IO errors were correctly printed to stderr, and propagated up to the
main loop for the server side, but the returned value was ignored. As a
consequence, the program for the listener side was no longer exiting
with an error code in case of IO issues.
Because of that, some issues might not have been seen. But very likely,
most issues either had an effect on the client side, or the file
transfer was not the expected one, e.g. the connection got reset before
the end. Still, it is better to fix this.
The main consequence of this issue is the error that was reported by the
selftests: the received and sent files were different, and the MIB
counters were not printed. Also, when such errors happened during the
'disconnect' tests, the program tried to continue until the timeout.
Now when an IO error is detected, the program exits directly with an
error.
Fixes: 05be5e273c ("selftests: mptcp: add disconnect tests")
Cc: stable@vger.kernel.org
Reviewed-by: Mat Martineau <martineau@kernel.org>
Reviewed-by: Geliang Tang <geliang@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20250912-net-mptcp-fix-sft-connect-v1-2-d40e77cbbf02@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When the MPTCP DATA FIN have been ACKed, there is no more MPTCP related
metadata to exchange, and all subflows can be safely shutdown.
Before this patch, the subflows were actually terminated at 'close()'
time. That's certainly fine most of the time, but not when the userspace
'shutdown()' a connection, without close()ing it. When doing so, the
subflows were staying in LAST_ACK state on one side -- and consequently
in FIN_WAIT2 on the other side -- until the 'close()' of the MPTCP
socket.
Now, when the DATA FIN have been ACKed, all subflows are shutdown. A
consequence of this is that the TCP 'FIN' flag can be set earlier now,
but the end result is the same. This affects the packetdrill tests
looking at the end of the MPTCP connections, but for a good reason.
Note that tcp_shutdown() will check the subflow state, so no need to do
that again before calling it.
Fixes: 3721b9b646 ("mptcp: Track received DATA_FIN sequence number and add related helpers")
Cc: stable@vger.kernel.org
Fixes: 16a9a9da17 ("mptcp: Add helper to process acks of DATA_FIN")
Reviewed-by: Mat Martineau <martineau@kernel.org>
Reviewed-by: Geliang Tang <geliang@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20250912-net-mptcp-fix-sft-connect-v1-1-d40e77cbbf02@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
After commit 5c3bf6cba7 ("bonding: assign random address if device
address is same as bond"), bonding will erroneously randomize the MAC
address of the first interface added to the bond if fail_over_mac =
follow.
Correct this by additionally testing for the bond being empty before
randomizing the MAC.
Fixes: 5c3bf6cba7 ("bonding: assign random address if device address is same as bond")
Reported-by: Qiuling Ren <qren@redhat.com>
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Link: https://patch.msgid.link/20250910024336.400253-1-liuhangbin@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
We need to increment i_fastreg_wrs before we bail out from
rds_ib_post_reg_frmr().
We have a fixed budget of how many FRWR operations that can be
outstanding using the dedicated QP used for memory registrations and
de-registrations. This budget is enforced by the atomic_t
i_fastreg_wrs. If we bail out early in rds_ib_post_reg_frmr(), we will
"leak" the possibility of posting an FRWR operation, and if that
accumulates, no FRWR operation can be carried out.
Fixes: 1659185fb4 ("RDS: IB: Support Fastreg MR (FRMR) memory registration mode")
Fixes: 3a2886cca7 ("net/rds: Keep track of and wait for FRWR segments in use upon shutdown")
Cc: stable@vger.kernel.org
Signed-off-by: Håkon Bugge <haakon.bugge@oracle.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Link: https://patch.msgid.link/20250911133336.451212-1-haakon.bugge@oracle.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ilya Maximets says:
====================
net: dst_metadata: fix DF flag extraction on tunnel rx
Two patches here, first fixes the issue where tunnel core doesn't
actually extract DF bit from the outer IP header, even though both
OVS and TC flower allow matching on it. More details in the commit
message.
The second is a selftest for openvswitch that reproduces the issue,
but also just adds some basic coverage for the tunnel metadata
extraction and related openvswitch uAPI.
====================
Link: https://patch.msgid.link/20250909165440.229890-1-i.maximets@ovn.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This test ensures that upon receiving decapsulated packets from a
tunnel interface in openvswitch, the tunnel metadata fields are
properly populated. This partially covers interoperability of the
kernel tunnel ports and openvswitch tunnels (LWT) and parsing and
formatting of the tunnel metadata fields of the openvswitch netlink
uAPI. Doing so, this test also ensures that fields and flags are
properly extracted during decapsulation by the tunnel core code,
serving as a regression test for the previously fixed issue with the
DF bit not being extracted from the outer IP header.
The ovs-dpctl.py script already supports all that is necessary for
the tunnel ports for this test, so we only need to adjust the
ovs_add_if() function to pass the '-t' port type argument in order
to be able to create tunnel ports in the openvswitch datapath.
Reviewed-by: Aaron Conole <aconole@redhat.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Link: https://patch.msgid.link/20250909165440.229890-3-i.maximets@ovn.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Both OVS and TC flower allow extracting and matching on the DF bit of
the outer IP header via OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT in the
OVS_KEY_ATTR_TUNNEL and TCA_FLOWER_KEY_FLAGS_TUNNEL_DONT_FRAGMENT in
the TCA_FLOWER_KEY_ENC_FLAGS respectively. Flow dissector extracts
this information as FLOW_DIS_F_TUNNEL_DONT_FRAGMENT from the tunnel
info key.
However, the IP_TUNNEL_DONT_FRAGMENT_BIT in the tunnel key is never
actually set, because the tunneling code doesn't actually extract it
from the IP header. OAM and CRIT_OPT are extracted by the tunnel
implementation code, same code also sets the KEY flag, if present.
UDP tunnel core takes care of setting the CSUM flag if the checksum
is present in the UDP header, but the DONT_FRAGMENT is not handled at
any layer.
Fix that by checking the bit and setting the corresponding flag while
populating the tunnel info in the IP layer where it belongs.
Not using __assign_bit as we don't really need to clear the bit in a
just initialized field. It also doesn't seem like using __assign_bit
will make the code look better.
Clearly, users didn't rely on this functionality for anything very
important until now. The reason why this doesn't break OVS logic is
that it only matches on what kernel previously parsed out and if kernel
consistently reports this bit as zero, OVS will only match on it to be
zero, which sort of works. But it is still a bug that the uAPI reports
and allows matching on the field that is not actually checked in the
packet. And this is causing misleading -df reporting in OVS datapath
flows, while the tunnel traffic actually has the bit set in most cases.
This may also cause issues if a hardware properly implements support
for tunnel flag matching as it will disagree with the implementation
in a software path of TC flower.
Fixes: 7d5437c709 ("openvswitch: Add tunneling interface.")
Fixes: 1d17568e74 ("net/sched: cls_flower: add support for matching tunnel control flags")
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20250909165440.229890-2-i.maximets@ovn.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In the protection override dump path, the firmware can return far too
many GRC elements, resulting in attempting to write past the end of the
previously-kmalloc'ed dump buffer.
This will result in a kernel panic with reason:
BUG: unable to handle kernel paging request at ADDRESS
where "ADDRESS" is just past the end of the protection override dump
buffer. The start address of the buffer is:
p_hwfn->cdev->dbg_features[DBG_FEATURE_PROTECTION_OVERRIDE].dump_buf
and the size of the buffer is buf_size in the same data structure.
The panic can be arrived at from either the qede Ethernet driver path:
[exception RIP: qed_grc_dump_addr_range+0x108]
qed_protection_override_dump at ffffffffc02662ed [qed]
qed_dbg_protection_override_dump at ffffffffc0267792 [qed]
qed_dbg_feature at ffffffffc026aa8f [qed]
qed_dbg_all_data at ffffffffc026b211 [qed]
qed_fw_fatal_reporter_dump at ffffffffc027298a [qed]
devlink_health_do_dump at ffffffff82497f61
devlink_health_report at ffffffff8249cf29
qed_report_fatal_error at ffffffffc0272baf [qed]
qede_sp_task at ffffffffc045ed32 [qede]
process_one_work at ffffffff81d19783
or the qedf storage driver path:
[exception RIP: qed_grc_dump_addr_range+0x108]
qed_protection_override_dump at ffffffffc068b2ed [qed]
qed_dbg_protection_override_dump at ffffffffc068c792 [qed]
qed_dbg_feature at ffffffffc068fa8f [qed]
qed_dbg_all_data at ffffffffc0690211 [qed]
qed_fw_fatal_reporter_dump at ffffffffc069798a [qed]
devlink_health_do_dump at ffffffff8aa95e51
devlink_health_report at ffffffff8aa9ae19
qed_report_fatal_error at ffffffffc0697baf [qed]
qed_hw_err_notify at ffffffffc06d32d7 [qed]
qed_spq_post at ffffffffc06b1011 [qed]
qed_fcoe_destroy_conn at ffffffffc06b2e91 [qed]
qedf_cleanup_fcport at ffffffffc05e7597 [qedf]
qedf_rport_event_handler at ffffffffc05e7bf7 [qedf]
fc_rport_work at ffffffffc02da715 [libfc]
process_one_work at ffffffff8a319663
Resolve this by clamping the firmware's return value to the maximum
number of legal elements the firmware should return.
Fixes: d52c89f120 ("qed*: Utilize FW 8.37.2.0")
Signed-off-by: Jamie Bainbridge <jamie.bainbridge@gmail.com>
Link: https://patch.msgid.link/f8e1182934aa274c18d0682a12dbaf347595469c.1757485536.git.jamie.bainbridge@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add a helper to validate the VF ID and use it in the VF ndo ops to
prevent accessing out-of-range entries.
Without this check, users can run commands such as:
# ip link show dev enp135s0
2: enp135s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:00:00:01:01:00 brd ff:ff:ff:ff:ff:ff
vf 0 link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff, spoof checking on, link-state enable, trust off
vf 1 link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff, spoof checking on, link-state enable, trust off
# ip link set dev enp135s0 vf 4 mac 00:00:00:00:00:14
# echo $?
0
even though VF 4 does not exist, which results in silent success instead
of returning an error.
Fixes: 8a241ef9b9 ("octeon_ep: add ndo ops for VFs in PF driver")
Signed-off-by: Kamal Heib <kheib@redhat.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250911223610.1803144-1-kheib@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The DPLL_CLOCK_QUALITY_LEVEL_ITU_OPT1_EPRC is not reported via netlink
due to bug in dpll_msg_add_clock_quality_level(). The usage of
DPLL_CLOCK_QUALITY_LEVEL_MAX for both DECLARE_BITMAP() and
for_each_set_bit() is not correct because these macros requires bitmap
size and not the highest valid bit in the bitmap.
Use correct bitmap size to fix this issue.
Fixes: a1afb959ad ("dpll: add clock quality level attribute and op")
Signed-off-by: Ivan Vecera <ivecera@redhat.com>
Reviewed-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Link: https://patch.msgid.link/20250912093331.862333-1-ivecera@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A NULL pointer dereference can occur in tcp_ao_finish_connect() during a
connect() system call on a socket with a TCP-AO key added and TCP_REPAIR
enabled.
The function is called with skb being NULL and attempts to dereference it
on tcp_hdr(skb)->seq without a prior skb validation.
Fix this by checking if skb is NULL before dereferencing it.
The commentary is taken from bpf_skops_established(), which is also called
in the same flow. Unlike the function being patched,
bpf_skops_established() validates the skb before dereferencing it.
int main(void){
struct sockaddr_in sockaddr;
struct tcp_ao_add tcp_ao;
int sk;
int one = 1;
memset(&sockaddr,'\0',sizeof(sockaddr));
memset(&tcp_ao,'\0',sizeof(tcp_ao));
sk = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP);
sockaddr.sin_family = AF_INET;
memcpy(tcp_ao.alg_name,"cmac(aes128)",12);
memcpy(tcp_ao.key,"ABCDEFGHABCDEFGH",16);
tcp_ao.keylen = 16;
memcpy(&tcp_ao.addr,&sockaddr,sizeof(sockaddr));
setsockopt(sk, IPPROTO_TCP, TCP_AO_ADD_KEY, &tcp_ao,
sizeof(tcp_ao));
setsockopt(sk, IPPROTO_TCP, TCP_REPAIR, &one, sizeof(one));
sockaddr.sin_family = AF_INET;
sockaddr.sin_port = htobe16(123);
inet_aton("127.0.0.1", &sockaddr.sin_addr);
connect(sk,(struct sockaddr *)&sockaddr,sizeof(sockaddr));
return 0;
}
$ gcc tcp-ao-nullptr.c -o tcp-ao-nullptr -Wall
$ unshare -Urn
BUG: kernel NULL pointer dereference, address: 00000000000000b6
PGD 1f648d067 P4D 1f648d067 PUD 1982e8067 PMD 0
Oops: Oops: 0000 [#1] SMP NOPTI
Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop
Reference Platform, BIOS 6.00 11/12/2020
RIP: 0010:tcp_ao_finish_connect (net/ipv4/tcp_ao.c:1182)
Fixes: 7c2ffaf21b ("net/tcp: Calculate TCP-AO traffic keys")
Signed-off-by: Anderson Nascimento <anderson@allelesecurity.com>
Reviewed-by: Dmitry Safonov <0x7f454c46@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20250911230743.2551-3-anderson@allelesecurity.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Network drivers sometimes return -EOPNOTSUPP from their get_ts_info()
method, and this should not cause the reporting of PHY timestamping
information to be prohibited. Handle this error code, and also
arrange for ethtool_net_get_ts_info_by_phc() to return -EOPNOTSUPP
when the method is not implemented.
This allows e.g. PHYs connected to DSA switches which support
timestamping to report their timestamping capabilities.
Fixes: b9e3f7dc9e ("net: ethtool: tsinfo: Enhance tsinfo to support several hwtstamp by net topology")
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Kory Maincent <kory.maincent@bootlin.com>
Link: https://patch.msgid.link/E1uwiW3-00000004jRF-3CnC@rmk-PC.armlinux.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Starting with commit c50e747596 ("dpaa2-switch: Fix error checking in
dpaa2_switch_seed_bp()"), the probing of a second DPSW object errors out
like below.
fsl_dpaa2_switch dpsw.1: fsl_mc_driver_probe failed: -12
fsl_dpaa2_switch dpsw.1: probe with driver fsl_dpaa2_switch failed with error -12
The aforementioned commit brought to the surface the fact that seeding
buffers into the buffer pool destined for control traffic is not
successful and an access violation recoverable error can be seen in the
MC firmware log:
[E, qbman_rec_isr:391, QBMAN] QBMAN recoverable event 0x1000000
This happens because the driver incorrectly used the ID of the DPBP
object instead of the hardware buffer pool ID when trying to release
buffers into it.
This is because any DPSW object uses two buffer pools, one managed by
the Linux driver and destined for control traffic packet buffers and the
other one managed by the MC firmware and destined only for offloaded
traffic. And since the buffer pool managed by the MC firmware does not
have an external facing DPBP equivalent, any subsequent DPBP objects
created after the first DPSW will have a DPBP id different to the
underlying hardware buffer ID.
The issue was not caught earlier because these two numbers can be
identical when all DPBP objects are created before the DPSW objects are.
This is the case when the DPL file is used to describe the entire DPAA2
object layout and objects are created at boot time and it's also true
for the first DPSW being created dynamically using ls-addsw.
Fix this by using the buffer pool ID instead of the DPBP id when
releasing buffers into the pool.
Fixes: 2877e4f7e1 ("staging: dpaa2-switch: setup buffer pool and RX path rings")
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Link: https://patch.msgid.link/20250910144825.2416019-1-ioana.ciornei@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
napi_stop_kthread waits for the NAPI_STATE_SCHED_THREADED to be unset
before stopping the kthread. But it uses test_bit with the
NAPIF_STATE_SCHED_THREADED and that might stop the kthread early before
the flag is unset.
Use the NAPI_* variant of the NAPI state bits in test_bit instead.
Tested:
./tools/testing/selftests/net/nl_netdev.py
TAP version 13
1..7
ok 1 nl_netdev.empty_check
ok 2 nl_netdev.lo_check
ok 3 nl_netdev.page_pool_check
ok 4 nl_netdev.napi_list_check
ok 5 nl_netdev.dev_set_threaded
ok 6 nl_netdev.napi_set_threaded
ok 7 nl_netdev.nsim_rxq_reset_down
# Totals: pass:7 fail:0 xfail:0 xpass:0 skip:0 error:0
./tools/testing/selftests/drivers/net/napi_threaded.py
TAP version 13
1..2
ok 1 napi_threaded.change_num_queues
ok 2 napi_threaded.enable_dev_threaded_disable_napi_threaded
# Totals: pass:2 fail:0 xfail:0 xpass:0 skip:0 error:0
Fixes: 689883de94 ("net: stop napi kthreads when THREADED napi is disabled")
Signed-off-by: Samiullah Khawaja <skhawaja@google.com>
Link: https://patch.msgid.link/20250910203716.1016546-1-skhawaja@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pull networking fixes from Paolo Abeni:
"Including fixes from CAN, netfilter and wireless.
We have an IPv6 routing regression with the relevant fix still a WiP.
This includes a last-minute revert to avoid more problems.
Current release - new code bugs:
- wifi: nl80211: completely disable per-link stats for now
Previous releases - regressions:
- dev_ioctl: take ops lock in hwtstamp lower paths
- netfilter:
- fix spurious set lookup failures
- fix lockdep splat due to missing annotation
- genetlink: fix genl_bind() invoking bind() after -EPERM
- phy: transfer phy_config_inband() locking responsibility to phylink
- can: xilinx_can: fix use-after-free of transmitted SKB
- hsr: fix lock warnings
- eth:
- igb: fix NULL pointer dereference in ethtool loopback test
- i40e: fix Jumbo Frame support after iPXE boot
- macsec: sync features on RTM_NEWLINK
Previous releases - always broken:
- tunnels: reset the GSO metadata before reusing the skb
- mptcp: make sync_socket_options propagate SOCK_KEEPOPEN
- can: j1939: implement NETDEV_UNREGISTER notification hanidler
- wifi: ath12k: fix WMI TLV header misalignment"
* tag 'net-6.17-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (47 commits)
Revert "net: usb: asix: ax88772: drop phylink use in PM to avoid MDIO runtime PM wakeups"
hsr: hold rcu and dev lock for hsr_get_port_ndev
hsr: use hsr_for_each_port_rtnl in hsr_port_get_hsr
hsr: use rtnl lock when iterating over ports
wifi: nl80211: completely disable per-link stats for now
net: usb: asix: ax88772: drop phylink use in PM to avoid MDIO runtime PM wakeups
net: ethtool: fix wrong type used in struct kernel_ethtool_ts_info
MAINTAINERS: add Phil as netfilter reviewer
netfilter: nf_tables: restart set lookup on base_seq change
netfilter: nf_tables: make nft_set_do_lookup available unconditionally
netfilter: nf_tables: place base_seq in struct net
netfilter: nft_set_rbtree: continue traversal if element is inactive
netfilter: nft_set_pipapo: don't check genbit from packetpath lookups
netfilter: nft_set_bitmap: fix lockdep splat due to missing annotation
can: rcar_can: rcar_can_resume(): fix s2ram with PSCI
can: xilinx_can: xcan_write_frame(): fix use-after-free of transmitted SKB
can: j1939: j1939_local_ecu_get(): undo increment when j1939_local_ecu_get() fails
can: j1939: j1939_sk_bind(): call j1939_priv_put() immediately when j1939_local_ecu_get() failed
can: j1939: implement NETDEV_UNREGISTER notification handler
selftests: can: enable CONFIG_CAN_VCAN as a module
...
Pull s390 fixes from Alexander Gordeev:
- ptep_modify_prot_start() may be called in a loop, which might lead to
the preempt_count overflow due to the unnecessary preemption
disabling. Do not disable preemption to prevent the overflow
- Events of type PERF_TYPE_HARDWARE are not tested for sampling and
return -EOPNOTSUPP eventually.
Instead, deny all sampling events by CPUMF counter facility and
return -ENOENT to allow other PMUs to be tried
- The PAI PMU driver returns -EINVAL if an event out of its range. That
aborts a search for an alternative PMU driver.
Instead, return -ENOENT to allow other PMUs to be tried
* tag 's390-6.17-4' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
s390/cpum_cf: Deny all sampling events by counter PMU
s390/pai: Deny all events not handled by this PMU
s390/mm: Prevent possible preempt_count overflow
Pull power management fixes from Rafael Wysocki:
"These fix a nasty hibernation regression introduced during the 6.16
cycle, an issue related to energy model management occurring on Intel
hybrid systems where some CPUs are offline to start with, and two
regressions in the amd-pstate driver:
- Restore a pm_restrict_gfp_mask() call in hibernation_snapshot()
that was removed incorrectly during the 6.16 development cycle
(Rafael Wysocki)
- Introduce a function for registering a perf domain without
triggering a system-wide CPU capacity update and make the
intel_pstate driver use it to avoid reocurring unsuccessful
attempts to update capacities of all CPUs in the system (Rafael
Wysocki)
- Fix setting of CPPC.min_perf in the active mode with performance
governor in the amd-pstate driver to restore its expected behavior
changed recently (Gautham Shenoy)
- Avoid mistakenly setting EPP to 0 in the amd-pstate driver after
system resume as a result of recent code changes (Mario
Limonciello)"
* tag 'pm-6.17-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
PM: hibernate: Restrict GFP mask in hibernation_snapshot()
PM: EM: Add function for registering a PD without capacity update
cpufreq/amd-pstate: Fix a regression leading to EPP 0 after resume
cpufreq/amd-pstate: Fix setting of CPPC.min_perf in active mode for performance governor
Pull btrfs fixes from David Sterba:
- fix delayed inode tracking in xarray, eviction can race with
insertion and leave behind a disconnected inode
- on systems with large page (64K) and small block size (4K) fix
compression read that can return partially filled folio
- slightly relax compression option format for backward compatibility,
allow to specify level for LZO although there's only one
- fix simple quota accounting of compressed extents
- validate minimum device size in 'device add'
- update maintainers' entry
* tag 'for-6.17-rc5-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
btrfs: don't allow adding block device of less than 1 MB
MAINTAINERS: update btrfs entry
btrfs: fix subvolume deletion lockup caused by inodes xarray race
btrfs: fix corruption reading compressed range when block size is smaller than page size
btrfs: accept and ignore compression level for lzo
btrfs: fix squota compressed stats leak
Pull bpf fixes from Alexei Starovoitov:
"A number of fixes accumulated due to summer vacations
- Fix out-of-bounds dynptr write in bpf_crypto_crypt() kfunc which
was misidentified as a security issue (Daniel Borkmann)
- Update the list of BPF selftests maintainers (Eduard Zingerman)
- Fix selftests warnings with icecc compiler (Ilya Leoshkevich)
- Disable XDP/cpumap direct return optimization (Jesper Dangaard
Brouer)
- Fix unexpected get_helper_proto() result in unusual configuration
BPF_SYSCALL=y and BPF_EVENTS=n (Jiri Olsa)
- Allow fallback to interpreter when JIT support is limited (KaFai
Wan)
- Fix rqspinlock and choose trylock fallback for NMI waiters. Pick
the simplest fix. More involved fix is targeted bpf-next (Kumar
Kartikeya Dwivedi)
- Fix cleanup when tcp_bpf_send_verdict() fails to allocate
psock->cork (Kuniyuki Iwashima)
- Disallow bpf_timer in PREEMPT_RT for now. Proper solution is being
discussed for bpf-next. (Leon Hwang)
- Fix XSK cq descriptor production (Maciej Fijalkowski)
- Tell memcg to use allow_spinning=false path in bpf_timer_init() to
avoid lockup in cgroup_file_notify() (Peilin Ye)
- Fix bpf_strnstr() to handle suffix match cases (Rong Tao)"
* tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
selftests/bpf: Skip timer cases when bpf_timer is not supported
bpf: Reject bpf_timer for PREEMPT_RT
tcp_bpf: Call sk_msg_free() when tcp_bpf_send_verdict() fails to allocate psock->cork.
bpf: Tell memcg to use allow_spinning=false path in bpf_timer_init()
bpf: Allow fall back to interpreter for programs with stack size <= 512
rqspinlock: Choose trylock fallback for NMI waiters
xsk: Fix immature cq descriptor production
bpf: Update the list of BPF selftests maintainers
selftests/bpf: Add tests for bpf_strnstr
selftests/bpf: Fix "expression result unused" warnings with icecc
bpf: Fix bpf_strnstr() to handle suffix match cases better
selftests/bpf: Extend crypto_sanity selftest with invalid dst buffer
bpf: Fix out-of-bounds dynptr write in bpf_crypto_crypt
bpf: Check the helper function is valid in get_helper_proto
bpf, cpumap: Disable page_pool direct xdp_return need larger scope
Merge a hibernation regression fix and an fix related to energy model
management for 6.17-rc6
* pm-sleep:
PM: hibernate: Restrict GFP mask in hibernation_snapshot()
* pm-em:
PM: EM: Add function for registering a PD without capacity update
Johannes Berg says:
====================
Some more fixes:
- iwlwifi: fix 130/1030 devices
- ath12k: fix alignment, power save
- virt_wifi: fix crash
- cfg80211: disable per-link stats due
to buffer size issues
* tag 'wireless-2025-09-11' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless:
wifi: nl80211: completely disable per-link stats for now
wifi: virt_wifi: Fix page fault on connect
wifi: cfg80211: Fix "no buffer space available" error in nl80211_get_station() for MLO
wifi: iwlwifi: fix 130/1030 configs
wifi: ath12k: fix WMI TLV header misalignment
wifi: ath12k: Fix missing station power save configuration
====================
Link: https://patch.msgid.link/20250911100345.20025-3-johannes@sipsolutions.net
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
hsr_port_get_hsr() iterates over ports using hsr_for_each_port(),
but many of its callers do not hold the required RCU lock.
Switch to hsr_for_each_port_rtnl(), since most callers already hold
the rtnl lock. After review, all callers are covered by either the rtnl
lock or the RCU lock, except hsr_dev_xmit(). Fix this by adding an
RCU read lock there.
Fixes: c5a7591172 ("net/hsr: Use list_head (and rcu) instead of array for slave devices.")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250905091533.377443-3-liuhangbin@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
hsr_for_each_port is called in many places without holding the RCU read
lock, this may trigger warnings on debug kernels. Most of the callers
are actually hold rtnl lock. So add a new helper hsr_for_each_port_rtnl
to allow callers in suitable contexts to iterate ports safely without
explicit RCU locking.
This patch only fixed the callers that is hold rtnl lock. Other caller
issues will be fixed in later patches.
Fixes: c5a7591172 ("net/hsr: Use list_head (and rcu) instead of array for slave devices.")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250905091533.377443-2-liuhangbin@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
After commit 8cc71fc3b8 ("wifi: cfg80211: Fix "no buffer
space available" error in nl80211_get_station() for MLO"),
the per-link data is only included in station dumps, where
the size limit is somewhat less of an issue. However, it's
still an issue, depending on how many links a station has
and how much per-link data there is. Thus, for now, disable
per-link statistics entirely.
A complete fix will need to take this into account, make it
opt-in by userspace, and change the dump format to be able
to split a single station's data across multiple netlink
dump messages, which all together is too much development
for a fix.
Fixes: 82d7f841d9 ("wifi: cfg80211: extend to embed link level statistics in NL message")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Pull misc fixes from Andrew Morton:
"20 hotfixes. 15 are cc:stable and the remainder address post-6.16
issues or aren't considered necessary for -stable kernels. 14 of these
fixes are for MM.
This includes
- kexec fixes from Breno for a recently introduced
use-uninitialized bug
- DAMON fixes from Quanmin Yan to avoid div-by-zero crashes
which can occur if the operator uses poorly-chosen insmod
parameters
and misc singleton fixes"
* tag 'mm-hotfixes-stable-2025-09-10-20-00' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm:
MAINTAINERS: add tree entry to numa memblocks and emulation block
mm/damon/sysfs: fix use-after-free in state_show()
proc: fix type confusion in pde_set_flags()
compiler-clang.h: define __SANITIZE_*__ macros only when undefined
mm/vmalloc, mm/kasan: respect gfp mask in kasan_populate_vmalloc()
ocfs2: fix recursive semaphore deadlock in fiemap call
mm/memory-failure: fix VM_BUG_ON_PAGE(PagePoisoned(page)) when unpoison memory
mm/mremap: fix regression in vrm->new_addr check
percpu: fix race on alloc failed warning limit
mm/memory-failure: fix redundant updates for already poisoned pages
s390: kexec: initialize kexec_buf struct
riscv: kexec: initialize kexec_buf struct
arm64: kexec: initialize kexec_buf struct in load_other_segments()
mm/damon/reclaim: avoid divide-by-zero in damon_reclaim_apply_parameters()
mm/damon/lru_sort: avoid divide-by-zero in damon_lru_sort_apply_parameters()
mm/damon/core: set quota->charged_from to jiffies at first charge window
mm/hugetlb: add missing hugetlb_lock in __unmap_hugepage_range()
init/main.c: fix boot time tracing crash
mm/memory_hotplug: fix hwpoisoned large folio handling in do_migrate_range()
mm/khugepaged: fix the address passed to notifier on testing young
Pull vmescape mitigation fixes from Dave Hansen:
"Mitigate vmscape issue with indirect branch predictor flushes.
vmscape is a vulnerability that essentially takes Spectre-v2 and
attacks host userspace from a guest. It particularly affects
hypervisors like QEMU.
Even if a hypervisor may not have any sensitive data like disk
encryption keys, guest-userspace may be able to attack the
guest-kernel using the hypervisor as a confused deputy.
There are many ways to mitigate vmscape using the existing Spectre-v2
defenses like IBRS variants or the IBPB flushes. This series focuses
solely on IBPB because it works universally across vendors and all
vulnerable processors. Further work doing vendor and model-specific
optimizations can build on top of this if needed / wanted.
Do the normal issue mitigation dance:
- Add the CPU bug boilerplate
- Add a list of vulnerable CPUs
- Use IBPB to flush the branch predictors after running guests"
* tag 'vmscape-for-linus-20250904' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/vmscape: Add old Intel CPUs to affected list
x86/vmscape: Warn when STIBP is disabled with SMT
x86/bugs: Move cpu_bugs_smt_update() down
x86/vmscape: Enable the mitigation
x86/vmscape: Add conditional IBPB mitigation
x86/vmscape: Enumerate VMSCAPE bug
Documentation/hw-vuln: Add VMSCAPE documentation
Florian Westpha says:
====================
netfilter pull request nf-25-09-10
First patch adds a lockdep annotation for a false-positive splat.
Last patch adds formal reviewer tag for Phil Sutter to MAINTAINERS.
Rest of the patches resolve spurious false negative results during set
lookups while another CPU is processing a transaction.
This has been broken at least since v4.18 when an unconditional
synchronize_rcu call was removed from the commit phase of nf_tables.
Quoting from Stefan Hanreichs original report:
It seems like we've found an issue with atomicity when reloading
nftables rulesets. Sometimes there is a small window where rules
containing sets do not seem to apply to incoming traffic, due to the set
apparently being empty for a short amount of time when flushing / adding
elements.
Exanple ruleset:
table ip filter {
set match {
type ipv4_addr
flags interval
elements = { 0.0.0.0-192.168.2.19, 192.168.2.21-255.255.255.255 }
}
chain pre {
type filter hook prerouting priority filter; policy accept;
ip saddr @match accept
counter comment "must never match"
}
}
Reproducer transaction:
while true:
nft -f -<<EOF
flush set ip filter match
create element ip filter match { \
0.0.0.0-192.168.2.19, 192.168.2.21-255.255.255.255 }
EOF
done
Then create traffic. to/from e.g. 192.168.2.1 to 192.168.3.10.
Once in a while the counter will increment even though the
'ip saddr @match' rule should have accepted the packet.
See individual patches for details.
Thanks to Stefan Hanreich for an initial description and reproducer for
this bug and to Pablo Neira Ayuso for reviewing earlier iterations of
the patchset.
* tag 'nf-25-09-10-v2' of https://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
MAINTAINERS: add Phil as netfilter reviewer
netfilter: nf_tables: restart set lookup on base_seq change
netfilter: nf_tables: make nft_set_do_lookup available unconditionally
netfilter: nf_tables: place base_seq in struct net
netfilter: nft_set_rbtree: continue traversal if element is inactive
netfilter: nft_set_pipapo: don't check genbit from packetpath lookups
netfilter: nft_set_bitmap: fix lockdep splat due to missing annotation
====================
Link: https://patch.msgid.link/20250910190308.13356-1-fw@strlen.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Marc Kleine-Budde says:
====================
pull-request: can 2025-09-10
The 1st patch is by Alex Tran and fixes the Documentation of the
struct bcm_msg_head.
Davide Caratti's patch enabled the VCAN driver as a module for the
Linux self tests.
Tetsuo Handa contributes 3 patches that fix various problems in the
CAN j1939 protocol.
Anssi Hannula's patch fixes a potential use-after-free in the
xilinx_can driver.
Geert Uytterhoeven's patch fixes the rcan_can's suspend to RAM on
R-Car Gen3 using PSCI.
* tag 'linux-can-fixes-for-6.17-20250910' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:
can: rcar_can: rcar_can_resume(): fix s2ram with PSCI
can: xilinx_can: xcan_write_frame(): fix use-after-free of transmitted SKB
can: j1939: j1939_local_ecu_get(): undo increment when j1939_local_ecu_get() fails
can: j1939: j1939_sk_bind(): call j1939_priv_put() immediately when j1939_local_ecu_get() failed
can: j1939: implement NETDEV_UNREGISTER notification handler
selftests: can: enable CONFIG_CAN_VCAN as a module
docs: networking: can: change bcm_msg_head frames member to support flexible array
====================
Link: https://patch.msgid.link/20250910162907.948454-1-mkl@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2025-09-09 (igb, i40e)
For igb:
Tianyu Xu removes passing of, no longer needed, NAPI id to avoid NULL
pointer dereference on ethtool loopback testing.
Kohei Enju corrects reporting/testing of link state when interface is
down.
For i40e:
Michal Schmidt corrects value being passed to free_irq().
Jake sets hardware maximum frame size on probe to ensure
expected/consistent state.
* '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
i40e: fix Jumbo Frame support after iPXE boot
i40e: fix IRQ freeing in i40e_vsi_request_irq_msix error path
igb: fix link test skipping when interface is admin down
igb: Fix NULL pointer dereference in ethtool loopback test
====================
Link: https://patch.msgid.link/20250909203236.3603960-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Drop phylink_{suspend,resume}() from ax88772 PM callbacks.
MDIO bus accesses have their own runtime-PM handling and will try to
wake the device if it is suspended. Such wake attempts must not happen
from PM callbacks while the device PM lock is held. Since phylink
{sus|re}sume may trigger MDIO, it must not be called in PM context.
No extra phylink PM handling is required for this driver:
- .ndo_open/.ndo_stop control the phylink start/stop lifecycle.
- ethtool/phylib entry points run in process context, not PM.
- phylink MAC ops program the MAC on link changes after resume.
Fixes: e0bffe3e68 ("net: asix: ax88772: migrate to phylink")
Reported-by: Hubert Wiśniewski <hubert.wisniewski.25632@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Tested-by: Hubert Wiśniewski <hubert.wisniewski.25632@gmail.com>
Tested-by: Xu Yang <xu.yang_2@nxp.com>
Link: https://patch.msgid.link/20250908112619.2900723-1-o.rempel@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In C, enumerated types do not have a defined size, apart from being
compatible with one of the standard types. This allows an ABI /
compiler to choose the type of an enum depending on the values it
needs to store, and storing larger values in it can lead to undefined
behaviour.
The tx_type and rx_filters members of struct kernel_ethtool_ts_info
are defined as enumerated types, but are bit arrays, where each bit
is defined by the enumerated type. This means they typically store
values in excess of the maximum value of the enumerated type, in
fact (1 << max_value) and thus must not be declared using the
enumated type.
Fix both of these to use u32, as per the corresponding __u32 UAPI type.
Fixes: 2111375b85 ("net: Add struct kernel_ethtool_ts_info")
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Kory Maincent <kory.maincent@bootlin.com>
Link: https://patch.msgid.link/E1uvMEK-00000003Amd-2pWR@rmk-PC.armlinux.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pull NFS client fixes from Trond Myklebust:
"Stable patches:
- Revert "SUNRPC: Don't allow waiting for exiting tasks" as it is
breaking ltp tests
Bugfixes:
- Another set of fixes to the tracking of NFSv4 server capabilities
when crossing filesystem boundaries
- Localio fix to restore credentials and prevent triggering a
BUG_ON()
- Fix to prevent flapping of the localio on/off trigger
- Protections against 'eof page pollution' as demonstrated in
xfstests generic/363
- Series of patches to ensure correct ordering of O_DIRECT i/o and
truncate, fallocate and copy functions
- Fix a NULL pointer check in flexfiles reads that regresses 6.17
- Correct a typo that breaks flexfiles layout segment processing"
* tag 'nfs-for-6.17-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
NFSv4/flexfiles: Fix layout merge mirror check.
SUNRPC: call xs_sock_process_cmsg for all cmsg
Revert "SUNRPC: Don't allow waiting for exiting tasks"
NFS: Fix the marking of the folio as up to date
NFS: nfs_invalidate_folio() must observe the offset and size arguments
NFSv4.2: Serialise O_DIRECT i/o and copy range
NFSv4.2: Serialise O_DIRECT i/o and clone range
NFSv4.2: Serialise O_DIRECT i/o and fallocate()
NFS: Serialise O_DIRECT i/o and truncate()
NFSv4.2: Protect copy offload and clone against 'eof page pollution'
NFS: Protect against 'eof page pollution'
flexfiles/pNFS: fix NULL checks on result of ff_layout_choose_ds_for_read
nfs/localio: avoid bouncing LOCALIO if nfs_client_is_local()
nfs/localio: restore creds before releasing pageio data
NFSv4: Clear the NFS_CAP_XATTR flag if not supported by the server
NFSv4: Clear NFS_CAP_OPEN_XOR and NFS_CAP_DELEGTIME if not supported
NFSv4: Clear the NFS_CAP_FS_LOCATIONS flag if it is not set
NFSv4: Don't clear capabilities that won't be reset
Leon Hwang says:
====================
bpf: Reject bpf_timer for PREEMPT_RT
While running './test_progs -t timer' to validate the test case from
"selftests/bpf: Introduce experimental bpf_in_interrupt()"[0] for
PREEMPT_RT, I encountered a kernel warning:
BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48
To address this, reject bpf_timer usage in the verifier when
PREEMPT_RT is enabled, and skip the corresponding timer selftests.
Changes:
v2 -> v3:
* Drop skipping test case 'timer_interrupt'.
* Address comments from Alexei:
* Respin targeting bpf tree.
* Trim commit log.
v1 -> v2:
* Skip test case 'timer_interrupt'.
Links:
[0] https://lore.kernel.org/bpf/20250903140438.59517-1-leon.hwang@linux.dev/
====================
Link: https://patch.msgid.link/20250910125740.52172-1-leon.hwang@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
When enable CONFIG_PREEMPT_RT, the kernel will warn when run timer
selftests by './test_progs -t timer':
BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48
In order to avoid such warning, reject bpf_timer in verifier when
PREEMPT_RT is enabled.
Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
Link: https://lore.kernel.org/r/20250910125740.52172-2-leon.hwang@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Pull tracing fixes from Steven Rostedt:
- Remove redundant __GFP_NOWARN flag is kmalloc
As now __GFP_NOWARN is part of __GFP_NOWAIT, it can be removed from
kmalloc as it is redundant.
- Use copy_from_user_nofault() instead of _inatomic() for trace markers
The trace_marker files are written to to allow user space to quickly
write into the tracing ring buffer.
Back in 2016, the get_user_pages_fast() and the kmap() logic was
replaced by a __copy_from_user_inatomic(), but didn't properly
disable page faults around it.
Since the time this was added, copy_from_user_nofault() was added
which does the required page fault disabling for us.
- Fix the assembly markup in the ftrace direct sample code
The ftrace direct sample code (which is also used for selftests), had
the size directive between the "leave" and the "ret" instead of after
the ret. This caused objtool to think the code was unreachable.
- Only call unregister_pm_notifier() on outer most fgraph registration
There was an error path in register_ftrace_graph() that did not call
unregister_pm_notifier() on error, so it was added in the error path.
The problem with that fix, is that register_pm_notifier() is only
called by the initial user of fgraph. If that succeeds, but another
fgraph registration were to fail, then unregister_pm_notifier() would
be called incorrectly.
- Fix a crash in osnoise when zero size cpumask is passed in
If a zero size CPU mask is passed in, the kmalloc() would return
ZERO_SIZE_PTR which is not checked, and the code would continue
thinking it had real memory and crash. If zero is passed in as the
size of the write, simply return 0.
- Fix possible warning in trace_pid_write()
If while processing a series of numbers passed to the "set_event_pid"
file, and one of the updates fails to allocate (triggered by a fault
injection), it can cause a warning to trigger. Check the return value
of the call to trace_pid_list_set() and break out early with an error
code if it fails.
* tag 'trace-v6.17-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
tracing: Silence warning when chunk allocation fails in trace_pid_write
tracing/osnoise: Fix null-ptr-deref in bitmap_parselist()
trace/fgraph: Fix error handling
ftrace/samples: Fix function size computation
tracing: Fix tracing_marker may trigger page fault during preempt_disable
trace: Remove redundant __GFP_NOWARN
Commit 12ffc3b151 ("PM: Restrict swap use to later in the suspend
sequence") incorrectly removed a pm_restrict_gfp_mask() call from
hibernation_snapshot(), so memory allocations involving swap are not
prevented from being carried out in this code path any more which may
lead to serious breakage.
The symptoms of such breakage have become visible after adding a
shrink_shmem_memory() call to hibernation_snapshot() in commit
2640e81947 ("PM: hibernate: shrink shmem pages after dev_pm_ops.prepare()")
which caused this problem to be much more likely to manifest itself.
However, since commit 2640e81947 was initially present in the DRM
tree that did not include commit 12ffc3b151, the symptoms of this
issue were not visible until merge commit 260f6f4fda ("Merge tag
'drm-next-2025-07-30' of https://gitlab.freedesktop.org/drm/kernel")
that exposed it through an entirely reasonable merge conflict
resolution.
Fixes: 12ffc3b151 ("PM: Restrict swap use to later in the suspend sequence")
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=220555
Reported-by: Todd Brandt <todd.e.brandt@linux.intel.com>
Tested-by: Todd Brandt <todd.e.brandt@linux.intel.com>
Cc: 6.16+ <stable@vger.kernel.org> # 6.16+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Phil has contributed to netfilter with features, fixes and patch reviews
for a long time. Make this more formal and add Reviewer tag.
Acked-by: Jozsef Kadlecsik <kadlec@netfilter.org>
Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
The hash, hash_fast, rhash and bitwise sets may indicate no result even
though a matching element exists during a short time window while other
cpu is finalizing the transaction.
This happens when the hash lookup/bitwise lookup function has picked up
the old genbit, right before it was toggled by nf_tables_commit(), but
then the same cpu managed to unlink the matching old element from the
hash table:
cpu0 cpu1
has added new elements to clone
has marked elements as being
inactive in new generation
perform lookup in the set
enters commit phase:
A) observes old genbit
increments base_seq
I) increments the genbit
II) removes old element from the set
B) finds matching element
C) returns no match: found
element is not valid in old
generation
Next lookup observes new genbit and
finds matching e2.
Consider a packet matching element e1, e2.
cpu0 processes following transaction:
1. remove e1
2. adds e2, which has same key as e1.
P matches both e1 and e2. Therefore, cpu1 should always find a match
for P. Due to above race, this is not the case:
cpu1 observed the old genbit. e2 will not be considered once it is found.
The element e1 is not found anymore if cpu0 managed to unlink it from the
hlist before cpu1 found it during list traversal.
The situation only occurs for a brief time period, lookups happening
after I) observe new genbit and return e2.
This problem exists in all set types except nft_set_pipapo, so fix it once
in nft_lookup rather than each set ops individually.
Sample the base sequence counter, which gets incremented right before the
genbit is changed.
Then, if no match is found, retry the lookup if the base sequence was
altered in between.
If the base sequence hasn't changed:
- No update took place: no-match result is expected.
This is the common case. or:
- nf_tables_commit() hasn't progressed to genbit update yet.
Old elements were still visible and nomatch result is expected, or:
- nf_tables_commit updated the genbit:
We picked up the new base_seq, so the lookup function also picked
up the new genbit, no-match result is expected.
If the old genbit was observed, then nft_lookup also picked up the old
base_seq: nft_lookup_should_retry() returns true and relookup is performed
in the new generation.
This problem was added when the unconditional synchronize_rcu() call
that followed the current/next generation bit toggle was removed.
Thanks to Pablo Neira Ayuso for reviewing an earlier version of this
patchset, for suggesting re-use of existing base_seq and placement of
the restart loop in nft_set_do_lookup().
Fixes: 0cbc06b3fa ("netfilter: nf_tables: remove synchronize_rcu in commit phase")
Signed-off-by: Florian Westphal <fw@strlen.de>
This function was added for retpoline mitigation and is replaced by a
static inline helper if mitigations are not enabled.
Enable this helper function unconditionally so next patch can add a lookup
restart mechanism to fix possible false negatives while transactions are
in progress.
Adding lookup restarts in nft_lookup_eval doesn't work as nft_objref would
then need the same copypaste loop.
This patch is separate to ease review of the actual bug fix.
Suggested-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
This will soon be read from packet path around same time as the gencursor.
Both gencursor and base_seq get incremented almost at the same time, so
it makes sense to place them in the same structure.
This doesn't increase struct net size on 64bit due to padding.
Signed-off-by: Florian Westphal <fw@strlen.de>
When the rbtree lookup function finds a match in the rbtree, it sets the
range start interval to a potentially inactive element.
Then, after tree lookup, if the matching element is inactive, it returns
NULL and suppresses a matching result.
This is wrong and leads to false negative matches when a transaction has
already entered the commit phase.
cpu0 cpu1
has added new elements to clone
has marked elements as being
inactive in new generation
perform lookup in the set
enters commit phase:
I) increments the genbit
A) observes new genbit
B) finds matching range
C) returns no match: found
range invalid in new generation
II) removes old elements from the tree
C New nft_lookup happening now
will find matching element,
because it is no longer
obscured by old, inactive one.
Consider a packet matching range r1-r2:
cpu0 processes following transaction:
1. remove r1-r2
2. add r1-r3
P is contained in both ranges. Therefore, cpu1 should always find a match
for P. Due to above race, this is not the case:
cpu1 does find r1-r2, but then ignores it due to the genbit indicating
the range has been removed. It does NOT test for further matches.
The situation persists for all lookups until after cpu0 hits II) after
which r1-r3 range start node is tested for the first time.
Move the "interval start is valid" check ahead so that tree traversal
continues if the starting interval is not valid in this generation.
Thanks to Stefan Hanreich for providing an initial reproducer for this
bug.
Reported-by: Stefan Hanreich <s.hanreich@proxmox.com>
Fixes: c1eda3c639 ("netfilter: nft_rbtree: ignore inactive matching element with no descendants")
Signed-off-by: Florian Westphal <fw@strlen.de>
The pipapo set type is special in that it has two copies of its
datastructure: one live copy containing only valid elements and one
on-demand clone used during transaction where adds/deletes happen.
This clone is not visible to the datapath.
This is unlike all other set types in nftables, those all link new
elements into their live hlist/tree.
For those sets, the lookup functions must skip the new elements while the
transaction is ongoing to ensure consistency.
As the clone is shallow, removal does have an effect on the packet path:
once the transaction enters the commit phase the 'gencursor' bit that
determines which elements are active and which elements should be ignored
(because they are no longer valid) is flipped.
This causes the datapath lookup to ignore these elements if they are found
during lookup.
This opens up a small race window where pipapo has an inconsistent view of
the dataset from when the transaction-cpu flipped the genbit until the
transaction-cpu calls nft_pipapo_commit() to swap live/clone pointers:
cpu0 cpu1
has added new elements to clone
has marked elements as being
inactive in new generation
perform lookup in the set
enters commit phase:
I) increments the genbit
A) observes new genbit
removes elements from the clone so
they won't be found anymore
B) lookup in datastructure
can't see new elements yet,
but old elements are ignored
-> Only matches elements that
were not changed in the
transaction
II) calls nft_pipapo_commit(), clone
and live pointers are swapped.
C New nft_lookup happening now
will find matching elements.
Consider a packet matching range r1-r2:
cpu0 processes following transaction:
1. remove r1-r2
2. add r1-r3
P is contained in both ranges. Therefore, cpu1 should always find a match
for P. Due to above race, this is not the case:
cpu1 does find r1-r2, but then ignores it due to the genbit indicating
the range has been removed.
At the same time, r1-r3 is not visible yet, because it can only be found
in the clone.
The situation persists for all lookups until after cpu0 hits II).
The fix is easy: Don't check the genbit from pipapo lookup functions.
This is possible because unlike the other set types, the new elements are
not reachable from the live copy of the dataset.
The clone/live pointer swap is enough to avoid matching on old elements
while at the same time all new elements are exposed in one go.
After this change, step B above returns a match in r1-r2.
This is fine: r1-r2 only becomes truly invalid the moment they get freed.
This happens after a synchronize_rcu() call and rcu read lock is held
via netfilter hook traversal (nf_hook_slow()).
Cc: Stefano Brivio <sbrivio@redhat.com>
Fixes: 3c4287f620 ("nf_tables: Add set type for arbitrary concatenation of ranges")
Signed-off-by: Florian Westphal <fw@strlen.de>
