The referenced fix is incomplete. It correctly computes
bond_dev->gso_partial_features across slaves, but unfortunately
netdev_fix_features discards gso_partial_features from the feature set
if NETIF_F_GSO_PARTIAL isn't set in bond_dev->features.
This is visible with ethtool -k bond0 | grep esp:
tx-esp-segmentation: off [requested on]
esp-hw-offload: on
esp-tx-csum-hw-offload: on
This patch reworks the bonding GSO offload support by:
- making aggregating gso_partial_features across slaves similar to the
other feature sets (this part is a no-op).
- advertising the default partial gso features on empty bond devs, same
as with other feature sets (also a no-op).
- adding NETIF_F_GSO_PARTIAL to hw_enc_features filtered across slaves.
- adding NETIF_F_GSO_PARTIAL to features in bond_setup()
With all of these, 'ethtool -k bond0 | grep esp' now reports:
tx-esp-segmentation: on
esp-hw-offload: on
esp-tx-csum-hw-offload: on
Fixes: 4861333b42 ("bonding: add ESP offload features when slaves support")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: Cosmin Ratiu <cratiu@nvidia.com>
Acked-by: Jay Vosburgh <jv@jvosburgh.net>
Link: https://patch.msgid.link/20250127104147.759658-1-cratiu@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
When Tx/Rx FIFO size is not specified in advance, the driver checks if
the value is zero and sets the hardware capability value in functions
where that value is used.
Consolidate the check and settings into function stmmac_hw_init() and
remove redundant other statements.
If FIFO size is zero and the hardware capability also doesn't have upper
limit values, return with an error message.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Reviewed-by: Yanteng Si <si.yanteng@linux.dev>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Tx/Rx FIFO size is specified by the parameter "{tx,rx}-fifo-depth" from
stmmac_platform layer.
However, these values are constrained by upper limits determined by the
capabilities of each hardware feature. There is a risk that the upper
bits will be truncated due to the calculation, so it's appropriate to
limit them to the upper limit values and display a warning message.
This only works if the hardware capability has the upper limit values.
Fixes: e7877f52fd ("stmmac: Read tx-fifo-depth and rx-fifo-depth from the devicetree")
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Reviewed-by: Yanteng Si <si.yanteng@linux.dev>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The number of MTL queues to use is specified by the parameter
"snps,{tx,rx}-queues-to-use" from stmmac_platform layer.
However, the maximum numbers of queues are constrained by upper limits
determined by the capability of each hardware feature. It's appropriate
to limit the values not to exceed the upper limit values and display
a warning message.
This only works if the hardware capability has the upper limit values.
Fixes: d976a525c3 ("net: stmmac: multiple queues dt configuration")
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Reviewed-by: Yanteng Si <si.yanteng@linux.dev>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Clarify that the "NCM" implementation in `ipheth` is very limited, as
iOS devices aren't compatible with the CDC NCM specification in regular
tethering mode.
For a standards-compliant implementation, one shall turn to
the `cdc_ncm` module.
Cc: stable@vger.kernel.org # 6.5.x
Signed-off-by: Foster Snowhill <forst@pen.gy>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Fix an out-of-bounds DPE read, limit the number of processed DPEs to
the amount that fits into the fixed-size NDP16 header.
Fixes: a2d274c62e ("usbnet: ipheth: add CDC NCM support")
Cc: stable@vger.kernel.org
Signed-off-by: Foster Snowhill <forst@pen.gy>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Originally, the total NCM header size was computed as the sum of two
vaguely labelled constants. While accurate, it wasn't particularly clear
where they were coming from.
Use sizes of existing NCM structs where available. Define the total
NDP16 size based on the maximum amount of DPEs that can fit into the
iOS-specific fixed-size header.
This change does not fix any particular issue. Rather, it introduces
intermediate constants that will simplify subsequent commits.
It should also make it clearer for the reader where the constant values
come from.
Cc: stable@vger.kernel.org # 6.5.x
Signed-off-by: Foster Snowhill <forst@pen.gy>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Introduce an rx_error label to reduce repetitions in the header
signature checks.
Store wDatagramIndex and wDatagramLength after endianness conversion to
avoid repeated le16_to_cpu() calls.
Rewrite the loop to return on a null trailing DPE, which is required
by the CDC NCM spec. In case it is missing, fall through to rx_error.
This change does not fix any particular issue. Its purpose is to
simplify a subsequent commit that fixes a potential OoB read by limiting
the maximum amount of processed DPEs.
Cc: stable@vger.kernel.org # 6.5.x
Signed-off-by: Foster Snowhill <forst@pen.gy>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Original code allowed for the start of NDP16 to be anywhere within the
URB based on the `wNdpIndex` value in NTH16. Only the start position of
NDP16 was checked, so it was possible for even the fixed-length part
of NDP16 to extend past the end of URB, leading to an out-of-bounds
read.
On iOS devices, the NDP16 header always directly follows NTH16. Rely on
and check for this specific format.
This, along with NCM-specific minimal URB length check that already
exists, will ensure that the fixed-length part of NDP16 plus a set
amount of DPEs fit within the URB.
Note that this commit alone does not fully address the OoB read.
The limit on the amount of DPEs needs to be enforced separately.
Fixes: a2d274c62e ("usbnet: ipheth: add CDC NCM support")
Cc: stable@vger.kernel.org
Signed-off-by: Foster Snowhill <forst@pen.gy>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
By definition, a DPE points at the start of a network frame/datagram.
Thus it makes no sense for it to point at anything that's part of the
NCM header. It is not a security issue, but merely an indication of
a malformed DPE.
Enforce that all DPEs point at the data portion of the URB, past the
NCM header.
Fixes: a2d274c62e ("usbnet: ipheth: add CDC NCM support")
Cc: stable@vger.kernel.org
Signed-off-by: Foster Snowhill <forst@pen.gy>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Originally, it was possible for the DPE length check to overflow if
wDatagramIndex + wDatagramLength > U16_MAX. This could lead to an OoB
read.
Move the wDatagramIndex term to the other side of the inequality.
An existing condition ensures that wDatagramIndex < urb->actual_length.
Fixes: a2d274c62e ("usbnet: ipheth: add CDC NCM support")
Cc: stable@vger.kernel.org
Signed-off-by: Foster Snowhill <forst@pen.gy>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2025-01-24 (idpf, ice, iavf)
For idpf:
Emil adds memory barrier when accessing control queue descriptors and
restores call to idpf_vc_xn_shutdown() when resetting.
Manoj Vishwanathan expands transaction lock to properly protect xn->salt
value and adds additional debugging information.
Marco Leogrande converts workqueues to be unbound.
For ice:
Przemek fixes incorrect size use for array.
Mateusz removes reporting of invalid parameter and value.
For iavf:
Michal adjusts some VLAN changes to occur without a PF call to avoid
timing issues with the calls.
* '200GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
iavf: allow changing VLAN state without calling PF
ice: remove invalid parameter of equalizer
ice: fix ice_parser_rt::bst_key array size
idpf: add more info during virtchnl transaction timeout/salt mismatch
idpf: convert workqueues to unbound
idpf: Acquire the lock before accessing the xn->salt
idpf: fix transaction timeouts on reset
idpf: add read memory barrier when checking descriptor done bit
====================
Link: https://patch.msgid.link/20250124213213.1328775-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
dm is netdev private data and it cannot be
used after free_netdev() call. Using dm after free_netdev()
can cause UAF bug. Fix it by moving free_netdev() at the end of the
function.
This is similar to the issue fixed in commit
ad297cd2db ("net: qcom/emac: fix UAF in emac_remove").
This bug is detected by our static analysis tool.
Fixes: cf9e60aa69 ("net: davicom: Fix regulator not turned off on driver removal")
Signed-off-by: Chenyuan Yang <chenyuan0y@gmail.com>
CC: Uwe Kleine-König <u.kleine-koenig@baylibre.com>
Link: https://patch.msgid.link/20250123214213.623518-1-chenyuan0y@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In application note (AN13663) for TJA1120, on page 30, there's a figure
with average PHY startup timing values following software reset.
The time it takes for SMI to become operational after software reset
ranges roughly from 500 us to 1500 us.
This commit adds 2000 us delay after MDIO write which triggers software
reset. Without this delay, soft_reset function returns an error and
prevents successful PHY init.
Cc: stable@vger.kernel.org
Fixes: b050f2f15e ("phy: nxp-c45: add driver for tja1103")
Signed-off-by: Milos Reljin <milos_reljin@outlook.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://patch.msgid.link/AM8P250MB0124D258E5A71041AF2CC322E1E32@AM8P250MB0124.EURP250.PROD.OUTLOOK.COM
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Syzbot reports:
BUG: KMSAN: uninit-value in nsim_get_ringparam+0xa8/0xe0 drivers/net/netdevsim/ethtool.c:77
nsim_get_ringparam+0xa8/0xe0 drivers/net/netdevsim/ethtool.c:77
ethtool_set_ringparam+0x268/0x570 net/ethtool/ioctl.c:2072
__dev_ethtool net/ethtool/ioctl.c:3209 [inline]
dev_ethtool+0x126d/0x2a40 net/ethtool/ioctl.c:3398
dev_ioctl+0xb0e/0x1280 net/core/dev_ioctl.c:759
This is the SET path, where we call GET to either check user request
against max values, or check if any of the settings will change.
The logic in netdevsim is trying to report the default (ENABLED)
if user has not requested any specific setting. The user setting
is recorded in dev->cfg, don't depend on kernel_ringparam being
pre-populated with it.
Fixes: 928459bbda ("net: ethtool: populate the default HDS params in the core")
Reported-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot+b3bcd80232d00091e061@syzkaller.appspotmail.com
Tested-by: syzbot+b3bcd80232d00091e061@syzkaller.appspotmail.com
Link: https://patch.msgid.link/20250123221410.1067678-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pull non-MM updates from Andrew Morton:
"Mainly individually changelogged singleton patches. The patch series
in this pull are:
- "lib min_heap: Improve min_heap safety, testing, and documentation"
from Kuan-Wei Chiu provides various tightenings to the min_heap
library code
- "xarray: extract __xa_cmpxchg_raw" from Tamir Duberstein preforms
some cleanup and Rust preparation in the xarray library code
- "Update reference to include/asm-<arch>" from Geert Uytterhoeven
fixes pathnames in some code comments
- "Converge on using secs_to_jiffies()" from Easwar Hariharan uses
the new secs_to_jiffies() in various places where that is
appropriate
- "ocfs2, dlmfs: convert to the new mount API" from Eric Sandeen
switches two filesystems to the new mount API
- "Convert ocfs2 to use folios" from Matthew Wilcox does that
- "Remove get_task_comm() and print task comm directly" from Yafang
Shao removes now-unneeded calls to get_task_comm() in various
places
- "squashfs: reduce memory usage and update docs" from Phillip
Lougher implements some memory savings in squashfs and performs
some maintainability work
- "lib: clarify comparison function requirements" from Kuan-Wei Chiu
tightens the sort code's behaviour and adds some maintenance work
- "nilfs2: protect busy buffer heads from being force-cleared" from
Ryusuke Konishi fixes an issues in nlifs when the fs is presented
with a corrupted image
- "nilfs2: fix kernel-doc comments for function return values" from
Ryusuke Konishi fixes some nilfs kerneldoc
- "nilfs2: fix issues with rename operations" from Ryusuke Konishi
addresses some nilfs BUG_ONs which syzbot was able to trigger
- "minmax.h: Cleanups and minor optimisations" from David Laight does
some maintenance work on the min/max library code
- "Fixes and cleanups to xarray" from Kemeng Shi does maintenance
work on the xarray library code"
* tag 'mm-nonmm-stable-2025-01-24-23-16' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (131 commits)
ocfs2: use str_yes_no() and str_no_yes() helper functions
include/linux/lz4.h: add some missing macros
Xarray: use xa_mark_t in xas_squash_marks() to keep code consistent
Xarray: remove repeat check in xas_squash_marks()
Xarray: distinguish large entries correctly in xas_split_alloc()
Xarray: move forward index correctly in xas_pause()
Xarray: do not return sibling entries from xas_find_marked()
ipc/util.c: complete the kernel-doc function descriptions
gcov: clang: use correct function param names
latencytop: use correct kernel-doc format for func params
minmax.h: remove some #defines that are only expanded once
minmax.h: simplify the variants of clamp()
minmax.h: move all the clamp() definitions after the min/max() ones
minmax.h: use BUILD_BUG_ON_MSG() for the lo < hi test in clamp()
minmax.h: reduce the #define expansion of min(), max() and clamp()
minmax.h: update some comments
minmax.h: add whitespace around operators and after commas
nilfs2: do not update mtime of renamed directory that is not moved
nilfs2: handle errors that nilfs_prepare_chunk() may return
CREDITS: fix spelling mistake
...
Pull pci updates from Bjorn Helgaas:
"Enumeration:
- Batch sizing of multiple BARs while memory decoding is disabled
instead of disabling/enabling decoding for each BAR individually;
this optimizes virtualized environments where toggling decoding
enable is expensive (Alex Williamson)
- Add host bridge .enable_device() and .disable_device() hooks for
bridges that need to configure things like Requester ID to StreamID
mapping when enabling devices (Frank Li)
- Extend struct pci_ecam_ops with .enable_device() and
.disable_device() hooks so drivers that use pci_host_common_probe()
instead of their own .probe() have a way to set the
.enable_device() callbacks (Marc Zyngier)
- Drop 'No bus range found' message so we don't complain when DTs
don't specify the default 'bus-range = <0x00 0xff>' (Bjorn Helgaas)
- Rename the drivers/pci/of_property.c struct of_pci_range to
of_pci_range_entry to avoid confusion with the global of_pci_range
in include/linux/of_address.h (Bjorn Helgaas)
Driver binding:
- Update resource request API documentation to encourage callers to
supply a driver name when requesting resources (Philipp Stanner)
- Export pci_intx_unmanaged() and pcim_intx() (always managed) so
callers of pci_intx() (which is sometimes managed) can explicitly
choose the one they need (Philipp Stanner)
- Convert drivers from pci_intx() to always-managed pcim_intx() or
never-managed pci_intx_unmanaged(): amd_sfh, ata (ahci, ata_piix,
pata_rdc, sata_sil24, sata_sis, sata_uli, sata_vsc), bnx2x, bna,
ntb, qtnfmac, rtsx, tifm_7xx1, vfio, xen-pciback (Philipp Stanner)
- Remove pci_intx_unmanaged() since pci_intx() is now always
unmanaged and pcim_intx() is always managed (Philipp Stanner)
Error handling:
- Unexport pcie_read_tlp_log() to encourage drivers to use PCI core
logging rather than building their own (Ilpo Järvinen)
- Move TLP Log handling to its own file (Ilpo Järvinen)
- Store number of supported End-End TLP Prefixes always so we can
read the correct number of DWORDs from the TLP Prefix Log (Ilpo
Järvinen)
- Read TLP Prefixes in addition to the Header Log in
pcie_read_tlp_log() (Ilpo Järvinen)
- Add pcie_print_tlp_log() to consolidate printing of TLP Header and
Prefix Log (Ilpo Järvinen)
- Quirk the Intel Raptor Lake-P PIO log size to accommodate vendor
BIOSes that don't configure it correctly (Takashi Iwai)
ASPM:
- Save parent L1 PM Substates config so when we restore it along with
an endpoint's config, the parent info isn't junk (Jian-Hong Pan)
Power management:
- Avoid D3 for Root Ports on TUXEDO Sirius Gen1 with old BIOS because
the system can't wake up from suspend (Werner Sembach)
Endpoint framework:
- Destroy the EPC device in devm_pci_epc_destroy(), which previously
didn't call devres_release() (Zijun Hu)
- Finish virtual EP removal in pci_epf_remove_vepf(), which
previously caused a subsequent pci_epf_add_vepf() to fail with
-EBUSY (Zijun Hu)
- Write BAR_MASK before iATU registers in pci_epc_set_bar() so we
don't depend on the BAR_MASK reset value being larger than the
requested BAR size (Niklas Cassel)
- Prevent changing BAR size/flags in pci_epc_set_bar() to prevent
reads from bypassing the iATU if we reduced the BAR size (Niklas
Cassel)
- Verify address alignment when programming iATU so we don't attempt
to write bits that are read-only because of the BAR size, which
could lead to directing accesses to the wrong address (Niklas
Cassel)
- Implement artpec6 pci_epc_features so we can rely on all drivers
supporting it so we can use it in EPC core code (Niklas Cassel)
- Check for BARs of fixed size to prevent endpoint drivers from
trying to change their size (Niklas Cassel)
- Verify that requested BAR size is a power of two when endpoint
driver sets the BAR (Niklas Cassel)
Endpoint framework tests:
- Clear pci-epf-test dma_chan_rx, not dma_chan_tx, after freeing
dma_chan_rx (Mohamed Khalfella)
- Correct the DMA MEMCPY test so it doesn't fail if the Endpoint
supports both DMA_PRIVATE and DMA_MEMCPY (Manivannan Sadhasivam)
- Add pci-epf-test and pci_endpoint_test support for capabilities
(Niklas Cassel)
- Add Endpoint test for consecutive BARs (Niklas Cassel)
- Remove redundant comparison from Endpoint BAR test because a > 1MB
BAR can always be exactly covered by iterating with a 1MB buffer
(Hans Zhang)
- Move and convert PCI Endpoint tests from tools/pci to Kselftests
(Manivannan Sadhasivam)
Apple PCIe controller driver:
- Convert StreamID mapping configuration from a bus notifier to the
.enable_device() and .disable_device() callbacks (Marc Zyngier)
Freescale i.MX6 PCIe controller driver:
- Add Requester ID to StreamID mapping configuration when enabling
devices (Frank Li)
- Use DWC core suspend/resume functions for imx6 (Frank Li)
- Add suspend/resume support for i.MX8MQ, i.MX8Q, and i.MX95 (Richard
Zhu)
- Add DT compatible string 'fsl,imx8q-pcie-ep' and driver support for
i.MX8Q series (i.MX8QM, i.MX8QXP, and i.MX8DXL) Endpoints (Frank
Li)
- Add DT binding for optional i.MX95 Refclk and driver support to
enable it if the platform hasn't enabled it (Richard Zhu)
- Configure PHY based on controller being in Root Complex or Endpoint
mode (Frank Li)
- Rely on dbi2 and iATU base addresses from DT via
dw_pcie_get_resources() instead of hardcoding them (Richard Zhu)
- Deassert apps_reset in imx_pcie_deassert_core_reset() since it is
asserted in imx_pcie_assert_core_reset() (Richard Zhu)
- Add missing reference clock enable or disable logic for IMX6SX,
IMX7D, IMX8MM (Richard Zhu)
- Remove redundant imx7d_pcie_init_phy() since
imx7d_pcie_enable_ref_clk() does the same thing (Richard Zhu)
Freescale Layerscape PCIe controller driver:
- Simplify by using syscon_regmap_lookup_by_phandle_args() instead
of syscon_regmap_lookup_by_phandle() followed by
of_property_read_u32_array() (Krzysztof Kozlowski)
Marvell MVEBU PCIe controller driver:
- Add MODULE_DEVICE_TABLE() to enable module autoloading (Liao Chen)
MediaTek PCIe Gen3 controller driver:
- Use clk_bulk_prepare_enable() instead of separate
clk_bulk_prepare() and clk_bulk_enable() (Lorenzo Bianconi)
- Rearrange reset assert/deassert so they're both done in the
*_power_up() callbacks (Lorenzo Bianconi)
- Document that Airoha EN7581 requires PHY init and power-on before
PHY reset deassert, unlike other MediaTek Gen3 controllers (Lorenzo
Bianconi)
- Move Airoha EN7581 post-reset delay from the en7581 clock .enable()
method to mtk_pcie_en7581_power_up() (Lorenzo Bianconi)
- Sleep instead of delay during Airoha EN7581 power-up, since this is
a non-atomic context (Lorenzo Bianconi)
- Skip PERST# assertion on Airoha EN7581 during probe and
suspend/resume to avoid a hardware defect (Lorenzo Bianconi)
- Enable async probe to reduce system startup time (Douglas Anderson)
Microchip PolarFlare PCIe controller driver:
- Set up the inbound address translation based on whether the
platform allows coherent or non-coherent DMA (Daire McNamara)
- Update DT binding such that platforms are DMA-coherent by default
and must specify 'dma-noncoherent' if needed (Conor Dooley)
Mobiveil PCIe controller driver:
- Convert mobiveil-pcie.txt to YAML and update 'interrupt-names'
and 'reg-names' (Frank Li)
Qualcomm PCIe controller driver:
- Add DT SM8550 and SM8650 optional 'global' interrupt for link
events (Neil Armstrong)
- Add DT 'compatible' strings for IPQ5424 PCIe controller (Manikanta
Mylavarapu)
- If 'global' IRQ is supported for detection of Link Up events, tell
DWC core not to wait for link up (Krishna chaitanya chundru)
Renesas R-Car PCIe controller driver:
- Avoid passing stack buffer as resource name (King Dix)
Rockchip PCIe controller driver:
- Simplify clock and reset handling by using bulk interfaces (Anand
Moon)
- Pass typed rockchip_pcie (not void) pointer to
rockchip_pcie_disable_clocks() (Anand Moon)
- Return -ENOMEM, not success, when pci_epc_mem_alloc_addr() fails
(Dan Carpenter)
Rockchip DesignWare PCIe controller driver:
- Use dll_link_up IRQ to detect Link Up and enumerate devices so
users don't have to manually rescan (Niklas Cassel)
- Tell DWC core not to wait for link up since the 'sys' interrupt is
required and detects Link Up events (Niklas Cassel)
Synopsys DesignWare PCIe controller driver:
- Don't wait for link up in DWC core if driver can detect Link Up
event (Krishna chaitanya chundru)
- Update ICC and OPP votes after Link Up events (Krishna chaitanya
chundru)
- Always stop link in dw_pcie_suspend_noirq(), which is required at
least for i.MX8QM to re-establish link on resume (Richard Zhu)
- Drop racy and unnecessary LTSSM state check before sending
PME_TURN_OFF message in dw_pcie_suspend_noirq() (Richard Zhu)
- Add struct of_pci_range.parent_bus_addr for devices that need their
immediate parent bus address, not the CPU address, e.g., to program
an internal Address Translation Unit (iATU) (Frank Li)
TI DRA7xx PCIe controller driver:
- Simplify by using syscon_regmap_lookup_by_phandle_args() instead of
syscon_regmap_lookup_by_phandle() followed by
of_parse_phandle_with_fixed_args() or of_property_read_u32_index()
(Krzysztof Kozlowski)
Xilinx Versal CPM PCIe controller driver:
- Add DT binding and driver support for Xilinx Versal CPM5
(Thippeswamy Havalige)
MicroSemi Switchtec management driver:
- Add Microchip PCI100X device IDs (Rakesh Babu Saladi)
Miscellaneous:
- Move reset related sysfs code from pci.c to pci-sysfs.c where other
similar code lives (Ilpo Järvinen)
- Simplify reset_method_store() memory management by using __free()
instead of explicit kfree() cleanup (Ilpo Järvinen)
- Constify struct bin_attribute for sysfs, VPD, P2PDMA, and the IBM
ACPI hotplug driver (Thomas Weißschuh)
- Remove redundant PCI_VSEC_HDR and PCI_VSEC_HDR_LEN_SHIFT (Dongdong
Zhang)
- Correct documentation of the 'config_acs=' kernel parameter
(Akihiko Odaki)"
* tag 'pci-v6.14-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci: (111 commits)
PCI: Batch BAR sizing operations
dt-bindings: PCI: microchip,pcie-host: Allow dma-noncoherent
PCI: microchip: Set inbound address translation for coherent or non-coherent mode
Documentation: Fix pci=config_acs= example
PCI: Remove redundant PCI_VSEC_HDR and PCI_VSEC_HDR_LEN_SHIFT
PCI: Don't include 'pm_wakeup.h' directly
selftests: pci_endpoint: Migrate to Kselftest framework
selftests: Move PCI Endpoint tests from tools/pci to Kselftests
misc: pci_endpoint_test: Fix IOCTL return value
dt-bindings: PCI: qcom: Document the IPQ5424 PCIe controller
dt-bindings: PCI: qcom,pcie-sm8550: Document 'global' interrupt
dt-bindings: PCI: mobiveil: Convert mobiveil-pcie.txt to YAML
PCI: switchtec: Add Microchip PCI100X device IDs
misc: pci_endpoint_test: Remove redundant 'remainder' test
misc: pci_endpoint_test: Add consecutive BAR test
misc: pci_endpoint_test: Add support for capabilities
PCI: endpoint: pci-epf-test: Add support for capabilities
PCI: endpoint: pci-epf-test: Fix check for DMA MEMCPY test
PCI: endpoint: pci-epf-test: Set dma_chan_rx pointer to NULL on error
PCI: dwc: Simplify config resource lookup
...
Pull rdma updates from Jason Gunthorpe:
"Lighter that normal, but the now usual collection of driver fixes and
small improvements:
- Small fixes and minor improvements to cxgb4, bnxt_re, rxe, srp,
efa, cxgb4
- Update mlx4 to use the new umem APIs, avoiding direct use of
scatterlist
- Support ROCEv2 in erdma
- Remove various uncalled functions, constify bin_attribute
- Provide core infrastructure to catch netdev events and route them
to drivers, consolidating duplicated driver code
- Fix rare race condition crashes in mlx5 ODP flows"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (63 commits)
RDMA/mlx5: Fix implicit ODP use after free
RDMA/mlx5: Fix a race for an ODP MR which leads to CQE with error
RDMA/qib: Constify 'struct bin_attribute'
RDMA/hfi1: Constify 'struct bin_attribute'
RDMA/rxe: Fix the warning "__rxe_cleanup+0x12c/0x170 [rdma_rxe]"
RDMA/cxgb4: Notify rdma stack for IB_EVENT_QP_LAST_WQE_REACHED event
RDMA/bnxt_re: Allocate dev_attr information dynamically
RDMA/bnxt_re: Pass the context for ulp_irq_stop
RDMA/bnxt_re: Add support to handle DCB_CONFIG_CHANGE event
RDMA/bnxt_re: Query firmware defaults of CC params during probe
RDMA/bnxt_re: Add Async event handling support
bnxt_en: Add ULP call to notify async events
RDMA/mlx5: Fix indirect mkey ODP page count
MAINTAINERS: Update the bnxt_re maintainers
RDMA/hns: Clean up the legacy CONFIG_INFINIBAND_HNS
RDMA/rtrs: Add missing deinit() call
RDMA/efa: Align interrupt related fields to same type
RDMA/bnxt_re: Fix to drop reference to the mmap entry in case of error
RDMA/mlx5: Fix link status down event for MPV
RDMA/erdma: Support create_ah/destroy_ah in non-sleepable contexts
...
First case:
> ip l a l $VF name vlanx type vlan id 100
> ip l d vlanx
> ip l a l $VF name vlanx type vlan id 100
As workqueue can be execute after sometime, there is a window to have
call trace like that:
- iavf_del_vlan
- iavf_add_vlan
- iavf_del_vlans (wq)
It means that our VLAN 100 will change the state from IAVF_VLAN_ACTIVE
to IAVF_VLAN_REMOVE (iavf_del_vlan). After that in iavf_add_vlan state
won't be changed because VLAN 100 is on the filter list. The final
result is that the VLAN 100 filter isn't added in hardware (no
iavf_add_vlans call).
To fix that change the state if the filter wasn't removed yet directly
to active. It is save as IAVF_VLAN_REMOVE means that virtchnl message
wasn't sent yet.
Second case:
> ip l a l $VF name vlanx type vlan id 100
Any type of VF reset ex. change trust
> ip l s $PF vf $VF_NUM trust on
> ip l d vlanx
> ip l a l $VF name vlanx type vlan id 100
In case of reset iavf driver is responsible for readding all filters
that are being used. To do that all VLAN filters state are changed to
IAVF_VLAN_ADD. Here is even longer window for changing VLAN state from
kernel side, as workqueue isn't called immediately. We can have call
trace like that:
- changing to IAVF_VLAN_ADD (after reset)
- iavf_del_vlan (called from kernel ops)
- iavf_del_vlans (wq)
Not exsisitng VLAN filters will be removed from hardware. It isn't a
bug, ice driver will handle it fine. However, we can have call trace
like that:
- changing to IAVF_VLAN_ADD (after reset)
- iavf_del_vlan (called from kernel ops)
- iavf_add_vlan (called from kernel ops)
- iavf_del_vlans (wq)
With fix for previous case we end up with no VLAN filters in hardware.
We have to remove VLAN filters if the state is IAVF_VLAN_ADD and delete
VLAN was called. It is save as IAVF_VLAN_ADD means that virtchnl message
wasn't sent yet.
Fixes: 0c0da0e951 ("iavf: refactor VLAN filter states")
Signed-off-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
It occurred that in the commit 70838938e8 ("ice: Implement driver
functionality to dump serdes equalizer values") the invalid DRATE parameter
for reading has been added. The output of the command:
$ ethtool -d <ethX>
returns the garbage value in the place where DRATE value should be
stored.
Remove mentioned parameter to prevent return of corrupted data to
userspace.
Fixes: 70838938e8 ("ice: Implement driver functionality to dump serdes equalizer values")
Signed-off-by: Mateusz Polchlopek <mateusz.polchlopek@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Tested-by: Rinitha S <sx.rinitha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
In the US country code, to avoid including 6 GHz rules in the 5 GHz rules
list, the number of 5 GHz rules is set to a default constant value of 4
(REG_US_5G_NUM_REG_RULES). However, if there are more than 4 valid 5 GHz
rules, the current logic will bypass the legitimate 6 GHz rules.
For example, if there are 5 valid 5 GHz rules and 1 valid 6 GHz rule, the
current logic will only consider 4 of the 5 GHz rules, treating the last
valid rule as a 6 GHz rule. Consequently, the actual 6 GHz rule is never
processed, leading to the eventual disabling of 6 GHz channels.
To fix this issue, instead of hardcoding the value to 4, use a helper
function to determine the number of 6 GHz rules present in the 5 GHz rules
list and ignore only those rules.
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.3.1-00173-QCAHKSWPL_SILICONZ-1
Cc: stable@vger.kernel.org
Fixes: d889913205 ("wifi: ath12k: driver for Qualcomm Wi-Fi 7 devices")
Signed-off-by: Aditya Kumar Singh <aditya.kumar.singh@oss.qualcomm.com>
Link: https://patch.msgid.link/20250123-fix_6ghz_rules_handling-v1-1-d734bfa58ff4@oss.qualcomm.com
Signed-off-by: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
Add more information related to the transaction like cookie, vc_op,
salt when transaction times out and include similar information
when transaction salt does not match.
Info output for transaction timeout:
-------------------
(op:5015 cookie:45fe vc_op:5015 salt:45 timeout:60000ms)
-------------------
before it was:
-------------------
(op 5015, 60000ms)
-------------------
Signed-off-by: Manoj Vishwanathan <manojvishy@google.com>
Signed-off-by: Brian Vazquez <brianvv@google.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
When a workqueue is created with `WQ_UNBOUND`, its work items are
served by special worker-pools, whose host workers are not bound to
any specific CPU. In the default configuration (i.e. when
`queue_delayed_work` and friends do not specify which CPU to run the
work item on), `WQ_UNBOUND` allows the work item to be executed on any
CPU in the same node of the CPU it was enqueued on. While this
solution potentially sacrifices locality, it avoids contention with
other processes that might dominate the CPU time of the processor the
work item was scheduled on.
This is not just a theoretical problem: in a particular scenario
misconfigured process was hogging most of the time from CPU0, leaving
less than 0.5% of its CPU time to the kworker. The IDPF workqueues
that were using the kworker on CPU0 suffered large completion delays
as a result, causing performance degradation, timeouts and eventual
system crash.
Tested:
* I have also run a manual test to gauge the performance
improvement. The test consists of an antagonist process
(`./stress --cpu 2`) consuming as much of CPU 0 as possible. This
process is run under `taskset 01` to bind it to CPU0, and its
priority is changed with `chrt -pQ 9900 10000 ${pid}` and
`renice -n -20 ${pid}` after start.
Then, the IDPF driver is forced to prefer CPU0 by editing all calls
to `queue_delayed_work`, `mod_delayed_work`, etc... to use CPU 0.
Finally, `ktraces` for the workqueue events are collected.
Without the current patch, the antagonist process can force
arbitrary delays between `workqueue_queue_work` and
`workqueue_execute_start`, that in my tests were as high as
`30ms`. With the current patch applied, the workqueue can be
migrated to another unloaded CPU in the same node, and, keeping
everything else equal, the maximum delay I could see was `6us`.
Fixes: 0fe45467a1 ("idpf: add create vport and netdev configuration")
Signed-off-by: Marco Leogrande <leogrande@google.com>
Signed-off-by: Manoj Vishwanathan <manojvishy@google.com>
Signed-off-by: Brian Vazquez <brianvv@google.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Restore the call to idpf_vc_xn_shutdown() at the beginning of
idpf_vc_core_deinit() provided the function is not called on remove.
In the reset path the mailbox is destroyed, leading to all transactions
timing out.
Fixes: 09d0fb5cb3 ("idpf: deinit virtchnl transaction manager after vport and vectors")
Reviewed-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Add read memory barrier to ensure the order of operations when accessing
control queue descriptors. Specifically, we want to avoid cases where loads
can be reordered:
1. Load #1 is dispatched to read descriptor flags.
2. Load #2 is dispatched to read some other field from the descriptor.
3. Load #2 completes, accessing memory/cache at a point in time when the DD
flag is zero.
4. NIC DMA overwrites the descriptor, now the DD flag is one.
5. Any fields loaded before step 4 are now inconsistent with the actual
descriptor state.
Add read memory barrier between steps 1 and 2, so that load #2 is not
executed until load #1 has completed.
Fixes: 8077c72756 ("idpf: add controlq init and reset checks")
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
Suggested-by: Lance Richardson <rlance@google.com>
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
When using temperature measurement on Marvell 88Q2XXX devices and the
reset-gpios property is set in DT, the device does a hardware reset when
interface is brought down and up again. That means that the content of
the register MDIO_MMD_PCS_MV_TEMP_SENSOR2 is reset to default and that
leads to permanent deactivation of the temperature measurement, because
activation is done in mv88q2xxx_probe. To fix this move activation of
temperature measurement to mv88q222x_config_init.
Fixes: a557a92e68 ("net: phy: marvell-88q2xxx: add support for temperature sensor")
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Dimitri Fedrau <dima.fedrau@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://patch.msgid.link/20250118-marvell-88q2xxx-fix-hwmon-v2-1-402e62ba2dcb@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
When unload hclge driver, it tries to disable sriov first for each
ae_dev node from hnae3_ae_dev_list. If user unloads hns3 driver at
the time, because it removes all the ae_dev nodes, and it may cause
oops.
But we can't simply use hnae3_common_lock for this. Because in the
process flow of pci_disable_sriov(), it will trigger the remove flow
of VF, which will also take hnae3_common_lock.
To fixes it, introduce a new mutex to protect the unload process.
Fixes: 0dd8a25f35 ("net: hns3: disable sriov before unload hclge layer")
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Link: https://patch.msgid.link/20250118094741.3046663-1-shaojijie@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Fix wrong GDM4 register definition, in Airoha SDK GDM4 is defined at
offset 0x2400 but this doesn't make sense as it does conflict with the
CDM4 that is in the same location.
Following the pattern where each GDM base is at the FWD_CFG, currently
GDM4 base offset is set to 0x2500. This is correct but REG_GDM4_FWD_CFG
and REG_GDM4_SRC_PORT_SET are still using the SDK reference with the
0x2400 offset. Fix these 2 define by subtracting 0x100 to each register
to reflect the real address location.
Fixes: 23020f0493 ("net: airoha: Introduce ethernet support for EN7581 SoC")
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Acked-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Link: https://patch.msgid.link/20250120154148.13424-1-ansuelsmth@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pull networking updates from Paolo Abeni:
"This is slightly smaller than usual, with the most interesting work
being still around RTNL scope reduction.
Core:
- More core refactoring to reduce the RTNL lock contention, including
preparatory work for the per-network namespace RTNL lock, replacing
RTNL lock with a per device-one to protect NAPI-related net device
data and moving synchronize_net() calls outside such lock.
- Extend drop reasons usage, adding net scheduler, AF_UNIX, bridge
and more specific TCP coverage.
- Reduce network namespace tear-down time by removing per-subsystems
synchronize_net() in tipc and sched.
- Add flow label selector support for fib rules, allowing traffic
redirection based on such header field.
Netfilter:
- Do not remove netdev basechain when last device is gone, allowing
netdev basechains without devices.
- Revisit the flowtable teardown strategy, dealing better with fin,
reset and re-open events.
- Scale-up IP-vs connection dumping by avoiding linear search on each
restart.
Protocols:
- A significant XDP socket refactor, consolidating and optimizing
several helpers into the core
- Better scaling of ICMP rate-limiting, by removing false-sharing in
inet peers handling.
- Introduces netlink notifications for multicast IPv4 and IPv6
address changes.
- Add ipsec support for IP-TFS/AggFrag encapsulation, allowing
aggregation and fragmentation of the inner IP.
- Add sysctl to configure TIME-WAIT reuse delay for TCP sockets, to
avoid local port exhaustion issues when the average connection
lifetime is very short.
- Support updating keys (re-keying) for connections using kernel TLS
(for TLS 1.3 only).
- Support ipv4-mapped ipv6 address clients in smc-r v2.
- Add support for jumbo data packet transmission in RxRPC sockets,
gluing multiple data packets in a single UDP packet.
- Support RxRPC RACK-TLP to manage packet loss and retransmission in
conjunction with the congestion control algorithm.
Driver API:
- Introduce a unified and structured interface for reporting PHY
statistics, exposing consistent data across different H/W via
ethtool.
- Make timestamping selectable, allow the user to select the desired
hwtstamp provider (PHY or MAC) administratively.
- Add support for configuring a header-data-split threshold (HDS)
value via ethtool, to deal with partial or buggy H/W
implementation.
- Consolidate DSA drivers Energy Efficiency Ethernet support.
- Add EEE management to phylink, making use of the phylib
implementation.
- Add phylib support for in-band capabilities negotiation.
- Simplify how phylib-enabled mac drivers expose the supported
interfaces.
Tests and tooling:
- Make the YNL tool package-friendly to make it easier to deploy it
separately from the kernel.
- Increase TCP selftest coverage importing several packetdrill
test-cases.
- Regenerate the ethtool uapi header from the YNL spec, to ease
maintenance and future development.
- Add YNL support for decoding the link types used in net self-tests,
allowing a single build to run both net and drivers/net.
Drivers:
- Ethernet high-speed NICs:
- nVidia/Mellanox (mlx5):
- add cross E-Switch QoS support
- add SW Steering support for ConnectX-8
- implement support for HW-Managed Flow Steering, improving the
rule deletion/insertion rate
- support for multi-host LAG
- Intel (ixgbe, ice, igb):
- ice: add support for devlink health events
- ixgbe: add initial support for E610 chipset variant
- igb: add support for AF_XDP zero-copy
- Meta:
- add support for basic RSS config
- allow changing the number of channels
- add hardware monitoring support
- Broadcom (bnxt):
- implement TCP data split and HDS threshold ethtool support,
enabling Device Memory TCP.
- Marvell Octeon:
- implement egress ipsec offload support for the cn10k family
- Hisilicon (HIBMC):
- implement unicast MAC filtering
- Ethernet NICs embedded and virtual:
- Convert UDP tunnel drivers to NETDEV_PCPU_STAT_DSTATS, avoiding
contented atomic operations for drop counters
- Freescale:
- quicc: phylink conversion
- enetc: support Tx and Rx checksum offload and improve TSO
performances
- MediaTek:
- airoha: introduce support for ETS and HTB Qdisc offload
- Microchip:
- lan78XX USB: preparation work for phylink conversion
- Synopsys (stmmac):
- support DWMAC IP on NXP Automotive SoCs S32G2xx/S32G3xx/S32R45
- refactor EEE support to leverage the new driver API
- optimize DMA and cache access to increase raw RX performances
by 40%
- TI:
- icssg-prueth: add multicast filtering support for VLAN
interface
- netkit:
- add ability to configure head/tailroom
- VXLAN:
- accepts packets with user-defined reserved bit
- Ethernet switches:
- Microchip:
- lan969x: add RGMII support
- lan969x: improve TX and RX performance using the FDMA engine
- nVidia/Mellanox:
- move Tx header handling to PCI driver, to ease XDP support
- Ethernet PHYs:
- Texas Instruments DP83822:
- add support for GPIO2 clock output
- Realtek:
- 8169: add support for RTL8125D rev.b
- rtl822x: add hwmon support for the temperature sensor
- Microchip:
- add support for RDS PTP hardware
- consolidate periodic output signal generation
- CAN:
- several DT-bindings to DT schema conversions
- tcan4x5x:
- add HW standby support
- support nWKRQ voltage selection
- kvaser:
- allowing Bus Error Reporting runtime configuration
- WiFi:
- the on-going Multi-Link Operation (MLO) effort continues,
affecting both the stack and in drivers
- mac80211/cfg80211:
- Emergency Preparedness Communication Services (EPCS) station
mode support
- support for adding and removing station links for MLO
- add support for WiFi 7/EHT mesh over 320 MHz channels
- report Tx power info for each link
- RealTek (rtw88):
- enable USB Rx aggregation and USB 3 to improve performance
- LED support
- RealTek (rtw89):
- refactor power save to support Multi-Link Operations
- add support for RTL8922AE-VS variant
- MediaTek (mt76):
- single wiphy multiband support (preparation for MLO)
- p2p device support
- add TP-Link TXE50UH USB adapter support
- Qualcomm (ath10k):
- support for the QCA6698AQ IP core
- Qualcomm (ath12k):
- enable MLO for QCN9274
- Bluetooth:
- Allow sysfs to trigger hdev reset, to allow recovering devices
not responsive from user-space
- MediaTek: add support for MT7922, MT7925, MT7921e devices
- Realtek: add support for RTL8851BE devices
- Qualcomm: add support for WCN785x devices
- ISO: allow BIG re-sync"
* tag 'net-next-6.14' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1386 commits)
net/rose: prevent integer overflows in rose_setsockopt()
net: phylink: fix regression when binding a PHY
net: ethernet: ti: am65-cpsw: streamline TX queue creation and cleanup
net: ethernet: ti: am65-cpsw: streamline RX queue creation and cleanup
net: ethernet: ti: am65-cpsw: ensure proper channel cleanup in error path
ipv6: Convert inet6_rtm_deladdr() to per-netns RTNL.
ipv6: Convert inet6_rtm_newaddr() to per-netns RTNL.
ipv6: Move lifetime validation to inet6_rtm_newaddr().
ipv6: Set cfg.ifa_flags before device lookup in inet6_rtm_newaddr().
ipv6: Pass dev to inet6_addr_add().
ipv6: Convert inet6_ioctl() to per-netns RTNL.
ipv6: Hold rtnl_net_lock() in addrconf_init() and addrconf_cleanup().
ipv6: Hold rtnl_net_lock() in addrconf_dad_work().
ipv6: Hold rtnl_net_lock() in addrconf_verify_work().
ipv6: Convert net.ipv6.conf.${DEV}.XXX sysctl to per-netns RTNL.
ipv6: Add __in6_dev_get_rtnl_net().
net: stmmac: Drop redundant skb_mark_for_recycle() for SKB frags
net: mii: Fix the Speed display when the network cable is not connected
sysctl net: Remove macro checks for CONFIG_SYSCTL
eth: bnxt: update header sizing defaults
...
Pull kthread updates from Frederic Weisbecker:
"Kthreads affinity follow either of 4 existing different patterns:
1) Per-CPU kthreads must stay affine to a single CPU and never
execute relevant code on any other CPU. This is currently handled
by smpboot code which takes care of CPU-hotplug operations.
Affinity here is a correctness constraint.
2) Some kthreads _have_ to be affine to a specific set of CPUs and
can't run anywhere else. The affinity is set through
kthread_bind_mask() and the subsystem takes care by itself to
handle CPU-hotplug operations. Affinity here is assumed to be a
correctness constraint.
3) Per-node kthreads _prefer_ to be affine to a specific NUMA node.
This is not a correctness constraint but merely a preference in
terms of memory locality. kswapd and kcompactd both fall into this
category. The affinity is set manually like for any other task and
CPU-hotplug is supposed to be handled by the relevant subsystem so
that the task is properly reaffined whenever a given CPU from the
node comes up. Also care should be taken so that the node affinity
doesn't cross isolated (nohz_full) cpumask boundaries.
4) Similar to the previous point except kthreads have a _preferred_
affinity different than a node. Both RCU boost kthreads and RCU
exp kworkers fall into this category as they refer to "RCU nodes"
from a distinctly distributed tree.
Currently the preferred affinity patterns (3 and 4) have at least 4
identified users, with more or less success when it comes to handle
CPU-hotplug operations and CPU isolation. Each of which do it in its
own ad-hoc way.
This is an infrastructure proposal to handle this with the following
API changes:
- kthread_create_on_node() automatically affines the created kthread
to its target node unless it has been set as per-cpu or bound with
kthread_bind[_mask]() before the first wake-up.
- kthread_affine_preferred() is a new function that can be called
right after kthread_create_on_node() to specify a preferred
affinity different than the specified node.
When the preferred affinity can't be applied because the possible
targets are offline or isolated (nohz_full), the kthread is affine to
the housekeeping CPUs (which means to all online CPUs most of the time
or only the non-nohz_full CPUs when nohz_full= is set).
kswapd, kcompactd, RCU boost kthreads and RCU exp kworkers have been
converted, along with a few old drivers.
Summary of the changes:
- Consolidate a bunch of ad-hoc implementations of
kthread_run_on_cpu()
- Introduce task_cpu_fallback_mask() that defines the default last
resort affinity of a task to become nohz_full aware
- Add some correctness check to ensure kthread_bind() is always
called before the first kthread wake up.
- Default affine kthread to its preferred node.
- Convert kswapd / kcompactd and remove their halfway working ad-hoc
affinity implementation
- Implement kthreads preferred affinity
- Unify kthread worker and kthread API's style
- Convert RCU kthreads to the new API and remove the ad-hoc affinity
implementation"
* tag 'kthread-for-6.14-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/frederic/linux-dynticks:
kthread: modify kernel-doc function name to match code
rcu: Use kthread preferred affinity for RCU exp kworkers
treewide: Introduce kthread_run_worker[_on_cpu]()
kthread: Unify kthread_create_on_cpu() and kthread_create_worker_on_cpu() automatic format
rcu: Use kthread preferred affinity for RCU boost
kthread: Implement preferred affinity
mm: Create/affine kswapd to its preferred node
mm: Create/affine kcompactd to its preferred node
kthread: Default affine kthread to its preferred NUMA node
kthread: Make sure kthread hasn't started while binding it
sched,arm64: Handle CPU isolation on last resort fallback rq selection
arm64: Exclude nohz_full CPUs from 32bits el0 support
lib: test_objpool: Use kthread_run_on_cpu()
kallsyms: Use kthread_run_on_cpu()
soc/qman: test: Use kthread_run_on_cpu()
arm/bL_switcher: Use kthread_run_on_cpu()
We are missing netif_napi_del() and am65_cpsw_nuss_free_tx/rx_chns()
in error path when am65_cpsw_nuss_init_tx/rx_chns() is used anywhere
other than at probe(). i.e. am65_cpsw_nuss_update_tx_rx_chns and
am65_cpsw_nuss_resume()
As reported, in am65_cpsw_nuss_update_tx_rx_chns(),
if am65_cpsw_nuss_init_tx_chns() partially fails then
devm_add_action(dev, am65_cpsw_nuss_free_tx_chns,..) is added
but the cleanup via am65_cpsw_nuss_free_tx_chns() will not run.
Same issue exists for am65_cpsw_nuss_init_tx/rx_chns() failures
in am65_cpsw_nuss_resume() as well.
This would otherwise require more instances of devm_add/remove_action
and is clearly more of a distraction than any benefit.
So, drop devm_add/remove_action for am65_cpsw_nuss_free_tx/rx_chns()
and call am65_cpsw_nuss_free_tx/rx_chns() and netif_napi_del()
where required.
Reported-by: Siddharth Vadapalli <s-vadapalli@ti.com>
Closes: https://lore.kernel.org/all/m4rhkzcr7dlylxr54udyt6lal5s2q4krrvmyay6gzgzhcu4q2c@r34snfumzqxy/
Signed-off-by: Roger Quadros <rogerq@kernel.org>
Link: https://patch.msgid.link/20250117-am65-cpsw-streamline-v2-1-91a29c97e569@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
