-2.0 +# +# Copyright 2026 Google LLC. +# +# This test verifies TCP flow failover between ECMP routes +# upon carrier loss on the active device. +# +# socat -----------------------------> socat +# | +# .-- veth-c1 -|- veth-s1 --. +# dummy0 -| | |-- dummy0 +# '-- veth-c2 -|- veth-s2 --' +# | +# + +REQUIRE_JQ=no +REQUIRE_MZ=no +NUM_NETIFS=0 + +source forwarding/lib.sh + +CLIENT_IP="10.0.59.1" +SERVER_IP="10.0.92.1" +CLIENT_IP6="2001:db8:5a9a::1" +SERVER_IP6="2001:db8:9292::1" + +setup_server() +{ + IP="ip -n $server" + NS_EXEC="ip netns exec $server" + + $IP link add dummy0 type dummy + $IP link set dummy0 up + + $IP -4 addr add $SERVER_IP/32 dev dummy0 + $IP -6 addr add $SERVER_IP6/128 dev dummy0 nodad + + $IP link set veth-s1 up + $IP link set veth-s2 up + + $IP -4 addr add 192.168.1.2/24 dev veth-s1 + $IP -4 addr add 192.168.2.2/24 dev veth-s2 + + $IP -4 route add $CLIENT_IP/32 \ + nexthop via 192.168.1.1 dev veth-s1 weight 1 \ + nexthop via 192.168.2.1 dev veth-s2 weight 1 + + $IP -6 addr add 2001:db8:1::2/64 dev veth-s1 nodad + $IP -6 addr add 2001:db8:2::2/64 dev veth-s2 nodad + + $IP -6 route add $CLIENT_IP6/128 \ + nexthop via 2001:db8:1::1 dev veth-s1 weight 1 \ + nexthop via 2001:db8:2::1 dev veth-s2 weight 1 +} + +setup_client() +{ + IP="ip -n $client" + NS_EXEC="ip netns exec $client" + + $IP link add dummy0 type dummy + $IP link set dummy0 up + + $IP -4 addr add $CLIENT_IP/32 dev dummy0 + $IP -6 addr add $CLIENT_IP6/128 dev dummy0 nodad + + $IP link set veth-c1 up + $IP link set veth-c2 up + + $IP -4 addr add 192.168.1.1/24 dev veth-c1 + $IP -4 addr add 192.168.2.1/24 dev veth-c2 + + $IP -4 route add $SERVER_IP/32 \ + nexthop via 192.168.1.2 dev veth-c1 weight 1 \ + nexthop via 192.168.2.2 dev veth-c2 weight 1 + + $IP -6 addr add 2001:db8:1::1/64 dev veth-c1 nodad + $IP -6 addr add 2001:db8:2::1/64 dev veth-c2 nodad + + $IP -6 route add $SERVER_IP6/128 \ + nexthop via 2001:db8:1::2 dev veth-c1 weight 1 \ + nexthop via 2001:db8:2::2 dev veth-c2 weight 1 + + # By default, tcp_retries1=3 triggers a route refresh + # after 3 retransmits (~5s). Ensure this never occurs + # for test stability. + $NS_EXEC sysctl -qw net.ipv4.tcp_retries1=100 + + # When NETDEV_CHANGE is issued for a dev tied to an ECMP + # route, RTNH_F_LINKDOWN is flagged and the sernum is + # bumped to invalidate the route via sk_dst_check(). + # + # Without ignore_routes_with_linkdown=1, subsequent + # lookups may still select the same RTNH_F_LINKDOWN route. + $NS_EXEC sysctl -qw net.ipv4.conf.veth-c1.ignore_routes_with_linkdown=1 + $NS_EXEC sysctl -qw net.ipv4.conf.veth-c2.ignore_routes_with_linkdown=1 + + $NS_EXEC sysctl -qw net.ipv6.conf.veth-c1.ignore_routes_with_linkdown=1 + $NS_EXEC sysctl -qw net.ipv6.conf.veth-c2.ignore_routes_with_linkdown=1 +} + +setup() +{ + setup_ns client server + + ip -n "$client" link add veth-c1 type veth peer veth-s1 netns "$server" + ip -n "$client" link add veth-c2 type veth peer veth-s2 netns "$server" + + setup_server + setup_client +} + +cleanup() +{ + cleanup_all_ns > /dev/null 2>&1 +} + +tcp_ecmp_failover() +{ + local pf=$1; shift + local server_ip=$1; shift + local client_ip=$1; shift + + RET=0 + + tcpdump_start veth-s1 "$server" + tcpdump_start veth-s2 "$server" + + ip netns exec "$server" \ + socat -u TCP-LISTEN:8080,pf="$pf",bind="$server_ip",reuseaddr /dev/null & + server_pid=$! + + # Wait for server to start listening. + # Sometimes client fails without this sleep. + sleep 1 + + ip netns exec "$client" \ + socat -u /dev/zero TCP:"$server_ip":8080,pf="$pf",bind="$client_ip" & + client_pid=$! + + # To capture enough packets. + sleep 3 + + tcpdump_stop veth-s1 + tcpdump_stop veth-s2 + + pkts_s1=$(tcpdump_show veth-s1 | wc -l) + pkts_s2=$(tcpdump_show veth-s2 | wc -l) + + tcpdump_cleanup veth-s1 + tcpdump_cleanup veth-s2 + + # Detect the device chosen by the client + if [ "$pkts_s1" -gt "$pkts_s2" ]; then + veth_down=veth-s1 + veth_up=veth-s2 + else + veth_down=veth-s2 + veth_up=veth-s1 + fi + + # Taking down $veth_down causes its peer to lose carrier, + # triggering NETDEV_CHANGE. This flags RTNH_F_LINKDOWN + # and bumps the sernum for the route associated with that + # peer, invalidating the cached dst in the TCP socket. + # + # Consequently, sk_dst_check() fails, forcing the subsequent + # lookup to select the remaining healthy route via $veth_up. + ip -n "$server" link set "$veth_down" down + + tcpdump_start "$veth_up" "$server" + + # To capture enough packets. + sleep 3 + + tcpdump_stop "$veth_up" + + kill -9 "$client_pid" > /dev/null 2>&1 + kill -9 "$server_pid" > /dev/null 2>&1 + wait 2> /dev/null + + pkts=$(tcpdump_show $veth_up | wc -l) + + tcpdump_cleanup "$veth_up" + + if [ "$pkts" -lt 1000 ]; then + RET=$ksft_fail + fi +} + +test_ipv4() +{ + setup + tcp_ecmp_failover IPv4 $SERVER_IP $CLIENT_IP + log_test "TCP IPv4 failover" + cleanup +} + +test_ipv6() +{ + setup + tcp_ecmp_failover IPv6 "[$SERVER_IP6]" "[$CLIENT_IP6]" + log_test "TCP IPv6 failover" + cleanup +} + +require_command socat +require_command tcpdump + +trap cleanup EXIT + +test_ipv4 +test_ipv6 + +exit "$EXIT_STATUS" From b1f1e80620deb49daf63c2e677046599b693dc1f Mon Sep 17 00:00:00 2001 From: DaeMyung Kang Date: Tue, 28 Apr 2026 23:08:54 +0900 Subject: [PATCH 169/525] ksmbd: centralize ksmbd_conn final release to plug transport leak ksmbd_conn_free() is one of four sites that can observe the last refcount drop of a struct ksmbd_conn. The other three fs/smb/server/connection.c ksmbd_conn_r_count_dec() fs/smb/server/oplock.c __free_opinfo() fs/smb/server/vfs_cache.c session_fd_check() end the conn with a bare kfree(), skipping ida_destroy(&conn->async_ida) and conn->transport->ops->free_transport(conn->transport). Whenever one of them is the last putter, the embedded async_ida and the entire transport struct leak -- for TCP, that is also the struct socket and the kvec iov. __free_opinfo() being a final putter is not theoretical. opinfo_put() queues the callback via call_rcu(&opinfo->rcu, free_opinfo_rcu), so ksmbd_server_terminate_conn() can deposit N opinfo releases in RCU and have ksmbd_conn_free() run in the handler thread before any of them fire. ksmbd_conn_free() then observes refcnt > 0 and short-circuits; the last RCU-delivered __free_opinfo() falls onto its bare kfree(conn) branch and the transport is lost. A/B validation in a QEMU/virtme guest, mounting //127.0.0.1/testshare: each iteration holds 8 files open via sleep processes, force-closes TCP with "ss -K sport = :445", kills the holders, lazy-umounts; repeated 10 times, then ksmbd shutdown and kmemleak scan. state conn_alloc conn_free tcp_free opi_rcu kmemleak ---------- ---------- --------- -------- ------- -------- pre-patch 20 20 10 160 7 with patch 20 20 20 160 0 Pre-patch conn_free=20 with tcp_free=10 directly demonstrates the bare-kfree paths skipping transport cleanup; kmemleak backtraces point into struct tcp_transport / iov. With this patch tcp_free matches conn_free at 20/20 and kmemleak is clean. Move the per-struct final release into __ksmbd_conn_release_work() and route the three bare-kfree final-put sites through a new ksmbd_conn_put(). Those sites now pair ida_destroy() and free_transport() with kfree(conn) regardless of which holder happens to release the last reference. stop_sessions() only triggers the transport shutdown and does not itself drop the last conn reference, so it is unaffected. The centralized release reaches sock_release() -> tcp_close() -> lock_sock_nested() (might_sleep) from every final putter, including __free_opinfo() invoked from an RCU softirq callback, which trips CONFIG_DEBUG_ATOMIC_SLEEP. Defer the release to a dedicated ksmbd_conn_wq workqueue so ksmbd_conn_put() is safe from any non-sleeping context. Make ksmbd_file own a strong connection reference while fp->conn is non-NULL so durable-preserve and final-close paths cannot dereference a stale connection. ksmbd_open_fd() and ksmbd_reopen_durable_fd() take the reference via ksmbd_conn_get() (the latter also reorders the fp->conn / fp->tcon assignments before __open_id() so the published fp is never observed with fp->conn == NULL); session_fd_check() and __ksmbd_close_fd() drop it via ksmbd_conn_put(). With that invariant, session_fd_check() can take a local conn pointer once and use it across the m_op_list and lock_list iterations even though op->conn puts may otherwise drop the last reference. At module exit the workqueue is flushed and destroyed after rcu_barrier(), so any release queued by a trailing RCU callback is drained before the inode hash and module text go away. Fixes: ee426bfb9d09 ("ksmbd: add refcnt to ksmbd_conn struct") Signed-off-by: DaeMyung Kang Acked-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/connection.c | 101 +++++++++++++++++++++++++++++++------ fs/smb/server/connection.h | 6 +++ fs/smb/server/oplock.c | 7 +-- fs/smb/server/server.c | 12 +++++ fs/smb/server/vfs_cache.c | 60 ++++++++++++++++++---- 5 files changed, 156 insertions(+), 30 deletions(-) diff --git a/fs/smb/server/connection.c b/fs/smb/server/connection.c index c5aac4946cbe..95654a808162 100644 --- a/fs/smb/server/connection.c +++ b/fs/smb/server/connection.c @@ -79,6 +79,81 @@ static int create_proc_clients(void) { return 0; } static void delete_proc_clients(void) {} #endif +static struct workqueue_struct *ksmbd_conn_wq; + +int ksmbd_conn_wq_init(void) +{ + ksmbd_conn_wq = alloc_workqueue("ksmbd-conn-release", + WQ_UNBOUND | WQ_MEM_RECLAIM, 0); + if (!ksmbd_conn_wq) + return -ENOMEM; + return 0; +} + +void ksmbd_conn_wq_destroy(void) +{ + if (ksmbd_conn_wq) { + destroy_workqueue(ksmbd_conn_wq); + ksmbd_conn_wq = NULL; + } +} + +/* + * __ksmbd_conn_release_work() - perform the final, once-per-struct cleanup + * of a ksmbd_conn whose refcount has just dropped to zero. + * + * This is the common release path used by ksmbd_conn_put() for the embedded + * state that outlives the connection thread: async_ida and the attached + * transport (which owns the socket and iov for TCP). Called from a workqueue + * so that sleep-allowed teardown (sock_release -> tcp_close -> + * lock_sock_nested) never runs from an RCU softirq callback (free_opinfo_rcu) + * or any other non-sleeping putter context. + */ +static void __ksmbd_conn_release_work(struct work_struct *work) +{ + struct ksmbd_conn *conn = + container_of(work, struct ksmbd_conn, release_work); + + ida_destroy(&conn->async_ida); + conn->transport->ops->free_transport(conn->transport); + kfree(conn); +} + +/** + * ksmbd_conn_get() - take a reference on @conn and return it. + * + * Returns @conn unchanged so callers can write + * "fp->conn = ksmbd_conn_get(work->conn);" in one expression. Returns NULL + * if @conn is NULL. + */ +struct ksmbd_conn *ksmbd_conn_get(struct ksmbd_conn *conn) +{ + if (!conn) + return NULL; + + atomic_inc(&conn->refcnt); + return conn; +} + +/** + * ksmbd_conn_put() - drop a reference and, if it was the last, queue the + * release onto ksmbd_conn_wq so it runs from process context. + * + * Callable from any context including RCU softirq callbacks and non-sleeping + * locks; the actual release is deferred to the workqueue. ksmbd_conn_wq is + * created in ksmbd_server_init() before any conn can be allocated and is + * destroyed in ksmbd_server_exit() after rcu_barrier(), so it is always + * non-NULL while a conn reference is held. + */ +void ksmbd_conn_put(struct ksmbd_conn *conn) +{ + if (!conn) + return; + + if (atomic_dec_and_test(&conn->refcnt)) + queue_work(ksmbd_conn_wq, &conn->release_work); +} + /** * ksmbd_conn_free() - free resources of the connection instance * @@ -93,23 +168,19 @@ void ksmbd_conn_free(struct ksmbd_conn *conn) hash_del(&conn->hlist); up_write(&conn_list_lock); + /* + * request_buf / preauth_info / mechToken are only ever accessed by the + * connection handler thread that owns @conn. ksmbd_conn_free() is + * called from the transport free_transport() path when that thread is + * exiting, so it is safe to release them unconditionally even when + * ksmbd_conn_put() below is not the final putter (oplock / ksmbd_file + * holders only retain the conn pointer, not these per-thread buffers). + */ xa_destroy(&conn->sessions); kvfree(conn->request_buf); kfree(conn->preauth_info); kfree(conn->mechToken); - if (atomic_dec_and_test(&conn->refcnt)) { - /* - * async_ida is embedded in struct ksmbd_conn, so pair - * ida_destroy() with the final kfree() rather than with - * the unconditional field teardown above. This keeps - * the IDA valid for the entire lifetime of the struct, - * even while other refcount holders (oplock / vfs - * durable handles) still reference the connection. - */ - ida_destroy(&conn->async_ida); - conn->transport->ops->free_transport(conn->transport); - kfree(conn); - } + ksmbd_conn_put(conn); } /** @@ -136,6 +207,7 @@ struct ksmbd_conn *ksmbd_conn_alloc(void) conn->um = ERR_PTR(-EOPNOTSUPP); if (IS_ERR(conn->um)) conn->um = NULL; + INIT_WORK(&conn->release_work, __ksmbd_conn_release_work); atomic_set(&conn->req_running, 0); atomic_set(&conn->r_count, 0); atomic_set(&conn->refcnt, 1); @@ -512,8 +584,7 @@ void ksmbd_conn_r_count_dec(struct ksmbd_conn *conn) if (!atomic_dec_return(&conn->r_count) && waitqueue_active(&conn->r_count_q)) wake_up(&conn->r_count_q); - if (atomic_dec_and_test(&conn->refcnt)) - kfree(conn); + ksmbd_conn_put(conn); } int ksmbd_conn_transport_init(void) diff --git a/fs/smb/server/connection.h b/fs/smb/server/connection.h index de2d46941c93..e074be942582 100644 --- a/fs/smb/server/connection.h +++ b/fs/smb/server/connection.h @@ -16,6 +16,7 @@ #include #include #include +#include #include "smb_common.h" #include "ksmbd_work.h" @@ -120,6 +121,7 @@ struct ksmbd_conn { bool binding; atomic_t refcnt; bool is_aapl; + struct work_struct release_work; }; struct ksmbd_conn_ops { @@ -164,6 +166,10 @@ void ksmbd_conn_wait_idle(struct ksmbd_conn *conn); int ksmbd_conn_wait_idle_sess_id(struct ksmbd_conn *curr_conn, u64 sess_id); struct ksmbd_conn *ksmbd_conn_alloc(void); void ksmbd_conn_free(struct ksmbd_conn *conn); +struct ksmbd_conn *ksmbd_conn_get(struct ksmbd_conn *conn); +void ksmbd_conn_put(struct ksmbd_conn *conn); +int ksmbd_conn_wq_init(void); +void ksmbd_conn_wq_destroy(void); bool ksmbd_conn_lookup_dialect(struct ksmbd_conn *c); int ksmbd_conn_write(struct ksmbd_work *work); int ksmbd_conn_rdma_read(struct ksmbd_conn *conn, diff --git a/fs/smb/server/oplock.c b/fs/smb/server/oplock.c index cd3f28b0e7cb..8feca02ddbf2 100644 --- a/fs/smb/server/oplock.c +++ b/fs/smb/server/oplock.c @@ -30,7 +30,6 @@ static DEFINE_RWLOCK(lease_list_lock); static struct oplock_info *alloc_opinfo(struct ksmbd_work *work, u64 id, __u16 Tid) { - struct ksmbd_conn *conn = work->conn; struct ksmbd_session *sess = work->sess; struct oplock_info *opinfo; @@ -39,7 +38,7 @@ static struct oplock_info *alloc_opinfo(struct ksmbd_work *work, return NULL; opinfo->sess = sess; - opinfo->conn = conn; + opinfo->conn = ksmbd_conn_get(work->conn); opinfo->level = SMB2_OPLOCK_LEVEL_NONE; opinfo->op_state = OPLOCK_STATE_NONE; opinfo->pending_break = 0; @@ -50,7 +49,6 @@ static struct oplock_info *alloc_opinfo(struct ksmbd_work *work, init_waitqueue_head(&opinfo->oplock_brk); atomic_set(&opinfo->refcount, 1); atomic_set(&opinfo->breaking_cnt, 0); - atomic_inc(&opinfo->conn->refcnt); return opinfo; } @@ -132,8 +130,7 @@ static void __free_opinfo(struct oplock_info *opinfo) { if (opinfo->is_lease) free_lease(opinfo); - if (opinfo->conn && atomic_dec_and_test(&opinfo->conn->refcnt)) - kfree(opinfo->conn); + ksmbd_conn_put(opinfo->conn); kfree(opinfo); } diff --git a/fs/smb/server/server.c b/fs/smb/server/server.c index 58ef02c423fc..5d799b2d4c62 100644 --- a/fs/smb/server/server.c +++ b/fs/smb/server/server.c @@ -596,8 +596,14 @@ static int __init ksmbd_server_init(void) if (ret) goto err_crypto_destroy; + ret = ksmbd_conn_wq_init(); + if (ret) + goto err_workqueue_destroy; + return 0; +err_workqueue_destroy: + ksmbd_workqueue_destroy(); err_crypto_destroy: ksmbd_crypto_destroy(); err_release_inode_hash: @@ -623,6 +629,12 @@ static void __exit ksmbd_server_exit(void) { ksmbd_server_shutdown(); rcu_barrier(); + /* + * ksmbd_conn_put() defers the final release onto ksmbd_conn_wq, + * so drain it after rcu_barrier() has fired any pending RCU + * callbacks that may have queued a release. + */ + ksmbd_conn_wq_destroy(); ksmbd_release_inode_hash(); } diff --git a/fs/smb/server/vfs_cache.c b/fs/smb/server/vfs_cache.c index 3551f01a3fa0..4a18107937cc 100644 --- a/fs/smb/server/vfs_cache.c +++ b/fs/smb/server/vfs_cache.c @@ -475,6 +475,17 @@ static void __ksmbd_close_fd(struct ksmbd_file_table *ft, struct ksmbd_file *fp) kfree(smb_lock); } + /* + * Drop fp's strong reference on conn (taken in ksmbd_open_fd() / + * ksmbd_reopen_durable_fd()). Durable fps that reached the + * scavenger have already had fp->conn cleared by session_fd_check(), + * in which case there is nothing to drop here. + */ + if (fp->conn) { + ksmbd_conn_put(fp->conn); + fp->conn = NULL; + } + if (ksmbd_stream_fd(fp)) kfree(fp->stream.name); kfree(fp->owner.name); @@ -752,7 +763,14 @@ struct ksmbd_file *ksmbd_open_fd(struct ksmbd_work *work, struct file *filp) atomic_set(&fp->refcount, 1); fp->filp = filp; - fp->conn = work->conn; + /* + * fp owns a strong reference on fp->conn for as long as fp->conn is + * non-NULL, so session_fd_check() and __ksmbd_close_fd() never + * dereference a dangling pointer. Paired with ksmbd_conn_put() in + * session_fd_check() (durable preserve), in __ksmbd_close_fd() + * (final close), and on the error paths below. + */ + fp->conn = ksmbd_conn_get(work->conn); fp->tcon = work->tcon; fp->volatile_id = KSMBD_NO_FID; fp->persistent_id = KSMBD_NO_FID; @@ -774,6 +792,8 @@ struct ksmbd_file *ksmbd_open_fd(struct ksmbd_work *work, struct file *filp) return fp; err_out: + /* fp->conn was set and refcounted before every branch here. */ + ksmbd_conn_put(fp->conn); kmem_cache_free(filp_cache, fp); return ERR_PTR(ret); } @@ -1062,25 +1082,32 @@ static bool session_fd_check(struct ksmbd_tree_connect *tcon, if (!is_reconnectable(fp)) return false; + if (WARN_ON_ONCE(!fp->conn)) + return false; + if (ksmbd_vfs_copy_durable_owner(fp, user)) return false; + /* + * fp owns a strong reference on fp->conn (taken in ksmbd_open_fd() + * / ksmbd_reopen_durable_fd()), so conn stays valid for the whole + * body of this function regardless of any op->conn puts below. + */ conn = fp->conn; ci = fp->f_ci; down_write(&ci->m_lock); list_for_each_entry_rcu(op, &ci->m_op_list, op_entry) { if (op->conn != conn) continue; - if (op->conn && atomic_dec_and_test(&op->conn->refcnt)) - kfree(op->conn); + ksmbd_conn_put(op->conn); op->conn = NULL; } up_write(&ci->m_lock); list_for_each_entry_safe(smb_lock, tmp_lock, &fp->lock_list, flist) { - spin_lock(&fp->conn->llist_lock); + spin_lock(&conn->llist_lock); list_del_init(&smb_lock->clist); - spin_unlock(&fp->conn->llist_lock); + spin_unlock(&conn->llist_lock); } fp->conn = NULL; @@ -1091,6 +1118,8 @@ static bool session_fd_check(struct ksmbd_tree_connect *tcon, fp->durable_scavenger_timeout = jiffies_to_msecs(jiffies) + fp->durable_timeout; + /* Drop fp's own reference on conn. */ + ksmbd_conn_put(conn); return true; } @@ -1178,15 +1207,27 @@ int ksmbd_reopen_durable_fd(struct ksmbd_work *work, struct ksmbd_file *fp) old_f_state = fp->f_state; fp->f_state = FP_NEW; + + /* + * Initialize fp's connection binding before publishing fp into the + * session's file table. If __open_id() is ordered first, a + * concurrent teardown that iterates the table can observe a valid + * volatile_id with fp->conn == NULL and preserve a + * partially-initialized fp. fp owns a strong reference on the new + * conn (see ksmbd_open_fd()); undo it on __open_id() failure. + */ + fp->conn = ksmbd_conn_get(conn); + fp->tcon = work->tcon; + __open_id(&work->sess->file_table, fp, OPEN_ID_TYPE_VOLATILE_ID); if (!has_file_id(fp->volatile_id)) { + fp->conn = NULL; + fp->tcon = NULL; + ksmbd_conn_put(conn); fp->f_state = old_f_state; return -EBADF; } - fp->conn = conn; - fp->tcon = work->tcon; - list_for_each_entry(smb_lock, &fp->lock_list, flist) { spin_lock(&conn->llist_lock); list_add_tail(&smb_lock->clist, &conn->lock_list); @@ -1198,8 +1239,7 @@ int ksmbd_reopen_durable_fd(struct ksmbd_work *work, struct ksmbd_file *fp) list_for_each_entry_rcu(op, &ci->m_op_list, op_entry) { if (op->conn) continue; - op->conn = fp->conn; - atomic_inc(&op->conn->refcnt); + op->conn = ksmbd_conn_get(fp->conn); } up_write(&ci->m_lock); From a42896bebfcc287ed1e61d820a888e33b1eb80ce Mon Sep 17 00:00:00 2001 From: DaeMyung Kang Date: Tue, 28 Apr 2026 23:08:55 +0900 Subject: [PATCH 170/525] ksmbd: harden file lifetime during session teardown __close_file_table_ids() is the per-session teardown that closes every fp belonging to a session (or to one tree connect on that session) by walking the session's volatile-id idr. The current loop has three related problems on busy or racing workloads: * Sleeping under ft->lock. The session-teardown skip callback, session_fd_check(), already sleeps in ksmbd_vfs_copy_durable_owner() -> kstrdup(GFP_KERNEL) and down_write(&fp->f_ci->m_lock) (a rw_semaphore). Running the callback inside write_lock(&ft->lock) trips CONFIG_DEBUG_ATOMIC_SLEEP / CONFIG_PROVE_LOCKING on a durable-fd workload. * Refcount accounting blind to f_state. The unconditional atomic_dec_and_test(&fp->refcount) does not distinguish FP_INITED (idr-owned reference still intact) from FP_CLOSED (an earlier ksmbd_close_fd() already consumed the idr-owned reference while leaving fp in the idr because a holder kept refcount non-zero). When the latter races with teardown the same path over-decrements into a holder reference and ksmbd_fd_put() later UAFs that holder. * FP_NEW window. Between __open_id() publishing fp into the session idr and ksmbd_update_fstate(..., FP_INITED) committing the transition at the end of smb2_open(), an fp is in FP_NEW and an intervening teardown that takes a transient reference and unpublishes the volatile id leaves the original idr-owned reference orphaned -- the opener is unaware that fp has been unpublished, returns success to the client, and the fp leaks at refcount = 1. Refactor __close_file_table_ids() to take a transient reference on fp and unpublish fp from the session idr *under ft->lock* before calling skip() outside the lock. A transient ref protects lifetime but not concurrent field mutation, so the idr_remove() is what keeps __ksmbd_lookup_fd() through this session's idr from granting a new ksmbd_fp_get() reference to an fp whose fp->conn / fp->tcon / fp->volatile_id / op->conn / lock_list links are about to be rewritten by session_fd_check(). Durable reconnect is unaffected because it reaches fp through the global durable table (ksmbd_lookup_durable_fd -> global_ft). Decide n_to_drop together with any FP_INITED -> FP_CLOSED transition under ft->lock so teardown and ksmbd_close_fd() never both consume the idr-owned reference. See ksmbd_mark_fp_closed() for the per-state accounting. For the FP_NEW path to be safe, the opener has to learn that fp was unpublished: ksmbd_update_fstate() now returns -ENOENT when an FP_NEW -> FP_INITED transition finds f_state already advanced or the volatile id cleared (both committed by teardown under ft->lock); smb2_open() propagates that as STATUS_OBJECT_NAME_INVALID and drops the original reference via ksmbd_fd_put(). The list removal cannot be left for a deferred final putter because fp->volatile_id has already been cleared and __ksmbd_remove_fd() will intentionally skip both idr_remove() and list_del_init(). Move the m_fp_list unlink in __ksmbd_remove_fd() above the volatile-id check so that an FP_NEW fp that happened to be added to m_fp_list (smb2_open() adds fp->node before ksmbd_update_fstate() runs) is still cleaned up on the deferred putter path; list_del_init() on an empty node is a no-op and remains safe for fps that were never added. Add a defensive guard in session_fd_check() that refuses non-FP_INITED fps so that even if a teardown reaches an FP_NEW fp it falls into the close branch (where the n_to_drop = 1 accounting keeps the opener's reference alive) instead of the durable-preserve branch (which mutates fp->conn / fp->tcon). Validation on a debug kernel additionally built with CONFIG_DEBUG_LIST and CONFIG_DEBUG_OBJECTS_WORK used a same-session two-tcon workload (open/write storm on one tcon, 50 tree disconnects on the other) and reported no list-corruption, work_struct ODEBUG, sleep-in-atomic, lockdep or kmemleak reports. Reverting only the __close_file_table_ids() hunk while keeping a forced-is_reconnectable() harness produced the expected sleep-in-atomic at vfs_cache.c:1095, confirming the ft->lock-out-of-sleepable-skip discipline. KASAN-enabled direct SMB2 coverage with durable handles enabled exercised ksmbd_close_tree_conn_fds(), ksmbd_close_session_fds(), the FP_NEW failure path, tree_conn_fd_check(), and a non-zero session_fd_check() durable-preserve return. This produced no KASAN, DEBUG_LIST, ODEBUG, or WARNING reports. Fixes: f44158485826 ("cifsd: add file operations") Signed-off-by: DaeMyung Kang Acked-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/smb2pdu.c | 6 +- fs/smb/server/vfs_cache.c | 179 +++++++++++++++++++++++++++++++++----- fs/smb/server/vfs_cache.h | 4 +- 3 files changed, 164 insertions(+), 25 deletions(-) diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index 47b7af631f7b..62d4399a993d 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -3767,8 +3767,10 @@ int smb2_open(struct ksmbd_work *work) err_out2: if (!rc) { - ksmbd_update_fstate(&work->sess->file_table, fp, FP_INITED); - rc = ksmbd_iov_pin_rsp(work, (void *)rsp, iov_len); + rc = ksmbd_update_fstate(&work->sess->file_table, fp, + FP_INITED); + if (!rc) + rc = ksmbd_iov_pin_rsp(work, (void *)rsp, iov_len); } if (rc) { if (rc == -EINVAL) diff --git a/fs/smb/server/vfs_cache.c b/fs/smb/server/vfs_cache.c index 4a18107937cc..dc4037ef1834 100644 --- a/fs/smb/server/vfs_cache.c +++ b/fs/smb/server/vfs_cache.c @@ -431,13 +431,13 @@ static void ksmbd_remove_durable_fd(struct ksmbd_file *fp) static void __ksmbd_remove_fd(struct ksmbd_file_table *ft, struct ksmbd_file *fp) { - if (!has_file_id(fp->volatile_id)) - return; - down_write(&fp->f_ci->m_lock); list_del_init(&fp->node); up_write(&fp->f_ci->m_lock); + if (!has_file_id(fp->volatile_id)) + return; + write_lock(&ft->lock); idr_remove(ft->idr, fp->volatile_id); write_unlock(&ft->lock); @@ -798,15 +798,58 @@ struct ksmbd_file *ksmbd_open_fd(struct ksmbd_work *work, struct file *filp) return ERR_PTR(ret); } -void ksmbd_update_fstate(struct ksmbd_file_table *ft, struct ksmbd_file *fp, - unsigned int state) +/** + * ksmbd_update_fstate() - update an fp state under the file-table lock + * @ft: file table that publishes @fp's volatile id + * @fp: file pointer to update + * @state: new state + * + * Return: 0 on success. The FP_NEW -> FP_INITED transition is special: + * -ENOENT if teardown already unpublished @fp by advancing the state or + * clearing the volatile id. Other state updates preserve the historical + * fire-and-forget behavior. + */ +int ksmbd_update_fstate(struct ksmbd_file_table *ft, struct ksmbd_file *fp, + unsigned int state) { + int ret; + if (!fp) - return; + return -ENOENT; write_lock(&ft->lock); - fp->f_state = state; + if (state == FP_INITED && + (fp->f_state != FP_NEW || !has_file_id(fp->volatile_id))) { + ret = -ENOENT; + } else { + fp->f_state = state; + ret = 0; + } write_unlock(&ft->lock); + + return ret; +} + +/* + * ksmbd_mark_fp_closed() - mark fp closed under ft->lock and return how many + * refs the teardown path owns. + * + * FP_INITED has a normal idr-owned reference, so teardown owns both that + * reference and the transient lookup reference. FP_NEW is still owned by the + * in-flight opener/reopener, which will drop the original reference after + * ksmbd_update_fstate(..., FP_INITED) observes the cleared volatile id. + * FP_CLOSED on entry means an earlier ksmbd_close_fd() already consumed the + * idr-owned ref. + */ +static int ksmbd_mark_fp_closed(struct ksmbd_file *fp) +{ + if (fp->f_state == FP_INITED) { + set_close_state_blocked_works(fp); + fp->f_state = FP_CLOSED; + return 2; + } + + return 1; } static int @@ -814,7 +857,8 @@ __close_file_table_ids(struct ksmbd_session *sess, struct ksmbd_tree_connect *tcon, bool (*skip)(struct ksmbd_tree_connect *tcon, struct ksmbd_file *fp, - struct ksmbd_user *user)) + struct ksmbd_user *user), + bool skip_preserves_fp) { struct ksmbd_file_table *ft = &sess->file_table; struct ksmbd_file *fp; @@ -822,32 +866,120 @@ __close_file_table_ids(struct ksmbd_session *sess, int num = 0; while (1) { + int n_to_drop; + write_lock(&ft->lock); fp = idr_get_next(ft->idr, &id); if (!fp) { write_unlock(&ft->lock); break; } - - if (skip(tcon, fp, sess->user) || - !atomic_dec_and_test(&fp->refcount)) { + if (!atomic_inc_not_zero(&fp->refcount)) { id++; write_unlock(&ft->lock); continue; } - set_close_state_blocked_works(fp); - idr_remove(ft->idr, fp->volatile_id); - fp->volatile_id = KSMBD_NO_FID; - write_unlock(&ft->lock); + if (skip_preserves_fp) { + /* + * Session teardown: skip() is session_fd_check(), + * which may sleep and mutates fp->conn / fp->tcon / + * fp->volatile_id when it chooses to preserve fp + * for durable reconnect. Unpublish fp from the + * session idr here, under ft->lock, so that + * __ksmbd_lookup_fd() through this session cannot + * grant a new ksmbd_fp_get() reference to an fp + * whose fields are about to be rewritten outside + * the lock. Durable reconnect still reaches fp via + * global_ft. + */ + idr_remove(ft->idr, id); + fp->volatile_id = KSMBD_NO_FID; + write_unlock(&ft->lock); + if (skip(tcon, fp, sess->user)) { + /* + * session_fd_check() has converted fp to + * durable-preserve state and cleared its + * per-conn fields. fp is already unpublished + * above; the original idr-owned ref keeps it + * alive for the durable scavenger. Drop only + * the transient ref. atomic_dec() is safe -- + * atomic_inc_not_zero() succeeded on a + * positive value and we added one more, so + * refcount cannot be zero here. + */ + atomic_dec(&fp->refcount); + id++; + continue; + } + + /* + * Keep the close-state decision under the same lock + * observed by ksmbd_update_fstate(), which is how an + * in-flight FP_NEW opener learns that teardown has + * cleared its volatile id. + */ + write_lock(&ft->lock); + n_to_drop = ksmbd_mark_fp_closed(fp); + write_unlock(&ft->lock); + } else { + /* + * Tree teardown: skip() is tree_conn_fd_check(), a + * cheap pointer compare that doesn't sleep and has + * no side effects, so keep the skip decision plus + * the unpublish-and-mark-closed sequence atomic + * under ft->lock. fps belonging to other tree + * connects (skip() == true) stay fully published in + * the session idr with no lock window. + */ + if (skip(tcon, fp, sess->user)) { + atomic_dec(&fp->refcount); + write_unlock(&ft->lock); + id++; + continue; + } + idr_remove(ft->idr, id); + fp->volatile_id = KSMBD_NO_FID; + n_to_drop = ksmbd_mark_fp_closed(fp); + write_unlock(&ft->lock); + } + + /* + * fp->volatile_id is already cleared to prevent stale idr + * removal from a deferred final close. Remove fp from + * m_fp_list here because __ksmbd_remove_fd() will skip the + * list unlink when volatile_id is KSMBD_NO_FID. + */ down_write(&fp->f_ci->m_lock); list_del_init(&fp->node); up_write(&fp->f_ci->m_lock); - __ksmbd_close_fd(ft, fp); - - num++; + /* + * Drop the references this iteration owns: + * + * n_to_drop == 2: we observed FP_INITED and committed + * the FP_CLOSED transition ourselves, so we own the + * transient (+1) and the still-intact idr-owned ref. + * + * n_to_drop == 1: either a prior ksmbd_close_fd() + * already consumed the idr-owned ref, or fp was still + * FP_NEW and the in-flight opener/reopener must keep + * the original reference until ksmbd_update_fstate() + * observes the cleared volatile id. + * + * If we end up as the final putter, finalize fp and + * account the open_files_count decrement via the caller's + * atomic_sub(num, ...). Otherwise the remaining user's + * ksmbd_fd_put() reaches __put_fd_final(), which does its + * own atomic_dec(&open_files_count), so we must not count + * this fp here -- doing so would double-decrement the + * connection-wide counter. + */ + if (atomic_sub_and_test(n_to_drop, &fp->refcount)) { + __ksmbd_close_fd(NULL, fp); + num++; + } id++; } @@ -1082,6 +1214,9 @@ static bool session_fd_check(struct ksmbd_tree_connect *tcon, if (!is_reconnectable(fp)) return false; + if (fp->f_state != FP_INITED) + return false; + if (WARN_ON_ONCE(!fp->conn)) return false; @@ -1127,7 +1262,8 @@ void ksmbd_close_tree_conn_fds(struct ksmbd_work *work) { int num = __close_file_table_ids(work->sess, work->tcon, - tree_conn_fd_check); + tree_conn_fd_check, + false); atomic_sub(num, &work->conn->stats.open_files_count); } @@ -1136,7 +1272,8 @@ void ksmbd_close_session_fds(struct ksmbd_work *work) { int num = __close_file_table_ids(work->sess, work->tcon, - session_fd_check); + session_fd_check, + true); atomic_sub(num, &work->conn->stats.open_files_count); } @@ -1268,7 +1405,7 @@ void ksmbd_destroy_file_table(struct ksmbd_session *sess) if (!ft->idr) return; - __close_file_table_ids(sess, NULL, session_fd_check); + __close_file_table_ids(sess, NULL, session_fd_check, true); idr_destroy(ft->idr); kfree(ft->idr); ft->idr = NULL; diff --git a/fs/smb/server/vfs_cache.h b/fs/smb/server/vfs_cache.h index 866f32c10d4d..e6871266a94b 100644 --- a/fs/smb/server/vfs_cache.h +++ b/fs/smb/server/vfs_cache.h @@ -172,8 +172,8 @@ int ksmbd_close_inode_fds(struct ksmbd_work *work, struct inode *inode); int ksmbd_init_global_file_table(void); void ksmbd_free_global_file_table(void); void ksmbd_set_fd_limit(unsigned long limit); -void ksmbd_update_fstate(struct ksmbd_file_table *ft, struct ksmbd_file *fp, - unsigned int state); +int ksmbd_update_fstate(struct ksmbd_file_table *ft, struct ksmbd_file *fp, + unsigned int state); bool ksmbd_vfs_compare_durable_owner(struct ksmbd_file *fp, struct ksmbd_user *user); From bf736184d063da1a552ffeff0481813599a182cc Mon Sep 17 00:00:00 2001 From: DaeMyung Kang Date: Tue, 28 Apr 2026 23:08:56 +0900 Subject: [PATCH 171/525] ksmbd: close durable scavenger races against m_fp_list lookups ksmbd_durable_scavenger() has two related races against any walker that iterates f_ci->m_fp_list, including ksmbd_lookup_fd_inode() (used by ksmbd_vfs_rename) and the share-mode checks in fs/smb/server/smb_common.c. (1) fp->node list-head reuse. Durable-preserved handles can remain linked on f_ci->m_fp_list after session teardown so share-mode checks still see them while the handle is reconnectable. The scavenger collected expired handles by adding fp->node to a local scavenger_list after removing them from the global durable idr. Because fp->node is the same list_head used by m_fp_list, list_add(&fp->node, &scavenger_list) overwrites the m_fp_list links and corrupts both lists. CONFIG_DEBUG_LIST can report this on the share-mode walk path. (2) Refcount race against m_fp_list walkers. The scavenger qualifies an expired durable handle with atomic_read(&fp->refcount) > 1 and fp->conn under global_ft.lock, removes fp from global_ft, then drops global_ft.lock before unlinking fp from m_fp_list and freeing it. During that gap fp is still linked on m_fp_list with f_state == FP_INITED. ksmbd_lookup_fd_inode() under m_lock read calls ksmbd_fp_get() (atomic_inc_not_zero on refcount that is still 1) and takes a live reference; the scavenger then unlinks and frees fp while the holder owns a reference, leading to UAF on the holder's subsequent ksmbd_fd_put() and on any field reads performed by a concurrent share-mode walker that iterates m_fp_list without taking ksmbd_fp_get() (smb_check_perm_dleases-like paths). Fix both: * Stop reusing fp->node as a scavenger-private list node. Remove one expired handle from global_ft under global_ft.lock, take an explicit transient reference, drop the lock, unlink fp->node from m_fp_list under f_ci->m_lock, then drop both the durable lifetime and transient references with atomic_sub_and_test(2, &fp->refcount). If the scavenger is the last putter the close runs there; otherwise an in-flight holder that already raced through the m_fp_list lookup owns the final close via its ksmbd_fd_put() path. The one-at-a-time disposal can rescan the durable idr when multiple handles expire in the same pass, but durable scavenging is a background expiration path and the final full scan recomputes min_timeout before the next wait. * Clear fp->persistent_id inside __ksmbd_remove_durable_fd() right after idr_remove(), so a delayed final close from a holder that snatched fp does not re-issue idr_remove() on a persistent id that idr_alloc_cyclic() in ksmbd_open_durable_fd() may have already handed out to a brand-new durable handle. * Bypass the per-conn open_files_count decrement in __put_fd_final() when fp is detached from any session table (fp->conn cleared by session_fd_check() at durable preserve -- paired with the volatile_id clear at unpublish, so checking fp->conn alone is sufficient). The walker that owns the final close runs from an unrelated work->conn whose stats.open_files_count never tracked this durable fp; without this guard the holder would underflow that unrelated counter. The two races are folded into one patch because patch (1) alone cleans up the corrupted list but leaves a deterministic UAF window for m_fp_list walkers that the transient-reference and persistent_id discipline in (2) close; bisecting onto an intermediate state would land on a UAF that pre-patch chaos merely made less reproducible. Validation: * CONFIG_DEBUG_LIST coverage for the list_head reuse path. * KASAN-enabled direct SMB2 durable-handle coverage that exercised ksmbd_durable_scavenger() and non-NULL ksmbd_lookup_fd_inode() returns while durable handles expired under concurrent rename lookups, with no KASAN, UAF, list-corruption, ODEBUG, or WARNING reports. * checkpatch --strict * make -j$(nproc) M=fs/smb/server Fixes: d484d621d40f ("ksmbd: add durable scavenger timer") Signed-off-by: DaeMyung Kang Acked-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/vfs_cache.c | 102 ++++++++++++++++++++++++++++---------- 1 file changed, 76 insertions(+), 26 deletions(-) diff --git a/fs/smb/server/vfs_cache.c b/fs/smb/server/vfs_cache.c index dc4037ef1834..354c4d8a1cfb 100644 --- a/fs/smb/server/vfs_cache.c +++ b/fs/smb/server/vfs_cache.c @@ -418,6 +418,14 @@ static void __ksmbd_remove_durable_fd(struct ksmbd_file *fp) return; idr_remove(global_ft.idr, fp->persistent_id); + /* + * Clear persistent_id so a later __ksmbd_close_fd() that runs from a + * delayed putter (e.g. when a concurrent ksmbd_lookup_fd_inode() + * walker held the final reference) does not re-issue idr_remove() on + * an id that idr_alloc_cyclic() may have already handed out to a new + * durable handle. + */ + fp->persistent_id = KSMBD_NO_FID; } static void ksmbd_remove_durable_fd(struct ksmbd_file *fp) @@ -521,6 +529,20 @@ static struct ksmbd_file *__ksmbd_lookup_fd(struct ksmbd_file_table *ft, static void __put_fd_final(struct ksmbd_work *work, struct ksmbd_file *fp) { + /* + * Detached durable fp -- session_fd_check() cleared fp->conn at + * preserve, so this fp is no longer tracked by any conn's + * stats.open_files_count. This happens when + * ksmbd_scavenger_dispose_dh() hands the final close off to an + * m_fp_list walker (e.g. ksmbd_lookup_fd_inode()) whose work->conn + * is unrelated to the conn that originally opened the handle; close + * via the NULL-ft path so we do not underflow that unrelated + * counter. + */ + if (!fp->conn) { + __ksmbd_close_fd(NULL, fp); + return; + } __ksmbd_close_fd(&work->sess->file_table, fp); atomic_dec(&work->conn->stats.open_files_count); } @@ -1033,24 +1055,37 @@ static bool ksmbd_durable_scavenger_alive(void) return true; } -static void ksmbd_scavenger_dispose_dh(struct list_head *head) +static void ksmbd_scavenger_dispose_dh(struct ksmbd_file *fp) { - while (!list_empty(head)) { - struct ksmbd_file *fp; + /* + * Durable-preserved fp can remain linked on f_ci->m_fp_list for + * share-mode checks. Unlink it before final close; fp->node is not + * available as a scavenger-private list node because re-adding it to + * another list corrupts m_fp_list. + */ + down_write(&fp->f_ci->m_lock); + list_del_init(&fp->node); + up_write(&fp->f_ci->m_lock); - fp = list_first_entry(head, struct ksmbd_file, node); - list_del_init(&fp->node); + /* + * Drop both the durable lifetime reference and the transient reference + * taken by the scavenger under global_ft.lock. If a concurrent + * ksmbd_lookup_fd_inode() (or any other m_fp_list walker) snatched fp + * before the unlink above, that holder owns the final close via + * ksmbd_fd_put() -> __ksmbd_close_fd(). Otherwise the scavenger is + * the last putter and finalises fp here. + */ + if (atomic_sub_and_test(2, &fp->refcount)) __ksmbd_close_fd(NULL, fp); - } } static int ksmbd_durable_scavenger(void *dummy) { struct ksmbd_file *fp = NULL; + struct ksmbd_file *expired_fp; unsigned int id; unsigned int min_timeout = 1; bool found_fp_timeout; - LIST_HEAD(scavenger_list); unsigned long remaining_jiffies; __module_get(THIS_MODULE); @@ -1060,8 +1095,6 @@ static int ksmbd_durable_scavenger(void *dummy) if (try_to_freeze()) continue; - found_fp_timeout = false; - remaining_jiffies = wait_event_timeout(dh_wq, ksmbd_durable_scavenger_alive() == false, __msecs_to_jiffies(min_timeout)); @@ -1070,23 +1103,39 @@ static int ksmbd_durable_scavenger(void *dummy) else min_timeout = DURABLE_HANDLE_MAX_TIMEOUT; - write_lock(&global_ft.lock); - idr_for_each_entry(global_ft.idr, fp, id) { - if (!fp->durable_timeout) - continue; + do { + expired_fp = NULL; + found_fp_timeout = false; - if (atomic_read(&fp->refcount) > 1 || - fp->conn) - continue; - - found_fp_timeout = true; - if (fp->durable_scavenger_timeout <= - jiffies_to_msecs(jiffies)) { - __ksmbd_remove_durable_fd(fp); - list_add(&fp->node, &scavenger_list); - } else { + write_lock(&global_ft.lock); + idr_for_each_entry(global_ft.idr, fp, id) { unsigned long durable_timeout; + if (!fp->durable_timeout) + continue; + + if (atomic_read(&fp->refcount) > 1 || + fp->conn) + continue; + + found_fp_timeout = true; + if (fp->durable_scavenger_timeout <= + jiffies_to_msecs(jiffies)) { + __ksmbd_remove_durable_fd(fp); + /* + * Take a transient reference so fp + * cannot be freed by an in-flight + * ksmbd_lookup_fd_inode() that found + * it through f_ci->m_fp_list while we + * drop global_ft.lock and reach the + * m_fp_list unlink in + * ksmbd_scavenger_dispose_dh(). + */ + atomic_inc(&fp->refcount); + expired_fp = fp; + break; + } + durable_timeout = fp->durable_scavenger_timeout - jiffies_to_msecs(jiffies); @@ -1094,10 +1143,11 @@ static int ksmbd_durable_scavenger(void *dummy) if (min_timeout > durable_timeout) min_timeout = durable_timeout; } - } - write_unlock(&global_ft.lock); + write_unlock(&global_ft.lock); - ksmbd_scavenger_dispose_dh(&scavenger_list); + if (expired_fp) + ksmbd_scavenger_dispose_dh(expired_fp); + } while (expired_fp); if (found_fp_timeout == false) break; From a74668eb2c0b866d7ac4823be6006ab2e227bc03 Mon Sep 17 00:00:00 2001 From: Shuhao Fu Date: Wed, 29 Apr 2026 16:59:56 +0800 Subject: [PATCH 172/525] ksmbd: fail share config requests when path allocation fails Non-pipe shares must have a duplicated backing path before they can be published. share_config_request() currently calls kstrndup() for that path, but if the allocation fails it leaves ret unchanged. If veto list parsing succeeds and share->name exists, the partially built share is still inserted into the global share table with share->path left NULL. A later share-root SMB2 create uses tree_conn->share_conf->path as the lookup root. If the share was published with path == NULL, that request passes a NULL pathname into do_getname_kernel()/strlen() and can crash the ksmbd worker. Set ret = -ENOMEM when path duplication fails so the incomplete share is destroyed before publication. Fixes: e2f34481b24d ("cifsd: add server-side procedures for SMB3") Signed-off-by: Shuhao Fu Acked-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/mgmt/share_config.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/fs/smb/server/mgmt/share_config.c b/fs/smb/server/mgmt/share_config.c index 53f44ff4d376..6f97f8d39657 100644 --- a/fs/smb/server/mgmt/share_config.c +++ b/fs/smb/server/mgmt/share_config.c @@ -167,7 +167,10 @@ static struct ksmbd_share_config *share_config_request(struct ksmbd_work *work, share->path = kstrndup(ksmbd_share_config_path(resp), path_len, KSMBD_DEFAULT_GFP); - if (share->path) { + if (!share->path) { + ret = -ENOMEM; + } else { + ret = 0; share->path_sz = strlen(share->path); while (share->path_sz > 1 && share->path[share->path_sz - 1] == '/') @@ -179,9 +182,10 @@ static struct ksmbd_share_config *share_config_request(struct ksmbd_work *work, share->force_directory_mode = resp->force_directory_mode; share->force_uid = resp->force_uid; share->force_gid = resp->force_gid; - ret = parse_veto_list(share, - KSMBD_SHARE_CONFIG_VETO_LIST(resp), - resp->veto_list_sz); + if (!ret) + ret = parse_veto_list(share, + KSMBD_SHARE_CONFIG_VETO_LIST(resp), + resp->veto_list_sz); if (!ret && share->path) { if (__ksmbd_override_fsids(work, share)) { kill_share(share); From 6fd7dd4e44d7840cb1ba0c3a895e9f576af3fe5c Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Fri, 1 May 2026 08:34:55 +0900 Subject: [PATCH 173/525] ksmbd: fix kernel-doc warnings from ksmbd_conn_get/put() The kernel test robot reported W=1 build warnings for ksmbd_conn_get() and ksmbd_conn_put() due to missing parameter descriptions. Add the @conn description to fix these warnings. Reported-by: kernel test robot Signed-off-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/connection.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/smb/server/connection.c b/fs/smb/server/connection.c index 95654a808162..8347495dbc62 100644 --- a/fs/smb/server/connection.c +++ b/fs/smb/server/connection.c @@ -122,6 +122,8 @@ static void __ksmbd_conn_release_work(struct work_struct *work) /** * ksmbd_conn_get() - take a reference on @conn and return it. * + * @conn: connection instance to get a reference to + * * Returns @conn unchanged so callers can write * "fp->conn = ksmbd_conn_get(work->conn);" in one expression. Returns NULL * if @conn is NULL. @@ -139,6 +141,8 @@ struct ksmbd_conn *ksmbd_conn_get(struct ksmbd_conn *conn) * ksmbd_conn_put() - drop a reference and, if it was the last, queue the * release onto ksmbd_conn_wq so it runs from process context. * + * @conn: connection instance to put a reference to + * * Callable from any context including RCU softirq callbacks and non-sleeping * locks; the actual release is deferred to the workqueue. ksmbd_conn_wq is * created in ksmbd_server_init() before any conn can be allocated and is From 996454bc0da84d5a1dedb1a7861823087e01a7ae Mon Sep 17 00:00:00 2001 From: Shota Zaizen Date: Tue, 28 Apr 2026 19:02:55 +0900 Subject: [PATCH 174/525] ksmbd: validate inherited ACE SID length smb_inherit_dacl() walks the parent directory DACL loaded from the security descriptor xattr. It verifies that each ACE contains the fixed SID header before using it, but does not verify that the variable-length SID described by sid.num_subauth is fully contained in the ACE. A malformed inheritable ACE can advertise more subauthorities than are present in the ACE. compare_sids() may then read past the ACE. smb_set_ace() also clamps the copied destination SID, but used the unchecked source SID count to compute the inherited ACE size. That could advance the temporary inherited ACE buffer pointer and nt_size accounting past the allocated buffer. Fix this by validating the parent ACE SID count and SID length before using the SID during inheritance. Compute the inherited ACE size from the copied SID so the size matches the bounded destination SID. Reject the inherited DACL if size accumulation would overflow smb_acl.size or the security descriptor allocation size. Fixes: e2f34481b24d ("cifsd: add server-side procedures for SMB3") Signed-off-by: Shota Zaizen Acked-by: Namjae Jeon Signed-off-by: Steve French --- fs/smb/server/smbacl.c | 66 +++++++++++++++++++++++++++++++++--------- 1 file changed, 52 insertions(+), 14 deletions(-) diff --git a/fs/smb/server/smbacl.c b/fs/smb/server/smbacl.c index 4bbc2c27e680..c1d1f34581d6 100644 --- a/fs/smb/server/smbacl.c +++ b/fs/smb/server/smbacl.c @@ -1068,7 +1068,26 @@ static void smb_set_ace(struct smb_ace *ace, const struct smb_sid *sid, u8 type, ace->flags = flags; ace->access_req = access_req; smb_copy_sid(&ace->sid, sid); - ace->size = cpu_to_le16(1 + 1 + 2 + 4 + 1 + 1 + 6 + (sid->num_subauth * 4)); + ace->size = cpu_to_le16(1 + 1 + 2 + 4 + 1 + 1 + 6 + + (ace->sid.num_subauth * 4)); +} + +static int smb_append_inherited_ace(struct smb_ace **ace, int *nt_size, + u16 *ace_cnt, const struct smb_sid *sid, + u8 type, u8 flags, __le32 access_req) +{ + int ace_size; + + smb_set_ace(*ace, sid, type, flags, access_req); + ace_size = le16_to_cpu((*ace)->size); + /* pdacl->size is __le16 and includes struct smb_acl. */ + if (check_add_overflow(*nt_size, ace_size, nt_size) || + *nt_size > U16_MAX - (int)sizeof(struct smb_acl)) + return -EINVAL; + + (*ace_cnt)++; + *ace = (struct smb_ace *)((char *)*ace + ace_size); + return 0; } int smb_inherit_dacl(struct ksmbd_conn *conn, @@ -1157,6 +1176,12 @@ int smb_inherit_dacl(struct ksmbd_conn *conn, CIFS_SID_BASE_SIZE) break; + if (parent_aces->sid.num_subauth > SID_MAX_SUB_AUTHORITIES || + pace_size < offsetof(struct smb_ace, sid) + + CIFS_SID_BASE_SIZE + + sizeof(__le32) * parent_aces->sid.num_subauth) + break; + aces_size -= pace_size; flags = parent_aces->flags; @@ -1186,22 +1211,24 @@ int smb_inherit_dacl(struct ksmbd_conn *conn, } if (is_dir && creator && flags & CONTAINER_INHERIT_ACE) { - smb_set_ace(aces, psid, parent_aces->type, inherited_flags, - parent_aces->access_req); - nt_size += le16_to_cpu(aces->size); - ace_cnt++; - aces = (struct smb_ace *)((char *)aces + le16_to_cpu(aces->size)); + rc = smb_append_inherited_ace(&aces, &nt_size, &ace_cnt, + psid, parent_aces->type, + inherited_flags, + parent_aces->access_req); + if (rc) + goto free_aces_base; flags |= INHERIT_ONLY_ACE; psid = creator; } else if (is_dir && !(parent_aces->flags & NO_PROPAGATE_INHERIT_ACE)) { psid = &parent_aces->sid; } - smb_set_ace(aces, psid, parent_aces->type, flags | inherited_flags, - parent_aces->access_req); - nt_size += le16_to_cpu(aces->size); - aces = (struct smb_ace *)((char *)aces + le16_to_cpu(aces->size)); - ace_cnt++; + rc = smb_append_inherited_ace(&aces, &nt_size, &ace_cnt, psid, + parent_aces->type, + flags | inherited_flags, + parent_aces->access_req); + if (rc) + goto free_aces_base; pass: parent_aces = (struct smb_ace *)((char *)parent_aces + pace_size); } @@ -1211,7 +1238,7 @@ int smb_inherit_dacl(struct ksmbd_conn *conn, struct smb_acl *pdacl; struct smb_sid *powner_sid = NULL, *pgroup_sid = NULL; int powner_sid_size = 0, pgroup_sid_size = 0, pntsd_size; - int pntsd_alloc_size; + size_t pntsd_alloc_size; if (parent_pntsd->osidoffset) { powner_sid = (struct smb_sid *)((char *)parent_pntsd + @@ -1224,8 +1251,19 @@ int smb_inherit_dacl(struct ksmbd_conn *conn, pgroup_sid_size = 1 + 1 + 6 + (pgroup_sid->num_subauth * 4); } - pntsd_alloc_size = sizeof(struct smb_ntsd) + powner_sid_size + - pgroup_sid_size + sizeof(struct smb_acl) + nt_size; + if (check_add_overflow(sizeof(struct smb_ntsd), + (size_t)powner_sid_size, + &pntsd_alloc_size) || + check_add_overflow(pntsd_alloc_size, + (size_t)pgroup_sid_size, + &pntsd_alloc_size) || + check_add_overflow(pntsd_alloc_size, sizeof(struct smb_acl), + &pntsd_alloc_size) || + check_add_overflow(pntsd_alloc_size, (size_t)nt_size, + &pntsd_alloc_size)) { + rc = -EINVAL; + goto free_aces_base; + } pntsd = kzalloc(pntsd_alloc_size, KSMBD_DEFAULT_GFP); if (!pntsd) { From 0c7ae130698e70107430254e79fbe996b4d37ab5 Mon Sep 17 00:00:00 2001 From: Paul Chaignon Date: Sat, 2 May 2026 12:12:40 +0200 Subject: [PATCH 175/525] tools/headers: Regenerate stddef.h to fix BPF selftests MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit With commit dacbfc167808 ("crypto: af_alg - Annotate struct af_alg_iv with __counted_by"), two selftests, test_tag and crypto_sanity, now indirectly rely on the __counted_by macro. On systems with commit dacbfc167808 in the installed UAPI headers, the selftests build fails with: In file included from tools/testing/selftests/bpf/prog_tests/crypto_sanity.c:7: /usr/include/linux/if_alg.h:45:22: error: expected ‘:’, ‘,’, ‘;’, ‘}’ or ‘__attribute__’ before ‘__counted_by’ 45 | __u8 iv[] __counted_by(ivlen); | ^~~~~~~~~~~~ This patch fixes it by regenerating stddef.h in tools/include using the instructions from commit a778f5d46b62 ("tools/headers: Pull in stddef.h to uapi to fix BPF selftests build in CI"). Fixes: dacbfc167808 ("crypto: af_alg - Annotate struct af_alg_iv with __counted_by") Signed-off-by: Paul Chaignon Reviewed-by: Alan Maguire Tested-by: Ihor Solodrai Link: https://lore.kernel.org/r/8da8ef16055aa452d940668ed5359ce54adc6b0b.1777715500.git.paul.chaignon@gmail.com Signed-off-by: Alexei Starovoitov --- tools/include/uapi/linux/stddef.h | 26 +++++++++++++++++++++++++- 1 file changed, 25 insertions(+), 1 deletion(-) diff --git a/tools/include/uapi/linux/stddef.h b/tools/include/uapi/linux/stddef.h index c53cde425406..457498259494 100644 --- a/tools/include/uapi/linux/stddef.h +++ b/tools/include/uapi/linux/stddef.h @@ -3,7 +3,6 @@ #define _LINUX_STDDEF_H - #ifndef __always_inline #define __always_inline __inline__ #endif @@ -36,6 +35,11 @@ struct __struct_group_tag(TAG) { MEMBERS } ATTRS NAME; \ } ATTRS +#ifdef __cplusplus +/* sizeof(struct{}) is 1 in C++, not 0, can't use C version of the macro. */ +#define __DECLARE_FLEX_ARRAY(T, member) \ + T member[0] +#else /** * __DECLARE_FLEX_ARRAY() - Declare a flexible array usable in a union * @@ -52,3 +56,23 @@ TYPE NAME[]; \ } #endif + +#ifndef __counted_by +#define __counted_by(m) +#endif + +#ifndef __counted_by_le +#define __counted_by_le(m) +#endif + +#ifndef __counted_by_be +#define __counted_by_be(m) +#endif + +#ifndef __counted_by_ptr +#define __counted_by_ptr(m) +#endif + +#define __kernel_nonstring + +#endif /* _LINUX_STDDEF_H */ From ddca6da148b8ced3e6d3d7fb3b2e5b4ed6359dc2 Mon Sep 17 00:00:00 2001 From: "Maciej W. Rozycki" Date: Mon, 27 Apr 2026 11:23:35 +0100 Subject: [PATCH 176/525] MAINTAINERS: Add self for the DEC LANCE network driver Like with the rest of DECstation and TURBOchannel hardware I have been handling the DEC LANCE network driver for some 25 years now anyway. Signed-off-by: Maciej W. Rozycki Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/alpine.DEB.2.21.2604271113520.28583@angie.orcam.me.uk Signed-off-by: Jakub Kicinski --- MAINTAINERS | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 27a073f53cea..ec8661b446fb 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -7077,6 +7077,12 @@ T: git git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git core/debugobjec F: include/linux/debugobjects.h F: lib/debugobjects.c +DEC LANCE NETWORK DRIVER +M: "Maciej W. Rozycki" +L: netdev@vger.kernel.org +S: Maintained +F: drivers/net/ethernet/amd/declance.c + DECSTATION PLATFORM SUPPORT M: "Maciej W. Rozycki" L: linux-mips@vger.kernel.org From 1a57efe250a13906396c2a4792f0090f142f9844 Mon Sep 17 00:00:00 2001 From: Holger Brunck Date: Wed, 29 Apr 2026 13:42:07 +0200 Subject: [PATCH 177/525] net: wan: fsl_ucc_hdlc: fix uhdlc_memclean Unmapping of uf_regs is done from ucc_fast_free and doesn't need to be done explicitly. If already unmapped ucc_fast_free will crash. Fixes: c19b6d246a35 ("drivers/net: support hdlc function for QE-UCC") Signed-off-by: Holger Brunck Reviewed-by: Simon Horman Signed-off-by: Jakub Kicinski --- drivers/net/wan/fsl_ucc_hdlc.c | 5 ----- 1 file changed, 5 deletions(-) diff --git a/drivers/net/wan/fsl_ucc_hdlc.c b/drivers/net/wan/fsl_ucc_hdlc.c index 3bd57527b1be..8155e92af14e 100644 --- a/drivers/net/wan/fsl_ucc_hdlc.c +++ b/drivers/net/wan/fsl_ucc_hdlc.c @@ -773,11 +773,6 @@ static void uhdlc_memclean(struct ucc_hdlc_private *priv) kfree(priv->tx_skbuff); priv->tx_skbuff = NULL; - if (priv->uf_regs) { - iounmap(priv->uf_regs); - priv->uf_regs = NULL; - } - if (priv->uccf) { ucc_fast_free(priv->uccf); priv->uccf = NULL; From 851bba8068d15f5a386da544096f7ed6bc16e551 Mon Sep 17 00:00:00 2001 From: Holger Brunck Date: Wed, 29 Apr 2026 13:42:08 +0200 Subject: [PATCH 178/525] net: wan: fsl_ucc_hdlc: fix ucc_hdlc_remove If the driver is used in a non tdm mode priv->utdm is a NULL pointer. Therefore we need to check this pointer first before checking si_regs. Fixes: c19b6d246a35 ("drivers/net: support hdlc function for QE-UCC") Signed-off-by: Holger Brunck Reviewed-by: Simon Horman Signed-off-by: Jakub Kicinski --- drivers/net/wan/fsl_ucc_hdlc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/wan/fsl_ucc_hdlc.c b/drivers/net/wan/fsl_ucc_hdlc.c index 8155e92af14e..15bfb78381d4 100644 --- a/drivers/net/wan/fsl_ucc_hdlc.c +++ b/drivers/net/wan/fsl_ucc_hdlc.c @@ -1250,12 +1250,12 @@ static void ucc_hdlc_remove(struct platform_device *pdev) uhdlc_memclean(priv); - if (priv->utdm->si_regs) { + if (priv->utdm && priv->utdm->si_regs) { iounmap(priv->utdm->si_regs); priv->utdm->si_regs = NULL; } - if (priv->utdm->siram) { + if (priv->utdm && priv->utdm->siram) { iounmap(priv->utdm->siram); priv->utdm->siram = NULL; } From 383d0fb8946921b4914ea0f360342e221d419d40 Mon Sep 17 00:00:00 2001 From: Gregory Fuchedgi Date: Wed, 29 Apr 2026 14:54:14 -0700 Subject: [PATCH 179/525] amd-xgbe: fix PTP addend overflow causing frozen clock XGBE_PTP_ACT_CLK_FREQ and XGBE_V2_PTP_ACT_CLK_FREQ were 10x too large (500MHz/1GHz instead of 50MHz/100MHz), causing the computed addend to overflow the 32-bit tstamp_addend. In the general case this would result in the clock advancing at the wrong rate. For v2 (PCI), ptpclk_rate is hardcoded to 125MHz, so the addend formula (ACT_CLK_FREQ << 32) / ptpclk_rate yields exactly 8 * 2^32, and when stored to the 32-bit tstamp_addend the value is zero. With addend = 0 the hardware accumulator never overflows and the PTP clock is fully stopped. For v1 (platform), ptpclk_rate is read from ACPI/DT so the exact overflow behavior depends on the firmware-reported frequency. Define the constants as NSEC_PER_SEC / SSINC so the relationship is explicit and cannot drift out of sync. Fixes: fbd47be098b5 ("amd-xgbe: add hardware PTP timestamping support") Tested-by: Gregory Fuchedgi Signed-off-by: Gregory Fuchedgi Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260429-fix-xgbe-ptp-addend-v1-1-fca5b0ca5e62@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/amd/xgbe/xgbe.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/amd/xgbe/xgbe.h b/drivers/net/ethernet/amd/xgbe/xgbe.h index 60b7e53206d1..3d3b09010d48 100644 --- a/drivers/net/ethernet/amd/xgbe/xgbe.h +++ b/drivers/net/ethernet/amd/xgbe/xgbe.h @@ -135,11 +135,11 @@ */ #define XGBE_TSTAMP_SSINC 20 #define XGBE_TSTAMP_SNSINC 0 -#define XGBE_PTP_ACT_CLK_FREQ 500000000 +#define XGBE_PTP_ACT_CLK_FREQ (NSEC_PER_SEC / XGBE_TSTAMP_SSINC) #define XGBE_V2_TSTAMP_SSINC 0xA #define XGBE_V2_TSTAMP_SNSINC 0 -#define XGBE_V2_PTP_ACT_CLK_FREQ 1000000000 +#define XGBE_V2_PTP_ACT_CLK_FREQ (NSEC_PER_SEC / XGBE_V2_TSTAMP_SSINC) /* Define maximum supported values */ #define XGBE_MAX_PPS_OUT 4 From 458d5615272d3de535748342eb68ca492343048c Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Thu, 30 Apr 2026 11:29:55 -0400 Subject: [PATCH 180/525] net/sched: sch_red: Replace direct dequeue call with peek and qdisc_dequeue_peeked When red qdisc has children (eg qfq qdisc) whose peek() callback is qdisc_peek_dequeued(), we could get a kernel panic. When the parent of such qdiscs (eg illustrated in patch #3 as tbf) wants to retrieve an skb from its child (red in this case), it will do the following: 1a. do a peek() - and when sensing there's an skb the child can offer, then - the child in this case(red) calls its child's (qfq) peek. qfq does the right thing and will return the gso_skb queue packet. Note: if there wasnt a gso_skb entry then qfq will store it there. 1b. invoke a dequeue() on the child (red). And herein lies the problem. - red will call the child's dequeue() which will essentially just try to grab something of qfq's queue. [ 78.667668][ T363] KASAN: null-ptr-deref in range [0x0000000000000048-0x000000000000004f] [ 78.667927][ T363] CPU: 1 UID: 0 PID: 363 Comm: ping Not tainted 7.1.0-rc1-00033-g46f74a3f7d57-dirty #790 PREEMPT(full) [ 78.668263][ T363] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 78.668486][ T363] RIP: 0010:qfq_dequeue+0x446/0xc90 [sch_qfq] [ 78.668718][ T363] Code: 54 c0 e8 dd 90 00 f1 48 c7 c7 e0 03 54 c0 48 89 de e8 ce 90 00 f1 48 8d 7b 48 b8 ff ff 37 00 48 89 fa 48 c1 e0 2a 48 c1 ea 03 <80> 3c 02 00 74 05 e8 ef a1 e1 f1 48 8b 7b 48 48 8d 54 24 58 48 8d [ 78.669312][ T363] RSP: 0018:ffff88810de573e0 EFLAGS: 00010216 [ 78.669533][ T363] RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000000000 [ 78.669790][ T363] RDX: 0000000000000009 RSI: 0000000000000004 RDI: 0000000000000048 [ 78.670044][ T363] RBP: ffff888110dc4000 R08: ffffffffb1b0885a R09: fffffbfff6ba9078 [ 78.670297][ T363] R10: 0000000000000003 R11: ffff888110e31c80 R12: 0000001880000000 [ 78.670560][ T363] R13: ffff888110dc4150 R14: ffff888110dc42b8 R15: 0000000000000200 [ 78.670814][ T363] FS: 00007f66a8f09c40(0000) GS:ffff888163428000(0000) knlGS:0000000000000000 [ 78.671110][ T363] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 78.671324][ T363] CR2: 000055db4c6a30a8 CR3: 000000010da67000 CR4: 0000000000750ef0 [ 78.671585][ T363] PKRU: 55555554 [ 78.671713][ T363] Call Trace: [ 78.671843][ T363] [ 78.671936][ T363] ? __pfx_qfq_dequeue+0x10/0x10 [sch_qfq] [ 78.672148][ T363] ? __pfx__printk+0x10/0x10 [ 78.672322][ T363] ? srso_alias_return_thunk+0x5/0xfbef5 [ 78.672496][ T363] ? lockdep_hardirqs_on_prepare+0xa8/0x1a0 [ 78.672706][ T363] ? srso_alias_return_thunk+0x5/0xfbef5 [ 78.672875][ T363] ? trace_hardirqs_on+0x19/0x1a0 [ 78.673047][ T363] red_dequeue+0x65/0x270 [sch_red] [ 78.673217][ T363] ? srso_alias_return_thunk+0x5/0xfbef5 [ 78.673385][ T363] tbf_dequeue.cold+0xb0/0x70c [sch_tbf] [ 78.673566][ T363] __qdisc_run+0x169/0x1900 The right thing to do in #1b is to grab the skb off gso_skb queue. This patchset fixes that issue by changing #1b to use qdisc_dequeue_peeked() method instead. Fixes: 77be155cba4e ("pkt_sched: Add peek emulation for non-work-conserving qdiscs.") Reported-by: Manas Reported-by: Rakshit Awasthi Signed-off-by: Jamal Hadi Salim Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20260430152957.194015-2-jhs@mojatatu.com Signed-off-by: Jakub Kicinski --- net/sched/sch_red.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_red.c b/net/sched/sch_red.c index 432b8a3000a5..4d0e44a2e7c6 100644 --- a/net/sched/sch_red.c +++ b/net/sched/sch_red.c @@ -162,7 +162,7 @@ static struct sk_buff *red_dequeue(struct Qdisc *sch) struct red_sched_data *q = qdisc_priv(sch); struct Qdisc *child = q->qdisc; - skb = child->dequeue(child); + skb = qdisc_dequeue_peeked(child); if (skb) { qdisc_bstats_update(sch, skb); qdisc_qstats_backlog_dec(sch, skb); From 1b9bc71153b01dbde8045b9edede4240f4f5520e Mon Sep 17 00:00:00 2001 From: Victor Nogueria Date: Thu, 30 Apr 2026 11:29:56 -0400 Subject: [PATCH 181/525] net/sched: sch_sfb: Replace direct dequeue call with peek and qdisc_dequeue_peeked When sfb has children (eg qfq qdisc) whose peek() callback is qdisc_peek_dequeued(), we could get a kernel panic. When the parent of such qdiscs (eg illustrated in patch #3 as tbf) wants to retrieve an skb from its child (sfb in this case), it will do the following: 1a. do a peek() - and when sensing there's an skb the child can offer, then - the child in this case(sfb) calls its child's (qfq) peek. qfq does the right thing and will return the gso_skb queue packet. Note: if there wasnt a gso_skb entry then qfq will store it there. 1b. invoke a dequeue() on the child (sfb). And herein lies the problem. - sfb will call the child's dequeue() which will essentially just try to grab something of qfq's queue. [ 127.594489][ T453] KASAN: null-ptr-deref in range [0x0000000000000048-0x000000000000004f] [ 127.594741][ T453] CPU: 2 UID: 0 PID: 453 Comm: ping Not tainted 7.1.0-rc1-00035-gac961974495b-dirty #793 PREEMPT(full) [ 127.595059][ T453] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 127.595254][ T453] RIP: 0010:qfq_dequeue+0x35c/0x1650 [sch_qfq] [ 127.595461][ T453] Code: 00 fc ff df 80 3c 02 00 0f 85 17 0e 00 00 4c 8d 73 48 48 89 9d b8 02 00 00 48 b8 00 00 00 00 00 fc ff df 4c 89 f2 48 c1 ea 03 <80> 3c 02 00 0f 85 76 0c 00 00 48 b8 00 00 00 00 00 fc ff df 4c 8b [ 127.596081][ T453] RSP: 0018:ffff88810e5af440 EFLAGS: 00010216 [ 127.596337][ T453] RAX: dffffc0000000000 RBX: 0000000000000000 RCX: dffffc0000000000 [ 127.596623][ T453] RDX: 0000000000000009 RSI: 0000001880000000 RDI: ffff888104fd82b0 [ 127.596917][ T453] RBP: ffff888104fd8000 R08: ffff888104fd8280 R09: 1ffff110211893a3 [ 127.597165][ T453] R10: 1ffff110211893a6 R11: 1ffff110211893a7 R12: 0000001880000000 [ 127.597404][ T453] R13: ffff888104fd82b8 R14: 0000000000000048 R15: 0000000040000000 [ 127.597644][ T453] FS: 00007fc380cbfc40(0000) GS:ffff88816f2a8000(0000) knlGS:0000000000000000 [ 127.597956][ T453] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 127.598160][ T453] CR2: 00005610aa9890a8 CR3: 000000010369e000 CR4: 0000000000750ef0 [ 127.598390][ T453] PKRU: 55555554 [ 127.598509][ T453] Call Trace: [ 127.598629][ T453] [ 127.598718][ T453] ? mark_held_locks+0x40/0x70 [ 127.598890][ T453] ? srso_alias_return_thunk+0x5/0xfbef5 [ 127.599053][ T453] sfb_dequeue+0x88/0x4d0 [ 127.599174][ T453] ? ktime_get+0x137/0x230 [ 127.599328][ T453] ? srso_alias_return_thunk+0x5/0xfbef5 [ 127.599480][ T453] ? qdisc_peek_dequeued+0x7b/0x350 [sch_qfq] [ 127.599670][ T453] ? srso_alias_return_thunk+0x5/0xfbef5 [ 127.599831][ T453] tbf_dequeue+0x6b1/0x1098 [sch_tbf] [ 127.599988][ T453] __qdisc_run+0x169/0x1900 The right thing to do in #1b is to grab the skb off gso_skb queue. This patchset fixes that issue by changing #1b to use qdisc_dequeue_peeked() method instead. Fixes: e13e02a3c68d ("net_sched: SFB flow scheduler") Signed-off-by: Victor Nogueria Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20260430152957.194015-3-jhs@mojatatu.com Signed-off-by: Jakub Kicinski --- net/sched/sch_sfb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_sfb.c b/net/sched/sch_sfb.c index bd5ef561030f..d3ee8e5479b3 100644 --- a/net/sched/sch_sfb.c +++ b/net/sched/sch_sfb.c @@ -441,7 +441,7 @@ static struct sk_buff *sfb_dequeue(struct Qdisc *sch) struct Qdisc *child = q->qdisc; struct sk_buff *skb; - skb = child->dequeue(q->qdisc); + skb = qdisc_dequeue_peeked(child); if (skb) { qdisc_bstats_update(sch, skb); From 3a3a30c14d7f04206916824a79878cfefac6c8e2 Mon Sep 17 00:00:00 2001 From: Victor Nogueira Date: Thu, 30 Apr 2026 11:29:57 -0400 Subject: [PATCH 182/525] selftests/tc-testing: Add tests that force red and sfb to dequeue from child's gso_skb Create 4 test cases: - Force red to dequeue from its child's gso_skb with qfq leaf - Force sfb to dequeue from its child's gso_skb with qfq leaf - Force red to dequeue from its child's gso_skb with dualpi2 leaf - Force sfb to dequeue from its child's gso_skb with dualpi2 leaf All of them have tbf followed by red (or sfb) followed by qfq (or dualpi2). Since tbf calls its child's peek followed by qdisc_dequeue_peeked, it will force red/sfb to call their child's peek. In this case, since the child (qfq/dualpi2) has qdisc_peek_dequeued as its peek callback, the packet will be stored in its gso_skb queue. During the subsequent call to qdisc_dequeue_peeked, red/sfb will have to dequeue from the child's gso_skb to retrieve the packet. Not doing so will cause a NULL ptr deref which was happening before a recent fix. Acked-by: Jamal Hadi Salim Signed-off-by: Victor Nogueira Link: https://patch.msgid.link/20260430152957.194015-4-jhs@mojatatu.com Signed-off-by: Jakub Kicinski --- .../tc-testing/tc-tests/infra/qdiscs.json | 148 ++++++++++++++++++ 1 file changed, 148 insertions(+) diff --git a/tools/testing/selftests/tc-testing/tc-tests/infra/qdiscs.json b/tools/testing/selftests/tc-testing/tc-tests/infra/qdiscs.json index eefadd0546d3..b1f856cf62c1 100644 --- a/tools/testing/selftests/tc-testing/tc-tests/infra/qdiscs.json +++ b/tools/testing/selftests/tc-testing/tc-tests/infra/qdiscs.json @@ -1136,5 +1136,153 @@ "teardown": [ "$TC qdisc del dev $DUMMY handle 1: root" ] + }, + { + "id": "7a5f", + "name": "Force red to dequeue from its child's gso_skb with qfq leaf", + "category": [ + "qdisc", + "tbf", + "red", + "qfq" + ], + "plugins": { + "requires": "nsPlugin" + }, + "setup": [ + "$IP link set dev $DUMMY up || true", + "$IP addr add 10.10.11.10/24 dev $DUMMY || true", + "$TC qdisc add dev $DUMMY root handle 1: tbf rate 88bit burst 1661b peakrate 2257333 minburst 1024 limit 7b", + "$TC qdisc add dev $DUMMY parent 1: handle 2: red limit 757 min 16 max 24 avpkt 16", + "$TC qdisc add dev $DUMMY parent 2: handle 3: qfq", + "$TC class add dev $DUMMY classid 3:1 parent 3: qfq maxpkt 512 weight 1", + "$TC filter add dev $DUMMY parent 3: protocol ip prio 1 matchall classid 3:1 action ok" + ], + "cmdUnderTest": "ping -c 1 10.10.10.1 -W0.01 -I$DUMMY || true", + "expExitCode": "0", + "verifyCmd": "$TC -s -j qdisc ls dev $DUMMY parent 1:", + "matchJSON": [ + { + "kind": "red", + "handle": "2:", + "bytes": 98, + "packets": 1, + "backlog": 0, + "qlen": 0 + } + ], + "teardown": [ + "$TC qdisc del dev $DUMMY handle 1: root" + ] + }, + { + "id": "cdae", + "name": "Force sfb to dequeue from its child's gso_skb with qfq leaf", + "category": [ + "qdisc", + "tbf", + "sfb", + "qfq" + ], + "plugins": { + "requires": "nsPlugin" + }, + "setup": [ + "$IP link set dev $DUMMY up || true", + "$IP addr add 10.10.11.10/24 dev $DUMMY || true", + "$TC qdisc add dev $DUMMY root handle 1: tbf rate 88bit burst 1661b peakrate 2257333 minburst 1024 limit 7b", + "$TC qdisc add dev $DUMMY parent 1: handle 2: sfb", + "$TC qdisc add dev $DUMMY parent 2: handle 3: qfq", + "$TC class add dev $DUMMY classid 3:1 parent 3: qfq maxpkt 512 weight 1", + "$TC filter add dev $DUMMY parent 3: protocol ip prio 1 matchall classid 3:1 action ok" + ], + "cmdUnderTest": "ping -c 1 10.10.10.1 -W0.01 -I$DUMMY || true", + "expExitCode": "0", + "verifyCmd": "$TC -s -j qdisc ls dev $DUMMY parent 1:", + "matchJSON": [ + { + "kind": "sfb", + "handle": "2:", + "bytes": 98, + "packets": 1, + "backlog": 0, + "qlen": 0 + } + ], + "teardown": [ + "$TC qdisc del dev $DUMMY handle 1: root" + ] + }, + { + "id": "291d", + "name": "Force red to dequeue from its child's gso_skb with dualpi2 leaf", + "category": [ + "qdisc", + "tbf", + "red", + "dualpi2" + ], + "plugins": { + "requires": "nsPlugin" + }, + "setup": [ + "$IP link set dev $DUMMY up || true", + "$IP addr add 10.10.11.10/24 dev $DUMMY || true", + "$TC qdisc add dev $DUMMY root handle 1: tbf rate 88bit burst 1661b peakrate 2257333 minburst 1024 limit 7b", + "$TC qdisc add dev $DUMMY parent 1: handle 2: red limit 757 min 16 max 24 avpkt 16", + "$TC qdisc add dev $DUMMY parent 2: handle 3: dualpi2" + ], + "cmdUnderTest": "ping -c 1 10.10.10.1 -W0.01 -I$DUMMY || true", + "expExitCode": "0", + "verifyCmd": "$TC -s -j qdisc ls dev $DUMMY parent 1:", + "matchJSON": [ + { + "kind": "red", + "handle": "2:", + "bytes": 98, + "packets": 1, + "backlog": 0, + "qlen": 0 + } + ], + "teardown": [ + "$TC qdisc del dev $DUMMY handle 1: root" + ] + }, + { + "id": "9c6d", + "name": "Force sfb to dequeue from its child's gso_skb with dualpi2 leaf", + "category": [ + "qdisc", + "tbf", + "sfb", + "dualpi2" + ], + "plugins": { + "requires": "nsPlugin" + }, + "setup": [ + "$IP link set dev $DUMMY up || true", + "$IP addr add 10.10.11.10/24 dev $DUMMY || true", + "$TC qdisc add dev $DUMMY root handle 1: tbf rate 88bit burst 1661b peakrate 2257333 minburst 1024 limit 7b", + "$TC qdisc add dev $DUMMY parent 1: handle 2: sfb", + "$TC qdisc add dev $DUMMY parent 2: handle 3: dualpi2" + ], + "cmdUnderTest": "ping -c 1 10.10.10.1 -W0.01 -I$DUMMY || true", + "expExitCode": "0", + "verifyCmd": "$TC -s -j qdisc ls dev $DUMMY parent 1:", + "matchJSON": [ + { + "kind": "sfb", + "handle": "2:", + "bytes": 98, + "packets": 1, + "backlog": 0, + "qlen": 0 + } + ], + "teardown": [ + "$TC qdisc del dev $DUMMY handle 1: root" + ] } ] From 1d324c2f43f70c965f25c58cc3611c779adbe47e Mon Sep 17 00:00:00 2001 From: Maoyi Xie Date: Thu, 30 Apr 2026 18:33:18 +0800 Subject: [PATCH 183/525] ip6_gre: Use cached t->net in ip6erspan_changelink(). After commit 5e72ce3e3980 ("net: ipv6: Use link netns in newlink() of rtnl_link_ops"), ip6erspan_newlink() correctly resolves the per-netns ip6gre hash via link_net. ip6erspan_changelink() was not converted in that series and still uses dev_net(dev), which diverges from the device's creation netns after IFLA_NET_NS_FD migration. This re-inserts the tunnel into the wrong per-netns hash. The original netns keeps a stale entry. When that netns is later destroyed, ip6gre_exit_rtnl_net() walks the stale entry, producing a slab-use-after-free reported by KASAN, followed by a kernel BUG at net/core/dev.c (LIST_POISON1) in unregister_netdevice_many_notify(). Reachable from an unprivileged user namespace (unshare --user --map-root-user --net). ip6gre_changelink() earlier in the same file already uses the cached t->net; only ip6erspan_changelink() has the wrong shape. Fixes: 2d665034f239 ("net: ip6_gre: Fix ip6erspan hlen calculation") Cc: stable@vger.kernel.org # v5.15+ Signed-off-by: Maoyi Xie Reviewed-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20260430103318.3206018-1-maoyi.xie@ntu.edu.sg Signed-off-by: Jakub Kicinski --- net/ipv6/ip6_gre.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c index 63fc8556b475..365b4059eb20 100644 --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@ -2262,10 +2262,11 @@ static int ip6erspan_changelink(struct net_device *dev, struct nlattr *tb[], struct nlattr *data[], struct netlink_ext_ack *extack) { - struct ip6gre_net *ign = net_generic(dev_net(dev), ip6gre_net_id); + struct ip6_tnl *t = netdev_priv(dev); struct __ip6_tnl_parm p; - struct ip6_tnl *t; + struct ip6gre_net *ign; + ign = net_generic(t->net, ip6gre_net_id); t = ip6gre_changelink_common(dev, tb, data, &p, extack); if (IS_ERR(t)) return PTR_ERR(t); From 70f780edcd1e86350202d8a409de026b2d2e2067 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:34 -0300 Subject: [PATCH 184/525] RDMA/ionic: Fix typo in format string Applying the corrupted patch by hand mangled the format string, put the s in the right place. Cc: stable@vger.kernel.org Fixes: 654a27f25530 ("RDMA/ionic: bound node_desc sysfs read with %.64s") Link: https://patch.msgid.link/r/1-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Reported-by: Brad Spengler Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/ionic/ionic_ibdev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/ionic/ionic_ibdev.c b/drivers/infiniband/hw/ionic/ionic_ibdev.c index 0382a64839d2..73a616ae3502 100644 --- a/drivers/infiniband/hw/ionic/ionic_ibdev.c +++ b/drivers/infiniband/hw/ionic/ionic_ibdev.c @@ -185,7 +185,7 @@ static ssize_t hca_type_show(struct device *device, struct ionic_ibdev *dev = rdma_device_to_drv_device(device, struct ionic_ibdev, ibdev); - return sysfs_emit(buf, "%s.64\n", dev->ibdev.node_desc); + return sysfs_emit(buf, "%.64s\n", dev->ibdev.node_desc); } static DEVICE_ATTR_RO(hca_type); From 641858d52f2372124d9312a407e2124915d846ee Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:35 -0300 Subject: [PATCH 185/525] RDMA/mlx5: Restore zero-init to mlx5_ib_modify_qp() ucmd Sashiko points out the check for inlen==0 got missed, the ={} was not redundant, put it back. Fixes: a9cd442a5347 ("RDMA: Remove redundant = {} for udata req structs") Link: https://patch.msgid.link/r/2-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mlx5/qp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index 1611a704c1b3..8fd05532c09c 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -4697,7 +4697,7 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, struct mlx5_ib_dev *dev = to_mdev(ibqp->device); struct mlx5_ib_modify_qp_resp resp = {}; struct mlx5_ib_qp *qp = to_mqp(ibqp); - struct mlx5_ib_modify_qp ucmd; + struct mlx5_ib_modify_qp ucmd = {}; enum ib_qp_type qp_type; enum ib_qp_state cur_state, new_state; int err = -EINVAL; From 45e8ebc9ede73543c55d597bb53b6bbb7e8b7327 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:36 -0300 Subject: [PATCH 186/525] RDMA/mlx5: Add missing store/release for lock elision pattern mlx5 has a common pattern implementing a device-global singleton resource where it checks the resource pointer for !NULL and then skips obtaining the lock. This is not ordered properly as observing !NULL doesn't mean that all the data under that pointer is also visible on this CPU when the lock is not taken. Use a release/acquire pairing to explicitly manage this. Pointed out by sashiko, Codex found more cases. Fixes: 5895e70f2e6e ("IB/mlx5: Allocate resources just before first QP/SRQ is created") Fixes: 638420115cc4 ("IB/mlx5: Create UMR QP just before first reg_mr occurs") Link: https://sashiko.dev/#/patchset/SYBPR01MB7881E1E0970268BD69C0BA75AF2B2%40SYBPR01MB7881.ausprd01.prod.outlook.com Link: https://patch.msgid.link/r/3-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Assisted-by: Codex:GPT-5.5 Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mlx5/main.c | 8 ++++---- drivers/infiniband/hw/mlx5/umr.c | 4 ++-- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index 8115ae869ef2..61078281953d 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -3310,7 +3310,7 @@ int mlx5_ib_dev_res_cq_init(struct mlx5_ib_dev *dev) * devr->c0 is set once, never changed until device unload. * Avoid taking the mutex if initialization is already done. */ - if (devr->c0) + if (smp_load_acquire(&devr->c0)) return 0; mutex_lock(&devr->cq_lock); @@ -3336,7 +3336,7 @@ int mlx5_ib_dev_res_cq_init(struct mlx5_ib_dev *dev) } devr->p0 = pd; - devr->c0 = cq; + smp_store_release(&devr->c0, cq); unlock: mutex_unlock(&devr->cq_lock); @@ -3354,7 +3354,7 @@ int mlx5_ib_dev_res_srq_init(struct mlx5_ib_dev *dev) * devr->s1 is set once, never changed until device unload. * Avoid taking the mutex if initialization is already done. */ - if (devr->s1) + if (smp_load_acquire(&devr->s1)) return 0; mutex_lock(&devr->srq_lock); @@ -3396,7 +3396,7 @@ int mlx5_ib_dev_res_srq_init(struct mlx5_ib_dev *dev) } devr->s0 = s0; - devr->s1 = s1; + smp_store_release(&devr->s1, s1); unlock: mutex_unlock(&devr->srq_lock); diff --git a/drivers/infiniband/hw/mlx5/umr.c b/drivers/infiniband/hw/mlx5/umr.c index 29488fba21a0..f2139474be37 100644 --- a/drivers/infiniband/hw/mlx5/umr.c +++ b/drivers/infiniband/hw/mlx5/umr.c @@ -147,7 +147,7 @@ int mlx5r_umr_resource_init(struct mlx5_ib_dev *dev) * UMR qp is set once, never changed until device unload. * Avoid taking the mutex if initialization is already done. */ - if (dev->umrc.qp) + if (smp_load_acquire(&dev->umrc.qp)) return 0; mutex_lock(&dev->umrc.init_lock); @@ -185,7 +185,7 @@ int mlx5r_umr_resource_init(struct mlx5_ib_dev *dev) sema_init(&dev->umrc.sem, MAX_UMR_WR); mutex_init(&dev->umrc.lock); dev->umrc.state = MLX5_UMR_STATE_ACTIVE; - dev->umrc.qp = qp; + smp_store_release(&dev->umrc.qp, qp); mutex_unlock(&dev->umrc.init_lock); return 0; From 6dd2d4ad9c8429523b1c220c5132bd551c006425 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:37 -0300 Subject: [PATCH 187/525] RDMA/mana: Validate rx_hash_key_len Sashiko points out that rx_hash_key_len comes from a uAPI structure and is blindly passed to memcpy, allowing the userspace to trash kernel memory. Bounds check it so the memcpy cannot overflow. Cc: stable@vger.kernel.org Fixes: 0266a177631d ("RDMA/mana_ib: Add a driver for Microsoft Azure Network Adapter") Link: https://sashiko.dev/#/patchset/0-v2-1c49eeb88c48%2B91-rdma_udata_rep_jgg%40nvidia.com?part=1 Link: https://patch.msgid.link/r/4-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Reviewed-by: Long Li Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mana/qp.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/infiniband/hw/mana/qp.c b/drivers/infiniband/hw/mana/qp.c index 645581359cee..f7bb0d1f0f80 100644 --- a/drivers/infiniband/hw/mana/qp.c +++ b/drivers/infiniband/hw/mana/qp.c @@ -21,6 +21,9 @@ static int mana_ib_cfg_vport_steering(struct mana_ib_dev *dev, gc = mdev_to_gc(dev); + if (rx_hash_key_len > sizeof(req->hashkey)) + return -EINVAL; + req_buf_size = struct_size(req, indir_tab, MANA_INDIRECT_TABLE_DEF_SIZE); req = kzalloc(req_buf_size, GFP_KERNEL); if (!req) From 159f2efabc89d3f931d38f2d35876535d4abf0a3 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:38 -0300 Subject: [PATCH 188/525] RDMA/mana: Remove user triggerable WARN_ON() in mana_ib_create_qp_rss() Sashiko points out that the user can specify WQs sharing the same CQ as a part of the uAPI and this will trigger the WARN_ON() then go on to corrupt the kernel. Just reject it outright and fail the QP creation. Cc: stable@vger.kernel.org Fixes: c15d7802a424 ("RDMA/mana_ib: Add CQ interrupt support for RAW QP") Link: https://sashiko.dev/#/patchset/0-v2-1c49eeb88c48%2B91-rdma_udata_rep_jgg%40nvidia.com?part=1 Link: https://patch.msgid.link/r/5-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Reviewed-by: Long Li Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mana/cq.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/mana/cq.c b/drivers/infiniband/hw/mana/cq.c index f4cbe21763bf..2d682428ef20 100644 --- a/drivers/infiniband/hw/mana/cq.c +++ b/drivers/infiniband/hw/mana/cq.c @@ -137,8 +137,9 @@ int mana_ib_install_cq_cb(struct mana_ib_dev *mdev, struct mana_ib_cq *cq) if (cq->queue.id >= gc->max_num_cqs) return -EINVAL; - /* Create CQ table entry */ - WARN_ON(gc->cq_table[cq->queue.id]); + /* Create CQ table entry, sharing a CQ between WQs is not supported */ + if (gc->cq_table[cq->queue.id]) + return -EINVAL; if (cq->queue.kmem) gdma_cq = cq->queue.kmem; else From 34ecf795692ee57c393109f4a24ccc313091e137 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:39 -0300 Subject: [PATCH 189/525] RDMA/mana: Fix mana_destroy_wq_obj() cleanup in mana_ib_create_qp_rss() Sashiko points out there are two bugs here in the error unwind flow, both related to how the WQ table is unwound. First there is a double i-- on the first failure path due to the while loop having a i--, remove it. Second if mana_ib_install_cq_cb() fails then mana_create_wq_obj() is not undone due to the above i--. Cc: stable@vger.kernel.org Fixes: c15d7802a424 ("RDMA/mana_ib: Add CQ interrupt support for RAW QP") Link: https://sashiko.dev/#/patchset/0-v2-1c49eeb88c48%2B91-rdma_udata_rep_jgg%40nvidia.com?part=1 Link: https://patch.msgid.link/r/6-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Reviewed-by: Long Li Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mana/qp.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/drivers/infiniband/hw/mana/qp.c b/drivers/infiniband/hw/mana/qp.c index f7bb0d1f0f80..8e1f052d0ec9 100644 --- a/drivers/infiniband/hw/mana/qp.c +++ b/drivers/infiniband/hw/mana/qp.c @@ -176,11 +176,8 @@ static int mana_ib_create_qp_rss(struct ib_qp *ibqp, struct ib_pd *pd, ret = mana_create_wq_obj(mpc, mpc->port_handle, GDMA_RQ, &wq_spec, &cq_spec, &wq->rx_object); - if (ret) { - /* Do cleanup starting with index i-1 */ - i--; + if (ret) goto fail; - } /* The GDMA regions are now owned by the WQ object */ wq->queue.gdma_region = GDMA_INVALID_DMA_REGION; @@ -200,8 +197,10 @@ static int mana_ib_create_qp_rss(struct ib_qp *ibqp, struct ib_pd *pd, /* Create CQ table entry */ ret = mana_ib_install_cq_cb(mdev, cq); - if (ret) + if (ret) { + mana_destroy_wq_obj(mpc, GDMA_RQ, wq->rx_object); goto fail; + } } resp.num_entries = i; From 6aaa978c6b6218cfac15fe1dab17c76fe229ce3f Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:40 -0300 Subject: [PATCH 190/525] RDMA/mana: Fix error unwind in mana_ib_create_qp_rss() Sashiko points out that mana_ib_cfg_vport_steering() is leaked, the normal destroy path cleans it up. Cc: stable@vger.kernel.org Fixes: 0266a177631d ("RDMA/mana_ib: Add a driver for Microsoft Azure Network Adapter") Link: https://sashiko.dev/#/patchset/0-v1-e911b76a94d1%2B65d95-rdma_udata_rep_jgg%40nvidia.com?part=4 Link: https://patch.msgid.link/r/7-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Reviewed-by: Long Li Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mana/qp.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mana/qp.c b/drivers/infiniband/hw/mana/qp.c index 8e1f052d0ec9..0fbcf449c134 100644 --- a/drivers/infiniband/hw/mana/qp.c +++ b/drivers/infiniband/hw/mana/qp.c @@ -217,13 +217,15 @@ static int mana_ib_create_qp_rss(struct ib_qp *ibqp, struct ib_pd *pd, ibdev_dbg(&mdev->ib_dev, "Failed to copy to udata create rss-qp, %d\n", ret); - goto fail; + goto err_disable_vport_rx; } kfree(mana_ind_table); return 0; +err_disable_vport_rx: + mana_disable_vport_rx(mpc); fail: while (i-- > 0) { ibwq = ind_tbl->ind_tbl[i]; From ea4e4b168a531c522d2850719816b6f583b1738b Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:41 -0300 Subject: [PATCH 191/525] RDMA/ocrdma: Clarify the mm_head searching The intention of this code is to find matching entries exactly, the driver never creates phys_addr's with different lens so the current expression is not a bug, but it doesn't make sense and confuses review tooling. Search for exact match instead. Link: https://sashiko.dev/#/patchset/0-v1-e911b76a94d1%2B65d95-rdma_udata_rep_jgg%40nvidia.com?part=4 Link: https://patch.msgid.link/r/8-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/ocrdma/ocrdma_verbs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c index c17e2a54dbca..463c9a5703fc 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c @@ -215,7 +215,7 @@ static void ocrdma_del_mmap(struct ocrdma_ucontext *uctx, u64 phy_addr, mutex_lock(&uctx->mm_list_lock); list_for_each_entry_safe(mm, tmp, &uctx->mm_head, entry) { - if (len != mm->key.len && phy_addr != mm->key.phy_addr) + if (len != mm->key.len || phy_addr != mm->key.phy_addr) continue; list_del(&mm->entry); @@ -233,7 +233,7 @@ static bool ocrdma_search_mmap(struct ocrdma_ucontext *uctx, u64 phy_addr, mutex_lock(&uctx->mm_list_lock); list_for_each_entry(mm, &uctx->mm_head, entry) { - if (len != mm->key.len && phy_addr != mm->key.phy_addr) + if (len != mm->key.len || phy_addr != mm->key.phy_addr) continue; found = true; From 34fbf48cf3b410d2a6e8c586fa952a36331ca5ba Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:42 -0300 Subject: [PATCH 192/525] RDMA/ocrdma: Don't NULL deref uctx on errors in ocrdma_copy_pd_uresp() Sashiko points out that pd->uctx isn't initialized until late in the function so all these error flow references are NULL and will crash. Use the uctx that isn't NULL. Cc: stable@vger.kernel.org Fixes: fe2caefcdf58 ("RDMA/ocrdma: Add driver for Emulex OneConnect IBoE RDMA adapter") Link: https://sashiko.dev/#/patchset/0-v1-e911b76a94d1%2B65d95-rdma_udata_rep_jgg%40nvidia.com?part=4 Link: https://patch.msgid.link/r/9-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/ocrdma/ocrdma_verbs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c index 463c9a5703fc..a88cc5d84af8 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c @@ -620,9 +620,9 @@ static int ocrdma_copy_pd_uresp(struct ocrdma_dev *dev, struct ocrdma_pd *pd, ucopy_err: if (pd->dpp_enabled) - ocrdma_del_mmap(pd->uctx, dpp_page_addr, PAGE_SIZE); + ocrdma_del_mmap(uctx, dpp_page_addr, PAGE_SIZE); dpp_map_err: - ocrdma_del_mmap(pd->uctx, db_page_addr, db_page_size); + ocrdma_del_mmap(uctx, db_page_addr, db_page_size); return status; } From e38e86995df27f1f854063dab1f0c6a513db3faf Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:43 -0300 Subject: [PATCH 193/525] RDMA/vmw_pvrdma: Fix double free on pvrdma_alloc_ucontext() error path Sashiko points out that pvrdma_uar_free() is already called within pvrdma_dealloc_ucontext(), so calling it before triggers a double free. Cc: stable@vger.kernel.org Fixes: 29c8d9eba550 ("IB: Add vmw_pvrdma driver") Link: https://sashiko.dev/#/patchset/0-v1-e911b76a94d1%2B65d95-rdma_udata_rep_jgg%40nvidia.com?part=4 Link: https://patch.msgid.link/r/10-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c index bcd43dc30e21..c7c2b41060e5 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c @@ -322,7 +322,7 @@ int pvrdma_alloc_ucontext(struct ib_ucontext *uctx, struct ib_udata *udata) uresp.qp_tab_size = vdev->dsr->caps.max_qp; ret = ib_copy_to_udata(udata, &uresp, sizeof(uresp)); if (ret) { - pvrdma_uar_free(vdev, &context->uar); + /* pvrdma_dealloc_ucontext() also frees the UAR */ pvrdma_dealloc_ucontext(&context->ibucontext); return -EFAULT; } From c54c7e4cb679c0aaa1cb489b9c3f2cd98e63a44c Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:44 -0300 Subject: [PATCH 194/525] RDMA/mlx4: Fix resource leak on error in mlx4_ib_create_srq() Sashiko points out that mlx4_srq_alloc() was not undone during error unwind, add the missing call to mlx4_srq_free(). Cc: stable@vger.kernel.org Fixes: 225c7b1feef1 ("IB/mlx4: Add a driver Mellanox ConnectX InfiniBand adapters") Link: https://sashiko.dev/#/patchset/0-v1-e911b76a94d1%2B65d95-rdma_udata_rep_jgg%40nvidia.com?part=8 Link: https://patch.msgid.link/r/11-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/mlx4/srq.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mlx4/srq.c b/drivers/infiniband/hw/mlx4/srq.c index 5b23e5f8b84a..767840736d58 100644 --- a/drivers/infiniband/hw/mlx4/srq.c +++ b/drivers/infiniband/hw/mlx4/srq.c @@ -194,13 +194,15 @@ int mlx4_ib_create_srq(struct ib_srq *ib_srq, if (udata) if (ib_copy_to_udata(udata, &srq->msrq.srqn, sizeof (__u32))) { err = -EFAULT; - goto err_wrid; + goto err_srq; } init_attr->attr.max_wr = srq->msrq.max - 1; return 0; +err_srq: + mlx4_srq_free(dev->dev, &srq->msrq); err_wrid: if (udata) mlx4_ib_db_unmap_user(ucontext, &srq->db); From c9341307ea16b9395c2e4c9c94d8499d91fe31d0 Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:45 -0300 Subject: [PATCH 195/525] RDMA/mlx4: Fix mis-use of RCU in mlx4_srq_event() Sashiko points out the radix_tree itself is RCU safe, but nothing ever frees the mlx4_srq struct with RCU, and it isn't even accessed within the RCU critical section. It also will crash if an event is delivered before the srq object is finished initializing. Use the spinlock since it isn't easy to make RCU work, use refcount_inc_not_zero() to protect against partially initialized objects, and order the refcount_set() to be after the srq is fully initialized. Cc: stable@vger.kernel.org Fixes: 30353bfc43a1 ("net/mlx4_core: Use RCU to perform radix tree lookup for SRQ") Link: https://sashiko.dev/#/patchset/0-v2-1c49eeb88c48%2B91-rdma_udata_rep_jgg%40nvidia.com?part=5 Link: https://patch.msgid.link/r/12-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Signed-off-by: Jason Gunthorpe --- drivers/net/ethernet/mellanox/mlx4/srq.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/srq.c b/drivers/net/ethernet/mellanox/mlx4/srq.c index dd890f5d7b72..8711689120f3 100644 --- a/drivers/net/ethernet/mellanox/mlx4/srq.c +++ b/drivers/net/ethernet/mellanox/mlx4/srq.c @@ -44,13 +44,14 @@ void mlx4_srq_event(struct mlx4_dev *dev, u32 srqn, int event_type) { struct mlx4_srq_table *srq_table = &mlx4_priv(dev)->srq_table; struct mlx4_srq *srq; + unsigned long flags; - rcu_read_lock(); + spin_lock_irqsave(&srq_table->lock, flags); srq = radix_tree_lookup(&srq_table->tree, srqn & (dev->caps.num_srqs - 1)); - rcu_read_unlock(); - if (srq) - refcount_inc(&srq->refcount); - else { + if (!srq || !refcount_inc_not_zero(&srq->refcount)) + srq = NULL; + spin_unlock_irqrestore(&srq_table->lock, flags); + if (!srq) { mlx4_warn(dev, "Async event for bogus SRQ %08x\n", srqn); return; } @@ -203,8 +204,8 @@ int mlx4_srq_alloc(struct mlx4_dev *dev, u32 pdn, u32 cqn, u16 xrcd, if (err) goto err_radix; - refcount_set(&srq->refcount, 1); init_completion(&srq->free); + refcount_set_release(&srq->refcount, 1); return 0; From 48973c6c938737bb900d15dc82b91dfe3586cb0f Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:46 -0300 Subject: [PATCH 196/525] RDMA/hns: Fix xarray race in hns_roce_create_srq() Sashiko points out that once the srq memory is stored into the xarray by alloc_srqc() it can immediately be looked up by: xa_lock(&srq_table->xa); srq = xa_load(&srq_table->xa, srqn & (hr_dev->caps.num_srqs - 1)); if (srq) refcount_inc(&srq->refcount); xa_unlock(&srq_table->xa); Which will fail refcount debug because the refcount is 0 and then crash: srq->event(srq, event_type); Because event is NULL. Use refcount_inc_not_zero() instead to ensure a partially prepared srq is never retrieved from the event handler and fix the ordering of the initialization so refcount becomes 1 only after it is fully ready. All the initialization must be done before calling free_srqc() since it depends on the completion and refcount. Fixes: 9a4435375cd1 ("IB/hns: Add driver files for hns RoCE driver") Link: https://sashiko.dev/#/patchset/0-v1-e911b76a94d1%2B65d95-rdma_udata_rep_jgg%40nvidia.com?part=3 Link: https://patch.msgid.link/r/13-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Reviewed-by: Junxian Huang Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/hns/hns_roce_srq.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/infiniband/hw/hns/hns_roce_srq.c b/drivers/infiniband/hw/hns/hns_roce_srq.c index cb848e8e6bbd..8b94cbdfa54d 100644 --- a/drivers/infiniband/hw/hns/hns_roce_srq.c +++ b/drivers/infiniband/hw/hns/hns_roce_srq.c @@ -16,8 +16,8 @@ void hns_roce_srq_event(struct hns_roce_dev *hr_dev, u32 srqn, int event_type) xa_lock(&srq_table->xa); srq = xa_load(&srq_table->xa, srqn & (hr_dev->caps.num_srqs - 1)); - if (srq) - refcount_inc(&srq->refcount); + if (srq && !refcount_inc_not_zero(&srq->refcount)) + srq = NULL; xa_unlock(&srq_table->xa); if (!srq) { @@ -470,6 +470,10 @@ int hns_roce_create_srq(struct ib_srq *ib_srq, if (ret) goto err_srqn; + srq->event = hns_roce_ib_srq_event; + init_completion(&srq->free); + refcount_set_release(&srq->refcount, 1); + if (udata) { resp.cap_flags = srq->cap_flags; resp.srqn = srq->srqn; @@ -480,10 +484,6 @@ int hns_roce_create_srq(struct ib_srq *ib_srq, } } - srq->event = hns_roce_ib_srq_event; - refcount_set(&srq->refcount, 1); - init_completion(&srq->free); - return 0; err_srqc: From 7d51783d82fea000a9ce96fa1dcf3e0a8cedc4fb Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:47 -0300 Subject: [PATCH 197/525] RDMA/hns: Fix xarray race in hns_roce_create_qp_common() Similar to the SRQ case the hr_qp is stored in the xarray before it is fully initialized. Unlike the SRQ case the error unwinds do not wait for the completion so keep the refcount 0 until the function succeeds. Fixes: 9a4435375cd1 ("IB/hns: Add driver files for hns RoCE driver") Link: https://patch.msgid.link/r/14-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Suggested-by: Junxian Huang Reviewed-by: Junxian Huang Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/hns/hns_roce_qp.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/infiniband/hw/hns/hns_roce_qp.c b/drivers/infiniband/hw/hns/hns_roce_qp.c index a27ea85bb063..f94ba98871f0 100644 --- a/drivers/infiniband/hw/hns/hns_roce_qp.c +++ b/drivers/infiniband/hw/hns/hns_roce_qp.c @@ -47,8 +47,8 @@ static struct hns_roce_qp *hns_roce_qp_lookup(struct hns_roce_dev *hr_dev, xa_lock_irqsave(&hr_dev->qp_table_xa, flags); qp = __hns_roce_qp_lookup(hr_dev, qpn); - if (qp) - refcount_inc(&qp->refcount); + if (qp && !refcount_inc_not_zero(&qp->refcount)) + qp = NULL; xa_unlock_irqrestore(&hr_dev->qp_table_xa, flags); if (!qp) @@ -1251,8 +1251,8 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev, hr_qp->ibqp.qp_num = hr_qp->qpn; hr_qp->event = hns_roce_ib_qp_event; - refcount_set(&hr_qp->refcount, 1); init_completion(&hr_qp->free); + refcount_set_release(&hr_qp->refcount, 1); return 0; From 0c99acbc8b6c6dd526ae475a48ee1897b61072fb Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 28 Apr 2026 13:17:48 -0300 Subject: [PATCH 198/525] RDMA/hns: Fix unlocked call to hns_roce_qp_remove() Sashiko points out that hns_roce_qp_remove() requires the caller to hold locks. The error flow in hns_roce_create_qp_common() doesn't hold those locks for the error unwind so it risks corrupting memory. Grab the same locks the other two callers use. Cc: stable@vger.kernel.org Fixes: e088a685eae9 ("RDMA/hns: Support rq record doorbell for the user space") Link: https://sashiko.dev/#/patchset/0-v2-1c49eeb88c48%2B91-rdma_udata_rep_jgg%40nvidia.com?part=9 Link: https://patch.msgid.link/r/15-v1-41f3135e5565+9d2-rdma_ai_fixes1_jgg@nvidia.com Reviewed-by: Junxian Huang Signed-off-by: Jason Gunthorpe --- drivers/infiniband/hw/hns/hns_roce_qp.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/infiniband/hw/hns/hns_roce_qp.c b/drivers/infiniband/hw/hns/hns_roce_qp.c index f94ba98871f0..bf04ee84a943 100644 --- a/drivers/infiniband/hw/hns/hns_roce_qp.c +++ b/drivers/infiniband/hw/hns/hns_roce_qp.c @@ -1171,6 +1171,7 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev, struct hns_roce_ib_create_qp_resp resp = {}; struct ib_device *ibdev = &hr_dev->ib_dev; struct hns_roce_ib_create_qp ucmd = {}; + unsigned long flags; int ret; mutex_init(&hr_qp->mutex); @@ -1257,7 +1258,13 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev, return 0; err_flow_ctrl: + spin_lock_irqsave(&hr_dev->qp_list_lock, flags); + hns_roce_lock_cqs(init_attr->send_cq ? to_hr_cq(init_attr->send_cq) : NULL, + init_attr->recv_cq ? to_hr_cq(init_attr->recv_cq) : NULL); hns_roce_qp_remove(hr_dev, hr_qp); + hns_roce_unlock_cqs(init_attr->send_cq ? to_hr_cq(init_attr->send_cq) : NULL, + init_attr->recv_cq ? to_hr_cq(init_attr->recv_cq) : NULL); + spin_unlock_irqrestore(&hr_dev->qp_list_lock, flags); err_store: free_qpc(hr_dev, hr_qp); err_qpc: From 046111a1a35a1720748f254377d3d1663664ea61 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 30 Apr 2026 06:16:09 +0000 Subject: [PATCH 199/525] net/sched: sch_cake: annotate data-races in cake_dump_class_stats (I) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit cake_dump_class_stats() runs without qdisc spinlock being held. In this first patch, I add READ_ONCE()/WRITE_ONCE() annotations for: - flow->head - flow->dropped - b->backlogs[] Fixes: 046f6fd5daef ("sched: Add Common Applications Kept Enhanced (cake) qdisc") Signed-off-by: Eric Dumazet Acked-by: Toke Høiland-Jørgensen Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260430061610.3503483-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/sched/sch_cake.c | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/net/sched/sch_cake.c b/net/sched/sch_cake.c index 13c6d1869a14..806eb73d6a05 100644 --- a/net/sched/sch_cake.c +++ b/net/sched/sch_cake.c @@ -914,7 +914,7 @@ static struct sk_buff *dequeue_head(struct cake_flow *flow) struct sk_buff *skb = flow->head; if (skb) { - flow->head = skb->next; + WRITE_ONCE(flow->head, skb->next); skb_mark_not_on_list(skb); } @@ -926,7 +926,7 @@ static struct sk_buff *dequeue_head(struct cake_flow *flow) static void flow_queue_add(struct cake_flow *flow, struct sk_buff *skb) { if (!flow->head) - flow->head = skb; + WRITE_ONCE(flow->head, skb); else flow->tail->next = skb; flow->tail = skb; @@ -1357,7 +1357,7 @@ static struct sk_buff *cake_ack_filter(struct cake_sched_data *q, if (elig_ack_prev) elig_ack_prev->next = elig_ack->next; else - flow->head = elig_ack->next; + WRITE_ONCE(flow->head, elig_ack->next); skb_mark_not_on_list(elig_ack); @@ -1595,11 +1595,11 @@ static unsigned int cake_drop(struct Qdisc *sch, struct sk_buff **to_free) len = qdisc_pkt_len(skb); q->buffer_used -= skb->truesize; - b->backlogs[idx] -= len; WRITE_ONCE(b->tin_backlog, b->tin_backlog - len); + WRITE_ONCE(b->backlogs[idx], b->backlogs[idx] - len); sch->qstats.backlog -= len; - flow->dropped++; + WRITE_ONCE(flow->dropped, flow->dropped + 1); WRITE_ONCE(b->tin_dropped, b->tin_dropped + 1); if (q->config->rate_flags & CAKE_FLAG_INGRESS) @@ -1824,11 +1824,11 @@ static s32 cake_enqueue(struct sk_buff *skb, struct Qdisc *sch, } /* stats */ - b->backlogs[idx] += slen; sch->qstats.backlog += slen; q->avg_window_bytes += slen; WRITE_ONCE(b->bytes, b->bytes + slen); WRITE_ONCE(b->tin_backlog, b->tin_backlog + slen); + WRITE_ONCE(b->backlogs[idx], b->backlogs[idx] + slen); qdisc_tree_reduce_backlog(sch, 1-numsegs, len-slen); consume_skb(skb); @@ -1861,11 +1861,11 @@ static s32 cake_enqueue(struct sk_buff *skb, struct Qdisc *sch, /* stats */ WRITE_ONCE(b->packets, b->packets + 1); - b->backlogs[idx] += len - ack_pkt_len; sch->qstats.backlog += len - ack_pkt_len; q->avg_window_bytes += len - ack_pkt_len; WRITE_ONCE(b->bytes, b->bytes + len - ack_pkt_len); WRITE_ONCE(b->tin_backlog, b->tin_backlog + len - ack_pkt_len); + WRITE_ONCE(b->backlogs[idx], b->backlogs[idx] + len - ack_pkt_len); } if (q->overflow_timeout) @@ -1977,7 +1977,7 @@ static struct sk_buff *cake_dequeue_one(struct Qdisc *sch) if (flow->head) { skb = dequeue_head(flow); len = qdisc_pkt_len(skb); - b->backlogs[q->cur_flow] -= len; + WRITE_ONCE(b->backlogs[q->cur_flow], b->backlogs[q->cur_flow] - len); WRITE_ONCE(b->tin_backlog, b->tin_backlog - len); sch->qstats.backlog -= len; q->buffer_used -= skb->truesize; @@ -2235,7 +2235,7 @@ static struct sk_buff *cake_dequeue(struct Qdisc *sch) flow->deficit -= len; b->tin_deficit -= len; } - flow->dropped++; + WRITE_ONCE(flow->dropped, flow->dropped + 1); WRITE_ONCE(b->tin_dropped, b->tin_dropped + 1); qdisc_tree_reduce_backlog(sch, 1, qdisc_pkt_len(skb)); qdisc_qstats_drop(sch); @@ -3137,7 +3137,7 @@ static int cake_dump_class_stats(struct Qdisc *sch, unsigned long cl, flow = &b->flows[idx % CAKE_QUEUES]; - if (flow->head) { + if (READ_ONCE(flow->head)) { sch_tree_lock(sch); skb = flow->head; while (skb) { @@ -3146,8 +3146,8 @@ static int cake_dump_class_stats(struct Qdisc *sch, unsigned long cl, } sch_tree_unlock(sch); } - qs.backlog = b->backlogs[idx % CAKE_QUEUES]; - qs.drops = flow->dropped; + qs.backlog = READ_ONCE(b->backlogs[idx % CAKE_QUEUES]); + qs.drops = READ_ONCE(flow->dropped); } if (gnet_stats_copy_queue(d, NULL, &qs, qs.qlen) < 0) return -1; From 67dc6c56b871617deac85b9f72500b69b1fdf835 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 30 Apr 2026 06:16:10 +0000 Subject: [PATCH 200/525] net/sched: sch_cake: annotate data-races in cake_dump_class_stats (II) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit cake_dump_class_stats() runs without qdisc spinlock being held. In this second patch, I add READ_ONCE()/WRITE_ONCE() annotations for: - flow->deficit - flow->cvars.dropping - flow->cvars.count - flow->cvars.p_drop - flow->cvars.blue_timer - flow->cvars.drop_next Fixes: 046f6fd5daef ("sched: Add Common Applications Kept Enhanced (cake) qdisc") Signed-off-by: Eric Dumazet Acked-by: Toke Høiland-Jørgensen Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260430061610.3503483-3-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/sched/sch_cake.c | 129 +++++++++++++++++++++++-------------------- 1 file changed, 70 insertions(+), 59 deletions(-) diff --git a/net/sched/sch_cake.c b/net/sched/sch_cake.c index 806eb73d6a05..5862933be8d7 100644 --- a/net/sched/sch_cake.c +++ b/net/sched/sch_cake.c @@ -399,14 +399,14 @@ static void cake_configure_rates(struct Qdisc *sch, u64 rate, bool rate_adjust); * Here, invsqrt is a fixed point number (< 1.0), 32bit mantissa, aka Q0.32 */ -static void cobalt_newton_step(struct cobalt_vars *vars) +static void cobalt_newton_step(struct cobalt_vars *vars, u32 count) { u32 invsqrt, invsqrt2; u64 val; invsqrt = vars->rec_inv_sqrt; invsqrt2 = ((u64)invsqrt * invsqrt) >> 32; - val = (3LL << 32) - ((u64)vars->count * invsqrt2); + val = (3LL << 32) - ((u64)count * invsqrt2); val >>= 2; /* avoid overflow in following multiply */ val = (val * invsqrt) >> (32 - 2 + 1); @@ -414,12 +414,12 @@ static void cobalt_newton_step(struct cobalt_vars *vars) vars->rec_inv_sqrt = val; } -static void cobalt_invsqrt(struct cobalt_vars *vars) +static void cobalt_invsqrt(struct cobalt_vars *vars, u32 count) { - if (vars->count < REC_INV_SQRT_CACHE) - vars->rec_inv_sqrt = inv_sqrt_cache[vars->count]; + if (count < REC_INV_SQRT_CACHE) + vars->rec_inv_sqrt = inv_sqrt_cache[count]; else - cobalt_newton_step(vars); + cobalt_newton_step(vars, count); } static void cobalt_vars_init(struct cobalt_vars *vars) @@ -449,16 +449,19 @@ static bool cobalt_queue_full(struct cobalt_vars *vars, bool up = false; if (ktime_to_ns(ktime_sub(now, vars->blue_timer)) > p->target) { - up = !vars->p_drop; - vars->p_drop += p->p_inc; - if (vars->p_drop < p->p_inc) - vars->p_drop = ~0; - vars->blue_timer = now; + u32 p_drop = vars->p_drop; + + up = !p_drop; + p_drop += p->p_inc; + if (p_drop < p->p_inc) + p_drop = ~0; + WRITE_ONCE(vars->p_drop, p_drop); + WRITE_ONCE(vars->blue_timer, now); } - vars->dropping = true; - vars->drop_next = now; + WRITE_ONCE(vars->dropping, true); + WRITE_ONCE(vars->drop_next, now); if (!vars->count) - vars->count = 1; + WRITE_ONCE(vars->count, 1); return up; } @@ -475,20 +478,20 @@ static bool cobalt_queue_empty(struct cobalt_vars *vars, if (vars->p_drop && ktime_to_ns(ktime_sub(now, vars->blue_timer)) > p->target) { if (vars->p_drop < p->p_dec) - vars->p_drop = 0; + WRITE_ONCE(vars->p_drop, 0); else - vars->p_drop -= p->p_dec; - vars->blue_timer = now; + WRITE_ONCE(vars->p_drop, vars->p_drop - p->p_dec); + WRITE_ONCE(vars->blue_timer, now); down = !vars->p_drop; } - vars->dropping = false; + WRITE_ONCE(vars->dropping, false); if (vars->count && ktime_to_ns(ktime_sub(now, vars->drop_next)) >= 0) { - vars->count--; - cobalt_invsqrt(vars); - vars->drop_next = cobalt_control(vars->drop_next, - p->interval, - vars->rec_inv_sqrt); + WRITE_ONCE(vars->count, vars->count - 1); + cobalt_invsqrt(vars, vars->count); + WRITE_ONCE(vars->drop_next, + cobalt_control(vars->drop_next, p->interval, + vars->rec_inv_sqrt)); } return down; @@ -507,6 +510,7 @@ static enum qdisc_drop_reason cobalt_should_drop(struct cobalt_vars *vars, bool next_due, over_target; ktime_t schedule; u64 sojourn; + u32 count; /* The 'schedule' variable records, in its sign, whether 'now' is before or * after 'drop_next'. This allows 'drop_next' to be updated before the next @@ -528,21 +532,22 @@ static enum qdisc_drop_reason cobalt_should_drop(struct cobalt_vars *vars, over_target = sojourn > p->target && sojourn > p->mtu_time * bulk_flows * 2 && sojourn > p->mtu_time * 4; - next_due = vars->count && ktime_to_ns(schedule) >= 0; + count = vars->count; + next_due = count && ktime_to_ns(schedule) >= 0; vars->ecn_marked = false; if (over_target) { if (!vars->dropping) { - vars->dropping = true; - vars->drop_next = cobalt_control(now, - p->interval, - vars->rec_inv_sqrt); + WRITE_ONCE(vars->dropping, true); + WRITE_ONCE(vars->drop_next, + cobalt_control(now, p->interval, + vars->rec_inv_sqrt)); } - if (!vars->count) - vars->count = 1; + if (!count) + count = 1; } else if (vars->dropping) { - vars->dropping = false; + WRITE_ONCE(vars->dropping, false); } if (next_due && vars->dropping) { @@ -550,23 +555,23 @@ static enum qdisc_drop_reason cobalt_should_drop(struct cobalt_vars *vars, if (!(vars->ecn_marked = INET_ECN_set_ce(skb))) reason = QDISC_DROP_CONGESTED; - vars->count++; - if (!vars->count) - vars->count--; - cobalt_invsqrt(vars); - vars->drop_next = cobalt_control(vars->drop_next, - p->interval, - vars->rec_inv_sqrt); + count++; + if (!count) + count--; + cobalt_invsqrt(vars, count); + WRITE_ONCE(vars->drop_next, + cobalt_control(vars->drop_next, p->interval, + vars->rec_inv_sqrt)); schedule = ktime_sub(now, vars->drop_next); } else { while (next_due) { - vars->count--; - cobalt_invsqrt(vars); - vars->drop_next = cobalt_control(vars->drop_next, - p->interval, - vars->rec_inv_sqrt); + count--; + cobalt_invsqrt(vars, count); + WRITE_ONCE(vars->drop_next, + cobalt_control(vars->drop_next, p->interval, + vars->rec_inv_sqrt)); schedule = ktime_sub(now, vars->drop_next); - next_due = vars->count && ktime_to_ns(schedule) >= 0; + next_due = count && ktime_to_ns(schedule) >= 0; } } @@ -575,11 +580,12 @@ static enum qdisc_drop_reason cobalt_should_drop(struct cobalt_vars *vars, get_random_u32() < vars->p_drop) reason = QDISC_DROP_FLOOD_PROTECTION; + WRITE_ONCE(vars->count, count); /* Overload the drop_next field as an activity timeout */ - if (!vars->count) - vars->drop_next = ktime_add_ns(now, p->interval); + if (!count) + WRITE_ONCE(vars->drop_next, ktime_add_ns(now, p->interval)); else if (ktime_to_ns(schedule) > 0 && reason == QDISC_DROP_UNSPEC) - vars->drop_next = now; + WRITE_ONCE(vars->drop_next, now); return reason; } @@ -1924,7 +1930,7 @@ static s32 cake_enqueue(struct sk_buff *skb, struct Qdisc *sch, flow->set = CAKE_SET_SPARSE; WRITE_ONCE(b->sparse_flow_count, b->sparse_flow_count + 1); - flow->deficit = cake_get_flow_quantum(b, flow, q->config->flow_mode); + WRITE_ONCE(flow->deficit, cake_get_flow_quantum(b, flow, q->config->flow_mode)); } else if (flow->set == CAKE_SET_SPARSE_WAIT) { /* this flow was empty, accounted as a sparse flow, but actually * in the bulk rotation. @@ -2166,7 +2172,8 @@ static struct sk_buff *cake_dequeue(struct Qdisc *sch) } } - flow->deficit += cake_get_flow_quantum(b, flow, q->config->flow_mode); + WRITE_ONCE(flow->deficit, + flow->deficit + cake_get_flow_quantum(b, flow, q->config->flow_mode)); list_move_tail(&flow->flowchain, &b->old_flows); goto retry; @@ -2232,7 +2239,7 @@ static struct sk_buff *cake_dequeue(struct Qdisc *sch) if (q->config->rate_flags & CAKE_FLAG_INGRESS) { len = cake_advance_shaper(q, b, skb, now, true); - flow->deficit -= len; + WRITE_ONCE(flow->deficit, flow->deficit - len); b->tin_deficit -= len; } WRITE_ONCE(flow->dropped, flow->dropped + 1); @@ -2259,7 +2266,7 @@ static struct sk_buff *cake_dequeue(struct Qdisc *sch) delay < b->base_delay ? 2 : 8)); len = cake_advance_shaper(q, b, skb, now, false); - flow->deficit -= len; + WRITE_ONCE(flow->deficit, flow->deficit - len); b->tin_deficit -= len; if (ktime_after(q->time_next_packet, now) && sch->q.qlen) { @@ -3153,6 +3160,8 @@ static int cake_dump_class_stats(struct Qdisc *sch, unsigned long cl, return -1; if (flow) { ktime_t now = ktime_get(); + bool dropping; + u32 p_drop; stats = nla_nest_start_noflag(d->skb, TCA_STATS_APP); if (!stats) @@ -3167,21 +3176,23 @@ static int cake_dump_class_stats(struct Qdisc *sch, unsigned long cl, goto nla_put_failure; \ } while (0) - PUT_STAT_S32(DEFICIT, flow->deficit); - PUT_STAT_U32(DROPPING, flow->cvars.dropping); - PUT_STAT_U32(COBALT_COUNT, flow->cvars.count); - PUT_STAT_U32(P_DROP, flow->cvars.p_drop); - if (flow->cvars.p_drop) { + PUT_STAT_S32(DEFICIT, READ_ONCE(flow->deficit)); + dropping = READ_ONCE(flow->cvars.dropping); + PUT_STAT_U32(DROPPING, dropping); + PUT_STAT_U32(COBALT_COUNT, READ_ONCE(flow->cvars.count)); + p_drop = READ_ONCE(flow->cvars.p_drop); + PUT_STAT_U32(P_DROP, p_drop); + if (p_drop) { PUT_STAT_S32(BLUE_TIMER_US, ktime_to_us( ktime_sub(now, - flow->cvars.blue_timer))); + READ_ONCE(flow->cvars.blue_timer)))); } - if (flow->cvars.dropping) { + if (dropping) { PUT_STAT_S32(DROP_NEXT_US, ktime_to_us( ktime_sub(now, - flow->cvars.drop_next))); + READ_ONCE(flow->cvars.drop_next)))); } if (nla_nest_end(d->skb, stats) < 0) From 7e7be31bfdb066c1c780dcd6b1224078fc54063f Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 29 Apr 2026 15:29:38 -0700 Subject: [PATCH 201/525] net: tls: fix silent data drop under pipe back-pressure tls_sw_splice_read() uses len when advancing rxm->offset / rxm->full_len after skb_splice_bits(), rather than copied (the actual number of bytes successfully spliced into the pipe). When the destination pipe cannot accept all the requested bytes, splice_to_pipe() returns fewer bytes than len, and 'len - copied' of data is effectively skipped over. Fixes: e062fe99cccd ("tls: splice_read: fix accessing pre-processed records") Link: https://patch.msgid.link/20260429222944.2139041-2-kuba@kernel.org Signed-off-by: Jakub Kicinski --- net/tls/tls_sw.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c index 798243eabb1f..2590e855f6a5 100644 --- a/net/tls/tls_sw.c +++ b/net/tls/tls_sw.c @@ -2317,9 +2317,9 @@ ssize_t tls_sw_splice_read(struct socket *sock, loff_t *ppos, if (copied < 0) goto splice_requeue; - if (chunk < rxm->full_len) { - rxm->offset += len; - rxm->full_len -= len; + if (copied < rxm->full_len) { + rxm->offset += copied; + rxm->full_len -= copied; goto splice_requeue; } From bd3a4795d5744f59a1f485379f1303e5e606f377 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 29 Apr 2026 15:29:39 -0700 Subject: [PATCH 202/525] selftests: tls: add test for data loss on small pipe Add selftest for data loss on short splice. Link: https://patch.msgid.link/20260429222944.2139041-3-kuba@kernel.org Signed-off-by: Jakub Kicinski --- tools/testing/selftests/net/tls.c | 43 +++++++++++++++++++++++++++++++ 1 file changed, 43 insertions(+) diff --git a/tools/testing/selftests/net/tls.c b/tools/testing/selftests/net/tls.c index 9e2ccea13d70..30a236b8e9f7 100644 --- a/tools/testing/selftests/net/tls.c +++ b/tools/testing/selftests/net/tls.c @@ -946,6 +946,49 @@ TEST_F(tls, peek_and_splice) EXPECT_EQ(memcmp(mem_send, mem_recv, send_len), 0); } +TEST_F(tls, splice_to_pipe_small) +{ + int send_len = TLS_PAYLOAD_MAX_LEN; + char mem_send[TLS_PAYLOAD_MAX_LEN]; + char mem_recv[TLS_PAYLOAD_MAX_LEN]; + size_t total = 0; + int p[2]; + + memrnd(mem_send, sizeof(mem_send)); + + ASSERT_GE(pipe(p), 0); + + /* Shrink pipe to 1 page (typically 4096 bytes) to force multiple + * splice iterations for a 16384-byte TLS record. + */ + EXPECT_GE(fcntl(p[1], F_SETPIPE_SZ, 4096), 4096); + + EXPECT_EQ(send(self->fd, mem_send, send_len, 0), send_len); + + while (total < (size_t)send_len) { + ssize_t spliced, drained; + + spliced = splice(self->cfd, NULL, p[1], NULL, + send_len - total, 0); + EXPECT_GT(spliced, 0); + if (spliced <= 0) + break; + + drained = read(p[0], mem_recv + total, spliced); + EXPECT_EQ(drained, spliced); + if (drained <= 0) + break; + + total += drained; + } + + EXPECT_EQ(total, (size_t)send_len); + EXPECT_EQ(memcmp(mem_send, mem_recv, send_len), 0); + + close(p[0]); + close(p[1]); +} + #define MAX_FRAGS 48 TEST_F(tls, splice_short) { From 0a69ac25bd596d50823d530d0a2004336668c0df Mon Sep 17 00:00:00 2001 From: Eliot Courtney Date: Fri, 1 May 2026 19:49:37 +0900 Subject: [PATCH 203/525] rust: drm: fix unsound initialization in drm::Device::new If pinned initialization of drm::Device::Data fails, it calls drm::Device::release via drm_dev_put. This materializes a reference to &drm::Device, but it's not fully constructed yet, because initializing `data` failed. It should not be dropped either. Instead, if pinned initialization fails, make sure drm::Device::release isn't called. Fixes: 2e9fdbe5ec7a ("rust: drm: device: drop_in_place() the drm::Device in release()") Signed-off-by: Eliot Courtney Reviewed-by: Gary Guo Link: https://patch.msgid.link/20260501-fix-drm-1-v2-1-5c4f681837bc@nvidia.com Signed-off-by: Danilo Krummrich --- rust/kernel/drm/device.rs | 22 ++++++++++++++++------ 1 file changed, 16 insertions(+), 6 deletions(-) diff --git a/rust/kernel/drm/device.rs b/rust/kernel/drm/device.rs index adbafe8db54d..403fc35353c7 100644 --- a/rust/kernel/drm/device.rs +++ b/rust/kernel/drm/device.rs @@ -119,13 +119,20 @@ pub fn new(dev: &device::Device, data: impl PinInit) -> Result()); + // Use a temporary vtable without a `release` callback until `data` is initialized, so + // init failure can release the DRM device without dropping uninitialized fields. + let alloc_vtable = bindings::drm_driver { + release: None, + ..Self::VTABLE + }; + // SAFETY: - // - `VTABLE`, as a `const` is pinned to the read-only section of the compilation, + // - `alloc_vtable` reference remains valid until no longer used, // - `dev` is valid by its type invarants, let raw_drm: *mut Self = unsafe { bindings::__drm_dev_alloc( dev.as_raw(), - &Self::VTABLE, + &alloc_vtable, layout.size(), mem::offset_of!(Self, dev), ) @@ -133,6 +140,10 @@ pub fn new(dev: &device::Device, data: impl PinInit) -> Result) -> Result Date: Mon, 27 Apr 2026 14:25:40 +0200 Subject: [PATCH 204/525] KVM: x86: check for nEPT/nNPT in slow flush hypercalls Checking is_guest_mode(vcpu) is incorrect, because translate_nested_gpa() is only valid if an L2 guest is running *with nested EPT/NPT enabled*. Instead use the same condition as translate_nested_gpa() itself. Cc: stable@vger.kernel.org Reviewed-by: Sean Christopherson Fixes: aee738236dca ("KVM: x86: Prepare kvm_hv_flush_tlb() to handle L2's GPAs", 2022-11-18) Link: https://patch.msgid.link/20260503200905.106077-1-pbonzini@redhat.com/ Signed-off-by: Paolo Bonzini --- arch/x86/kvm/hyperv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c index 9b140bbdc1d8..4438ecac9a89 100644 --- a/arch/x86/kvm/hyperv.c +++ b/arch/x86/kvm/hyperv.c @@ -2040,7 +2040,7 @@ static u64 kvm_hv_flush_tlb(struct kvm_vcpu *vcpu, struct kvm_hv_hcall *hc) * flush). Translate the address here so the memory can be uniformly * read with kvm_read_guest(). */ - if (!hc->fast && is_guest_mode(vcpu)) { + if (!hc->fast && mmu_is_nested(vcpu)) { hc->ingpa = translate_nested_gpa(vcpu, hc->ingpa, 0, NULL); if (unlikely(hc->ingpa == INVALID_GPA)) return HV_STATUS_INVALID_HYPERCALL_INPUT; From 33fd0ccd2590b470b65adcca288615ad3b5e3e06 Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Sun, 3 May 2026 19:19:32 +0200 Subject: [PATCH 205/525] KVM: x86: Do IRR scan in __kvm_apic_update_irr even if PIR is empty Fall back to apic_find_highest_vector() when PID.ON is set but PIR turns out to be empty, to correctly report the highest pending interrupt from the existing IRR. In a nested VM stress test, the following WARNING fires in vmx_check_nested_events() when kvm_cpu_has_interrupt() reports a pending interrupt but the subsequent kvm_apic_has_interrupt() (which invokes vmx_sync_pir_to_irr() again) returns -1: WARNING: CPU: 99 PID: 57767 at arch/x86/kvm/vmx/nested.c:4449 vmx_check_nested_events+0x6bf/0x6e0 [kvm_intel] Call Trace: kvm_check_and_inject_events vcpu_enter_guest.constprop.0 vcpu_run kvm_arch_vcpu_ioctl_run kvm_vcpu_ioctl __x64_sys_ioctl do_syscall_64 entry_SYSCALL_64_after_hwframe The root cause is a race between vmx_sync_pir_to_irr() on the target vCPU and __vmx_deliver_posted_interrupt() on a sender vCPU. The sender performs two individually-atomic operations that are not a single transaction: 1. pi_test_and_set_pir(vector) -- sets the PIR bit 2. pi_test_and_set_on() -- sets PID.ON The following interleaving triggers the bug: Sender vCPU (IPI): Target vCPU (1st sync_pir_to_irr): B1: set PIR[vector] A1: pi_clear_on() A2: pi_harvest_pir() -> sees B1 bit A3: xchg() -> consumes bit, PIR=0 (1st sync returns correct max_irr) B2: set PID.ON = 1 Target vCPU (2nd sync_pir_to_irr): C1: pi_test_on() -> TRUE (from B2) C2: pi_clear_on() -> ON=0 C3: pi_harvest_pir() -> PIR empty C4: *max_irr = -1, early return IRR NOT SCANNED The interrupt is not lost (it resides in the IRR from the first sync and is recovered on the next vcpu_enter_guest() iteration), but the incorrect max_irr causes a spurious WARNING and a wasted L2 VM-Enter/VM-Exit cycle. Fixes: b41f8638b9d3 ("KVM: VMX: Isolate pure loads from atomic XCHG when processing PIR") Reported-by: Farrah Chen Analyzed-by: Chenyi Qiang Cc: stable@vger.kernel.org Reviewed-by: Sean Christopherson Link: https://lore.kernel.org/kvm/20260428070349.1633238-1-chenyi.qiang@intel.com/T/ Link: https://patch.msgid.link/20260503201703.108231-2-pbonzini@redhat.com/ Signed-off-by: Paolo Bonzini --- arch/x86/kvm/lapic.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c index e3ec4d8607c1..5ee14d6bc288 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -669,12 +669,14 @@ bool __kvm_apic_update_irr(unsigned long *pir, void *regs, int *max_irr) u32 irr_val, prev_irr_val; int max_updated_irr; + if (!pi_harvest_pir(pir, pir_vals)) { + *max_irr = apic_find_highest_vector(regs + APIC_IRR); + return false; + } + max_updated_irr = -1; *max_irr = -1; - if (!pi_harvest_pir(pir, pir_vals)) - return false; - for (i = vec = 0; i <= 7; i++, vec += 32) { u32 *p_irr = (u32 *)(regs + APIC_IRR + i * 0x10); From 0aec99f9bf0213f7910688472e1ccf517c0e8b5a Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Tue, 28 Apr 2026 08:50:43 -0700 Subject: [PATCH 206/525] KVM: x86: Fix misleading variable names and add more comments for PIR=>IRR flow Rename kvm_apic_update_irr()'s "irr_updated" and vmx_sync_pir_to_irr()'s "got_posted_interrupt" to a more accurate "max_irr_is_from_pir", as neither "irr_updated" nor "got_posted_interrupt" is accurate. __kvm_apic_update_irr() and thus kvm_apic_update_irr() specifically return true if and only if the highest priority IRQ, i.e. max_irr, is a "new" pending IRQ from the PIR. I.e. it's possible for the IRR to be updated, i.e. for a posted IRQ to be "got", *without* the APIs returning true. Expand vmx_sync_pir_to_irr()'s comment to explain why it's necessary to set KVM_REQ_EVENT only if a "new" IRQ was found, and to explain why it's safe to do so only if a new IRQ is also the highest priority pending IRQ. No functional change intended. Signed-off-by: Sean Christopherson Link: https://patch.msgid.link/20260503201703.108231-3-pbonzini@redhat.com/ Signed-off-by: Paolo Bonzini --- arch/x86/kvm/lapic.c | 16 ++++++++-------- arch/x86/kvm/vmx/vmx.c | 40 ++++++++++++++++++++++++++++++++-------- 2 files changed, 40 insertions(+), 16 deletions(-) diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c index 5ee14d6bc288..4078e624ca66 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -667,14 +667,14 @@ bool __kvm_apic_update_irr(unsigned long *pir, void *regs, int *max_irr) u32 *__pir = (void *)pir_vals; u32 i, vec; u32 irr_val, prev_irr_val; - int max_updated_irr; + int max_new_irr; if (!pi_harvest_pir(pir, pir_vals)) { *max_irr = apic_find_highest_vector(regs + APIC_IRR); return false; } - max_updated_irr = -1; + max_new_irr = -1; *max_irr = -1; for (i = vec = 0; i <= 7; i++, vec += 32) { @@ -690,25 +690,25 @@ bool __kvm_apic_update_irr(unsigned long *pir, void *regs, int *max_irr) !try_cmpxchg(p_irr, &prev_irr_val, irr_val)); if (prev_irr_val != irr_val) - max_updated_irr = __fls(irr_val ^ prev_irr_val) + vec; + max_new_irr = __fls(irr_val ^ prev_irr_val) + vec; } if (irr_val) *max_irr = __fls(irr_val) + vec; } - return ((max_updated_irr != -1) && - (max_updated_irr == *max_irr)); + return max_new_irr != -1 && max_new_irr == *max_irr; } EXPORT_SYMBOL_FOR_KVM_INTERNAL(__kvm_apic_update_irr); bool kvm_apic_update_irr(struct kvm_vcpu *vcpu, unsigned long *pir, int *max_irr) { struct kvm_lapic *apic = vcpu->arch.apic; - bool irr_updated = __kvm_apic_update_irr(pir, apic->regs, max_irr); + bool max_irr_is_from_pir; - if (unlikely(!apic->apicv_active && irr_updated)) + max_irr_is_from_pir = __kvm_apic_update_irr(pir, apic->regs, max_irr); + if (unlikely(!apic->apicv_active && max_irr_is_from_pir)) apic->irr_pending = true; - return irr_updated; + return max_irr_is_from_pir; } EXPORT_SYMBOL_FOR_KVM_INTERNAL(kvm_apic_update_irr); diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index a29896a9ef14..5c2c33a5f7dc 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -7029,8 +7029,8 @@ static void vmx_set_rvi(int vector) int vmx_sync_pir_to_irr(struct kvm_vcpu *vcpu) { struct vcpu_vt *vt = to_vt(vcpu); + bool max_irr_is_from_pir; int max_irr; - bool got_posted_interrupt; if (KVM_BUG_ON(!enable_apicv, vcpu->kvm)) return -EIO; @@ -7042,17 +7042,22 @@ int vmx_sync_pir_to_irr(struct kvm_vcpu *vcpu) * But on x86 this is just a compiler barrier anyway. */ smp_mb__after_atomic(); - got_posted_interrupt = - kvm_apic_update_irr(vcpu, vt->pi_desc.pir, &max_irr); + max_irr_is_from_pir = kvm_apic_update_irr(vcpu, vt->pi_desc.pir, + &max_irr); } else { max_irr = kvm_lapic_find_highest_irr(vcpu); - got_posted_interrupt = false; + max_irr_is_from_pir = false; } /* - * Newly recognized interrupts are injected via either virtual interrupt - * delivery (RVI) or KVM_REQ_EVENT. Virtual interrupt delivery is - * disabled in two cases: + * If APICv is enabled and L2 is not active, then update the Requesting + * Virtual Interrupt (RVI) portion of vmcs01.GUEST_INTR_STATUS with the + * highest priority IRR to deliver the IRQ via Virtual Interrupt + * Delivery. Note, this is required even if the highest priority IRQ + * was already pending in the IRR, as RVI isn't updated in lockstep with + * the IRR (unlike apic->irr_pending). + * + * For the cases where Virtual Interrupt Delivery can't be used: * * 1) If L2 is running and the vCPU has a new pending interrupt. If L1 * wants to exit on interrupts, KVM_REQ_EVENT is needed to synthesize a @@ -7063,10 +7068,29 @@ int vmx_sync_pir_to_irr(struct kvm_vcpu *vcpu) * 2) If APICv is disabled for this vCPU, assigned devices may still * attempt to post interrupts. The posted interrupt vector will cause * a VM-Exit and the subsequent entry will call sync_pir_to_irr. + * + * In both cases, set KVM_REQ_EVENT if and only if the highest priority + * pending IRQ came from the PIR, as setting KVM_REQ_EVENT if any IRQ + * is pending may put the vCPU into an infinite loop, e.g. if the IRQ + * is blocked, then it will stay pending until an IRQ window is opened. + * + * Note! It's possible that one or more IRQs were moved from the PIR + * to the IRR _without_ max_irr_is_from_pir being true! I.e. if there + * was a higher priority IRQ already pending in the IRR. Not setting + * KVM_REQ_EVENT in this case is intentional and safe. If APICv is + * inactive, or L2 is running with exit-on-interrupt off (in vmcs12), + * i.e. without nested virtual interrupt delivery, then there's no need + * to request an IRQ window as the lower priority IRQ only needs to be + * delivered when the higher priority IRQ is dismissed from the ISR, + * i.e. on the next EOI, and EOIs are always intercepted if APICv is + * disabled or if L2 is running without nested VID. If L2 is running + * exit-on-interrupt on (in vmcs12), then the higher priority IRQ will + * trigger a nested VM-Exit, at which point KVM will re-evaluate L1's + * pending IRQs. */ if (!is_guest_mode(vcpu) && kvm_vcpu_apicv_active(vcpu)) vmx_set_rvi(max_irr); - else if (got_posted_interrupt) + else if (max_irr_is_from_pir) kvm_make_request(KVM_REQ_EVENT, vcpu); return max_irr; From 0cb2af2ea66ad8ff195c156ea690f11216285bdf Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Wed, 15 Apr 2026 00:52:34 +0000 Subject: [PATCH 207/525] KVM: x86: Fix shadow paging use-after-free due to unexpected GFN The shadow MMU computes GFNs for direct shadow pages using sp->gfn plus the SPTE index. This assumption breaks for shadow paging if the guest page tables are modified between VM entries (similar to commit aad885e77496, "KVM: x86/mmu: Drop/zap existing present SPTE even when creating an MMIO SPTE", 2026-03-27). The flow is as follows: - a PDE is installed for a 2MB mapping, and a page in that area is accessed. KVM creates a kvm_mmu_page consisting of 512 4KB pages; the kvm_mmu_page is marked by FNAME(fetch) as direct-mapped because the guest's mapping is a huge page (and thus contiguous). - the PDE mapping is changed from outside the guest. - the guest accesses another page in the same 2MB area. KVM installs a new leaf SPTE and rmap entry; the SPTE uses the "correct" GFN (i.e. based on the new mapping, as changed in the previous step) but that GFN is outside of the [sp->gfn, sp->gfn + 511] range; therefore the rmap entry cannot be found and removed when the kvm_mmu_page is zapped. - the memslot that covers the first 2MB mapping is deleted, and the kvm_mmu_page for the now-invalid GPA is zapped. However, rmap_remove() only looks at the [sp->gfn, sp->gfn + 511] range established in step 1, and fails to find the rmap entry that was recorded by step 3. - any operation that causes an rmap walk for the same page accessed by step 3 then walks a stale rmap and dereferences a freed kvm_mmu_page. This includes dirty logging or MMU notifier invalidations (e.g., from MADV_DONTNEED). The underlying issue is that KVM's walking of shadow PTEs assumes that if a SPTE is present when KVM wants to install a non-leaf SPTE, then the existing kvm_mmu_page must be for the correct gfn. Because the only way for the gfn to be wrong is if KVM messed up and failed to zap a SPTE... which shouldn't happen, but *actually* only happens in response to a guest write. That bug dates back literally forever, as even the first version of KVM assumes that the GFN matches and walks into the "wrong" shadow page. However, that was only an imprecision until 2032a93d66fa ("KVM: MMU: Don't allocate gfns page for direct mmu pages") came along. Fix it by checking for a target gfn mismatch and zapping the existing SPTE. That way the old SP and rmap entries are gone, KVM installs the rmap in the right location, and everyone is happy. Fixes: 2032a93d66fa ("KVM: MMU: Don't allocate gfns page for direct mmu pages") Fixes: 6aa8b732ca01 ("kvm: userspace interface") Reported-by: Alexander Bulekov Reported-by: Fred Griffoul Cc: stable@vger.kernel.org Signed-off-by: Sean Christopherson Link: https://patch.msgid.link/20260503201029.106481-1-pbonzini@redhat.com/ Signed-off-by: Paolo Bonzini --- arch/x86/kvm/mmu/mmu.c | 35 ++++++++++++++--------------------- 1 file changed, 14 insertions(+), 21 deletions(-) diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c index 24fbc9ea502a..892246204435 100644 --- a/arch/x86/kvm/mmu/mmu.c +++ b/arch/x86/kvm/mmu/mmu.c @@ -182,6 +182,8 @@ static struct kmem_cache *pte_list_desc_cache; struct kmem_cache *mmu_page_header_cache; static void mmu_spte_set(u64 *sptep, u64 spte); +static int mmu_page_zap_pte(struct kvm *kvm, struct kvm_mmu_page *sp, + u64 *spte, struct list_head *invalid_list); struct kvm_mmu_role_regs { const unsigned long cr0; @@ -1287,19 +1289,6 @@ static void drop_spte(struct kvm *kvm, u64 *sptep) rmap_remove(kvm, sptep); } -static void drop_large_spte(struct kvm *kvm, u64 *sptep, bool flush) -{ - struct kvm_mmu_page *sp; - - sp = sptep_to_sp(sptep); - WARN_ON_ONCE(sp->role.level == PG_LEVEL_4K); - - drop_spte(kvm, sptep); - - if (flush) - kvm_flush_remote_tlbs_sptep(kvm, sptep); -} - /* * Write-protect on the specified @sptep, @pt_protect indicates whether * spte write-protection is caused by protecting shadow page table. @@ -2466,7 +2455,8 @@ static struct kvm_mmu_page *kvm_mmu_get_child_sp(struct kvm_vcpu *vcpu, { union kvm_mmu_page_role role; - if (is_shadow_present_pte(*sptep) && !is_large_pte(*sptep)) + if (is_shadow_present_pte(*sptep) && !is_large_pte(*sptep) && + spte_to_child_sp(*sptep) && spte_to_child_sp(*sptep)->gfn == gfn) return ERR_PTR(-EEXIST); role = kvm_mmu_child_role(sptep, direct, access); @@ -2544,13 +2534,16 @@ static void __link_shadow_page(struct kvm *kvm, BUILD_BUG_ON(VMX_EPT_WRITABLE_MASK != PT_WRITABLE_MASK); - /* - * If an SPTE is present already, it must be a leaf and therefore - * a large one. Drop it, and flush the TLB if needed, before - * installing sp. - */ - if (is_shadow_present_pte(*sptep)) - drop_large_spte(kvm, sptep, flush); + if (is_shadow_present_pte(*sptep)) { + struct kvm_mmu_page *parent_sp; + LIST_HEAD(invalid_list); + + parent_sp = sptep_to_sp(sptep); + WARN_ON_ONCE(parent_sp->role.level == PG_LEVEL_4K); + + mmu_page_zap_pte(kvm, parent_sp, sptep, &invalid_list); + kvm_mmu_remote_flush_or_zap(kvm, &invalid_list, true); + } spte = make_nonleaf_spte(sp->spt, sp_ad_disabled(sp)); From 4808e5cc4fe1f40c4d690d0ce5df0f413fb02938 Mon Sep 17 00:00:00 2001 From: Huacai Chen Date: Mon, 4 May 2026 09:00:00 +0800 Subject: [PATCH 208/525] LoongArch: Make CONFIG_64BIT as the default option CONFIG_64BIT is the mandatory option before v7.0, but in v7.1-rc1 both CONFIG_32BIT and CONFIG_64BIT are selectable and CONFIG_32BIT became the default option. This breaks existing configurations, so explicitly make CONFIG_64BIT as the default option to keep existing behavior. Signed-off-by: Huacai Chen --- arch/loongarch/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/loongarch/Kconfig b/arch/loongarch/Kconfig index 3b042dbb2c41..606597da46b8 100644 --- a/arch/loongarch/Kconfig +++ b/arch/loongarch/Kconfig @@ -220,6 +220,7 @@ menu "Kernel type and options" choice prompt "Kernel type" + default 64BIT # Keep existing behavior config 32BIT bool "32-bit kernel" From 5643c6b2c8308b206cb01cbfd0e6ac80f9f1bc9a Mon Sep 17 00:00:00 2001 From: Huacai Chen Date: Mon, 4 May 2026 09:00:01 +0800 Subject: [PATCH 209/525] LoongArch: Specify -m32/-m64 explicitly for 32BIT/64BIT Clang/LLVM build needs -m32/-m64 to switch triple variants (i.e. the --target=xxx parameter). Otherwise we get build errors for CONFIG_32BIT. GCC doesn't support -m32/-m64 now, but maybe support in future, so use cc-option to specify them. Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202604232041.ESJDwVG4-lkp@intel.com/ Suggested-by: Nathan Chancellor Signed-off-by: Huacai Chen --- arch/loongarch/Makefile | 2 ++ arch/loongarch/vdso/Makefile | 2 ++ 2 files changed, 4 insertions(+) diff --git a/arch/loongarch/Makefile b/arch/loongarch/Makefile index 47516aeea9d2..54fcfa1eac1f 100644 --- a/arch/loongarch/Makefile +++ b/arch/loongarch/Makefile @@ -55,9 +55,11 @@ endif ifdef CONFIG_32BIT tool-archpref = $(32bit-tool-archpref) UTS_MACHINE := loongarch32 +cflags-y += $(call cc-option,-m32) else tool-archpref = $(64bit-tool-archpref) UTS_MACHINE := loongarch64 +cflags-y += $(call cc-option,-m64) endif ifneq ($(SUBARCH),$(ARCH)) diff --git a/arch/loongarch/vdso/Makefile b/arch/loongarch/vdso/Makefile index 42aa96249828..9c9181bb4071 100644 --- a/arch/loongarch/vdso/Makefile +++ b/arch/loongarch/vdso/Makefile @@ -12,6 +12,8 @@ obj-vdso-$(CONFIG_GENERIC_GETTIMEOFDAY) += vgettimeofday.o ccflags-vdso := \ $(filter -I%,$(KBUILD_CFLAGS)) \ $(filter -E%,$(KBUILD_CFLAGS)) \ + $(filter -m32,$(KBUILD_CFLAGS)) \ + $(filter -m64,$(KBUILD_CFLAGS)) \ $(filter -march=%,$(KBUILD_CFLAGS)) \ $(filter -m%-float,$(KBUILD_CFLAGS)) \ $(CLANG_FLAGS) \ From 98b8aebb14fdc0133939fd8fe07d0d98333dc976 Mon Sep 17 00:00:00 2001 From: Huacai Chen Date: Mon, 4 May 2026 09:00:01 +0800 Subject: [PATCH 210/525] LoongArch: Fix SYM_SIGFUNC_START definition for 32BIT The SYM_SIGFUNC_START definition should match sigcontext that the length of GPRs are 8 bytes for both 32BIT and 64BIT. So replace SZREG with 8 to fix it. Cc: stable@vger.kernel.org Fixes: e4878c37f6679fde ("LoongArch: vDSO: Emit GNU_EH_FRAME correctly") Suggested-by: Xi Ruoyao Signed-off-by: Huacai Chen --- arch/loongarch/include/asm/linkage.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/loongarch/include/asm/linkage.h b/arch/loongarch/include/asm/linkage.h index a1bd6a3ee03a..ae937d1708b2 100644 --- a/arch/loongarch/include/asm/linkage.h +++ b/arch/loongarch/include/asm/linkage.h @@ -69,7 +69,7 @@ 9, 10, 11, 12, 13, 14, 15, 16, \ 17, 18, 19, 20, 21, 22, 23, 24, \ 25, 26, 27, 28, 29, 30, 31; \ - .cfi_offset \num, SC_REGS + \num * SZREG; \ + .cfi_offset \num, SC_REGS + \num * 8; \ .endr; \ \ nop; \ From 49f33840dcc907d21313d369e34872880846b61c Mon Sep 17 00:00:00 2001 From: Huacai Chen Date: Mon, 4 May 2026 09:00:20 +0800 Subject: [PATCH 211/525] LoongArch: Use per-root-bridge PCIH flag to skip mem resource fixup When firmware enables 64-bit PCI host bridge support, some root bridges already provide valid 64-bit mem resource windows through ACPI. In this case, the LoongArch-specific mem resource high-bits fixup in acpi_prepare_root_resources() should not be applied unconditionally. Otherwise, the kernel may override the native resource layout derived from firmware, and later BAR assignment can fail to place device BARs into the intended 64-bit address space correctly. Add a per-root-bridge ACPI flag, PCIH, and evaluate it from the current root bridge device scope. When PCIH is set, skip the mem resource high- bits fixup path and let the kernel use the firmware-provided resource description directly. When PCIH is absent or cleared, keep the existing behavior and continue filling the high address bits from the host bridge address. This makes the behavior per-root-bridge configurable and avoids breaking valid 64-bit BAR space allocation on bridges whose 64-bit windows have already been fully described by firmware. Cc: stable@vger.kernel.org Suggested-by: Chao Li Tested-by: Dongyan Qian Signed-off-by: Dongyan Qian Signed-off-by: Huacai Chen --- arch/loongarch/pci/acpi.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/arch/loongarch/pci/acpi.c b/arch/loongarch/pci/acpi.c index 0dde3ddcd544..b02698a338ee 100644 --- a/arch/loongarch/pci/acpi.c +++ b/arch/loongarch/pci/acpi.c @@ -61,11 +61,16 @@ static void acpi_release_root_info(struct acpi_pci_root_info *ci) static int acpi_prepare_root_resources(struct acpi_pci_root_info *ci) { int status; + unsigned long long pci_h = 0; struct resource_entry *entry, *tmp; struct acpi_device *device = ci->bridge; status = acpi_pci_probe_root_resources(ci); if (status > 0) { + acpi_evaluate_integer(device->handle, "PCIH", NULL, &pci_h); + if (pci_h) + return status; + resource_list_for_each_entry_safe(entry, tmp, &ci->resources) { if (entry->res->flags & IORESOURCE_MEM) { entry->offset = ci->root->mcfg_addr & GENMASK_ULL(63, 40); From 8dfa2f8780e486d05b9a0ffce70b8f5fbd62053e Mon Sep 17 00:00:00 2001 From: Wentao Guan Date: Mon, 4 May 2026 09:00:20 +0800 Subject: [PATCH 212/525] LoongArch: Fix potential ADE in loongson_gpu_fixup_dma_hang() The switch case in loongson_gpu_fixup_dma_hang() may not DC2 or DC3, and readl(crtc_reg) will access with random address, because the "device" is from "base+PCI_DEVICE_ID", "base" is from "pdev->devfn+1". This is wrong when my platform inserts a discrete GPU: lspci -tv -[0000:00]-+-00.0 Loongson Technology LLC Hyper Transport Bridge Controller ... +-06.0 Loongson Technology LLC LG100 GPU +-06.2 Loongson Technology LLC Device 7a37 ... Add a default switch case to fix the panic as below: Kernel ade access[#1]: CPU: 0 PID: 1 Comm: swapper/0 Not tainted 6.6.136-loong64-desktop-hwe+ #4 pc 90000000017e5534 ra 90000000017e54c0 tp 90000001002f8000 sp 90000001002fb6c0 a0 80000efe00003100 a1 0000000000003100 a2 0000000000000000 a3 0000000000000002 a4 90000001002fb6b4 a5 900000087cdb58fd a6 90000000027af000 a7 0000000000000001 t0 00000000000085b9 t1 000000000000ffff t2 0000000000000000 t3 0000000000000000 t4 fffffffffffffffd t5 00000000fffb6d9c t6 0000000000083b00 t7 00000000000070c0 t8 900000087cdb4d94 u0 900000087cdb58fd s9 90000001002fb826 s0 90000000031c12c8 s1 7fffffffffffff00 s2 90000000031c12d0 s3 0000000000002710 s4 0000000000000000 s5 0000000000000000 s6 9000000100053000 s7 7fffffffffffff00 s8 90000000030d4000 ra: 90000000017e54c0 loongson_gpu_fixup_dma_hang+0x40/0x210 ERA: 90000000017e5534 loongson_gpu_fixup_dma_hang+0xb4/0x210 CRMD: 000000b0 (PLV0 -IE -DA +PG DACF=CC DACM=CC -WE) PRMD: 00000004 (PPLV0 +PIE -PWE) EUEN: 00000000 (-FPE -SXE -ASXE -BTE) ECFG: 00071c1d (LIE=0,2-4,10-12 VS=7) ESTAT: 00480000 [ADEM] (IS= ECode=8 EsubCode=1) BADV: 7fffffffffffff00 PRID: 0014d000 (Loongson-64bit, Loongson-3A6000-HV) Modules linked in: Process swapper/0 (pid: 1, threadinfo=(____ptrval____), task=(____ptrval____)) Stack : 0000000000000006 90000001002fb778 90000001002fb704 0000000000000007 0000000016a65700 90000000017e5690 000000000000ffff ffffffffffffffff 900000000209f7c0 9000000100053000 900000000209f7a8 9000000000eebc08 0000000000000000 0000000000000000 0000000000000006 90000001002fb778 90000001000530b8 90000000027af000 0000000000000000 9000000100054000 9000000100053000 9000000000ebb70c 9000000100004c00 9000000004000001 90000001002fb7e4 bae765461f31cb12 0000000000000000 0000000000000000 0000000000000006 90000000027af000 0000000000000030 90000000027af000 900000087cd6f800 9000000100053000 0000000000000000 9000000000ebc560 7a2500147cdaf720 bae765461f31cb12 0000000000000001 0000000000000030 ... Call Trace: [<90000000017e5534>] loongson_gpu_fixup_dma_hang+0xb4/0x210 [<9000000000eebc08>] pci_fixup_device+0x108/0x280 [<9000000000ebb70c>] pci_setup_device+0x24c/0x690 [<9000000000ebc560>] pci_scan_single_device+0xe0/0x140 [<9000000000ebc684>] pci_scan_slot+0xc4/0x280 [<9000000000ebdd00>] pci_scan_child_bus_extend+0x60/0x3f0 [<9000000000f5bc94>] acpi_pci_root_create+0x2b4/0x420 [<90000000017e5e74>] pci_acpi_scan_root+0x2d4/0x440 [<9000000000f5b02c>] acpi_pci_root_add+0x21c/0x3a0 [<9000000000f4ee54>] acpi_bus_attach+0x1a4/0x3c0 [<90000000010e200c>] device_for_each_child+0x6c/0xe0 [<9000000000f4bbf4>] acpi_dev_for_each_child+0x44/0x70 [<9000000000f4ef40>] acpi_bus_attach+0x290/0x3c0 [<90000000010e200c>] device_for_each_child+0x6c/0xe0 [<9000000000f4bbf4>] acpi_dev_for_each_child+0x44/0x70 [<9000000000f4ef40>] acpi_bus_attach+0x290/0x3c0 [<9000000000f5211c>] acpi_bus_scan+0x6c/0x280 [<900000000189c028>] acpi_scan_init+0x194/0x310 [<900000000189bc6c>] acpi_init+0xcc/0x140 [<9000000000220cdc>] do_one_initcall+0x4c/0x310 [<90000000018618fc>] kernel_init_freeable+0x258/0x2d4 [<900000000184326c>] kernel_init+0x28/0x13c [<9000000000222008>] ret_from_kernel_thread+0xc/0xa4 Cc: stable@vger.kernel.org Fixes: 95db0c9f526d ("LoongArch: Workaround LS2K/LS7A GPU DMA hang bug") Link: https://gist.github.com/opsiff/ebf2dac51b4013d22462f2124c55f807 Link: https://gist.github.com/opsiff/a62f2a73db0492b3c49bf223a339b133 Signed-off-by: Wentao Guan Signed-off-by: Huacai Chen --- arch/loongarch/pci/pci.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/loongarch/pci/pci.c b/arch/loongarch/pci/pci.c index d233ea2218fe..f33c7ea1443d 100644 --- a/arch/loongarch/pci/pci.c +++ b/arch/loongarch/pci/pci.c @@ -132,6 +132,9 @@ static void loongson_gpu_fixup_dma_hang(struct pci_dev *pdev, bool on) crtc_reg = regbase; crtc_offset = 0x400; break; + default: + iounmap(regbase); + return; } for (i = 0; i < CRTC_NUM_MAX; i++, crtc_reg += crtc_offset) { From 7e2c41bc62e436f465ee1ff7ebc14e35c99d95fb Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Thomas=20Wei=C3=9Fschuh?= Date: Mon, 4 May 2026 09:00:20 +0800 Subject: [PATCH 213/525] LoongArch: vDSO: Drop custom __arch_vdso_hres_capable() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The custom definition is identical to the generic fallback one. So remove it. Signed-off-by: Thomas Weißschuh Signed-off-by: Huacai Chen --- arch/loongarch/include/asm/vdso/gettimeofday.h | 6 ------ 1 file changed, 6 deletions(-) diff --git a/arch/loongarch/include/asm/vdso/gettimeofday.h b/arch/loongarch/include/asm/vdso/gettimeofday.h index bae76767c693..18ba403e1ed9 100644 --- a/arch/loongarch/include/asm/vdso/gettimeofday.h +++ b/arch/loongarch/include/asm/vdso/gettimeofday.h @@ -85,12 +85,6 @@ static __always_inline u64 __arch_get_hw_counter(s32 clock_mode, return count; } -static inline bool loongarch_vdso_hres_capable(void) -{ - return true; -} -#define __arch_vdso_hres_capable loongarch_vdso_hres_capable - #endif /* CONFIG_GENERIC_GETTIMEOFDAY */ #endif /* !__ASSEMBLER__ */ From 5203012fa6045aac4b69d4e7c212e16dcf38ef10 Mon Sep 17 00:00:00 2001 From: Xianglai Li Date: Mon, 4 May 2026 09:00:37 +0800 Subject: [PATCH 214/525] LoongArch: KVM: Compile switch.S directly into the kernel If we directly compile the switch.S file into the kernel, the address of the kvm_exc_entry function will definitely be within the DMW memory area. Therefore, we will no longer need to perform a copy relocation of the kvm_exc_entry. So this patch compiles switch.S directly into the kernel, and then remove the copy relocation execution logic for the kvm_exc_entry function. Cc: stable@vger.kernel.org Signed-off-by: Xianglai Li Signed-off-by: Huacai Chen --- arch/loongarch/Kbuild | 2 +- arch/loongarch/include/asm/asm-prototypes.h | 20 ++++++++++++ arch/loongarch/include/asm/kvm_host.h | 3 -- arch/loongarch/kvm/Makefile | 3 +- arch/loongarch/kvm/main.c | 35 ++------------------- arch/loongarch/kvm/switch.S | 20 +++++++++--- 6 files changed, 41 insertions(+), 42 deletions(-) diff --git a/arch/loongarch/Kbuild b/arch/loongarch/Kbuild index beb8499dd8ed..1c7a0dbe5e72 100644 --- a/arch/loongarch/Kbuild +++ b/arch/loongarch/Kbuild @@ -3,7 +3,7 @@ obj-y += mm/ obj-y += net/ obj-y += vdso/ -obj-$(CONFIG_KVM) += kvm/ +obj-$(subst m,y,$(CONFIG_KVM)) += kvm/ # for cleaning subdir- += boot diff --git a/arch/loongarch/include/asm/asm-prototypes.h b/arch/loongarch/include/asm/asm-prototypes.h index 704066b4f736..de0c17f3f49c 100644 --- a/arch/loongarch/include/asm/asm-prototypes.h +++ b/arch/loongarch/include/asm/asm-prototypes.h @@ -20,3 +20,23 @@ asmlinkage void noinstr __no_stack_protector ret_from_kernel_thread(struct task_ struct pt_regs *regs, int (*fn)(void *), void *fn_arg); + +struct kvm_run; +struct kvm_vcpu; +struct loongarch_fpu; + +void kvm_exc_entry(void); +int kvm_enter_guest(struct kvm_run *run, struct kvm_vcpu *vcpu); + +void kvm_save_fpu(struct loongarch_fpu *fpu); +void kvm_restore_fpu(struct loongarch_fpu *fpu); + +#ifdef CONFIG_CPU_HAS_LSX +void kvm_save_lsx(struct loongarch_fpu *fpu); +void kvm_restore_lsx(struct loongarch_fpu *fpu); +#endif + +#ifdef CONFIG_CPU_HAS_LASX +void kvm_save_lasx(struct loongarch_fpu *fpu); +void kvm_restore_lasx(struct loongarch_fpu *fpu); +#endif diff --git a/arch/loongarch/include/asm/kvm_host.h b/arch/loongarch/include/asm/kvm_host.h index 130cedbb6b39..776bc487a705 100644 --- a/arch/loongarch/include/asm/kvm_host.h +++ b/arch/loongarch/include/asm/kvm_host.h @@ -87,7 +87,6 @@ struct kvm_context { struct kvm_world_switch { int (*exc_entry)(void); int (*enter_guest)(struct kvm_run *run, struct kvm_vcpu *vcpu); - unsigned long page_order; }; #define MAX_PGTABLE_LEVELS 4 @@ -359,8 +358,6 @@ void kvm_exc_entry(void); int kvm_enter_guest(struct kvm_run *run, struct kvm_vcpu *vcpu); extern unsigned long vpid_mask; -extern const unsigned long kvm_exception_size; -extern const unsigned long kvm_enter_guest_size; extern struct kvm_world_switch *kvm_loongarch_ops; #define SW_GCSR (1 << 0) diff --git a/arch/loongarch/kvm/Makefile b/arch/loongarch/kvm/Makefile index ae469edec99c..a4d044da3aa7 100644 --- a/arch/loongarch/kvm/Makefile +++ b/arch/loongarch/kvm/Makefile @@ -7,11 +7,12 @@ include $(srctree)/virt/kvm/Makefile.kvm obj-$(CONFIG_KVM) += kvm.o +obj-y += switch.o + kvm-y += exit.o kvm-y += interrupt.o kvm-y += main.o kvm-y += mmu.o -kvm-y += switch.o kvm-y += timer.o kvm-y += tlb.o kvm-y += vcpu.o diff --git a/arch/loongarch/kvm/main.c b/arch/loongarch/kvm/main.c index 76ebff2faedd..f105a86143f5 100644 --- a/arch/loongarch/kvm/main.c +++ b/arch/loongarch/kvm/main.c @@ -348,8 +348,7 @@ void kvm_arch_disable_virtualization_cpu(void) static int kvm_loongarch_env_init(void) { - int cpu, order, ret; - void *addr; + int cpu, ret; struct kvm_context *context; vmcs = alloc_percpu(struct kvm_context); @@ -365,30 +364,8 @@ static int kvm_loongarch_env_init(void) return -ENOMEM; } - /* - * PGD register is shared between root kernel and kvm hypervisor. - * So world switch entry should be in DMW area rather than TLB area - * to avoid page fault reenter. - * - * In future if hardware pagetable walking is supported, we won't - * need to copy world switch code to DMW area. - */ - order = get_order(kvm_exception_size + kvm_enter_guest_size); - addr = (void *)__get_free_pages(GFP_KERNEL, order); - if (!addr) { - free_percpu(vmcs); - vmcs = NULL; - kfree(kvm_loongarch_ops); - kvm_loongarch_ops = NULL; - return -ENOMEM; - } - - memcpy(addr, kvm_exc_entry, kvm_exception_size); - memcpy(addr + kvm_exception_size, kvm_enter_guest, kvm_enter_guest_size); - flush_icache_range((unsigned long)addr, (unsigned long)addr + kvm_exception_size + kvm_enter_guest_size); - kvm_loongarch_ops->exc_entry = addr; - kvm_loongarch_ops->enter_guest = addr + kvm_exception_size; - kvm_loongarch_ops->page_order = order; + kvm_loongarch_ops->exc_entry = (void *)kvm_exc_entry; + kvm_loongarch_ops->enter_guest = (void *)kvm_enter_guest; vpid_mask = read_csr_gstat(); vpid_mask = (vpid_mask & CSR_GSTAT_GIDBIT) >> CSR_GSTAT_GIDBIT_SHIFT; @@ -428,16 +405,10 @@ static int kvm_loongarch_env_init(void) static void kvm_loongarch_env_exit(void) { - unsigned long addr; - if (vmcs) free_percpu(vmcs); if (kvm_loongarch_ops) { - if (kvm_loongarch_ops->exc_entry) { - addr = (unsigned long)kvm_loongarch_ops->exc_entry; - free_pages(addr, kvm_loongarch_ops->page_order); - } kfree(kvm_loongarch_ops); } diff --git a/arch/loongarch/kvm/switch.S b/arch/loongarch/kvm/switch.S index f1768b7a6194..1d3ba7190154 100644 --- a/arch/loongarch/kvm/switch.S +++ b/arch/loongarch/kvm/switch.S @@ -4,9 +4,11 @@ */ #include +#include #include #include #include +#include #include #include @@ -100,8 +102,13 @@ * - is still in guest mode, such as pgd table/vmid registers etc, * - will fix with hw page walk enabled in future * load kvm_vcpu from reserved CSR KVM_VCPU_KS, and save a2 to KVM_TEMP_KS + * + * PGD register is shared between root kernel and kvm hypervisor. + * So world switch entry should be in DMW area rather than TLB area + * to avoid page fault re-enter. */ .text + .p2align PAGE_SHIFT .cfi_sections .debug_frame SYM_CODE_START(kvm_exc_entry) UNWIND_HINT_UNDEFINED @@ -190,8 +197,8 @@ ret_to_host: kvm_restore_host_gpr a2 jr ra -SYM_INNER_LABEL(kvm_exc_entry_end, SYM_L_LOCAL) SYM_CODE_END(kvm_exc_entry) +EXPORT_SYMBOL_FOR_KVM(kvm_exc_entry) /* * int kvm_enter_guest(struct kvm_run *run, struct kvm_vcpu *vcpu) @@ -215,8 +222,8 @@ SYM_FUNC_START(kvm_enter_guest) /* Save kvm_vcpu to kscratch */ csrwr a1, KVM_VCPU_KS kvm_switch_to_guest -SYM_INNER_LABEL(kvm_enter_guest_end, SYM_L_LOCAL) SYM_FUNC_END(kvm_enter_guest) +EXPORT_SYMBOL_FOR_KVM(kvm_enter_guest) SYM_FUNC_START(kvm_save_fpu) fpu_save_csr a0 t1 @@ -224,6 +231,7 @@ SYM_FUNC_START(kvm_save_fpu) fpu_save_cc a0 t1 t2 jr ra SYM_FUNC_END(kvm_save_fpu) +EXPORT_SYMBOL_FOR_KVM(kvm_save_fpu) SYM_FUNC_START(kvm_restore_fpu) fpu_restore_double a0 t1 @@ -231,6 +239,7 @@ SYM_FUNC_START(kvm_restore_fpu) fpu_restore_cc a0 t1 t2 jr ra SYM_FUNC_END(kvm_restore_fpu) +EXPORT_SYMBOL_FOR_KVM(kvm_restore_fpu) #ifdef CONFIG_CPU_HAS_LSX SYM_FUNC_START(kvm_save_lsx) @@ -239,6 +248,7 @@ SYM_FUNC_START(kvm_save_lsx) lsx_save_data a0 t1 jr ra SYM_FUNC_END(kvm_save_lsx) +EXPORT_SYMBOL_FOR_KVM(kvm_save_lsx) SYM_FUNC_START(kvm_restore_lsx) lsx_restore_data a0 t1 @@ -246,6 +256,7 @@ SYM_FUNC_START(kvm_restore_lsx) fpu_restore_csr a0 t1 t2 jr ra SYM_FUNC_END(kvm_restore_lsx) +EXPORT_SYMBOL_FOR_KVM(kvm_restore_lsx) #endif #ifdef CONFIG_CPU_HAS_LASX @@ -255,6 +266,7 @@ SYM_FUNC_START(kvm_save_lasx) lasx_save_data a0 t1 jr ra SYM_FUNC_END(kvm_save_lasx) +EXPORT_SYMBOL_FOR_KVM(kvm_save_lasx) SYM_FUNC_START(kvm_restore_lasx) lasx_restore_data a0 t1 @@ -262,10 +274,8 @@ SYM_FUNC_START(kvm_restore_lasx) fpu_restore_csr a0 t1 t2 jr ra SYM_FUNC_END(kvm_restore_lasx) +EXPORT_SYMBOL_FOR_KVM(kvm_restore_lasx) #endif - .section ".rodata" -SYM_DATA(kvm_exception_size, .quad kvm_exc_entry_end - kvm_exc_entry) -SYM_DATA(kvm_enter_guest_size, .quad kvm_enter_guest_end - kvm_enter_guest) #ifdef CONFIG_CPU_HAS_LBT STACK_FRAME_NON_STANDARD kvm_restore_fpu From b323a441da602dfdfc24f30d3190cac786ffebf2 Mon Sep 17 00:00:00 2001 From: Xianglai Li Date: Mon, 4 May 2026 09:00:37 +0800 Subject: [PATCH 215/525] LoongArch: KVM: Fix "unreliable stack" for kvm_exc_entry Insert the appropriate UNWIND hint into the kvm_exc_entry assembly function to guide the generation of correct ORC table entries, thereby solving the timeout problem ("unreliable stack") while loading the livepatch-sample module on a physical machine running virtual machines with multiple vcpus. Cc: stable@vger.kernel.org Signed-off-by: Xianglai Li Signed-off-by: Huacai Chen --- arch/loongarch/kvm/switch.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/loongarch/kvm/switch.S b/arch/loongarch/kvm/switch.S index 1d3ba7190154..936e4ae3e408 100644 --- a/arch/loongarch/kvm/switch.S +++ b/arch/loongarch/kvm/switch.S @@ -111,7 +111,7 @@ .p2align PAGE_SHIFT .cfi_sections .debug_frame SYM_CODE_START(kvm_exc_entry) - UNWIND_HINT_UNDEFINED + UNWIND_HINT_END_OF_STACK csrwr a2, KVM_TEMP_KS csrrd a2, KVM_VCPU_KS addi.d a2, a2, KVM_VCPU_ARCH From b3e31a6650d4cab63f0814c37c0b360372c6ee9e Mon Sep 17 00:00:00 2001 From: Qiang Ma Date: Mon, 4 May 2026 09:00:37 +0800 Subject: [PATCH 216/525] LoongArch: KVM: Cap KVM_CAP_NR_VCPUS by KVM_CAP_MAX_VCPUS It doesn't make sense to return the recommended maximum number of vCPUs which exceeds the maximum possible number of vCPUs. Other architectures have already done this, such as commit 57a2e13ebdda ("KVM: MIPS: Cap KVM_CAP_NR_VCPUS by KVM_CAP_MAX_VCPUS") Cc: stable@vger.kernel.org Reviewed-by: Bibo Mao Signed-off-by: Qiang Ma Signed-off-by: Huacai Chen --- arch/loongarch/kvm/vm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/loongarch/kvm/vm.c b/arch/loongarch/kvm/vm.c index 8cc5ee1c53ef..1317c718f896 100644 --- a/arch/loongarch/kvm/vm.c +++ b/arch/loongarch/kvm/vm.c @@ -125,7 +125,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) r = 1; break; case KVM_CAP_NR_VCPUS: - r = num_online_cpus(); + r = min_t(unsigned int, num_online_cpus(), KVM_MAX_VCPUS); break; case KVM_CAP_MAX_VCPUS: r = KVM_MAX_VCPUS; From f26faae96c411a70641e4d21b759475caa6122d5 Mon Sep 17 00:00:00 2001 From: Tao Cui Date: Mon, 4 May 2026 09:00:38 +0800 Subject: [PATCH 217/525] LoongArch: KVM: Fix missing EMULATE_FAIL in kvm_emu_mmio_read() In the ldptr (0x24...0x27) opcode decoding path, the default case only breaks out but without setting "ret" value to EMULATE_FAIL. This leaves run->mmio.len uninitialized (stale from a previous MMIO operation) while "ret" value remains EMULATE_DO_MMIO, causing the code to proceed with an incorrect MMIO length. Add "ret = EMULATE_FAIL" to match the other default branches in the same function (e.g. the 0x28...0x2e and 0x38 cases). Cc: stable@vger.kernel.org Reviewed-by: Bibo Mao Signed-off-by: Tao Cui Signed-off-by: Huacai Chen --- arch/loongarch/kvm/exit.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/loongarch/kvm/exit.c b/arch/loongarch/kvm/exit.c index da0ad89f2eb7..3b95cd0f989b 100644 --- a/arch/loongarch/kvm/exit.c +++ b/arch/loongarch/kvm/exit.c @@ -390,6 +390,7 @@ int kvm_emu_mmio_read(struct kvm_vcpu *vcpu, larch_inst inst) run->mmio.len = 8; break; default: + ret = EMULATE_FAIL; break; } break; From 81e18777d61440511451866c7c80b34a8bdd6b33 Mon Sep 17 00:00:00 2001 From: Tao Cui Date: Mon, 4 May 2026 09:00:38 +0800 Subject: [PATCH 218/525] LoongArch: KVM: Use kvm_set_pte() in kvm_flush_pte() kvm_flush_pte() is the only caller that directly assigns *pte instead of using the kvm_set_pte() wrapper. Use the wrapper for consistency with the rest of the file. No functional change intended. Cc: stable@vger.kernel.org Reviewed-by: Bibo Mao Signed-off-by: Tao Cui Signed-off-by: Huacai Chen --- arch/loongarch/kvm/mmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/loongarch/kvm/mmu.c b/arch/loongarch/kvm/mmu.c index a7fa458e3360..e104897aa532 100644 --- a/arch/loongarch/kvm/mmu.c +++ b/arch/loongarch/kvm/mmu.c @@ -95,7 +95,7 @@ static int kvm_flush_pte(kvm_pte_t *pte, phys_addr_t addr, kvm_ptw_ctx *ctx) else kvm->stat.pages--; - *pte = ctx->invalid_entry; + kvm_set_pte(pte, ctx->invalid_entry); return 1; } From 6debfff78584f0adedf7355fe5263198a3fc6b19 Mon Sep 17 00:00:00 2001 From: Bibo Mao Date: Mon, 4 May 2026 09:00:48 +0800 Subject: [PATCH 219/525] LoongArch: KVM: Move AVEC interrupt injection into switch loop When AVEC interrupt controller is emulated in user space, AVEC interrupt is injected by software like SIP0/SIP1/TI/IPI interrupts. Here also move the AVEC interrupt injection in switch loop. Cc: stable@vger.kernel.org Signed-off-by: Bibo Mao Signed-off-by: Huacai Chen --- arch/loongarch/kvm/interrupt.c | 20 +++++++++----------- 1 file changed, 9 insertions(+), 11 deletions(-) diff --git a/arch/loongarch/kvm/interrupt.c b/arch/loongarch/kvm/interrupt.c index 32930959f7c2..53dac8ab8fb1 100644 --- a/arch/loongarch/kvm/interrupt.c +++ b/arch/loongarch/kvm/interrupt.c @@ -33,13 +33,12 @@ static int kvm_irq_deliver(struct kvm_vcpu *vcpu, unsigned int priority) if (priority < EXCCODE_INT_NUM) irq = priority_to_irq[priority]; - if (kvm_guest_has_msgint(&vcpu->arch) && (priority == INT_AVEC)) { - dmsintc_inject_irq(vcpu); - set_gcsr_estat(irq); - return 1; - } - switch (priority) { + case INT_AVEC: + if (!kvm_guest_has_msgint(&vcpu->arch)) + break; + dmsintc_inject_irq(vcpu); + fallthrough; case INT_TI: case INT_IPI: case INT_SWI0: @@ -66,12 +65,11 @@ static int kvm_irq_clear(struct kvm_vcpu *vcpu, unsigned int priority) if (priority < EXCCODE_INT_NUM) irq = priority_to_irq[priority]; - if (kvm_guest_has_msgint(&vcpu->arch) && (priority == INT_AVEC)) { - clear_gcsr_estat(irq); - return 1; - } - switch (priority) { + case INT_AVEC: + if (!kvm_guest_has_msgint(&vcpu->arch)) + break; + fallthrough; case INT_TI: case INT_IPI: case INT_SWI0: From 2433f3f5724b3af569d9fb411ba728629524738b Mon Sep 17 00:00:00 2001 From: Bibo Mao Date: Mon, 4 May 2026 09:00:48 +0800 Subject: [PATCH 220/525] LoongArch: KVM: Fix HW timer interrupt lost when inject interrupt by software With passthrough HW timer, timer interrupt is injected by HW. When inject emulated CPU interrupt by software such SIP0/SIP1/IPI, HW timer interrupt may be lost. Here check whether there is timer tick value inversion before and after injecting emulated CPU interrupt by software, timer enabling by reading timer cfg register is skipped. If the timer tick value is detected with changing, then timer should be enabled. And inject a timer interrupt by software if there is. Cc: Fixes: f45ad5b8aa93 ("LoongArch: KVM: Implement vcpu interrupt operations"). Signed-off-by: Bibo Mao Signed-off-by: Huacai Chen --- arch/loongarch/kvm/interrupt.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/arch/loongarch/kvm/interrupt.c b/arch/loongarch/kvm/interrupt.c index 53dac8ab8fb1..a18c60dffbba 100644 --- a/arch/loongarch/kvm/interrupt.c +++ b/arch/loongarch/kvm/interrupt.c @@ -28,6 +28,7 @@ static unsigned int priority_to_irq[EXCCODE_INT_NUM] = { static int kvm_irq_deliver(struct kvm_vcpu *vcpu, unsigned int priority) { unsigned int irq = 0; + unsigned long old, new; clear_bit(priority, &vcpu->arch.irq_pending); if (priority < EXCCODE_INT_NUM) @@ -43,7 +44,13 @@ static int kvm_irq_deliver(struct kvm_vcpu *vcpu, unsigned int priority) case INT_IPI: case INT_SWI0: case INT_SWI1: + old = kvm_read_hw_gcsr(LOONGARCH_CSR_TVAL); set_gcsr_estat(irq); + new = kvm_read_hw_gcsr(LOONGARCH_CSR_TVAL); + + /* Inject TI if TVAL inverted */ + if (new > old) + set_gcsr_estat(CPU_TIMER); break; case INT_HWI0 ... INT_HWI7: @@ -60,6 +67,7 @@ static int kvm_irq_deliver(struct kvm_vcpu *vcpu, unsigned int priority) static int kvm_irq_clear(struct kvm_vcpu *vcpu, unsigned int priority) { unsigned int irq = 0; + unsigned long old, new; clear_bit(priority, &vcpu->arch.irq_clear); if (priority < EXCCODE_INT_NUM) @@ -74,7 +82,13 @@ static int kvm_irq_clear(struct kvm_vcpu *vcpu, unsigned int priority) case INT_IPI: case INT_SWI0: case INT_SWI1: + old = kvm_read_hw_gcsr(LOONGARCH_CSR_TVAL); clear_gcsr_estat(irq); + new = kvm_read_hw_gcsr(LOONGARCH_CSR_TVAL); + + /* Inject TI if TVAL inverted */ + if (new > old) + set_gcsr_estat(CPU_TIMER); break; case INT_HWI0 ... INT_HWI7: From 5a873d77ba792410a796595a917be6a440f9b7d2 Mon Sep 17 00:00:00 2001 From: Bibo Mao Date: Mon, 4 May 2026 09:00:48 +0800 Subject: [PATCH 221/525] LoongArch: KVM: Move unconditional delay into timer clear scenery When timer interrupt arrives in guest kernel, guest kernel clears the timer interrupt and program timer with the next incoming event. During this stage, timer tick is -1 and timer interrupt status is disabled in ESTAT register. KVM hypervisor need write zero with timer tick register and wait timer interrupt injection from HW side, and then clear timer interrupt. So there is 2 cycle delay in KVM hypervisor to emulate such scenery, and the delay is unnecessary if there is no need to clear the timer interrupt. Here move 2 cycle delay into timer clear scenery and add timer ESTAT checking after delay, and set max timer expire value if timer interrupt does not arrive still. Cc: stable@vger.kernel.org Signed-off-by: Bibo Mao Signed-off-by: Huacai Chen --- arch/loongarch/kvm/timer.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/arch/loongarch/kvm/timer.c b/arch/loongarch/kvm/timer.c index 29c2aaba63c3..8356fce0043f 100644 --- a/arch/loongarch/kvm/timer.c +++ b/arch/loongarch/kvm/timer.c @@ -96,15 +96,21 @@ void kvm_restore_timer(struct kvm_vcpu *vcpu) * and set CSR TVAL with -1 */ write_gcsr_timertick(0); - __delay(2); /* Wait cycles until timer interrupt injected */ /* * Writing CSR_TINTCLR_TI to LOONGARCH_CSR_TINTCLR will clear * timer interrupt, and CSR TVAL keeps unchanged with -1, it * avoids spurious timer interrupt */ - if (!(estat & CPU_TIMER)) + if (!(estat & CPU_TIMER)) { + __delay(2); /* Wait cycles until timer interrupt injected */ + + /* Write TVAL with max value if no TI shot */ + estat = kvm_read_hw_gcsr(LOONGARCH_CSR_ESTAT); + if (!(estat & CPU_TIMER)) + write_gcsr_timertick(CSR_TCFG_VAL); gcsr_write(CSR_TINTCLR_TI, LOONGARCH_CSR_TINTCLR); + } return; } From d68ce834f8cf6cb2e77f3331df65166b35466b53 Mon Sep 17 00:00:00 2001 From: Shyam Prasad N Date: Tue, 28 Apr 2026 21:37:47 +0530 Subject: [PATCH 222/525] cifs: abort open_cached_dir if we don't request leases It is possible that SMB2_open_init may not set lease context based on the requested oplock level. This can happen when leases have been temporarily or permanently disabled. When this happens, we will have open_cached_dir making an open without lease context and the response will anyway be rejected by open_cached_dir (thereby forcing a close to discard this open). That's unnecessary two round-trips to the server. This change adds a check before making the open request to the server to make sure that SMB2_open_init did add the expected lease context to the open in open_cached_dir. Cc: Reviewed-by: Bharath SM Signed-off-by: Shyam Prasad N Signed-off-by: Steve French --- fs/smb/client/cached_dir.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/fs/smb/client/cached_dir.c b/fs/smb/client/cached_dir.c index 02791ec3c5a1..88d5e9a32f28 100644 --- a/fs/smb/client/cached_dir.c +++ b/fs/smb/client/cached_dir.c @@ -286,6 +286,14 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, &rqst[0], &oplock, &oparms, utf16_path); if (rc) goto oshr_free; + + if (oplock != SMB2_OPLOCK_LEVEL_II) { + rc = -EINVAL; + cifs_dbg(FYI, "%s: Oplock level %d not suitable for cached directory\n", + __func__, oplock); + goto oshr_free; + } + smb2_set_next_command(tcon, &rqst[0]); memset(&qi_iov, 0, sizeof(qi_iov)); From 5e489c6c47a2ac15edbaca153b9348e42c1eacab Mon Sep 17 00:00:00 2001 From: Bjoern Doebel Date: Thu, 30 Apr 2026 08:57:17 +0000 Subject: [PATCH 223/525] smb: client: use kzalloc to zero-initialize security descriptor buffer Commit 62e7dd0a39c2d ("smb: common: change the data type of num_aces to le16") split struct smb_acl's __le32 num_aces field into __le16 num_aces and __le16 reserved. The reserved field corresponds to Sbz2 in the MS-DTYP ACL wire format, which must be zero [1]. When building an ACL descriptor in build_sec_desc(), we are using a kmalloc()'ed descriptor buffer and writing the fields explicitly using le16() writes now. This never writes to the 2 byte reserved field, leaving it as uninitialized heap data. When the reserved field happens to contain non-zero slab garbage, Samba rejects the security descriptor with "ndr_pull_security_descriptor failed: Range Error", causing chmod to fail with EINVAL. Change kmalloc() to kzalloc() to ensure the entire buffer is zero-initialized. Fixes: 62e7dd0a39c2d ("smb: common: change the data type of num_aces to le16") Cc: stable@vger.kernel.org Signed-off-by: Bjoern Doebel Assisted-by: Kiro:claude-opus-4.6 [1] https://learn.microsoft.com/en-us/openspecs/windows_protocols/ms-dtyp/20233ed8-a6c6-4097-aafa-dd545ed24428 Signed-off-by: Steve French --- fs/smb/client/cifsacl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/smb/client/cifsacl.c b/fs/smb/client/cifsacl.c index ec5d47779304..a2750f1e3d90 100644 --- a/fs/smb/client/cifsacl.c +++ b/fs/smb/client/cifsacl.c @@ -1732,7 +1732,7 @@ id_mode_to_cifs_acl(struct inode *inode, const char *path, __u64 *pnmode, * descriptor parameters, and security descriptor itself */ nsecdesclen = max_t(u32, nsecdesclen, DEFAULT_SEC_DESC_LEN); - pnntsd = kmalloc(nsecdesclen, GFP_KERNEL); + pnntsd = kzalloc(nsecdesclen, GFP_KERNEL); if (!pnntsd) { kfree(pntsd); cifs_put_tlink(tlink); From 84d5d76c4e8e2750fa17869b7272f189d2bdd40b Mon Sep 17 00:00:00 2001 From: Matthew Brost Date: Fri, 1 May 2026 23:53:38 -0700 Subject: [PATCH 224/525] drm/ttm: Fix GPU MM stats during pool shrinking TTM pool shrinking frees pages by calling __free_pages() directly, which bypasses updates to NR_GPU_ACTIVE and leaves GPU MM accounting out of sync. Introduce a helper, __free_pages_gpu_account(), and use it for all page frees in ttm_pool.c so GPU MM statistics are updated consistently. Reported-by: Kenneth Crudup Fixes: ae80122f3896 ("drm/ttm: use gpu mm stats to track gpu memory allocations. (v4)") Cc: Christian Koenig Cc: Huang Rui Cc: Matthew Auld Cc: David Airlie Cc: dri-devel@lists.freedesktop.org Signed-off-by: Matthew Brost Tested-by: Kenneth Crudup Reviewed-by: Dave Airlie Link: https://patch.msgid.link/20260502065338.2720646-1-matthew.brost@intel.com --- drivers/gpu/drm/ttm/ttm_pool.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/ttm/ttm_pool.c b/drivers/gpu/drm/ttm/ttm_pool.c index 26a3689e5fd9..278bbe7a11ad 100644 --- a/drivers/gpu/drm/ttm/ttm_pool.c +++ b/drivers/gpu/drm/ttm/ttm_pool.c @@ -206,6 +206,14 @@ static struct page *ttm_pool_alloc_page(struct ttm_pool *pool, gfp_t gfp_flags, return NULL; } +static void __free_pages_gpu_account(struct page *p, unsigned int order, + bool reclaim) +{ + mod_lruvec_page_state(p, reclaim ? NR_GPU_RECLAIM : NR_GPU_ACTIVE, + -(1 << order)); + __free_pages(p, order); +} + /* Reset the caching and pages of size 1 << order */ static void ttm_pool_free_page(struct ttm_pool *pool, enum ttm_caching caching, unsigned int order, struct page *p, bool reclaim) @@ -223,9 +231,7 @@ static void ttm_pool_free_page(struct ttm_pool *pool, enum ttm_caching caching, #endif if (!pool || !ttm_pool_uses_dma_alloc(pool)) { - mod_lruvec_page_state(p, reclaim ? NR_GPU_RECLAIM : NR_GPU_ACTIVE, - -(1 << order)); - __free_pages(p, order); + __free_pages_gpu_account(p, order, reclaim); return; } @@ -606,7 +612,7 @@ static int ttm_pool_restore_commit(struct ttm_pool_tt_restore *restore, */ ttm_pool_split_for_swap(restore->pool, p); copy_highpage(restore->alloced_page + i, p); - __free_pages(p, 0); + __free_pages_gpu_account(p, 0, false); } restore->restored_pages++; @@ -1068,7 +1074,7 @@ long ttm_pool_backup(struct ttm_pool *pool, struct ttm_tt *tt, if (flags->purge) { shrunken += num_pages; page->private = 0; - __free_pages(page, order); + __free_pages_gpu_account(page, order, false); memset(tt->pages + i, 0, num_pages * sizeof(*tt->pages)); } @@ -1109,7 +1115,7 @@ long ttm_pool_backup(struct ttm_pool *pool, struct ttm_tt *tt, } handle = shandle; tt->pages[i] = ttm_backup_handle_to_page_ptr(handle); - put_page(page); + __free_pages_gpu_account(page, 0, false); shrunken++; } From b8c2e9e27636b92dc96c12f16894cbc60c58a306 Mon Sep 17 00:00:00 2001 From: Yufan Chen Date: Mon, 4 May 2026 01:56:10 +0800 Subject: [PATCH 225/525] io_uring/napi: clear tracked NAPI entries on unregister IORING_UNREGISTER_NAPI disables NAPI busy polling, but it currently leaves any previously tracked NAPI IDs on the ring context. The normal wait path only checks whether the list is empty before entering the busy poll helper, so an unregistered ring can still observe stale entries and run an unexpected busy poll pass. Make unregister switch the context to inactive and free the tracked entries. Do the same inactive transition while changing the tracking strategy, and recheck the expected tracking mode under napi_lock before inserting a newly learned NAPI ID. This prevents a racing poll path from repopulating the list after unregister or reconfiguration. Also make the busy poll dispatcher ignore inactive mode explicitly. Signed-off-by: Yufan Chen Fixes: 6bf90bd8c58a ("io_uring/napi: add static napi tracking strategy") Link: https://patch.msgid.link/20260503175610.35521-1-yufan.chen@linux.dev Signed-off-by: Jens Axboe --- io_uring/napi.c | 27 ++++++++++++++++++++------- io_uring/napi.h | 8 +++++--- 2 files changed, 25 insertions(+), 10 deletions(-) diff --git a/io_uring/napi.c b/io_uring/napi.c index 8d68366a4b90..bfc771445912 100644 --- a/io_uring/napi.c +++ b/io_uring/napi.c @@ -38,7 +38,8 @@ static inline ktime_t net_to_ktime(unsigned long t) return ns_to_ktime(t << 10); } -int __io_napi_add_id(struct io_ring_ctx *ctx, unsigned int napi_id) +int __io_napi_add_id(struct io_ring_ctx *ctx, unsigned int napi_id, + unsigned int mode) { struct hlist_head *hash_list; struct io_napi_entry *e; @@ -69,6 +70,11 @@ int __io_napi_add_id(struct io_ring_ctx *ctx, unsigned int napi_id) * kfree() */ spin_lock(&ctx->napi_lock); + if (unlikely(READ_ONCE(ctx->napi_track_mode) != mode)) { + spin_unlock(&ctx->napi_lock); + kfree(e); + return -EINVAL; + } if (unlikely(io_napi_hash_find(hash_list, napi_id))) { spin_unlock(&ctx->napi_lock); kfree(e); @@ -196,9 +202,14 @@ __io_napi_do_busy_loop(struct io_ring_ctx *ctx, bool (*loop_end)(void *, unsigned long), void *loop_end_arg) { - if (READ_ONCE(ctx->napi_track_mode) == IO_URING_NAPI_TRACKING_STATIC) + switch (READ_ONCE(ctx->napi_track_mode)) { + case IO_URING_NAPI_TRACKING_STATIC: return static_tracking_do_busy_loop(ctx, loop_end, loop_end_arg); - return dynamic_tracking_do_busy_loop(ctx, loop_end, loop_end_arg); + case IO_URING_NAPI_TRACKING_DYNAMIC: + return dynamic_tracking_do_busy_loop(ctx, loop_end, loop_end_arg); + default: + return false; + } } static void io_napi_blocking_busy_loop(struct io_ring_ctx *ctx, @@ -273,13 +284,13 @@ static int io_napi_register_napi(struct io_ring_ctx *ctx, default: return -EINVAL; } - /* clean the napi list for new settings */ + WRITE_ONCE(ctx->napi_track_mode, IO_URING_NAPI_TRACKING_INACTIVE); io_napi_free(ctx); - WRITE_ONCE(ctx->napi_track_mode, napi->op_param); /* cap NAPI at 10 msec of spin time */ napi->busy_poll_to = min(10000, napi->busy_poll_to); WRITE_ONCE(ctx->napi_busy_poll_dt, napi->busy_poll_to * NSEC_PER_USEC); WRITE_ONCE(ctx->napi_prefer_busy_poll, !!napi->prefer_busy_poll); + WRITE_ONCE(ctx->napi_track_mode, napi->op_param); return 0; } @@ -315,7 +326,8 @@ int io_register_napi(struct io_ring_ctx *ctx, void __user *arg) case IO_URING_NAPI_STATIC_ADD_ID: if (curr.op_param != IO_URING_NAPI_TRACKING_STATIC) return -EINVAL; - return __io_napi_add_id(ctx, napi.op_param); + return __io_napi_add_id(ctx, napi.op_param, + IO_URING_NAPI_TRACKING_STATIC); case IO_URING_NAPI_STATIC_DEL_ID: if (curr.op_param != IO_URING_NAPI_TRACKING_STATIC) return -EINVAL; @@ -343,9 +355,10 @@ int io_unregister_napi(struct io_ring_ctx *ctx, void __user *arg) if (arg && copy_to_user(arg, &curr, sizeof(curr))) return -EFAULT; + WRITE_ONCE(ctx->napi_track_mode, IO_URING_NAPI_TRACKING_INACTIVE); WRITE_ONCE(ctx->napi_busy_poll_dt, 0); WRITE_ONCE(ctx->napi_prefer_busy_poll, false); - WRITE_ONCE(ctx->napi_track_mode, IO_URING_NAPI_TRACKING_INACTIVE); + io_napi_free(ctx); return 0; } diff --git a/io_uring/napi.h b/io_uring/napi.h index fa742f42e09b..e0aecccc5065 100644 --- a/io_uring/napi.h +++ b/io_uring/napi.h @@ -15,7 +15,8 @@ void io_napi_free(struct io_ring_ctx *ctx); int io_register_napi(struct io_ring_ctx *ctx, void __user *arg); int io_unregister_napi(struct io_ring_ctx *ctx, void __user *arg); -int __io_napi_add_id(struct io_ring_ctx *ctx, unsigned int napi_id); +int __io_napi_add_id(struct io_ring_ctx *ctx, unsigned int napi_id, + unsigned int mode); void __io_napi_busy_loop(struct io_ring_ctx *ctx, struct io_wait_queue *iowq); int io_napi_sqpoll_busy_poll(struct io_ring_ctx *ctx); @@ -43,13 +44,14 @@ static inline void io_napi_add(struct io_kiocb *req) { struct io_ring_ctx *ctx = req->ctx; struct socket *sock; + unsigned int mode = IO_URING_NAPI_TRACKING_DYNAMIC; - if (READ_ONCE(ctx->napi_track_mode) != IO_URING_NAPI_TRACKING_DYNAMIC) + if (READ_ONCE(ctx->napi_track_mode) != mode) return; sock = sock_from_file(req->file); if (sock && sock->sk) - __io_napi_add_id(ctx, READ_ONCE(sock->sk->sk_napi_id)); + __io_napi_add_id(ctx, READ_ONCE(sock->sk->sk_napi_id), mode); } #else From 04fe9aeb4f3c0999e6715385664c677469dfd8f4 Mon Sep 17 00:00:00 2001 From: Yufan Chen Date: Mon, 4 May 2026 01:57:10 +0800 Subject: [PATCH 226/525] io_uring/eventfd: reset deferred signal state Recursive eventfd wakeups must defer io_uring eventfd signaling because eventfd_signal_mask() rejects reentry from eventfd wakeup handlers. The io_ev_fd ops bit tracks an outstanding deferred signal so that the same rcu_head is not queued twice. That bit is only set today. Once the first deferred callback runs, later recursive notifications still see the bit set and skip queueing another deferred signal. This can leave new completions without a matching eventfd wake after the first recursive deferral. Clear the pending bit before issuing the deferred signal. If the wakeup path recurses while the callback runs, a new signal can be queued for the next RCU grace period while the current callback keeps its reference until it returns. Signed-off-by: Yufan Chen Fixes: 60b6c075e8eb ("io_uring/eventfd: move to more idiomatic RCU free usage") Link: https://patch.msgid.link/20260503175710.37209-1-yufan.chen@linux.dev Signed-off-by: Jens Axboe --- io_uring/eventfd.c | 1 + 1 file changed, 1 insertion(+) diff --git a/io_uring/eventfd.c b/io_uring/eventfd.c index 3da028500f76..d656cc2a0b9b 100644 --- a/io_uring/eventfd.c +++ b/io_uring/eventfd.c @@ -43,6 +43,7 @@ static void io_eventfd_do_signal(struct rcu_head *rcu) { struct io_ev_fd *ev_fd = container_of(rcu, struct io_ev_fd, rcu); + atomic_andnot(BIT(IO_EVENTFD_OP_SIGNAL_BIT), &ev_fd->ops); eventfd_signal_mask(ev_fd->cq_ev_fd, EPOLL_URING_WAKE); io_eventfd_put(ev_fd); } From 0cfff13c94cb5fa818bb374945ff280e08dc1bb9 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 4 May 2026 08:54:27 +0200 Subject: [PATCH 227/525] wifi: mac80211: tests: mark HT check strict The HT check now only applies in strict mode since APs were found to be broken. Mark it as such. Fixes: 711a9c018ad2 ("wifi: mac80211: skip ieee80211_verify_sta_ht_mcs_support check in non-strict mode") Signed-off-by: Johannes Berg --- net/mac80211/tests/chan-mode.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/mac80211/tests/chan-mode.c b/net/mac80211/tests/chan-mode.c index adc069065e73..fa370831d617 100644 --- a/net/mac80211/tests/chan-mode.c +++ b/net/mac80211/tests/chan-mode.c @@ -65,6 +65,7 @@ static const struct determine_chan_mode_case { .ht_capa_mask = { .mcs.rx_mask[0] = 0xf7, }, + .strict = true, }, { .desc = "Masking out a RX rate in VHT capabilities", .conn_mode = IEEE80211_CONN_MODE_EHT, From 65493f27a6008bf84bd11bd41c5e1ea6b0bf3c3d Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Thu, 30 Apr 2026 10:44:15 -0700 Subject: [PATCH 228/525] wifi: cw1200: Revert "Fix locking in error paths" Revert commit d98c24617a83 ("wifi: cw1200: Fix locking in error paths") because it introduces a locking bug instead of fixing a locking bug. cw1200_wow_resume() unlocks priv->conf_mutex. Hence, adding mutex_unlock(&priv->conf_mutex) just after cw1200_wow_resume() is wrong. Reported-by: Ben Hutchings Closes: https://lore.kernel.org/all/408661f69f263266b028713e1412ba36d457e63d.camel@decadent.org.uk/ Fixes: d98c24617a83 ("wifi: cw1200: Fix locking in error paths") Signed-off-by: Bart Van Assche Link: https://patch.msgid.link/20260430174418.1845431-1-bvanassche@acm.org Signed-off-by: Johannes Berg --- drivers/net/wireless/st/cw1200/pm.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/net/wireless/st/cw1200/pm.c b/drivers/net/wireless/st/cw1200/pm.c index 84eb15d729c7..120f0379f81d 100644 --- a/drivers/net/wireless/st/cw1200/pm.c +++ b/drivers/net/wireless/st/cw1200/pm.c @@ -264,14 +264,12 @@ int cw1200_wow_suspend(struct ieee80211_hw *hw, struct cfg80211_wowlan *wowlan) wiphy_err(priv->hw->wiphy, "PM request failed: %d. WoW is disabled.\n", ret); cw1200_wow_resume(hw); - mutex_unlock(&priv->conf_mutex); return -EBUSY; } /* Force resume if event is coming from the device. */ if (atomic_read(&priv->bh_rx)) { cw1200_wow_resume(hw); - mutex_unlock(&priv->conf_mutex); return -EAGAIN; } From 8bc6b14aab669f0c301febcf4aa5249b9393bf51 Mon Sep 17 00:00:00 2001 From: Marco Crivellari Date: Thu, 30 Apr 2026 11:08:10 +0200 Subject: [PATCH 229/525] i2c: testunit: Replace system_long_wq with system_dfl_long_wq Currently the code enqueue work items using {queue|mod}_delayed_work(), using system_long_wq. This workqueue should be used when long works are expected, but it is a per-cpu workqueue. This is important because queue_delayed_work() queue the work using: queue_delayed_work_on(WORK_CPU_UNBOUND, ...); Note that WORK_CPU_UNBOUND = NR_CPUS. This would end up calling __queue_delayed_work() that does: if (housekeeping_enabled(HK_TYPE_TIMER)) { // [....] } else { if (likely(cpu == WORK_CPU_UNBOUND)) add_timer_global(timer); else add_timer_on(timer, cpu); } So when cpu == WORK_CPU_UNBOUND the timer is global and is not using a specific CPU. Later, when __queue_work() is called: if (req_cpu == WORK_CPU_UNBOUND) { if (wq->flags & WQ_UNBOUND) cpu = wq_select_unbound_cpu(raw_smp_processor_id()); else cpu = raw_smp_processor_id(); } Because the wq is not unbound, it takes the CPU where the timer fired and enqueue the work on that CPU. The consequence of all of this is that the work can run anywhere, depending on where the timer fired. Recently, a new unbound workqueue specific for long running work has been added: c116737e972e ("workqueue: Add system_dfl_long_wq for long unbound works") So change system_long_wq with system_dfl_long_wq so that the work may benefit from scheduler task placement. Signed-off-by: Marco Crivellari [wsa: remove FIXME as well] Signed-off-by: Wolfram Sang --- drivers/i2c/i2c-slave-testunit.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/i2c/i2c-slave-testunit.c b/drivers/i2c/i2c-slave-testunit.c index 6de4307050dd..871c58461ebc 100644 --- a/drivers/i2c/i2c-slave-testunit.c +++ b/drivers/i2c/i2c-slave-testunit.c @@ -15,7 +15,7 @@ #include #include #include -#include /* FIXME: is system_long_wq the best choice? */ +#include #define TU_VERSION_MAX_LENGTH 128 @@ -124,7 +124,7 @@ static int i2c_slave_testunit_slave_cb(struct i2c_client *client, case I2C_SLAVE_STOP: if (tu->reg_idx == TU_NUM_REGS) { set_bit(TU_FLAG_IN_PROCESS, &tu->flags); - queue_delayed_work(system_long_wq, &tu->worker, + queue_delayed_work(system_dfl_long_wq, &tu->worker, msecs_to_jiffies(10 * tu->regs[TU_REG_DELAY])); } From 9a937ca22741eebe2bf10b18657b8b9aed9c009b Mon Sep 17 00:00:00 2001 From: Ronald Claveau Date: Fri, 24 Apr 2026 16:17:33 +0200 Subject: [PATCH 230/525] dt-bindings: i2c: amlogic: Add compatible for T7 SOC Add the T7 SOC compatible which fallback to AXG compatible. Acked-by: Rob Herring (Arm) Signed-off-by: Ronald Claveau Signed-off-by: Wolfram Sang --- .../devicetree/bindings/i2c/amlogic,meson6-i2c.yaml | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/Documentation/devicetree/bindings/i2c/amlogic,meson6-i2c.yaml b/Documentation/devicetree/bindings/i2c/amlogic,meson6-i2c.yaml index c4cc8af18280..7b59b60b62e5 100644 --- a/Documentation/devicetree/bindings/i2c/amlogic,meson6-i2c.yaml +++ b/Documentation/devicetree/bindings/i2c/amlogic,meson6-i2c.yaml @@ -16,10 +16,15 @@ allOf: properties: compatible: - enum: - - amlogic,meson6-i2c # Meson6, Meson8 and compatible SoCs - - amlogic,meson-gxbb-i2c # GXBB and compatible SoCs - - amlogic,meson-axg-i2c # AXG and compatible SoCs + oneOf: + - items: + - enum: + - amlogic,t7-i2c + - const: amlogic,meson-axg-i2c + - enum: + - amlogic,meson6-i2c # Meson6, Meson8 and compatible SoCs + - amlogic,meson-gxbb-i2c # GXBB and compatible SoCs + - amlogic,meson-axg-i2c # AXG and compatible SoCs reg: maxItems: 1 From 02b7cd683892d8d8464815d69a3a76579909a726 Mon Sep 17 00:00:00 2001 From: Marek Vasut Date: Sat, 2 May 2026 17:31:54 +0200 Subject: [PATCH 231/525] i2c: stm32f7: reinit_completion() per transfer not per msg Currently, the driver may repeatedly call reinit_completion() during transfer which contains multiple messages, while another thread is waiting for the completion. This happens during transfer with more than 1 message, invoked via stm32f7_i2c_xfer_core() -> stm32f7_i2c_xfer_msg(). After invoking the stm32f7_i2c_xfer_msg() to start transfer, stm32f7_i2c_xfer_core() calls wait_for_completion_timeout() to wait for completion of the transfer of all messages. When the first message transfer completes, the hard IRQ handler triggers, and detects transfer completion, which leads to stm32f7_i2c_isr_event_thread() IRQ thread being started. The stm32f7_i2c_isr_event_thread() calls stm32f7_i2c_xfer_msg() in case there are more messages. Without this change, the second and later stm32f7_i2c_xfer_msg() would call reinit_completion() on the completion which is still being waited for in stm32f7_i2c_xfer_core(). Fix this by moving the reinit_completion() into stm32f7_i2c_xfer_core(), together with wait_for_completion_timeout(). Since stm32f7_i2c_xfer_core() now waits for completion of the entire transfer, increase the default timeout. This fixes sporadic transfer timeouts on STM32MP25xx during kernel boot. Fixes: aeb068c57214 ("i2c: i2c-stm32f7: add driver") Signed-off-by: Marek Vasut [wsa: reworded commit subject] Signed-off-by: Wolfram Sang --- drivers/i2c/busses/i2c-stm32f7.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/i2c/busses/i2c-stm32f7.c b/drivers/i2c/busses/i2c-stm32f7.c index 70cb5822bf17..53d9df70ebe4 100644 --- a/drivers/i2c/busses/i2c-stm32f7.c +++ b/drivers/i2c/busses/i2c-stm32f7.c @@ -895,8 +895,6 @@ static void stm32f7_i2c_xfer_msg(struct stm32f7_i2c_dev *i2c_dev, f7_msg->result = 0; f7_msg->stop = (i2c_dev->msg_id >= i2c_dev->msg_num - 1); - reinit_completion(&i2c_dev->complete); - cr1 = readl_relaxed(base + STM32F7_I2C_CR1); cr2 = readl_relaxed(base + STM32F7_I2C_CR2); @@ -1728,6 +1726,8 @@ static int stm32f7_i2c_xfer_core(struct i2c_adapter *i2c_adap, if (ret) goto pm_free; + reinit_completion(&i2c_dev->complete); + stm32f7_i2c_xfer_msg(i2c_dev, msgs); if (!i2c_dev->atomic) @@ -2253,7 +2253,7 @@ static int stm32f7_i2c_probe(struct platform_device *pdev) snprintf(adap->name, sizeof(adap->name), "STM32F7 I2C(%pa)", &res->start); adap->owner = THIS_MODULE; - adap->timeout = 2 * HZ; + adap->timeout = 8 * HZ; adap->retries = 3; adap->algo = &stm32f7_i2c_algo; adap->dev.parent = &pdev->dev; From 10161b4a791d5c4b7ea16512c1ddb133c3f8f953 Mon Sep 17 00:00:00 2001 From: Weinan Liu Date: Thu, 30 Apr 2026 23:28:51 +0000 Subject: [PATCH 232/525] iommu/amd: Fix precedence order in set_dte_passthrough() Bitwise OR | operator has a higher precedence than the ternary ?: operatior. It will be incorrectly evaluated as: new->data[1] |= (FIELD_PREP(...) | dev_data->ats_enabled) ? DTE_FLAG_IOTLB : 0; Wrap the conditional operation in parentheses to enforce the correct evaluation order. Fixes: 93eee2a49c1b ("iommu/amd: Refactor logic to program the host page table in DTE") Signed-off-by: Weinan Liu Reviewed-by: Jason Gunthorpe Reviewed-by: Vasant Hegde Signed-off-by: Joerg Roedel --- drivers/iommu/amd/iommu.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/iommu/amd/iommu.c b/drivers/iommu/amd/iommu.c index 157505d96fdd..f78e23f03938 100644 --- a/drivers/iommu/amd/iommu.c +++ b/drivers/iommu/amd/iommu.c @@ -2149,7 +2149,8 @@ static void set_dte_passthrough(struct iommu_dev_data *dev_data, new->data[0] |= DTE_FLAG_TV | DTE_FLAG_IR | DTE_FLAG_IW; new->data[1] |= FIELD_PREP(DTE_DOMID_MASK, domain->id) | - (dev_data->ats_enabled) ? DTE_FLAG_IOTLB : 0; + (dev_data->ats_enabled ? DTE_FLAG_IOTLB : 0); + } static void set_dte_entry(struct amd_iommu *iommu, From c5f25f5800f56f1754d9eeb3ced7c1e08c29119a Mon Sep 17 00:00:00 2001 From: Janne Grunau Date: Fri, 20 Mar 2026 13:23:24 +0100 Subject: [PATCH 233/525] dt-bindings: i2c: apple,i2c: Add t8122 compatible The i2c block on the Apple silicon t8122 (M3) SoC is compatible with the existing driver. Add "apple,t8122-i2c" as SoC specific compatible under "apple,t8103-i2c" used by the deriver. Signed-off-by: Janne Grunau Acked-by: Andi Shyti Acked-by: Rob Herring (Arm) Signed-off-by: Wolfram Sang --- Documentation/devicetree/bindings/i2c/apple,i2c.yaml | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/Documentation/devicetree/bindings/i2c/apple,i2c.yaml b/Documentation/devicetree/bindings/i2c/apple,i2c.yaml index 500a965bdb7a..9e59200ad37b 100644 --- a/Documentation/devicetree/bindings/i2c/apple,i2c.yaml +++ b/Documentation/devicetree/bindings/i2c/apple,i2c.yaml @@ -22,7 +22,9 @@ properties: compatible: oneOf: - items: - - const: apple,t6020-i2c + - enum: + - apple,t6020-i2c + - apple,t8122-i2c - const: apple,t8103-i2c - items: - enum: From 8de779dc40d35d39fa07387b6f921eb11df0f511 Mon Sep 17 00:00:00 2001 From: Rajat Gupta Date: Sun, 3 May 2026 20:51:10 -0700 Subject: [PATCH 234/525] fbdev: udlfb: add vm_ops to dlfb_ops_mmap to prevent use-after-free dlfb_ops_mmap() uses remap_pfn_range() to map vmalloc framebuffer pages to userspace but sets no vm_ops on the VMA. This means the kernel cannot track active mmaps. When dlfb_realloc_framebuffer() replaces the backing buffer via FBIOPUT_VSCREENINFO, existing mmap PTEs are not invalidated. On USB disconnect, dlfb_ops_destroy() calls vfree() on the old pages while userspace PTEs still reference them, resulting in a use-after-free: the process retains read/write access to freed kernel pages. Add vm_operations_struct with open/close callbacks that maintain an atomic mmap_count on struct dlfb_data. In dlfb_realloc_framebuffer(), check mmap_count and return -EBUSY if the buffer is currently mapped, preventing buffer replacement while userspace holds stale PTEs. Tested with PoC using dummy_hcd + raw_gadget USB device emulation. Signed-off-by: Rajat Gupta Acked-by: Greg Kroah-Hartman Cc: stable@vger.kernel.org Signed-off-by: Helge Deller --- drivers/video/fbdev/udlfb.c | 31 ++++++++++++++++++++++++++++++- include/video/udlfb.h | 1 + 2 files changed, 31 insertions(+), 1 deletion(-) diff --git a/drivers/video/fbdev/udlfb.c b/drivers/video/fbdev/udlfb.c index c341d76bc564..fdbb8671a810 100644 --- a/drivers/video/fbdev/udlfb.c +++ b/drivers/video/fbdev/udlfb.c @@ -321,12 +321,32 @@ static int dlfb_set_video_mode(struct dlfb_data *dlfb, return retval; } +static void dlfb_vm_open(struct vm_area_struct *vma) +{ + struct dlfb_data *dlfb = vma->vm_private_data; + + atomic_inc(&dlfb->mmap_count); +} + +static void dlfb_vm_close(struct vm_area_struct *vma) +{ + struct dlfb_data *dlfb = vma->vm_private_data; + + atomic_dec(&dlfb->mmap_count); +} + +static const struct vm_operations_struct dlfb_vm_ops = { + .open = dlfb_vm_open, + .close = dlfb_vm_close, +}; + static int dlfb_ops_mmap(struct fb_info *info, struct vm_area_struct *vma) { unsigned long start = vma->vm_start; unsigned long size = vma->vm_end - vma->vm_start; unsigned long offset = vma->vm_pgoff << PAGE_SHIFT; unsigned long page, pos; + struct dlfb_data *dlfb = info->par; if (info->fbdefio) return fb_deferred_io_mmap(info, vma); @@ -358,6 +378,9 @@ static int dlfb_ops_mmap(struct fb_info *info, struct vm_area_struct *vma) size = 0; } + vma->vm_ops = &dlfb_vm_ops; + vma->vm_private_data = dlfb; + atomic_inc(&dlfb->mmap_count); return 0; } @@ -1176,7 +1199,6 @@ static void dlfb_deferred_vfree(struct dlfb_data *dlfb, void *mem) /* * Assumes &info->lock held by caller - * Assumes no active clients have framebuffer open */ static int dlfb_realloc_framebuffer(struct dlfb_data *dlfb, struct fb_info *info, u32 new_len) { @@ -1188,6 +1210,13 @@ static int dlfb_realloc_framebuffer(struct dlfb_data *dlfb, struct fb_info *info new_len = PAGE_ALIGN(new_len); if (new_len > old_len) { + if (atomic_read(&dlfb->mmap_count) > 0) { + dev_warn(info->dev, + "refusing realloc: %d active mmaps\n", + atomic_read(&dlfb->mmap_count)); + return -EBUSY; + } + /* * Alloc system memory for virtual framebuffer */ diff --git a/include/video/udlfb.h b/include/video/udlfb.h index 58fb5732831a..ab34790d57ec 100644 --- a/include/video/udlfb.h +++ b/include/video/udlfb.h @@ -56,6 +56,7 @@ struct dlfb_data { spinlock_t damage_lock; struct work_struct damage_work; struct fb_ops ops; + atomic_t mmap_count; /* blit-only rendering path metrics, exposed through sysfs */ atomic_t bytes_rendered; /* raw pixel-bytes driver asked to render */ atomic_t bytes_identical; /* saved effort with backbuffer comparison */ From 9998e388be9930c106eb5904c23ecf2162407527 Mon Sep 17 00:00:00 2001 From: Niels Franke Date: Sat, 18 Apr 2026 07:37:19 +0200 Subject: [PATCH 235/525] i2c: acpi: Add ELAN0678 to i2c_acpi_force_100khz_device_ids The ELAN0678 touchpad (04F3:3195) found in the Lenovo ThinkPad X13 exhibits excessive smoothing when the I2C bus runs at 400KHz, making the touchpad feel sluggish when plugged into AC power. This is the same issue previously fixed for ELAN06FA. The device's ACPI table (Lenovo TP-R22) specifies 0x00061A80 (400KHz) for the I2cSerialBusV2 descriptor. Forcing the bus to 100KHz eliminates the sluggish behavior. Signed-off-by: Niels Franke Acked-by: Mika Westerberg [wsa: kept the sorting] Signed-off-by: Wolfram Sang --- drivers/i2c/i2c-core-acpi.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/i2c/i2c-core-acpi.c b/drivers/i2c/i2c-core-acpi.c index 2cbd31f77667..28c0e4884a7f 100644 --- a/drivers/i2c/i2c-core-acpi.c +++ b/drivers/i2c/i2c-core-acpi.c @@ -371,6 +371,7 @@ static const struct acpi_device_id i2c_acpi_force_100khz_device_ids[] = { * a 400KHz frequency. The root cause of the issue is not known. */ { "DLL0945", 0 }, + { "ELAN0678", 0 }, { "ELAN06FA", 0 }, {} }; From 32c91e8ee039777d0b95b914633fc6a42607959c Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Fri, 24 Apr 2026 12:49:10 +0200 Subject: [PATCH 236/525] staging: vme_user: fix root device leak on init failure Make sure to deregister and free the root device in case module initialisation fails. Fixes: 658bcdae9c67 ("vme: Adding Fake VME driver") Cc: stable@vger.kernel.org # 4.9 Cc: Martyn Welch Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260424104910.2619349-1-johan@kernel.org Signed-off-by: Greg Kroah-Hartman --- drivers/staging/vme_user/vme_fake.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/staging/vme_user/vme_fake.c b/drivers/staging/vme_user/vme_fake.c index be4ad47ed526..8abaa3165fbb 100644 --- a/drivers/staging/vme_user/vme_fake.c +++ b/drivers/staging/vme_user/vme_fake.c @@ -1230,6 +1230,8 @@ static int __init fake_init(void) err_driver: kfree(fake_bridge); err_struct: + root_device_unregister(vme_root); + return retval; } From 617eb7c0961a8dfcfc811844a6396e406b2923ea Mon Sep 17 00:00:00 2001 From: Mingyu Wang <25181214217@stu.xidian.edu.cn> Date: Mon, 27 Apr 2026 10:57:45 +0800 Subject: [PATCH 237/525] i2c: dev: prevent integer overflow in I2C_TIMEOUT ioctl While fuzzing with Syzkaller, a persistent `schedule_timeout: wrong timeout value` warning was observed, accompanied by SMBus controller state machine corruption. The I2C_TIMEOUT ioctl accepts a user-provided timeout in multiples of 10 ms. The user argument is checked against INT_MAX, but it is subsequently multiplied by 10 before being passed to msecs_to_jiffies(). A malicious user can pass a large value (e.g., 429496729) that passes the `arg > INT_MAX` check but overflows when multiplied by 10. This results in a truncated 32-bit unsigned value that bypasses the internal `(int)m < 0` check in `msecs_to_jiffies()`. The truncated value is then assigned to `client->adapter->timeout` (a signed 32-bit int), which is reinterpreted as a negative number. When passed to wait_for_completion_timeout(), this negative value undergoes sign extension to a 64-bit unsigned long, triggering the `schedule_timeout` warning and causing premature returns. This leaves the SMBus state machine in an unrecoverable state, constituting a local Denial of Service (DoS). Fix this by bounding the user argument to `INT_MAX / 10`. Signed-off-by: Mingyu Wang <25181214217@stu.xidian.edu.cn> [wsa: move the comment as well] Signed-off-by: Wolfram Sang --- drivers/i2c/i2c-dev.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c index 7bbe0263411e..ccaac5e29f90 100644 --- a/drivers/i2c/i2c-dev.c +++ b/drivers/i2c/i2c-dev.c @@ -487,12 +487,13 @@ static long i2cdev_ioctl(struct file *file, unsigned int cmd, unsigned long arg) client->adapter->retries = arg; break; case I2C_TIMEOUT: - if (arg > INT_MAX) + /* + * For historical reasons, user-space sets the timeout value in + * units of 10 ms. + */ + if (arg > INT_MAX / 10) return -EINVAL; - /* For historical reasons, user-space sets the timeout - * value in units of 10 ms. - */ client->adapter->timeout = msecs_to_jiffies(arg * 10); break; default: From bc851db06045a40c18233dd76ef0562d7f8bb6db Mon Sep 17 00:00:00 2001 From: Shyam Sunder Reddy Padira Date: Tue, 14 Apr 2026 12:43:06 +0530 Subject: [PATCH 238/525] staging: rtl8723bs: os_dep: avoid NULL pointer dereference in rtw_cbuf_alloc The return value of kzalloc_flex() is used without ensuring that the allocation succeeded, and the pointer is dereferenced unconditionally. Guard the access to the allocated structure to avoid a potential NULL pointer dereference if the allocation fails. Fixes: 980cd426a257 ("staging: rtl8723bs: replace rtw_zmalloc() with kzalloc()") Cc: stable Signed-off-by: Shyam Sunder Reddy Padira Reviewed-by: Dan Carpenter Link: https://patch.msgid.link/20260414071308.4781-2-shyamsunderreddypadira@gmail.com Signed-off-by: Greg Kroah-Hartman --- drivers/staging/rtl8723bs/os_dep/osdep_service.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/staging/rtl8723bs/os_dep/osdep_service.c b/drivers/staging/rtl8723bs/os_dep/osdep_service.c index 7959daeabc6f..4cfdf7c62344 100644 --- a/drivers/staging/rtl8723bs/os_dep/osdep_service.c +++ b/drivers/staging/rtl8723bs/os_dep/osdep_service.c @@ -194,7 +194,8 @@ struct rtw_cbuf *rtw_cbuf_alloc(u32 size) struct rtw_cbuf *cbuf; cbuf = kzalloc_flex(*cbuf, bufs, size); - cbuf->size = size; + if (cbuf) + cbuf->size = size; return cbuf; } From 37b0dc5e279f35036fb638d1e187197b6c05a76d Mon Sep 17 00:00:00 2001 From: Hongling Zeng Date: Sun, 3 May 2026 12:17:44 +0800 Subject: [PATCH 239/525] parisc: Fix IRQ leak in LASI driver When request_irq() succeeds but gsc_common_setup() fails later, the IRQ is never released. Fix this by adding proper error handling with goto labels to ensure resources are released in LIFO order. Detected by Smatch: drivers/parisc/lasi.c:216 lasi_init_chip() warn: 'lasi->gsc_irq.irq' from request_irq() not released on lines: 207. Reported-by: kernel test robot Reported-by: Dan Carpenter Closes: https://lore.kernel.org/r/202604180957.4QdAIxP6-lkp@intel.com/ Signed-off-by: Hongling Zeng Cc: stable@vger.kernel.org Signed-off-by: Helge Deller --- drivers/parisc/lasi.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/drivers/parisc/lasi.c b/drivers/parisc/lasi.c index ef6125d83878..a5b80cd5cc37 100644 --- a/drivers/parisc/lasi.c +++ b/drivers/parisc/lasi.c @@ -193,8 +193,7 @@ static int __init lasi_init_chip(struct parisc_device *dev) ret = request_irq(lasi->gsc_irq.irq, gsc_asic_intr, 0, "lasi", lasi); if (ret < 0) { - kfree(lasi); - return ret; + goto err_free; } /* enable IRQ's for devices below LASI */ @@ -203,8 +202,7 @@ static int __init lasi_init_chip(struct parisc_device *dev) /* Done init'ing, register this driver */ ret = gsc_common_setup(dev, lasi); if (ret) { - kfree(lasi); - return ret; + goto err_irq; } gsc_fixup_irqs(dev, lasi, lasi_choose_irq); @@ -214,6 +212,12 @@ static int __init lasi_init_chip(struct parisc_device *dev) SYS_OFF_PRIO_DEFAULT, lasi_power_off, lasi); return ret; + +err_irq: + free_irq(lasi->gsc_irq.irq, lasi); +err_free: + kfree(lasi); + return ret; } static struct parisc_device_id lasi_tbl[] __initdata = { From b47bc7c022ddab7c79a84dd5f3f0d07fe09ec786 Mon Sep 17 00:00:00 2001 From: "Nikola Z. Ivanov" Date: Wed, 15 Apr 2026 23:50:21 +0300 Subject: [PATCH 240/525] i2c: Compare the return value of gpiod_get_direction against GPIO_LINE_DIRECTION_OUT The GPIO_LINE_DIRECTION_* definitions have just recently been exposed to gpio consumers.h by breaking them out in a separate defs.h file. Use this to validate the gpio direction instead of the hard-coded literal. Signed-off-by: Nikola Z. Ivanov Signed-off-by: Wolfram Sang --- drivers/i2c/i2c-core-base.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/i2c/i2c-core-base.c b/drivers/i2c/i2c-core-base.c index 9c46147e3506..a2132d70fb36 100644 --- a/drivers/i2c/i2c-core-base.c +++ b/drivers/i2c/i2c-core-base.c @@ -445,8 +445,7 @@ static int i2c_init_recovery(struct i2c_adapter *adap) bri->set_scl = set_scl_gpio_value; if (bri->sda_gpiod) { bri->get_sda = get_sda_gpio_value; - /* FIXME: add proper flag instead of '0' once available */ - if (gpiod_get_direction(bri->sda_gpiod) == 0) + if (gpiod_get_direction(bri->sda_gpiod) == GPIO_LINE_DIRECTION_OUT) bri->set_sda = set_sda_gpio_value; } } else if (bri->recover_bus == i2c_generic_scl_recovery) { From 088f65e206087bf903743bd18417261d7a4c9644 Mon Sep 17 00:00:00 2001 From: Ivan Hu Date: Thu, 30 Apr 2026 15:41:07 +0800 Subject: [PATCH 241/525] x86/efi: Fix graceful fault handling after FPU softirq changes Since commit d02198550423 ("x86/fpu: Improve crypto performance by making kernel-mode FPU reliably usable in softirqs"), kernel_fpu_begin() calls fpregs_lock() which uses local_bh_disable() instead of the previous preempt_disable(). This sets SOFTIRQ_OFFSET in preempt_count during the entire EFI runtime service call, causing in_interrupt() to return true in normal task context. The graceful page fault handler efi_crash_gracefully_on_page_fault() uses in_interrupt() to bail out for faults in real interrupt context. With SOFTIRQ_OFFSET now set, the handler always bails out, leaving EFI firmware page faults unhandled. This escalates to die() which also sees in_interrupt() as true and calls panic("Fatal exception in interrupt"), resulting in a hard system freeze. On systems with buggy firmware that triggers page faults during EFI runtime calls (e.g., accessing unmapped memory in GetTime()), this causes an unrecoverable hang instead of the expected graceful EFI_ABORTED recovery. Fix by replacing in_interrupt() with !in_task(). This preserves the original intent of bailing for interrupts or NMI faults, while no longer falsely triggering from the FPU code path's local_bh_disable(). Fixes: d02198550423 ("x86/fpu: Improve crypto performance by making kernel-mode FPU reliably usable in softirqs") Cc: Signed-off-by: Ivan Hu [ardb: Sashiko spotted that using 'in_hardirq() || in_nmi()' leaves a window where a softirq may be taken before fpregs_lock() is called, but after efi_rts_work.efi_rts_id has been assigned, and any page faults occurring in that window will then be misidentified as having been caused by the firmware. Instead, use !in_task(), which incorporates in_serving_softirq(). ] Signed-off-by: Ard Biesheuvel --- arch/x86/platform/efi/quirks.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/platform/efi/quirks.c b/arch/x86/platform/efi/quirks.c index df24ffc6105d..c8f5e094ed9d 100644 --- a/arch/x86/platform/efi/quirks.c +++ b/arch/x86/platform/efi/quirks.c @@ -770,7 +770,7 @@ void efi_crash_gracefully_on_page_fault(unsigned long phys_addr) * If we get an interrupt/NMI while processing an EFI runtime service * then this is a regular OOPS, not an EFI failure. */ - if (in_interrupt()) + if (!in_task()) return; /* From 6036b5067a8199ba7a2dc7b377d4b9dd276d5f9e Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Wed, 15 Apr 2026 01:23:39 +0800 Subject: [PATCH 242/525] i2c: stub: Reject I2C block transfers with invalid length The I2C_SMBUS_I2C_BLOCK_DATA case in stub_xfer() uses data->block[0] as the transfer length. The existing check only clamps it to avoid overrunning the chip->words[256] register array, but does not validate it against I2C_SMBUS_BLOCK_MAX (32), which is the limit of the union i2c_smbus_data.block buffer (34 bytes total). The driver is a development/test tool (CONFIG_I2C_STUB=m, not built by default) that must be loaded with a chip_addr= parameter. A local user with access to /dev/i2c-* can issue an I2C_SMBUS ioctl with I2C_SMBUS_I2C_BLOCK_DATA and data->block[0] > 32, causing stub_xfer() to read or write past the end of the union i2c_smbus_data.block buffer: BUG: KASAN: stack-out-of-bounds in stub_xfer (drivers/i2c/i2c-stub.c:223) Read of size 1 at addr ffff88800abcfd92 by task exploit/81 Call Trace: stub_xfer (drivers/i2c/i2c-stub.c:223) __i2c_smbus_xfer (drivers/i2c/i2c-core-smbus.c:593) i2c_smbus_xfer (drivers/i2c/i2c-core-smbus.c:536) i2cdev_ioctl_smbus (drivers/i2c/i2c-dev.c:391) i2cdev_ioctl (drivers/i2c/i2c-dev.c:478) __x64_sys_ioctl (fs/ioctl.c:583) do_syscall_64 (arch/x86/entry/syscall_64.c:94) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) The bug exists because i2c-stub implements .smbus_xfer directly, bypassing the I2C_SMBUS_BLOCK_MAX validation in i2c_smbus_xfer_emulated(). The I2C_SMBUS_BLOCK_DATA case in the same function correctly validates against I2C_SMBUS_BLOCK_MAX, but the I2C_SMBUS_I2C_BLOCK_DATA case does not. Fix by rejecting transfers with data->block[0] == 0 or data->block[0] > I2C_SMBUS_BLOCK_MAX with -EINVAL, consistent with both the I2C_SMBUS_BLOCK_DATA case in the same function and the I2C_SMBUS_I2C_BLOCK_DATA validation in i2c_smbus_xfer_emulated(). Fixes: 4710317891e4 ("i2c-stub: Implement I2C block support") Reported-by: Xiang Mei Signed-off-by: Weiming Shi Reviewed-by: Jean Delvare Signed-off-by: Wolfram Sang --- drivers/i2c/i2c-stub.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/i2c/i2c-stub.c b/drivers/i2c/i2c-stub.c index fbb0db41b10e..04314e3ed24c 100644 --- a/drivers/i2c/i2c-stub.c +++ b/drivers/i2c/i2c-stub.c @@ -214,6 +214,11 @@ static s32 stub_xfer(struct i2c_adapter *adap, u16 addr, unsigned short flags, * We ignore banks here, because banked chips don't use I2C * block transfers */ + if (data->block[0] == 0 || + data->block[0] > I2C_SMBUS_BLOCK_MAX) { + ret = -EINVAL; + break; + } if (data->block[0] > 256 - command) /* Avoid overrun */ data->block[0] = 256 - command; len = data->block[0]; From 359b626d362127451dbe8a687ac5c240f896ae2e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?C=C3=A1ssio=20Gabriel?= Date: Fri, 1 May 2026 14:45:14 -0300 Subject: [PATCH 243/525] ALSA: pcmtest: Return -EFAULT on pattern read copy failure MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit pattern_write() reports -EFAULT when copy_from_user() fails, but pattern_read() converts copy_to_user() failures into a zero-length read. That makes a userspace buffer fault look like EOF instead of reporting the actual error. Return -EFAULT from pattern_read() when copying the pattern data to userspace fails, and update the file offset only after a successful copy. Fixes: 315a3d57c64c ("ALSA: Implement the new Virtual PCM Test Driver") Signed-off-by: Cássio Gabriel Link: https://patch.msgid.link/20260501-alsa-pcmtest-pattern-read-efault-v1-1-53e1e8c11dda@gmail.com Signed-off-by: Takashi Iwai --- sound/drivers/pcmtest.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/sound/drivers/pcmtest.c b/sound/drivers/pcmtest.c index 5bfec4c7bf71..7f93557b51ec 100644 --- a/sound/drivers/pcmtest.c +++ b/sound/drivers/pcmtest.c @@ -679,9 +679,9 @@ static ssize_t pattern_read(struct file *file, char __user *u_buff, size_t len, return 0; if (copy_to_user(u_buff, patt_buf->buf + *off, to_read)) - to_read = 0; - else - *off += to_read; + return -EFAULT; + + *off += to_read; return to_read; } From e366ce8b22ec68189ffea2bb8009f7b20d549b0f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig=20=28The=20Capable=20Hub=29?= Date: Thu, 30 Apr 2026 17:45:24 +0200 Subject: [PATCH 244/525] ASoC: codecs: ab8500: Remove suspicious code MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit anc_configure() passed values from drvdata->anc_fir_values[], drvdata->anc_iir_values[] and drvdata->sid_fir_values[] as register offset to snd_soc_component_read(). The content of these arrays are user controllable via the component controls "ANC FIR Coefficients", "ANC IIR Coefficients" and "Sidetone FIR Coefficients" which I assume are supposed to hold register values, not register offsets. Without a datasheet for that component and given that before commit a201aef1a88b ("ASoC: codecs: ab8500: Fix casting of private data") the arrays overlapped with driver control structures and thus didn't work properly since 2012, drop that functionality and let someone repair it who has an actual need for it. With the core functionally removed several code parts become essentially unused and are removed, too. Reported-by: Sashiko (gemini/gemini-3.1-pro-preview) Link: https://sashiko.dev/#/patchset/20260428192255.2294705-2-u.kleine-koenig%40baylibre.com Fixes: 679d7abdc754 ("ASoC: codecs: Add AB8500 codec-driver") Signed-off-by: Uwe Kleine-König (The Capable Hub) Link: https://patch.msgid.link/20260430154524.338912-2-u.kleine-koenig@baylibre.com Signed-off-by: Mark Brown --- sound/soc/codecs/ab8500-codec.c | 304 +------------------------------- 1 file changed, 3 insertions(+), 301 deletions(-) diff --git a/sound/soc/codecs/ab8500-codec.c b/sound/soc/codecs/ab8500-codec.c index 8ab2e60f80b4..6e8ef9cd1b31 100644 --- a/sound/soc/codecs/ab8500-codec.c +++ b/sound/soc/codecs/ab8500-codec.c @@ -60,19 +60,6 @@ low before proceeding with the configuration sequence */ #define AB8500_ANC_SM_DELAY 2000 -#define AB8500_FILTER_CONTROL(xname, xcount, xmin, xmax) \ -{ .iface = SNDRV_CTL_ELEM_IFACE_MIXER, .name = (xname), \ - .info = filter_control_info, \ - .get = filter_control_get, .put = filter_control_put, \ - .private_value = (unsigned long)&(struct filter_control) \ - {.count = xcount, .min = xmin, .max = xmax} } - -struct filter_control { - long min, max; - unsigned int count; - long value[128]; -}; - /* Sidetone states */ static const char * const enum_sid_state[] = { "Unconfigured", @@ -85,45 +72,13 @@ enum sid_state { SID_FIR_CONFIGURED = 2, }; -static const char * const enum_anc_state[] = { - "Unconfigured", - "Apply FIR and IIR", - "FIR and IIR are configured", - "Apply FIR", - "FIR is configured", - "Apply IIR", - "IIR is configured" -}; -enum anc_state { - ANC_UNCONFIGURED = 0, - ANC_APPLY_FIR_IIR = 1, - ANC_FIR_IIR_CONFIGURED = 2, - ANC_APPLY_FIR = 3, - ANC_FIR_CONFIGURED = 4, - ANC_APPLY_IIR = 5, - ANC_IIR_CONFIGURED = 6 -}; - -/* Analog microphones */ -enum amic_idx { - AMIC_IDX_1A, - AMIC_IDX_1B, - AMIC_IDX_2 -}; - /* Private data for AB8500 device-driver */ struct ab8500_codec_drvdata { struct regmap *regmap; struct mutex ctrl_lock; /* Sidetone */ - long *sid_fir_values; enum sid_state sid_status; - - /* ANC */ - long *anc_fir_values; - long *anc_iir_values; - enum anc_state anc_status; }; static inline const char *amic_micbias_str(enum amic_micbias micbias) @@ -1024,89 +979,6 @@ static const struct snd_soc_dapm_route ab8500_dapm_routes_mic2_vamicx[] = { {"MIC2 V-AMICx Enable", NULL, "V-AMIC2"}, }; -/* ANC FIR-coefficients configuration sequence */ -static void anc_fir(struct snd_soc_component *component, - unsigned int bnk, unsigned int par, unsigned int val) -{ - if (par == 0 && bnk == 0) - snd_soc_component_update_bits(component, AB8500_ANCCONF1, - BIT(AB8500_ANCCONF1_ANCFIRUPDATE), - BIT(AB8500_ANCCONF1_ANCFIRUPDATE)); - - snd_soc_component_write(component, AB8500_ANCCONF5, val >> 8 & 0xff); - snd_soc_component_write(component, AB8500_ANCCONF6, val & 0xff); - - if (par == AB8500_ANC_FIR_COEFFS - 1 && bnk == 1) - snd_soc_component_update_bits(component, AB8500_ANCCONF1, - BIT(AB8500_ANCCONF1_ANCFIRUPDATE), 0); -} - -/* ANC IIR-coefficients configuration sequence */ -static void anc_iir(struct snd_soc_component *component, unsigned int bnk, - unsigned int par, unsigned int val) -{ - if (par == 0) { - if (bnk == 0) { - snd_soc_component_update_bits(component, AB8500_ANCCONF1, - BIT(AB8500_ANCCONF1_ANCIIRINIT), - BIT(AB8500_ANCCONF1_ANCIIRINIT)); - usleep_range(AB8500_ANC_SM_DELAY, AB8500_ANC_SM_DELAY*2); - snd_soc_component_update_bits(component, AB8500_ANCCONF1, - BIT(AB8500_ANCCONF1_ANCIIRINIT), 0); - usleep_range(AB8500_ANC_SM_DELAY, AB8500_ANC_SM_DELAY*2); - } else { - snd_soc_component_update_bits(component, AB8500_ANCCONF1, - BIT(AB8500_ANCCONF1_ANCIIRUPDATE), - BIT(AB8500_ANCCONF1_ANCIIRUPDATE)); - } - } else if (par > 3) { - snd_soc_component_write(component, AB8500_ANCCONF7, 0); - snd_soc_component_write(component, AB8500_ANCCONF8, val >> 16 & 0xff); - } - - snd_soc_component_write(component, AB8500_ANCCONF7, val >> 8 & 0xff); - snd_soc_component_write(component, AB8500_ANCCONF8, val & 0xff); - - if (par == AB8500_ANC_IIR_COEFFS - 1 && bnk == 1) - snd_soc_component_update_bits(component, AB8500_ANCCONF1, - BIT(AB8500_ANCCONF1_ANCIIRUPDATE), 0); -} - -/* ANC IIR-/FIR-coefficients configuration sequence */ -static void anc_configure(struct snd_soc_component *component, - bool apply_fir, bool apply_iir) -{ - struct ab8500_codec_drvdata *drvdata = dev_get_drvdata(component->dev); - unsigned int bnk, par, val; - - dev_dbg(component->dev, "%s: Enter.\n", __func__); - - if (apply_fir) - snd_soc_component_update_bits(component, AB8500_ANCCONF1, - BIT(AB8500_ANCCONF1_ENANC), 0); - - snd_soc_component_update_bits(component, AB8500_ANCCONF1, - BIT(AB8500_ANCCONF1_ENANC), BIT(AB8500_ANCCONF1_ENANC)); - - if (apply_fir) - for (bnk = 0; bnk < AB8500_NR_OF_ANC_COEFF_BANKS; bnk++) - for (par = 0; par < AB8500_ANC_FIR_COEFFS; par++) { - val = snd_soc_component_read(component, - drvdata->anc_fir_values[par]); - anc_fir(component, bnk, par, val); - } - - if (apply_iir) - for (bnk = 0; bnk < AB8500_NR_OF_ANC_COEFF_BANKS; bnk++) - for (par = 0; par < AB8500_ANC_IIR_COEFFS; par++) { - val = snd_soc_component_read(component, - drvdata->anc_iir_values[par]); - anc_iir(component, bnk, par, val); - } - - dev_dbg(component->dev, "%s: Exit.\n", __func__); -} - /* * Control-events */ @@ -1130,7 +1002,7 @@ static int sid_status_control_put(struct snd_kcontrol *kcontrol, { struct snd_soc_component *component = snd_kcontrol_chip(kcontrol); struct ab8500_codec_drvdata *drvdata = dev_get_drvdata(component->dev); - unsigned int param, sidconf, val; + unsigned int param, sidconf; int status = 1; dev_dbg(component->dev, "%s: Enter\n", __func__); @@ -1159,9 +1031,8 @@ static int sid_status_control_put(struct snd_kcontrol *kcontrol, snd_soc_component_write(component, AB8500_SIDFIRADR, 0); for (param = 0; param < AB8500_SID_FIR_COEFFS; param++) { - val = snd_soc_component_read(component, drvdata->sid_fir_values[param]); - snd_soc_component_write(component, AB8500_SIDFIRCOEF1, val >> 8 & 0xff); - snd_soc_component_write(component, AB8500_SIDFIRCOEF2, val & 0xff); + snd_soc_component_write(component, AB8500_SIDFIRCOEF1, 0); + snd_soc_component_write(component, AB8500_SIDFIRCOEF2, 0); } snd_soc_component_update_bits(component, AB8500_SIDFIRADR, @@ -1180,136 +1051,6 @@ static int sid_status_control_put(struct snd_kcontrol *kcontrol, return status; } -static int anc_status_control_get(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct snd_soc_component *component = snd_kcontrol_chip(kcontrol); - struct ab8500_codec_drvdata *drvdata = dev_get_drvdata(component->dev); - - mutex_lock(&drvdata->ctrl_lock); - ucontrol->value.enumerated.item[0] = drvdata->anc_status; - mutex_unlock(&drvdata->ctrl_lock); - - return 0; -} - -static int anc_status_control_put(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct snd_soc_component *component = snd_kcontrol_chip(kcontrol); - struct snd_soc_dapm_context *dapm = snd_soc_component_to_dapm(component); - struct ab8500_codec_drvdata *drvdata = dev_get_drvdata(component->dev); - struct device *dev = component->dev; - bool apply_fir, apply_iir; - unsigned int req; - int status; - - dev_dbg(dev, "%s: Enter.\n", __func__); - - mutex_lock(&drvdata->ctrl_lock); - - req = ucontrol->value.enumerated.item[0]; - if (req >= ARRAY_SIZE(enum_anc_state)) { - status = -EINVAL; - goto cleanup; - } - if (req != ANC_APPLY_FIR_IIR && req != ANC_APPLY_FIR && - req != ANC_APPLY_IIR) { - dev_err(dev, "%s: ERROR: Unsupported status to set '%s'!\n", - __func__, enum_anc_state[req]); - status = -EINVAL; - goto cleanup; - } - apply_fir = req == ANC_APPLY_FIR || req == ANC_APPLY_FIR_IIR; - apply_iir = req == ANC_APPLY_IIR || req == ANC_APPLY_FIR_IIR; - - status = snd_soc_dapm_force_enable_pin(dapm, "ANC Configure Input"); - if (status < 0) { - dev_err(dev, - "%s: ERROR: Failed to enable power (status = %d)!\n", - __func__, status); - goto cleanup; - } - snd_soc_dapm_sync(dapm); - - anc_configure(component, apply_fir, apply_iir); - - if (apply_fir) { - if (drvdata->anc_status == ANC_IIR_CONFIGURED) - drvdata->anc_status = ANC_FIR_IIR_CONFIGURED; - else if (drvdata->anc_status != ANC_FIR_IIR_CONFIGURED) - drvdata->anc_status = ANC_FIR_CONFIGURED; - } - if (apply_iir) { - if (drvdata->anc_status == ANC_FIR_CONFIGURED) - drvdata->anc_status = ANC_FIR_IIR_CONFIGURED; - else if (drvdata->anc_status != ANC_FIR_IIR_CONFIGURED) - drvdata->anc_status = ANC_IIR_CONFIGURED; - } - - status = snd_soc_dapm_disable_pin(dapm, "ANC Configure Input"); - snd_soc_dapm_sync(dapm); - -cleanup: - mutex_unlock(&drvdata->ctrl_lock); - - if (status < 0) - dev_err(dev, "%s: Unable to configure ANC! (status = %d)\n", - __func__, status); - - dev_dbg(dev, "%s: Exit.\n", __func__); - - return (status < 0) ? status : 1; -} - -static int filter_control_info(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_info *uinfo) -{ - struct filter_control *fc = - (struct filter_control *)kcontrol->private_value; - - uinfo->type = SNDRV_CTL_ELEM_TYPE_INTEGER; - uinfo->count = fc->count; - uinfo->value.integer.min = fc->min; - uinfo->value.integer.max = fc->max; - - return 0; -} - -static int filter_control_get(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct snd_soc_component *component = snd_kcontrol_chip(kcontrol); - struct ab8500_codec_drvdata *drvdata = snd_soc_component_get_drvdata(component); - struct filter_control *fc = - (struct filter_control *)kcontrol->private_value; - unsigned int i; - - mutex_lock(&drvdata->ctrl_lock); - for (i = 0; i < fc->count; i++) - ucontrol->value.integer.value[i] = fc->value[i]; - mutex_unlock(&drvdata->ctrl_lock); - - return 0; -} - -static int filter_control_put(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct snd_soc_component *component = snd_kcontrol_chip(kcontrol); - struct ab8500_codec_drvdata *drvdata = snd_soc_component_get_drvdata(component); - struct filter_control *fc = - (struct filter_control *)kcontrol->private_value; - unsigned int i; - - mutex_lock(&drvdata->ctrl_lock); - for (i = 0; i < fc->count; i++) - fc->value[i] = ucontrol->value.integer.value[i]; - mutex_unlock(&drvdata->ctrl_lock); - - return 0; -} - /* * Controls - Non-DAPM ASoC */ @@ -1597,7 +1338,6 @@ static SOC_ENUM_SINGLE_DECL(soc_enum_bfifomast, static SOC_ENUM_SINGLE_EXT_DECL(soc_enum_sidstate, enum_sid_state); /* ANC */ -static SOC_ENUM_SINGLE_EXT_DECL(soc_enum_ancstate, enum_anc_state); static struct snd_kcontrol_new ab8500_ctrls[] = { /* Charge pump */ @@ -1873,8 +1613,6 @@ static struct snd_kcontrol_new ab8500_ctrls[] = { AB8500_FIFOCONF6_BFIFOSAMPLE_MAX, 0), /* ANC */ - SOC_ENUM_EXT("ANC Status", soc_enum_ancstate, - anc_status_control_get, anc_status_control_put), SOC_SINGLE_XR_SX("ANC Warp Delay Shift", AB8500_ANCCONF2, 1, AB8500_ANCCONF2_SHIFT, AB8500_ANCCONF2_MIN, AB8500_ANCCONF2_MAX, 0), @@ -1895,21 +1633,6 @@ static struct snd_kcontrol_new ab8500_ctrls[] = { AB8500_SIDFIRADR, AB8500_SIDFIRADR_FIRSIDSET, 0), }; -static struct snd_kcontrol_new ab8500_filter_controls[] = { - AB8500_FILTER_CONTROL("ANC FIR Coefficients", AB8500_ANC_FIR_COEFFS, - AB8500_ANC_FIR_COEFF_MIN, AB8500_ANC_FIR_COEFF_MAX), - AB8500_FILTER_CONTROL("ANC IIR Coefficients", AB8500_ANC_IIR_COEFFS, - AB8500_ANC_IIR_COEFF_MIN, AB8500_ANC_IIR_COEFF_MAX), - AB8500_FILTER_CONTROL("Sidetone FIR Coefficients", - AB8500_SID_FIR_COEFFS, AB8500_SID_FIR_COEFF_MIN, - AB8500_SID_FIR_COEFF_MAX) -}; -enum ab8500_filter { - AB8500_FILTER_ANC_FIR = 0, - AB8500_FILTER_ANC_IIR = 1, - AB8500_FILTER_SID_FIR = 2, -}; - /* * Extended interface for codec-driver */ @@ -2454,7 +2177,6 @@ static int ab8500_codec_probe(struct snd_soc_component *component) struct device_node *np = dev->of_node; struct ab8500_codec_drvdata *drvdata = dev_get_drvdata(dev); struct ab8500_codec_platform_data codec_pdata; - struct filter_control *fc; int status; dev_dbg(dev, "%s: Enter.\n", __func__); @@ -2486,25 +2208,6 @@ static int ab8500_codec_probe(struct snd_soc_component *component) snd_soc_component_write(component, AB8500_SHORTCIRCONF, BIT(AB8500_SHORTCIRCONF_HSZCDDIS)); - /* Add filter controls */ - status = snd_soc_add_component_controls(component, ab8500_filter_controls, - ARRAY_SIZE(ab8500_filter_controls)); - if (status < 0) { - dev_err(dev, - "%s: failed to add ab8500 filter controls (%d).\n", - __func__, status); - return status; - } - fc = (struct filter_control *) - ab8500_filter_controls[AB8500_FILTER_ANC_FIR].private_value; - drvdata->anc_fir_values = (long *)fc->value; - fc = (struct filter_control *) - ab8500_filter_controls[AB8500_FILTER_ANC_IIR].private_value; - drvdata->anc_iir_values = (long *)fc->value; - fc = (struct filter_control *) - ab8500_filter_controls[AB8500_FILTER_SID_FIR].private_value; - drvdata->sid_fir_values = (long *)fc->value; - snd_soc_dapm_disable_pin(dapm, "ANC Configure Input"); mutex_init(&drvdata->ctrl_lock); @@ -2538,7 +2241,6 @@ static int ab8500_codec_driver_probe(struct platform_device *pdev) if (!drvdata) return -ENOMEM; drvdata->sid_status = SID_UNCONFIGURED; - drvdata->anc_status = ANC_UNCONFIGURED; dev_set_drvdata(&pdev->dev, drvdata); drvdata->regmap = devm_regmap_init(&pdev->dev, NULL, &pdev->dev, From 8acd2d7e0889ac62bc102bd7b648cd7bee04f902 Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Fri, 24 Apr 2026 20:25:18 +0900 Subject: [PATCH 245/525] drm/qxl: Fix missing KMS poll cleanup drm_kms_helper_poll_init() initializes the output polling work and enables polling for the DRM device. qxl enables polling before calling drm_dev_register(), but the drm_dev_register() failure path tears down the modeset and device state without disabling the polling helper. The remove path also unregisters and shuts down the DRM device without first disabling the polling helper. Add matching drm_kms_helper_poll_fini() calls in both paths so the delayed polling work is cancelled before qxl tears down the associated modeset/device state. Signed-off-by: Myeonghun Pak Reviewed-by: Thomas Zimmermann Fixes: 5ff91e442652 ("qxl: use drm helper hotplug support") Signed-off-by: Thomas Zimmermann Link: https://patch.msgid.link/20260424112543.57819-1-mhun512@gmail.com --- drivers/gpu/drm/qxl/qxl_drv.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/qxl/qxl_drv.c b/drivers/gpu/drm/qxl/qxl_drv.c index 2bbb1168a3ff..1e6a2392d7c6 100644 --- a/drivers/gpu/drm/qxl/qxl_drv.c +++ b/drivers/gpu/drm/qxl/qxl_drv.c @@ -118,12 +118,13 @@ qxl_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent) /* Complete initialization. */ ret = drm_dev_register(&qdev->ddev, ent->driver_data); if (ret) - goto modeset_cleanup; + goto poll_fini; drm_client_setup(&qdev->ddev, NULL); return 0; -modeset_cleanup: +poll_fini: + drm_kms_helper_poll_fini(&qdev->ddev); qxl_modeset_fini(qdev); unload: qxl_device_fini(qdev); @@ -154,6 +155,7 @@ qxl_pci_remove(struct pci_dev *pdev) { struct drm_device *dev = pci_get_drvdata(pdev); + drm_kms_helper_poll_fini(dev); drm_dev_unregister(dev); drm_atomic_helper_shutdown(dev); if (pci_is_vga(pdev) && pdev->revision < 5) From c28c22c8cfbd43f2ad71a157324d9fbebc0d0f2e Mon Sep 17 00:00:00 2001 From: Francesco Lavra Date: Tue, 10 Feb 2026 18:35:45 +0100 Subject: [PATCH 246/525] drm/fb-helper: Fix clipping when damage area spans a single scanline When the damage area resulting from a dirty memory range spans a single scanline, the width of the rectangle is calculated dynamically because it may not coincide with the framebuffer width. If the dirty range ends exactly at the end of the scanline, the `bit_end` variable is incorrectly assigned a 0 value, which results in a bogus clip rectangle where the x2 coordinate is 0. This prevents the dirty scanline from being flushed to the hardware. Change the calculation of the `bit_end` value to fix the x2 coordinate value in the above edge case. Fixes: ded74cafeea9 ("drm/fb-helper: Clip damage area horizontally") Signed-off-by: Francesco Lavra Reviewed-by: Thomas Zimmermann Signed-off-by: Thomas Zimmermann Link: https://patch.msgid.link/20260210173545.733937-1-flavra@baylibre.com --- drivers/gpu/drm/drm_fb_helper.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c index a80a335f4148..1541fc8a9ac2 100644 --- a/drivers/gpu/drm/drm_fb_helper.c +++ b/drivers/gpu/drm/drm_fb_helper.c @@ -490,7 +490,7 @@ static void drm_fb_helper_memory_range_to_clip(struct fb_info *info, off_t off, * the number of horizontal pixels that need an update. */ off_t bit_off = (off % line_length) * 8; - off_t bit_end = (end % line_length) * 8; + off_t bit_end = bit_off + len * 8; x1 = bit_off / info->var.bits_per_pixel; x2 = DIV_ROUND_UP(bit_end, info->var.bits_per_pixel); From ec0611868f2fcf29e4c2bebdc6702d3e1f272fec Mon Sep 17 00:00:00 2001 From: Troy Mitchell Date: Wed, 29 Apr 2026 17:00:50 +0800 Subject: [PATCH 247/525] ASoC: spacemit: fix RX DMA params not set when TX is running When TX is already running (SSCR_SSE is set), the hw_params callback returns early before setting up DMA parameters for the RX stream. This prevents the capture path from configuring its DMA data properly. Move the SSCR_SSE check after DMA parameter setup and format constraints, so both TX and RX streams get their DMA configuration regardless of whether the hardware is already enabled. The early return now only skips the register writes that would disrupt an active stream. Fixes: fce217449075 ("ASoC: spacemit: add i2s support for K1 SoC") Signed-off-by: Troy Mitchell Link: https://patch.msgid.link/20260429-k1-i2s-fix-v2-1-8d67835aaddc@linux.spacemit.com Signed-off-by: Mark Brown --- sound/soc/spacemit/k1_i2s.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/sound/soc/spacemit/k1_i2s.c b/sound/soc/spacemit/k1_i2s.c index 43481f387c44..5420ca2aefbd 100644 --- a/sound/soc/spacemit/k1_i2s.c +++ b/sound/soc/spacemit/k1_i2s.c @@ -148,10 +148,6 @@ static int spacemit_i2s_hw_params(struct snd_pcm_substream *substream, u32 val; int ret; - val = readl(i2s->base + SSCR); - if (val & SSCR_SSE) - return 0; - dma_data = &i2s->playback_dma_data; if (substream->stream == SNDRV_PCM_STREAM_CAPTURE) @@ -199,6 +195,9 @@ static int spacemit_i2s_hw_params(struct snd_pcm_substream *substream, } val = readl(i2s->base + SSCR); + if (val & SSCR_SSE) + return 0; + val &= ~SSCR_DW_32BYTE; val |= data_width; writel(val, i2s->base + SSCR); From c64e77490b7e5d9dec738850f18878edb07e0f13 Mon Sep 17 00:00:00 2001 From: Richard Fitzgerald Date: Wed, 29 Apr 2026 11:53:15 +0100 Subject: [PATCH 248/525] ASoC: cs35l56: Fix hibernate write in runtime resume error path The error path of cs35l56_runtime_resume_common() should only write the hibernation sequence if can_hibernate is true. Something has already gone badly wrong if we ever reach the error path. But triggering hibernate on hardware that does not support it is likely to make the situation unrecoverable without a full reboot because there might not be any hardware signal to exit hibernate. Fixes: a47cf4dac7dc ("ASoC: cs35l56: Change hibernate sequence to use allow auto hibernate") Signed-off-by: Richard Fitzgerald Link: https://patch.msgid.link/20260429105315.2438298-1-rf@opensource.cirrus.com Signed-off-by: Mark Brown --- sound/soc/codecs/cs35l56-shared.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/sound/soc/codecs/cs35l56-shared.c b/sound/soc/codecs/cs35l56-shared.c index 033e56d5e9db..ea724101cdd1 100644 --- a/sound/soc/codecs/cs35l56-shared.c +++ b/sound/soc/codecs/cs35l56-shared.c @@ -851,9 +851,11 @@ int cs35l56_runtime_resume_common(struct cs35l56_base *cs35l56_base, bool is_sou err: regcache_cache_only(cs35l56_base->regmap, true); - regmap_multi_reg_write_bypassed(cs35l56_base->regmap, - cs35l56_hibernate_seq, - ARRAY_SIZE(cs35l56_hibernate_seq)); + if (cs35l56_base->can_hibernate) { + regmap_multi_reg_write_bypassed(cs35l56_base->regmap, + cs35l56_hibernate_seq, + ARRAY_SIZE(cs35l56_hibernate_seq)); + } return ret; } From 0e60d96616640ffcf51b81a87c71e30d92385a93 Mon Sep 17 00:00:00 2001 From: Bob Song Date: Thu, 30 Apr 2026 09:49:20 +0800 Subject: [PATCH 249/525] ASoC: amd: yc: Add DMI quirk for MSI Bravo 15 C7VE The laptop requires a quirk ID to enable its internal microphone. Add it to the DMI quirk table. Reported-by: gannovera Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218402 Signed-off-by: Bob Song Link: https://patch.msgid.link/20260430014920.141276-1-songxiebing@kylinos.cn Signed-off-by: Mark Brown --- sound/soc/amd/yc/acp6x-mach.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/sound/soc/amd/yc/acp6x-mach.c b/sound/soc/amd/yc/acp6x-mach.c index c5cf45881416..dde1a144e4d4 100644 --- a/sound/soc/amd/yc/acp6x-mach.c +++ b/sound/soc/amd/yc/acp6x-mach.c @@ -479,6 +479,13 @@ static const struct dmi_system_id yc_acp_quirk_table[] = { DMI_MATCH(DMI_PRODUCT_NAME, "Bravo 15 B7ED"), } }, + { + .driver_data = &acp6x_card, + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "Micro-Star International Co., Ltd."), + DMI_MATCH(DMI_PRODUCT_NAME, "Bravo 15 C7VE"), + } + }, { .driver_data = &acp6x_card, .matches = { From 0f9bfb84b3f0fe1406b2555fc11b45283ea21644 Mon Sep 17 00:00:00 2001 From: Richard Fitzgerald Date: Thu, 30 Apr 2026 11:11:34 +0100 Subject: [PATCH 250/525] ASoC: cs35l56: Fix out-of-bounds in dev_err() in cs35l56_read_onchip_spkid() Remove the incorrect use of onchip_spkid_gpios[i] in the dev_err() after regmap_read() of CS35L56_GPIO_STATUS1 returns an error. This dev_err() was incorrectly copy-pasted from one inside the for-loop, where i was valid. The read of CS35L56_GPIO_STATUS1 isn't for a specific GPIO register, so the use of onchip_spkid_gpios[i] in the error message is both irrelevant and out-of-bounds here. Fixes: 4d1e3e2c404d ("ASoC: cs35l56: Support for reading speaker ID from on-chip GPIOs") Signed-off-by: Richard Fitzgerald Link: https://patch.msgid.link/20260430101134.2655938-1-rf@opensource.cirrus.com Signed-off-by: Mark Brown --- sound/soc/codecs/cs35l56-shared.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/sound/soc/codecs/cs35l56-shared.c b/sound/soc/codecs/cs35l56-shared.c index ea724101cdd1..795e2764d67e 100644 --- a/sound/soc/codecs/cs35l56-shared.c +++ b/sound/soc/codecs/cs35l56-shared.c @@ -1730,8 +1730,7 @@ int cs35l56_read_onchip_spkid(struct cs35l56_base *cs35l56_base) ret = regmap_read(regmap, CS35L56_GPIO_STATUS1, &val); if (ret) { - dev_err(cs35l56_base->dev, "GPIO%d status read failed: %d\n", - cs35l56_base->onchip_spkid_gpios[i] + 1, ret); + dev_err(cs35l56_base->dev, "GPIO status read failed: %d\n", ret); return ret; } From d63c219b7ff39f897da10c160a2edef76320f16c Mon Sep 17 00:00:00 2001 From: Tommaso Soncin Date: Wed, 29 Apr 2026 18:08:57 +0200 Subject: [PATCH 251/525] ASoC: amd: yc: Add HP OMEN Gaming Laptop 16-ap0xxx product line in quirk table Add a DMI quirk for the HP OMEN Gaming Laptop 16-ap0xxx line fixing the issue where the internal microphone was not detected. Cc: stable@vger.kernel.org Signed-off-by: Tommaso Soncin Link: https://patch.msgid.link/20260429160858.538986-1-soncintommaso@gmail.com Signed-off-by: Mark Brown --- sound/soc/amd/yc/acp6x-mach.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/sound/soc/amd/yc/acp6x-mach.c b/sound/soc/amd/yc/acp6x-mach.c index dde1a144e4d4..7a637d6b5576 100644 --- a/sound/soc/amd/yc/acp6x-mach.c +++ b/sound/soc/amd/yc/acp6x-mach.c @@ -59,6 +59,13 @@ static const struct dmi_system_id yc_acp_quirk_table[] = { DMI_MATCH(DMI_PRODUCT_NAME, "HP Laptop 15-fc0xxx"), } }, + { + .driver_data = &acp6x_card, + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "HP"), + DMI_MATCH(DMI_PRODUCT_NAME, "OMEN Gaming Laptop 16-ap0xxx"), + } + }, { .driver_data = &acp6x_card, .matches = { @@ -675,6 +682,13 @@ static const struct dmi_system_id yc_acp_quirk_table[] = { DMI_MATCH(DMI_BOARD_NAME, "8EE4"), } }, + { + .driver_data = &acp6x_card, + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "HP"), + DMI_MATCH(DMI_BOARD_NAME, "8E35"), + } + }, { .driver_data = &acp6x_card, .matches = { From 56d5a9eaf60af5c824a33a83e1468aa143627a62 Mon Sep 17 00:00:00 2001 From: Derek Fang Date: Thu, 30 Apr 2026 20:10:43 +0800 Subject: [PATCH 252/525] ASoC: sdw_utils: avoid the SDCA companion function not supported failure Treat the companion amp as generic AMP until full support for companion amp is added. Signed-off-by: Derek Fang Reviewed-by: Charles Keepax Signed-off-by: Bard Liao Link: https://patch.msgid.link/20260430121043.552241-1-yung-chuan.liao@linux.intel.com Signed-off-by: Mark Brown --- sound/soc/sdw_utils/soc_sdw_utils.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/sdw_utils/soc_sdw_utils.c b/sound/soc/sdw_utils/soc_sdw_utils.c index 1637cc3f3d59..849ae876d7a4 100644 --- a/sound/soc/sdw_utils/soc_sdw_utils.c +++ b/sound/soc/sdw_utils/soc_sdw_utils.c @@ -1608,6 +1608,7 @@ int asoc_sdw_get_dai_type(u32 type) switch (type) { case SDCA_FUNCTION_TYPE_SMART_AMP: case SDCA_FUNCTION_TYPE_SIMPLE_AMP: + case SDCA_FUNCTION_TYPE_COMPANION_AMP: return SOC_SDW_DAI_TYPE_AMP; case SDCA_FUNCTION_TYPE_SMART_MIC: case SDCA_FUNCTION_TYPE_SIMPLE_MIC: From e8446a4a574d19f0fb39c06af15dbc5165079474 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?C=C3=A1ssio=20Gabriel?= Date: Tue, 28 Apr 2026 00:07:08 -0300 Subject: [PATCH 253/525] ASoC: fsl_xcvr: Fix event generation for cached controls MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ALSA controls should return 1 from a put callback when the control value changes. fsl_xcvr_capds_put() and fsl_xcvr_tx_cs_put() both update cached control data but always return 0, so ALSA suppresses change notifications for the Capabilities Data Structure and playback IEC958 channel status controls. Compare the old and new cached values before copying the new data, and return whether the control value changed. Fixes: 28564486866f ("ASoC: fsl_xcvr: Add XCVR ASoC CPU DAI driver") Signed-off-by: Cássio Gabriel Link: https://patch.msgid.link/20260428-asoc-fsl-xcvr-event-generation-v1-1-f21cf0812c4f@gmail.com Signed-off-by: Mark Brown --- sound/soc/fsl/fsl_xcvr.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-) diff --git a/sound/soc/fsl/fsl_xcvr.c b/sound/soc/fsl/fsl_xcvr.c index ee16cf681488..6677d3bf36ec 100644 --- a/sound/soc/fsl/fsl_xcvr.c +++ b/sound/soc/fsl/fsl_xcvr.c @@ -228,10 +228,14 @@ static int fsl_xcvr_capds_put(struct snd_kcontrol *kcontrol, { struct snd_soc_dai *dai = snd_kcontrol_chip(kcontrol); struct fsl_xcvr *xcvr = snd_soc_dai_get_drvdata(dai); + int changed; - memcpy(xcvr->cap_ds, ucontrol->value.bytes.data, FSL_XCVR_CAPDS_SIZE); + changed = memcmp(xcvr->cap_ds, ucontrol->value.bytes.data, + sizeof(xcvr->cap_ds)) != 0; + memcpy(xcvr->cap_ds, ucontrol->value.bytes.data, + sizeof(xcvr->cap_ds)); - return 0; + return changed; } static struct snd_kcontrol_new fsl_xcvr_earc_capds_kctl = { @@ -1040,10 +1044,15 @@ static int fsl_xcvr_tx_cs_put(struct snd_kcontrol *kcontrol, { struct snd_soc_dai *dai = snd_kcontrol_chip(kcontrol); struct fsl_xcvr *xcvr = snd_soc_dai_get_drvdata(dai); + int changed; - memcpy(xcvr->tx_iec958.status, ucontrol->value.iec958.status, 24); + changed = memcmp(xcvr->tx_iec958.status, + ucontrol->value.iec958.status, + sizeof(xcvr->tx_iec958.status)) != 0; + memcpy(xcvr->tx_iec958.status, ucontrol->value.iec958.status, + sizeof(xcvr->tx_iec958.status)); - return 0; + return changed; } static struct snd_kcontrol_new fsl_xcvr_rx_ctls[] = { From 24e0fd8b852062d5e8a740f7945eaa26818adce8 Mon Sep 17 00:00:00 2001 From: John Madieu Date: Fri, 1 May 2026 13:59:49 +0000 Subject: [PATCH 254/525] spi: imx: Fix precedence bug in spi_imx_dma_max_wml_find() The watermark search in spi_imx_dma_max_wml_find() reads: if (!dma_data->dma_len % (i * bytes_per_word)) break; The unary ! binds tighter than %, so this parses as: if ((!dma_data->dma_len) % (i * bytes_per_word)) break; !dma_data->dma_len is 0 or 1, and `0 % x == 0` for any x; `1 % x` is 0 unless x == 1. The condition is therefore false in every case except dma_len != 0 with i * bytes_per_word == 1, i.e. i == 1 and bytes_per_word == 1. The loop almost always falls through to its end, leaving i == 0, which the post-loop fallback rewrites to 1: if (i == 0) i = 1; So spi_imx->wml ends up at 1 for essentially every DMA transfer, defeating the entire purpose of the function. The DMA engine then requests service after every single FIFO word instead of using multi-word bursts, hurting throughput on every DMA-capable variant. Add the missing parentheses so the modulo is computed first, then negated: if (!(dma_data->dma_len % (i * bytes_per_word))) break; Fixes: faa8e404ad8e ("spi: imx: support dynamic burst length for ECSPI DMA mode") Signed-off-by: John Madieu Reviewed-by: Frank Li Link: https://patch.msgid.link/20260501135951.2416527-2-john.madieu@gmail.com Signed-off-by: Mark Brown --- drivers/spi/spi-imx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/spi/spi-imx.c b/drivers/spi/spi-imx.c index e5c907c45b87..7ae8078c10ef 100644 --- a/drivers/spi/spi-imx.c +++ b/drivers/spi/spi-imx.c @@ -1836,7 +1836,7 @@ static void spi_imx_dma_max_wml_find(struct spi_imx_data *spi_imx, unsigned int i; for (i = spi_imx->devtype_data->fifo_size / 2; i > 0; i--) { - if (!dma_data->dma_len % (i * bytes_per_word)) + if (!(dma_data->dma_len % (i * bytes_per_word))) break; } /* Use 1 as wml in case no available burst length got */ From f5b5548255040ec3bef05bcb1e9c9c3614dfa7db Mon Sep 17 00:00:00 2001 From: John Madieu Date: Fri, 1 May 2026 13:59:50 +0000 Subject: [PATCH 255/525] spi: imx: Fix UAF on package-1 prepare failure in spi_imx_dma_data_prepare() When transfer->len exceeds MX51_ECSPI_CTRL_MAX_BURST and is not a multiple of it, spi_imx_dma_data_prepare() splits the transfer into two DMA packages. If preparing the second package fails: ret = spi_imx_dma_tx_data_handle(spi_imx, &spi_imx->dma_data[1], transfer->tx_buf + spi_imx->dma_data[0].data_len, false); if (ret) { kfree(spi_imx->dma_data[0].dma_tx_buf); kfree(spi_imx->dma_data[0].dma_rx_buf); kfree(spi_imx->dma_data); } } return 0; the function frees the package-0 buffers and the dma_data array, then falls through to `return 0`, telling the caller the prepare succeeded. The caller then dereferences the freed dma_data array, producing a use-after-free. Return the error from the failure path so the caller takes its existing failure branch. Fixes: faa8e404ad8e ("spi: imx: support dynamic burst length for ECSPI DMA mode") Signed-off-by: John Madieu Reviewed-by: Frank Li Link: https://patch.msgid.link/20260501135951.2416527-3-john.madieu@gmail.com Signed-off-by: Mark Brown --- drivers/spi/spi-imx.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/spi/spi-imx.c b/drivers/spi/spi-imx.c index 7ae8078c10ef..4e3dbd01d619 100644 --- a/drivers/spi/spi-imx.c +++ b/drivers/spi/spi-imx.c @@ -1709,6 +1709,7 @@ static int spi_imx_dma_data_prepare(struct spi_imx_data *spi_imx, kfree(spi_imx->dma_data[0].dma_tx_buf); kfree(spi_imx->dma_data[0].dma_rx_buf); kfree(spi_imx->dma_data); + return ret; } } From 894e04b7116297a6529e0c4ed90e3eb160939805 Mon Sep 17 00:00:00 2001 From: John Madieu Date: Fri, 1 May 2026 13:59:51 +0000 Subject: [PATCH 256/525] spi: imx: Propagate prepare_transfer() error from spi_imx_setupxfer() spi_imx_setupxfer() calls the per-variant prepare_transfer() callback and returns 0 unconditionally: spi_imx->devtype_data->prepare_transfer(spi_imx, spi, t); return 0; mx51_ecspi_prepare_transfer() can return -EINVAL when the requested word_delay does not fit in MX51_ECSPI_PERIOD_MASK. The error is detected after a partial set of register writes (CTRL: BL, clkdiv, SMC), so the controller is left in a partially-configured state and the transfer is then submitted as if setup succeeded. Propagate the return value. The other variants' prepare_transfer callbacks all return 0, so this is a no-op for them. Fixes: a3bb4e663df3 ("spi: imx: support word delay") Signed-off-by: John Madieu Reviewed-by: Frank Li Link: https://patch.msgid.link/20260501135951.2416527-4-john.madieu@gmail.com Signed-off-by: Mark Brown --- drivers/spi/spi-imx.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/spi/spi-imx.c b/drivers/spi/spi-imx.c index 4e3dbd01d619..480d1e8b281f 100644 --- a/drivers/spi/spi-imx.c +++ b/drivers/spi/spi-imx.c @@ -1382,9 +1382,7 @@ static int spi_imx_setupxfer(struct spi_device *spi, spi_imx->target_burst = t->len; } - spi_imx->devtype_data->prepare_transfer(spi_imx, spi, t); - - return 0; + return spi_imx->devtype_data->prepare_transfer(spi_imx, spi, t); } static void spi_imx_sdma_exit(struct spi_imx_data *spi_imx) From 7672749e1496215e8683ce57cf323119033954cf Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Thu, 30 Apr 2026 11:10:18 +0100 Subject: [PATCH 257/525] spi: microchip-core-qspi: control built-in cs manually The coreQSPI IP supports only a single chip select, which is automagically operated by the hardware - set low when the transmit buffer first gets written to and set high when the number of bytes written to the TOTALBYTES field of the FRAMES register have been sent on the bus. Additional devices must use GPIOs for their chip selects. It was reported to me that if there are two devices attached to this QSPI controller that the in-built chip select is set low while linux tries to access the device attached to the GPIO. This went undetected as the boards that connected multiple devices to the SPI controller all exclusively used GPIOs for chip selects, not relying on the built-in chip select at all. It turns out that this was because the built-in chip select, when controlled automagically, is set low when active and high when inactive, thereby ruling out its use for active-high devices or devices that need to transmit with the chip select disabled. Modify the driver so that it controls chip select directly, retaining the behaviour for mem_ops of setting the chip select active for the entire duration of the transfer in the exec_op callback. For regular transfers, implement the set_cs callback for the core to use. As part of this, the existing setup callback, mchp_coreqspi_setup_op(), is removed. Modifying the CLKIDLE field is not safe to do during operation when there are multiple devices, so this code is removed entirely. Setting the MASTER and ENABLE fields is something that can be done once at probe, it doesn't need to be re-run for each device. Instead the new setup callback sets the built-in chip select to its inactive state for active-low devices, as the reset value of the chip select in software controlled mode is low. Fixes: 8f9cf02c88528 ("spi: microchip-core-qspi: Add regular transfers") Fixes: 8596124c4c1bc ("spi: microchip-core-qspi: Add support for microchip fpga qspi controllers") CC: stable@vger.kernel.org Signed-off-by: Conor Dooley Link: https://patch.msgid.link/20260430-hamstring-busload-f941d0347b5e@spud Signed-off-by: Mark Brown --- drivers/spi/spi-microchip-core-qspi.c | 79 ++++++++++++++++++++++----- 1 file changed, 64 insertions(+), 15 deletions(-) diff --git a/drivers/spi/spi-microchip-core-qspi.c b/drivers/spi/spi-microchip-core-qspi.c index eab059fb0bc2..ffa0f33a0ae0 100644 --- a/drivers/spi/spi-microchip-core-qspi.c +++ b/drivers/spi/spi-microchip-core-qspi.c @@ -74,6 +74,13 @@ #define STATUS_FLAGSX4 BIT(8) #define STATUS_MASK GENMASK(8, 0) +/* + * QSPI Direct Access register defines + */ +#define DIRECT_ACCESS_EN_SSEL BIT(0) +#define DIRECT_ACCESS_OP_SSEL BIT(1) +#define DIRECT_ACCESS_OP_SSEL_SHIFT 1 + #define BYTESUPPER_MASK GENMASK(31, 16) #define BYTESLOWER_MASK GENMASK(15, 0) @@ -158,6 +165,38 @@ static int mchp_coreqspi_set_mode(struct mchp_coreqspi *qspi, const struct spi_m return 0; } +static void mchp_coreqspi_set_cs(struct spi_device *spi, bool enable) +{ + struct mchp_coreqspi *qspi = spi_controller_get_devdata(spi->controller); + u32 val; + + val = readl(qspi->regs + REG_DIRECT_ACCESS); + + val &= ~DIRECT_ACCESS_OP_SSEL; + val |= !enable << DIRECT_ACCESS_OP_SSEL_SHIFT; + + writel(val, qspi->regs + REG_DIRECT_ACCESS); +} + +static int mchp_coreqspi_setup(struct spi_device *spi) +{ + struct mchp_coreqspi *qspi = spi_controller_get_devdata(spi->controller); + u32 val; + + /* + * Active low devices need to be specifically set to their inactive + * states during probe. + */ + if (spi->mode & SPI_CS_HIGH) + return 0; + + val = readl(qspi->regs + REG_DIRECT_ACCESS); + val |= DIRECT_ACCESS_OP_SSEL; + writel(val, qspi->regs + REG_DIRECT_ACCESS); + + return 0; +} + static inline void mchp_coreqspi_read_op(struct mchp_coreqspi *qspi) { u32 control, data; @@ -380,19 +419,6 @@ static int mchp_coreqspi_setup_clock(struct mchp_coreqspi *qspi, struct spi_devi return 0; } -static int mchp_coreqspi_setup_op(struct spi_device *spi_dev) -{ - struct spi_controller *ctlr = spi_dev->controller; - struct mchp_coreqspi *qspi = spi_controller_get_devdata(ctlr); - u32 control = readl_relaxed(qspi->regs + REG_CONTROL); - - control |= (CONTROL_MASTER | CONTROL_ENABLE); - control &= ~CONTROL_CLKIDLE; - writel_relaxed(control, qspi->regs + REG_CONTROL); - - return 0; -} - static inline void mchp_coreqspi_config_op(struct mchp_coreqspi *qspi, const struct spi_mem_op *op) { u32 idle_cycles = 0; @@ -483,6 +509,7 @@ static int mchp_coreqspi_exec_op(struct spi_mem *mem, const struct spi_mem_op *o reinit_completion(&qspi->data_completion); mchp_coreqspi_config_op(qspi, op); + mchp_coreqspi_set_cs(mem->spi, true); if (op->cmd.opcode) { qspi->txbuf = &opcode; qspi->rxbuf = NULL; @@ -523,6 +550,7 @@ static int mchp_coreqspi_exec_op(struct spi_mem *mem, const struct spi_mem_op *o err = -ETIMEDOUT; error: + mchp_coreqspi_set_cs(mem->spi, false); mutex_unlock(&qspi->op_lock); mchp_coreqspi_disable_ints(qspi); @@ -686,6 +714,7 @@ static int mchp_coreqspi_probe(struct platform_device *pdev) struct device *dev = &pdev->dev; struct device_node *np = dev->of_node; int ret; + u32 num_cs, val; ctlr = devm_spi_alloc_host(&pdev->dev, sizeof(*qspi)); if (!ctlr) @@ -718,10 +747,18 @@ static int mchp_coreqspi_probe(struct platform_device *pdev) return ret; } + /* + * The IP core only has a single CS, any more have to be provided via + * gpios + */ + if (of_property_read_u32(pdev->dev.of_node, "num-cs", &num_cs)) + num_cs = 1; + + ctlr->num_chipselect = num_cs; + ctlr->bits_per_word_mask = SPI_BPW_MASK(8); ctlr->mem_ops = &mchp_coreqspi_mem_ops; ctlr->mem_caps = &mchp_coreqspi_mem_caps; - ctlr->setup = mchp_coreqspi_setup_op; ctlr->mode_bits = SPI_CPOL | SPI_CPHA | SPI_RX_DUAL | SPI_RX_QUAD | SPI_TX_DUAL | SPI_TX_QUAD; ctlr->dev.of_node = np; @@ -729,9 +766,21 @@ static int mchp_coreqspi_probe(struct platform_device *pdev) ctlr->prepare_message = mchp_coreqspi_prepare_message; ctlr->unprepare_message = mchp_coreqspi_unprepare_message; ctlr->transfer_one = mchp_coreqspi_transfer_one; - ctlr->num_chipselect = 2; + ctlr->setup = mchp_coreqspi_setup; + ctlr->set_cs = mchp_coreqspi_set_cs; ctlr->use_gpio_descriptors = true; + val = readl_relaxed(qspi->regs + REG_CONTROL); + val |= (CONTROL_MASTER | CONTROL_ENABLE); + writel_relaxed(val, qspi->regs + REG_CONTROL); + + /* + * Put cs into software controlled mode + */ + val = readl_relaxed(qspi->regs + REG_DIRECT_ACCESS); + val |= DIRECT_ACCESS_EN_SSEL; + writel(val, qspi->regs + REG_DIRECT_ACCESS); + ret = spi_register_controller(ctlr); if (ret) return dev_err_probe(&pdev->dev, ret, From eb56deaabf127e8985fc91fa6c97bf8a3b062844 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Thu, 30 Apr 2026 11:10:19 +0100 Subject: [PATCH 258/525] spi: microchip-core-qspi: don't attempt to transmit during emulated read-only dual/quad operations The core will deal with reads by creating clock cycles itself, there's no need to generate clock cycles by transmitting garbage data at the driver level. Further, transmitting garbage data just bricks the transfer since QSPI doesn't have a dedicated master-out line like MOSI in regular SPI. I'm not entirely sure if the transfer is bricked because of the garbage data being transmitted on the bus or because the core loses track of whether it is supposed to be sending or receiving data. Fixes: 8f9cf02c88528 ("spi: microchip-core-qspi: Add regular transfers") CC: stable@vger.kernel.org Signed-off-by: Conor Dooley Link: https://patch.msgid.link/20260430-freezing-saloon-95b1f3d9dad0@spud Signed-off-by: Mark Brown --- drivers/spi/spi-microchip-core-qspi.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/drivers/spi/spi-microchip-core-qspi.c b/drivers/spi/spi-microchip-core-qspi.c index ffa0f33a0ae0..70215a407b5a 100644 --- a/drivers/spi/spi-microchip-core-qspi.c +++ b/drivers/spi/spi-microchip-core-qspi.c @@ -690,18 +690,28 @@ static int mchp_coreqspi_transfer_one(struct spi_controller *ctlr, struct spi_de struct spi_transfer *t) { struct mchp_coreqspi *qspi = spi_controller_get_devdata(ctlr); + bool dual_quad = false; qspi->tx_len = t->len; + if (t->tx_nbits == SPI_NBITS_QUAD || t->rx_nbits == SPI_NBITS_QUAD || + t->tx_nbits == SPI_NBITS_DUAL || + t->rx_nbits == SPI_NBITS_DUAL) + dual_quad = true; + if (t->tx_buf) qspi->txbuf = (u8 *)t->tx_buf; if (!t->rx_buf) { mchp_coreqspi_write_op(qspi); - } else { + } else if (!dual_quad) { qspi->rxbuf = (u8 *)t->rx_buf; qspi->rx_len = t->len; mchp_coreqspi_write_read_op(qspi); + } else { + qspi->rxbuf = (u8 *)t->rx_buf; + qspi->rx_len = t->len; + mchp_coreqspi_read_op(qspi); } return 0; From 0b2eb1f8473eddeff5317e521498329581432f89 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Thu, 30 Apr 2026 11:10:20 +0100 Subject: [PATCH 259/525] spi: microchip-core-qspi: remove some inline markings Remove inline markings from a number of functions that are called as part of mem ops callbacks. None of them are either particularly trivial or sensitive to overhead of a function call. Just let the compiler decide what to do with them. Signed-off-by: Conor Dooley Link: https://patch.msgid.link/20260430-serpent-stimulate-59fb860ef429@spud Signed-off-by: Mark Brown --- drivers/spi/spi-microchip-core-qspi.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/spi/spi-microchip-core-qspi.c b/drivers/spi/spi-microchip-core-qspi.c index 70215a407b5a..4dee0fea1df8 100644 --- a/drivers/spi/spi-microchip-core-qspi.c +++ b/drivers/spi/spi-microchip-core-qspi.c @@ -197,7 +197,7 @@ static int mchp_coreqspi_setup(struct spi_device *spi) return 0; } -static inline void mchp_coreqspi_read_op(struct mchp_coreqspi *qspi) +static void mchp_coreqspi_read_op(struct mchp_coreqspi *qspi) { u32 control, data; @@ -233,7 +233,7 @@ static inline void mchp_coreqspi_read_op(struct mchp_coreqspi *qspi) } } -static inline void mchp_coreqspi_write_op(struct mchp_coreqspi *qspi) +static void mchp_coreqspi_write_op(struct mchp_coreqspi *qspi) { u32 control, data; @@ -261,7 +261,7 @@ static inline void mchp_coreqspi_write_op(struct mchp_coreqspi *qspi) } } -static inline void mchp_coreqspi_write_read_op(struct mchp_coreqspi *qspi) +static void mchp_coreqspi_write_read_op(struct mchp_coreqspi *qspi) { u32 control, data; @@ -419,7 +419,7 @@ static int mchp_coreqspi_setup_clock(struct mchp_coreqspi *qspi, struct spi_devi return 0; } -static inline void mchp_coreqspi_config_op(struct mchp_coreqspi *qspi, const struct spi_mem_op *op) +static void mchp_coreqspi_config_op(struct mchp_coreqspi *qspi, const struct spi_mem_op *op) { u32 idle_cycles = 0; int total_bytes, cmd_bytes, frames, ctrl; From 05c5078de822148e7cb84968a8783ddfcb6c9ef1 Mon Sep 17 00:00:00 2001 From: Nicolas Escande Date: Wed, 22 Apr 2026 18:32:58 +0200 Subject: [PATCH 260/525] wifi: ath12k: fix leak in some ath12k_wmi_xxx() functions Some wmi functions were using plain 'return ath12k_wmi_cmd_send(...)' without explicitly handling the error code. This leads to leaking the skb in case of error. Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.3.1-00218-QCAHKSWPL_SILICONZ-1 Fixes: 66a9448b1b89 ("wifi: ath12k: implement hardware data filter") Fixes: 593174170919 ("wifi: ath12k: implement WoW enable and wakeup commands") Fixes: 4a3c212eee0e ("wifi: ath12k: add basic WoW functionalities") Fixes: 16f474d6d49d ("wifi: ath12k: add WoW net-detect functionality") Fixes: 1666108c74c4 ("wifi: ath12k: support ARP and NS offload") Fixes: aab4ae566fa1 ("wifi: ath12k: support GTK rekey offload") Fixes: 7af01e569529 ("wifi: ath12k: handle keepalive during WoWLAN suspend and resume") Signed-off-by: Nicolas Escande Reviewed-by: Baochen Qiang Reviewed-by: Vasanthakumar Thiagarajan Link: https://patch.msgid.link/20260422163258.3013872-1-nico.escande@gmail.com Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath12k/wmi.c | 103 ++++++++++++++++++++++---- 1 file changed, 88 insertions(+), 15 deletions(-) diff --git a/drivers/net/wireless/ath/ath12k/wmi.c b/drivers/net/wireless/ath/ath12k/wmi.c index 65a05a9520ff..75c87edd2a8a 100644 --- a/drivers/net/wireless/ath/ath12k/wmi.c +++ b/drivers/net/wireless/ath/ath12k/wmi.c @@ -10251,7 +10251,7 @@ int ath12k_wmi_hw_data_filter_cmd(struct ath12k *ar, struct wmi_hw_data_filter_a { struct wmi_hw_data_filter_cmd *cmd; struct sk_buff *skb; - int len; + int ret, len; len = sizeof(*cmd); skb = ath12k_wmi_alloc_skb(ar->wmi->wmi_ab, len); @@ -10275,7 +10275,13 @@ int ath12k_wmi_hw_data_filter_cmd(struct ath12k *ar, struct wmi_hw_data_filter_a "wmi hw data filter enable %d filter_bitmap 0x%x\n", arg->enable, arg->hw_filter_bitmap); - return ath12k_wmi_cmd_send(ar->wmi, skb, WMI_HW_DATA_FILTER_CMDID); + ret = ath12k_wmi_cmd_send(ar->wmi, skb, WMI_HW_DATA_FILTER_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_HW_DATA_FILTER_CMDID\n"); + dev_kfree_skb(skb); + } + + return ret; } int ath12k_wmi_wow_host_wakeup_ind(struct ath12k *ar) @@ -10283,6 +10289,7 @@ int ath12k_wmi_wow_host_wakeup_ind(struct ath12k *ar) struct wmi_wow_host_wakeup_cmd *cmd; struct sk_buff *skb; size_t len; + int ret; len = sizeof(*cmd); skb = ath12k_wmi_alloc_skb(ar->wmi->wmi_ab, len); @@ -10295,14 +10302,20 @@ int ath12k_wmi_wow_host_wakeup_ind(struct ath12k *ar) ath12k_dbg(ar->ab, ATH12K_DBG_WMI, "wmi tlv wow host wakeup ind\n"); - return ath12k_wmi_cmd_send(ar->wmi, skb, WMI_WOW_HOSTWAKEUP_FROM_SLEEP_CMDID); + ret = ath12k_wmi_cmd_send(ar->wmi, skb, WMI_WOW_HOSTWAKEUP_FROM_SLEEP_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_WOW_HOSTWAKEUP_FROM_SLEEP_CMDID\n"); + dev_kfree_skb(skb); + } + + return ret; } int ath12k_wmi_wow_enable(struct ath12k *ar) { struct wmi_wow_enable_cmd *cmd; struct sk_buff *skb; - int len; + int ret, len; len = sizeof(*cmd); skb = ath12k_wmi_alloc_skb(ar->wmi->wmi_ab, len); @@ -10317,7 +10330,13 @@ int ath12k_wmi_wow_enable(struct ath12k *ar) cmd->pause_iface_config = cpu_to_le32(WOW_IFACE_PAUSE_ENABLED); ath12k_dbg(ar->ab, ATH12K_DBG_WMI, "wmi tlv wow enable\n"); - return ath12k_wmi_cmd_send(ar->wmi, skb, WMI_WOW_ENABLE_CMDID); + ret = ath12k_wmi_cmd_send(ar->wmi, skb, WMI_WOW_ENABLE_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_WOW_ENABLE_CMDID\n"); + dev_kfree_skb(skb); + } + + return ret; } int ath12k_wmi_wow_add_wakeup_event(struct ath12k *ar, u32 vdev_id, @@ -10327,6 +10346,7 @@ int ath12k_wmi_wow_add_wakeup_event(struct ath12k *ar, u32 vdev_id, struct wmi_wow_add_del_event_cmd *cmd; struct sk_buff *skb; size_t len; + int ret; len = sizeof(*cmd); skb = ath12k_wmi_alloc_skb(ar->wmi->wmi_ab, len); @@ -10343,7 +10363,13 @@ int ath12k_wmi_wow_add_wakeup_event(struct ath12k *ar, u32 vdev_id, ath12k_dbg(ar->ab, ATH12K_DBG_WMI, "wmi tlv wow add wakeup event %s enable %d vdev_id %d\n", wow_wakeup_event(event), enable, vdev_id); - return ath12k_wmi_cmd_send(ar->wmi, skb, WMI_WOW_ENABLE_DISABLE_WAKE_EVENT_CMDID); + ret = ath12k_wmi_cmd_send(ar->wmi, skb, WMI_WOW_ENABLE_DISABLE_WAKE_EVENT_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_WOW_ENABLE_DISABLE_WAKE_EVENT_CMDID\n"); + dev_kfree_skb(skb); + } + + return ret; } int ath12k_wmi_wow_add_pattern(struct ath12k *ar, u32 vdev_id, u32 pattern_id, @@ -10356,6 +10382,7 @@ int ath12k_wmi_wow_add_pattern(struct ath12k *ar, u32 vdev_id, u32 pattern_id, struct sk_buff *skb; void *ptr; size_t len; + int ret; len = sizeof(*cmd) + sizeof(*tlv) + /* array struct */ @@ -10435,7 +10462,13 @@ int ath12k_wmi_wow_add_pattern(struct ath12k *ar, u32 vdev_id, u32 pattern_id, ath12k_dbg_dump(ar->ab, ATH12K_DBG_WMI, NULL, "wow bitmask: ", bitmap->bitmaskbuf, pattern_len); - return ath12k_wmi_cmd_send(ar->wmi, skb, WMI_WOW_ADD_WAKE_PATTERN_CMDID); + ret = ath12k_wmi_cmd_send(ar->wmi, skb, WMI_WOW_ADD_WAKE_PATTERN_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_WOW_ADD_WAKE_PATTERN_CMDID\n"); + dev_kfree_skb(skb); + } + + return ret; } int ath12k_wmi_wow_del_pattern(struct ath12k *ar, u32 vdev_id, u32 pattern_id) @@ -10443,6 +10476,7 @@ int ath12k_wmi_wow_del_pattern(struct ath12k *ar, u32 vdev_id, u32 pattern_id) struct wmi_wow_del_pattern_cmd *cmd; struct sk_buff *skb; size_t len; + int ret; len = sizeof(*cmd); skb = ath12k_wmi_alloc_skb(ar->wmi->wmi_ab, len); @@ -10459,7 +10493,13 @@ int ath12k_wmi_wow_del_pattern(struct ath12k *ar, u32 vdev_id, u32 pattern_id) ath12k_dbg(ar->ab, ATH12K_DBG_WMI, "wmi tlv wow del pattern vdev_id %d pattern_id %d\n", vdev_id, pattern_id); - return ath12k_wmi_cmd_send(ar->wmi, skb, WMI_WOW_DEL_WAKE_PATTERN_CMDID); + ret = ath12k_wmi_cmd_send(ar->wmi, skb, WMI_WOW_DEL_WAKE_PATTERN_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_WOW_DEL_WAKE_PATTERN_CMDID\n"); + dev_kfree_skb(skb); + } + + return ret; } static struct sk_buff * @@ -10595,6 +10635,7 @@ int ath12k_wmi_wow_config_pno(struct ath12k *ar, u32 vdev_id, struct wmi_pno_scan_req_arg *pno_scan) { struct sk_buff *skb; + int ret; if (pno_scan->enable) skb = ath12k_wmi_op_gen_config_pno_start(ar, vdev_id, pno_scan); @@ -10604,7 +10645,13 @@ int ath12k_wmi_wow_config_pno(struct ath12k *ar, u32 vdev_id, if (IS_ERR_OR_NULL(skb)) return -ENOMEM; - return ath12k_wmi_cmd_send(ar->wmi, skb, WMI_NETWORK_LIST_OFFLOAD_CONFIG_CMDID); + ret = ath12k_wmi_cmd_send(ar->wmi, skb, WMI_NETWORK_LIST_OFFLOAD_CONFIG_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_NETWORK_LIST_OFFLOAD_CONFIG_CMDID\n"); + dev_kfree_skb(skb); + } + + return ret; } static void ath12k_wmi_fill_ns_offload(struct ath12k *ar, @@ -10717,6 +10764,7 @@ int ath12k_wmi_arp_ns_offload(struct ath12k *ar, void *buf_ptr; size_t len; u8 ns_cnt, ns_ext_tuples = 0; + int ret; ns_cnt = offload->ipv6_count; @@ -10752,7 +10800,13 @@ int ath12k_wmi_arp_ns_offload(struct ath12k *ar, if (ns_ext_tuples) ath12k_wmi_fill_ns_offload(ar, offload, &buf_ptr, enable, 1); - return ath12k_wmi_cmd_send(ar->wmi, skb, WMI_SET_ARP_NS_OFFLOAD_CMDID); + ret = ath12k_wmi_cmd_send(ar->wmi, skb, WMI_SET_ARP_NS_OFFLOAD_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_SET_ARP_NS_OFFLOAD_CMDID\n"); + dev_kfree_skb(skb); + } + + return ret; } int ath12k_wmi_gtk_rekey_offload(struct ath12k *ar, @@ -10762,7 +10816,7 @@ int ath12k_wmi_gtk_rekey_offload(struct ath12k *ar, struct wmi_gtk_rekey_offload_cmd *cmd; struct sk_buff *skb; __le64 replay_ctr; - int len; + int ret, len; len = sizeof(*cmd); skb = ath12k_wmi_alloc_skb(ar->wmi->wmi_ab, len); @@ -10789,7 +10843,13 @@ int ath12k_wmi_gtk_rekey_offload(struct ath12k *ar, ath12k_dbg(ar->ab, ATH12K_DBG_WMI, "offload gtk rekey vdev: %d %d\n", arvif->vdev_id, enable); - return ath12k_wmi_cmd_send(ar->wmi, skb, WMI_GTK_OFFLOAD_CMDID); + ret = ath12k_wmi_cmd_send(ar->wmi, skb, WMI_GTK_OFFLOAD_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_GTK_OFFLOAD_CMDID offload\n"); + dev_kfree_skb(skb); + } + + return ret; } int ath12k_wmi_gtk_rekey_getinfo(struct ath12k *ar, @@ -10797,7 +10857,7 @@ int ath12k_wmi_gtk_rekey_getinfo(struct ath12k *ar, { struct wmi_gtk_rekey_offload_cmd *cmd; struct sk_buff *skb; - int len; + int ret, len; len = sizeof(*cmd); skb = ath12k_wmi_alloc_skb(ar->wmi->wmi_ab, len); @@ -10811,7 +10871,13 @@ int ath12k_wmi_gtk_rekey_getinfo(struct ath12k *ar, ath12k_dbg(ar->ab, ATH12K_DBG_WMI, "get gtk rekey vdev_id: %d\n", arvif->vdev_id); - return ath12k_wmi_cmd_send(ar->wmi, skb, WMI_GTK_OFFLOAD_CMDID); + ret = ath12k_wmi_cmd_send(ar->wmi, skb, WMI_GTK_OFFLOAD_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_GTK_OFFLOAD_CMDID getinfo\n"); + dev_kfree_skb(skb); + } + + return ret; } int ath12k_wmi_sta_keepalive(struct ath12k *ar, @@ -10822,6 +10888,7 @@ int ath12k_wmi_sta_keepalive(struct ath12k *ar, struct wmi_sta_keepalive_cmd *cmd; struct sk_buff *skb; size_t len; + int ret; len = sizeof(*cmd) + sizeof(*arp); skb = ath12k_wmi_alloc_skb(wmi->wmi_ab, len); @@ -10849,7 +10916,13 @@ int ath12k_wmi_sta_keepalive(struct ath12k *ar, "wmi sta keepalive vdev %d enabled %d method %d interval %d\n", arg->vdev_id, arg->enabled, arg->method, arg->interval); - return ath12k_wmi_cmd_send(wmi, skb, WMI_STA_KEEPALIVE_CMDID); + ret = ath12k_wmi_cmd_send(wmi, skb, WMI_STA_KEEPALIVE_CMDID); + if (ret) { + ath12k_warn(ar->ab, "failed to send WMI_STA_KEEPALIVE_CMDID\n"); + dev_kfree_skb(skb); + } + + return ret; } int ath12k_wmi_mlo_setup(struct ath12k *ar, struct wmi_mlo_setup_arg *mlo_params) From 81594a12d5cecb3ab35b603a00037c7c3ee87ab2 Mon Sep 17 00:00:00 2001 From: Rameshkumar Sundaram Date: Mon, 27 Apr 2026 16:00:11 +0530 Subject: [PATCH 261/525] wifi: ath12k: initialize RSSI dBm conversion event state Currently, the RSSI dBm conversion event handler leaves struct ath12k_wmi_rssi_dbm_conv_info_arg uninitialized on the stack before calling the TLV parser. If one of the optional sub-TLVs is absent, the corresponding *_present flag retains stack garbage and later gets read in ath12k_wmi_update_rssi_offsets(). With UBSAN enabled this triggers an invalid-load report for _Bool: UBSAN: invalid-load in drivers/net/wireless/ath/ath12k/wmi.c:9682:15 load of value 9 is not a valid value for type '_Bool' Call Trace: ath12k_wmi_rssi_dbm_conversion_params_info_event.cold+0x72/0x85 [ath12k] ath12k_wmi_op_rx+0x1871/0x2ab0 [ath12k] ath12k_htc_rx_completion_handler+0x44b/0x810 [ath12k] ath12k_ce_recv_process_cb+0x554/0x9f0 [ath12k] ath12k_ce_per_engine_service+0xbe/0xf0 [ath12k] ath12k_pci_ce_workqueue+0x69/0x120 [ath12k] Initialize the parsed event state to zero before passing it to the TLV parser so missing sub-TLVs correctly leave the presence flags false. Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.4.1-00199-QCAHKSWPL_SILICONZ-1 Fixes: 0314ee81a91d ("wifi: ath12k: handle WMI event for real noise floor calculation") Signed-off-by: Rameshkumar Sundaram Reviewed-by: Baochen Qiang Link: https://patch.msgid.link/20260427103011.2983269-1-rameshkumar.sundaram@oss.qualcomm.com Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath12k/wmi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath12k/wmi.c b/drivers/net/wireless/ath/ath12k/wmi.c index 75c87edd2a8a..b5e904a55aea 100644 --- a/drivers/net/wireless/ath/ath12k/wmi.c +++ b/drivers/net/wireless/ath/ath12k/wmi.c @@ -9778,7 +9778,7 @@ static void ath12k_wmi_rssi_dbm_conversion_params_info_event(struct ath12k_base *ab, struct sk_buff *skb) { - struct ath12k_wmi_rssi_dbm_conv_info_arg rssi_info; + struct ath12k_wmi_rssi_dbm_conv_info_arg rssi_info = {}; struct ath12k *ar; s32 noise_floor; u32 pdev_id; From 0e1308803d2c3fd365a6d21e6be355ec1e28eaaf Mon Sep 17 00:00:00 2001 From: Baochen Qiang Date: Mon, 27 Apr 2026 13:51:41 +0800 Subject: [PATCH 262/525] wifi: ath12k: fix peer_id usage in normal RX path ath12k_dp_rx_deliver_msdu() currently uses hal_rx_desc_data::peer_id parsed from mpdu_start descriptor to do peer lookup. However In an A-MSDU aggregation scenario, hardware only populates mpdu_start descriptor for the first sub-msdu, but not the following ones. In that case peer_id could be invalid, leading to peer lookup failure: ath12k_wifi7_pci 0000:06:00.0: rx skb 00000000c391c041 len 1532 peer (null) 0 ucast sn 0 eht320 rate_idx 12 vht_nss 2 freq 6105 band 3 flag 0x40d1a fcs-err 0 mic-err 0 amsdu-more 0 As a result pubsta is NULL and parts of ieee80211_rx_status structure are left uninitialized, which may cause unexpected behavior. Fix it by switching the normal RX path to use ath12k_skb_rxcb::peer_id which is parsed from REO ring's rx_mpdu_desc and is always valid. hal_rx_desc_data::peer_id is still used in ath12k_wifi7_dp_rx_frag_h_mpdu(), which is safe since A-MSDU aggregation does not occur for fragmented frames. Similarly, ath12k_skb_rxcb::peer_id may be overwritten by hal_rx_desc_data::peer_id in ath12k_wifi7_dp_rx_h_mpdu(), which only handles non-aggregated multicast/broadcast traffic. Tested-on: WCN7850 hw2.0 PCI WLAN.HMT.1.1.c5-00302-QCAHMTSWPL_V1.0_V2.0_SILICONZ-1.115823.3 Fixes: 11157e0910fd ("wifi: ath12k: Use ath12k_dp_peer in per packet Tx & Rx paths") Signed-off-by: Baochen Qiang Reviewed-by: Rameshkumar Sundaram Link: https://patch.msgid.link/20260427-ath12k-fix-peer-id-source-v1-1-b5f701fb8e88@oss.qualcomm.com Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath12k/dp_rx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath12k/dp_rx.c b/drivers/net/wireless/ath/ath12k/dp_rx.c index 25557dea5826..b108ccd0f637 100644 --- a/drivers/net/wireless/ath/ath12k/dp_rx.c +++ b/drivers/net/wireless/ath/ath12k/dp_rx.c @@ -1340,7 +1340,7 @@ void ath12k_dp_rx_deliver_msdu(struct ath12k_pdev_dp *dp_pdev, struct napi_struc bool is_mcbc = rxcb->is_mcbc; bool is_eapol = rxcb->is_eapol; - peer = ath12k_dp_peer_find_by_peerid(dp_pdev, rx_info->peer_id); + peer = ath12k_dp_peer_find_by_peerid(dp_pdev, rxcb->peer_id); pubsta = peer ? peer->sta : NULL; From d748603f12baff112caa3ab7d39f50100f010dbd Mon Sep 17 00:00:00 2001 From: "Jiri Slaby (SUSE)" Date: Tue, 9 Dec 2025 11:04:59 +0100 Subject: [PATCH 263/525] wifi: ath5k: do not access array OOB Vincent reports: > The ath5k driver seems to do an array-index-out-of-bounds access as > shown by the UBSAN kernel message: > UBSAN: array-index-out-of-bounds in drivers/net/wireless/ath/ath5k/base.c:1741:20 > index 4 is out of range for type 'ieee80211_tx_rate [4]' > ... > Call Trace: > > dump_stack_lvl+0x5d/0x80 > ubsan_epilogue+0x5/0x2b > __ubsan_handle_out_of_bounds.cold+0x46/0x4b > ath5k_tasklet_tx+0x4e0/0x560 [ath5k] > tasklet_action_common+0xb5/0x1c0 It is real. 'ts->ts_final_idx' can be 3 on 5212, so: info->status.rates[ts->ts_final_idx + 1].idx = -1; with the array defined as: struct ieee80211_tx_rate rates[IEEE80211_TX_MAX_RATES]; while the size is: #define IEEE80211_TX_MAX_RATES 4 is indeed bogus. Set this 'idx = -1' sentinel only if the array index is less than the array size. As mac80211 will not look at rates beyond the size (IEEE80211_TX_MAX_RATES). Note: The effect of the OOB write is negligible. It just overwrites the next member of info->status, i.e. ack_signal. Signed-off-by: Jiri Slaby (SUSE) Reported-by: Vincent Danjean Link: https://lore.kernel.org/all/aQYUkIaT87ccDCin@eldamar.lan Closes: https://bugs.debian.org/1119093 Fixes: 6d7b97b23e11 ("ath5k: fix tx status reporting issues") Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20251209100459.2253198-1-jirislaby@kernel.org Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath5k/base.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath5k/base.c b/drivers/net/wireless/ath/ath5k/base.c index 05c9c07591fc..6ca31d4ea437 100644 --- a/drivers/net/wireless/ath/ath5k/base.c +++ b/drivers/net/wireless/ath/ath5k/base.c @@ -1738,7 +1738,8 @@ ath5k_tx_frame_completed(struct ath5k_hw *ah, struct sk_buff *skb, } info->status.rates[ts->ts_final_idx].count = ts->ts_final_retry; - info->status.rates[ts->ts_final_idx + 1].idx = -1; + if (ts->ts_final_idx + 1 < IEEE80211_TX_MAX_RATES) + info->status.rates[ts->ts_final_idx + 1].idx = -1; if (unlikely(ts->ts_status)) { ah->stats.ack_fail++; From 2a46a9356ba7b1bdd741c8b41e5374edcd960557 Mon Sep 17 00:00:00 2001 From: "Kory Maincent (TI)" Date: Tue, 28 Apr 2026 11:04:56 +0200 Subject: [PATCH 264/525] drm/bridge: tda998x: Use __be32 for audio port OF property pointer of_get_property() returns a pointer to big-endian (__be32) data, but port_data in tda998x_get_audio_ports() was declared as const u32 *, causing a sparse endianness type mismatch warning. Fix the declaration to use const __be32 *. Fixes: 7e567624dc5a4 ("drm/i2c: tda998x: Register ASoC hdmi-codec and add audio DT binding") Cc: stable@vger.kernel.org Signed-off-by: Kory Maincent (TI) Reviewed-by: Russell King (Oracle) Link: https://patch.msgid.link/20260428090457.121894-1-kory.maincent@bootlin.com Signed-off-by: Luca Ceresoli --- drivers/gpu/drm/bridge/tda998x_drv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/bridge/tda998x_drv.c b/drivers/gpu/drm/bridge/tda998x_drv.c index d9b388165de1..779b976f601c 100644 --- a/drivers/gpu/drm/bridge/tda998x_drv.c +++ b/drivers/gpu/drm/bridge/tda998x_drv.c @@ -1762,7 +1762,7 @@ static const struct drm_bridge_funcs tda998x_bridge_funcs = { static int tda998x_get_audio_ports(struct tda998x_priv *priv, struct device_node *np) { - const u32 *port_data; + const __be32 *port_data; u32 size; int i; From b5d0ad616ca8dd8c7b6b24dc13012e342278a085 Mon Sep 17 00:00:00 2001 From: "Kory Maincent (TI)" Date: Fri, 17 Apr 2026 17:54:45 +0200 Subject: [PATCH 265/525] drm/bridge: tda998x: Return NULL instead of 0 in tda998x_edid_read() tda998x_edid_read() returns a const struct drm_edid pointer, but when tda998x_edid_delay_wait() fails (process killed while waiting for the HPD timeout), the integer literal 0 is returned instead of NULL, triggering a sparse warning: "Using plain integer as NULL pointer" Replace 0 with NULL to fix the sparse warning. Fixes: c76a8be4feec ("drm/bridge: tda998x: Add support for DRM_BRIDGE_ATTACH_NO_CONNECTOR") Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202604172257.Imo6GOH9-lkp@intel.com/ Signed-off-by: Kory Maincent (TI) Reviewed-by: Luca Ceresoli Link: https://patch.msgid.link/20260417155446.1068893-1-kory.maincent@bootlin.com Signed-off-by: Luca Ceresoli --- drivers/gpu/drm/bridge/tda998x_drv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/bridge/tda998x_drv.c b/drivers/gpu/drm/bridge/tda998x_drv.c index 779b976f601c..6c427bc75896 100644 --- a/drivers/gpu/drm/bridge/tda998x_drv.c +++ b/drivers/gpu/drm/bridge/tda998x_drv.c @@ -1293,7 +1293,7 @@ static const struct drm_edid *tda998x_edid_read(struct tda998x_priv *priv, * can't handle signals gracefully. */ if (tda998x_edid_delay_wait(priv)) - return 0; + return NULL; if (priv->rev == TDA19988) reg_clear(priv, REG_TX4, TX4_PD_RAM); From f80785888f7c0980a49545b87a80e3817c9ed7c6 Mon Sep 17 00:00:00 2001 From: Anton Swart Date: Sun, 3 May 2026 23:15:17 +0200 Subject: [PATCH 266/525] ALSA: usb-audio: Add quirk flags for AlphaTheta EUPHONIA The AlphaTheta EUPHONIA (VID 0x2b73, PID 0x0047) is a USB Audio Class 2 DJ mixer that requires implicit feedback for full-duplex operation. The capture endpoint (0x83 IN, interface 2) acts as the implicit feedback source for the playback endpoint (0x03 OUT, interface 1), and the device firmware does not send isochronous data on the capture endpoint unless the host is simultaneously sending data on the playback endpoint, i.e. playback must be started first. Without QUIRK_FLAG_PLAYBACK_FIRST the kernel waits for capture URBs before submitting playback URBs, creating a deadlock: the device waits for playback data and the host waits for capture data. Without QUIRK_FLAG_GENERIC_IMPLICIT_FB the kernel does not detect the implicit feedback relationship between the two interfaces. The same flag combination is already used for the Behringer UMC202HD, UMC204HD and UMC404HD (0x1397:0x0507/0x0508/0x0509), which exhibit the identical implicit-feedback topology. Tested on Raspberry Pi 5 with kernel 6.12.75; continuous full-duplex streaming at 96 kHz / 24-bit, zero XRUNs. Signed-off-by: Anton Swart Link: https://patch.msgid.link/20260503211517.14332-1-anton.swart.jhb@gmail.com Signed-off-by: Takashi Iwai --- sound/usb/quirks.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/sound/usb/quirks.c b/sound/usb/quirks.c index f77ff05dff0b..b17bdae09bc3 100644 --- a/sound/usb/quirks.c +++ b/sound/usb/quirks.c @@ -2460,6 +2460,8 @@ static const struct usb_audio_quirk_flags_table quirk_flags_table[] = { QUIRK_FLAG_GENERIC_IMPLICIT_FB), DEVICE_FLG(0x2b53, 0x0031, /* Fiero SC-01 (firmware v1.1.0) */ QUIRK_FLAG_GENERIC_IMPLICIT_FB), + DEVICE_FLG(0x2b73, 0x0047, /* AlphaTheta EUPHONIA */ + QUIRK_FLAG_PLAYBACK_FIRST | QUIRK_FLAG_GENERIC_IMPLICIT_FB), DEVICE_FLG(0x2d95, 0x8011, /* VIVO USB-C HEADSET */ QUIRK_FLAG_CTL_MSG_DELAY_1M), DEVICE_FLG(0x2d95, 0x8021, /* VIVO USB-C-XE710 HEADSET */ From 0749daa8eb5ab90334aaad3b0671efd7150d43b1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?C=C3=A1ssio=20Gabriel?= Date: Sun, 3 May 2026 21:55:52 -0300 Subject: [PATCH 267/525] ALSA: firewire-tascam: Do not drop unread control events MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit tscm_hwdep_read_queue() copies as many queued control events as fit in the userspace buffer. When the buffer is smaller than the current contiguous queue segment, length is rounded down to the number of bytes that can be copied. However, after copying that shortened length, the code advances pull_pos to the original tail_pos, marking the whole contiguous segment as consumed. Any events between the copied portion and tail_pos are lost. Limit tail_pos to the position after the entries actually copied before updating pull_pos. When the whole segment fits, this is equivalent to the old tail_pos update; when the buffer is smaller, the remaining events stay queued for the next read. Fixes: a8c0d13267a4 ("ALSA: firewire-tascam: notify events of change of state for userspace applications") Cc: stable@vger.kernel.org Suggested-by: Takashi Sakamoto Signed-off-by: Cássio Gabriel Reviewed-by: Takashi Sakamoto Co-developed-by: Takashi Sakamoto Signed-off-by: Takashi Sakamoto Signed-off-by: Takashi Iwai Link: https://patch.msgid.link/20260503-alsa-firewire-tascam-read-queue-v2-1-126c6efd7642@gmail.com --- sound/firewire/tascam/tascam-hwdep.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/firewire/tascam/tascam-hwdep.c b/sound/firewire/tascam/tascam-hwdep.c index 867b4ea1096e..6270263e7bf4 100644 --- a/sound/firewire/tascam/tascam-hwdep.c +++ b/sound/firewire/tascam/tascam-hwdep.c @@ -73,6 +73,7 @@ static long tscm_hwdep_read_queue(struct snd_tscm *tscm, char __user *buf, length = rounddown(remained, sizeof(*entries)); if (length == 0) break; + tail_pos = head_pos + length / sizeof(*entries); spin_unlock_irq(&tscm->lock); if (copy_to_user(pos, &entries[head_pos], length)) From db4cd7d6c571c5ec6b52a6e80ec59ad99c4aa1d6 Mon Sep 17 00:00:00 2001 From: Rong Zhang Date: Mon, 4 May 2026 19:38:05 +0800 Subject: [PATCH 268/525] ALSA: usb-audio: Add quirk flags for JBL Pebbles JBL Pebbles is a pair of desktop speakers with UAC interface. Its Playback and Capture mixers use linear volume with val = 0/999/1 and 0/3996/4. Meanwhile, the reported sample rates are truncated to multiples of 0x100 (i.e., 44100 => 44032), resulting in noisy kmsg, as a warning message is printed each time a stream is opened. Add a quirk table entry matching VID/PID=0x05fc/0x0231 and applying linear volume and sample rate quirk flags, so that it can work properly. Also note that the volume control knob on device is an incremental encoder. It does nothing but sends KEY_VOLUMEUP and KEY_VOLUMEDOWN per rotation, controlling the UAC Playback volume mixer indirectly. Hence, the linear volume quirk flags also enable the volume control knob to function properly. Quirky device sample: usb 5-1.1: new full-speed USB device number 12 using xhci_hcd usb 5-1.1: New USB device found, idVendor=05fc, idProduct=0231, bcdDevice= 1.00 usb 5-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=3 usb 5-1.1: Product: JBL Pebbles usb 5-1.1: Manufacturer: Harman International Industries usb 5-1.1: SerialNumber: 1.0.0 usb-storage 5-1.1:1.0: USB Mass Storage device detected scsi host0: usb-storage 5-1.1:1.0 usb 5-1.1: Found last interface = 1 usb 5-1.1: 2:1: add audio endpoint 0x5 usb 5-1.1: Creating new data endpoint #5 usb 5-1.1: 2:1 Set sample rate 44100, clock 0 usb 5-1.1: current rate 44032 is different from the runtime rate 44100 usb 5-1.1: 3:1: add audio endpoint 0x84 usb 5-1.1: Creating new data endpoint #84 usb 5-1.1: 3:1 Set sample rate 44100, clock 0 usb 5-1.1: current rate 44032 is different from the runtime rate 44100 usb 5-1.1: [2] FU [PCM Playback Switch] ch = 1, val = 0/1/1 usb 5-1.1: Warning! Unlikely big volume step count (=999), linear volume or wrong cval->res? usb 5-1.1: [2] FU [PCM Playback Volume] ch = 2, val = 0/999/1 usb 5-1.1: [5] FU [Mic Capture Switch] ch = 1, val = 0/1/1 usb 5-1.1: Warning! Unlikely big volume step count (=999), linear volume or wrong cval->res? usb 5-1.1: [5] FU [Mic Capture Volume] ch = 2, val = 0/3996/4 input: Harman International Industries JBL Pebbles as /devices/pci0000:00/0000:00:08.3/0000:67:00.3/usb5/5-1/5-1.1/5-1.1:1.4/0003:05FC:0231.0018/input/input55 hid-generic 0003:05FC:0231.0018: input,hidraw2: USB HID v2.01 Device [Harman International Industries JBL Pebbles] on usb-0000:67:00.3-1.1/input4 Signed-off-by: Rong Zhang Link: https://patch.msgid.link/20260504-uac-jbl-pebbles-v1-1-c888d592a286@rong.moe Signed-off-by: Takashi Iwai --- sound/usb/quirks.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/sound/usb/quirks.c b/sound/usb/quirks.c index b17bdae09bc3..17983d9774f8 100644 --- a/sound/usb/quirks.c +++ b/sound/usb/quirks.c @@ -2277,6 +2277,9 @@ static const struct usb_audio_quirk_flags_table quirk_flags_table[] = { QUIRK_FLAG_ALIGN_TRANSFER), DEVICE_FLG(0x05e1, 0x0480, /* Hauppauge Woodbury */ QUIRK_FLAG_SHARE_MEDIA_DEVICE | QUIRK_FLAG_ALIGN_TRANSFER), + DEVICE_FLG(0x05fc, 0x0231, /* JBL Pebbles */ + QUIRK_FLAG_MIXER_PLAYBACK_LINEAR_VOL | QUIRK_FLAG_MIXER_CAPTURE_LINEAR_VOL | + QUIRK_FLAG_GET_SAMPLE_RATE), DEVICE_FLG(0x0624, 0x3d3f, /* AB13X USB Audio */ QUIRK_FLAG_FORCE_IFACE_RESET | QUIRK_FLAG_IFACE_DELAY), DEVICE_FLG(0x0644, 0x8043, /* TEAC UD-501/UD-501V2/UD-503/NT-503 */ From dc1e0172be54e742bccb28d5f14c0c395e28c098 Mon Sep 17 00:00:00 2001 From: Fernando Antunez Antonio Date: Mon, 4 May 2026 07:33:26 -0600 Subject: [PATCH 269/525] ALSA: hda/realtek: Fix mute and mic-mute LEDs for HP Envy X360 15-fh0xxx This enables the mute and mic-mute LEDs on the HP Envy X360 15-fh0xxx 2-in-1 laptops. The quirk 'ALC245_FIXUP_HP_ENVY_X360_15_FH0XXX' has been created and is now enabled for this device. This is my first patch, and I'm still getting to grips with the code, so there's probably a better way to implement this fix. I apologize for any inconvenience caused by the constant release of new versions of this patch. Signed-off-by: Fernando Antunez Antonio Link: https://patch.msgid.link/20260504-hpenvy-muteled-fix-v3-1-5567fd9b3d25@gmail.com Signed-off-by: Takashi Iwai --- sound/hda/codecs/realtek/alc269.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/sound/hda/codecs/realtek/alc269.c b/sound/hda/codecs/realtek/alc269.c index e061673d3c2e..a7525ed83e2b 100644 --- a/sound/hda/codecs/realtek/alc269.c +++ b/sound/hda/codecs/realtek/alc269.c @@ -4137,6 +4137,7 @@ enum { ALC245_FIXUP_CS35L41_I2C_2_MUTE_LED, ALC236_FIXUP_HP_DMIC, ALC256_FIXUP_HONOR_MRB_XXX_M1020_AUDIO, + ALC245_FIXUP_HP_ENVY_X360_15_FH0XXX, }; /* A special fixup for Lenovo C940 and Yoga Duet 7; @@ -6693,6 +6694,12 @@ static const struct hda_fixup alc269_fixups[] = { { 0x1b, 0x90170110 }, { } } + }, + [ALC245_FIXUP_HP_ENVY_X360_15_FH0XXX] = { + .type = HDA_FIXUP_FUNC, + .v.func = cs35l41_fixup_i2c_two, + .chained = true, + .chain_id = ALC245_FIXUP_HP_X360_MUTE_LEDS } }; @@ -7115,7 +7122,7 @@ static const struct hda_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x103c, 0x8be6, "HP Envy 16", ALC287_FIXUP_CS35L41_I2C_2), SND_PCI_QUIRK(0x103c, 0x8be7, "HP Envy 17", ALC287_FIXUP_CS35L41_I2C_2), SND_PCI_QUIRK(0x103c, 0x8be8, "HP Envy 17", ALC287_FIXUP_CS35L41_I2C_2), - SND_PCI_QUIRK(0x103c, 0x8be9, "HP Envy 15", ALC287_FIXUP_CS35L41_I2C_2), + SND_PCI_QUIRK(0x103c, 0x8be9, "HP Envy x360 2-in-1 Laptop 15-fh0xxx", ALC245_FIXUP_HP_ENVY_X360_15_FH0XXX), SND_PCI_QUIRK(0x103c, 0x8bf0, "HP", ALC236_FIXUP_HP_GPIO_LED), SND_PCI_QUIRK(0x103c, 0x8c15, "HP Spectre x360 2-in-1 Laptop 14-eu0xxx", ALC245_FIXUP_HP_SPECTRE_X360_EU0XXX), SND_PCI_QUIRK(0x103c, 0x8c16, "HP Spectre x360 2-in-1 Laptop 16-aa0xxx", ALC245_FIXUP_HP_SPECTRE_X360_16_AA0XXX), From f3c57c9c2a49a21d784b7c04a2c883bffc070659 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?C=C3=A1ssio=20Gabriel?= Date: Mon, 4 May 2026 11:08:45 -0300 Subject: [PATCH 270/525] ALSA: usb-audio: midi2: Restart output URBs on resume MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit USB MIDI 2.0 suspend saves the endpoint running state, clears it and kills all endpoint URBs. Resume restores the running state, but only restarts input endpoints. For a running output endpoint, this leaves the endpoint marked running with an empty URB queue. Output transfer progress depends on either the rawmidi trigger path starting the queue or an output completion refilling it. After suspend there is no completion left, and output data that remains queued in the raw UMP or legacy rawmidi buffer can stay stalled until userspace happens to trigger the stream again. Restore the saved state with atomic accessors, keep input endpoints restarted as before, and restart output endpoints that were running before suspend. Clear the saved suspend state after restoring it. Fixes: ff49d1df79ae ("ALSA: usb-audio: USB MIDI 2.0 UMP support") Cc: stable@vger.kernel.org Signed-off-by: Cássio Gabriel Link: https://patch.msgid.link/20260504-usb-midi2-output-resume-v1-1-c089cc8ad3c6@gmail.com Signed-off-by: Takashi Iwai --- sound/usb/midi2.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/sound/usb/midi2.c b/sound/usb/midi2.c index 3546ba926cb3..2785600d2312 100644 --- a/sound/usb/midi2.c +++ b/sound/usb/midi2.c @@ -227,7 +227,7 @@ static void kill_midi_urbs(struct snd_usb_midi2_endpoint *ep, bool suspending) if (!ep) return; if (suspending) - ep->suspended = ep->running; + atomic_set(&ep->suspended, atomic_read(&ep->running)); atomic_set(&ep->running, 0); for (i = 0; i < ep->num_urbs; i++) { if (!ep->urbs[i].urb) @@ -1188,10 +1188,11 @@ void snd_usb_midi_v2_suspend_all(struct snd_usb_audio *chip) static void resume_midi2_endpoint(struct snd_usb_midi2_endpoint *ep) { - ep->running = ep->suspended; - if (ep->direction == STR_IN) + atomic_set(&ep->running, atomic_read(&ep->suspended)); + atomic_set(&ep->suspended, 0); + + if (ep->direction == STR_IN || atomic_read(&ep->running)) submit_io_urbs(ep); - /* FIXME: does it all? */ } void snd_usb_midi_v2_resume_all(struct snd_usb_audio *chip) From 320e55722ca466a7d40dd69e1aea982cb6189006 Mon Sep 17 00:00:00 2001 From: Nicola Lunghi Date: Mon, 4 May 2026 16:45:20 +0200 Subject: [PATCH 271/525] ALSA: usb-audio: add clock quirk for Motu 1248 The Motu 1248 (and probably other older Motu AVB interfaces) take more than 2 seconds to switch clock. During the clock switching process the device return that the clock is not valid. This is similar to what already implemented for the Microbook II interface. Add the Motu 1248 usb id to the existing Motu quirk. Signed-off-by: Nicola Lunghi Link: https://patch.msgid.link/20260504144520.699522-2-nick83ola@gmail.com Signed-off-by: Takashi Iwai --- sound/usb/clock.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/sound/usb/clock.c b/sound/usb/clock.c index 842ba5b801ea..2e0c18e35281 100644 --- a/sound/usb/clock.c +++ b/sound/usb/clock.c @@ -208,11 +208,18 @@ static bool uac_clock_source_is_valid_quirk(struct snd_usb_audio *chip, } /* - * MOTU MicroBook IIc - * Sample rate changes takes more than 2 seconds for this device. Clock - * validity request returns false during that period. + * Quirk for older MOTU AVB / hybrid interfaces + * + * These devices take more than 2 seconds to switch sample rate or + * clock source. During this period the clock validity request + * returns false, causing ALSA to fail prematurely. + * + * Affected models (all use vendor 0x07fd): + * - MicroBook IIc → 0x0004 + * - 1248, 624, 8A, UltraLite AVB, 8M, 16A, ... → 0x0005 */ - if (chip->usb_id == USB_ID(0x07fd, 0x0004)) { + if (chip->usb_id == USB_ID(0x07fd, 0x0004) || /* MicroBook IIc */ + chip->usb_id == USB_ID(0x07fd, 0x0005)) { /* 1248 / 624 / 8A / UltraLite AVB / ... */ count = 0; while ((!ret) && (count < 50)) { From 93618edf753838a727dbff63c7c291dee22d656b Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 1 May 2026 08:31:22 -1000 Subject: [PATCH 272/525] cgroup: Defer css percpu_ref kill on rmdir until cgroup is depopulated A chain of commits going back to v7.0 reworked rmdir to satisfy the controller invariant that a subsystem's ->css_offline() must not run while tasks are still doing kernel-side work in the cgroup. [1] d245698d727a ("cgroup: Defer task cgroup unlink until after the task is done switching out") [2] a72f73c4dd9b ("cgroup: Don't expose dead tasks in cgroup") [3] 1b164b876c36 ("cgroup: Wait for dying tasks to leave on rmdir") [4] 4c56a8ac6869 ("cgroup: Fix cgroup_drain_dying() testing the wrong condition") [5] 13e786b64bd3 ("cgroup: Increment nr_dying_subsys_* from rmdir context") [1] moved task cset unlink from do_exit() to finish_task_switch() so a task's cset link drops only after the task has fully stopped scheduling. That made tasks past exit_signals() linger on cset->tasks until their final context switch, which led to a series of problems as what userspace expected to see after rmdir diverged from what the kernel needs to wait for. [2]-[5] tried to bridge that divergence: [2] filtered the exiting tasks from cgroup.procs; [3] had rmdir(2) sleep in TASK_UNINTERRUPTIBLE for them; [4] fixed the wait's condition; [5] made nr_dying_subsys_* visible synchronously. The cgroup_drain_dying() wait in [3] turned out to be a dead end. When the rmdir caller is also the reaper of a zombie that pins a pidns teardown (e.g. host PID 1 systemd reaping orphan pids that were re-parented to it during the same teardown), rmdir blocks in TASK_UNINTERRUPTIBLE waiting for those pids to free, the pids can't free because PID 1 is the reaper and it's stuck in rmdir, and the system A-A deadlocks. No internal lock ordering breaks this; the wait itself is the bug. The css killing side that drove the original reorder, however, can be made cleanly asynchronous: ->css_offline() is already async, run from css_killed_work_fn() driven by percpu_ref_kill_and_confirm(). The fix is to make that chain start only after all tasks have left the cgroup. rmdir's user-visible side then returns as soon as cgroup.procs and friends are empty, while ->css_offline() still runs only after the cgroup is fully drained. Verified by the original reproducer (pidns teardown + zombie reaper, runs under vng) which hangs vanilla and succeeds here, and by per-commit deterministic repros for [2], [3], [4], [5] with a boot parameter that widens the post-exit_signals() window so each state is reliably reachable. Some stress tests on top of that. cgroup_apply_control_disable() has the same shape of pre-existing race: when a controller is disabled via subtree_control, kill_css() ran synchronously while tasks past exit_signals() could still be linked to the cgroup's csets, and ->css_offline() could fire before they drained. This patch preserves the existing synchronous behavior at that call site (kill_css_sync() + kill_css_finish() back-to-back) and a follow-up patch will defer kill_css_finish() there using a per-css trigger. This seems like the right approach and I don't see problems with it. The changes are somewhat invasive but not excessively so, so backporting to -stable should be okay. If something does turn out to be wrong, the fallback is to revert the entire chain ([1]-[5]) and rework in the development branch instead. v2: Pin cgrp across the deferred destroy work with explicit cgroup_get()/cgroup_put() around queue_work() and the work_fn. v1 wasn't actually broken (ordered cgroup_offline_wq + queue_work order in cgroup_task_dead() saved it) but the explicit ref removes the dependency on those non-obvious invariants. Also note the pre-existing cgroup_apply_control_disable() race in the description; a follow-up will defer kill_css_finish() there. Fixes: 1b164b876c36 ("cgroup: Wait for dying tasks to leave on rmdir") Cc: stable@vger.kernel.org # v7.0+ Reported-and-tested-by: Martin Pitt Link: https://lore.kernel.org/all/afHNg2VX2jy9bW7y@piware.de/ Link: https://lore.kernel.org/all/35e0670adb4abeab13da2c321582af9f@kernel.org/ Signed-off-by: Tejun Heo Acked-by: Sebastian Andrzej Siewior --- include/linux/cgroup-defs.h | 4 +- kernel/cgroup/cgroup.c | 244 +++++++++++++++++------------------- 2 files changed, 116 insertions(+), 132 deletions(-) diff --git a/include/linux/cgroup-defs.h b/include/linux/cgroup-defs.h index f42563739d2e..50a784da7a81 100644 --- a/include/linux/cgroup-defs.h +++ b/include/linux/cgroup-defs.h @@ -611,8 +611,8 @@ struct cgroup { /* used to wait for offlining of csses */ wait_queue_head_t offline_waitq; - /* used by cgroup_rmdir() to wait for dying tasks to leave */ - wait_queue_head_t dying_populated_waitq; + /* defers killing csses after removal until cgroup is depopulated */ + struct work_struct finish_destroy_work; /* used to schedule release agent */ struct work_struct release_agent_work; diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index c928dea9dea6..bd10a7e2f9c5 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -264,10 +264,12 @@ static void cgroup_finalize_control(struct cgroup *cgrp, int ret); static void css_task_iter_skip(struct css_task_iter *it, struct task_struct *task); static int cgroup_destroy_locked(struct cgroup *cgrp); +static void cgroup_finish_destroy(struct cgroup *cgrp); +static void kill_css_sync(struct cgroup_subsys_state *css); +static void kill_css_finish(struct cgroup_subsys_state *css); static struct cgroup_subsys_state *css_create(struct cgroup *cgrp, struct cgroup_subsys *ss); static void css_release(struct percpu_ref *ref); -static void kill_css(struct cgroup_subsys_state *css); static int cgroup_addrm_files(struct cgroup_subsys_state *css, struct cgroup *cgrp, struct cftype cfts[], bool is_add); @@ -797,6 +799,16 @@ static void cgroup_update_populated(struct cgroup *cgrp, bool populated) if (was_populated == cgroup_is_populated(cgrp)) break; + /* + * Subtree just emptied below an offlined cgrp. Fire deferred + * destroy. The transition is one-shot. + */ + if (was_populated && !css_is_online(&cgrp->self)) { + cgroup_get(cgrp); + WARN_ON_ONCE(!queue_work(cgroup_offline_wq, + &cgrp->finish_destroy_work)); + } + cgroup1_check_for_release(cgrp); TRACE_CGROUP_PATH(notify_populated, cgrp, cgroup_is_populated(cgrp)); @@ -2039,6 +2051,16 @@ static int cgroup_reconfigure(struct fs_context *fc) return 0; } +static void cgroup_finish_destroy_work_fn(struct work_struct *work) +{ + struct cgroup *cgrp = container_of(work, struct cgroup, finish_destroy_work); + + cgroup_lock(); + cgroup_finish_destroy(cgrp); + cgroup_unlock(); + cgroup_put(cgrp); +} + static void init_cgroup_housekeeping(struct cgroup *cgrp) { struct cgroup_subsys *ss; @@ -2065,7 +2087,7 @@ static void init_cgroup_housekeeping(struct cgroup *cgrp) #endif init_waitqueue_head(&cgrp->offline_waitq); - init_waitqueue_head(&cgrp->dying_populated_waitq); + INIT_WORK(&cgrp->finish_destroy_work, cgroup_finish_destroy_work_fn); INIT_WORK(&cgrp->release_agent_work, cgroup1_release_agent); } @@ -3375,7 +3397,8 @@ static void cgroup_apply_control_disable(struct cgroup *cgrp) if (css->parent && !(cgroup_ss_mask(dsct) & (1 << ss->id))) { - kill_css(css); + kill_css_sync(css); + kill_css_finish(css); } else if (!css_visible(css)) { css_clear_dir(css); if (ss->css_reset) @@ -5514,7 +5537,7 @@ static struct cftype cgroup_psi_files[] = { * css destruction is four-stage process. * * 1. Destruction starts. Killing of the percpu_ref is initiated. - * Implemented in kill_css(). + * Implemented in kill_css_finish(). * * 2. When the percpu_ref is confirmed to be visible as killed on all CPUs * and thus css_tryget_online() is guaranteed to fail, the css can be @@ -5993,7 +6016,7 @@ int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, umode_t mode) /* * This is called when the refcnt of a css is confirmed to be killed. * css_tryget_online() is now guaranteed to fail. Tell the subsystem to - * initiate destruction and put the css ref from kill_css(). + * initiate destruction and put the css ref from kill_css_finish(). */ static void css_killed_work_fn(struct work_struct *work) { @@ -6025,15 +6048,12 @@ static void css_killed_ref_fn(struct percpu_ref *ref) } /** - * kill_css - destroy a css - * @css: css to destroy + * kill_css_sync - synchronous half of css teardown + * @css: css being killed * - * This function initiates destruction of @css by removing cgroup interface - * files and putting its base reference. ->css_offline() will be invoked - * asynchronously once css_tryget_online() is guaranteed to fail and when - * the reference count reaches zero, @css will be released. + * See cgroup_destroy_locked(). */ -static void kill_css(struct cgroup_subsys_state *css) +static void kill_css_sync(struct cgroup_subsys_state *css) { struct cgroup_subsys *ss = css->ss; @@ -6056,24 +6076,6 @@ static void kill_css(struct cgroup_subsys_state *css) */ css_clear_dir(css); - /* - * Killing would put the base ref, but we need to keep it alive - * until after ->css_offline(). - */ - css_get(css); - - /* - * cgroup core guarantees that, by the time ->css_offline() is - * invoked, no new css reference will be given out via - * css_tryget_online(). We can't simply call percpu_ref_kill() and - * proceed to offlining css's because percpu_ref_kill() doesn't - * guarantee that the ref is seen as killed on all CPUs on return. - * - * Use percpu_ref_kill_and_confirm() to get notifications as each - * css is confirmed to be seen as killed on all CPUs. - */ - percpu_ref_kill_and_confirm(&css->refcnt, css_killed_ref_fn); - css->cgroup->nr_dying_subsys[ss->id]++; /* * Parent css and cgroup cannot be freed until after the freeing @@ -6086,44 +6088,88 @@ static void kill_css(struct cgroup_subsys_state *css) } /** - * cgroup_destroy_locked - the first stage of cgroup destruction + * kill_css_finish - deferred half of css teardown + * @css: css being killed + * + * See cgroup_destroy_locked(). + */ +static void kill_css_finish(struct cgroup_subsys_state *css) +{ + lockdep_assert_held(&cgroup_mutex); + + /* + * Skip on re-entry: cgroup_apply_control_disable() may have killed @css + * earlier. cgroup_destroy_locked() can still walk it because + * offline_css() (which NULLs cgrp->subsys[ssid]) runs async. + */ + if (percpu_ref_is_dying(&css->refcnt)) + return; + + /* + * Killing would put the base ref, but we need to keep it alive until + * after ->css_offline(). + */ + css_get(css); + + /* + * cgroup core guarantees that, by the time ->css_offline() is invoked, + * no new css reference will be given out via css_tryget_online(). We + * can't simply call percpu_ref_kill() and proceed to offlining css's + * because percpu_ref_kill() doesn't guarantee that the ref is seen as + * killed on all CPUs on return. + * + * Use percpu_ref_kill_and_confirm() to get notifications as each css is + * confirmed to be seen as killed on all CPUs. + */ + percpu_ref_kill_and_confirm(&css->refcnt, css_killed_ref_fn); +} + +/** + * cgroup_destroy_locked - destroy @cgrp (called on rmdir) * @cgrp: cgroup to be destroyed * - * css's make use of percpu refcnts whose killing latency shouldn't be - * exposed to userland and are RCU protected. Also, cgroup core needs to - * guarantee that css_tryget_online() won't succeed by the time - * ->css_offline() is invoked. To satisfy all the requirements, - * destruction is implemented in the following two steps. + * Tear down @cgrp on behalf of rmdir. Constraints: * - * s1. Verify @cgrp can be destroyed and mark it dying. Remove all - * userland visible parts and start killing the percpu refcnts of - * css's. Set up so that the next stage will be kicked off once all - * the percpu refcnts are confirmed to be killed. + * - Userspace: rmdir must succeed when cgroup.procs and friends are empty. * - * s2. Invoke ->css_offline(), mark the cgroup dead and proceed with the - * rest of destruction. Once all cgroup references are gone, the - * cgroup is RCU-freed. + * - Kernel: subsystem ->css_offline() must not run while any task in @cgrp's + * subtree is still doing kernel work. A task hidden from cgroup.procs (past + * exit_signals() with signal->live cleared) can still schedule, allocate, and + * consume resources until its final context switch. Dying descendants in the + * subtree can host such tasks too. * - * This function implements s1. After this step, @cgrp is gone as far as - * the userland is concerned and a new cgroup with the same name may be - * created. As cgroup doesn't care about the names internally, this - * doesn't cause any problem. + * - Kernel: css_tryget_online() must fail by the time ->css_offline() runs. + * + * The destruction runs in three parts: + * + * - This function: synchronous user-visible state teardown plus kill_css_sync() + * on each subsystem css. + * + * - cgroup_finish_destroy(): kicks the percpu_ref kill via kill_css_finish() on + * each subsystem css. Fires once @cgrp's subtree is fully drained, either + * inline here or from cgroup_update_populated(). + * + * - The percpu_ref kill chain: css_killed_ref_fn -> css_killed_work_fn -> + * ->css_offline() -> release/free. + * + * Return 0 on success, -EBUSY if a userspace-visible task or an online child + * remains. */ static int cgroup_destroy_locked(struct cgroup *cgrp) - __releases(&cgroup_mutex) __acquires(&cgroup_mutex) { struct cgroup *tcgrp, *parent = cgroup_parent(cgrp); struct cgroup_subsys_state *css; struct cgrp_cset_link *link; + struct css_task_iter it; + struct task_struct *task; int ssid, ret; lockdep_assert_held(&cgroup_mutex); - /* - * Only migration can raise populated from zero and we're already - * holding cgroup_mutex. - */ - if (cgroup_is_populated(cgrp)) + css_task_iter_start(&cgrp->self, 0, &it); + task = css_task_iter_next(&it); + css_task_iter_end(&it); + if (task) return -EBUSY; /* @@ -6147,9 +6193,8 @@ static int cgroup_destroy_locked(struct cgroup *cgrp) link->cset->dead = true; spin_unlock_irq(&css_set_lock); - /* initiate massacre of all css's */ for_each_css(css, ssid, cgrp) - kill_css(css); + kill_css_sync(css); /* clear and remove @cgrp dir, @cgrp has an extra ref on its kn */ css_clear_dir(&cgrp->self); @@ -6180,79 +6225,27 @@ static int cgroup_destroy_locked(struct cgroup *cgrp) /* put the base reference */ percpu_ref_kill(&cgrp->self.refcnt); + if (!cgroup_is_populated(cgrp)) + cgroup_finish_destroy(cgrp); + return 0; }; /** - * cgroup_drain_dying - wait for dying tasks to leave before rmdir - * @cgrp: the cgroup being removed + * cgroup_finish_destroy - deferred half of @cgrp destruction + * @cgrp: cgroup whose subtree just became empty * - * cgroup.procs and cgroup.threads use css_task_iter which filters out - * PF_EXITING tasks so that userspace doesn't see tasks that have already been - * reaped via waitpid(). However, cgroup_has_tasks() - which tests whether the - * cgroup has non-empty css_sets - is only updated when dying tasks pass through - * cgroup_task_dead() in finish_task_switch(). This creates a window where - * cgroup.procs reads empty but cgroup_has_tasks() is still true, making rmdir - * fail with -EBUSY from cgroup_destroy_locked() even though userspace sees no - * tasks. - * - * This function aligns cgroup_has_tasks() with what userspace can observe. If - * cgroup_has_tasks() but the task iterator sees nothing (all remaining tasks are - * PF_EXITING), we wait for cgroup_task_dead() to finish processing them. As the - * window between PF_EXITING and cgroup_task_dead() is short, the wait is brief. - * - * This function only concerns itself with this cgroup's own dying tasks. - * Whether the cgroup has children is cgroup_destroy_locked()'s problem. - * - * Each cgroup_task_dead() kicks the waitqueue via cset->cgrp_links, and we - * retry the full check from scratch. - * - * Must be called with cgroup_mutex held. + * See cgroup_destroy_locked() for the rationale. */ -static int cgroup_drain_dying(struct cgroup *cgrp) - __releases(&cgroup_mutex) __acquires(&cgroup_mutex) +static void cgroup_finish_destroy(struct cgroup *cgrp) { - struct css_task_iter it; - struct task_struct *task; - DEFINE_WAIT(wait); + struct cgroup_subsys_state *css; + int ssid; lockdep_assert_held(&cgroup_mutex); -retry: - if (!cgroup_has_tasks(cgrp)) - return 0; - /* Same iterator as cgroup.threads - if any task is visible, it's busy */ - css_task_iter_start(&cgrp->self, 0, &it); - task = css_task_iter_next(&it); - css_task_iter_end(&it); - - if (task) - return -EBUSY; - - /* - * All remaining tasks are PF_EXITING and will pass through - * cgroup_task_dead() shortly. Wait for a kick and retry. - * - * cgroup_has_tasks() can't transition from false to true while we're - * holding cgroup_mutex, but the true to false transition happens - * under css_set_lock (via cgroup_task_dead()). We must retest and - * prepare_to_wait() under css_set_lock. Otherwise, the transition - * can happen between our first test and prepare_to_wait(), and we - * sleep with no one to wake us. - */ - spin_lock_irq(&css_set_lock); - if (!cgroup_has_tasks(cgrp)) { - spin_unlock_irq(&css_set_lock); - return 0; - } - prepare_to_wait(&cgrp->dying_populated_waitq, &wait, - TASK_UNINTERRUPTIBLE); - spin_unlock_irq(&css_set_lock); - mutex_unlock(&cgroup_mutex); - schedule(); - finish_wait(&cgrp->dying_populated_waitq, &wait); - mutex_lock(&cgroup_mutex); - goto retry; + for_each_css(css, ssid, cgrp) + kill_css_finish(css); } int cgroup_rmdir(struct kernfs_node *kn) @@ -6264,12 +6257,9 @@ int cgroup_rmdir(struct kernfs_node *kn) if (!cgrp) return 0; - ret = cgroup_drain_dying(cgrp); - if (!ret) { - ret = cgroup_destroy_locked(cgrp); - if (!ret) - TRACE_CGROUP_PATH(rmdir, cgrp); - } + ret = cgroup_destroy_locked(cgrp); + if (!ret) + TRACE_CGROUP_PATH(rmdir, cgrp); cgroup_kn_unlock(kn); return ret; @@ -7029,7 +7019,6 @@ void cgroup_task_exit(struct task_struct *tsk) static void do_cgroup_task_dead(struct task_struct *tsk) { - struct cgrp_cset_link *link; struct css_set *cset; unsigned long flags; @@ -7043,11 +7032,6 @@ static void do_cgroup_task_dead(struct task_struct *tsk) if (thread_group_leader(tsk) && atomic_read(&tsk->signal->live)) list_add_tail(&tsk->cg_list, &cset->dying_tasks); - /* kick cgroup_drain_dying() waiters, see cgroup_rmdir() */ - list_for_each_entry(link, &cset->cgrp_links, cgrp_link) - if (waitqueue_active(&link->cgrp->dying_populated_waitq)) - wake_up(&link->cgrp->dying_populated_waitq); - if (dl_task(tsk)) dec_dl_tasks_cs(tsk); From 360f61bc29085d65a24dbaaf707c802553a239fd Mon Sep 17 00:00:00 2001 From: Peter Rosin Date: Thu, 30 Apr 2026 06:09:58 +0200 Subject: [PATCH 273/525] MAINTAINERS: Update mail for Peter Rosin I'm resigning from my position at Axentia. Signed-off-by: Peter Rosin Signed-off-by: Wolfram Sang --- .mailmap | 1 + MAINTAINERS | 24 +++++++++++------------- 2 files changed, 12 insertions(+), 13 deletions(-) diff --git a/.mailmap b/.mailmap index b78aa092b4bb..eec4a740f7ca 100644 --- a/.mailmap +++ b/.mailmap @@ -682,6 +682,7 @@ Peter A Jonsson Peter Hilber Peter Oruba Peter Oruba +Peter Rosin Pierre-Louis Bossart Pratyush Anand Pratyush Yadav diff --git a/MAINTAINERS b/MAINTAINERS index 882214b0e7db..1b0c453bca46 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4299,18 +4299,16 @@ F: Documentation/devicetree/bindings/leds/backlight/awinic,aw99706.yaml F: drivers/video/backlight/aw99706.c AXENTIA ARM DEVICES -M: Peter Rosin L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) -S: Maintained +S: Orphan F: arch/arm/boot/dts/microchip/at91-linea.dtsi F: arch/arm/boot/dts/microchip/at91-natte.dtsi F: arch/arm/boot/dts/microchip/at91-nattis-2-natte-2.dts F: arch/arm/boot/dts/microchip/at91-tse850-3.dts AXENTIA ASOC DRIVERS -M: Peter Rosin L: linux-sound@vger.kernel.org -S: Maintained +S: Orphan F: Documentation/devicetree/bindings/sound/axentia,* F: sound/soc/atmel/tse850-pcm5142.c @@ -12046,7 +12044,7 @@ F: Documentation/i2c/busses/i2c-nvidia-gpu.rst F: drivers/i2c/busses/i2c-nvidia-gpu.c I2C MUXES -M: Peter Rosin +M: Peter Rosin L: linux-i2c@vger.kernel.org S: Maintained F: Documentation/devicetree/bindings/i2c/i2c-arb* @@ -12447,7 +12445,7 @@ F: drivers/iio/industrialio-backend.c F: include/linux/iio/backend.h IIO DIGITAL POTENTIOMETER DAC -M: Peter Rosin +M: Peter Rosin L: linux-iio@vger.kernel.org S: Maintained F: Documentation/ABI/testing/sysfs-bus-iio-dac-dpot-dac @@ -12455,7 +12453,7 @@ F: Documentation/devicetree/bindings/iio/dac/dpot-dac.yaml F: drivers/iio/dac/dpot-dac.c IIO ENVELOPE DETECTOR -M: Peter Rosin +M: Peter Rosin L: linux-iio@vger.kernel.org S: Maintained F: Documentation/ABI/testing/sysfs-bus-iio-adc-envelope-detector @@ -12471,7 +12469,7 @@ F: include/linux/iio/iio-gts-helper.h F: drivers/iio/test/iio-test-gts.c IIO MULTIPLEXER -M: Peter Rosin +M: Peter Rosin L: linux-iio@vger.kernel.org S: Maintained F: Documentation/devicetree/bindings/iio/multiplexer/io-channel-mux.yaml @@ -12502,7 +12500,7 @@ F: include/linux/iio/ F: tools/iio/ IIO UNIT CONVERTER -M: Peter Rosin +M: Peter Rosin L: linux-iio@vger.kernel.org S: Maintained F: Documentation/devicetree/bindings/iio/afe/current-sense-amplifier.yaml @@ -15718,7 +15716,7 @@ F: Documentation/devicetree/bindings/media/i2c/maxim,max96717.yaml F: drivers/media/i2c/max96717.c MAX9860 MONO AUDIO VOICE CODEC DRIVER -M: Peter Rosin +M: Peter Rosin L: linux-sound@vger.kernel.org S: Maintained F: Documentation/devicetree/bindings/sound/max9860.txt @@ -15933,7 +15931,7 @@ F: Documentation/devicetree/bindings/net/can/microchip,mcp251xfd.yaml F: drivers/net/can/spi/mcp251xfd/ MCP4018 AND MCP4531 MICROCHIP DIGITAL POTENTIOMETER DRIVERS -M: Peter Rosin +M: Peter Rosin L: linux-iio@vger.kernel.org S: Maintained F: Documentation/ABI/testing/sysfs-bus-iio-potentiometer-mcp4531 @@ -18238,7 +18236,7 @@ F: include/linux/mmc/ F: include/uapi/linux/mmc/ MULTIPLEXER SUBSYSTEM -M: Peter Rosin +M: Peter Rosin S: Odd Fixes F: Documentation/ABI/testing/sysfs-class-mux* F: Documentation/devicetree/bindings/mux/ @@ -19347,7 +19345,7 @@ F: include/dt-bindings/display/tda998x.h K: "nxp,tda998x" NXP TFA9879 DRIVER -M: Peter Rosin +M: Peter Rosin L: linux-sound@vger.kernel.org S: Maintained F: Documentation/devicetree/bindings/sound/trivial-codec.yaml From 60f21a2649308bbd84919ba6656d5ccd660953cf Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 27 Apr 2026 14:16:34 -1000 Subject: [PATCH 274/525] cgroup, sched_ext: Include exiting tasks in cgroup iter a72f73c4dd9b ("cgroup: Don't expose dead tasks in cgroup") made css_task_iter_advance() skip exiting tasks so cgroup.procs stays consistent with waitpid() visibility. Unfortunately, this broke scx_task_iter. scx_task_iter walks either scx_tasks (global) or a cgroup subtree via css_task_iter() and the two modes are expected to cover the same set of tasks. After the above change the cgroup-scoped mode silently skips tasks past exit_signals() that are still on scx_tasks. scx_sub_enable_workfn()'s abort path is one of the symptoms: an exiting SCX_TASK_SUB_INIT task can race past the cgroup iter leaking __scx_init_task() state. Other iterations share the same gap. Add CSS_TASK_ITER_WITH_DEAD to opt out of the skip and use it from scx_task_iter(). Fixes: b0e4c2f8a0f0 ("sched_ext: Implement cgroup subtree iteration for scx_task_iter") Reported-by: Cheng-Yang Chou Signed-off-by: Tejun Heo --- include/linux/cgroup.h | 1 + kernel/cgroup/cgroup.c | 8 +++++--- kernel/sched/ext.c | 6 ++++-- 3 files changed, 10 insertions(+), 5 deletions(-) diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h index e52160e85af4..f6d037a30fd8 100644 --- a/include/linux/cgroup.h +++ b/include/linux/cgroup.h @@ -53,6 +53,7 @@ struct kernel_clone_args; enum css_task_iter_flags { CSS_TASK_ITER_PROCS = (1U << 0), /* walk only threadgroup leaders */ CSS_TASK_ITER_THREADED = (1U << 1), /* walk all threaded css_sets in the domain */ + CSS_TASK_ITER_WITH_DEAD = (1U << 2), /* include exiting tasks */ CSS_TASK_ITER_SKIPPED = (1U << 16), /* internal flags */ }; diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 1f084ee71443..e51ce4cd3739 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -5059,10 +5059,12 @@ static void css_task_iter_advance(struct css_task_iter *it) task = list_entry(it->task_pos, struct task_struct, cg_list); /* - * Hide tasks that are exiting but not yet removed. Keep zombie - * leaders with live threads visible. + * Hide tasks that are exiting but not yet removed by default. Keep + * zombie leaders with live threads visible. Usages that need to walk + * every existing task can opt out via CSS_TASK_ITER_WITH_DEAD. */ - if ((task->flags & PF_EXITING) && !atomic_read(&task->signal->live)) + if (!(it->flags & CSS_TASK_ITER_WITH_DEAD) && + (task->flags & PF_EXITING) && !atomic_read(&task->signal->live)) goto repeat; if (it->flags & CSS_TASK_ITER_PROCS) { diff --git a/kernel/sched/ext.c b/kernel/sched/ext.c index 9483be03a4ca..dc5d4787296b 100644 --- a/kernel/sched/ext.c +++ b/kernel/sched/ext.c @@ -766,7 +766,8 @@ static void scx_task_iter_start(struct scx_task_iter *iter, struct cgroup *cgrp) lockdep_assert_held(&cgroup_mutex); iter->cgrp = cgrp; iter->css_pos = css_next_descendant_pre(NULL, &iter->cgrp->self); - css_task_iter_start(iter->css_pos, 0, &iter->css_iter); + css_task_iter_start(iter->css_pos, CSS_TASK_ITER_WITH_DEAD, + &iter->css_iter); return; } #endif @@ -866,7 +867,8 @@ static struct task_struct *scx_task_iter_next(struct scx_task_iter *iter) iter->css_pos = css_next_descendant_pre(iter->css_pos, &iter->cgrp->self); if (iter->css_pos) - css_task_iter_start(iter->css_pos, 0, &iter->css_iter); + css_task_iter_start(iter->css_pos, CSS_TASK_ITER_WITH_DEAD, + &iter->css_iter); } return NULL; } From ff9eda4ea906b1f02fc260ddc42d2d9bd736a49c Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 27 Apr 2026 14:16:35 -1000 Subject: [PATCH 275/525] sched_ext: Skip past-sched_ext_dead() tasks in scx_task_iter_next_locked() scx_task_iter's cgroup-scoped mode can return tasks whose sched_ext_dead() has already completed: cgroup_task_dead() removes from cset->tasks after sched_ext_dead() in finish_task_switch() and is irq-work deferred on PREEMPT_RT. The global mode is fine - sched_ext_dead() removes from scx_tasks via list_del_init() first. Callers (sub-sched enable prep/abort/apply, scx_sub_disable(), scx_fail_parent()) assume returned tasks are still on @sch and trip WARN_ON_ONCE() or operate on torn-down state otherwise. Set %SCX_TASK_OFF_TASKS in sched_ext_dead() under @p's rq lock and have scx_task_iter_next_locked() skip flagged tasks under the same lock. Setter and reader serialize on the per-task rq lock - no race. Signed-off-by: Tejun Heo --- include/linux/sched/ext.h | 1 + kernel/sched/ext.c | 35 ++++++++++++++++++++++++++--------- 2 files changed, 27 insertions(+), 9 deletions(-) diff --git a/include/linux/sched/ext.h b/include/linux/sched/ext.h index 1a3af2ea2a79..adb9a4de068a 100644 --- a/include/linux/sched/ext.h +++ b/include/linux/sched/ext.h @@ -101,6 +101,7 @@ enum scx_ent_flags { SCX_TASK_DEQD_FOR_SLEEP = 1 << 3, /* last dequeue was for SLEEP */ SCX_TASK_SUB_INIT = 1 << 4, /* task being initialized for a sub sched */ SCX_TASK_IMMED = 1 << 5, /* task is on local DSQ with %SCX_ENQ_IMMED */ + SCX_TASK_OFF_TASKS = 1 << 6, /* removed from scx_tasks by sched_ext_dead() */ /* * Bits 8 and 9 are used to carry task state: diff --git a/kernel/sched/ext.c b/kernel/sched/ext.c index dc5d4787296b..3f0d8aeaed81 100644 --- a/kernel/sched/ext.c +++ b/kernel/sched/ext.c @@ -928,16 +928,27 @@ static struct task_struct *scx_task_iter_next_locked(struct scx_task_iter *iter) * * Test for idle_sched_class as only init_tasks are on it. */ - if (p->sched_class != &idle_sched_class) - break; + if (p->sched_class == &idle_sched_class) + continue; + + iter->rq = task_rq_lock(p, &iter->rf); + iter->locked_task = p; + + /* + * cgroup_task_dead() removes the dead tasks from cset->tasks + * after sched_ext_dead() and cgroup iteration may see tasks + * which already finished sched_ext_dead(). %SCX_TASK_OFF_TASKS + * is set by sched_ext_dead() under @p's rq lock. Test it to + * avoid visiting tasks which are already dead from SCX POV. + */ + if (p->scx.flags & SCX_TASK_OFF_TASKS) { + __scx_task_iter_rq_unlock(iter); + continue; + } + + return p; } - if (!p) - return NULL; - - iter->rq = task_rq_lock(p, &iter->rf); - iter->locked_task = p; - - return p; + return NULL; } /** @@ -3850,6 +3861,11 @@ void sched_ext_dead(struct task_struct *p) /* * @p is off scx_tasks and wholly ours. scx_root_enable()'s READY -> * ENABLED transitions can't race us. Disable ops for @p. + * + * %SCX_TASK_OFF_TASKS synchronizes against cgroup task iteration - see + * scx_task_iter_next_locked(). NONE tasks need no marking: cgroup + * iteration is only used from sub-sched paths, which require root + * enabled. Root enable transitions every live task to at least READY. */ if (scx_get_task_state(p) != SCX_TASK_NONE) { struct rq_flags rf; @@ -3857,6 +3873,7 @@ void sched_ext_dead(struct task_struct *p) rq = task_rq_lock(p, &rf); scx_disable_and_exit_task(scx_task_sched(p), p); + p->scx.flags |= SCX_TASK_OFF_TASKS; task_rq_unlock(rq, p, &rf); } } From 84ae1840260fece9b6b70d3872b79384bbe5a90b Mon Sep 17 00:00:00 2001 From: Osama Abdelkader Date: Thu, 23 Apr 2026 22:06:19 +0200 Subject: [PATCH 276/525] drm/sti: remove bridge when sti_hda component_add fails MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Use devm_drm_bridge_add() so the bridge is released if probe fails after registration, and drop the manual drm_bridge_remove() in remove(). Check the return value of devm_drm_bridge_add(). Signed-off-by: Osama Abdelkader Fixes: d28726efc637 ("drm/sti: hda: add bridge before attaching") Cc: stable@vger.kernel.org Reviewed-by: Luca Ceresoli Acked-by: Raphaël Gallais-Pou Link: https://patch.msgid.link/20260423200622.325076-1-osama.abdelkader@gmail.com Signed-off-by: Raphael Gallais-Pou --- drivers/gpu/drm/sti/sti_hda.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/sti/sti_hda.c b/drivers/gpu/drm/sti/sti_hda.c index b7397827889c..360a88ca8f0c 100644 --- a/drivers/gpu/drm/sti/sti_hda.c +++ b/drivers/gpu/drm/sti/sti_hda.c @@ -741,6 +741,7 @@ static int sti_hda_probe(struct platform_device *pdev) struct device *dev = &pdev->dev; struct sti_hda *hda; struct resource *res; + int ret; DRM_INFO("%s\n", __func__); @@ -779,7 +780,9 @@ static int sti_hda_probe(struct platform_device *pdev) return PTR_ERR(hda->clk_hddac); } - drm_bridge_add(&hda->bridge); + ret = devm_drm_bridge_add(dev, &hda->bridge); + if (ret) + return ret; platform_set_drvdata(pdev, hda); @@ -788,10 +791,7 @@ static int sti_hda_probe(struct platform_device *pdev) static void sti_hda_remove(struct platform_device *pdev) { - struct sti_hda *hda = platform_get_drvdata(pdev); - component_del(&pdev->dev, &sti_hda_ops); - drm_bridge_remove(&hda->bridge); } static const struct of_device_id hda_of_match[] = { From b34c82777a2c0648ee053595f4b290fd5249b093 Mon Sep 17 00:00:00 2001 From: David Carlier Date: Thu, 30 Apr 2026 10:27:47 +0100 Subject: [PATCH 277/525] sched_ext: idle: Recheck prev_cpu after narrowing allowed mask scx_select_cpu_dfl() narrows @allowed to @cpus_allowed & @p->cpus_ptr when the BPF caller supplies a @cpus_allowed that differs from @p->cpus_ptr and @p doesn't have full affinity. However, @is_prev_allowed was computed against the original (wider) @cpus_allowed, so the prev_cpu fast paths could pick a @prev_cpu that is in @cpus_allowed but not in @p->cpus_ptr, violating the intended invariant that the returned CPU is always usable by @p. The kernel masks this via the SCX_EV_SELECT_CPU_FALLBACK fallback, but the behavior contradicts the documented contract. Move the @is_prev_allowed evaluation past the narrowing block so it tests against the final @allowed mask. Fixes: ee9a4e92799d ("sched_ext: idle: Properly handle invalid prev_cpu during idle selection") Cc: stable@vger.kernel.org # v6.16+ Assisted-by: Claude Signed-off-by: David Carlier Reviewed-by: Andrea Righi Signed-off-by: Tejun Heo --- kernel/sched/ext_idle.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/kernel/sched/ext_idle.c b/kernel/sched/ext_idle.c index 7468560a6d80..6e1980763270 100644 --- a/kernel/sched/ext_idle.c +++ b/kernel/sched/ext_idle.c @@ -465,12 +465,6 @@ s32 scx_select_cpu_dfl(struct task_struct *p, s32 prev_cpu, u64 wake_flags, preempt_disable(); - /* - * Check whether @prev_cpu is still within the allowed set. If not, - * we can still try selecting a nearby CPU. - */ - is_prev_allowed = cpumask_test_cpu(prev_cpu, allowed); - /* * Determine the subset of CPUs usable by @p within @cpus_allowed. */ @@ -487,6 +481,12 @@ s32 scx_select_cpu_dfl(struct task_struct *p, s32 prev_cpu, u64 wake_flags, } } + /* + * Check whether @prev_cpu is still within the allowed set. If not, + * we can still try selecting a nearby CPU. + */ + is_prev_allowed = cpumask_test_cpu(prev_cpu, allowed); + /* * This is necessary to protect llc_cpus. */ From d8769544bde51b0ac980d10f8fe9f9fed6c95995 Mon Sep 17 00:00:00 2001 From: "T.J. Mercier" Date: Thu, 30 Apr 2026 13:11:42 -0700 Subject: [PATCH 278/525] docs: cgroup-v1: Update charge-commit section Commit 1d8f136a421f ("memcg/hugetlb: remove memcg hugetlb try-commit-cancel protocol") removed mem_cgroup_commit_charge() and mem_cgroup_cancel_charge(), but the docs still refer to those functions. There is no longer any charge cancellation. Update the docs to match the code. Signed-off-by: T.J. Mercier Signed-off-by: Tejun Heo --- Documentation/admin-guide/cgroup-v1/memcg_test.rst | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/Documentation/admin-guide/cgroup-v1/memcg_test.rst b/Documentation/admin-guide/cgroup-v1/memcg_test.rst index 9f8e27355cba..7c7cd457cf69 100644 --- a/Documentation/admin-guide/cgroup-v1/memcg_test.rst +++ b/Documentation/admin-guide/cgroup-v1/memcg_test.rst @@ -47,21 +47,19 @@ Please note that implementation details can be changed. Called when swp_entry's refcnt goes down to 0. A charge against swap disappears. -3. charge-commit-cancel +3. charge-commit ======================= Memcg pages are charged in two steps: - mem_cgroup_try_charge() - - mem_cgroup_commit_charge() or mem_cgroup_cancel_charge() + - commit_charge() At try_charge(), there are no flags to say "this page is charged". at this point, usage += PAGE_SIZE. At commit(), the page is associated with the memcg. - At cancel(), simply usage -= PAGE_SIZE. - Under below explanation, we assume CONFIG_SWAP=y. 4. Anonymous From a200cdbf95932631ec338d08a6e9e31b34c4e8a6 Mon Sep 17 00:00:00 2001 From: Qingfang Deng Date: Mon, 27 Apr 2026 12:00:11 +0800 Subject: [PATCH 279/525] ovpn: reset MAC header before passing skb up After decapsulating a packet, the skb->mac_header still points to the outer transport header. Fix this by calling skb_reset_mac_header() in ovpn_netdev_write() to ensure the MAC header points to the beginning of the inner IP/network packet, as expected by the rest of the stack. Reported-by: Minqiang Chen Fixes: 8534731dbf2d ("ovpn: implement packet processing") Signed-off-by: Qingfang Deng Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/io.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ovpn/io.c b/drivers/net/ovpn/io.c index db43a1f8a07a..d92bb87be2b2 100644 --- a/drivers/net/ovpn/io.c +++ b/drivers/net/ovpn/io.c @@ -85,6 +85,7 @@ static void ovpn_netdev_write(struct ovpn_peer *peer, struct sk_buff *skb) skb_scrub_packet(skb, true); /* network header reset in ovpn_decrypt_post() */ + skb_reset_mac_header(skb); skb_reset_transport_header(skb); skb_reset_inner_headers(skb); From c539cb30f93f119566f2ae9d016cce11f188d780 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Wed, 25 Mar 2026 17:49:18 +0100 Subject: [PATCH 280/525] ovpn: ensure packet delivery happens with BH disabled ovpn injects decrypted packets into the netdev RX path through ovpn_netdev_write() which invokes gro_cells_receive() and dev_dstats_rx_add(). ovpn_netdev_write() is normally called in softirq context, however, in case of TCP connections it may also be invoked process context. When this happens gro_cells_receive() will throw a warning: [ 230.183747][ T12] WARNING: net/core/gro_cells.c:30 at gro_cells_receive+0x708/0xaa0, CPU#1: kworker/u16:0/12 and lockdep will also report a potential inconsistent lock state: WARNING: inconsistent lock state 7.0.0-rc4+ #246 Tainted: G W -------------------------------- inconsistent {IN-SOFTIRQ-W} -> {SOFTIRQ-ON-W} usage. because attempts to acquire gro_cells->bh_lock by both contexts may lead to a deadlock. At the same time, dev_dstats_rx_add() does not expect to race with a softirq (which may happen when invoked in process context), because the latter may access its per-cpu state and corrupt it. Fix all this by invoking local_bh_disable/enable() around gro_cells_receive() and dev_dstats_rx_add() to ensure that bottom halves are always disabled before calling both of them. Fixes: 11851cbd60ea ("ovpn: implement TCP transport") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/io.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/ovpn/io.c b/drivers/net/ovpn/io.c index d92bb87be2b2..22c555dd962e 100644 --- a/drivers/net/ovpn/io.c +++ b/drivers/net/ovpn/io.c @@ -91,12 +91,18 @@ static void ovpn_netdev_write(struct ovpn_peer *peer, struct sk_buff *skb) /* cause packet to be "received" by the interface */ pkt_len = skb->len; + /* we may get here in process context in case of TCP connections, + * therefore we have to disable BHs to ensure gro_cells_receive() + * and dev_dstats_rx_add() do not get corrupted or enter deadlock + */ + local_bh_disable(); ret = gro_cells_receive(&peer->ovpn->gro_cells, skb); if (likely(ret == NET_RX_SUCCESS)) { /* update RX stats with the size of decrypted packet */ ovpn_peer_stats_increment_rx(&peer->vpn_stats, pkt_len); dev_dstats_rx_add(peer->ovpn->dev, pkt_len); } + local_bh_enable(); } void ovpn_decrypt_post(void *data, int ret) From 201ba706318d460a2ea660e3652610be62532a70 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Wed, 29 Apr 2026 10:00:16 +0200 Subject: [PATCH 281/525] selftests: ovpn: reduce ping count in test.sh The second stage of test.sh ("run baseline data traffic") performs a basic connectivity check with ping -qfc 500 -w 3. On slower CI instances this is too strict for TCP: the RTT is high enough that 500 echo requests do not reliably complete within 3 seconds, so the stage flakes and the test fails even though the ovpn setup is healthy. Reduce the packet count to 100 for both the plain and 3000-byte pings in that stage. This still verifies peer setup, key exchange, routing, and data-path traffic, without making the basic connectivity check depend on timing out under load. Fixes: 959bc330a439 ("testing/selftests: add test tool and scripts for ovpn module") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- tools/testing/selftests/net/ovpn/test.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/net/ovpn/test.sh b/tools/testing/selftests/net/ovpn/test.sh index b50dbe45a4d0..c06e3135fbef 100755 --- a/tools/testing/selftests/net/ovpn/test.sh +++ b/tools/testing/selftests/net/ovpn/test.sh @@ -98,10 +98,10 @@ ovpn_run_basic_traffic() { sleep 0.3 ovpn_cmd_ok "send baseline traffic to peer ${p}" \ ip netns exec ovpn_peer0 \ - ping -qfc 500 -w 3 5.5.5.$((p + 1)) + ping -qfc 100 -w 3 5.5.5.$((p + 1)) ovpn_cmd_ok "send large-payload traffic to peer ${p}" \ ip netns exec ovpn_peer0 \ - ping -qfc 500 -s 3000 -w 3 5.5.5.$((p + 1)) + ping -qfc 100 -s 3000 -w 3 5.5.5.$((p + 1)) wait "${tcpdump_pid1}" || return 1 wait "${tcpdump_pid2}" || return 1 From afbd961305eb483515650ccfcb7743608e7add78 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Thu, 30 Apr 2026 10:44:13 +0300 Subject: [PATCH 282/525] ipvs: fixes for the new ip_vs_status info Sashiko reports some problems for the recently added /proc/net/ip_vs_status: * ip_vs_status_show() as a table reader may run long after the conn_tab and svc_table table are released. While ip_vs_conn_flush() properly changes the conn_tab_changes counter when conn_tab is removed, ip_vs_del_service() and ip_vs_flush() were missing such change for the svc_table_changes counter. As result, readers like ip_vs_dst_event() and ip_vs_status_show() may continue to use a freed table after a cond_resched_rcu() call. * While counting the buckets in ip_vs_status_show() make sure we traverse only the needed number of entries in the chain. This also prevents possible overflow of the 'count' variable. * Add check for 'loops' to prevent infinite loops while restarting the traversal on table change. * While IP_VS_CONN_TAB_MAX_BITS is 20 on 32-bit platforms and there is no risk to overflow when multiplying the number of conn_tab buckets to 100, prefer the div_u64() helper to make the following dividing safer. * Use 0440 permissions for ip_vs_status to restrict the info only to root due to the exported information for hash distribution. Link: https://sashiko.dev/#/patchset/20260410112352.23599-1-fw%40strlen.de Fixes: 9a9ccef907a7 ("ipvs: add ip_vs_status info") Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_ctl.c | 51 ++++++++++++++++++++++++---------- 1 file changed, 36 insertions(+), 15 deletions(-) diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index 6632daa87ded..27e50afe9a54 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -2032,6 +2032,9 @@ static int ip_vs_del_service(struct ip_vs_service *svc) cancel_delayed_work_sync(&ipvs->svc_resize_work); if (t) { rcu_assign_pointer(ipvs->svc_table, NULL); + /* Inform readers that table is removed */ + smp_mb__before_atomic(); + atomic_inc(&ipvs->svc_table_changes); while (1) { p = rcu_dereference_protected(t->new_tbl, 1); call_rcu(&t->rcu_head, ip_vs_rht_rcu_free); @@ -2078,6 +2081,9 @@ static int ip_vs_flush(struct netns_ipvs *ipvs, bool cleanup) t = rcu_dereference_protected(ipvs->svc_table, 1); if (t) { rcu_assign_pointer(ipvs->svc_table, NULL); + /* Inform readers that table is removed */ + smp_mb__before_atomic(); + atomic_inc(&ipvs->svc_table_changes); while (1) { p = rcu_dereference_protected(t->new_tbl, 1); call_rcu(&t->rcu_head, ip_vs_rht_rcu_free); @@ -3004,7 +3010,8 @@ static int ip_vs_status_show(struct seq_file *seq, void *v) int old_gen, new_gen; u32 counts[8]; u32 bucket; - int count; + u32 count; + int loops; u32 sum1; u32 sum; int i; @@ -3020,6 +3027,7 @@ static int ip_vs_status_show(struct seq_file *seq, void *v) if (!atomic_read(&ipvs->conn_count)) goto after_conns; old_gen = atomic_read(&ipvs->conn_tab_changes); + loops = 0; repeat_conn: smp_rmb(); /* ipvs->conn_tab and conn_tab_changes */ @@ -3032,8 +3040,11 @@ static int ip_vs_status_show(struct seq_file *seq, void *v) resched_score++; ip_vs_rht_walk_bucket_rcu(t, bucket, head) { count = 0; - hlist_bl_for_each_entry_rcu(hn, e, head, node) + hlist_bl_for_each_entry_rcu(hn, e, head, node) { count++; + if (count >= ARRAY_SIZE(counts) - 1) + break; + } } resched_score += count; if (resched_score >= 100) { @@ -3042,37 +3053,41 @@ static int ip_vs_status_show(struct seq_file *seq, void *v) new_gen = atomic_read(&ipvs->conn_tab_changes); /* New table installed ? */ if (old_gen != new_gen) { + /* Too many changes? */ + if (++loops >= 5) + goto after_conns; old_gen = new_gen; goto repeat_conn; } } - counts[min(count, (int)ARRAY_SIZE(counts) - 1)]++; + counts[count]++; } } for (sum = 0, i = 0; i < ARRAY_SIZE(counts); i++) sum += counts[i]; sum1 = sum - counts[0]; - seq_printf(seq, "Conn buckets empty:\t%u (%lu%%)\n", - counts[0], (unsigned long)counts[0] * 100 / max(sum, 1U)); + seq_printf(seq, "Conn buckets empty:\t%u (%llu%%)\n", + counts[0], div_u64((u64)counts[0] * 100U, max(sum, 1U))); for (i = 1; i < ARRAY_SIZE(counts); i++) { if (!counts[i]) continue; - seq_printf(seq, "Conn buckets len-%d:\t%u (%lu%%)\n", + seq_printf(seq, "Conn buckets len-%d:\t%u (%llu%%)\n", i, counts[i], - (unsigned long)counts[i] * 100 / max(sum1, 1U)); + div_u64((u64)counts[i] * 100U, max(sum1, 1U))); } after_conns: t = rcu_dereference(ipvs->svc_table); count = ip_vs_get_num_services(ipvs); - seq_printf(seq, "Services:\t%d\n", count); + seq_printf(seq, "Services:\t%u\n", count); seq_printf(seq, "Service buckets:\t%d (%d bits, lfactor %d)\n", t ? t->size : 0, t ? t->bits : 0, t ? t->lfactor : 0); if (!count) goto after_svc; old_gen = atomic_read(&ipvs->svc_table_changes); + loops = 0; repeat_svc: smp_rmb(); /* ipvs->svc_table and svc_table_changes */ @@ -3086,8 +3101,11 @@ static int ip_vs_status_show(struct seq_file *seq, void *v) ip_vs_rht_walk_bucket_rcu(t, bucket, head) { count = 0; hlist_bl_for_each_entry_rcu(svc, e, head, - s_list) + s_list) { count++; + if (count >= ARRAY_SIZE(counts) - 1) + break; + } } resched_score += count; if (resched_score >= 100) { @@ -3096,24 +3114,27 @@ static int ip_vs_status_show(struct seq_file *seq, void *v) new_gen = atomic_read(&ipvs->svc_table_changes); /* New table installed ? */ if (old_gen != new_gen) { + /* Too many changes? */ + if (++loops >= 5) + goto after_svc; old_gen = new_gen; goto repeat_svc; } } - counts[min(count, (int)ARRAY_SIZE(counts) - 1)]++; + counts[count]++; } } for (sum = 0, i = 0; i < ARRAY_SIZE(counts); i++) sum += counts[i]; sum1 = sum - counts[0]; - seq_printf(seq, "Service buckets empty:\t%u (%lu%%)\n", - counts[0], (unsigned long)counts[0] * 100 / max(sum, 1U)); + seq_printf(seq, "Service buckets empty:\t%u (%llu%%)\n", + counts[0], div_u64((u64)counts[0] * 100U, max(sum, 1U))); for (i = 1; i < ARRAY_SIZE(counts); i++) { if (!counts[i]) continue; - seq_printf(seq, "Service buckets len-%d:\t%u (%lu%%)\n", + seq_printf(seq, "Service buckets len-%d:\t%u (%llu%%)\n", i, counts[i], - (unsigned long)counts[i] * 100 / max(sum1, 1U)); + div_u64((u64)counts[i] * 100U, max(sum1, 1U))); } after_svc: @@ -5039,7 +5060,7 @@ int __net_init ip_vs_control_net_init(struct netns_ipvs *ipvs) ipvs->net->proc_net, ip_vs_stats_percpu_show, NULL)) goto err_percpu; - if (!proc_create_net_single("ip_vs_status", 0, ipvs->net->proc_net, + if (!proc_create_net_single("ip_vs_status", 0440, ipvs->net->proc_net, ip_vs_status_show, NULL)) goto err_status; #endif From f2da9a96abb4b7a64626e931cedd85f05d5498ca Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Thu, 30 Apr 2026 10:44:14 +0300 Subject: [PATCH 283/525] ipvs: fix races around the conn_lfactor and svc_lfactor sysctl vars Sashiko warns that the new sysctls vars can be changed after the hash tables are destroyed and their respective resizing works canceled, leading to mod_delayed_work() being called for canceled works. Solve this in different ways. conn_tab can be present even without services and is destroyed only on netns exit, so use disable_delayed_work_sync() to disable the work instead of adding more synchronization mechanisms. As for the svc_table, it is destroyed when the services are deleted, so we must be sure that netns exit is not called yet (the check for 'enable') and the work is not canceled by checking all under same mutex lock. Also, use WRITE_ONCE when updating the sysctl vars as we already read them with READ_ONCE. Link: https://sashiko.dev/#/patchset/20260410112352.23599-1-fw%40strlen.de Fixes: 8d7de5477e47 ("ipvs: add conn_lfactor and svc_lfactor sysctl vars") Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_conn.c | 2 +- net/netfilter/ipvs/ip_vs_ctl.c | 12 +++++++++--- 2 files changed, 10 insertions(+), 4 deletions(-) diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c index 2082bfb2d93c..84a4921a7865 100644 --- a/net/netfilter/ipvs/ip_vs_conn.c +++ b/net/netfilter/ipvs/ip_vs_conn.c @@ -1835,7 +1835,7 @@ static void ip_vs_conn_flush(struct netns_ipvs *ipvs) if (!rcu_dereference_protected(ipvs->conn_tab, 1)) return; - cancel_delayed_work_sync(&ipvs->conn_resize_work); + disable_delayed_work_sync(&ipvs->conn_resize_work); if (!atomic_read(&ipvs->conn_count)) goto unreg; diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index 27e50afe9a54..caec516856e9 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -2469,7 +2469,7 @@ static int ipvs_proc_conn_lfactor(const struct ctl_table *table, int write, if (val < -8 || val > 8) { ret = -EINVAL; } else { - *valp = val; + WRITE_ONCE(*valp, val); if (rcu_access_pointer(ipvs->conn_tab)) mod_delayed_work(system_unbound_wq, &ipvs->conn_resize_work, 0); @@ -2496,10 +2496,16 @@ static int ipvs_proc_svc_lfactor(const struct ctl_table *table, int write, if (val < -8 || val > 8) { ret = -EINVAL; } else { - *valp = val; - if (rcu_access_pointer(ipvs->svc_table)) + mutex_lock(&ipvs->service_mutex); + WRITE_ONCE(*valp, val); + /* Make sure the services are present */ + if (rcu_access_pointer(ipvs->svc_table) && + READ_ONCE(ipvs->enable) && + !test_bit(IP_VS_WORK_SVC_NORESIZE, + &ipvs->work_flags)) mod_delayed_work(system_unbound_wq, &ipvs->svc_resize_work, 0); + mutex_unlock(&ipvs->service_mutex); } } return ret; From d493d9de1c21313cf62be0f6e1a4d48385fa7beb Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Thu, 30 Apr 2026 10:44:15 +0300 Subject: [PATCH 284/525] ipvs: fix the spin_lock usage for RT build syzbot reports for sleeping function called from invalid context [1]. The recently added code for resizable hash tables uses hlist_bl bit locks in combination with spin_lock for the connection fields (cp->lock). Fix the following problems: * avoid using spin_lock(&cp->lock) under locked bit lock because it sleeps on PREEMPT_RT * as the recent changes call ip_vs_conn_hash() only for newly allocated connection, the spin_lock can be removed there because the connection is still not linked to table and does not need cp->lock protection. * the lock can be removed also from ip_vs_conn_unlink() where we are the last connection user. * the last place that is fixed is ip_vs_conn_fill_cport() where now the cp->lock is locked before the other locks to ensure other packets do not modify the cp->flags in non-atomic way. Here we make sure cport and flags are changed only once if two or more packets race to fill the cport. Also, we fill cport early, so that if we race with resizing there will be valid cport key for the hashing. Add a warning if too many hash table changes occur for our RCU read-side critical section which is error condition but minor because the connection still can expire gracefully. Still, restore the cport to 0 to allow retransmitted packet to properly fill the cport. Problems reported by Sashiko. [1]: BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48 in_atomic(): 1, irqs_disabled(): 0, non_block: 0, pid: 16, name: ktimers/0 preempt_count: 2, expected: 0 RCU nest depth: 3, expected: 3 8 locks held by ktimers/0/16: #0: ffffffff8de5f260 (local_bh){.+.+}-{1:3}, at: __local_bh_disable_ip+0x3c/0x420 kernel/softirq.c:163 #1: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: __local_bh_disable_ip+0x3c/0x420 kernel/softirq.c:163 #2: ffff8880b8826360 (&base->expiry_lock){+...}-{3:3}, at: spin_lock include/linux/spinlock_rt.h:45 [inline] #2: ffff8880b8826360 (&base->expiry_lock){+...}-{3:3}, at: timer_base_lock_expiry kernel/time/timer.c:1502 [inline] #2: ffff8880b8826360 (&base->expiry_lock){+...}-{3:3}, at: __run_timer_base+0x120/0x9f0 kernel/time/timer.c:2384 #3: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: rcu_lock_acquire include/linux/rcupdate.h:300 [inline] #3: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: rcu_read_lock include/linux/rcupdate.h:838 [inline] #3: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: __rt_spin_lock kernel/locking/spinlock_rt.c:50 [inline] #3: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: rt_spin_lock+0x1e0/0x400 kernel/locking/spinlock_rt.c:57 #4: ffffc90000157a80 ((&cp->timer)){+...}-{0:0}, at: call_timer_fn+0xd4/0x5e0 kernel/time/timer.c:1745 #5: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: rcu_lock_acquire include/linux/rcupdate.h:300 [inline] #5: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: rcu_read_lock include/linux/rcupdate.h:838 [inline] #5: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: ip_vs_conn_unlink net/netfilter/ipvs/ip_vs_conn.c:315 [inline] #5: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: ip_vs_conn_expire+0x257/0x2390 net/netfilter/ipvs/ip_vs_conn.c:1260 #6: ffffffff8de5f260 (local_bh){.+.+}-{1:3}, at: __local_bh_disable_ip+0x3c/0x420 kernel/softirq.c:163 #7: ffff888068d4c3f0 (&cp->lock#2){+...}-{3:3}, at: spin_lock include/linux/spinlock_rt.h:45 [inline] #7: ffff888068d4c3f0 (&cp->lock#2){+...}-{3:3}, at: ip_vs_conn_unlink net/netfilter/ipvs/ip_vs_conn.c:324 [inline] #7: ffff888068d4c3f0 (&cp->lock#2){+...}-{3:3}, at: ip_vs_conn_expire+0xd4a/0x2390 net/netfilter/ipvs/ip_vs_conn.c:1260 Preemption disabled at: [] bit_spin_lock include/linux/bit_spinlock.h:38 [inline] [] hlist_bl_lock+0x18/0x110 include/linux/list_bl.h:149 CPU: 0 UID: 0 PID: 16 Comm: ktimers/0 Tainted: G W L syzkaller #0 PREEMPT_{RT,(full)} Tainted: [W]=WARN, [L]=SOFTLOCKUP Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 03/18/2026 Call Trace: dump_stack_lvl+0xe8/0x150 lib/dump_stack.c:120 __might_resched+0x329/0x480 kernel/sched/core.c:9162 __rt_spin_lock kernel/locking/spinlock_rt.c:48 [inline] rt_spin_lock+0xc2/0x400 kernel/locking/spinlock_rt.c:57 spin_lock include/linux/spinlock_rt.h:45 [inline] ip_vs_conn_unlink net/netfilter/ipvs/ip_vs_conn.c:324 [inline] ip_vs_conn_expire+0xd4a/0x2390 net/netfilter/ipvs/ip_vs_conn.c:1260 call_timer_fn+0x192/0x5e0 kernel/time/timer.c:1748 expire_timers kernel/time/timer.c:1799 [inline] __run_timers kernel/time/timer.c:2374 [inline] __run_timer_base+0x6a3/0x9f0 kernel/time/timer.c:2386 run_timer_base kernel/time/timer.c:2395 [inline] run_timer_softirq+0xb7/0x170 kernel/time/timer.c:2405 handle_softirqs+0x1de/0x6d0 kernel/softirq.c:622 __do_softirq kernel/softirq.c:656 [inline] run_ktimerd+0x69/0x100 kernel/softirq.c:1151 smpboot_thread_fn+0x541/0xa50 kernel/smpboot.c:160 kthread+0x388/0x470 kernel/kthread.c:436 ret_from_fork+0x514/0xb70 arch/x86/kernel/process.c:158 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245 Reported-by: syzbot+504e778ddaecd36fdd17@syzkaller.appspotmail.com Link: https://sashiko.dev/#/patchset/20260415200216.79699-1-ja%40ssi.bg Link: https://sashiko.dev/#/patchset/20260420165539.85174-4-ja%40ssi.bg Link: https://sashiko.dev/#/patchset/20260422135823.50489-4-ja%40ssi.bg Fixes: 2fa7cc9c7025 ("ipvs: switch to per-net connection table") Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_conn.c | 72 ++++++++++++++++++--------------- 1 file changed, 40 insertions(+), 32 deletions(-) diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c index 84a4921a7865..9ea6b4fa78bf 100644 --- a/net/netfilter/ipvs/ip_vs_conn.c +++ b/net/netfilter/ipvs/ip_vs_conn.c @@ -267,27 +267,20 @@ static inline int ip_vs_conn_hash(struct ip_vs_conn *cp) hash_key2 = hash_key; use2 = false; } + conn_tab_lock(t, cp, hash_key, hash_key2, use2, true /* new_hash */, &head, &head2); - spin_lock(&cp->lock); - if (!(cp->flags & IP_VS_CONN_F_HASHED)) { - cp->flags |= IP_VS_CONN_F_HASHED; - WRITE_ONCE(cp->hn0.hash_key, hash_key); - WRITE_ONCE(cp->hn1.hash_key, hash_key2); - refcount_inc(&cp->refcnt); - hlist_bl_add_head_rcu(&cp->hn0.node, head); - if (use2) - hlist_bl_add_head_rcu(&cp->hn1.node, head2); - ret = 1; - } else { - pr_err("%s(): request for already hashed, called from %pS\n", - __func__, __builtin_return_address(0)); - ret = 0; - } + cp->flags |= IP_VS_CONN_F_HASHED; + WRITE_ONCE(cp->hn0.hash_key, hash_key); + WRITE_ONCE(cp->hn1.hash_key, hash_key2); + refcount_inc(&cp->refcnt); + hlist_bl_add_head_rcu(&cp->hn0.node, head); + if (use2) + hlist_bl_add_head_rcu(&cp->hn1.node, head2); - spin_unlock(&cp->lock); conn_tab_unlock(head, head2); + ret = 1; /* Schedule resizing if load increases */ if (atomic_read(&ipvs->conn_count) > t->u_thresh && @@ -321,7 +314,6 @@ static inline bool ip_vs_conn_unlink(struct ip_vs_conn *cp) conn_tab_lock(t, cp, hash_key, hash_key2, use2, false /* new_hash */, &head, &head2); - spin_lock(&cp->lock); if (cp->flags & IP_VS_CONN_F_HASHED) { /* Decrease refcnt and unlink conn only if we are last user */ @@ -334,7 +326,6 @@ static inline bool ip_vs_conn_unlink(struct ip_vs_conn *cp) } } - spin_unlock(&cp->lock); conn_tab_unlock(head, head2); rcu_read_unlock(); @@ -637,6 +628,7 @@ void ip_vs_conn_fill_cport(struct ip_vs_conn *cp, __be16 cport) struct ip_vs_conn_hnode *hn; u32 hash_key, hash_key_new; struct ip_vs_conn_param p; + bool by_me = false; int ntbl; int dir; @@ -664,8 +656,16 @@ void ip_vs_conn_fill_cport(struct ip_vs_conn *cp, __be16 cport) t = rcu_dereference(t->new_tbl); ntbl++; /* We are lost? */ - if (ntbl >= 2) + if (ntbl >= 2) { + spin_lock_bh(&cp->lock); + if (cp->flags & IP_VS_CONN_F_NO_CPORT && by_me) + cp->cport = 0; + /* hn1 will be rehashed on next packet */ + spin_unlock_bh(&cp->lock); + IP_VS_ERR_RL("%s(): Too many ht changes for dir %d\n", + __func__, dir); return; + } } /* Rehashing during resize? Use the recent table for adds */ @@ -683,10 +683,13 @@ void ip_vs_conn_fill_cport(struct ip_vs_conn *cp, __be16 cport) if (head > head2 && t == t2) swap(head, head2); + /* Protect the cp->flags modification */ + spin_lock_bh(&cp->lock); + /* Lock seqcount only for the old bucket, even if we are on new table * because it affects the del operation, not the adding. */ - spin_lock_bh(&t->lock[hash_key & t->lock_mask].l); + spin_lock(&t->lock[hash_key & t->lock_mask].l); preempt_disable_nested(); write_seqcount_begin(&t->seqc[hash_key & t->seqc_mask]); @@ -704,14 +707,23 @@ void ip_vs_conn_fill_cport(struct ip_vs_conn *cp, __be16 cport) hlist_bl_unlock(head); write_seqcount_end(&t->seqc[hash_key & t->seqc_mask]); preempt_enable_nested(); - spin_unlock_bh(&t->lock[hash_key & t->lock_mask].l); + spin_unlock(&t->lock[hash_key & t->lock_mask].l); + spin_unlock_bh(&cp->lock); hash_key = hash_key_new; goto retry; } - spin_lock(&cp->lock); - if ((cp->flags & IP_VS_CONN_F_NO_CPORT) && - (cp->flags & IP_VS_CONN_F_HASHED)) { + /* Fill cport once, even if multiple packets try to do it */ + if (cp->flags & IP_VS_CONN_F_NO_CPORT && (!cp->cport || by_me)) { + /* If we race with resizing make sure cport is set for dir 1 */ + if (!cp->cport) { + cp->cport = cport; + by_me = true; + } + if (!dir) { + atomic_dec(&ipvs->no_cport_conns[af_id]); + cp->flags &= ~IP_VS_CONN_F_NO_CPORT; + } /* We do not recalc hash_key_r under lock, we assume the * parameters in cp do not change, i.e. cport is * the only possible change. @@ -726,21 +738,17 @@ void ip_vs_conn_fill_cport(struct ip_vs_conn *cp, __be16 cport) hlist_bl_del_rcu(&hn->node); hlist_bl_add_head_rcu(&hn->node, head_new); } - if (!dir) { - atomic_dec(&ipvs->no_cport_conns[af_id]); - cp->flags &= ~IP_VS_CONN_F_NO_CPORT; - cp->cport = cport; - } } - spin_unlock(&cp->lock); if (head != head2) hlist_bl_unlock(head2); hlist_bl_unlock(head); write_seqcount_end(&t->seqc[hash_key & t->seqc_mask]); preempt_enable_nested(); - spin_unlock_bh(&t->lock[hash_key & t->lock_mask].l); - if (dir--) + spin_unlock(&t->lock[hash_key & t->lock_mask].l); + + spin_unlock_bh(&cp->lock); + if (dir-- && by_me) goto next_dir; } From fbe1e01e818ee6db86ff947599bf0bea96de7e71 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Thu, 30 Apr 2026 10:44:16 +0300 Subject: [PATCH 285/525] ipvs: do not leak dest after get from dest trash Sashiko warns about leaked dest if ip_vs_start_estimator() fails in ip_vs_add_dest(). Add ip_vs_trash_put_dest() to put back the dest into dest trash. Link: https://sashiko.dev/#/patchset/20260428175725.72050-1-ja%40ssi.bg Fixes: 705dd3444081 ("ipvs: use kthreads for stats estimation") Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_ctl.c | 37 ++++++++++++++++++++++------------ 1 file changed, 24 insertions(+), 13 deletions(-) diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index caec516856e9..d81077c2457a 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -1102,6 +1102,24 @@ ip_vs_trash_get_dest(struct ip_vs_service *svc, int dest_af, return dest; } +/* Put destination in trash */ +static void ip_vs_trash_put_dest(struct netns_ipvs *ipvs, + struct ip_vs_dest *dest, unsigned long istart, + bool cleanup) +{ + spin_lock_bh(&ipvs->dest_trash_lock); + IP_VS_DBG_BUF(3, "Moving dest %s:%u into trash, dest->refcnt=%d\n", + IP_VS_DBG_ADDR(dest->af, &dest->addr), ntohs(dest->port), + refcount_read(&dest->refcnt)); + if (list_empty(&ipvs->dest_trash) && !cleanup) + mod_timer(&ipvs->dest_trash_timer, + jiffies + (IP_VS_DEST_TRASH_PERIOD >> 1)); + /* dest lives in trash with reference */ + list_add(&dest->t_list, &ipvs->dest_trash); + dest->idle_start = istart; + spin_unlock_bh(&ipvs->dest_trash_lock); +} + static void ip_vs_dest_rcu_free(struct rcu_head *head) { struct ip_vs_dest *dest; @@ -1461,9 +1479,12 @@ ip_vs_add_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) ntohs(dest->vport)); ret = ip_vs_start_estimator(svc->ipvs, &dest->stats); + /* On error put back dest into the trash */ if (ret < 0) - return ret; - __ip_vs_update_dest(svc, dest, udest, 1); + ip_vs_trash_put_dest(svc->ipvs, dest, dest->idle_start, + false); + else + __ip_vs_update_dest(svc, dest, udest, 1); } else { /* * Allocate and initialize the dest structure @@ -1533,17 +1554,7 @@ static void __ip_vs_del_dest(struct netns_ipvs *ipvs, struct ip_vs_dest *dest, */ ip_vs_rs_unhash(dest); - spin_lock_bh(&ipvs->dest_trash_lock); - IP_VS_DBG_BUF(3, "Moving dest %s:%u into trash, dest->refcnt=%d\n", - IP_VS_DBG_ADDR(dest->af, &dest->addr), ntohs(dest->port), - refcount_read(&dest->refcnt)); - if (list_empty(&ipvs->dest_trash) && !cleanup) - mod_timer(&ipvs->dest_trash_timer, - jiffies + (IP_VS_DEST_TRASH_PERIOD >> 1)); - /* dest lives in trash with reference */ - list_add(&dest->t_list, &ipvs->dest_trash); - dest->idle_start = 0; - spin_unlock_bh(&ipvs->dest_trash_lock); + ip_vs_trash_put_dest(ipvs, dest, 0, cleanup); /* Queue up delayed work to expire all no destination connections. * No-op when CONFIG_SYSCTL is disabled. From 2fd109238925d53c44ea409df0558844af7877b8 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Thu, 30 Apr 2026 10:44:17 +0300 Subject: [PATCH 286/525] ipvs: fix races around est_mutex and est_cpulist Sashiko reports for races and possible crash around the usage of est_cpulist_valid and sysctl_est_cpulist. The problem is that we do not lock est_mutex in some places which can lead to wrong write ordering and as result problems when calling cpumask_weight() and cpumask_empty(). Fix them by moving the est_max_threads read/write under locked est_mutex. Do the same for one ip_vs_est_reload_start() call to protect the cpumask_empty() usage of sysctl_est_cpulist. To remove the chance of deadlock while stopping the estimation kthreads, keep the data structure for kthread 0 even after last estimator is removed and do not hold mutexes while stopping this task. Now we will use a new flag 'needed' to know when kthread 0 should run. The kthreads above 0 do not use mutexes, so stop them under est_mutex because their kthread data still can be destroyed if they do not serve estimators. Now all kthreads will be started by the est_reload_work to properly serialize the stop/start for kthread 0. Reduce the use of service_mutex in ip_vs_est_calc_phase() because under est_mutex we can safely walk est_kt_arr to stop the kthreads above slot 0. As ip_vs_stop_estimator() for tot_stats should be called under service_mutex, do it early in the netns exit path in ip_vs_flush() to avoid locking the mutex again later. It still should be called in ip_vs_control_net_cleanup_sysctl() when we are called during netns init error. Use -2 for ktid as indicator if estimator was already stopped. Finally, fix use-after-free for kd->est_row in ip_vs_est_calc_phase(). est->ktrow should simply switch to a delay value while estimator is linked to est_temp_list. Link: https://sashiko.dev/#/patchset/20260331165015.2777765-1-longman%40redhat.com Link: https://sashiko.dev/#/patchset/20260420171308.87192-1-ja%40ssi.bg Link: https://sashiko.dev/#/patchset/20260422125123.40658-1-ja%40ssi.bg Link: https://sashiko.dev/#/patchset/20260424175858.54752-1-ja%40ssi.bg Link: https://sashiko.dev/#/patchset/20260425103918.7447-1-ja%40ssi.bg Fixes: f0be83d54217 ("ipvs: add est_cpulist and est_nice sysctl vars") Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- include/net/ip_vs.h | 11 ++++- net/netfilter/ipvs/ip_vs_ctl.c | 51 +++++++++++++++++---- net/netfilter/ipvs/ip_vs_est.c | 83 ++++++++++++++++++++-------------- 3 files changed, 100 insertions(+), 45 deletions(-) diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index 72d325c81313..d28ad8a0541f 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -491,6 +491,7 @@ struct ip_vs_est_kt_data { DECLARE_BITMAP(avail, IPVS_EST_NTICKS); /* tick has space for ests */ unsigned long est_timer; /* estimation timer (jiffies) */ struct ip_vs_stats *calc_stats; /* Used for calculation */ + int needed; /* task is needed */ int tick_len[IPVS_EST_NTICKS]; /* est count */ int id; /* ktid per netns */ int chain_max; /* max ests per tick chain */ @@ -1884,11 +1885,19 @@ int ip_vs_start_estimator(struct netns_ipvs *ipvs, struct ip_vs_stats *stats); void ip_vs_stop_estimator(struct netns_ipvs *ipvs, struct ip_vs_stats *stats); void ip_vs_zero_estimator(struct ip_vs_stats *stats); void ip_vs_read_estimator(struct ip_vs_kstats *dst, struct ip_vs_stats *stats); -void ip_vs_est_reload_start(struct netns_ipvs *ipvs); +void ip_vs_est_reload_start(struct netns_ipvs *ipvs, bool restart); int ip_vs_est_kthread_start(struct netns_ipvs *ipvs, struct ip_vs_est_kt_data *kd); void ip_vs_est_kthread_stop(struct ip_vs_est_kt_data *kd); +static inline void ip_vs_stop_estimator_tot_stats(struct netns_ipvs *ipvs) +{ +#ifdef CONFIG_SYSCTL + ip_vs_stop_estimator(ipvs, &ipvs->tot_stats->s); + ipvs->tot_stats->s.est.ktid = -2; +#endif +} + static inline void ip_vs_est_stopped_recalc(struct netns_ipvs *ipvs) { #ifdef CONFIG_SYSCTL diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index d81077c2457a..5c9f8e0e238f 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -261,12 +261,28 @@ static void est_reload_work_handler(struct work_struct *work) if (!kd) continue; /* New config ? Stop kthread tasks */ - if (genid != genid_done) - ip_vs_est_kthread_stop(kd); + if (genid != genid_done) { + if (!id) { + /* Only we can stop kt 0 but not under mutex */ + mutex_unlock(&ipvs->est_mutex); + ip_vs_est_kthread_stop(kd); + mutex_lock(&ipvs->est_mutex); + if (!READ_ONCE(ipvs->enable)) + goto unlock; + /* kd for kt 0 is never destroyed */ + } else { + ip_vs_est_kthread_stop(kd); + } + } if (!kd->task && !ip_vs_est_stopped(ipvs)) { + bool start; + /* Do not start kthreads above 0 in calc phase */ - if ((!id || !ipvs->est_calc_phase) && - ip_vs_est_kthread_start(ipvs, kd) < 0) + if (id) + start = !ipvs->est_calc_phase; + else + start = kd->needed; + if (start && ip_vs_est_kthread_start(ipvs, kd) < 0) repeat = true; } } @@ -1823,11 +1839,16 @@ ip_vs_add_service(struct netns_ipvs *ipvs, struct ip_vs_service_user_kern *u, *svc_p = svc; if (!READ_ONCE(ipvs->enable)) { + mutex_lock(&ipvs->est_mutex); + /* Now there is a service - full throttle */ WRITE_ONCE(ipvs->enable, 1); + ipvs->est_max_threads = ip_vs_est_max_threads(ipvs); + /* Start estimation for first time */ - ip_vs_est_reload_start(ipvs); + ip_vs_est_reload_start(ipvs, true); + mutex_unlock(&ipvs->est_mutex); } return 0; @@ -2103,6 +2124,11 @@ static int ip_vs_flush(struct netns_ipvs *ipvs, bool cleanup) t = p; } } + /* Stop the tot_stats estimator early under service_mutex + * to avoid locking it again later. + */ + if (cleanup) + ip_vs_stop_estimator_tot_stats(ipvs); return 0; } @@ -2348,7 +2374,7 @@ static int ipvs_proc_est_cpumask_set(const struct ctl_table *table, /* est_max_threads may depend on cpulist size */ ipvs->est_max_threads = ip_vs_est_max_threads(ipvs); ipvs->est_calc_phase = 1; - ip_vs_est_reload_start(ipvs); + ip_vs_est_reload_start(ipvs, true); unlock: mutex_unlock(&ipvs->est_mutex); @@ -2428,7 +2454,7 @@ static int ipvs_proc_est_nice(const struct ctl_table *table, int write, mutex_lock(&ipvs->est_mutex); if (*valp != val) { *valp = val; - ip_vs_est_reload_start(ipvs); + ip_vs_est_reload_start(ipvs, true); } mutex_unlock(&ipvs->est_mutex); } @@ -2455,7 +2481,7 @@ static int ipvs_proc_run_estimation(const struct ctl_table *table, int write, mutex_lock(&ipvs->est_mutex); if (*valp != val) { *valp = val; - ip_vs_est_reload_start(ipvs); + ip_vs_est_reload_start(ipvs, true); } mutex_unlock(&ipvs->est_mutex); } @@ -5005,7 +5031,14 @@ static void __net_exit ip_vs_control_net_cleanup_sysctl(struct netns_ipvs *ipvs) cancel_delayed_work_sync(&ipvs->defense_work); cancel_work_sync(&ipvs->defense_work.work); unregister_net_sysctl_table(ipvs->sysctl_hdr); - ip_vs_stop_estimator(ipvs, &ipvs->tot_stats->s); + if (ipvs->tot_stats->s.est.ktid != -2) { + /* Not stopped yet? This happens only on netns init error and + * we even do not need to lock the service_mutex for this case. + */ + mutex_lock(&ipvs->service_mutex); + ip_vs_stop_estimator(ipvs, &ipvs->tot_stats->s); + mutex_unlock(&ipvs->service_mutex); + } if (ipvs->est_cpulist_valid) free_cpumask_var(ipvs->sysctl_est_cpulist); diff --git a/net/netfilter/ipvs/ip_vs_est.c b/net/netfilter/ipvs/ip_vs_est.c index 433ba3cab58c..ab09f5182951 100644 --- a/net/netfilter/ipvs/ip_vs_est.c +++ b/net/netfilter/ipvs/ip_vs_est.c @@ -68,6 +68,11 @@ and the limit of estimators per kthread - est_add_ktid: ktid where to add new ests, can point to empty slot where we should add kt data + - data protected by service_mutex: est_temp_list, est_add_ktid, + est_kt_count(R/W), est_kt_arr(R/W), est_genid_done, kd->needed(R/W) + - data protected by est_mutex: est_genid, est_max_threads, sysctl_est_cpulist, + est_cpulist_valid, sysctl_est_nice, est_stopped, sysctl_run_estimation, + est_kt_count(R), est_kt_arr(R), kd->needed(R), kd->task (id > 0) */ static struct lock_class_key __ipvs_est_key; @@ -227,14 +232,17 @@ static int ip_vs_estimation_kthread(void *data) } /* Schedule stop/start for kthread tasks */ -void ip_vs_est_reload_start(struct netns_ipvs *ipvs) +void ip_vs_est_reload_start(struct netns_ipvs *ipvs, bool restart) { + lockdep_assert_held(&ipvs->est_mutex); + /* Ignore reloads before first service is added */ if (!READ_ONCE(ipvs->enable)) return; ip_vs_est_stopped_recalc(ipvs); - /* Bump the kthread configuration genid */ - atomic_inc(&ipvs->est_genid); + /* Bump the kthread configuration genid if stopping is requested */ + if (restart) + atomic_inc(&ipvs->est_genid); queue_delayed_work(system_long_wq, &ipvs->est_reload_work, 0); } @@ -304,12 +312,17 @@ static int ip_vs_est_add_kthread(struct netns_ipvs *ipvs) void *arr = NULL; int i; - if ((unsigned long)ipvs->est_kt_count >= ipvs->est_max_threads && - READ_ONCE(ipvs->enable) && ipvs->est_max_threads) - return -EINVAL; - mutex_lock(&ipvs->est_mutex); + /* Allow kt 0 data to be created before the services are added + * and limit the kthreads when services are present. + */ + if ((unsigned long)ipvs->est_kt_count >= ipvs->est_max_threads && + READ_ONCE(ipvs->enable) && ipvs->est_max_threads) { + ret = -EINVAL; + goto out; + } + for (i = 0; i < id; i++) { if (!ipvs->est_kt_arr[i]) break; @@ -333,6 +346,7 @@ static int ip_vs_est_add_kthread(struct netns_ipvs *ipvs) kd->est_timer = jiffies; kd->id = id; ip_vs_est_set_params(ipvs, kd); + kd->needed = 1; /* Pre-allocate stats used in calc phase */ if (!id && !kd->calc_stats) { @@ -341,12 +355,8 @@ static int ip_vs_est_add_kthread(struct netns_ipvs *ipvs) goto out; } - /* Start kthread tasks only when services are present */ - if (READ_ONCE(ipvs->enable) && !ip_vs_est_stopped(ipvs)) { - ret = ip_vs_est_kthread_start(ipvs, kd); - if (ret < 0) - goto out; - } + /* Request kthread to be started */ + ip_vs_est_reload_start(ipvs, false); if (arr) ipvs->est_kt_count++; @@ -482,12 +492,11 @@ static int ip_vs_enqueue_estimator(struct netns_ipvs *ipvs, /* Start estimation for stats */ int ip_vs_start_estimator(struct netns_ipvs *ipvs, struct ip_vs_stats *stats) { + struct ip_vs_est_kt_data *kd = ipvs->est_kt_count > 0 ? + ipvs->est_kt_arr[0] : NULL; struct ip_vs_estimator *est = &stats->est; int ret; - if (!ipvs->est_max_threads && READ_ONCE(ipvs->enable)) - ipvs->est_max_threads = ip_vs_est_max_threads(ipvs); - est->ktid = -1; est->ktrow = IPVS_EST_NTICKS - 1; /* Initial delay */ @@ -496,8 +505,15 @@ int ip_vs_start_estimator(struct netns_ipvs *ipvs, struct ip_vs_stats *stats) * will not allocate much memory, just for kt 0. */ ret = 0; - if (!ipvs->est_kt_count || !ipvs->est_kt_arr[0]) + if (!kd) { ret = ip_vs_est_add_kthread(ipvs); + } else if (!kd->needed) { + mutex_lock(&ipvs->est_mutex); + /* We have job for the kt 0 task */ + kd->needed = 1; + ip_vs_est_reload_start(ipvs, true); + mutex_unlock(&ipvs->est_mutex); + } if (ret >= 0) hlist_add_head(&est->list, &ipvs->est_temp_list); else @@ -578,16 +594,14 @@ void ip_vs_stop_estimator(struct netns_ipvs *ipvs, struct ip_vs_stats *stats) } end_kt0: - /* kt 0 is freed after all other kthreads and chains are empty */ + /* kt 0 task is stopped after all other kt slots and chains are empty */ if (ipvs->est_kt_count == 1 && hlist_empty(&ipvs->est_temp_list)) { kd = ipvs->est_kt_arr[0]; - if (!kd || !kd->est_count) { + if (kd && !kd->est_count) { mutex_lock(&ipvs->est_mutex); - if (kd) { - ip_vs_est_kthread_destroy(kd); - ipvs->est_kt_arr[0] = NULL; - } - ipvs->est_kt_count--; + /* Keep the kt0 data but request kthread_stop */ + kd->needed = 0; + ip_vs_est_reload_start(ipvs, true); mutex_unlock(&ipvs->est_mutex); ipvs->est_add_ktid = 0; } @@ -647,9 +661,9 @@ static int ip_vs_est_calc_limits(struct netns_ipvs *ipvs, int *chain_max) u64 val; INIT_HLIST_HEAD(&chain); - mutex_lock(&ipvs->service_mutex); + mutex_lock(&ipvs->est_mutex); kd = ipvs->est_kt_arr[0]; - mutex_unlock(&ipvs->service_mutex); + mutex_unlock(&ipvs->est_mutex); s = kd ? kd->calc_stats : NULL; if (!s) goto out; @@ -748,16 +762,16 @@ static void ip_vs_est_calc_phase(struct netns_ipvs *ipvs) if (!ip_vs_est_calc_limits(ipvs, &chain_max)) return; - mutex_lock(&ipvs->service_mutex); - /* Stop all other tasks, so that we can immediately move the * estimators to est_temp_list without RCU grace period */ mutex_lock(&ipvs->est_mutex); for (id = 1; id < ipvs->est_kt_count; id++) { /* netns clean up started, abort */ - if (!READ_ONCE(ipvs->enable)) - goto unlock2; + if (kthread_should_stop() || !READ_ONCE(ipvs->enable)) { + mutex_unlock(&ipvs->est_mutex); + return; + } kd = ipvs->est_kt_arr[id]; if (!kd) continue; @@ -765,9 +779,11 @@ static void ip_vs_est_calc_phase(struct netns_ipvs *ipvs) } mutex_unlock(&ipvs->est_mutex); + mutex_lock(&ipvs->service_mutex); + /* Move all estimators to est_temp_list but carefully, * all estimators and kthread data can be released while - * we reschedule. Even for kthread 0. + * we reschedule. */ step = 0; @@ -849,9 +865,7 @@ static void ip_vs_est_calc_phase(struct netns_ipvs *ipvs) ip_vs_stop_estimator(ipvs, stats); /* Tasks are stopped, move without RCU grace period */ est->ktid = -1; - est->ktrow = row - kd->est_row; - if (est->ktrow < 0) - est->ktrow += IPVS_EST_NTICKS; + est->ktrow = delay; hlist_add_head(&est->list, &ipvs->est_temp_list); /* kd freed ? */ if (last) @@ -889,7 +903,6 @@ static void ip_vs_est_calc_phase(struct netns_ipvs *ipvs) if (genid == atomic_read(&ipvs->est_genid)) ipvs->est_calc_phase = 0; -unlock2: mutex_unlock(&ipvs->est_mutex); unlock: From 4ee52b7021a7cb9356f8b9aff5631c68512a9e1b Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Thu, 30 Apr 2026 10:44:18 +0300 Subject: [PATCH 287/525] ipvs: fix shift-out-of-bounds in ip_vs_rht_desired_size Calling roundup_pow_of_two() with 0 has undefined result: UBSAN: shift-out-of-bounds in ./include/linux/log2.h:57:13 shift exponent 64 is too large for 64-bit type 'unsigned long' CPU: 1 UID: 0 PID: 77 Comm: kworker/u8:4 Not tainted syzkaller #0 PREEMPT(full) Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 04/18/2026 Workqueue: events_unbound conn_resize_work_handler Call Trace: dump_stack_lvl+0xe8/0x150 lib/dump_stack.c:120 ubsan_epilogue+0xa/0x30 lib/ubsan.c:233 __ubsan_handle_shift_out_of_bounds+0x385/0x410 lib/ubsan.c:494 __roundup_pow_of_two include/linux/log2.h:57 [inline] ip_vs_rht_desired_size+0x2cf/0x410 net/netfilter/ipvs/ip_vs_core.c:240 ip_vs_conn_desired_size net/netfilter/ipvs/ip_vs_conn.c:765 [inline] conn_resize_work_handler+0x1b6/0x14c0 net/netfilter/ipvs/ip_vs_conn.c:822 process_one_work kernel/workqueue.c:3302 [inline] process_scheduled_works+0xb5d/0x1860 kernel/workqueue.c:3385 worker_thread+0xa53/0xfc0 kernel/workqueue.c:3466 kthread+0x388/0x470 kernel/kthread.c:436 ret_from_fork+0x514/0xb70 arch/x86/kernel/process.c:158 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245 Reported-by: syzbot+217f1db9c791e27fe54a@syzkaller.appspotmail.com Fixes: b655388111cf ("ipvs: add resizable hash tables") Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index f5b7a2047291..d40b404c1bf6 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -237,7 +237,7 @@ int ip_vs_rht_desired_size(struct netns_ipvs *ipvs, struct ip_vs_rht *t, int n, { if (!t) return 1 << min_bits; - n = roundup_pow_of_two(n); + n = n > 0 ? roundup_pow_of_two(n) : 1; if (lfactor < 0) { int factor = min(-lfactor, max_bits); From aa6065206987278291c09d0c6aebed687114c925 Mon Sep 17 00:00:00 2001 From: Waiman Long Date: Thu, 30 Apr 2026 10:44:19 +0300 Subject: [PATCH 288/525] ipvs: Guard access of HK_TYPE_KTHREAD cpumask with RCU The ip_vs_ctl.c file and the associated ip_vs.h file are the only places in the kernel where HK_TYPE_KTHREAD cpumask is being retrieved and used. Now that HK_TYPE_KTHREAD/HK_TYPE_DOMAIN cpumask can be changed at run time. We need to use RCU to guard access to this cpumask to avoid a potential UAF problem as the returned cpumask may be freed before it is being used. We can replace HK_TYPE_KTHREAD by HK_TYPE_DOMAIN as they are aliases of each other, but keeping the HK_TYPE_KTHREAD name can highlight the fact that it is the kthread initiated by ipvs that is being controlled. Fixes: 03ff73510169 ("cpuset: Update HK_TYPE_DOMAIN cpumask from cpuset") Signed-off-by: Waiman Long Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- include/net/ip_vs.h | 20 ++++++++++++++++---- net/netfilter/ipvs/ip_vs_ctl.c | 13 ++++++++----- 2 files changed, 24 insertions(+), 9 deletions(-) diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index d28ad8a0541f..02762ce73a0c 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -1412,7 +1412,7 @@ static inline int sysctl_run_estimation(struct netns_ipvs *ipvs) return ipvs->sysctl_run_estimation; } -static inline const struct cpumask *sysctl_est_cpulist(struct netns_ipvs *ipvs) +static inline const struct cpumask *__sysctl_est_cpulist(struct netns_ipvs *ipvs) { if (ipvs->est_cpulist_valid) return ipvs->sysctl_est_cpulist; @@ -1530,7 +1530,7 @@ static inline int sysctl_run_estimation(struct netns_ipvs *ipvs) return 1; } -static inline const struct cpumask *sysctl_est_cpulist(struct netns_ipvs *ipvs) +static inline const struct cpumask *__sysctl_est_cpulist(struct netns_ipvs *ipvs) { return housekeeping_cpumask(HK_TYPE_KTHREAD); } @@ -1565,6 +1565,18 @@ static inline int sysctl_svc_lfactor(struct netns_ipvs *ipvs) return READ_ONCE(ipvs->sysctl_svc_lfactor); } +static inline bool sysctl_est_cpulist_empty(struct netns_ipvs *ipvs) +{ + guard(rcu)(); + return cpumask_empty(__sysctl_est_cpulist(ipvs)); +} + +static inline unsigned int sysctl_est_cpulist_weight(struct netns_ipvs *ipvs) +{ + guard(rcu)(); + return cpumask_weight(__sysctl_est_cpulist(ipvs)); +} + /* IPVS core functions * (from ip_vs_core.c) */ @@ -1904,7 +1916,7 @@ static inline void ip_vs_est_stopped_recalc(struct netns_ipvs *ipvs) /* Stop tasks while cpulist is empty or if disabled with flag */ ipvs->est_stopped = !sysctl_run_estimation(ipvs) || (ipvs->est_cpulist_valid && - cpumask_empty(sysctl_est_cpulist(ipvs))); + sysctl_est_cpulist_empty(ipvs)); #endif } @@ -1920,7 +1932,7 @@ static inline bool ip_vs_est_stopped(struct netns_ipvs *ipvs) static inline int ip_vs_est_max_threads(struct netns_ipvs *ipvs) { unsigned int limit = IPVS_EST_CPU_KTHREADS * - cpumask_weight(sysctl_est_cpulist(ipvs)); + sysctl_est_cpulist_weight(ipvs); return max(1U, limit); } diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index 5c9f8e0e238f..c7c7f6a7a9f6 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -2394,11 +2394,14 @@ static int ipvs_proc_est_cpumask_get(const struct ctl_table *table, mutex_lock(&ipvs->est_mutex); - if (ipvs->est_cpulist_valid) - mask = *valp; - else - mask = (struct cpumask *)housekeeping_cpumask(HK_TYPE_KTHREAD); - ret = scnprintf(buffer, size, "%*pbl\n", cpumask_pr_args(mask)); + /* HK_TYPE_KTHREAD cpumask needs RCU protection */ + scoped_guard(rcu) { + if (ipvs->est_cpulist_valid) + mask = *valp; + else + mask = (struct cpumask *)housekeeping_cpumask(HK_TYPE_KTHREAD); + ret = scnprintf(buffer, size, "%*pbl\n", cpumask_pr_args(mask)); + } mutex_unlock(&ipvs->est_mutex); From 8f78b749f3da0f43990490b4c1193b5ede3eec0a Mon Sep 17 00:00:00 2001 From: Waiman Long Date: Thu, 30 Apr 2026 10:44:20 +0300 Subject: [PATCH 289/525] sched/isolation: Make HK_TYPE_KTHREAD an alias of HK_TYPE_DOMAIN Since commit 041ee6f3727a ("kthread: Rely on HK_TYPE_DOMAIN for preferred affinity management"), kthreads default to use the HK_TYPE_DOMAIN cpumask. IOW, it is no longer affected by the setting of the nohz_full boot kernel parameter. That means HK_TYPE_KTHREAD should now be an alias of HK_TYPE_DOMAIN instead of HK_TYPE_KERNEL_NOISE to correctly reflect the current kthread behavior. Make the change as HK_TYPE_KTHREAD is still being used in some networking code. Fixes: 041ee6f3727a ("kthread: Rely on HK_TYPE_DOMAIN for preferred affinity management") Signed-off-by: Waiman Long Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- include/linux/sched/isolation.h | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/include/linux/sched/isolation.h b/include/linux/sched/isolation.h index dc3975ff1b2e..cf0fd03dd7a2 100644 --- a/include/linux/sched/isolation.h +++ b/include/linux/sched/isolation.h @@ -20,6 +20,11 @@ enum hk_type { HK_TYPE_KERNEL_NOISE, HK_TYPE_MAX, + /* + * HK_TYPE_KTHREAD is now an alias of HK_TYPE_DOMAIN + */ + HK_TYPE_KTHREAD = HK_TYPE_DOMAIN, + /* * The following housekeeping types are only set by the nohz_full * boot commandline option. So they can share the same value. @@ -29,7 +34,6 @@ enum hk_type { HK_TYPE_RCU = HK_TYPE_KERNEL_NOISE, HK_TYPE_MISC = HK_TYPE_KERNEL_NOISE, HK_TYPE_WQ = HK_TYPE_KERNEL_NOISE, - HK_TYPE_KTHREAD = HK_TYPE_KERNEL_NOISE }; #ifdef CONFIG_CPU_ISOLATION From d82ba05263c69fa2437fe93e4e561cc40f4c03af Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Fri, 1 May 2026 07:39:41 +0000 Subject: [PATCH 290/525] af_unix: Set gc_in_progress to true in unix_gc(). Igor Ushakov reported that unix_gc() could run with gc_in_progress being false if the work is scheduled while running: Thread 1 Thread 2 Thread 3 -------- -------- -------- unix_schedule_gc() unix_schedule_gc() `- if (!gc_in_progress) `- if (!gc_in_progress) |- gc_in_progress = true | `- queue_work() | unix_gc() <----------------/ | | |- gc_in_progress = true ... `- queue_work() | | `- gc_in_progress = false | | unix_gc() <---------------------------------------------' | ... /* gc_in_progress == false */ | `- gc_in_progress = false unix_peek_fpl() relies on gc_in_progress not to confuse GC by MSG_PEEK. Let's set gc_in_progress to true in unix_gc(). Fixes: 8b90a9f819dc ("af_unix: Run GC on only one CPU.") Reported-by: Igor Ushakov Signed-off-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20260501073945.1884564-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- net/unix/garbage.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/net/unix/garbage.c b/net/unix/garbage.c index a7967a345827..0783555e2526 100644 --- a/net/unix/garbage.c +++ b/net/unix/garbage.c @@ -607,6 +607,8 @@ static void unix_gc(struct work_struct *work) struct sk_buff_head hitlist; struct sk_buff *skb; + WRITE_ONCE(gc_in_progress, true); + spin_lock(&unix_gc_lock); if (unix_graph_state == UNIX_GRAPH_NOT_CYCLIC) { @@ -649,10 +651,8 @@ void unix_schedule_gc(struct user_struct *user) READ_ONCE(user->unix_inflight) < UNIX_INFLIGHT_SANE_USER) return; - if (!READ_ONCE(gc_in_progress)) { - WRITE_ONCE(gc_in_progress, true); + if (!READ_ONCE(gc_in_progress)) queue_work(system_dfl_wq, &unix_gc_work); - } if (user && READ_ONCE(unix_graph_cyclic_sccs)) flush_work(&unix_gc_work); From 76b93a8107574006b25495664304ea9237494d70 Mon Sep 17 00:00:00 2001 From: Breno Leitao Date: Fri, 1 May 2026 02:58:41 -0700 Subject: [PATCH 291/525] netpoll: pass buffer size to egress_dev() to avoid MAC truncation egress_dev() formats np->dev_mac via snprintf() but receives buf as a bare char *, so it cannot derive the buffer size from the pointer. The size argument was hardcoded to MAC_ADDR_STR_LEN (3 * ETH_ALEN - 1 = 17), which is silly wrong in two ways: 1) misleading kernel log output on the MAC-selected target path (np->dev_name[0] == '\0'); for example "aa:bb:cc:dd:ee:ff doesn't exist, aborting" was logged as "aa:bb:cc:dd:ee:f doesn't exist, aborting". 2) the second argument of snprintf is the size of the buffer, not the size of what you want to write. Add a bufsz parameter to egress_dev() and pass sizeof(buf) from each caller, matching the standard snprintf() idiom and removing the hardcoded size from the helper. Every caller already declares "char buf[MAC_ADDR_STR_LEN + 1]" so the formatted MAC continues to fit. Tested by booting with netconsole=6665@/aa:bb:cc:dd:ee:ff,6666@10.0.0.1/00:11:22:33:44:55 on a kernel without a matching device. Pre-fix dmesg shows "aa:bb:cc:dd:ee:f doesn't exist, aborting"; post-fix shows the full "aa:bb:cc:dd:ee:ff doesn't exist, aborting". Fixes: f8a10bed32f5 ("netconsole: allow selection of egress interface via MAC address") Cc: stable@vger.kernel.org Signed-off-by: Breno Leitao Link: https://patch.msgid.link/20260501-netpoll_snprintf_fix-v1-1-84b0566e6597@debian.org Signed-off-by: Jakub Kicinski --- net/core/netpoll.c | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-) diff --git a/net/core/netpoll.c b/net/core/netpoll.c index 4381e0fc25bf..84faace50ac2 100644 --- a/net/core/netpoll.c +++ b/net/core/netpoll.c @@ -608,14 +608,16 @@ EXPORT_SYMBOL_GPL(__netpoll_setup); /* * Returns a pointer to a string representation of the identifier used * to select the egress interface for the given netpoll instance. buf - * must be a buffer of length at least MAC_ADDR_STR_LEN + 1. + * is used to format np->dev_mac when np->dev_name is empty; bufsz must + * be at least MAC_ADDR_STR_LEN + 1 to fit the formatted MAC address + * and its NUL terminator. */ -static char *egress_dev(struct netpoll *np, char *buf) +static char *egress_dev(struct netpoll *np, char *buf, size_t bufsz) { if (np->dev_name[0]) return np->dev_name; - snprintf(buf, MAC_ADDR_STR_LEN, "%pM", np->dev_mac); + snprintf(buf, bufsz, "%pM", np->dev_mac); return buf; } @@ -645,7 +647,7 @@ static int netpoll_take_ipv6(struct netpoll *np, struct net_device *ndev) if (!IS_ENABLED(CONFIG_IPV6)) { np_err(np, "IPv6 is not supported %s, aborting\n", - egress_dev(np, buf)); + egress_dev(np, buf, sizeof(buf))); return -EINVAL; } @@ -667,7 +669,7 @@ static int netpoll_take_ipv6(struct netpoll *np, struct net_device *ndev) } if (err) { np_err(np, "no IPv6 address for %s, aborting\n", - egress_dev(np, buf)); + egress_dev(np, buf, sizeof(buf))); return err; } @@ -687,14 +689,14 @@ static int netpoll_take_ipv4(struct netpoll *np, struct net_device *ndev) in_dev = __in_dev_get_rtnl(ndev); if (!in_dev) { np_err(np, "no IP address for %s, aborting\n", - egress_dev(np, buf)); + egress_dev(np, buf, sizeof(buf))); return -EDESTADDRREQ; } ifa = rtnl_dereference(in_dev->ifa_list); if (!ifa) { np_err(np, "no IP address for %s, aborting\n", - egress_dev(np, buf)); + egress_dev(np, buf, sizeof(buf))); return -EDESTADDRREQ; } @@ -736,7 +738,8 @@ int netpoll_setup(struct netpoll *np) ndev = dev_getbyhwaddr(net, ARPHRD_ETHER, np->dev_mac); if (!ndev) { - np_err(np, "%s doesn't exist, aborting\n", egress_dev(np, buf)); + np_err(np, "%s doesn't exist, aborting\n", + egress_dev(np, buf, sizeof(buf))); err = -ENODEV; goto unlock; } @@ -744,14 +747,14 @@ int netpoll_setup(struct netpoll *np) if (netdev_master_upper_dev_get(ndev)) { np_err(np, "%s is a slave device, aborting\n", - egress_dev(np, buf)); + egress_dev(np, buf, sizeof(buf))); err = -EBUSY; goto put; } if (!netif_running(ndev)) { np_info(np, "device %s not up yet, forcing it\n", - egress_dev(np, buf)); + egress_dev(np, buf, sizeof(buf))); err = dev_open(ndev, NULL); if (err) { From 915f9860fe1c9f7eb6c48c299b2db64fd57ef32f Mon Sep 17 00:00:00 2001 From: James Calligeros Date: Sun, 3 May 2026 22:23:23 +1000 Subject: [PATCH 292/525] ASoC: tas2764: Deal with bogus initial temperature register value The TAS2764 datasheet specifies that the chip initialises the temperature register such that the temperature reading is 2.6 *C, ostensibly to prevent tripping the chip's protection circuitry. The chip is not capable of representing 2.6 *C however, and the register is actually initialised to 0. The ADC does not start sampling until the chip is powered up, and the last sampled temperature persists in the register during software shutdown. Therefore, any reading returning 0 is almost certain to be from before the ADC has actually started sampling, meaning that it is invalid. Return -ENODATA early if the temperature has not yet been sampled by the chip, and indicate a fault condition using HWMON_T_FAULT. Fixes: 186dfc85f9a8 ("ASoC: tas2764: expose die temp to hwmon") Signed-off-by: James Calligeros Signed-off-by: Mark Brown --- sound/soc/codecs/tas2764.c | 35 +++++++++++++++++++++++++++-------- 1 file changed, 27 insertions(+), 8 deletions(-) diff --git a/sound/soc/codecs/tas2764.c b/sound/soc/codecs/tas2764.c index 6aab6d2b7419..55211266927d 100644 --- a/sound/soc/codecs/tas2764.c +++ b/sound/soc/codecs/tas2764.c @@ -684,18 +684,33 @@ static int tas2764_read_die_temp(struct tas2764_priv *tas2764, long *result) * As per datasheet, subtract 93 from raw value to get degrees * Celsius. hwmon wants millidegrees. * - * NOTE: The chip will initialise the TAS2764_TEMP register to - * 2.6 *C to avoid triggering temperature protection. Since the - * ADC is powered down during software shutdown, this value will - * persist until the chip is fully powered up (e.g. the PCM it's - * attached to is opened). The ADC will power down again when - * the chip is put back into software shutdown, with the last - * value sampled persisting in the ADC's register. + * NOTE: The TAS2764 datasheet mentions initialising TAS2764_TEMP + * such that the temperature is 2.6 *C, however the register + * is actually initialised to 0. The ADC is also powered down during + * software shutdown. The last sampled temperature will persist + * in the register while the amp is in this power state. */ + if (reg == 0) + return -ENODATA; + *result = (reg - 93) * 1000; return 0; } +static int tas2764_hwmon_is_fault(struct tas2764_priv *tas2764, long *result) +{ + int ret; + long temp; + + ret = tas2764_read_die_temp(tas2764, &temp); + if (ret == -ENODATA) { + *result = true; + return 0; + } + + return ret; +} + static umode_t tas2764_hwmon_is_visible(const void *data, enum hwmon_sensor_types type, u32 attr, int channel) @@ -705,6 +720,7 @@ static umode_t tas2764_hwmon_is_visible(const void *data, switch (attr) { case hwmon_temp_input: + case hwmon_temp_fault: return 0444; default: break; @@ -724,6 +740,9 @@ static int tas2764_hwmon_read(struct device *dev, case hwmon_temp_input: ret = tas2764_read_die_temp(tas2764, val); break; + case hwmon_temp_fault: + ret = tas2764_hwmon_is_fault(tas2764, val); + break; default: ret = -EOPNOTSUPP; break; @@ -733,7 +752,7 @@ static int tas2764_hwmon_read(struct device *dev, } static const struct hwmon_channel_info *const tas2764_hwmon_info[] = { - HWMON_CHANNEL_INFO(temp, HWMON_T_INPUT), + HWMON_CHANNEL_INFO(temp, HWMON_T_INPUT | HWMON_T_FAULT), NULL }; From d0771f4995d3285756bf496cf6e346df99481f83 Mon Sep 17 00:00:00 2001 From: James Calligeros Date: Sun, 3 May 2026 22:23:24 +1000 Subject: [PATCH 293/525] ASoC: tas2770: Deal with bogus initial temperature value TAS2770 initialises the temperature readout registers to 0. This value persists until the chip is fully powered up and the ADC starts sampling. The ADC then persists the last sampled temperature during software shutdown. The ADC should therefore never return 0 in normal operating conditions, so return -ENODATA and mark it as a fault condition using HWMON_T_FAULT. Fixes: ff73e2780169 ("ASoC: tas2770: expose die temp to hwmon") Signed-off-by: James Calligeros Signed-off-by: Mark Brown --- sound/soc/codecs/tas2770.c | 23 ++++++++++++++++++++++- 1 file changed, 22 insertions(+), 1 deletion(-) diff --git a/sound/soc/codecs/tas2770.c b/sound/soc/codecs/tas2770.c index 50501bcbe916..dbda9f327535 100644 --- a/sound/soc/codecs/tas2770.c +++ b/sound/soc/codecs/tas2770.c @@ -633,10 +633,27 @@ static int tas2770_read_die_temp(struct tas2770_priv *tas2770, long *result) * value read back from its registers will be the last value sampled * before entering software shutdown. */ + if (reading == 0) + return -ENODATA; + *result = (reading - (93 * 16)) * 1000 / 16; return 0; } +static int tas2770_hwmon_is_fault(struct tas2770_priv *tas2770, long *result) +{ + int ret; + long temp; + + ret = tas2770_read_die_temp(tas2770, &temp); + if (ret == -ENODATA) { + *result = true; + return 0; + } + + return ret; +} + static umode_t tas2770_hwmon_is_visible(const void *data, enum hwmon_sensor_types type, u32 attr, int channel) @@ -646,6 +663,7 @@ static umode_t tas2770_hwmon_is_visible(const void *data, switch (attr) { case hwmon_temp_input: + case hwmon_temp_fault: return 0444; default: break; @@ -665,6 +683,9 @@ static int tas2770_hwmon_read(struct device *dev, case hwmon_temp_input: ret = tas2770_read_die_temp(tas2770, val); break; + case hwmon_temp_fault: + ret = tas2770_hwmon_is_fault(tas2770, val); + break; default: ret = -EOPNOTSUPP; break; @@ -674,7 +695,7 @@ static int tas2770_hwmon_read(struct device *dev, } static const struct hwmon_channel_info *const tas2770_hwmon_info[] = { - HWMON_CHANNEL_INFO(temp, HWMON_T_INPUT), + HWMON_CHANNEL_INFO(temp, HWMON_T_INPUT | HWMON_T_FAULT), NULL }; From 36bdc0e815b4e8a05b9028d8ef8a25e1ead35cc1 Mon Sep 17 00:00:00 2001 From: Markus Baier Date: Fri, 1 May 2026 18:39:41 +0200 Subject: [PATCH 294/525] net: usb: asix: ax88772: re-add usbnet_link_change() in phylink callbacks Commit e0bffe3e6894 ("net: asix: ax88772: migrate to phylink") replaced the asix_adjust_link() PHY callback with phylink's mac_link_up() and mac_link_down() handlers, but did not carry over the usbnet_link_change() notification that commit 805206e66fab ("net: asix: fix "can't send until first packet is send" issue") had added. As a result, the original symptom returns: when the link comes up, usbnet is never notified, so the RX URB submission stays dormant until some other event (e.g. a transmitted packet triggering the status endpoint interrupt) wakes it up. This is reproducible with the Apple A1277 USB Ethernet Adapter (05ac:1402, AX88772A based) on a Banana Pro using a static IPv4 configuration. After bringing the interface up, no incoming packets are received until the first outgoing frame triggers usbnet's RX path. Restore the link change notification, gated on a carrier transition so the call remains idempotent if the status endpoint also reports the change later. Fixes: e0bffe3e6894 ("net: asix: ax88772: migrate to phylink") Signed-off-by: Markus Baier Tested-by: Oleksij Rempel Link: https://patch.msgid.link/20260501163941.107668-1-Markus.Baier@soslab.tu-darmstadt.de Signed-off-by: Jakub Kicinski --- drivers/net/usb/asix_devices.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/net/usb/asix_devices.c b/drivers/net/usb/asix_devices.c index df0bcfedddbc..293ef80c4e30 100644 --- a/drivers/net/usb/asix_devices.c +++ b/drivers/net/usb/asix_devices.c @@ -756,6 +756,7 @@ static void ax88772_mac_link_down(struct phylink_config *config, struct usbnet *dev = netdev_priv(to_net_dev(config->dev)); asix_write_medium_mode(dev, 0, 0); + usbnet_link_change(dev, false, false); } static void ax88772_mac_link_up(struct phylink_config *config, @@ -786,6 +787,7 @@ static void ax88772_mac_link_up(struct phylink_config *config, m |= AX_MEDIUM_RFC; asix_write_medium_mode(dev, m, 0); + usbnet_link_change(dev, true, false); } static const struct phylink_mac_ops ax88772_phylink_mac_ops = { From 059b7dbd20a6f0c539a45ddff1573cb8946685b5 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 30 Apr 2026 12:26:52 +0000 Subject: [PATCH 295/525] vsock/virtio: fix potential unbounded skb queue MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit virtio_transport_inc_rx_pkt() checks vvs->rx_bytes + len > vvs->buf_alloc. virtio_transport_recv_enqueue() skips coalescing for packets with VIRTIO_VSOCK_SEQ_EOM. If fed with packets with len == 0 and VIRTIO_VSOCK_SEQ_EOM, a very large number of packets can be queued because vvs->rx_bytes stays at 0. Fix this by estimating the skb metadata size: (Number of skbs in the queue) * SKB_TRUESIZE(0) Fixes: 077706165717 ("virtio/vsock: don't use skbuff state to account credit") Signed-off-by: Eric Dumazet Cc: Arseniy Krasnov Cc: Stefan Hajnoczi Cc: Stefano Garzarella Cc: "Michael S. Tsirkin" Cc: Jason Wang Cc: Xuan Zhuo Cc: "Eugenio Pérez" Cc: virtualization@lists.linux.dev Link: https://patch.msgid.link/20260430122653.554058-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/vmw_vsock/virtio_transport_common.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c index 416d533f493d..9b8014516f4f 100644 --- a/net/vmw_vsock/virtio_transport_common.c +++ b/net/vmw_vsock/virtio_transport_common.c @@ -447,7 +447,9 @@ static int virtio_transport_send_pkt_info(struct vsock_sock *vsk, static bool virtio_transport_inc_rx_pkt(struct virtio_vsock_sock *vvs, u32 len) { - if (vvs->buf_used + len > vvs->buf_alloc) + u64 skb_overhead = (skb_queue_len(&vvs->rx_queue) + 1) * SKB_TRUESIZE(0); + + if (skb_overhead + vvs->buf_used + len > vvs->buf_alloc) return false; vvs->rx_bytes += len; From c4a99a921949cddc590b22bb14eeb23dffcc3ba6 Mon Sep 17 00:00:00 2001 From: Shardul Bankar Date: Fri, 1 May 2026 21:35:34 +0200 Subject: [PATCH 296/525] mptcp: use MPJoinSynAckHMacFailure for SynAck HMAC failure In subflow_finish_connect(), HMAC validation of the server's HMAC in SYN/ACK + MP_JOIN increments MPTCP_MIB_JOINACKMAC ("HMAC was wrong on ACK + MP_JOIN") on failure. The function processes the SYN/ACK, not the ACK; the matching MPTCP_MIB_JOINSYNACKMAC counter ("HMAC was wrong on SYN/ACK + MP_JOIN") exists but is not incremented anywhere in the tree. The mirror site on the server, subflow_syn_recv_sock(), already uses JOINACKMAC correctly for ACK HMAC failure. Use JOINSYNACKMAC at the SYN/ACK validation site so each counter reflects the packet whose HMAC actually failed. Suggested-by: Matthieu Baerts (NGI0) Fixes: fc518953bc9c ("mptcp: add and use MIB counter infrastructure") Cc: stable@vger.kernel.org Signed-off-by: Shardul Bankar Reviewed-by: Matthieu Baerts (NGI0) Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260501-net-mptcp-misc-fixes-7-1-rc3-v1-1-b70118df778e@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/subflow.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index e2cb9d23e4a0..bda6862264ca 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -581,7 +581,7 @@ static void subflow_finish_connect(struct sock *sk, const struct sk_buff *skb) subflow->backup); if (!subflow_thmac_valid(subflow)) { - MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_JOINACKMAC); + MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_JOINSYNACKMAC); subflow->reset_reason = MPTCP_RST_EMPTCP; goto do_reset; } From a6da02d4c00fdda2417e42ad2b762a9209e6cc49 Mon Sep 17 00:00:00 2001 From: Shardul Bankar Date: Fri, 1 May 2026 21:35:35 +0200 Subject: [PATCH 297/525] mptcp: use MPTCP_RST_EMPTCP for ACK HMAC validation failure When HMAC validation fails on a received ACK + MP_JOIN in subflow_syn_recv_sock(), the subflow is reset with reason MPTCP_RST_EPROHIBIT ("Administratively prohibited"). This is incorrect: HMAC validation failure is an MPTCP protocol-level error, not an administrative policy denial. The mirror site on the client, in subflow_finish_connect(), already uses MPTCP_RST_EMPTCP ("MPTCP-specific error") for the same kind of HMAC failure on the SYN/ACK + MP_JOIN. Use the same reason on the server side for symmetry and accuracy. Suggested-by: Matthieu Baerts (NGI0) Fixes: 443041deb5ef ("mptcp: fix NULL pointer in can_accept_new_subflow") Cc: stable@vger.kernel.org Signed-off-by: Shardul Bankar Reviewed-by: Matthieu Baerts (NGI0) Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260501-net-mptcp-misc-fixes-7-1-rc3-v1-2-b70118df778e@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/subflow.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index bda6862264ca..d562e149606f 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -908,7 +908,7 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk, if (!subflow_hmac_valid(subflow_req, &mp_opt)) { SUBFLOW_REQ_INC_STATS(req, MPTCP_MIB_JOINACKMAC); - subflow_add_reset_reason(skb, MPTCP_RST_EPROHIBIT); + subflow_add_reset_reason(skb, MPTCP_RST_EMPTCP); goto dispose_child; } From 6254a16d6f0c672e3809ca5d7c9a28a55d71f764 Mon Sep 17 00:00:00 2001 From: Paolo Abeni Date: Fri, 1 May 2026 21:35:36 +0200 Subject: [PATCH 298/525] mptcp: fix rx timestamp corruption on fastopen The skb cb offset containing the timestamp presence flag is cleared before loading such information. Cache such value before MPTCP CB initialization. Fixes: 36b122baf6a8 ("mptcp: add subflow_v(4,6)_send_synack()") Cc: stable@vger.kernel.org Signed-off-by: Paolo Abeni Reviewed-by: Matthieu Baerts (NGI0) Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260501-net-mptcp-misc-fixes-7-1-rc3-v1-3-b70118df778e@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/fastopen.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/mptcp/fastopen.c b/net/mptcp/fastopen.c index 82ec15bcfd7f..082c46c0f50e 100644 --- a/net/mptcp/fastopen.c +++ b/net/mptcp/fastopen.c @@ -12,6 +12,7 @@ void mptcp_fastopen_subflow_synack_set_params(struct mptcp_subflow_context *subf struct sock *sk, *ssk; struct sk_buff *skb; struct tcp_sock *tp; + bool has_rxtstamp; /* on early fallback the subflow context is deleted by * subflow_syn_recv_sock() @@ -40,12 +41,13 @@ void mptcp_fastopen_subflow_synack_set_params(struct mptcp_subflow_context *subf */ tp->copied_seq += skb->len; subflow->ssn_offset += skb->len; + has_rxtstamp = TCP_SKB_CB(skb)->has_rxtstamp; /* Only the sequence delta is relevant */ MPTCP_SKB_CB(skb)->map_seq = -skb->len; MPTCP_SKB_CB(skb)->end_seq = 0; MPTCP_SKB_CB(skb)->offset = 0; - MPTCP_SKB_CB(skb)->has_rxtstamp = TCP_SKB_CB(skb)->has_rxtstamp; + MPTCP_SKB_CB(skb)->has_rxtstamp = has_rxtstamp; MPTCP_SKB_CB(skb)->cant_coalesce = 1; mptcp_data_lock(sk); From 70ece9d7021c54cf40c72b31b066e9088f5f75f5 Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Fri, 1 May 2026 21:35:37 +0200 Subject: [PATCH 299/525] mptcp: sockopt: increase seq in mptcp_setsockopt_all_sf mptcp_setsockopt_all_sf() was missing a call to sockopt_seq_inc(). This is required not to cause missing synchronization for newer subflows created later on. This helper is called each time a socket option is set on subflows, and future ones will need to inherit this option after their creation. Fixes: 51c5fd09e1b4 ("mptcp: add TCP_MAXSEG sockopt support") Cc: stable@vger.kernel.org Suggested-by: Paolo Abeni Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260501-net-mptcp-misc-fixes-7-1-rc3-v1-4-b70118df778e@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/sockopt.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/mptcp/sockopt.c b/net/mptcp/sockopt.c index 0efe40be2fde..1cf608e7357b 100644 --- a/net/mptcp/sockopt.c +++ b/net/mptcp/sockopt.c @@ -812,6 +812,10 @@ static int mptcp_setsockopt_all_sf(struct mptcp_sock *msk, int level, if (ret) break; } + + if (!ret) + sockopt_seq_inc(msk); + return ret; } From ac0841d7d202073415c808bda7848502163b87dd Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Sat, 2 May 2026 12:41:02 +0000 Subject: [PATCH 300/525] net: prevent possible UAF in rtnl_prop_list_size() I was mistaken by synchronize_rcu() [1] call in netdev_name_node_alt_destroy(), giving a false sense of RCU safety at delete times. We have to use list_del_rcu() to not confuse potential readers in rtnl_prop_list_size(). [1] This synchronize_rcu() call was later removed in commit 723de3ebef03 ("net: free altname using an RCU callback"). Fixes: 9f30831390ed ("net: add rcu safety to rtnl_prop_list_size()") Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260502124102.499204-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/core/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/dev.c b/net/core/dev.c index 06c195906231..8bfa8313ef62 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -371,7 +371,7 @@ static void netdev_name_node_alt_free(struct rcu_head *head) static void __netdev_name_node_alt_destroy(struct netdev_name_node *name_node) { netdev_name_node_del(name_node); - list_del(&name_node->list); + list_del_rcu(&name_node->list); call_rcu(&name_node->rcu, netdev_name_node_alt_free); } From 30cb24f97d44f6b81c14b85c5323de62eef1fb7f Mon Sep 17 00:00:00 2001 From: David Carlier Date: Sat, 2 May 2026 15:19:45 +0100 Subject: [PATCH 301/525] psp: strip variable-length PSP header in psp_dev_rcv() psp_dev_rcv() unconditionally removes a fixed PSP_ENCAP_HLEN, even when psph->hdrlen indicates that the PSP header carries optional fields. A frame whose PSP header advertises a non-zero VC or any extension would therefore be silently mis-decapsulated: option bytes would spill into the inner packet head and downstream parsing would fail on a corrupted skb. Compute the full PSP header length from psph->hdrlen, pull the optional bytes into the linear region, and strip the whole header when decapsulating. Optional fields (VC, ...) are still ignored, just discarded with the rest of the header instead of leaking. crypt_offset and the VIRT flag are intentionally not validated here - callers know their device's PSP implementation and can decide. Both in-tree callers gate on hardware-validated PSP, so this is a correctness fix rather than a reachable corruption path under current configurations. Fixes: 0eddb8023cee ("psp: provide decapsulation and receive helper for drivers") Reviewed-by: Willem de Bruijn Reviewed-by: Daniel Zahka Cc: stable@vger.kernel.org Signed-off-by: David Carlier Link: https://patch.msgid.link/20260502141945.14484-1-devnexen@gmail.com Signed-off-by: Jakub Kicinski --- net/psp/psp_main.c | 42 +++++++++++++++++++++++++++++++----------- 1 file changed, 31 insertions(+), 11 deletions(-) diff --git a/net/psp/psp_main.c b/net/psp/psp_main.c index 9508b6c38003..e45549f08eef 100644 --- a/net/psp/psp_main.c +++ b/net/psp/psp_main.c @@ -263,15 +263,16 @@ EXPORT_SYMBOL(psp_dev_encapsulate); /* Receive handler for PSP packets. * - * Presently it accepts only already-authenticated packets and does not - * support optional fields, such as virtualization cookies. The caller should - * ensure that skb->data is pointing to the mac header, and that skb->mac_len - * is set. This function does not currently adjust skb->csum (CHECKSUM_COMPLETE - * is not supported). + * Accepts only already-authenticated packets. The full PSP header is + * stripped according to psph->hdrlen; any optional fields it advertises + * (virtualization cookies, etc.) are ignored and discarded along with the + * rest of the header. The caller should ensure that skb->data is pointing + * to the mac header, and that skb->mac_len is set. This function does not + * currently adjust skb->csum (CHECKSUM_COMPLETE is not supported). */ int psp_dev_rcv(struct sk_buff *skb, u16 dev_id, u8 generation, bool strip_icv) { - int l2_hlen = 0, l3_hlen, encap; + int l2_hlen = 0, l3_hlen, encap, psp_hlen; struct psp_skb_ext *pse; struct psphdr *psph; struct ethhdr *eth; @@ -312,18 +313,36 @@ int psp_dev_rcv(struct sk_buff *skb, u16 dev_id, u8 generation, bool strip_icv) if (unlikely(uh->dest != htons(PSP_DEFAULT_UDP_PORT))) return -EINVAL; - pse = skb_ext_add(skb, SKB_EXT_PSP); - if (!pse) + psph = (struct psphdr *)(skb->data + l2_hlen + l3_hlen + + sizeof(struct udphdr)); + + /* Strip the full PSP header per psph->hdrlen; VC/options are pulled + * into the linear region only so they can be discarded with the + * rest of the header. + */ + psp_hlen = (psph->hdrlen + 1) * 8; + + if (unlikely(psp_hlen < sizeof(struct psphdr))) + return -EINVAL; + + if (psp_hlen > sizeof(struct psphdr) && + !pskb_may_pull(skb, l2_hlen + l3_hlen + + sizeof(struct udphdr) + psp_hlen)) return -EINVAL; psph = (struct psphdr *)(skb->data + l2_hlen + l3_hlen + sizeof(struct udphdr)); + + pse = skb_ext_add(skb, SKB_EXT_PSP); + if (!pse) + return -EINVAL; + pse->spi = psph->spi; pse->dev_id = dev_id; pse->generation = generation; pse->version = FIELD_GET(PSPHDR_VERFL_VERSION, psph->verfl); - encap = PSP_ENCAP_HLEN; + encap = sizeof(struct udphdr) + psp_hlen; encap += strip_icv ? PSP_TRL_SIZE : 0; if (proto == htons(ETH_P_IP)) { @@ -340,8 +359,9 @@ int psp_dev_rcv(struct sk_buff *skb, u16 dev_id, u8 generation, bool strip_icv) ipv6h->payload_len = htons(ntohs(ipv6h->payload_len) - encap); } - memmove(skb->data + PSP_ENCAP_HLEN, skb->data, l2_hlen + l3_hlen); - skb_pull(skb, PSP_ENCAP_HLEN); + memmove(skb->data + sizeof(struct udphdr) + psp_hlen, + skb->data, l2_hlen + l3_hlen); + skb_pull(skb, sizeof(struct udphdr) + psp_hlen); if (strip_icv) pskb_trim(skb, skb->len - PSP_TRL_SIZE); From a6039776c7994dd0b9a4acce23a3f897d1688cbf Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Sat, 2 May 2026 18:07:47 +0000 Subject: [PATCH 302/525] ipmr: Add __rcu to netns_ipv4.mrt. kernel test robot reported this Sparse warning: $ make C=1 net/ipv4/ipmr.o net/ipv4/ipmr.c:312:24: error: incompatible types in comparison expression (different address spaces): net/ipv4/ipmr.c:312:24: struct mr_table [noderef] __rcu * net/ipv4/ipmr.c:312:24: struct mr_table * Let's add __rcu annotation to netns_ipv4.mrt. Fixes: b3b6babf4751 ("ipmr: Free mr_table after RCU grace period.") Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202605030032.glNApko7-lkp@intel.com/ Signed-off-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20260502180755.359554-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/netns/ipv4.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h index 80ccd4dda8e0..6e27c56514df 100644 --- a/include/net/netns/ipv4.h +++ b/include/net/netns/ipv4.h @@ -275,7 +275,7 @@ struct netns_ipv4 { #ifdef CONFIG_IP_MROUTE #ifndef CONFIG_IP_MROUTE_MULTIPLE_TABLES - struct mr_table *mrt; + struct mr_table __rcu *mrt; #else struct list_head mr_tables; struct fib_rules_ops *mr_rules_ops; From 07d99587396024932e02474c3a5bede71d108454 Mon Sep 17 00:00:00 2001 From: Daniel Golle Date: Sat, 2 May 2026 11:55:02 +0100 Subject: [PATCH 303/525] net: dsa: mt7530: fix .get_stats64 sleeping in atomic context The .get_stats64 callback runs in atomic context, but on MDIO-connected switches every register read acquires the MDIO bus mutex, which can sleep: [ 12.645973] BUG: sleeping function called from invalid context at kernel/locking/mutex.c:609 [ 12.654442] in_atomic(): 0, irqs_disabled(): 0, non_block: 0, pid: 759, name: grep [ 12.663377] preempt_count: 0, expected: 0 [ 12.667410] RCU nest depth: 1, expected: 0 [ 12.671511] INFO: lockdep is turned off. [ 12.675441] CPU: 0 UID: 0 PID: 759 Comm: grep Tainted: G S W 7.0.0+ #0 PREEMPT [ 12.675453] Tainted: [S]=CPU_OUT_OF_SPEC, [W]=WARN [ 12.675456] Hardware name: Bananapi BPI-R64 (DT) [ 12.675459] Call trace: [ 12.675462] show_stack+0x14/0x1c (C) [ 12.675477] dump_stack_lvl+0x68/0x8c [ 12.675487] dump_stack+0x14/0x1c [ 12.675495] __might_resched+0x14c/0x220 [ 12.675504] __might_sleep+0x44/0x80 [ 12.675511] __mutex_lock+0x50/0xb10 [ 12.675523] mutex_lock_nested+0x20/0x30 [ 12.675532] mt7530_get_stats64+0x40/0x2ac [ 12.675542] dsa_user_get_stats64+0x2c/0x40 [ 12.675553] dev_get_stats+0x44/0x1e0 [ 12.675564] dev_seq_printf_stats+0x24/0xe0 [ 12.675575] dev_seq_show+0x14/0x3c [ 12.675583] seq_read_iter+0x37c/0x480 [ 12.675595] seq_read+0xd0/0xec [ 12.675605] proc_reg_read+0x94/0xe4 [ 12.675615] vfs_read+0x98/0x29c [ 12.675625] ksys_read+0x54/0xdc [ 12.675633] __arm64_sys_read+0x18/0x20 [ 12.675642] invoke_syscall.constprop.0+0x54/0xec [ 12.675653] do_el0_svc+0x3c/0xb4 [ 12.675662] el0_svc+0x38/0x200 [ 12.675670] el0t_64_sync_handler+0x98/0xdc [ 12.675679] el0t_64_sync+0x158/0x15c For MDIO-connected switches, poll MIB counters asynchronously using a delayed workqueue every second and let .get_stats64 return the cached values under a spinlock. A mod_delayed_work() call on each read triggers an immediate refresh so counters stay responsive when queried more frequently. MMIO-connected switches (MT7988, EN7581, AN7583) are not affected because their regmap does not sleep, so they continue to read MIB counters directly in .get_stats64. Fixes: 88c810f35ed5 ("net: dsa: mt7530: implement .get_stats64") Signed-off-by: Daniel Golle Acked-by: Chester A. Unal Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/6940b913da2c29156f0feff74b678d3c526ee84c.1777719253.git.daniel@makrotopia.org Signed-off-by: Jakub Kicinski --- drivers/net/dsa/mt7530.c | 75 ++++++++++++++++++++++++++++++++++++++-- drivers/net/dsa/mt7530.h | 8 +++++ 2 files changed, 80 insertions(+), 3 deletions(-) diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c index b9423389c2ef..44d670904ad8 100644 --- a/drivers/net/dsa/mt7530.c +++ b/drivers/net/dsa/mt7530.c @@ -25,6 +25,9 @@ #include "mt7530.h" +#define MT7530_STATS_POLL_INTERVAL (1 * HZ) +#define MT7530_STATS_RATE_LIMIT (HZ / 10) + static struct mt753x_pcs *pcs_to_mt753x_pcs(struct phylink_pcs *pcs) { return container_of(pcs, struct mt753x_pcs, pcs); @@ -906,10 +909,9 @@ static void mt7530_get_rmon_stats(struct dsa_switch *ds, int port, *ranges = mt7530_rmon_ranges; } -static void mt7530_get_stats64(struct dsa_switch *ds, int port, - struct rtnl_link_stats64 *storage) +static void mt7530_read_port_stats64(struct mt7530_priv *priv, int port, + struct rtnl_link_stats64 *storage) { - struct mt7530_priv *priv = ds->priv; uint64_t data; /* MIB counter doesn't provide a FramesTransmittedOK but instead @@ -951,6 +953,54 @@ static void mt7530_get_stats64(struct dsa_switch *ds, int port, &storage->rx_crc_errors); } +static void mt7530_stats_refresh(struct mt7530_priv *priv) +{ + struct rtnl_link_stats64 stats = {}; + struct dsa_port *dp; + int port; + + dsa_switch_for_each_user_port(dp, priv->ds) { + port = dp->index; + + mt7530_read_port_stats64(priv, port, &stats); + + spin_lock_bh(&priv->stats_lock); + priv->ports[port].stats = stats; + priv->stats_last = jiffies; + spin_unlock_bh(&priv->stats_lock); + } +} + +static void mt7530_stats_poll(struct work_struct *work) +{ + struct mt7530_priv *priv = container_of(work, struct mt7530_priv, + stats_work.work); + + mt7530_stats_refresh(priv); + schedule_delayed_work(&priv->stats_work, + MT7530_STATS_POLL_INTERVAL); +} + +static void mt7530_get_stats64(struct dsa_switch *ds, int port, + struct rtnl_link_stats64 *storage) +{ + struct mt7530_priv *priv = ds->priv; + bool refresh; + + if (priv->bus) { + spin_lock_bh(&priv->stats_lock); + *storage = priv->ports[port].stats; + refresh = time_after(jiffies, priv->stats_last + + MT7530_STATS_RATE_LIMIT); + spin_unlock_bh(&priv->stats_lock); + if (refresh) + mod_delayed_work(system_percpu_wq, + &priv->stats_work, 0); + } else { + mt7530_read_port_stats64(priv, port, storage); + } +} + static void mt7530_get_eth_ctrl_stats(struct dsa_switch *ds, int port, struct ethtool_eth_ctrl_stats *ctrl_stats) { @@ -3137,9 +3187,24 @@ mt753x_setup(struct dsa_switch *ds) if (ret && priv->irq_domain) mt7530_free_mdio_irq(priv); + if (!ret && priv->bus) { + mt7530_stats_refresh(priv); + schedule_delayed_work(&priv->stats_work, + MT7530_STATS_POLL_INTERVAL); + } + return ret; } +static void +mt753x_teardown(struct dsa_switch *ds) +{ + struct mt7530_priv *priv = ds->priv; + + if (priv->bus) + cancel_delayed_work_sync(&priv->stats_work); +} + static int mt753x_set_mac_eee(struct dsa_switch *ds, int port, struct ethtool_keee *e) { @@ -3257,6 +3322,7 @@ static int mt7988_setup(struct dsa_switch *ds) static const struct dsa_switch_ops mt7530_switch_ops = { .get_tag_protocol = mtk_get_tag_protocol, .setup = mt753x_setup, + .teardown = mt753x_teardown, .preferred_default_local_cpu_port = mt753x_preferred_default_local_cpu_port, .get_strings = mt7530_get_strings, .get_ethtool_stats = mt7530_get_ethtool_stats, @@ -3395,6 +3461,9 @@ mt7530_probe_common(struct mt7530_priv *priv) priv->ds->ops = &mt7530_switch_ops; priv->ds->phylink_mac_ops = &mt753x_phylink_mac_ops; mutex_init(&priv->reg_mutex); + spin_lock_init(&priv->stats_lock); + INIT_DELAYED_WORK(&priv->stats_work, mt7530_stats_poll); + dev_set_drvdata(dev, priv); return 0; diff --git a/drivers/net/dsa/mt7530.h b/drivers/net/dsa/mt7530.h index 3e0090bed298..dd33b0df3419 100644 --- a/drivers/net/dsa/mt7530.h +++ b/drivers/net/dsa/mt7530.h @@ -796,6 +796,7 @@ struct mt7530_fdb { * @pvid: The VLAN specified is to be considered a PVID at ingress. Any * untagged frames will be assigned to the related VLAN. * @sgmii_pcs: Pointer to PCS instance for SerDes ports + * @stats: Cached port statistics for MDIO-connected switches */ struct mt7530_port { bool enable; @@ -803,6 +804,7 @@ struct mt7530_port { u32 pm; u16 pvid; struct phylink_pcs *sgmii_pcs; + struct rtnl_link_stats64 stats; }; /* Port 5 mode definitions of the MT7530 switch */ @@ -875,6 +877,9 @@ struct mt753x_info { * @create_sgmii: Pointer to function creating SGMII PCS instance(s) * @active_cpu_ports: Holding the active CPU ports * @mdiodev: The pointer to the MDIO device structure + * @stats_lock: Protects cached per-port stats from concurrent access + * @stats_work: Delayed work for polling MIB counters on MDIO switches + * @stats_last: Jiffies timestamp of last MIB counter poll */ struct mt7530_priv { struct device *dev; @@ -900,6 +905,9 @@ struct mt7530_priv { int (*create_sgmii)(struct mt7530_priv *priv); u8 active_cpu_ports; struct mdio_device *mdiodev; + spinlock_t stats_lock; /* protects cached stats counters */ + struct delayed_work stats_work; + unsigned long stats_last; }; struct mt7530_hw_vlan_entry { From f4c50a4034e62ab75f1d5cdd191dd5f9c77fdff4 Mon Sep 17 00:00:00 2001 From: Kuan-Ting Chen Date: Mon, 4 May 2026 23:27:12 +0800 Subject: [PATCH 304/525] xfrm: esp: avoid in-place decrypt on shared skb frags MSG_SPLICE_PAGES can attach pages from a pipe directly to an skb. TCP marks such skbs with SKBFL_SHARED_FRAG after skb_splice_from_iter(), so later paths that may modify packet data can first make a private copy. The IPv4/IPv6 datagram append paths did not set this flag when splicing pages into UDP skbs. That leaves an ESP-in-UDP packet made from shared pipe pages looking like an ordinary uncloned nonlinear skb. ESP input then takes the no-COW fast path for uncloned skbs without a frag_list and decrypts in place over data that is not owned privately by the skb. Mark IPv4/IPv6 datagram splice frags with SKBFL_SHARED_FRAG, matching TCP. Also make ESP input fall back to skb_cow_data() when the flag is present, so ESP does not decrypt externally backed frags in place. Private nonlinear skb frags still use the existing fast path. This intentionally does not change ESP output. In esp_output_head(), the path that appends the ESP trailer to existing skb tailroom without calling skb_cow_data() is not reachable for nonlinear skbs: skb_tailroom() returns zero when skb->data_len is nonzero, while ESP tailen is positive. Thus ESP output will either use the separate destination-frag path or fall back to skb_cow_data(). Fixes: cac2661c53f3 ("esp4: Avoid skb_cow_data whenever possible") Fixes: 03e2a30f6a27 ("esp6: Avoid skb_cow_data whenever possible") Fixes: 7da0dde68486 ("ip, udp: Support MSG_SPLICE_PAGES") Fixes: 6d8192bd69bb ("ip6, udp6: Support MSG_SPLICE_PAGES") Reported-by: Hyunwoo Kim Reported-by: Kuan-Ting Chen Tested-by: Hyunwoo Kim Cc: stable@vger.kernel.org Signed-off-by: Kuan-Ting Chen Signed-off-by: Steffen Klassert --- net/ipv4/esp4.c | 3 ++- net/ipv4/ip_output.c | 2 ++ net/ipv6/esp6.c | 3 ++- net/ipv6/ip6_output.c | 2 ++ 4 files changed, 8 insertions(+), 2 deletions(-) diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c index 6dfc0bcdef65..6a5febbdbee4 100644 --- a/net/ipv4/esp4.c +++ b/net/ipv4/esp4.c @@ -873,7 +873,8 @@ static int esp_input(struct xfrm_state *x, struct sk_buff *skb) nfrags = 1; goto skip_cow; - } else if (!skb_has_frag_list(skb)) { + } else if (!skb_has_frag_list(skb) && + !skb_has_shared_frag(skb)) { nfrags = skb_shinfo(skb)->nr_frags; nfrags++; diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index e4790cc7b5c2..5bcd73cbdb41 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -1233,6 +1233,8 @@ static int __ip_append_data(struct sock *sk, if (err < 0) goto error; copy = err; + if (!(flags & MSG_NO_SHARED_FRAGS)) + skb_shinfo(skb)->flags |= SKBFL_SHARED_FRAG; wmem_alloc_delta += copy; } else if (!zc) { int i = skb_shinfo(skb)->nr_frags; diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c index 9f75313734f8..9c06c5a1419d 100644 --- a/net/ipv6/esp6.c +++ b/net/ipv6/esp6.c @@ -915,7 +915,8 @@ static int esp6_input(struct xfrm_state *x, struct sk_buff *skb) nfrags = 1; goto skip_cow; - } else if (!skb_has_frag_list(skb)) { + } else if (!skb_has_frag_list(skb) && + !skb_has_shared_frag(skb)) { nfrags = skb_shinfo(skb)->nr_frags; nfrags++; diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index 7e92909ab5be..1f2a33fbed6e 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -1794,6 +1794,8 @@ static int __ip6_append_data(struct sock *sk, if (err < 0) goto error; copy = err; + if (!(flags & MSG_NO_SHARED_FRAGS)) + skb_shinfo(skb)->flags |= SKBFL_SHARED_FRAG; wmem_alloc_delta += copy; } else if (!zc) { int i = skb_shinfo(skb)->nr_frags; From aab3d205a086233c612fee86009265451793e0c2 Mon Sep 17 00:00:00 2001 From: Juha-Pekka Heikkila Date: Mon, 27 Apr 2026 19:57:15 +0300 Subject: [PATCH 305/525] drm/i915/display: enable ccs modifiers on dg2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Since Xe driver aux ccs enablement dg2 ccs modifiers have been disabled on i915 driver. Here allow dg2 to use ccs again for framebuffers. Fixes: 6a99e91a6ca8 ("drm/i915/display: Detect AuxCCS support via display parent interface") Signed-off-by: Juha-Pekka Heikkila Reviewed-by: Ville Syrjälä Signed-off-by: Mika Kahola Link: https://patch.msgid.link/20260427165715.864721-1-juhapekka.heikkila@gmail.com (cherry picked from commit aee13ba1448213975f36942ba5d1ce693eb5c002) Signed-off-by: Tvrtko Ursulin --- drivers/gpu/drm/i915/i915_driver.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_driver.c b/drivers/gpu/drm/i915/i915_driver.c index 385a634c3ed0..d9be7a5a239c 100644 --- a/drivers/gpu/drm/i915/i915_driver.c +++ b/drivers/gpu/drm/i915/i915_driver.c @@ -750,9 +750,8 @@ static bool has_auxccs(struct drm_device *drm) { struct drm_i915_private *i915 = to_i915(drm); - return IS_GRAPHICS_VER(i915, 9, 12) || - IS_ALDERLAKE_P(i915) || - IS_METEORLAKE(i915); + return IS_GRAPHICS_VER(i915, 9, 12) && + !HAS_FLAT_CCS(i915); } static bool has_fenced_regions(struct drm_device *drm) From 2c340aab5485ebe9e33c01437dd4815ef33c8df5 Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Fri, 1 May 2026 09:16:38 +0200 Subject: [PATCH 306/525] x86/efi: Restore IRQ state in EFI page fault handler The kernel's softirq API does not permit re-enabling softirqs while IRQs are disabled. The reason for this is that local_bh_enable() will not only re-enable delivery of softirqs over the back of IRQs, it will also handle any pending softirqs immediately, regardless of whether IRQs are enabled at that point. For this reason, commit d02198550423 ("x86/fpu: Improve crypto performance by making kernel-mode FPU reliably usable in softirqs") disables softirqs only when IRQs are enabled, as it is not permitted otherwise, but also unnecessary, given that asynchronous softirq delivery never happens to begin with while IRQs are disabled. However, this does mean that entering a kernel mode FPU section with IRQs enabled and leaving it with IRQs disabled leads to problems, as identified by Sashiko [0]: the EFI page fault handler is called from page_fault_oops() with IRQs disabled, and thus ends the kernel mode FPU section with IRQs disabled as well, regardless of whether IRQs were enabled when it was started. This may result in schedule() being called with a non-zero preempt_count, causing a BUG(). So take care to re-enable IRQs when handling any EFI page faults if they were taken with IRQs enabled. [0] https://sashiko.dev/#/patchset/20260430074107.27051-1-ivan.hu%40canonical.com Cc: Eric Biggers Cc: Ivan Hu Cc: x86@kernel.org Cc: Fixes: d02198550423 ("x86/fpu: Improve crypto performance by making kernel-mode FPU reliably usable in softirqs") Reviewed-by: Eric Biggers Signed-off-by: Ard Biesheuvel --- arch/x86/include/asm/efi.h | 3 ++- arch/x86/mm/fault.c | 2 +- arch/x86/platform/efi/quirks.c | 11 ++++++++++- 3 files changed, 13 insertions(+), 3 deletions(-) diff --git a/arch/x86/include/asm/efi.h b/arch/x86/include/asm/efi.h index dc8fe1361c18..be58b7f5c806 100644 --- a/arch/x86/include/asm/efi.h +++ b/arch/x86/include/asm/efi.h @@ -137,7 +137,8 @@ extern void __init efi_dump_pagetable(void); extern void __init efi_apply_memmap_quirks(void); extern int __init efi_reuse_config(u64 tables, int nr_tables); extern void efi_delete_dummy_variable(void); -extern void efi_crash_gracefully_on_page_fault(unsigned long phys_addr); +extern void efi_crash_gracefully_on_page_fault(unsigned long phys_addr, + const struct pt_regs *regs); extern void efi_unmap_boot_services(void); void arch_efi_call_virt_setup(void); diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index f0e77e084482..63de8e8684f2 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -686,7 +686,7 @@ page_fault_oops(struct pt_regs *regs, unsigned long error_code, * avoid hanging the system. */ if (IS_ENABLED(CONFIG_EFI)) - efi_crash_gracefully_on_page_fault(address); + efi_crash_gracefully_on_page_fault(address, regs); /* Only not-present faults should be handled by KFENCE. */ if (!(error_code & X86_PF_PROT) && diff --git a/arch/x86/platform/efi/quirks.c b/arch/x86/platform/efi/quirks.c index c8f5e094ed9d..90a065fcb1fa 100644 --- a/arch/x86/platform/efi/quirks.c +++ b/arch/x86/platform/efi/quirks.c @@ -761,7 +761,8 @@ int efi_capsule_setup_info(struct capsule_info *cap_info, void *kbuff, * @return: Returns, if the page fault is not handled. This function * will never return if the page fault is handled successfully. */ -void efi_crash_gracefully_on_page_fault(unsigned long phys_addr) +void efi_crash_gracefully_on_page_fault(unsigned long phys_addr, + const struct pt_regs *regs) { if (!IS_ENABLED(CONFIG_X86_64)) return; @@ -810,6 +811,14 @@ void efi_crash_gracefully_on_page_fault(unsigned long phys_addr) return; } + /* + * The API does not permit entering a kernel mode FPU section with + * interrupts enabled and leaving it with interrupts disabled. So + * re-enable interrupts now if they were enabled when the page fault + * occurred. + */ + local_irq_restore(regs->flags); + /* * Before calling EFI Runtime Service, the kernel has switched the * calling process to efi_mm. Hence, switch back to task_mm. From 212ec34e4e726e8cd4af7bea4740db24de8a9dab Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 4 May 2026 08:34:32 -0600 Subject: [PATCH 307/525] block: only read from sqe on initial invocation of blkdev_uring_cmd() This passthrough helper currently only supports discards. Part of that command is the start and length, which is read from the SQE. It does so on every invocation, where it really should just make it stable on the first invocation. This avoids needing to copy the SQE upfront, as we only really need those two 8b values stored in our per-req payload. Cc: stable@vger.kernel.org # 6.17+ Signed-off-by: Jens Axboe --- block/ioctl.c | 24 +++++++++++++++--------- 1 file changed, 15 insertions(+), 9 deletions(-) diff --git a/block/ioctl.c b/block/ioctl.c index fc3be0549aa7..ab2c9ed79946 100644 --- a/block/ioctl.c +++ b/block/ioctl.c @@ -857,6 +857,8 @@ long compat_blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg) #endif struct blk_iou_cmd { + u64 start; + u64 len; int res; bool nowait; }; @@ -946,23 +948,27 @@ int blkdev_uring_cmd(struct io_uring_cmd *cmd, unsigned int issue_flags) { struct block_device *bdev = I_BDEV(cmd->file->f_mapping->host); struct blk_iou_cmd *bic = io_uring_cmd_to_pdu(cmd, struct blk_iou_cmd); - const struct io_uring_sqe *sqe = cmd->sqe; u32 cmd_op = cmd->cmd_op; - uint64_t start, len; - if (unlikely(sqe->ioprio || sqe->__pad1 || sqe->len || - sqe->rw_flags || sqe->file_index)) - return -EINVAL; + /* Read what we need from the SQE on the first issue */ + if (!(issue_flags & IORING_URING_CMD_REISSUE)) { + const struct io_uring_sqe *sqe = cmd->sqe; + + if (unlikely(sqe->ioprio || sqe->__pad1 || sqe->len || + sqe->rw_flags || sqe->file_index)) + return -EINVAL; + + bic->start = READ_ONCE(sqe->addr); + bic->len = READ_ONCE(sqe->addr3); + } bic->res = 0; bic->nowait = issue_flags & IO_URING_F_NONBLOCK; - start = READ_ONCE(sqe->addr); - len = READ_ONCE(sqe->addr3); - switch (cmd_op) { case BLOCK_URING_CMD_DISCARD: - return blkdev_cmd_discard(cmd, bdev, start, len, bic->nowait); + return blkdev_cmd_discard(cmd, bdev, bic->start, bic->len, + bic->nowait); } return -EINVAL; } From 5f8719945244dd65b5fa06195f4600db62581610 Mon Sep 17 00:00:00 2001 From: Juergen Gross Date: Tue, 5 May 2026 10:06:53 +0200 Subject: [PATCH 308/525] x86/xen: Fix a potential problem in xen_e820_resolve_conflicts() When fixing a conflict in xen_e820_resolve_conflicts(), the loop over the E820 map entries needs to be restarted, as the E820 map will have been modified by the fix. Otherwise entries might be skipped by accident. Fixes: be35d91c8880 ("xen: tolerate ACPI NVS memory overlapping with Xen allocated memory") Signed-off-by: Juergen Gross Signed-off-by: Ingo Molnar Cc: xen-devel@lists.xenproject.org Link: https://patch.msgid.link/20260505080653.197775-1-jgross@suse.com --- arch/x86/xen/setup.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c index ac8021c3a997..bb95a05259b8 100644 --- a/arch/x86/xen/setup.c +++ b/arch/x86/xen/setup.c @@ -695,17 +695,22 @@ static void __init xen_e820_resolve_conflicts(phys_addr_t start, return; end = start + size; - entry = xen_e820_table.entries; + mapcnt = 0; - for (mapcnt = 0; mapcnt < xen_e820_table.nr_entries; mapcnt++) { + while (mapcnt < xen_e820_table.nr_entries) { + entry = xen_e820_table.entries + mapcnt; if (entry->addr >= end) return; if (entry->addr + entry->size > start && - entry->type == E820_TYPE_NVS) + entry->type == E820_TYPE_NVS) { xen_e820_swap_entry_with_ram(entry); + /* E820 map has been changed, restart loop! */ + mapcnt = 0; + continue; + } - entry++; + mapcnt++; } } From 3780c41460a9ad6d5d4c09a416765c6cc285033b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ma=C3=ADra=20Canal?= Date: Thu, 2 Apr 2026 16:32:35 -0300 Subject: [PATCH 309/525] drm/etnaviv: Fix armed job not being pushed to the DRM scheduler MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When xa_alloc_cyclic() failed in etnaviv_sched_push_job(), the error path skipped drm_sched_entity_push_job(). This is a violation of the DRM scheduler contract, as once a job has been armed with drm_sched_job_arm(), it must be pushed with drm_sched_entity_push_job(). From the DRM scheduler documentation, """ drm_sched_job_arm() is a point of no return since it initializes the fences and their sequence number etc. Once that function has been called, you *must* submit it with drm_sched_entity_push_job() and cannot simply abort it by calling drm_sched_job_cleanup(). """ Fix this by splitting the fence ID allocation into two phases: first, alloc an xarray slot before arming the job (which can fail), then fill in the actual fence with xa_store() after arming. This way, allocation failures are handled before the job is armed, and once armed, the job is always pushed to the scheduler. This also fixes a double call to drm_sched_job_cleanup(), as both etnaviv_sched_push_job() and its caller would call it on failure. Fixes: 764be12345c3 ("drm/etnaviv: convert user fence tracking to XArray") Signed-off-by: Maíra Canal Link: https://patch.msgid.link/20260402193424.2023318-1-mcanal@igalia.com Signed-off-by: Christian Gmeiner --- drivers/gpu/drm/etnaviv/etnaviv_sched.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/etnaviv/etnaviv_sched.c b/drivers/gpu/drm/etnaviv/etnaviv_sched.c index df4232d7e135..3cc50d697c89 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_sched.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_sched.c @@ -116,16 +116,18 @@ int etnaviv_sched_push_job(struct etnaviv_gem_submit *submit) */ mutex_lock(&gpu->sched_lock); + ret = xa_alloc_cyclic(&gpu->user_fences, &submit->out_fence_id, + NULL, xa_limit_32b, &gpu->next_user_fence, + GFP_KERNEL); + if (ret < 0) + goto out_unlock; + drm_sched_job_arm(&submit->sched_job); submit->out_fence = dma_fence_get(&submit->sched_job.s_fence->finished); - ret = xa_alloc_cyclic(&gpu->user_fences, &submit->out_fence_id, - submit->out_fence, xa_limit_32b, - &gpu->next_user_fence, GFP_KERNEL); - if (ret < 0) { - drm_sched_job_cleanup(&submit->sched_job); - goto out_unlock; - } + + xa_store(&gpu->user_fences, submit->out_fence_id, + submit->out_fence, GFP_KERNEL); /* the scheduler holds on to the job now */ kref_get(&submit->refcount); From 4a142520d166f91627f27a7017525a228137c808 Mon Sep 17 00:00:00 2001 From: Jakov Novak Date: Mon, 4 May 2026 18:23:57 +0200 Subject: [PATCH 310/525] wifi: libertas: notify firmware load wait on disconnect Currently, when the firmware is not fully loaded and if_usb_disconnect is called, if_usb_prog_firmware gets stuck waiting for cardp->surprise_removed or cardp->fwdnldover while lbs_remove_card also waits for the firmware loading to be completed, which never happens. This caused the reported syzbot bug. To address this, the wake_up function call can be added in the if_usb_disconnect function which notifies the if_usb_prog_firmware thread and resolves the firmware loading. Fixes: 954ee164f4f4 ("[PATCH] libertas: reorganize and simplify init sequence") Reported-and-tested-by: syzbot+c99d17aa44dbdba16ad2@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=c99d17aa44dbdba16ad2 Signed-off-by: Jakov Novak Link: https://patch.msgid.link/20260504162356.17250-2-jakovnovak30@gmail.com [fix subject] Signed-off-by: Johannes Berg --- drivers/net/wireless/marvell/libertas/if_usb.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/wireless/marvell/libertas/if_usb.c b/drivers/net/wireless/marvell/libertas/if_usb.c index a00d53350fa9..5cc0c5cac257 100644 --- a/drivers/net/wireless/marvell/libertas/if_usb.c +++ b/drivers/net/wireless/marvell/libertas/if_usb.c @@ -310,6 +310,7 @@ static void if_usb_disconnect(struct usb_interface *intf) struct lbs_private *priv = cardp->priv; cardp->surprise_removed = 1; + wake_up(&cardp->fw_wq); if (priv) { lbs_stop_card(priv); From e9e334f8063a991b4f648b8dbb8dac44cf810540 Mon Sep 17 00:00:00 2001 From: Dipayaan Roy Date: Wed, 29 Apr 2026 20:57:52 -0700 Subject: [PATCH 311/525] net: mana: check xdp_rxq registration before unreg in mana_destroy_rxq() When mana_create_rxq() fails at mana_create_wq_obj() or any step before xdp_rxq_info_reg() is called, the error path jumps to `out:` which calls mana_destroy_rxq(). mana_destroy_rxq() unconditionally calls xdp_rxq_info_unreg() on xilinx xdp_rxq that was never registered, triggering a WARN_ON in net/core/xdp.c: mana 7870:00:00.0: HWC: Failed hw_channel req: 0xc000009a mana 7870:00:00.0 eth7: Failed to create RXQ: err = -71 Driver BUG WARNING: CPU: 442 PID: 491615 at ../net/core/xdp.c:150 xdp_rxq_info_unreg+0x44/0x70 Modules linked in: tcp_bbr xsk_diag udp_diag raw_diag unix_diag af_packet_diag netlink_diag nf_tables nfnetlink tcp_diag inet_diag binfmt_misc rpcsec_gss_krb5 nfsv3 nfs_acl auth_rpcgss nfsv4 dns_resolver nfs lockd ext4 grace crc16 iscsi_tcp mbcache fscache libiscsi_tcp jbd2 netfs rpcrdma af_packet sunrpc rdma_ucm ib_iser rdma_cm iw_cm iscsi_ibft ib_cm iscsi_boot_sysfs libiscsi rfkill scsi_transport_iscsi mana_ib ib_uverbs ib_core mana hyperv_drm(X) drm_shmem_helper intel_rapl_msr drm_kms_helper intel_rapl_common syscopyarea nls_iso8859_1 sysfillrect intel_uncore_frequency_common nls_cp437 vfat fat nfit sysimgblt libnvdimm hv_netvsc(X) hv_utils(X) fb_sys_fops hv_balloon(X) joydev fuse drm dm_mod configfs ip_tables x_tables xfs libcrc32c sd_mod nvme nvme_core nvme_common t10_pi crc64_rocksoft_generic crc64_rocksoft crc64 hid_generic serio_raw pci_hyperv(X) hv_storvsc(X) scsi_transport_fc hyperv_keyboard(X) hid_hyperv(X) pci_hyperv_intf(X) crc32_pclmul crc32c_intel ghash_clmulni_intel aesni_intel crypto_simd cryptd hv_vmbus(X) softdog sg scsi_mod efivarfs Supported: Yes, External CPU: 442 PID: 491615 Comm: ethtool Kdump: loaded Tainted: G X 5.14.21-150500.55.136-default #1 SLE15-SP5 a627be1b53abbfd64ad16b2685e4308c52847f42 Hardware name: Microsoft Corporation Virtual Machine/Virtual Machine, BIOS Hyper-V UEFI Release v4.1 07/25/2025 RIP: 0010:xdp_rxq_info_unreg+0x44/0x70 Code: e8 91 fe ff ff c7 43 0c 02 00 00 00 48 c7 03 00 00 00 00 5b c3 cc cc cc cc e9 58 3a 1c 00 48 c7 c7 f6 5f 19 97 e8 5c a4 7e ff <0f> 0b 83 7b 0c 01 74 ca 48 c7 c7 d9 5f 19 97 e8 48 a4 7e ff 0f 0b RSP: 0018:ff3df6c8f7207818 EFLAGS: 00010286 RAX: 0000000000000000 RBX: ff30d89f94808a80 RCX: 0000000000000027 RDX: 0000000000000000 RSI: 0000000000000002 RDI: ff30d94bdcca2908 RBP: 0000000000080000 R08: ffffffff98ed11a0 R09: ff3df6c8f72077a0 R10: dead000000000100 R11: 000000000000000a R12: 0000000000000000 R13: 0000000000002000 R14: 0000000000040000 R15: ff30d89f94800000 FS: 00007fe6d8432b80(0000) GS:ff30d94bdcc80000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fe6d81a89b1 CR3: 00000b3b6d578001 CR4: 0000000000371ee0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe07f0 DR7: 0000000000000400 Call Trace: mana_destroy_rxq+0x5b/0x2f0 [mana 267acf7006bcb696095bba4d810643d1db3b9e94] mana_create_rxq.isra.55+0x3db/0x720 [mana 267acf7006bcb696095bba4d810643d1db3b9e94] ? simple_lookup+0x36/0x50 ? current_time+0x42/0x80 ? __d_free_external+0x30/0x30 mana_alloc_queues+0x32a/0x470 [mana 267acf7006bcb696095bba4d810643d1db3b9e94] ? _raw_spin_unlock+0xa/0x30 ? d_instantiate.part.29+0x2e/0x40 ? _raw_spin_unlock+0xa/0x30 ? debugfs_create_dir+0xe4/0x140 mana_attach+0x5c/0xf0 [mana 267acf7006bcb696095bba4d810643d1db3b9e94] mana_set_ringparam+0xd5/0x1a0 [mana 267acf7006bcb696095bba4d810643d1db3b9e94] ethnl_set_rings+0x292/0x320 genl_family_rcv_msg_doit.isra.15+0x11b/0x150 genl_rcv_msg+0xe3/0x1e0 ? rings_prepare_data+0x80/0x80 ? genl_family_rcv_msg_doit.isra.15+0x150/0x150 netlink_rcv_skb+0x50/0x100 genl_rcv+0x24/0x40 netlink_unicast+0x1b6/0x280 netlink_sendmsg+0x365/0x4d0 sock_sendmsg+0x5f/0x70 __sys_sendto+0x112/0x140 __x64_sys_sendto+0x24/0x30 do_syscall_64+0x5b/0x80 ? handle_mm_fault+0xd7/0x290 ? do_user_addr_fault+0x2d8/0x740 ? exc_page_fault+0x67/0x150 entry_SYSCALL_64_after_hwframe+0x6b/0xd5 RIP: 0033:0x7fe6d8122f06 Code: 00 00 00 00 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 41 89 ca 64 8b 04 25 18 00 00 00 85 c0 75 11 b8 2c 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 72 f3 c3 41 57 41 56 4d 89 c7 41 55 41 54 41 RSP: 002b:00007fff2b66b068 EFLAGS: 00000246 ORIG_RAX: 000000000000002c RAX: ffffffffffffffda RBX: 000055771123d2a0 RCX: 00007fe6d8122f06 RDX: 0000000000000034 RSI: 000055771123d3b0 RDI: 0000000000000003 RBP: 00007fff2b66b100 R08: 00007fe6d8203360 R09: 000000000000000c R10: 0000000000000000 R11: 0000000000000246 R12: 000055771123d350 R13: 000055771123d340 R14: 0000000000000000 R15: 00007fff2b66b2b0 Guard the xdp_rxq_info_unreg() call with xdp_rxq_info_is_reg() so that mana_destroy_rxq() is safe to call regardless of how far initialization progressed. Fixes: ed5356b53f07 ("net: mana: Add XDP support") Reviewed-by: Haiyang Zhang Signed-off-by: Dipayaan Roy Link: https://patch.msgid.link/20260430035935.1859220-2-dipayanroy@linux.microsoft.com Reviewed-by: Simon Horman Signed-off-by: Paolo Abeni --- drivers/net/ethernet/microsoft/mana/mana_en.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/microsoft/mana/mana_en.c b/drivers/net/ethernet/microsoft/mana/mana_en.c index a654b3699c4c..dfb4ba9f7664 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_en.c +++ b/drivers/net/ethernet/microsoft/mana/mana_en.c @@ -2520,7 +2520,9 @@ static void mana_destroy_rxq(struct mana_port_context *apc, napi_disable_locked(napi); netif_napi_del_locked(napi); } - xdp_rxq_info_unreg(&rxq->xdp_rxq); + + if (xdp_rxq_info_is_reg(&rxq->xdp_rxq)) + xdp_rxq_info_unreg(&rxq->xdp_rxq); mana_destroy_wq_obj(apc, GDMA_RQ, rxq->rxobj); From 2a1c691182823a5c149d502ac153e249ee697b4a Mon Sep 17 00:00:00 2001 From: Dipayaan Roy Date: Wed, 29 Apr 2026 20:57:53 -0700 Subject: [PATCH 312/525] net: mana: Skip WQ object destruction for uninitialized RXQ In mana_destroy_rxq(), mana_destroy_wq_obj() is called unconditionally even when the WQ object was never created (rxobj is still INVALID_MANA_HANDLE). When mana_create_rxq() fails before mana_create_wq_obj() succeeds, the error path calls mana_destroy_rxq() which sends a bogus destroy command to the hardware: mana 7870:00:00.0: HWC: Failed hw_channel req: 0x1d mana 7870:00:00.0: Failed to send mana message: -71, 0x1d mana 7870:00:00.0 eth7: Failed to destroy WQ object: -71 Guard mana_destroy_wq_obj() with an INVALID_MANA_HANDLE check so that mana_destroy_rxq() is safe to call at any stage of RXQ initialization. Fixes: ca9c54d2d6a5 ("net: mana: Add a driver for Microsoft Azure Network Adapter (MANA)") Reviewed-by: Haiyang Zhang Signed-off-by: Dipayaan Roy Link: https://patch.msgid.link/20260430035935.1859220-3-dipayanroy@linux.microsoft.com Reviewed-by: Simon Horman Signed-off-by: Paolo Abeni --- drivers/net/ethernet/microsoft/mana/mana_en.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/microsoft/mana/mana_en.c b/drivers/net/ethernet/microsoft/mana/mana_en.c index dfb4ba9f7664..f2a6ea162dc3 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_en.c +++ b/drivers/net/ethernet/microsoft/mana/mana_en.c @@ -2524,7 +2524,8 @@ static void mana_destroy_rxq(struct mana_port_context *apc, if (xdp_rxq_info_is_reg(&rxq->xdp_rxq)) xdp_rxq_info_unreg(&rxq->xdp_rxq); - mana_destroy_wq_obj(apc, GDMA_RQ, rxq->rxobj); + if (rxq->rxobj != INVALID_MANA_HANDLE) + mana_destroy_wq_obj(apc, GDMA_RQ, rxq->rxobj); mana_deinit_cq(apc, &rxq->rx_cq); From 3985c9a56da49af8b2e45cb1fa55c03c89b1d471 Mon Sep 17 00:00:00 2001 From: Dipayaan Roy Date: Wed, 29 Apr 2026 20:57:54 -0700 Subject: [PATCH 313/525] net: mana: remove double CQ cleanup in mana_create_rxq error path MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In mana_create_rxq(), the error cleanup path calls mana_destroy_rxq() followed by mana_deinit_cq(). This is incorrect for two reasons: 1. mana_destroy_rxq() already calls mana_deinit_cq() internally, so the CQ's GDMA queue is destroyed twice. 2. mana_destroy_rxq() frees the rxq via kfree(rxq) before returning. The subsequent mana_deinit_cq(apc, cq) then operates on freed memory since cq points to &rxq->rx_cq, which is embedded in the already-freed rxq structure — a use-after-free. Remove the redundant mana_deinit_cq() call from the error path since mana_destroy_rxq() already handles CQ cleanup. mana_deinit_cq() is itself safe for an uninitialized CQ as it checks for a NULL gdma_cq before proceeding. Fixes: ca9c54d2d6a5 ("net: mana: Add a driver for Microsoft Azure Network Adapter (MANA)") Reviewed-by: Haiyang Zhang Signed-off-by: Dipayaan Roy Reviewed-by: Aditya Garg Link: https://patch.msgid.link/20260430035935.1859220-4-dipayanroy@linux.microsoft.com Reviewed-by: Simon Horman Signed-off-by: Paolo Abeni --- drivers/net/ethernet/microsoft/mana/mana_en.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/net/ethernet/microsoft/mana/mana_en.c b/drivers/net/ethernet/microsoft/mana/mana_en.c index f2a6ea162dc3..9afc786b297a 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_en.c +++ b/drivers/net/ethernet/microsoft/mana/mana_en.c @@ -2799,9 +2799,6 @@ static struct mana_rxq *mana_create_rxq(struct mana_port_context *apc, mana_destroy_rxq(apc, rxq, false); - if (cq) - mana_deinit_cq(apc, cq); - return NULL; } From c69df06e4e26e50611190ce04eab92c5cc261b61 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 26 Mar 2026 12:28:21 +0100 Subject: [PATCH 314/525] perf/core: Fix deadlock in perf_mmap() failure path Ian noted that commit 77de62ad3de3 ("perf/core: Fix refcount bug and potential UAF in perf_mmap") would cause a deadlock due to event->mmap_mutex recursion. This happens because we're now calling perf_mmap_close() under mmap_mutex, while that function itself can also take mmap_mutex. Solve this by noting that perf_mmap_close() is far more complicated than we need at this particular point, since it deals with scenarios that cannot happen in this particular case. Replace the call to perf_mmap_close() with a very narrow undo for the case of first-exposure. If this is not the first mmap(), there is no race and it is fine to drop the lock and call perf_mmap_close() to handle to more complicated scenarios. Note: move the rb->mmap_user (namespace) handling into the rb init/free code such that it does not complicate the mmap handling. Fixes: 77de62ad3de3 ("perf/core: Fix refcount bug and potential UAF in perf_mmap") Reported-by: Ian Rogers Closes: https://patch.msgid.link/CAP-5%3DfVJyVMZw%3DDqP53Kxg58nUmJ_0bxoaeOKAbC03BVc11HaA%40mail.gmail.com Signed-off-by: Peter Zijlstra (Intel) Link: https://patch.msgid.link/20260326112821.GK3738786@noisy.programming.kicks-ass.net --- kernel/events/core.c | 70 +++++++++++++++++++++++++++++-------- kernel/events/internal.h | 1 + kernel/events/ring_buffer.c | 2 ++ 3 files changed, 58 insertions(+), 15 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index 6d1f8bad7e1c..7935d5663944 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -7006,6 +7006,7 @@ static void perf_mmap_open(struct vm_area_struct *vma) } static void perf_pmu_output_stop(struct perf_event *event); +static void perf_mmap_unaccount(struct vm_area_struct *vma, struct perf_buffer *rb); /* * A buffer can be mmap()ed multiple times; either directly through the same @@ -7021,8 +7022,6 @@ static void perf_mmap_close(struct vm_area_struct *vma) mapped_f unmapped = get_mapped(event, event_unmapped); struct perf_buffer *rb = ring_buffer_get(event); struct user_struct *mmap_user = rb->mmap_user; - int mmap_locked = rb->mmap_locked; - unsigned long size = perf_data_size(rb); bool detach_rest = false; /* FIXIES vs perf_pmu_unregister() */ @@ -7117,11 +7116,7 @@ static void perf_mmap_close(struct vm_area_struct *vma) * Aside from that, this buffer is 'fully' detached and unmapped, * undo the VM accounting. */ - - atomic_long_sub((size >> PAGE_SHIFT) + 1 - mmap_locked, - &mmap_user->locked_vm); - atomic64_sub(mmap_locked, &vma->vm_mm->pinned_vm); - free_uid(mmap_user); + perf_mmap_unaccount(vma, rb); out_put: ring_buffer_put(rb); /* could be last */ @@ -7261,6 +7256,15 @@ static void perf_mmap_account(struct vm_area_struct *vma, long user_extra, long atomic64_add(extra, &vma->vm_mm->pinned_vm); } +static void perf_mmap_unaccount(struct vm_area_struct *vma, struct perf_buffer *rb) +{ + struct user_struct *user = rb->mmap_user; + + atomic_long_sub((perf_data_size(rb) >> PAGE_SHIFT) + 1 - rb->mmap_locked, + &user->locked_vm); + atomic64_sub(rb->mmap_locked, &vma->vm_mm->pinned_vm); +} + static int perf_mmap_rb(struct vm_area_struct *vma, struct perf_event *event, unsigned long nr_pages) { @@ -7323,8 +7327,6 @@ static int perf_mmap_rb(struct vm_area_struct *vma, struct perf_event *event, if (!rb) return -ENOMEM; - refcount_set(&rb->mmap_count, 1); - rb->mmap_user = get_current_user(); rb->mmap_locked = extra; ring_buffer_attach(event, rb); @@ -7474,16 +7476,54 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma) mapped(event, vma->vm_mm); /* - * Try to map it into the page table. On fail, invoke - * perf_mmap_close() to undo the above, as the callsite expects - * full cleanup in this case and therefore does not invoke - * vmops::close(). + * Try to map it into the page table. On fail undo the above, + * as the callsite expects full cleanup in this case and + * therefore does not invoke vmops::close(). */ ret = map_range(event->rb, vma); - if (ret) - perf_mmap_close(vma); + if (likely(!ret)) + return 0; + + /* Error path */ + + /* + * If this is the first mmap(), then event->mmap_count should + * be stable at 1. It is only modified by: + * perf_mmap_{open,close}() and perf_mmap(). + * + * The former are not possible because this mmap() hasn't been + * successful yet, and the latter is serialized by + * event->mmap_mutex which we still hold (note that mmap_lock + * is not strictly sufficient here, because the event fd can + * be passed to another process through trivial means like + * fork(), leading to concurrent mmap() from different mm). + * + * Make sure to remove event->rb before releasing + * event->mmap_mutex, such that any concurrent mmap() will not + * attempt use this failed buffer. + */ + if (refcount_read(&event->mmap_count) == 1) { + /* + * Minimal perf_mmap_close(); there can't be AUX or + * other events on account of this being the first. + */ + mapped = get_mapped(event, event_unmapped); + if (mapped) + mapped(event, vma->vm_mm); + perf_mmap_unaccount(vma, event->rb); + ring_buffer_attach(event, NULL); /* drops last rb->refcount */ + refcount_set(&event->mmap_count, 0); + return ret; + } + + /* + * Otherwise this is an already existing buffer, and there is + * no race vs first exposure, so fall-through and call + * perf_mmap_close(). + */ } + perf_mmap_close(vma); return ret; } diff --git a/kernel/events/internal.h b/kernel/events/internal.h index d9cc57083091..c03c4f2eea57 100644 --- a/kernel/events/internal.h +++ b/kernel/events/internal.h @@ -67,6 +67,7 @@ static inline void rb_free_rcu(struct rcu_head *rcu_head) struct perf_buffer *rb; rb = container_of(rcu_head, struct perf_buffer, rcu_head); + free_uid(rb->mmap_user); rb_free(rb); } diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c index 3e7de2661417..9fe92161715e 100644 --- a/kernel/events/ring_buffer.c +++ b/kernel/events/ring_buffer.c @@ -340,6 +340,8 @@ ring_buffer_init(struct perf_buffer *rb, long watermark, int flags) rb->paused = 1; mutex_init(&rb->aux_mutex); + rb->mmap_user = get_current_user(); + refcount_set(&rb->mmap_count, 1); } void perf_aux_output_flag(struct perf_output_handle *handle, u64 flags) From 5ad732a56be46aabf158c16aa0c095291727aaef Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 30 Apr 2026 08:25:54 +0800 Subject: [PATCH 315/525] perf/x86/intel: Improve validation and configuration of ACR masks Currently there are several issues on the user space ACR mask validation and configuration. - The validation for user space ACR mask (attr.config2) is incomplete, e.g., the ACR mask could include the index which belongs to another ACR events group, but it's not validated. - An early return on an invalid ACR mask caused all subsequent ACR groups to be skipped. - The stale hardware ACR mask (hw.config1) is not cleared before setting new hardware ACR mask. The following changes address all of the above issues. - Figure out the event index group of an ACR group. Any bits in the user-space mask not present in the index group are now dropped. - Instead of an early return on invalid bits, drop only the invalid portions and continue iterating through all ACR events to ensure full configuration. - Explicitly clear the stale hardware ACR mask for each event prior to writing the new configuration. Besides, a non-leader event member of ACR group could be disabled in theory. This could cause bit-shifting errors in the acr_mask of remaining group members. But since ACR sampling requires all events to be active, this should not be a big concern in real use case. Add a "FIXME" comment to notice this risk. Fixes: ec980e4facef ("perf/x86/intel: Support auto counter reload") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260430002558.712334-2-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/core.c | 32 +++++++++++++++++++++++++------- 1 file changed, 25 insertions(+), 7 deletions(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index d9488ade0f8e..f8deb67b3c51 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -3332,23 +3332,41 @@ static void intel_pmu_enable_event(struct perf_event *event) static void intel_pmu_acr_late_setup(struct cpu_hw_events *cpuc) { struct perf_event *event, *leader; - int i, j, idx; + int i, j, k, bit, idx; + /* + * FIXME: ACR mask parsing relies on cpuc->event_list[] (active events only). + * Disabling an ACR event causes bit-shifting errors in the acr_mask of + * remaining group members. As ACR sampling requires all events to be active, + * this limitation is acceptable for now. Revisit if independent event toggling + * is required. + */ for (i = 0; i < cpuc->n_events; i++) { leader = cpuc->event_list[i]; if (!is_acr_event_group(leader)) continue; - /* The ACR events must be contiguous. */ + /* Find the last event of the ACR group. */ for (j = i; j < cpuc->n_events; j++) { event = cpuc->event_list[j]; if (event->group_leader != leader->group_leader) break; - for_each_set_bit(idx, (unsigned long *)&event->attr.config2, X86_PMC_IDX_MAX) { - if (i + idx >= cpuc->n_events || - !is_acr_event_group(cpuc->event_list[i + idx])) - return; - __set_bit(cpuc->assign[i + idx], (unsigned long *)&event->hw.config1); + } + + /* + * Translate the user-space ACR mask (attr.config2) into the physical + * counter bitmask (hw.config1) for each ACR event in the group. + * NOTE: ACR event contiguity is guaranteed by intel_pmu_hw_config(). + */ + for (k = i; k < j; k++) { + event = cpuc->event_list[k]; + event->hw.config1 = 0; + for_each_set_bit(bit, (unsigned long *)&event->attr.config2, X86_PMC_IDX_MAX) { + idx = i + bit; + /* Event index of ACR group must locate in [i, j). */ + if (idx >= j || !is_acr_event_group(cpuc->event_list[idx])) + continue; + __set_bit(cpuc->assign[idx], (unsigned long *)&event->hw.config1); } } i = j - 1; From 8ba0b706a485b1e607594cf4210786d517ad1611 Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 30 Apr 2026 08:25:55 +0800 Subject: [PATCH 316/525] perf/x86/intel: Always reprogram ACR events to prevent stale masks Members of an ACR group are logically linked via a bitmask of their hardware counter indices. If some members of the group are assigned new hardware counters during rescheduling, even events that keep their original counter index must be updated with a new mask. Without this, an event will continue to use a stale acr_mask that references the old indices of its group peers. Ensure all ACR events are reprogrammed during the scheduling path to maintain consistency across the group. Fixes: ec980e4facef ("perf/x86/intel: Support auto counter reload") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260430002558.712334-3-dapeng1.mi@linux.intel.com --- arch/x86/events/core.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c index 810ab21ffd99..4b9e105309c6 100644 --- a/arch/x86/events/core.c +++ b/arch/x86/events/core.c @@ -1294,13 +1294,16 @@ int x86_perf_rdpmc_index(struct perf_event *event) return event->hw.event_base_rdpmc; } -static inline int match_prev_assignment(struct hw_perf_event *hwc, +static inline int match_prev_assignment(struct perf_event *event, struct cpu_hw_events *cpuc, int i) { + struct hw_perf_event *hwc = &event->hw; + return hwc->idx == cpuc->assign[i] && - hwc->last_cpu == smp_processor_id() && - hwc->last_tag == cpuc->tags[i]; + hwc->last_cpu == smp_processor_id() && + hwc->last_tag == cpuc->tags[i] && + !is_acr_event_group(event); } static void x86_pmu_start(struct perf_event *event, int flags); @@ -1346,7 +1349,7 @@ static void x86_pmu_enable(struct pmu *pmu) * - no other event has used the counter since */ if (hwc->idx == -1 || - match_prev_assignment(hwc, cpuc, i)) + match_prev_assignment(event, cpuc, i)) continue; /* @@ -1367,7 +1370,7 @@ static void x86_pmu_enable(struct pmu *pmu) event = cpuc->event_list[i]; hwc = &event->hw; - if (!match_prev_assignment(hwc, cpuc, i)) + if (!match_prev_assignment(event, cpuc, i)) x86_assign_hw_event(event, cpuc, i); else if (i < n_running) continue; From 1271aeccc307066315b2d3b0d5af2510e27018b5 Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 30 Apr 2026 08:25:56 +0800 Subject: [PATCH 317/525] perf/x86/intel: Disable PMI for self-reloaded ACR events On platforms with Auto Counter Reload (ACR) support, such as NVL, a "NMI received for unknown reason 30" warning is observed when running multiple events in a group with ACR enabled: $ perf record -e '{instructions/period=20000,acr_mask=0x2/u,\ cycles/period=40000,acr_mask=0x3/u}' ./test The warning occurs because the Performance Monitoring Interrupt (PMI) is enabled for the self-reloaded event (the cycles event in this case). According to the Intel SDM, the overflow bit (IA32_PERF_GLOBAL_STATUS.PMCn_OVF) is never set for self-reloaded events. Since the bit is not set, the perf NMI handler cannot identify the source of the interrupt, leading to the "unknown reason" message. Furthermore, enabling PMI for self-reloaded events is unnecessary and can lead to extraneous records that pollute the user's requested data. Disable the interrupt bit for all events configured with ACR self-reload. Fixes: ec980e4facef ("perf/x86/intel: Support auto counter reload") Reported-by: Andi Kleen Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260430002558.712334-4-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/core.c | 17 +++++++++++++---- arch/x86/events/perf_event.h | 10 ++++++++++ 2 files changed, 23 insertions(+), 4 deletions(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index f8deb67b3c51..ead6d95cec6a 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -3118,11 +3118,11 @@ static void intel_pmu_enable_fixed(struct perf_event *event) intel_set_masks(event, idx); /* - * Enable IRQ generation (0x8), if not PEBS, - * and enable ring-3 counting (0x2) and ring-0 counting (0x1) - * if requested: + * Enable IRQ generation (0x8), if not PEBS or self-reloaded + * ACR event, and enable ring-3 counting (0x2) and ring-0 + * counting (0x1) if requested: */ - if (!event->attr.precise_ip) + if (!event->attr.precise_ip && !is_acr_self_reload_event(event)) bits |= INTEL_FIXED_0_ENABLE_PMI; if (hwc->config & ARCH_PERFMON_EVENTSEL_USR) bits |= INTEL_FIXED_0_USER; @@ -3306,6 +3306,15 @@ static void intel_pmu_enable_event(struct perf_event *event) intel_set_masks(event, idx); static_call_cond(intel_pmu_enable_acr_event)(event); static_call_cond(intel_pmu_enable_event_ext)(event); + /* + * For self-reloaded ACR event, don't enable PMI since + * HW won't set overflow bit in GLOBAL_STATUS. Otherwise, + * the PMI would be recognized as a suspicious NMI. + */ + if (is_acr_self_reload_event(event)) + hwc->config &= ~ARCH_PERFMON_EVENTSEL_INT; + else if (!event->attr.precise_ip) + hwc->config |= ARCH_PERFMON_EVENTSEL_INT; __x86_pmu_enable_event(hwc, enable_mask); break; case INTEL_PMC_IDX_FIXED ... INTEL_PMC_IDX_FIXED_BTS - 1: diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h index fad87d3c8b2c..524668dcf4cc 100644 --- a/arch/x86/events/perf_event.h +++ b/arch/x86/events/perf_event.h @@ -137,6 +137,16 @@ static inline bool is_acr_event_group(struct perf_event *event) return check_leader_group(event->group_leader, PERF_X86_EVENT_ACR); } +static inline bool is_acr_self_reload_event(struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + if (hwc->idx < 0) + return false; + + return test_bit(hwc->idx, (unsigned long *)&hwc->config1); +} + struct amd_nb { int nb_id; /* NorthBridge id */ int refcnt; /* reference count */ From aa4384bc8f4360167f3c3d5322121fe892289ea2 Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 30 Apr 2026 08:25:57 +0800 Subject: [PATCH 318/525] perf/x86/intel: Enable auto counter reload for DMR MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Panther cove µarch starts to support auto counter reload (ACR), but the static_call intel_pmu_enable_acr_event() is not updated for the Panther Cove µarch used by DMR. It leads to the auto counter reload is not really enabled on DMR. Update static_call intel_pmu_enable_acr_event() in intel_pmu_init_pnc(). Fixes: d345b6bb8860 ("perf/x86/intel: Add core PMU support for DMR") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260430002558.712334-5-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index ead6d95cec6a..dd1e3aa75ee9 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -7531,6 +7531,7 @@ static __always_inline void intel_pmu_init_pnc(struct pmu *pmu) hybrid(pmu, event_constraints) = intel_pnc_event_constraints; hybrid(pmu, pebs_constraints) = intel_pnc_pebs_event_constraints; hybrid(pmu, extra_regs) = intel_pnc_extra_regs; + static_call_update(intel_pmu_enable_acr_event, intel_pmu_enable_acr); } static __always_inline void intel_pmu_init_skt(struct pmu *pmu) From 78538047717bdeabe8481ef611c9131e455e61df Mon Sep 17 00:00:00 2001 From: Richard Fitzgerald Date: Tue, 5 May 2026 11:51:22 +0100 Subject: [PATCH 319/525] ASoC: wm_adsp_fw_find_test: Redirect wm_adsp_release_firmware_files() Redirect wm_adsp_release_firmware_files() to a replacement function that handles the dummy firmware created by the tests. Use the same cleanup function to cleanup in the test exit() function. Also call it on each loop in wm_adsp_fw_find_test_find_firmware_byindex() to free the created strings before reusing priv->found_fw on the next loop. wm_adsp_release_firmware_files() will pass the struct firmware* pointers to release_firmware(). But the pointers created by the tests are dummies and must not be passed to release_firmware(). The test never invokes wm_adsp_release_firmware_files() so it wasn't redirected. But the error handling in wm_adsp_request_firmware_files() calls wm_adsp_release_firmware_files(). The redirected function makes this safe. Using the same cleanup function to perform cleanup from the test exit() handler and wm_adsp_fw_find_test_find_firmware_byindex() avoids the risk of duplicate cleanup code that all needs updating if there is any change to the cleanup requirements. This problem was found by https://sashiko.dev. Fixes: bf2d44d07de7 ("ASoC: wm_adsp: Add kunit test for firmware file search") Closes: https://sashiko.dev/#/patchset/20260326100853.1582886-1-rf%40opensource.cirrus.com Signed-off-by: Richard Fitzgerald Link: https://patch.msgid.link/20260505105123.3539778-2-rf@opensource.cirrus.com Signed-off-by: Mark Brown --- sound/soc/codecs/wm_adsp_fw_find_test.c | 56 ++++++++++++++++++++----- 1 file changed, 46 insertions(+), 10 deletions(-) diff --git a/sound/soc/codecs/wm_adsp_fw_find_test.c b/sound/soc/codecs/wm_adsp_fw_find_test.c index d0c7fb30a95d..516ba08eb5ba 100644 --- a/sound/soc/codecs/wm_adsp_fw_find_test.c +++ b/sound/soc/codecs/wm_adsp_fw_find_test.c @@ -45,6 +45,34 @@ struct wm_adsp_fw_find_test_params { /* Dummy struct firmware to return from wm_adsp_request_firmware_files */ static const struct firmware wm_adsp_find_test_dummy_firmware; +static void wm_adsp_fw_find_test_release_firmware_files_stub(struct wm_adsp_fw_files *fw) +{ + /* + * fw->wmfw.firmware and fw->coeff.firmware allocated by this KUnit + * test are dummies not allocated by the real request_firmware() call + * so they must not be passed to release_firmware(). + * This function replaces wm_adsp_release_firmware_files(). + */ + + if (!fw) + return; + + kfree(fw->wmfw.filename); + kfree(fw->coeff.filename); + + fw->wmfw.firmware = NULL; + fw->coeff.firmware = NULL; + fw->wmfw.filename = NULL; + fw->coeff.filename = NULL; +} + +static void wm_adsp_free_found_fw(struct kunit *test) +{ + struct wm_adsp_fw_find_test *priv = test->priv; + + wm_adsp_fw_find_test_release_firmware_files_stub(&priv->found_fw); +} + /* Simple lookup of a filename in a list of names */ static int wm_adsp_fw_find_test_firmware_request_simple_stub(const struct firmware **firmware, const char *filename, @@ -97,9 +125,14 @@ static void wm_adsp_fw_find_test_pick_file(struct kunit *test) kunit_activate_static_stub(test, wm_adsp_firmware_request, wm_adsp_fw_find_test_firmware_request_simple_stub); + kunit_activate_static_stub(test, + wm_adsp_release_firmware_files, + wm_adsp_fw_find_test_release_firmware_files_stub); ret = wm_adsp_request_firmware_files(dsp, &priv->found_fw); kunit_deactivate_static_stub(test, wm_adsp_firmware_request); + kunit_deactivate_static_stub(test, wm_adsp_release_firmware_files); + KUNIT_EXPECT_EQ_MSG(test, ret, (params->expect_wmfw || params->expect_bin) ? 0 : -ENOENT, "%s\n", priv->searched_fw_files); @@ -173,10 +206,13 @@ static void wm_adsp_fw_find_test_search_order(struct kunit *test) kunit_activate_static_stub(test, wm_adsp_firmware_request, wm_adsp_fw_find_test_firmware_request_stub); + kunit_activate_static_stub(test, + wm_adsp_release_firmware_files, + wm_adsp_fw_find_test_release_firmware_files_stub); wm_adsp_request_firmware_files(dsp, &priv->found_fw); - kunit_deactivate_static_stub(test, wm_adsp_firmware_request); + kunit_deactivate_static_stub(test, wm_adsp_release_firmware_files); KUNIT_EXPECT_STREQ(test, priv->searched_fw_files, params->expected_searches); @@ -201,6 +237,7 @@ static void wm_adsp_fw_find_test_find_firmware_byindex(struct kunit *test) dsp->cs_dsp.name = "cs1234"; dsp->part = "dsp1"; + for (dsp->fw = 0;; dsp->fw++) { fw_name = wm_adsp_get_fwf_name_by_index(dsp->fw); if (!fw_name) @@ -209,14 +246,21 @@ static void wm_adsp_fw_find_test_find_firmware_byindex(struct kunit *test) kunit_activate_static_stub(test, wm_adsp_firmware_request, wm_adsp_fw_find_test_firmware_request_stub); + kunit_activate_static_stub(test, + wm_adsp_release_firmware_files, + wm_adsp_fw_find_test_release_firmware_files_stub); wm_adsp_request_firmware_files(dsp, &priv->found_fw); + kunit_deactivate_static_stub(test, wm_adsp_firmware_request); + kunit_deactivate_static_stub(test, wm_adsp_release_firmware_files); KUNIT_EXPECT_NOT_NULL_MSG(test, strstr(priv->searched_fw_files, fw_name), "fw#%d Did not find '%s' in '%s'\n", dsp->fw, fw_name, priv->searched_fw_files); + + wm_adsp_free_found_fw(test); } } @@ -255,15 +299,7 @@ static int wm_adsp_fw_find_test_case_init(struct kunit *test) static void wm_adsp_fw_find_test_case_exit(struct kunit *test) { - struct wm_adsp_fw_find_test *priv = test->priv; - - /* - * priv->found_wmfw_firmware and priv->found_bin_firmware are - * dummies not allocated by the real request_firmware() call they - * must not be passed to release_firmware(). - */ - kfree(priv->found_fw.wmfw.filename); - kfree(priv->found_fw.coeff.filename); + wm_adsp_free_found_fw(test); } static void wm_adsp_fw_find_test_param_desc(const struct wm_adsp_fw_find_test_params *param, From af64f790969973b325efda7264d6860167623cdd Mon Sep 17 00:00:00 2001 From: Richard Fitzgerald Date: Tue, 5 May 2026 11:51:23 +0100 Subject: [PATCH 320/525] ASoC: wm_adsp_fw_find_test: Clear searched_fw_files in find-by-index test In wm_adsp_fw_find_test_find_firmware_byindex() the content of priv->searched_fw_files must be cleared before starting the next iteration. The files searched for are appended to priv->searched_fw_files, so if it is not cleared on each iteration it will still contain the searches from the previous iteration. Fixes: bf2d44d07de7 ("ASoC: wm_adsp: Add kunit test for firmware file search") Signed-off-by: Richard Fitzgerald Link: https://patch.msgid.link/20260505105123.3539778-3-rf@opensource.cirrus.com Signed-off-by: Mark Brown --- sound/soc/codecs/wm_adsp_fw_find_test.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/codecs/wm_adsp_fw_find_test.c b/sound/soc/codecs/wm_adsp_fw_find_test.c index 516ba08eb5ba..ae686dc4fa94 100644 --- a/sound/soc/codecs/wm_adsp_fw_find_test.c +++ b/sound/soc/codecs/wm_adsp_fw_find_test.c @@ -261,6 +261,7 @@ static void wm_adsp_fw_find_test_find_firmware_byindex(struct kunit *test) dsp->fw, fw_name, priv->searched_fw_files); wm_adsp_free_found_fw(test); + memset(priv->searched_fw_files, 0, sizeof(priv->searched_fw_files)); } } From 50987d4e6c55929aa2d4d3976e74ccbae22d5017 Mon Sep 17 00:00:00 2001 From: Chen Ni Date: Fri, 27 Mar 2026 10:17:28 +0800 Subject: [PATCH 321/525] drm/panel: himax-hx83121a: Fix incorrect error check for devm_drm_panel_alloc() Check devm_drm_panel_alloc() return value for ERR_PTR instead of NULL. devm_drm_panel_alloc() returns an ERR_PTR on failure, never NULL. Using a NULL check skips the error path and may cause a NULL pointer dereference. Fixes: a7c61963b727 ("drm/panel: Add Himax HX83121A panel driver") Signed-off-by: Chen Ni Reviewed-by: Pengyu Luo Signed-off-by: Neil Armstrong Link: https://patch.msgid.link/20260327021728.647182-1-nichen@iscas.ac.cn --- drivers/gpu/drm/panel/panel-himax-hx83121a.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/panel/panel-himax-hx83121a.c b/drivers/gpu/drm/panel/panel-himax-hx83121a.c index ebe643ba4184..bed79aa06f46 100644 --- a/drivers/gpu/drm/panel/panel-himax-hx83121a.c +++ b/drivers/gpu/drm/panel/panel-himax-hx83121a.c @@ -596,8 +596,8 @@ static int himax_probe(struct mipi_dsi_device *dsi) ctx = devm_drm_panel_alloc(dev, struct himax, panel, &himax_panel_funcs, DRM_MODE_CONNECTOR_DSI); - if (!ctx) - return -ENOMEM; + if (IS_ERR(ctx)) + return PTR_ERR(ctx); ret = devm_regulator_bulk_get_const(&dsi->dev, ARRAY_SIZE(himax_supplies), From defab7b01e0848e004077d7d8dcc04d305ea1a27 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 13 Apr 2026 09:10:19 +0200 Subject: [PATCH 322/525] drm/panel: hx83121a: select DRM_DISPLAY_DSC_HELPER Like a number of other panel drivers, this newly merged driver needs DRM_DISPLAY_DSC_HELPER to be enabled: arm-linux-gnueabi-ld: drivers/gpu/drm/panel/panel-himax-hx83121a.o: in function `himax_prepare': panel-himax-hx83121a.c:(.text+0x1024): undefined reference to `drm_dsc_pps_payload_pack' Fixes: a7c61963b727 ("drm/panel: Add Himax HX83121A panel driver") Signed-off-by: Arnd Bergmann Reviewed-by: Neil Armstrong Reviewed-by: David Heidelberg Signed-off-by: Neil Armstrong Link: https://patch.msgid.link/20260413071043.3829868-1-arnd@kernel.org --- drivers/gpu/drm/panel/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/panel/Kconfig b/drivers/gpu/drm/panel/Kconfig index d6863b28ddc5..d592f4f4b939 100644 --- a/drivers/gpu/drm/panel/Kconfig +++ b/drivers/gpu/drm/panel/Kconfig @@ -208,6 +208,7 @@ config DRM_PANEL_HIMAX_HX83121A depends on OF depends on DRM_MIPI_DSI depends on BACKLIGHT_CLASS_DEVICE + select DRM_DISPLAY_DSC_HELPER select DRM_KMS_HELPER help Say Y here if you want to enable support for Himax HX83121A-based From c67e8787f6743101c90c7a9c4bb7cf6f1f739f83 Mon Sep 17 00:00:00 2001 From: Christian Van Date: Sat, 25 Apr 2026 01:39:48 -0400 Subject: [PATCH 323/525] drm/panel: feiyang-fy07024di26a30d: return display-on error mipi_dsi_dcs_set_display_on() returns an error code, but feiyang_enable() currently ignores it and always reports success. Return the DCS command result so callers can observe enable failures. Signed-off-by: Christian Van Reviewed-by: Neil Armstrong Signed-off-by: Neil Armstrong Link: https://patch.msgid.link/20260425053948.117714-1-cvan20191@gmail.com --- drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c b/drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c index 4f8d6d8c07e4..dbdb7e3cb7b6 100644 --- a/drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c +++ b/drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c @@ -98,9 +98,7 @@ static int feiyang_enable(struct drm_panel *panel) /* T12 (video & logic signal rise + backlight rise) T12 >= 200ms */ msleep(200); - mipi_dsi_dcs_set_display_on(ctx->dsi); - - return 0; + return mipi_dsi_dcs_set_display_on(ctx->dsi); } static int feiyang_disable(struct drm_panel *panel) From 570cf799e87ae805eacfab3b4ba66676b5fccdb6 Mon Sep 17 00:00:00 2001 From: Icenowy Zheng Date: Sun, 3 May 2026 17:17:08 +0800 Subject: [PATCH 324/525] drm/panel: boe-tv101wum-nl6: restore MODE_LPM after sending disable cmds When preparing the panel, it seems that it always expects commands to be transferred in LP mode. However, the disable function removes the MIPI_DSI_MODE_LPM flag, and no other function re-adds it. As the unprepare function contains no DSI commands, re-adding the flag just after disabling the panel should be safe. Add the code re-adding the flag after the two commands for disabling the panel are sent. This fixes error messages shown in kernel log when unblanking on mt8183-kukui-kodama-sku32 device. Cc: stable@vger.kernel.org Fixes: a869b9db7adf ("drm/panel: support for boe tv101wum-nl6 wuxga dsi video mode panel") Signed-off-by: Icenowy Zheng Reviewed-by: Neil Armstrong Reviewed-by: Douglas Anderson Signed-off-by: Neil Armstrong Link: https://patch.msgid.link/20260503091708.1079962-1-zhengxingda@iscas.ac.cn --- drivers/gpu/drm/panel/panel-boe-tv101wum-nl6.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/panel/panel-boe-tv101wum-nl6.c b/drivers/gpu/drm/panel/panel-boe-tv101wum-nl6.c index d5fe105bdbdd..658ce64c71eb 100644 --- a/drivers/gpu/drm/panel/panel-boe-tv101wum-nl6.c +++ b/drivers/gpu/drm/panel/panel-boe-tv101wum-nl6.c @@ -1324,6 +1324,8 @@ static int boe_panel_disable(struct drm_panel *panel) mipi_dsi_dcs_set_display_off_multi(&ctx); mipi_dsi_dcs_enter_sleep_mode_multi(&ctx); + boe->dsi->mode_flags |= MIPI_DSI_MODE_LPM; + mipi_dsi_msleep(&ctx, 150); return ctx.accum_err; From 2d4e80271f784aa0c7b17676e9762c7e8156be1c Mon Sep 17 00:00:00 2001 From: Icenowy Zheng Date: Sun, 26 Apr 2026 00:57:51 +0800 Subject: [PATCH 325/525] drm/panel: himax-hx83102: restore MODE_LPM after sending disable cmds When preparing the panel, it seems that it always expects commands to be transferred in LP mode. However, the disable function removes the MIPI_DSI_MODE_LPM flag, and no other function re-adds it. As the unprepare function contains no DSI commands, re-adding the flag just after disabling the panel should be safe. Add the code re-adding the flag after the two commands for disabling the panel are sent. This fixes screen unblanking (after blanking once) on mt8188-geralt-ciri-sku1 device. Cc: stable@vger.kernel.org # 6.11+ Fixes: 0ef94554dc40 ("drm/panel: himax-hx83102: Break out as separate driver") Signed-off-by: Icenowy Zheng Reviewed-by: Neil Armstrong Reviewed-by: Douglas Anderson Signed-off-by: Neil Armstrong Link: https://patch.msgid.link/20260425165751.1716569-1-zhengxingda@iscas.ac.cn --- drivers/gpu/drm/panel/panel-himax-hx83102.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/panel/panel-himax-hx83102.c b/drivers/gpu/drm/panel/panel-himax-hx83102.c index 8b2a68ee851e..a5e5c9ea7a73 100644 --- a/drivers/gpu/drm/panel/panel-himax-hx83102.c +++ b/drivers/gpu/drm/panel/panel-himax-hx83102.c @@ -937,6 +937,8 @@ static int hx83102_disable(struct drm_panel *panel) mipi_dsi_dcs_set_display_off_multi(&dsi_ctx); mipi_dsi_dcs_enter_sleep_mode_multi(&dsi_ctx); + dsi->mode_flags |= MIPI_DSI_MODE_LPM; + mipi_dsi_msleep(&dsi_ctx, 150); return dsi_ctx.accum_err; From 8cf5dd235eff6008cb04c3d8064d2acfa90616f1 Mon Sep 17 00:00:00 2001 From: Prasanna Kumar T S M Date: Wed, 1 Apr 2026 04:18:56 -0700 Subject: [PATCH 326/525] EDAC/versalnet: Fix device name memory leak The device name allocated via kzalloc() in init_one_mc() is assigned to dev->init_name but never freed on the normal removal path. device_register() copies init_name and then sets dev->init_name to NULL, so the name pointer becomes unreachable from the device. Thus leaking memory. Use a stack-local char array instead of using kzalloc() for name. Fixes: d5fe2fec6c40 ("EDAC: Add a driver for the AMD Versal NET DDR controller") Signed-off-by: Prasanna Kumar T S M Signed-off-by: Borislav Petkov (AMD) Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260401111856.2342975-1-ptsm@linux.microsoft.com --- drivers/edac/versalnet_edac.c | 10 ++-------- 1 file changed, 2 insertions(+), 8 deletions(-) diff --git a/drivers/edac/versalnet_edac.c b/drivers/edac/versalnet_edac.c index ec1315582414..97ec05d68bbb 100644 --- a/drivers/edac/versalnet_edac.c +++ b/drivers/edac/versalnet_edac.c @@ -777,9 +777,9 @@ static int init_one_mc(struct mc_priv *priv, struct platform_device *pdev, int i u32 num_chans, rank, dwidth, config; struct edac_mc_layer layers[2]; struct mem_ctl_info *mci; + char name[MC_NAME_LEN]; struct device *dev; enum dev_type dt; - char *name; int rc; config = priv->adec[CONF + i * ADEC_NUM]; @@ -813,13 +813,9 @@ static int init_one_mc(struct mc_priv *priv, struct platform_device *pdev, int i layers[1].is_virt_csrow = false; rc = -ENOMEM; - name = kzalloc(MC_NAME_LEN, GFP_KERNEL); - if (!name) - return rc; - dev = kzalloc(sizeof(*dev), GFP_KERNEL); if (!dev) - goto err_name_free; + return rc; mci = edac_mc_alloc(i, ARRAY_SIZE(layers), layers, sizeof(struct mc_priv)); if (!mci) { @@ -858,8 +854,6 @@ static int init_one_mc(struct mc_priv *priv, struct platform_device *pdev, int i edac_mc_free(mci); err_dev_free: kfree(dev); -err_name_free: - kfree(name); return rc; } From 83861c48ba122f85cc8384780764b3a791341678 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Thu, 30 Apr 2026 23:32:50 +0200 Subject: [PATCH 327/525] openvswitch: vport: fix race between tunnel creation and linking When a tunnel vport is created it first creates the tunnel device, e.g., with geneve_dev_create_fb(), then it calls ovs_netdev_link() to take a reference and link it to the device that represents openvswitch datapath. The creation of the device is happening under RTNL, but then RTNL is released and re-acquired to find the device by name. It is technically possible for the tunnel device to be re-named or deleted within that window while RTNL is not held, and some other device created in its place. This will cause a non-tunnel device to be referenced in the vport and tunnel-specific functions used on it, e.g. vxlan_get_options() that directly casts the private netdev data into a struct vxlan_dev causing an invalid memory access: BUG: KASAN: slab-use-after-free in vxlan_get_options+0x323/0x3a0 vxlan_get_options+0x323/0x3a0 ovs_vport_cmd_new+0x6e3/0xd30 Fix that by taking a reference to the just created device before releasing RTNL. This ensures that the device in the vport is always the one that was just created. The search by name is only needed for a standard vport-netdev that links pre-existing devices, so that functionality and device type checks are moved to netdev_create(). It is also awkward that ovs_netdev_link() takes ownership of the vport and destroys it on failure. It doesn't know the type of the port it is dealing with, so we need to pass down the indicator that it's a tunnel, so the link can be properly deleted on failure. It's possible to refactor the logic to make the ovs_netdev_link() do only the linking part and let the callers perform a proper destruction, but it will be much more code for each legacy tunnel port type, so it is not worth it for the bug fix. Fixes: 614732eaa12d ("openvswitch: Use regular VXLAN net_device device") Reported-by: Yuan Tan Reported-by: Yifan Wu Reported-by: Juefei Pu Reported-by: Xin Liu Reported-by: Yang Yang Signed-off-by: Ilya Maximets Acked-by: Eelco Chaudron Link: https://patch.msgid.link/20260430213349.407991-1-i.maximets@ovn.org Signed-off-by: Paolo Abeni --- net/openvswitch/vport-geneve.c | 5 ++- net/openvswitch/vport-gre.c | 5 ++- net/openvswitch/vport-netdev.c | 58 ++++++++++++++++++++-------------- net/openvswitch/vport-netdev.h | 2 +- net/openvswitch/vport-vxlan.c | 5 ++- 5 files changed, 48 insertions(+), 27 deletions(-) diff --git a/net/openvswitch/vport-geneve.c b/net/openvswitch/vport-geneve.c index b10e1602c6b1..cb5ea4424ffc 100644 --- a/net/openvswitch/vport-geneve.c +++ b/net/openvswitch/vport-geneve.c @@ -97,6 +97,9 @@ static struct vport *geneve_tnl_create(const struct vport_parms *parms) goto error; } + vport->dev = dev; + netdev_hold(vport->dev, &vport->dev_tracker, GFP_KERNEL); + rtnl_unlock(); return vport; error: @@ -111,7 +114,7 @@ static struct vport *geneve_create(const struct vport_parms *parms) if (IS_ERR(vport)) return vport; - return ovs_netdev_link(vport, parms->name); + return ovs_netdev_link(vport, true); } static struct vport_ops ovs_geneve_vport_ops = { diff --git a/net/openvswitch/vport-gre.c b/net/openvswitch/vport-gre.c index 4014c9b5eb79..6cb5a697b396 100644 --- a/net/openvswitch/vport-gre.c +++ b/net/openvswitch/vport-gre.c @@ -63,6 +63,9 @@ static struct vport *gre_tnl_create(const struct vport_parms *parms) return ERR_PTR(err); } + vport->dev = dev; + netdev_hold(vport->dev, &vport->dev_tracker, GFP_KERNEL); + rtnl_unlock(); return vport; } @@ -75,7 +78,7 @@ static struct vport *gre_create(const struct vport_parms *parms) if (IS_ERR(vport)) return vport; - return ovs_netdev_link(vport, parms->name); + return ovs_netdev_link(vport, true); } static struct vport_ops ovs_gre_vport_ops = { diff --git a/net/openvswitch/vport-netdev.c b/net/openvswitch/vport-netdev.c index 12055af832dc..a92ca8b37f96 100644 --- a/net/openvswitch/vport-netdev.c +++ b/net/openvswitch/vport-netdev.c @@ -73,37 +73,21 @@ static struct net_device *get_dpdev(const struct datapath *dp) return local->dev; } -struct vport *ovs_netdev_link(struct vport *vport, const char *name) +struct vport *ovs_netdev_link(struct vport *vport, bool tunnel) { int err; - vport->dev = dev_get_by_name(ovs_dp_get_net(vport->dp), name); - if (!vport->dev) { + if (WARN_ON_ONCE(!vport->dev)) { err = -ENODEV; goto error_free_vport; } - /* Ensure that the device exists and that the provided - * name is not one of its aliases. - */ - if (strcmp(name, ovs_vport_name(vport))) { - err = -ENODEV; - goto error_put; - } - netdev_tracker_alloc(vport->dev, &vport->dev_tracker, GFP_KERNEL); - if (vport->dev->flags & IFF_LOOPBACK || - (vport->dev->type != ARPHRD_ETHER && - vport->dev->type != ARPHRD_NONE) || - ovs_is_internal_dev(vport->dev)) { - err = -EINVAL; - goto error_put; - } rtnl_lock(); err = netdev_master_upper_dev_link(vport->dev, get_dpdev(vport->dp), NULL, NULL, NULL); if (err) - goto error_unlock; + goto error_put_unlock; err = netdev_rx_handler_register(vport->dev, netdev_frame_hook, vport); @@ -119,10 +103,11 @@ struct vport *ovs_netdev_link(struct vport *vport, const char *name) error_master_upper_dev_unlink: netdev_upper_dev_unlink(vport->dev, get_dpdev(vport->dp)); -error_unlock: - rtnl_unlock(); -error_put: +error_put_unlock: + if (tunnel && vport->dev->reg_state == NETREG_REGISTERED) + rtnl_delete_link(vport->dev, 0, NULL); netdev_put(vport->dev, &vport->dev_tracker); + rtnl_unlock(); error_free_vport: ovs_vport_free(vport); return ERR_PTR(err); @@ -132,12 +117,39 @@ EXPORT_SYMBOL_GPL(ovs_netdev_link); static struct vport *netdev_create(const struct vport_parms *parms) { struct vport *vport; + int err; vport = ovs_vport_alloc(0, &ovs_netdev_vport_ops, parms); if (IS_ERR(vport)) return vport; - return ovs_netdev_link(vport, parms->name); + vport->dev = dev_get_by_name(ovs_dp_get_net(vport->dp), parms->name); + if (!vport->dev) { + err = -ENODEV; + goto error_free_vport; + } + netdev_tracker_alloc(vport->dev, &vport->dev_tracker, GFP_KERNEL); + + /* Ensure that the provided name is not an alias. */ + if (strcmp(parms->name, ovs_vport_name(vport))) { + err = -ENODEV; + goto error_put; + } + + if (vport->dev->flags & IFF_LOOPBACK || + (vport->dev->type != ARPHRD_ETHER && + vport->dev->type != ARPHRD_NONE) || + ovs_is_internal_dev(vport->dev)) { + err = -EINVAL; + goto error_put; + } + + return ovs_netdev_link(vport, false); +error_put: + netdev_put(vport->dev, &vport->dev_tracker); +error_free_vport: + ovs_vport_free(vport); + return ERR_PTR(err); } static void vport_netdev_free(struct rcu_head *rcu) diff --git a/net/openvswitch/vport-netdev.h b/net/openvswitch/vport-netdev.h index c5d83a43bfc4..6c0d7366f986 100644 --- a/net/openvswitch/vport-netdev.h +++ b/net/openvswitch/vport-netdev.h @@ -13,7 +13,7 @@ struct vport *ovs_netdev_get_vport(struct net_device *dev); -struct vport *ovs_netdev_link(struct vport *vport, const char *name); +struct vport *ovs_netdev_link(struct vport *vport, bool tunnel); void ovs_netdev_detach_dev(struct vport *); int __init ovs_netdev_init(void); diff --git a/net/openvswitch/vport-vxlan.c b/net/openvswitch/vport-vxlan.c index 0b881b043bcf..c1b37b50d29e 100644 --- a/net/openvswitch/vport-vxlan.c +++ b/net/openvswitch/vport-vxlan.c @@ -126,6 +126,9 @@ static struct vport *vxlan_tnl_create(const struct vport_parms *parms) goto error; } + vport->dev = dev; + netdev_hold(vport->dev, &vport->dev_tracker, GFP_KERNEL); + rtnl_unlock(); return vport; error: @@ -140,7 +143,7 @@ static struct vport *vxlan_create(const struct vport_parms *parms) if (IS_ERR(vport)) return vport; - return ovs_netdev_link(vport, parms->name); + return ovs_netdev_link(vport, true); } static struct vport_ops ovs_vxlan_netdev_vport_ops = { From aa69918bd418e700309fdd08509dba324fb24296 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Fri, 1 May 2026 01:38:37 +0200 Subject: [PATCH 328/525] openvswitch: vport: fix self-deadlock on release of tunnel ports vports are used concurrently and protected by RCU, so netdev_put() must happen after the RCU grace period. So, either in an RCU call or after the synchronize_net(). The rtnl_delete_link() must happen under RTNL and so can't be executed in RCU context. Calling synchronize_net() while holding RTNL is not a good idea for performance and system stability under load in general, so calling netdev_put() in RCU call is the right solution here. However, when the device is deleted, rtnl_unlock() will call netdev_run_todo() and block until all the references are gone. In the current code this means that we never reach the call_rcu() and the vport is never freed and the reference is never released, causing a self-deadlock on device removal. Fix that by moving the rcu_call() before the rtnl_unlock(), so the scheduled RCU callback will be executed when synchronize_net() is called from the rtnl_unlock()->netdev_run_todo() while the RTNL itself is already released. Fixes: 6931d21f87bc ("openvswitch: defer tunnel netdev_put to RCU release") Cc: stable@vger.kernel.org Acked-by: Eelco Chaudron Signed-off-by: Ilya Maximets Acked-by: Aaron Conole Link: https://patch.msgid.link/20260430233848.440994-2-i.maximets@ovn.org Signed-off-by: Paolo Abeni --- net/openvswitch/vport-netdev.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/net/openvswitch/vport-netdev.c b/net/openvswitch/vport-netdev.c index a92ca8b37f96..c42642075685 100644 --- a/net/openvswitch/vport-netdev.c +++ b/net/openvswitch/vport-netdev.c @@ -208,9 +208,13 @@ void ovs_netdev_tunnel_destroy(struct vport *vport) */ if (vport->dev->reg_state == NETREG_REGISTERED) rtnl_delete_link(vport->dev, 0, NULL); - rtnl_unlock(); + /* We can't put the device reference yet, since it can still be in + * use, but rtnl_unlock()->netdev_run_todo() will block until all + * the references are released, so the RCU call must be before it. + */ call_rcu(&vport->rcu, vport_netdev_free); + rtnl_unlock(); } EXPORT_SYMBOL_GPL(ovs_netdev_tunnel_destroy); From 05416ada37aa4efe93f25b0532f551d424fb7b3d Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Fri, 1 May 2026 01:38:38 +0200 Subject: [PATCH 329/525] selftests: openvswitch: add tests for tunnel vport refcounting There were a few issues found with the tunnel vport types around the vport destruction code. Add some basic tests, so at least we know that they can be properly added and removed without obvious issues. The test creates OVS datapath, adds a non-LWT tunnel port, makes sure they are created, and then removes the datapath and waits for all the ports to be gone. The dpctl script had a few bugs in the none-lwt tunnel creation code, so fixing them as well to make the testing possible: - The type of the --lwt option changed in order to properly disable it. - Removed byte order conversion for the port numbers, as the value supposed to be in the host order. - Added missing 'gre' choice for the tunnel type. Signed-off-by: Ilya Maximets Acked-by: Eelco Chaudron Acked-by: Aaron Conole Link: https://patch.msgid.link/20260430233848.440994-3-i.maximets@ovn.org Signed-off-by: Paolo Abeni --- .../selftests/net/openvswitch/openvswitch.sh | 37 +++++++++++++++++++ .../selftests/net/openvswitch/ovs-dpctl.py | 19 +++++++--- 2 files changed, 50 insertions(+), 6 deletions(-) diff --git a/tools/testing/selftests/net/openvswitch/openvswitch.sh b/tools/testing/selftests/net/openvswitch/openvswitch.sh index b327d3061ed5..3cdd953f6813 100755 --- a/tools/testing/selftests/net/openvswitch/openvswitch.sh +++ b/tools/testing/selftests/net/openvswitch/openvswitch.sh @@ -26,6 +26,7 @@ tests=" netlink_checks ovsnl: validate netlink attrs and settings upcall_interfaces ovs: test the upcall interfaces tunnel_metadata ovs: test extraction of tunnel metadata + tunnel_refcount ovs: test tunnel vport reference cleanup drop_reason drop: test drop reasons are emitted psample psample: Sampling packets with psample" @@ -830,6 +831,42 @@ test_tunnel_metadata() { return 0 } +test_tunnel_refcount() { + sbxname="test_tunnel_refcount" + sbx_add "${sbxname}" || return 1 + + ovs_sbx "${sbxname}" ip netns add trefns || return 1 + on_exit "ovs_sbx ${sbxname} ip netns del trefns" + + for tun_type in gre vxlan geneve; do + info "testing ${tun_type} tunnel vport refcount" + + ovs_sbx "${sbxname}" ip netns exec trefns \ + python3 $ovs_base/ovs-dpctl.py \ + add-dp dp-${tun_type} || return 1 + + ovs_sbx "${sbxname}" ip netns exec trefns \ + python3 $ovs_base/ovs-dpctl.py \ + add-if --no-lwt -t ${tun_type} \ + dp-${tun_type} ovs-${tun_type}0 || return 1 + + ovs_wait ip -netns trefns link show \ + ovs-${tun_type}0 >/dev/null 2>&1 || return 1 + + info "deleting dp - may hang if reference counting is broken" + ovs_sbx "${sbxname}" ip netns exec trefns \ + python3 $ovs_base/ovs-dpctl.py \ + del-dp dp-${tun_type} & + + dev_removed() { + ! ip -netns trefns link show "$1" >/dev/null 2>&1 + } + ovs_wait dev_removed dp-${tun_type} || return 1 + ovs_wait dev_removed ovs-${tun_type}0 || return 1 + done + return 0 +} + run_test() { ( tname="$1" diff --git a/tools/testing/selftests/net/openvswitch/ovs-dpctl.py b/tools/testing/selftests/net/openvswitch/ovs-dpctl.py index 848f61fdcee0..bbe35e2718d2 100644 --- a/tools/testing/selftests/net/openvswitch/ovs-dpctl.py +++ b/tools/testing/selftests/net/openvswitch/ovs-dpctl.py @@ -11,7 +11,6 @@ import logging import math import multiprocessing import re -import socket import struct import sys import time @@ -2069,7 +2068,7 @@ class OvsVport(GenericNetlinkSocket): elif vport_type == "internal": return OvsVport.OVS_VPORT_TYPE_INTERNAL elif vport_type == "gre": - return OvsVport.OVS_VPORT_TYPE_INTERNAL + return OvsVport.OVS_VPORT_TYPE_GRE elif vport_type == "vxlan": return OvsVport.OVS_VPORT_TYPE_VXLAN elif vport_type == "geneve": @@ -2121,6 +2120,7 @@ class OvsVport(GenericNetlinkSocket): ) TUNNEL_DEFAULTS = [("geneve", 6081), + ("gre", 0), ("vxlan", 4789)] for tnl in TUNNEL_DEFAULTS: @@ -2129,9 +2129,13 @@ class OvsVport(GenericNetlinkSocket): dport = tnl[1] if not lwt: + if tnl[0] == "gre": + # GRE tunnels have no options. + break + vportopt = OvsVport.ovs_vport_msg.vportopts() vportopt["attrs"].append( - ["OVS_TUNNEL_ATTR_DST_PORT", socket.htons(dport)] + ["OVS_TUNNEL_ATTR_DST_PORT", dport] ) msg["attrs"].append( ["OVS_VPORT_ATTR_OPTIONS", vportopt] @@ -2145,6 +2149,9 @@ class OvsVport(GenericNetlinkSocket): geneve_port=dport, geneve_collect_metadata=True, geneve_udp_zero_csum6_rx=1) + elif tnl[0] == "gre": + ipr.link("add", ifname=vport_ifname, kind="gretap", + gre_collect_metadata=True) elif tnl[0] == "vxlan": ipr.link("add", ifname=vport_ifname, kind=tnl[0], vxlan_learning=0, vxlan_collect_metadata=1, @@ -2563,7 +2570,7 @@ def print_ovsdp_full(dp_lookup_rep, ifindex, ndb=NDB(), vpl=OvsVport()): if vpo: dpo = vpo.get_attr("OVS_TUNNEL_ATTR_DST_PORT") if dpo: - opts += " tnl-dport:%s" % socket.ntohs(dpo) + opts += " tnl-dport:%s" % dpo print( " port %d: %s (%s%s)" % ( @@ -2632,7 +2639,7 @@ def main(argv): "--ptype", type=str, default="netdev", - choices=["netdev", "internal", "geneve", "vxlan"], + choices=["netdev", "internal", "gre", "geneve", "vxlan"], help="Interface type (default netdev)", ) addifcmd.add_argument( @@ -2645,7 +2652,7 @@ def main(argv): addifcmd.add_argument( "-l", "--lwt", - type=bool, + action=argparse.BooleanOptionalAction, default=True, help="Use LWT infrastructure instead of vport (default true)." ) From 44b550d88b267320459d518c0743a241ab2108fa Mon Sep 17 00:00:00 2001 From: Nan Li Date: Fri, 1 May 2026 09:08:44 +0800 Subject: [PATCH 330/525] net/rds: handle zerocopy send cleanup before the message is queued A zerocopy send can fail after user pages have been pinned but before the message is attached to the sending socket. The purge path currently infers zerocopy state from rm->m_rs, so an unqueued message can be cleaned up as if it owned normal payload pages. However, zerocopy ownership is really determined by the presence of op_mmp_znotifier, regardless of whether the message has reached the socket queue. Capture op_mmp_znotifier up front in rds_message_purge() and use it as the cleanup discriminator. If the message is already associated with a socket, keep the existing completion path. Otherwise, drop the pinned page accounting directly and release the notifier before putting the payload pages. This keeps early send failure cleanup consistent with the zerocopy lifetime rules without changing the normal queued completion path. Fixes: 0cebaccef3ac ("rds: zerocopy Tx support.") Cc: stable@kernel.org Reported-by: Yuan Tan Reported-by: Yifan Wu Reported-by: Juefei Pu Reported-by: Xin Liu Co-developed-by: Xiao Liu Signed-off-by: Xiao Liu Signed-off-by: Nan Li Signed-off-by: Ren Wei Reviewed-by: Allison Henderson Link: https://patch.msgid.link/d2ea98a6313d5467bac00f7c9fef8c7acddb9258.1777550074.git.tonanli66@gmail.com Signed-off-by: Paolo Abeni --- net/rds/message.c | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) diff --git a/net/rds/message.c b/net/rds/message.c index eaa6f22601a4..25fedcb3cd00 100644 --- a/net/rds/message.c +++ b/net/rds/message.c @@ -131,24 +131,34 @@ static void rds_rm_zerocopy_callback(struct rds_sock *rs, */ static void rds_message_purge(struct rds_message *rm) { + struct rds_znotifier *znotifier; unsigned long i, flags; - bool zcopy = false; + bool zcopy; if (unlikely(test_bit(RDS_MSG_PAGEVEC, &rm->m_flags))) return; spin_lock_irqsave(&rm->m_rs_lock, flags); + znotifier = rm->data.op_mmp_znotifier; + rm->data.op_mmp_znotifier = NULL; + zcopy = !!znotifier; + if (rm->m_rs) { struct rds_sock *rs = rm->m_rs; - if (rm->data.op_mmp_znotifier) { - zcopy = true; - rds_rm_zerocopy_callback(rs, rm->data.op_mmp_znotifier); + if (znotifier) { + rds_rm_zerocopy_callback(rs, znotifier); rds_wake_sk_sleep(rs); - rm->data.op_mmp_znotifier = NULL; } sock_put(rds_rs_to_sk(rs)); rm->m_rs = NULL; + } else if (znotifier) { + /* + * Zerocopy can fail before the message is queued on the + * socket, so there is no rs to carry the notification. + */ + mm_unaccount_pinned_pages(&znotifier->z_mmp); + kfree(rds_info_from_znotifier(znotifier)); } spin_unlock_irqrestore(&rm->m_rs_lock, flags); From 95084f1883a760e0d4290698346759d58e2b944a Mon Sep 17 00:00:00 2001 From: Dipayaan Roy Date: Thu, 30 Apr 2026 19:47:12 -0700 Subject: [PATCH 331/525] net: mana: Fix crash from unvalidated SHM offset read from BAR0 during FLR During Function Level Reset recovery, the MANA driver reads hardware BAR0 registers that may temporarily contain garbage values. The SHM (Shared Memory) offset read from GDMA_REG_SHM_OFFSET is used to compute gc->shm_base, which is later dereferenced via readl() in mana_smc_poll_register(). If the hardware returns an unaligned or out-of-range value, the driver must not blindly use it, as this would propagate the hardware error into a kernel crash. The following crash was observed on an arm64 Hyper-V guest running kernel 6.17.0-3013-azure during VF reset recovery triggered by HWC timeout. [13291.785274] Unable to handle kernel paging request at virtual address ffff8000a200001b [13291.785311] Mem abort info: [13291.785332] ESR = 0x0000000096000021 [13291.785343] EC = 0x25: DABT (current EL), IL = 32 bits [13291.785355] SET = 0, FnV = 0 [13291.785363] EA = 0, S1PTW = 0 [13291.785372] FSC = 0x21: alignment fault [13291.785382] Data abort info: [13291.785391] ISV = 0, ISS = 0x00000021, ISS2 = 0x00000000 [13291.785404] CM = 0, WnR = 0, TnD = 0, TagAccess = 0 [13291.785412] GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0 [13291.785421] swapper pgtable: 4k pages, 48-bit VAs, pgdp=00000014df3a1000 [13291.785432] [ffff8000a200001b] pgd=1000000100438403, p4d=1000000100438403, pud=1000000100439403, pmd=0068000fc2000711 [13291.785703] Internal error: Oops: 0000000096000021 [#1] SMP [13291.830975] Modules linked in: tls qrtr mana_ib ib_uverbs ib_core xt_owner xt_tcpudp xt_conntrack nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 nft_compat nf_tables cfg80211 8021q garp mrp stp llc binfmt_misc joydev serio_raw nls_iso8859_1 hid_generic aes_ce_blk aes_ce_cipher polyval_ce ghash_ce sm4_ce_gcm sm4_ce_ccm sm4_ce sm4_ce_cipher hid_hyperv sm4 sm3_ce sha3_ce hv_netvsc hid vmgenid hyperv_keyboard hyperv_drm sch_fq_codel nvme_fabrics efi_pstore dm_multipath nfnetlink vsock_loopback vmw_vsock_virtio_transport_common hv_sock vmw_vsock_vmci_transport vmw_vmci vsock dmi_sysfs ip_tables x_tables autofs4 [13291.862630] CPU: 122 UID: 0 PID: 61796 Comm: kworker/122:2 Tainted: G W 6.17.0-3013-azure #13-Ubuntu VOLUNTARY [13291.869902] Tainted: [W]=WARN [13291.871901] Hardware name: Microsoft Corporation Virtual Machine/Virtual Machine, BIOS Hyper-V UEFI Release v4.1 01/08/2026 [13291.878086] Workqueue: events mana_serv_func [13291.880718] pstate: 62400005 (nZCv daif +PAN -UAO +TCO -DIT -SSBS BTYPE=--) [13291.884835] pc : mana_smc_poll_register+0x48/0xb0 [13291.887902] lr : mana_smc_setup_hwc+0x70/0x1c0 [13291.890493] sp : ffff8000ab79bbb0 [13291.892364] x29: ffff8000ab79bbb0 x28: ffff00410c8b5900 x27: ffff00410d630680 [13291.896252] x26: ffff004171f9fd80 x25: 000000016ed55000 x24: 000000017f37e000 [13291.899990] x23: 0000000000000000 x22: 000000016ed55000 x21: 0000000000000000 [13291.904497] x20: ffff8000a200001b x19: 0000000000004e20 x18: ffff8000a6183050 [13291.908308] x17: 0000000000000000 x16: 0000000000000000 x15: 000000000000000a [13291.912542] x14: 0000000000000004 x13: 0000000000000000 x12: 0000000000000000 [13291.916298] x11: 0000000000000000 x10: 0000000000000001 x9 : ffffc45006af1bd8 [13291.920945] x8 : ffff000151129000 x7 : 0000000000000000 x6 : 0000000000000000 [13291.925293] x5 : 000000015f214000 x4 : 000000017217a000 x3 : 000000016ed50000 [13291.930436] x2 : 000000016ed55000 x1 : 0000000000000000 x0 : ffff8000a1ffffff [13291.934342] Call trace: [13291.935736] mana_smc_poll_register+0x48/0xb0 (P) [13291.938611] mana_smc_setup_hwc+0x70/0x1c0 [13291.941113] mana_hwc_create_channel+0x1a0/0x3a0 [13291.944283] mana_gd_setup+0x16c/0x398 [13291.946584] mana_gd_resume+0x24/0x70 [13291.948917] mana_do_service+0x13c/0x1d0 [13291.951583] mana_serv_func+0x34/0x68 [13291.953732] process_one_work+0x168/0x3d0 [13291.956745] worker_thread+0x2ac/0x480 [13291.959104] kthread+0xf8/0x110 [13291.961026] ret_from_fork+0x10/0x20 [13291.963560] Code: d2807d00 9417c551 71000673 54000220 (b9400281) [13291.967299] ---[ end trace 0000000000000000 ]--- Disassembly of mana_smc_poll_register() around the crash site: Disassembly of section .text: 00000000000047c8 : 47c8: d503201f nop 47cc: d503201f nop 47d0: d503233f paciasp 47d4: f800865e str x30, [x18], #8 47d8: a9bd7bfd stp x29, x30, [sp, #-48]! 47dc: 910003fd mov x29, sp 47e0: a90153f3 stp x19, x20, [sp, #16] 47e4: 91007014 add x20, x0, #0x1c 47e8: 5289c413 mov w19, #0x4e20 47ec: f90013f5 str x21, [sp, #32] 47f0: 12001c35 and w21, w1, #0xff 47f4: 14000008 b 4814 47f8: 36f801e1 tbz w1, #31, 4834 47fc: 52800042 mov w2, #0x2 4800: d280fa01 mov x1, #0x7d0 4804: d2807d00 mov x0, #0x3e8 4808: 94000000 bl 0 480c: 71000673 subs w19, w19, #0x1 4810: 54000200 b.eq 4850 4814: b9400281 ldr w1, [x20] <-- **** CRASHED HERE ***** 4818: d50331bf dmb oshld 481c: 2a0103e2 mov w2, w1 ... From the crash signature x20 = ffff8000a200001b, this address ends in 0x1b which is not 4-byte aligned, so the 'ldr w1, [x20]' instruction (readl) triggers the arm64 alignment fault (FSC = 0x21). The root cause is in mana_gd_init_vf_regs(), which computes: gc->shm_base = gc->bar0_va + mana_gd_r64(gc, GDMA_REG_SHM_OFFSET); The offset is used without any validation. The same problem exists in mana_gd_init_pf_regs() for sriov_base_off and sriov_shm_off. Fix this by validating all offsets before use: - VF: check shm_off is within BAR0, properly aligned to 4 bytes (readl requirement), and leaves room for the full 256-bit (32-byte) SMC aperture. - PF: check sriov_base_off is within BAR0, aligned to 8 bytes (readq requirement), and leaves room to safely read the sriov_shm_off register at sriov_base_off + GDMA_PF_REG_SHM_OFF. Then check sriov_shm_off leaves room for the full SMC aperture. All arithmetic uses subtraction rather than addition to avoid integer overflow on garbage values. Define SMC_APERTURE_SIZE (32 bytes, derived from the 256-bit aperture width) Return -EPROTO on invalid values. The existing recovery path in mana_serv_reset() already handles -EPROTO by falling through to PCI device rescan, giving the hardware another chance to present valid register values after reset. Fixes: 9bf66036d686 ("net: mana: Handle hardware recovery events when probing the device") Signed-off-by: Dipayaan Roy Link: https://patch.msgid.link/afQUMClyjmBVfD+u@linuxonhyperv3.guj3yctzbm1etfxqx2vob5hsef.xx.internal.cloudapp.net Signed-off-by: Paolo Abeni --- .../net/ethernet/microsoft/mana/gdma_main.c | 40 ++++++++++++++++--- .../net/ethernet/microsoft/mana/shm_channel.c | 5 --- include/net/mana/shm_channel.h | 6 +++ 3 files changed, 41 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/microsoft/mana/gdma_main.c b/drivers/net/ethernet/microsoft/mana/gdma_main.c index 098fbda0d128..d8e816882f02 100644 --- a/drivers/net/ethernet/microsoft/mana/gdma_main.c +++ b/drivers/net/ethernet/microsoft/mana/gdma_main.c @@ -43,8 +43,9 @@ static u64 mana_gd_r64(struct gdma_context *g, u64 offset) static int mana_gd_init_pf_regs(struct pci_dev *pdev) { struct gdma_context *gc = pci_get_drvdata(pdev); - void __iomem *sriov_base_va; + u64 remaining_barsize; u64 sriov_base_off; + u64 sriov_shm_off; gc->db_page_size = mana_gd_r32(gc, GDMA_PF_REG_DB_PAGE_SIZE) & 0xFFFF; @@ -73,10 +74,28 @@ static int mana_gd_init_pf_regs(struct pci_dev *pdev) gc->phys_db_page_base = gc->bar0_pa + gc->db_page_off; sriov_base_off = mana_gd_r64(gc, GDMA_SRIOV_REG_CFG_BASE_OFF); + if (sriov_base_off >= gc->bar0_size || + gc->bar0_size - sriov_base_off < + GDMA_PF_REG_SHM_OFF + sizeof(u64) || + !IS_ALIGNED(sriov_base_off, sizeof(u64))) { + dev_err(gc->dev, + "SRIOV base offset 0x%llx out of range or unaligned (BAR0 size 0x%llx)\n", + sriov_base_off, (u64)gc->bar0_size); + return -EPROTO; + } - sriov_base_va = gc->bar0_va + sriov_base_off; - gc->shm_base = sriov_base_va + - mana_gd_r64(gc, sriov_base_off + GDMA_PF_REG_SHM_OFF); + remaining_barsize = gc->bar0_size - sriov_base_off; + sriov_shm_off = mana_gd_r64(gc, sriov_base_off + GDMA_PF_REG_SHM_OFF); + if (sriov_shm_off >= remaining_barsize || + remaining_barsize - sriov_shm_off < SMC_APERTURE_SIZE || + !IS_ALIGNED(sriov_shm_off, sizeof(u32))) { + dev_err(gc->dev, + "SRIOV SHM offset 0x%llx out of range or unaligned (BAR0 size 0x%llx)\n", + sriov_shm_off, (u64)gc->bar0_size); + return -EPROTO; + } + + gc->shm_base = gc->bar0_va + sriov_base_off + sriov_shm_off; return 0; } @@ -84,6 +103,7 @@ static int mana_gd_init_pf_regs(struct pci_dev *pdev) static int mana_gd_init_vf_regs(struct pci_dev *pdev) { struct gdma_context *gc = pci_get_drvdata(pdev); + u64 shm_off; gc->db_page_size = mana_gd_r32(gc, GDMA_REG_DB_PAGE_SIZE) & 0xFFFF; @@ -111,7 +131,17 @@ static int mana_gd_init_vf_regs(struct pci_dev *pdev) gc->db_page_base = gc->bar0_va + gc->db_page_off; gc->phys_db_page_base = gc->bar0_pa + gc->db_page_off; - gc->shm_base = gc->bar0_va + mana_gd_r64(gc, GDMA_REG_SHM_OFFSET); + shm_off = mana_gd_r64(gc, GDMA_REG_SHM_OFFSET); + if (shm_off >= gc->bar0_size || + gc->bar0_size - shm_off < SMC_APERTURE_SIZE || + !IS_ALIGNED(shm_off, sizeof(u32))) { + dev_err(gc->dev, + "SHM offset 0x%llx out of range or unaligned (BAR0 size 0x%llx)\n", + shm_off, (u64)gc->bar0_size); + return -EPROTO; + } + + gc->shm_base = gc->bar0_va + shm_off; return 0; } diff --git a/drivers/net/ethernet/microsoft/mana/shm_channel.c b/drivers/net/ethernet/microsoft/mana/shm_channel.c index 0f1679ebad96..d21b5db06e50 100644 --- a/drivers/net/ethernet/microsoft/mana/shm_channel.c +++ b/drivers/net/ethernet/microsoft/mana/shm_channel.c @@ -61,11 +61,6 @@ union smc_proto_hdr { }; }; /* HW DATA */ -#define SMC_APERTURE_BITS 256 -#define SMC_BASIC_UNIT (sizeof(u32)) -#define SMC_APERTURE_DWORDS (SMC_APERTURE_BITS / (SMC_BASIC_UNIT * 8)) -#define SMC_LAST_DWORD (SMC_APERTURE_DWORDS - 1) - static int mana_smc_poll_register(void __iomem *base, bool reset) { void __iomem *ptr = base + SMC_LAST_DWORD * SMC_BASIC_UNIT; diff --git a/include/net/mana/shm_channel.h b/include/net/mana/shm_channel.h index 5199b41497ff..dbabcfb95daf 100644 --- a/include/net/mana/shm_channel.h +++ b/include/net/mana/shm_channel.h @@ -4,6 +4,12 @@ #ifndef _SHM_CHANNEL_H #define _SHM_CHANNEL_H +#define SMC_APERTURE_BITS 256 +#define SMC_BASIC_UNIT (sizeof(u32)) +#define SMC_APERTURE_DWORDS (SMC_APERTURE_BITS / (SMC_BASIC_UNIT * 8)) +#define SMC_LAST_DWORD (SMC_APERTURE_DWORDS - 1) +#define SMC_APERTURE_SIZE (SMC_APERTURE_BITS / 8) + struct shm_channel { struct device *dev; void __iomem *base; From b9eac6a9d93c952c4b7775a24d5c7a1bbf4c3c00 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 25 Apr 2026 00:47:54 +0200 Subject: [PATCH 332/525] rseq: Revert to historical performance killing behaviour The recent RSEQ optimization work broke the TCMalloc abuse of the RSEQ ABI as it not longer unconditionally updates the CPU, node, mm_cid fields, which are documented as read only for user space. Due to the observed behavior of the kernel it was possible for TCMalloc to overwrite the cpu_id_start field for their own purposes and rely on the kernel to update it unconditionally after each context switch and before signal delivery. The RSEQ ABI only guarantees that these fields are updated when the data changes, i.e. the task is migrated or the MMCID of the task changes due to switching from or to per CPU ownership mode. The optimization work eliminated the unconditional updates and reduced them to the documented ABI guarantees, which results in a massive performance win for syscall, scheduling heavy work loads, which in turn breaks the TCMalloc expectations. There have been several options discussed to restore the TCMalloc functionality while preserving the optimization benefits. They all end up in a series of hard to maintain workarounds, which in the worst case introduce overhead for everyone, e.g. in the scheduler. The requirements of TCMalloc and the optimization work are diametral and the required work arounds are a maintainence burden. They end up as fragile constructs, which are blocking further optimization work and are pretty much guaranteed to cause more subtle issues down the road. The optimization work heavily depends on the generic entry code, which is not used by all architectures yet. So the rework preserved the original mechanism moslty unmodified to keep the support for architectures, which handle rseq in their own exit to user space loop. That code is currently optimized out by the compiler on architectures which use the generic entry code. This allows to revert back to the original behaviour by replacing the compile time constant conditions with a runtime condition where required, which disables the optimization and the dependend time slice extension feature until the run-time condition can be enabled in the RSEQ registration code on a per task basis again. The following changes are required to restore the original behavior, which makes TCMalloc work again: 1) Replace the compile time constant conditionals with runtime conditionals where appropriate to prevent the compiler from optimizing the legacy mode out 2) Enforce unconditional update of IDs on context switch for the non-optimized v1 mode 3) Enforce update of IDs in the pre signal delivery path for the non-optimized v1 mode 4) Enforce update of IDs in the membarrier(RSEQ) IPI for the non-optimized v1 mode 5) Make time slice and future extensions depend on optimized v2 mode This brings back the full performance problems, but preserves the v2 optimization code and for generic entry code using architectures also the TIF_RSEQ optimization which avoids a full evaluation of the exit to user mode loop in many cases. Fixes: 566d8015f7ee ("rseq: Avoid CPU/MM CID updates when no event pending") Reported-by: Mathias Stearn Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Closes: https://lore.kernel.org/CAHnCjA25b+nO2n5CeifknSKHssJpPrjnf+dtr7UgzRw4Zgu=oA@mail.gmail.com Link: https://patch.msgid.link/20260428224427.517051752%40kernel.org Cc: stable@vger.kernel.org --- include/linux/rseq.h | 35 ++++++++++++++++++++++----------- include/linux/rseq_entry.h | 39 +++++++++++++++++++++++++++---------- include/linux/rseq_types.h | 9 ++++++++- kernel/rseq.c | 40 +++++++++++++++++++++++++++++++------- kernel/sched/membarrier.c | 11 ++++++++++- 5 files changed, 104 insertions(+), 30 deletions(-) diff --git a/include/linux/rseq.h b/include/linux/rseq.h index f446909551df..7ef79b25e714 100644 --- a/include/linux/rseq.h +++ b/include/linux/rseq.h @@ -9,6 +9,11 @@ void __rseq_handle_slowpath(struct pt_regs *regs); +static __always_inline bool rseq_v2(struct task_struct *t) +{ + return IS_ENABLED(CONFIG_GENERIC_IRQ_ENTRY) && likely(t->rseq.event.has_rseq > 1); +} + /* Invoked from resume_user_mode_work() */ static inline void rseq_handle_slowpath(struct pt_regs *regs) { @@ -16,8 +21,7 @@ static inline void rseq_handle_slowpath(struct pt_regs *regs) if (current->rseq.event.slowpath) __rseq_handle_slowpath(regs); } else { - /* '&' is intentional to spare one conditional branch */ - if (current->rseq.event.sched_switch & current->rseq.event.has_rseq) + if (current->rseq.event.sched_switch && current->rseq.event.has_rseq) __rseq_handle_slowpath(regs); } } @@ -30,9 +34,9 @@ void __rseq_signal_deliver(int sig, struct pt_regs *regs); */ static inline void rseq_signal_deliver(struct ksignal *ksig, struct pt_regs *regs) { - if (IS_ENABLED(CONFIG_GENERIC_IRQ_ENTRY)) { - /* '&' is intentional to spare one conditional branch */ - if (current->rseq.event.has_rseq & current->rseq.event.user_irq) + if (rseq_v2(current)) { + /* has_rseq is implied in rseq_v2() */ + if (current->rseq.event.user_irq) __rseq_signal_deliver(ksig->sig, regs); } else { if (current->rseq.event.has_rseq) @@ -50,15 +54,22 @@ static __always_inline void rseq_sched_switch_event(struct task_struct *t) { struct rseq_event *ev = &t->rseq.event; - if (IS_ENABLED(CONFIG_GENERIC_IRQ_ENTRY)) { + /* + * Only apply the user_irq optimization for RSEQ ABI V2 registrations. + * Legacy users like TCMalloc rely on the original ABI V1 behaviour + * which updates IDs on every context swtich. + */ + if (rseq_v2(t)) { /* - * Avoid a boat load of conditionals by using simple logic - * to determine whether NOTIFY_RESUME needs to be raised. + * Avoid a boat load of conditionals by using simple logic to + * determine whether TIF_NOTIFY_RESUME or TIF_RSEQ needs to be + * raised. * - * It's required when the CPU or MM CID has changed or - * the entry was from user space. + * It's required when the CPU or MM CID has changed or the entry + * was via interrupt from user space. ev->has_rseq does not have + * to be evaluated here because rseq_v2() implies has_rseq. */ - bool raise = (ev->user_irq | ev->ids_changed) & ev->has_rseq; + bool raise = ev->user_irq | ev->ids_changed; if (raise) { ev->sched_switch = true; @@ -66,6 +77,7 @@ static __always_inline void rseq_sched_switch_event(struct task_struct *t) } } else { if (ev->has_rseq) { + t->rseq.event.ids_changed = true; t->rseq.event.sched_switch = true; rseq_raise_notify_resume(t); } @@ -161,6 +173,7 @@ static inline unsigned int rseq_alloc_align(void) } #else /* CONFIG_RSEQ */ +static inline bool rseq_v2(struct task_struct *t) { return false; } static inline void rseq_handle_slowpath(struct pt_regs *regs) { } static inline void rseq_signal_deliver(struct ksignal *ksig, struct pt_regs *regs) { } static inline void rseq_sched_switch_event(struct task_struct *t) { } diff --git a/include/linux/rseq_entry.h b/include/linux/rseq_entry.h index f11ebd34f8b9..934db41ec782 100644 --- a/include/linux/rseq_entry.h +++ b/include/linux/rseq_entry.h @@ -111,6 +111,20 @@ static __always_inline void rseq_slice_clear_grant(struct task_struct *t) t->rseq.slice.state.granted = false; } +/* + * Open coded, so it can be invoked within a user access region. + * + * This clears the user space state of the time slice extensions field only when + * the task has registered the optimized RSEQ_ABI V2. Some legacy registrations, + * e.g. TCMalloc, have conflicting non-ABI fields in struct RSEQ, which would be + * overwritten by an unconditional write. + */ +#define rseq_slice_clear_user(rseq, efault) \ +do { \ + if (rseq_slice_extension_enabled()) \ + unsafe_put_user(0U, &rseq->slice_ctrl.all, efault); \ +} while (0) + static __always_inline bool __rseq_grant_slice_extension(bool work_pending) { struct task_struct *curr = current; @@ -230,6 +244,7 @@ static __always_inline bool rseq_slice_extension_enabled(void) { return false; } static __always_inline bool rseq_arm_slice_extension_timer(void) { return false; } static __always_inline void rseq_slice_clear_grant(struct task_struct *t) { } static __always_inline bool rseq_grant_slice_extension(unsigned long ti_work, unsigned long mask) { return false; } +#define rseq_slice_clear_user(rseq, efault) do { } while (0) #endif /* !CONFIG_RSEQ_SLICE_EXTENSION */ bool rseq_debug_update_user_cs(struct task_struct *t, struct pt_regs *regs, unsigned long csaddr); @@ -517,11 +532,9 @@ bool rseq_set_ids_get_csaddr(struct task_struct *t, struct rseq_ids *ids, if (csaddr) unsafe_get_user(*csaddr, &rseq->rseq_cs, efault); - /* Open coded, so it's in the same user access region */ - if (rseq_slice_extension_enabled()) { - /* Unconditionally clear it, no point in conditionals */ - unsafe_put_user(0U, &rseq->slice_ctrl.all, efault); - } + /* RSEQ ABI V2 only operations */ + if (rseq_v2(t)) + rseq_slice_clear_user(rseq, efault); } rseq_slice_clear_grant(t); @@ -612,6 +625,14 @@ static __always_inline bool rseq_exit_user_update(struct pt_regs *regs, struct t * interrupts disabled */ guard(pagefault)(); + /* + * This optimization is only valid when the task registered for the + * optimized RSEQ_ABI_V2 variant. Some legacy users rely on the original + * RSEQ implementation behaviour which unconditionally updated the IDs. + * rseq_sched_switch_event() ensures that legacy registrations always + * have both sched_switch and ids_changed set, which is compatible with + * the historical TIF_NOTIFY_RESUME behaviour. + */ if (likely(!t->rseq.event.ids_changed)) { struct rseq __user *rseq = t->rseq.usrptr; /* @@ -623,11 +644,9 @@ static __always_inline bool rseq_exit_user_update(struct pt_regs *regs, struct t scoped_user_rw_access(rseq, efault) { unsafe_get_user(csaddr, &rseq->rseq_cs, efault); - /* Open coded, so it's in the same user access region */ - if (rseq_slice_extension_enabled()) { - /* Unconditionally clear it, no point in conditionals */ - unsafe_put_user(0U, &rseq->slice_ctrl.all, efault); - } + /* RSEQ ABI V2 only operations */ + if (rseq_v2(t)) + rseq_slice_clear_user(rseq, efault); } rseq_slice_clear_grant(t); diff --git a/include/linux/rseq_types.h b/include/linux/rseq_types.h index 0b42045988db..a469c1870849 100644 --- a/include/linux/rseq_types.h +++ b/include/linux/rseq_types.h @@ -9,6 +9,12 @@ #ifdef CONFIG_RSEQ struct rseq; +/* + * rseq_event::has_rseq contains the ABI version number so preserving it + * in AND operations requires a mask. + */ +#define RSEQ_HAS_RSEQ_VERSION_MASK 0xff + /** * struct rseq_event - Storage for rseq related event management * @all: Compound to initialize and clear the data efficiently @@ -17,7 +23,8 @@ struct rseq; * exit to user * @ids_changed: Indicator that IDs need to be updated * @user_irq: True on interrupt entry from user mode - * @has_rseq: True if the task has a rseq pointer installed + * @has_rseq: Greater than 0 if the task has a rseq pointer installed. + * Contains the RSEQ version number * @error: Compound error code for the slow path to analyze * @fatal: User space data corrupted or invalid * @slowpath: Indicator that slow path processing via TIF_NOTIFY_RESUME diff --git a/kernel/rseq.c b/kernel/rseq.c index 586f58f652c6..aa25753ea135 100644 --- a/kernel/rseq.c +++ b/kernel/rseq.c @@ -253,11 +253,14 @@ static bool rseq_handle_cs(struct task_struct *t, struct pt_regs *regs) static void rseq_slowpath_update_usr(struct pt_regs *regs) { /* - * Preserve rseq state and user_irq state. The generic entry code - * clears user_irq on the way out, the non-generic entry - * architectures are not having user_irq. + * Preserve has_rseq and user_irq state. The generic entry code clears + * user_irq on the way out, the non-generic entry architectures are not + * setting user_irq. */ - const struct rseq_event evt_mask = { .has_rseq = true, .user_irq = true, }; + const struct rseq_event evt_mask = { + .has_rseq = RSEQ_HAS_RSEQ_VERSION_MASK, + .user_irq = true, + }; struct task_struct *t = current; struct rseq_ids ids; u32 node_id; @@ -330,8 +333,9 @@ void __rseq_handle_slowpath(struct pt_regs *regs) void __rseq_signal_deliver(int sig, struct pt_regs *regs) { rseq_stat_inc(rseq_stats.signal); + /* - * Don't update IDs, they are handled on exit to user if + * Don't update IDs yet, they are handled on exit to user if * necessary. The important thing is to abort a critical section of * the interrupted context as after this point the instruction * pointer in @regs points to the signal handler. @@ -344,6 +348,13 @@ void __rseq_signal_deliver(int sig, struct pt_regs *regs) current->rseq.event.error = 0; force_sigsegv(sig); } + + /* + * In legacy mode, force the update of IDs before returning to user + * space to stay compatible. + */ + if (!rseq_v2(current)) + rseq_force_update(); } /* @@ -408,6 +419,7 @@ static bool rseq_reset_ids(void) SYSCALL_DEFINE4(rseq, struct rseq __user *, rseq, u32, rseq_len, int, flags, u32, sig) { u32 rseqfl = 0; + u8 version = 1; if (flags & RSEQ_FLAG_UNREGISTER) { if (flags & ~RSEQ_FLAG_UNREGISTER) @@ -461,7 +473,11 @@ SYSCALL_DEFINE4(rseq, struct rseq __user *, rseq, u32, rseq_len, int, flags, u32 if (!access_ok(rseq, rseq_len)) return -EFAULT; - if (IS_ENABLED(CONFIG_RSEQ_SLICE_EXTENSION)) { + /* + * The version check effectivly disables time slice extensions until the + * RSEQ ABI V2 registration are implemented. + */ + if (IS_ENABLED(CONFIG_RSEQ_SLICE_EXTENSION) && version > 1) { if (rseq_slice_extension_enabled()) { rseqfl |= RSEQ_CS_FLAG_SLICE_EXT_AVAILABLE; if (flags & RSEQ_FLAG_SLICE_EXT_DEFAULT_ON) @@ -484,7 +500,15 @@ SYSCALL_DEFINE4(rseq, struct rseq __user *, rseq, u32, rseq_len, int, flags, u32 unsafe_put_user(RSEQ_CPU_ID_UNINITIALIZED, &rseq->cpu_id, efault); unsafe_put_user(0U, &rseq->node_id, efault); unsafe_put_user(0U, &rseq->mm_cid, efault); - unsafe_put_user(0U, &rseq->slice_ctrl.all, efault); + + /* + * All fields past mm_cid are only valid for non-legacy v2 + * registrations. + */ + if (version > 1) { + if (IS_ENABLED(CONFIG_RSEQ_SLICE_EXTENSION)) + unsafe_put_user(0U, &rseq->slice_ctrl.all, efault); + } } /* @@ -712,6 +736,8 @@ int rseq_slice_extension_prctl(unsigned long arg2, unsigned long arg3) return -ENOTSUPP; if (!current->rseq.usrptr) return -ENXIO; + if (!rseq_v2(current)) + return -ENOTSUPP; /* No change? */ if (enable == !!current->rseq.slice.state.enabled) diff --git a/kernel/sched/membarrier.c b/kernel/sched/membarrier.c index 623445603725..226a6329f3e9 100644 --- a/kernel/sched/membarrier.c +++ b/kernel/sched/membarrier.c @@ -199,7 +199,16 @@ static void ipi_rseq(void *info) * is negligible. */ smp_mb(); - rseq_sched_switch_event(current); + /* + * Legacy mode requires that IDs are written and the critical section is + * evaluated. V2 optimized mode handles the critical section and IDs are + * only updated if they change as a consequence of preemption after + * return from this IPI. + */ + if (rseq_v2(current)) + rseq_sched_switch_event(current); + else + rseq_force_update(); } static void ipi_sync_rq_state(void *info) From 02b44d943b3adddc3a15c1da97045e205b7d14c1 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 25 Apr 2026 15:46:06 +0200 Subject: [PATCH 333/525] selftests/rseq: Skip tests if time slice extensions are not available Don't fail, skip the test if the extensions are not enabled at compile or runtime. Fixes: 830969e7821a ("selftests/rseq: Implement time slice extension test") Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Link: https://patch.msgid.link/20260428224427.597838491%40kernel.org Cc: stable@vger.kernel.org --- tools/testing/selftests/rseq/slice_test.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/tools/testing/selftests/rseq/slice_test.c b/tools/testing/selftests/rseq/slice_test.c index 357122dcb487..77e668ff74d7 100644 --- a/tools/testing/selftests/rseq/slice_test.c +++ b/tools/testing/selftests/rseq/slice_test.c @@ -124,6 +124,13 @@ FIXTURE_SETUP(slice_ext) { cpu_set_t affinity; + if (rseq_register_current_thread()) + SKIP(return, "RSEQ not supported\n"); + + if (prctl(PR_RSEQ_SLICE_EXTENSION, PR_RSEQ_SLICE_EXTENSION_SET, + PR_RSEQ_SLICE_EXT_ENABLE, 0, 0)) + SKIP(return, "Time slice extension not supported\n"); + ASSERT_EQ(sched_getaffinity(0, sizeof(affinity), &affinity), 0); /* Pin it on a single CPU. Avoid CPU 0 */ @@ -137,11 +144,6 @@ FIXTURE_SETUP(slice_ext) break; } - ASSERT_EQ(rseq_register_current_thread(), 0); - - ASSERT_EQ(prctl(PR_RSEQ_SLICE_EXTENSION, PR_RSEQ_SLICE_EXTENSION_SET, - PR_RSEQ_SLICE_EXT_ENABLE, 0, 0), 0); - self->noise_params.noise_nsecs = variant->noise_nsecs; self->noise_params.sleep_nsecs = variant->sleep_nsecs; self->noise_params.run = 1; From d97cb2ef0b221b068e90b6058aa97faa0626bdab Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sun, 26 Apr 2026 18:13:54 +0200 Subject: [PATCH 334/525] selftests/rseq: Make registration flexible for legacy and optimized mode rseq_register_current_thread() either uses the glibc registered RSEQ region or registers it's own region with the legacy size of 32 bytes. That worked so far, but becomes a problem when the kernel implements a distinction between legacy and performance optimized behavior based on the registration size as that does not allow to test both modes with the self test suite. Add two arguments to the function. One to enforce that the registration is not using libc provided mode and one to tell the registration to use the legacy size and not the kernel advertised size. Rename it and make the original one a inline wrapper which preserves the existing behavior. Fixes: 566d8015f7ee ("rseq: Avoid CPU/MM CID updates when no event pending") Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Link: https://patch.msgid.link/20260428224427.677889423%40kernel.org Cc: stable@vger.kernel.org --- tools/testing/selftests/rseq/rseq-abi.h | 7 ++++- tools/testing/selftests/rseq/rseq.c | 39 ++++++++++++------------- tools/testing/selftests/rseq/rseq.h | 8 ++++- 3 files changed, 31 insertions(+), 23 deletions(-) diff --git a/tools/testing/selftests/rseq/rseq-abi.h b/tools/testing/selftests/rseq/rseq-abi.h index ecef315204b2..5f4ea2152c2f 100644 --- a/tools/testing/selftests/rseq/rseq-abi.h +++ b/tools/testing/selftests/rseq/rseq-abi.h @@ -191,10 +191,15 @@ struct rseq_abi { */ struct rseq_abi_slice_ctrl slice_ctrl; + /* + * Place holder to push the size above 32 bytes. + */ + __u8 __reserved; + /* * Flexible array member at end of structure, after last feature field. */ char end[]; -} __attribute__((aligned(4 * sizeof(__u64)))); +} __attribute__((aligned(256))); #endif /* _RSEQ_ABI_H */ diff --git a/tools/testing/selftests/rseq/rseq.c b/tools/testing/selftests/rseq/rseq.c index a736727b83c1..be0d0a97031e 100644 --- a/tools/testing/selftests/rseq/rseq.c +++ b/tools/testing/selftests/rseq/rseq.c @@ -56,6 +56,7 @@ ptrdiff_t rseq_offset; * unsuccessful. */ unsigned int rseq_size = -1U; +static unsigned int rseq_alloc_size; /* Flags used during rseq registration. */ unsigned int rseq_flags; @@ -115,29 +116,17 @@ bool rseq_available(void) } } -/* The rseq areas need to be at least 32 bytes. */ -static -unsigned int get_rseq_min_alloc_size(void) -{ - unsigned int alloc_size = rseq_size; - - if (alloc_size < ORIG_RSEQ_ALLOC_SIZE) - alloc_size = ORIG_RSEQ_ALLOC_SIZE; - return alloc_size; -} - /* * Return the feature size supported by the kernel. * * Depending on the value returned by getauxval(AT_RSEQ_FEATURE_SIZE): * - * 0: Return ORIG_RSEQ_FEATURE_SIZE (20) + * 0: Return ORIG_RSEQ_FEATURE_SIZE (20) * > 0: Return the value from getauxval(AT_RSEQ_FEATURE_SIZE). * * It should never return a value below ORIG_RSEQ_FEATURE_SIZE. */ -static -unsigned int get_rseq_kernel_feature_size(void) +static unsigned int get_rseq_kernel_feature_size(void) { unsigned long auxv_rseq_feature_size, auxv_rseq_align; @@ -152,15 +141,24 @@ unsigned int get_rseq_kernel_feature_size(void) return ORIG_RSEQ_FEATURE_SIZE; } -int rseq_register_current_thread(void) +int __rseq_register_current_thread(bool nolibc, bool legacy) { + unsigned int size; int rc; if (!rseq_ownership) { /* Treat libc's ownership as a successful registration. */ - return 0; + return nolibc ? -EBUSY : 0; } - rc = sys_rseq(&__rseq.abi, get_rseq_min_alloc_size(), 0, RSEQ_SIG); + + /* The minimal allocation size is 32, which is the legacy allocation size */ + size = get_rseq_kernel_feature_size(); + if (legacy || size < ORIG_RSEQ_ALLOC_SIZE) + rseq_alloc_size = ORIG_RSEQ_ALLOC_SIZE; + else + rseq_alloc_size = size; + + rc = sys_rseq(&__rseq.abi, rseq_alloc_size, 0, RSEQ_SIG); if (rc) { /* * After at least one thread has registered successfully @@ -179,9 +177,8 @@ int rseq_register_current_thread(void) * The first thread to register sets the rseq_size to mimic the libc * behavior. */ - if (RSEQ_READ_ONCE(rseq_size) == 0) { - RSEQ_WRITE_ONCE(rseq_size, get_rseq_kernel_feature_size()); - } + if (RSEQ_READ_ONCE(rseq_size) == 0) + RSEQ_WRITE_ONCE(rseq_size, size); return 0; } @@ -194,7 +191,7 @@ int rseq_unregister_current_thread(void) /* Treat libc's ownership as a successful unregistration. */ return 0; } - rc = sys_rseq(&__rseq.abi, get_rseq_min_alloc_size(), RSEQ_ABI_FLAG_UNREGISTER, RSEQ_SIG); + rc = sys_rseq(&__rseq.abi, rseq_alloc_size, RSEQ_ABI_FLAG_UNREGISTER, RSEQ_SIG); if (rc) return -1; return 0; diff --git a/tools/testing/selftests/rseq/rseq.h b/tools/testing/selftests/rseq/rseq.h index f51a5fdb0444..c62ebb9290c0 100644 --- a/tools/testing/selftests/rseq/rseq.h +++ b/tools/testing/selftests/rseq/rseq.h @@ -8,6 +8,7 @@ #ifndef RSEQ_H #define RSEQ_H +#include #include #include #include @@ -142,7 +143,12 @@ static inline struct rseq_abi *rseq_get_abi(void) * succeed. A restartable sequence executed from a non-registered * thread will always fail. */ -int rseq_register_current_thread(void); +int __rseq_register_current_thread(bool nolibc, bool legacy); + +static inline int rseq_register_current_thread(void) +{ + return __rseq_register_current_thread(false, false); +} /* * Unregister rseq for current thread. From 9b4e3495d1bd2469bf94b74930c153c2d534ddb7 Mon Sep 17 00:00:00 2001 From: Felix Kuehling Date: Mon, 20 Apr 2026 11:55:57 -0400 Subject: [PATCH 335/525] drm/amdkfd: Make all TLB-flushes heavy-weight With only one sequence number we cannot track the need for legacy vs heavy-weight flushes reliably. Always use heavy-weight. Signed-off-by: Felix Kuehling Reviewed-by: Philip Yang Signed-off-by: Alex Deucher (cherry picked from commit c1a3ff1d327820cd9a52bc1056b98681fc088949) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 4 ++-- drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 6 +++--- drivers/gpu/drm/amd/amdkfd/kfd_priv.h | 6 +++--- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 4 ++-- 4 files changed, 10 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index f829d65a79b4..f95bf6d95534 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -1360,7 +1360,7 @@ static int kfd_ioctl_map_memory_to_gpu(struct file *filep, peer_pdd = kfd_process_device_data_by_id(p, devices_arr[i]); if (WARN_ON_ONCE(!peer_pdd)) continue; - kfd_flush_tlb(peer_pdd, TLB_FLUSH_LEGACY); + kfd_flush_tlb(peer_pdd); } kfree(devices_arr); @@ -1455,7 +1455,7 @@ static int kfd_ioctl_unmap_memory_from_gpu(struct file *filep, if (WARN_ON_ONCE(!peer_pdd)) continue; if (flush_tlb) - kfd_flush_tlb(peer_pdd, TLB_FLUSH_HEAVYWEIGHT); + kfd_flush_tlb(peer_pdd); /* Remove dma mapping after tlb flush to avoid IO_PAGE_FAULT */ err = amdgpu_amdkfd_gpuvm_dmaunmap_mem(mem, peer_pdd->drm_priv); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index ab3b2e7be9bd..9185ebe4c079 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -572,7 +572,7 @@ static int allocate_vmid(struct device_queue_manager *dqm, qpd->vmid, qpd->page_table_base); /* invalidate the VM context after pasid and vmid mapping is set up */ - kfd_flush_tlb(qpd_to_pdd(qpd), TLB_FLUSH_LEGACY); + kfd_flush_tlb(qpd_to_pdd(qpd)); if (dqm->dev->kfd2kgd->set_scratch_backing_va) dqm->dev->kfd2kgd->set_scratch_backing_va(dqm->dev->adev, @@ -610,7 +610,7 @@ static void deallocate_vmid(struct device_queue_manager *dqm, if (flush_texture_cache_nocpsch(q->device, qpd)) dev_err(dev, "Failed to flush TC\n"); - kfd_flush_tlb(qpd_to_pdd(qpd), TLB_FLUSH_LEGACY); + kfd_flush_tlb(qpd_to_pdd(qpd)); /* Release the vmid mapping */ set_pasid_vmid_mapping(dqm, 0, qpd->vmid); @@ -1284,7 +1284,7 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm, dqm->dev->adev, qpd->vmid, qpd->page_table_base); - kfd_flush_tlb(pdd, TLB_FLUSH_LEGACY); + kfd_flush_tlb(pdd); } /* Take a safe reference to the mm_struct, which may otherwise diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index 163d665a6074..7b5b12206919 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -1554,13 +1554,13 @@ void kfd_signal_reset_event(struct kfd_node *dev); void kfd_signal_poison_consumed_event(struct kfd_node *dev, u32 pasid); void kfd_signal_process_terminate_event(struct kfd_process *p); -static inline void kfd_flush_tlb(struct kfd_process_device *pdd, - enum TLB_FLUSH_TYPE type) +static inline void kfd_flush_tlb(struct kfd_process_device *pdd) { struct amdgpu_device *adev = pdd->dev->adev; struct amdgpu_vm *vm = drm_priv_to_vm(pdd->drm_priv); - amdgpu_vm_flush_compute_tlb(adev, vm, type, pdd->dev->xcc_mask); + amdgpu_vm_flush_compute_tlb(adev, vm, TLB_FLUSH_HEAVYWEIGHT, + pdd->dev->xcc_mask); } static inline bool kfd_flush_tlb_after_unmap(struct kfd_dev *dev) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index 38085a0a0f58..35ec67d9739b 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -1424,7 +1424,7 @@ svm_range_unmap_from_gpus(struct svm_range *prange, unsigned long start, if (r) break; } - kfd_flush_tlb(pdd, TLB_FLUSH_HEAVYWEIGHT); + kfd_flush_tlb(pdd); } return r; @@ -1571,7 +1571,7 @@ svm_range_map_to_gpus(struct svm_range *prange, unsigned long offset, } } - kfd_flush_tlb(pdd, TLB_FLUSH_LEGACY); + kfd_flush_tlb(pdd); } return r; From 7bbfb2559bcec39d1a4e1182d931a2046112c352 Mon Sep 17 00:00:00 2001 From: "John B. Moore" Date: Tue, 28 Apr 2026 11:35:12 -0500 Subject: [PATCH 336/525] drm/amdgpu/gfx9: drop unnecessary 64-bit fence flag check in KIQ MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Remove the BUG_ON(flags & AMDGPU_FENCE_FLAG_64BIT) assertion from gfx_v9_0_ring_emit_fence_kiq(). The KIQ hardware supports 64-bit fence writes; the 32-bit writeback address constraint is an upper-layer convention, not a hardware limitation. The check serves no purpose and should not be present. Found by code inspection while investigating related BUG_ON assertions in the GFX and compute ring emission paths. Reviewed-by: Christian König Signed-off-by: John B. Moore Signed-off-by: Alex Deucher (cherry picked from commit 1b1101a46a426bb4328116bb5273c326a2780389) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 95be105671ec..86c7c2a429b7 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -5660,9 +5660,6 @@ static void gfx_v9_0_ring_emit_fence_kiq(struct amdgpu_ring *ring, u64 addr, { struct amdgpu_device *adev = ring->adev; - /* we only allocate 32bit for each seq wb address */ - BUG_ON(flags & AMDGPU_FENCE_FLAG_64BIT); - /* write fence seq to the "addr" */ amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) | From 2a561b361b7681509710f3cfc3d95d54c87ac69f Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 27 Apr 2026 11:38:58 -0400 Subject: [PATCH 337/525] drm/amdgpu/pm: add missing revision check for CI MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The ci_populate_all_memory_levels() workaround only applies to revision 0 SKUs. Link: https://gitlab.freedesktop.org/drm/amd/-/work_items/1816 Fixes: 9f4b35411cfe ("drm/amd/powerplay: add CI asics support to smumgr (v3)") Reviewed-by: Timur Kristóf Reviewed-by: Kent Russell Signed-off-by: Alex Deucher (cherry picked from commit 1db15ba8f72f400bbad8ae0ce24fafc43429d4bd) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/pm/powerplay/smumgr/ci_smumgr.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/powerplay/smumgr/ci_smumgr.c b/drivers/gpu/drm/amd/pm/powerplay/smumgr/ci_smumgr.c index 731355bdb9bc..0a3a0722b5c9 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/smumgr/ci_smumgr.c +++ b/drivers/gpu/drm/amd/pm/powerplay/smumgr/ci_smumgr.c @@ -1333,8 +1333,9 @@ static int ci_populate_all_memory_levels(struct pp_hwmgr *hwmgr) dev_id = adev->pdev->device; - if ((dpm_table->mclk_table.count >= 2) - && ((dev_id == 0x67B0) || (dev_id == 0x67B1))) { + if ((dpm_table->mclk_table.count >= 2) && + ((dev_id == 0x67B0) || (dev_id == 0x67B1)) && + (adev->pdev->revision == 0)) { smu_data->smc_state_table.MemoryLevel[1].MinVddci = smu_data->smc_state_table.MemoryLevel[0].MinVddci; smu_data->smc_state_table.MemoryLevel[1].MinMvdd = From 1987c79b4fe5789dfa14423e78b5c25f6acf3e9d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 28 Apr 2026 10:42:49 -0400 Subject: [PATCH 338/525] drm/amdgpu/pm: align Hawaii mclk workaround with radeon MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Align the hawaii mclk workaround with radeon and windows. Link: https://gitlab.freedesktop.org/drm/amd/-/work_items/1816 Fixes: 9f4b35411cfe ("drm/amd/powerplay: add CI asics support to smumgr (v3)") Reviewed-by: Timur Kristóf Reviewed-by: Kent Russell Signed-off-by: Alex Deucher (cherry picked from commit 9649528b637f668c5af9f2b83ca4ad8576ae2121) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/pm/powerplay/smumgr/ci_smumgr.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/powerplay/smumgr/ci_smumgr.c b/drivers/gpu/drm/amd/pm/powerplay/smumgr/ci_smumgr.c index 0a3a0722b5c9..3650e7beeb67 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/smumgr/ci_smumgr.c +++ b/drivers/gpu/drm/amd/pm/powerplay/smumgr/ci_smumgr.c @@ -1336,10 +1336,10 @@ static int ci_populate_all_memory_levels(struct pp_hwmgr *hwmgr) if ((dpm_table->mclk_table.count >= 2) && ((dev_id == 0x67B0) || (dev_id == 0x67B1)) && (adev->pdev->revision == 0)) { - smu_data->smc_state_table.MemoryLevel[1].MinVddci = - smu_data->smc_state_table.MemoryLevel[0].MinVddci; - smu_data->smc_state_table.MemoryLevel[1].MinMvdd = - smu_data->smc_state_table.MemoryLevel[0].MinMvdd; + smu_data->smc_state_table.MemoryLevel[1].MinVddc = + smu_data->smc_state_table.MemoryLevel[0].MinVddc; + smu_data->smc_state_table.MemoryLevel[1].MinVddcPhases = + smu_data->smc_state_table.MemoryLevel[0].MinVddcPhases; } smu_data->smc_state_table.MemoryLevel[0].ActivityLevel = 0x1F; CONVERT_FROM_HOST_TO_SMC_US(smu_data->smc_state_table.MemoryLevel[0].ActivityLevel); From 17223816498f7b117d138d18eb0eba63604dc74e Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 27 Apr 2026 11:40:25 -0400 Subject: [PATCH 339/525] drm/radeon: add missing revision check for CI MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The memory level workarounds only apply to revision 0 SKUs. Link: https://gitlab.freedesktop.org/drm/amd/-/work_items/1816 Fixes: 127e056e2a82 ("drm/radeon: fix mclk vddc configuration for cards for hawaii") Fixes: 21b8a369046f ("drm/radeon: fix dram timing for certain hawaii boards") Fixes: 90b2fee35cb9 ("drm/radeon: fix dpm mc init for certain hawaii boards") Reviewed-by: Timur Kristóf Reviewed-by: Kent Russell Signed-off-by: Alex Deucher (cherry picked from commit 4d8dcc14311515077062b5740f39f427075de5c9) Cc: stable@vger.kernel.org --- drivers/gpu/drm/radeon/ci_dpm.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/radeon/ci_dpm.c b/drivers/gpu/drm/radeon/ci_dpm.c index 22321eb95b7d..703848fac189 100644 --- a/drivers/gpu/drm/radeon/ci_dpm.c +++ b/drivers/gpu/drm/radeon/ci_dpm.c @@ -2461,7 +2461,8 @@ static void ci_register_patching_mc_arb(struct radeon_device *rdev, if (patch && ((rdev->pdev->device == 0x67B0) || - (rdev->pdev->device == 0x67B1))) { + (rdev->pdev->device == 0x67B1)) && + (rdev->pdev->revision == 0)) { if ((memory_clock > 100000) && (memory_clock <= 125000)) { tmp2 = (((0x31 * engine_clock) / 125000) - 1) & 0xff; *dram_timimg2 &= ~0x00ff0000; @@ -3304,7 +3305,8 @@ static int ci_populate_all_memory_levels(struct radeon_device *rdev) pi->smc_state_table.MemoryLevel[0].EnabledForActivity = 1; if ((dpm_table->mclk_table.count >= 2) && - ((rdev->pdev->device == 0x67B0) || (rdev->pdev->device == 0x67B1))) { + ((rdev->pdev->device == 0x67B0) || (rdev->pdev->device == 0x67B1)) && + (rdev->pdev->revision == 0)) { pi->smc_state_table.MemoryLevel[1].MinVddc = pi->smc_state_table.MemoryLevel[0].MinVddc; pi->smc_state_table.MemoryLevel[1].MinVddcPhases = @@ -4493,7 +4495,8 @@ static int ci_register_patching_mc_seq(struct radeon_device *rdev, if (patch && ((rdev->pdev->device == 0x67B0) || - (rdev->pdev->device == 0x67B1))) { + (rdev->pdev->device == 0x67B1)) && + (rdev->pdev->revision == 0)) { for (i = 0; i < table->last; i++) { if (table->last >= SMU7_DISCRETE_MC_REGISTER_ARRAY_SIZE) return -EINVAL; From 78d2e624fa073c14970aa097adcf3ea31c157a66 Mon Sep 17 00:00:00 2001 From: "John B. Moore" Date: Mon, 27 Apr 2026 16:06:28 -0500 Subject: [PATCH 340/525] drm/amdgpu/sdma4: replace BUG_ON with WARN_ON in fence emission MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit sdma_v4_0_ring_emit_fence() contains two BUG_ON(addr & 0x3) assertions that verify fence writeback addresses are dword-aligned. These assertions can be reached from unprivileged userspace via crafted DRM_IOCTL_AMDGPU_CS submissions, causing a fatal kernel panic in a scheduler worker thread. Replace both BUG_ON() calls with WARN_ON() to log the condition without crashing the kernel. A misaligned fence address at this point indicates a driver bug, but crashing the kernel is never the correct response when the assertion is reachable from userspace. The CS IOCTL path is the correct place to filter invalid submissions; the ring emission callback is too late to do anything about it. Fixes: 2130f89ced2c ("drm/amdgpu: add SDMA v4.0 implementation (v2)") Reviewed-by: Christian König Signed-off-by: John B. Moore Signed-off-by: Alex Deucher (cherry picked from commit b90250bd933afd1ba94d86d6b13821997b22b18e) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index 44f0f23e1148..e64f2f6df9a9 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -889,7 +889,7 @@ static void sdma_v4_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 se /* write the fence */ amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_FENCE)); /* zero in first two bits */ - BUG_ON(addr & 0x3); + WARN_ON(addr & 0x3); amdgpu_ring_write(ring, lower_32_bits(addr)); amdgpu_ring_write(ring, upper_32_bits(addr)); amdgpu_ring_write(ring, lower_32_bits(seq)); @@ -899,7 +899,7 @@ static void sdma_v4_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 se addr += 4; amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_FENCE)); /* zero in first two bits */ - BUG_ON(addr & 0x3); + WARN_ON(addr & 0x3); amdgpu_ring_write(ring, lower_32_bits(addr)); amdgpu_ring_write(ring, upper_32_bits(addr)); amdgpu_ring_write(ring, upper_32_bits(seq)); From e6c2e6c2e1fa066968a16aca1cb66cd1bdde7741 Mon Sep 17 00:00:00 2001 From: Philip Yang Date: Mon, 27 Apr 2026 09:30:23 -0400 Subject: [PATCH 341/525] drm/amdgpu: zero-initialize GART table on allocation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit GART TLB is flushed after unmapping but not after mapping. Since amdgpu_bo_create_kernel() does not zero-initialize the buffer, when a single PTE is written the TLB may speculatively load other uninitialized entries from the same cacheline. Those garbage entries can appear valid, and a subsequent write to another PTE in the same cacheline may cause the GPU to use a stale garbage PTE from the TLB. Fix this by calling memset_io() to zero-initialize the GART table with gart_pte_flags immediately after allocation. Using AMDGPU_GEM_CREATE_VRAM_CLEARED, SDMA-based clear will not work since SDMA needs GART to be initialized to work. Suggested-by: Felix Kuehling Signed-off-by: Philip Yang Reviewed-by: Christian König Signed-off-by: Alex Deucher (cherry picked from commit d9af8263b82b6eaa60c5718e0c6631c5037e4b24) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c index bc772ca3dab7..b6f849d51c2e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c @@ -262,12 +262,19 @@ void amdgpu_gart_table_ram_free(struct amdgpu_device *adev) */ int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev) { + int r; + if (adev->gart.bo != NULL) return 0; - return amdgpu_bo_create_kernel(adev, adev->gart.table_size, PAGE_SIZE, - AMDGPU_GEM_DOMAIN_VRAM, &adev->gart.bo, - NULL, (void *)&adev->gart.ptr); + r = amdgpu_bo_create_kernel(adev, adev->gart.table_size, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, &adev->gart.bo, + NULL, (void *)&adev->gart.ptr); + if (r) + return r; + + memset_io(adev->gart.ptr, adev->gart.gart_pte_flags, adev->gart.table_size); + return 0; } /** From 81665e35f143d93adef654f3be1360def9196e72 Mon Sep 17 00:00:00 2001 From: Xiaogang Chen Date: Fri, 24 Apr 2026 13:47:01 -0500 Subject: [PATCH 342/525] drm/amdkfd: Check if there are kfd porcesses using adev by kfd_processes_count During gpu hot-unplug need check if there are kfd porcesses still using the being removed gpu before clean resources of the device. Current driver checks if kfd_processes_table is empty. kfd processes are not terminated after removed from kfd_processes_table immediately. They are still alive and may access the device until kfd_process_wq work queue got ran. Check kfd->kfd_processes_count value that is updated after kfd process got uninitialized when its ref becomes zero. Fixes: 6cca686dfce7 ("drm/amdkfd: kfd driver supports hot unplug/replug amdgpu devices") Signed-off-by: Xiaogang Chen Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher (cherry picked from commit d12d05c4bc4c15585130af43e897923ff292df7b) --- drivers/gpu/drm/amd/amdkfd/kfd_device.c | 33 +------------------------ 1 file changed, 1 insertion(+), 32 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index 8ff97bf7d95a..b7f8f7ff8198 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c @@ -1737,37 +1737,6 @@ bool kgd2kfd_vmfault_fast_path(struct amdgpu_device *adev, struct amdgpu_iv_entr return false; } -/* check if there is kfd process still uses adev */ -static bool kgd2kfd_check_device_idle(struct amdgpu_device *adev) -{ - struct kfd_process *p; - struct hlist_node *p_temp; - unsigned int temp; - struct kfd_node *dev; - - mutex_lock(&kfd_processes_mutex); - - if (hash_empty(kfd_processes_table)) { - mutex_unlock(&kfd_processes_mutex); - return true; - } - - /* check if there is device still use adev */ - hash_for_each_safe(kfd_processes_table, temp, p_temp, p, kfd_processes) { - for (int i = 0; i < p->n_pdds; i++) { - dev = p->pdds[i]->dev; - if (dev->adev == adev) { - mutex_unlock(&kfd_processes_mutex); - return false; - } - } - } - - mutex_unlock(&kfd_processes_mutex); - - return true; -} - /** kgd2kfd_teardown_processes - gracefully tear down existing * kfd processes that use adev * @@ -1800,7 +1769,7 @@ void kgd2kfd_teardown_processes(struct amdgpu_device *adev) mutex_unlock(&kfd_processes_mutex); /* wait all kfd processes use adev terminate */ - while (!kgd2kfd_check_device_idle(adev)) + while (!!atomic_read(&adev->kfd.dev->kfd_processes_count)) cond_resched(); } From 6da7b1242da4455b11c24ce667d1cab1a348c8ea Mon Sep 17 00:00:00 2001 From: Sunil Khatri Date: Mon, 4 May 2026 18:21:17 +0530 Subject: [PATCH 343/525] drm/amdgpu/userq: fix access to stale wptr mapping MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Use drm_exec to take both locks i.e vm root bo and wptr_obj bo to access the mapping data properly. This fixes the security issue of unmap the wptr_obj while a queue creation is in progress and passing other bo at same address. Signed-off-by: Sunil Khatri Reviewed-by: Christian König Signed-off-by: Alex Deucher (cherry picked from commit 1fc6c8ab45dbee096469c08c13f6099d57a52d6c) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/mes_userqueue.c | 97 +++++++++------------- 1 file changed, 38 insertions(+), 59 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c b/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c index 2fc39a6938f6..5b4121ddc78c 100644 --- a/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c +++ b/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c @@ -30,34 +30,6 @@ #define AMDGPU_USERQ_PROC_CTX_SZ PAGE_SIZE #define AMDGPU_USERQ_GANG_CTX_SZ PAGE_SIZE -static int -mes_userq_map_gtt_bo_to_gart(struct amdgpu_bo *bo) -{ - int ret; - - ret = amdgpu_bo_reserve(bo, true); - if (ret) { - DRM_ERROR("Failed to reserve bo. ret %d\n", ret); - goto err_reserve_bo_failed; - } - - ret = amdgpu_ttm_alloc_gart(&bo->tbo); - if (ret) { - DRM_ERROR("Failed to bind bo to GART. ret %d\n", ret); - goto err_map_bo_gart_failed; - } - - amdgpu_bo_unreserve(bo); - bo = amdgpu_bo_ref(bo); - - return 0; - -err_map_bo_gart_failed: - amdgpu_bo_unreserve(bo); -err_reserve_bo_failed: - return ret; -} - static int mes_userq_create_wptr_mapping(struct amdgpu_device *adev, struct amdgpu_userq_mgr *uq_mgr, @@ -65,55 +37,62 @@ mes_userq_create_wptr_mapping(struct amdgpu_device *adev, uint64_t wptr) { struct amdgpu_bo_va_mapping *wptr_mapping; - struct amdgpu_vm *wptr_vm; struct amdgpu_userq_obj *wptr_obj = &queue->wptr_obj; + struct amdgpu_bo *obj; + struct amdgpu_vm *vm = queue->vm; + struct drm_exec exec; int ret; - wptr_vm = queue->vm; - ret = amdgpu_bo_reserve(wptr_vm->root.bo, false); - if (ret) - return ret; - wptr &= AMDGPU_GMC_HOLE_MASK; - wptr_mapping = amdgpu_vm_bo_lookup_mapping(wptr_vm, wptr >> PAGE_SHIFT); - amdgpu_bo_unreserve(wptr_vm->root.bo); - if (!wptr_mapping) { - DRM_ERROR("Failed to lookup wptr bo\n"); - return -EINVAL; + + drm_exec_init(&exec, DRM_EXEC_IGNORE_DUPLICATES, 2); + drm_exec_until_all_locked(&exec) { + ret = amdgpu_vm_lock_pd(vm, &exec, 1); + drm_exec_retry_on_contention(&exec); + if (unlikely(ret)) + goto fail_lock; + + wptr_mapping = amdgpu_vm_bo_lookup_mapping(vm, wptr >> PAGE_SHIFT); + if (!wptr_mapping) { + ret = -EINVAL; + goto fail_lock; + } + + obj = wptr_mapping->bo_va->base.bo; + ret = drm_exec_lock_obj(&exec, &obj->tbo.base); + drm_exec_retry_on_contention(&exec); + if (unlikely(ret)) + goto fail_lock; } - wptr_obj->obj = wptr_mapping->bo_va->base.bo; + wptr_obj->obj = amdgpu_bo_ref(wptr_mapping->bo_va->base.bo); if (wptr_obj->obj->tbo.base.size > PAGE_SIZE) { - DRM_ERROR("Requested GART mapping for wptr bo larger than one page\n"); - return -EINVAL; - } - - ret = mes_userq_map_gtt_bo_to_gart(wptr_obj->obj); - if (ret) { - DRM_ERROR("Failed to map wptr bo to GART\n"); - return ret; - } - - ret = amdgpu_bo_reserve(wptr_obj->obj, true); - if (ret) { - DRM_ERROR("Failed to reserve wptr bo\n"); - return ret; + ret = -EINVAL; + goto fail_map; } /* TODO use eviction fence instead of pinning. */ ret = amdgpu_bo_pin(wptr_obj->obj, AMDGPU_GEM_DOMAIN_GTT); if (ret) { - drm_file_err(uq_mgr->file, "[Usermode queues] Failed to pin wptr bo\n"); - goto unresv_bo; + DRM_ERROR("Failed to pin wptr bo. ret %d\n", ret); + goto fail_map; + } + + ret = amdgpu_ttm_alloc_gart(&wptr_obj->obj->tbo); + if (ret) { + DRM_ERROR("Failed to bind bo to GART. ret %d\n", ret); + goto fail_map; } queue->wptr_obj.gpu_addr = amdgpu_bo_gpu_offset(wptr_obj->obj); - amdgpu_bo_unreserve(wptr_obj->obj); + drm_exec_fini(&exec); return 0; -unresv_bo: - amdgpu_bo_unreserve(wptr_obj->obj); +fail_map: + amdgpu_bo_unref(&wptr_obj->obj); +fail_lock: + drm_exec_fini(&exec); return ret; } From 4e02e0afa95f691dc7cc17538cdd648089a843f0 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Mon, 13 Oct 2025 15:26:02 +0200 Subject: [PATCH 344/525] drm/amdgpu: nuke amdgpu_userq_fence_slab v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit As preparation for independent fences remove the extra slab, kmalloc should do just fine. v2: use GFP_KERNEL instead of GFP_ATOMIC Signed-off-by: Christian König Reviewed-by: Prike Liang Reviewed-by: Sunil Khatri Signed-off-by: Alex Deucher (cherry picked from commit 0d831487b5be0ae59cac865a0aa87b0acc3dc717) --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 13 ++------- .../gpu/drm/amd/amdgpu/amdgpu_userq_fence.c | 28 +++---------------- .../gpu/drm/amd/amdgpu/amdgpu_userq_fence.h | 3 -- 3 files changed, 7 insertions(+), 37 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 46aae3fad4bf..60debd543e44 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -3149,11 +3149,7 @@ static int __init amdgpu_init(void) r = amdgpu_sync_init(); if (r) - goto error_sync; - - r = amdgpu_userq_fence_slab_init(); - if (r) - goto error_fence; + return r; amdgpu_register_atpx_handler(); amdgpu_acpi_detect(); @@ -3161,7 +3157,7 @@ static int __init amdgpu_init(void) /* Ignore KFD init failures when CONFIG_HSA_AMD is not set. */ r = amdgpu_amdkfd_init(); if (r && r != -ENOENT) - goto error_fence; + goto error_fini_sync; if (amdgpu_pp_feature_mask & PP_OVERDRIVE_MASK) { add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_STILL_OK); @@ -3172,10 +3168,8 @@ static int __init amdgpu_init(void) /* let modprobe override vga console setting */ return pci_register_driver(&amdgpu_kms_pci_driver); -error_fence: +error_fini_sync: amdgpu_sync_fini(); - -error_sync: return r; } @@ -3186,7 +3180,6 @@ static void __exit amdgpu_exit(void) amdgpu_unregister_atpx_handler(); amdgpu_acpi_release(); amdgpu_sync_fini(); - amdgpu_userq_fence_slab_fini(); mmu_notifier_synchronize(); amdgpu_xcp_drv_release(); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c index da39ac862f37..e2d5f04296e1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.c @@ -32,29 +32,9 @@ #include "amdgpu.h" #include "amdgpu_userq_fence.h" -static const struct dma_fence_ops amdgpu_userq_fence_ops; -static struct kmem_cache *amdgpu_userq_fence_slab; - #define AMDGPU_USERQ_MAX_HANDLES (1U << 16) -int amdgpu_userq_fence_slab_init(void) -{ - amdgpu_userq_fence_slab = kmem_cache_create("amdgpu_userq_fence", - sizeof(struct amdgpu_userq_fence), - 0, - SLAB_HWCACHE_ALIGN, - NULL); - if (!amdgpu_userq_fence_slab) - return -ENOMEM; - - return 0; -} - -void amdgpu_userq_fence_slab_fini(void) -{ - rcu_barrier(); - kmem_cache_destroy(amdgpu_userq_fence_slab); -} +static const struct dma_fence_ops amdgpu_userq_fence_ops; static inline struct amdgpu_userq_fence *to_amdgpu_userq_fence(struct dma_fence *f) { @@ -231,7 +211,7 @@ void amdgpu_userq_fence_driver_put(struct amdgpu_userq_fence_driver *fence_drv) static int amdgpu_userq_fence_alloc(struct amdgpu_userq_fence **userq_fence) { - *userq_fence = kmem_cache_alloc(amdgpu_userq_fence_slab, GFP_ATOMIC); + *userq_fence = kmalloc(sizeof(**userq_fence), GFP_KERNEL); return *userq_fence ? 0 : -ENOMEM; } @@ -342,7 +322,7 @@ static void amdgpu_userq_fence_free(struct rcu_head *rcu) amdgpu_userq_fence_driver_put(fence_drv); kvfree(userq_fence->fence_drv_array); - kmem_cache_free(amdgpu_userq_fence_slab, userq_fence); + kfree(userq_fence); } static void amdgpu_userq_fence_release(struct dma_fence *f) @@ -545,7 +525,7 @@ int amdgpu_userq_signal_ioctl(struct drm_device *dev, void *data, r = amdgpu_userq_fence_create(queue, userq_fence, wptr, &fence); if (r) { mutex_unlock(&userq_mgr->userq_mutex); - kmem_cache_free(amdgpu_userq_fence_slab, userq_fence); + kfree(userq_fence); goto put_gobj_write; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.h index d56246ad8c26..d355a0eecc07 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq_fence.h @@ -58,9 +58,6 @@ struct amdgpu_userq_fence_driver { char timeline_name[TASK_COMM_LEN]; }; -int amdgpu_userq_fence_slab_init(void); -void amdgpu_userq_fence_slab_fini(void); - void amdgpu_userq_fence_driver_get(struct amdgpu_userq_fence_driver *fence_drv); void amdgpu_userq_fence_driver_put(struct amdgpu_userq_fence_driver *fence_drv); int amdgpu_userq_fence_driver_alloc(struct amdgpu_device *adev, From 26f6654a9a60eb4d241f42a0ec85412e8821480b Mon Sep 17 00:00:00 2001 From: Osama Abdelkader Date: Thu, 23 Apr 2026 22:06:20 +0200 Subject: [PATCH 345/525] drm/exynos: remove bridge when component_add fails MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Use devm_drm_bridge_add() so the bridge is released if probe fails after registration, and drop the manual drm_bridge_remove() in remove(). Check the return value of devm_drm_bridge_add(). Signed-off-by: Osama Abdelkader Fixes: 576d72fbfb45 ("drm/exynos: mic: add a bridge at probe") Cc: stable@vger.kernel.org Reviewed-by: Raphaël Gallais-Pou Reviewed-by: Luca Ceresoli Link: https://patch.msgid.link/20260423200622.325076-2-osama.abdelkader@gmail.com Signed-off-by: Luca Ceresoli --- drivers/gpu/drm/exynos/exynos_drm_mic.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/exynos/exynos_drm_mic.c b/drivers/gpu/drm/exynos/exynos_drm_mic.c index 29a8366513fa..e68c954ec3e6 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_mic.c +++ b/drivers/gpu/drm/exynos/exynos_drm_mic.c @@ -423,7 +423,9 @@ static int exynos_mic_probe(struct platform_device *pdev) mic->bridge.of_node = dev->of_node; - drm_bridge_add(&mic->bridge); + ret = devm_drm_bridge_add(dev, &mic->bridge); + if (ret) + goto err; pm_runtime_enable(dev); @@ -443,12 +445,8 @@ static int exynos_mic_probe(struct platform_device *pdev) static void exynos_mic_remove(struct platform_device *pdev) { - struct exynos_mic *mic = platform_get_drvdata(pdev); - component_del(&pdev->dev, &exynos_mic_component_ops); pm_runtime_disable(&pdev->dev); - - drm_bridge_remove(&mic->bridge); } static const struct of_device_id exynos_mic_of_match[] = { From ac8eb3e18f41e2cc8492cc1d358bcb786c850270 Mon Sep 17 00:00:00 2001 From: Benjamin Berg Date: Tue, 5 May 2026 15:15:40 +0200 Subject: [PATCH 346/525] wifi: mac80211: use safe list iteration in radar detect work The call to ieee80211_dfs_cac_cancel can cause the iterated chanctx to be freed and removed from the list. Guard against this to avoid a slab-use-after-free error. Cc: stable@vger.kernel.org Fixes: bca8bc0399ac ("wifi: mac80211: handle ieee80211_radar_detected() for MLO") Signed-off-by: Benjamin Berg Link: https://patch.msgid.link/20260505151539.236d63a1b736.I35dbb9e96a2d4a480be208770fdd99ba3b817b79@changeid Signed-off-by: Johannes Berg --- net/mac80211/util.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/mac80211/util.c b/net/mac80211/util.c index b093bc203c81..2529b01e2cd5 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -3700,11 +3700,11 @@ void ieee80211_dfs_radar_detected_work(struct wiphy *wiphy, struct ieee80211_local *local = container_of(work, struct ieee80211_local, radar_detected_work); struct cfg80211_chan_def chandef; - struct ieee80211_chanctx *ctx; + struct ieee80211_chanctx *ctx, *tmp; lockdep_assert_wiphy(local->hw.wiphy); - list_for_each_entry(ctx, &local->chanctx_list, list) { + list_for_each_entry_safe(ctx, tmp, &local->chanctx_list, list) { if (ctx->replace_state == IEEE80211_CHANCTX_REPLACES_OTHER) continue; From 644132a48f4e28a1d949d162160869286f3e75de Mon Sep 17 00:00:00 2001 From: Stephen Smalley Date: Tue, 5 May 2026 08:49:48 -0400 Subject: [PATCH 347/525] selinux: prune /sys/fs/selinux/checkreqprot commit a7e4676e8e2cb ("selinux: remove the 'checkreqprot' functionality") removed the ability to modify the checkreqprot setting but left everything except the updating of the checkreqprot value intact. Aside from unnecessary processing, this could produce a local DoS from log spam and incorrectly calls selinux_ima_measure_state() on each write even though no state has changed. Prune it to just log an error message once and return count (i.e. all bytes written successfully) so that userspace never breaks. Cc: stable@vger.kernel.org Signed-off-by: Stephen Smalley Signed-off-by: Paul Moore --- security/selinux/selinuxfs.c | 47 ++++++------------------------------ 1 file changed, 7 insertions(+), 40 deletions(-) diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c index 83aa765a09f9..6f74f87cb2b0 100644 --- a/security/selinux/selinuxfs.c +++ b/security/selinux/selinuxfs.c @@ -689,46 +689,13 @@ static ssize_t sel_read_checkreqprot(struct file *filp, char __user *buf, static ssize_t sel_write_checkreqprot(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { - char *page; - ssize_t length; - unsigned int new_value; - - length = avc_has_perm(current_sid(), SECINITSID_SECURITY, - SECCLASS_SECURITY, SECURITY__SETCHECKREQPROT, - NULL); - if (length) - return length; - - if (count >= PAGE_SIZE) - return -ENOMEM; - - /* No partial writes. */ - if (*ppos != 0) - return -EINVAL; - - page = memdup_user_nul(buf, count); - if (IS_ERR(page)) - return PTR_ERR(page); - - if (sscanf(page, "%u", &new_value) != 1) { - length = -EINVAL; - goto out; - } - length = count; - - if (new_value) { - char comm[sizeof(current->comm)]; - - strscpy(comm, current->comm); - pr_err("SELinux: %s (%d) set checkreqprot to 1. This is no longer supported.\n", - comm, current->pid); - } - - selinux_ima_measure_state(); - -out: - kfree(page); - return length; + /* + * Setting checkreqprot is no longer supported, see + * https://github.com/SELinuxProject/selinux-kernel/wiki/DEPRECATE-checkreqprot + */ + pr_err_once("SELinux: %s (%d) wrote to checkreqprot. This is no longer supported.\n", + current->comm, current->pid); + return count; } static const struct file_operations sel_checkreqprot_ops = { .read = sel_read_checkreqprot, From 19cfa0099024bb9cd40f6d950caa7f47ff8e77f6 Mon Sep 17 00:00:00 2001 From: Stephen Smalley Date: Tue, 5 May 2026 08:49:49 -0400 Subject: [PATCH 348/525] selinux: prune /sys/fs/selinux/disable Commit f22f9aaf6c3d ("selinux: remove the runtime disable functionality") removed the underlying SELinux runtime disable functionality but left everything else intact and started logging an error message to warn any residual users. Prune it to just log an error message once and to return count (i.e. all bytes written successfully) to avoid breaking userspace. This also fixes a local DoS from logspam. Cc: stable@vger.kernel.org Signed-off-by: Stephen Smalley Signed-off-by: Paul Moore --- security/selinux/selinuxfs.c | 36 +++++++----------------------------- 1 file changed, 7 insertions(+), 29 deletions(-) diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c index 6f74f87cb2b0..343303b73d6f 100644 --- a/security/selinux/selinuxfs.c +++ b/security/selinux/selinuxfs.c @@ -272,35 +272,13 @@ static ssize_t sel_write_disable(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { - char *page; - ssize_t length; - int new_value; - - if (count >= PAGE_SIZE) - return -ENOMEM; - - /* No partial writes. */ - if (*ppos != 0) - return -EINVAL; - - page = memdup_user_nul(buf, count); - if (IS_ERR(page)) - return PTR_ERR(page); - - if (sscanf(page, "%d", &new_value) != 1) { - length = -EINVAL; - goto out; - } - length = count; - - if (new_value) { - pr_err("SELinux: https://github.com/SELinuxProject/selinux-kernel/wiki/DEPRECATE-runtime-disable\n"); - pr_err("SELinux: Runtime disable is not supported, use selinux=0 on the kernel cmdline.\n"); - } - -out: - kfree(page); - return length; + /* + * Setting disable is no longer supported, see + * https://github.com/SELinuxProject/selinux-kernel/wiki/DEPRECATE-runtime-disable + */ + pr_err_once("SELinux: %s (%d) wrote to disable. This is no longer supported.\n", + current->comm, current->pid); + return count; } static const struct file_operations sel_disable_ops = { From ad1ac3d740cc6b858a99ab9c45c8c0574be7d1d3 Mon Sep 17 00:00:00 2001 From: Stephen Smalley Date: Tue, 5 May 2026 08:49:50 -0400 Subject: [PATCH 349/525] selinux: prune /sys/fs/selinux/user Remove the previously deprecated /sys/fs/selinux/user interface aside from a residual stub for userspace compatibility. Commit d7b6918e22c7 ("selinux: Deprecate /sys/fs/selinux/user") started the deprecation process for /sys/fs/selinux/user: The selinuxfs "user" node allows userspace to request a list of security contexts that can be reached for a given SELinux user from a given starting context. This was used by libselinux when various login-style programs requested contexts for users, but libselinux stopped using it in 2020. Kernel support will be removed no sooner than Dec 2025. A pr_warn() message has been in place since Linux v6.13, and a 5 second sleep was introduced since Linux v6.17 to help make it more noticeable. We are now past the stated deadline of Dec 2025, so remove the underlying functionality and replace it with a stub that returns a '0\0' buffer to avoid breaking userspace. This also avoids a local DoS from logspam and an uninterruptible sleep delay. Cc: stable@vger.kernel.org Signed-off-by: Stephen Smalley Signed-off-by: Paul Moore --- .../{obsolete => removed}/sysfs-selinux-user | 0 security/selinux/include/security.h | 2 - security/selinux/selinuxfs.c | 68 +--------- security/selinux/ss/services.c | 125 ------------------ 4 files changed, 5 insertions(+), 190 deletions(-) rename Documentation/ABI/{obsolete => removed}/sysfs-selinux-user (100%) diff --git a/Documentation/ABI/obsolete/sysfs-selinux-user b/Documentation/ABI/removed/sysfs-selinux-user similarity index 100% rename from Documentation/ABI/obsolete/sysfs-selinux-user rename to Documentation/ABI/removed/sysfs-selinux-user diff --git a/security/selinux/include/security.h b/security/selinux/include/security.h index d1f16d7f684d..0babb8992181 100644 --- a/security/selinux/include/security.h +++ b/security/selinux/include/security.h @@ -312,8 +312,6 @@ int security_context_to_sid_default(const char *scontext, u32 scontext_len, int security_context_to_sid_force(const char *scontext, u32 scontext_len, u32 *sid); -int security_get_user_sids(u32 fromsid, const char *username, u32 **sids, u32 *nel); - int security_port_sid(u8 protocol, u16 port, u32 *out_sid); int security_ib_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *out_sid); diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c index 343303b73d6f..6ed669309132 100644 --- a/security/selinux/selinuxfs.c +++ b/security/selinux/selinuxfs.c @@ -1018,69 +1018,11 @@ static ssize_t sel_write_relabel(struct file *file, char *buf, size_t size) static ssize_t sel_write_user(struct file *file, char *buf, size_t size) { - char *con = NULL, *user = NULL, *ptr; - u32 sid, *sids = NULL; - ssize_t length; - char *newcon; - int rc; - u32 i, len, nsids; - - pr_warn_ratelimited("SELinux: %s (%d) wrote to /sys/fs/selinux/user!" - " This will not be supported in the future; please update your" - " userspace.\n", current->comm, current->pid); - ssleep(5); - - length = avc_has_perm(current_sid(), SECINITSID_SECURITY, - SECCLASS_SECURITY, SECURITY__COMPUTE_USER, - NULL); - if (length) - goto out; - - length = -ENOMEM; - con = kzalloc(size + 1, GFP_KERNEL); - if (!con) - goto out; - - length = -ENOMEM; - user = kzalloc(size + 1, GFP_KERNEL); - if (!user) - goto out; - - length = -EINVAL; - if (sscanf(buf, "%s %s", con, user) != 2) - goto out; - - length = security_context_str_to_sid(con, &sid, GFP_KERNEL); - if (length) - goto out; - - length = security_get_user_sids(sid, user, &sids, &nsids); - if (length) - goto out; - - length = sprintf(buf, "%u", nsids) + 1; - ptr = buf + length; - for (i = 0; i < nsids; i++) { - rc = security_sid_to_context(sids[i], &newcon, &len); - if (rc) { - length = rc; - goto out; - } - if ((length + len) >= SIMPLE_TRANSACTION_LIMIT) { - kfree(newcon); - length = -ERANGE; - goto out; - } - memcpy(ptr, newcon, len); - kfree(newcon); - ptr += len; - length += len; - } -out: - kfree(sids); - kfree(user); - kfree(con); - return length; + pr_err_once("SELinux: %s (%d) wrote to user. This is no longer supported.\n", + current->comm, current->pid); + buf[0] = '0'; + buf[1] = 0; + return 2; } static ssize_t sel_write_member(struct file *file, char *buf, size_t size) diff --git a/security/selinux/ss/services.c b/security/selinux/ss/services.c index e8e7ccbd1e44..143021c5e326 100644 --- a/security/selinux/ss/services.c +++ b/security/selinux/ss/services.c @@ -2746,131 +2746,6 @@ int security_node_sid(u16 domain, return rc; } -#define SIDS_NEL 25 - -/** - * security_get_user_sids - Obtain reachable SIDs for a user. - * @fromsid: starting SID - * @username: username - * @sids: array of reachable SIDs for user - * @nel: number of elements in @sids - * - * Generate the set of SIDs for legal security contexts - * for a given user that can be reached by @fromsid. - * Set *@sids to point to a dynamically allocated - * array containing the set of SIDs. Set *@nel to the - * number of elements in the array. - */ - -int security_get_user_sids(u32 fromsid, - const char *username, - u32 **sids, - u32 *nel) -{ - struct selinux_policy *policy; - struct policydb *policydb; - struct sidtab *sidtab; - struct context *fromcon, usercon; - u32 *mysids = NULL, *mysids2, sid; - u32 i, j, mynel, maxnel = SIDS_NEL; - struct user_datum *user; - struct role_datum *role; - struct ebitmap_node *rnode, *tnode; - int rc; - - *sids = NULL; - *nel = 0; - - if (!selinux_initialized()) - return 0; - - mysids = kcalloc(maxnel, sizeof(*mysids), GFP_KERNEL); - if (!mysids) - return -ENOMEM; - -retry: - mynel = 0; - rcu_read_lock(); - policy = rcu_dereference(selinux_state.policy); - policydb = &policy->policydb; - sidtab = policy->sidtab; - - context_init(&usercon); - - rc = -EINVAL; - fromcon = sidtab_search(sidtab, fromsid); - if (!fromcon) - goto out_unlock; - - rc = -EINVAL; - user = symtab_search(&policydb->p_users, username); - if (!user) - goto out_unlock; - - usercon.user = user->value; - - ebitmap_for_each_positive_bit(&user->roles, rnode, i) { - role = policydb->role_val_to_struct[i]; - usercon.role = i + 1; - ebitmap_for_each_positive_bit(&role->types, tnode, j) { - usercon.type = j + 1; - - if (mls_setup_user_range(policydb, fromcon, user, - &usercon)) - continue; - - rc = sidtab_context_to_sid(sidtab, &usercon, &sid); - if (rc == -ESTALE) { - rcu_read_unlock(); - goto retry; - } - if (rc) - goto out_unlock; - if (mynel < maxnel) { - mysids[mynel++] = sid; - } else { - rc = -ENOMEM; - maxnel += SIDS_NEL; - mysids2 = kcalloc(maxnel, sizeof(*mysids2), GFP_ATOMIC); - if (!mysids2) - goto out_unlock; - memcpy(mysids2, mysids, mynel * sizeof(*mysids2)); - kfree(mysids); - mysids = mysids2; - mysids[mynel++] = sid; - } - } - } - rc = 0; -out_unlock: - rcu_read_unlock(); - if (rc || !mynel) { - kfree(mysids); - return rc; - } - - rc = -ENOMEM; - mysids2 = kcalloc(mynel, sizeof(*mysids2), GFP_KERNEL); - if (!mysids2) { - kfree(mysids); - return rc; - } - for (i = 0, j = 0; i < mynel; i++) { - struct av_decision dummy_avd; - rc = avc_has_perm_noaudit(fromsid, mysids[i], - SECCLASS_PROCESS, /* kernel value */ - PROCESS__TRANSITION, AVC_STRICT, - &dummy_avd); - if (!rc) - mysids2[j++] = mysids[i]; - cond_resched(); - } - kfree(mysids); - *sids = mysids2; - *nel = j; - return 0; -} - /** * __security_genfs_sid - Helper to obtain a SID for a file in a filesystem * @policy: policy From a02cd6805562305f936e807da83e253b719dd965 Mon Sep 17 00:00:00 2001 From: Stephen Smalley Date: Tue, 5 May 2026 10:06:38 -0400 Subject: [PATCH 350/525] selinux: allow multiple opens of /sys/fs/selinux/policy Currently there can only be a single open of /sys/fs/selinux/policy at any time. This allows any process to block any other process from reading the kernel policy. The original motivation seems to have been a mix of preventing an inconsistent view of the policy size and preventing userspace from allocating kernel memory without bound, but this is arguably equally bad. Eliminate the policy_opened flag and shrink the critical section that the policy mutex is held. While we are making changes here, drop a couple of extraneous BUG_ONs. Cc: stable@vger.kernel.org Link: https://lore.kernel.org/selinux/20100726193414.19538.64028.stgit@paris.rdu.redhat.com/ Signed-off-by: Stephen Smalley Signed-off-by: Paul Moore --- security/selinux/selinuxfs.c | 27 ++++----------------------- 1 file changed, 4 insertions(+), 23 deletions(-) diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c index 6ed669309132..a43a38a3ae25 100644 --- a/security/selinux/selinuxfs.c +++ b/security/selinux/selinuxfs.c @@ -76,7 +76,6 @@ struct selinux_fs_info { int *bool_pending_values; struct dentry *class_dir; unsigned long last_class_ino; - bool policy_opened; unsigned long last_ino; struct super_block *sb; }; @@ -340,44 +339,31 @@ struct policy_load_memory { static int sel_open_policy(struct inode *inode, struct file *filp) { - struct selinux_fs_info *fsi = inode->i_sb->s_fs_info; struct policy_load_memory *plm = NULL; int rc; - BUG_ON(filp->private_data); - - mutex_lock(&selinux_state.policy_mutex); - rc = avc_has_perm(current_sid(), SECINITSID_SECURITY, SECCLASS_SECURITY, SECURITY__READ_POLICY, NULL); if (rc) - goto err; + return rc; - rc = -EBUSY; - if (fsi->policy_opened) - goto err; - - rc = -ENOMEM; plm = kzalloc_obj(*plm); if (!plm) - goto err; + return -ENOMEM; + mutex_lock(&selinux_state.policy_mutex); rc = security_read_policy(&plm->data, &plm->len); if (rc) goto err; - if ((size_t)i_size_read(inode) != plm->len) { inode_lock(inode); i_size_write(inode, plm->len); inode_unlock(inode); } - - fsi->policy_opened = 1; + mutex_unlock(&selinux_state.policy_mutex); filp->private_data = plm; - mutex_unlock(&selinux_state.policy_mutex); - return 0; err: mutex_unlock(&selinux_state.policy_mutex); @@ -390,13 +376,8 @@ static int sel_open_policy(struct inode *inode, struct file *filp) static int sel_release_policy(struct inode *inode, struct file *filp) { - struct selinux_fs_info *fsi = inode->i_sb->s_fs_info; struct policy_load_memory *plm = filp->private_data; - BUG_ON(!plm); - - fsi->policy_opened = 0; - vfree(plm->data); kfree(plm); From 868f31e4061eca8c3cd607d79d954d5e54f204aa Mon Sep 17 00:00:00 2001 From: Stephen Smalley Date: Thu, 30 Apr 2026 14:36:52 -0400 Subject: [PATCH 351/525] selinux: shrink critical section in sel_write_load() Currently sel_write_load() takes the policy mutex earlier than necessary. Move the taking of the mutex later. This avoids holding it unnecessarily across the vmalloc() and copy_from_user() of the policy data. Cc: stable@vger.kernel.org Signed-off-by: Stephen Smalley Signed-off-by: Paul Moore --- security/selinux/selinuxfs.c | 18 ++++++++---------- 1 file changed, 8 insertions(+), 10 deletions(-) diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c index a43a38a3ae25..25ca7d714014 100644 --- a/security/selinux/selinuxfs.c +++ b/security/selinux/selinuxfs.c @@ -553,34 +553,31 @@ static ssize_t sel_write_load(struct file *file, const char __user *buf, if (!count) return -EINVAL; - mutex_lock(&selinux_state.policy_mutex); - length = avc_has_perm(current_sid(), SECINITSID_SECURITY, SECCLASS_SECURITY, SECURITY__LOAD_POLICY, NULL); if (length) - goto out; + return length; data = vmalloc(count); - if (!data) { - length = -ENOMEM; - goto out; - } + if (!data) + return -ENOMEM; if (copy_from_user(data, buf, count) != 0) { length = -EFAULT; goto out; } + mutex_lock(&selinux_state.policy_mutex); length = security_load_policy(data, count, &load_state); if (length) { pr_warn_ratelimited("SELinux: failed to load policy\n"); - goto out; + goto out_unlock; } fsi = file_inode(file)->i_sb->s_fs_info; length = sel_make_policy_nodes(fsi, load_state.policy); if (length) { pr_warn_ratelimited("SELinux: failed to initialize selinuxfs\n"); selinux_policy_cancel(&load_state); - goto out; + goto out_unlock; } selinux_policy_commit(&load_state); @@ -590,8 +587,9 @@ static ssize_t sel_write_load(struct file *file, const char __user *buf, from_kuid(&init_user_ns, audit_get_loginuid(current)), audit_get_sessionid(current)); -out: +out_unlock: mutex_unlock(&selinux_state.policy_mutex); +out: vfree(data); return length; } From 60a1e131a811b68703da58fd805ab359b704ab03 Mon Sep 17 00:00:00 2001 From: Gustavo Sousa Date: Thu, 16 Apr 2026 15:17:19 -0300 Subject: [PATCH 352/525] drm/xe/hdcp: Add NULL check for media_gt in intel_hdcp_gsc_check_status() When media GT is disabled via configfs, there is no allocation for media_gt, which is kept as NULL. In such scenario, intel_hdcp_gsc_check_status() results in a kernel pagefault error due to >->uc.gsc being evaluated as an invalid memory address. Fix that by introducing a NULL check on media_gt and bailing out early if so. While at it, also drop the NULL check for gsc, since it can't be NULL if media_gt is not NULL. v2: - Get address for gsc only after checking that gt is not NULL. (Shuicheng) - Drop the NULL check for gsc. (Shuicheng) v3: - Add "Fixes" and "Cc: " tags. (Matt) Fixes: 4af50beb4e0f ("drm/xe: Use gsc_proxy_init_done to check proxy status") Cc: # v6.10+ Reviewed-by: Matt Roper Reviewed-by: Shuicheng Lin Link: https://patch.msgid.link/20260416-check-for-null-media_gt-in-intel_hdcp_gsc_check_status-v2-1-9adb9fd3b621@intel.com Signed-off-by: Gustavo Sousa (cherry picked from commit bfaf87e84ca3ca3f6e275f9ae56da47a8b55ffd1) Signed-off-by: Matthew Brost --- drivers/gpu/drm/xe/display/xe_hdcp_gsc.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/xe/display/xe_hdcp_gsc.c b/drivers/gpu/drm/xe/display/xe_hdcp_gsc.c index 29c72aa4b0d2..33494b86205d 100644 --- a/drivers/gpu/drm/xe/display/xe_hdcp_gsc.c +++ b/drivers/gpu/drm/xe/display/xe_hdcp_gsc.c @@ -37,9 +37,17 @@ static bool intel_hdcp_gsc_check_status(struct drm_device *drm) struct xe_device *xe = to_xe_device(drm); struct xe_tile *tile = xe_device_get_root_tile(xe); struct xe_gt *gt = tile->media_gt; - struct xe_gsc *gsc = >->uc.gsc; + struct xe_gsc *gsc; - if (!gsc || !xe_uc_fw_is_available(&gsc->fw)) { + if (!gt) { + drm_dbg_kms(&xe->drm, + "not checking GSC status for HDCP2.x: media GT not present or disabled\n"); + return false; + } + + gsc = >->uc.gsc; + + if (!xe_uc_fw_is_available(&gsc->fw)) { drm_dbg_kms(&xe->drm, "GSC Components not ready for HDCP2.x\n"); return false; From d01012c740bbb298b957e30cc0848e482c6f486f Mon Sep 17 00:00:00 2001 From: Shuicheng Lin Date: Tue, 28 Apr 2026 20:14:48 +0000 Subject: [PATCH 353/525] drm/xe/pf: Fix EAGAIN sign in pf_migration_consume() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit PTR_ERR() returns a negative value, so comparing against the positive EAGAIN is always true for ERR_PTR(-EAGAIN), causing pf_migration_consume() to bail out instead of continuing to the remaining GTs. On multi-GT platforms this can skip GTs that already have data ready. Compare against -EAGAIN to match the intent (and the following line that correctly uses -EAGAIN). While at it, gate PTR_ERR() with IS_ERR(). v2: add IS_ERR() guard before PTR_ERR(). (Gustavo) Fixes: 67df4a5cbc58 ("drm/xe/pf: Add data structures and handlers for migration rings") Cc: Michał Winiarski Reviewed-by: Gustavo Sousa Link: https://patch.msgid.link/20260428201448.3999428-1-shuicheng.lin@intel.com Signed-off-by: Shuicheng Lin (cherry picked from commit 9d770e72e1edb54beacfce5f402edb51632811e3) Signed-off-by: Matthew Brost --- drivers/gpu/drm/xe/xe_sriov_pf_migration.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_sriov_pf_migration.c b/drivers/gpu/drm/xe/xe_sriov_pf_migration.c index 6c4b16409cc9..150a241110fb 100644 --- a/drivers/gpu/drm/xe/xe_sriov_pf_migration.c +++ b/drivers/gpu/drm/xe/xe_sriov_pf_migration.c @@ -149,10 +149,11 @@ pf_migration_consume(struct xe_device *xe, unsigned int vfid) for_each_gt(gt, xe, gt_id) { data = xe_gt_sriov_pf_migration_save_consume(gt, vfid); - if (data && PTR_ERR(data) != EAGAIN) + if (!data) + continue; + if (!IS_ERR(data) || PTR_ERR(data) != -EAGAIN) return data; - if (PTR_ERR(data) == -EAGAIN) - more_data = true; + more_data = true; } if (!more_data) From b87951a0ae9f95ca6590bf0939edced7d36929dd Mon Sep 17 00:00:00 2001 From: Shuicheng Lin Date: Wed, 29 Apr 2026 19:22:59 +0000 Subject: [PATCH 354/525] drm/xe/pf: Fix MMIO access using PF view instead of VF view during migration MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit pf_migration_mmio_save() and pf_migration_mmio_restore() initialize a local VF-specific MMIO view via xe_mmio_init_vf_view() but then pass >->mmio (the PF base) to all xe_mmio_read32()/xe_mmio_write32() calls instead of the local &mmio. This causes the PF own SW flag registers to be saved/restored rather than the target VF registers, silently corrupting migration state. Use the VF MMIO view for all register accesses, matching the correct pattern used in pf_clear_vf_scratch_regs(). Fixes: b7c1b990f719 ("drm/xe/pf: Handle MMIO migration data as part of PF control") Cc: Michał Winiarski Assisted-by: Claude:claude-opus-4.6 Reviewed-by: Michal Wajdeczko Reviewed-by: Stuart Summers Link: https://patch.msgid.link/20260429192259.4009211-1-shuicheng.lin@intel.com Signed-off-by: Shuicheng Lin (cherry picked from commit 7d9c39cfb31ff389490ca1308767c2807a9829a6) Signed-off-by: Matthew Brost --- drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c b/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c index 87a164efcc33..01fe03b9efe8 100644 --- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c +++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c @@ -385,10 +385,10 @@ static int pf_migration_mmio_save(struct xe_gt *gt, unsigned int vfid, void *buf if (xe_gt_is_media_type(gt)) for (n = 0; n < MED_VF_SW_FLAG_COUNT; n++) - regs[n] = xe_mmio_read32(>->mmio, MED_VF_SW_FLAG(n)); + regs[n] = xe_mmio_read32(&mmio, MED_VF_SW_FLAG(n)); else for (n = 0; n < VF_SW_FLAG_COUNT; n++) - regs[n] = xe_mmio_read32(>->mmio, VF_SW_FLAG(n)); + regs[n] = xe_mmio_read32(&mmio, VF_SW_FLAG(n)); return 0; } @@ -407,10 +407,10 @@ static int pf_migration_mmio_restore(struct xe_gt *gt, unsigned int vfid, if (xe_gt_is_media_type(gt)) for (n = 0; n < MED_VF_SW_FLAG_COUNT; n++) - xe_mmio_write32(>->mmio, MED_VF_SW_FLAG(n), regs[n]); + xe_mmio_write32(&mmio, MED_VF_SW_FLAG(n), regs[n]); else for (n = 0; n < VF_SW_FLAG_COUNT; n++) - xe_mmio_write32(>->mmio, VF_SW_FLAG(n), regs[n]); + xe_mmio_write32(&mmio, VF_SW_FLAG(n), regs[n]); return 0; } From b29987dfd943e655df6e3b641ecffad5cc1509c2 Mon Sep 17 00:00:00 2001 From: Satyanarayana K V P Date: Mon, 4 May 2026 09:49:26 +0000 Subject: [PATCH 355/525] drm/xe/guc: Exclude indirect ring state page from ADS engine state size The engine state size reported to GuC via ADS should only include the engine state portion and should not include the indirect ring state page that comes after it in the context image. The GuC uses this size to overwrite the engine state in the LRC on watchdog resets and we don't want it to overwrite the indirect ring state as well. Fixes: d6219e1cd5e3 ("drm/xe: Add Indirect Ring State support") Suggested-by: Daniele Ceraolo Spurio Signed-off-by: Satyanarayana K V P Cc: Michal Wajdeczko Cc: Matthew Brost Reviewed-by: Matthew Brost Reviewed-by: Daniele Ceraolo Spurio Signed-off-by: Daniele Ceraolo Spurio Link: https://patch.msgid.link/20260504094924.3760713-4-satyanarayana.k.v.p@intel.com (cherry picked from commit 3ec5f003f6c377beda8bd5438941f5a7795e1848) Signed-off-by: Matthew Brost --- drivers/gpu/drm/xe/xe_guc_ads.c | 5 +---- drivers/gpu/drm/xe/xe_lrc.c | 11 +++++++++-- drivers/gpu/drm/xe/xe_lrc.h | 2 +- 3 files changed, 11 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_guc_ads.c b/drivers/gpu/drm/xe/xe_guc_ads.c index 81b5f01b1f65..2b835d48b565 100644 --- a/drivers/gpu/drm/xe/xe_guc_ads.c +++ b/drivers/gpu/drm/xe/xe_guc_ads.c @@ -512,12 +512,9 @@ static void guc_golden_lrc_init(struct xe_guc_ads *ads) * that starts after the execlists LRC registers. This is * required to allow the GuC to restore just the engine state * when a watchdog reset occurs. - * We calculate the engine state size by removing the size of - * what comes before it in the context image (which is identical - * on all engines). */ ads_blob_write(ads, ads.eng_state_size[guc_class], - real_size - xe_lrc_skip_size(xe)); + xe_lrc_engine_state_size(gt, class)); ads_blob_write(ads, ads.golden_context_lrca[guc_class], addr_ggtt); diff --git a/drivers/gpu/drm/xe/xe_lrc.c b/drivers/gpu/drm/xe/xe_lrc.c index c725cde4508d..4af9f0d7c6f3 100644 --- a/drivers/gpu/drm/xe/xe_lrc.c +++ b/drivers/gpu/drm/xe/xe_lrc.c @@ -746,9 +746,16 @@ size_t xe_lrc_reg_size(struct xe_device *xe) return 80 * sizeof(u32); } -size_t xe_lrc_skip_size(struct xe_device *xe) +/** + * xe_lrc_engine_state_size() - Get size of the engine state within LRC + * @gt: the &xe_gt struct instance + * @class: Hardware engine class + * + * Returns: Size of the engine state + */ +size_t xe_lrc_engine_state_size(struct xe_gt *gt, enum xe_engine_class class) { - return LRC_PPHWSP_SIZE + xe_lrc_reg_size(xe); + return xe_gt_lrc_hang_replay_size(gt, class) - xe_lrc_reg_size(gt_to_xe(gt)); } static inline u32 __xe_lrc_seqno_offset(struct xe_lrc *lrc) diff --git a/drivers/gpu/drm/xe/xe_lrc.h b/drivers/gpu/drm/xe/xe_lrc.h index e7c975f9e2d9..5440663183f6 100644 --- a/drivers/gpu/drm/xe/xe_lrc.h +++ b/drivers/gpu/drm/xe/xe_lrc.h @@ -130,7 +130,7 @@ u32 xe_lrc_parallel_ggtt_addr(struct xe_lrc *lrc); struct iosys_map xe_lrc_parallel_map(struct xe_lrc *lrc); size_t xe_lrc_reg_size(struct xe_device *xe); -size_t xe_lrc_skip_size(struct xe_device *xe); +size_t xe_lrc_engine_state_size(struct xe_gt *gt, enum xe_engine_class class); void xe_lrc_dump_default(struct drm_printer *p, struct xe_gt *gt, From 901a7d9e2f280a9e76e6c58406a519cb11ad5ff8 Mon Sep 17 00:00:00 2001 From: Alyssa Ross Date: Sun, 3 May 2026 21:25:16 +0200 Subject: [PATCH 356/525] ipv6: default IPV6_SIT to m This basically defaulted to m until recently, since IPV6 defaulted to m. Since IPV6 was changed to a boolean with a default of y, IPV6_SIT started defaulting to built-in as well. This results in a surprise sit0 device by default for defconfig (and defconfig-derived config) users at boot. For me, this broke an (admittedly non-robust) script. Preserve the behaviour of most configs by avoiding building this module, that's probably overall seldom used compared to IPv6 as a whole, into the kernel. Fixes: 309b905deee59 ("ipv6: convert CONFIG_IPV6 to built-in only and clean up Kconfigs") Signed-off-by: Alyssa Ross Reviewed-by: Fernando Fernandez Mancera Link: https://patch.msgid.link/20260503192515.290900-2-hi@alyssa.is Signed-off-by: Jakub Kicinski --- net/ipv6/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/ipv6/Kconfig b/net/ipv6/Kconfig index c024aa77f25b..c3806c6ac96f 100644 --- a/net/ipv6/Kconfig +++ b/net/ipv6/Kconfig @@ -164,7 +164,7 @@ config IPV6_SIT select INET_TUNNEL select NET_IP_TUNNEL select IPV6_NDISC_NODETYPE - default y + default m help Tunneling means encapsulating data of one protocol type within another protocol and sending it over a channel that understands the @@ -172,7 +172,7 @@ config IPV6_SIT into IPv4 packets. This is useful if you want to connect two IPv6 networks over an IPv4-only path. - Saying M here will produce a module called sit. If unsure, say Y. + Saying M here will produce a module called sit. If unsure, say M. config IPV6_SIT_6RD bool "IPv6: IPv6 Rapid Deployment (6RD)" From 5ad509c1fdad4bf0993b72d1b3d462f036d8a0d8 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Mon, 4 May 2026 06:43:13 +0000 Subject: [PATCH 357/525] ipv6: Fix null-ptr-deref in fib6_mtu(). syzbot reported null-ptr-deref in fib6_mtu(). [0] When res->f6i->fib6_pmtu is 0 in fib6_mtu(), it fetches MTU from __in6_dev_get(nh->fib_nh_dev)->cnf.mtu6. However, __in6_dev_get() could return NULL when the device is being unregistered. Let's return 0 MTU if __in6_dev_get() returns NULL in fib6_mtu(). [0]: Oops: general protection fault, probably for non-canonical address 0xdffffc00000000bc: 0000 [#1] SMP KASAN NOPTI KASAN: null-ptr-deref in range [0x00000000000005e0-0x00000000000005e7] CPU: 0 UID: 0 PID: 7890 Comm: syz.2.502 Tainted: G L syzkaller #0 PREEMPT(full) Tainted: [L]=SOFTLOCKUP Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2 04/01/2014 RIP: 0010:fib6_mtu net/ipv6/route.c:1648 [inline] RIP: 0010:rt6_insert_exception+0x9eb/0x10a0 net/ipv6/route.c:1753 Code: 3b 14 cf f7 45 85 f6 0f 85 1d 02 00 00 e8 7d 19 cf f7 48 8d bb e0 05 00 00 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <0f> b6 14 02 48 89 f8 83 e0 07 83 c0 03 38 d0 7c 08 84 d2 0f 85 89 RSP: 0000:ffffc9000610f120 EFLAGS: 00010202 RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffffc9000c001000 RDX: 00000000000000bc RSI: ffffffff8a38bc83 RDI: 00000000000005e0 RBP: ffff888052f06000 R08: 0000000000000005 R09: 0000000000000000 R10: 0000000000000001 R11: 0000000000000000 R12: ffff888042d16c00 R13: ffff888042d16cc8 R14: 0000000000000001 R15: 0000000000000500 FS: 0000000000000000(0000) GS:ffff88809717d000(0063) knlGS:00000000f540db40 CS: 0010 DS: 002b ES: 002b CR0: 0000000080050033 CR2: 00000000f73c6d50 CR3: 000000006eff0000 CR4: 0000000000352ef0 Call Trace: __ip6_rt_update_pmtu+0x555/0xd60 net/ipv6/route.c:2982 ip6_update_pmtu+0x34f/0x3b0 net/ipv6/route.c:3014 icmpv6_err+0x2a2/0x3f0 net/ipv6/icmp.c:82 icmpv6_notify+0x35e/0x820 net/ipv6/icmp.c:1087 icmpv6_rcv+0x10bf/0x1ae0 net/ipv6/icmp.c:1228 ip6_protocol_deliver_rcu+0xf97/0x1500 net/ipv6/ip6_input.c:478 ip6_input_finish+0x1e4/0x4a0 net/ipv6/ip6_input.c:529 NF_HOOK include/linux/netfilter.h:318 [inline] NF_HOOK include/linux/netfilter.h:312 [inline] ip6_input+0x105/0x2f0 net/ipv6/ip6_input.c:540 ip6_mc_input+0x513/0xf50 net/ipv6/ip6_input.c:630 dst_input include/net/dst.h:480 [inline] ip6_rcv_finish net/ipv6/ip6_input.c:119 [inline] NF_HOOK include/linux/netfilter.h:318 [inline] NF_HOOK include/linux/netfilter.h:312 [inline] ipv6_rcv+0x34c/0x3d0 net/ipv6/ip6_input.c:351 __netif_receive_skb_one_core+0x12d/0x1e0 net/core/dev.c:6202 __netif_receive_skb+0x1f/0x120 net/core/dev.c:6315 netif_receive_skb_internal net/core/dev.c:6401 [inline] netif_receive_skb+0x13b/0x7f0 net/core/dev.c:6460 tun_rx_batched.isra.0+0x3f6/0x750 drivers/net/tun.c:1511 tun_get_user+0x1e31/0x3c20 drivers/net/tun.c:1955 tun_chr_write_iter+0xdc/0x200 drivers/net/tun.c:2001 new_sync_write fs/read_write.c:595 [inline] vfs_write+0x6ac/0x1070 fs/read_write.c:688 ksys_write+0x12a/0x250 fs/read_write.c:740 do_syscall_32_irqs_on arch/x86/entry/syscall_32.c:83 [inline] do_int80_emulation+0x141/0x700 arch/x86/entry/syscall_32.c:172 asm_int80_emulation+0x1a/0x20 arch/x86/include/asm/idtentry.h:621 RIP: 0023:0xf715616b Code: 57 56 53 8b 44 24 14 f6 00 08 75 23 8b 44 24 18 8b 5c 24 1c 8b 4c 24 20 8b 54 24 24 8b 74 24 28 8b 7c 24 2c 8b 6c 24 30 cd 80 <5b> 5e 5f 5d c3 5b 5e 5f 5d e9 f7 a1 ff ff 66 90 66 90 66 90 90 53 RSP: 002b:00000000f540d44c EFLAGS: 00000246 ORIG_RAX: 0000000000000004 RAX: ffffffffffffffda RBX: 00000000000000c8 RCX: 0000000080000640 RDX: 000000000000007a RSI: 0000000000000000 RDI: 0000000000000000 RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000292 R12: 0000000000000000 R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000 Fixes: dcd1f572954f ("net/ipv6: Remove fib6_idev") Reported-by: syzbot+01f005f9c6387ca6f6dd@syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/69f83f22.170a0220.13cc2.0004.GAE@google.com/ Signed-off-by: Kuniyuki Iwashima Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260504064316.3820775-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- net/ipv6/route.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 0dc0316530ca..e3d355d1fbd6 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1645,6 +1645,10 @@ static unsigned int fib6_mtu(const struct fib6_result *res) rcu_read_lock(); idev = __in6_dev_get(dev); + if (!idev) { + rcu_read_unlock(); + return 0; + } mtu = READ_ONCE(idev->cnf.mtu6); rcu_read_unlock(); } From 07f44433355f70fa97d4c44b4c0d2e86adc082fb Mon Sep 17 00:00:00 2001 From: Michael Chan Date: Mon, 4 May 2026 14:06:08 +0530 Subject: [PATCH 358/525] bnxt_en: Delay for 5 seconds after AER DPC for all chips The FW on all chips is requiring a 5-second delay after Downstream Port Containment (DPC) AER. The previously added 900 msec delay was not long enough in all cases because the chip's CRS (Configuration Request Retry Status) mechanism is not always reliable. Fixes: d5ab32e9b02d ("bnxt_en: Add delay to handle Downstream Port Containment (DPC) AER") Reviewed-by: Kalesh AP Signed-off-by: Michael Chan Signed-off-by: Pavan Chebbi Link: https://patch.msgid.link/20260504083611.1383776-2-pavan.chebbi@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 8c55874f44ca..3db951d0c690 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -17360,9 +17360,14 @@ static pci_ers_result_t bnxt_io_slot_reset(struct pci_dev *pdev) netdev_info(bp->dev, "PCI Slot Reset\n"); - if (!(bp->flags & BNXT_FLAG_CHIP_P5_PLUS) && - test_bit(BNXT_STATE_PCI_CHANNEL_IO_FROZEN, &bp->state)) - msleep(900); + if (test_bit(BNXT_STATE_PCI_CHANNEL_IO_FROZEN, &bp->state)) { + /* After DPC, the chip should return CRS when the vendor ID + * config register is read until it is ready. On all chips, + * this is not happening reliably so add a 5-second delay as a + * workaround. + */ + msleep(5000); + } netdev_lock(netdev); From 54c28fab2fa5afd681c9c4b10f4f6da1efdd397a Mon Sep 17 00:00:00 2001 From: Michael Chan Date: Mon, 4 May 2026 14:06:09 +0530 Subject: [PATCH 359/525] bnxt_en: Set bp->max_tpa according to what the FW supports Fix the logic to set bp->max_tpa no higher than what the FW supports. On P5 chips, some older FW sets max_tpa very low so we override it to prevent performance regressions with the older FW. Fixes: 79632e9ba386 ("bnxt_en: Expand bnxt_tpa_info struct to support 57500 chips.") Reviewed-by: Kalesh AP Reviewed-by: Colin Winegarden Reviewed-by: Rukhsana Ansari Signed-off-by: Michael Chan Signed-off-by: Pavan Chebbi Link: https://patch.msgid.link/20260504083611.1383776-3-pavan.chebbi@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 3db951d0c690..008c34cff7b4 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -3825,7 +3825,10 @@ static int bnxt_alloc_tpa_info(struct bnxt *bp) if (bp->flags & BNXT_FLAG_CHIP_P5_PLUS) { if (!bp->max_tpa_v2) return 0; - bp->max_tpa = max_t(u16, bp->max_tpa_v2, MAX_TPA_P5); + bp->max_tpa = min_t(u16, bp->max_tpa_v2, MAX_TPA_P5); + /* Older P5 FW sets max_tpa_v2 low by mistake except NPAR */ + if (bp->max_tpa <= 32 && BNXT_CHIP_P5(bp) && !BNXT_NPAR(bp)) + bp->max_tpa = MAX_TPA_P5; } for (i = 0; i < bp->rx_nr_rings; i++) { From 16517bc98a56004274472cc9949194cb4d2ad0b7 Mon Sep 17 00:00:00 2001 From: Kalesh AP Date: Mon, 4 May 2026 14:06:10 +0530 Subject: [PATCH 360/525] bnxt_en: Check return value of bnxt_hwrm_vnic_cfg When the bnxt RDMA driver is loaded, it calls bnxt_register_dev(). As part of this, driver sends HWRM_VNIC_CFG firmware command to configure the VNIC to operate in dual VNIC mode. Currently the driver ignores the result of this firmware command. The RDMA driver must know the result since it affects its functioning. Check return value of call to bnxt_hwrm_vnic_cfg() in bnxt_register_dev() and return failure on error. Fixes: a588e4580a7e ("bnxt_en: Add interface to support RDMA driver.") Reviewed-by: Michael Chan Signed-off-by: Kalesh AP Signed-off-by: Pavan Chebbi Link: https://patch.msgid.link/20260504083611.1383776-4-pavan.chebbi@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c index 052bf69cfa4c..5c751933da6a 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c @@ -175,8 +175,14 @@ int bnxt_register_dev(struct bnxt_en_dev *edev, ulp->handle = handle; rcu_assign_pointer(ulp->ulp_ops, ulp_ops); - if (test_bit(BNXT_STATE_OPEN, &bp->state)) - bnxt_hwrm_vnic_cfg(bp, &bp->vnic_info[BNXT_VNIC_DEFAULT]); + if (test_bit(BNXT_STATE_OPEN, &bp->state)) { + rc = bnxt_hwrm_vnic_cfg(bp, &bp->vnic_info[BNXT_VNIC_DEFAULT]); + if (rc) { + netdev_err(dev, "Failed to configure dual VNIC mode\n"); + RCU_INIT_POINTER(ulp->ulp_ops, NULL); + goto exit; + } + } edev->ulp_tbl->msix_requested = bnxt_get_ulp_msix_num(bp); From bd279e104e5f5400307d56116a36756b35ab345a Mon Sep 17 00:00:00 2001 From: Pavan Chebbi Date: Mon, 4 May 2026 14:06:11 +0530 Subject: [PATCH 361/525] bnxt_en: Use absolute target ns from ptp_clock_request There is no need to calculate the target PHC cycles required to make phase adjustment on the PPS OUT signal. This is because the application supplies absolute n_sec value in the future and is already the actual desired target value. Remove the unnecessary code. Fixes: 9e518f25802c ("bnxt_en: 1PPS functions to configure TSIO pins") Reviewed-by: Kalesh AP Cc: Richard Cochran Signed-off-by: Pavan Chebbi Reviewed-by: Vadim Fedorenko Tested-by: Vadim Fedorenko Link: https://patch.msgid.link/20260504083611.1383776-5-pavan.chebbi@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt_ptp.c | 29 ++++--------------- 1 file changed, 5 insertions(+), 24 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ptp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ptp.c index 53f336db4fcc..5d41dc1bc782 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ptp.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ptp.c @@ -419,31 +419,13 @@ void bnxt_ptp_reapply_pps(struct bnxt *bp) } } -static int bnxt_get_target_cycles(struct bnxt_ptp_cfg *ptp, u64 target_ns, - u64 *cycles_delta) -{ - u64 cycles_now; - u64 nsec_now, nsec_delta; - int rc; - - rc = bnxt_refclk_read(ptp->bp, NULL, &cycles_now); - if (rc) - return rc; - - nsec_now = bnxt_timecounter_cyc2time(ptp, cycles_now); - - nsec_delta = target_ns - nsec_now; - *cycles_delta = div64_u64(nsec_delta << ptp->cc.shift, ptp->cc.mult); - return 0; -} - static int bnxt_ptp_perout_cfg(struct bnxt_ptp_cfg *ptp, struct ptp_clock_request *rq) { struct hwrm_func_ptp_cfg_input *req; struct bnxt *bp = ptp->bp; struct timespec64 ts; - u64 target_ns, delta; + u64 target_ns; u16 enables; int rc; @@ -451,10 +433,6 @@ static int bnxt_ptp_perout_cfg(struct bnxt_ptp_cfg *ptp, ts.tv_nsec = rq->perout.start.nsec; target_ns = timespec64_to_ns(&ts); - rc = bnxt_get_target_cycles(ptp, target_ns, &delta); - if (rc) - return rc; - rc = hwrm_req_init(bp, req, HWRM_FUNC_PTP_CFG); if (rc) return rc; @@ -468,7 +446,10 @@ static int bnxt_ptp_perout_cfg(struct bnxt_ptp_cfg *ptp, req->ptp_freq_adj_dll_phase = 0; req->ptp_freq_adj_ext_period = cpu_to_le32(NSEC_PER_SEC); req->ptp_freq_adj_ext_up = 0; - req->ptp_freq_adj_ext_phase_lower = cpu_to_le32(delta); + req->ptp_freq_adj_ext_phase_lower = + cpu_to_le32(lower_32_bits(target_ns)); + req->ptp_freq_adj_ext_phase_upper = + cpu_to_le32(upper_32_bits(target_ns)); return hwrm_req_send(bp, req); } From f83e07b29246f468bc7c99f98ca1897843fa8167 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 4 May 2026 16:38:42 +0000 Subject: [PATCH 362/525] net/sched: sch_fq_codel: annotate data-races from fq_codel_dump_class_stats() fq_codel_dump_class_stats() acquires qdisc spinlock only when requested to follow flow->head chain. As we did in sch_cake recently, add the missing READ_ONCE()/WRITE_ONCE() annotations. Fixes: edb09eb17ed8 ("net: sched: do not acquire qdisc spinlock in qdisc/class stats dump") Signed-off-by: Eric Dumazet Reviewed-by: Jamal Hadi Salim Link: https://patch.msgid.link/20260504163842.1162001-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/sched/sch_fq_codel.c | 39 ++++++++++++++++++++------------------- 1 file changed, 20 insertions(+), 19 deletions(-) diff --git a/net/sched/sch_fq_codel.c b/net/sched/sch_fq_codel.c index 0664b2f2d6f2..24db54684e8a 100644 --- a/net/sched/sch_fq_codel.c +++ b/net/sched/sch_fq_codel.c @@ -117,7 +117,7 @@ static inline struct sk_buff *dequeue_head(struct fq_codel_flow *flow) { struct sk_buff *skb = flow->head; - flow->head = skb->next; + WRITE_ONCE(flow->head, skb->next); skb_mark_not_on_list(skb); return skb; } @@ -127,7 +127,7 @@ static inline void flow_queue_add(struct fq_codel_flow *flow, struct sk_buff *skb) { if (flow->head == NULL) - flow->head = skb; + WRITE_ONCE(flow->head, skb); else flow->tail->next = skb; flow->tail = skb; @@ -173,8 +173,8 @@ static unsigned int fq_codel_drop(struct Qdisc *sch, unsigned int max_packets, } while (++i < max_packets && len < threshold); /* Tell codel to increase its signal strength also */ - flow->cvars.count += i; - q->backlogs[idx] -= len; + WRITE_ONCE(flow->cvars.count, flow->cvars.count + i); + WRITE_ONCE(q->backlogs[idx], q->backlogs[idx] - len); q->memory_usage -= mem; sch->qstats.drops += i; sch->qstats.backlog -= len; @@ -204,13 +204,13 @@ static int fq_codel_enqueue(struct sk_buff *skb, struct Qdisc *sch, codel_set_enqueue_time(skb); flow = &q->flows[idx]; flow_queue_add(flow, skb); - q->backlogs[idx] += qdisc_pkt_len(skb); + WRITE_ONCE(q->backlogs[idx], q->backlogs[idx] + qdisc_pkt_len(skb)); qdisc_qstats_backlog_inc(sch, skb); if (list_empty(&flow->flowchain)) { list_add_tail(&flow->flowchain, &q->new_flows); q->new_flow_count++; - flow->deficit = q->quantum; + WRITE_ONCE(flow->deficit, q->quantum); } get_codel_cb(skb)->mem_usage = skb->truesize; q->memory_usage += get_codel_cb(skb)->mem_usage; @@ -263,7 +263,8 @@ static struct sk_buff *dequeue_func(struct codel_vars *vars, void *ctx) flow = container_of(vars, struct fq_codel_flow, cvars); if (flow->head) { skb = dequeue_head(flow); - q->backlogs[flow - q->flows] -= qdisc_pkt_len(skb); + WRITE_ONCE(q->backlogs[flow - q->flows], + q->backlogs[flow - q->flows] - qdisc_pkt_len(skb)); q->memory_usage -= get_codel_cb(skb)->mem_usage; sch->q.qlen--; sch->qstats.backlog -= qdisc_pkt_len(skb); @@ -296,7 +297,7 @@ static struct sk_buff *fq_codel_dequeue(struct Qdisc *sch) flow = list_first_entry(head, struct fq_codel_flow, flowchain); if (flow->deficit <= 0) { - flow->deficit += q->quantum; + WRITE_ONCE(flow->deficit, flow->deficit + q->quantum); list_move_tail(&flow->flowchain, &q->old_flows); goto begin; } @@ -314,7 +315,7 @@ static struct sk_buff *fq_codel_dequeue(struct Qdisc *sch) goto begin; } qdisc_bstats_update(sch, skb); - flow->deficit -= qdisc_pkt_len(skb); + WRITE_ONCE(flow->deficit, flow->deficit - qdisc_pkt_len(skb)); if (q->cstats.drop_count) { qdisc_tree_reduce_backlog(sch, q->cstats.drop_count, @@ -328,7 +329,7 @@ static struct sk_buff *fq_codel_dequeue(struct Qdisc *sch) static void fq_codel_flow_purge(struct fq_codel_flow *flow) { rtnl_kfree_skbs(flow->head, flow->tail); - flow->head = NULL; + WRITE_ONCE(flow->head, NULL); } static void fq_codel_reset(struct Qdisc *sch) @@ -656,21 +657,21 @@ static int fq_codel_dump_class_stats(struct Qdisc *sch, unsigned long cl, memset(&xstats, 0, sizeof(xstats)); xstats.type = TCA_FQ_CODEL_XSTATS_CLASS; - xstats.class_stats.deficit = flow->deficit; + xstats.class_stats.deficit = READ_ONCE(flow->deficit); xstats.class_stats.ldelay = - codel_time_to_us(flow->cvars.ldelay); - xstats.class_stats.count = flow->cvars.count; - xstats.class_stats.lastcount = flow->cvars.lastcount; - xstats.class_stats.dropping = flow->cvars.dropping; - if (flow->cvars.dropping) { - codel_tdiff_t delta = flow->cvars.drop_next - + codel_time_to_us(READ_ONCE(flow->cvars.ldelay)); + xstats.class_stats.count = READ_ONCE(flow->cvars.count); + xstats.class_stats.lastcount = READ_ONCE(flow->cvars.lastcount); + xstats.class_stats.dropping = READ_ONCE(flow->cvars.dropping); + if (xstats.class_stats.dropping) { + codel_tdiff_t delta = READ_ONCE(flow->cvars.drop_next) - codel_get_time(); xstats.class_stats.drop_next = (delta >= 0) ? codel_time_to_us(delta) : -codel_time_to_us(-delta); } - if (flow->head) { + if (READ_ONCE(flow->head)) { sch_tree_lock(sch); skb = flow->head; while (skb) { @@ -679,7 +680,7 @@ static int fq_codel_dump_class_stats(struct Qdisc *sch, unsigned long cl, } sch_tree_unlock(sch); } - qs.backlog = q->backlogs[idx]; + qs.backlog = READ_ONCE(q->backlogs[idx]); qs.drops = 0; } if (gnet_stats_copy_queue(d, NULL, &qs, qs.qlen) < 0) From 54d54f33813d7911555226b4220737177a2ba8d6 Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Sat, 14 Mar 2026 18:54:30 +0530 Subject: [PATCH 363/525] powerpc/pseries/htmdump: Free the global buffers in htmdump module exit htmdump modules uses global memory buffers to capture details like capabilities, status of specified HTM, read the trace buffer. These are initialized during module init and hence needs to be freed in module exit. Patch adds freeing of the memory in module exit. The change also includes minor clean up for the variable name. The read call back for the debugfs interface file saves filp->private_data to local variable name which is same as global variable name for the memory buffers. Rename these local variable names. Signed-off-by: Athira Rajeev Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260314132432.25581-1-atrajeev@linux.ibm.com --- arch/powerpc/platforms/pseries/htmdump.c | 31 +++++++++++++----------- 1 file changed, 17 insertions(+), 14 deletions(-) diff --git a/arch/powerpc/platforms/pseries/htmdump.c b/arch/powerpc/platforms/pseries/htmdump.c index 742ec52c9d4d..93f0cc2dc7fb 100644 --- a/arch/powerpc/platforms/pseries/htmdump.c +++ b/arch/powerpc/platforms/pseries/htmdump.c @@ -86,7 +86,7 @@ static ssize_t htm_return_check(long rc) static ssize_t htmdump_read(struct file *filp, char __user *ubuf, size_t count, loff_t *ppos) { - void *htm_buf = filp->private_data; + void *htm_buf_data = filp->private_data; unsigned long page, read_size, available; loff_t offset; long rc, ret; @@ -100,7 +100,7 @@ static ssize_t htmdump_read(struct file *filp, char __user *ubuf, * - last three values are address, size and offset */ rc = htm_hcall_wrapper(htmflags, nodeindex, nodalchipindex, coreindexonchip, - htmtype, H_HTM_OP_DUMP_DATA, virt_to_phys(htm_buf), + htmtype, H_HTM_OP_DUMP_DATA, virt_to_phys(htm_buf_data), PAGE_SIZE, page); ret = htm_return_check(rc); @@ -112,7 +112,7 @@ static ssize_t htmdump_read(struct file *filp, char __user *ubuf, available = PAGE_SIZE; read_size = min(count, available); *ppos += read_size; - return simple_read_from_buffer(ubuf, count, &offset, htm_buf, available); + return simple_read_from_buffer(ubuf, count, &offset, htm_buf_data, available); } static const struct file_operations htmdump_fops = { @@ -226,7 +226,7 @@ static int htmstart_get(void *data, u64 *val) static ssize_t htmstatus_read(struct file *filp, char __user *ubuf, size_t count, loff_t *ppos) { - void *htm_status_buf = filp->private_data; + void *htm_status_data = filp->private_data; long rc, ret; u64 *num_entries; u64 to_copy; @@ -238,7 +238,7 @@ static ssize_t htmstatus_read(struct file *filp, char __user *ubuf, * - last three values as addr, size and offset */ rc = htm_hcall_wrapper(htmflags, nodeindex, nodalchipindex, coreindexonchip, - htmtype, H_HTM_OP_STATUS, virt_to_phys(htm_status_buf), + htmtype, H_HTM_OP_STATUS, virt_to_phys(htm_status_data), PAGE_SIZE, 0); ret = htm_return_check(rc); @@ -255,13 +255,13 @@ static ssize_t htmstatus_read(struct file *filp, char __user *ubuf, * So total count to copy is: * 32 bytes (for first 7 fields) + (number of HTM entries * entry size) */ - num_entries = htm_status_buf + 0x10; + num_entries = htm_status_data + 0x10; if (htmtype == 0x2) htmstatus_flag = 0x8; else htmstatus_flag = 0x6; to_copy = 32 + (be64_to_cpu(*num_entries) * htmstatus_flag); - return simple_read_from_buffer(ubuf, count, ppos, htm_status_buf, to_copy); + return simple_read_from_buffer(ubuf, count, ppos, htm_status_data, to_copy); } static const struct file_operations htmstatus_fops = { @@ -273,7 +273,7 @@ static const struct file_operations htmstatus_fops = { static ssize_t htminfo_read(struct file *filp, char __user *ubuf, size_t count, loff_t *ppos) { - void *htm_info_buf = filp->private_data; + void *htm_info_data = filp->private_data; long rc, ret; u64 *num_entries; u64 to_copy; @@ -284,7 +284,7 @@ static ssize_t htminfo_read(struct file *filp, char __user *ubuf, * - last three values as addr, size and offset */ rc = htm_hcall_wrapper(htmflags, nodeindex, nodalchipindex, coreindexonchip, - htmtype, H_HTM_OP_DUMP_SYSPROC_CONF, virt_to_phys(htm_info_buf), + htmtype, H_HTM_OP_DUMP_SYSPROC_CONF, virt_to_phys(htm_info_data), PAGE_SIZE, 0); ret = htm_return_check(rc); @@ -301,15 +301,15 @@ static ssize_t htminfo_read(struct file *filp, char __user *ubuf, * So total count to copy is: * 32 bytes (for first 5 fields) + (number of HTM entries * entry size) */ - num_entries = htm_info_buf + 0x10; + num_entries = htm_info_data + 0x10; to_copy = 32 + (be64_to_cpu(*num_entries) * 16); - return simple_read_from_buffer(ubuf, count, ppos, htm_info_buf, to_copy); + return simple_read_from_buffer(ubuf, count, ppos, htm_info_data, to_copy); } static ssize_t htmcaps_read(struct file *filp, char __user *ubuf, size_t count, loff_t *ppos) { - void *htm_caps_buf = filp->private_data; + void *htm_caps_data = filp->private_data; long rc, ret; /* @@ -319,7 +319,7 @@ static ssize_t htmcaps_read(struct file *filp, char __user *ubuf, * and zero */ rc = htm_hcall_wrapper(htmflags, nodeindex, nodalchipindex, coreindexonchip, - htmtype, H_HTM_OP_CAPABILITIES, virt_to_phys(htm_caps_buf), + htmtype, H_HTM_OP_CAPABILITIES, virt_to_phys(htm_caps_data), 0x80, 0); ret = htm_return_check(rc); @@ -328,7 +328,7 @@ static ssize_t htmcaps_read(struct file *filp, char __user *ubuf, return ret; } - return simple_read_from_buffer(ubuf, count, ppos, htm_caps_buf, 0x80); + return simple_read_from_buffer(ubuf, count, ppos, htm_caps_data, 0x80); } static const struct file_operations htminfo_fops = { @@ -482,6 +482,9 @@ static void __exit htmdump_exit(void) { debugfs_remove_recursive(htmdump_debugfs_dir); kfree(htm_buf); + kfree(htm_status_buf); + kfree(htm_info_buf); + kfree(htm_caps_buf); } module_init(htmdump_init); From 4ec6ade73f1626a74d46e61ff247cf2b1b96446b Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Sat, 14 Mar 2026 18:54:31 +0530 Subject: [PATCH 364/525] powerpc/pseries/htmdump: Fix the offset value used in processor configuration dump H_HTM call is invoked using three parameters specifying the address of the buffer, size of the buffer and offset where to read from. offset used was always zero. "offset" is value from output buffer header that points to next entry to dump. zero is the first entry to dump. next entry is read from the output bufferbyte offset 0x8. Update htminfo_read() function to use right offset. Return when offset points to -1 Fixes: dea7384e14e7 ("powerpc/pseries/htmdump: Add htm info support to htmdump module") Signed-off-by: Athira Rajeev Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260314132432.25581-2-atrajeev@linux.ibm.com --- arch/powerpc/platforms/pseries/htmdump.c | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/platforms/pseries/htmdump.c b/arch/powerpc/platforms/pseries/htmdump.c index 93f0cc2dc7fb..34978a794eba 100644 --- a/arch/powerpc/platforms/pseries/htmdump.c +++ b/arch/powerpc/platforms/pseries/htmdump.c @@ -277,15 +277,26 @@ static ssize_t htminfo_read(struct file *filp, char __user *ubuf, long rc, ret; u64 *num_entries; u64 to_copy; + loff_t offset = 0; + u64 info_offset = 0; /* * Invoke H_HTM call with: * - operation as htm status (H_HTM_OP_STATUS) * - last three values as addr, size and offset + * "offset" is value from output buffer header + * that points to next entry to dump. 0 is the first + * entry to dump. next entry is read from the output + * bufferbyte offset 0x8. */ + if (*ppos) { + info_offset = *(u64 *)(htm_info_data + 0x8); + if (info_offset == -1) + return 0; + } rc = htm_hcall_wrapper(htmflags, nodeindex, nodalchipindex, coreindexonchip, htmtype, H_HTM_OP_DUMP_SYSPROC_CONF, virt_to_phys(htm_info_data), - PAGE_SIZE, 0); + PAGE_SIZE, be64_to_cpu(info_offset)); ret = htm_return_check(rc); if (ret <= 0) { @@ -303,7 +314,9 @@ static ssize_t htminfo_read(struct file *filp, char __user *ubuf, */ num_entries = htm_info_data + 0x10; to_copy = 32 + (be64_to_cpu(*num_entries) * 16); - return simple_read_from_buffer(ubuf, count, ppos, htm_info_data, to_copy); + + *ppos += to_copy; + return simple_read_from_buffer(ubuf, count, &offset, htm_info_data, to_copy); } static ssize_t htmcaps_read(struct file *filp, char __user *ubuf, From 0031424cbc0a6eafc56a8f9201a4a69d3649981e Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Sat, 14 Mar 2026 18:54:32 +0530 Subject: [PATCH 365/525] powerpc/pseries/htmdump: Fix the offset value used in htm status dump H_HTM call is invoked using three parameters specifying the address of the buffer, size of the buffer and offset where to read from. offset used was always zero. "offset" is value from output buffer header that points to next entry to dump. zero is the first entry to dump. next entry is read from the output bufferbyte offset 0x8. Update htmstatus_read() function to use right offset. Return when offset points to -1 Fixes: 627cf584f4c3 ("powerpc/pseries/htmdump: Add htm status support to htmdump module") Signed-off-by: Athira Rajeev Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260314132432.25581-3-atrajeev@linux.ibm.com --- arch/powerpc/platforms/pseries/htmdump.c | 19 ++++++++++++++++--- 1 file changed, 16 insertions(+), 3 deletions(-) diff --git a/arch/powerpc/platforms/pseries/htmdump.c b/arch/powerpc/platforms/pseries/htmdump.c index 34978a794eba..76444c6c5cc1 100644 --- a/arch/powerpc/platforms/pseries/htmdump.c +++ b/arch/powerpc/platforms/pseries/htmdump.c @@ -231,15 +231,26 @@ static ssize_t htmstatus_read(struct file *filp, char __user *ubuf, u64 *num_entries; u64 to_copy; int htmstatus_flag; + loff_t offset = 0; + u64 status_offset = 0; /* * Invoke H_HTM call with: * - operation as htm status (H_HTM_OP_STATUS) - * - last three values as addr, size and offset + * - last three values as addr, size and offset. + * "offset" is value from output buffer header + * that points to next entry to dump. 0 is the first + * entry to dump. next entry is read from the output + * bufferbyte offset 0x8. */ + if (*ppos) { + status_offset = *(u64 *)(htm_status_data + 0x8); + if (status_offset == -1) + return 0; + } rc = htm_hcall_wrapper(htmflags, nodeindex, nodalchipindex, coreindexonchip, htmtype, H_HTM_OP_STATUS, virt_to_phys(htm_status_data), - PAGE_SIZE, 0); + PAGE_SIZE, be64_to_cpu(status_offset)); ret = htm_return_check(rc); if (ret <= 0) { @@ -261,7 +272,9 @@ static ssize_t htmstatus_read(struct file *filp, char __user *ubuf, else htmstatus_flag = 0x6; to_copy = 32 + (be64_to_cpu(*num_entries) * htmstatus_flag); - return simple_read_from_buffer(ubuf, count, ppos, htm_status_data, to_copy); + *ppos += to_copy; + + return simple_read_from_buffer(ubuf, count, &offset, htm_status_data, to_copy); } static const struct file_operations htmstatus_fops = { From 0342545a29bc2edfbe132c68c68b51c92a12444c Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Sat, 14 Mar 2026 18:59:53 +0530 Subject: [PATCH 366/525] powerpc/pseries/htmdump: Add memory configuration dump support to htmdump module H_HTM (Hardware Trace Macro) hypervisor call has capability to capture SystemMemory Configuration. This information helps to understand the address mapping for the partitions in the system. Support dumping system memory configuration from Hardware Trace Macro (HTM) function via debugfs interface. Under debugfs folder "/sys/kernel/debug/powerpc/htmdump", add file "htmsystem_mem". The interface allows only read of this file which will present the content of HTM buffer from the hcall. The 16th offset of HTM buffer has value for the number of entries for array of processors. Use this information to copy data to the debugfs file Signed-off-by: Athira Rajeev Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260314132953.27269-1-atrajeev@linux.ibm.com --- arch/powerpc/platforms/pseries/htmdump.c | 70 ++++++++++++++++++++++++ 1 file changed, 70 insertions(+) diff --git a/arch/powerpc/platforms/pseries/htmdump.c b/arch/powerpc/platforms/pseries/htmdump.c index 76444c6c5cc1..489a80e87082 100644 --- a/arch/powerpc/platforms/pseries/htmdump.c +++ b/arch/powerpc/platforms/pseries/htmdump.c @@ -16,6 +16,7 @@ static void *htm_buf; static void *htm_status_buf; static void *htm_info_buf; static void *htm_caps_buf; +static void *htm_mem_buf; static u32 nodeindex; static u32 nodalchipindex; static u32 coreindexonchip; @@ -115,12 +116,72 @@ static ssize_t htmdump_read(struct file *filp, char __user *ubuf, return simple_read_from_buffer(ubuf, count, &offset, htm_buf_data, available); } +static ssize_t htmsystem_mem_read(struct file *filp, char __user *ubuf, + size_t count, loff_t *ppos) +{ + void *htm_mem_data = filp->private_data; + long rc, ret; + u64 *num_entries; + u64 to_copy = 0; + loff_t offset = 0; + u64 mem_offset = 0; + + /* + * Invoke H_HTM call with: + * - operation as htm status (H_HTM_OP_STATUS) + * - last three values as addr, size and offset. "offset" + * is value from output buffer header that points to next + * entry to dump. 0 is the first entry to dump. next entry + * is read from the output bufferbyte offset 0x8. + * + * When first time hcall is invoked, mem_offset should be + * zero because zero is the first entry. + * In the next hcall, offset of next entry to read from is + * picked from output buffer header itself. So don't fill + * mem_offset for first read. + * + * If there is no further data to read in next iteration, + * offset value from output buffer header will point to -1. + */ + if (*ppos) { + mem_offset = *(u64 *)(htm_mem_data + 0x8); + if (mem_offset == -1) + return 0; + } + rc = htm_hcall_wrapper(htmflags, nodeindex, nodalchipindex, coreindexonchip, + htmtype, H_HTM_OP_DUMP_SYSMEM_CONF, virt_to_phys(htm_mem_data), + PAGE_SIZE, be64_to_cpu(mem_offset)); + ret = htm_return_check(rc); + if (ret <= 0) { + pr_debug("H_HTM hcall returned for op: H_HTM_OP_DUMP_SYSMEM_CONF with hcall returning %ld\n", ret); + return ret; + } + + /* + * HTM system mem buffer, start of buffer + 0x10 gives the + * number of HTM entries in the buffer. + * So total count to copy is: + * 32 bytes (for first 5 fields) + (number of HTM entries * entry size) + */ + num_entries = htm_mem_data + 0x10; + to_copy = 32 + (be64_to_cpu(*num_entries) * 32); + + *ppos += to_copy; + return simple_read_from_buffer(ubuf, count, &offset, htm_mem_data, to_copy); +} + static const struct file_operations htmdump_fops = { .llseek = NULL, .read = htmdump_read, .open = simple_open, }; +static const struct file_operations htmsystem_mem_fops = { + .llseek = NULL, + .read = htmsystem_mem_read, + .open = simple_open, +}; + static int htmconfigure_set(void *data, u64 val) { long rc, ret; @@ -483,9 +544,17 @@ static int htmdump_init_debugfs(void) return -ENOMEM; } + /* Memory to present HTM system memory configuration */ + htm_mem_buf = kmalloc(PAGE_SIZE, GFP_KERNEL); + if (!htm_mem_buf) { + pr_err("Failed to allocate htm mem buf\n"); + return -ENOMEM; + } + debugfs_create_file("htmstatus", 0400, htmdump_debugfs_dir, htm_status_buf, &htmstatus_fops); debugfs_create_file("htminfo", 0400, htmdump_debugfs_dir, htm_info_buf, &htminfo_fops); debugfs_create_file("htmcaps", 0400, htmdump_debugfs_dir, htm_caps_buf, &htmcaps_fops); + debugfs_create_file("htmsystem_mem", 0400, htmdump_debugfs_dir, htm_mem_buf, &htmsystem_mem_fops); return 0; } @@ -511,6 +580,7 @@ static void __exit htmdump_exit(void) kfree(htm_status_buf); kfree(htm_info_buf); kfree(htm_caps_buf); + kfree(htm_mem_buf); } module_init(htmdump_init); From 7a4f0846ee6cc8cf44ae0046ed42e3259d1dd45b Mon Sep 17 00:00:00 2001 From: "Ritesh Harjani (IBM)" Date: Fri, 1 May 2026 09:41:40 +0530 Subject: [PATCH 367/525] pseries/papr-hvpipe: Fix race with interrupt handler While executing ->ioctl handler or ->release handler, if an interrupt fires on the same cpu, then we can enter into a deadlock. This patch fixes both these handlers to take spin_lock_irq{save|restore} versions of the lock to prevent this deadlock. Cc: stable@vger.kernel.org Fixes: 814ef095f12c9 ("powerpc/pseries: Add papr-hvpipe char driver for HVPIPE interfaces") Signed-off-by: Ritesh Harjani (IBM) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/e4ed435c44fc191f2eb23c7907ba6f72f193e6aa.1777606826.git.ritesh.list@gmail.com --- arch/powerpc/platforms/pseries/papr-hvpipe.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/arch/powerpc/platforms/pseries/papr-hvpipe.c b/arch/powerpc/platforms/pseries/papr-hvpipe.c index 14ae480d060a..c41d45e1986d 100644 --- a/arch/powerpc/platforms/pseries/papr-hvpipe.c +++ b/arch/powerpc/platforms/pseries/papr-hvpipe.c @@ -444,13 +444,14 @@ static int papr_hvpipe_handle_release(struct inode *inode, struct file *file) { struct hvpipe_source_info *src_info; + unsigned long flags; /* * Hold the lock, remove source from src_list, reset the * hvpipe status and release the lock to prevent any race * with message event IRQ. */ - spin_lock(&hvpipe_src_list_lock); + spin_lock_irqsave(&hvpipe_src_list_lock, flags); src_info = file->private_data; list_del(&src_info->list); file->private_data = NULL; @@ -461,10 +462,10 @@ static int papr_hvpipe_handle_release(struct inode *inode, */ if (src_info->hvpipe_status & HVPIPE_MSG_AVAILABLE) { src_info->hvpipe_status = 0; - spin_unlock(&hvpipe_src_list_lock); + spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); hvpipe_rtas_recv_msg(NULL, 0); } else - spin_unlock(&hvpipe_src_list_lock); + spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); kfree(src_info); return 0; @@ -480,20 +481,21 @@ static const struct file_operations papr_hvpipe_handle_ops = { static int papr_hvpipe_dev_create_handle(u32 srcID) { struct hvpipe_source_info *src_info __free(kfree) = NULL; + unsigned long flags; - spin_lock(&hvpipe_src_list_lock); + spin_lock_irqsave(&hvpipe_src_list_lock, flags); /* * Do not allow more than one process communicates with * each source. */ src_info = hvpipe_find_source(srcID); if (src_info) { - spin_unlock(&hvpipe_src_list_lock); + spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); pr_err("pid(%d) is already using the source(%d)\n", src_info->tsk->pid, srcID); return -EALREADY; } - spin_unlock(&hvpipe_src_list_lock); + spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); src_info = kzalloc_obj(*src_info, GFP_KERNEL_ACCOUNT); if (!src_info) @@ -510,18 +512,18 @@ static int papr_hvpipe_dev_create_handle(u32 srcID) return fdf.err; retain_and_null_ptr(src_info); - spin_lock(&hvpipe_src_list_lock); + spin_lock_irqsave(&hvpipe_src_list_lock, flags); /* * If two processes are executing ioctl() for the same * source ID concurrently, prevent the second process to * acquire FD. */ if (hvpipe_find_source(srcID)) { - spin_unlock(&hvpipe_src_list_lock); + spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); return -EALREADY; } list_add(&src_info->list, &hvpipe_src_list); - spin_unlock(&hvpipe_src_list_lock); + spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); return fd_publish(fdf); } From cefeed44296261173a806bef988b26bc565da4be Mon Sep 17 00:00:00 2001 From: "Ritesh Harjani (IBM)" Date: Fri, 1 May 2026 09:41:41 +0530 Subject: [PATCH 368/525] pseries/papr-hvpipe: Prevent kernel stack memory leak to userspace The hdr variable is allocated on the stack and only hdr.version and hdr.flags are initialized explicitly. Because the struct papr_hvpipe_hdr contains reserved padding bytes (reserved[3] and reserved2[40]), these could leak the uninitialized bytes to userspace after copy_to_user(). This patch fixes that by initializing the whole struct to 0. Cc: stable@vger.kernel.org Fixes: cebdb522fd3ed ("powerpc/pseries: Receive payload with ibm,receive-hvpipe-msg RTAS") Signed-off-by: Ritesh Harjani (IBM) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/7bfe03b65a282c856ed8182d1871bb973c0b78f2.1777606826.git.ritesh.list@gmail.com --- arch/powerpc/platforms/pseries/papr-hvpipe.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/platforms/pseries/papr-hvpipe.c b/arch/powerpc/platforms/pseries/papr-hvpipe.c index c41d45e1986d..3392874ebdf6 100644 --- a/arch/powerpc/platforms/pseries/papr-hvpipe.c +++ b/arch/powerpc/platforms/pseries/papr-hvpipe.c @@ -327,7 +327,7 @@ static ssize_t papr_hvpipe_handle_read(struct file *file, { struct hvpipe_source_info *src_info = file->private_data; - struct papr_hvpipe_hdr hdr; + struct papr_hvpipe_hdr hdr = {}; long ret; /* From 1b9f7aafa44f5ce852c00509104d10fd9eb0f402 Mon Sep 17 00:00:00 2001 From: "Ritesh Harjani (IBM)" Date: Fri, 1 May 2026 09:41:42 +0530 Subject: [PATCH 369/525] pseries/papr-hvpipe: Fix null ptr deref in papr_hvpipe_dev_create_handle() commit 6d3789d347a7 ("papr-hvpipe: convert papr_hvpipe_dev_create_handle() to FD_PREPARE()"), changed the create handle to FD_PREPARE(), but it caused kernel null-ptr-deref because after call to retain_and_null_ptr(src_info), src_info is re-used for adding it to the global list. Getting the following kernel panic in papr_hvpipe_dev_create_handle() when trying to add src_info to the list. Kernel attempted to write user page (0) - exploit attempt? (uid: 0) BUG: Kernel NULL pointer dereference on write at 0x00000000 Faulting instruction address: 0xc0000000001b44a0 Oops: Kernel access of bad area, sig: 11 [#1] ... Call Trace: papr_hvpipe_dev_ioctl+0x1f4/0x48c (unreliable) sys_ioctl+0x528/0x1064 system_call_exception+0x128/0x360 system_call_vectored_common+0x15c/0x2ec Now, the error handling with FD_PREPARE's file cleanup and __free(kfree) auto cleanup is getting too convoluted. This is mainly because we need to ensure only 1 user get the srcID handle. To simplify this, we allocate prepare the src_info in the beginning and add it to the global list under a spinlock after checking that no duplicates exist. This simplify the error handling where if the FD_ADD fails, we can simply remove the src_info from the list and consume any pending msg in hvpipe to be cleared, after src_info became visible in the global list. Cc: stable@vger.kernel.org Fixes: 6d3789d347a7 ("papr-hvpipe: convert papr_hvpipe_dev_create_handle() to FD_PREPARE()") Reported-by: Haren Myneni Signed-off-by: Ritesh Harjani (IBM) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/31ad94bc89d44156ee700c5bd006cb47a748e3cb.1777606826.git.ritesh.list@gmail.com --- arch/powerpc/platforms/pseries/papr-hvpipe.c | 57 ++++++++++---------- 1 file changed, 30 insertions(+), 27 deletions(-) diff --git a/arch/powerpc/platforms/pseries/papr-hvpipe.c b/arch/powerpc/platforms/pseries/papr-hvpipe.c index 3392874ebdf6..402781299497 100644 --- a/arch/powerpc/platforms/pseries/papr-hvpipe.c +++ b/arch/powerpc/platforms/pseries/papr-hvpipe.c @@ -480,23 +480,10 @@ static const struct file_operations papr_hvpipe_handle_ops = { static int papr_hvpipe_dev_create_handle(u32 srcID) { - struct hvpipe_source_info *src_info __free(kfree) = NULL; + struct hvpipe_source_info *src_info; + int fd; unsigned long flags; - spin_lock_irqsave(&hvpipe_src_list_lock, flags); - /* - * Do not allow more than one process communicates with - * each source. - */ - src_info = hvpipe_find_source(srcID); - if (src_info) { - spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); - pr_err("pid(%d) is already using the source(%d)\n", - src_info->tsk->pid, srcID); - return -EALREADY; - } - spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); - src_info = kzalloc_obj(*src_info, GFP_KERNEL_ACCOUNT); if (!src_info) return -ENOMEM; @@ -505,26 +492,42 @@ static int papr_hvpipe_dev_create_handle(u32 srcID) src_info->tsk = current; init_waitqueue_head(&src_info->recv_wqh); - FD_PREPARE(fdf, O_RDONLY | O_CLOEXEC, - anon_inode_getfile("[papr-hvpipe]", &papr_hvpipe_handle_ops, - (void *)src_info, O_RDWR)); - if (fdf.err) - return fdf.err; - - retain_and_null_ptr(src_info); - spin_lock_irqsave(&hvpipe_src_list_lock, flags); /* - * If two processes are executing ioctl() for the same - * source ID concurrently, prevent the second process to - * acquire FD. + * Do not allow more than one process communicates with + * each source. */ + spin_lock_irqsave(&hvpipe_src_list_lock, flags); if (hvpipe_find_source(srcID)) { spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); + pr_err("pid(%d) could not get the source(%d)\n", + src_info->tsk->pid, srcID); + kfree(src_info); return -EALREADY; } list_add(&src_info->list, &hvpipe_src_list); spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); - return fd_publish(fdf); + + fd = FD_ADD(O_RDONLY | O_CLOEXEC, + anon_inode_getfile("[papr-hvpipe]", &papr_hvpipe_handle_ops, + (void *)src_info, O_RDWR)); + if (fd < 0) { + spin_lock_irqsave(&hvpipe_src_list_lock, flags); + list_del(&src_info->list); + spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); + /* + * if we fail to add FD, that means no userspace program is + * polling. In that case if there is a msg pending because the + * interrupt was fired after the src_info was added to the + * global list, then let's consume it here, to unblock the + * hvpipe + */ + if (src_info->hvpipe_status & HVPIPE_MSG_AVAILABLE) + hvpipe_rtas_recv_msg(NULL, 0); + kfree(src_info); + return fd; + } + + return fd; } /* From 713e468cdbc2277db6ce949c32c1acbd83501733 Mon Sep 17 00:00:00 2001 From: "Ritesh Harjani (IBM)" Date: Fri, 1 May 2026 09:41:43 +0530 Subject: [PATCH 370/525] pseries/papr-hvpipe: Fix & simplify error handling in papr_hvpipe_init() Remove such 3 levels of nesting patterns to check success return values from function calls. ret = enable_hvpipe_IRQ() if (!ret) ret = set_hvpipe_sys_param(1) if (!ret) ret = misc_register() Instead just bail out to "out*:" labels, in case of any error. This simplifies the init flow. While at it let's also fix the following error handling logic: We have already enabled interrupt sources and enabled hvpipe to received interrupts, if misc_register() fails, we will destroy the workqueue, but the HMC might send us a msg via hvpipe which will call, queue work on the workqueue which might be destroyed. So instead, let's reverse the order of enabling set_hvpipe_sys_param(1) and in case of an error let's remove the misc dev by calling misc_deregister(). Cc: stable@vger.kernel.org Fixes: 39a08a4f94980 ("powerpc/pseries: Enable hvpipe with ibm,set-system-parameter RTAS") Signed-off-by: Ritesh Harjani (IBM) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/f2141eafb80e7780395e03aa9a22e8a37be80513.1777606826.git.ritesh.list@gmail.com --- arch/powerpc/platforms/pseries/papr-hvpipe.c | 28 ++++++++++++-------- 1 file changed, 17 insertions(+), 11 deletions(-) diff --git a/arch/powerpc/platforms/pseries/papr-hvpipe.c b/arch/powerpc/platforms/pseries/papr-hvpipe.c index 402781299497..800649f309a5 100644 --- a/arch/powerpc/platforms/pseries/papr-hvpipe.c +++ b/arch/powerpc/platforms/pseries/papr-hvpipe.c @@ -780,23 +780,29 @@ static int __init papr_hvpipe_init(void) } ret = enable_hvpipe_IRQ(); - if (!ret) { - ret = set_hvpipe_sys_param(1); - if (!ret) - ret = misc_register(&papr_hvpipe_dev); - } + if (ret) + goto out_wq; - if (!ret) { - pr_info("hvpipe feature is enabled\n"); - hvpipe_feature = true; - return 0; - } + ret = misc_register(&papr_hvpipe_dev); + if (ret) + goto out_wq; - pr_err("hvpipe feature is not enabled %d\n", ret); + ret = set_hvpipe_sys_param(1); + if (ret) + goto out_misc; + + pr_info("hvpipe feature is enabled\n"); + hvpipe_feature = true; + return 0; + +out_misc: + misc_deregister(&papr_hvpipe_dev); +out_wq: destroy_workqueue(papr_hvpipe_wq); out: kfree(papr_hvpipe_work); papr_hvpipe_work = NULL; + pr_err("hvpipe feature is not enabled %d\n", ret); return ret; } machine_device_initcall(pseries, papr_hvpipe_init); From d48654bd8b1a75f662e224d257db54de475120dc Mon Sep 17 00:00:00 2001 From: "Ritesh Harjani (IBM)" Date: Fri, 1 May 2026 09:41:44 +0530 Subject: [PATCH 371/525] pseries/papr-hvpipe: Fix the usage of copy_to_user() copy_to_user() return bytes_not_copied to the user buffer. If there was an error writing bytes into the user buffer, i.e. if copy_to_user returns a non-zero value, then we should simply return -EFAULT from the ->read() call. Otherwise, in the non-patched version, we may end up mixing "bytes_not_copied + bytes_copied (HVPIPE_HDR_LEN)" as the return value to the user in ->read() call Also let's make sure we clear the hvpipe_status flag, if we have consumed the hvpipe msg by making the rtas call. ret = -EFAULT means copy_to_user has failed but that still means that the msg was read from the hvpipe, hence for both cases, success & -EFAULT, we should clear the HVPIPE_MSG_AVAILABLE flag in hvpipe_status. Cc: stable@vger.kernel.org Fixes: cebdb522fd3edd1 ("powerpc/pseries: Receive payload with ibm,receive-hvpipe-msg RTAS") Signed-off-by: Ritesh Harjani (IBM) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/8fda3212a1ad48879c174e92f67472d9b9f1c3b7.1777606826.git.ritesh.list@gmail.com --- arch/powerpc/platforms/pseries/papr-hvpipe.c | 23 ++++++++++++-------- 1 file changed, 14 insertions(+), 9 deletions(-) diff --git a/arch/powerpc/platforms/pseries/papr-hvpipe.c b/arch/powerpc/platforms/pseries/papr-hvpipe.c index 800649f309a5..c007560d2d8c 100644 --- a/arch/powerpc/platforms/pseries/papr-hvpipe.c +++ b/arch/powerpc/platforms/pseries/papr-hvpipe.c @@ -206,10 +206,11 @@ static int hvpipe_rtas_recv_msg(char __user *buf, int size) bytes_written, size); bytes_written = size; } - ret = copy_to_user(buf, + if (copy_to_user(buf, rtas_work_area_raw_buf(work_area), - bytes_written); - if (!ret) + bytes_written)) + ret = -EFAULT; + else ret = bytes_written; } } else { @@ -328,7 +329,7 @@ static ssize_t papr_hvpipe_handle_read(struct file *file, struct hvpipe_source_info *src_info = file->private_data; struct papr_hvpipe_hdr hdr = {}; - long ret; + ssize_t ret = 0; /* * Return -ENXIO during migration @@ -376,7 +377,7 @@ static ssize_t papr_hvpipe_handle_read(struct file *file, ret = copy_to_user(buf, &hdr, HVPIPE_HDR_LEN); if (ret) - return ret; + return -EFAULT; /* * Message event has payload, so get the payload with @@ -385,19 +386,23 @@ static ssize_t papr_hvpipe_handle_read(struct file *file, if (hdr.flags & HVPIPE_MSG_AVAILABLE) { ret = hvpipe_rtas_recv_msg(buf + HVPIPE_HDR_LEN, size - HVPIPE_HDR_LEN); - if (ret > 0) { + /* + * Always clear MSG_AVAILABLE once the RTAS call has drained + * the message, regardless of whether copy_to_user succeeded. + */ + if (ret >= 0 || ret == -EFAULT) src_info->hvpipe_status &= ~HVPIPE_MSG_AVAILABLE; - ret += HVPIPE_HDR_LEN; - } } else if (hdr.flags & HVPIPE_LOST_CONNECTION) { /* * Hypervisor is closing the pipe for the specific * source. So notify user space. */ src_info->hvpipe_status &= ~HVPIPE_LOST_CONNECTION; - ret = HVPIPE_HDR_LEN; } + if (ret >= 0) + ret += HVPIPE_HDR_LEN; + return ret; } From 2eeac577480848801b35885b3a8201aa35f46236 Mon Sep 17 00:00:00 2001 From: "Ritesh Harjani (IBM)" Date: Fri, 1 May 2026 09:41:45 +0530 Subject: [PATCH 372/525] pseries/papr-hvpipe: Simplify spin unlock usage in papr_hvpipe_handle_release() Once the src_info is removed from the global list, no one can access it. This simplies the usage of spin_unlock_irqrestore() in papr_hvpipe_handle_release() Signed-off-by: Ritesh Harjani (IBM) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/4a980331557af3d10aada8576aaa16cddc691c65.1777606826.git.ritesh.list@gmail.com --- arch/powerpc/platforms/pseries/papr-hvpipe.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/arch/powerpc/platforms/pseries/papr-hvpipe.c b/arch/powerpc/platforms/pseries/papr-hvpipe.c index c007560d2d8c..5aa37f6ad8c9 100644 --- a/arch/powerpc/platforms/pseries/papr-hvpipe.c +++ b/arch/powerpc/platforms/pseries/papr-hvpipe.c @@ -460,6 +460,7 @@ static int papr_hvpipe_handle_release(struct inode *inode, src_info = file->private_data; list_del(&src_info->list); file->private_data = NULL; + spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); /* * If the pipe for this specific source has any pending * payload, issue recv HVPIPE RTAS so that pipe will not @@ -467,10 +468,8 @@ static int papr_hvpipe_handle_release(struct inode *inode, */ if (src_info->hvpipe_status & HVPIPE_MSG_AVAILABLE) { src_info->hvpipe_status = 0; - spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); hvpipe_rtas_recv_msg(NULL, 0); - } else - spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); + } kfree(src_info); return 0; From 4e2d83c80495a9327141e8636f25dde13155f14f Mon Sep 17 00:00:00 2001 From: "Ritesh Harjani (IBM)" Date: Fri, 1 May 2026 09:41:46 +0530 Subject: [PATCH 373/525] pseries/papr-hvpipe: Kill task_struct pointer from struct hvpipe_source_info We don't really use task_struct pointer for anything meaningful. So just kill it for now, and we can bring back later if we need this for any future debug purposes. Signed-off-by: Ritesh Harjani (IBM) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/895e061e45cdc95db36fa7f27aa1922b81eed867.1777606826.git.ritesh.list@gmail.com --- arch/powerpc/platforms/pseries/papr-hvpipe.c | 5 ++--- arch/powerpc/platforms/pseries/papr-hvpipe.h | 1 - 2 files changed, 2 insertions(+), 4 deletions(-) diff --git a/arch/powerpc/platforms/pseries/papr-hvpipe.c b/arch/powerpc/platforms/pseries/papr-hvpipe.c index 5aa37f6ad8c9..46159f1c1cf1 100644 --- a/arch/powerpc/platforms/pseries/papr-hvpipe.c +++ b/arch/powerpc/platforms/pseries/papr-hvpipe.c @@ -493,7 +493,6 @@ static int papr_hvpipe_dev_create_handle(u32 srcID) return -ENOMEM; src_info->srcID = srcID; - src_info->tsk = current; init_waitqueue_head(&src_info->recv_wqh); /* @@ -503,8 +502,8 @@ static int papr_hvpipe_dev_create_handle(u32 srcID) spin_lock_irqsave(&hvpipe_src_list_lock, flags); if (hvpipe_find_source(srcID)) { spin_unlock_irqrestore(&hvpipe_src_list_lock, flags); - pr_err("pid(%d) could not get the source(%d)\n", - src_info->tsk->pid, srcID); + pr_err("pid(%s:%d) could not get the source(%d)\n", + current->comm, task_pid_nr(current), srcID); kfree(src_info); return -EALREADY; } diff --git a/arch/powerpc/platforms/pseries/papr-hvpipe.h b/arch/powerpc/platforms/pseries/papr-hvpipe.h index c343f4230865..4bdf7bb2fc4d 100644 --- a/arch/powerpc/platforms/pseries/papr-hvpipe.h +++ b/arch/powerpc/platforms/pseries/papr-hvpipe.h @@ -21,7 +21,6 @@ struct hvpipe_source_info { u32 srcID; u32 hvpipe_status; wait_queue_head_t recv_wqh; /* wake up poll() waitq */ - struct task_struct *tsk; }; /* From fe53d2ae82c06aa2d6402624af01e8f8ddfcd5b3 Mon Sep 17 00:00:00 2001 From: "Ritesh Harjani (IBM)" Date: Fri, 1 May 2026 09:41:47 +0530 Subject: [PATCH 374/525] pseries/papr-hvpipe: Refactor and simplify hvpipe_rtas_recv_msg() Simplify hvpipe_rtas_recv_msg() by removing three levels of nesting... if (!ret) if (buf) if (size < bytes_written) ... this refactoring of the function bails out to "out:" label first, in case of any error. This simplifies the init flow. Signed-off-by: Ritesh Harjani (IBM) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/bbe7ddf8b8e25c9be8fc5e2c4aea9e5fca128bf4.1777606826.git.ritesh.list@gmail.com --- arch/powerpc/platforms/pseries/papr-hvpipe.c | 52 ++++++++++---------- 1 file changed, 26 insertions(+), 26 deletions(-) diff --git a/arch/powerpc/platforms/pseries/papr-hvpipe.c b/arch/powerpc/platforms/pseries/papr-hvpipe.c index 46159f1c1cf1..3688b2be0445 100644 --- a/arch/powerpc/platforms/pseries/papr-hvpipe.c +++ b/arch/powerpc/platforms/pseries/papr-hvpipe.c @@ -190,34 +190,34 @@ static int hvpipe_rtas_recv_msg(char __user *buf, int size) return -ENOMEM; } - ret = rtas_ibm_receive_hvpipe_msg(work_area, &srcID, - &bytes_written); - if (!ret) { - /* - * Recv HVPIPE RTAS is successful. - * When releasing FD or no one is waiting on the - * specific source, issue recv HVPIPE RTAS call - * so that pipe is not blocked - this func is called - * with NULL buf. - */ - if (buf) { - if (size < bytes_written) { - pr_err("Received the payload size = %d, but the buffer size = %d\n", - bytes_written, size); - bytes_written = size; - } - if (copy_to_user(buf, - rtas_work_area_raw_buf(work_area), - bytes_written)) - ret = -EFAULT; - else - ret = bytes_written; - } - } else { - pr_err("ibm,receive-hvpipe-msg failed with %d\n", - ret); + /* + * Recv HVPIPE RTAS is successful. + * When releasing FD or no one is waiting on the + * specific source, issue recv HVPIPE RTAS call + * so that pipe is not blocked - this func is called + * with NULL buf. + */ + ret = rtas_ibm_receive_hvpipe_msg(work_area, &srcID, &bytes_written); + if (ret) { + pr_err("ibm,receive-hvpipe-msg failed with %d\n", ret); + goto out; } + if (!buf) + goto out; + + if (size < bytes_written) { + pr_err("Received the payload size = %d, but the buffer size = %d\n", + bytes_written, size); + bytes_written = size; + } + + if (copy_to_user(buf, rtas_work_area_raw_buf(work_area), bytes_written)) + ret = -EFAULT; + else + ret = bytes_written; + +out: rtas_work_area_free(work_area); return ret; } From 629d1a901de57490d29a495273df11e64993ec04 Mon Sep 17 00:00:00 2001 From: "Ritesh Harjani (IBM)" Date: Fri, 1 May 2026 09:41:48 +0530 Subject: [PATCH 375/525] pseries/papr-hvpipe: Fix style and checkpatch issues in enable_hvpipe_IRQ() While at it let's also fix the similar style issue in enable_hvpipe_IRQ() function. This also fixes a minor checkpatch warning which I got due to an extra space before " ==". Signed-off-by: Ritesh Harjani (IBM) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/1174f60d0ae128e773dbefd11dd8d46d69e7f50e.1777606826.git.ritesh.list@gmail.com --- arch/powerpc/platforms/pseries/papr-hvpipe.c | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) diff --git a/arch/powerpc/platforms/pseries/papr-hvpipe.c b/arch/powerpc/platforms/pseries/papr-hvpipe.c index 3688b2be0445..0c40bdde45e2 100644 --- a/arch/powerpc/platforms/pseries/papr-hvpipe.c +++ b/arch/powerpc/platforms/pseries/papr-hvpipe.c @@ -693,20 +693,19 @@ static int __init enable_hvpipe_IRQ(void) struct device_node *np; hvpipe_check_exception_token = rtas_function_token(RTAS_FN_CHECK_EXCEPTION); - if (hvpipe_check_exception_token == RTAS_UNKNOWN_SERVICE) + if (hvpipe_check_exception_token == RTAS_UNKNOWN_SERVICE) return -ENODEV; /* hvpipe events */ np = of_find_node_by_path("/event-sources/ibm,hvpipe-msg-events"); - if (np != NULL) { - request_event_sources_irqs(np, hvpipe_event_interrupt, - "HPIPE_EVENT"); - of_node_put(np); - } else { - pr_err("Can not enable hvpipe event IRQ\n"); + if (!np) { + pr_err("No device node found, could not enable hvpipe event IRQ\n"); return -ENODEV; } + request_event_sources_irqs(np, hvpipe_event_interrupt, "HPIPE_EVENT"); + of_node_put(np); + return 0; } From b3a97f9484080c6e71db9e803e3cc1bb372a9bc7 Mon Sep 17 00:00:00 2001 From: Sourabh Jain Date: Tue, 7 Apr 2026 18:13:44 +0530 Subject: [PATCH 376/525] powerpc/kdump: fix KASAN sanitization flag for core_$(BITS).o MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit KASAN instrumentation is intended to be disabled for the kexec core code, but the existing Makefile entry misses the object suffix. As a result, the flag is not applied correctly to core_$(BITS).o. So when KASAN is enabled, kexec_copy_flush and copy_segments in kexec/core_64.c are instrumented, which can result in accesses to shadow memory via normal address translation paths. Since these run with the MMU disabled, such accesses may trigger page faults (bad_page_fault) that cannot be handled in the kdump path, ultimately causing a hang and preventing the kdump kernel from booting. The same is true for kexec as well, since the same functions are used there. Update the entry to include the “.o” suffix so that KASAN instrumentation is properly disabled for this object file. Fixes: 2ab2d5794f14 ("powerpc/kasan: Disable address sanitization in kexec paths") Reported-by: Venkat Rao Bagalkote Closes: https://lore.kernel.org/all/1dee8891-8bcc-46b4-93f3-fc3a774abd5b@linux.ibm.com/ Cc: stable@vger.kernel.org Reviewed-by: Ritesh Harjani (IBM) Tested-by: Venkat Rao Bagalkote Acked-by: Mahesh Salgaonkar Reviewed-by: Aboorva Devarajan Tested-by: Aboorva Devarajan Signed-off-by: Sourabh Jain Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260407124349.1698552-1-sourabhjain@linux.ibm.com --- arch/powerpc/kexec/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/kexec/Makefile b/arch/powerpc/kexec/Makefile index 470eb0453e17..ec7a0eed75dc 100644 --- a/arch/powerpc/kexec/Makefile +++ b/arch/powerpc/kexec/Makefile @@ -16,4 +16,4 @@ GCOV_PROFILE_core_$(BITS).o := n KCOV_INSTRUMENT_core_$(BITS).o := n UBSAN_SANITIZE_core_$(BITS).o := n KASAN_SANITIZE_core.o := n -KASAN_SANITIZE_core_$(BITS) := n +KASAN_SANITIZE_core_$(BITS).o := n From 38e989d504fc52900a3786b7144fb53cd67e0389 Mon Sep 17 00:00:00 2001 From: Sourabh Jain Date: Tue, 7 Apr 2026 18:13:45 +0530 Subject: [PATCH 377/525] powerpc/vmx: avoid KASAN instrumentation in enter_vmx_ops() for kexec The kexec sequence invokes enter_vmx_ops() via copy_page() with the MMU disabled. In this context, code must not rely on normal virtual address translations or trigger page faults. With KASAN enabled, functions get instrumented and may access shadow memory using regular address translation. When executed with the MMU off, this can lead to page faults (bad_page_fault) from which the kernel cannot recover in the kexec path, resulting in a hang. The kexec path sets preempt_count to HARDIRQ_OFFSET before entering the MMU-off copy sequence. current_thread_info()->preempt_count = HARDIRQ_OFFSET kexec_sequence(..., copy_with_mmu_off = 1) -> kexec_copy_flush(image) copy_segments() -> copy_page(dest, addr) bl enter_vmx_ops() if (in_interrupt()) return 0 beq .Lnonvmx_copy Since kexec sets preempt_count to HARDIRQ_OFFSET, in_interrupt() evaluates to true and enter_vmx_ops() returns early. As in_interrupt() (and preempt_count()) are always inlined, mark enter_vmx_ops() with __no_sanitize_address to avoid KASAN instrumentation and shadow memory access with MMU disabled, helping kexec boot fine with KASAN enabled. Reported-by: Aboorva Devarajan Reviewed-by: Aboorva Devarajan Tested-by: Aboorva Devarajan Reviewed-by: Ritesh Harjani (IBM) Signed-off-by: Sourabh Jain Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260407124349.1698552-2-sourabhjain@linux.ibm.com --- arch/powerpc/lib/vmx-helper.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/arch/powerpc/lib/vmx-helper.c b/arch/powerpc/lib/vmx-helper.c index 554b248002b4..57e897b60db8 100644 --- a/arch/powerpc/lib/vmx-helper.c +++ b/arch/powerpc/lib/vmx-helper.c @@ -52,7 +52,14 @@ int exit_vmx_usercopy(void) } EXPORT_SYMBOL(exit_vmx_usercopy); -int enter_vmx_ops(void) +/* + * Can be called from kexec copy_page() path with MMU off. The kexec + * code sets preempt_count to HARDIRQ_OFFSET so we return early here. + * Since in_interrupt() is always inline, __no_sanitize_address on this + * function is sufficient to avoid KASAN shadow memory accesses in real + * mode. + */ +int __no_sanitize_address enter_vmx_ops(void) { if (in_interrupt()) return 0; From 0e7c074cfcd9bd93765505f9eb8b42f03ed2a744 Mon Sep 17 00:00:00 2001 From: Pavitra Jha Date: Fri, 1 May 2026 07:07:12 -0400 Subject: [PATCH 378/525] net: wwan: t7xx: validate port_count against message length in t7xx_port_enum_msg_handler t7xx_port_enum_msg_handler() uses the modem-supplied port_count field as a loop bound over port_msg->data[] without checking that the message buffer contains sufficient data. A modem sending port_count=65535 in a 12-byte buffer triggers a slab-out-of-bounds read of up to 262140 bytes. Add a sizeof(*port_msg) check before accessing the port message header fields to guard against undersized messages. Add a struct_size() check after extracting port_count and before the loop. In t7xx_parse_host_rt_data(), guard the rt_feature header read with a remaining-buffer check before accessing data_len, validate feat_data_len against the actual remaining buffer to prevent OOB reads and signed integer overflow on offset. Pass msg_len from both call sites: skb->len at the DPMAIF path after skb_pull(), and the validated feat_data_len at the handshake path. Fixes: da45d2566a1d ("net: wwan: t7xx: Add control port") Cc: stable@vger.kernel.org Signed-off-by: Pavitra Jha Link: https://patch.msgid.link/20260501110713.145563-1-jhapavitra98@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/wwan/t7xx/t7xx_modem_ops.c | 20 +++++++++++++++++--- drivers/net/wwan/t7xx/t7xx_port_ctrl_msg.c | 18 ++++++++++++++++-- drivers/net/wwan/t7xx/t7xx_port_proxy.h | 2 +- 3 files changed, 34 insertions(+), 6 deletions(-) diff --git a/drivers/net/wwan/t7xx/t7xx_modem_ops.c b/drivers/net/wwan/t7xx/t7xx_modem_ops.c index 7968e208dd37..adb29d30c63f 100644 --- a/drivers/net/wwan/t7xx/t7xx_modem_ops.c +++ b/drivers/net/wwan/t7xx/t7xx_modem_ops.c @@ -457,8 +457,20 @@ static int t7xx_parse_host_rt_data(struct t7xx_fsm_ctl *ctl, struct t7xx_sys_inf offset = sizeof(struct feature_query); for (i = 0; i < FEATURE_COUNT && offset < data_length; i++) { + size_t remaining = data_length - offset; + size_t feat_data_len, feat_total; + + if (remaining < sizeof(*rt_feature)) + break; + rt_feature = data + offset; - offset += sizeof(*rt_feature) + le32_to_cpu(rt_feature->data_len); + feat_data_len = le32_to_cpu(rt_feature->data_len); + + if (feat_data_len > remaining - sizeof(*rt_feature)) + break; + + feat_total = sizeof(*rt_feature) + feat_data_len; + offset += feat_total; ft_spt_cfg = FIELD_GET(FEATURE_MSK, core->feature_set[i]); if (ft_spt_cfg != MTK_FEATURE_MUST_BE_SUPPORTED) @@ -468,8 +480,10 @@ static int t7xx_parse_host_rt_data(struct t7xx_fsm_ctl *ctl, struct t7xx_sys_inf if (ft_spt_st != MTK_FEATURE_MUST_BE_SUPPORTED) return -EINVAL; - if (i == RT_ID_MD_PORT_ENUM || i == RT_ID_AP_PORT_ENUM) - t7xx_port_enum_msg_handler(ctl->md, rt_feature->data); + if (i == RT_ID_MD_PORT_ENUM || i == RT_ID_AP_PORT_ENUM) { + t7xx_port_enum_msg_handler(ctl->md, rt_feature->data, + feat_data_len); + } } return 0; diff --git a/drivers/net/wwan/t7xx/t7xx_port_ctrl_msg.c b/drivers/net/wwan/t7xx/t7xx_port_ctrl_msg.c index ae632ef96698..f869e4ed9ee9 100644 --- a/drivers/net/wwan/t7xx/t7xx_port_ctrl_msg.c +++ b/drivers/net/wwan/t7xx/t7xx_port_ctrl_msg.c @@ -117,6 +117,7 @@ static int fsm_ee_message_handler(struct t7xx_port *port, struct t7xx_fsm_ctl *c * t7xx_port_enum_msg_handler() - Parse the port enumeration message to create/remove nodes. * @md: Modem context. * @msg: Message. + * @msg_len: Length of @msg in bytes. * * Used to control create/remove device node. * @@ -124,12 +125,18 @@ static int fsm_ee_message_handler(struct t7xx_port *port, struct t7xx_fsm_ctl *c * * 0 - Success. * * -EFAULT - Message check failure. */ -int t7xx_port_enum_msg_handler(struct t7xx_modem *md, void *msg) +int t7xx_port_enum_msg_handler(struct t7xx_modem *md, void *msg, size_t msg_len) { struct device *dev = &md->t7xx_dev->pdev->dev; unsigned int version, port_count, i; struct port_msg *port_msg = msg; + if (msg_len < sizeof(*port_msg)) { + dev_err(dev, "Port enum msg too short for header: need %zu, have %zu\n", + sizeof(*port_msg), msg_len); + return -EINVAL; + } + version = FIELD_GET(PORT_MSG_VERSION, le32_to_cpu(port_msg->info)); if (version != PORT_ENUM_VER || le32_to_cpu(port_msg->head_pattern) != PORT_ENUM_HEAD_PATTERN || @@ -141,6 +148,13 @@ int t7xx_port_enum_msg_handler(struct t7xx_modem *md, void *msg) } port_count = FIELD_GET(PORT_MSG_PRT_CNT, le32_to_cpu(port_msg->info)); + + if (msg_len < struct_size(port_msg, data, port_count)) { + dev_err(dev, "Port enum msg too short: need %zu, have %zu\n", + struct_size(port_msg, data, port_count), msg_len); + return -EINVAL; + } + for (i = 0; i < port_count; i++) { u32 port_info = le32_to_cpu(port_msg->data[i]); unsigned int ch_id; @@ -191,7 +205,7 @@ static int control_msg_handler(struct t7xx_port *port, struct sk_buff *skb) case CTL_ID_PORT_ENUM: skb_pull(skb, sizeof(*ctrl_msg_h)); - ret = t7xx_port_enum_msg_handler(ctl->md, (struct port_msg *)skb->data); + ret = t7xx_port_enum_msg_handler(ctl->md, (struct port_msg *)skb->data, skb->len); if (!ret) ret = port_ctl_send_msg_to_md(port, CTL_ID_PORT_ENUM, 0); else diff --git a/drivers/net/wwan/t7xx/t7xx_port_proxy.h b/drivers/net/wwan/t7xx/t7xx_port_proxy.h index f0918b36e899..7c3190bf0fcf 100644 --- a/drivers/net/wwan/t7xx/t7xx_port_proxy.h +++ b/drivers/net/wwan/t7xx/t7xx_port_proxy.h @@ -103,7 +103,7 @@ void t7xx_port_proxy_reset(struct port_proxy *port_prox); void t7xx_port_proxy_uninit(struct port_proxy *port_prox); int t7xx_port_proxy_init(struct t7xx_modem *md); void t7xx_port_proxy_md_status_notify(struct port_proxy *port_prox, unsigned int state); -int t7xx_port_enum_msg_handler(struct t7xx_modem *md, void *msg); +int t7xx_port_enum_msg_handler(struct t7xx_modem *md, void *msg, size_t msg_len); int t7xx_port_proxy_chl_enable_disable(struct port_proxy *port_prox, unsigned int ch_id, bool en_flag); void t7xx_port_proxy_set_cfg(struct t7xx_modem *md, enum port_cfg_id cfg_id); From da107152c43915211e1fc0319b9526f4241abca2 Mon Sep 17 00:00:00 2001 From: "Christophe Leroy (CS GROUP)" Date: Tue, 21 Apr 2026 08:26:08 +0200 Subject: [PATCH 379/525] powerpc/8xx: Fix interrupt mask in cpm1_gpiochip_add16() Allthough fsl,cpm1-gpio-irq-mask always contains a 16 bits value, it is a standard u32 OF property as documented in Documentation/devicetree/bindings/soc/fsl/cpm_qe/gpio.txt The driver erroneously uses of_property_read_u16() leading to a mask which is always 0. Fix it by using of_property_read_u32() instead. Fixes: 726bd223105c ("powerpc/8xx: Adding support of IRQ in MPC8xx GPIO") Signed-off-by: Christophe Leroy (CS GROUP) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/bb0b6d6c4543238c38d5d29a776d0674a8c0c180.1776752750.git.chleroy@kernel.org --- arch/powerpc/platforms/8xx/cpm1.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/platforms/8xx/cpm1.c b/arch/powerpc/platforms/8xx/cpm1.c index 7433be7d66ee..f00734f0590c 100644 --- a/arch/powerpc/platforms/8xx/cpm1.c +++ b/arch/powerpc/platforms/8xx/cpm1.c @@ -477,7 +477,7 @@ int cpm1_gpiochip_add16(struct device *dev) struct device_node *np = dev->of_node; struct cpm1_gpio16_chip *cpm1_gc; struct gpio_chip *gc; - u16 mask; + u32 mask; cpm1_gc = devm_kzalloc(dev, sizeof(*cpm1_gc), GFP_KERNEL); if (!cpm1_gc) @@ -485,7 +485,7 @@ int cpm1_gpiochip_add16(struct device *dev) spin_lock_init(&cpm1_gc->lock); - if (!of_property_read_u16(np, "fsl,cpm1-gpio-irq-mask", &mask)) { + if (!of_property_read_u32(np, "fsl,cpm1-gpio-irq-mask", &mask)) { int i, j; for (i = 0, j = 0; i < 16; i++) From 131717e656b379addb95af2dcb2d90c723bae24b Mon Sep 17 00:00:00 2001 From: Shivani Nittor Date: Tue, 21 Apr 2026 20:36:28 +0530 Subject: [PATCH 380/525] powerpc/perf: Update check for PERF_SAMPLE_DATA_SRC marked events The core-book3s PMU sampling code validates the SIER TYPE field when PERF_SAMPLE_DATA_SRC is requested. The SIER TYPE field indicates the instruction type and is only valid for random sampling (marked events). To handle cases observed where SIER TYPE could be zero even for marked events,validation was added to drop such samples and increment event->lost_samples. However, this validation was applied to all samples, including continuous sampling. In continuous sampling mode, the PMU does not set the SIER TYPE field, so it remains zero. As a result, valid continuous samples were incorrectly treated as invalid and dropped. Fixed this by gating the SIER TYPE validation with mark_event, so the check runs only for marked (random) events. Continuous samples now skip this check and are recorded normally in the final data recording path. Fixes: 2ffb26afa642 ("arch/powerpc/perf: Check the instruction type before creating sample with perf_mem_data_src") Signed-off-by: Shivani Nittor Reviewed-by: Mukesh Kumar Chaurasiya (IBM) Reviewed-by: Athira Rajeev [Maddy: Fixed reviewed-by tag] Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260421150628.96500-1-shivani@linux.ibm.com --- arch/powerpc/perf/core-book3s.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/perf/core-book3s.c b/arch/powerpc/perf/core-book3s.c index 8b0081441f85..2e6adf5b95c4 100644 --- a/arch/powerpc/perf/core-book3s.c +++ b/arch/powerpc/perf/core-book3s.c @@ -2242,6 +2242,7 @@ static void record_and_restart(struct perf_event *event, unsigned long val, const u64 last_period = event->hw.last_period; s64 prev, delta, left; int record = 0; + int mark_event = regs->dsisr & MMCRA_SAMPLE_ENABLE; if (event->hw.state & PERF_HES_STOPPED) { write_pmc(event->hw.idx, 0); @@ -2304,9 +2305,9 @@ static void record_and_restart(struct perf_event *event, unsigned long val, * In ISA v3.0 and before values "0" and "7" are considered reserved. * In ISA v3.1, value "7" has been used to indicate "larx/stcx". * Drop the sample if "type" has reserved values for this field with a - * ISA version check. + * ISA version check for marked events. */ - if (event->attr.sample_type & PERF_SAMPLE_DATA_SRC && + if (mark_event && event->attr.sample_type & PERF_SAMPLE_DATA_SRC && ppmu->get_mem_data_src) { val = (regs->dar & SIER_TYPE_MASK) >> SIER_TYPE_SHIFT; if (val == 0 || (val == 7 && !cpu_has_feature(CPU_FTR_ARCH_31))) { From ae9582cd0b9ccc4a121af300df68fd27f72e9822 Mon Sep 17 00:00:00 2001 From: Cosmin Ratiu Date: Mon, 4 May 2026 21:10:58 +0300 Subject: [PATCH 381/525] net/mlx5e: psp: Fix invalid access on PSP dev registration fail priv->psp->psp is initialized with the PSP device as returned by psp_dev_create(). This could also return an error, in which case a future psp_dev_unregister() will result in unpleasantness. Avoid that by using a local variable and only saving the PSP device when registration succeeds. In case psp_dev_create() fails, priv->psp and steering structs are left in place, but they will be inert. The unchecked access of priv->psp in mlx5e_psp_offload_handle_rx_skb() won't happen because without a PSP device, there can be no SAs added and therefore no packets will be successfully decrypted and be handed off to the SW handler. Fixes: 89ee2d92f66c ("net/mlx5e: Support PSP offload functionality") Signed-off-by: Cosmin Ratiu Reviewed-by: Dragos Tatulea Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260504181100.269334-2-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- .../mellanox/mlx5/core/en_accel/psp.c | 26 ++++++++++++------- 1 file changed, 17 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c index 6a50b6dec0fa..1ff818fb48df 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c @@ -1070,29 +1070,37 @@ static struct psp_dev_ops mlx5_psp_ops = { void mlx5e_psp_unregister(struct mlx5e_priv *priv) { - if (!priv->psp || !priv->psp->psp) + struct mlx5e_psp *psp = priv->psp; + + if (!psp || !psp->psp) return; - psp_dev_unregister(priv->psp->psp); + psp_dev_unregister(psp->psp); + psp->psp = NULL; } void mlx5e_psp_register(struct mlx5e_priv *priv) { + struct mlx5e_psp *psp = priv->psp; + struct psp_dev *psd; + /* FW Caps missing */ if (!priv->psp) return; - priv->psp->caps.assoc_drv_spc = sizeof(u32); - priv->psp->caps.versions = 1 << PSP_VERSION_HDR0_AES_GCM_128; + psp->caps.assoc_drv_spc = sizeof(u32); + psp->caps.versions = 1 << PSP_VERSION_HDR0_AES_GCM_128; if (MLX5_CAP_PSP(priv->mdev, psp_crypto_esp_aes_gcm_256_encrypt) && MLX5_CAP_PSP(priv->mdev, psp_crypto_esp_aes_gcm_256_decrypt)) - priv->psp->caps.versions |= 1 << PSP_VERSION_HDR0_AES_GCM_256; + psp->caps.versions |= 1 << PSP_VERSION_HDR0_AES_GCM_256; - priv->psp->psp = psp_dev_create(priv->netdev, &mlx5_psp_ops, - &priv->psp->caps, NULL); - if (IS_ERR(priv->psp->psp)) + psd = psp_dev_create(priv->netdev, &mlx5_psp_ops, &psp->caps, NULL); + if (IS_ERR(psd)) { mlx5_core_err(priv->mdev, "PSP failed to register due to %pe\n", - priv->psp->psp); + psd); + return; + } + psp->psp = psd; } int mlx5e_psp_init(struct mlx5e_priv *priv) From 50690733db59fbb3de9fa811b606af324eeb4e37 Mon Sep 17 00:00:00 2001 From: Cosmin Ratiu Date: Mon, 4 May 2026 21:10:59 +0300 Subject: [PATCH 382/525] net/mlx5e: psp: Expose only a fully initialized priv->psp Currently, during PSP init, priv->psp is initialized to an incompletely built psp struct. Additionally, on fs init failure priv->psp is reset to NULL. Change this so that only a fully initialized priv->psp is set, which makes the code easier to reason about in failure scenarios. Fixes: af2196f49480 ("net/mlx5e: Implement PSP operations .assoc_add and .assoc_del") Signed-off-by: Cosmin Ratiu Reviewed-by: Dragos Tatulea Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260504181100.269334-3-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c index 1ff818fb48df..d9adb993e64d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/psp.c @@ -1139,22 +1139,18 @@ int mlx5e_psp_init(struct mlx5e_priv *priv) if (!psp) return -ENOMEM; - priv->psp = psp; fs = mlx5e_accel_psp_fs_init(priv); if (IS_ERR(fs)) { err = PTR_ERR(fs); - goto out_err; + kfree(psp); + return err; } psp->fs = fs; + priv->psp = psp; mlx5_core_dbg(priv->mdev, "PSP attached to netdevice\n"); return 0; - -out_err: - priv->psp = NULL; - kfree(psp); - return err; } void mlx5e_psp_cleanup(struct mlx5e_priv *priv) From c4a5c46199b5addf0157934da3aa89c33eb02a6d Mon Sep 17 00:00:00 2001 From: Cosmin Ratiu Date: Mon, 4 May 2026 21:11:00 +0300 Subject: [PATCH 383/525] net/mlx5e: psp: Hook PSP dev reg/unreg to profile enable/disable devlink reload while PSP connections are active does: mlx5_unload_one_devl_locked() -> mlx5_detach_device() -> _mlx5e_suspend() -> mlx5e_detach_netdev() -> profile->cleanup_rx -> profile->cleanup_tx -> mlx5e_destroy_mdev_resources() -> mlx5_core_dealloc_pd() fails: ... mlx5_core 0000:08:00.0: mlx5_cmd_out_err:821:(pid 19722): DEALLOC_PD(0x801) op_mod(0x0) failed, status bad resource state(0x9), syndrome (0xef0c8a), err(-22) ... The reason for failure is the existence of TX keys, which are removed by the PSP dev unregistration happening in: profile->cleanup() -> mlx5e_psp_unregister() -> mlx5e_psp_cleanup() -> psp_dev_unregister() ...but this isn't invoked in the devlink reload flow, only when changing the NIC profile (e.g. when transitioning to switchdev mode) or on dev teardown. Move PSP device registration into mlx5e_nic_enable(), and unregistration into the corresponding mlx5e_nic_disable(). These functions are called during netdev attach/detach after RX & TX are set up. This ensures that the keys will be gone by the time the PD is destroyed. Fixes: 89ee2d92f66c ("net/mlx5e: Support PSP offload functionality") Signed-off-by: Cosmin Ratiu Reviewed-by: Dragos Tatulea Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260504181100.269334-4-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 5a46870c4b74..8e9443caa933 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -6023,7 +6023,6 @@ static int mlx5e_nic_init(struct mlx5_core_dev *mdev, if (take_rtnl) rtnl_lock(); - mlx5e_psp_register(priv); /* update XDP supported features */ mlx5e_set_xdp_feature(priv); @@ -6036,7 +6035,6 @@ static int mlx5e_nic_init(struct mlx5_core_dev *mdev, static void mlx5e_nic_cleanup(struct mlx5e_priv *priv) { mlx5e_health_destroy_reporters(priv); - mlx5e_psp_unregister(priv); mlx5e_ktls_cleanup(priv); mlx5e_psp_cleanup(priv); mlx5e_fs_cleanup(priv->fs); @@ -6160,6 +6158,7 @@ static void mlx5e_nic_enable(struct mlx5e_priv *priv) mlx5e_fs_init_l2_addr(priv->fs, netdev); mlx5e_ipsec_init(priv); + mlx5e_psp_register(priv); err = mlx5e_macsec_init(priv); if (err) @@ -6230,6 +6229,7 @@ static void mlx5e_nic_disable(struct mlx5e_priv *priv) mlx5_lag_remove_netdev(mdev, priv->netdev); mlx5_vxlan_reset_to_default(mdev->vxlan); mlx5e_macsec_cleanup(priv); + mlx5e_psp_unregister(priv); mlx5e_ipsec_cleanup(priv); } From 4052b932041614675fa2dbf48f725557c23ebf05 Mon Sep 17 00:00:00 2001 From: Thomas Zimmermann Date: Wed, 1 Apr 2026 10:30:03 +0200 Subject: [PATCH 384/525] arch/powerpc: Drop CONFIG_FIRMWARE_EDID from defconfig files CONFIG_FIRMWARE_EDID=y depends on X86 or EFI_GENERIC_STUB. Neither is true here, so drop the lines from the defconfig files. Signed-off-by: Thomas Zimmermann Reviewed-by: Christophe Leroy (CS GROUP) Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260401083023.214426-1-tzimmermann@suse.de --- arch/powerpc/configs/amigaone_defconfig | 1 - arch/powerpc/configs/chrp32_defconfig | 1 - arch/powerpc/configs/g5_defconfig | 1 - arch/powerpc/configs/pasemi_defconfig | 1 - arch/powerpc/configs/powernv_defconfig | 1 - arch/powerpc/configs/ppc64_defconfig | 1 - arch/powerpc/configs/ppc64e_defconfig | 1 - arch/powerpc/configs/skiroot_defconfig | 1 - 8 files changed, 8 deletions(-) diff --git a/arch/powerpc/configs/amigaone_defconfig b/arch/powerpc/configs/amigaone_defconfig index 69ef3dc31c4b..7a515390646b 100644 --- a/arch/powerpc/configs/amigaone_defconfig +++ b/arch/powerpc/configs/amigaone_defconfig @@ -76,7 +76,6 @@ CONFIG_SERIAL_8250_CONSOLE=y # CONFIG_HW_RANDOM is not set # CONFIG_HWMON is not set CONFIG_FB=y -CONFIG_FIRMWARE_EDID=y CONFIG_FB_TILEBLITTING=y CONFIG_FB_RADEON=y CONFIG_FB_3DFX=y diff --git a/arch/powerpc/configs/chrp32_defconfig b/arch/powerpc/configs/chrp32_defconfig index b799c95480ae..66eae5b7e16c 100644 --- a/arch/powerpc/configs/chrp32_defconfig +++ b/arch/powerpc/configs/chrp32_defconfig @@ -76,7 +76,6 @@ CONFIG_SERIAL_8250_CONSOLE=y CONFIG_NVRAM=y # CONFIG_HWMON is not set CONFIG_FB=y -CONFIG_FIRMWARE_EDID=y CONFIG_FB_OF=y CONFIG_FB_MATROX=y CONFIG_FB_MATROX_MILLENIUM=y diff --git a/arch/powerpc/configs/g5_defconfig b/arch/powerpc/configs/g5_defconfig index 04bbb37f5978..e9996711b362 100644 --- a/arch/powerpc/configs/g5_defconfig +++ b/arch/powerpc/configs/g5_defconfig @@ -121,7 +121,6 @@ CONFIG_I2C_CHARDEV=y CONFIG_AGP=m CONFIG_AGP_UNINORTH=m CONFIG_FB=y -CONFIG_FIRMWARE_EDID=y CONFIG_FB_TILEBLITTING=y CONFIG_FB_OF=y CONFIG_FB_NVIDIA=y diff --git a/arch/powerpc/configs/pasemi_defconfig b/arch/powerpc/configs/pasemi_defconfig index 8bbf51b38480..89bcbeb05067 100644 --- a/arch/powerpc/configs/pasemi_defconfig +++ b/arch/powerpc/configs/pasemi_defconfig @@ -98,7 +98,6 @@ CONFIG_SENSORS_LM85=y CONFIG_SENSORS_LM90=y CONFIG_DRM=y CONFIG_DRM_RADEON=y -CONFIG_FIRMWARE_EDID=y CONFIG_FB_TILEBLITTING=y CONFIG_FB_VGA16=y CONFIG_FB_NVIDIA=y diff --git a/arch/powerpc/configs/powernv_defconfig b/arch/powerpc/configs/powernv_defconfig index cc9802420237..5d32c2767a65 100644 --- a/arch/powerpc/configs/powernv_defconfig +++ b/arch/powerpc/configs/powernv_defconfig @@ -196,7 +196,6 @@ CONFIG_I2C_CHARDEV=y # CONFIG_PTP_1588_CLOCK is not set CONFIG_DRM=y CONFIG_DRM_AST=y -CONFIG_FIRMWARE_EDID=y CONFIG_FB_OF=y CONFIG_FB_MATROX=m CONFIG_FB_MATROX_MILLENIUM=y diff --git a/arch/powerpc/configs/ppc64_defconfig b/arch/powerpc/configs/ppc64_defconfig index 3bf518e3a573..6316ca4df25d 100644 --- a/arch/powerpc/configs/ppc64_defconfig +++ b/arch/powerpc/configs/ppc64_defconfig @@ -249,7 +249,6 @@ CONFIG_I2C_CHARDEV=y CONFIG_I2C_AMD8111=y CONFIG_I2C_PASEMI=y CONFIG_FB=y -CONFIG_FIRMWARE_EDID=y CONFIG_FB_OF=y CONFIG_FB_MATROX=y CONFIG_FB_MATROX_MILLENIUM=y diff --git a/arch/powerpc/configs/ppc64e_defconfig b/arch/powerpc/configs/ppc64e_defconfig index 0fd49f67331f..20cc17dce94d 100644 --- a/arch/powerpc/configs/ppc64e_defconfig +++ b/arch/powerpc/configs/ppc64e_defconfig @@ -118,7 +118,6 @@ CONFIG_SERIAL_8250_CONSOLE=y CONFIG_I2C_CHARDEV=y CONFIG_I2C_AMD8111=y CONFIG_FB=y -CONFIG_FIRMWARE_EDID=y CONFIG_FB_OF=y CONFIG_FB_MATROX=y CONFIG_FB_MATROX_MILLENIUM=y diff --git a/arch/powerpc/configs/skiroot_defconfig b/arch/powerpc/configs/skiroot_defconfig index ff1bed4b6d2c..005536ee75bb 100644 --- a/arch/powerpc/configs/skiroot_defconfig +++ b/arch/powerpc/configs/skiroot_defconfig @@ -214,7 +214,6 @@ CONFIG_SENSORS_IBMPOWERNV=m CONFIG_DRM=m CONFIG_DRM_AST=m CONFIG_FB=y -CONFIG_FIRMWARE_EDID=y # CONFIG_VGA_CONSOLE is not set CONFIG_FRAMEBUFFER_CONSOLE=y CONFIG_LOGO=y From 3abcedfdfd3125431ed404fa75724118beac630b Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Mon, 4 May 2026 21:02:03 +0300 Subject: [PATCH 385/525] net/mlx5: SD: Serialize init/cleanup mlx5_sd_init() / mlx5_sd_cleanup() may run from multiple PFs in the same Socket-Direct group. This can cause the SD bring-up/tear-down sequence to be executed more than once or interleaved across PFs. Protect SD init/cleanup with mlx5_devcom_comp_lock() and track the SD group state on the primary device. Skip init if the primary is already UP, and skip cleanup unless the primary is UP. The state check on cleanup is needed because sd_register() drops the devcom comp lock between marking the comp ready and assigning primary_dev on each peer. A concurrent cleanup that acquires the lock in this window would observe devcom_is_ready==true while primary_dev is still NULL (causing mlx5_sd_get_primary() to return NULL) or while the FW alias setup performed by mlx5_sd_init()'s body has not yet run (causing sd_cmd_unset_primary() to dereference a NULL tx_ft). Gate the cleanup body on primary_sd->state == MLX5_SD_STATE_UP, which is set only at the very end of mlx5_sd_init() under the same comp lock - so observing UP guarantees primary_dev, secondaries[], tx_ft, and dfs are all populated. Also bail explicitly if mlx5_sd_get_primary() returns NULL, in case state is checked on a peer whose primary_dev hasn't been assigned yet. In addition, move mlx5_devcom_comp_set_ready(false) from sd_unregister() into the cleanup's locked section, including the !primary and state != UP early-exit paths, so the device cannot unregister and free its struct mlx5_sd while devcom is still marked ready. A concurrent init acquiring the devcom lock will now observe devcom is no longer ready and bail out immediately. Fixes: 381978d28317 ("net/mlx5e: Create single netdev per SD group") Signed-off-by: Shay Drory Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260504180206.268568-2-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/mellanox/mlx5/core/lib/sd.c | 56 +++++++++++++++++-- 1 file changed, 50 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c index 762c783156b4..ec42685bdece 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c @@ -18,6 +18,7 @@ struct mlx5_sd { u8 host_buses; struct mlx5_devcom_comp_dev *devcom; struct dentry *dfs; + u8 state; bool primary; union { struct { /* primary */ @@ -31,6 +32,11 @@ struct mlx5_sd { }; }; +enum mlx5_sd_state { + MLX5_SD_STATE_DOWN = 0, + MLX5_SD_STATE_UP, +}; + static int mlx5_sd_get_host_buses(struct mlx5_core_dev *dev) { struct mlx5_sd *sd = mlx5_get_sd(dev); @@ -270,9 +276,6 @@ static void sd_unregister(struct mlx5_core_dev *dev) { struct mlx5_sd *sd = mlx5_get_sd(dev); - mlx5_devcom_comp_lock(sd->devcom); - mlx5_devcom_comp_set_ready(sd->devcom, false); - mlx5_devcom_comp_unlock(sd->devcom); mlx5_devcom_unregister_component(sd->devcom); } @@ -426,6 +429,7 @@ int mlx5_sd_init(struct mlx5_core_dev *dev) struct mlx5_core_dev *primary, *pos, *to; struct mlx5_sd *sd = mlx5_get_sd(dev); u8 alias_key[ACCESS_KEY_LEN]; + struct mlx5_sd *primary_sd; int err, i; err = sd_init(dev); @@ -440,10 +444,17 @@ int mlx5_sd_init(struct mlx5_core_dev *dev) if (err) goto err_sd_cleanup; + mlx5_devcom_comp_lock(sd->devcom); if (!mlx5_devcom_comp_is_ready(sd->devcom)) - return 0; + goto out; primary = mlx5_sd_get_primary(dev); + if (!primary) + goto out; + + primary_sd = mlx5_get_sd(primary); + if (primary_sd->state != MLX5_SD_STATE_DOWN) + goto out; for (i = 0; i < ACCESS_KEY_LEN; i++) alias_key[i] = get_random_u8(); @@ -472,6 +483,9 @@ int mlx5_sd_init(struct mlx5_core_dev *dev) sd->group_id, mlx5_devcom_comp_get_size(sd->devcom)); sd_print_group(primary); + primary_sd->state = MLX5_SD_STATE_UP; +out: + mlx5_devcom_comp_unlock(sd->devcom); return 0; err_unset_secondaries: @@ -481,6 +495,15 @@ int mlx5_sd_init(struct mlx5_core_dev *dev) sd_cmd_unset_primary(primary); debugfs_remove_recursive(sd->dfs); err_sd_unregister: + mlx5_sd_for_each_secondary(i, primary, pos) { + struct mlx5_sd *peer_sd = mlx5_get_sd(pos); + + primary_sd->secondaries[i - 1] = NULL; + peer_sd->primary_dev = NULL; + } + primary_sd->primary = false; + mlx5_devcom_comp_set_ready(sd->devcom, false); + mlx5_devcom_comp_unlock(sd->devcom); sd_unregister(dev); err_sd_cleanup: sd_cleanup(dev); @@ -491,22 +514,43 @@ void mlx5_sd_cleanup(struct mlx5_core_dev *dev) { struct mlx5_sd *sd = mlx5_get_sd(dev); struct mlx5_core_dev *primary, *pos; + struct mlx5_sd *primary_sd; int i; if (!sd) return; + mlx5_devcom_comp_lock(sd->devcom); if (!mlx5_devcom_comp_is_ready(sd->devcom)) - goto out; + goto out_unlock; primary = mlx5_sd_get_primary(dev); + if (!primary) + goto out_ready_false; + + primary_sd = mlx5_get_sd(primary); + if (primary_sd->state != MLX5_SD_STATE_UP) + goto out_clear_peers; + mlx5_sd_for_each_secondary(i, primary, pos) sd_cmd_unset_secondary(pos); sd_cmd_unset_primary(primary); debugfs_remove_recursive(sd->dfs); sd_info(primary, "group id %#x, uncombined\n", sd->group_id); -out: + primary_sd->state = MLX5_SD_STATE_DOWN; +out_clear_peers: + mlx5_sd_for_each_secondary(i, primary, pos) { + struct mlx5_sd *peer_sd = mlx5_get_sd(pos); + + primary_sd->secondaries[i - 1] = NULL; + peer_sd->primary_dev = NULL; + } + primary_sd->primary = false; +out_ready_false: + mlx5_devcom_comp_set_ready(sd->devcom, false); +out_unlock: + mlx5_devcom_comp_unlock(sd->devcom); sd_unregister(dev); sd_cleanup(dev); } From 05217e4ffbb229e7218cf318e0033780abadb624 Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Mon, 4 May 2026 21:02:04 +0300 Subject: [PATCH 386/525] net/mlx5: SD, Keep multi-pf debugfs entries on primary mlx5_sd_init() creates the "multi-pf" debugfs directory under the primary device debugfs root, but stored the dentry in the calling device's sd struct. When sd_cleanup() run on a different PF, this leads to using the wrong sd->dfs for removing entries, which results in memory leak and an error in when re-creating the SD.[1] Fix it by explicitly storing the debugfs dentry in the primary device sd struct and use it for all per-group files. [1] debugfs: 'multi-pf' already exists in '0000:08:00.1' Fixes: 4375130bf527 ("net/mlx5: SD, Add debugfs") Signed-off-by: Shay Drory Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260504180206.268568-3-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/mellanox/mlx5/core/lib/sd.c | 19 +++++++++++++------ 1 file changed, 13 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c index ec42685bdece..89b7e4d67303 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c @@ -463,9 +463,13 @@ int mlx5_sd_init(struct mlx5_core_dev *dev) if (err) goto err_sd_unregister; - sd->dfs = debugfs_create_dir("multi-pf", mlx5_debugfs_get_dev_root(primary)); - debugfs_create_x32("group_id", 0400, sd->dfs, &sd->group_id); - debugfs_create_file("primary", 0400, sd->dfs, primary, &dev_fops); + primary_sd->dfs = + debugfs_create_dir("multi-pf", + mlx5_debugfs_get_dev_root(primary)); + debugfs_create_x32("group_id", 0400, primary_sd->dfs, + &primary_sd->group_id); + debugfs_create_file("primary", 0400, primary_sd->dfs, primary, + &dev_fops); mlx5_sd_for_each_secondary(i, primary, pos) { char name[32]; @@ -475,7 +479,8 @@ int mlx5_sd_init(struct mlx5_core_dev *dev) goto err_unset_secondaries; snprintf(name, sizeof(name), "secondary_%d", i - 1); - debugfs_create_file(name, 0400, sd->dfs, pos, &dev_fops); + debugfs_create_file(name, 0400, primary_sd->dfs, pos, + &dev_fops); } @@ -493,7 +498,8 @@ int mlx5_sd_init(struct mlx5_core_dev *dev) mlx5_sd_for_each_secondary_to(i, primary, to, pos) sd_cmd_unset_secondary(pos); sd_cmd_unset_primary(primary); - debugfs_remove_recursive(sd->dfs); + debugfs_remove_recursive(primary_sd->dfs); + primary_sd->dfs = NULL; err_sd_unregister: mlx5_sd_for_each_secondary(i, primary, pos) { struct mlx5_sd *peer_sd = mlx5_get_sd(pos); @@ -535,7 +541,8 @@ void mlx5_sd_cleanup(struct mlx5_core_dev *dev) mlx5_sd_for_each_secondary(i, primary, pos) sd_cmd_unset_secondary(pos); sd_cmd_unset_primary(primary); - debugfs_remove_recursive(sd->dfs); + debugfs_remove_recursive(primary_sd->dfs); + primary_sd->dfs = NULL; sd_info(primary, "group id %#x, uncombined\n", sd->group_id); primary_sd->state = MLX5_SD_STATE_DOWN; From 3564222cfdde83a2d760b80192155a3ada1c9bdd Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Mon, 4 May 2026 21:02:05 +0300 Subject: [PATCH 387/525] net/mlx5e: SD, Fix missing cleanup on probe error When _mlx5e_probe() fails, the preceding successful mlx5_sd_init() is not undone. Auxiliary bus probe failure skips binding, so mlx5e_remove() is never called for that adev and the matching mlx5_sd_cleanup() never runs - leaking the per-dev SD struct. Call mlx5_sd_cleanup() on the probe error path to balance mlx5_sd_init(). A similar gap exists on the resume path: mlx5_sd_init() and mlx5_sd_cleanup() are currently bundled with both probe/remove and suspend/resume, even though only the FW alias state actually needs to follow the suspend/resume lifecycle - the sd struct allocation and devcom membership are software state that should track the full bound lifetime. As a result, a failed resume can leave a still-bound device with sd == NULL, which mlx5_sd_get_adev() can't distinguish from a non-SD device. Fixing this requires sd_suspend/resume APIs which will only destroy FW resources and is left for a follow-up series. Fixes: 381978d28317 ("net/mlx5e: Create single netdev per SD group") Signed-off-by: Shay Drory Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260504180206.268568-4-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 8e9443caa933..62b70334a13d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -6775,8 +6775,8 @@ static int mlx5e_resume(struct auxiliary_device *adev) actual_adev = mlx5_sd_get_adev(mdev, adev, edev->idx); if (actual_adev) - return _mlx5e_resume(actual_adev); - return 0; + err = _mlx5e_resume(actual_adev); + return err; } static int _mlx5e_suspend(struct auxiliary_device *adev, bool pre_netdev_reg) @@ -6912,9 +6912,16 @@ static int mlx5e_probe(struct auxiliary_device *adev, return err; actual_adev = mlx5_sd_get_adev(mdev, adev, edev->idx); - if (actual_adev) - return _mlx5e_probe(actual_adev); + if (actual_adev) { + err = _mlx5e_probe(actual_adev); + if (err) + goto sd_cleanup; + } return 0; + +sd_cleanup: + mlx5_sd_cleanup(mdev); + return err; } static void _mlx5e_remove(struct auxiliary_device *adev) From d466ddda5500b6b8ae060909d2317811f2c32a6a Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Mon, 4 May 2026 21:02:06 +0300 Subject: [PATCH 388/525] net/mlx5e: SD, Fix race condition in secondary device probe/remove When utilizing Socket-Direct single netdev functionality the driver resolves the actual auxiliary device using mlx5_sd_get_adev(). However, the current implementation returns the primary ETH auxiliary device without holding the device lock, leading to a potential race condition where the ETH device could be unbound or removed concurrently during probe, suspend, resume, or remove operations.[1] Fix this by introducing mlx5_sd_put_adev() and updating mlx5_sd_get_adev() so that secondaries devices would get a ref and acquire the device lock of the returned auxiliary device. After the lock is acquired, a second devcom check is needed[2]. In addition, update The callers to pair the get operation with the new put operation, ensuring the lock is held while the auxiliary device is being operated on and released afterwards. The "primary" designation is determined once in sd_register(). It's set before devcom is marked ready, and it never changes after that. In Addition, The primary path never locks a secondary: When the primary device invoke mlx5_sd_get_adev(), it sees dev == primary and returns. no additional lock is taken. Therefore lock ordering is always: secondary_lock -> primary_lock. The reverse never happens, so ABBA deadlock is impossible. [1] for example: BUG: kernel NULL pointer dereference, address: 0000000000000370 PGD 0 P4D 0 Oops: Oops: 0000 [#1] SMP CPU: 4 UID: 0 PID: 3945 Comm: bash Not tainted 6.19.0-rc3+ #1 NONE Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014 RIP: 0010:mlx5e_dcbnl_dscp_app+0x23/0x100 [mlx5_core] Call Trace: mlx5e_remove+0x82/0x12a [mlx5_core] device_release_driver_internal+0x194/0x1f0 bus_remove_device+0xc6/0x140 device_del+0x159/0x3c0 ? devl_param_driverinit_value_get+0x29/0x80 mlx5_rescan_drivers_locked+0x92/0x160 [mlx5_core] mlx5_unregister_device+0x34/0x50 [mlx5_core] mlx5_uninit_one+0x43/0xb0 [mlx5_core] remove_one+0x4e/0xc0 [mlx5_core] pci_device_remove+0x39/0xa0 device_release_driver_internal+0x194/0x1f0 unbind_store+0x99/0xa0 kernfs_fop_write_iter+0x12e/0x1e0 vfs_write+0x215/0x3d0 ksys_write+0x5f/0xd0 do_syscall_64+0x55/0xe90 entry_SYSCALL_64_after_hwframe+0x4b/0x53 [2] CPU0 (primary) CPU1 (secondary) ========================================================================== mlx5e_remove() (device_lock held) mlx5e_remove() (2nd device_lock held) mlx5_sd_get_adev() mlx5_devcom_comp_is_ready() => true device_lock(primary) mlx5_sd_get_adev() ==> ret adev _mlx5e_remove() mlx5_sd_cleanup() // mlx5e_remove finished // releasing device_lock //need another check here... mlx5_devcom_comp_is_ready() => false Fixes: 381978d28317 ("net/mlx5e: Create single netdev per SD group") Signed-off-by: Shay Drory Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260504180206.268568-5-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/mellanox/mlx5/core/en_main.c | 11 +++++- .../net/ethernet/mellanox/mlx5/core/lib/sd.c | 39 +++++++++++++++++-- .../net/ethernet/mellanox/mlx5/core/lib/sd.h | 2 + 3 files changed, 48 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 62b70334a13d..8f2b3abe0092 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -6774,8 +6774,10 @@ static int mlx5e_resume(struct auxiliary_device *adev) return err; actual_adev = mlx5_sd_get_adev(mdev, adev, edev->idx); - if (actual_adev) + if (actual_adev) { err = _mlx5e_resume(actual_adev); + mlx5_sd_put_adev(actual_adev, adev); + } return err; } @@ -6815,6 +6817,8 @@ static int mlx5e_suspend(struct auxiliary_device *adev, pm_message_t state) err = _mlx5e_suspend(actual_adev, false); mlx5_sd_cleanup(mdev); + if (actual_adev) + mlx5_sd_put_adev(actual_adev, adev); return err; } @@ -6916,11 +6920,14 @@ static int mlx5e_probe(struct auxiliary_device *adev, err = _mlx5e_probe(actual_adev); if (err) goto sd_cleanup; + mlx5_sd_put_adev(actual_adev, adev); } return 0; sd_cleanup: mlx5_sd_cleanup(mdev); + if (actual_adev) + mlx5_sd_put_adev(actual_adev, adev); return err; } @@ -6973,6 +6980,8 @@ static void mlx5e_remove(struct auxiliary_device *adev) _mlx5e_remove(actual_adev); mlx5_sd_cleanup(mdev); + if (actual_adev) + mlx5_sd_put_adev(actual_adev, adev); } static const struct auxiliary_device_id mlx5e_id_table[] = { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c index 89b7e4d67303..6e199161b008 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.c @@ -562,22 +562,55 @@ void mlx5_sd_cleanup(struct mlx5_core_dev *dev) sd_cleanup(dev); } +/* Lock order: + * primary: actual_adev_lock -> SD devcom comp lock + * secondary: SD devcom comp lock -> (drop) -> actual_adev_lock + * The two locks are never held together, so no ABBA. + */ struct auxiliary_device *mlx5_sd_get_adev(struct mlx5_core_dev *dev, struct auxiliary_device *adev, int idx) { struct mlx5_sd *sd = mlx5_get_sd(dev); struct mlx5_core_dev *primary; + struct mlx5_adev *primary_adev; if (!sd) return adev; - if (!mlx5_devcom_comp_is_ready(sd->devcom)) + mlx5_devcom_comp_lock(sd->devcom); + if (!mlx5_devcom_comp_is_ready(sd->devcom)) { + mlx5_devcom_comp_unlock(sd->devcom); return NULL; + } primary = mlx5_sd_get_primary(dev); - if (dev == primary) + if (!primary || dev == primary) { + mlx5_devcom_comp_unlock(sd->devcom); return adev; + } - return &primary->priv.adev[idx]->adev; + primary_adev = primary->priv.adev[idx]; + get_device(&primary_adev->adev.dev); + mlx5_devcom_comp_unlock(sd->devcom); + + device_lock(&primary_adev->adev.dev); + /* Primary may have completed remove between dropping devcom and + * acquiring device_lock; recheck. + */ + if (!mlx5_devcom_comp_is_ready(sd->devcom)) { + device_unlock(&primary_adev->adev.dev); + put_device(&primary_adev->adev.dev); + return NULL; + } + return &primary_adev->adev; +} + +void mlx5_sd_put_adev(struct auxiliary_device *actual_adev, + struct auxiliary_device *adev) +{ + if (actual_adev != adev) { + device_unlock(&actual_adev->dev); + put_device(&actual_adev->dev); + } } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.h b/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.h index 137efaf9aabc..9bfd5b9756b5 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/sd.h @@ -15,6 +15,8 @@ struct mlx5_core_dev *mlx5_sd_ch_ix_get_dev(struct mlx5_core_dev *primary, int c struct auxiliary_device *mlx5_sd_get_adev(struct mlx5_core_dev *dev, struct auxiliary_device *adev, int idx); +void mlx5_sd_put_adev(struct auxiliary_device *actual_adev, + struct auxiliary_device *adev); int mlx5_sd_init(struct mlx5_core_dev *dev); void mlx5_sd_cleanup(struct mlx5_core_dev *dev); From 525cb7ba6661074c1c5cc3772bccc6afab6791ef Mon Sep 17 00:00:00 2001 From: Tzung-Bi Shih Date: Tue, 5 May 2026 05:34:03 +0000 Subject: [PATCH 389/525] platform/chrome: cros_ec_typec: Init mutex in Thunderbolt registration cros_typec_register_thunderbolt() missed initializing the `adata->lock` mutex. This leads to a NULL dereference when the mutex is later acquired (e.g. in cros_typec_altmode_work()). Initialize the mutex in cros_typec_register_thunderbolt() to fix the issue. Cc: stable@vger.kernel.org Fixes: 3b00be26b16a ("platform/chrome: cros_ec_typec: Thunderbolt support") Reviewed-by: Benson Leung Reviewed-by: Abhishek Pandit-Subedi Link: https://lore.kernel.org/r/20260505053403.3335740-1-tzungbi@kernel.org Signed-off-by: Tzung-Bi Shih --- drivers/platform/chrome/cros_typec_altmode.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/platform/chrome/cros_typec_altmode.c b/drivers/platform/chrome/cros_typec_altmode.c index 557340b53af0..66c546bf89b5 100644 --- a/drivers/platform/chrome/cros_typec_altmode.c +++ b/drivers/platform/chrome/cros_typec_altmode.c @@ -359,6 +359,7 @@ cros_typec_register_thunderbolt(struct cros_typec_port *port, } INIT_WORK(&adata->work, cros_typec_altmode_work); + mutex_init(&adata->lock); adata->alt = alt; adata->port = port; adata->ap_mode_entry = true; From 60c71369ee356d11ad845ddeb28ceb70ec6cd70e Mon Sep 17 00:00:00 2001 From: Nathan Chancellor Date: Wed, 11 Mar 2026 17:39:56 -0700 Subject: [PATCH 390/525] powerpc/vdso: Drop -DCC_USING_PATCHABLE_FUNCTION_ENTRY from 32-bit flags with clang After commit 73cdf24e81e4 ("powerpc64: make clang cross-build friendly"), building 64-bit little endian + CONFIG_COMPAT=y with clang results in many warnings along the lines of: $ cat arch/powerpc/configs/compat.config CONFIG_COMPAT=y $ make -skj"$(nproc)" ARCH=powerpc LLVM=1 ppc64le_defconfig compat.config arch/powerpc/kernel/vdso/ ... In file included from :4: In file included from lib/vdso/gettimeofday.c:6: In file included from include/vdso/datapage.h:15: In file included from include/vdso/cache.h:5: arch/powerpc/include/asm/cache.h:77:8: warning: unknown attribute 'patchable_function_entry' ignored [-Wunknown-attributes] 77 | static inline u32 l1_icache_bytes(void) | ^~~~~~ include/linux/compiler_types.h:235:58: note: expanded from macro 'inline' 235 | #define inline inline __gnu_inline __inline_maybe_unused notrace | ^~~~~~~ include/linux/compiler_types.h:215:34: note: expanded from macro 'notrace' 215 | #define notrace __attribute__((patchable_function_entry(0, 0))) | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ... arch/powerpc/Makefile adds -DCC_USING_PATCHABLE_FUNCTION_ENTRY to KBUILD_CPPFLAGS, which is inherited by the 32-bit vDSO. However, the 32-bit little endian target does not support '-fpatchable-function-entry', resulting in the warnings above. Remove -DCC_USING_PATCHABLE_FUNCTION_ENTRY from the 32-bit vDSO flags when building with clang to avoid the warnings. Fixes: 73cdf24e81e4 ("powerpc64: make clang cross-build friendly") Signed-off-by: Nathan Chancellor Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260311-ppc-vdso-drop-cc-using-pfe-define-clang-v1-1-66c790e22650@kernel.org --- arch/powerpc/kernel/vdso/Makefile | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/arch/powerpc/kernel/vdso/Makefile b/arch/powerpc/kernel/vdso/Makefile index 8834dfe9d727..368759f81708 100644 --- a/arch/powerpc/kernel/vdso/Makefile +++ b/arch/powerpc/kernel/vdso/Makefile @@ -62,6 +62,12 @@ CC32FLAGSREMOVE += -fno-stack-clash-protection # 32-bit one. clang validates the values passed to these arguments during # parsing, even when -fno-stack-protector is passed afterwards. CC32FLAGSREMOVE += -mstack-protector-guard% +# ftrace is disabled for the vdso but arch/powerpc/Makefile adds this define to +# KBUILD_CPPFLAGS, which enables use of the 'patchable_function_entry' +# attribute in the 'inline' define via 'notrace'. This attribute is not +# supported for the powerpcle target, resulting in many instances of +# -Wunknown-attributes. +CC32FLAGSREMOVE += -DCC_USING_PATCHABLE_FUNCTION_ENTRY endif LD32FLAGS := -Wl,-soname=linux-vdso32.so.1 AS32FLAGS := -D__VDSO32__ From 8333e4916040e529bd5b56b82d573aba51e88a14 Mon Sep 17 00:00:00 2001 From: Krzysztof Kozlowski Date: Tue, 17 Mar 2026 14:08:24 +0100 Subject: [PATCH 391/525] powerpc/ps3: Drop redundant result assignment Return value of ps3_start_probe_thread() is not used, so code can be simplified to fix W=1 clang warnings: arch/powerpc/platforms/ps3/device-init.c:953:6: error: variable 'result' set but not used [-Werror,-Wunused-but-set-variable] Signed-off-by: Krzysztof Kozlowski Reviewed-by: Geert Uytterhoeven Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260317130823.240279-3-krzysztof.kozlowski@oss.qualcomm.com --- arch/powerpc/platforms/ps3/device-init.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/arch/powerpc/platforms/ps3/device-init.c b/arch/powerpc/platforms/ps3/device-init.c index 12c473768c39..9109c218a060 100644 --- a/arch/powerpc/platforms/ps3/device-init.c +++ b/arch/powerpc/platforms/ps3/device-init.c @@ -950,8 +950,6 @@ static int __init ps3_start_probe_thread(enum ps3_bus_type bus_type) static int __init ps3_register_devices(void) { - int result; - if (!firmware_has_feature(FW_FEATURE_PS3_LV1)) return -ENODEV; @@ -959,7 +957,7 @@ static int __init ps3_register_devices(void) /* ps3_repository_dump_bus_info(); */ - result = ps3_start_probe_thread(PS3_BUS_TYPE_STORAGE); + ps3_start_probe_thread(PS3_BUS_TYPE_STORAGE); ps3_register_vuart_devices(); From f583bd5f64d40e083dde5bb22846c4d93e59d471 Mon Sep 17 00:00:00 2001 From: Krzysztof Kozlowski Date: Tue, 17 Mar 2026 14:08:25 +0100 Subject: [PATCH 392/525] powerpc/pasemi: Drop redundant res assignment Return value of pas_add_bridge() is not used, so code can be simplified to fix W=1 clang warnings: arch/powerpc/platforms/pasemi/pci.c:275:6: error: variable 'res' set but not used [-Werror,-Wunused-but-set-variable] Signed-off-by: Krzysztof Kozlowski Signed-off-by: Madhavan Srinivasan Link: https://patch.msgid.link/20260317130823.240279-4-krzysztof.kozlowski@oss.qualcomm.com --- arch/powerpc/platforms/pasemi/pci.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/powerpc/platforms/pasemi/pci.c b/arch/powerpc/platforms/pasemi/pci.c index 60f990a336c4..2df955274652 100644 --- a/arch/powerpc/platforms/pasemi/pci.c +++ b/arch/powerpc/platforms/pasemi/pci.c @@ -272,13 +272,12 @@ void __init pas_pci_init(void) { struct device_node *root = of_find_node_by_path("/"); struct device_node *np; - int res; pci_set_flags(PCI_SCAN_ALL_PCIE_DEVS); np = of_find_compatible_node(root, NULL, "pasemi,rootbus"); if (np) { - res = pas_add_bridge(np); + pas_add_bridge(np); of_node_put(np); } of_node_put(root); From d73a9a63f9f7f7c17637731fd28daf3665992d1e Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Sat, 2 May 2026 23:07:15 +0300 Subject: [PATCH 393/525] xsk: reject sw-csum UMEM binding to IFF_TX_SKB_NO_LINEAR devices skb_checksum_help() is a common helper that writes the folded 16-bit checksum back via skb->data + csum_start + csum_offset, i.e. it relies on the skb's linear head and fails (with WARN_ONCE and -EINVAL) when skb_headlen() is 0. AF_XDP generic xmit takes two very different paths depending on the netdev. Drivers that advertise IFF_TX_SKB_NO_LINEAR (e.g. virtio_net) skip the "copy payload into a linear head" step on purpose as a performance optimisation: xsk_build_skb_zerocopy() only attaches UMEM pages as frags and never calls skb_put(), so skb_headlen() stays 0 for the whole skb. For these skbs there is simply no linear area for skb_checksum_help() to write the csum into - the sw-csum fallback is structurally inapplicable. The patch tries to catch this and reject the combination with error at setup time. Rejecting at bind() converts this silent per-packet failure into a synchronous, actionable -EOPNOTSUPP at setup time. HW csum and launch_time metadata on IFF_TX_SKB_NO_LINEAR drivers are unaffected because they do not call skb_checksum_help(). Without the patch, every descriptor carrying 'XDP_TX_METADATA | XDP_TXMD_FLAGS_CHECKSUM' produces: 1) a WARN_ONCE "offset (N) >= skb_headlen() (0)" from skb_checksum_help(), 2) sendmsg() returning -EINVAL without consuming the descriptor (invalid_descs is not incremented), 3) a wedged TX ring: __xsk_generic_xmit() does not advance the consumer on non-EOVERFLOW errors, so the next sendmsg() re-reads the same descriptor and re-hits the same WARN until the socket is closed. Closes: https://lore.kernel.org/all/20260419045822.843BFC2BCAF@smtp.kernel.org/#t Acked-by: Stanislav Fomichev Signed-off-by: Jason Xing Acked-by: Stanislav Fomichev Signed-off-by: Jason Xing Reviewed-by: Alexander Lobakin Fixes: 30c3055f9c0d ("xsk: wrap generic metadata handling onto separate function") Link: https://patch.msgid.link/20260502200722.53960-2-kerneljasonxing@gmail.com Signed-off-by: Jakub Kicinski --- net/xdp/xsk_buff_pool.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/xdp/xsk_buff_pool.c b/net/xdp/xsk_buff_pool.c index cd7bc50872f6..d981cfdd8535 100644 --- a/net/xdp/xsk_buff_pool.c +++ b/net/xdp/xsk_buff_pool.c @@ -175,6 +175,9 @@ int xp_assign_dev(struct xsk_buff_pool *pool, if (force_zc && force_copy) return -EINVAL; + if (pool->tx_sw_csum && (netdev->priv_flags & IFF_TX_SKB_NO_LINEAR)) + return -EOPNOTSUPP; + if (xsk_get_pool_from_qid(netdev, queue_id)) return -EBUSY; From 0bb7a9caf5c1d6e25ba376ea6b39261ad28550f4 Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Sat, 2 May 2026 23:07:16 +0300 Subject: [PATCH 394/525] xsk: free the skb when hitting the upper bound MAX_SKB_FRAGS Fix it by explicitly adding kfree_skb() before returning back to its caller. How to reproduce it in virtio_net: 1. the current skb is the first one (which means xs->skb is NULL) and hit the limit MAX_SKB_FRAGS. 2. xsk_build_skb_zerocopy() returns -EOVERFLOW. 3. the caller xsk_build_skb() clears skb by using 'skb = NULL;'. This is why bug can be triggered. 4. there is no chance to free this skb anymore. Note that if in this case the xs->skb is not NULL, xsk_build_skb() will call xsk_drop_skb(xs->skb) to do the right thing. Fixes: cf24f5a5feea ("xsk: add support for AF_XDP multi-buffer on Tx path") Acked-by: Stanislav Fomichev Signed-off-by: Jason Xing Reviewed-by: Alexander Lobakin Link: https://patch.msgid.link/20260502200722.53960-3-kerneljasonxing@gmail.com Signed-off-by: Jakub Kicinski --- net/xdp/xsk.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index 887abed25466..d706b1e0bf60 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -856,8 +856,11 @@ static struct sk_buff *xsk_build_skb_zerocopy(struct xdp_sock *xs, addr = buffer - pool->addrs; for (copied = 0, i = skb_shinfo(skb)->nr_frags; copied < len; i++) { - if (unlikely(i >= MAX_SKB_FRAGS)) + if (unlikely(i >= MAX_SKB_FRAGS)) { + if (!xs->skb) + kfree_skb(skb); return ERR_PTR(-EOVERFLOW); + } page = pool->umem->pgs[addr >> PAGE_SHIFT]; get_page(page); From 8cd3c1c6e7d9a1f0954159ec5f2fdaa7f6a48bd8 Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Sat, 2 May 2026 23:07:17 +0300 Subject: [PATCH 395/525] xsk: handle NULL dereference of the skb without frags issue When a first descriptor (xs->skb == NULL) triggers -EOVERFLOW in xsk_build_skb_zerocopy() (e.g., MAX_SKB_FRAGS exceeded), the free_err -EOVERFLOW handler unconditionally dereferences xs->skb via xsk_inc_num_desc(xs->skb) and xsk_drop_skb(xs->skb), causing a NULL pointer dereference. Fix this by guarding the existing xsk_inc_num_desc()/xsk_drop_skb() calls with an xs->skb check (for the continuation case), and add an else branch for the first-descriptor case that manually cancels the one reserved CQ slot and increments invalid_descs by one to account for the single invalid descriptor. Fixes: cf24f5a5feea ("xsk: add support for AF_XDP multi-buffer on Tx path") Acked-by: Stanislav Fomichev Signed-off-by: Jason Xing Reviewed-by: Alexander Lobakin Link: https://patch.msgid.link/20260502200722.53960-4-kerneljasonxing@gmail.com Signed-off-by: Jakub Kicinski --- net/xdp/xsk.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index d706b1e0bf60..06ee260f3afc 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -976,9 +976,14 @@ static struct sk_buff *xsk_build_skb(struct xdp_sock *xs, kfree_skb(skb); if (err == -EOVERFLOW) { - /* Drop the packet */ - xsk_inc_num_desc(xs->skb); - xsk_drop_skb(xs->skb); + if (xs->skb) { + /* Drop the packet */ + xsk_inc_num_desc(xs->skb); + xsk_drop_skb(xs->skb); + } else { + xsk_cq_cancel_locked(xs->pool, 1); + xs->tx->invalid_descs++; + } xskq_cons_release(xs->tx); } else { /* Let application retry */ From 0f3776583d282550dbafe6082a914efcf9094d59 Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Sat, 2 May 2026 23:07:18 +0300 Subject: [PATCH 396/525] xsk: fix use-after-free of xs->skb in xsk_build_skb() free_err path When xsk_build_skb() processes multi-buffer packets in copy mode, the first descriptor stores data into the skb linear area without adding any frags, so nr_frags stays at 0. The caller then sets xs->skb = skb to accumulate subsequent descriptors. If a continuation descriptor fails (e.g. alloc_page returns NULL with -EAGAIN), we jump to free_err where the condition: if (skb && !skb_shinfo(skb)->nr_frags) kfree_skb(skb); evaluates to true because nr_frags is still 0 (the first descriptor used the linear area, not frags). This frees the skb while xs->skb still points to it, creating a dangling pointer. On the next transmit attempt or socket close, xs->skb is dereferenced, causing a use-after-free or double-free. Fix by using a !xs->skb check to handle first frag situation, ensuring we only free skbs that were freshly allocated in this call (xs->skb is NULL) and never free an in-progress multi-buffer skb that the caller still references. Closes: https://lore.kernel.org/all/20260415082654.21026-4-kerneljasonxing@gmail.com/ Fixes: 6b9c129c2f93 ("xsk: remove @first_frag from xsk_build_skb()") Acked-by: Stanislav Fomichev Signed-off-by: Jason Xing Reviewed-by: Alexander Lobakin Link: https://patch.msgid.link/20260502200722.53960-5-kerneljasonxing@gmail.com Signed-off-by: Jakub Kicinski --- net/xdp/xsk.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index 06ee260f3afc..55378c3855d5 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -972,7 +972,7 @@ static struct sk_buff *xsk_build_skb(struct xdp_sock *xs, return skb; free_err: - if (skb && !skb_shinfo(skb)->nr_frags) + if (skb && !xs->skb) kfree_skb(skb); if (err == -EOVERFLOW) { From 3dec153ae484e3b2ddac841156e197ba54c8df94 Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Sat, 2 May 2026 23:07:19 +0300 Subject: [PATCH 397/525] xsk: prevent CQ desync when freeing half-built skbs in xsk_build_skb() Once xsk_skb_init_misc() has been called on an skb, its destructor is set to xsk_destruct_skb(), which submits the descriptor address(es) to the completion queue and advances the CQ producer. If such an skb is subsequently freed via kfree_skb() along an error path - before the skb has ever been handed to the driver - the destructor still runs and submits a bogus, half-initialized address to the CQ. Postpone the init phase when we believe the allocation of first frag is successfully completed. Before this init, skb can be safely freed by kfree_skb(). Closes: https://lore.kernel.org/all/20260419045822.843BFC2BCAF@smtp.kernel.org/ Fixes: c30d084960cf ("xsk: avoid overwriting skb fields for multi-buffer traffic") Acked-by: Stanislav Fomichev Signed-off-by: Jason Xing Reviewed-by: Alexander Lobakin Link: https://patch.msgid.link/20260502200722.53960-6-kerneljasonxing@gmail.com Signed-off-by: Jakub Kicinski --- net/xdp/xsk.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index 55378c3855d5..af3c5752bb63 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -819,8 +819,6 @@ static struct sk_buff *xsk_build_skb_zerocopy(struct xdp_sock *xs, return ERR_PTR(err); skb_reserve(skb, hr); - - xsk_skb_init_misc(skb, xs, desc->addr); if (desc->options & XDP_TX_METADATA) { err = xsk_skb_metadata(skb, buffer, desc, pool, hr); if (unlikely(err)) @@ -917,7 +915,6 @@ static struct sk_buff *xsk_build_skb(struct xdp_sock *xs, if (unlikely(err)) goto free_err; - xsk_skb_init_misc(skb, xs, desc->addr); if (desc->options & XDP_TX_METADATA) { err = xsk_skb_metadata(skb, buffer, desc, xs->pool, hr); @@ -967,6 +964,8 @@ static struct sk_buff *xsk_build_skb(struct xdp_sock *xs, } } + if (!xs->skb) + xsk_skb_init_misc(skb, xs, desc->addr); xsk_inc_num_desc(skb); return skb; From 8c2cff50afdd2b53c7cc2ca2297301c0ffd3e802 Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Sat, 2 May 2026 23:07:20 +0300 Subject: [PATCH 398/525] xsk: avoid skb leak in XDP_TX_METADATA case Fix it by explicitly adding kfree_skb() before returning back to its caller. How to reproduce it in virtio_net: 1. the current skb is the first one (which means no frag and xs->skb is NULL) and users enable metadata feature. 2. xsk_skb_metadata() returns a error code. 3. the caller xsk_build_skb() clears skb by using 'skb = NULL;'. 4. there is no chance to free this skb anymore. Closes: https://lore.kernel.org/all/20260415085204.3F87AC19424@smtp.kernel.org/ Fixes: 30c3055f9c0d ("xsk: wrap generic metadata handling onto separate function") Acked-by: Stanislav Fomichev Signed-off-by: Jason Xing Reviewed-by: Alexander Lobakin Link: https://patch.msgid.link/20260502200722.53960-7-kerneljasonxing@gmail.com Signed-off-by: Jakub Kicinski --- net/xdp/xsk.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index af3c5752bb63..770ba4695a9d 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -821,8 +821,10 @@ static struct sk_buff *xsk_build_skb_zerocopy(struct xdp_sock *xs, skb_reserve(skb, hr); if (desc->options & XDP_TX_METADATA) { err = xsk_skb_metadata(skb, buffer, desc, pool, hr); - if (unlikely(err)) + if (unlikely(err)) { + kfree_skb(skb); return ERR_PTR(err); + } } } else { struct xsk_addrs *xsk_addr; From e0f229025a8e774a695017a376c4a01279c0e66e Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Sat, 2 May 2026 23:07:21 +0300 Subject: [PATCH 399/525] xsk: fix xsk_addrs slab leak on multi-buffer error path When xsk_build_skb() / xsk_build_skb_zerocopy() sees the first continuation descriptor, it promotes destructor_arg from an inlined address to a freshly allocated xsk_addrs (num_descs = 1). The counter is bumped to >= 2 only at the very end of a successful build (by calling xsk_inc_num_desc()). If the build fails in between (e.g. alloc_page() returns NULL with -EAGAIN, or the MAX_SKB_FRAGS overflow hits), we jump to free_err, skip calling xsk_inc_num_desc() to increment num_descs and leave the half-built skb attached to xs->skb for the app to retry. The skb now has 1) destructor_arg = a real xsk_addrs pointer, 2) num_descs = 1 If the app never retries and just close()s the socket, xsk_release() calls xsk_drop_skb() -> xsk_consume_skb(), which decides whether to free xsk_addrs by testing num_descs > 1: if (unlikely(num_descs > 1)) kmem_cache_free(xsk_tx_generic_cache, destructor_arg); Because num_descs is exactly 1 the branch is skipped and the xsk_addrs object is leaked to the xsk_tx_generic_cache slab. Fix it by directly testing if destructor_arg is still addr. Or else it is modified and used to store the newly allocated memory from xsk_tx_generic_cache regardless of increment of num_desc, which we need to handle. Closes: https://lore.kernel.org/all/20260419045824.D9E5EC2BCAF@smtp.kernel.org/ Fixes: 0ebc27a4c67d ("xsk: avoid data corruption on cq descriptor number") Acked-by: Stanislav Fomichev Signed-off-by: Jason Xing Reviewed-by: Alexander Lobakin Link: https://patch.msgid.link/20260502200722.53960-8-kerneljasonxing@gmail.com Signed-off-by: Jakub Kicinski --- net/xdp/xsk.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index 770ba4695a9d..079abd4bcb69 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -685,7 +685,7 @@ static void xsk_cq_submit_addr_locked(struct xsk_buff_pool *pool, spin_lock_irqsave(&pool->cq_prod_lock, flags); idx = xskq_get_prod(pool->cq); - if (unlikely(num_descs > 1)) { + if (unlikely(!xsk_skb_destructor_is_addr(skb))) { xsk_addr = (struct xsk_addrs *)skb_shinfo(skb)->destructor_arg; for (i = 0; i < num_descs; i++) { @@ -740,7 +740,7 @@ static void xsk_consume_skb(struct sk_buff *skb) u32 num_descs = xsk_get_num_desc(skb); struct xsk_addrs *xsk_addr; - if (unlikely(num_descs > 1)) { + if (unlikely(!xsk_skb_destructor_is_addr(skb))) { xsk_addr = (struct xsk_addrs *)skb_shinfo(skb)->destructor_arg; kmem_cache_free(xsk_tx_generic_cache, xsk_addr); } From 203cee647f551abc87b992045cd920b117ff990a Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Sat, 2 May 2026 23:07:22 +0300 Subject: [PATCH 400/525] xsk: fix u64 descriptor address truncation on 32-bit architectures In copy mode TX, xsk_skb_destructor_set_addr() stores the 64-bit descriptor address into skb_shinfo(skb)->destructor_arg (void *) via a uintptr_t cast: skb_shinfo(skb)->destructor_arg = (void *)((uintptr_t)addr | 0x1UL); On 32-bit architectures uintptr_t is 32 bits, so the upper 32 bits of the descriptor address are silently dropped. In XDP_ZEROCOPY unaligned mode the chunk offset is encoded in bits 48-63 of the descriptor address (XSK_UNALIGNED_BUF_OFFSET_SHIFT = 48), meaning the offset is lost entirely. The completion queue then returns a truncated address to userspace, making buffer recycling impossible. Fix this by handling the 32-bit case directly in xsk_skb_destructor_set_addr(): when !CONFIG_64BIT, allocate an xsk_addrs struct (the same path already used for multi-descriptor SKBs) to store the full u64 address. The existing tagged-pointer logic in xsk_skb_destructor_is_addr() stays unchanged: slab pointers returned from kmem_cache_zalloc() are always word-aligned and therefore have bit 0 clear, which correctly identifies them as a struct pointer rather than an inline tagged address on every architecture. Factor the shared kmem_cache_zalloc + destructor_arg assignment into __xsk_addrs_alloc() and add a wrapper xsk_addrs_alloc() that handles the inline-to-list upgrade (is_addr check + get_addr + num_descs = 1). The three former open-coded kmem_cache_zalloc call sites now reduce to a single call each. Propagate the -ENOMEM from xsk_skb_destructor_set_addr() through xsk_skb_init_misc() so the caller can clean up the skb via kfree_skb() before skb->destructor is installed. The overhead is one extra kmem_cache_zalloc per first descriptor on 32-bit only; 64-bit builds are completely unchanged. Closes: https://lore.kernel.org/all/20260419045824.D9E5EC2BCAF@smtp.kernel.org/ Fixes: 0ebc27a4c67d ("xsk: avoid data corruption on cq descriptor number") Signed-off-by: Jason Xing Acked-by: Stanislav Fomichev Reviewed-by: Alexander Lobakin Link: https://patch.msgid.link/20260502200722.53960-9-kerneljasonxing@gmail.com Signed-off-by: Jakub Kicinski --- net/xdp/xsk.c | 88 ++++++++++++++++++++++++++++++++------------------- 1 file changed, 56 insertions(+), 32 deletions(-) diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index 079abd4bcb69..5e5786cd9af5 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -646,9 +646,42 @@ static u64 xsk_skb_destructor_get_addr(struct sk_buff *skb) return (u64)((uintptr_t)skb_shinfo(skb)->destructor_arg & ~0x1UL); } -static void xsk_skb_destructor_set_addr(struct sk_buff *skb, u64 addr) +static struct xsk_addrs *__xsk_addrs_alloc(struct sk_buff *skb, u64 addr) { - skb_shinfo(skb)->destructor_arg = (void *)((uintptr_t)addr | 0x1UL); + struct xsk_addrs *xsk_addr; + + xsk_addr = kmem_cache_zalloc(xsk_tx_generic_cache, GFP_KERNEL); + if (unlikely(!xsk_addr)) + return NULL; + + xsk_addr->addrs[0] = addr; + skb_shinfo(skb)->destructor_arg = (void *)xsk_addr; + return xsk_addr; +} + +static struct xsk_addrs *xsk_addrs_alloc(struct sk_buff *skb) +{ + struct xsk_addrs *xsk_addr; + + if (!xsk_skb_destructor_is_addr(skb)) + return (struct xsk_addrs *)skb_shinfo(skb)->destructor_arg; + + xsk_addr = __xsk_addrs_alloc(skb, xsk_skb_destructor_get_addr(skb)); + if (likely(xsk_addr)) + xsk_addr->num_descs = 1; + return xsk_addr; +} + +static int xsk_skb_destructor_set_addr(struct sk_buff *skb, u64 addr) +{ + if (IS_ENABLED(CONFIG_64BIT)) { + skb_shinfo(skb)->destructor_arg = (void *)((uintptr_t)addr | 0x1UL); + return 0; + } + + if (unlikely(!__xsk_addrs_alloc(skb, addr))) + return -ENOMEM; + return 0; } static void xsk_inc_num_desc(struct sk_buff *skb) @@ -724,14 +757,20 @@ void xsk_destruct_skb(struct sk_buff *skb) sock_wfree(skb); } -static void xsk_skb_init_misc(struct sk_buff *skb, struct xdp_sock *xs, - u64 addr) +static int xsk_skb_init_misc(struct sk_buff *skb, struct xdp_sock *xs, + u64 addr) { + int err; + + err = xsk_skb_destructor_set_addr(skb, addr); + if (unlikely(err)) + return err; + skb->dev = xs->dev; skb->priority = READ_ONCE(xs->sk.sk_priority); skb->mark = READ_ONCE(xs->sk.sk_mark); skb->destructor = xsk_destruct_skb; - xsk_skb_destructor_set_addr(skb, addr); + return 0; } static void xsk_consume_skb(struct sk_buff *skb) @@ -829,18 +868,9 @@ static struct sk_buff *xsk_build_skb_zerocopy(struct xdp_sock *xs, } else { struct xsk_addrs *xsk_addr; - if (xsk_skb_destructor_is_addr(skb)) { - xsk_addr = kmem_cache_zalloc(xsk_tx_generic_cache, - GFP_KERNEL); - if (!xsk_addr) - return ERR_PTR(-ENOMEM); - - xsk_addr->num_descs = 1; - xsk_addr->addrs[0] = xsk_skb_destructor_get_addr(skb); - skb_shinfo(skb)->destructor_arg = (void *)xsk_addr; - } else { - xsk_addr = (struct xsk_addrs *)skb_shinfo(skb)->destructor_arg; - } + xsk_addr = xsk_addrs_alloc(skb); + if (!xsk_addr) + return ERR_PTR(-ENOMEM); /* in case of -EOVERFLOW that could happen below, * xsk_consume_skb() will release this node as whole skb @@ -929,19 +959,10 @@ static struct sk_buff *xsk_build_skb(struct xdp_sock *xs, struct page *page; u8 *vaddr; - if (xsk_skb_destructor_is_addr(skb)) { - xsk_addr = kmem_cache_zalloc(xsk_tx_generic_cache, - GFP_KERNEL); - if (!xsk_addr) { - err = -ENOMEM; - goto free_err; - } - - xsk_addr->num_descs = 1; - xsk_addr->addrs[0] = xsk_skb_destructor_get_addr(skb); - skb_shinfo(skb)->destructor_arg = (void *)xsk_addr; - } else { - xsk_addr = (struct xsk_addrs *)skb_shinfo(skb)->destructor_arg; + xsk_addr = xsk_addrs_alloc(skb); + if (!xsk_addr) { + err = -ENOMEM; + goto free_err; } if (unlikely(nr_frags == (MAX_SKB_FRAGS - 1) && xp_mb_desc(desc))) { @@ -966,8 +987,11 @@ static struct sk_buff *xsk_build_skb(struct xdp_sock *xs, } } - if (!xs->skb) - xsk_skb_init_misc(skb, xs, desc->addr); + if (!xs->skb) { + err = xsk_skb_init_misc(skb, xs, desc->addr); + if (unlikely(err)) + goto free_err; + } xsk_inc_num_desc(skb); return skb; From bd3c45dd01283ada23b0a388c578dcf5600deb8a Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Thu, 23 Apr 2026 18:53:49 +0200 Subject: [PATCH 401/525] timers/migration: Fix another hotplug activation race The hotplug control CPU is assumed to be active in the hierarchy but that doesn't imply that the root is active. If the current CPU is not the one that activated the current hierarchy, and the CPU performing this duty is still halfway through the tree, the root may still be observed inactive. And this can break the activation of a new root as in the following scenario: 1) Initially, the whole system has 64 CPUs and only CPU 63 is awake. [GRP1:0] active / | \ / | \ [GRP0:0] [...] [GRP0:7] idle idle active / | \ | CPU 0 CPU 1 ... CPU 63 idle idle active 2) CPU 63 goes idle _but_ due to a #VMEXIT it hasn't yet reached the [GRP1:0]->parent dereference (that would be NULL and stop the walk) in __walk_groups_from(). [GRP1:0] idle / | \ / | \ [GRP0:0] [...] [GRP0:7] idle idle idle / | \ | CPU 0 CPU 1 ... CPU 63 idle idle idle 3) CPU 1 wakes up, activates GRP0:0 but didn't yet manage to propagate up to GRP1:0 due to yet another #VMEXIT. [GRP1:0] idle / | \ / | \ [GRP0:0] [...] [GRP0:7] active idle idle / | \ | CPU 0 CPU 1 ... CPU 63 idle active idle 3) CPU 0 wakes up and doesn't need to walk above GRP0:0 as it's CPU 1 role. [GRP1:0] idle / | \ / | \ [GRP0:0] [...] [GRP0:7] active idle idle / | \ | CPU 0 CPU 1 ... CPU 63 active active idle 4) CPU 0 boots CPU 64. It creates a new root for it. [GRP2:0] idle / \ / \ [GRP1:0] [GRP1:1] idle idle / | \ \ / | \ \ [GRP0:0] [...] [GRP0:7] [GRP0:8] active idle idle idle / | \ | | CPU 0 CPU 1 ... CPU 63 CPU 64 active active idle offline 5) CPU 0 activates the new root, but note that GRP1:0 is still idle, waiting for CPU 1 to resume from #VMEXIT and activate it. [GRP2:0] active / \ / \ [GRP1:0] [GRP1:1] idle idle / | \ \ / | \ \ [GRP0:0] [...] [GRP0:7] [GRP0:8] active idle idle idle / | \ | | CPU 0 CPU 1 ... CPU 63 CPU 64 active active idle offline 6) CPU 63 resumes after #VMEXIT and sees the new GRP1:0 parent. Therefore it propagates the stale inactive state of GRP1:0 up to GRP2:0. [GRP2:0] idle / \ / \ [GRP1:0] [GRP1:1] idle idle / | \ \ / | \ \ [GRP0:0] [...] [GRP0:7] [GRP0:8] active idle idle idle / | \ | | CPU 0 CPU 1 ... CPU 63 CPU 64 active active idle offline 7) CPU 1 resumes after #VMEXIT and finally activates GRP1:0. But it doesn't observe its parent link because no ordering enforced that. Therefore GRP2:0 is spuriously left idle. [GRP2:0] idle / \ / \ [GRP1:0] [GRP1:1] active idle / | \ \ / | \ \ [GRP0:0] [...] [GRP0:7] [GRP0:8] active idle idle idle / | \ | | CPU 0 CPU 1 ... CPU 63 CPU 64 active active idle offline Such races are highly theoretical and the problem would solve itself once the old root ever becomes idle again. But it still leaves a taste of discomfort. Fix it with enforcing a fully ordered atomic read of the old root state before propagating the activate state up to the new root. It has a two directions ordering effect: * Acquire + release of the latest old root state: If the hotplug control CPU is not the one that woke up the old root, make sure to acquire its active state and propagate it upwards through the ordered chain of activation (the acquire pairs with the cmpxchg() in tmigr_active_up() and subsequent releases will pair with atomic_read_acquire() and smp_mb__after_atomic() in tmigr_inactive_up()). * Release: If the hotplug control CPU is not the one that must wake up the old root, but the CPU covering that is lagging behind its duty, publish the links from the old root to the new parents. This way the lagging CPU will propagate the active state itself. Fixes: 7ee988770326 ("timers: Implement the hierarchical pull model") Signed-off-by: Frederic Weisbecker Signed-off-by: Thomas Gleixner Link: https://patch.msgid.link/20260423165354.95152-2-frederic@kernel.org --- kernel/time/timer_migration.c | 40 +++++++++++++++++++++++++---------- 1 file changed, 29 insertions(+), 11 deletions(-) diff --git a/kernel/time/timer_migration.c b/kernel/time/timer_migration.c index 155eeaea4113..1d0d3a4058d5 100644 --- a/kernel/time/timer_migration.c +++ b/kernel/time/timer_migration.c @@ -1860,19 +1860,37 @@ static int tmigr_setup_groups(unsigned int cpu, unsigned int node, * child to the new parents. So tmigr_active_up() activates the * new parents while walking up from the old root to the new. * - * * It is ensured that @start is active, as this setup path is - * executed in hotplug prepare callback. This is executed by an - * already connected and !idle CPU. Even if all other CPUs go idle, - * the CPU executing the setup will be responsible up to current top - * level group. And the next time it goes inactive, it will release - * the new childmask and parent to subsequent walkers through this - * @child. Therefore propagate active state unconditionally. + * * It is ensured that @start is active, (or on the way to be activated + * by another CPU that woke up before the current one) as this setup path + * is executed in hotplug prepare callback. This is executed by an already + * connected and !idle CPU in the hierarchy. + * + * * The below RmW atomic operation ensures that: + * + * 1) If the old root has been completely activated, the latest state is + * acquired (the below implicit acquire pairs with the implicit release + * from cmpxchg() in tmigr_active_up()). + * + * 2) If the old root is still on the way to be activated, the lagging behind + * CPU performing the activation will acquire the links up to the new root. + * (The below implicit release pairs with the implicit acquire from cmpxchg() + * in tmigr_active_up()). + * + * 3) Every subsequent CPU below the old root will acquire the new links while + * walking through the old root (The below implicit release pairs with the + * implicit acquire from cmpxchg() in either tmigr_active_up()) or + * tmigr_inactive_up(). */ - state.state = atomic_read(&start->migr_state); - WARN_ON_ONCE(!state.active); + state.state = atomic_fetch_or(0, &start->migr_state); WARN_ON_ONCE(!start->parent); - data.childmask = start->groupmask; - __walk_groups_from(tmigr_active_up, &data, start, start->parent); + /* + * If the state of the old root is inactive, another CPU is on its way to activate + * it and propagate to the new root. + */ + if (state.active) { + data.childmask = start->groupmask; + __walk_groups_from(tmigr_active_up, &data, start, start->parent); + } } /* Root update */ From 92429ca999db99febced82f23362a71b2ba4c1d8 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?C=C3=A1ssio=20Gabriel?= Date: Wed, 6 May 2026 00:15:48 -0300 Subject: [PATCH 402/525] ALSA: seq: Fix UMP group 16 filtering MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The sequencer UAPI defines group_filter as an unsigned int bitmap. Bit 0 filters groupless messages and bits 1-16 filter UMP groups 1-16. The internal snd_seq_client storage is only unsigned short, so bit 16 is truncated when userspace sets the filter. The same truncation affects the automatic UMP client filter used to avoid delivery to inactive groups, so events for group 16 cannot be filtered. Store the internal bitmap as unsigned int and keep both userspace-provided and automatically generated values limited to the defined UAPI bits. Fixes: d2b706077792 ("ALSA: seq: Add UMP group filter") Cc: stable@vger.kernel.org Signed-off-by: Cássio Gabriel Link: https://patch.msgid.link/20260506-alsa-seq-ump-group16-filter-v1-1-b75160bf6993@gmail.com Signed-off-by: Takashi Iwai --- sound/core/seq/seq_clientmgr.c | 2 +- sound/core/seq/seq_clientmgr.h | 5 ++++- sound/core/seq/seq_ump_client.c | 2 +- 3 files changed, 6 insertions(+), 3 deletions(-) diff --git a/sound/core/seq/seq_clientmgr.c b/sound/core/seq/seq_clientmgr.c index 75a7a2af9d8c..5719637575a9 100644 --- a/sound/core/seq/seq_clientmgr.c +++ b/sound/core/seq/seq_clientmgr.c @@ -1253,7 +1253,7 @@ static int snd_seq_ioctl_set_client_info(struct snd_seq_client *client, if (client->user_pversion >= SNDRV_PROTOCOL_VERSION(1, 0, 3)) client->midi_version = client_info->midi_version; memcpy(client->event_filter, client_info->event_filter, 32); - client->group_filter = client_info->group_filter; + client->group_filter = client_info->group_filter & SND_SEQ_GROUP_FILTER_MASK; /* notify the change */ snd_seq_system_client_ev_client_change(client->number); diff --git a/sound/core/seq/seq_clientmgr.h b/sound/core/seq/seq_clientmgr.h index ece02c58db70..feea8bb7d987 100644 --- a/sound/core/seq/seq_clientmgr.h +++ b/sound/core/seq/seq_clientmgr.h @@ -14,6 +14,9 @@ /* client manager */ +#define SND_SEQ_GROUP_FILTER_MASK GENMASK(SNDRV_UMP_MAX_GROUPS, 0) +#define SND_SEQ_GROUP_FILTER_GROUPS GENMASK(SNDRV_UMP_MAX_GROUPS, 1) + struct snd_seq_user_client { struct file *file; /* file struct of client */ /* ... */ @@ -40,7 +43,7 @@ struct snd_seq_client { int number; /* client number */ unsigned int filter; /* filter flags */ DECLARE_BITMAP(event_filter, 256); - unsigned short group_filter; + unsigned int group_filter; snd_use_lock_t use_lock; int event_lost; /* ports */ diff --git a/sound/core/seq/seq_ump_client.c b/sound/core/seq/seq_ump_client.c index fdc76f23e03f..9079ccfdc866 100644 --- a/sound/core/seq/seq_ump_client.c +++ b/sound/core/seq/seq_ump_client.c @@ -369,7 +369,7 @@ static void setup_client_group_filter(struct seq_ump_client *client) cptr = snd_seq_kernel_client_get(client->seq_client); if (!cptr) return; - filter = ~(1U << 0); /* always allow groupless messages */ + filter = SND_SEQ_GROUP_FILTER_GROUPS; /* always allow groupless messages */ for (p = 0; p < SNDRV_UMP_MAX_GROUPS; p++) { if (client->ump->groups[p].active) filter &= ~(1U << (p + 1)); From 01801e20d69346e1e6cec0d908f1cea3a49e51b5 Mon Sep 17 00:00:00 2001 From: Rodrigo Faria Date: Tue, 5 May 2026 19:55:18 +0100 Subject: [PATCH 403/525] ALSA: hda/realtek: Add mute LED fixup for HP Pavilion 15-cs1xxx Add a new fixup for the mute LED on the HP Pavilion 15-cs1xxx series using the VREF on NID 0x1b. The BIOS on these models (tested up to F.32) incorrectly reports the mute LED on NID 0x18 via DMI OEM strings, which lacks VREF capabilities. This fixup overrides the LED pin to the correct NID 0x1b. Signed-off-by: Rodrigo Faria Link: https://patch.msgid.link/20260505185518.23625-1-rodrigofilipefaria@gmail.com Signed-off-by: Takashi Iwai --- sound/hda/codecs/realtek/alc269.c | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/sound/hda/codecs/realtek/alc269.c b/sound/hda/codecs/realtek/alc269.c index a7525ed83e2b..11d0ea8ed859 100644 --- a/sound/hda/codecs/realtek/alc269.c +++ b/sound/hda/codecs/realtek/alc269.c @@ -1669,6 +1669,21 @@ static void alc295_fixup_hp_mute_led_coefbit11(struct hda_codec *codec, } } +/* Override wrong pin to NID 0x1b (F.32 BIOS reports 0x18 via DMI OEM string) + * on HP pavilion 15-cs1xxx laptops + */ +static void alc295_fixup_hp_pavilion_mute_led_1b(struct hda_codec *codec, + const struct hda_fixup *fix, + int action) +{ + struct alc_spec *spec = codec->spec; + + alc269_fixup_hp_mute_led(codec, fix, action); + + if (action == HDA_FIXUP_ACT_PRE_PROBE) + spec->mute_led_nid = 0x1b; +} + static void alc233_fixup_lenovo_coef_micmute_led(struct hda_codec *codec, const struct hda_fixup *fix, int action) { @@ -3870,6 +3885,7 @@ enum { ALC290_FIXUP_SUBWOOFER, ALC290_FIXUP_SUBWOOFER_HSJACK, ALC295_FIXUP_HP_MUTE_LED_COEFBIT11, + ALC295_FIXUP_HP_PAVILION_MUTE_LED_1B, ALC269_FIXUP_THINKPAD_ACPI, ALC269_FIXUP_LENOVO_XPAD_ACPI, ALC269_FIXUP_DMIC_THINKPAD_ACPI, @@ -5715,6 +5731,10 @@ static const struct hda_fixup alc269_fixups[] = { .type = HDA_FIXUP_FUNC, .v.func = alc295_fixup_hp_mute_led_coefbit11, }, + [ALC295_FIXUP_HP_PAVILION_MUTE_LED_1B] = { + .type = HDA_FIXUP_FUNC, + .v.func = alc295_fixup_hp_pavilion_mute_led_1b, + }, [ALC298_FIXUP_SAMSUNG_AMP] = { .type = HDA_FIXUP_FUNC, .v.func = alc298_fixup_samsung_amp, @@ -6931,6 +6951,7 @@ static const struct hda_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x103c, 0x8537, "HP ProBook 440 G6", ALC236_FIXUP_HP_MUTE_LED_MICMUTE_VREF), SND_PCI_QUIRK(0x103c, 0x8548, "HP EliteBook x360 830 G6", ALC285_FIXUP_HP_GPIO_LED), SND_PCI_QUIRK(0x103c, 0x854a, "HP EliteBook 830 G6", ALC285_FIXUP_HP_GPIO_LED), + SND_PCI_QUIRK(0x103c, 0x856a, "HP Pavilion 15-cs1xxx", ALC295_FIXUP_HP_PAVILION_MUTE_LED_1B), SND_PCI_QUIRK(0x103c, 0x85c6, "HP Pavilion x360 Convertible 14-dy1xxx", ALC295_FIXUP_HP_MUTE_LED_COEFBIT11), SND_PCI_QUIRK(0x103c, 0x85de, "HP Envy x360 13-ar0xxx", ALC285_FIXUP_HP_ENVY_X360), SND_PCI_QUIRK(0x103c, 0x8603, "HP Omen 17-cb0xxx", ALC285_FIXUP_HP_MUTE_LED), From 5337213381df578058e2e41da93cbd0e4639935f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?C=C3=A1ssio=20Gabriel?= Date: Wed, 6 May 2026 00:34:47 -0300 Subject: [PATCH 404/525] ALSA: core: Serialize deferred fasync state checks MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit snd_fasync_helper() updates fasync->on under snd_fasync_lock, and snd_fasync_work_fn() now also evaluates fasync->on under the same lock. snd_kill_fasync() still tests the flag before taking the lock, leaving an unsynchronized read against FASYNC enable/disable updates. Move the enabled-state check into the locked section. Also clear fasync->on under snd_fasync_lock in snd_fasync_free() before unlinking the pending entry. Together with the locked sender-side check, this publishes teardown before flushing the deferred work and prevents a racing sender from requeueing the entry after free has started. Fixes: ef34a0ae7a26 ("ALSA: core: Add async signal helpers") Fixes: 8146cd333d23 ("ALSA: core: Fix potential data race at fasync handling") Cc: stable@vger.kernel.org Signed-off-by: Cássio Gabriel Link: https://patch.msgid.link/20260506-alsa-core-fasync-on-lock-v1-1-ea48c77d6ca4@gmail.com Signed-off-by: Takashi Iwai --- sound/core/misc.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/sound/core/misc.c b/sound/core/misc.c index 5aca09edf971..833124c8e4fa 100644 --- a/sound/core/misc.c +++ b/sound/core/misc.c @@ -148,9 +148,11 @@ EXPORT_SYMBOL_GPL(snd_fasync_helper); void snd_kill_fasync(struct snd_fasync *fasync, int signal, int poll) { - if (!fasync || !fasync->on) + if (!fasync) return; guard(spinlock_irqsave)(&snd_fasync_lock); + if (!fasync->on) + return; fasync->signal = signal; fasync->poll = poll; list_move(&fasync->list, &snd_fasync_list); @@ -163,8 +165,10 @@ void snd_fasync_free(struct snd_fasync *fasync) if (!fasync) return; - scoped_guard(spinlock_irq, &snd_fasync_lock) + scoped_guard(spinlock_irq, &snd_fasync_lock) { + fasync->on = 0; list_del_init(&fasync->list); + } flush_work(&snd_fasync_work); kfree(fasync); From 2bcbb163162789d3488562073dbb99d9bd71a762 Mon Sep 17 00:00:00 2001 From: Rosen Penev Date: Tue, 5 May 2026 20:18:54 -0700 Subject: [PATCH 405/525] ALSA: sparc/dbri: add missing fallthrough Fixes compiler error with probably newer compilers: sound/sparc/dbri.c:595:2: error: unannotated fall-through between switch labels [-Werror,-Wimplicit-fallthrough] 595 | case 1: | ^ sound/sparc/dbri.c:595:2: note: insert 'break;' to avoid fall-through 595 | case 1: | ^ | break; Signed-off-by: Rosen Penev Link: https://patch.msgid.link/20260506031854.780411-1-rosenp@gmail.com Signed-off-by: Takashi Iwai --- sound/sparc/dbri.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/sparc/dbri.c b/sound/sparc/dbri.c index 75f82a92ff44..2f5f62079fa4 100644 --- a/sound/sparc/dbri.c +++ b/sound/sparc/dbri.c @@ -592,6 +592,7 @@ static __u32 reverse_bytes(__u32 b, int len) fallthrough; case 2: b = ((b & 0xaaaaaaaa) >> 1) | ((b & 0x55555555) << 1); + fallthrough; case 1: case 0: break; From 283fc9e44ff5b5ac967439b4951b80bd4299f4e4 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 5 May 2026 15:15:34 +0200 Subject: [PATCH 406/525] wifi: mac80211: remove station if connection prep fails If connection preparation fails for MLO connections, then the interface is completely reset to non-MLD. In this case, we must not keep the station since it's related to the link of the vif being removed. Delete an existing station. Any "new_sta" is already being removed, so that doesn't need changes. This fixes a use-after-free/double-free in debugfs if that's enabled, because a vif going from MLD (and to MLD, but that's not relevant here) recreates its entire debugfs. Cc: stable@vger.kernel.org Fixes: 81151ce462e5 ("wifi: mac80211: support MLO authentication/association with one link") Reviewed-by: Miriam Rachel Korenblit Link: https://patch.msgid.link/20260505151533.c4e52deb06ad.Iafe56cec7de8512626169496b134bce3a6c17010@changeid Signed-off-by: Johannes Berg --- net/mac80211/mlme.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 298ebff6bbf8..0a0f27836d57 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -9149,7 +9149,7 @@ static int ieee80211_prep_connection(struct ieee80211_sub_if_data *sdata, struct ieee80211_bss *bss = (void *)cbss->priv; struct sta_info *new_sta = NULL; struct ieee80211_link_data *link; - bool have_sta = false; + struct sta_info *have_sta = NULL; bool mlo; int err; u16 new_links; @@ -9168,11 +9168,8 @@ static int ieee80211_prep_connection(struct ieee80211_sub_if_data *sdata, mlo = false; } - if (assoc) { - rcu_read_lock(); + if (assoc) have_sta = sta_info_get(sdata, ap_mld_addr); - rcu_read_unlock(); - } if (mlo && !have_sta && WARN_ON(sdata->vif.valid_links || sdata->vif.active_links)) @@ -9336,6 +9333,8 @@ static int ieee80211_prep_connection(struct ieee80211_sub_if_data *sdata, out_release_chan: ieee80211_link_release_channel(link); out_err: + if (mlo && have_sta) + WARN_ON(__sta_info_destroy(have_sta)); ieee80211_vif_set_links(sdata, 0, 0); return err; } From 0f3c0a197309717d74729568f88957d448847937 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 5 May 2026 13:38:37 +0200 Subject: [PATCH 407/525] wifi: nl80211: fix NL80211_PMSR_FTM_REQ_ATTR_FTMS_PER_BURST usage This is documented as a u8 and has a policy of NLA_U8, but uses nla_get_u32() which means it's completely broken on big-endian. Fix it to use nla_get_u8(). Fixes: 9bb7e0f24e7e ("cfg80211: add peer measurement with FTM initiator API") Link: https://patch.msgid.link/20260505113837.260159-2-johannes@sipsolutions.net Signed-off-by: Johannes Berg --- net/wireless/pmsr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/wireless/pmsr.c b/net/wireless/pmsr.c index 4c8ea0583f94..d6cd0de64d1f 100644 --- a/net/wireless/pmsr.c +++ b/net/wireless/pmsr.c @@ -88,7 +88,7 @@ static int pmsr_parse_ftm(struct cfg80211_registered_device *rdev, out->ftm.ftms_per_burst = 0; if (tb[NL80211_PMSR_FTM_REQ_ATTR_FTMS_PER_BURST]) out->ftm.ftms_per_burst = - nla_get_u32(tb[NL80211_PMSR_FTM_REQ_ATTR_FTMS_PER_BURST]); + nla_get_u8(tb[NL80211_PMSR_FTM_REQ_ATTR_FTMS_PER_BURST]); if (capa->ftm.max_ftms_per_burst && (out->ftm.ftms_per_burst > capa->ftm.max_ftms_per_burst || From 15994bb0cbb8fc4879da7552ddd08c1896261c39 Mon Sep 17 00:00:00 2001 From: Maoyi Xie Date: Wed, 6 May 2026 14:48:53 +0800 Subject: [PATCH 408/525] wifi: nl80211: require CAP_NET_ADMIN over the target netns in SET_WIPHY_NETNS NL80211_CMD_SET_WIPHY_NETNS dispatches with GENL_UNS_ADMIN_PERM, which verifies that the caller has CAP_NET_ADMIN for the source netns. It doesn't verify that the caller has CAP_NET_ADMIN over the target netns selected by NL80211_ATTR_NETNS_FD or NL80211_ATTR_PID. This diverges from the convention enforced in net/core/rtnetlink.c::rtnl_get_net_ns_capable(): /* For now, the caller is required to have CAP_NET_ADMIN in * the user namespace owning the target net ns. */ if (!sk_ns_capable(sk, net->user_ns, CAP_NET_ADMIN)) return ERR_PTR(-EACCES); A user with CAP_NET_ADMIN in their own user namespace can therefore push a wiphy into an arbitrary netns (including init_net) over which they have no privilege. Mirror the rtnetlink convention by requiring CAP_NET_ADMIN in the target netns before calling cfg80211_switch_netns(). Signed-off-by: Maoyi Xie Link: https://patch.msgid.link/20260506064854.2207105-2-maoyixie.tju@gmail.com Signed-off-by: Johannes Berg --- net/wireless/nl80211.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 67088804dcc7..db546dd93d08 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -13867,6 +13867,19 @@ static int nl80211_wiphy_netns(struct sk_buff *skb, struct genl_info *info) if (IS_ERR(net)) return PTR_ERR(net); + /* + * The caller already has CAP_NET_ADMIN over the source netns + * (enforced by GENL_UNS_ADMIN_PERM on the genl op). Mirror the + * convention used by net/core/rtnetlink.c::rtnl_get_net_ns_capable() + * and require CAP_NET_ADMIN over the target netns as well, so that + * a caller that is privileged in their own user namespace cannot + * push a wiphy into a netns where they have no privilege. + */ + if (!ns_capable(net->user_ns, CAP_NET_ADMIN)) { + put_net(net); + return -EPERM; + } + err = 0; /* check if anything to do */ From 79240f3f6d766b342b57c32397d643e1cfa26b81 Mon Sep 17 00:00:00 2001 From: Maoyi Xie Date: Wed, 6 May 2026 14:48:54 +0800 Subject: [PATCH 409/525] wifi: nl80211: re-check wiphy netns in nl80211_prepare_wdev_dump() continuation NL80211_CMD_GET_SCAN is implemented as a multi-call dumpit. The first invocation of nl80211_prepare_wdev_dump() validates the requested wdev against the caller's netns via __cfg80211_wdev_from_attrs(). Subsequent invocations look up the same wiphy by its global index and do not check that the wiphy is still in the caller's netns. Add the same filter to the continuation path. If the wiphy's netns no longer matches the caller's, return -ENODEV and the netlink dump machinery terminates the walk cleanly. Signed-off-by: Maoyi Xie Link: https://patch.msgid.link/20260506064854.2207105-3-maoyixie.tju@gmail.com Signed-off-by: Johannes Berg --- net/wireless/nl80211.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index db546dd93d08..7db9cd433801 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -1276,6 +1276,18 @@ static int nl80211_prepare_wdev_dump(struct netlink_callback *cb, rtnl_unlock(); return -ENODEV; } + + /* + * The first invocation validated the wdev's netns against + * the caller via __cfg80211_wdev_from_attrs(). The wiphy + * may have moved netns between dumpit invocations (via + * NL80211_CMD_SET_WIPHY_NETNS), so re-check here. + */ + if (!net_eq(wiphy_net(wiphy), sock_net(cb->skb->sk))) { + rtnl_unlock(); + return -ENODEV; + } + *rdev = wiphy_to_rdev(wiphy); *wdev = NULL; From 86f33ca9bea30cf011f2b1edad4593faea9c6e98 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Wed, 6 May 2026 16:22:38 +0800 Subject: [PATCH 410/525] ublk: validate physical_bs_shift, io_min_shift and io_opt_shift ublk_validate_params() checks logical_bs_shift is within [9, PAGE_SHIFT] but has no upper bound for physical_bs_shift, io_min_shift, or io_opt_shift. A malicious userspace can set any of these to a large value (e.g., 44), causing undefined behavior from `1 << shift` in ublk_ctrl_start_dev() since the result is stored in 32-bit unsigned int. Cap all three at ilog2(SZ_256M) (28). 256M is big enough to cover all practical block sizes, and originates from the maximum physical block size possible in NVMe (lba_size * (1 + npwg), where npwg is 16-bit). Also zero out ub->params with memset() when copy_from_user() fails or ublk_validate_params() returns error, so that no stale or partial params survive for a subsequent START_DEV to consume. Fixes: 71f28f3136af ("ublk_drv: add io_uring based userspace block driver") Signed-off-by: Ming Lei Link: https://patch.msgid.link/20260506082238.22363-1-tom.leiming@gmail.com Signed-off-by: Jens Axboe --- drivers/block/ublk_drv.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c index d10460d29e4a..57ec900f0ce0 100644 --- a/drivers/block/ublk_drv.c +++ b/drivers/block/ublk_drv.c @@ -900,6 +900,20 @@ static int ublk_validate_params(const struct ublk_device *ub) if (p->logical_bs_shift > PAGE_SHIFT || p->logical_bs_shift < 9) return -EINVAL; + /* + * 256M is a reasonable upper bound for physical block size, + * io_min and io_opt; it aligns with the maximum physical + * block size possible in NVMe. + */ + if (p->physical_bs_shift > ilog2(SZ_256M)) + return -EINVAL; + + if (p->io_min_shift > ilog2(SZ_256M)) + return -EINVAL; + + if (p->io_opt_shift > ilog2(SZ_256M)) + return -EINVAL; + if (p->logical_bs_shift > p->physical_bs_shift) return -EINVAL; @@ -4992,13 +5006,15 @@ static int ublk_ctrl_set_params(struct ublk_device *ub, */ ret = -EACCES; } else if (copy_from_user(&ub->params, argp, ph.len)) { + /* zero out partial copy so no stale params survive */ + memset(&ub->params, 0, sizeof(ub->params)); ret = -EFAULT; } else { /* clear all we don't support yet */ ub->params.types &= UBLK_PARAM_TYPE_ALL; ret = ublk_validate_params(ub); if (ret) - ub->params.types = 0; + memset(&ub->params, 0, sizeof(ub->params)); } mutex_unlock(&ub->mutex); From 9cc6bac1bebf8310d2950d1411a91479e86d69a1 Mon Sep 17 00:00:00 2001 From: Maoyi Xie Date: Mon, 4 May 2026 23:37:54 +0800 Subject: [PATCH 411/525] io_uring/timeout: honour caller's time namespace for IORING_TIMEOUT_ABS io_uring's IORING_OP_TIMEOUT and IORING_OP_LINK_TIMEOUT accept a timespec from the caller via io_parse_user_time(). With IORING_TIMEOUT_ABS, the timestamp is an absolute deadline on the selected clock. The clock is CLOCK_MONOTONIC by default. CLOCK_BOOTTIME and CLOCK_REALTIME are also selectable. A submitter inside a CLONE_NEWTIME time namespace observes CLOCK_MONOTONIC and CLOCK_BOOTTIME shifted by the namespace's offsets relative to the host. Every other ABS timer interface in the kernel converts the caller's absolute time to host view via timens_ktime_to_host() before arming an hrtimer: kernel/time/posix-timers.c -- timer_settime(TIMER_ABSTIME) kernel/time/posix-stubs.c -- clock_nanosleep(TIMER_ABSTIME) kernel/time/alarmtimer.c -- alarm_timer_nsleep(TIMER_ABSTIME) fs/timerfd.c -- timerfd_settime(TFD_TIMER_ABSTIME) io_parse_user_time() does not. As a result, an absolute timeout submitted from within a time namespace is interpreted in host view. That is generally a different point in time. It may already be in the past, causing the timer to fire immediately, or far in the future, causing the timer not to fire when expected. Reproducer: in unshare --user --time, with a -10s monotonic offset, submit IORING_OP_TIMEOUT with IORING_TIMEOUT_ABS and deadline = now + 1s. The CQE is delivered after <1ms instead of the expected ~1s. Apply timens_ktime_to_host() to the parsed time when IORING_TIMEOUT_ABS is set. Split the existing clock id resolver in io_timeout_get_clock() into a flags only helper io_flags_to_clock(), so io_parse_user_time() can resolve the clock without a struct io_timeout_data. timens_ktime_to_host() is a no-op for clocks not affected by time namespaces, e.g. CLOCK_REALTIME. It is also a no-op for callers in the initial time namespace. The fast path is unchanged. SQPOLL is also covered. The SQPOLL kernel thread is created via create_io_thread() with CLONE_THREAD and no CLONE_NEW* flag. copy_namespaces() therefore shares the submitter's nsproxy by reference. Inside the SQPOLL kthread, current->nsproxy->time_ns is the submitter's time_ns. timens_ktime_to_host() resolves correctly. Suggested-by: Pavel Begunkov Suggested-by: Jens Axboe Signed-off-by: Maoyi Xie Link: https://patch.msgid.link/20260504153755.1293932-2-maoyi.xie@ntu.edu.sg Signed-off-by: Jens Axboe --- io_uring/timeout.c | 35 ++++++++++++++++++++++------------- 1 file changed, 22 insertions(+), 13 deletions(-) diff --git a/io_uring/timeout.c b/io_uring/timeout.c index 4cfdfc519770..e2595cae2b07 100644 --- a/io_uring/timeout.c +++ b/io_uring/timeout.c @@ -3,6 +3,7 @@ #include #include #include +#include #include @@ -35,6 +36,22 @@ struct io_timeout_rem { bool ltimeout; }; +static clockid_t io_flags_to_clock(unsigned flags) +{ + switch (flags & IORING_TIMEOUT_CLOCK_MASK) { + case IORING_TIMEOUT_BOOTTIME: + return CLOCK_BOOTTIME; + case IORING_TIMEOUT_REALTIME: + return CLOCK_REALTIME; + default: + /* can't happen, vetted at prep time */ + WARN_ON_ONCE(1); + fallthrough; + case 0: + return CLOCK_MONOTONIC; + } +} + static int io_parse_user_time(ktime_t *time, u64 arg, unsigned flags) { struct timespec64 ts; @@ -43,7 +60,7 @@ static int io_parse_user_time(ktime_t *time, u64 arg, unsigned flags) *time = ns_to_ktime(arg); if (*time < 0) return -EINVAL; - return 0; + goto out; } if (get_timespec64(&ts, u64_to_user_ptr(arg))) @@ -51,6 +68,9 @@ static int io_parse_user_time(ktime_t *time, u64 arg, unsigned flags) if (ts.tv_sec < 0 || ts.tv_nsec < 0) return -EINVAL; *time = timespec64_to_ktime(ts); +out: + if (flags & IORING_TIMEOUT_ABS) + *time = timens_ktime_to_host(io_flags_to_clock(flags), *time); return 0; } @@ -399,18 +419,7 @@ static enum hrtimer_restart io_link_timeout_fn(struct hrtimer *timer) static clockid_t io_timeout_get_clock(struct io_timeout_data *data) { - switch (data->flags & IORING_TIMEOUT_CLOCK_MASK) { - case IORING_TIMEOUT_BOOTTIME: - return CLOCK_BOOTTIME; - case IORING_TIMEOUT_REALTIME: - return CLOCK_REALTIME; - default: - /* can't happen, vetted at prep time */ - WARN_ON_ONCE(1); - fallthrough; - case 0: - return CLOCK_MONOTONIC; - } + return io_flags_to_clock(data->flags); } static int io_linked_timeout_update(struct io_ring_ctx *ctx, __u64 user_data, From 45d2b37a37ab98484693533496395c610a2cab96 Mon Sep 17 00:00:00 2001 From: Maoyi Xie Date: Mon, 4 May 2026 23:37:55 +0800 Subject: [PATCH 412/525] io_uring/wait: honour caller's time namespace for IORING_ENTER_ABS_TIMER io_uring_enter() with IORING_ENTER_ABS_TIMER takes an absolute timespec from the caller via ext_arg->ts. It arms an ABS mode hrtimer in __io_cqring_wait_schedule(). The conversion path in io_uring/wait.c parses ext_arg->ts inline rather than going through io_parse_user_time(). It therefore does not pick up the time namespace conversion added by the previous patch. Apply timens_ktime_to_host() to the parsed time on the IORING_ENTER_ABS_TIMER branch. This mirrors the IORING_TIMEOUT_ABS fix in io_parse_user_time(). Use ctx->clockid as the clock id. ctx->clockid is set either at ring creation or via IORING_REGISTER_CLOCK. timens_ktime_to_host() is a no-op for clocks not affected by time namespaces. It is also a no-op for callers in the initial time namespace. The fast path is unchanged. Reproducer: in unshare --user --time, with a -10s monotonic offset, call io_uring_enter with min_complete=1, IORING_ENTER_ABS_TIMER, and ts = now + 1s. The call returns -ETIME after <1ms instead of after the expected ~1s. Suggested-by: Pavel Begunkov Suggested-by: Jens Axboe Signed-off-by: Maoyi Xie Link: https://patch.msgid.link/20260504153755.1293932-3-maoyi.xie@ntu.edu.sg Signed-off-by: Jens Axboe --- io_uring/wait.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/io_uring/wait.c b/io_uring/wait.c index 91df86ce0d18..ec01e78a216d 100644 --- a/io_uring/wait.c +++ b/io_uring/wait.c @@ -5,6 +5,7 @@ #include #include #include +#include #include @@ -229,7 +230,10 @@ int io_cqring_wait(struct io_ring_ctx *ctx, int min_events, u32 flags, if (ext_arg->ts_set) { iowq.timeout = timespec64_to_ktime(ext_arg->ts); - if (!(flags & IORING_ENTER_ABS_TIMER)) + if (flags & IORING_ENTER_ABS_TIMER) + iowq.timeout = timens_ktime_to_host(ctx->clockid, + iowq.timeout); + else iowq.timeout = ktime_add(iowq.timeout, start_time); } From 5cbb61bf4168859d97c068d88d364f4f1f440325 Mon Sep 17 00:00:00 2001 From: Breno Leitao Date: Tue, 5 May 2026 09:02:13 -0700 Subject: [PATCH 413/525] arm64/fpsimd: ptrace: zero target's fpsimd_state, not the tracer's sve_set_common() is the backend for PTRACE_SETREGSET(NT_ARM_SVE) and PTRACE_SETREGSET(NT_ARM_SSVE). Every write in the function operates on the tracee (target) - except a single memset that uses current instead, zeroing the tracer's saved V0-V31 / FPSR / FPCR shadow on every ptrace SETREGSET call. The memset is meant to give the tracee a defined zero register image before the user-supplied payload is copied in (for partial writes, header-only writes, and FPSIMD<->SVE format switches). Aiming it at current both denies the tracee that clean slate and silently corrupts the tracer. The corruption of the tracer's saved FPSIMD state is not always observable. Where the tracer's state is live on a CPU, this may be reused without loading the corrupted state from memory, and will eventually be written back over the corrupted state. Where the tracer's state is saved in SVE_PT_REGS_SVE format, only the FPSR and FPCR are clobbered, and the effective copy of the vectors is in the task's sve_state. Reproducible on an arm64 kernel with SVE: a single-threaded tracer that loads a known pattern into V0-V31, issues PTRACE_SETREGSET(NT_ARM_SVE) on a child, and reads V0-V31 back observes them all zeroed within tens of thousands of iterations when a sibling thread keeps stealing the FPSIMD CPU binding. Fixes: 316283f276eb ("arm64/fpsimd: ptrace: Consistently handle partial writes to NT_ARM_(S)SVE") Cc: Signed-off-by: Breno Leitao Acked-by: Mark Rutland Signed-off-by: Catalin Marinas --- arch/arm64/kernel/ptrace.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm64/kernel/ptrace.c b/arch/arm64/kernel/ptrace.c index ba5eab23fd90..4d08598e2891 100644 --- a/arch/arm64/kernel/ptrace.c +++ b/arch/arm64/kernel/ptrace.c @@ -983,8 +983,8 @@ static int sve_set_common(struct task_struct *target, } /* Always zero V regs, FPSR, and FPCR */ - memset(¤t->thread.uw.fpsimd_state, 0, - sizeof(current->thread.uw.fpsimd_state)); + memset(&target->thread.uw.fpsimd_state, 0, + sizeof(target->thread.uw.fpsimd_state)); /* Registers: FPSIMD-only case */ From bee87cf0f1248c0f20710d7a79df41fe892d9f88 Mon Sep 17 00:00:00 2001 From: Richard Fitzgerald Date: Tue, 5 May 2026 17:11:23 +0100 Subject: [PATCH 414/525] ASoC: cs35l56: Don't use devres to unregister component Manually call snd_soc_unregister_component() from cs35l56_remove() instead of using devres cleanup. This ensures that the component is destroyed before cs35l56_remove() starts cleanup of anything the component code could be using. Devres cleanup happens after the driver remove() callback, so if snd_soc_register_component() is used, it will not be destroyed until after cs35l56_remove() has returned. But there is some cleanup that must be done in cs35l56_remove(), or wrapped in a custom devres cleanup handler to ensure correct ordering. The simplest option is to call snd_soc_unregister_component() at the start of cs35l56_remove(). Fixes: e49611252900 ("ASoC: cs35l56: Add driver for Cirrus Logic CS35L56") Closes: https://sashiko.dev/#/patchset/20260501103002.2843735-1-rf%40opensource.cirrus.com Signed-off-by: Richard Fitzgerald Link: https://patch.msgid.link/20260505161124.3621000-2-rf@opensource.cirrus.com Signed-off-by: Mark Brown --- sound/soc/codecs/cs35l56.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/sound/soc/codecs/cs35l56.c b/sound/soc/codecs/cs35l56.c index 378017fcea10..e847bb32af2e 100644 --- a/sound/soc/codecs/cs35l56.c +++ b/sound/soc/codecs/cs35l56.c @@ -1956,9 +1956,9 @@ int cs35l56_common_probe(struct cs35l56_private *cs35l56) goto err; } - ret = devm_snd_soc_register_component(cs35l56->base.dev, - &soc_component_dev_cs35l56, - cs35l56_dai, ARRAY_SIZE(cs35l56_dai)); + ret = snd_soc_register_component(cs35l56->base.dev, + &soc_component_dev_cs35l56, + cs35l56_dai, ARRAY_SIZE(cs35l56_dai)); if (ret < 0) { dev_err_probe(cs35l56->base.dev, ret, "Register codec failed\n"); goto err; @@ -2057,6 +2057,8 @@ EXPORT_SYMBOL_NS_GPL(cs35l56_init, "SND_SOC_CS35L56_CORE"); void cs35l56_remove(struct cs35l56_private *cs35l56) { + snd_soc_unregister_component(cs35l56->base.dev); + cs35l56->base.init_done = false; /* From fd4d83e1437d6395021b21531e187c8a67ac21b0 Mon Sep 17 00:00:00 2001 From: Richard Fitzgerald Date: Tue, 5 May 2026 17:11:24 +0100 Subject: [PATCH 415/525] ASoC: cs35l56: Destroy workqueue in probe error path The error path in cs35l56_common_probe() should call destroy_workqueue() on the workqueue that was created by cs35l56_dsp_init(). Fixes: e49611252900 ("ASoC: cs35l56: Add driver for Cirrus Logic CS35L56") Signed-off-by: Richard Fitzgerald Link: https://patch.msgid.link/20260505161124.3621000-3-rf@opensource.cirrus.com Signed-off-by: Mark Brown --- sound/soc/codecs/cs35l56.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/sound/soc/codecs/cs35l56.c b/sound/soc/codecs/cs35l56.c index e847bb32af2e..849d70ca23d6 100644 --- a/sound/soc/codecs/cs35l56.c +++ b/sound/soc/codecs/cs35l56.c @@ -1970,6 +1970,9 @@ int cs35l56_common_probe(struct cs35l56_private *cs35l56) gpiod_set_value_cansleep(cs35l56->base.reset_gpio, 0); regulator_bulk_disable(ARRAY_SIZE(cs35l56->supplies), cs35l56->supplies); + if (cs35l56->dsp_wq) + destroy_workqueue(cs35l56->dsp_wq); + return ret; } EXPORT_SYMBOL_NS_GPL(cs35l56_common_probe, "SND_SOC_CS35L56_CORE"); From 628497e6d925d43efb56e3ffecef0a9d217926b3 Mon Sep 17 00:00:00 2001 From: Fenglin Wu Date: Wed, 6 May 2026 02:28:51 -0700 Subject: [PATCH 416/525] regulator: qcom-rpmh: Fix index for pmh0101 ldo16 The wrong index is assigned to pmh0101 ldo16, which results incorrect rpmh resource being used when the regulator device is voted. Fix it. Fixes: 65efe5404d15 ("regulator: rpmh-regulator: Add RPMH regulator support for Glymur") Signed-off-by: Fenglin Wu Reviewed-by: Konrad Dybcio Link: https://patch.msgid.link/20260506-fix_pmh0101_ldo16_index-v1-1-cdc8708b01f4@oss.qualcomm.com Signed-off-by: Mark Brown --- drivers/regulator/qcom-rpmh-regulator.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/regulator/qcom-rpmh-regulator.c b/drivers/regulator/qcom-rpmh-regulator.c index 6e4cb2871fca..0dcb50bf5c35 100644 --- a/drivers/regulator/qcom-rpmh-regulator.c +++ b/drivers/regulator/qcom-rpmh-regulator.c @@ -1512,7 +1512,7 @@ static const struct rpmh_vreg_init_data pmh0101_vreg_data[] = { RPMH_VREG("ldo13", LDO, 13, &pmic5_pldo530_mvp150, "vdd-l2-l13-l14"), RPMH_VREG("ldo14", LDO, 14, &pmic5_pldo530_mvp150, "vdd-l2-l13-l14"), RPMH_VREG("ldo15", LDO, 15, &pmic5_nldo530, "vdd-l15"), - RPMH_VREG("ldo16", LDO, 15, &pmic5_pldo530_mvp600, "vdd-l5-l16"), + RPMH_VREG("ldo16", LDO, 16, &pmic5_pldo530_mvp600, "vdd-l5-l16"), RPMH_VREG("ldo17", LDO, 17, &pmic5_pldo515_mv, "vdd-l17"), RPMH_VREG("ldo18", LDO, 18, &pmic5_nldo530, "vdd-l18"), RPMH_VREG("bob1", BOB, 1, &pmic5_bob, "vdd-bob1"), From 4bacec2317527ba04b7172145848f1c206999ea1 Mon Sep 17 00:00:00 2001 From: Jiawei Liu Date: Wed, 6 May 2026 14:24:12 +0800 Subject: [PATCH 417/525] spi: ch341: correct company name in MODULE_DESCRIPTION The company name "QiHeng Electronics" is incorrect. The correct legal name is "Nanjing Qinheng Microelectronics Co., Ltd.". Update the module description accordingly. Signed-off-by: Jiawei Liu Link: https://patch.msgid.link/20260506062412.371034-1-ljw@wch.cn Signed-off-by: Mark Brown --- drivers/spi/spi-ch341.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/spi/spi-ch341.c b/drivers/spi/spi-ch341.c index 3eaa8f176f63..6448a44a8b67 100644 --- a/drivers/spi/spi-ch341.c +++ b/drivers/spi/spi-ch341.c @@ -250,5 +250,5 @@ static struct usb_driver ch341a_usb_driver = { module_usb_driver(ch341a_usb_driver); MODULE_AUTHOR("Johannes Thumshirn "); -MODULE_DESCRIPTION("QiHeng Electronics ch341 USB2SPI"); +MODULE_DESCRIPTION("Nanjing Qinheng Microelectronics CH341 USB2SPI driver"); MODULE_LICENSE("GPL v2"); From 0c8c88b8eb82a2a41bec5f17c076d6312dc40316 Mon Sep 17 00:00:00 2001 From: Colin Walters Date: Tue, 5 May 2026 15:42:57 -0700 Subject: [PATCH 418/525] ovl: fix verity lazy-load guard broken by fsverity_active() semantic change Commit f77f281b6118 ("fsverity: use a hashtable to find the fsverity_info") made fsverity_active() check whether the inode has the verity flag, rather than whether the inode's fsverity_info is loaded. This broke ovl_ensure_verity_loaded(), which wants to load the fsverity_info for any verity inodes that haven't had it loaded yet. Therefore, to check that the fsverity_info hasn't yet been loaded, use fsverity_get_info(inode) == NULL instead of !fsverity_active(inode). Also, since fsverity_get_info() now involves a hash table lookup, put the more lightweight IS_VERITY() flag check first. Fixes: f77f281b6118 ("fsverity: use a hashtable to find the fsverity_info") Cc: stable@vger.kernel.org Link: https://github.com/bootc-dev/bootc/issues/2174 Signed-off-by: Colin Walters Acked-by: Amir Goldstein Link: https://patch.msgid.link/20260505224257.23213-1-ebiggers@kernel.org Signed-off-by: Eric Biggers --- fs/overlayfs/util.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/overlayfs/util.c b/fs/overlayfs/util.c index 7b86a6bac644..b41f4788e4f0 100644 --- a/fs/overlayfs/util.c +++ b/fs/overlayfs/util.c @@ -1354,7 +1354,7 @@ int ovl_ensure_verity_loaded(const struct path *datapath) struct inode *inode = d_inode(datapath->dentry); struct file *filp; - if (!fsverity_active(inode) && IS_VERITY(inode)) { + if (IS_VERITY(inode) && fsverity_get_info(inode) == NULL) { /* * If this inode was not yet opened, the verity info hasn't been * loaded yet, so we need to do that here to force it into memory. From fdf4eb632683bfc2840acebe62716cb468d43e10 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sun, 26 Apr 2026 17:51:07 +0200 Subject: [PATCH 419/525] selftests/rseq: Validate legacy behavior The RSEQ legacy mode behavior requires that the ID fields in the rseq region are unconditionally updated on every context switch and before signal delivery even if not required by the ABI specification. To ensure that this behavior is preserved for legacy users in the future, add a test which validates that with a sleep() and a signal sent to self. Provide a run script which prevents GLIBC from registering a RSEQ region, so that the test can register it's own legacy sized region. Fixes: 566d8015f7ee ("rseq: Avoid CPU/MM CID updates when no event pending") Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Link: https://patch.msgid.link/20260428224427.764705536%40kernel.org Cc: stable@vger.kernel.org --- tools/testing/selftests/rseq/Makefile | 5 +- tools/testing/selftests/rseq/legacy_check.c | 65 +++++++++++++++++++ .../selftests/rseq/run_legacy_check.sh | 4 ++ 3 files changed, 72 insertions(+), 2 deletions(-) create mode 100644 tools/testing/selftests/rseq/legacy_check.c create mode 100755 tools/testing/selftests/rseq/run_legacy_check.sh diff --git a/tools/testing/selftests/rseq/Makefile b/tools/testing/selftests/rseq/Makefile index 0d1947c0d623..0293a2f17f51 100644 --- a/tools/testing/selftests/rseq/Makefile +++ b/tools/testing/selftests/rseq/Makefile @@ -22,9 +22,10 @@ TEST_GEN_PROGS_EXTENDED = librseq.so \ param_test_compare_twice \ param_test_mm_cid \ param_test_mm_cid_compare_twice \ - syscall_errors_test + syscall_errors_test \ + legacy_check -TEST_PROGS = run_param_test.sh run_syscall_errors_test.sh +TEST_PROGS = run_param_test.sh run_syscall_errors_test.sh run_legacy_check.sh TEST_FILES := settings diff --git a/tools/testing/selftests/rseq/legacy_check.c b/tools/testing/selftests/rseq/legacy_check.c new file mode 100644 index 000000000000..3f7de4e28303 --- /dev/null +++ b/tools/testing/selftests/rseq/legacy_check.c @@ -0,0 +1,65 @@ +// SPDX-License-Identifier: GPL-2.0 +#ifndef _GNU_SOURCE +#define _GNU_SOURCE +#endif + +#include +#include +#include +#include + +#include "rseq.h" + +#include "../kselftest_harness.h" + +FIXTURE(legacy) +{ +}; + +static int cpu_id_in_sigfn = -1; + +static void sigfn(int sig) +{ + struct rseq_abi *rs = rseq_get_abi(); + + cpu_id_in_sigfn = rs->cpu_id_start; +} + +FIXTURE_SETUP(legacy) +{ + int res = __rseq_register_current_thread(true, true); + + switch (res) { + case -ENOSYS: + SKIP(return, "RSEQ not enabled\n"); + case -EBUSY: + SKIP(return, "GLIBC owns RSEQ. Disable GLIBC RSEQ registration\n"); + default: + ASSERT_EQ(res, 0); + } + + ASSERT_NE(signal(SIGUSR1, sigfn), SIG_ERR); +} + +FIXTURE_TEARDOWN(legacy) +{ +} + +TEST_F(legacy, legacy_test) +{ + struct rseq_abi *rs = rseq_get_abi(); + + ASSERT_NE(rs, NULL); + + /* Overwrite rs::cpu_id_start */ + rs->cpu_id_start = -1; + sleep(1); + ASSERT_NE(rs->cpu_id_start, -1); + + rs->cpu_id_start = -1; + ASSERT_EQ(raise(SIGUSR1), 0); + ASSERT_NE(rs->cpu_id_start, -1); + ASSERT_NE(cpu_id_in_sigfn, -1); +} + +TEST_HARNESS_MAIN diff --git a/tools/testing/selftests/rseq/run_legacy_check.sh b/tools/testing/selftests/rseq/run_legacy_check.sh new file mode 100755 index 000000000000..5577b46ea092 --- /dev/null +++ b/tools/testing/selftests/rseq/run_legacy_check.sh @@ -0,0 +1,4 @@ +#!/bin/bash +# SPDX-License-Identifier: GPL-2.0 + +GLIBC_TUNABLES="${GLIBC_TUNABLES:-}:glibc.pthread.rseq=0" ./legacy_check From 82f572449cfe75f12ea985986da60e11f308f77d Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sun, 26 Apr 2026 16:21:02 +0200 Subject: [PATCH 420/525] rseq: Implement read only ABI enforcement for optimized RSEQ V2 mode The optimized RSEQ V2 mode requires that user space adheres to the ABI specification and does not modify the read-only fields cpu_id_start, cpu_id, node_id and mm_cid behind the kernel's back. While the kernel does not rely on these fields, the adherence to this is a fundamental prerequisite to allow multiple entities, e.g. libraries, in an application to utilize the full potential of RSEQ without stepping on each other toes. Validate this adherence on every update of these fields. If the kernel detects that user space modified the fields, the application is force terminated. Fixes: d6200245c75e ("rseq: Allow registering RSEQ with slice extension") Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Link: https://patch.msgid.link/20260428224427.845230956%40kernel.org Cc: stable@vger.kernel.org --- include/linux/rseq_entry.h | 83 ++++++++++++++------------------------ include/linux/rseq_types.h | 4 +- kernel/rseq.c | 5 +-- 3 files changed, 35 insertions(+), 57 deletions(-) diff --git a/include/linux/rseq_entry.h b/include/linux/rseq_entry.h index 934db41ec782..2d0295df5107 100644 --- a/include/linux/rseq_entry.h +++ b/include/linux/rseq_entry.h @@ -248,7 +248,6 @@ static __always_inline bool rseq_grant_slice_extension(unsigned long ti_work, un #endif /* !CONFIG_RSEQ_SLICE_EXTENSION */ bool rseq_debug_update_user_cs(struct task_struct *t, struct pt_regs *regs, unsigned long csaddr); -bool rseq_debug_validate_ids(struct task_struct *t); static __always_inline void rseq_note_user_irq_entry(void) { @@ -368,43 +367,6 @@ bool rseq_debug_update_user_cs(struct task_struct *t, struct pt_regs *regs, return false; } -/* - * On debug kernels validate that user space did not mess with it if the - * debug branch is enabled. - */ -bool rseq_debug_validate_ids(struct task_struct *t) -{ - struct rseq __user *rseq = t->rseq.usrptr; - u32 cpu_id, uval, node_id; - - /* - * On the first exit after registering the rseq region CPU ID is - * RSEQ_CPU_ID_UNINITIALIZED and node_id in user space is 0! - */ - node_id = t->rseq.ids.cpu_id != RSEQ_CPU_ID_UNINITIALIZED ? - cpu_to_node(t->rseq.ids.cpu_id) : 0; - - scoped_user_read_access(rseq, efault) { - unsafe_get_user(cpu_id, &rseq->cpu_id_start, efault); - if (cpu_id != t->rseq.ids.cpu_id) - goto die; - unsafe_get_user(uval, &rseq->cpu_id, efault); - if (uval != cpu_id) - goto die; - unsafe_get_user(uval, &rseq->node_id, efault); - if (uval != node_id) - goto die; - unsafe_get_user(uval, &rseq->mm_cid, efault); - if (uval != t->rseq.ids.mm_cid) - goto die; - } - return true; -die: - t->rseq.event.fatal = true; -efault: - return false; -} - #endif /* RSEQ_BUILD_SLOW_PATH */ /* @@ -514,20 +476,32 @@ rseq_update_user_cs(struct task_struct *t, struct pt_regs *regs, unsigned long c * faults in task context are fatal too. */ static rseq_inline -bool rseq_set_ids_get_csaddr(struct task_struct *t, struct rseq_ids *ids, - u32 node_id, u64 *csaddr) +bool rseq_set_ids_get_csaddr(struct task_struct *t, struct rseq_ids *ids, u64 *csaddr) { struct rseq __user *rseq = t->rseq.usrptr; - if (static_branch_unlikely(&rseq_debug_enabled)) { - if (!rseq_debug_validate_ids(t)) - return false; - } - scoped_user_rw_access(rseq, efault) { + /* Validate the R/O fields for debug and optimized mode */ + if (static_branch_unlikely(&rseq_debug_enabled) || rseq_v2(t)) { + u32 cpu_id, uval; + + unsafe_get_user(cpu_id, &rseq->cpu_id_start, efault); + if (cpu_id != t->rseq.ids.cpu_id) + goto die; + unsafe_get_user(uval, &rseq->cpu_id, efault); + if (uval != cpu_id) + goto die; + unsafe_get_user(uval, &rseq->node_id, efault); + if (uval != t->rseq.ids.node_id) + goto die; + unsafe_get_user(uval, &rseq->mm_cid, efault); + if (uval != t->rseq.ids.mm_cid) + goto die; + } + unsafe_put_user(ids->cpu_id, &rseq->cpu_id_start, efault); unsafe_put_user(ids->cpu_id, &rseq->cpu_id, efault); - unsafe_put_user(node_id, &rseq->node_id, efault); + unsafe_put_user(ids->node_id, &rseq->node_id, efault); unsafe_put_user(ids->mm_cid, &rseq->mm_cid, efault); if (csaddr) unsafe_get_user(*csaddr, &rseq->rseq_cs, efault); @@ -539,10 +513,13 @@ bool rseq_set_ids_get_csaddr(struct task_struct *t, struct rseq_ids *ids, rseq_slice_clear_grant(t); /* Cache the new values */ - t->rseq.ids.cpu_cid = ids->cpu_cid; + t->rseq.ids = *ids; rseq_stat_inc(rseq_stats.ids); rseq_trace_update(t, ids); return true; + +die: + t->rseq.event.fatal = true; efault: return false; } @@ -552,11 +529,11 @@ bool rseq_set_ids_get_csaddr(struct task_struct *t, struct rseq_ids *ids, * is in a critical section. */ static rseq_inline bool rseq_update_usr(struct task_struct *t, struct pt_regs *regs, - struct rseq_ids *ids, u32 node_id) + struct rseq_ids *ids) { u64 csaddr; - if (!rseq_set_ids_get_csaddr(t, ids, node_id, &csaddr)) + if (!rseq_set_ids_get_csaddr(t, ids, &csaddr)) return false; /* @@ -659,12 +636,12 @@ static __always_inline bool rseq_exit_user_update(struct pt_regs *regs, struct t } struct rseq_ids ids = { - .cpu_id = task_cpu(t), - .mm_cid = task_mm_cid(t), + .cpu_id = task_cpu(t), + .mm_cid = task_mm_cid(t), + .node_id = cpu_to_node(ids.cpu_id), }; - u32 node_id = cpu_to_node(ids.cpu_id); - return rseq_update_usr(t, regs, &ids, node_id); + return rseq_update_usr(t, regs, &ids); efault: return false; } diff --git a/include/linux/rseq_types.h b/include/linux/rseq_types.h index a469c1870849..85739a63e85e 100644 --- a/include/linux/rseq_types.h +++ b/include/linux/rseq_types.h @@ -66,8 +66,9 @@ struct rseq_event { * compiler emit a single compare on 64-bit * @cpu_id: The CPU ID which was written last to user space * @mm_cid: The MM CID which was written last to user space + * @node_id: The node ID which was written last to user space * - * @cpu_id and @mm_cid are updated when the data is written to user space. + * @cpu_id, @mm_cid and @node_id are updated when the data is written to user space. */ struct rseq_ids { union { @@ -77,6 +78,7 @@ struct rseq_ids { u32 mm_cid; }; }; + u32 node_id; }; /** diff --git a/kernel/rseq.c b/kernel/rseq.c index aa25753ea135..101612027f6a 100644 --- a/kernel/rseq.c +++ b/kernel/rseq.c @@ -263,7 +263,6 @@ static void rseq_slowpath_update_usr(struct pt_regs *regs) }; struct task_struct *t = current; struct rseq_ids ids; - u32 node_id; bool event; if (unlikely(t->flags & PF_EXITING)) @@ -299,9 +298,9 @@ static void rseq_slowpath_update_usr(struct pt_regs *regs) if (!event) return; - node_id = cpu_to_node(ids.cpu_id); + ids.node_id = cpu_to_node(ids.cpu_id); - if (unlikely(!rseq_update_usr(t, regs, &ids, node_id))) { + if (unlikely(!rseq_update_usr(t, regs, &ids))) { /* * Clear the errors just in case this might survive magically, but * leave the rest intact. From 99428157dcf32fdac97355aa1cc1364dbc9e073c Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sun, 26 Apr 2026 10:01:56 +0200 Subject: [PATCH 421/525] rseq: Reenable performance optimizations conditionally Due to the incompatibility with TCMalloc the RSEQ optimizations and extended features (time slice extensions) have been disabled and made run-time conditional. The original RSEQ implementation, which TCMalloc depends on, registers a 32 byte region (ORIG_RSEG_SIZE). This region has a 32 byte alignment requirement. The extension safe newer variant exposes the kernel RSEQ feature size via getauxval(AT_RSEQ_FEATURE_SIZE) and the alignment requirement via getauxval(AT_RSEQ_ALIGN). The alignment requirement is that the registered RSEQ region is aligned to the next power of two of the feature size. The kernel currently has a feature size of 33 bytes, which means the alignment requirement is 64 bytes. The TCMalloc RSEQ region is embedded into a cache line aligned data structure starting at offset 32 bytes so that bytes 28-31 and the cpu_id_start field at bytes 32-35 form a 64-bit little endian pointer with the top-most bit (63 set) to check whether the kernel has overwritten cpu_id_start with an actual CPU id value, which is guaranteed to not have the top most bit set. As this is part of their performance tuned magic, it's a pretty safe assumption, that TCMalloc won't use a larger RSEQ size. This allows the kernel to declare that registrations with a size greater than the original size of 32 bytes, which is the cases since time slice extensions got introduced, as RSEQ ABI v2 with the following differences to the original behaviour: 1) Unconditional updates of the user read only fields (CPU, node, MMCID) are removed. Those fields are only updated on registration, task migration and MMCID changes. 2) Unconditional evaluation of the criticial section pointer is removed. It's only evaluated when user space was interrupted and was scheduled out or before delivering a signal in the interrupted context. 3) The read/only requirement of the ID fields is enforced. When the kernel detects that userspace manipulated the fields, the process is terminated. This ensures that multiple entities (libraries) can utilize RSEQ without interfering. 4) Todays extended RSEQ feature (time slice extensions) and future extensions are only enabled in the v2 enabled mode. Registrations with the original size of 32 bytes operate in backwards compatible legacy mode without performance improvements and extended features. Unfortunately that also affects users of older GLIBC versions which register the original size of 32 bytes and do not evaluate the kernel required size in the auxiliary vector AT_RSEQ_FEATURE_SIZE. That's the result of the lack of enforcement in the original implementation and the unwillingness of a single entity to cooperate with the larger ecosystem for many years. Implement the required registration changes by restructuring the spaghetti code and adding the size/version check. Also add documentation about the differences of legacy and optimized RSEQ V2 mode. Thanks to Mathieu for pointing out the ORIG_RSEQ_SIZE constraints! Fixes: d6200245c75e ("rseq: Allow registering RSEQ with slice extension") Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Link: https://patch.msgid.link/20260428224427.927160119%40kernel.org Cc: stable@vger.kernel.org --- Documentation/userspace-api/rseq.rst | 94 ++++++++++++++++- kernel/rseq.c | 144 ++++++++++++++++----------- 2 files changed, 178 insertions(+), 60 deletions(-) diff --git a/Documentation/userspace-api/rseq.rst b/Documentation/userspace-api/rseq.rst index 3cd27a3c7c7e..8549a6c61531 100644 --- a/Documentation/userspace-api/rseq.rst +++ b/Documentation/userspace-api/rseq.rst @@ -24,6 +24,97 @@ Quick access to CPU number, node ID Allows to implement per CPU data efficiently. Documentation is in code and selftests. :( +Optimized RSEQ V2 +----------------- + +On architectures which utilize the generic entry code and generic TIF bits +the kernel supports runtime optimizations for RSEQ, which also enable +enhanced features like scheduler time slice extensions. + +To enable them a task has to register the RSEQ region with at least the +length advertised by getauxval(AT_RSEQ_FEATURE_SIZE). + +If existing binaries register with RSEQ_ORIG_SIZE (32 bytes), the kernel +keeps the legacy low performance mode enabled to fulfil the expectations +of existing users regarding the original RSEQ implementation behaviour. + +The following table documents the ABI and behavioral guarantees of the +legacy and the optimized V2 mode. + +.. list-table:: RSEQ modes + :header-rows: 1 + + * - Nr + - What + + - Legacy + - Optimized V2 + + * - 1 + - The cpu_id_start, cpu_id, node_id and mm_cid fields (User mode read + only) + .. Legacy + - Updated by the kernel unconditionally after each context switch and + before signal delivery + .. Optimized V2 + - Updated by the kernel if and only if they change, i.e. if the task + is migrated or mm_cid changes + + * - 2 + - The rseq_cs critical section field + .. Legacy + - Evaluated and handled unconditionally after each context switch and + before signal delivery + .. Optimized V2 + - Evaluated and handled conditionally only when user space was + interrupted and was scheduled out or before delivering a signal in + the interrupted context. + + * - 3 + - Read only fields + .. Legacy + - No strict enforcement except in debug mode + .. Optimized V2 + - Strict enforcement + + * - 4 + - membarrier(...RSEQ) + .. Legacy + - All running threads of the process are interrupted and the ID fields + are rewritten and eventually active critical sections are aborted + before they return to user space. All threads which are scheduled + out whether voluntary or not are covered by #1/#2 above. + .. Optimized V2 + - All running threads of the process are interrupted and eventually + active critical sections are aborted before these threads return to + user space. The ID fields are only updated if changed as a + consequence of the interrupt. All threads which are scheduled out + whether voluntary or not are covered by #1/#2 above. + + * - 5 + - Time slice extensions + .. Legacy + - Not supported + .. Optimized V2 + - Supported + +The legacy mode is obviously less performant as it does unconditional +updates and critical section checks even if not strictly required by the +ABI contract. That can't be changed anymore as some users depend on that +observed behavior, which in turn enables them to violate the ABI and +overwrite the cpu_id_start field for their own purposes. This is obviously +discouraged as it renders RSEQ incompatible with the intended usage and +breaks the expectation of other libraries in the same application. + +The ABI compliant optimized v2 mode, which respects the read only fields, +does not require unconditional updates and therefore is way more +performant. The kernel validates the read only fields for compliance. If +user space modifies them, the process is killed. Compliant usage allows +multiple libraries in the same application to benefit from the RSEQ +functionality without disturbing each other. The ABI compliant optimized v2 +mode also enables extended RSEQ features like time slice extensions. + + Scheduler time slice extensions ------------------------------- @@ -37,7 +128,8 @@ The prerequisites for this functionality are: * Enabled at boot time (default is enabled) - * A rseq userspace pointer has been registered for the thread + * A rseq userspace pointer has been registered for the thread in + optimized V2 mode The thread has to enable the functionality via prctl(2):: diff --git a/kernel/rseq.c b/kernel/rseq.c index 101612027f6a..e75e3a5e312c 100644 --- a/kernel/rseq.c +++ b/kernel/rseq.c @@ -412,70 +412,23 @@ static bool rseq_reset_ids(void) /* The original rseq structure size (including padding) is 32 bytes. */ #define ORIG_RSEQ_SIZE 32 -/* - * sys_rseq - setup restartable sequences for caller thread. - */ -SYSCALL_DEFINE4(rseq, struct rseq __user *, rseq, u32, rseq_len, int, flags, u32, sig) +static long rseq_register(struct rseq __user * rseq, u32 rseq_len, int flags, u32 sig) { u32 rseqfl = 0; u8 version = 1; - if (flags & RSEQ_FLAG_UNREGISTER) { - if (flags & ~RSEQ_FLAG_UNREGISTER) - return -EINVAL; - /* Unregister rseq for current thread. */ - if (current->rseq.usrptr != rseq || !current->rseq.usrptr) - return -EINVAL; - if (rseq_len != current->rseq.len) - return -EINVAL; - if (current->rseq.sig != sig) - return -EPERM; - if (!rseq_reset_ids()) - return -EFAULT; - rseq_reset(current); - return 0; - } - - if (unlikely(flags & ~(RSEQ_FLAG_SLICE_EXT_DEFAULT_ON))) - return -EINVAL; - - if (current->rseq.usrptr) { - /* - * If rseq is already registered, check whether - * the provided address differs from the prior - * one. - */ - if (current->rseq.usrptr != rseq || rseq_len != current->rseq.len) - return -EINVAL; - if (current->rseq.sig != sig) - return -EPERM; - /* Already registered. */ - return -EBUSY; - } - - /* - * If there was no rseq previously registered, ensure the provided rseq - * is properly aligned, as communcated to user-space through the ELF - * auxiliary vector AT_RSEQ_ALIGN. If rseq_len is the original rseq - * size, the required alignment is the original struct rseq alignment. - * - * The rseq_len is required to be greater or equal to the original rseq - * size. In order to be valid, rseq_len is either the original rseq size, - * or large enough to contain all supported fields, as communicated to - * user-space through the ELF auxiliary vector AT_RSEQ_FEATURE_SIZE. - */ - if (rseq_len < ORIG_RSEQ_SIZE || - (rseq_len == ORIG_RSEQ_SIZE && !IS_ALIGNED((unsigned long)rseq, ORIG_RSEQ_SIZE)) || - (rseq_len != ORIG_RSEQ_SIZE && (!IS_ALIGNED((unsigned long)rseq, rseq_alloc_align()) || - rseq_len < offsetof(struct rseq, end)))) - return -EINVAL; if (!access_ok(rseq, rseq_len)) return -EFAULT; /* - * The version check effectivly disables time slice extensions until the - * RSEQ ABI V2 registration are implemented. + * Architectures, which use the generic IRQ entry code (at least) enable + * registrations with a size greater than the original v1 fixed sized + * @rseq_len, which has been validated already to utilize the optimized + * v2 ABI mode which also enables extended RSEQ features beyond MMCID. */ + if (IS_ENABLED(CONFIG_GENERIC_IRQ_ENTRY) && rseq_len > ORIG_RSEQ_SIZE) + version = 2; + if (IS_ENABLED(CONFIG_RSEQ_SLICE_EXTENSION) && version > 1) { if (rseq_slice_extension_enabled()) { rseqfl |= RSEQ_CS_FLAG_SLICE_EXT_AVAILABLE; @@ -523,11 +476,10 @@ SYSCALL_DEFINE4(rseq, struct rseq __user *, rseq, u32, rseq_len, int, flags, u32 #endif /* - * If rseq was previously inactive, and has just been - * registered, ensure the cpu_id_start and cpu_id fields - * are updated before returning to user-space. + * Ensure the cpu_id_start and cpu_id fields are updated before + * returning to user-space. */ - current->rseq.event.has_rseq = true; + current->rseq.event.has_rseq = version; rseq_force_update(); return 0; @@ -535,6 +487,80 @@ SYSCALL_DEFINE4(rseq, struct rseq __user *, rseq, u32, rseq_len, int, flags, u32 return -EFAULT; } +static long rseq_unregister(struct rseq __user * rseq, u32 rseq_len, int flags, u32 sig) +{ + if (flags & ~RSEQ_FLAG_UNREGISTER) + return -EINVAL; + if (current->rseq.usrptr != rseq || !current->rseq.usrptr) + return -EINVAL; + if (rseq_len != current->rseq.len) + return -EINVAL; + if (current->rseq.sig != sig) + return -EPERM; + if (!rseq_reset_ids()) + return -EFAULT; + rseq_reset(current); + return 0; +} + +static long rseq_reregister(struct rseq __user * rseq, u32 rseq_len, u32 sig) +{ + /* + * If rseq is already registered, check whether the provided address + * differs from the prior one. + */ + if (current->rseq.usrptr != rseq || rseq_len != current->rseq.len) + return -EINVAL; + if (current->rseq.sig != sig) + return -EPERM; + /* Already registered. */ + return -EBUSY; +} + +static bool rseq_length_valid(struct rseq __user *rseq, unsigned int rseq_len) +{ + /* + * Ensure the provided rseq is properly aligned, as communicated to + * user-space through the ELF auxiliary vector AT_RSEQ_ALIGN. If + * rseq_len is the original rseq size, the required alignment is the + * original struct rseq alignment. + * + * In order to be valid, rseq_len is either the original rseq size, or + * large enough to contain all supported fields, as communicated to + * user-space through the ELF auxiliary vector AT_RSEQ_FEATURE_SIZE. + */ + if (rseq_len < ORIG_RSEQ_SIZE) + return false; + + if (rseq_len == ORIG_RSEQ_SIZE) + return IS_ALIGNED((unsigned long)rseq, ORIG_RSEQ_SIZE); + + return IS_ALIGNED((unsigned long)rseq, rseq_alloc_align()) && + rseq_len >= offsetof(struct rseq, end); +} + +#define RSEQ_FLAGS_SUPPORTED (RSEQ_FLAG_SLICE_EXT_DEFAULT_ON) + +/* + * sys_rseq - Register or unregister restartable sequences for the caller thread. + */ +SYSCALL_DEFINE4(rseq, struct rseq __user *, rseq, u32, rseq_len, int, flags, u32, sig) +{ + if (flags & RSEQ_FLAG_UNREGISTER) + return rseq_unregister(rseq, rseq_len, flags, sig); + + if (unlikely(flags & ~RSEQ_FLAGS_SUPPORTED)) + return -EINVAL; + + if (current->rseq.usrptr) + return rseq_reregister(rseq, rseq_len, sig); + + if (!rseq_length_valid(rseq, rseq_len)) + return -EINVAL; + + return rseq_register(rseq, rseq_len, flags, sig); +} + #ifdef CONFIG_RSEQ_SLICE_EXTENSION struct slice_timer { struct hrtimer timer; From e744060076871eebc2647b24420b550ff44b2b65 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 25 Apr 2026 14:48:23 +0200 Subject: [PATCH 422/525] selftests/rseq: Expand for optimized RSEQ ABI v2 Update the selftests so they are executed for legacy (32 bytes RSEQ region) and optimized RSEQ ABI v2 mode. Fixes: d6200245c75e ("rseq: Allow registering RSEQ with slice extension") Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Link: https://patch.msgid.link/20260428224428.009121296%40kernel.org Cc: stable@vger.kernel.org --- tools/testing/selftests/rseq/Makefile | 11 ++++-- .../testing/selftests/rseq/check_optimized.c | 17 ++++++++ tools/testing/selftests/rseq/param_test.c | 25 +++++++----- .../testing/selftests/rseq/run_param_test.sh | 39 +++++++++++++++++++ .../selftests/rseq/run_timeslice_test.sh | 14 +++++++ tools/testing/selftests/rseq/slice_test.c | 2 +- 6 files changed, 95 insertions(+), 13 deletions(-) create mode 100644 tools/testing/selftests/rseq/check_optimized.c create mode 100755 tools/testing/selftests/rseq/run_timeslice_test.sh diff --git a/tools/testing/selftests/rseq/Makefile b/tools/testing/selftests/rseq/Makefile index 0293a2f17f51..50d69e22ee7a 100644 --- a/tools/testing/selftests/rseq/Makefile +++ b/tools/testing/selftests/rseq/Makefile @@ -15,7 +15,7 @@ LDLIBS += -lpthread -ldl OVERRIDE_TARGETS = 1 TEST_GEN_PROGS = basic_test basic_percpu_ops_test basic_percpu_ops_mm_cid_test \ - param_test_benchmark param_test_mm_cid_benchmark slice_test + param_test_benchmark param_test_mm_cid_benchmark TEST_GEN_PROGS_EXTENDED = librseq.so \ param_test \ @@ -23,9 +23,11 @@ TEST_GEN_PROGS_EXTENDED = librseq.so \ param_test_mm_cid \ param_test_mm_cid_compare_twice \ syscall_errors_test \ - legacy_check + legacy_check \ + slice_test \ + check_optimized -TEST_PROGS = run_param_test.sh run_syscall_errors_test.sh run_legacy_check.sh +TEST_PROGS = run_param_test.sh run_syscall_errors_test.sh run_legacy_check.sh run_timeslice_test.sh TEST_FILES := settings @@ -66,3 +68,6 @@ $(OUTPUT)/syscall_errors_test: syscall_errors_test.c $(TEST_GEN_PROGS_EXTENDED) $(OUTPUT)/slice_test: slice_test.c $(TEST_GEN_PROGS_EXTENDED) rseq.h rseq-*.h $(CC) $(CFLAGS) $< $(LDLIBS) -lrseq -o $@ + +$(OUTPUT)/check_optimized: check_optimized.c $(TEST_GEN_PROGS_EXTENDED) rseq.h rseq-*.h + $(CC) $(CFLAGS) $< $(LDLIBS) -lrseq -o $@ diff --git a/tools/testing/selftests/rseq/check_optimized.c b/tools/testing/selftests/rseq/check_optimized.c new file mode 100644 index 000000000000..a13e3f2c8fc6 --- /dev/null +++ b/tools/testing/selftests/rseq/check_optimized.c @@ -0,0 +1,17 @@ +// SPDX-License-Identifier: LGPL-2.1 +#define _GNU_SOURCE +#include +#include +#include +#include +#include +#include + +#include "rseq.h" + +int main(int argc, char **argv) +{ + if (__rseq_register_current_thread(true, false)) + return -1; + return 0; +} diff --git a/tools/testing/selftests/rseq/param_test.c b/tools/testing/selftests/rseq/param_test.c index 05d03e679e06..e1e98dbabe4b 100644 --- a/tools/testing/selftests/rseq/param_test.c +++ b/tools/testing/selftests/rseq/param_test.c @@ -38,7 +38,7 @@ static int opt_modulo, verbose; static int opt_yield, opt_signal, opt_sleep, opt_disable_rseq, opt_threads = 200, opt_disable_mod = 0, opt_test = 's'; - +static bool opt_rseq_legacy; static long long opt_reps = 5000; static __thread __attribute__((tls_model("initial-exec"))) @@ -281,9 +281,12 @@ unsigned int yield_mod_cnt, nr_abort; } \ } +#define rseq_no_glibc true + #else #define printf_verbose(fmt, ...) +#define rseq_no_glibc false #endif /* BENCHMARK */ @@ -481,7 +484,7 @@ void *test_percpu_spinlock_thread(void *arg) long long i, reps; if (!opt_disable_rseq && thread_data->reg && - rseq_register_current_thread()) + __rseq_register_current_thread(rseq_no_glibc, opt_rseq_legacy)) abort(); reps = thread_data->reps; for (i = 0; i < reps; i++) { @@ -558,7 +561,7 @@ void *test_percpu_inc_thread(void *arg) long long i, reps; if (!opt_disable_rseq && thread_data->reg && - rseq_register_current_thread()) + __rseq_register_current_thread(rseq_no_glibc, opt_rseq_legacy)) abort(); reps = thread_data->reps; for (i = 0; i < reps; i++) { @@ -712,7 +715,7 @@ void *test_percpu_list_thread(void *arg) long long i, reps; struct percpu_list *list = (struct percpu_list *)arg; - if (!opt_disable_rseq && rseq_register_current_thread()) + if (!opt_disable_rseq && __rseq_register_current_thread(rseq_no_glibc, opt_rseq_legacy)) abort(); reps = opt_reps; @@ -895,7 +898,7 @@ void *test_percpu_buffer_thread(void *arg) long long i, reps; struct percpu_buffer *buffer = (struct percpu_buffer *)arg; - if (!opt_disable_rseq && rseq_register_current_thread()) + if (!opt_disable_rseq && __rseq_register_current_thread(rseq_no_glibc, opt_rseq_legacy)) abort(); reps = opt_reps; @@ -1105,7 +1108,7 @@ void *test_percpu_memcpy_buffer_thread(void *arg) long long i, reps; struct percpu_memcpy_buffer *buffer = (struct percpu_memcpy_buffer *)arg; - if (!opt_disable_rseq && rseq_register_current_thread()) + if (!opt_disable_rseq && __rseq_register_current_thread(rseq_no_glibc, opt_rseq_legacy)) abort(); reps = opt_reps; @@ -1258,7 +1261,7 @@ void *test_membarrier_worker_thread(void *arg) const int iters = opt_reps; int i; - if (rseq_register_current_thread()) { + if (__rseq_register_current_thread(rseq_no_glibc, opt_rseq_legacy)) { fprintf(stderr, "Error: rseq_register_current_thread(...) failed(%d): %s\n", errno, strerror(errno)); abort(); @@ -1323,7 +1326,7 @@ void *test_membarrier_manager_thread(void *arg) intptr_t expect_a = 0, expect_b = 0; int cpu_a = 0, cpu_b = 0; - if (rseq_register_current_thread()) { + if (__rseq_register_current_thread(rseq_no_glibc, opt_rseq_legacy)) { fprintf(stderr, "Error: rseq_register_current_thread(...) failed(%d): %s\n", errno, strerror(errno)); abort(); @@ -1475,6 +1478,7 @@ static void show_usage(int argc, char **argv) printf(" [-D M] Disable rseq for each M threads\n"); printf(" [-T test] Choose test: (s)pinlock, (l)ist, (b)uffer, (m)emcpy, (i)ncrement, membarrie(r)\n"); printf(" [-M] Push into buffer and memcpy buffer with memory barriers.\n"); + printf(" [-O] Test with optimized RSEQ\n"); printf(" [-v] Verbose output.\n"); printf(" [-h] Show this help.\n"); printf("\n"); @@ -1602,6 +1606,9 @@ int main(int argc, char **argv) case 'M': opt_mo = RSEQ_MO_RELEASE; break; + case 'L': + opt_rseq_legacy = true; + break; default: show_usage(argc, argv); goto error; @@ -1618,7 +1625,7 @@ int main(int argc, char **argv) if (set_signal_handler()) goto error; - if (!opt_disable_rseq && rseq_register_current_thread()) + if (!opt_disable_rseq && __rseq_register_current_thread(rseq_no_glibc, opt_rseq_legacy)) goto error; if (!opt_disable_rseq && !rseq_validate_cpu_id()) { fprintf(stderr, "Error: cpu id getter unavailable\n"); diff --git a/tools/testing/selftests/rseq/run_param_test.sh b/tools/testing/selftests/rseq/run_param_test.sh index 8d31426ab41f..69a3fa049929 100755 --- a/tools/testing/selftests/rseq/run_param_test.sh +++ b/tools/testing/selftests/rseq/run_param_test.sh @@ -34,6 +34,11 @@ REPS=1000 SLOW_REPS=100 NR_THREADS=$((6*${NR_CPUS})) +# Prevent GLIBC from registering RSEQ so the selftest can run in legacy and +# performance optimized mode. +GLIBC_TUNABLES="${GLIBC_TUNABLES:-}:glibc.pthread.rseq=0" +export GLIBC_TUNABLES + function do_tests() { local i=0 @@ -103,6 +108,40 @@ function inject_blocking() NR_LOOPS= } +echo "Testing in legacy RSEQ mode" +echo "Yield injection (25%)" +inject_blocking -m 4 -y -L + +echo "Yield injection (50%)" +inject_blocking -m 2 -y -L + +echo "Yield injection (100%)" +inject_blocking -m 1 -y -L + +echo "Kill injection (25%)" +inject_blocking -m 4 -k -L + +echo "Kill injection (50%)" +inject_blocking -m 2 -k -L + +echo "Kill injection (100%)" +inject_blocking -m 1 -k -L + +echo "Sleep injection (1ms, 25%)" +inject_blocking -m 4 -s 1 -L + +echo "Sleep injection (1ms, 50%)" +inject_blocking -m 2 -s 1 -L + +echo "Sleep injection (1ms, 100%)" +inject_blocking -m 1 -s 1 -L + +./check_optimized || { + echo "Skipping optimized RSEQ mode test. Not supported"; + exit 0 +} + +echo "Testing in optimized RSEQ mode" echo "Yield injection (25%)" inject_blocking -m 4 -y diff --git a/tools/testing/selftests/rseq/run_timeslice_test.sh b/tools/testing/selftests/rseq/run_timeslice_test.sh new file mode 100755 index 000000000000..551ebed71ec6 --- /dev/null +++ b/tools/testing/selftests/rseq/run_timeslice_test.sh @@ -0,0 +1,14 @@ +#!/bin/bash +# SPDX-License-Identifier: GPL-2.0+ + +# Prevent GLIBC from registering RSEQ so the selftest can run in legacy +# and performance optimized mode. +GLIBC_TUNABLES="${GLIBC_TUNABLES:-}:glibc.pthread.rseq=0" +export GLIBC_TUNABLES + +./check_optimized || { + echo "Skipping optimized RSEQ mode test. Not supported"; + exit 0 +} + +./slice_test diff --git a/tools/testing/selftests/rseq/slice_test.c b/tools/testing/selftests/rseq/slice_test.c index 77e668ff74d7..e402d4440bc2 100644 --- a/tools/testing/selftests/rseq/slice_test.c +++ b/tools/testing/selftests/rseq/slice_test.c @@ -124,7 +124,7 @@ FIXTURE_SETUP(slice_ext) { cpu_set_t affinity; - if (rseq_register_current_thread()) + if (__rseq_register_current_thread(true, false)) SKIP(return, "RSEQ not supported\n"); if (prctl(PR_RSEQ_SLICE_EXTENSION, PR_RSEQ_SLICE_EXTENSION_SET, From b6eee96843e8d088200f01b035da98e72067c5fe Mon Sep 17 00:00:00 2001 From: Zhan Xusheng Date: Fri, 1 May 2026 12:40:06 +0200 Subject: [PATCH 423/525] sched/fair: Fix overflow in vruntime_eligible() Zhan Xusheng reported running into sporadic a s64 mult overflow in vruntime_eligible(). When constructing a worst case scenario: If you have cgroups, then you can have an entity of weight 2 (per calc_group_shares()), and its vlag should then be bounded by: (slice+TICK_NSEC) * NICE_0_LOAD, which is around 44 bits as per the comment on entity_key(). The other extreme is 100*NICE_0_LOAD, thus you get: {key, weight}[] := { puny: { (slice + TICK_NSEC) * NICE_0_LOAD, 2 }, max: { 0, 100*NICE_0_LOAD }, } The avg_vruntime() would end up being very close to 0 (which is zero_vruntime), so no real help making that more accurate. vruntime_eligible(puny) ends up with: avg = 2 * puny.key (+ 0) load = 2 + 100 * NICE_0_LOAD avg >= puny.key * load And that is: (slice + TICK_NSEC) * NICE_0_LOAD * NICE_0_LOAD * 100, which will overflow s64. Zhan suggested using __builtin_mul_overflow(), however after staring at compiler output for various architectures using godbolt, it seems that using an __int128 multiplication often results in better code. Specifically, a number of architectures already compute the __int128 product to determine the overflow. Eg. arm64 already has the 'smulh' instruction used. By explicitly doing an __int128 multiply, it will emit the 'mul; smulh' pattern, which modern cores can fuse (armv8-a clang-22.1.0). x86_64 has less branches (no OF handling). Since Linux has ARCH_SUPPORTS_INT128 to gate __int128 usage, also provide the __builtin_mul_overflow() variant as a fallback. [peterz: Changelog and __int128 bits] Fixes: 556146ce5e94 ("sched/fair: Avoid overflow in enqueue_entity()") Reported-by: Zhan Xusheng Closes: https://patch.msgid.link/20260415145742.10359-1-zhanxusheng%40xiaomi.com Signed-off-by: Zhan Xusheng Signed-off-by: Peter Zijlstra (Intel) Link: https://patch.msgid.link/20260505103155.GN3102924%40noisy.programming.kicks-ass.net --- kernel/sched/fair.c | 39 ++++++++++++++++++++++++++++++++++----- 1 file changed, 34 insertions(+), 5 deletions(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 728965851842..b91c8b294229 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -882,11 +882,11 @@ bool update_entity_lag(struct cfs_rq *cfs_rq, struct sched_entity *se) * * lag_i >= 0 -> V >= v_i * - * \Sum (v_i - v)*w_i - * V = ------------------ + v + * \Sum (v_i - v0)*w_i + * V = ------------------- + v0 * \Sum w_i * - * lag_i >= 0 -> \Sum (v_i - v)*w_i >= (v_i - v)*(\Sum w_i) + * lag_i >= 0 -> \Sum (v_i - v0)*w_i >= (v_i - v0)*(\Sum w_i) * * Note: using 'avg_vruntime() > se->vruntime' is inaccurate due * to the loss in precision caused by the division. @@ -894,7 +894,7 @@ bool update_entity_lag(struct cfs_rq *cfs_rq, struct sched_entity *se) static int vruntime_eligible(struct cfs_rq *cfs_rq, u64 vruntime) { struct sched_entity *curr = cfs_rq->curr; - s64 avg = cfs_rq->sum_w_vruntime; + s64 key, avg = cfs_rq->sum_w_vruntime; long load = cfs_rq->sum_weight; if (curr && curr->on_rq) { @@ -904,7 +904,36 @@ static int vruntime_eligible(struct cfs_rq *cfs_rq, u64 vruntime) load += weight; } - return avg >= vruntime_op(vruntime, "-", cfs_rq->zero_vruntime) * load; + key = vruntime_op(vruntime, "-", cfs_rq->zero_vruntime); + + /* + * The worst case term for @key includes 'NSEC_TICK * NICE_0_LOAD' + * and @load obviously includes NICE_0_LOAD. NSEC_TICK is around 24 + * bits, while NICE_0_LOAD is 20 on 64bit and 10 otherwise. + * + * This gives that on 64bit the product will be at least 64bit which + * overflows s64, while on 32bit it will only be 44bits and should fit + * comfortably. + */ +#ifdef CONFIG_64BIT +#ifdef CONFIG_ARCH_SUPPORTS_INT128 + /* This often results in simpler code than __builtin_mul_overflow(). */ + return avg >= (__int128)key * load; +#else + s64 rhs; + /* + * On overflow, the sign of key tells us the correct answer: a large + * positive key means vruntime >> V, so not eligible; a large negative + * key means vruntime << V, so eligible. + */ + if (check_mul_overflow(key, load, &rhs)) + return key <= 0; + + return avg >= rhs; +#endif +#else /* 32bit */ + return avg >= key * load; +#endif } int entity_eligible(struct cfs_rq *cfs_rq, struct sched_entity *se) From 9f6d929ee2c6f0266edb564bcd2bd47fd6e884a8 Mon Sep 17 00:00:00 2001 From: Vincent Guittot Date: Sun, 3 May 2026 12:45:03 +0200 Subject: [PATCH 424/525] sched/fair: Fix wakeup_preempt_fair() for not waking up task Make sure to only call pick_next_entity() on an non-empty cfs_rq. The assumption that p is always enqueued and not delayed, is only true for wakeup. If p was moved while delayed, pick_next_entity() will dequeue it and the cfs might become empty. Test if there are still queued tasks before trying again to determine if p could be the next one to be picked. There are at least 2 cases: When cfs becomes idle, it tries to pull tasks but if those pulled tasks are delayed, they will be dequeued when attached to cfs. attach_tasks() -> attach_task() -> wakeup_preempt(rq, p, 0); A misfit task running on cfs A triggers a load balance to be pulled on a better cpu, the load balance on cfs B starts an active load balance to pulled the running misfit task. If there is a delayed dequeue task on cfs A, it can be pulled instead of the previously running misfit task. attach_one_task() -> attach_task() -> wakeup_preempt(rq, p, 0); Fixes: ac8e69e69363 ("sched/fair: Fix wakeup_preempt_fair() vs delayed dequeue") Signed-off-by: Vincent Guittot Signed-off-by: Peter Zijlstra (Intel) Link: https://patch.msgid.link/20260503104503.1732682-1-vincent.guittot@linaro.org --- kernel/sched/fair.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index b91c8b294229..3ebec186f982 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -9174,9 +9174,10 @@ static void wakeup_preempt_fair(struct rq *rq, struct task_struct *p, int wake_f /* * Because p is enqueued, nse being null can only mean that we - * dequeued a delayed task. + * dequeued a delayed task. If there are still entities queued in + * cfs, check if the next one will be p. */ - if (!nse) + if (!nse && cfs_rq->nr_queued) goto pick; if (sched_feat(RUN_TO_PARITY)) From b819db93d73f4593636299e229914052b89e3ef2 Mon Sep 17 00:00:00 2001 From: Pauli Virtanen Date: Sun, 12 Apr 2026 21:47:42 +0300 Subject: [PATCH 425/525] Bluetooth: SCO: fix sleeping under spinlock in sco_conn_ready sco_conn_ready calls sleeping functions under conn->lock spinlock. The critical section can be reduced: conn->hcon is modified only with hdev->lock held. It is guaranteed to be held in sco_conn_ready, so conn->lock is not needed to guard it. Move taking conn->lock after lock_sock(parent). This also follows the lock ordering lock_sock() > conn->lock elsewhere in the file. Fixes: 27c24fda62b60 ("Bluetooth: switch to lock_sock in SCO") Signed-off-by: Pauli Virtanen Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/sco.c | 20 +++++++++----------- 1 file changed, 9 insertions(+), 11 deletions(-) diff --git a/net/bluetooth/sco.c b/net/bluetooth/sco.c index 18826d4b9c0b..3a5479538e85 100644 --- a/net/bluetooth/sco.c +++ b/net/bluetooth/sco.c @@ -1377,26 +1377,24 @@ static void sco_conn_ready(struct sco_conn *conn) sk->sk_state_change(sk); release_sock(sk); } else { - sco_conn_lock(conn); - - if (!conn->hcon) { - sco_conn_unlock(conn); + if (!conn->hcon) return; - } + + lockdep_assert_held(&conn->hcon->hdev->lock); parent = sco_get_sock_listen(&conn->hcon->src); - if (!parent) { - sco_conn_unlock(conn); + if (!parent) return; - } lock_sock(parent); + sco_conn_lock(conn); + sk = sco_sock_alloc(sock_net(parent), NULL, BTPROTO_SCO, GFP_ATOMIC, 0); if (!sk) { - release_sock(parent); sco_conn_unlock(conn); + release_sock(parent); return; } @@ -1417,9 +1415,9 @@ static void sco_conn_ready(struct sco_conn *conn) /* Wake up parent */ parent->sk_data_ready(parent); - release_sock(parent); - sco_conn_unlock(conn); + + release_sock(parent); } } From 0beddb0c380bed5f5b8e61ddbe14635bb73d0b41 Mon Sep 17 00:00:00 2001 From: David Carlier Date: Sun, 12 Apr 2026 21:29:16 +0100 Subject: [PATCH 426/525] Bluetooth: hci_conn: fix potential UAF in create_big_sync Add hci_conn_valid() check in create_big_sync() to detect stale connections before proceeding with BIG creation. Handle the resulting -ECANCELED in create_big_complete() and re-validate the connection under hci_dev_lock() before dereferencing, matching the pattern used by create_le_conn_complete() and create_pa_complete(). Keep the hci_conn object alive across the async boundary by taking a reference via hci_conn_get() when queueing create_big_sync(), and dropping it in the completion callback. The refcount and the lock are complementary: the refcount keeps the object allocated, while hci_dev_lock() serializes hci_conn_hash_del()'s list_del_rcu() on hdev->conn_hash, as required by hci_conn_del(). hci_conn_put() is called outside hci_dev_unlock() so the final put (which resolves to kfree() via bt_link_release) does not run under hdev->lock, though the release path would be safe either way. Without this, create_big_complete() would unconditionally dereference the conn pointer on error, causing a use-after-free via hci_connect_cfm() and hci_conn_del(). Fixes: eca0ae4aea66 ("Bluetooth: Add initial implementation of BIS connections") Cc: stable@vger.kernel.org Co-developed-by: Luiz Augusto von Dentz Signed-off-by: Luiz Augusto von Dentz Signed-off-by: David Carlier Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_conn.c | 19 ++++++++++++++++++- 1 file changed, 18 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index 3a0592599086..96e345fcf303 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -2130,6 +2130,9 @@ static int create_big_sync(struct hci_dev *hdev, void *data) u32 flags = 0; int err; + if (!hci_conn_valid(hdev, conn)) + return -ECANCELED; + if (qos->bcast.out.phys == BIT(1)) flags |= MGMT_ADV_FLAG_SEC_2M; @@ -2204,11 +2207,24 @@ static void create_big_complete(struct hci_dev *hdev, void *data, int err) bt_dev_dbg(hdev, "conn %p", conn); + if (err == -ECANCELED) + goto done; + + hci_dev_lock(hdev); + + if (!hci_conn_valid(hdev, conn)) + goto unlock; + if (err) { bt_dev_err(hdev, "Unable to create BIG: %d", err); hci_connect_cfm(conn, err); hci_conn_del(conn); } + +unlock: + hci_dev_unlock(hdev); +done: + hci_conn_put(conn); } struct hci_conn *hci_bind_bis(struct hci_dev *hdev, bdaddr_t *dst, __u8 sid, @@ -2336,10 +2352,11 @@ struct hci_conn *hci_connect_bis(struct hci_dev *hdev, bdaddr_t *dst, BT_BOUND, &data); /* Queue start periodic advertising and create BIG */ - err = hci_cmd_sync_queue(hdev, create_big_sync, conn, + err = hci_cmd_sync_queue(hdev, create_big_sync, hci_conn_get(conn), create_big_complete); if (err < 0) { hci_conn_drop(conn); + hci_conn_put(conn); return ERR_PTR(err); } From 5ddb8014261137cadaf83ab5617a588d80a22586 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Fri, 10 Apr 2026 15:29:52 -0400 Subject: [PATCH 427/525] Bluetooth: hci_event: Fix OOB read and infinite loop in hci_le_create_big_complete_evt hci_le_create_big_complete_evt() iterates over BT_BOUND connections for a BIG handle using a while loop, accessing ev->bis_handle[i++] on each iteration. However, there is no check that i stays within ev->num_bis before the array access. When a controller sends a LE_Create_BIG_Complete event with fewer bis_handle entries than there are BT_BOUND connections for that BIG, or with num_bis=0, the loop reads beyond the valid bis_handle[] flex array into adjacent heap memory. Since the out-of-bounds values typically exceed HCI_CONN_HANDLE_MAX (0x0EFF), hci_conn_set_handle() rejects them and the connection remains in BT_BOUND state. The same connection is then found again by hci_conn_hash_lookup_big_state(), creating an infinite loop with hci_dev_lock held. Fix this by terminating the BIG if in case not all BIS could be setup properly. Fixes: a0bfde167b50 ("Bluetooth: ISO: Add support for connecting multiple BISes") Cc: stable@vger.kernel.org Signed-off-by: ZhiTao Ou Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_event.c | 29 ++++++++++++++++++++++++++--- 1 file changed, 26 insertions(+), 3 deletions(-) diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index b2ee6b6a0f56..1b3b9131affa 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -7118,9 +7118,29 @@ static void hci_le_create_big_complete_evt(struct hci_dev *hdev, void *data, continue; } - if (hci_conn_set_handle(conn, - __le16_to_cpu(ev->bis_handle[i++]))) + if (ev->num_bis <= i) { + bt_dev_err(hdev, + "Not enough BIS handles for BIG 0x%2.2x", + ev->handle); + ev->status = HCI_ERROR_UNSPECIFIED; + hci_connect_cfm(conn, ev->status); + hci_conn_del(conn); continue; + } + + if (hci_conn_set_handle(conn, + __le16_to_cpu(ev->bis_handle[i++]))) { + bt_dev_err(hdev, + "Failed to set BIS handle for BIG 0x%2.2x", + ev->handle); + /* Force error so BIG gets terminated as not all BIS + * could be connected. + */ + ev->status = HCI_ERROR_UNSPECIFIED; + hci_connect_cfm(conn, ev->status); + hci_conn_del(conn); + continue; + } conn->state = BT_CONNECTED; set_bit(HCI_CONN_BIG_CREATED, &conn->flags); @@ -7129,7 +7149,10 @@ static void hci_le_create_big_complete_evt(struct hci_dev *hdev, void *data, hci_iso_setup_path(conn); } - if (!ev->status && !i) + /* If there is an unexpected error or if no BISes have been connected + * for the BIG, terminate it. + */ + if (ev->status == HCI_ERROR_UNSPECIFIED || (!ev->status && !i)) /* If no BISes have been connected for the BIG, * terminate. This is in case all bound connections * have been closed before the BIG creation From 72b8deccff17a7644e0367e1aaf1a36cfb014324 Mon Sep 17 00:00:00 2001 From: Dudu Lu Date: Wed, 15 Apr 2026 17:39:53 +0800 Subject: [PATCH 428/525] Bluetooth: bnep: fix incorrect length parsing in bnep_rx_frame() extension handling In bnep_rx_frame(), the BNEP_FILTER_NET_TYPE_SET and BNEP_FILTER_MULTI_ADDR_SET extension header parsing has two bugs: 1) The 2-byte length field is read with *(u16 *)(skb->data + 1), which performs a native-endian read. The BNEP protocol specifies this field in big-endian (network byte order), and the same file correctly uses get_unaligned_be16() for the identical fields in bnep_ctrl_set_netfilter() and bnep_ctrl_set_mcfilter(). 2) The length is multiplied by 2, but unlike BNEP_SETUP_CONN_REQ where the length byte counts UUID pairs (requiring * 2 for two UUIDs per entry), the filter extension length field already represents the total data size in bytes. This is confirmed by bnep_ctrl_set_netfilter() which reads the same field as a byte count and divides by 4 to get the number of filter entries. The bogus * 2 means skb_pull advances twice as far as it should, either dropping valid data from the next header or causing the pull to fail entirely when the doubled length exceeds the remaining skb. Fix by splitting the pull into two steps: first use skb_pull_data() to safely pull and validate the 3-byte fixed header (ctrl type + length), then pull the variable-length data using the properly decoded length. Fixes: bf8b9a9cb77b ("Bluetooth: bnep: Add support to extended headers of control frames") Signed-off-by: Dudu Lu Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/bnep/core.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/net/bluetooth/bnep/core.c b/net/bluetooth/bnep/core.c index d44987d4515c..853c8d7644b5 100644 --- a/net/bluetooth/bnep/core.c +++ b/net/bluetooth/bnep/core.c @@ -330,11 +330,18 @@ static int bnep_rx_frame(struct bnep_session *s, struct sk_buff *skb) goto badframe; break; case BNEP_FILTER_MULTI_ADDR_SET: - case BNEP_FILTER_NET_TYPE_SET: - /* Pull: ctrl type (1 b), len (2 b), data (len bytes) */ - if (!skb_pull(skb, 3 + *(u16 *)(skb->data + 1) * 2)) + case BNEP_FILTER_NET_TYPE_SET: { + u8 *hdr; + + /* Pull ctrl type (1 b) + len (2 b) */ + hdr = skb_pull_data(skb, 3); + if (!hdr) + goto badframe; + /* Pull data (len bytes); length is big-endian */ + if (!skb_pull(skb, get_unaligned_be16(&hdr[1]))) goto badframe; break; + } default: kfree_skb(skb); return 0; From 4f42363c814f28fe3f59847c35acf1ed033bedd4 Mon Sep 17 00:00:00 2001 From: Dudu Lu Date: Wed, 15 Apr 2026 18:43:55 +0800 Subject: [PATCH 429/525] Bluetooth: l2cap: fix MPS check in l2cap_ecred_reconf_req The L2CAP specification states that if more than one channel is being reconfigured, the MPS shall not be decreased. The current check has two issues: 1) The comparison uses >= (greater-than-or-equal), which incorrectly rejects reconfiguration requests where the MPS stays the same. Since the spec says MPS "shall be greater than or equal to the current MPS", only a strict decrease (remote_mps > mps) should be rejected. Keeping the same MPS is valid. 2) The multi-channel guard uses `&& i` (loop index) to approximate "more than one channel", but this incorrectly allows MPS decrease for the first channel (i==0) even when multiple channels are being reconfigured. Replace with `&& num_scid > 1` which correctly checks whether the request covers more than one channel. Fixes: 7accb1c4321a ("Bluetooth: L2CAP: Fix invalid response to L2CAP_ECRED_RECONF_REQ") Signed-off-by: Dudu Lu Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/l2cap_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c index 77dec104a9c3..b15374b951fa 100644 --- a/net/bluetooth/l2cap_core.c +++ b/net/bluetooth/l2cap_core.c @@ -5428,7 +5428,7 @@ static inline int l2cap_ecred_reconf_req(struct l2cap_conn *conn, * configured, the MPS field may be less than the current MPS * of that channel. */ - if (chan[i]->remote_mps >= mps && i) { + if (chan[i]->remote_mps > mps && num_scid > 1) { BT_ERR("chan %p decreased MPS %u -> %u", chan[i], chan[i]->remote_mps, mps); result = L2CAP_RECONF_INVALID_MPS; From 91b5a598b5285da794b72619f31777b62dd336f8 Mon Sep 17 00:00:00 2001 From: Mikhail Gavrilov Date: Wed, 15 Apr 2026 02:52:37 +0500 Subject: [PATCH 430/525] Bluetooth: l2cap: defer conn param update to avoid conn->lock/hdev->lock inversion When a BLE peripheral sends an L2CAP Connection Parameter Update Request the processing path is: process_pending_rx() [takes conn->lock] l2cap_le_sig_channel() l2cap_conn_param_update_req() hci_le_conn_update() [takes hdev->lock] Meanwhile other code paths take the locks in the opposite order: l2cap_chan_connect() [takes hdev->lock] ... mutex_lock(&conn->lock) l2cap_conn_ready() [hdev->lock via hci_cb_list_lock] ... mutex_lock(&conn->lock) This is a classic AB/BA deadlock which lockdep reports as a circular locking dependency when connecting a BLE MIDI keyboard (Carry-On FC-49). Fix this by making hci_le_conn_update() defer the HCI command through hci_cmd_sync_queue() so it no longer needs to take hdev->lock in the caller context. The sync callback uses __hci_cmd_sync_status_sk() to wait for the HCI_EV_LE_CONN_UPDATE_COMPLETE event, then updates the stored connection parameters (hci_conn_params) and notifies userspace (mgmt_new_conn_param) only after the controller has confirmed the update. A reference on hci_conn is held via hci_conn_get()/hci_conn_put() for the lifetime of the queued work to prevent use-after-free, and hci_conn_valid() is checked before proceeding in case the connection was removed while the work was pending. The hci_dev_lock is held across hci_conn_valid() and all conn field accesses to prevent a concurrent disconnect from invalidating the connection mid-use. Fixes: f044eb0524a0 ("Bluetooth: Store latency and supervision timeout in connection params") Signed-off-by: Mikhail Gavrilov Reviewed-by: Paul Menzel Signed-off-by: Luiz Augusto von Dentz --- include/net/bluetooth/hci_core.h | 2 +- net/bluetooth/hci_conn.c | 105 +++++++++++++++++++++++++------ net/bluetooth/l2cap_core.c | 12 +--- 3 files changed, 89 insertions(+), 30 deletions(-) diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index a7bffb908c1e..aa600fbf9a53 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -2495,7 +2495,7 @@ void mgmt_adv_monitor_device_lost(struct hci_dev *hdev, u16 handle, bdaddr_t *bdaddr, u8 addr_type); int hci_abort_conn(struct hci_conn *conn, u8 reason); -u8 hci_le_conn_update(struct hci_conn *conn, u16 min, u16 max, u16 latency, +void hci_le_conn_update(struct hci_conn *conn, u16 min, u16 max, u16 latency, u16 to_multiplier); void hci_le_start_enc(struct hci_conn *conn, __le16 ediv, __le64 rand, __u8 ltk[16], __u8 key_size); diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index 96e345fcf303..17b46ad6a349 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -480,40 +480,107 @@ bool hci_setup_sync(struct hci_conn *conn, __u16 handle) return hci_setup_sync_conn(conn, handle); } -u8 hci_le_conn_update(struct hci_conn *conn, u16 min, u16 max, u16 latency, - u16 to_multiplier) +struct le_conn_update_data { + struct hci_conn *conn; + u16 min; + u16 max; + u16 latency; + u16 to_multiplier; +}; + +static int le_conn_update_sync(struct hci_dev *hdev, void *data) { - struct hci_dev *hdev = conn->hdev; + struct le_conn_update_data *d = data; + struct hci_conn *conn = d->conn; struct hci_conn_params *params; struct hci_cp_le_conn_update cp; + u16 timeout; + u8 store_hint; + int err; + /* Verify connection is still alive and read conn fields under + * the same lock to prevent a concurrent disconnect from freeing + * or reusing the connection while we build the HCI command. + */ + hci_dev_lock(hdev); + + if (!hci_conn_valid(hdev, conn)) { + hci_dev_unlock(hdev); + return -ECANCELED; + } + + memset(&cp, 0, sizeof(cp)); + cp.handle = cpu_to_le16(conn->handle); + cp.conn_interval_min = cpu_to_le16(d->min); + cp.conn_interval_max = cpu_to_le16(d->max); + cp.conn_latency = cpu_to_le16(d->latency); + cp.supervision_timeout = cpu_to_le16(d->to_multiplier); + cp.min_ce_len = cpu_to_le16(0x0000); + cp.max_ce_len = cpu_to_le16(0x0000); + timeout = conn->conn_timeout; + + hci_dev_unlock(hdev); + + err = __hci_cmd_sync_status_sk(hdev, HCI_OP_LE_CONN_UPDATE, + sizeof(cp), &cp, + HCI_EV_LE_CONN_UPDATE_COMPLETE, + timeout, NULL); + if (err) + return err; + + /* Update stored connection parameters after the controller has + * confirmed the update via the LE Connection Update Complete event. + */ hci_dev_lock(hdev); params = hci_conn_params_lookup(hdev, &conn->dst, conn->dst_type); if (params) { - params->conn_min_interval = min; - params->conn_max_interval = max; - params->conn_latency = latency; - params->supervision_timeout = to_multiplier; + params->conn_min_interval = d->min; + params->conn_max_interval = d->max; + params->conn_latency = d->latency; + params->supervision_timeout = d->to_multiplier; + store_hint = 0x01; + } else { + store_hint = 0x00; } hci_dev_unlock(hdev); - memset(&cp, 0, sizeof(cp)); - cp.handle = cpu_to_le16(conn->handle); - cp.conn_interval_min = cpu_to_le16(min); - cp.conn_interval_max = cpu_to_le16(max); - cp.conn_latency = cpu_to_le16(latency); - cp.supervision_timeout = cpu_to_le16(to_multiplier); - cp.min_ce_len = cpu_to_le16(0x0000); - cp.max_ce_len = cpu_to_le16(0x0000); + mgmt_new_conn_param(hdev, &conn->dst, conn->dst_type, store_hint, + d->min, d->max, d->latency, d->to_multiplier); - hci_send_cmd(hdev, HCI_OP_LE_CONN_UPDATE, sizeof(cp), &cp); + return 0; +} - if (params) - return 0x01; +static void le_conn_update_complete(struct hci_dev *hdev, void *data, int err) +{ + struct le_conn_update_data *d = data; - return 0x00; + hci_conn_put(d->conn); + kfree(d); +} + +void hci_le_conn_update(struct hci_conn *conn, u16 min, u16 max, u16 latency, + u16 to_multiplier) +{ + struct le_conn_update_data *d; + + d = kzalloc_obj(*d); + if (!d) + return; + + hci_conn_get(conn); + d->conn = conn; + d->min = min; + d->max = max; + d->latency = latency; + d->to_multiplier = to_multiplier; + + if (hci_cmd_sync_queue(conn->hdev, le_conn_update_sync, d, + le_conn_update_complete) < 0) { + hci_conn_put(conn); + kfree(d); + } } void hci_le_start_enc(struct hci_conn *conn, __le16 ediv, __le64 rand, diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c index b15374b951fa..7701528f1167 100644 --- a/net/bluetooth/l2cap_core.c +++ b/net/bluetooth/l2cap_core.c @@ -4706,16 +4706,8 @@ static inline int l2cap_conn_param_update_req(struct l2cap_conn *conn, l2cap_send_cmd(conn, cmd->ident, L2CAP_CONN_PARAM_UPDATE_RSP, sizeof(rsp), &rsp); - if (!err) { - u8 store_hint; - - store_hint = hci_le_conn_update(hcon, min, max, latency, - to_multiplier); - mgmt_new_conn_param(hcon->hdev, &hcon->dst, hcon->dst_type, - store_hint, min, max, latency, - to_multiplier); - - } + if (!err) + hci_le_conn_update(hcon, min, max, latency, to_multiplier); return 0; } From 2ff1a41a912de8517b4482e946dd951b7d80edbf Mon Sep 17 00:00:00 2001 From: Siwei Zhang Date: Wed, 15 Apr 2026 16:51:36 -0400 Subject: [PATCH 431/525] Bluetooth: L2CAP: Fix null-ptr-deref in l2cap_sock_state_change_cb() Add the same NULL guard already present in l2cap_sock_resume_cb() and l2cap_sock_ready_cb(). Fixes: 89bc500e41fc ("Bluetooth: Add state tracking to struct l2cap_chan") Cc: stable@kernel.org Signed-off-by: Siwei Zhang Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/l2cap_sock.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c index 71e8c1b45bce..fb3cb70a5a39 100644 --- a/net/bluetooth/l2cap_sock.c +++ b/net/bluetooth/l2cap_sock.c @@ -1657,6 +1657,9 @@ static void l2cap_sock_state_change_cb(struct l2cap_chan *chan, int state, { struct sock *sk = chan->data; + if (!sk) + return; + sk->sk_state = state; if (err) From 78a88d43dab8d23aeef934ed8ce34d40e6b3d613 Mon Sep 17 00:00:00 2001 From: Siwei Zhang Date: Wed, 15 Apr 2026 16:53:36 -0400 Subject: [PATCH 432/525] Bluetooth: L2CAP: Fix null-ptr-deref in l2cap_sock_get_sndtimeo_cb() Add the same NULL guard already present in l2cap_sock_resume_cb() and l2cap_sock_ready_cb(). Fixes: 8d836d71e222 ("Bluetooth: Access sk_sndtimeo indirectly in l2cap_core.c") Cc: stable@kernel.org Signed-off-by: Siwei Zhang Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/l2cap_sock.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c index fb3cb70a5a39..879c9f90269a 100644 --- a/net/bluetooth/l2cap_sock.c +++ b/net/bluetooth/l2cap_sock.c @@ -1761,6 +1761,9 @@ static long l2cap_sock_get_sndtimeo_cb(struct l2cap_chan *chan) { struct sock *sk = chan->data; + if (!sk) + return 0; + return READ_ONCE(sk->sk_sndtimeo); } From 0a120d96166301d7a95be75b52f843837dbd1219 Mon Sep 17 00:00:00 2001 From: Siwei Zhang Date: Wed, 15 Apr 2026 16:49:59 -0400 Subject: [PATCH 433/525] Bluetooth: L2CAP: Fix null-ptr-deref in l2cap_sock_new_connection_cb() Add the same NULL guard already present in l2cap_sock_resume_cb() and l2cap_sock_ready_cb(). Fixes: 80808e431e1e ("Bluetooth: Add l2cap_chan_ops abstraction") Cc: stable@kernel.org Signed-off-by: Siwei Zhang Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/l2cap_sock.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c index 879c9f90269a..cf590a67d364 100644 --- a/net/bluetooth/l2cap_sock.c +++ b/net/bluetooth/l2cap_sock.c @@ -1498,6 +1498,9 @@ static struct l2cap_chan *l2cap_sock_new_connection_cb(struct l2cap_chan *chan) { struct sock *sk, *parent = chan->data; + if (!parent) + return NULL; + lock_sock(parent); /* Check for backlog size */ From 4e37f6452d586b95c346a9abdd2fb80b67794f39 Mon Sep 17 00:00:00 2001 From: Pauli Virtanen Date: Sat, 18 Apr 2026 18:41:12 +0300 Subject: [PATCH 434/525] Bluetooth: SCO: hold sk properly in sco_conn_ready sk deref in sco_conn_ready must be done either under conn->lock, or holding a refcount, to avoid concurrent close. conn->sk and parent sk is currently accessed without either, and without checking parent->sk_state: [Task 1] [Task 2] sco_sock_release sco_conn_ready sk = conn->sk lock_sock(sk) conn->sk = NULL lock_sock(sk) release_sock(sk) sco_sock_kill(sk) UAF on sk deref and similarly for access to sco_get_sock_listen() return value. Fix possible UAF by holding sk refcount in sco_conn_ready() and making sco_get_sock_listen() increase refcount. Also recheck after lock_sock that the socket is still valid. Adjust conn->sk locking so it's protected also by lock_sock() of the associated socket if any. Fixes: 27c24fda62b60 ("Bluetooth: switch to lock_sock in SCO") Signed-off-by: Pauli Virtanen Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/sco.c | 44 ++++++++++++++++++++++++++++++++------------ 1 file changed, 32 insertions(+), 12 deletions(-) diff --git a/net/bluetooth/sco.c b/net/bluetooth/sco.c index 3a5479538e85..eba44525d41d 100644 --- a/net/bluetooth/sco.c +++ b/net/bluetooth/sco.c @@ -472,9 +472,13 @@ static struct sock *sco_get_sock_listen(bdaddr_t *src) sk1 = sk; } + sk = sk ? sk : sk1; + if (sk) + sock_hold(sk); + read_unlock(&sco_sk_list.lock); - return sk ? sk : sk1; + return sk; } static void sco_sock_destruct(struct sock *sk) @@ -515,11 +519,13 @@ static void sco_sock_kill(struct sock *sk) BT_DBG("sk %p state %d", sk, sk->sk_state); /* Sock is dead, so set conn->sk to NULL to avoid possible UAF */ + lock_sock(sk); if (sco_pi(sk)->conn) { sco_conn_lock(sco_pi(sk)->conn); sco_pi(sk)->conn->sk = NULL; sco_conn_unlock(sco_pi(sk)->conn); } + release_sock(sk); /* Kill poor orphan */ bt_sock_unlink(&sco_sk_list, sk); @@ -1365,17 +1371,28 @@ static int sco_sock_release(struct socket *sock) static void sco_conn_ready(struct sco_conn *conn) { - struct sock *parent; - struct sock *sk = conn->sk; + struct sock *parent, *sk; + + sco_conn_lock(conn); + sk = sco_sock_hold(conn); + sco_conn_unlock(conn); BT_DBG("conn %p", conn); if (sk) { lock_sock(sk); - sco_sock_clear_timer(sk); - sk->sk_state = BT_CONNECTED; - sk->sk_state_change(sk); + + /* conn->sk may have become NULL if racing with sk close, but + * due to held hdev->lock, it can't become different sk. + */ + if (conn->sk) { + sco_sock_clear_timer(sk); + sk->sk_state = BT_CONNECTED; + sk->sk_state_change(sk); + } + release_sock(sk); + sock_put(sk); } else { if (!conn->hcon) return; @@ -1390,13 +1407,15 @@ static void sco_conn_ready(struct sco_conn *conn) sco_conn_lock(conn); + /* hdev->lock guarantees conn->sk == NULL still here */ + + if (parent->sk_state != BT_LISTEN) + goto release; + sk = sco_sock_alloc(sock_net(parent), NULL, BTPROTO_SCO, GFP_ATOMIC, 0); - if (!sk) { - sco_conn_unlock(conn); - release_sock(parent); - return; - } + if (!sk) + goto release; sco_sock_init(sk, parent); @@ -1415,9 +1434,10 @@ static void sco_conn_ready(struct sco_conn *conn) /* Wake up parent */ parent->sk_data_ready(parent); +release: sco_conn_unlock(conn); - release_sock(parent); + sock_put(parent); } } From 5917dd39db2bfc8b1b4c6ea8ed99adb4badef707 Mon Sep 17 00:00:00 2001 From: Sai Teja Aluvala Date: Mon, 20 Apr 2026 23:07:35 +0530 Subject: [PATCH 435/525] Bluetooth: btintel_pcie: treat boot stage bit 12 as warning CSR boot stage register bit 12 is documented as a device warning, not a fatal error. Rename the bit definition accordingly and stop including it in btintel_pcie_in_error(). This keeps warning-only boot stage values from being classified as errors while preserving abort-handler state as the actual error condition. Fixes: 190377500fde ("Bluetooth: btintel_pcie: Dump debug registers on error") Signed-off-by: Kiran K Signed-off-by: Sai Teja Aluvala Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btintel_pcie.c | 13 ++++++++++--- drivers/bluetooth/btintel_pcie.h | 2 +- 2 files changed, 11 insertions(+), 4 deletions(-) diff --git a/drivers/bluetooth/btintel_pcie.c b/drivers/bluetooth/btintel_pcie.c index 2f59c0d6f9ec..a3643e67b33f 100644 --- a/drivers/bluetooth/btintel_pcie.c +++ b/drivers/bluetooth/btintel_pcie.c @@ -289,6 +289,9 @@ static inline void btintel_pcie_dump_debug_registers(struct hci_dev *hdev) skb_put_data(skb, buf, strlen(buf)); data->boot_stage_cache = reg; + if (reg & BTINTEL_PCIE_CSR_BOOT_STAGE_DEVICE_WARNING) + bt_dev_warn(hdev, "Controller device warning (boot_stage: 0x%8.8x)", reg); + reg = btintel_pcie_rd_reg32(data, BTINTEL_PCIE_CSR_IPC_STATUS_REG); snprintf(buf, sizeof(buf), "ipc status: 0x%8.8x", reg); skb_put_data(skb, buf, strlen(buf)); @@ -880,8 +883,11 @@ static inline bool btintel_pcie_in_lockdown(struct btintel_pcie_data *data) static inline bool btintel_pcie_in_error(struct btintel_pcie_data *data) { - return (data->boot_stage_cache & BTINTEL_PCIE_CSR_BOOT_STAGE_DEVICE_ERR) || - (data->boot_stage_cache & BTINTEL_PCIE_CSR_BOOT_STAGE_ABORT_HANDLER); + if (data->boot_stage_cache & BTINTEL_PCIE_CSR_BOOT_STAGE_DEVICE_WARNING) + bt_dev_warn(data->hdev, "Controller device warning (boot_stage: 0x%8.8x)", + data->boot_stage_cache); + + return data->boot_stage_cache & BTINTEL_PCIE_CSR_BOOT_STAGE_ABORT_HANDLER; } static void btintel_pcie_msix_gp1_handler(struct btintel_pcie_data *data) @@ -914,7 +920,8 @@ static void btintel_pcie_msix_gp0_handler(struct btintel_pcie_data *data) data->img_resp_cache = reg; if (btintel_pcie_in_error(data)) { - bt_dev_err(data->hdev, "Controller in error state"); + bt_dev_err(data->hdev, "Controller in error state (boot_stage: 0x%8.8x)", + data->boot_stage_cache); btintel_pcie_dump_debug_registers(data->hdev); return; } diff --git a/drivers/bluetooth/btintel_pcie.h b/drivers/bluetooth/btintel_pcie.h index 3c7bb708362d..f922abd1e7d8 100644 --- a/drivers/bluetooth/btintel_pcie.h +++ b/drivers/bluetooth/btintel_pcie.h @@ -48,7 +48,7 @@ #define BTINTEL_PCIE_CSR_BOOT_STAGE_OPFW (BIT(2)) #define BTINTEL_PCIE_CSR_BOOT_STAGE_ROM_LOCKDOWN (BIT(10)) #define BTINTEL_PCIE_CSR_BOOT_STAGE_IML_LOCKDOWN (BIT(11)) -#define BTINTEL_PCIE_CSR_BOOT_STAGE_DEVICE_ERR (BIT(12)) +#define BTINTEL_PCIE_CSR_BOOT_STAGE_DEVICE_WARNING (BIT(12)) #define BTINTEL_PCIE_CSR_BOOT_STAGE_ABORT_HANDLER (BIT(13)) #define BTINTEL_PCIE_CSR_BOOT_STAGE_DEVICE_HALTED (BIT(14)) #define BTINTEL_PCIE_CSR_BOOT_STAGE_MAC_ACCESS_ON (BIT(16)) From 902fe40bce7059722f7ffa1c378e577675cf1918 Mon Sep 17 00:00:00 2001 From: Aurelien DESBRIERES Date: Tue, 21 Apr 2026 15:53:31 +0200 Subject: [PATCH 436/525] Bluetooth: hci_uart: Fix NULL deref in recv callbacks when priv is uninitialized When a fault is injected during hci_uart line discipline setup, the proto open() callback may fail leaving hu->priv as NULL. A subsequent TIOCSTI ioctl can trigger the recv() callback before priv is initialized, causing a NULL pointer dereference. Fix all four affected HCI UART protocol drivers by adding a NULL check on hu->priv at the start of their recv() callbacks: h4, h5, ath and bcsp. Reported-by: syzbot+ff30eeab8e07b37d524e@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=ff30eeab8e07b37d524e Signed-off-by: Aurelien DESBRIERES Assisted-by: Claude:claude-sonnet-4-6 Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/hci_ath.c | 3 +++ drivers/bluetooth/hci_bcsp.c | 3 +++ drivers/bluetooth/hci_h4.c | 3 +++ drivers/bluetooth/hci_h5.c | 3 +++ 4 files changed, 12 insertions(+) diff --git a/drivers/bluetooth/hci_ath.c b/drivers/bluetooth/hci_ath.c index fa679ad0acdf..8201fa7f61e8 100644 --- a/drivers/bluetooth/hci_ath.c +++ b/drivers/bluetooth/hci_ath.c @@ -191,6 +191,9 @@ static int ath_recv(struct hci_uart *hu, const void *data, int count) { struct ath_struct *ath = hu->priv; + if (!ath) + return -ENODEV; + ath->rx_skb = h4_recv_buf(hu, ath->rx_skb, data, count, ath_recv_pkts, ARRAY_SIZE(ath_recv_pkts)); if (IS_ERR(ath->rx_skb)) { diff --git a/drivers/bluetooth/hci_bcsp.c b/drivers/bluetooth/hci_bcsp.c index b386f91d8b46..db56eead27ce 100644 --- a/drivers/bluetooth/hci_bcsp.c +++ b/drivers/bluetooth/hci_bcsp.c @@ -585,6 +585,9 @@ static int bcsp_recv(struct hci_uart *hu, const void *data, int count) if (!test_bit(HCI_UART_REGISTERED, &hu->flags)) return -EUNATCH; + if (!bcsp) + return -ENODEV; + BT_DBG("hu %p count %d rx_state %d rx_count %ld", hu, count, bcsp->rx_state, bcsp->rx_count); diff --git a/drivers/bluetooth/hci_h4.c b/drivers/bluetooth/hci_h4.c index a889a66a326f..767372707498 100644 --- a/drivers/bluetooth/hci_h4.c +++ b/drivers/bluetooth/hci_h4.c @@ -109,6 +109,9 @@ static int h4_recv(struct hci_uart *hu, const void *data, int count) { struct h4_struct *h4 = hu->priv; + if (!h4) + return -ENODEV; + h4->rx_skb = h4_recv_buf(hu, h4->rx_skb, data, count, h4_recv_pkts, ARRAY_SIZE(h4_recv_pkts)); if (IS_ERR(h4->rx_skb)) { diff --git a/drivers/bluetooth/hci_h5.c b/drivers/bluetooth/hci_h5.c index cfdf75dc2847..d35383718212 100644 --- a/drivers/bluetooth/hci_h5.c +++ b/drivers/bluetooth/hci_h5.c @@ -587,6 +587,9 @@ static int h5_recv(struct hci_uart *hu, const void *data, int count) struct h5 *h5 = hu->priv; const unsigned char *ptr = data; + if (!h5) + return -ENODEV; + BT_DBG("%s pending %zu count %d", hu->hdev->name, h5->rx_pending, count); From ca40d481079c05c6891a14a798c79596fd2d5f0c Mon Sep 17 00:00:00 2001 From: SeungJu Cheon Date: Tue, 21 Apr 2026 11:51:21 +0900 Subject: [PATCH 437/525] Bluetooth: ISO: Fix data-race on dst in iso_sock_connect() iso_sock_connect() copies the destination address into iso_pi(sk)->dst under lock_sock, then releases the lock and reads it back with bacmp() to decide between the CIS and BIS connect paths: lock_sock(sk); bacpy(&iso_pi(sk)->dst, &sa->iso_bdaddr); iso_pi(sk)->dst_type = sa->iso_bdaddr_type; release_sock(sk); if (bacmp(&iso_pi(sk)->dst, BDADDR_ANY)) // <- no lock held This read after release_sock() races with any concurrent write to iso_pi(sk)->dst on the same socket. Fix by reading the destination address directly from the local sockaddr argument (sa->iso_bdaddr) instead of iso_pi(sk)->dst. Since sa is a function-local argument, reading it requires no locking and avoids the race. This patch addresses only the bacmp() race in iso_sock_connect(); other unprotected iso_pi(sk) accesses are fixed separately in the next patch. KCSAN report: BUG: KCSAN: data-race in memcmp+0x39/0xb0 race at unknown origin, with read to 0xffff8f96ea66dde3 of 1 bytes by task 549 on cpu 1: memcmp+0x39/0xb0 iso_sock_connect+0x275/0xb40 __sys_connect_file+0xbd/0xe0 __sys_connect+0xe0/0x110 __x64_sys_connect+0x40/0x50 x64_sys_call+0xcad/0x1c60 do_syscall_64+0x133/0x590 entry_SYSCALL_64_after_hwframe+0x77/0x7f value changed: 0x00 -> 0xee Reported by Kernel Concurrency Sanitizer on: CPU: 1 UID: 0 PID: 549 Comm: iso_race_combin Not tainted 7.0.0-08391-g1d51b370a0f8 #40 PREEMPT(lazy) Fixes: ccf74f2390d6 ("Bluetooth: Add BTPROTO_ISO socket type") Signed-off-by: SeungJu Cheon Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/iso.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index be145e2736b7..290a1b9a9daa 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -1193,7 +1193,7 @@ static int iso_sock_connect(struct socket *sock, struct sockaddr_unsized *addr, release_sock(sk); - if (bacmp(&iso_pi(sk)->dst, BDADDR_ANY)) + if (bacmp(&sa->iso_bdaddr, BDADDR_ANY)) err = iso_connect_cis(sk); else err = iso_connect_bis(sk); From f958c7805b18e9d69f6b322b231ecee46ec6f331 Mon Sep 17 00:00:00 2001 From: SeungJu Cheon Date: Tue, 21 Apr 2026 11:51:22 +0900 Subject: [PATCH 438/525] Bluetooth: ISO: Fix data-race on iso_pi(sk) in socket and HCI event paths Several iso_pi(sk) fields (qos, qos_user_set, bc_sid, base, base_len, sync_handle, bc_num_bis) are written under lock_sock in iso_sock_setsockopt() and iso_sock_bind(), but read and written under hci_dev_lock only in two other paths: - iso_connect_bis() / iso_connect_cis(), invoked from connect(2), read qos/base/bc_sid and reset qos to default_qos on the qos_user_set validation failure -- all without lock_sock. - iso_connect_ind(), invoked from hci_rx_work, writes sync_handle, bc_sid, qos.bcast.encryption, bc_num_bis, base and base_len on PA_SYNC_ESTABLISHED / PAST_RECEIVED / BIG_INFO_ADV_REPORT / PER_ADV_REPORT events. The BIG_INFO handler additionally passes &iso_pi(sk)->qos together with sync_handle / bc_num_bis / bc_bis to hci_conn_big_create_sync() while setsockopt may be mutating them. Acquire lock_sock around the affected accesses in both paths. The locking order hci_dev_lock -> lock_sock matches the existing iso_conn_big_sync() precedent, whose comment documents the same requirement for hci_conn_big_create_sync(). The HCI connect/bind helpers do not wait for command completion -- they enqueue work via hci_cmd_sync_queue{,_once}() / hci_le_create_cis_pending() and return -- so the added hold time is comparable to iso_conn_big_sync(). KCSAN report: BUG: KCSAN: data-race in iso_connect_cis / iso_sock_setsockopt read to 0xffffa3ae8ce3cdc8 of 1 bytes by task 335 on cpu 0: iso_connect_cis+0x49f/0xa20 iso_sock_connect+0x60e/0xb40 __sys_connect_file+0xbd/0xe0 __sys_connect+0xe0/0x110 __x64_sys_connect+0x40/0x50 x64_sys_call+0xcad/0x1c60 do_syscall_64+0x133/0x590 entry_SYSCALL_64_after_hwframe+0x77/0x7f write to 0xffffa3ae8ce3cdc8 of 60 bytes by task 334 on cpu 1: iso_sock_setsockopt+0x69a/0x930 do_sock_setsockopt+0xc3/0x170 __sys_setsockopt+0xd1/0x130 __x64_sys_setsockopt+0x64/0x80 x64_sys_call+0x1547/0x1c60 do_syscall_64+0x133/0x590 entry_SYSCALL_64_after_hwframe+0x77/0x7f Reported by Kernel Concurrency Sanitizer on: CPU: 1 UID: 0 PID: 334 Comm: iso_setup_race Not tainted 7.0.0-10949-g8541d8f725c6 #44 PREEMPT(lazy) The iso_connect_ind() races were found by inspection. Fixes: ccf74f2390d6 ("Bluetooth: Add BTPROTO_ISO socket type") Signed-off-by: SeungJu Cheon Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/iso.c | 54 +++++++++++++++++++++++++-------------------- 1 file changed, 30 insertions(+), 24 deletions(-) diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index 290a1b9a9daa..7cb2864fe872 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -347,6 +347,7 @@ static int iso_connect_bis(struct sock *sk) return -EHOSTUNREACH; hci_dev_lock(hdev); + lock_sock(sk); if (!bis_capable(hdev)) { err = -EOPNOTSUPP; @@ -399,13 +400,9 @@ static int iso_connect_bis(struct sock *sk) goto unlock; } - lock_sock(sk); - err = iso_chan_add(conn, sk, NULL); - if (err) { - release_sock(sk); + if (err) goto unlock; - } /* Update source addr of the socket */ bacpy(&iso_pi(sk)->src, &hcon->src); @@ -421,9 +418,8 @@ static int iso_connect_bis(struct sock *sk) iso_sock_set_timer(sk, READ_ONCE(sk->sk_sndtimeo)); } - release_sock(sk); - unlock: + release_sock(sk); hci_dev_unlock(hdev); hci_dev_put(hdev); return err; @@ -444,6 +440,7 @@ static int iso_connect_cis(struct sock *sk) return -EHOSTUNREACH; hci_dev_lock(hdev); + lock_sock(sk); if (!cis_central_capable(hdev)) { err = -EOPNOTSUPP; @@ -498,13 +495,9 @@ static int iso_connect_cis(struct sock *sk) goto unlock; } - lock_sock(sk); - err = iso_chan_add(conn, sk, NULL); - if (err) { - release_sock(sk); + if (err) goto unlock; - } /* Update source addr of the socket */ bacpy(&iso_pi(sk)->src, &hcon->src); @@ -520,9 +513,8 @@ static int iso_connect_cis(struct sock *sk) iso_sock_set_timer(sk, READ_ONCE(sk->sk_sndtimeo)); } - release_sock(sk); - unlock: + release_sock(sk); hci_dev_unlock(hdev); hci_dev_put(hdev); return err; @@ -2256,8 +2248,10 @@ int iso_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags) sk = iso_get_sock(hdev, &hdev->bdaddr, bdaddr, BT_LISTEN, iso_match_sid, ev1); if (sk && !ev1->status) { + lock_sock(sk); iso_pi(sk)->sync_handle = le16_to_cpu(ev1->handle); iso_pi(sk)->bc_sid = ev1->sid; + release_sock(sk); } goto done; @@ -2268,8 +2262,10 @@ int iso_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags) sk = iso_get_sock(hdev, &hdev->bdaddr, bdaddr, BT_LISTEN, iso_match_sid_past, ev1a); if (sk && !ev1a->status) { + lock_sock(sk); iso_pi(sk)->sync_handle = le16_to_cpu(ev1a->sync_handle); iso_pi(sk)->bc_sid = ev1a->sid; + release_sock(sk); } goto done; @@ -2296,27 +2292,35 @@ int iso_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags) ev2); if (sk) { - int err; - struct hci_conn *hcon = iso_pi(sk)->conn->hcon; + int err = 0; + bool big_sync; + struct hci_conn *hcon; + lock_sock(sk); + + hcon = iso_pi(sk)->conn->hcon; iso_pi(sk)->qos.bcast.encryption = ev2->encryption; if (ev2->num_bis < iso_pi(sk)->bc_num_bis) iso_pi(sk)->bc_num_bis = ev2->num_bis; - if (!test_bit(BT_SK_DEFER_SETUP, &bt_sk(sk)->flags) && - !test_and_set_bit(BT_SK_BIG_SYNC, &iso_pi(sk)->flags)) { + big_sync = !test_bit(BT_SK_DEFER_SETUP, &bt_sk(sk)->flags) && + !test_and_set_bit(BT_SK_BIG_SYNC, &iso_pi(sk)->flags); + + if (big_sync) err = hci_conn_big_create_sync(hdev, hcon, &iso_pi(sk)->qos, iso_pi(sk)->sync_handle, iso_pi(sk)->bc_num_bis, iso_pi(sk)->bc_bis); - if (err) { - bt_dev_err(hdev, "hci_le_big_create_sync: %d", - err); - sock_put(sk); - sk = NULL; - } + + release_sock(sk); + + if (big_sync && err) { + bt_dev_err(hdev, "hci_le_big_create_sync: %d", + err); + sock_put(sk); + sk = NULL; } } @@ -2370,8 +2374,10 @@ int iso_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags) if (!base || base_len > BASE_MAX_LENGTH) goto done; + lock_sock(sk); memcpy(iso_pi(sk)->base, base, base_len); iso_pi(sk)->base_len = base_len; + release_sock(sk); } else { /* This is a PA data fragment. Keep pa_data_len set to 0 * until all data has been reassembled. From 634a4408c0615c523cf7531790f4f14a422b9206 Mon Sep 17 00:00:00 2001 From: Tristan Madani Date: Tue, 21 Apr 2026 11:14:54 +0000 Subject: [PATCH 439/525] Bluetooth: btmtk: validate WMT event SKB length before struct access btmtk_usb_hci_wmt_sync() casts the WMT event response SKB data to struct btmtk_hci_wmt_evt (7 bytes) and struct btmtk_hci_wmt_evt_funcc (9 bytes) without first checking that the SKB contains enough data. A short firmware response causes out-of-bounds reads from SKB tailroom. Use skb_pull_data() to validate and advance past the base WMT event header. For the FUNC_CTRL case, pull the additional status field bytes before accessing them. Fixes: d019930b0049 ("Bluetooth: btmtk: move btusb_mtk_hci_wmt_sync to btmtk.c") Cc: stable@vger.kernel.org Signed-off-by: Tristan Madani Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btmtk.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/drivers/bluetooth/btmtk.c b/drivers/bluetooth/btmtk.c index 6fb6ca274808..f70c1b0f8990 100644 --- a/drivers/bluetooth/btmtk.c +++ b/drivers/bluetooth/btmtk.c @@ -695,8 +695,13 @@ static int btmtk_usb_hci_wmt_sync(struct hci_dev *hdev, if (data->evt_skb == NULL) goto err_free_wc; - /* Parse and handle the return WMT event */ - wmt_evt = (struct btmtk_hci_wmt_evt *)data->evt_skb->data; + wmt_evt = skb_pull_data(data->evt_skb, sizeof(*wmt_evt)); + if (!wmt_evt) { + bt_dev_err(hdev, "WMT event too short (%u bytes)", + data->evt_skb->len); + err = -EINVAL; + goto err_free_skb; + } if (wmt_evt->whdr.op != hdr->op) { bt_dev_err(hdev, "Wrong op received %d expected %d", wmt_evt->whdr.op, hdr->op); @@ -712,6 +717,12 @@ static int btmtk_usb_hci_wmt_sync(struct hci_dev *hdev, status = BTMTK_WMT_PATCH_DONE; break; case BTMTK_WMT_FUNC_CTRL: + if (!skb_pull_data(data->evt_skb, + sizeof(wmt_evt_funcc->status))) { + err = -EINVAL; + goto err_free_skb; + } + wmt_evt_funcc = (struct btmtk_hci_wmt_evt_funcc *)wmt_evt; if (be16_to_cpu(wmt_evt_funcc->status) == 0x404) status = BTMTK_WMT_ON_DONE; From 21bd244b6de5d2fe1063c23acc93fbdd2b20d112 Mon Sep 17 00:00:00 2001 From: Michael Bommarito Date: Tue, 21 Apr 2026 13:08:44 -0400 Subject: [PATCH 440/525] Bluetooth: virtio_bt: clamp rx length before skb_put virtbt_rx_work() calls skb_put(skb, len) where len comes directly from virtqueue_get_buf() with no validation against the buffer we posted to the device. The RX skb is allocated in virtbt_add_inbuf() and exposed to virtio as exactly 1000 bytes via sg_init_one(). Checking len against skb_tailroom(skb) is not sufficient because alloc_skb() can leave more tailroom than the 1000 bytes actually handed to the device. A malicious or buggy backend can therefore report used.len between 1001 and skb_tailroom(skb), causing skb_put() to include uninitialized kernel heap bytes that were never written by the device. The same path also accepts len == 0, in which case skb_put(skb, 0) leaves the skb empty but virtbt_rx_handle() still reads the pkt_type byte from skb->data, consuming uninitialized memory. Define VIRTBT_RX_BUF_SIZE once and reuse it in alloc_skb() and sg_init_one(), and gate virtbt_rx_work() on that same constant so the bound checked matches the buffer actually exposed to the device. Reject used.len == 0 in the same gate so an empty completion can no longer reach virtbt_rx_handle(). Use bt_dev_err_ratelimited() because the length value comes from an untrusted backend that can otherwise flood the kernel log. Same class of bug as commit c04db81cd028 ("net/9p: Fix buffer overflow in USB transport layer"), which hardened the USB 9p transport against unchecked device-reported length. Fixes: 160fbcf3bfb9 ("Bluetooth: virtio_bt: Use skb_put to set length") Cc: stable@vger.kernel.org Cc: Soenke Huster Signed-off-by: Michael Bommarito Assisted-by: Claude:claude-opus-4-7 Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/virtio_bt.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/drivers/bluetooth/virtio_bt.c b/drivers/bluetooth/virtio_bt.c index 76d61af8a275..2c5c39356a1c 100644 --- a/drivers/bluetooth/virtio_bt.c +++ b/drivers/bluetooth/virtio_bt.c @@ -12,6 +12,7 @@ #include #define VERSION "0.1" +#define VIRTBT_RX_BUF_SIZE 1000 enum { VIRTBT_VQ_TX, @@ -33,11 +34,11 @@ static int virtbt_add_inbuf(struct virtio_bluetooth *vbt) struct sk_buff *skb; int err; - skb = alloc_skb(1000, GFP_KERNEL); + skb = alloc_skb(VIRTBT_RX_BUF_SIZE, GFP_KERNEL); if (!skb) return -ENOMEM; - sg_init_one(sg, skb->data, 1000); + sg_init_one(sg, skb->data, VIRTBT_RX_BUF_SIZE); err = virtqueue_add_inbuf(vq, sg, 1, skb, GFP_KERNEL); if (err < 0) { @@ -227,8 +228,15 @@ static void virtbt_rx_work(struct work_struct *work) if (!skb) return; - skb_put(skb, len); - virtbt_rx_handle(vbt, skb); + if (!len || len > VIRTBT_RX_BUF_SIZE) { + bt_dev_err_ratelimited(vbt->hdev, + "rx reply len %u outside [1, %u]\n", + len, VIRTBT_RX_BUF_SIZE); + kfree_skb(skb); + } else { + skb_put(skb, len); + virtbt_rx_handle(vbt, skb); + } if (virtbt_add_inbuf(vbt) < 0) return; From daf23014e5d975e72ea9c02b5160d3fcf070ea47 Mon Sep 17 00:00:00 2001 From: Michael Bommarito Date: Tue, 21 Apr 2026 13:08:45 -0400 Subject: [PATCH 441/525] Bluetooth: virtio_bt: validate rx pkt_type header length virtbt_rx_handle() reads the leading pkt_type byte from the RX skb and forwards the remainder to hci_recv_frame() for every event/ACL/SCO/ISO type, without checking that the remaining payload is at least the fixed HCI header for that type. After the preceding patch bounds the backend-supplied used.len to [1, VIRTBT_RX_BUF_SIZE], a one-byte completion still reaches hci_recv_frame() with skb->len already pulled to 0. If the byte happened to be HCI_ACLDATA_PKT, the ACL-vs-ISO classification fast-path in hci_dev_classify_pkt_type() dereferences hci_acl_hdr(skb)->handle whenever the HCI device has an active CIS_LINK, BIS_LINK, or PA_LINK connection, reading two bytes of uninitialized RX-buffer data. The same hazard exists for every packet type the driver accepts because none of the switch cases in virtbt_rx_handle() check skb->len against the per-type minimum HCI header size before handing the frame to the core. After stripping pkt_type, require skb->len to cover the fixed header size for the selected type (event 2, ACL 4, SCO 3, ISO 4) before calling hci_recv_frame(); drop ratelimited otherwise. Unknown pkt_type values still take the original kfree_skb() default path. Use bt_dev_err_ratelimited() because both the length and pkt_type values come from an untrusted backend that can otherwise flood the kernel log. Fixes: 160fbcf3bfb9 ("Bluetooth: virtio_bt: Use skb_put to set length") Cc: stable@vger.kernel.org Cc: Soenke Huster Signed-off-by: Michael Bommarito Assisted-by: Claude:claude-opus-4-7 Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/virtio_bt.c | 23 ++++++++++++++++++++--- 1 file changed, 20 insertions(+), 3 deletions(-) diff --git a/drivers/bluetooth/virtio_bt.c b/drivers/bluetooth/virtio_bt.c index 2c5c39356a1c..140ab55c9fc5 100644 --- a/drivers/bluetooth/virtio_bt.c +++ b/drivers/bluetooth/virtio_bt.c @@ -198,6 +198,7 @@ static int virtbt_shutdown_generic(struct hci_dev *hdev) static void virtbt_rx_handle(struct virtio_bluetooth *vbt, struct sk_buff *skb) { + size_t min_hdr; __u8 pkt_type; pkt_type = *((__u8 *) skb->data); @@ -205,16 +206,32 @@ static void virtbt_rx_handle(struct virtio_bluetooth *vbt, struct sk_buff *skb) switch (pkt_type) { case HCI_EVENT_PKT: + min_hdr = sizeof(struct hci_event_hdr); + break; case HCI_ACLDATA_PKT: + min_hdr = sizeof(struct hci_acl_hdr); + break; case HCI_SCODATA_PKT: + min_hdr = sizeof(struct hci_sco_hdr); + break; case HCI_ISODATA_PKT: - hci_skb_pkt_type(skb) = pkt_type; - hci_recv_frame(vbt->hdev, skb); + min_hdr = sizeof(struct hci_iso_hdr); break; default: kfree_skb(skb); - break; + return; } + + if (skb->len < min_hdr) { + bt_dev_err_ratelimited(vbt->hdev, + "rx pkt_type 0x%02x payload %u < hdr %zu\n", + pkt_type, skb->len, min_hdr); + kfree_skb(skb); + return; + } + + hci_skb_pkt_type(skb) = pkt_type; + hci_recv_frame(vbt->hdev, skb); } static void virtbt_rx_work(struct work_struct *work) From 8f59d17b18a78fdfdbb67d693b3d3eb03db184e0 Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Thu, 23 Apr 2026 23:31:00 +0800 Subject: [PATCH 442/525] Bluetooth: RFCOMM: pull credit byte with skb_pull_data() rfcomm_recv_data() treats the first payload byte as a credit field when the UIH frame carries PF and credit-based flow control is enabled. After the header has been stripped, the PF/CFC path consumes that byte with a direct skb->data dereference followed by skb_pull(). A malformed short frame can reach this path without a byte available. Use skb_pull_data() so the length check and pull happen together before the returned credit byte is consumed. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Pengpeng Hou Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/rfcomm/core.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/net/bluetooth/rfcomm/core.c b/net/bluetooth/rfcomm/core.c index 611a9a94151e..d11bd5337d57 100644 --- a/net/bluetooth/rfcomm/core.c +++ b/net/bluetooth/rfcomm/core.c @@ -1715,9 +1715,12 @@ static int rfcomm_recv_data(struct rfcomm_session *s, u8 dlci, int pf, struct sk } if (pf && d->cfc) { - u8 credits = *(u8 *) skb->data; skb_pull(skb, 1); + u8 *credits = skb_pull_data(skb, 1); - d->tx_credits += credits; + if (!credits) + goto drop; + + d->tx_credits += *credits; if (d->tx_credits) clear_bit(RFCOMM_TX_THROTTLED, &d->flags); } From 72d97cae2a83cecf6f47208646675ecd066d0a3e Mon Sep 17 00:00:00 2001 From: Jann Horn Date: Wed, 29 Apr 2026 15:40:46 +0200 Subject: [PATCH 443/525] Bluetooth: hci_event: fix memset typo hci_le_big_sync_established_evt() currently does: conn->num_bis = 0; memset(conn->bis, 0, sizeof(conn->num_bis)); sizeof(conn->num_bis) is wrong - it would make sense to either use conn->num_bis (before setting that to 0) or sizeof(conn->bis). Fix it by using sizeof(conn->bis), the least intrusive change. Luckily, nothing actually depends on this memset() working properly: Nothing seems to ever read from conn->bis beyond conn->num_bis, and when conn->num_bis is increased, the corresponding elements of conn->bis are initialized. So I think this line could also just be removed. This is a purely theoretical fix and should have no impact on actual behavior. Fixes: 42ecf1947135 ("Bluetooth: ISO: Do not emit LE BIG Create Sync if previous is pending") Signed-off-by: Jann Horn Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_event.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 1b3b9131affa..eea2f810aafa 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -7191,7 +7191,7 @@ static void hci_le_big_sync_established_evt(struct hci_dev *hdev, void *data, clear_bit(HCI_CONN_CREATE_BIG_SYNC, &conn->flags); conn->num_bis = 0; - memset(conn->bis, 0, sizeof(conn->num_bis)); + memset(conn->bis, 0, sizeof(conn->bis)); for (i = 0; i < ev->num_bis; i++) { u16 handle = le16_to_cpu(ev->bis[i]); From c5d415596cb6fbdf6334b06cc87a1a5a268d8725 Mon Sep 17 00:00:00 2001 From: Michael Bommarito Date: Sat, 2 May 2026 12:43:03 -0400 Subject: [PATCH 444/525] Bluetooth: HIDP: serialise l2cap_unregister_user via hidp_session_sem Commit dbf666e4fc9b ("Bluetooth: HIDP: Fix possible UAF") made hidp_session_remove() drop the L2CAP reference and set session->conn = NULL once the session is considered removed, and added a bare if (session->conn) guard around the kthread-exit l2cap_unregister_user() call in hidp_session_thread(). The sibling ioctl site in hidp_connection_del() still reads session->conn unlocked and unguarded, and the kthread-exit guard itself is a lockless double-read. hidp_session_find() drops hidp_session_sem before returning, so hidp_session_remove() can null session->conn between the lookup and the call in hidp_connection_del(). Worse, since commit 752a6c9596dd ("Bluetooth: L2CAP: Fix use-after-free in l2cap_unregister_user") takes mutex_lock(&conn->lock) inside l2cap_unregister_user(), a stale non-NULL snapshot also UAFs on conn->lock. v1 only added an if (session->conn) guard at the ioctl site, which doesn't address either race; Luiz suggested snapshotting session->conn under the sem and clearing it before the call. Taking hidp_session_sem across l2cap_unregister_user() would be wrong: l2cap_conn_del() already establishes the lock order conn->lock -> hidp_session_sem via l2cap_unregister_all_users() -> user->remove == hidp_session_remove(), so taking hidp_session_sem before conn->lock would AB/BA deadlock. Factor a helper hidp_session_unregister_conn() that under down_write(&hidp_session_sem) snapshots session->conn and clears the member, then outside the sem calls l2cap_unregister_user() and l2cap_conn_put() on the snapshot. Call it from both hidp_connection_del() and hidp_session_thread()'s exit path. At most one consumer wins the write-sem; later callers observe session->conn == NULL and skip the unregister and put, so the reference hidp_session_new() took via l2cap_conn_get() is consumed exactly once. session_free() already tolerates a NULL session->conn. Fixes: dbf666e4fc9b ("Bluetooth: HIDP: Fix possible UAF") Suggested-by: Luiz Augusto von Dentz Link: https://lore.kernel.org/all/20260422011437.176643-1-michael.bommarito@gmail.com/ Signed-off-by: Michael Bommarito Assisted-by: Claude:claude-opus-4-7 Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hidp/core.c | 27 ++++++++++++++++++++++++--- 1 file changed, 24 insertions(+), 3 deletions(-) diff --git a/net/bluetooth/hidp/core.c b/net/bluetooth/hidp/core.c index 7bcf8c5ceaee..976f91eeb745 100644 --- a/net/bluetooth/hidp/core.c +++ b/net/bluetooth/hidp/core.c @@ -1035,6 +1035,28 @@ static struct hidp_session *hidp_session_find(const bdaddr_t *bdaddr) return session; } +/* + * Consume session->conn: clear the member under hidp_session_sem, then + * l2cap_unregister_user() and l2cap_conn_put() the snapshot outside the + * sem. At most one caller wins; later callers see NULL and skip. The + * reference is the one hidp_session_new() took via l2cap_conn_get(). + */ +static void hidp_session_unregister_conn(struct hidp_session *session) +{ + struct l2cap_conn *conn; + + down_write(&hidp_session_sem); + conn = session->conn; + if (conn) + session->conn = NULL; + up_write(&hidp_session_sem); + + if (conn) { + l2cap_unregister_user(conn, &session->user); + l2cap_conn_put(conn); + } +} + /* * Start session synchronously * This starts a session thread and waits until initialization @@ -1311,8 +1333,7 @@ static int hidp_session_thread(void *arg) * Instead, this call has the same semantics as if user-space tried to * delete the session. */ - if (session->conn) - l2cap_unregister_user(session->conn, &session->user); + hidp_session_unregister_conn(session); hidp_session_put(session); @@ -1418,7 +1439,7 @@ int hidp_connection_del(struct hidp_conndel_req *req) HIDP_CTRL_VIRTUAL_CABLE_UNPLUG, NULL, 0); else - l2cap_unregister_user(session->conn, &session->user); + hidp_session_unregister_conn(session); hidp_session_put(session); From 0e1368a28dd5231ae0dbe240dfe0ff2657de5647 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 6 May 2026 17:22:05 -0700 Subject: [PATCH 445/525] selftests: drv-net: fix sort order of makefile and config Recent changes added configs and tests in the wrong spot. Link: https://lore.kernel.org/20260506170435.34984dfc@kernel.org Signed-off-by: Jakub Kicinski --- tools/testing/selftests/drivers/net/hw/Makefile | 2 +- tools/testing/selftests/drivers/net/hw/config | 8 ++++---- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/tools/testing/selftests/drivers/net/hw/Makefile b/tools/testing/selftests/drivers/net/hw/Makefile index 3b6ff4708005..82809d5b2478 100644 --- a/tools/testing/selftests/drivers/net/hw/Makefile +++ b/tools/testing/selftests/drivers/net/hw/Makefile @@ -30,8 +30,8 @@ TEST_PROGS = \ gro_hw.py \ hw_stats_l3.sh \ hw_stats_l3_gre.sh \ - ipsec_vxlan.py \ iou-zcrx.py \ + ipsec_vxlan.py \ irq.py \ loopback.sh \ nic_timestamp.py \ diff --git a/tools/testing/selftests/drivers/net/hw/config b/tools/testing/selftests/drivers/net/hw/config index ae0168c2bbe6..8c132ace2b8d 100644 --- a/tools/testing/selftests/drivers/net/hw/config +++ b/tools/testing/selftests/drivers/net/hw/config @@ -3,6 +3,10 @@ CONFIG_FAIL_FUNCTION=y CONFIG_FAULT_INJECTION=y CONFIG_FAULT_INJECTION_DEBUG_FS=y CONFIG_FUNCTION_ERROR_INJECTION=y +CONFIG_INET6_ESP=y +CONFIG_INET6_ESP_OFFLOAD=y +CONFIG_INET_ESP=y +CONFIG_INET_ESP_OFFLOAD=y CONFIG_IO_URING=y CONFIG_IPV6=y CONFIG_IPV6_GRE=y @@ -12,10 +16,6 @@ CONFIG_NET_IPGRE=y CONFIG_NET_IPGRE_DEMUX=y CONFIG_NETKIT=y CONFIG_NET_SCH_INGRESS=y -CONFIG_INET6_ESP=y -CONFIG_INET6_ESP_OFFLOAD=y -CONFIG_INET_ESP=y -CONFIG_INET_ESP_OFFLOAD=y CONFIG_UDMABUF=y CONFIG_VXLAN=y CONFIG_XFRM_USER=y From 7aaa8f5e45a92678256c1e17f1fa2c2f45c61dd1 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 May 2026 13:00:56 +0000 Subject: [PATCH 446/525] ipv6: fix potential UAF caused by ip6_forward_proxy_check() ip6_forward_proxy_check() calls pskb_may_pull() which might re-allocate skb->head. Reload ipv6_hdr() after the pskb_may_pull() call to avoid using the freed memory. Fixes: e21e0b5f19ac ("[IPV6] NDISC: Handle NDP messages to proxied addresses.") Reported-by: Damiano Melotti Signed-off-by: Eric Dumazet Reviewed-by: David Ahern Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260505130056.2927197-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/ipv6/ip6_output.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index 1f2a33fbed6e..c14adcdd4396 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -468,6 +468,7 @@ static int ip6_forward_proxy_check(struct sk_buff *skb) default: break; } + hdr = ipv6_hdr(skb); } /* @@ -582,6 +583,8 @@ int ip6_forward(struct sk_buff *skb) if (READ_ONCE(net->ipv6.devconf_all->proxy_ndp) && pneigh_lookup(&nd_tbl, net, &hdr->daddr, skb->dev)) { int proxied = ip6_forward_proxy_check(skb); + + hdr = ipv6_hdr(skb); if (proxied > 0) { /* It's tempting to decrease the hop limit * here by 1, as we do at the end of the From 7ce3f1bedaac88880594720ba0f687da3bd7fc8a Mon Sep 17 00:00:00 2001 From: Daniel Zahka Date: Tue, 5 May 2026 03:42:23 -0700 Subject: [PATCH 447/525] netdevsim: psp: only call nsim_psp_uninit() on PFs VFs go through nsim_init_netdevsim_vf() which never calls nsim_psp_init(), so ns->psp.dev stays NULL. nsim_psp_uninit() guards with !IS_ERR(ns->psp.dev), so destroying a VF reaches psp_dev_unregister(NULL) and dereferences NULL on the first mutex_lock(&psd->lock): BUG: kernel NULL pointer dereference, address: 0000000000000020 RIP: 0010:mutex_lock+0x1c/0x30 Call Trace: psp_dev_unregister+0x2a/0x1a0 nsim_psp_uninit+0x1f/0x40 [netdevsim] nsim_destroy+0x61/0x1e0 [netdevsim] __nsim_dev_port_del+0x47/0x90 [netdevsim] nsim_drv_configure_vfs+0xc9/0x130 [netdevsim] nsim_bus_dev_numvfs_store+0x79/0xb0 [netdevsim] Gate nsim_psp_uninit() on nsim_dev_port_is_pf(), matching the pattern already used for nsim_exit_netdevsim() and the bpf/ipsec/macsec/queue teardowns. Reproducer: modprobe netdevsim echo "10 1" > /sys/bus/netdevsim/new_device echo 1 > /sys/bus/netdevsim/devices/netdevsim10/sriov_numvfs devlink dev eswitch set netdevsim/netdevsim10 mode switchdev echo 0 > /sys/bus/netdevsim/devices/netdevsim10/sriov_numvfs Fixes: f857478d6206 ("netdevsim: a basic test PSP implementation") Signed-off-by: Daniel Zahka Reviewed-by: Willem de Bruijn Link: https://patch.msgid.link/20260505-psd-rcu-v1-1-a8f69ec1ab96@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/netdevsim/netdev.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/netdevsim/netdev.c b/drivers/net/netdevsim/netdev.c index a05af192caf3..a750768912b5 100644 --- a/drivers/net/netdevsim/netdev.c +++ b/drivers/net/netdevsim/netdev.c @@ -1182,7 +1182,8 @@ void nsim_destroy(struct netdevsim *ns) unregister_netdevice_notifier_dev_net(ns->netdev, &ns->nb, &ns->nn); - nsim_psp_uninit(ns); + if (nsim_dev_port_is_pf(ns->nsim_dev_port)) + nsim_psp_uninit(ns); rtnl_lock(); peer = rtnl_dereference(ns->peer); From 24c96a42006ee27a078ec8c631c906dea8a3ca6d Mon Sep 17 00:00:00 2001 From: Daniel Zahka Date: Tue, 5 May 2026 03:42:24 -0700 Subject: [PATCH 448/525] netdevsim: psp: serialize calls to nsim_psp_uninit() The debugfs write handler, nsim_psp_rereg_write(), can race against nsim_destroy() and against itself, causing nsim_psp_uninit() to run more than once concurrently. Two complementary changes serialize all callers: 1. Delete the psp_rereg debugfs file from nsim_psp_uninit() before doing the actual teardown. debugfs_remove() drains any in-flight writers and prevents new ones from starting. 2. Add a mutex around the body of nsim_psp_rereg_write() so that two concurrent userspace writers cannot both enter the teardown path at once. The teardown work itself is moved into a new __nsim_psp_uninit() that the rereg handler calls under the mutex, while the public nsim_psp_uninit() wraps it with the debugfs_remove()/mutex_destroy() pair so nsim_destroy() doesn't have to know about the psp internals. Fixes: f857478d6206 ("netdevsim: a basic test PSP implementation") Signed-off-by: Daniel Zahka Reviewed-by: Willem de Bruijn Link: https://patch.msgid.link/20260505-psd-rcu-v1-2-a8f69ec1ab96@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/netdevsim/netdevsim.h | 2 ++ drivers/net/netdevsim/psp.c | 17 ++++++++++++++--- 2 files changed, 16 insertions(+), 3 deletions(-) diff --git a/drivers/net/netdevsim/netdevsim.h b/drivers/net/netdevsim/netdevsim.h index 7e129dddbbe7..e373ffc26b0c 100644 --- a/drivers/net/netdevsim/netdevsim.h +++ b/drivers/net/netdevsim/netdevsim.h @@ -121,6 +121,8 @@ struct netdevsim { u64_stats_t tx_bytes; struct u64_stats_sync syncp; struct psp_dev *dev; + struct dentry *rereg; + struct mutex rereg_lock; u32 spi; u32 assoc_cnt; } psp; diff --git a/drivers/net/netdevsim/psp.c b/drivers/net/netdevsim/psp.c index 0b4d717253b0..86d84b7e566b 100644 --- a/drivers/net/netdevsim/psp.c +++ b/drivers/net/netdevsim/psp.c @@ -209,13 +209,20 @@ static struct psp_dev_caps nsim_psp_caps = { .assoc_drv_spc = sizeof(void *), }; -void nsim_psp_uninit(struct netdevsim *ns) +static void __nsim_psp_uninit(struct netdevsim *ns) { if (!IS_ERR(ns->psp.dev)) psp_dev_unregister(ns->psp.dev); WARN_ON(ns->psp.assoc_cnt); } +void nsim_psp_uninit(struct netdevsim *ns) +{ + debugfs_remove(ns->psp.rereg); + mutex_destroy(&ns->psp.rereg_lock); + __nsim_psp_uninit(ns); +} + static ssize_t nsim_psp_rereg_write(struct file *file, const char __user *data, size_t count, loff_t *ppos) @@ -223,11 +230,13 @@ nsim_psp_rereg_write(struct file *file, const char __user *data, size_t count, struct netdevsim *ns = file->private_data; int err; - nsim_psp_uninit(ns); + mutex_lock(&ns->psp.rereg_lock); + __nsim_psp_uninit(ns); ns->psp.dev = psp_dev_create(ns->netdev, &nsim_psp_ops, &nsim_psp_caps, ns); err = PTR_ERR_OR_ZERO(ns->psp.dev); + mutex_unlock(&ns->psp.rereg_lock); return err ?: count; } @@ -249,6 +258,8 @@ int nsim_psp_init(struct netdevsim *ns) if (err) return err; - debugfs_create_file("psp_rereg", 0200, ddir, ns, &nsim_psp_rereg_fops); + mutex_init(&ns->psp.rereg_lock); + ns->psp.rereg = debugfs_create_file("psp_rereg", 0200, ddir, ns, + &nsim_psp_rereg_fops); return 0; } From 07bdec3fc737aac7f4c273aafa803d353174c43e Mon Sep 17 00:00:00 2001 From: Daniel Zahka Date: Tue, 5 May 2026 03:42:25 -0700 Subject: [PATCH 449/525] netdevsim: psp: rcu protect psp_dev reference There are two issues with the way psp_dev is used in nsim_do_psp(): 1. There is no check for IS_ERR() on the peers psp_dev, before dereferencing. 2. The refcount on this psp_dev can be dropped by nsim_psp_rereg_write() To fix this, we can make netdevsim's reference to its psp_dev an rcu reference, and then nsim_do_psp() can read the fields it needs from an rcu critical section. Fixes: f857478d6206 ("netdevsim: a basic test PSP implementation") Signed-off-by: Daniel Zahka Reviewed-by: Willem de Bruijn Link: https://patch.msgid.link/20260505-psd-rcu-v1-3-a8f69ec1ab96@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/netdevsim/netdevsim.h | 2 +- drivers/net/netdevsim/psp.c | 54 ++++++++++++++++++++----------- 2 files changed, 36 insertions(+), 20 deletions(-) diff --git a/drivers/net/netdevsim/netdevsim.h b/drivers/net/netdevsim/netdevsim.h index e373ffc26b0c..d909c4160ea1 100644 --- a/drivers/net/netdevsim/netdevsim.h +++ b/drivers/net/netdevsim/netdevsim.h @@ -120,7 +120,7 @@ struct netdevsim { u64_stats_t tx_packets; u64_stats_t tx_bytes; struct u64_stats_sync syncp; - struct psp_dev *dev; + struct psp_dev __rcu *dev; struct dentry *rereg; struct mutex rereg_lock; u32 spi; diff --git a/drivers/net/netdevsim/psp.c b/drivers/net/netdevsim/psp.c index 86d84b7e566b..6936ecb8173e 100644 --- a/drivers/net/netdevsim/psp.c +++ b/drivers/net/netdevsim/psp.c @@ -19,6 +19,7 @@ nsim_do_psp(struct sk_buff *skb, struct netdevsim *ns, struct netdevsim *peer_ns, struct skb_ext **psp_ext) { enum skb_drop_reason rc = 0; + struct psp_dev *peer_psd; struct psp_assoc *pas; struct net *net; void **ptr; @@ -48,7 +49,8 @@ nsim_do_psp(struct sk_buff *skb, struct netdevsim *ns, } /* Now pretend we just received this frame */ - if (peer_ns->psp.dev->config.versions & (1 << pas->version)) { + peer_psd = rcu_dereference(peer_ns->psp.dev); + if (peer_psd && peer_psd->config.versions & (1 << pas->version)) { bool strip_icv = false; u8 generation; @@ -61,8 +63,7 @@ nsim_do_psp(struct sk_buff *skb, struct netdevsim *ns, skb_ext_reset(skb); skb->mac_len = ETH_HLEN; - if (psp_dev_rcv(skb, peer_ns->psp.dev->id, generation, - strip_icv)) { + if (psp_dev_rcv(skb, peer_psd->id, generation, strip_icv)) { rc = SKB_DROP_REASON_PSP_OUTPUT; goto out_unlock; } @@ -209,10 +210,18 @@ static struct psp_dev_caps nsim_psp_caps = { .assoc_drv_spc = sizeof(void *), }; -static void __nsim_psp_uninit(struct netdevsim *ns) +static void __nsim_psp_uninit(struct netdevsim *ns, bool teardown) { - if (!IS_ERR(ns->psp.dev)) - psp_dev_unregister(ns->psp.dev); + struct psp_dev *psd; + + psd = rcu_dereference_protected(ns->psp.dev, + teardown || + lockdep_is_held(&ns->psp.rereg_lock)); + if (psd) { + rcu_assign_pointer(ns->psp.dev, NULL); + synchronize_rcu(); + psp_dev_unregister(psd); + } WARN_ON(ns->psp.assoc_cnt); } @@ -220,7 +229,7 @@ void nsim_psp_uninit(struct netdevsim *ns) { debugfs_remove(ns->psp.rereg); mutex_destroy(&ns->psp.rereg_lock); - __nsim_psp_uninit(ns); + __nsim_psp_uninit(ns, true); } static ssize_t @@ -228,16 +237,23 @@ nsim_psp_rereg_write(struct file *file, const char __user *data, size_t count, loff_t *ppos) { struct netdevsim *ns = file->private_data; - int err; + struct psp_dev *psd; + ssize_t ret; mutex_lock(&ns->psp.rereg_lock); - __nsim_psp_uninit(ns); + __nsim_psp_uninit(ns, false); - ns->psp.dev = psp_dev_create(ns->netdev, &nsim_psp_ops, - &nsim_psp_caps, ns); - err = PTR_ERR_OR_ZERO(ns->psp.dev); + psd = psp_dev_create(ns->netdev, &nsim_psp_ops, &nsim_psp_caps, ns); + if (IS_ERR(psd)) { + ret = PTR_ERR(psd); + goto out; + } + + rcu_assign_pointer(ns->psp.dev, psd); + ret = count; +out: mutex_unlock(&ns->psp.rereg_lock); - return err ?: count; + return ret; } static const struct file_operations nsim_psp_rereg_fops = { @@ -250,13 +266,13 @@ static const struct file_operations nsim_psp_rereg_fops = { int nsim_psp_init(struct netdevsim *ns) { struct dentry *ddir = ns->nsim_dev_port->ddir; - int err; + struct psp_dev *psd; - ns->psp.dev = psp_dev_create(ns->netdev, &nsim_psp_ops, - &nsim_psp_caps, ns); - err = PTR_ERR_OR_ZERO(ns->psp.dev); - if (err) - return err; + psd = psp_dev_create(ns->netdev, &nsim_psp_ops, &nsim_psp_caps, ns); + if (IS_ERR(psd)) + return PTR_ERR(psd); + + rcu_assign_pointer(ns->psp.dev, psd); mutex_init(&ns->psp.rereg_lock); ns->psp.rereg = debugfs_create_file("psp_rereg", 0200, ddir, ns, From 701ea57feaabdea403cf299ee5cd0445083bc0ac Mon Sep 17 00:00:00 2001 From: Shitalkumar Gandhi Date: Tue, 5 May 2026 18:02:36 +0530 Subject: [PATCH 450/525] net: rtsn: fix mdio_node leak in rtsn_mdio_alloc() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit of_get_child_by_name() takes a reference. The rtsn_reset() and rtsn_change_mode() failure paths jump to out_free_bus and leak mdio_node. Add out_put_node to drop it before falling through. Fixes: b0d3969d2b4d ("net: ethernet: rtsn: Add support for Renesas Ethernet-TSN") Signed-off-by: Shitalkumar Gandhi Reviewed-by: Geert Uytterhoeven Reviewed-by: Andrew Lunn Reviewed-by: Niklas Söderlund Link: https://patch.msgid.link/20260505123236.406000-1-shitalkumar.gandhi@cambiumnetworks.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/renesas/rtsn.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/renesas/rtsn.c b/drivers/net/ethernet/renesas/rtsn.c index 03a2669f0518..ee8381b60b8d 100644 --- a/drivers/net/ethernet/renesas/rtsn.c +++ b/drivers/net/ethernet/renesas/rtsn.c @@ -797,11 +797,11 @@ static int rtsn_mdio_alloc(struct rtsn_private *priv) /* Enter config mode before registering the MDIO bus */ ret = rtsn_reset(priv); if (ret) - goto out_free_bus; + goto out_put_node; ret = rtsn_change_mode(priv, OCR_OPC_CONFIG); if (ret) - goto out_free_bus; + goto out_put_node; rtsn_modify(priv, MPIC, MPIC_PSMCS_MASK | MPIC_PSMHT_MASK, MPIC_PSMCS_DEFAULT | MPIC_PSMHT_DEFAULT); @@ -824,6 +824,8 @@ static int rtsn_mdio_alloc(struct rtsn_private *priv) return 0; +out_put_node: + of_node_put(mdio_node); out_free_bus: mdiobus_free(mii); return ret; From 67ef49047d312be692c8c439145f4514174e517f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 May 2026 13:32:33 +0000 Subject: [PATCH 451/525] inetpeer: add a missing read_seqretry() in inet_getpeer() When performing a lockless lookup over the inet_peer rbtree, if a matching node is found, inet_getpeer() returns it immediately without validating the seqlock sequence. This missing check introduces a race condition: Trigger Path: When a host receives an incoming fragmented IPv4 packet, ip4_frag_init() (in net/ipv4/ip_fragment.c) calls inet_getpeer_v4() to track the peer. The Race: If the packet is from a new source IP, CPU A acquires the write_seqlock, allocates a new inet_peer node (p), sets its IP address (daddr), and links it to the rbtree (rb_link_node). Uninitialized Access: Due to the lack of memory barriers between rb_link_node and the initialization of the rest of the struct (like refcount_set(&p->refcnt, 1)), CPU A can make the node visible to readers before its refcnt is initialized. This is especially true on weakly-ordered architectures like ARM64 where the CPU can reorder the memory stores. Lockless Reader: Concurrently, CPU B processes a second fragmented packet from the same source IP. CPU B does a lockless lookup, finds the newly inserted node, and returns it immediately. Use-After-Free (UAF): CPU B reads p->refcnt as uninitialized garbage (left over from previous kmalloc-128/192 allocations). If the garbage is > 0, refcount_inc_not_zero(&p->refcnt) succeeds. CPU A then executes refcount_set(&p->refcnt, 1), overwriting CPU B's increment. When CPU B finishes with the fragment queue, it calls inet_putpeer(), which drops the refcount to 0 and frees the node via RCU. The node is now freed but remains linked in the rbtree, resulting in a Use-After-Free in the rbtree. Fixes: b145425f269a ("inetpeer: remove AVL implementation in favor of RB tree") Reported-by: Damiano Melotti Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260505133233.3039575-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/ipv4/inetpeer.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/ipv4/inetpeer.c b/net/ipv4/inetpeer.c index d8083b9033c2..5b957a831e7c 100644 --- a/net/ipv4/inetpeer.c +++ b/net/ipv4/inetpeer.c @@ -179,7 +179,8 @@ struct inet_peer *inet_getpeer(struct inet_peer_base *base, seq = read_seqbegin(&base->lock); p = lookup(daddr, base, seq, NULL, &gc_cnt, &parent, &pp); - if (p) + /* Make sure tree was not modified during our lookup. */ + if (p && !read_seqretry(&base->lock, seq)) return p; /* retry an exact lookup, taking the lock before. From 770b136ff9bf3e319d19875da59c4f7f4853da3a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 May 2026 09:11:33 +0000 Subject: [PATCH 452/525] net/sched: sch_sfq: annotate data-races from sfq_dump_class_stats() sfq_dump_class_stats() runs locklessly, add needed READ_ONCE() and WRITE_ONCE() annotations. Fixes: edb09eb17ed8 ("net: sched: do not acquire qdisc spinlock in qdisc/class stats dump") Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260505091133.2452510-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/sched/sch_sfq.c | 48 +++++++++++++++++++++++---------------------- 1 file changed, 25 insertions(+), 23 deletions(-) diff --git a/net/sched/sch_sfq.c b/net/sched/sch_sfq.c index c3f3181dba54..f39822babf88 100644 --- a/net/sched/sch_sfq.c +++ b/net/sched/sch_sfq.c @@ -225,7 +225,8 @@ static inline void sfq_dec(struct sfq_sched_data *q, sfq_index x) sfq_unlink(q, x, n, p); - d = q->slots[x].qlen--; + d = q->slots[x].qlen; + WRITE_ONCE(q->slots[x].qlen, d - 1); if (n == p && q->cur_depth == d) q->cur_depth--; sfq_link(q, x); @@ -238,7 +239,8 @@ static inline void sfq_inc(struct sfq_sched_data *q, sfq_index x) sfq_unlink(q, x, n, p); - d = ++q->slots[x].qlen; + d = q->slots[x].qlen + 1; + WRITE_ONCE(q->slots[x].qlen, d); if (q->cur_depth < d) q->cur_depth = d; sfq_link(q, x); @@ -298,7 +300,7 @@ static unsigned int sfq_drop(struct Qdisc *sch, struct sk_buff **to_free) drop: skb = q->headdrop ? slot_dequeue_head(slot) : slot_dequeue_tail(slot); len = qdisc_pkt_len(skb); - slot->backlog -= len; + WRITE_ONCE(slot->backlog, slot->backlog - len); sfq_dec(q, x); sch->q.qlen--; qdisc_qstats_backlog_dec(sch, skb); @@ -314,7 +316,7 @@ static unsigned int sfq_drop(struct Qdisc *sch, struct sk_buff **to_free) q->tail = NULL; /* no more active slots */ else q->tail->next = slot->next; - q->ht[slot->hash] = SFQ_EMPTY_SLOT; + WRITE_ONCE(q->ht[slot->hash], SFQ_EMPTY_SLOT); goto drop; } @@ -364,10 +366,10 @@ sfq_enqueue(struct sk_buff *skb, struct Qdisc *sch, struct sk_buff **to_free) x = q->dep[0].next; /* get a free slot */ if (x >= SFQ_MAX_FLOWS) return qdisc_drop_reason(skb, sch, to_free, QDISC_DROP_MAXFLOWS); - q->ht[hash] = x; + WRITE_ONCE(q->ht[hash], x); slot = &q->slots[x]; slot->hash = hash; - slot->backlog = 0; /* should already be 0 anyway... */ + WRITE_ONCE(slot->backlog, 0); /* should already be 0 anyway... */ red_set_vars(&slot->vars); goto enqueue; } @@ -426,7 +428,7 @@ sfq_enqueue(struct sk_buff *skb, struct Qdisc *sch, struct sk_buff **to_free) head = slot_dequeue_head(slot); delta = qdisc_pkt_len(head) - qdisc_pkt_len(skb); sch->qstats.backlog -= delta; - slot->backlog -= delta; + WRITE_ONCE(slot->backlog, slot->backlog - delta); qdisc_drop_reason(head, sch, to_free, QDISC_DROP_FLOW_LIMIT); slot_queue_add(slot, skb); @@ -436,7 +438,7 @@ sfq_enqueue(struct sk_buff *skb, struct Qdisc *sch, struct sk_buff **to_free) enqueue: qdisc_qstats_backlog_inc(sch, skb); - slot->backlog += qdisc_pkt_len(skb); + WRITE_ONCE(slot->backlog, slot->backlog + qdisc_pkt_len(skb)); slot_queue_add(slot, skb); sfq_inc(q, x); if (slot->qlen == 1) { /* The flow is new */ @@ -452,7 +454,7 @@ sfq_enqueue(struct sk_buff *skb, struct Qdisc *sch, struct sk_buff **to_free) */ q->tail = slot; /* We could use a bigger initial quantum for new flows */ - slot->allot = q->quantum; + WRITE_ONCE(slot->allot, q->quantum); } if (++sch->q.qlen <= q->limit) return NET_XMIT_SUCCESS; @@ -489,7 +491,7 @@ sfq_dequeue(struct Qdisc *sch) slot = &q->slots[a]; if (slot->allot <= 0) { q->tail = slot; - slot->allot += q->quantum; + WRITE_ONCE(slot->allot, slot->allot + q->quantum); goto next_slot; } skb = slot_dequeue_head(slot); @@ -497,10 +499,10 @@ sfq_dequeue(struct Qdisc *sch) qdisc_bstats_update(sch, skb); sch->q.qlen--; qdisc_qstats_backlog_dec(sch, skb); - slot->backlog -= qdisc_pkt_len(skb); + WRITE_ONCE(slot->backlog, slot->backlog - qdisc_pkt_len(skb)); /* Is the slot empty? */ if (slot->qlen == 0) { - q->ht[slot->hash] = SFQ_EMPTY_SLOT; + WRITE_ONCE(q->ht[slot->hash], SFQ_EMPTY_SLOT); next_a = slot->next; if (a == next_a) { q->tail = NULL; /* no more active slots */ @@ -508,7 +510,7 @@ sfq_dequeue(struct Qdisc *sch) } q->tail->next = next_a; } else { - slot->allot -= qdisc_pkt_len(skb); + WRITE_ONCE(slot->allot, slot->allot - qdisc_pkt_len(skb)); } return skb; } @@ -549,9 +551,9 @@ static void sfq_rehash(struct Qdisc *sch) sfq_dec(q, i); __skb_queue_tail(&list, skb); } - slot->backlog = 0; + WRITE_ONCE(slot->backlog, 0); red_set_vars(&slot->vars); - q->ht[slot->hash] = SFQ_EMPTY_SLOT; + WRITE_ONCE(q->ht[slot->hash], SFQ_EMPTY_SLOT); } q->tail = NULL; @@ -570,7 +572,7 @@ static void sfq_rehash(struct Qdisc *sch) dropped++; continue; } - q->ht[hash] = x; + WRITE_ONCE(q->ht[hash], x); slot = &q->slots[x]; slot->hash = hash; } @@ -581,7 +583,7 @@ static void sfq_rehash(struct Qdisc *sch) slot->vars.qavg = red_calc_qavg(q->red_parms, &slot->vars, slot->backlog); - slot->backlog += qdisc_pkt_len(skb); + WRITE_ONCE(slot->backlog, slot->backlog + qdisc_pkt_len(skb)); sfq_inc(q, x); if (slot->qlen == 1) { /* The flow is new */ if (q->tail == NULL) { /* It is the first flow */ @@ -591,7 +593,7 @@ static void sfq_rehash(struct Qdisc *sch) q->tail->next = x; } q->tail = slot; - slot->allot = q->quantum; + WRITE_ONCE(slot->allot, q->quantum); } } sch->q.qlen -= dropped; @@ -905,16 +907,16 @@ static int sfq_dump_class_stats(struct Qdisc *sch, unsigned long cl, struct gnet_dump *d) { struct sfq_sched_data *q = qdisc_priv(sch); - sfq_index idx = q->ht[cl - 1]; + sfq_index idx = READ_ONCE(q->ht[cl - 1]); struct gnet_stats_queue qs = { 0 }; struct tc_sfq_xstats xstats = { 0 }; if (idx != SFQ_EMPTY_SLOT) { const struct sfq_slot *slot = &q->slots[idx]; - xstats.allot = slot->allot; - qs.qlen = slot->qlen; - qs.backlog = slot->backlog; + xstats.allot = READ_ONCE(slot->allot); + qs.qlen = READ_ONCE(slot->qlen); + qs.backlog = READ_ONCE(slot->backlog); } if (gnet_stats_copy_queue(d, NULL, &qs, qs.qlen) < 0) return -1; @@ -930,7 +932,7 @@ static void sfq_walk(struct Qdisc *sch, struct qdisc_walker *arg) return; for (i = 0; i < q->divisor; i++) { - if (q->ht[i] == SFQ_EMPTY_SLOT) { + if (READ_ONCE(q->ht[i]) == SFQ_EMPTY_SLOT) { arg->count++; continue; } From c8f7244c8cccaaed4e6c9fe4b8a07e101d0423e5 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 May 2026 15:39:27 +0000 Subject: [PATCH 453/525] tcp: tcp_child_process() related UAF tcp_child_process( .. child ...) currently calls sock_put(child). Unfortunately @child (named @nsk in callers) can be used after this point to send a RST packet. To fix this UAF, I remove the sock_put() from tcp_child_process() and let the callers handle this after it is safe. Remove @rsk variable in tcp_v4_do_rcv() and change tcp_v6_do_rcv() so that both functions look the same. Fixes: cfb6eeb4c860 ("[TCP]: MD5 Signature Option (RFC2385) support.") Reported-by: Damiano Melotti Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20260505153927.3435532-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/ipv4/tcp_ipv4.c | 14 ++++++-------- net/ipv4/tcp_minisocks.c | 2 +- net/ipv6/tcp_ipv6.c | 13 ++++++++----- 3 files changed, 15 insertions(+), 14 deletions(-) diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 8fc24c3743c5..c0526cc03980 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1827,7 +1827,6 @@ INDIRECT_CALLABLE_DECLARE(struct dst_entry *ipv4_dst_check(struct dst_entry *, int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb) { enum skb_drop_reason reason; - struct sock *rsk; reason = psp_sk_rx_policy_check(sk, skb); if (reason) @@ -1863,24 +1862,21 @@ int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb) return 0; if (nsk != sk) { reason = tcp_child_process(sk, nsk, skb); - if (reason) { - rsk = nsk; + sock_put(nsk); + if (reason) goto reset; - } return 0; } } else sock_rps_save_rxhash(sk, skb); reason = tcp_rcv_state_process(sk, skb); - if (reason) { - rsk = sk; + if (reason) goto reset; - } return 0; reset: - tcp_v4_send_reset(rsk, skb, sk_rst_convert_drop_reason(reason)); + tcp_v4_send_reset(sk, skb, sk_rst_convert_drop_reason(reason)); discard: sk_skb_reason_drop(sk, skb, reason); /* Be careful here. If this function gets more complicated and @@ -2193,8 +2189,10 @@ int tcp_v4_rcv(struct sk_buff *skb) rst_reason = sk_rst_convert_drop_reason(drop_reason); tcp_v4_send_reset(nsk, skb, rst_reason); + sock_put(nsk); goto discard_and_relse; } + sock_put(nsk); sock_put(sk); return 0; } diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c index 199f0b579e89..e6092c3ac840 100644 --- a/net/ipv4/tcp_minisocks.c +++ b/net/ipv4/tcp_minisocks.c @@ -1012,6 +1012,6 @@ enum skb_drop_reason tcp_child_process(struct sock *parent, struct sock *child, } bh_unlock_sock(child); - sock_put(child); + return reason; } diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 2c3f7a739709..51583aef0643 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1617,12 +1617,13 @@ int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) if (sk->sk_state == TCP_LISTEN) { struct sock *nsk = tcp_v6_cookie_check(sk, skb); + if (!nsk) + return 0; if (nsk != sk) { - if (nsk) { - reason = tcp_child_process(sk, nsk, skb); - if (reason) - goto reset; - } + reason = tcp_child_process(sk, nsk, skb); + sock_put(nsk); + if (reason) + goto reset; return 0; } } else @@ -1827,8 +1828,10 @@ INDIRECT_CALLABLE_SCOPE int tcp_v6_rcv(struct sk_buff *skb) rst_reason = sk_rst_convert_drop_reason(drop_reason); tcp_v6_send_reset(nsk, skb, rst_reason); + sock_put(nsk); goto discard_and_relse; } + sock_put(nsk); sock_put(sk); return 0; } From b12014d2d36eaed4e4bec5f1ac7e91110eeb100d Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:49 +0200 Subject: [PATCH 454/525] mptcp: pm: kernel: correctly retransmit ADD_ADDR ID 0 When adding the ADD_ADDR to the list, the address including the IP, port and ID are copied. On the other hand, when the endpoint corresponds to the one from the initial subflow, the ID is set to 0, as specified by the MPTCP protocol. The issue is that the ID was reset after having copied the ID in the ADD_ADDR entry. So the retransmission was done, but using a different ID than the initial one. Fixes: 8b8ed1b429f8 ("mptcp: pm: reuse ID 0 after delete and re-add") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-1-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/pm_kernel.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/net/mptcp/pm_kernel.c b/net/mptcp/pm_kernel.c index c9f1e5af3cd3..fc818b63752e 100644 --- a/net/mptcp/pm_kernel.c +++ b/net/mptcp/pm_kernel.c @@ -347,6 +347,8 @@ static void mptcp_pm_create_subflow_or_signal_addr(struct mptcp_sock *msk) /* check first for announce */ if (msk->pm.add_addr_signaled < endp_signal_max) { + u8 endp_id; + /* due to racing events on both ends we can reach here while * previous add address is still running: if we invoke now * mptcp_pm_announce_addr(), that will fail and the @@ -360,19 +362,20 @@ static void mptcp_pm_create_subflow_or_signal_addr(struct mptcp_sock *msk) if (!select_signal_address(pernet, msk, &local)) goto subflow; + /* Special case for ID0: set the correct ID */ + endp_id = local.addr.id; + if (endp_id == msk->mpc_endpoint_id) + local.addr.id = 0; + /* If the alloc fails, we are on memory pressure, not worth * continuing, and trying to create subflows. */ if (!mptcp_pm_alloc_anno_list(msk, &local.addr)) return; - __clear_bit(local.addr.id, msk->pm.id_avail_bitmap); + __clear_bit(endp_id, msk->pm.id_avail_bitmap); msk->pm.add_addr_signaled++; - /* Special case for ID0: set the correct ID */ - if (local.addr.id == msk->mpc_endpoint_id) - local.addr.id = 0; - mptcp_pm_announce_addr(msk, &local.addr, false); mptcp_pm_addr_send_ack(msk); From 03f324f3f1f7619a47b9c91282cb12775ab0a2f1 Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:50 +0200 Subject: [PATCH 455/525] mptcp: pm: ADD_ADDR rtx: allow ID 0 ADD_ADDR can be sent for the ID 0, which corresponds to the local address and port linked to the initial subflow. Indeed, this address could be removed, and re-added later on, e.g. what is done in the "delete re-add signal" MPTCP Join selftests. So no reason to ignore it. Fixes: 00cfd77b9063 ("mptcp: retransmit ADD_ADDR when timeout") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-2-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/pm.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c index 57a456690406..5056eb8db24e 100644 --- a/net/mptcp/pm.c +++ b/net/mptcp/pm.c @@ -337,9 +337,6 @@ static void mptcp_pm_add_timer(struct timer_list *timer) if (inet_sk_state_load(sk) == TCP_CLOSE) return; - if (!entry->addr.id) - return; - if (mptcp_pm_should_add_signal_addr(msk)) { sk_reset_timer(sk, timer, jiffies + TCP_RTO_MAX / 8); goto out; From 5cd6e0ad79d2615264f63929f8b457ad97ae550d Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:51 +0200 Subject: [PATCH 456/525] mptcp: pm: ADD_ADDR rtx: fix potential data-race This mptcp_pm_add_timer() helper is executed as a timer callback in softirq context. To avoid any data races, the socket lock needs to be held with bh_lock_sock(). If the socket is in use, retry again soon after, similar to what is done with the keepalive timer. Fixes: 00cfd77b9063 ("mptcp: retransmit ADD_ADDR when timeout") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-3-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/pm.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c index 5056eb8db24e..3912128d9b86 100644 --- a/net/mptcp/pm.c +++ b/net/mptcp/pm.c @@ -337,6 +337,13 @@ static void mptcp_pm_add_timer(struct timer_list *timer) if (inet_sk_state_load(sk) == TCP_CLOSE) return; + bh_lock_sock(sk); + if (sock_owned_by_user(sk)) { + /* Try again later. */ + sk_reset_timer(sk, timer, jiffies + HZ / 20); + goto out; + } + if (mptcp_pm_should_add_signal_addr(msk)) { sk_reset_timer(sk, timer, jiffies + TCP_RTO_MAX / 8); goto out; @@ -365,6 +372,7 @@ static void mptcp_pm_add_timer(struct timer_list *timer) mptcp_pm_subflow_established(msk); out: + bh_unlock_sock(sk); __sock_put(sk); } From 9634cb35af17019baec21ca648516ce376fa10e6 Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:52 +0200 Subject: [PATCH 457/525] mptcp: pm: ADD_ADDR rtx: always decrease sk refcount When an ADD_ADDR is retransmitted, the sk is held in sk_reset_timer(). It should then be released in all cases at the end. Some (unlikely) checks were returning directly instead of calling sock_put() to decrease the refcount. Jump to a new 'exit' label to call __sock_put() (which will become sock_put() in the next commit) to fix this potential leak. While at it, drop the '!msk' check which cannot happen because it is never reset, and explicitly mark the remaining one as "unlikely". Fixes: 00cfd77b9063 ("mptcp: retransmit ADD_ADDR when timeout") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-4-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/pm.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c index 3912128d9b86..2a01bf1b5bfd 100644 --- a/net/mptcp/pm.c +++ b/net/mptcp/pm.c @@ -331,11 +331,8 @@ static void mptcp_pm_add_timer(struct timer_list *timer) pr_debug("msk=%p\n", msk); - if (!msk) - return; - - if (inet_sk_state_load(sk) == TCP_CLOSE) - return; + if (unlikely(inet_sk_state_load(sk) == TCP_CLOSE)) + goto exit; bh_lock_sock(sk); if (sock_owned_by_user(sk)) { @@ -373,6 +370,7 @@ static void mptcp_pm_add_timer(struct timer_list *timer) out: bh_unlock_sock(sk); +exit: __sock_put(sk); } From b7b9a461569734d33d3259d58d2507adfac107ed Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:53 +0200 Subject: [PATCH 458/525] mptcp: pm: ADD_ADDR rtx: free sk if last When an ADD_ADDR is retransmitted, the sk is held in sk_reset_timer(), and released at the end. If at that moment, it was the last reference being held, the sk would not be freed. sock_put() should then be called instead of __sock_put(). But that's not enough: if it is the last reference, sock_put() will call sk_free(), which will end up calling sk_stop_timer_sync() on the same timer, and waiting indefinitely to finish. So it is needed to mark that the timer is done at the end of the timer handler when it has not been rescheduled, not to call sk_stop_timer_sync() on "itself". Fixes: 00cfd77b9063 ("mptcp: retransmit ADD_ADDR when timeout") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-5-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/pm.c | 30 +++++++++++++++++++----------- 1 file changed, 19 insertions(+), 11 deletions(-) diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c index 2a01bf1b5bfd..8899327e59a1 100644 --- a/net/mptcp/pm.c +++ b/net/mptcp/pm.c @@ -16,6 +16,7 @@ struct mptcp_pm_add_entry { struct list_head list; struct mptcp_addr_info addr; u8 retrans_times; + bool timer_done; struct timer_list add_timer; struct mptcp_sock *sock; struct rcu_head rcu; @@ -327,22 +328,22 @@ static void mptcp_pm_add_timer(struct timer_list *timer) add_timer); struct mptcp_sock *msk = entry->sock; struct sock *sk = (struct sock *)msk; - unsigned int timeout; + unsigned int timeout = 0; pr_debug("msk=%p\n", msk); - if (unlikely(inet_sk_state_load(sk) == TCP_CLOSE)) - goto exit; - bh_lock_sock(sk); + if (unlikely(inet_sk_state_load(sk) == TCP_CLOSE)) + goto out; + if (sock_owned_by_user(sk)) { /* Try again later. */ - sk_reset_timer(sk, timer, jiffies + HZ / 20); + timeout = HZ / 20; goto out; } if (mptcp_pm_should_add_signal_addr(msk)) { - sk_reset_timer(sk, timer, jiffies + TCP_RTO_MAX / 8); + timeout = TCP_RTO_MAX / 8; goto out; } @@ -360,8 +361,9 @@ static void mptcp_pm_add_timer(struct timer_list *timer) } if (entry->retrans_times < ADD_ADDR_RETRANS_MAX) - sk_reset_timer(sk, timer, - jiffies + (timeout << entry->retrans_times)); + timeout <<= entry->retrans_times; + else + timeout = 0; spin_unlock_bh(&msk->pm.lock); @@ -369,9 +371,13 @@ static void mptcp_pm_add_timer(struct timer_list *timer) mptcp_pm_subflow_established(msk); out: + if (timeout) + sk_reset_timer(sk, timer, jiffies + timeout); + else + /* if sock_put calls sk_free: avoid waiting for this timer */ + entry->timer_done = true; bh_unlock_sock(sk); -exit: - __sock_put(sk); + sock_put(sk); } struct mptcp_pm_add_entry * @@ -434,6 +440,7 @@ bool mptcp_pm_alloc_anno_list(struct mptcp_sock *msk, timer_setup(&add_entry->add_timer, mptcp_pm_add_timer, 0); reset_timer: + add_entry->timer_done = false; timeout = mptcp_adjust_add_addr_timeout(msk); if (timeout) sk_reset_timer(sk, &add_entry->add_timer, jiffies + timeout); @@ -454,7 +461,8 @@ static void mptcp_pm_free_anno_list(struct mptcp_sock *msk) spin_unlock_bh(&msk->pm.lock); list_for_each_entry_safe(entry, tmp, &free_list, list) { - sk_stop_timer_sync(sk, &entry->add_timer); + if (!entry->timer_done) + sk_stop_timer_sync(sk, &entry->add_timer); kfree_rcu(entry, rcu); } } From 3cf12492891c4b5ff54dda404a2de4ec54c9e1b5 Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:54 +0200 Subject: [PATCH 459/525] mptcp: pm: ADD_ADDR rtx: resched blocked ADD_ADDR quicker When an ADD_ADDR needs to be retransmitted and another one has already been prepared -- e.g. multiple ADD_ADDRs have been sent in a row and need to be retransmitted later -- this additional retransmission will need to wait. In this case, the timer was reset to TCP_RTO_MAX / 8, which is ~15 seconds. This delay is unnecessary long: it should just be rescheduled at the next opportunity, e.g. after the retransmission timeout. Without this modification, some issues can be seen from time to time in the selftests when multiple ADD_ADDRs are sent, and the host takes time to process them, e.g. the "signal addresses, ADD_ADDR timeout" MPTCP Join selftest, especially with a debug kernel config. Note that on older kernels, 'timeout' is not available. It should be enough to replace it by one second (HZ). Fixes: 00cfd77b9063 ("mptcp: retransmit ADD_ADDR when timeout") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-6-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/pm.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c index 8899327e59a1..29d1bb6a69cf 100644 --- a/net/mptcp/pm.c +++ b/net/mptcp/pm.c @@ -342,13 +342,8 @@ static void mptcp_pm_add_timer(struct timer_list *timer) goto out; } - if (mptcp_pm_should_add_signal_addr(msk)) { - timeout = TCP_RTO_MAX / 8; - goto out; - } - timeout = mptcp_adjust_add_addr_timeout(msk); - if (!timeout) + if (!timeout || mptcp_pm_should_add_signal_addr(msk)) goto out; spin_lock_bh(&msk->pm.lock); From c6d395e2de1306b5fef0344a3c3835fbbfaa18be Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:55 +0200 Subject: [PATCH 460/525] mptcp: pm: ADD_ADDR rtx: skip inactive subflows When looking at the maximum RTO amongst the subflows, inactive subflows were taken into account: that includes stale ones, and the initial one if it has been already been closed. Unusable subflows are now simply skipped. Stale ones are used as an alternative: if there are only stale ones, to take their maximum RTO and avoid to eventually fallback to net.mptcp.add_addr_timeout, which is set to 2 minutes by default. Fixes: 30549eebc4d8 ("mptcp: make ADD_ADDR retransmission timeout adaptive") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-7-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/pm.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c index 29d1bb6a69cf..8a5dba7fe66e 100644 --- a/net/mptcp/pm.c +++ b/net/mptcp/pm.c @@ -306,18 +306,28 @@ static unsigned int mptcp_adjust_add_addr_timeout(struct mptcp_sock *msk) const struct net *net = sock_net((struct sock *)msk); unsigned int rto = mptcp_get_add_addr_timeout(net); struct mptcp_subflow_context *subflow; - unsigned int max = 0; + unsigned int max = 0, max_stale = 0; mptcp_for_each_subflow(msk, subflow) { struct sock *ssk = mptcp_subflow_tcp_sock(subflow); struct inet_connection_sock *icsk = inet_csk(ssk); - if (icsk->icsk_rto > max) + if (!__mptcp_subflow_active(subflow)) + continue; + + if (unlikely(subflow->stale)) { + if (icsk->icsk_rto > max_stale) + max_stale = icsk->icsk_rto; + } else if (icsk->icsk_rto > max) { max = icsk->icsk_rto; + } } - if (max && max < rto) - rto = max; + if (max) + return min(max, rto); + + if (max_stale) + return min(max_stale, rto); return rto; } From 62a9b19dce77e72426f049fb99b9d1d032b9a8ea Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:56 +0200 Subject: [PATCH 461/525] mptcp: pm: ADD_ADDR rtx: return early if no retrans No need to iterate over all subflows if there is no retransmission needed. Exit early in this case then. Fixes: 30549eebc4d8 ("mptcp: make ADD_ADDR retransmission timeout adaptive") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-8-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/pm.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c index 8a5dba7fe66e..4a6e5ab30d80 100644 --- a/net/mptcp/pm.c +++ b/net/mptcp/pm.c @@ -308,6 +308,9 @@ static unsigned int mptcp_adjust_add_addr_timeout(struct mptcp_sock *msk) struct mptcp_subflow_context *subflow; unsigned int max = 0, max_stale = 0; + if (!rto) + return 0; + mptcp_for_each_subflow(msk, subflow) { struct sock *ssk = mptcp_subflow_tcp_sock(subflow); struct inet_connection_sock *icsk = inet_csk(ssk); From 166b78344031bf7ac9f55cb5282776cfd85f220e Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:57 +0200 Subject: [PATCH 462/525] mptcp: pm: prio: skip closed subflows When sending an MP_PRIO, closed subflows need to be skipped. This fixes the case where the initial subflow got closed, re-opened later, then an MP_PRIO is needed for the same local address. Note that explicit MP_PRIO cannot be sent during the 3WHS, so it is fine to use __mptcp_subflow_active(). Fixes: 067065422fcd ("mptcp: add the outgoing MP_PRIO support") Cc: stable@vger.kernel.org Fixes: b29fcfb54cd7 ("mptcp: full disconnect implementation") Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-9-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- net/mptcp/pm.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c index 4a6e5ab30d80..3c152bf66cd5 100644 --- a/net/mptcp/pm.c +++ b/net/mptcp/pm.c @@ -284,6 +284,9 @@ int mptcp_pm_mp_prio_send_ack(struct mptcp_sock *msk, struct sock *ssk = mptcp_subflow_tcp_sock(subflow); struct mptcp_addr_info local, remote; + if (!__mptcp_subflow_active(subflow)) + continue; + mptcp_local_address((struct sock_common *)ssk, &local); if (!mptcp_addresses_equal(&local, addr, addr->port)) continue; From 65db7b27b90e2ea8d4966935aa9a50b6a60c31ac Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:58 +0200 Subject: [PATCH 463/525] selftests: mptcp: check output: catch cmd errors Using '${?}' inside the if-statement to check the returned value from the command that was evaluated as part of the if-statement is not correct: here, '${?}' will be linked to the previous instruction, not the one that is expected here (${cmd}). Instead, simply mark the error, except if an error is expected. If that's the case, 1 can be passed as the 4th argument of this helper. Three checks from pm_netlink.sh expect an error. While at it, improve the error message when the command unexpectedly fails or succeeds. Note that we could expect a specific returned value, but the checks currently expecting an error can be used with 'ip mptcp' or 'pm_nl_ctl', and these two tools don't return the same error code. Fixes: 2d0c1d27ea4e ("selftests: mptcp: add mptcp_lib_check_output helper") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-10-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- tools/testing/selftests/net/mptcp/mptcp_lib.sh | 16 ++++++++++------ tools/testing/selftests/net/mptcp/pm_netlink.sh | 10 ++++++---- 2 files changed, 16 insertions(+), 10 deletions(-) diff --git a/tools/testing/selftests/net/mptcp/mptcp_lib.sh b/tools/testing/selftests/net/mptcp/mptcp_lib.sh index 5fea7e7df628..989a5975dcea 100644 --- a/tools/testing/selftests/net/mptcp/mptcp_lib.sh +++ b/tools/testing/selftests/net/mptcp/mptcp_lib.sh @@ -474,20 +474,24 @@ mptcp_lib_wait_local_port_listen() { wait_local_port_listen "${@}" "tcp" } +# $1: error file, $2: cmd, $3: expected msg, [$4: expected error] mptcp_lib_check_output() { local err="${1}" local cmd="${2}" local expected="${3}" + local exp_error="${4:-0}" local cmd_ret=0 local out - if ! out=$(${cmd} 2>"${err}"); then - cmd_ret=${?} - fi + out=$(${cmd} 2>"${err}") || cmd_ret=1 - if [ ${cmd_ret} -ne 0 ]; then - mptcp_lib_pr_fail "command execution '${cmd}' stderr" - cat "${err}" + if [ "${cmd_ret}" != "${exp_error}" ]; then + mptcp_lib_pr_fail "unexpected returned code for '${cmd}', info:" + if [ "${exp_error}" = 0 ]; then + cat "${err}" + else + echo "${out}" + fi return 2 elif [ "${out}" = "${expected}" ]; then return 0 diff --git a/tools/testing/selftests/net/mptcp/pm_netlink.sh b/tools/testing/selftests/net/mptcp/pm_netlink.sh index 123d9d7a0278..b69f30fcb91e 100755 --- a/tools/testing/selftests/net/mptcp/pm_netlink.sh +++ b/tools/testing/selftests/net/mptcp/pm_netlink.sh @@ -122,10 +122,12 @@ check() local cmd="$1" local expected="$2" local msg="$3" + local exp_error="$4" local rc=0 mptcp_lib_print_title "$msg" - mptcp_lib_check_output "${err}" "${cmd}" "${expected}" || rc=${?} + mptcp_lib_check_output "${err}" "${cmd}" "${expected}" "${exp_error}" || + rc=${?} if [ ${rc} -eq 2 ]; then mptcp_lib_result_fail "${msg} # error ${rc}" ret=${KSFT_FAIL} @@ -158,13 +160,13 @@ check "show_endpoints" \ "3,10.0.1.3,signal backup")" "dump addrs" del_endpoint 2 -check "get_endpoint 2" "" "simple del addr" +check "get_endpoint 2" "" "simple del addr" 1 check "show_endpoints" \ "$(format_endpoints "1,10.0.1.1" \ "3,10.0.1.3,signal backup")" "dump addrs after del" add_endpoint 10.0.1.3 2>/dev/null -check "get_endpoint 4" "" "duplicate addr" +check "get_endpoint 4" "" "duplicate addr" 1 add_endpoint 10.0.1.4 flags signal check "get_endpoint 4" "$(format_endpoints "4,10.0.1.4,signal")" "id addr increment" @@ -173,7 +175,7 @@ for i in $(seq 5 9); do add_endpoint "10.0.1.${i}" flags signal >/dev/null 2>&1 done check "get_endpoint 9" "$(format_endpoints "9,10.0.1.9,signal")" "hard addr limit" -check "get_endpoint 10" "" "above hard addr limit" +check "get_endpoint 10" "" "above hard addr limit" 1 del_endpoint 9 for i in $(seq 10 255); do From 53705ddfa18408f8e1f064331b6387509fa19f7f Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Tue, 5 May 2026 17:00:59 +0200 Subject: [PATCH 464/525] selftests: mptcp: pm: restrict 'unknown' check to pm_nl_ctl When pm_netlink.sh is executed with '-i', 'ip mptcp' is used instead of 'pm_nl_ctl'. IPRoute2 doesn't support the 'unknown' flag, which has only been added to 'pm_nl_ctl' for this specific check: to ensure that the kernel ignores such unsupported flag. No reason to add this flag to 'ip mptcp'. Then, this check should be skipped when 'ip mptcp' is used. Fixes: 0cef6fcac24d ("selftests: mptcp: ip_mptcp option for more scripts") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20260505-net-mptcp-pm-fixes-7-1-rc3-v1-11-fca8091060a4@kernel.org Signed-off-by: Jakub Kicinski --- tools/testing/selftests/net/mptcp/pm_netlink.sh | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/tools/testing/selftests/net/mptcp/pm_netlink.sh b/tools/testing/selftests/net/mptcp/pm_netlink.sh index b69f30fcb91e..04594dfc22b1 100755 --- a/tools/testing/selftests/net/mptcp/pm_netlink.sh +++ b/tools/testing/selftests/net/mptcp/pm_netlink.sh @@ -194,9 +194,13 @@ check "show_endpoints" \ flush_endpoint check "show_endpoints" "" "flush addrs" -add_endpoint 10.0.1.1 flags unknown -check "show_endpoints" "$(format_endpoints "1,10.0.1.1")" "ignore unknown flags" -flush_endpoint +# "unknown" flag is only supported by pm_nl_ctl +if ! mptcp_lib_is_ip_mptcp; then + add_endpoint 10.0.1.1 flags unknown + check "show_endpoints" "$(format_endpoints "1,10.0.1.1")" \ + "ignore unknown flags" + flush_endpoint +fi set_limits 9 1 2>/dev/null check "get_limits" "${default_limits}" "rcv addrs above hard limit" From b266bacba796ff5c4dcd2ae2fc08aacf7ab39153 Mon Sep 17 00:00:00 2001 From: Andreas Haarmann-Thiemann Date: Tue, 5 May 2026 23:52:17 +0200 Subject: [PATCH 465/525] net: ethernet: cortina: Drop half-assembled SKB MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In gmac_rx() (drivers/net/ethernet/cortina/gemini.c), when gmac_get_queue_page() returns NULL for the second page of a multi-page fragment, the driver logs an error and continues — but does not free the partially assembled skb that was being assembled via napi_build_skb() / napi_get_frags(). Free the in-progress partially assembled skb via napi_free_frags() and increase the number of dropped frames appropriately and assign the skb pointer NULL to make sure it is not lingering around, matching the pattern already used elsewhere in the driver. Fixes: 4d5ae32f5e1e ("net: ethernet: Add a driver for Gemini gigabit ethernet") Signed-off-by: Andreas Haarmann-Thiemann Signed-off-by: Linus Walleij Reviewed-by: Alexander Lobakin Link: https://patch.msgid.link/20260505-gemini-ethernet-fix-v2-1-997c31d06079@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/cortina/gemini.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/net/ethernet/cortina/gemini.c b/drivers/net/ethernet/cortina/gemini.c index 4824232f4890..065cbbf52686 100644 --- a/drivers/net/ethernet/cortina/gemini.c +++ b/drivers/net/ethernet/cortina/gemini.c @@ -1491,6 +1491,11 @@ static unsigned int gmac_rx(struct net_device *netdev, unsigned int budget) gpage = gmac_get_queue_page(geth, port, mapping + PAGE_SIZE); if (!gpage) { dev_err(geth->dev, "could not find mapping\n"); + if (skb) { + napi_free_frags(&port->napi); + port->stats.rx_dropped++; + skb = NULL; + } continue; } page = gpage->page; From 5772f6535227ebd104065d80afa8ed3478d34c5c Mon Sep 17 00:00:00 2001 From: David Gow Date: Thu, 16 Apr 2026 14:57:43 +0800 Subject: [PATCH 466/525] x86/boot/e820: Re-enable BIOS fallback if e820 table is empty In commit: 157266edcc56 ("x86/boot/e820: Simplify append_e820_table() and remove restriction on single-entry tables") the check on the number of entries in the e820 table was removed. The intention was to support single-entry maps, but by removing the check entirely, we also skip the fallback (to, e.g., the BIOS 88h function). This means that if no E820 map is passed in from the bootloader (which is the case on some bootloaders, like linld), we end up with an empty memory map, and the kernel fails to boot (either by deadlocking on OOM, or by failing to allocate the real mode trampoline, or similar). Re-instate the check in append_e820_table(), but only check that nr_entries is non-zero. This allows e820__memory_setup_default() to fall back to other memory size sources, and doesn't affect e820__memory_setup_extended(), as the latter ignores the return value from append_e820_table(). In doing so, we also update the return values to be proper error codes, with -ENOENT for this case (there are no entries), and -EINVAL for the case where an entry appears invalid. Given none of the callers check the actual value -- just whether it's nonzero -- this is largely aesthetic in practice. Tested against linld, and the kernel boots again fine. [ mingo: Readability edits to the comment and the changelog. ] Fixes: 157266edcc56 ("x86/boot/e820: Simplify append_e820_table() and remove restriction on single-entry tables") Signed-off-by: David Gow Signed-off-by: Ingo Molnar Reviewed-by: Andy Shevchenko Cc: stable@vger.kernel.org Cc: Arnd Bergmann Cc: "H. Peter Anvin" Link: https://patch.msgid.link/20260416065746.1896647-1-david@davidgow.net --- arch/x86/kernel/e820.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 2a9992758933..eb72537bc0b1 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -450,6 +450,10 @@ __init static int append_e820_table(struct boot_e820_entry *entries, u32 nr_entr { struct boot_e820_entry *entry = entries; + /* If there aren't any entries, we'll want to fall back to another source: */ + if (!nr_entries) + return -ENOENT; + while (nr_entries) { u64 start = entry->addr; u64 size = entry->size; @@ -458,7 +462,7 @@ __init static int append_e820_table(struct boot_e820_entry *entries, u32 nr_entr /* Ignore the remaining entries on 64-bit overflow: */ if (start > end && likely(size)) - return -1; + return -EINVAL; e820__range_add(start, size, type); From b15838b03cd0c6cf35651cfde62d17f14bb1d566 Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Fri, 24 Apr 2026 21:34:28 +0900 Subject: [PATCH 467/525] drm/bochs: Drop manual put on probe error path bochs_pci_probe() allocates the DRM device with devm_drm_dev_alloc(), which registers a devres action to drop the initial DRM device reference on driver detach or probe failure. The error path currently calls drm_dev_put() manually. If probe then returns an error, devres will run the registered release action and put the same device again, after the first put may already have released it. Return the probe error directly and let devres own the final put. Signed-off-by: Myeonghun Pak Fixes: 04826f588682 ("drm/bochs: Allocate DRM device in struct bochs_device") Signed-off-by: Thomas Zimmermann Reviewed-by: Thomas Zimmermann Link: https://patch.msgid.link/20260424123506.32275-1-mhun512@gmail.com --- drivers/gpu/drm/tiny/bochs.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/tiny/bochs.c b/drivers/gpu/drm/tiny/bochs.c index 222e4ae1abbd..5d8dc5efec77 100644 --- a/drivers/gpu/drm/tiny/bochs.c +++ b/drivers/gpu/drm/tiny/bochs.c @@ -761,25 +761,21 @@ static int bochs_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent ret = pcim_enable_device(pdev); if (ret) - goto err_free_dev; + return ret; pci_set_drvdata(pdev, dev); ret = bochs_load(bochs); if (ret) - goto err_free_dev; + return ret; ret = drm_dev_register(dev, 0); if (ret) - goto err_free_dev; + return ret; drm_client_setup(dev, NULL); return ret; - -err_free_dev: - drm_dev_put(dev); - return ret; } static void bochs_pci_remove(struct pci_dev *pdev) From 3051cd060fa496df42954291fa2306ed2eab4ecc Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Wed, 6 May 2026 01:55:11 +0800 Subject: [PATCH 468/525] i2c: smbus: reject oversized block transfers in the common path The SMBus block transfer length data->block[0] is validated in i2c_smbus_xfer_emulated() but that check runs too late for tracepoints and is skipped entirely when the adapter provides a native smbus_xfer implementation. This allows user-controlled oversized block lengths to reach tracepoint memcpy calls and driver callbacks unchecked. Add an early validation in __i2c_smbus_xfer() that rejects block transfers whose caller-supplied length is zero or exceeds I2C_SMBUS_BLOCK_MAX before any tracepoint fires or driver callback runs. data->block[0] is filled in by the device on SMBus block reads, so the check is scoped to operations where the length is actually supplied by the caller. This is consistent with the existing -EINVAL convention in the emulated path and protects all downstream consumers at once: the smbus_write tracepoint, all native smbus_xfer driver implementations, and the emulated path. Two distinct bugs are fixed by this change: Bug 1: smbus_write tracepoint OOB (include/trace/events/smbus.h) trace_smbus_write() fires before any validation and copies data->block[0]+1 bytes into a 34-byte event buffer. With block[0]=0xfe the tracepoint copies 255 bytes, overflowing by 221. BUG: KASAN: stack-out-of-bounds in trace_event_raw_event_smbus_write+0x27c/0x530 Read of size 255 at addr ffff88800d98fcf8 by task poc_smbus/91 Call Trace: __asan_memcpy+0x23/0x80 trace_event_raw_event_smbus_write+0x27c/0x530 __i2c_smbus_xfer+0x43a/0xa40 i2c_smbus_xfer+0x19e/0x340 i2cdev_ioctl_smbus+0x38f/0x7f0 i2cdev_ioctl+0x35e/0x680 __x64_sys_ioctl+0x147/0x1e0 do_syscall_64+0xcf/0x15a0 entry_SYSCALL_64_after_hwframe+0x76/0x7e Bug 2: i2c-stub I2C_SMBUS_I2C_BLOCK_DATA OOB (drivers/i2c/i2c-stub.c) stub_xfer() implements .smbus_xfer directly and only clamps block[0] against 256-command, not I2C_SMBUS_BLOCK_MAX. With block[0]=0xff and command=0 the loop accesses block[1+i] for i up to 254, far past the 34-byte union. UBSAN: array-index-out-of-bounds in drivers/i2c/i2c-stub.c:223:44 index 34 is out of range for type '__u8 [34]' Call Trace: __ubsan_handle_out_of_bounds+0xd7/0x120 stub_xfer+0x1971/0x198f [i2c_stub] __i2c_smbus_xfer+0x306/0xa40 i2c_smbus_xfer+0x19e/0x340 i2cdev_ioctl_smbus+0x38f/0x7f0 i2cdev_ioctl+0x35e/0x680 __x64_sys_ioctl+0x147/0x1e0 do_syscall_64+0xcf/0x15a0 entry_SYSCALL_64_after_hwframe+0x76/0x7e Both traces reproduced on v7.0-rc6+i2c/for-current with KASAN+UBSAN. Fixes: 8a325997d95d ("i2c: Add message transfer tracepoints for SMBUS [ver #2]") Fixes: 4710317891e4 ("i2c-stub: Implement I2C block support") Reported-by: Xiang Mei Signed-off-by: Weiming Shi Signed-off-by: Wolfram Sang --- drivers/i2c/i2c-core-smbus.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/drivers/i2c/i2c-core-smbus.c b/drivers/i2c/i2c-core-smbus.c index 71eb1ef56f0c..ad6acb5ebadc 100644 --- a/drivers/i2c/i2c-core-smbus.c +++ b/drivers/i2c/i2c-core-smbus.c @@ -566,6 +566,18 @@ s32 __i2c_smbus_xfer(struct i2c_adapter *adapter, u16 addr, if (res) return res; + /* Reject invalid caller-supplied block lengths before any + * tracepoint or native smbus_xfer callback runs. + */ + if (data && + (protocol == I2C_SMBUS_I2C_BLOCK_DATA || + protocol == I2C_SMBUS_BLOCK_PROC_CALL || + (protocol == I2C_SMBUS_BLOCK_DATA && + read_write == I2C_SMBUS_WRITE)) && + (data->block[0] == 0 || + data->block[0] > I2C_SMBUS_BLOCK_MAX)) + return -EINVAL; + /* If enabled, the following two tracepoints are conditional on * read_write and protocol. */ From 593dfd40a94ca0ab20297ea4629d94268deed0ed Mon Sep 17 00:00:00 2001 From: Bobby Eshleman Date: Mon, 4 May 2026 18:42:11 -0700 Subject: [PATCH 469/525] eth: fbnic: fix double-free of PCS on phylink creation failure fbnic_phylink_create() stores the newly allocated PCS in fbn->pcs and then calls phylink_create(). When phylink_create() fails, the error path correctly destroys the PCS via xpcs_destroy_pcs(), but the caller, fbnic_netdev_alloc(), responds by invoking fbnic_netdev_free() which calls fbnic_phylink_destroy(). That function finds fbn->pcs non-NULL and calls xpcs_destroy_pcs() a second time on the already-freed object, triggering a refcount underflow use-after-free: [ 1.934973] fbnic 0000:01:00.0: Failed to create Phylink interface, err: -22 [ 1.935103] ------------[ cut here ]------------ [ 1.935179] refcount_t: underflow; use-after-free. [ 1.935252] WARNING: lib/refcount.c:28 at refcount_warn_saturate+0x59/0x90, CPU#0: swapper/0/1 [ 1.935389] Modules linked in: [ 1.935484] CPU: 0 UID: 0 PID: 1 Comm: swapper/0 Not tainted 7.0.0-virtme-04244-g1f5ffc672165-dirty #1 PREEMPT(lazy) [ 1.935661] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014 [ 1.935826] RIP: 0010:refcount_warn_saturate+0x59/0x90 [ 1.935931] Code: 44 48 8d 3d 49 f9 a7 01 67 48 0f b9 3a e9 bf 1e 96 00 48 8d 3d 48 f9 a7 01 67 48 0f b9 3a c3 cc cc cc cc 48 8d 3d 47 f9 a7 01 <67> 48 0f b9 3a c3 cc cc cc cc 48 8d 3d 46 f9 a7 01 67 48 0f b9 3a [ 1.936274] RSP: 0000:ffffd0d440013c58 EFLAGS: 00010246 [ 1.936376] RAX: 0000000000000000 RBX: ffff8f39c188c278 RCX: 000000000000002b [ 1.936524] RDX: ffff8f39c004f000 RSI: 0000000000000003 RDI: ffffffff96abab00 [ 1.936692] RBP: ffff8f39c188c240 R08: ffffffff96988e88 R09: 00000000ffffdfff [ 1.936835] R10: ffffffff96878ea0 R11: 0000000000000187 R12: 0000000000000000 [ 1.936970] R13: ffff8f39c0cef0c8 R14: ffff8f39c1ac01c0 R15: 0000000000000000 [ 1.937114] FS: 0000000000000000(0000) GS:ffff8f3ba08b4000(0000) knlGS:0000000000000000 [ 1.937273] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 1.937382] CR2: ffff8f3b3ffff000 CR3: 0000000172642001 CR4: 0000000000372ef0 [ 1.937540] Call Trace: [ 1.937619] [ 1.937698] xpcs_destroy_pcs+0x25/0x40 [ 1.937783] fbnic_netdev_alloc+0x1e5/0x200 [ 1.937859] fbnic_probe+0x230/0x370 [ 1.937939] local_pci_probe+0x3e/0x90 [ 1.938013] pci_device_probe+0xbb/0x1e0 [ 1.938091] ? sysfs_do_create_link_sd+0x6d/0xe0 [ 1.938188] really_probe+0xc1/0x2b0 [ 1.938282] __driver_probe_device+0x73/0x120 [ 1.938371] driver_probe_device+0x1e/0xe0 [ 1.938466] __driver_attach+0x8d/0x190 [ 1.938560] ? __pfx___driver_attach+0x10/0x10 [ 1.938663] bus_for_each_dev+0x7b/0xd0 [ 1.938758] bus_add_driver+0xe8/0x210 [ 1.938854] driver_register+0x60/0x120 [ 1.938929] ? __pfx_fbnic_init_module+0x10/0x10 [ 1.939026] fbnic_init_module+0x25/0x60 [ 1.939109] do_one_initcall+0x49/0x220 [ 1.939202] ? rdinit_setup+0x20/0x40 [ 1.939304] kernel_init_freeable+0x1b0/0x310 [ 1.939449] ? __pfx_kernel_init+0x10/0x10 [ 1.939560] kernel_init+0x1a/0x1c0 [ 1.939640] ret_from_fork+0x1ed/0x240 [ 1.939730] ? __pfx_kernel_init+0x10/0x10 [ 1.939805] ret_from_fork_asm+0x1a/0x30 [ 1.939886] [ 1.939927] ---[ end trace 0000000000000000 ]--- [ 1.940184] fbnic 0000:01:00.0: Netdev allocation failed Instead of calling fbnic_phylink_destroy(), the prior initialization of netdev should just be unrolled with free_netdev() and clearing fbd->netdev. Clearing fbd->netdev to NULL avoids UAF in init_failure_mode where callers guard by checking !fbd->netdev, such as fbnic_mdio_read_pmd(). These callers remain active even after a failed probe, so fdb->netdev still needs to be cleared. Fixes: d0fe7104c795 ("fbnic: Replace use of internal PCS w/ Designware XPCS") Signed-off-by: Bobby Eshleman Link: https://patch.msgid.link/20260504-fbnic-pcs-fix-v2-1-de45192821d9@meta.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/meta/fbnic/fbnic_netdev.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_netdev.c b/drivers/net/ethernet/meta/fbnic/fbnic_netdev.c index c406a3b56b37..4dea2bb58d2f 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_netdev.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_netdev.c @@ -826,7 +826,8 @@ struct net_device *fbnic_netdev_alloc(struct fbnic_dev *fbd) netif_tx_stop_all_queues(netdev); if (fbnic_phylink_create(netdev)) { - fbnic_netdev_free(fbd); + free_netdev(netdev); + fbd->netdev = NULL; return NULL; } From 08f566e8f83bb70f04ad5aba5be352c490a01c8a Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Tue, 5 May 2026 15:21:53 +0200 Subject: [PATCH 470/525] veth: fix OOB txq access in veth_poll() with asymmetric queue counts XDP redirect into a veth device (via bpf_redirect()) calls veth_xdp_xmit(), which enqueues frames into the peer's ptr_ring using smp_processor_id() % peer->real_num_rx_queues as the ring index. With an asymmetric veth pair where the peer has fewer TX queues than RX queues, that index can exceed peer->real_num_tx_queues. veth_poll() then resolves peer_txq for the ring via: peer_txq = peer_dev ? netdev_get_tx_queue(peer_dev, queue_idx) : NULL; where queue_idx = rq->xdp_rxq.queue_index. When queue_idx exceeds peer_dev->real_num_tx_queues this is an out-of-bounds (OOB) access into the peer's netdev_queue array, triggering DEBUG_NET_WARN_ON_ONCE in netdev_get_tx_queue(). The normal ndo_start_xmit path is not affected: the stack clamps skb->queue_mapping via netdev_cap_txqueue() before invoking ndo_start_xmit, so rxq in veth_xmit() never exceeds real_num_tx_queues. Fix veth_poll() by clamping: only dereference peer_txq when queue_idx is within bounds, otherwise set it to NULL. The out-of-range rings are fed exclusively via XDP redirect (veth_xdp_xmit), never via ndo_start_xmit (veth_xmit), so the peer txq was never stopped and there is nothing to wake; NULL is the correct fallback. Reported-by: Sashiko Closes: https://lore.kernel.org/all/20260502071828.616C3C19425@smtp.kernel.org/ Fixes: dc82a33297fc ("veth: apply qdisc backpressure on full ptr_ring to reduce TX drops") Signed-off-by: Jesper Dangaard Brouer Link: https://patch.msgid.link/20260505132159.241305-2-hawk@kernel.org Signed-off-by: Paolo Abeni --- drivers/net/veth.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/veth.c b/drivers/net/veth.c index e35df717e65e..0cfb19b760dd 100644 --- a/drivers/net/veth.c +++ b/drivers/net/veth.c @@ -972,7 +972,8 @@ static int veth_poll(struct napi_struct *napi, int budget) /* NAPI functions as RCU section */ peer_dev = rcu_dereference_check(priv->peer, rcu_read_lock_bh_held()); - peer_txq = peer_dev ? netdev_get_tx_queue(peer_dev, queue_idx) : NULL; + peer_txq = (peer_dev && queue_idx < peer_dev->real_num_tx_queues) ? + netdev_get_tx_queue(peer_dev, queue_idx) : NULL; xdp_set_return_frame_no_direct(); done = veth_xdp_rcv(rq, budget, &bq, &stats); From d119775f2bad827edc28071c061fdd4a91f889a5 Mon Sep 17 00:00:00 2001 From: Jiexun Wang Date: Wed, 6 May 2026 22:08:23 +0800 Subject: [PATCH 471/525] af_unix: Reject SIOCATMARK on non-stream sockets SIOCATMARK reports whether the receive queue is at the urgent mark for MSG_OOB. In AF_UNIX, MSG_OOB is supported only for SOCK_STREAM sockets. SOCK_DGRAM and SOCK_SEQPACKET reject MSG_OOB in sendmsg() and recvmsg(), so they should not support SIOCATMARK either. Return -EOPNOTSUPP for non-stream sockets before checking the receive queue. Fixes: 314001f0bf92 ("af_unix: Add OOB support") Cc: stable@kernel.org Reported-by: Yuan Tan Reported-by: Yifan Wu Reported-by: Juefei Pu Reported-by: Xin Liu Suggested-by: Kuniyuki Iwashima Signed-off-by: Jiexun Wang Signed-off-by: Ren Wei Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20260506140825.2987635-1-n05ec@lzu.edu.cn Signed-off-by: Jakub Kicinski --- net/unix/af_unix.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index e2d787ca3e74..1cbf36ea043b 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -3323,6 +3323,9 @@ static int unix_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg) struct sk_buff *skb; int answ = 0; + if (sk->sk_type != SOCK_STREAM) + return -EOPNOTSUPP; + mutex_lock(&u->iolock); skb = skb_peek(&sk->sk_receive_queue); From 9032f7676935a13fd402608223d326c5f62da9c0 Mon Sep 17 00:00:00 2001 From: "D. Wythe" Date: Wed, 6 May 2026 09:41:05 +0800 Subject: [PATCH 472/525] net/smc: fix missing sk_err when TCP handshake fails In smc_connect_work(), when the underlying TCP handshake fails, the error code (rc) must be propagated to sk_err to ensure userspace can correctly retrieve the error status via SO_ERROR. Currently, the code only handles a restricted set of error codes (e.g., EPIPE, ECONNREFUSED). If other errors occurs, such as EHOSTUNREACH, sk_err remains unset (zero). This affects applications that rely on SO_ERROR to determine connect outcome. For example, higher versions of Go's netpoller treats SO_ERROR == 0 combined with a failed getpeername() as a spurious wakeup and re-enters epoll_wait(). Under ET mode, no further edge will be generated since the socket is already in a terminal state, causing the connect to hang indefinitely or until a user-specified timeout, if one is set. Fixes: 50717a37db03 ("net/smc: nonblocking connect rework") Signed-off-by: D. Wythe Reviewed-by: Dust Li Link: https://patch.msgid.link/20260506014105.27093-1-alibuda@linux.alibaba.com Signed-off-by: Jakub Kicinski --- net/smc/af_smc.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index 1a565095376a..185dbed7de5d 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -1628,12 +1628,8 @@ static void smc_connect_work(struct work_struct *work) lock_sock(&smc->sk); if (rc != 0 || smc->sk.sk_err) { smc->sk.sk_state = SMC_CLOSED; - if (rc == -EPIPE || rc == -EAGAIN) - smc->sk.sk_err = EPIPE; - else if (rc == -ECONNREFUSED) - smc->sk.sk_err = ECONNREFUSED; - else if (signal_pending(current)) - smc->sk.sk_err = -sock_intr_errno(timeo); + if (!smc->sk.sk_err) + smc->sk.sk_err = (rc == -EAGAIN) ? EPIPE : -rc; sock_put(&smc->sk); /* passive closing */ goto out; } From 32cd651d14fc72a93703ea2384cb5cd8998523a8 Mon Sep 17 00:00:00 2001 From: Justin Chen Date: Tue, 5 May 2026 10:39:26 -0700 Subject: [PATCH 473/525] net: phy: broadcom: Save PHY counters during suspend The PHY counters can be lost if the PHY is reset during suspend. We need to save the values into the shadow counters or the accounting will be incorrect over multiple suspend and resume cycles. Fixes: 820ee17b8d3b ("net: phy: broadcom: Add support code for reading PHY counters") Signed-off-by: Justin Chen Reviewed-by: Florian Fainelli Link: https://patch.msgid.link/20260505173926.2870069-1-justin.chen@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/phy/bcm-phy-lib.c | 9 +++++++++ drivers/net/phy/bcm-phy-lib.h | 1 + drivers/net/phy/bcm7xxx.c | 14 ++++++++++++++ drivers/net/phy/broadcom.c | 5 +++++ 4 files changed, 29 insertions(+) diff --git a/drivers/net/phy/bcm-phy-lib.c b/drivers/net/phy/bcm-phy-lib.c index 5198d66dbbc0..b64beade8dd9 100644 --- a/drivers/net/phy/bcm-phy-lib.c +++ b/drivers/net/phy/bcm-phy-lib.c @@ -563,6 +563,15 @@ void bcm_phy_get_stats(struct phy_device *phydev, u64 *shadow, } EXPORT_SYMBOL_GPL(bcm_phy_get_stats); +void bcm_phy_update_stats_shadow(struct phy_device *phydev, u64 *shadow) +{ + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(bcm_phy_hw_stats); i++) + bcm_phy_get_stat(phydev, shadow, i); +} +EXPORT_SYMBOL_GPL(bcm_phy_update_stats_shadow); + void bcm_phy_r_rc_cal_reset(struct phy_device *phydev) { /* Reset R_CAL/RC_CAL Engine */ diff --git a/drivers/net/phy/bcm-phy-lib.h b/drivers/net/phy/bcm-phy-lib.h index bceddbc860eb..bba94ce96195 100644 --- a/drivers/net/phy/bcm-phy-lib.h +++ b/drivers/net/phy/bcm-phy-lib.h @@ -85,6 +85,7 @@ int bcm_phy_get_sset_count(struct phy_device *phydev); void bcm_phy_get_strings(struct phy_device *phydev, u8 *data); void bcm_phy_get_stats(struct phy_device *phydev, u64 *shadow, struct ethtool_stats *stats, u64 *data); +void bcm_phy_update_stats_shadow(struct phy_device *phydev, u64 *shadow); void bcm_phy_r_rc_cal_reset(struct phy_device *phydev); int bcm_phy_28nm_a0b0_afe_config_init(struct phy_device *phydev); int bcm_phy_enable_jumbo(struct phy_device *phydev); diff --git a/drivers/net/phy/bcm7xxx.c b/drivers/net/phy/bcm7xxx.c index 00e8fa14aa77..71a163f62c0e 100644 --- a/drivers/net/phy/bcm7xxx.c +++ b/drivers/net/phy/bcm7xxx.c @@ -807,6 +807,17 @@ static void bcm7xxx_28nm_get_phy_stats(struct phy_device *phydev, bcm_phy_get_stats(phydev, priv->stats, stats, data); } +static int bcm7xxx_28nm_suspend(struct phy_device *phydev) +{ + struct bcm7xxx_phy_priv *priv = phydev->priv; + + mutex_lock(&phydev->lock); + bcm_phy_update_stats_shadow(phydev, priv->stats); + mutex_unlock(&phydev->lock); + + return genphy_suspend(phydev); +} + static int bcm7xxx_28nm_probe(struct phy_device *phydev) { struct bcm7xxx_phy_priv *priv; @@ -849,6 +860,7 @@ static int bcm7xxx_28nm_probe(struct phy_device *phydev) .flags = PHY_IS_INTERNAL, \ .config_init = bcm7xxx_28nm_config_init, \ .resume = bcm7xxx_28nm_resume, \ + .suspend = bcm7xxx_28nm_suspend, \ .get_tunable = bcm7xxx_28nm_get_tunable, \ .set_tunable = bcm7xxx_28nm_set_tunable, \ .get_sset_count = bcm_phy_get_sset_count, \ @@ -866,6 +878,7 @@ static int bcm7xxx_28nm_probe(struct phy_device *phydev) .flags = PHY_IS_INTERNAL, \ .config_init = bcm7xxx_28nm_ephy_config_init, \ .resume = bcm7xxx_28nm_ephy_resume, \ + .suspend = bcm7xxx_28nm_suspend, \ .get_sset_count = bcm_phy_get_sset_count, \ .get_strings = bcm_phy_get_strings, \ .get_stats = bcm7xxx_28nm_get_phy_stats, \ @@ -902,6 +915,7 @@ static int bcm7xxx_28nm_probe(struct phy_device *phydev) .config_aneg = genphy_config_aneg, \ .read_status = genphy_read_status, \ .resume = bcm7xxx_16nm_ephy_resume, \ + .suspend = bcm7xxx_28nm_suspend, \ } static struct phy_driver bcm7xxx_driver[] = { diff --git a/drivers/net/phy/broadcom.c b/drivers/net/phy/broadcom.c index bf0c6a04481e..d1a4edb34ad2 100644 --- a/drivers/net/phy/broadcom.c +++ b/drivers/net/phy/broadcom.c @@ -592,8 +592,13 @@ static int bcm54xx_set_wakeup_irq(struct phy_device *phydev, bool state) static int bcm54xx_suspend(struct phy_device *phydev) { + struct bcm54xx_phy_priv *priv = phydev->priv; int ret = 0; + mutex_lock(&phydev->lock); + bcm_phy_update_stats_shadow(phydev, priv->stats); + mutex_unlock(&phydev->lock); + bcm54xx_ptp_stop(phydev); /* Acknowledge any Wake-on-LAN interrupt prior to suspend */ From 019c892e46544af0ae94ec833f79aa903c837666 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Wed, 6 May 2026 06:59:53 +0000 Subject: [PATCH 474/525] ipmr: Call ipmr_fib_lookup() under RCU. Yi Lai reported RCU splat in reg_vif_xmit() below. [0] When CONFIG_IP_MROUTE_MULTIPLE_TABLES=n, ipmr_fib_lookup() uses rcu_dereference() without explicit rcu_read_lock(). Although rcu_read_lock_bh() is already held by the caller __dev_queue_xmit(), lockdep requires explicit rcu_read_lock() for rcu_dereference(). Let's move up rcu_read_lock() in reg_vif_xmit() to cover ipmr_fib_lookup(). [0]: WARNING: suspicious RCU usage 7.1.0-rc2-next-20260504-9d0d467c3572 #1 Not tainted ----------------------------- net/ipv4/ipmr.c:329 suspicious rcu_dereference_check() usage! other info that might help us debug this: rcu_scheduler_active = 2, debug_locks = 1 2 locks held by syz.2.17/1779: #0: ffffffff87896440 (rcu_read_lock_bh){....}-{1:3}, at: local_bh_disable include/linux/bottom_half.h:20 [inline] #0: ffffffff87896440 (rcu_read_lock_bh){....}-{1:3}, at: rcu_read_lock_bh include/linux/rcupdate.h:891 [inline] #0: ffffffff87896440 (rcu_read_lock_bh){....}-{1:3}, at: __dev_queue_xmit+0x239/0x4140 net/core/dev.c:4792 #1: ffff88801a199d18 (_xmit_PIMREG#2){+...}-{3:3}, at: spin_lock include/linux/spinlock.h:342 [inline] #1: ffff88801a199d18 (_xmit_PIMREG#2){+...}-{3:3}, at: __netif_tx_lock include/linux/netdevice.h:4795 [inline] #1: ffff88801a199d18 (_xmit_PIMREG#2){+...}-{3:3}, at: __dev_queue_xmit+0x1d5d/0x4140 net/core/dev.c:4865 stack backtrace: CPU: 1 UID: 0 PID: 1779 Comm: syz.2.17 Not tainted 7.1.0-rc2-next-20260504-9d0d467c3572 #1 PREEMPT(lazy) Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014 Call Trace: __dump_stack lib/dump_stack.c:94 [inline] dump_stack_lvl+0x121/0x150 lib/dump_stack.c:120 dump_stack+0x19/0x20 lib/dump_stack.c:129 lockdep_rcu_suspicious+0x15b/0x1f0 kernel/locking/lockdep.c:6878 ipmr_fib_lookup net/ipv4/ipmr.c:329 [inline] reg_vif_xmit+0x2ee/0x3c0 net/ipv4/ipmr.c:540 __netdev_start_xmit include/linux/netdevice.h:5382 [inline] netdev_start_xmit include/linux/netdevice.h:5391 [inline] xmit_one net/core/dev.c:3889 [inline] dev_hard_start_xmit+0x170/0x700 net/core/dev.c:3905 __dev_queue_xmit+0x1df1/0x4140 net/core/dev.c:4871 dev_queue_xmit include/linux/netdevice.h:3423 [inline] packet_xmit+0x252/0x370 net/packet/af_packet.c:276 packet_snd net/packet/af_packet.c:3082 [inline] packet_sendmsg+0x39ad/0x5650 net/packet/af_packet.c:3114 sock_sendmsg_nosec net/socket.c:797 [inline] __sock_sendmsg net/socket.c:812 [inline] ____sys_sendmsg+0xa21/0xba0 net/socket.c:2716 ___sys_sendmsg+0x121/0x1c0 net/socket.c:2770 __sys_sendmsg+0x177/0x220 net/socket.c:2802 __do_sys_sendmsg net/socket.c:2807 [inline] __se_sys_sendmsg net/socket.c:2805 [inline] __x64_sys_sendmsg+0x80/0xc0 net/socket.c:2805 x64_sys_call+0x1d9c/0x21c0 arch/x86/include/generated/asm/syscalls_64.h:47 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0xc1/0x1020 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7f37e563ee5d Code: ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 93 af 1b 00 f7 d8 64 89 01 48 RSP: 002b:00007ffe5caa7fa8 EFLAGS: 00000246 ORIG_RAX: 000000000000002e RAX: ffffffffffffffda RBX: 00000000005c5fa0 RCX: 00007f37e563ee5d RDX: 0000000000000000 RSI: 00002000000012c0 RDI: 0000000000000004 RBP: 00000000005c5fa0 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000 R13: 0000000000000000 R14: 00000000005c5fac R15: 00000000005c5fa0 Fixes: b3b6babf4751 ("ipmr: Free mr_table after RCU grace period.") Reported-by: syzkaller Reported-by: Yi Lai Closes: https://lore.kernel.org/netdev/afrY34dLXNUboevf@ly-workstation/ Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260506065955.1695753-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- net/ipv4/ipmr.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c index 05fb6eefe0be..2628cd3a93a6 100644 --- a/net/ipv4/ipmr.c +++ b/net/ipv4/ipmr.c @@ -537,15 +537,16 @@ static netdev_tx_t reg_vif_xmit(struct sk_buff *skb, struct net_device *dev) }; int err; + rcu_read_lock(); err = ipmr_fib_lookup(net, &fl4, &mrt); if (err < 0) { + rcu_read_unlock(); kfree_skb(skb); return err; } DEV_STATS_ADD(dev, tx_bytes, skb->len); DEV_STATS_INC(dev, tx_packets); - rcu_read_lock(); /* Pairs with WRITE_ONCE() in vif_add() and vif_delete() */ ipmr_cache_report(mrt, skb, READ_ONCE(mrt->mroute_reg_vif_num), From ecddc523cfdb85b3e132f13e293224ebfdfab564 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Wed, 6 May 2026 07:04:42 +0000 Subject: [PATCH 475/525] tcp: Fix dst leak in tcp_v6_connect(). If a socket is bound to a wildcard address, tcp_v[46]_connect() updates it with a non-wildcard address based on the route lookup. After bhash2 was introduced in the cited commit, we must call inet_bhash2_update_saddr() to update the bhash2 entry as well. If inet_bhash2_update_saddr() fails, we must release the refcount for dst by ip_route_connect() or ip6_dst_lookup_flow(). While tcp_v4_connect() calls ip_rt_put() in the error path, tcp_v6_connect() does not call dst_release(). Let's call dst_release() when inet_bhash2_update_saddr() fails in tcp_v6_connect(). Fixes: 28044fc1d495 ("net: Add a bhash2 table hashed by port and address") Reported-by: Damiano Melotti Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20260506070443.1699879-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- net/ipv6/tcp_ipv6.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 51583aef0643..d13d49bfef19 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -288,8 +288,10 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr_unsized *uaddr, saddr = &fl6->saddr; err = inet_bhash2_update_saddr(sk, saddr, AF_INET6); - if (err) + if (err) { + dst_release(dst); goto failure; + } } /* set the source address */ From dedf6c90386d99b878763c183a08b61d3ce4824e Mon Sep 17 00:00:00 2001 From: Joey Lu Date: Wed, 6 May 2026 16:46:13 +0800 Subject: [PATCH 476/525] net: stmmac: dwmac-nuvoton: fix NULL pointer dereference in nvt_set_phy_intf_sel() priv->dev was never initialized after devm_kzalloc() allocates the private data structure. When nvt_set_phy_intf_sel() is later invoked via the phylink interface_select callback, it calls nvt_gmac_get_delay(priv->dev, ...) which dereferences the NULL pointer. Fix this by assigning priv->dev = dev immediately after allocation. Fixes: 4d7c557f58ef ("net: stmmac: dwmac-nuvoton: Add dwmac glue for Nuvoton MA35 family") Signed-off-by: Joey Lu Link: https://patch.msgid.link/20260506084614.192894-2-a0987203069@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/dwmac-nuvoton.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-nuvoton.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-nuvoton.c index e2240b68ad98..2ab6ecac6422 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-nuvoton.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-nuvoton.c @@ -100,6 +100,8 @@ static int nvt_gmac_probe(struct platform_device *pdev) if (!priv) return dev_err_probe(dev, -ENOMEM, "Failed to allocate private data\n"); + priv->dev = dev; + priv->regmap = syscon_regmap_lookup_by_phandle_args(dev->of_node, "nuvoton,sys", 1, &priv->macid); if (IS_ERR(priv->regmap)) From b131dc93f7bf1b1461f5bde0c06c4c2384aa5b58 Mon Sep 17 00:00:00 2001 From: Daniel Machon Date: Wed, 6 May 2026 09:25:38 +0200 Subject: [PATCH 477/525] net: sparx5: fix wrong chip ids for TSN SKUs The TSN SKUs in enum spx5_target_chiptype have incorrect IDs: SPX5_TARGET_CT_7546TSN = 0x47546, SPX5_TARGET_CT_7549TSN = 0x47549, SPX5_TARGET_CT_7552TSN = 0x47552, SPX5_TARGET_CT_7556TSN = 0x47556, SPX5_TARGET_CT_7558TSN = 0x47558, The value read back from the chip is GCB_CHIP_ID_PART_ID, which is a GENMASK(27, 12) field, i.e. at most 16 bits wide. It can never match these IDs, so probing a TSN part fails with a "Target not supported" error. Fix the enum to use the actual 16-bit part IDs returned by the hardware: 0x0546, 0x0549, 0x0552, 0x0556 and 0x0558. Reported-by: Andrew Lunn Fixes: 3cfa11bac9bb ("net: sparx5: add the basic sparx5 driver") Signed-off-by: Daniel Machon Link: https://patch.msgid.link/20260506-misc-fixes-sparx5-lan969x-v2-3-fb236aa96908@microchip.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/microchip/sparx5/sparx5_main.h | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_main.h b/drivers/net/ethernet/microchip/sparx5/sparx5_main.h index 6a745bb71b5c..eb57b86fbe22 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_main.h +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_main.h @@ -31,11 +31,11 @@ enum spx5_target_chiptype { SPX5_TARGET_CT_7552 = 0x7552, /* SparX-5-128 Enterprise */ SPX5_TARGET_CT_7556 = 0x7556, /* SparX-5-160 Enterprise */ SPX5_TARGET_CT_7558 = 0x7558, /* SparX-5-200 Enterprise */ - SPX5_TARGET_CT_7546TSN = 0x47546, /* SparX-5-64i Industrial */ - SPX5_TARGET_CT_7549TSN = 0x47549, /* SparX-5-90i Industrial */ - SPX5_TARGET_CT_7552TSN = 0x47552, /* SparX-5-128i Industrial */ - SPX5_TARGET_CT_7556TSN = 0x47556, /* SparX-5-160i Industrial */ - SPX5_TARGET_CT_7558TSN = 0x47558, /* SparX-5-200i Industrial */ + SPX5_TARGET_CT_7546TSN = 0x0546, /* SparX-5-64i Industrial */ + SPX5_TARGET_CT_7549TSN = 0x0549, /* SparX-5-90i Industrial */ + SPX5_TARGET_CT_7552TSN = 0x0552, /* SparX-5-128i Industrial */ + SPX5_TARGET_CT_7556TSN = 0x0556, /* SparX-5-160i Industrial */ + SPX5_TARGET_CT_7558TSN = 0x0558, /* SparX-5-200i Industrial */ SPX5_TARGET_CT_LAN9694 = 0x9694, /* lan969x-40 */ SPX5_TARGET_CT_LAN9691VAO = 0x9691, /* lan969x-40-VAO */ SPX5_TARGET_CT_LAN9694TSN = 0x9695, /* lan969x-40-TSN */ From 41ae14071cd7f6a7770e2fe1f8a0859d4c2c6ba4 Mon Sep 17 00:00:00 2001 From: Daniel Machon Date: Wed, 6 May 2026 09:25:39 +0200 Subject: [PATCH 478/525] net: sparx5: configure serdes for 1000BASE-X in sparx5_port_init() sparx5_port_init() only invokes sparx5_serdes_set() and the associated shadow-device enable and low-speed device switch for SGMII and QSGMII. On any port with a high-speed primary device (DEV5G/DEV10G/DEV25G) configured for 1000BASE-X the serdes is therefore left uninitialized, the DEV2G5 shadow is never enabled, and the port stays pointed at its high-speed device rather than the DEV2G5. The PCS1G block looks healthy in isolation, but no frames reach the link partner. Add 1000BASE-X to the check so the same three steps run. Note: the same issue might apply to 2500BASE-X, but that will, eventually, be addressed in a separate commit. Reported-by: Andrew Lunn Fixes: 946e7fd5053a ("net: sparx5: add port module support") Signed-off-by: Daniel Machon Link: https://patch.msgid.link/20260506-misc-fixes-sparx5-lan969x-v2-4-fb236aa96908@microchip.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/microchip/sparx5/sparx5_port.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_port.c b/drivers/net/ethernet/microchip/sparx5/sparx5_port.c index 04bc8fffaf96..62c49893de3c 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_port.c +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_port.c @@ -1128,7 +1128,8 @@ int sparx5_port_init(struct sparx5 *sparx5, DEV2G5_PCS1G_SD_CFG(port->portno)); if (conf->portmode == PHY_INTERFACE_MODE_QSGMII || - conf->portmode == PHY_INTERFACE_MODE_SGMII) { + conf->portmode == PHY_INTERFACE_MODE_SGMII || + conf->portmode == PHY_INTERFACE_MODE_1000BASEX) { err = sparx5_serdes_set(sparx5, port, conf); if (err) return err; From 5be7a0cef3229fb3b63a07c0d289daf752545424 Mon Sep 17 00:00:00 2001 From: Piyush Sachdeva Date: Thu, 7 May 2026 22:22:13 +0530 Subject: [PATCH 479/525] smb: client: Use FullSessionKey for AES-256 encryption key derivation When Kerberos authentication is used with AES-256 encryption (AES-256-CCM or AES-256-GCM), the SMB3 encryption and decryption keys must be derived using the full session key (Session.FullSessionKey) rather than just the first 16 bytes (Session.SessionKey). Per MS-SMB2 section 3.2.5.3.1, when Connection.Dialect is "3.1.1" and Connection.CipherId is AES-256-CCM or AES-256-GCM, Session.FullSessionKey must be set to the full cryptographic key from the GSS authentication context. The encryption and decryption key derivation (SMBC2SCipherKey, SMBS2CCipherKey) must use this FullSessionKey as the KDF input. The signing key derivation continues to use Session.SessionKey (first 16 bytes) in all cases. Previously, generate_key() hardcoded SMB2_NTLMV2_SESSKEY_SIZE (16) as the HMAC-SHA256 key input length for all derivations. When Kerberos with AES-256 provides a 32-byte session key, the KDF for encryption/decryption was using only the first 16 bytes, producing keys that did not match the server's, causing mount failures with sec=krb5 and require_gcm_256=1. Add a full_key_size parameter to generate_key() and pass the appropriate size from generate_smb3signingkey(): - Signing: always SMB2_NTLMV2_SESSKEY_SIZE (16 bytes) - Encryption/Decryption: ses->auth_key.len when AES-256, otherwise 16 Also fix cifs_dump_full_key() to report the actual session key length for AES-256 instead of hardcoded CIFS_SESS_KEY_SIZE, so that userspace tools like Wireshark receive the correct key for decryption. Cc: Reviewed-by: Bharath SM Signed-off-by: Piyush Sachdeva Signed-off-by: Piyush Sachdeva Signed-off-by: Steve French --- fs/smb/client/ioctl.c | 2 +- fs/smb/client/smb2transport.c | 35 ++++++++++++++++++++++++++--------- 2 files changed, 27 insertions(+), 10 deletions(-) diff --git a/fs/smb/client/ioctl.c b/fs/smb/client/ioctl.c index 9afab3237e54..17408bb8ab65 100644 --- a/fs/smb/client/ioctl.c +++ b/fs/smb/client/ioctl.c @@ -296,7 +296,7 @@ static int cifs_dump_full_key(struct cifs_tcon *tcon, struct smb3_full_key_debug break; case SMB2_ENCRYPTION_AES256_CCM: case SMB2_ENCRYPTION_AES256_GCM: - out.session_key_length = CIFS_SESS_KEY_SIZE; + out.session_key_length = ses->auth_key.len; out.server_in_key_length = out.server_out_key_length = SMB3_GCM256_CRYPTKEY_SIZE; break; default: diff --git a/fs/smb/client/smb2transport.c b/fs/smb/client/smb2transport.c index 41009039b4cb..e8eeff9e50d6 100644 --- a/fs/smb/client/smb2transport.c +++ b/fs/smb/client/smb2transport.c @@ -251,7 +251,8 @@ smb2_calc_signature(struct smb_rqst *rqst, struct TCP_Server_Info *server) } static void generate_key(struct cifs_ses *ses, struct kvec label, - struct kvec context, __u8 *key, unsigned int key_size) + struct kvec context, __u8 *key, unsigned int key_size, + unsigned int full_key_size) { unsigned char zero = 0x0; __u8 i[4] = {0, 0, 0, 1}; @@ -265,7 +266,7 @@ static void generate_key(struct cifs_ses *ses, struct kvec label, memset(key, 0x0, key_size); hmac_sha256_init_usingrawkey(&hmac_ctx, ses->auth_key.response, - SMB2_NTLMV2_SESSKEY_SIZE); + full_key_size); hmac_sha256_update(&hmac_ctx, i, 4); hmac_sha256_update(&hmac_ctx, label.iov_base, label.iov_len); hmac_sha256_update(&hmac_ctx, &zero, 1); @@ -298,6 +299,7 @@ generate_smb3signingkey(struct cifs_ses *ses, struct TCP_Server_Info *server, const struct derivation_triplet *ptriplet) { + unsigned int full_key_size = SMB2_NTLMV2_SESSKEY_SIZE; bool is_binding = false; int chan_index = 0; @@ -330,12 +332,24 @@ generate_smb3signingkey(struct cifs_ses *ses, if (is_binding) { generate_key(ses, ptriplet->signing.label, ptriplet->signing.context, - ses->chans[chan_index].signkey, - SMB3_SIGN_KEY_SIZE); + ses->chans[chan_index].signkey, SMB3_SIGN_KEY_SIZE, + SMB2_NTLMV2_SESSKEY_SIZE); } else { generate_key(ses, ptriplet->signing.label, - ptriplet->signing.context, - ses->smb3signingkey, SMB3_SIGN_KEY_SIZE); + ptriplet->signing.context, ses->smb3signingkey, + SMB3_SIGN_KEY_SIZE, SMB2_NTLMV2_SESSKEY_SIZE); + + /* + * Per MS-SMB2 3.2.5.3.1, signing key always uses Session.SessionKey + * (first 16 bytes). Encryption/decryption keys use + * Session.FullSessionKey when dialect is 3.1.1 and cipher is + * AES-256-CCM or AES-256-GCM, otherwise Session.SessionKey. + */ + + if (server->dialect == SMB311_PROT_ID && + (server->cipher_type == SMB2_ENCRYPTION_AES256_CCM || + server->cipher_type == SMB2_ENCRYPTION_AES256_GCM)) + full_key_size = ses->auth_key.len; /* safe to access primary channel, since it will never go away */ spin_lock(&ses->chan_lock); @@ -345,10 +359,13 @@ generate_smb3signingkey(struct cifs_ses *ses, generate_key(ses, ptriplet->encryption.label, ptriplet->encryption.context, - ses->smb3encryptionkey, SMB3_ENC_DEC_KEY_SIZE); + ses->smb3encryptionkey, SMB3_ENC_DEC_KEY_SIZE, + full_key_size); + generate_key(ses, ptriplet->decryption.label, ptriplet->decryption.context, - ses->smb3decryptionkey, SMB3_ENC_DEC_KEY_SIZE); + ses->smb3decryptionkey, SMB3_ENC_DEC_KEY_SIZE, + full_key_size); } #ifdef CONFIG_CIFS_DEBUG_DUMP_KEYS @@ -361,7 +378,7 @@ generate_smb3signingkey(struct cifs_ses *ses, &ses->Suid); cifs_dbg(VFS, "Cipher type %d\n", server->cipher_type); cifs_dbg(VFS, "Session Key %*ph\n", - SMB2_NTLMV2_SESSKEY_SIZE, ses->auth_key.response); + (int)ses->auth_key.len, ses->auth_key.response); cifs_dbg(VFS, "Signing Key %*ph\n", SMB3_SIGN_KEY_SIZE, ses->smb3signingkey); if ((server->cipher_type == SMB2_ENCRYPTION_AES256_CCM) || From 8cb6fc3231500233ddaf63cb7fd5435008d9ed5f Mon Sep 17 00:00:00 2001 From: Piyush Sachdeva Date: Thu, 7 May 2026 22:22:14 +0530 Subject: [PATCH 480/525] smb: client: Zero-pad short GSS session keys per MS-SMB2 Per MS-SMB2 section 3.2.5.3, Session.SessionKey is the first 16 bytes of the GSS cryptographic key, right-padded with zero bytes if the key is shorter than 16 bytes. SMB2_auth_kerberos() copies the GSS session key from the cifs.upcall response using kmemdup(msg->data, msg->sesskey_len, ...) and stores the GSS-reported length verbatim in ses->auth_key.len. generate_key() reads SMB2_NTLMV2_SESSKEY_SIZE bytes from this buffer when feeding the HMAC-SHA256 KDF for signing key derivation. If a GSS mechanism returns a session key shorter than 16 bytes (e.g. a deprecated single-DES Kerberos enctype with an 8-byte session key), the KDF call performs an out-of-bounds slab read and derives keys that do not match the server, which pads per the spec. Modern KDCs disable short-key enctypes by default, so this is latent rather than reachable in production, but it is still a kernel heap over-read. Allocate auth_key.response with kzalloc() at a length of max(msg->sesskey_len, SMB2_NTLMV2_SESSKEY_SIZE), copy the GSS key in, and rely on kzalloc()'s zero initialization for the spec-mandated padding. Set ses->auth_key.len to the padded length. Larger GSS keys (e.g. the 32-byte aes256-cts-hmac-sha1-96 session key) continue to be stored at their natural length, preserving the FullSessionKey path. Emit a cifs_dbg(VFS, ...) message when a short key is encountered to surface deprecated-enctype usage. NTLMv2 and NTLMSSP code paths produce a 16-byte session key by construction and are unaffected. Signed-off-by: Piyush Sachdeva Signed-off-by: Piyush Sachdeva Signed-off-by: Steve French --- fs/smb/client/smb2pdu.c | 23 ++++++++++++++++++----- 1 file changed, 18 insertions(+), 5 deletions(-) diff --git a/fs/smb/client/smb2pdu.c b/fs/smb/client/smb2pdu.c index cb61051f9af3..995fcdd30681 100644 --- a/fs/smb/client/smb2pdu.c +++ b/fs/smb/client/smb2pdu.c @@ -1713,17 +1713,30 @@ SMB2_auth_kerberos(struct SMB2_sess_data *sess_data) is_binding = (ses->ses_status == SES_GOOD); spin_unlock(&ses->ses_lock); + /* + * Per MS-SMB2 3.2.5.3, Session.SessionKey is the first 16 bytes of the + * GSS cryptographic key, right-padded with zero bytes if shorter. + * Allocate at least SMB2_NTLMV2_SESSKEY_SIZE bytes (zeroed) so the KDF + * input buffer is always valid for HMAC-SHA256 even with deprecated + * Kerberos enctypes that return a short session key. + */ + if (unlikely(msg->sesskey_len < SMB2_NTLMV2_SESSKEY_SIZE)) + cifs_dbg(VFS, + "short GSS session key (%u bytes); zero-padding per MS-SMB2 3.2.5.3\n", + msg->sesskey_len); + kfree_sensitive(ses->auth_key.response); - ses->auth_key.response = kmemdup(msg->data, - msg->sesskey_len, - GFP_KERNEL); + ses->auth_key.len = max_t(unsigned int, msg->sesskey_len, + SMB2_NTLMV2_SESSKEY_SIZE); + ses->auth_key.response = kzalloc(ses->auth_key.len, GFP_KERNEL); if (!ses->auth_key.response) { cifs_dbg(VFS, "%s: can't allocate (%u bytes) memory\n", - __func__, msg->sesskey_len); + __func__, ses->auth_key.len); + ses->auth_key.len = 0; rc = -ENOMEM; goto out_put_spnego_key; } - ses->auth_key.len = msg->sesskey_len; + memcpy(ses->auth_key.response, msg->data, msg->sesskey_len); sess_data->iov[1].iov_base = msg->data + msg->sesskey_len; sess_data->iov[1].iov_len = msg->secblob_len; From d62b8d236fab503c6fec1d3e9a38bea71feaca20 Mon Sep 17 00:00:00 2001 From: Zisen Ye Date: Sat, 2 May 2026 18:48:36 +0800 Subject: [PATCH 481/525] smb/client: fix out-of-bounds read in symlink_data() Since smb2_check_message() returns success without length validation for the symlink error response, in symlink_data() it is possible for iov->iov_len to be smaller than sizeof(struct smb2_err_rsp). If the buffer only contains the base SMB2 header (64 bytes), accessing err->ErrorContextCount (at offset 66) or err->ByteCount later in symlink_data() will cause an out-of-bounds read. Link: https://lore.kernel.org/linux-cifs/297d8d9b-adf7-42fd-a1c2-5b1f230032bc@chenxiaosong.com/ Fixes: 76894f3e2f71 ("cifs: improve symlink handling for smb2+") Cc: Stable@vger.kernel.org Signed-off-by: Zisen Ye Reviewed-by: ChenXiaoSong Signed-off-by: Steve French --- fs/smb/client/smb2misc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/smb/client/smb2misc.c b/fs/smb/client/smb2misc.c index 973fce3c959c..2a7355ce1a07 100644 --- a/fs/smb/client/smb2misc.c +++ b/fs/smb/client/smb2misc.c @@ -241,7 +241,8 @@ smb2_check_message(char *buf, unsigned int pdu_len, unsigned int len, if (len != calc_len) { /* create failed on symlink */ if (command == SMB2_CREATE_HE && - shdr->Status == STATUS_STOPPED_ON_SYMLINK) + shdr->Status == STATUS_STOPPED_ON_SYMLINK && + len > calc_len) return 0; /* Windows 7 server returns 24 bytes more */ if (calc_len + 24 == len && command == SMB2_OPLOCK_BREAK_HE) From 8d09328dfda089675e4c049f3f256064a1d1996b Mon Sep 17 00:00:00 2001 From: Zisen Ye Date: Wed, 6 May 2026 11:49:08 +0800 Subject: [PATCH 482/525] smb/client: fix out-of-bounds read in smb2_compound_op() If a server sends a truncated response but a large OutputBufferLength, and terminates the EA list early, check_wsl_eas() returns success without validating that the entire OutputBufferLength fits within iov_len. Then smb2_compound_op() does: memcpy(idata->wsl.eas, data[0], size[0]); Where size[0] is OutputBufferLength. If iov_len is smaller than size[0], memcpy can read beyond the end of the rsp_iov allocation and leak adjacent kernel heap memory. Link: https://lore.kernel.org/linux-cifs/d998240c-aca9-420d-9dbd-f5ba24af19e0@chenxiaosong.com/ Fixes: ea41367b2a60 ("smb: client: introduce SMB2_OP_QUERY_WSL_EA") Cc: stable@vger.kernel.org Signed-off-by: Zisen Ye Reviewed-by: ChenXiaoSong Signed-off-by: Steve French --- fs/smb/client/smb2inode.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/fs/smb/client/smb2inode.c b/fs/smb/client/smb2inode.c index 286912616c73..6c9c229b91f6 100644 --- a/fs/smb/client/smb2inode.c +++ b/fs/smb/client/smb2inode.c @@ -111,7 +111,7 @@ static int check_wsl_eas(struct kvec *rsp_iov) u32 outlen, next; u16 vlen; u8 nlen; - u8 *end; + u8 *ea_end, *iov_end; outlen = le32_to_cpu(rsp->OutputBufferLength); if (outlen < SMB2_WSL_MIN_QUERY_EA_RESP_SIZE || @@ -120,15 +120,19 @@ static int check_wsl_eas(struct kvec *rsp_iov) ea = (void *)((u8 *)rsp_iov->iov_base + le16_to_cpu(rsp->OutputBufferOffset)); - end = (u8 *)rsp_iov->iov_base + rsp_iov->iov_len; + ea_end = (u8 *)ea + outlen; + iov_end = (u8 *)rsp_iov->iov_base + rsp_iov->iov_len; + if (ea_end > iov_end) + return -EINVAL; + for (;;) { - if ((u8 *)ea > end - sizeof(*ea)) + if ((u8 *)ea > ea_end - sizeof(*ea)) return -EINVAL; nlen = ea->ea_name_length; vlen = le16_to_cpu(ea->ea_value_length); if (nlen != SMB2_WSL_XATTR_NAME_LEN || - (u8 *)ea->ea_data + nlen + 1 + vlen > end) + (u8 *)ea->ea_data + nlen + 1 + vlen > ea_end) return -EINVAL; switch (vlen) { From f98b48151cc502ada59d9778f0112d21f2586ca3 Mon Sep 17 00:00:00 2001 From: Michael Bommarito Date: Mon, 20 Apr 2026 10:47:47 -0400 Subject: [PATCH 483/525] smb: client: validate dacloffset before building DACL pointers parse_sec_desc(), build_sec_desc(), and the chown path in id_mode_to_cifs_acl() all add the server-supplied dacloffset to pntsd before proving a DACL header fits inside the returned security descriptor. On 32-bit builds a malicious server can return dacloffset near U32_MAX, wrap the derived DACL pointer below end_of_acl, and then slip past the later pointer-based bounds checks. build_sec_desc() and id_mode_to_cifs_acl() can then dereference DACL fields from the wrapped pointer in the chmod/chown rewrite paths. Validate dacloffset numerically before building any DACL pointer and reuse the same helper at the three DACL entry points. Fixes: bc3e9dd9d104 ("cifs: Change SIDs in ACEs while transferring file ownership.") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-opus-4-6 Signed-off-by: Michael Bommarito Signed-off-by: Steve French --- fs/smb/client/cifsacl.c | 35 ++++++++++++++++++++++++++++++++--- 1 file changed, 32 insertions(+), 3 deletions(-) diff --git a/fs/smb/client/cifsacl.c b/fs/smb/client/cifsacl.c index a2750f1e3d90..786dbbc43c5b 100644 --- a/fs/smb/client/cifsacl.c +++ b/fs/smb/client/cifsacl.c @@ -1264,6 +1264,17 @@ static int parse_sid(struct smb_sid *psid, char *end_of_acl) return 0; } +static bool dacl_offset_valid(unsigned int acl_len, __u32 dacloffset) +{ + if (acl_len < sizeof(struct smb_acl)) + return false; + + if (dacloffset < sizeof(struct smb_ntsd)) + return false; + + return dacloffset <= acl_len - sizeof(struct smb_acl); +} + /* Convert CIFS ACL to POSIX form */ static int parse_sec_desc(struct cifs_sb_info *cifs_sb, @@ -1284,7 +1295,6 @@ static int parse_sec_desc(struct cifs_sb_info *cifs_sb, group_sid_ptr = (struct smb_sid *)((char *)pntsd + le32_to_cpu(pntsd->gsidoffset)); dacloffset = le32_to_cpu(pntsd->dacloffset); - dacl_ptr = (struct smb_acl *)((char *)pntsd + dacloffset); cifs_dbg(NOISY, "revision %d type 0x%x ooffset 0x%x goffset 0x%x sacloffset 0x%x dacloffset 0x%x\n", pntsd->revision, pntsd->type, le32_to_cpu(pntsd->osidoffset), le32_to_cpu(pntsd->gsidoffset), @@ -1315,11 +1325,18 @@ static int parse_sec_desc(struct cifs_sb_info *cifs_sb, return rc; } - if (dacloffset) + if (dacloffset) { + if (!dacl_offset_valid(acl_len, dacloffset)) { + cifs_dbg(VFS, "Server returned illegal DACL offset\n"); + return -EINVAL; + } + + dacl_ptr = (struct smb_acl *)((char *)pntsd + dacloffset); parse_dacl(dacl_ptr, end_of_acl, owner_sid_ptr, group_sid_ptr, fattr, get_mode_from_special_sid); - else + } else { cifs_dbg(FYI, "no ACL\n"); /* BB grant all or default perms? */ + } return rc; } @@ -1342,6 +1359,11 @@ static int build_sec_desc(struct smb_ntsd *pntsd, struct smb_ntsd *pnntsd, dacloffset = le32_to_cpu(pntsd->dacloffset); if (dacloffset) { + if (!dacl_offset_valid(secdesclen, dacloffset)) { + cifs_dbg(VFS, "Server returned illegal DACL offset\n"); + return -EINVAL; + } + dacl_ptr = (struct smb_acl *)((char *)pntsd + dacloffset); rc = validate_dacl(dacl_ptr, end_of_acl); if (rc) @@ -1710,6 +1732,12 @@ id_mode_to_cifs_acl(struct inode *inode, const char *path, __u64 *pnmode, nsecdesclen = sizeof(struct smb_ntsd) + (sizeof(struct smb_sid) * 2); dacloffset = le32_to_cpu(pntsd->dacloffset); if (dacloffset) { + if (!dacl_offset_valid(secdesclen, dacloffset)) { + cifs_dbg(VFS, "Server returned illegal DACL offset\n"); + rc = -EINVAL; + goto id_mode_to_cifs_acl_exit; + } + dacl_ptr = (struct smb_acl *)((char *)pntsd + dacloffset); rc = validate_dacl(dacl_ptr, (char *)pntsd + secdesclen); if (rc) { @@ -1752,6 +1780,7 @@ id_mode_to_cifs_acl(struct inode *inode, const char *path, __u64 *pnmode, rc = ops->set_acl(pnntsd, nsecdesclen, inode, path, aclflag); cifs_dbg(NOISY, "set_cifs_acl rc: %d\n", rc); } +id_mode_to_cifs_acl_exit: cifs_put_tlink(tlink); kfree(pnntsd); From 3607422cdeebd1f0c259964bf33aaa9792e21930 Mon Sep 17 00:00:00 2001 From: Markus Stockhausen Date: Sat, 2 May 2026 19:32:06 +0200 Subject: [PATCH 484/525] hwmon: (lm75) Fix AS6200 and TMP112 setup and alarm handling The initialization of the AS6200 has two shortcomings - The device-add-commit states "Conversion mode: continuous" but the the lm75_params structure uses set_mask = 0x94c0. This activates single shot mode (bit 15). According to the datasheet "The device features a single shot measurement mode if the device is in sleep mode (SM=1)". This is quite contradictionary. - It is the only device that activates polarity active-high (bit 10) All this is paired with a undefined clear mask bug in function lm75_write_config() that was introduced with a later refactoring commit. [as6200] = { .config_reg_16bits = true, .set_mask = 0x94C0, -> .clr_mask not defined here .default_resolution = 12, ... static inline int lm75_write_config(struct lm75_data *data, u16 set_mask, u16 clr_mask) { return regmap_update_bits(data->regmap, LM75_REG_CONF, clr_mask | LM75_SHUTDOWN, set_mask); } regmap_update_bits() requires clr_mask to be a superset of set_mask. So basically all sensors with "wrong" masks like the AS6200 are not initialized as intended. Fix that by - Change the set_mask to 0xc010 to reflect the current active-low setup properly and to drive the sensor in continous mode. This takes into account that the config register is little endian and the first byte sent to the chip is the LSB. - Adapt the alarm handling so it can report the alarm correctly even if it is high active. This is done by comparing config register bit 5 and 10 (translated to 2 and 13). This commit does not introduce any ABI breakage as the mutliple bugs effectly drive the AS6200 in standard active-low mode. Fixes: 4b6358e1fe46 ("hwmon: (lm75) Add AMS AS6200 temperature sensor") Suggested-by: Guenter Roeck Signed-off-by: Markus Stockhausen Link: https://lore.kernel.org/r/20260502173207.3567876-2-markus.stockhausen@gmx.de [groeck: Update set_mask for as6200 further: As modeled, the upper bits contain the conversion rate, so the config register needs to be set to 0xc010 instead of 0x10c0 to reflect 8 samples/s and 4 consecutive faults. Fix the same problem for TMP112.] Signed-off-by: Guenter Roeck --- drivers/hwmon/lm75.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/hwmon/lm75.c b/drivers/hwmon/lm75.c index f1a1e5b888f6..a1bf4e9813ed 100644 --- a/drivers/hwmon/lm75.c +++ b/drivers/hwmon/lm75.c @@ -137,7 +137,7 @@ static const struct lm75_params device_params[] = { }, [as6200] = { .config_reg_16bits = true, - .set_mask = 0x94C0, /* 8 sample/s, 4 CF, positive polarity */ + .set_mask = 0xC010, /* 8 sample/s, 4 CF */ .default_resolution = 12, .default_sample_time = 125, .num_sample_times = 4, @@ -286,8 +286,8 @@ static const struct lm75_params device_params[] = { }, [tmp112] = { .config_reg_16bits = true, - .set_mask = 0x60C0, /* 12-bit mode, 8 samples / second */ - .clr_mask = 1 << 15, /* no one-shot mode*/ + .set_mask = 0xC060, /* 12-bit mode, 8 samples / second */ + .clr_mask = 1 << 7, /* no one-shot mode*/ .default_resolution = 12, .default_sample_time = 125, .num_sample_times = 4, @@ -416,7 +416,7 @@ static int lm75_read(struct device *dev, enum hwmon_sensor_types type, switch (data->kind) { case as6200: case tmp112: - *val = (regval >> 13) & 0x1; + *val = !!(regval & BIT(13)) == !!(regval & BIT(2)); break; default: return -EINVAL; From 05aaac8746c5786eaa779b163fae4ddcd5172707 Mon Sep 17 00:00:00 2001 From: Markus Stockhausen Date: Sat, 2 May 2026 19:32:07 +0200 Subject: [PATCH 485/525] hwmon: (lm75) Fix configuration register writes. Sensors configurations are defined by set and clear masks. These do not follow a consistent "clear mask is a superset of set mask" rule. This relaxed definition breaks lm75_write_config() static inline int lm75_write_config(struct lm75_data *data, u16 set_mask, u16 clr_mask) { return regmap_update_bits(data->regmap, LM75_REG_CONF, clr_mask | LM75_SHUTDOWN, set_mask); } Basically all bits from set_mask that are not defined in clr_mask are dropped. Fix that by enhancing the helper to always combine clr_mask and set_mask into the mask bits of regmap_update_bits(). Fixes: 6da24a25f766 ("hwmon: (lm75) Hide register size differences in regmap access functions") Suggested-by: Guenter Roeck Signed-off-by: Markus Stockhausen Link: https://lore.kernel.org/r/20260502173207.3567876-3-markus.stockhausen@gmx.de Signed-off-by: Guenter Roeck --- drivers/hwmon/lm75.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/hwmon/lm75.c b/drivers/hwmon/lm75.c index a1bf4e9813ed..c283443e363b 100644 --- a/drivers/hwmon/lm75.c +++ b/drivers/hwmon/lm75.c @@ -353,7 +353,7 @@ static inline int lm75_write_config(struct lm75_data *data, u16 set_mask, u16 clr_mask) { return regmap_update_bits(data->regmap, LM75_REG_CONF, - clr_mask | LM75_SHUTDOWN, set_mask); + clr_mask | set_mask | LM75_SHUTDOWN, set_mask); } static irqreturn_t lm75_alarm_handler(int irq, void *private) From 99076a17a112ac43cbd37f6883898ae649166303 Mon Sep 17 00:00:00 2001 From: Tabrez Ahmed Date: Sat, 2 May 2026 07:38:42 +0530 Subject: [PATCH 486/525] hwmon: (ads7871) Fix endianness bug in 16-bit register reads The ads7871_read_reg16() function relies on spi_w8r16() to read the 16-bit sensor output. The ADS7871 device transmits the Least Significant Byte (LSB) first. On Little-Endian architectures, spi_w8r16() correctly reconstructs the 16-bit value. However, on Big-Endian architectures, the byte swapping causes the first received byte (LSB) to be placed in the most significant byte of the u16, resulting in corrupted voltage readings. To fix this, cast the integer result of spi_w8r16() to a restricted __le16 type and convert it to the host CPU's native byte order using le16_to_cpu(). Negative error codes returned by the SPI core are caught and returned prior to the conversion to avoid mangling the error status. Reported-by: Sashiko Closes: https://sashiko.dev/#/patchset/20260418034601.90226-1-tabreztalks@gmail.com Fixes: e0c70b8078629 ("hwmon: add TI ads7871 a/d converter driver") Suggested-by: David Laight Signed-off-by: Tabrez Ahmed Link: https://lore.kernel.org/r/20260502020844.110038-2-tabreztalks@gmail.com Signed-off-by: Guenter Roeck --- drivers/hwmon/ads7871.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/hwmon/ads7871.c b/drivers/hwmon/ads7871.c index 9bfdf9e6bcd7..9ee3ce01f130 100644 --- a/drivers/hwmon/ads7871.c +++ b/drivers/hwmon/ads7871.c @@ -77,9 +77,13 @@ static int ads7871_read_reg8(struct spi_device *spi, int reg) static int ads7871_read_reg16(struct spi_device *spi, int reg) { int ret; + reg = reg | INST_READ_BM | INST_16BIT_BM; ret = spi_w8r16(spi, reg); - return ret; + if (ret < 0) + return ret; + + return le16_to_cpu((__force __le16)ret); } static int ads7871_write_reg8(struct spi_device *spi, int reg, u8 val) From 5e28b7b94408897e41c63477aabc9e1db439bc8c Mon Sep 17 00:00:00 2001 From: "Francis, David" Date: Tue, 28 Apr 2026 19:25:50 +0000 Subject: [PATCH 487/525] drm: Set old handle to NULL before prime swap in change_handle There was a potential race condition in change_handle. The ioctl briefly had a single object with two idr entries; a concurrent gem_close could delete the object and remove one of the handles while leaving the other one dangling, which could subsequently be dereferenced for a use-after-free. To fix this, do the same dance that gem_close itself does. (f6cd7daecff5 drm: Release driver references to handle before making it available again) First idr_replace the old handle to NULL. Later, if the prime operations are successful, actually close it. create_tail required a similar dance to avoid a similar problem. (bd46cece51a3 drm/gem: Fix race in drm_gem_handle_create_tail()) It idr_allocs the new handle with NULL, then swaps in the correct object later to avoid races. We don't need to do that here, since the only operations that could race are drm_prime, and change_handle holds the prime lock for the entire duration. v2: cleanups of error paths Signed-off-by: David Francis Co-authored-by: Dave Airlie Reported-by: Puttimet Thammasaeng Tested-by: Vitaly Prosyak Cc: Simona Vetter Cc: stable@vger.kernel.org Cc: Christian Koenig Fixes: 53096728b8910 ("drm: Add DRM prime interface to reassign GEM handle") Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_gem.c | 25 ++++++++++++++++++++++++- 1 file changed, 24 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c index d6424267260b..51a887cc7fd7 100644 --- a/drivers/gpu/drm/drm_gem.c +++ b/drivers/gpu/drm/drm_gem.c @@ -1019,7 +1019,7 @@ int drm_gem_change_handle_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv) { struct drm_gem_change_handle *args = data; - struct drm_gem_object *obj; + struct drm_gem_object *obj, *idrobj; int handle, ret; if (!drm_core_check_feature(dev, DRIVER_GEM)) @@ -1042,8 +1042,29 @@ int drm_gem_change_handle_ioctl(struct drm_device *dev, void *data, mutex_lock(&file_priv->prime.lock); spin_lock(&file_priv->table_lock); + + /* When create_tail allocs an obj idr, it needs to first alloc as NULL, + * then later replace with the correct object. This is not necessary + * here, because the only operations that could race are drm_prime + * bookkeeping, and we hold the prime lock. + */ ret = idr_alloc(&file_priv->object_idr, obj, handle, handle + 1, GFP_NOWAIT); + + if (ret < 0) { + spin_unlock(&file_priv->table_lock); + goto out_unlock; + } + + idrobj = idr_replace(&file_priv->object_idr, NULL, handle); + if (idrobj != obj) { + idr_replace(&file_priv->object_idr, idrobj, handle); + idr_remove(&file_priv->object_idr, args->new_handle); + spin_unlock(&file_priv->table_lock); + ret = -ENOENT; + goto out_unlock; + } + spin_unlock(&file_priv->table_lock); if (ret < 0) @@ -1055,6 +1076,8 @@ int drm_gem_change_handle_ioctl(struct drm_device *dev, void *data, if (ret < 0) { spin_lock(&file_priv->table_lock); idr_remove(&file_priv->object_idr, handle); + idrobj = idr_replace(&file_priv->object_idr, obj, handle); + WARN_ON(idrobj != NULL); spin_unlock(&file_priv->table_lock); goto out_unlock; } From f7700a4415afb3ac1767a556094e4ef8bd440e41 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Fri, 8 May 2026 20:37:46 +0800 Subject: [PATCH 488/525] ublk: fix use-after-free in ublk_cancel_cmd() When ublk_reset_ch_dev() clears io->cmd via ublk_queue_reinit() concurrently with ublk_cancel_cmd(), ublk_cancel_cmd() can read a stale pointer and pass it to io_uring_cmd_done(), causing a use-after-free. Fix by synchronizing the two paths with ubq->cancel_lock: - ublk_cancel_cmd(): read and clear io->cmd under cancel_lock, then call io_uring_cmd_done() on the saved local copy outside the lock. - ublk_reset_ch_dev(): hold cancel_lock across ublk_queue_reinit() so that io->cmd and io->flags are cleared atomically with respect to ublk_cancel_cmd(). Fixes: 216c8f5ef0f2 ("ublk: replace monitor with cancelable uring_cmd") Signed-off-by: Ming Lei Link: https://patch.msgid.link/20260508123746.242018-1-tom.leiming@gmail.com Signed-off-by: Jens Axboe --- drivers/block/ublk_drv.c | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c index 57ec900f0ce0..6d13f1481de0 100644 --- a/drivers/block/ublk_drv.c +++ b/drivers/block/ublk_drv.c @@ -2411,8 +2411,14 @@ static void ublk_reset_ch_dev(struct ublk_device *ub) { int i; - for (i = 0; i < ub->dev_info.nr_hw_queues; i++) - ublk_queue_reinit(ub, ublk_get_queue(ub, i)); + for (i = 0; i < ub->dev_info.nr_hw_queues; i++) { + struct ublk_queue *ubq = ublk_get_queue(ub, i); + + /* Sync with ublk_cancel_cmd() */ + spin_lock(&ubq->cancel_lock); + ublk_queue_reinit(ub, ubq); + spin_unlock(&ubq->cancel_lock); + } /* set to NULL, otherwise new tasks cannot mmap io_cmd_buf */ ub->mm = NULL; @@ -2753,6 +2759,7 @@ static void ublk_cancel_cmd(struct ublk_queue *ubq, unsigned tag, { struct ublk_io *io = &ubq->ios[tag]; struct ublk_device *ub = ubq->dev; + struct io_uring_cmd *cmd = NULL; struct request *req; bool done; @@ -2775,12 +2782,15 @@ static void ublk_cancel_cmd(struct ublk_queue *ubq, unsigned tag, spin_lock(&ubq->cancel_lock); done = !!(io->flags & UBLK_IO_FLAG_CANCELED); - if (!done) + if (!done) { io->flags |= UBLK_IO_FLAG_CANCELED; + cmd = io->cmd; + io->cmd = NULL; + } spin_unlock(&ubq->cancel_lock); - if (!done) - io_uring_cmd_done(io->cmd, UBLK_IO_RES_ABORT, issue_flags); + if (!done && cmd) + io_uring_cmd_done(cmd, UBLK_IO_RES_ABORT, issue_flags); } /* From 481c2265286ef302327c93403a8cf7b3fe4506d0 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Mon, 4 May 2026 21:04:48 +0000 Subject: [PATCH 489/525] bpf: tcp: Fix type confusion in bpf_tcp_sock(). bpf_tcp_sock() only checks if sk->sk_protocol is IPPROTO_TCP, but RAW socket can bypass it: socket(AF_INET, SOCK_RAW, IPPROTO_TCP) Calling bpf_setsockopt() in SOCKOPT prog triggers out-of-bounds access to another slab object. [0] Let's use sk_is_tcp(). [0]: BUG: KASAN: slab-out-of-bounds in sol_tcp_sockopt (net/core/filter.c:5519) Read of size 8 at addr ffff88801083d760 by task test_progs/1259 CPU: 1 UID: 0 PID: 1259 Comm: test_progs Tainted: G OE 7.0.0-11175-gb5c111f4967b #1 PREEMPT(full) Tainted: [O]=OOT_MODULE, [E]=UNSIGNED_MODULE Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.17.0-debian-1.17.0-1 04/01/2014 Call Trace: dump_stack_lvl (lib/dump_stack.c:94 lib/dump_stack.c:120) print_report (mm/kasan/report.c:378 mm/kasan/report.c:482) kasan_report (mm/kasan/report.c:595) sol_tcp_sockopt (net/core/filter.c:5519) __bpf_getsockopt (net/core/filter.c:5633) bpf_sk_getsockopt (net/core/filter.c:5654) bpf_prog_629ba00a1601e9f2__setsockopt+0x86/0x22c __cgroup_bpf_run_filter_setsockopt (./include/linux/bpf.h:1402 ./include/linux/filter.h:722 ./include/linux/filter.h:729 kernel/bpf/cgroup.c:81 kernel/bpf/cgroup.c:2026) do_sock_setsockopt (net/socket.c:2363) __x64_sys_setsockopt (net/socket.c:2406) do_syscall_64 (arch/x86/entry/syscall_64.c:63) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:121) RIP: 0033:0x7f85f82fe7de Code: 55 48 63 c9 48 63 ff 45 89 c9 48 89 e5 48 83 ec 08 6a 2c e8 34 69 f7 ff c9 c3 66 90 f3 0f 1e fa 49 89 ca b8 36 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 0a c3 66 0f 1f 84 00 00 00 00 00 48 8b 15 e1 RSP: 002b:00007ffe59dcecd8 EFLAGS: 00000202 ORIG_RAX: 0000000000000036 RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f85f82fe7de RDX: 000000000000001c RSI: 0000000000000006 RDI: 000000000000000d RBP: 00007ffe59dcef20 R08: 000000000000003c R09: 0000000000000000 R10: 00007ffe59dcef00 R11: 0000000000000202 R12: 00007ffe59dcf268 R13: 0000000000000003 R14: 00007f85f9da5000 R15: 000055b2f3201400 The buggy address belongs to the object at ffff88801083d280 which belongs to the cache RAW of size 1792 The buggy address is located 1248 bytes inside of allocated 1792-byte region [ffff88801083d280, ffff88801083d980) Fixes: 655a51e536c0 ("bpf: Add struct bpf_tcp_sock and BPF_FUNC_tcp_sock") Reported-by: Damiano Melotti Signed-off-by: Kuniyuki Iwashima Signed-off-by: Martin KaFai Lau Acked-by: Daniel Borkmann Link: https://patch.msgid.link/20260504210610.180150-2-kuniyu@google.com --- net/core/filter.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/filter.c b/net/core/filter.c index bc96c18df4e0..cd88633f8dc1 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -7475,7 +7475,7 @@ u32 bpf_tcp_sock_convert_ctx_access(enum bpf_access_type type, BPF_CALL_1(bpf_tcp_sock, struct sock *, sk) { - if (sk_fullsock(sk) && sk->sk_protocol == IPPROTO_TCP) + if (sk_fullsock(sk) && sk_is_tcp(sk)) return (unsigned long)sk; return (unsigned long)NULL; From d73549b8bb7fa6147666c579d66f72bf076f719f Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Mon, 4 May 2026 21:04:49 +0000 Subject: [PATCH 490/525] selftest: bpf: Add test for bpf_tcp_sock() and RAW socket. Let's extend sockopt_sk.c to cover bpf_tcp_sock() for the wrong socket type. Before: # ./test_progs -t sockopt_sk [ 151.948613] ================================================================== [ 151.951376] BUG: KASAN: slab-out-of-bounds in sol_tcp_sockopt+0xc7/0x8e0 [ 151.954159] Read of size 8 at addr ffff88801083d760 by task test_progs/1259 ... run_test:FAIL:getsetsockopt unexpected error: -1 (errno 0) #427 sockopt_sk:FAIL After: #427 sockopt_sk:OK While at it, missing free() is fixed up. Signed-off-by: Kuniyuki Iwashima Signed-off-by: Martin KaFai Lau Link: https://patch.msgid.link/20260504210610.180150-3-kuniyu@google.com --- .../selftests/bpf/prog_tests/sockopt_sk.c | 17 ++++++++++++++++- tools/testing/selftests/bpf/progs/sockopt_sk.c | 16 ++++++++++++++++ 2 files changed, 32 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/prog_tests/sockopt_sk.c b/tools/testing/selftests/bpf/prog_tests/sockopt_sk.c index 53637431ec5d..3a41c517b918 100644 --- a/tools/testing/selftests/bpf/prog_tests/sockopt_sk.c +++ b/tools/testing/selftests/bpf/prog_tests/sockopt_sk.c @@ -190,7 +190,7 @@ static int getsetsockopt(void) fd = socket(AF_NETLINK, SOCK_RAW, 0); if (fd < 0) { log_err("Failed to create AF_NETLINK socket"); - return -1; + goto err; } buf.u32 = 1; @@ -211,6 +211,21 @@ static int getsetsockopt(void) } ASSERT_EQ(optlen, 8, "Unexpected NETLINK_LIST_MEMBERSHIPS value"); + /* Trick bpf_tcp_sock() with IPPROTO_TCP */ + close(fd); + fd = socket(AF_INET, SOCK_RAW, IPPROTO_TCP); + if (!ASSERT_OK_FD(fd, "socket")) + goto err; + + /* The BPF prog intercepts this before the kernel sees it, any + * optlen works. Go with 4 bytes for simplicity. + */ + buf.u32 = 1; + optlen = sizeof(buf.u32); + err = setsockopt(fd, SOL_TCP, TCP_SAVED_SYN, &buf, optlen); + if (!ASSERT_ERR(err, "setsockopt(TCP_SAVED_SYN)")) + goto err; + free(big_buf); close(fd); return 0; diff --git a/tools/testing/selftests/bpf/progs/sockopt_sk.c b/tools/testing/selftests/bpf/progs/sockopt_sk.c index cb990a7d3d45..5e0b27e7855c 100644 --- a/tools/testing/selftests/bpf/progs/sockopt_sk.c +++ b/tools/testing/selftests/bpf/progs/sockopt_sk.c @@ -149,6 +149,20 @@ int _setsockopt(struct bpf_sockopt *ctx) if (sk && sk->family == AF_NETLINK) goto out; + if (sk && sk->family == AF_INET && sk->type == SOCK_RAW) { + struct bpf_tcp_sock *tp = bpf_tcp_sock(sk); + + if (tp) { + char saved_syn[60]; + + bpf_getsockopt(sk, SOL_TCP, TCP_SAVED_SYN, + &saved_syn, sizeof(saved_syn)); + goto consumed; + } + + goto out; + } + /* Make sure bpf_get_netns_cookie is callable. */ if (bpf_get_netns_cookie(NULL) == 0) @@ -224,6 +238,8 @@ int _setsockopt(struct bpf_sockopt *ctx) return 0; /* couldn't get sk storage */ storage->val = optval[0]; + +consumed: ctx->optlen = -1; /* BPF has consumed this option, don't call kernel * setsockopt handler. */ From 7995b216a731db657f356f6ae37a42f445b9a0ec Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Mon, 4 May 2026 21:04:50 +0000 Subject: [PATCH 491/525] mptcp: bpf: Fix type confusion in bpf_mptcp_sock_from_subflow() bpf_mptcp_sock_from_subflow() only checks if sk->sk_protocol is IPPROTO_TCP, but RAW socket can bypass it: socket(AF_INET, SOCK_RAW, IPPROTO_TCP) In this case, it would NOT be valid to call sk_is_mptcp() which will assume sk is a pointer to a struct tcp_sock, and wrongly checks for: tcp_sk(sk)->is_mptcp. Fixes: 3bc253c2e652 ("bpf: Add bpf_skc_to_mptcp_sock_proto") Signed-off-by: Matthieu Baerts (NGI0) Signed-off-by: Kuniyuki Iwashima Signed-off-by: Martin KaFai Lau Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20260504210610.180150-4-kuniyu@google.com --- net/mptcp/bpf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mptcp/bpf.c b/net/mptcp/bpf.c index 8a16672b94e2..4cc16cbeb328 100644 --- a/net/mptcp/bpf.c +++ b/net/mptcp/bpf.c @@ -14,7 +14,7 @@ struct mptcp_sock *bpf_mptcp_sock_from_subflow(struct sock *sk) { - if (sk && sk_fullsock(sk) && sk->sk_protocol == IPPROTO_TCP && sk_is_mptcp(sk)) + if (sk && sk_fullsock(sk) && sk_is_tcp(sk) && sk_is_mptcp(sk)) return mptcp_sk(mptcp_subflow_ctx(sk)->conn); return NULL; From decb84b8383ab7acff94db208ef7ed19f9c55e1f Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Mon, 4 May 2026 21:04:51 +0000 Subject: [PATCH 492/525] bpf: tcp: Fix type confusion in bpf_skc_to_tcp_sock(). bpf_skc_to_tcp_sock() only checks if sk->sk_protocol is IPPROTO_TCP, but RAW socket can bypass it: socket(AF_INET, SOCK_RAW, IPPROTO_TCP) Let's use sk_is_tcp(). Fixes: 478cfbdf5f13 ("bpf: Add bpf_skc_to_{tcp, tcp_timewait, tcp_request}_sock() helpers") Signed-off-by: Kuniyuki Iwashima Signed-off-by: Martin KaFai Lau Link: https://patch.msgid.link/20260504210610.180150-5-kuniyu@google.com --- net/core/filter.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/filter.c b/net/core/filter.c index cd88633f8dc1..7d945dc2cb92 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -11963,7 +11963,7 @@ const struct bpf_func_proto bpf_skc_to_tcp6_sock_proto = { BPF_CALL_1(bpf_skc_to_tcp_sock, struct sock *, sk) { - if (sk && sk_fullsock(sk) && sk->sk_protocol == IPPROTO_TCP) + if (sk && sk_fullsock(sk) && sk_is_tcp(sk)) return (unsigned long)sk; return (unsigned long)NULL; From 843064b0a77eed3d6d63ffc53aeaa359672b4e12 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Mon, 4 May 2026 21:04:52 +0000 Subject: [PATCH 493/525] bpf: tcp: Fix type confusion in bpf_skc_to_tcp6_sock(). bpf_skc_to_tcp6_sock() only checks if sk->sk_protocol is IPPROTO_TCP and sk->sk_family is AF_INET6, but RAW socket can bypass it: socket(AF_INET6, SOCK_RAW, IPPROTO_TCP) Let's check sk->sk_type too. Fixes: af7ec1383361 ("bpf: Add bpf_skc_to_tcp6_sock() helper") Signed-off-by: Kuniyuki Iwashima Signed-off-by: Martin KaFai Lau Link: https://patch.msgid.link/20260504210610.180150-6-kuniyu@google.com --- net/core/filter.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/filter.c b/net/core/filter.c index 7d945dc2cb92..684922efd481 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -11947,7 +11947,7 @@ BPF_CALL_1(bpf_skc_to_tcp6_sock, struct sock *, sk) */ BTF_TYPE_EMIT(struct tcp6_sock); if (sk && sk_fullsock(sk) && sk->sk_protocol == IPPROTO_TCP && - sk->sk_family == AF_INET6) + sk->sk_type == SOCK_STREAM && sk->sk_family == AF_INET6) return (unsigned long)sk; return (unsigned long)NULL; From 1c2958e4ab1ed4594db16425dbcab33c56ea8330 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Mon, 4 May 2026 21:04:53 +0000 Subject: [PATCH 494/525] bpf: tcp: Fix type confusion in sol_tcp_sockopt(). sol_tcp_sockopt() only checks if sk->sk_protocol is IPPROTO_TCP, but RAW socket can bypass it: socket(AF_INET, SOCK_RAW, IPPROTO_TCP) Let's use sk_is_tcp(). Note that initially sol_tcp_sockopt() checked sk->sk_prot->setsockopt. Fixes: 2ab42c7b871f ("bpf: Check the protocol of a sock to agree the calls to bpf_setsockopt().") Signed-off-by: Kuniyuki Iwashima Signed-off-by: Martin KaFai Lau Link: https://patch.msgid.link/20260504210610.180150-7-kuniyu@google.com --- net/core/filter.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/filter.c b/net/core/filter.c index 684922efd481..ef0877eefaa7 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -5481,7 +5481,7 @@ static int sol_tcp_sockopt(struct sock *sk, int optname, char *optval, int *optlen, bool getopt) { - if (sk->sk_protocol != IPPROTO_TCP) + if (!sk_is_tcp(sk)) return -EINVAL; switch (optname) { From 18fc650ccd7fe3376eca89203668cfb8268f60df Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Sun, 26 Apr 2026 01:26:43 +0000 Subject: [PATCH 495/525] bpf: Free reuseport cBPF prog after RCU grace period. Eulgyu Kim reported the splat below with a repro. [0] The repro sets up a UDP reuseport group with a cBPF prog and replaces it with a new one while another thread is sending a UDP packet to the group. The reuseport prog is freed by sk_reuseport_prog_free(). bpf_prog_put() is called for "e"BPF prog to destruct through multiple stages while cBPF prog is freed immediately by bpf_release_orig_filter() and bpf_prog_free(). If a reuseport prog is detached from the setsockopt() path (reuseport_attach_prog() or reuseport_detach_prog()), sk_reuseport_prog_free() is called without waiting for RCU readers to complete, resulting in various bugs. Let's defer freeing the reuseport cBPF prog after one RCU grace period. Note "e"BPF prog is safe as is unless the fast path starts to touch fields destroyed in bpf_prog_put_deferred() and __bpf_prog_put_noref(). [0]: BUG: KASAN: vmalloc-out-of-bounds in reuseport_select_sock+0xedc/0x1220 net/core/sock_reuseport.c:596 Read of size 4 at addr ffffc9000051e004 by task slowme/10208 CPU: 6 UID: 1000 PID: 10208 Comm: slowme Not tainted 7.0.0-geb7ac95ff75e #32 PREEMPT(full) Hardware name: QEMU Ubuntu 24.04 PC v2 (i440FX + PIIX, arch_caps fix, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014 Call Trace: dump_stack_lvl+0xe8/0x150 lib/dump_stack.c:120 print_address_description mm/kasan/report.c:378 [inline] print_report+0xca/0x240 mm/kasan/report.c:482 kasan_report+0x118/0x150 mm/kasan/report.c:595 reuseport_select_sock+0xedc/0x1220 net/core/sock_reuseport.c:596 udp4_lib_lookup2+0x3bc/0x950 net/ipv4/udp.c:495 __udp4_lib_lookup+0x768/0xe20 net/ipv4/udp.c:723 __udp4_lib_lookup_skb+0x297/0x390 net/ipv4/udp.c:752 __udp4_lib_rcv+0x1312/0x2620 net/ipv4/udp.c:2752 ip_protocol_deliver_rcu+0x282/0x440 net/ipv4/ip_input.c:207 ip_local_deliver_finish+0x3bb/0x6f0 net/ipv4/ip_input.c:241 NF_HOOK+0x30c/0x3a0 include/linux/netfilter.h:318 NF_HOOK+0x30c/0x3a0 include/linux/netfilter.h:318 __netif_receive_skb_one_core net/core/dev.c:6181 [inline] __netif_receive_skb net/core/dev.c:6294 [inline] process_backlog+0xaa4/0x1960 net/core/dev.c:6645 __napi_poll+0xae/0x340 net/core/dev.c:7709 napi_poll net/core/dev.c:7772 [inline] net_rx_action+0x5d7/0xf50 net/core/dev.c:7929 handle_softirqs+0x22b/0x870 kernel/softirq.c:622 do_softirq+0x76/0xd0 kernel/softirq.c:523 __local_bh_enable_ip+0xf8/0x130 kernel/softirq.c:450 local_bh_enable include/linux/bottom_half.h:33 [inline] rcu_read_unlock_bh include/linux/rcupdate.h:924 [inline] __dev_queue_xmit+0x1dd7/0x3710 net/core/dev.c:4890 neigh_output include/net/neighbour.h:556 [inline] ip_finish_output2+0xca9/0x1070 net/ipv4/ip_output.c:237 NF_HOOK_COND include/linux/netfilter.h:307 [inline] ip_output+0x29f/0x450 net/ipv4/ip_output.c:438 ip_send_skb+0x45/0xc0 net/ipv4/ip_output.c:1508 udp_send_skb+0xb04/0x1510 net/ipv4/udp.c:1195 udp_sendmsg+0x1a71/0x2350 net/ipv4/udp.c:1485 sock_sendmsg_nosec net/socket.c:727 [inline] __sock_sendmsg net/socket.c:742 [inline] __sys_sendto+0x554/0x680 net/socket.c:2206 __do_sys_sendto net/socket.c:2213 [inline] __se_sys_sendto net/socket.c:2209 [inline] __x64_sys_sendto+0xde/0x100 net/socket.c:2209 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x160/0xf80 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x415a2d Code: b3 66 2e 0f 1f 84 00 00 00 00 00 66 90 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007f6bc31e41e8 EFLAGS: 00000212 ORIG_RAX: 000000000000002c RAX: ffffffffffffffda RBX: 00007f6bc31e4cdc RCX: 0000000000415a2d RDX: 0000000000000001 RSI: 00007f6bc31e421f RDI: 0000000000000003 RBP: 00007f6bc31e4240 R08: 00007f6bc31e4220 R09: 0000000000000010 R10: 0000000000000000 R11: 0000000000000212 R12: 00007f6bc31e46c0 R13: ffffffffffffffb8 R14: 0000000000000000 R15: 00007ffc9b0d70b0 Fixes: 538950a1b752 ("soreuseport: setsockopt SO_ATTACH_REUSEPORT_[CE]BPF") Reported-by: Eulgyu Kim Reported-by: Taeyang Lee <0wn@theori.io> Signed-off-by: Kuniyuki Iwashima Signed-off-by: Daniel Borkmann Acked-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20260426012647.3233119-1-kuniyu@google.com --- net/core/filter.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/net/core/filter.c b/net/core/filter.c index ef0877eefaa7..70eef14edb99 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -1654,15 +1654,24 @@ int sk_reuseport_attach_bpf(u32 ufd, struct sock *sk) return err; } +static void sk_reuseport_prog_free_rcu(struct rcu_head *rcu) +{ + struct bpf_prog_aux *aux = container_of(rcu, struct bpf_prog_aux, rcu); + struct bpf_prog *prog = aux->prog; + + bpf_release_orig_filter(prog); + bpf_prog_free(prog); +} + void sk_reuseport_prog_free(struct bpf_prog *prog) { if (!prog) return; - if (prog->type == BPF_PROG_TYPE_SK_REUSEPORT) - bpf_prog_put(prog); + if (bpf_prog_was_classic(prog)) + call_rcu(&prog->aux->rcu, sk_reuseport_prog_free_rcu); else - bpf_prog_destroy(prog); + bpf_prog_put(prog); } static inline int __bpf_try_make_writable(struct sk_buff *skb, From 909f7bf9b080c10df3c3b38533906dbf09ff1d8b Mon Sep 17 00:00:00 2001 From: Lukas Wunner Date: Wed, 15 Apr 2026 17:56:06 +0200 Subject: [PATCH 496/525] PCI: Update saved_config_space upon resource assignment Bernd reports passthrough failure of a Digital Devices Cine S2 V6 DVB adapter plugged into an ASRock X570S PG Riptide board with BIOS version P5.41 (09/07/2023): ddbridge 0000:05:00.0: detected Digital Devices Cine S2 V6 DVB adapter ddbridge 0000:05:00.0: cannot read registers ddbridge 0000:05:00.0: fail BIOS assigns an incorrect BAR to the DVB adapter which doesn't fit into the upstream bridge window. The kernel corrects the BAR assignment: pci 0000:07:00.0: BAR 0 [mem 0xfffffffffc500000-0xfffffffffc50ffff 64bit]: can't claim; no compatible bridge window pci 0000:07:00.0: BAR 0 [mem 0xfc500000-0xfc50ffff 64bit]: assigned Correction of the BAR assignment happens in an x86-specific fs_initcall, pcibios_assign_resources(), after device enumeration in a subsys_initcall. This order was introduced at the behest of Linus in 2004: https://git.kernel.org/tglx/history/c/a06a30144bbc No other architecture performs such a late BAR correction. Bernd bisected the issue to commit a2f1e22390ac ("PCI/ERR: Ensure error recoverability at all times"), but it only occurs in the absence of commit 4d4c10f763d7 ("PCI: Explicitly put devices into D0 when initializing"). This combination exists in stable kernel v6.12.70, but not in mainline, hence Bernd cannot reproduce the issue with mainline. Since a2f1e22390ac, config space is saved on enumeration, prior to BAR correction. Upon passthrough, the corrected BAR is overwritten with the incorrect saved value by: vfio_pci_core_register_device() vfio_pci_set_power_state() pci_restore_state() But only if the device's current_state is PCI_UNKNOWN, as it was prior to commit 4d4c10f763d7. Since the commit, it is PCI_D0, which changes the behavior of vfio_pci_set_power_state() to no longer restore the state without saving it first. Alexandre is reporting the same issue as Bernd, but in his case, mainline is affected as well. The difference is that on Alexandre's system, the host kernel binds a driver to the device which is unbound prior to passthrough, whereas on Bernd's system no driver gets bound by the host kernel. Unbinding sets current_state to PCI_UNKNOWN in pci_device_remove(), so when vfio-pci is subsequently bound to the device, pci_restore_state() is once again called without invoking pci_save_state() first. To robustly fix the issue, always update saved_config_space upon resource assignment. Reported-by: Bernd Schumacher Closes: https://lore.kernel.org/r/acfZrlP0Ua_5D3U4@eldamar.lan/ Reported-by: Alexandre N. Closes: https://lore.kernel.org/r/dd3c3358-de0f-4a56-9c81-04aceaab4058@mailo.com/ Fixes: a2f1e22390ac ("PCI/ERR: Ensure error recoverability at all times") Signed-off-by: Lukas Wunner Signed-off-by: Bjorn Helgaas Tested-by: Bernd Schumacher Tested-by: Alexandre N. Cc: stable@vger.kernel.org # v6.12+ Link: https://patch.msgid.link/febc3f354e0c1f5a9f5b3ee9ffddaa44caccf651.1776268054.git.lukas@wunner.de --- drivers/pci/setup-res.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/pci/setup-res.c b/drivers/pci/setup-res.c index fbc05cda96ee..991d3ed543f5 100644 --- a/drivers/pci/setup-res.c +++ b/drivers/pci/setup-res.c @@ -102,6 +102,7 @@ static void pci_std_update_resource(struct pci_dev *dev, int resno) } pci_write_config_dword(dev, reg, new); + dev->saved_config_space[reg / 4] = new; pci_read_config_dword(dev, reg, &check); if ((new ^ check) & mask) { @@ -112,6 +113,7 @@ static void pci_std_update_resource(struct pci_dev *dev, int resno) if (res->flags & IORESOURCE_MEM_64) { new = region.start >> 16 >> 16; pci_write_config_dword(dev, reg + 4, new); + dev->saved_config_space[(reg + 4) / 4] = new; pci_read_config_dword(dev, reg + 4, &check); if (check != new) { pci_err(dev, "%s: error updating (high %#010x != %#010x)\n", From f45a49a2380a47332817b7248c61a0ebbc6f0d00 Mon Sep 17 00:00:00 2001 From: Samiullah Khawaja Date: Tue, 5 May 2026 23:43:27 +0000 Subject: [PATCH 497/525] PCI: Initialize temporary device in new_id_store() When setting new_id of a PCI device driver using sysfs a lockdep splat occurs. This is because new_id_store() builds a temporary pci_dev for pci_match_device(), which calls device_match_driver_override(). That depends on the driver_override.lock added by cb3d1049f4ea ("driver core: generalize driver_override in struct device"). The new driver_override.lock was not initialized in the temporary pci_dev, resulting in this lockdep splat. Initialize the temporary pci_dev to fix this. Repro: Build with CONFIG_LOCKDEP=y, boot with QEMU, and add a new ID: # echo "8086 10f5" > /sys/bus/pci/drivers/e1000e/new_id INFO: trying to register non-static key. The code is fine but needs lockdep annotation, or maybe you didn't initialize this object before use? turning off the locking correctness validator. CPU: 2 UID: 0 PID: 177 Comm: liveupdate-iomm Not tainted 7.0.0+ #9 PREEMPT(full) Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014 Call Trace: dump_stack_lvl+0x5d/0x80 register_lock_class+0x77e/0x790 lock_acquire+0xbf/0x2e0 pci_match_device+0x24/0x180 new_id_store+0x189/0x1d0 kernfs_fop_write_iter+0x14f/0x210 vfs_write+0x263/0x5e0 ksys_write+0x79/0xf0 do_syscall_64+0x117/0xf80 Fixes: 10a4206a2401 ("PCI: use generic driver_override infrastructure") Fixes: 8895d3bcb8ba ("PCI: Fail new_id for vendor/device values already built into driver") Signed-off-by: Samiullah Khawaja [bhelgaas: add commit log details and repro, trim backtrace] Signed-off-by: Bjorn Helgaas Reviewed-by: Danilo Krummrich Link: https://patch.msgid.link/20260505234327.716630-1-skhawaja@google.com --- drivers/pci/pci-driver.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c index d10ece0889f0..e3f59001785a 100644 --- a/drivers/pci/pci-driver.c +++ b/drivers/pci/pci-driver.c @@ -179,6 +179,11 @@ static const struct pci_device_id *pci_match_device(struct pci_driver *drv, return NULL; } +static void _pci_free_device(struct device *dev) +{ + kfree(to_pci_dev(dev)); +} + /** * new_id_store - sysfs frontend to pci_add_dynid() * @driver: target device driver @@ -214,11 +219,13 @@ static ssize_t new_id_store(struct device_driver *driver, const char *buf, pdev->subsystem_vendor = subvendor; pdev->subsystem_device = subdevice; pdev->class = class; + pdev->dev.release = _pci_free_device; + device_initialize(&pdev->dev); if (pci_match_device(pdrv, pdev)) retval = -EEXIST; - kfree(pdev); + put_device(&pdev->dev); if (retval) return retval; From bf5421b3d8d3d9216d045af89d4f7e2f75375554 Mon Sep 17 00:00:00 2001 From: Marek Vasut Date: Tue, 28 Apr 2026 07:19:54 +0200 Subject: [PATCH 498/525] MAINTAINERS: Update Marek Vasut email for PCIe R-Car Use up to date address. No functional change. Signed-off-by: Marek Vasut Signed-off-by: Bjorn Helgaas Link: https://patch.msgid.link/20260428052030.51101-1-marek.vasut+renesas@mailbox.org --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 2fb1c75afd16..db06875cfb46 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -20433,7 +20433,7 @@ F: drivers/pci/controller/plda/pcie-plda-host.c F: drivers/pci/controller/plda/pcie-plda.h PCI DRIVER FOR RENESAS R-CAR -M: Marek Vasut +M: Marek Vasut M: Yoshihiro Shimoda L: linux-pci@vger.kernel.org L: linux-renesas-soc@vger.kernel.org From 78e115d806b0eea83a0e4df182a689823aa80511 Mon Sep 17 00:00:00 2001 From: Hans Zhang <18255117159@163.com> Date: Fri, 8 May 2026 10:30:06 +0800 Subject: [PATCH 499/525] MAINTAINERS: Update Hans Zhang email for PCIe CIX Sky1 Update my email address as my work email account is no longer in use. Signed-off-by: Hans Zhang <18255117159@163.com> Signed-off-by: Bjorn Helgaas Link: https://patch.msgid.link/20260508023006.1787674-1-18255117159@163.com --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index db06875cfb46..b10ee9a3b5f5 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -20321,7 +20321,7 @@ F: Documentation/devicetree/bindings/pci/cdns,* F: drivers/pci/controller/cadence/*cadence* PCI DRIVER FOR CIX Sky1 -M: Hans Zhang +M: Hans Zhang <18255117159@163.com> L: linux-pci@vger.kernel.org S: Maintained F: Documentation/devicetree/bindings/pci/cix,sky1-pcie-*.yaml From 9ef40a09c5de18f0d275a451f5c2a7fd4f07158b Mon Sep 17 00:00:00 2001 From: Aksh Garg Date: Fri, 8 May 2026 11:39:51 +0530 Subject: [PATCH 500/525] MAINTAINERS: Add Aksh Garg as PCIe CADENCE reviewer I wish to contribute to the review process for Cadence PCIe IP drivers, hence add myself as a reviewer. Signed-off-by: Aksh Garg Signed-off-by: Bjorn Helgaas Link: https://patch.msgid.link/20260508060951.840233-1-a-garg7@ti.com --- MAINTAINERS | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index b10ee9a3b5f5..0cafa37693a6 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -20315,10 +20315,11 @@ F: Documentation/devicetree/bindings/pci/marvell,armada8k-pcie.yaml F: drivers/pci/controller/dwc/pcie-armada8k.c PCI DRIVER FOR CADENCE PCIE IP +R: Aksh Garg L: linux-pci@vger.kernel.org S: Orphan F: Documentation/devicetree/bindings/pci/cdns,* -F: drivers/pci/controller/cadence/*cadence* +F: drivers/pci/controller/cadence/ PCI DRIVER FOR CIX Sky1 M: Hans Zhang <18255117159@163.com> From 512809bb8a370d071f66fc53abe67368e171dec5 Mon Sep 17 00:00:00 2001 From: Paul Chaignon Date: Thu, 7 May 2026 20:22:06 +0200 Subject: [PATCH 501/525] bpf: Don't run arg-tracking analysis twice on main subprog Because subprog 0, the main subprog, is considered a global function, we end up running the arg-tracking dataflow analysis twice on it. That results in slightly longer verification but mostly in more verbose verifier logs. This patch fixes it by keeping only the iteration over global subprogs. When running over all of Cilium's programs with BPF_LOG_LEVEL2, this reduces verbosity by ~20% on average. Fixes: bf0c571f7feb6 ("bpf: introduce forward arg-tracking dataflow analysis") Signed-off-by: Paul Chaignon Acked-by: Eduard Zingerman Link: https://lore.kernel.org/r/e4d7b53d4963ef520541a782f5fc8108a168877c.1778176504.git.paul.chaignon@gmail.com Signed-off-by: Alexei Starovoitov --- kernel/bpf/liveness.c | 25 +++++++------------------ 1 file changed, 7 insertions(+), 18 deletions(-) diff --git a/kernel/bpf/liveness.c b/kernel/bpf/liveness.c index 332e6e003f27..58197d73b120 100644 --- a/kernel/bpf/liveness.c +++ b/kernel/bpf/liveness.c @@ -1914,26 +1914,15 @@ int bpf_compute_subprog_arg_access(struct bpf_verifier_env *env) return -ENOMEM; } - instance = call_instance(env, NULL, 0, 0); - if (IS_ERR(instance)) { - err = PTR_ERR(instance); - goto out; - } - err = analyze_subprog(env, NULL, info, instance, callsites); - if (err) - goto out; - /* - * Subprogs and callbacks that don't receive FP-derived arguments - * cannot access ancestor stack frames, so they were skipped during - * the recursive walk above. Async callbacks (timer, workqueue) are - * also not reachable from the main program's call graph. Analyze - * all unvisited subprogs as independent roots at depth 0. + * Analyze every subprog in reverse topological order (callers + * before callees) so that each subprog is analyzed before its + * callees, allowing the recursive walk inside analyze_subprog() + * to naturally reach callees that receive FP-derived args. * - * Use reverse topological order (callers before callees) so that - * each subprog is analyzed before its callees, allowing the - * recursive walk inside analyze_subprog() to naturally - * reach nested callees that also lack FP-derived args. + * Subprogs and callbacks that don't receive FP-derived arguments + * cannot access ancestor stack frames are analyzed independently. + * Async callbacks (timer, workqueue) are handled the same way. */ for (k = env->subprog_cnt - 1; k >= 0; k--) { int sub = env->subprog_topo_order[k]; From bf6d507f7e3c65751d52fd8caf1ea4e003922624 Mon Sep 17 00:00:00 2001 From: Linpu Yu Date: Fri, 8 May 2026 22:43:43 +0800 Subject: [PATCH 502/525] xskmap: reject TX-only AF_XDP sockets XSKMAP entries are used as redirect targets for incoming XDP frames. A TX-only AF_XDP socket lacks an Rx ring and cannot handle redirected traffic, but xsk_map_update_elem() currently allows such sockets to be inserted into the map. Redirecting packets to such a socket on the veth generic-XDP path causes a kernel crash in xsk_generic_rcv(). This became possible after xsk_is_setup_for_bpf_map() was removed from the XSKMAP update path, which allowed bound TX-only sockets to be inserted into the map. Reject TX-only sockets during XSKMAP updates to avoid the crash. They remain fully operational for pure Tx purposes outside XSKMAP. Fixes: 968be23ceaca ("xsk: Fix possible segfault at xskmap entry insertion") Reported-by: Juefei Pu Reported-by: Yuan Tan Reported-by: Xin Liu Signed-off-by: Yifan Wu Signed-off-by: Linpu Yu Reviewed-by: Jason Xing Link: https://lore.kernel.org/r/20260508144344.694-1-linpu5433@gmail.com Signed-off-by: Alexei Starovoitov --- net/xdp/xskmap.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/xdp/xskmap.c b/net/xdp/xskmap.c index afa457506274..3bff346308d0 100644 --- a/net/xdp/xskmap.c +++ b/net/xdp/xskmap.c @@ -184,6 +184,10 @@ static long xsk_map_update_elem(struct bpf_map *map, void *key, void *value, } xs = (struct xdp_sock *)sock->sk; + if (!READ_ONCE(xs->rx)) { + sockfd_put(sock); + return -ENOBUFS; + } map_entry = &m->xsk_map[i]; node = xsk_map_node_alloc(m, map_entry); From 3ac1a467e37683f602221e243fa3c59b0de81165 Mon Sep 17 00:00:00 2001 From: Junyoung Jang Date: Mon, 27 Apr 2026 02:25:05 +0900 Subject: [PATCH 503/525] bpf: Fix off-by-one boundary validation in arena direct-value access BPF_MAP_TYPE_ARENA accepts BPF_PSEUDO_MAP_VALUE offsets at exactly the end of the arena mapping (off == arena_size). The boundary check in arena_map_direct_value_addr() uses `>` instead of `>=`, which incorrectly allows a one-past-end pointer to be accepted. Change the condition to `>=` to correctly reject offsets that fall outside the valid arena user_vm range. Fixes: 317460317a02 ("bpf: Introduce bpf_arena.") Signed-off-by: Junyoung Jang Reviewed-by: Emil Tsalapatis Link: https://lore.kernel.org/r/20260426172505.1947915-1-graypanda.inzag@gmail.com Signed-off-by: Alexei Starovoitov --- kernel/bpf/arena.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/bpf/arena.c b/kernel/bpf/arena.c index 802656c6fd3c..49a8f7b1beef 100644 --- a/kernel/bpf/arena.c +++ b/kernel/bpf/arena.c @@ -511,7 +511,7 @@ static int arena_map_direct_value_addr(const struct bpf_map *map, u64 *imm, u32 { struct bpf_arena *arena = container_of(map, struct bpf_arena, map); - if ((u64)off > arena->user_vm_end - arena->user_vm_start) + if ((u64)off >= arena->user_vm_end - arena->user_vm_start) return -ERANGE; *imm = (unsigned long)arena->user_vm_start; return 0; From aa54b1d27fe0c2b78e664a34fd0fdf7cd1960d71 Mon Sep 17 00:00:00 2001 From: Hyunwoo Kim Date: Fri, 8 May 2026 17:53:09 +0900 Subject: [PATCH 504/525] rxrpc: Also unshare DATA/RESPONSE packets when paged frags are present The DATA-packet handler in rxrpc_input_call_event() and the RESPONSE handler in rxrpc_verify_response() copy the skb to a linear one before calling into the security ops only when skb_cloned() is true. An skb that is not cloned but still carries externally-owned paged fragments (e.g. SKBFL_SHARED_FRAG set by splice() into a UDP socket via __ip_append_data, or a chained skb_has_frag_list()) falls through to the in-place decryption path, which binds the frag pages directly into the AEAD/skcipher SGL via skb_to_sgvec(). Extend the gate to also unshare when skb_has_frag_list() or skb_has_shared_frag() is true. This catches the splice-loopback vector and other externally-shared frag sources while preserving the zero-copy fast path for skbs whose frags are kernel-private (e.g. NIC page_pool RX, GRO). The OOM/trace handling already in place is reused. Fixes: d0d5c0cd1e71 ("rxrpc: Use skb_unshare() rather than skb_cow_data()") Cc: stable@vger.kernel.org Signed-off-by: Hyunwoo Kim Reviewed-by: Jiayuan Chen Acked-by: David Howells Signed-off-by: Linus Torvalds --- net/rxrpc/call_event.c | 4 +++- net/rxrpc/conn_event.c | 3 ++- 2 files changed, 5 insertions(+), 2 deletions(-) diff --git a/net/rxrpc/call_event.c b/net/rxrpc/call_event.c index fdd683261226..2b19b252225e 100644 --- a/net/rxrpc/call_event.c +++ b/net/rxrpc/call_event.c @@ -334,7 +334,9 @@ bool rxrpc_input_call_event(struct rxrpc_call *call) if (sp->hdr.type == RXRPC_PACKET_TYPE_DATA && sp->hdr.securityIndex != 0 && - skb_cloned(skb)) { + (skb_cloned(skb) || + skb_has_frag_list(skb) || + skb_has_shared_frag(skb))) { /* Unshare the packet so that it can be * modified by in-place decryption. */ diff --git a/net/rxrpc/conn_event.c b/net/rxrpc/conn_event.c index a2130d25aaa9..442414d90ba1 100644 --- a/net/rxrpc/conn_event.c +++ b/net/rxrpc/conn_event.c @@ -245,7 +245,8 @@ static int rxrpc_verify_response(struct rxrpc_connection *conn, { int ret; - if (skb_cloned(skb)) { + if (skb_cloned(skb) || skb_has_frag_list(skb) || + skb_has_shared_frag(skb)) { /* Copy the packet if shared so that we can do in-place * decryption. */ From 5d6919055dec134de3c40167a490f33c74c12581 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 10 May 2026 14:08:09 -0700 Subject: [PATCH 505/525] Linux 7.1-rc3 --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index 93eded47ccd2..b7b80e84e1eb 100644 --- a/Makefile +++ b/Makefile @@ -2,7 +2,7 @@ VERSION = 7 PATCHLEVEL = 1 SUBLEVEL = 0 -EXTRAVERSION = -rc2 +EXTRAVERSION = -rc3 NAME = Baby Opossum Posse # *DOCUMENTATION* From 4af89d7d8552a1f0437521acf89fa51601bce973 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:28:50 +0200 Subject: [PATCH 506/525] spi: pic32: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-2-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-pic32.c | 11 +++-------- 1 file changed, 3 insertions(+), 8 deletions(-) diff --git a/drivers/spi/spi-pic32.c b/drivers/spi/spi-pic32.c index 70427e529945..972128271e4b 100644 --- a/drivers/spi/spi-pic32.c +++ b/drivers/spi/spi-pic32.c @@ -752,7 +752,7 @@ static int pic32_spi_probe(struct platform_device *pdev) struct pic32_spi *pic32s; int ret; - host = spi_alloc_host(&pdev->dev, sizeof(*pic32s)); + host = devm_spi_alloc_host(&pdev->dev, sizeof(*pic32s)); if (!host) return -ENOMEM; @@ -761,7 +761,7 @@ static int pic32_spi_probe(struct platform_device *pdev) ret = pic32_spi_hw_probe(pdev, pic32s); if (ret) - goto err_host; + return ret; host->dev.of_node = pdev->dev.of_node; host->mode_bits = SPI_MODE_3 | SPI_MODE_0 | SPI_CS_HIGH; @@ -833,8 +833,7 @@ static int pic32_spi_probe(struct platform_device *pdev) err_bailout: pic32_spi_dma_unprep(pic32s); -err_host: - spi_controller_put(host); + return ret; } @@ -842,14 +841,10 @@ static void pic32_spi_remove(struct platform_device *pdev) { struct pic32_spi *pic32s = platform_get_drvdata(pdev); - spi_controller_get(pic32s->host); - spi_unregister_controller(pic32s->host); pic32_spi_disable(pic32s); pic32_spi_dma_unprep(pic32s); - - spi_controller_put(pic32s->host); } static const struct of_device_id pic32_spi_of_match[] = { From dda3a77e1a32b329d3c543a1ac236106acf64ec5 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:28:51 +0200 Subject: [PATCH 507/525] spi: pic32-sqi: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-3-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-pic32-sqi.c | 28 ++++++++-------------------- 1 file changed, 8 insertions(+), 20 deletions(-) diff --git a/drivers/spi/spi-pic32-sqi.c b/drivers/spi/spi-pic32-sqi.c index 41662992dbe5..5d3921e29461 100644 --- a/drivers/spi/spi-pic32-sqi.c +++ b/drivers/spi/spi-pic32-sqi.c @@ -572,7 +572,7 @@ static int pic32_sqi_probe(struct platform_device *pdev) struct pic32_sqi *sqi; int ret; - host = spi_alloc_host(&pdev->dev, sizeof(*sqi)); + host = devm_spi_alloc_host(&pdev->dev, sizeof(*sqi)); if (!host) return -ENOMEM; @@ -580,31 +580,25 @@ static int pic32_sqi_probe(struct platform_device *pdev) sqi->host = host; sqi->regs = devm_platform_ioremap_resource(pdev, 0); - if (IS_ERR(sqi->regs)) { - ret = PTR_ERR(sqi->regs); - goto err_free_host; - } + if (IS_ERR(sqi->regs)) + return PTR_ERR(sqi->regs); /* irq */ sqi->irq = platform_get_irq(pdev, 0); - if (sqi->irq < 0) { - ret = sqi->irq; - goto err_free_host; - } + if (sqi->irq < 0) + return sqi->irq; /* clocks */ sqi->sys_clk = devm_clk_get_enabled(&pdev->dev, "reg_ck"); if (IS_ERR(sqi->sys_clk)) { - ret = PTR_ERR(sqi->sys_clk); dev_err(&pdev->dev, "no sys_clk ?\n"); - goto err_free_host; + return PTR_ERR(sqi->sys_clk); } sqi->base_clk = devm_clk_get_enabled(&pdev->dev, "spi_ck"); if (IS_ERR(sqi->base_clk)) { - ret = PTR_ERR(sqi->base_clk); dev_err(&pdev->dev, "no base clk ?\n"); - goto err_free_host; + return PTR_ERR(sqi->base_clk); } init_completion(&sqi->xfer_done); @@ -616,7 +610,7 @@ static int pic32_sqi_probe(struct platform_device *pdev) ret = ring_desc_ring_alloc(sqi); if (ret) { dev_err(&pdev->dev, "ring alloc failed\n"); - goto err_free_host; + return ret; } /* install irq handlers */ @@ -656,8 +650,6 @@ static int pic32_sqi_probe(struct platform_device *pdev) err_free_ring: ring_desc_ring_free(sqi); -err_free_host: - spi_controller_put(host); return ret; } @@ -665,15 +657,11 @@ static void pic32_sqi_remove(struct platform_device *pdev) { struct pic32_sqi *sqi = platform_get_drvdata(pdev); - spi_controller_get(sqi->host); - spi_unregister_controller(sqi->host); /* release resources */ free_irq(sqi->irq, sqi); ring_desc_ring_free(sqi); - - spi_controller_put(sqi->host); } static const struct of_device_id pic32_sqi_of_ids[] = { From 02efc5557c8e4202b1c5d260ec532986e9769897 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:28:52 +0200 Subject: [PATCH 508/525] spi: pl022: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Reviewed-by: Linus Walleij Link: https://patch.msgid.link/20260505072909.618363-4-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-pl022.c | 11 +++-------- 1 file changed, 3 insertions(+), 8 deletions(-) diff --git a/drivers/spi/spi-pl022.c b/drivers/spi/spi-pl022.c index 9c0211f94fd0..95652df5fd09 100644 --- a/drivers/spi/spi-pl022.c +++ b/drivers/spi/spi-pl022.c @@ -1868,7 +1868,7 @@ static int pl022_probe(struct amba_device *adev, const struct amba_id *id) } /* Allocate host with space for data */ - host = spi_alloc_host(dev, sizeof(struct pl022)); + host = devm_spi_alloc_host(dev, sizeof(struct pl022)); if (host == NULL) { dev_err(&adev->dev, "probe - cannot alloc SPI host\n"); return -ENOMEM; @@ -1907,7 +1907,7 @@ static int pl022_probe(struct amba_device *adev, const struct amba_id *id) status = amba_request_regions(adev, NULL); if (status) - goto err_no_ioregion; + return status; pl022->phybase = adev->res.start; pl022->virtbase = devm_ioremap(dev, adev->res.start, @@ -1984,8 +1984,7 @@ static int pl022_probe(struct amba_device *adev, const struct amba_id *id) err_no_clk: err_no_ioremap: amba_release_regions(adev); - err_no_ioregion: - spi_controller_put(host); + return status; } @@ -1997,8 +1996,6 @@ pl022_remove(struct amba_device *adev) if (!pl022) return; - spi_controller_get(pl022->host); - spi_unregister_controller(pl022->host); /* @@ -2012,8 +2009,6 @@ pl022_remove(struct amba_device *adev) pl022_dma_remove(pl022); amba_release_regions(adev); - - spi_controller_put(pl022->host); } #ifdef CONFIG_PM_SLEEP From 86e8160240af1143e0a9f185e45ac300fe0d93a6 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:28:53 +0200 Subject: [PATCH 509/525] spi: qup: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-5-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-qup.c | 15 ++++----------- 1 file changed, 4 insertions(+), 11 deletions(-) diff --git a/drivers/spi/spi-qup.c b/drivers/spi/spi-qup.c index 45d9b4cb75e4..4df01ef2e662 100644 --- a/drivers/spi/spi-qup.c +++ b/drivers/spi/spi-qup.c @@ -1071,11 +1071,9 @@ static int spi_qup_probe(struct platform_device *pdev) if (ret && ret != -ENODEV) return dev_err_probe(dev, ret, "invalid OPP table\n"); - host = spi_alloc_host(dev, sizeof(struct spi_qup)); - if (!host) { - dev_err(dev, "cannot allocate host\n"); + host = devm_spi_alloc_host(dev, sizeof(struct spi_qup)); + if (!host) return -ENOMEM; - } /* use num-cs unless not present or out of range */ if (of_property_read_u32(dev->of_node, "num-cs", &num_cs) || @@ -1108,7 +1106,7 @@ static int spi_qup_probe(struct platform_device *pdev) ret = spi_qup_init_dma(host, res->start); if (ret == -EPROBE_DEFER) - goto error; + return ret; else if (!ret) host->can_dma = spi_qup_can_dma; @@ -1206,8 +1204,7 @@ static int spi_qup_probe(struct platform_device *pdev) clk_disable_unprepare(iclk); error_dma: spi_qup_release_dma(host); -error: - spi_controller_put(host); + return ret; } @@ -1320,8 +1317,6 @@ static void spi_qup_remove(struct platform_device *pdev) struct spi_qup *controller = spi_controller_get_devdata(host); int ret; - spi_controller_get(host); - spi_unregister_controller(host); ret = pm_runtime_get_sync(&pdev->dev); @@ -1343,8 +1338,6 @@ static void spi_qup_remove(struct platform_device *pdev) pm_runtime_put_noidle(&pdev->dev); pm_runtime_disable(&pdev->dev); - - spi_controller_put(host); } static const struct of_device_id spi_qup_dt_match[] = { From 368d0e6c6f82a090b3fb080929f4478217e597e8 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:28:54 +0200 Subject: [PATCH 510/525] spi: rspi: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-6-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-rspi.c | 19 +++++-------------- 1 file changed, 5 insertions(+), 14 deletions(-) diff --git a/drivers/spi/spi-rspi.c b/drivers/spi/spi-rspi.c index a8180dece716..951e9a8af547 100644 --- a/drivers/spi/spi-rspi.c +++ b/drivers/spi/spi-rspi.c @@ -1171,14 +1171,10 @@ static void rspi_remove(struct platform_device *pdev) { struct rspi_data *rspi = platform_get_drvdata(pdev); - spi_controller_get(rspi->ctlr); - spi_unregister_controller(rspi->ctlr); rspi_release_dma(rspi->ctlr); pm_runtime_disable(&pdev->dev); - - spi_controller_put(rspi->ctlr); } static const struct spi_ops rspi_ops = { @@ -1294,7 +1290,7 @@ static int rspi_probe(struct platform_device *pdev) const struct spi_ops *ops; unsigned long clksrc; - ctlr = spi_alloc_host(&pdev->dev, sizeof(struct rspi_data)); + ctlr = devm_spi_alloc_host(&pdev->dev, sizeof(struct rspi_data)); if (ctlr == NULL) return -ENOMEM; @@ -1302,7 +1298,7 @@ static int rspi_probe(struct platform_device *pdev) if (ops) { ret = rspi_parse_dt(&pdev->dev, ctlr); if (ret) - goto error1; + return ret; } else { ops = (struct spi_ops *)pdev->id_entry->driver_data; ctlr->num_chipselect = 2; /* default */ @@ -1314,16 +1310,13 @@ static int rspi_probe(struct platform_device *pdev) rspi->ctlr = ctlr; rspi->addr = devm_platform_get_and_ioremap_resource(pdev, 0, &res); - if (IS_ERR(rspi->addr)) { - ret = PTR_ERR(rspi->addr); - goto error1; - } + if (IS_ERR(rspi->addr)) + return PTR_ERR(rspi->addr); rspi->clk = devm_clk_get(&pdev->dev, NULL); if (IS_ERR(rspi->clk)) { dev_err(&pdev->dev, "cannot get clock\n"); - ret = PTR_ERR(rspi->clk); - goto error1; + return PTR_ERR(rspi->clk); } rspi->pdev = pdev; @@ -1396,8 +1389,6 @@ static int rspi_probe(struct platform_device *pdev) rspi_release_dma(ctlr); error2: pm_runtime_disable(&pdev->dev); -error1: - spi_controller_put(ctlr); return ret; } From 042414e4da73e67d0e1e77ac1292e6aa15a54928 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:28:55 +0200 Subject: [PATCH 511/525] spi: sh-hspi: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-7-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-sh-hspi.c | 11 ++--------- 1 file changed, 2 insertions(+), 9 deletions(-) diff --git a/drivers/spi/spi-sh-hspi.c b/drivers/spi/spi-sh-hspi.c index 1e3ca718ca73..f840467cfdb2 100644 --- a/drivers/spi/spi-sh-hspi.c +++ b/drivers/spi/spi-sh-hspi.c @@ -224,15 +224,14 @@ static int hspi_probe(struct platform_device *pdev) return -EINVAL; } - ctlr = spi_alloc_host(&pdev->dev, sizeof(*hspi)); + ctlr = devm_spi_alloc_host(&pdev->dev, sizeof(*hspi)); if (!ctlr) return -ENOMEM; clk = clk_get(&pdev->dev, NULL); if (IS_ERR(clk)) { dev_err(&pdev->dev, "couldn't get clock\n"); - ret = -EINVAL; - goto error0; + return PTR_ERR(clk); } hspi = spi_controller_get_devdata(ctlr); @@ -269,8 +268,6 @@ static int hspi_probe(struct platform_device *pdev) pm_runtime_disable(&pdev->dev); error1: clk_put(clk); - error0: - spi_controller_put(ctlr); return ret; } @@ -279,15 +276,11 @@ static void hspi_remove(struct platform_device *pdev) { struct hspi_priv *hspi = platform_get_drvdata(pdev); - spi_controller_get(hspi->ctlr); - spi_unregister_controller(hspi->ctlr); pm_runtime_disable(&pdev->dev); clk_put(hspi->clk); - - spi_controller_put(hspi->ctlr); } static const struct of_device_id hspi_of_match[] = { From 354b0a4ad4eb50a947b1b7b143c01a01a37489e7 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:28:56 +0200 Subject: [PATCH 512/525] spi: sh-msiof: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-8-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-sh-msiof.c | 28 +++++++++------------------- 1 file changed, 9 insertions(+), 19 deletions(-) diff --git a/drivers/spi/spi-sh-msiof.c b/drivers/spi/spi-sh-msiof.c index f114b6313f4f..070e16bc764f 100644 --- a/drivers/spi/spi-sh-msiof.c +++ b/drivers/spi/spi-sh-msiof.c @@ -1215,9 +1215,9 @@ static int sh_msiof_spi_probe(struct platform_device *pdev) info->dtdl = 200; if (info->mode == MSIOF_SPI_TARGET) - ctlr = spi_alloc_target(dev, sizeof(struct sh_msiof_spi_priv)); + ctlr = devm_spi_alloc_target(dev, sizeof(struct sh_msiof_spi_priv)); else - ctlr = spi_alloc_host(dev, sizeof(struct sh_msiof_spi_priv)); + ctlr = devm_spi_alloc_host(dev, sizeof(struct sh_msiof_spi_priv)); if (ctlr == NULL) return -ENOMEM; @@ -1234,26 +1234,21 @@ static int sh_msiof_spi_probe(struct platform_device *pdev) p->clk = devm_clk_get(dev, NULL); if (IS_ERR(p->clk)) { dev_err(dev, "cannot get clock\n"); - ret = PTR_ERR(p->clk); - goto err1; + return PTR_ERR(p->clk); } i = platform_get_irq(pdev, 0); - if (i < 0) { - ret = i; - goto err1; - } + if (i < 0) + return i; p->mapbase = devm_platform_ioremap_resource(pdev, 0); - if (IS_ERR(p->mapbase)) { - ret = PTR_ERR(p->mapbase); - goto err1; - } + if (IS_ERR(p->mapbase)) + return PTR_ERR(p->mapbase); ret = devm_request_irq(dev, i, sh_msiof_spi_irq, 0, dev_name(dev), p); if (ret) { dev_err(dev, "unable to request irq\n"); - goto err1; + return ret; } p->pdev = pdev; @@ -1300,8 +1295,7 @@ static int sh_msiof_spi_probe(struct platform_device *pdev) err2: sh_msiof_release_dma(p); pm_runtime_disable(dev); - err1: - spi_controller_put(ctlr); + return ret; } @@ -1309,14 +1303,10 @@ static void sh_msiof_spi_remove(struct platform_device *pdev) { struct sh_msiof_spi_priv *p = platform_get_drvdata(pdev); - spi_controller_get(p->ctlr); - spi_unregister_controller(p->ctlr); sh_msiof_release_dma(p); pm_runtime_disable(&pdev->dev); - - spi_controller_put(p->ctlr); } static const struct platform_device_id spi_driver_ids[] = { From fd260013577d0a6f3b6a8b07d059c34fb710efac Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:28:57 +0200 Subject: [PATCH 513/525] spi: sifive: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-9-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-sifive.c | 39 +++++++++++---------------------------- 1 file changed, 11 insertions(+), 28 deletions(-) diff --git a/drivers/spi/spi-sifive.c b/drivers/spi/spi-sifive.c index 74a3e32fd2b5..cee4d92e46f4 100644 --- a/drivers/spi/spi-sifive.c +++ b/drivers/spi/spi-sifive.c @@ -296,7 +296,7 @@ static int sifive_spi_probe(struct platform_device *pdev) u32 cs_bits, max_bits_per_word; struct spi_controller *host; - host = spi_alloc_host(&pdev->dev, sizeof(struct sifive_spi)); + host = devm_spi_alloc_host(&pdev->dev, sizeof(struct sifive_spi)); if (!host) { dev_err(&pdev->dev, "out of memory\n"); return -ENOMEM; @@ -307,24 +307,19 @@ static int sifive_spi_probe(struct platform_device *pdev) platform_set_drvdata(pdev, host); spi->regs = devm_platform_ioremap_resource(pdev, 0); - if (IS_ERR(spi->regs)) { - ret = PTR_ERR(spi->regs); - goto put_host; - } + if (IS_ERR(spi->regs)) + return PTR_ERR(spi->regs); /* Spin up the bus clock before hitting registers */ spi->clk = devm_clk_get_enabled(&pdev->dev, NULL); if (IS_ERR(spi->clk)) { dev_err(&pdev->dev, "Unable to find bus clock\n"); - ret = PTR_ERR(spi->clk); - goto put_host; + return PTR_ERR(spi->clk); } irq = platform_get_irq(pdev, 0); - if (irq < 0) { - ret = irq; - goto put_host; - } + if (irq < 0) + return irq; /* Optional parameters */ ret = @@ -339,8 +334,7 @@ static int sifive_spi_probe(struct platform_device *pdev) if (!ret && max_bits_per_word < 8) { dev_err(&pdev->dev, "Only 8bit SPI words supported by the driver\n"); - ret = -EINVAL; - goto put_host; + return -EINVAL; } /* probe the number of CS lines */ @@ -350,15 +344,13 @@ static int sifive_spi_probe(struct platform_device *pdev) sifive_spi_write(spi, SIFIVE_SPI_REG_CSDEF, spi->cs_inactive); if (!cs_bits) { dev_err(&pdev->dev, "Could not auto probe CS lines\n"); - ret = -EINVAL; - goto put_host; + return -EINVAL; } num_cs = ilog2(cs_bits) + 1; if (num_cs > SIFIVE_SPI_MAX_CS) { dev_err(&pdev->dev, "Invalid number of spi targets\n"); - ret = -EINVAL; - goto put_host; + return -EINVAL; } /* Define our host */ @@ -386,7 +378,7 @@ static int sifive_spi_probe(struct platform_device *pdev) dev_name(&pdev->dev), spi); if (ret) { dev_err(&pdev->dev, "Unable to bind to interrupt\n"); - goto put_host; + return ret; } dev_info(&pdev->dev, "mapped; irq=%d, cs=%d\n", @@ -395,15 +387,10 @@ static int sifive_spi_probe(struct platform_device *pdev) ret = spi_register_controller(host); if (ret < 0) { dev_err(&pdev->dev, "spi_register_host failed\n"); - goto put_host; + return ret; } return 0; - -put_host: - spi_controller_put(host); - - return ret; } static void sifive_spi_remove(struct platform_device *pdev) @@ -411,14 +398,10 @@ static void sifive_spi_remove(struct platform_device *pdev) struct spi_controller *host = platform_get_drvdata(pdev); struct sifive_spi *spi = spi_controller_get_devdata(host); - spi_controller_get(host); - spi_unregister_controller(host); /* Disable all the interrupts just in case */ sifive_spi_write(spi, SIFIVE_SPI_REG_IE, 0); - - spi_controller_put(host); } static int sifive_spi_suspend(struct device *dev) From cd1cd2ff56bf106a71d3170d6b74d08386d99428 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:28:58 +0200 Subject: [PATCH 514/525] spi: slave-mt27xx: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-10-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-slave-mt27xx.c | 33 ++++++++++----------------------- 1 file changed, 10 insertions(+), 23 deletions(-) diff --git a/drivers/spi/spi-slave-mt27xx.c b/drivers/spi/spi-slave-mt27xx.c index 7aedeaa5889d..e60ab4c18bed 100644 --- a/drivers/spi/spi-slave-mt27xx.c +++ b/drivers/spi/spi-slave-mt27xx.c @@ -388,11 +388,9 @@ static int mtk_spi_slave_probe(struct platform_device *pdev) int irq, ret; const struct of_device_id *of_id; - ctlr = spi_alloc_target(&pdev->dev, sizeof(*mdata)); - if (!ctlr) { - dev_err(&pdev->dev, "failed to alloc spi target\n"); + ctlr = devm_spi_alloc_target(&pdev->dev, sizeof(*mdata)); + if (!ctlr) return -ENOMEM; - } ctlr->auto_runtime_pm = true; ctlr->mode_bits = SPI_CPOL | SPI_CPHA; @@ -406,8 +404,7 @@ static int mtk_spi_slave_probe(struct platform_device *pdev) of_id = of_match_node(mtk_spi_slave_of_match, pdev->dev.of_node); if (!of_id) { dev_err(&pdev->dev, "failed to probe of_node\n"); - ret = -EINVAL; - goto err_put_ctlr; + return -EINVAL; } mdata = spi_controller_get_devdata(ctlr); mdata->dev_comp = of_id->data; @@ -420,35 +417,31 @@ static int mtk_spi_slave_probe(struct platform_device *pdev) init_completion(&mdata->xfer_done); mdata->dev = &pdev->dev; mdata->base = devm_platform_ioremap_resource(pdev, 0); - if (IS_ERR(mdata->base)) { - ret = PTR_ERR(mdata->base); - goto err_put_ctlr; - } + if (IS_ERR(mdata->base)) + return PTR_ERR(mdata->base); irq = platform_get_irq(pdev, 0); - if (irq < 0) { - ret = irq; - goto err_put_ctlr; - } + if (irq < 0) + return irq; ret = devm_request_irq(&pdev->dev, irq, mtk_spi_slave_interrupt, IRQF_TRIGGER_NONE, dev_name(&pdev->dev), ctlr); if (ret) { dev_err(&pdev->dev, "failed to register irq (%d)\n", ret); - goto err_put_ctlr; + return ret; } mdata->spi_clk = devm_clk_get(&pdev->dev, "spi"); if (IS_ERR(mdata->spi_clk)) { ret = PTR_ERR(mdata->spi_clk); dev_err(&pdev->dev, "failed to get spi-clk: %d\n", ret); - goto err_put_ctlr; + return ret; } ret = clk_prepare_enable(mdata->spi_clk); if (ret < 0) { dev_err(&pdev->dev, "failed to enable spi_clk (%d)\n", ret); - goto err_put_ctlr; + return ret; } pm_runtime_enable(&pdev->dev); @@ -465,8 +458,6 @@ static int mtk_spi_slave_probe(struct platform_device *pdev) err_disable_runtime_pm: pm_runtime_disable(&pdev->dev); -err_put_ctlr: - spi_controller_put(ctlr); return ret; } @@ -475,13 +466,9 @@ static void mtk_spi_slave_remove(struct platform_device *pdev) { struct spi_controller *ctlr = platform_get_drvdata(pdev); - spi_controller_get(ctlr); - spi_unregister_controller(ctlr); pm_runtime_disable(&pdev->dev); - - spi_controller_put(ctlr); } #ifdef CONFIG_PM_SLEEP From d68627cc76cd895d07363c795a198d1edd687107 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:28:59 +0200 Subject: [PATCH 515/525] spi: sprd: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-11-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-sprd.c | 20 ++++++-------------- 1 file changed, 6 insertions(+), 14 deletions(-) diff --git a/drivers/spi/spi-sprd.c b/drivers/spi/spi-sprd.c index fd3fd0ce122c..d27b51698dbd 100644 --- a/drivers/spi/spi-sprd.c +++ b/drivers/spi/spi-sprd.c @@ -923,16 +923,14 @@ static int sprd_spi_probe(struct platform_device *pdev) int ret; pdev->id = of_alias_get_id(pdev->dev.of_node, "spi"); - sctlr = spi_alloc_host(&pdev->dev, sizeof(*ss)); + sctlr = devm_spi_alloc_host(&pdev->dev, sizeof(*ss)); if (!sctlr) return -ENOMEM; ss = spi_controller_get_devdata(sctlr); ss->base = devm_platform_get_and_ioremap_resource(pdev, 0, &res); - if (IS_ERR(ss->base)) { - ret = PTR_ERR(ss->base); - goto free_controller; - } + if (IS_ERR(ss->base)) + return PTR_ERR(ss->base); ss->phy_base = res->start; ss->dev = &pdev->dev; @@ -949,15 +947,15 @@ static int sprd_spi_probe(struct platform_device *pdev) platform_set_drvdata(pdev, sctlr); ret = sprd_spi_clk_init(pdev, ss); if (ret) - goto free_controller; + return ret; ret = sprd_spi_irq_init(pdev, ss); if (ret) - goto free_controller; + return ret; ret = sprd_spi_dma_init(pdev, ss); if (ret) - goto free_controller; + return ret; ret = clk_prepare_enable(ss->clk); if (ret) @@ -992,8 +990,6 @@ static int sprd_spi_probe(struct platform_device *pdev) clk_disable_unprepare(ss->clk); release_dma: sprd_spi_dma_release(ss); -free_controller: - spi_controller_put(sctlr); return ret; } @@ -1008,8 +1004,6 @@ static void sprd_spi_remove(struct platform_device *pdev) if (ret < 0) dev_err(ss->dev, "failed to resume SPI controller\n"); - spi_controller_get(sctlr); - spi_unregister_controller(sctlr); if (ret >= 0) { @@ -1019,8 +1013,6 @@ static void sprd_spi_remove(struct platform_device *pdev) } pm_runtime_put_noidle(&pdev->dev); pm_runtime_disable(&pdev->dev); - - spi_controller_put(sctlr); } static int __maybe_unused sprd_spi_runtime_suspend(struct device *dev) From d3cf5ebdf1c9fa909ae8de5be041eac2658c0c68 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:29:00 +0200 Subject: [PATCH 516/525] spi: st-ssc4: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-12-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-st-ssc4.c | 14 ++++---------- 1 file changed, 4 insertions(+), 10 deletions(-) diff --git a/drivers/spi/spi-st-ssc4.c b/drivers/spi/spi-st-ssc4.c index 9c8099fe6e19..df61094fc444 100644 --- a/drivers/spi/spi-st-ssc4.c +++ b/drivers/spi/spi-st-ssc4.c @@ -279,7 +279,7 @@ static int spi_st_probe(struct platform_device *pdev) int irq, ret = 0; u32 var; - host = spi_alloc_host(&pdev->dev, sizeof(*spi_st)); + host = devm_spi_alloc_host(&pdev->dev, sizeof(*spi_st)); if (!host) return -ENOMEM; @@ -296,13 +296,12 @@ static int spi_st_probe(struct platform_device *pdev) spi_st->clk = devm_clk_get(&pdev->dev, "ssc"); if (IS_ERR(spi_st->clk)) { dev_err(&pdev->dev, "Unable to request clock\n"); - ret = PTR_ERR(spi_st->clk); - goto put_host; + return PTR_ERR(spi_st->clk); } ret = clk_prepare_enable(spi_st->clk); if (ret) - goto put_host; + return ret; init_completion(&spi_st->done); @@ -361,8 +360,7 @@ static int spi_st_probe(struct platform_device *pdev) pm_runtime_disable(&pdev->dev); clk_disable: clk_disable_unprepare(spi_st->clk); -put_host: - spi_controller_put(host); + return ret; } @@ -371,16 +369,12 @@ static void spi_st_remove(struct platform_device *pdev) struct spi_controller *host = platform_get_drvdata(pdev); struct spi_st *spi_st = spi_controller_get_devdata(host); - spi_controller_get(host); - spi_unregister_controller(host); pm_runtime_disable(&pdev->dev); clk_disable_unprepare(spi_st->clk); - spi_controller_put(host); - pinctrl_pm_select_sleep_state(&pdev->dev); } From 02b36d644ded410e8f70cdd78d0c5d523252bafd Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:29:01 +0200 Subject: [PATCH 517/525] spi: sun4i: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-13-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-sun4i.c | 35 +++++++++++------------------------ 1 file changed, 11 insertions(+), 24 deletions(-) diff --git a/drivers/spi/spi-sun4i.c b/drivers/spi/spi-sun4i.c index b7fbb5270edb..d5c16392cd4d 100644 --- a/drivers/spi/spi-sun4i.c +++ b/drivers/spi/spi-sun4i.c @@ -434,32 +434,26 @@ static int sun4i_spi_probe(struct platform_device *pdev) struct sun4i_spi *sspi; int ret = 0, irq; - host = spi_alloc_host(&pdev->dev, sizeof(struct sun4i_spi)); - if (!host) { - dev_err(&pdev->dev, "Unable to allocate SPI Host\n"); + host = devm_spi_alloc_host(&pdev->dev, sizeof(struct sun4i_spi)); + if (!host) return -ENOMEM; - } platform_set_drvdata(pdev, host); sspi = spi_controller_get_devdata(host); sspi->base_addr = devm_platform_ioremap_resource(pdev, 0); - if (IS_ERR(sspi->base_addr)) { - ret = PTR_ERR(sspi->base_addr); - goto err_free_host; - } + if (IS_ERR(sspi->base_addr)) + return PTR_ERR(sspi->base_addr); irq = platform_get_irq(pdev, 0); - if (irq < 0) { - ret = -ENXIO; - goto err_free_host; - } + if (irq < 0) + return -ENXIO; ret = devm_request_irq(&pdev->dev, irq, sun4i_spi_handler, 0, "sun4i-spi", sspi); if (ret) { dev_err(&pdev->dev, "Cannot request IRQ\n"); - goto err_free_host; + return ret; } sspi->host = host; @@ -477,15 +471,13 @@ static int sun4i_spi_probe(struct platform_device *pdev) sspi->hclk = devm_clk_get(&pdev->dev, "ahb"); if (IS_ERR(sspi->hclk)) { dev_err(&pdev->dev, "Unable to acquire AHB clock\n"); - ret = PTR_ERR(sspi->hclk); - goto err_free_host; + return PTR_ERR(sspi->hclk); } sspi->mclk = devm_clk_get(&pdev->dev, "mod"); if (IS_ERR(sspi->mclk)) { dev_err(&pdev->dev, "Unable to acquire module clock\n"); - ret = PTR_ERR(sspi->mclk); - goto err_free_host; + return PTR_ERR(sspi->mclk); } init_completion(&sspi->done); @@ -497,7 +489,7 @@ static int sun4i_spi_probe(struct platform_device *pdev) ret = sun4i_spi_runtime_resume(&pdev->dev); if (ret) { dev_err(&pdev->dev, "Couldn't resume the device\n"); - goto err_free_host; + return ret; } pm_runtime_set_active(&pdev->dev); @@ -515,8 +507,7 @@ static int sun4i_spi_probe(struct platform_device *pdev) err_pm_disable: pm_runtime_disable(&pdev->dev); sun4i_spi_runtime_suspend(&pdev->dev); -err_free_host: - spi_controller_put(host); + return ret; } @@ -524,13 +515,9 @@ static void sun4i_spi_remove(struct platform_device *pdev) { struct spi_controller *host = platform_get_drvdata(pdev); - spi_controller_get(host); - spi_unregister_controller(host); pm_runtime_force_suspend(&pdev->dev); - - spi_controller_put(host); } static const struct of_device_id sun4i_spi_match[] = { From 9864636b1cd95dd2b3c702a2ff22e9d169f6523a Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:29:02 +0200 Subject: [PATCH 518/525] spi: sun6i: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-14-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-sun6i.c | 38 ++++++++++++-------------------------- 1 file changed, 12 insertions(+), 26 deletions(-) diff --git a/drivers/spi/spi-sun6i.c b/drivers/spi/spi-sun6i.c index 5ac73d324d06..4631e9c7ca1d 100644 --- a/drivers/spi/spi-sun6i.c +++ b/drivers/spi/spi-sun6i.c @@ -633,7 +633,7 @@ static int sun6i_spi_probe(struct platform_device *pdev) struct resource *mem; int ret = 0, irq; - host = spi_alloc_host(&pdev->dev, sizeof(struct sun6i_spi)); + host = devm_spi_alloc_host(&pdev->dev, sizeof(struct sun6i_spi)); if (!host) { dev_err(&pdev->dev, "Unable to allocate SPI Host\n"); return -ENOMEM; @@ -643,22 +643,18 @@ static int sun6i_spi_probe(struct platform_device *pdev) sspi = spi_controller_get_devdata(host); sspi->base_addr = devm_platform_get_and_ioremap_resource(pdev, 0, &mem); - if (IS_ERR(sspi->base_addr)) { - ret = PTR_ERR(sspi->base_addr); - goto err_free_host; - } + if (IS_ERR(sspi->base_addr)) + return PTR_ERR(sspi->base_addr); irq = platform_get_irq(pdev, 0); - if (irq < 0) { - ret = -ENXIO; - goto err_free_host; - } + if (irq < 0) + return -ENXIO; ret = devm_request_irq(&pdev->dev, irq, sun6i_spi_handler, 0, "sun6i-spi", sspi); if (ret) { dev_err(&pdev->dev, "Cannot request IRQ\n"); - goto err_free_host; + return ret; } sspi->host = host; @@ -679,15 +675,13 @@ static int sun6i_spi_probe(struct platform_device *pdev) sspi->hclk = devm_clk_get(&pdev->dev, "ahb"); if (IS_ERR(sspi->hclk)) { dev_err(&pdev->dev, "Unable to acquire AHB clock\n"); - ret = PTR_ERR(sspi->hclk); - goto err_free_host; + return PTR_ERR(sspi->hclk); } sspi->mclk = devm_clk_get(&pdev->dev, "mod"); if (IS_ERR(sspi->mclk)) { dev_err(&pdev->dev, "Unable to acquire module clock\n"); - ret = PTR_ERR(sspi->mclk); - goto err_free_host; + return PTR_ERR(sspi->mclk); } init_completion(&sspi->done); @@ -696,17 +690,14 @@ static int sun6i_spi_probe(struct platform_device *pdev) sspi->rstc = devm_reset_control_get_exclusive(&pdev->dev, NULL); if (IS_ERR(sspi->rstc)) { dev_err(&pdev->dev, "Couldn't get reset controller\n"); - ret = PTR_ERR(sspi->rstc); - goto err_free_host; + return PTR_ERR(sspi->rstc); } host->dma_tx = dma_request_chan(&pdev->dev, "tx"); if (IS_ERR(host->dma_tx)) { /* Check tx to see if we need defer probing driver */ - if (PTR_ERR(host->dma_tx) == -EPROBE_DEFER) { - ret = -EPROBE_DEFER; - goto err_free_host; - } + if (PTR_ERR(host->dma_tx) == -EPROBE_DEFER) + return -EPROBE_DEFER; dev_warn(&pdev->dev, "Failed to request TX DMA channel\n"); host->dma_tx = NULL; } @@ -759,8 +750,7 @@ static int sun6i_spi_probe(struct platform_device *pdev) err_free_dma_tx: if (host->dma_tx) dma_release_channel(host->dma_tx); -err_free_host: - spi_controller_put(host); + return ret; } @@ -768,8 +758,6 @@ static void sun6i_spi_remove(struct platform_device *pdev) { struct spi_controller *host = platform_get_drvdata(pdev); - spi_controller_get(host); - spi_unregister_controller(host); pm_runtime_force_suspend(&pdev->dev); @@ -778,8 +766,6 @@ static void sun6i_spi_remove(struct platform_device *pdev) dma_release_channel(host->dma_tx); if (host->dma_rx) dma_release_channel(host->dma_rx); - - spi_controller_put(host); } static const struct sun6i_spi_cfg sun6i_a31_spi_cfg = { From 5d5bbf177d18bdec01b9356f85d5883cbc6acf29 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:29:03 +0200 Subject: [PATCH 519/525] spi: syncuacer: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-15-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-synquacer.c | 24 +++++++----------------- 1 file changed, 7 insertions(+), 17 deletions(-) diff --git a/drivers/spi/spi-synquacer.c b/drivers/spi/spi-synquacer.c index 290c439897c4..c14225e39fd1 100644 --- a/drivers/spi/spi-synquacer.c +++ b/drivers/spi/spi-synquacer.c @@ -605,7 +605,7 @@ static int synquacer_spi_probe(struct platform_device *pdev) int ret; int rx_irq, tx_irq; - host = spi_alloc_host(&pdev->dev, sizeof(*sspi)); + host = devm_spi_alloc_host(&pdev->dev, sizeof(*sspi)); if (!host) return -ENOMEM; @@ -617,10 +617,8 @@ static int synquacer_spi_probe(struct platform_device *pdev) init_completion(&sspi->transfer_done); sspi->regs = devm_platform_ioremap_resource(pdev, 0); - if (IS_ERR(sspi->regs)) { - ret = PTR_ERR(sspi->regs); - goto put_spi; - } + if (IS_ERR(sspi->regs)) + return PTR_ERR(sspi->regs); sspi->clk_src_type = SYNQUACER_HSSPI_CLOCK_SRC_IHCLK; /* Default */ device_property_read_u32(&pdev->dev, "socionext,ihclk-rate", @@ -637,21 +635,19 @@ static int synquacer_spi_probe(struct platform_device *pdev) sspi->clk = devm_clk_get(sspi->dev, "iPCLK"); } else { dev_err(&pdev->dev, "specified wrong clock source\n"); - ret = -EINVAL; - goto put_spi; + return -EINVAL; } if (IS_ERR(sspi->clk)) { - ret = dev_err_probe(&pdev->dev, PTR_ERR(sspi->clk), - "clock not found\n"); - goto put_spi; + return dev_err_probe(&pdev->dev, PTR_ERR(sspi->clk), + "clock not found\n"); } ret = clk_prepare_enable(sspi->clk); if (ret) { dev_err(&pdev->dev, "failed to enable clock (%d)\n", ret); - goto put_spi; + return ret; } host->max_speed_hz = clk_get_rate(sspi->clk); @@ -726,8 +722,6 @@ static int synquacer_spi_probe(struct platform_device *pdev) pm_runtime_disable(sspi->dev); disable_clk: clk_disable_unprepare(sspi->clk); -put_spi: - spi_controller_put(host); return ret; } @@ -737,15 +731,11 @@ static void synquacer_spi_remove(struct platform_device *pdev) struct spi_controller *host = platform_get_drvdata(pdev); struct synquacer_spi *sspi = spi_controller_get_devdata(host); - spi_controller_get(host); - spi_unregister_controller(host); pm_runtime_disable(sspi->dev); clk_disable_unprepare(sspi->clk); - - spi_controller_put(host); } static int __maybe_unused synquacer_spi_suspend(struct device *dev) From 3068e7063cc42032b30e3eaef34066a86cb0aa68 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:29:04 +0200 Subject: [PATCH 520/525] spi: tegra114: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-16-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-tegra114.c | 34 ++++++++++++---------------------- 1 file changed, 12 insertions(+), 22 deletions(-) diff --git a/drivers/spi/spi-tegra114.c b/drivers/spi/spi-tegra114.c index b8b0ebe0fe93..aa44ffd09e61 100644 --- a/drivers/spi/spi-tegra114.c +++ b/drivers/spi/spi-tegra114.c @@ -1302,7 +1302,7 @@ static int tegra_spi_probe(struct platform_device *pdev) int ret, spi_irq; int bus_num; - host = spi_alloc_host(&pdev->dev, sizeof(*tspi)); + host = devm_spi_alloc_host(&pdev->dev, sizeof(*tspi)); if (!host) { dev_err(&pdev->dev, "host allocation failed\n"); return -ENOMEM; @@ -1336,36 +1336,31 @@ static int tegra_spi_probe(struct platform_device *pdev) tspi->soc_data = of_device_get_match_data(&pdev->dev); if (!tspi->soc_data) { dev_err(&pdev->dev, "unsupported tegra\n"); - ret = -ENODEV; - goto exit_free_host; + return -ENODEV; } tspi->base = devm_platform_get_and_ioremap_resource(pdev, 0, &r); - if (IS_ERR(tspi->base)) { - ret = PTR_ERR(tspi->base); - goto exit_free_host; - } + if (IS_ERR(tspi->base)) + return PTR_ERR(tspi->base); + tspi->phys = r->start; spi_irq = platform_get_irq(pdev, 0); - if (spi_irq < 0) { - ret = spi_irq; - goto exit_free_host; - } + if (spi_irq < 0) + return spi_irq; + tspi->irq = spi_irq; tspi->clk = devm_clk_get(&pdev->dev, "spi"); if (IS_ERR(tspi->clk)) { dev_err(&pdev->dev, "can not get clock\n"); - ret = PTR_ERR(tspi->clk); - goto exit_free_host; + return PTR_ERR(tspi->clk); } tspi->rst = devm_reset_control_get_exclusive(&pdev->dev, "spi"); if (IS_ERR(tspi->rst)) { dev_err(&pdev->dev, "can not get reset\n"); - ret = PTR_ERR(tspi->rst); - goto exit_free_host; + return PTR_ERR(tspi->rst); } tspi->max_buf_size = SPI_FIFO_DEPTH << 2; @@ -1373,7 +1368,7 @@ static int tegra_spi_probe(struct platform_device *pdev) ret = tegra_spi_init_dma_param(tspi, true); if (ret < 0) - goto exit_free_host; + return ret; ret = tegra_spi_init_dma_param(tspi, false); if (ret < 0) goto exit_rx_dma_free; @@ -1431,8 +1426,7 @@ static int tegra_spi_probe(struct platform_device *pdev) tegra_spi_deinit_dma_param(tspi, false); exit_rx_dma_free: tegra_spi_deinit_dma_param(tspi, true); -exit_free_host: - spi_controller_put(host); + return ret; } @@ -1441,8 +1435,6 @@ static void tegra_spi_remove(struct platform_device *pdev) struct spi_controller *host = platform_get_drvdata(pdev); struct tegra_spi_data *tspi = spi_controller_get_devdata(host); - spi_controller_get(host); - spi_unregister_controller(host); free_irq(tspi->irq, tspi); @@ -1456,8 +1448,6 @@ static void tegra_spi_remove(struct platform_device *pdev) pm_runtime_disable(&pdev->dev); if (!pm_runtime_status_suspended(&pdev->dev)) tegra_spi_runtime_suspend(&pdev->dev); - - spi_controller_put(host); } #ifdef CONFIG_PM_SLEEP From 3a14bf4f5453121cae3cbdfb6180317c5d74f424 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:29:05 +0200 Subject: [PATCH 521/525] spi: tegra20-sflash: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-17-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-tegra20-sflash.c | 24 ++++++++---------------- 1 file changed, 8 insertions(+), 16 deletions(-) diff --git a/drivers/spi/spi-tegra20-sflash.c b/drivers/spi/spi-tegra20-sflash.c index 9256729f2d49..2caa33f0a52c 100644 --- a/drivers/spi/spi-tegra20-sflash.c +++ b/drivers/spi/spi-tegra20-sflash.c @@ -427,11 +427,9 @@ static int tegra_sflash_probe(struct platform_device *pdev) return -ENODEV; } - host = spi_alloc_host(&pdev->dev, sizeof(*tsd)); - if (!host) { - dev_err(&pdev->dev, "host allocation failed\n"); + host = devm_spi_alloc_host(&pdev->dev, sizeof(*tsd)); + if (!host) return -ENOMEM; - } /* the spi->mode bits understood by this driver: */ host->mode_bits = SPI_CPOL | SPI_CPHA; @@ -450,14 +448,13 @@ static int tegra_sflash_probe(struct platform_device *pdev) host->max_speed_hz = 25000000; /* 25MHz */ tsd->base = devm_platform_ioremap_resource(pdev, 0); - if (IS_ERR(tsd->base)) { - ret = PTR_ERR(tsd->base); - goto exit_free_host; - } + if (IS_ERR(tsd->base)) + return PTR_ERR(tsd->base); ret = platform_get_irq(pdev, 0); if (ret < 0) - goto exit_free_host; + return ret; + tsd->irq = ret; ret = request_irq(tsd->irq, tegra_sflash_isr, 0, @@ -465,7 +462,7 @@ static int tegra_sflash_probe(struct platform_device *pdev) if (ret < 0) { dev_err(&pdev->dev, "Failed to register ISR for IRQ %d\n", tsd->irq); - goto exit_free_host; + return ret; } tsd->clk = devm_clk_get(&pdev->dev, NULL); @@ -518,8 +515,7 @@ static int tegra_sflash_probe(struct platform_device *pdev) tegra_sflash_runtime_suspend(&pdev->dev); exit_free_irq: free_irq(tsd->irq, tsd); -exit_free_host: - spi_controller_put(host); + return ret; } @@ -528,8 +524,6 @@ static void tegra_sflash_remove(struct platform_device *pdev) struct spi_controller *host = platform_get_drvdata(pdev); struct tegra_sflash_data *tsd = spi_controller_get_devdata(host); - spi_controller_get(host); - spi_unregister_controller(host); free_irq(tsd->irq, tsd); @@ -537,8 +531,6 @@ static void tegra_sflash_remove(struct platform_device *pdev) pm_runtime_disable(&pdev->dev); if (!pm_runtime_status_suspended(&pdev->dev)) tegra_sflash_runtime_suspend(&pdev->dev); - - spi_controller_put(host); } #ifdef CONFIG_PM_SLEEP From 76a24627b98ca712df8724985a4601c502b7875f Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:29:06 +0200 Subject: [PATCH 522/525] spi: ti-qspi: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-18-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-ti-qspi.c | 30 +++++++++--------------------- 1 file changed, 9 insertions(+), 21 deletions(-) diff --git a/drivers/spi/spi-ti-qspi.c b/drivers/spi/spi-ti-qspi.c index 1fbd710d616f..2a8548810f84 100644 --- a/drivers/spi/spi-ti-qspi.c +++ b/drivers/spi/spi-ti-qspi.c @@ -765,7 +765,7 @@ static int ti_qspi_probe(struct platform_device *pdev) int ret = 0, num_cs, irq; dma_cap_mask_t mask; - host = spi_alloc_host(&pdev->dev, sizeof(*qspi)); + host = devm_spi_alloc_host(&pdev->dev, sizeof(*qspi)); if (!host) return -ENOMEM; @@ -793,8 +793,7 @@ static int ti_qspi_probe(struct platform_device *pdev) r = platform_get_resource(pdev, IORESOURCE_MEM, 0); if (r == NULL) { dev_err(&pdev->dev, "missing platform data\n"); - ret = -ENODEV; - goto free_host; + return -ENODEV; } } @@ -812,28 +811,22 @@ static int ti_qspi_probe(struct platform_device *pdev) qspi->mmap_size = resource_size(res_mmap); irq = platform_get_irq(pdev, 0); - if (irq < 0) { - ret = irq; - goto free_host; - } + if (irq < 0) + return irq; mutex_init(&qspi->list_lock); qspi->base = devm_ioremap_resource(&pdev->dev, r); - if (IS_ERR(qspi->base)) { - ret = PTR_ERR(qspi->base); - goto free_host; - } + if (IS_ERR(qspi->base)) + return PTR_ERR(qspi->base); if (of_property_present(np, "syscon-chipselects")) { qspi->ctrl_base = syscon_regmap_lookup_by_phandle_args(np, "syscon-chipselects", 1, &qspi->ctrl_reg); - if (IS_ERR(qspi->ctrl_base)) { - ret = PTR_ERR(qspi->ctrl_base); - goto free_host; - } + if (IS_ERR(qspi->ctrl_base)) + return PTR_ERR(qspi->ctrl_base); } qspi->fclk = devm_clk_get(&pdev->dev, "fck"); @@ -895,8 +888,7 @@ static int ti_qspi_probe(struct platform_device *pdev) ti_qspi_dma_cleanup(qspi); pm_runtime_disable(&pdev->dev); -free_host: - spi_controller_put(host); + return ret; } @@ -904,16 +896,12 @@ static void ti_qspi_remove(struct platform_device *pdev) { struct ti_qspi *qspi = platform_get_drvdata(pdev); - spi_controller_get(qspi->host); - spi_unregister_controller(qspi->host); pm_runtime_put_sync(&pdev->dev); pm_runtime_disable(&pdev->dev); ti_qspi_dma_cleanup(qspi); - - spi_controller_put(qspi->host); } static const struct dev_pm_ops ti_qspi_pm_ops = { From f8689d5a9ee44a7cdd0ff469e5ee943c933fc830 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:29:07 +0200 Subject: [PATCH 523/525] spi: ti-qspi: cleanup registration error path Add a proper error path for when registration fails so that the probe tests for errors consistently. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-19-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-ti-qspi.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/spi/spi-ti-qspi.c b/drivers/spi/spi-ti-qspi.c index 2a8548810f84..6b407c7b5d33 100644 --- a/drivers/spi/spi-ti-qspi.c +++ b/drivers/spi/spi-ti-qspi.c @@ -882,9 +882,12 @@ static int ti_qspi_probe(struct platform_device *pdev) qspi->current_cs = -1; ret = spi_register_controller(host); - if (!ret) - return 0; + if (ret) + goto err_free_dma; + return 0; + +err_free_dma: ti_qspi_dma_cleanup(qspi); pm_runtime_disable(&pdev->dev); From 789986b1456497837c542f64b1a0c7d9dafd1b2a Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:29:08 +0200 Subject: [PATCH 524/525] spi: uniphier: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-20-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-uniphier.c | 33 +++++++++++---------------------- 1 file changed, 11 insertions(+), 22 deletions(-) diff --git a/drivers/spi/spi-uniphier.c b/drivers/spi/spi-uniphier.c index eac6c3e8908b..5f0abc59b716 100644 --- a/drivers/spi/spi-uniphier.c +++ b/drivers/spi/spi-uniphier.c @@ -649,7 +649,7 @@ static int uniphier_spi_probe(struct platform_device *pdev) int irq; int ret; - host = spi_alloc_host(&pdev->dev, sizeof(*priv)); + host = devm_spi_alloc_host(&pdev->dev, sizeof(*priv)); if (!host) return -ENOMEM; @@ -660,30 +660,26 @@ static int uniphier_spi_probe(struct platform_device *pdev) priv->is_save_param = false; priv->base = devm_platform_get_and_ioremap_resource(pdev, 0, &res); - if (IS_ERR(priv->base)) { - ret = PTR_ERR(priv->base); - goto out_host_put; - } + if (IS_ERR(priv->base)) + return PTR_ERR(priv->base); + priv->base_dma_addr = res->start; priv->clk = devm_clk_get_enabled(&pdev->dev, NULL); if (IS_ERR(priv->clk)) { dev_err(&pdev->dev, "failed to get clock\n"); - ret = PTR_ERR(priv->clk); - goto out_host_put; + return PTR_ERR(priv->clk); } irq = platform_get_irq(pdev, 0); - if (irq < 0) { - ret = irq; - goto out_host_put; - } + if (irq < 0) + return irq; ret = devm_request_irq(&pdev->dev, irq, uniphier_spi_handler, 0, "uniphier-spi", priv); if (ret) { dev_err(&pdev->dev, "failed to request IRQ\n"); - goto out_host_put; + return ret; } init_completion(&priv->xfer_done); @@ -710,10 +706,9 @@ static int uniphier_spi_probe(struct platform_device *pdev) host->dma_tx = dma_request_chan(&pdev->dev, "tx"); if (IS_ERR_OR_NULL(host->dma_tx)) { - if (PTR_ERR(host->dma_tx) == -EPROBE_DEFER) { - ret = -EPROBE_DEFER; - goto out_host_put; - } + if (PTR_ERR(host->dma_tx) == -EPROBE_DEFER) + return -EPROBE_DEFER; + host->dma_tx = NULL; dma_tx_burst = INT_MAX; } else { @@ -762,8 +757,6 @@ static int uniphier_spi_probe(struct platform_device *pdev) host->dma_tx = NULL; } -out_host_put: - spi_controller_put(host); return ret; } @@ -771,16 +764,12 @@ static void uniphier_spi_remove(struct platform_device *pdev) { struct spi_controller *host = platform_get_drvdata(pdev); - spi_controller_get(host); - spi_unregister_controller(host); if (host->dma_tx) dma_release_channel(host->dma_tx); if (host->dma_rx) dma_release_channel(host->dma_rx); - - spi_controller_put(host); } static const struct of_device_id uniphier_spi_match[] = { From be552efa43eec9972b490fce9a3dd9cb0da92a3c Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 May 2026 09:29:09 +0200 Subject: [PATCH 525/525] spi: zync-qspi: switch to managed controller allocation Switch to device managed controller allocation to simplify error handling and to avoid having to take another reference during deregistration. Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260505072909.618363-21-johan@kernel.org Signed-off-by: Mark Brown --- drivers/spi/spi-zynq-qspi.c | 40 +++++++++++-------------------------- 1 file changed, 12 insertions(+), 28 deletions(-) diff --git a/drivers/spi/spi-zynq-qspi.c b/drivers/spi/spi-zynq-qspi.c index 406fd9d5337e..d762aaf452af 100644 --- a/drivers/spi/spi-zynq-qspi.c +++ b/drivers/spi/spi-zynq-qspi.c @@ -637,7 +637,7 @@ static int zynq_qspi_probe(struct platform_device *pdev) struct zynq_qspi *xqspi; u32 num_cs; - ctlr = spi_alloc_host(&pdev->dev, sizeof(*xqspi)); + ctlr = devm_spi_alloc_host(&pdev->dev, sizeof(*xqspi)); if (!ctlr) return -ENOMEM; @@ -645,16 +645,13 @@ static int zynq_qspi_probe(struct platform_device *pdev) xqspi->dev = dev; platform_set_drvdata(pdev, ctlr); xqspi->regs = devm_platform_ioremap_resource(pdev, 0); - if (IS_ERR(xqspi->regs)) { - ret = PTR_ERR(xqspi->regs); - goto remove_ctlr; - } + if (IS_ERR(xqspi->regs)) + return PTR_ERR(xqspi->regs); xqspi->pclk = devm_clk_get_enabled(&pdev->dev, "pclk"); if (IS_ERR(xqspi->pclk)) { dev_err(&pdev->dev, "pclk clock not found.\n"); - ret = PTR_ERR(xqspi->pclk); - goto remove_ctlr; + return PTR_ERR(xqspi->pclk); } init_completion(&xqspi->data_completion); @@ -662,21 +659,18 @@ static int zynq_qspi_probe(struct platform_device *pdev) xqspi->refclk = devm_clk_get_enabled(&pdev->dev, "ref_clk"); if (IS_ERR(xqspi->refclk)) { dev_err(&pdev->dev, "ref_clk clock not found.\n"); - ret = PTR_ERR(xqspi->refclk); - goto remove_ctlr; + return PTR_ERR(xqspi->refclk); } xqspi->irq = platform_get_irq(pdev, 0); - if (xqspi->irq < 0) { - ret = xqspi->irq; - goto remove_ctlr; - } + if (xqspi->irq < 0) + return xqspi->irq; + ret = devm_request_irq(&pdev->dev, xqspi->irq, zynq_qspi_irq, 0, pdev->name, xqspi); if (ret != 0) { - ret = -ENXIO; dev_err(&pdev->dev, "request_irq failed\n"); - goto remove_ctlr; + return -ENXIO; } ret = of_property_read_u32(np, "num-cs", @@ -684,9 +678,8 @@ static int zynq_qspi_probe(struct platform_device *pdev) if (ret < 0) { ctlr->num_chipselect = 1; } else if (num_cs > ZYNQ_QSPI_MAX_NUM_CS) { - ret = -EINVAL; dev_err(&pdev->dev, "only 2 chip selects are available\n"); - goto remove_ctlr; + return -EINVAL; } else { ctlr->num_chipselect = num_cs; } @@ -705,15 +698,10 @@ static int zynq_qspi_probe(struct platform_device *pdev) ret = spi_register_controller(ctlr); if (ret) { dev_err(&pdev->dev, "failed to register controller\n"); - goto remove_ctlr; + return ret; } - return ret; - -remove_ctlr: - spi_controller_put(ctlr); - - return ret; + return 0; } /** @@ -731,13 +719,9 @@ static void zynq_qspi_remove(struct platform_device *pdev) struct spi_controller *ctlr = platform_get_drvdata(pdev); struct zynq_qspi *xqspi = spi_controller_get_devdata(ctlr); - spi_controller_get(ctlr); - spi_unregister_controller(ctlr); zynq_qspi_write(xqspi, ZYNQ_QSPI_ENABLE_OFFSET, 0); - - spi_controller_put(ctlr); } static const struct of_device_id zynq_qspi_of_match[] = {