Author: David Howells <dhowells@redhat.com>
Date: Fri Sep 11 10:25:04 2026 +0100
9p: Fix v9fs_issue_write() to update i_size and remote_i_size
commit c60ae98c5aa64021751b38ab1313b19d620bf640 upstream.
Fix v9fs_issue_write() to update i_size and remote_i_size to the new size
of the server file if we made it larger, using the start fpos and the count
returned by p9_client_write() to calculate the new minimum file size.
This assumes that if the 9P server makes a short write (say it hits
ENOSPC), a reduced count is returned.
Fixes: 5fb70e7275a6 ("netfs, 9p: Implement helpers for new write code")
Reported-by: Michael Mulqueen <mike@method-b.uk>
Closes: https://lore.kernel.org/r/fbb9e395-1e07-4212-8f70-23f3cd498074@method-b.uk/
Cc: stable@vger.kernel.org
Signed-off-by: David Howells <dhowells@redhat.com>
Message-ID: <2226525.1789118704@warthog.procyon.org.uk>
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Kuniyuki Iwashima <kuniyu@google.com>
Date: Sat Sep 12 03:07:51 2026 +0000
af_unix: Unify scc_index when finalising SCC in __unix_walk_scc().
[ Upstream commit 4a4263dfeabad72f95e8ab6e15146861fa4144dd ]
Commit bfdb01283ee8 ("af_unix: Assign a unique index to SCC.")
changed Tarjan's algorithm to update lowlink with lowlink,
which is called lowpoint (unix_vertex.scc_index).
unix_vertex_dead() assumes all vertices in an SCC share the same
lowpoint, but this is not always true if an SCC has two or more
back edges, depending on the order of DFS.
For example, the graph below has two back edges from B to A
and from C to B.
A --> B --> C
^ | ^ |
`----' `----'
If DFS walks through A -> B -> C -> B (-> C -> B) -> A (-> B -> A),
each index and scc_index will be updated as follows.
A --> B --> C C = (3, 3) (index, scc_index)
B = (2, 2)
A = (1, 1)
A ... B ... C C = (3, 2)<-.
^ | B = (2, 2) -'
`----' A = (1, 1)
A ... B ... C C = (3, 2)
^ | . . B = (2, 1)<-.
`----' .... A = (1, 1) -'
Then, unix_vertex_dead() thinks that B is passed to another
SCC with scc_index 2, and the SCC is not garbage-collected.
This does not happen if DFS walks in a different order below
or starts from B.
1 3
A --> B --> C
^ | ^ |
`----' `----'
2 4
Let's unify scc_index across the SCC when finalising it.
Note that updating v->index was previously done in unix_scc_dead(),
when called from __unix_walk_scc(), just to save one loop. Since
__unix_walk_scc() now iterates over the SCC anyway, the update is
moved back to __unix_walk_scc() and 'fast' argument is dropped.
Fixes: 4090fa373f0e ("af_unix: Replace garbage collection algorithm.")
Reported-by: James Burton <jamesburton@meta.com>
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260912030852.1467872-2-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Xiang Mei <xmei5@asu.edu>
Date: Mon Sep 14 00:43:24 2026 -0700
ALSA: 6fire: fix OOB write from device-reported iso length
[ Upstream commit 1589afe2d099d3e817873bc474676968d7080410 ]
usb6fire_pcm_in_urb_handler() sizes each outgoing isochronous packet as
(actual_length - 4) / (in_n_analog << 2) * (out_n_analog << 2) + 4, where
actual_length is the unsigned length the device reported for the matching
IN packet. A packet completed with status 0 and actual_length < 4 wraps
the subtraction to 0x7fffffec; a zero-length isochronous packet is legal
on the bus, and the preceding loop rejects only non-zero status. The sum
reaches memset() on out_urb->buffer, a 4832-byte object from
kcalloc(PCM_MAX_PACKET_SIZE, PCM_N_PACKETS_PER_URB).
Even without the wrap the result is out of bounds: at 88.2/96 kHz the
4-in/6-out scaling turns a full 420-byte IN packet into 628, so eight
packets span 5024 bytes of that buffer. usb_submit_urb() rejects an
over-long descriptor only after the memset() and the
usb6fire_pcm_playback() copy of user PCM data have run.
Guard the subtraction as the sibling usb6fire_pcm_capture() already does,
and limit the frame count to what fits in rt->out_packet_size, the OUT
endpoint's wMaxPacketSize. This bounds total_length by the buffer size
while keeping each packet length aligned to a whole output frame.
BUG: KASAN: out-of-bounds in usb6fire_pcm_in_urb_handler (sound/usb/6fire/pcm.c:338)
Write of size 18446744073709551456 at addr ffff88802a3d0000 by task vhci_rx/5018
Call Trace:
dump_stack_lvl (lib/dump_stack.c:94 lib/dump_stack.c:120)
print_report (mm/kasan/report.c:378 mm/kasan/report.c:482)
kasan_report (mm/kasan/report.c:595)
kasan_check_range (mm/kasan/generic.c:186 mm/kasan/generic.c:200)
__asan_memset (mm/kasan/shadow.c:84)
usb6fire_pcm_in_urb_handler (sound/usb/6fire/pcm.c:338)
__usb_hcd_giveback_urb (drivers/usb/core/hcd.c:1657)
usb_hcd_giveback_urb (drivers/usb/core/hcd.c:1741)
vhci_rx_loop (drivers/usb/usbip/vhci_rx.c:107 drivers/usb/usbip/vhci_rx.c:242)
kthread (kernel/kthread.c:436)
ret_from_fork (arch/x86/kernel/process.c:158)
ret_from_fork_asm (arch/x86/entry/entry_64.S:245)
Allocated by task 10:
__kmalloc_cache_noprof (mm/slub.c:5563)
usb6fire_pcm_init (sound/usb/6fire/pcm.c:560 sound/usb/6fire/pcm.c:595)
usb6fire_chip_probe (sound/usb/6fire/chip.c:133)
usb_probe_interface (drivers/usb/core/driver.c:399)
The buggy address belongs to the object at ffff88802a3d0000
which belongs to the cache kmalloc-8k of size 8192
The buggy address is located 0 bytes inside of
4832-byte region [ffff88802a3d0000, ffff88802a3d12e0)
Kernel panic - not syncing: Fatal exception in interrupt
Fixes: c6d43ba816d1 ("ALSA: usb/6fire - Driver for TerraTec DMX 6Fire USB")
Reported-by: co+855929c2df672879@bugs.sh
Closes: https://lore.kernel.org/all/gisnub8aWGLbyZLcDCSc7zWsHonMWGcyRgt5%40bugs.sh/
Assisted-by: LLM
Signed-off-by: Xiang Mei <xmei5@asu.edu>
Link: https://patch.msgid.link/20260914074324.3590843-1-xmei5@asu.edu
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Takashi Iwai <tiwai@suse.de>
Date: Thu Sep 10 17:52:23 2026 +0200
ALSA: bcd2000: Fix race between rawmidi and disconnect
[ Upstream commit 221253723dc58bb901c3f27a7659823e63fc598c ]
Although we tried to fix the potential UAF issues at USB disconnect on
bcd2000 driver, there is still an overlooked case -- namely, when a
rawmidi trigger callback has been already running at USB disconnect
handling, the in-flight function (e.g. bcd2000_midi_send()) could
still access the URB, because the previous URB NULL-check & clearance
was considered only for the URB complete callbacks, but not about the
parallel rawmidi operations.
For addressing the race, this patch introduced a new spinlock that
covers each rawmidi operation as well as the rawmidi handling in the
complete callback. The URB is cleared with the lock, so it guarantees
that the pending rawmidi task already finished or a NULL check is
effective.
Fixes: 459d3a64766f ("ALSA: bcd2000: clear the URB pointers on disconnect")
Link: https://patch.msgid.link/20260910155227.996210-1-tiwai@suse.de
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Takashi Iwai <tiwai@suse.de>
Date: Sat Sep 12 18:21:42 2026 +0200
ALSA: core: Fix potential UAF after asynchronous card release
commit fd95e68df6fe66344161a1329cbe5e5805e7b704 upstream.
Usually a sound driver releases the resources assigned to the card via
snd_card_free(), and it synchronizes with the whole release procedure.
However, when the card is released asynchronously via
snd_card_free_when_closed() like USB-audio driver, the situation is
slightly different; although the snd_card_disconnect() call at the
disconnection guarantees that any newer accesses will be gated, the
in-flight tasks might be still accessing to the underlying card->dev
device even after the disconnection, which would cause a
use-after-free in the end, as reported by fuzzers.
For addressing the bug above, this patch takes the refcount of
card->dev at initialization of the card object, and releases at its
destructor. This assures the availability of the card->dev in its
whole lifecycle.
Reported-by: Farhad Alemi <farhad.alemi@berkeley.edu>
Closes: https://lore.kernel.org/CA+0ovChexj4TrZL_2iG_P0WBEbZc5+73GfB3DkciQi=R8pZOnA@mail.gmail.com
Closes: https://lore.kernel.org/CA+0ovCgQUQNN=Z1tJTouiCsDaXR5M-3-SQEGk-cpPXQkM5Xh+w@mail.gmail.com
Cc: <stable@vger.kernel.org>
Link: https://patch.msgid.link/20260912162150.455144-1-tiwai@suse.de
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Slavin Liu <bolin.liu@seu.edu.cn>
Date: Sun Sep 13 20:51:54 2026 +0800
ALSA: hda: trace PCM open only after assigning a stream
[ Upstream commit c9e6e5f38bf75276605f1952b22285f5f3abcaff ]
Stream assignment can fail when hardware streams are exhausted.
Move the tracepoint after the NULL check because its payload accesses
the assigned stream tag.
Detected by static analysis and reviewed with AI-assisted source auditing.
Fixes: 184865085b88 ("ALSA: hda - rename hda_intel_trace.h to hda_controller_trace.h")
Assisted-by: LLM
Signed-off-by: Slavin Liu <bolin.liu@seu.edu.cn>
Link: https://patch.msgid.link/20260913125154.109944-1-bolin.liu@seu.edu.cn
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Nguyen Ngoc Thang <ngocthang2710.1999@gmail.com>
Date: Sun Sep 13 20:44:46 2026 +0700
ALSA: pcm: set timer->private_data before registering the PCM timer
[ Upstream commit 1e713f9bb2ac583521f06b0eb4e22440b1e3d078 ]
snd_pcm_timer_init() calls snd_device_register() to link the new
struct snd_timer into the global timer list while it still carries
hw.c_resolution = snd_pcm_timer_resolution (and hw.start/hw.stop),
and only afterwards sets timer->private_data = substream.
Once the timer is on the list under register_mutex, a concurrent
reader can already reach it through the same mutex and invoke these
callbacks. /proc/asound/timers does this via c_resolution(), and
snd_timer_open()+snd_timer_start() reach start()/stop() the same way.
All three dereference timer->private_data, which for this brief
window is NULL, giving a NULL-pointer dereference:
substream = timer->private_data;
return substream->runtime ? ... // substream is NULL
Move the private_data/private_free assignment before
snd_device_register() so the timer is never visible on the list
without its private_data set. On the snd_device_register() failure
path, private_free() (snd_pcm_timer_free()) can now run, but it only
does substream->timer = NULL, which is already NULL at that point
since substream->timer is set to the new timer just once, after a
successful registration -- so the failure path stays safe.
Reported-by: syzbot+19da64013c46df87f971@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=19da64013c46df87f971
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Nguyen Ngoc Thang <ngocthang2710.1999@gmail.com>
Link: https://patch.msgid.link/20260913134446.114724-1-ngocthang2710.1999@gmail.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Xiang Mei <xmei5@asu.edu>
Date: Sat Sep 12 13:05:30 2026 -0700
ALSA: usb-audio: Clamp implicit feedback packet count to URB capacity
commit 76a986c980bb502c7688d605ac7a67fd257a9a1b upstream.
data_ep_set_params() allocates each data URB for exactly u->packets
isochronous frames, so urb->iso_frame_desc[] has u->packets slots and
ctx->packets is the driver's only record of that limit. For an implicit
feedback sink, snd_usb_queue_pending_output_urbs() overwrites it with the
sync source's packet count, which is calculated independently from the
capture endpoint's parameters. When that count is larger,
prepare_playback_urb() and prepare_silent_urb() can write
iso_frame_desc[] past the allocation; their existing bounds limit payload
bytes, not the descriptor index.
The reproducer uses a high-speed UAC2 device declaring bInterval 1 for
implicit feedback capture (8 packets) and bInterval 4 for playback
(1 packet). On the first capture completion after the stream starts, it
accesses seven descriptors spanning 112 bytes beyond the one-packet URB:
BUG: KASAN: slab-out-of-bounds in prepare_playback_urb (sound/usb/pcm.c:1560)
Write of size 4 at addr ffff88801e696ad0 by task vhci_rx/178
prepare_playback_urb (sound/usb/pcm.c:1560)
prepare_outbound_urb (sound/usb/endpoint.c:340)
snd_usb_queue_pending_output_urbs (sound/usb/endpoint.c:501)
snd_complete_urb (sound/usb/endpoint.c:1834)
__usb_hcd_giveback_urb (drivers/usb/core/hcd.c:1657)
usb_hcd_giveback_urb (drivers/usb/core/hcd.c:1741)
vhci_rx_loop (drivers/usb/usbip/vhci_rx.c:107)
kthread (kernel/kthread.c:436)
The buggy address belongs to the object at ffff88801e696a00
which belongs to the cache kmalloc-256 of size 256
The buggy address is located 0 bytes to the right of
allocated 208-byte region [ffff88801e696a00, ffff88801e696ad0)
Record the allocated packet count per endpoint and clamp both the adopted
count and the packet-size copy to it. Fold the Format Type II delimiter
into urb_packs before the allocation loop so the recorded limit matches
every URB.
Fixes: cf044e441902 ("ALSA: usb-audio: Update the number of packets properly at receiving")
Reported-by: co+8eacd4fa193b1b28@bugs.sh
Closes: https://lore.kernel.org/all/22xPn8drvIUtYgVeQnBiNqXuevOTpBAjepLz%40bugs.sh/
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-5
Signed-off-by: Xiang Mei <xmei5@asu.edu>
Link: https://patch.msgid.link/20260912200530.1955491-1-xmei5@asu.edu
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Takashi Iwai <tiwai@suse.de>
Date: Thu Sep 3 18:04:39 2026 +0200
ALSA: usb: 6fire: Avoid embedded URBs
[ Upstream commit 9fe49dbc023e82dfaee7b245997d820d01742a9a ]
The USB 6fire driver uses URBs embedded in different structs for PCM,
MIDI and communication, and this is basically a buggy implementation
nowadays; since a URB is managed with a refcount, this may lead to a
UAF when the URB is released asynchronously.
For addressing the problem, this patch converts those embedded URBs to
ones that are properly allocated via usb_alloc_urb(). The
pcm_urb.packets[] is gone, as it's allocated by usb_alloc_urb(), hence
it's found in urb.iso_frame_desc[] instead.
The conversions are rather straightforward; each embedded struct urb
is changed to a pointer, and its callers are updated accordingly.
The resource for those structs are released in the common destructor
functions (usb6fire_comm_free(), etc), which are called at both the
init error path and the disconnect.
No functional changes, only compile-tested.
Link: https://lore.kernel.org/20260903130757.0668310a.michal.pecio@gmail.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Link: https://patch.msgid.link/20260903160458.1938392-4-tiwai@suse.de
Stable-dep-of: 1589afe2d099 ("ALSA: 6fire: fix OOB write from device-reported iso length")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Yuho Choi <oss.patchbox@gmail.com>
Date: Thu Sep 10 23:11:21 2026 -0400
ALSA: virtio: reset device before deleting virtqueues
commit 6c05d00af307560e6a9f1631d6270d3df5aa2272 upstream.
virtsnd_remove() and virtsnd_freeze() delete the virtqueues before
resetting the device. del_vqs() frees the vring backing, but does not
provide a generic device quiesce operation. In particular, modern
virtio-pci keeps enabled queues active until the device is reset.
Reset the device before deleting the virtqueues so it can no longer
access the vring memory when that memory is released. This also covers
probe failures after DRIVER_OK, which unwind through virtsnd_remove().
Fixes: de3a9980d8c3 ("ALSA: virtio: add virtio sound driver")
Fixes: 575483e90a32 ("ALSA: virtio: introduce device suspend/resume support")
Cc: stable@vger.kernel.org
Signed-off-by: Yuho Choi <oss.patchbox@gmail.com>
Link: https://patch.msgid.link/20260911031121.1542502-1-oss.patchbox@gmail.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Lucas Tanure <tanure@linux.com>
Date: Sat Aug 29 10:47:58 2026 +0100
arm64: dts: amlogic: t7: fix the pin groups of the vsync PWM
[ Upstream commit 406292fd75f95aa3010fec95b5beb5a8b7e3ba3a ]
The vsync PWM output can appear on either of two pins, but the
description named a single group that does not exist, so anything using
it would refuse to start.
Name the real groups instead, one entry per pin. No board describes this
output yet, so nothing changes today.
Fixes: 2a2a7b9701a7 ("arm64: dts: amlogic: t7: Add PWM pinctrl nodes")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Lucas Tanure <tanure@linux.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patch.msgid.link/20260829094758.23248-5-tanure@linux.com
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lucas Tanure <tanure@linux.com>
Date: Sat Aug 29 10:47:56 2026 +0100
arm64: dts: amlogic: t7: fix the pin groups of two PWM outputs
[ Upstream commit ae7be5c58a819259e5eb98a3abfa60842478f2ae ]
Two of the PWM outputs can each appear on more than one pin, but the
description named a single group that does not exist, so anything using
it refused to start.
Name the real groups instead, one entry per pin, the same way the other
multi-pin PWM output is already described.
Fixes: 2a2a7b9701a7 ("arm64: dts: amlogic: t7: Add PWM pinctrl nodes")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Lucas Tanure <tanure@linux.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patch.msgid.link/20260829094758.23248-3-tanure@linux.com
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lucas Tanure <tanure@linux.com>
Date: Sat Aug 29 10:47:57 2026 +0100
arm64: dts: amlogic: t7: khadas-vim4: add the PWM-driven supplies
[ Upstream commit 1e5a53bd16ac501e68463e84023d1bff543cc696 ]
The board powers its two CPU clusters, the GPU, the NPU, the memory and
the always-on domain from regulators steered by PWM outputs. None of
them were described, so Linux treated those outputs as unused and
switched them off part way through boot. The supplies then drifted away
from the levels the bootloader had set, which showed up as random hangs
and memory corruption.
Describe each supply so it has an owner and is left alone.
The voltage ranges are read off the feedback networks on the board
schematic. VDDNPU is deliberately not the range in Amlogic's own
device tree: this board fits a different feedback resistor, which
puts it about 40mV higher at both ends.
Fixes: 8c45bf9ae4ef ("arm64: dts: amlogic: t7: khadas-vim4: Add power regulators")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Lucas Tanure <tanure@linux.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patch.msgid.link/20260829094758.23248-4-tanure@linux.com
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lucas Tanure <tanure@linux.com>
Date: Sun Aug 23 12:53:35 2026 +0100
arm64: dts: amlogic: t7: khadas-vim4: allow the SD card to be power cycled
[ Upstream commit c793a084ab3088744516aef437b4239d0d98623f ]
SD cards start at 3.3V and switch to 1.8V to reach UHS-I speeds. Some
cards refuse that switch, and the SD specification says the only way
to recover is to power the card off and start again.
SD_3V3 is marked regulator-always-on, so the supply never goes off,
the card stays stuck half way through the switch, and the MMC core
retries forever:
mmc1: error -95 whilst initialising SD card
Drop regulator-always-on. regulator-boot-on still turns the supply on
at boot, and a card that refuses the switch now falls back to high
speed instead of failing to initialise.
Fixes: 8c45bf9ae4ef ("arm64: dts: amlogic: t7: khadas-vim4: Add power regulators")
Signed-off-by: Lucas Tanure <tanure@linux.com>
Assisted-by: Claude:claude-fable-5
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patch.msgid.link/20260823115335.102219-3-tanure@linux.com
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lucas Tanure <tanure@linux.com>
Date: Sun Aug 23 12:53:34 2026 +0100
arm64: dts: amlogic: t7: use the real UART pclk
[ Upstream commit 882de800bf583dd15969836088789edbf035a944 ]
uart_a listed the 24MHz crystal for all three of its clocks because the
T7 clock controller driver did not exist when these boards were added.
That leaves the real UART bus clock without a user, so the kernel
turns it off when it disables unused clocks at the end of boot, and
the board hangs.
Update the board DTS files to point uart_a's pclk to CLKID_SYS_UART_A
instead of the dummy crystal clock.
Fixes: 4fef056588f5 ("arm64: dts: amlogic-t7-a311d2-khadas-vim4: add initial device-tree")
Fixes: 6f048cc7a635 ("arm64: dts: add board AN400")
Signed-off-by: Lucas Tanure <tanure@linux.com>
Assisted-by: Claude:claude-fable-5
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patch.msgid.link/20260823115335.102219-2-tanure@linux.com
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Koichiro Den <den@valinux.co.jp>
Date: Fri Sep 11 16:30:58 2026 +0900
arm64: dts: renesas: r8a779f0: Set UFS lane count
commit 8dc2615d5702059b2b71fca6f93c0d7d10ae54cb upstream.
Since commit e72323f3b09f ("scsi: ufs: core: Configure only active lanes
during link"), the following error is observed on R-Car S4:
ufshcd-renesas e6860000.ufs: Tx lane mismatch [config,reported] [2,1]
ufshcd-renesas e6860000.ufs: link startup failed -67
ufshcd-renesas e6860000.ufs: error -ENOLINK: Initialization failed with error -67
ufshcd-renesas e6860000.ufs: probe with driver ufshcd-renesas failed with error -67
R-Car S4 has one UFS lane per direction, as described in section 152.1
of its hardware manual. Without lanes-per-direction, the UFS platform
driver defaults to two lanes.
Previously, the core used PA_CONNECTEDRXDATALANES and
PA_CONNECTEDTXDATALANES to configure the link without checking them
against lanes-per-direction, so the missing property did not prevent
initialization.
Explicitly set lanes-per-direction to 1, now that the validation is in
place.
Fixes: 5235d551779d ("arm64: dts: renesas: r8a779f0: Add UFS node")
Cc: stable@vger.kernel.org # 7.2+
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://patch.msgid.link/20260911073058.253000-1-den@valinux.co.jp
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Ovidiu Panait <ovidiu.panait.rb@renesas.com>
Date: Wed Jul 22 08:53:51 2026 +0000
arm64: dts: renesas: r9a09g047: Switch GBETH TX queue scheduling to WRR
[ Upstream commit 63016c3a91f2c458ca75869c8c782e899591f22d ]
The GBETH ethernet nodes don't specify a TX scheduling policy, so stmmac
falls back to Strict Priority. In this configuration the queue with the
highest priority gets all the traffic, starving the others under load.
Under sustained UDP TX load with multiple data streams, this starvation
triggers spurious adapter resets due to TX queue timeouts:
iperf3 -c <ip> -i0 -t60 --bind-dev end0 -u -b0 -P4
end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms
end0: Reset adapter.
Investigation shows that only the highest priority queue is advancing
while the others stall for more than 5 seconds, causing a netdev watchdog
reset.
Switch the TX scheduling policy to Weighted-Round-Robin (WRR) so that
traffic is processed across all queues, eliminating the stalls.
Fixes: 41ffbb1c42d3 ("arm64: dts: renesas: r9a09g047: Add GBETH nodes")
Signed-off-by: Ovidiu Panait <ovidiu.panait.rb@renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Tested-by: Tommaso Merciai <tommaso.merciai.xr@bp.renesas.com>
Link: https://patch.msgid.link/20260722085353.136986-4-ovidiu.panait.rb@renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ovidiu Panait <ovidiu.panait.rb@renesas.com>
Date: Wed Jul 22 08:53:50 2026 +0000
arm64: dts: renesas: r9a09g056: Switch GBETH TX queue scheduling to WRR
[ Upstream commit 66fcbdbeca0118b8aeac218b33fa18c394513543 ]
The GBETH ethernet nodes don't specify a TX scheduling policy, so stmmac
falls back to Strict Priority. In this configuration the queue with the
highest priority gets all the traffic, starving the others under load.
Under sustained UDP TX load with multiple data streams, this starvation
triggers spurious adapter resets due to TX queue timeouts:
iperf3 -c <ip> -i0 -t60 --bind-dev end0 -u -b0 -P4
end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms
end0: Reset adapter.
Investigation shows that only the highest priority queue is advancing
while the others stall for more than 5 seconds, causing a netdev watchdog
reset.
Switch the TX scheduling policy to Weighted-Round-Robin (WRR) so that
traffic is processed across all queues, eliminating the stalls.
Fixes: c8c8a57c5b40 ("arm64: dts: renesas: r9a09g056: Add GBETH nodes")
Signed-off-by: Ovidiu Panait <ovidiu.panait.rb@renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://patch.msgid.link/20260722085353.136986-3-ovidiu.panait.rb@renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ovidiu Panait <ovidiu.panait.rb@renesas.com>
Date: Wed Jul 22 08:53:49 2026 +0000
arm64: dts: renesas: r9a09g057: Switch GBETH TX queue scheduling to WRR
[ Upstream commit 33da68f61d25ef8411489d06514ff627c1f88152 ]
The GBETH ethernet nodes don't specify a TX scheduling policy, so stmmac
falls back to Strict Priority. In this configuration the queue with the
highest priority gets all the traffic, starving the others under load.
Under sustained UDP TX load with multiple data streams, this starvation
triggers spurious adapter resets due to TX queue timeouts:
iperf3 -c <ip> -i0 -t60 --bind-dev end0 -u -b0 -P4
end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms
end0: Reset adapter.
Investigation shows that only the highest priority queue is advancing
while the others stall for more than 5 seconds, causing a netdev watchdog
reset.
Switch the TX scheduling policy to Weighted-Round-Robin (WRR) so that
traffic is processed across all queues, eliminating the stalls.
Fixes: 050ee38d0002 ("arm64: dts: renesas: r9a09g057: Add GBETH nodes")
Signed-off-by: Ovidiu Panait <ovidiu.panait.rb@renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://patch.msgid.link/20260722085353.136986-2-ovidiu.panait.rb@renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ovidiu Panait <ovidiu.panait.rb@renesas.com>
Date: Wed Jul 22 08:53:52 2026 +0000
arm64: dts: renesas: r9a09g077: Switch GBETH TX queue scheduling to WRR
[ Upstream commit f9d9a1913c09366bf7b967d39575e06cb80128dc ]
The GBETH ethernet nodes don't specify a TX scheduling policy, so stmmac
falls back to Strict Priority. In this configuration the queue with the
highest priority gets all the traffic, starving the others under load.
Under sustained UDP TX load with multiple data streams, this starvation
triggers spurious adapter resets due to TX queue timeouts:
iperf3 -c <ip> -i0 -t60 --bind-dev end0 -u -b0 -P4
end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms
end0: Reset adapter.
Investigation shows that only the highest priority queue is advancing
while the others stall for more than 5 seconds, causing a netdev watchdog
reset.
Switch the TX scheduling policy to Weighted-Round-Robin (WRR) so that
traffic is processed across all queues, eliminating the stalls.
Fixes: 394c1e24a4cf ("arm64: dts: renesas: r9a09g077: Add GMAC nodes")
Signed-off-by: Ovidiu Panait <ovidiu.panait.rb@renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://patch.msgid.link/20260722085353.136986-5-ovidiu.panait.rb@renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ovidiu Panait <ovidiu.panait.rb@renesas.com>
Date: Wed Jul 22 08:53:53 2026 +0000
arm64: dts: renesas: r9a09g087: Switch GBETH TX queue scheduling to WRR
[ Upstream commit 2ac7bad110be6ebe478d6bd57821f7f5259a1f54 ]
The GBETH ethernet nodes don't specify a TX scheduling policy, so stmmac
falls back to Strict Priority. In this configuration the queue with the
highest priority gets all the traffic, starving the others under load.
Under sustained UDP TX load with multiple data streams, this starvation
triggers spurious adapter resets due to TX queue timeouts:
iperf3 -c <ip> -i0 -t60 --bind-dev end0 -u -b0 -P4
end0: NETDEV WATCHDOG: CPU: 1: transmit queue 1 timed out 5228 ms
end0: Reset adapter.
Investigation shows that only the highest priority queue is advancing
while the others stall for more than 5 seconds, causing a netdev watchdog
reset.
Switch the TX scheduling policy to Weighted-Round-Robin (WRR) so that
traffic is processed across all queues, eliminating the stalls.
Fixes: c4698a34993b ("arm64: dts: renesas: r9a09g087: Add GMAC nodes")
Signed-off-by: Ovidiu Panait <ovidiu.panait.rb@renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://patch.msgid.link/20260722085353.136986-6-ovidiu.panait.rb@renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Dinh Nguyen <dinguyen@kernel.org>
Date: Fri Aug 28 10:31:29 2026 +0800
arm64: dts: socfpga: change access permission from 755 to 644
commit 5f8cb07d7a8573753c5d31dc76b51cab8e5e9460 upstream.
These files have an incorrect access permission of 755 instead of 644.
Change them to the correct access permission of 644.
Fixes: 4bc04eb90b7c ("arm64: dts: socfpga: stratix10: Add emmc support")
Cc: <stable@vger.kernel.org> # v7.1+
Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Breno Leitao <leitao@debian.org>
Date: Thu Sep 10 06:53:27 2026 -0700
arm64: hibernate: clone only the linear map that exists at runtime
[ Upstream commit e4a6f57d22e079e23fafac51057fad534160b269 ]
This is similar to commit 1537e55728ec2 ("arm64: trans_pgd: clone only
the linear map that exists at runtime"), but in a different place.
swsusp_arch_resume() clones the kernel linear map with
trans_pgd_create_copy(..., PAGE_OFFSET, PAGE_END). PAGE_OFFSET comes
from the compile-time VA_BITS, so a CONFIG_ARM64_VA_BITS_52 kernel
booting on hardware without LPA2 -- vabits_actual is 48 and the fifth
level is folded -- hands the walk a 3.9PB window while its linear map
only spans the top 128TB.
On a VA_BITS_52 4k kernel with CONFIG_KASAN_GENERIC in a 4GB VM, I see:
swapper/0: page allocation failure: order:0, mode:0x920(GFP_ATOMIC|__GFP_ZERO)
hibernate_page_alloc+0x10/0x1c
swsusp_arch_resume+0x70/0x320
hibernation_restore+0xa4/0x138
software_resume+0x15c/0x270
PM: hibernation: Failed to load image, recovering.
PM: hibernation: resume failed (-12)
Fix it by copying the linear map that is the actual one, not the
compiled one.
Fixes: a6bbf5d4d9d1 ("arm64: mm: Add definitions to support 5 levels of paging")
Signed-off-by: Breno Leitao <leitao@debian.org>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Bradley Morgan <include@grrlz.net>
Date: Sun Aug 9 21:36:15 2026 +0000
arm64: hibernate: pass HVC_SET_VECTORS args to the resume hvc
commit 955d86e5f3b95b731991fdb84966c50b16314629 upstream.
swsusp_arch_suspend_exit() reinstalls the restored kernel's hyp stub
vectors with an hvc, but never passes the arguments. x0 is not set to
HVC_SET_VECTORS and x1 is not set to the vector address, so the stub
dispatch falls through and returns without writing vbar_el2. EL2 is
left pointing at the trans_pgd copy of the vectors, a page that
swsusp_free() releases right after resume.
Set the arguments up the same way __hyp_set_vectors() does.
Without this fix, Vladimir was able to trigger a hang when resuming from
hibernation with CONFIG_PAGE_POISONING=y and page_poison=on.
Fixes: 788bfdd97434 ("arm64: trans_pgd: hibernate: Add trans_pgd_copy_el2_vectors")
Cc: stable@vger.kernel.org
Signed-off-by: Bradley Morgan <include@grrlz.net>
Reviewed-by: Vladimir Murzin <vladimir.murzin@arm.com>
Tested-by: Vladimir Murzin <vladimir.murzin@arm.com>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Catalin Marinas <catalin.marinas@arm.com>
Date: Mon Sep 7 12:30:25 2026 +0100
arm64: mte: Fix PTRACE_{PEEK,POKE}MTETAGS error documentation
[ Upstream commit fdb9ebc7fb7788b8371fbef6c17dc5c8291e1429 ]
PTRACE_{PEEK,POKE}MTETAGS return -EIO rather than -EOPNOTSUPP (as
documented) when no tags are copied from/to a mapping without PROT_MTE.
This has been the behaviour since the interface was introduced, though
the original intent was to distinguish between address not being
accessible and mapped as untagged.
Update the documentation to match the implementation (de-facto ABI).
Since -EOPNOTSUPP was never returned, change the error assignment to
-EIO as well to avoid confusion.
Fixes: df9d7a22dd21 ("arm64: mte: Add Memory Tagging Extension documentation")
Fixes: 18ddbaa02b7a ("arm64: mte: ptrace: Add PTRACE_{PEEK,POKE}MTETAGS support")
Reported-by: Yury Khrustalev <yury.khrustalev@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Mark Rutland <mark.rutland@arm.com>
Date: Tue Sep 8 16:17:23 2026 +0100
arm64: percpu: Fix LSE operations on {8,16}-bit types
commit 8cf2093f5372952a9ebc805c418d45df7112cd14 upstream.
The assembly for __percpu_##name##_case_##sz() and
__percpu_##name##_return_case_##sz() doesn't use the 'sfx' macro
argument to form the LSE instruction. Without 'sfx', a W register
argument will imply a 32-bit memory location, and consequently
{8,16}-bit ops will erroneously read and write 32 bits of memory when
the LSE instruction is used.
Fix this by appending 'sfx' to 'op_lse' to LSE instruction. It is not
necessary (and not valid) to append 'sfx' to 'op_llsc', as 'op_llsc' is
a register-register operation which does not access memory (and does not
take a size suffix).
Fixes: 959bf2fd03b5 ("arm64: percpu: Rewrite per-cpu ops to allow use of LSE atomics")
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Jinjie Ruan <ruanjinjie@huawei.com>
Cc: Ada Couprie Diaz <ada.coupriediaz@arm.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Vladimir Murzin <vladimir.murzin@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Yang Shi <yang@os.amperecomputing.com>
Cc: stable@vger.kernel.org
Reviewed-by: Vladimir Murzin <vladimir.murzin@arm.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Mark Rutland <mark.rutland@arm.com>
Date: Tue Sep 8 16:17:22 2026 +0100
arm64: percpu: Fix this_cpu_and() mask generation
commit 44274c657256b4911de82f8104e9e22f054cf742 upstream.
The arm64 implementation of this_cpu_and(pcp, val) is built in terms of
ANDNOT operations, which requires the 'val' argument to be bitwise
negated. The bitwise negation is not implemented correctly, with two
bugs described below.
(1) The bitwise negation is performed as '~val' rather than '~(val)'.
This won't always generate the expected value when 'val' is an
expression.
For example, for this_cpu_and(pcp, 1 - 1):
* 'val' is '1 - 1' ===> (int) 0x00000000
* '~val' is '~1 - 1' ===> (int) 0xfffffffd
* '~(val)' is '~(1 - 1)' ===> (int) 0xffffffff
... and thus bit[1] of 'pcp' would be preserved unexpectedly by the
ANDNOT operation.
(2) The bitwise negation is performed on 'val' before it has been cast
to (at least) the width of 'pcp'. This won't always generate the
expected value for the upper bits.
For example, for this_cpu_and(pcp, zero), where 'pcp' is a u64 and
'zero' is a u32:
* 'zero' ===> (u32) 0x00000000
* '~(zero)' ===> (u32) 0xffffffff
* '(u64)~(zero)' ===> (u64) 0x00000000ffffffff
* '~((u64)(zero))' ===> (u64) 0xffffffffffffffff
... and thus bits[63:32] of 'pcp' would be preserved unexpectedly by
the ANDNOT operation.
Fix these issues by adding brackets around 'val', and by casting 'val'
to an appropriately-sized type before bitwise negation.
Fixes: 959bf2fd03b5 ("arm64: percpu: Rewrite per-cpu ops to allow use of LSE atomics")
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Jinjie Ruan <ruanjinjie@huawei.com>
Tested-by: Muhammad Usama Anjum <usama.anjum@arm.com>
Acked-by: Christopher Lameter (Ampere) <cl@gentwo.org>
Cc: Ada Couprie Diaz <ada.coupriediaz@arm.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Vladimir Murzin <vladimir.murzin@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Yang Shi <yang@os.amperecomputing.com>
Cc: stable@vger.kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Mark Rutland <mark.rutland@arm.com>
Date: Tue Sep 8 16:17:21 2026 +0100
arm64: percpu: Fix this_cpu_write() casting
commit 885bff055a0f251a51a0d4fd4f0a7b525582a3de upstream.
The arm64 implementation of this_cpu_write() casts 'val' to unsigned
long. This is necessary to handle cases where 'val' is a pointer type,
and to avoid spurious compiler warnings for the (unreachable!) cases
where the pointer type would be cast to a smaller integer type.
Unfortunately, the cast is applied to 'val' rather than '(val)', which
won't always generate the expected value when 'val' is an expression.
For example, for this_cpu_write(pcp, zero - 1), where 'pcp' is a u64 and
'zero' is a u32:
* 'zero' ===> (u32) 0x00000000
* 'zero - 1' ===> (u32) 0xffffffff
* '(unsigned long)zero - 1' ===> (u64) 0xffffffffffffffff
* '(unsigned long)(zero - 1)' ===> (u64) 0x00000000ffffffff
Fix this by adding brackets around 'val'.
Fixes: 959bf2fd03b5 ("arm64: percpu: Rewrite per-cpu ops to allow use of LSE atomics")
Reported-by: David Laight <david.laight.linux@gmail.com>
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: David Laight <david.laight.linux@gmail.com>
Reviewed-by: Jinjie Ruan <ruanjinjie@huawei.com>
Tested-by: Muhammad Usama Anjum <usama.anjum@arm.com>
Acked-by: Christopher Lameter (Ampere) <cl@gentwo.org>
Cc: Ada Couprie Diaz <ada.coupriediaz@arm.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Vladimir Murzin <vladimir.murzin@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Yang Shi <yang@os.amperecomputing.com>
Cc: stable@vger.kernel.org
Reviewed-by: Lorenzo Stoakes (ARM) <ljs@kernel.org>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Pengpeng Hou <pengpeng@iscas.ac.cn>
Date: Sun Aug 30 22:22:44 2026 +0800
ARM: socfpga: select the PL310 erratum 753970 workaround
[ Upstream commit cfc1e9a543e3589ba200795b6e7fd8ef4314efdf ]
ARCH_INTEL_SOCFPGA selects CACHE_L2X0 and several PL310 erratum
workarounds. The 753970 workaround is still conditioned on PL310, but that
Kconfig symbol no longer exists, so this one selection is always disabled.
Select PL310_ERRATA_753970 directly, consistently with the other PL310
workarounds required by the platform.
Fixes: fbc125afdc50 ("ARM: socfpga: Turn on ARM errata for L2 cache")
Signed-off-by: Pengpeng Hou <pengpeng@iscas.ac.cn>
Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Richard Fitzgerald <rf@opensource.cirrus.com>
Date: Thu Sep 10 12:44:57 2026 +0100
ASoC: Add codec_ch_mask to snd_soc_dai_link_ch_map
[ Upstream commit 88b14c0d0bab5c0f3e7c641f274e3c70210c0e36 ]
Add a codec_ch_mask member to snd_soc_dai_link_ch_map.
The CPU and codec channel masks are not necessarily the same, and are
quite likely different. SoundWire and I2S/TDM both support assigning
different sample slots to each codec, so for example channel 0 on each
codec could map to different channels at the CPU.
It is also possible for one TX channel to map to multiple RX channels.
So it isn't _always_ safe to assume that the total number of set bits
in the CPU ch_mask is the same as the total number of enabled channels
on the codec.
For example consider this mapping on a capture stream:
CPU0 CODEC0 cpu_ch_mask = 0x03
CPU1 CODEC0 cpu_ch_mask = 0x03
This could be either four TX channels on the codec split across two
receiving CPUs, or two TX channels on the codec duplicated to two CPUs.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://patch.msgid.link/20260910114500.1586637-3-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Stable-dep-of: 6b382bdfe26a ("ASoC: soc-pcm: Apply snd_soc_dai_link_ch_map.codec_ch_mask to codec params")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Vijendar Mukunda <Vijendar.Mukunda@amd.com>
Date: Thu Sep 10 21:46:46 2026 +0530
ASoC: amd: acp: bounds-check SoundWire link ID in machine drivers
[ Upstream commit 29218a4d11a31a8157389bc2b9e62dd768d7ea42 ]
Add a bounds check in create_sdw_dailink() to validate that the
SoundWire link ID derived from link_mask does not exceed the maximum
supported by the platform. If the link ID is out of range or link_mask
is zero, log an error and return -EINVAL to prevent accessing invalid
CPU pin ID tables.
Applied to both acp-sdw-sof-mach.c and acp-sdw-legacy-mach.c.
Fixes: 6d8348ddc56e ("ASoC: amd: acp: refactor SoundWire machine driver code")
Signed-off-by: Vijendar Mukunda <Vijendar.Mukunda@amd.com>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Link: https://patch.msgid.link/20260910161728.1452808-2-Vijendar.Mukunda@amd.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Vijendar Mukunda <Vijendar.Mukunda@amd.com>
Date: Thu Sep 10 21:46:48 2026 +0530
ASoC: amd: acp: fix ffs() operator precedence for SoundWire link ID
[ Upstream commit 27098aaf28b96ab4e6891709062c343566d4882b ]
ffs(link_mask - 1) computes ffs on (link_mask - 1) instead of
subtracting 1 from the result of ffs(link_mask). For a typical
power-of-2 link_mask this returns the wrong link ID, causing cpu_pin_id
lookup to select the incorrect SoundWire manager.
Fix the operator precedence to ffs(link_mask) - 1 in both
acp-sdw-sof-mach.c and acp-sdw-legacy-mach.c.
Fixes: 6d8348ddc56e ("ASoC: amd: acp: refactor SoundWire machine driver code")
Signed-off-by: Vijendar Mukunda <Vijendar.Mukunda@amd.com>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Link: https://patch.msgid.link/20260910161728.1452808-4-Vijendar.Mukunda@amd.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Vijendar Mukunda <Vijendar.Mukunda@amd.com>
Date: Thu Sep 10 21:46:47 2026 +0530
ASoC: amd: acp: refactor codec config count in SOF SoundWire machine driver
[ Upstream commit 0b7d55d3a91200f2b1ed710f525a944b0a7d6369 ]
num_devs was used both as the endpoint count and as the output for
asoc_sdw_parse_sdw_endpoints(), which overwrites it with the codec
configuration count. Introduce a separate num_confs variable to hold
the codec conf count so the two values remain distinct across
codec_conf allocation and card->num_configs assignment.
Fixes: 6d8348ddc56e ("ASoC: amd: acp: refactor SoundWire machine driver code")
Signed-off-by: Vijendar Mukunda <Vijendar.Mukunda@amd.com>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Link: https://patch.msgid.link/20260910161728.1452808-3-Vijendar.Mukunda@amd.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jiangshan Yi <yijiangshan@kylinos.cn>
Date: Mon Sep 14 18:47:12 2026 +0800
ASoC: codecs: rt712-sdca-dmic: fix uninitialized stream_config->type
commit 03a5699a0a04309c597683967aaaf25d1e555ea2 upstream.
stream_config is not initialized before being passed to
sdw_stream_add_slave(). The type field may contain garbage and is
later copied to stream->type by sdw_config_stream().
Zero-initialize stream_config so type defaults to SDW_STREAM_PCM.
While at it, use snd_sdw_params_to_config() helper instead of
open-coding the same logic.
Fixes: 63a511284c9e ("ASoC: rt712-sdca: Add RT712 SDCA driver for Mic topology")
Cc: stable@vger.kernel.org
Signed-off-by: Jiangshan Yi <yijiangshan@kylinos.cn>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.dev>
Link: https://patch.msgid.link/20260914104712.379574-1-yijiangshan@kylinos.cn
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Richard Fitzgerald <rf@opensource.cirrus.com>
Date: Mon Sep 14 13:26:11 2026 +0100
ASoC: cs-amp-lib: Prevent NULL pointer if efi variable is zero length
[ Upstream commit 3482062c786ce4233f8ed3224d824184f53ec154 ]
In cs_amp_alloc_get_efi_variable() the first call to
cs_amp_get_efi_variable() might return EFI_SUCCESS if the variable
exists with zero length. Trap this and return -ENOENT to prevent
returning an unexpected NULL pointer.
The first cs_amp_get_efi_variable() call was assumed to return
EFI_BUFFER_TOO_SMALL if the variable existed, but if instead it
returned EFI_SUCCESS this would be converted to 0 by
cs_amp_convert_efi_status() and then be returned as a NULL pointer.
Fixes: 00fd40bc7acec ("ASoC: cs-amp-lib: Support Dell SSIDExV2 UEFI variable")
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://patch.msgid.link/20260914122611.2783563-1-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: HyeongJun An <sammiee5311@gmail.com>
Date: Tue Sep 15 18:25:15 2026 +0900
ASoC: hdmi-codec: Report a change when the channel status moves
[ Upstream commit c17ae8c26eac16ad244daef44044d714f68a2ddc ]
The put() callback of "IEC958 Playback Default" stores all 24 channel
status bytes and then returns 0. The core notifies userspace only on a
positive return, so a write that changes what the get() callback hands
back is never announced, and a mixer holding the control open keeps
showing the old value.
Compare the stored bytes and return 1 when they move, the way
snd_hda_spdif_default_put() does.
The same shape is in img-spdif-out and uniperif_player.
No board with this codec was to hand. The change is a comparison of
driver state with no hardware behaviour in it, and mixer-test counts the
missing notification as event_missing.
Fixes: 7a8e1d44211e ("ASoC: hdmi-codec: Add iec958 controls")
Signed-off-by: HyeongJun An <sammiee5311@gmail.com>
Assisted-by: Claude:claude-opus-5
Link: https://patch.msgid.link/20260915092515.2638542-1-sammiee5311@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Richard Fitzgerald <rf@opensource.cirrus.com>
Date: Thu Sep 10 12:44:56 2026 +0100
ASoC: Rename snd_soc_dai_link_ch_map.ch_mask to cpu_ch_mask
[ Upstream commit 4d855d747521505b54457c96bc73577bf74b2374 ]
Rename the ch_mask member of snd_soc_dai_link_ch_map to cpu_ch_mask,
as that is what it is used for.
The CPU and codec channel masks are not necessarily the same, and are
quite likely different. SoundWire and I2S/TDM both support assigning
different sample slots to each codec, so for example channel 0 on each
codec could map to different channels at the CPU. So it's quite normal
that the channel mask at the CPU end is different for each codec, but
the codec channel masks are the same for each codec.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://patch.msgid.link/20260910114500.1586637-2-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Stable-dep-of: 6b382bdfe26a ("ASoC: soc-pcm: Apply snd_soc_dai_link_ch_map.codec_ch_mask to codec params")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Date: Fri Jun 26 05:45:10 2026 +0000
ASoC: sdw_utils: tidyup .count_sidecar
[ Upstream commit c97f0bf5f705b16d150f2b0d5ce0ee24eee4f68a ]
count_sidecar() is not using *card. Tidyup it.
Current code makes old style / new style conversion difficult.
To make future conversions easier to understand, this patch clean up the
code a little. but no functional change.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://patch.msgid.link/87jyrlety1.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Stable-dep-of: 0b7d55d3a912 ("ASoC: amd: acp: refactor codec config count in SOF SoundWire machine driver")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Date: Fri Jun 26 05:45:15 2026 +0000
ASoC: sdw_utils: tidyup asoc_sdw_parse_sdw_endpoints()
[ Upstream commit a1332be2a07090cf422507ec812ce2b9ba0a558a ]
We can avoid to use *card. Tidyup it.
Current code makes old style / new style conversion difficult.
To make future conversions easier to understand, this patch clean up the
code a little. but no functional change.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Reviewed-by: Vijendar Mukunda <Vijendar.Mukunda@amd.com>
Link: https://patch.msgid.link/87ik75etxw.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Stable-dep-of: 0b7d55d3a912 ("ASoC: amd: acp: refactor codec config count in SOF SoundWire machine driver")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Richard Fitzgerald <rf@opensource.cirrus.com>
Date: Thu Sep 10 12:44:58 2026 +0100
ASoC: soc-pcm: Apply snd_soc_dai_link_ch_map.codec_ch_mask to codec params
[ Upstream commit 6b382bdfe26a2232091bf743e454e6794295783e ]
In __soc_pcm_hw_params() if there is a snd_soc_dai_link_ch_map with
non-zero codec_ch_mask, use that channel mask to restrict which channels
are enabled on the codec. But only if there isn't a TDM mask.
It is possible that a snd_soc_dai_link_ch_map could include the same codec
multiple times on different CPUs so the for_each_rtd_ch_maps() loop
accumulates the channel masks for all entries of that codec.
If a TDM mask was also set, it takes priority and is used instead of any
possible snd_soc_dai_link_ch_map entries. (They cannot be ANDed together
because the bit positions are indicating different things: TDM is a bit
for each TDM slot, codec_ch_mask is a bit for each codec channel.)
This fixes a problem of incorrect TX channels enabled on the codec when
multiple codecs are aggregated on a single capture link. For example:
- Two CPUs with six 4-channel codecs.
- The machine driver chooses to assign one channel from each codec to
one channel on the CPU
- But the codec hw_params() would be passed a channel count of 6, which
(a) is more channels than the codec has and (b) allows enabling channels
that should not be driving the audio bus.
Fixes: ac950278b087 ("ASoC: add N cpus to M codecs dai link support")
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Link: https://patch.msgid.link/20260910114500.1586637-4-rf@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Sasha Levin <sashal@kernel.org>
Date: Sun Sep 13 13:31:32 2026 -0400
ASoC: ux500: Parenthesize MSP_{RX,TX}_CLKPOL_BIT() arguments
[ Upstream commit 11fc0048a6930f4fca44fe3bd16a0023e78846a2 ]
arm allmodconfig fails to build with gcc:
In file included from sound/soc/ux500/ux500_msp_i2s.c:20:
sound/soc/ux500/ux500_msp_i2s.h:151:38: error: suggest parentheses
around arithmetic in operand of '^' [-Werror=parentheses]
sound/soc/ux500/ux500_msp_i2s.c:204:21: note: in expansion of macro
'MSP_TX_CLKPOL_BIT'
cc1: all warnings being treated as errors
The macros never parenthesized their argument:
#define MSP_TX_CLKPOL_BIT(n) ((n & TCKPOL_MASK) << TCKPOL_SHIFT)
That went unnoticed while every caller passed a plain variable, but
configure_protocol() now passes an XOR expression, which binds as
"a ^ (b & MASK)" rather than "(a ^ b) & MASK", and gcc rightly
complains.
No functional change: tx_clk_pol and rx_clk_pol only ever hold
MSP_FALLING_EDGE (0) or MSP_RISING_EDGE (1), and bclk_inverted is a
bool, so masking before or after the XOR gives the same 0/1 result.
Parenthesize the argument anyway - it fixes the build and stops the
macros from silently mis-evaluating a future composite argument.
Fixes: 9ccbacf5a012 ("ASoC: ux500: Validate MSP DAI configuration")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202609051547.G9SJp8UQ-lkp@intel.com/
Assisted-by: LLM
Signed-off-by: Sasha Levin <sashal@kernel.org>
Reviewed-by: Linus Walleij <linusw@kernel.org>
Link: https://patch.msgid.link/20260913173132.1172003-1-sashal@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Niklas Cassel <cassel@kernel.org>
Date: Fri Sep 4 15:43:11 2026 +0200
ata: libahci: clear PxCLBU and PxFBU for AHCI_HFLAG_32BIT_ONLY
commit 82e47533221d4746947b74d2e79a478c36c6433a upstream.
A user reported that commit 105c42566a55 ("ata: ahci: force 32-bit DMA for
JMicron JMB582/JMB585") made the JMicron JMB585 unusable on his board.
The failure is seen as soon as the ahci driver is probed, and booting with
iommu=off does not solve the problem.
Looking at the AHCI specification, PxCLBU and PxFBU are both read only '0'
for HBAs that do not support 64-bit addressing.
For HBAs that do support 64-bit addressing, the registers are read write,
with a reset value that is Implementation Specific.
When using the AHCI_HFLAG_32BIT_ONLY flag, the HBA does support 64-bit
addressing, and a 32-bit DMA mask is set by simply clearing HOST_CAP_64.
Thus, in this case, we need to explicitly clear the registers to 0.
Fixes: 105c42566a55 ("ata: ahci: force 32-bit DMA for JMicron JMB582/JMB585")
Fixes: c7a42156d99b ("ahci: disable 64bit dma on sb600")
Cc: stable@vger.kernel.org
Reported-by: Roland Waltersson <roland.waltersson@netinsight.net>
Closes: https://lore.kernel.org/linux-ide/IA0PR17MB668730A4ECCD65F7A1DC3EDC9EB62@IA0PR17MB6687.namprd17.prod.outlook.com/
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20260904134310.1465051-2-cassel@kernel.org
Signed-off-by: Niklas Cassel <cassel@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Wentao Liang <vulab@iscas.ac.cn>
Date: Tue Sep 15 06:59:33 2026 +0000
ata: libahci_platform: Fix device reference leak in ahci_platform_get_resources()
commit 0d1cb83337f13af082afb68b28d3fdfe29cde7fb upstream.
of_find_device_by_node() takes a reference on the port platform device,
which is only used to look up its port regulator and is never released,
neither on success nor on the error paths. Drop the reference with
put_device() once the regulator has been obtained, which covers both the
success and error paths.
Fixes: c7d7ddee7e24 ("ata: libahci: Allow using multiple regulators")
Cc: stable@vger.kernel.org
Signed-off-by: Wentao Liang <vulab@iscas.ac.cn>
Link: https://lore.kernel.org/r/20260915065933.1733061-1-vulab@iscas.ac.cn
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Signed-off-by: Niklas Cassel <cassel@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Sai Teja Aluvala <aluvala.sai.teja@intel.com>
Date: Fri Sep 11 17:22:22 2026 +0530
Bluetooth: btintel_pcie: fix off-by-one bounds check in RX submit
[ Upstream commit 2ea5a87a5a7ae58cb2662b8a7d06f209383e1765 ]
btintel_pcie_submit_rx() used frbd_index > rxq->count to guard the
FRBD array access, allowing frbd_index == rxq->count to pass through
and index one element past the end of the array. Change the check to
>= rxq->count so every out-of-range index is rejected.
This issue was reported by Claude Mythos.
Fixes: c2b636b3f788 (Bluetooth: btintel_pcie: Add support for PCIe transport)
Signed-off-by: Sai Teja Aluvala <aluvala.sai.teja@intel.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Chandrashekar Devegowda <chandrashekar.devegowda@intel.com>
Date: Tue Sep 8 15:26:58 2026 +0530
Bluetooth: btintel_pcie: validate TX skb length in send_sync
[ Upstream commit 4b837ebd0ea21ae5cc26f02dc042edc6fe7b46b9 ]
btintel_pcie_prepare_tx() copies skb->len bytes into a fixed
BTINTEL_PCIE_BUFFER_SIZE (4096) DMA slot via an unchecked memcpy.
Oversized packets are currently rejected only in
btintel_pcie_send_frame(); any future caller of
btintel_pcie_send_sync() would silently overflow the DMA buffer.
Add the bounds check in btintel_pcie_send_sync() itself, right
before skb_push() and the DMA copy.
Assisted-by: Copilot:claude-sonnet-5 code-review code-generation
Fixes: 6e65a09f9275 ("Bluetooth: btintel_pcie: Add *setup* function to download firmware")
Signed-off-by: Chandrashekar Devegowda <chandrashekar.devegowda@intel.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Chris Lu <chris.lu@mediatek.com>
Date: Mon Sep 14 14:56:53 2026 +0800
Bluetooth: btmtk: fix wrong status for short WMT FUNC_CTRL events
[ Upstream commit 78b6abd6c7a7591aacdae657f813214dae4fcd3b ]
A too-short BTMTK_WMT_FUNC_CTRL event (WMT header only, no trailing
2-byte status word) is always treated as BTMTK_WMT_ON_UNDONE. This
short form is how firmware acks a plain enable/disable request, and
the actual result is carried in the header's own flag byte (0 =
success), not a separate status word. Decode it from there instead of
assuming failure.
Verified setup on MT7920, MT7921, MT7922 and MT7925: no regression.
Fixes: e3ac0d9f1a20 ("Bluetooth: btmtk: accept too short WMT FUNC_CTRL events")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Chris Lu <chris.lu@mediatek.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Chris Lu <chris.lu@mediatek.com>
Date: Mon Sep 14 14:56:54 2026 +0800
Bluetooth: btmtksdio, btmtkuart: validate WMT event length before struct access
[ Upstream commit 8879e3e0a84a86954c855caceead4867e74a9a27 ]
btmtksdio.c and btmtkuart.c cast a received WMT event straight to
struct btmtk_hci_wmt_evt and read its op/flag fields without checking
the event is long enough to contain them, unlike btmtk.c. The
FUNC_CTRL case then further casts to struct btmtk_hci_wmt_evt_funcc
and reads its 2-byte status field, again without a length check.
Firmware that sends a short or malformed WMT event makes both drivers
read past the end of the received SKB.
Mirror btmtk.c: validate the base WMT header with skb_pull_data()
before touching any of its fields, and when a FUNC_CTRL event turns
out to be the short, header-only form (a plain enable/disable ack
with no status word), decode the result from the header's own flag
byte instead (0 = success, otherwise failure).
Verified setup on MT7920, MT7921, MT7922 and MT7925: no regression.
Fixes: 9aebfd4a2200 ("Bluetooth: mediatek: add support for MediaTek MT7663S and MT7668S SDIO devices")
Fixes: e0b67035a90b ("Bluetooth: mediatek: update the common setup between MT7622 and other devices")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Chris Lu <chris.lu@mediatek.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Tzung-Bi Shih <tzungbi@kernel.org>
Date: Mon Sep 14 09:47:29 2026 +0000
Bluetooth: btmtksdio: Fix PM runtime reference leak in shutdown
[ Upstream commit 7b60ee5f46f2ee329de661f7c68b6818d8136220 ]
In btmtksdio_shutdown(), pm_runtime_get_sync() is called at the
beginning of the function. However, if sending the WMT function
control command fails later, the driver returns early.
It bypasses the corresponding pm_runtime_put_noidle() and
pm_runtime_disable() calls, leaking the PM usage counter and leaving PM
runtime enabled indefinitely.
Fall through to execute the PM runtime cleanup block even if WMT errors.
Fixes: 7f3c563c575e ("Bluetooth: btmtksdio: Add runtime PM support to SDIO based Bluetooth")
Signed-off-by: Tzung-Bi Shih <tzungbi@kernel.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Nicolas Thibert <nithibert@gmail.com>
Date: Tue Sep 8 10:01:08 2026 +0200
Bluetooth: btusb: fix NXP IW610 composite device handling
commit 2b50adefed9808a56d84d1de803cad882cc787fa upstream.
The NXP IW610 module exposes itself as a composite USB device
(0471:0215) with three interfaces: two real Bluetooth HCI interfaces
(class 0xe0) and one vendor-specific WiFi interface (class 0xff) used
by mwifiex-nxp.
The composite device's whole USB descriptor reports class 0xe0/01/01
(Bluetooth), so btusb_table's generic USB_DEVICE_INFO(0xe0, 0x01, 0x01)
entry matches every interface, not just the two real HCI ones -- btusb
ends up binding the WiFi interface too, and mwifiex-nxp never gets it.
Fix:
1. In btusb_table (the table the USB core actually matches against),
explicitly ignore the WiFi interface via BTUSB_IGNORE, ahead of the
generic entry.
2. In quirks_table, scope the existing BTUSB_MARVELL entry to the BT
interface class instead of matching the whole device by VID/PID
(harmless either way since quirks_table isn't consulted for initial
binding, but keep it correct).
Not upstream anywhere: checked NXP's own i.MX kernel fork
(nxp-imx/linux-imx), no IW610 references in btusb.c on any branch --
their reference designs wire this chip differently (WiFi over SDIO
per their release notes), so they never hit this.
Signed-off-by: Nicolas Thibert <nithibert@gmail.com>
Cc: stable@vger.kernel.org
Assisted-by: LLM (Claude Sonnet 5, Anthropic)
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Weiming Shi <bestswngs@gmail.com>
Date: Sun Sep 6 23:43:32 2026 +0800
Bluetooth: coredump: Quiesce dump work on unregister
[ Upstream commit d236517c264e41dc09833c708ef23bccb7a91219 ]
hci_devcd_handle_pkt_init() arms dump_timeout and coredump producers
queue dump_rx without holding an hdev reference. Unregister leaves both
works live, so disconnecting during an active dump lets them access hdev
after hci_release_dev() frees it.
Shut down coredump processing during unregister. Close the producer gate
under dump_q.lock before disabling both works, then free the active buffer
and queued packets under hci_dev_lock. Serializing the gate with enqueue
prevents controller-specific workers from adding packets after the final
purge.
Fixes: 9695ef876fd1 ("Bluetooth: Add support for hci devcoredump")
Reported-by: syzbot+b170dbf55520ebf5969a@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=b170dbf55520ebf5969a
Reported-by: Aby Sam Ross <abysamross@gmail.com>
Link: https://lore.kernel.org/r/20260322210849.68743-1-abysamross@gmail.com
Suggested-by: Aby Sam Ross <abysamross@gmail.com>
Reported-by: Tristan Madani <tristan@talencesecurity.com>
Link: https://lore.kernel.org/r/20260814231248.3096377-1-tristmd@gmail.com
Reported-by: Xiang Mei <xmei5@asu.edu>
Assisted-by: OpenAI Codex:gpt-5
Signed-off-by: Weiming Shi <bestswngs@gmail.com>
Reported-by: Xiang Mei <xmei5@asu.edu>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Aamir Ahmed <elb12345@hotmail.co.uk>
Date: Mon Sep 7 00:37:43 2026 +0100
Bluetooth: eir: validate service data length before reading UUID
commit e8241766794cf551d787fa3a77c0d54bbea6f6aa upstream.
eir_get_service_data() reads a 16-bit UUID from the service data using
get_unaligned_le16() without first checking that the data is long enough
to hold a UUID16 (2 bytes). If a malformed EIR entry has a service data
field with only 1 byte of payload (field_len=2), eir_get_data() returns
dlen=1. The subsequent get_unaligned_le16() then reads 1 byte past the
field boundary.
Additionally, if the corrupted UUID happens to match, the length
calculation "dlen - 2" underflows to SIZE_MAX since dlen is size_t.
Current callers either pass NULL for the length parameter or bounds-check
the returned length, but future callers may not.
Add a check that dlen >= sizeof(u16) and skip fields that are too short
to contain a valid UUID16.
Fixes: 8f9ae5b3ae80 ("Bluetooth: eir: Add helpers for managing service data")
Cc: stable@vger.kernel.org
Signed-off-by: Aamir Ahmed <elb12345@hotmail.co.uk>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Laxman Acharya Padhya <acharyalaxman8848@gmail.com>
Date: Mon Aug 24 21:42:36 2026 +0545
Bluetooth: hci_codec: validate vendor codec count length
commit d0795cfd6f655f4de84868a4f4bb41a03f037b3d upstream.
The Read Local Supported Codecs parsers consume the variable-sized
standard codec array before parsing the vendor codec count. Although the
initial reply-size check includes a vendor count byte in the fixed layout,
it does not guarantee that the byte remains after the standard codec array.
If a controller reply ends immediately after that array, calculating the
vendor codec array size reads vnd_codecs->num beyond the skb data. Use
skb_pull_data() to validate and consume each codec header before using its
count in both command variants.
Fixes: 8961987f3f5f ("Bluetooth: Enumerate local supported codec and cache details")
Fixes: 9ae664028a9e ("Bluetooth: Add support for Read Local Supported Codecs V2")
Cc: stable@vger.kernel.org
Suggested-by: Luiz Augusto von Dentz <luiz.dentz@gmail.com>
Signed-off-by: Laxman Acharya Padhya <acharyalaxman8848@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: ThangNN99 <ngocthang2710.1999@gmail.com>
Date: Sun Sep 6 22:21:27 2026 +0700
Bluetooth: hci_core: Fix queuing tx_work after workqueue is drained
[ Upstream commit 6610c6fe4b8936c232048e6049bf77c70a6f759c ]
hci_send_acl(), hci_send_sco() and hci_send_iso() queue hdev->tx_work
unconditionally. They can run from the L2CAP/SCO/ISO socket send path
while hci_dev_close_sync() is draining hdev->workqueue (HCIDEVDOWN
racing with a socket write). Since that queue_work() is not chained
work from the tx_work worker itself, __queue_work() sees the queue
marked __WQ_DRAINING, warns "cannot queue %ps on wq %s", and drops
the work:
WARNING: CPU: 1 PID: 5985 at kernel/workqueue.c:2352 __queue_work
Call Trace:
queue_work_on
l2cap_chan_send
l2cap_sock_sendmsg
...
hci_dev_close_sync() already sets HCI_CMD_DRAIN_WORKQUEUE before
draining, but only hci_cmd_work() and handle_cmd_cnt_and_timer()
check it before queuing. Route the tx_work producers through the
same guard via a shared hci_sched_tx() helper.
Fixes: 525daaea459f ("Bluetooth: hci_sync: Set HCI_CMD_DRAIN_WORKQUEUE during device close")
Reported-by: syzbot+b6919040d9958e2fc1ae@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=b6919040d9958e2fc1ae
Signed-off-by: ThangNN99 <ngocthang2710.1999@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ibrahim Abdelkader <iabdelka@qti.qualcomm.com>
Date: Wed Aug 19 14:54:25 2026 +0200
Bluetooth: hci_qca: Do not write to the serial port after it is closed
[ Upstream commit 4e93c65f87825e1e012bce56615320aeb123815d ]
hci_uart_close() closes the serdev port if HCI_QUIRK_NON_PERSISTENT_SETUP
is set (for example, for the WCN399x family). A failed hci_dev_open_sync()
following a successful qca_setup() calls hdev->close() but not
hdev->shutdown(), so the port is closed while power->vregs_on is left true.
qca_serdev_remove() then passes its power->vregs_on test and calls
qca_power_off(), which writes to the closed port unconditionally.
Seen on a WCN3988 by unbinding the driver after a controller failure. The
trace below is from a 7.0.0 based kernel, where qca_power_off() was still
named qca_power_shutdown():
Unable to handle kernel NULL pointer dereference at virtual address
0000000000000038
Call trace:
tty_set_termios+0x50/0x238 (P)
ttyport_set_baudrate+0x84/0xc0
serdev_device_set_baudrate+0x24/0x40
qca_power_shutdown+0x158/0x1fc [hci_uart]
qca_serdev_remove+0x54/0x68 [hci_uart]
serdev_drv_remove+0x1c/0x2c
device_remove+0x4c/0x80
device_release_driver_internal+0x1cc/0x224
device_driver_detach+0x18/0x24
unbind_store+0xb4/0xc0
Check HCI_UART_PROTO_READY, which hci_uart_close() clears in the same place
it closes the port, before writing to it. The regulator disable is left
unconditional so the controller is still powered down.
The dangling serport->tty that turns this into a use-after-free is
addressed in a separate patch.
Fixes: fa9ad876b8e0 ("Bluetooth: hci_qca: Add support for Qualcomm Bluetooth chip wcn3990")
Signed-off-by: Ibrahim Abdelkader <iabdelka@qti.qualcomm.com>
Reviewed-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
Signed-off-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Chengfeng Ye <nicoyip.dev@gmail.com>
Date: Sat Aug 22 01:43:50 2026 +0800
Bluetooth: hci_sync: Serialize local codec list cleanup
commit 9a10987a2f160a44a638c9a35994ca6e3089696e upstream.
hci_dev_close_sync() clears hdev->local_codecs after releasing hdev->lock.
Codec list additions and both traversals in sco_sock_getsockopt() use that
lock, but the close path does not. A close and BT_CODEC query can therefore
interleave as follows:
hci_dev_close_sync() sco_sock_getsockopt()
hci_dev_lock()
fetch codec entry
hci_codec_list_clear()
kfree(entry)
read entry->id
The reader then accesses an entry which the close path has freed. KASAN
reported:
BUG: KASAN: slab-use-after-free in sco_sock_getsockopt+0xfa0/0xfe0
Read of size 1 at addr ffff8881001c3450
Call Trace:
sco_sock_getsockopt+0xfa0/0xfe0
do_sock_getsockopt+0x537/0x7b0
__sys_getsockopt+0xf2/0x170
Allocated by task 92:
hci_codec_list_add.isra.0+0x2c/0x440
hci_read_codec_capabilities+0x224/0x590
hci_read_supported_codecs+0x2c2/0x640
Freed by task 92:
kfree+0x131/0x3c0
hci_codec_list_clear+0xd8/0x160
hci_dev_close_sync+0x92a/0xfa0
Take hdev->lock around the clear operation at its existing point in the
close path. This makes the clear wait for active readers and prevents a new
traversal until the list is empty without changing teardown ordering.
Fixes: b938790e7054 ("Bluetooth: hci_codec: Fix leaking content of local_codecs")
Cc: stable@vger.kernel.org
Signed-off-by: Chengfeng Ye <nicoyip.dev@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Date: Thu Sep 10 14:06:27 2026 -0400
Bluetooth: ISO: Fix parent socket leak in iso_conn_ready()
[ Upstream commit ca18ee413a7cb6f09885778039225e58bae0d607 ]
iso_get_sock() returns the parent socket with a reference held, which is
dropped by sock_put() once the child socket has been set up. The error
path taken when iso_sock_alloc() fails only calls release_sock() and
returns, leaking the reference and thus the parent socket itself.
Drop the reference on that path as well.
Fixes: fa224d0c094a ("Bluetooth: ISO: Reassociate a socket with an active BIS")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Date: Thu Sep 10 14:07:24 2026 -0400
Bluetooth: ISO: set BT_LISTEN before requesting a BIG sync
[ Upstream commit 296e7f3c5071cc02dc22e1566e759179fa1792ae ]
A BIS connection is matched to its parent socket by looking for a
socket in BT_LISTEN state with the same BIG handle:
iso_conn_ready()
if (test_bit(HCI_CONN_BIG_SYNC, &hcon->flags))
parent = iso_get_sock(hdev, &hcon->src, &hcon->dst,
BT_LISTEN, iso_match_big_hcon, hcon);
The socket was only moved to BT_LISTEN after iso_conn_big_sync()
returned, while the LE BIG Create Sync command has already been queued
by then. If the BIG sync is established before the state is updated,
which is easy to hit with an emulated controller as the command may
complete in a few hundred microseconds, no parent is found and the BIS
connections are never notified to the listening socket.
The user space is then left waiting for connections that never arrive,
e.g. bluetoothd never completes a MediaTransport1.Acquire of a
Broadcast Sink transport.
Move the socket to BT_LISTEN before requesting the BIG sync, so the
state is visible by the time the command is queued, and restore the
previous state if the request could not be started. Since the socket is
briefly visible as a listening socket, child sockets may have been
queued in the meantime, so drain the accept queue before restoring the
state: the cleanup paths of BT_CONNECT2/BT_CONNECTED don't do it and the
children would be left with a dangling parent pointer.
Fixes: fbdc4bc47268 ("Bluetooth: ISO: Use defer setup to separate PA sync and BIG sync")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Radek Podgorny <radek@podgorny.cz>
Date: Sun Sep 13 22:28:02 2026 +0200
Bluetooth: keep dst_type with dst when reusing an LE connection
commit 555cd2bd860e7c4bdc3f4e4405b05515b0d9bc87 upstream.
hci_connect_le() swaps the caller's identity address for the peer's
cached RPA when one is known, and stamps the matching
ADDR_LE_DEV_RANDOM on the local dst_type. On the conn-reuse path only
the address is copied into the connection:
if (conn) {
bacpy(&conn->dst, dst);
so conn->dst ends up holding an RPA while conn->dst_type still names the
identity it was resolved from, and hci_le_create_conn_sync() puts that
pair on air unchanged. An RPA declared as a public address is not
something any peer can answer.
Measured on a CYW43438 against a peer advertising an RPA the host holds
the IRK for, connecting to the identity address over a raw L2CAP socket.
The first attempt creates the connection, the second takes the reuse
path:
LE Create Connection 3C:78:95:78:37:C3 type public
LE Create Connection 5B:75:A2:26:D6:18 type public
LE Connection Complete: Unknown Connection Identifier (0x02)
The second address is the peer's RPA. btmon annotates it with an OUI
lookup rather than "(Resolvable)" precisely because the command declares
it public; the same bit pattern annotates as resolvable once the type is
right.
The mistyped pair is also why nothing downstream repairs it.
hci_bdaddr_is_rpa() tests the type before the address, so an RPA carrying
a public type is not recognised as one, and hci_find_irk_by_addr() then
searches for an identity address that does not match it either.
Copy the type along with the address.
The assignment used to be unconditional just below this block and covered
both paths; it moved into hci_conn_add_unset(), which the reuse path does
not go through.
Cc: stable@vger.kernel.org
Fixes: 14b06c3a88f7 ("Bluetooth: HCI: Always use the identity address when initializing a connection")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Radek Podgorny <radek@podgorny.cz>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Radek Podgorny <radek@podgorny.cz>
Date: Wed Sep 9 00:29:37 2026 +0200
Bluetooth: put the peer's on-air address on air when we cannot resolve
[ Upstream commit 4914c499896121ae8b9d5b90f0abc5c8287ff396 ]
An identity address only reaches a peer that is advertising an RPA if the
controller resolves it on our behalf. Where it cannot, the host has to put
the peer's on-air address on air itself.
hci_connect_le() still swaps the caller's identity address for the peer's
cached RPA before creating the connection, but __hci_conn_add() resolves
the RPA back to the identity address when it stores it, so the identity is
what goes out. Storing the identity is right when the controller
translates it on the way to the radio; without LL Privacy, or with this
peer absent from the resolving list, nothing does.
A peer advertising an RPA cannot answer its identity address, so the
attempt burns a full create-connection timeout. That is not merely a slow
connect: a controller without extended scanning cannot scan while it is
initiating, so every dead attempt also takes the scanner off the air for
the whole timeout.
Measured on a CYW43438, which reports neither LL Privacy nor extended
advertising (LE features 3f 00 00 08 00 00 00 00), against a peer
advertising a resolvable private address the host holds the IRK for, with
the connection requested on the peer's identity address:
before: LE Create Connection to the identity address, public type
1.61s -> 22.07s, then LE Create Connection Cancel
LE Connection Complete: Unknown Connection Identifier (0x02)
after: LE Create Connection to the peer's RPA, random type
LE Connection Complete: Success
Advertising reports reaching the host per second, same window, same five
unrelated devices on the adapter:
before 1s:2 [nothing from 2s through 21s] 22s:5 23s:3
after 0s:11 1s:5 2s:2 3s:5 4s:3 5s:4 ... 21s:2 22s:1 23s:2
One dead connect costs twenty seconds of scanning for every device on the
adapter, not just the one being dialled.
Keep the RPA in conn->dst unless the controller will translate the
identity address: address resolution enabled and the peer's identity
actually programmed into the resolving list. Testing ll_privacy_capable()
alone would not be enough: it reports the feature bit, not whether
resolution is switched on and not whether this peer is in the list.
Resolution is cleared with the other volatile flags on power-off and
switched off again while suspend pauses scanning, and a peer's IRK is only
programmed along the accept list path, so a direct-connect target, a peer
without HCI_CONN_FLAG_ADDRESS_RESOLUTION, and one that did not fit in a
full list are all absent from it.
With the peer programmed, the identity address stays in conn->dst and the
controller translates it: measured on an Intel controller, the host dials
the identity and LE Enhanced Connection Complete reports Resolved Public
with the peer's RPA in the separate peer resolvable private address field.
With the peer absent from the list the same setup dials the RPA itself.
Everything downstream already copes with an RPA in conn->dst: it is what
every outgoing LE connection stored before 14b06c3a88f7, the connection
complete event names the address that was dialled, and
le_conn_complete_evt() resolves it back to the identity once the link is
up. ISO links keep the unconditional conversion: they are created from an
existing ACL or a periodic sync and never dial this address themselves.
Keeping the RPA is only right while the peer is still using it, which is
why the preceding patch drops the cached RPA as soon as the peer is seen
advertising its identity address. Without that, a peer that turns privacy
off would be dialled on the address it abandoned rather than the one it
is answering on.
Fixes: 14b06c3a88f7 ("Bluetooth: HCI: Always use the identity address when initializing a connection")
Assisted-by: Claude:claude-opus-5
Assisted-by: Claude:claude-fable-5
Signed-off-by: Radek Podgorny <radek@podgorny.cz>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Juan Perdomo <jcperdomo100@gmail.com>
Date: Sat Sep 12 23:09:45 2026 -0400
Bluetooth: RFCOMM: avoid socket lock inversion in listener cleanup
[ Upstream commit 801fb950cae7048eb7d83b18857d1ca37b8cd5a4 ]
rfcomm_sock_cleanup_listen() closes unaccepted child sockets through
rfcomm_sock_close(), which takes the child socket lock before
rfcomm_dlc_close() acquires rfcomm_mutex. The RFCOMM worker takes these
locks in reverse order while handling connections and DLC state changes,
so lockdep reports a possible deadlock.
Close dequeued children without taking their socket lock. The accept queue
owns a reference to each child, and bt_accept_dequeue() locks the child
while unlinking it and clearing its parent pointer.
Dropping the child lock makes it important to prevent a concurrent
rfcomm_connect_ind() from enqueueing a new child after cleanup observes an
empty queue. Set a listening socket to BT_CLOSED while its lock is still
held, before dropping the lock and draining the queue. The state check in
rfcomm_connect_ind() then rejects new children once cleanup starts.
Reported-by: syzbot+0cece8fa7d83523f47a3@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=0cece8fa7d83523f47a3
Fixes: b7ce436a5d79 ("Bluetooth: switch to lock_sock in RFCOMM")
Signed-off-by: Juan Perdomo <jcperdomo100@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Filipe Manana <fdmanana@suse.com>
Date: Wed Sep 16 15:43:41 2026 +0100
btrfs: abort transaction on failure to update inode for hole punching and reflinking
[ Upstream commit 97fcd34aa9fd73cefe3120ac9a82ca9d7763922f ]
If we fail to update the inode we error out without aborting the
transaction, which can result in a persistent inconsistency if after
the failure the transaction is committed, as we have dropped file
extent items from a range and either punched a hole or insert a new file
extent item for that range (for reflinks).
So add the missing transaction abort.
Fixes: 2aaa66558172 ("Btrfs: add hole punching")
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Filipe Manana <fdmanana@suse.com>
Date: Wed Sep 16 16:49:37 2026 +0100
btrfs: check if there is space for chunk item when validating sys chunk array
[ Upstream commit aeab4c62875748ecfd390a47ac1d91ea7c9a6abb ]
We checked if have enough remaining space for a key before dereferencing a
key, but we then dereference a chunk item, to get the number of stripes,
without checking if there is space for the item. So add a check to see if
there is enough space for a chunk item before dereferencing the item to
extract the stripe count.
Fixes: 2a9bb78cfd36 ("btrfs: validate system chunk array at btrfs_validate_super()")
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Guanghui Yang <3497809730@qq.com>
Date: Wed Sep 16 05:16:38 2026 +0000
btrfs: clear free space tree creation state on rebuild failure
commit 3565893cc72cdf6b795cf6a33e7ff9605322334d upstream.
btrfs_rebuild_free_space_tree() sets BTRFS_FS_CREATING_FREE_SPACE_TREE
before rebuilding the free space tree. Several error paths return
without clearing this flag.
The transaction restart failure path can leave the flag set on a live
filesystem, causing delayed reference processing to be skipped. Clear it
on all free space tree rebuild failure paths. Keep
BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED set, since a failed rebuild leaves
the free space tree untrusted. Callers must fall back to extent-tree
caching.
Fixes: 882af9f13e83 ("btrfs: handle free space tree rebuild in multiple transactions")
CC: stable@vger.kernel.org # 6.14+
Assisted-by: LLM
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Guanghui Yang <3497809730@qq.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Anand Jain <asj@kernel.org>
Date: Sun Sep 13 02:06:28 2026 +0800
btrfs: derive f_fsid with dev_t only when temp_fsid is active
commit 72de4807ba84da485dda1a91572d66da9149e95a upstream.
Commit c2a74ed0494c ("btrfs: derive f_fsid from on-disk fsid and dev_t")
mixed dev_t into f_fsid for all single-device setups to avoid f_fsid
collisions with cloned filesystems.
However, doing this unconditionally breaks backward compatibility.
statfs(2) f_fsid changes after a kernel upgrade, and also can shift
across reboots or dev re-attaches as dev_t values change.
Fix this by only mixing dev_t when temp_fsid is active. This means for
non-temp_fsid setups or the original mount, we use the old method of
deriving fsid based on the UUID.
So in the case of a cloned Btrfs filesystem, we won't be able to
maintain the same fsid across mount recycle if the mount order changes.
Reported-by: Dave Hansen <dave.hansen@intel.com>
Link: https://lore.kernel.org/linux-btrfs/be0c08f5-2f31-40f5-8a3b-f2f58b3e00ff@intel.com
Fixes: c2a74ed0494c ("btrfs: derive f_fsid from on-disk fsid and dev_t")
CC: stable@vger.kernel.org # 7.2
Signed-off-by: Anand Jain <asj@kernel.org>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Filipe Manana <fdmanana@suse.com>
Date: Mon Sep 14 18:11:30 2026 +0100
btrfs: fix creation of compressed inline extents that don't save space
commit cc337324cc6be1ce0ae7e5a068dcb7e99ae1b59c upstream.
If the compressed data of an inline extent is larger than or equals to the
size of the uncompressed data, we are still allowing the creation of the
compressed inline extent, which does not result in any benefits, quite the
contrary as we waste metadata space and have to decompress when reading.
This is a recent regression introduced in commit 3eaf5f082c4c ("btrfs:
extract inlined creation into a dedicated delalloc helper").
It happens because we are passing the block size to btrfs_compress_bio(),
so we don't get -E2BIG from the compression code anymore, but we can not
pass i_size either, because if i_size is smaller than sector size, we
end up never creating lzo compressed inline extent for such small i_size
values. So refuse the compressed result at run_delalloc_inline() if
its size is not smaller than the uncompressed size (i_size).
Reported-by: Hanabishi <i.r.e.c.c.a.k.u.n+kernel.org@gmail.com>
Link: https://lore.kernel.org/linux-btrfs/c97652a5-ac6b-4de6-aa23-3cdebc01d00b@gmail.com/
Fixes: 3eaf5f082c4c ("btrfs: extract inlined creation into a dedicated delalloc helper")
CC: stable@vger.kernel.org # 7.1+
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Daniel Linjama <daniel@dev.linjama.com>
Date: Wed Sep 16 09:15:56 2026 +0300
btrfs: handle lack of space when cleaning up verity items
[ Upstream commit 76bf149cd0298544631e756670b89c399c7acbca ]
When enable_verity() hits the qgroup limit, rollback_verity() needs its
own metadata reservation. When the qgroup limit or lack of space refuses
the rollback, the whole filesystem is forced read-only even though the
qgroup limit was for one subvolume only. Also orphan cleanup at the next
mount fails the same way, so the leftover items are never removed: with
-EDQUOT the subvolume stays unreachable, and with -ENOSPC on a full
filesystem the next read-write mount fails.
Start transactions with btrfs_start_transaction_fallback_global_rsv() in
btrfs_orphan_cleanup(), drop_verity_items() and rollback_verity(). Those
calls only delete items and free the space in the end, so they may use
the global reserve and skip the qgroup limit, which avoids -ENOSPC and
-EDQUOT.
Fixes: 146054090b08 ("btrfs: initial fsverity support")
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Daniel Linjama <daniel@dev.linjama.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Hongling Zeng <zenghongling@kylinos.cn>
Date: Mon Aug 31 13:38:01 2026 +0800
btrfs: take commit root semaphore when iterating in mark_block_group_to_copy()
commit 0594e3423f4ba3137c734371169491f9a98e9af4 upstream.
mark_block_group_to_copy() iterates over the commit root with
skip_locking=true. A concurrent transaction commit can swap and free
the commit root during iteration, causing use-after-free when
accessing extent buffers.
Fix it by using path->need_commit_sem to protect the commit root search.
Fixes: 78ce9fc269af ("btrfs: zoned: mark block groups to copy for device-replace")
CC: stable@vger.kernel.org
Assisted-by: Codex:gpt-5.5
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Hongling Zeng <zenghongling@kylinos.cn>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Filipe Manana <fdmanana@suse.com>
Date: Thu Sep 10 17:37:48 2026 +0100
btrfs: tree-checker: print dev extent offset in error message
[ Upstream commit a1167d9420474ab9ed9efca99d86aeb6217c0265 ]
If a dev extent's offset is not sector size aligned, the error message is
printing the dev extent's objectid instead of the offset. This is a copy
paste error, as before this check we check the objectid field.
Fixes: 008e2512dc56 ("btrfs: tree-checker: add dev extent item checks")
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Michal Koutný <mkoutny@suse.com>
Date: Mon Sep 14 14:19:10 2026 +0200
cgroup: Avoid iteration of dying tasks with zero refcount
commit 057dac23d329d5c5ed62352f2659a39fd46c6d4a upstream.
The commit 260fbcb92bbea ("cgroup: Move dying_tasks cleanup from
cgroup_task_release() to cgroup_task_free()") extended the lifetime of
tasks on the dying_tasks list.
The iterators have provision to go through dying_tasks because of
dying threadgroup leaders or explicit CSS_TASK_ITER_WITH_DEAD, however,
it was expected that such tasks can obtain a new reference (that is
possible before cgroup_task_release()/put_task_struct_rcu_user()).
The tasks after cgroup_task_release() and before cgroup_task_free()
are subject to race when they may or may not have ->usage count > 0.
The race window is between css_task_iter_next() invocations
when css_set_lock is released and we may arrive at a new ->task_pos.
The iterator should not attempt to resurrect tasks whose ->usage count
dropped to zero. (When that happens, __put_task_struct_rcu_cb() is
already imminent and the returned task_struct would could be used
after free.)
As for the fix, we cannot simply check the signal->live count of a task
on the dying list because that won't distinguish regular zombies waiting
to be reaped from RCU remnant tasks that are going to be free'd.
Therefore add an extra check to rule out ->usage==0 tasks from any
iteration.
The repeat: loop in css_task_iter_advance() doesn't consider ->usage
count, so add a new loop to css_task_iter_next() to skip de-used tasks
on the dying_list.
Rough illustration of the possible race
R (reader of cgroup.procs) T (thread) L (group leader)
--------------------------------- -------------------------------- --------------------------------
L exits, signal->live > 0
cgroup_task_dead(L)
css_set_skip_task_iters() // skips only cset->tasks
list_add_tail(&L->cg_list, &cset->dying_tasks)
css_task_iter_next()
take css_set_lock
css_task_iter_advance()
leader && signal->live != 0
=> it->task_pos = &L->cg_list
release css_set_lock
T exits
--signal->live == 0
cgroup_task_dead(T) // css_set_lock
release_task(T)
cgroup_task_release(T)
release_task(L) // zap_leader
cgroup_task_release(L)
put_task_struct_rcu_user(L)
...RCU...
put_task_struct(L)
L->usage = 0
/* L still on dying_tasks */
...RCU...
__put_task_struct(L)
css_task_iter_next() // another iteration
take css_set_lock
it->task_pos = &L->cg_list
get_task_struct(L)
=> addition on 0
drop css_set_lock
cgroup_task_free(L)
css_set_skip_task_iters() // dying skip comes too late
free_task(L)
cgroup_procs_show()
task_pid_vnr(L)
Fixes: 260fbcb92bbea ("cgroup: Move dying_tasks cleanup from cgroup_task_release() to cgroup_task_free()")
Cc: stable@vger.kernel.org # v6.19+
Link: https://lists.debian.org/debian-kernel/2026/08/msg00220.html
Reported-by: Noah Elias Feldt <N.Feldt@mittwald.de>
Reported-by: Salvatore Bonaccorso <carnil@debian.org>
Tested-by: Salvatore Bonaccorso <carnil@debian.org>
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Wentao Liang <vulab@iscas.ac.cn>
Date: Thu Sep 17 16:34:39 2026 +0000
cifs: Fix server use-after-free in cifs_chan_skip_or_disable()
commit 717e0a25036b6c92cecace30913b2d874a4c22b8 upstream.
When a secondary channel is no longer supported by the server,
cifs_chan_skip_or_disable() drops the channel reference with
cifs_put_tcp_session() and then continues to use the server pointer by
calling cifs_signal_cifsd_for_reconnect() on it and reading its
primary_server pointer. cifs_put_tcp_session() can drop the last
reference of the channel and tear it down, so both the channel and the
primary server (whose reference is also dropped by
cifs_put_tcp_session()) can be freed before they are signaled for
reconnect.
Signal the channel and the primary server and capture the primary
server pointer before dropping the channel reference with
cifs_put_tcp_session().
Fixes: f591062bdbf4 ("cifs: handle servers that still advertise multichannel after disabling")
Cc: stable@vger.kernel.org
Signed-off-by: Wentao Liang <vulab@iscas.ac.cn>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Xixin Liu <liuxixin@kylinos.cn>
Date: Tue Jul 28 08:50:00 2026 +0800
clk: scpi: bound-check DVFS index in scpi_dvfs_recalc_rate
[ Upstream commit 70f4b78d560e592cbf3325b162424737d032fc1d ]
dvfs_get_idx() may return an out-of-range index if the SCP firmware is
buggy or returns a stale value. Only negative indexes were rejected, so a
large index walked past info->opps and could treat garbage as a clock rate
(KASAN OOB / wrong frequency to consumers). The missing upper bound dates
back to the original SCPI clock driver.
Treat indexes >= opp count as invalid and return 0, same as idx < 0.
Fixes: cd52c2a4b5c4 ("clk: add support for clocks provided by SCP(System Control Processor)")
Signed-off-by: Xixin Liu <liuxixin@kylinos.cn>
Link: https://patch.msgid.link/04f9ab766e07.v2.1785200642.git.liuxixin@kylinos.cn
Signed-off-by: Sudeep Holla <sudeep.holla@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Xixin Liu <liuxixin@kylinos.cn>
Date: Tue Jul 28 08:50:00 2026 +0800
clk: scpi: register scpi-cpufreq once and clear on failure
[ Upstream commit ab06cf8152dace327cd873188e4a036c4e0b5944 ]
scpi_clocks_probe() walks clock children and, for each DVFS provider, calls
platform_device_register_simple("scpi-cpufreq", -1, ...). Two related bugs:
Since all DVFS providers register the fixed scpi-cpufreq device using
PLATFORM_DEVID_NONE, a second registration fails with -EEXIST and
overwrites the pointer to the successfully registered device. The first
device can then no longer be unregistered.
Register the virtual device only once. If registration fails, reset the
pointer to NULL so a subsequent DVFS provider can retry and the global
pointer only represents a successfully registered device.
Fixes: 9490f01e2471 ("clk: scpi: add support for cpufreq virtual device")
Fixes: 67bcc2c5f1da ("clk: scpi: don't add cpufreq device if the scpi dvfs node is disabled")
Signed-off-by: Xixin Liu <liuxixin@kylinos.cn>
Link: https://patch.msgid.link/fd1b9199a9c3.v2.1785200642.git.liuxixin@kylinos.cn
(sudeep.holla: reworded the commit message to improve readability)
Signed-off-by: Sudeep Holla <sudeep.holla@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Christian König <christian.koenig@amd.com>
Date: Fri Sep 4 10:26:13 2026 +0200
dma-buf/dma-fence: fix checking signaling bit for timeline and driver name v3
commit 3ed11c671ff7ec58c8fd96410233c677df23f407 upstream.
The patch "dma-buf: dma-fence: Fix potential NULL pointer dereference"
changed the check to test for the ops pointer instead of the signaled
bit to avoid a potential NULL dereference when the ops pointer has been
cleared.
The problem is now that the ops pointer is cleared only when neither the
release nor the wait callback is implemented and this isn't true for a lot
of dma_fence implementations yet. So those implementations lost the RCU
protection after signaling of the returned string resulting in potential
use after free.
Add the signaling check additional to the ops pointer check so that we
have both the protection against NULL dereference as well as the RCU
protection after signaling for the returned string.
v2: improve comments to note RCU protection and explain why we check
both signaling state and ops pointer
v3: some comment improvements suggested by Philip
Signed-off-by: Christian König <christian.koenig@amd.com>
Fixes: 035219a760ed ("dma-buf: dma-fence: Fix potential NULL pointer dereference")
CC: stable@vger.kernel.org # 7.2+
Reported-by: Jonghyuk Kim(MalHyuk) <malhyuk97@gmail.com>
Tested-by: Jonghyuk Kim(MalHyuk) <malhyuk97@gmail.com>
Reviewed-by: Philipp Stanner <phasta@kernel.org>
Link: https://lore.kernel.org/r/20260914182740.1587-1-christian.koenig@amd.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: David Hu <xuehaohu@google.com>
Date: Tue Sep 1 17:08:48 2026 +0000
dma-buf: Fix silent overflow for phys vec to sgt
commit b344ca94e8cc85796f16ea25e2e5a8e0303fe813 upstream.
In case MMIO size is bigger than 4G and peer2peer DMA goes
through host bridge, we trigger a code path that assigns the
total linked IOVA (which is greater than 4G) to mapped_len.
Previously, `mapped_len` was declared as 32-bit `unsigned int`.
When accumulating `size_t` lengths, this leads to a silent wrap-around.
This truncation causes truncated lengths to be passed to functions
like `fill_sg_entry()`.
Fix this by changing `mapped_len` to `size_t` (64-bit). While
at it, fix similar potential overflow issues in `calc_sg_nents`
by using `check_add_overflow()` for `nents` and using
`unsigned int` for the loop iterator in `fill_sg_entry` to match.
Fixes: 3aa31a8bb11e ("dma-buf: provide phys_vec to scatter-gather mapping routine")
Cc: stable@vger.kernel.org
Cc: iommu@lists.linux.dev
Reviewed-by: Pranjal Shrivastava <praan@google.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: David Hu <xuehaohu@google.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Link: https://lore.kernel.org/r/20260901170849.4052816-2-dhu@x6u.co
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: David Hu <xuehaohu@google.com>
Date: Tue Sep 1 17:08:49 2026 +0000
dma-buf: Split sgl by largest page-aligned chunk
commit 06dd5e1ae8ce4e129791087c8c66594950f0ba03 upstream.
Currently, `fill_sg_entry()` splits the scatterlist using `UINT_MAX`.
This creates a non-page-aligned DMA length (`0xFFFFFFFF`) for the
first entry, resulting in non-page-aligned DMA addresses for all
subsequent entries.
While the underlying IOMMU mapping may be contiguous, hardware
DMA engines often require explicit address alignment (e.g., page,
cacheline, or storage sector boundaries). Passing unaligned
addresses and lengths can cause explicit failures in DMA descriptor
creation or silent data corruption if lower unaligned bits are
truncated.
In addition, a non-page-aligned sgl length will trigger an edge case
in `ib_umem_find_best_pgsz()`. In case of a discontinuity in later
buffers, we will have a `va` with lowest bit set to 1. That will lead
to `ib_umem_find_best_pgsz()` always return 0, and break the promise
to find best page size for the mapping on the NIC side.
Fix this by splitting the scatterlist by the largest possible page
aligned chunk within `UINT_MAX` (`ALIGN_DOWN(UINT_MAX, PAGE_SIZE)`).
This ensures all scatterlist DMA addresses and lengths remain page
aligned, while minimizing the total number of sgl entries.
Page-aligned entries allow the system to cleanly chunk payloads into
PCIe MaxPayloadSize (MPS) (e.g., 128 bytes, 256 bytes, 512 bytes).
As a result, this may help reduce TLP fragmentation in P2P transfers
and alleviate potential congestion within a logical PCIe switch
partition, especially when Relaxed Ordering is not possible due to
hardware constraints.
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Closes: https://lore.kernel.org/all/20260609165431.778061F00893@smtp.kernel.org/
Fixes: 3aa31a8bb11e ("dma-buf: provide phys_vec to scatter-gather mapping routine")
Cc: stable@vger.kernel.org
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David Hu <xuehaohu@google.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Link: https://lore.kernel.org/r/20260901170849.4052816-3-dhu@x6u.co
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Donggeun Yoo <donggeunyoo.kernel@gmail.com>
Date: Sat Sep 5 16:47:27 2026 +0900
dma-coherent: report a failed reserved memory assignment
[ Upstream commit 504981db4f69bdd28054fb98c96a3a67f7248dde ]
rmem_dma_device_init() drops the return value of
dma_assign_coherent_memory() and always reports success. That call fails
with -EBUSY when the device already has a coherent pool, and the file
allows only "*one* such region of memory" per device.
of_reserved_mem_device_init_by_idx() reads the zero as success. It logs
"assigned reserved memory node" for a region that was not assigned and
records the pairing, so of_reserved_mem_device_release() later runs
rmem_dma_device_release() for it. That clears dev->dma_mem without
looking at which region it was called for, dropping the pool the device
did get and leaving it on ordinary memory.
dma_declare_coherent_memory() checks the same call and releases the
memory on failure, and rmem_swiotlb_device_init() propagates its own
errors. Return the error here as well, so a device tree that assigns two
pools to one device fails the probe instead of half working.
Fixes: 7bfa5ab6fa1b ("drivers: dma-coherent: add initialization from device tree")
Signed-off-by: Donggeun Yoo <donggeunyoo.kernel@gmail.com>
Link: https://lore.kernel.org/r/20260905074727.108029-1-donggeunyoo.kernel@gmail.com
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Donggeun Yoo <donggeunyoo.kernel@gmail.com>
Date: Mon Sep 7 21:01:24 2026 +0900
dma-mapping: don't trace the DMA address when the allocation fails
[ Upstream commit 92c6a8d6470f7e7aa86c1f144818d8fa4fcbd5aa ]
dma_alloc_attrs() passes *dma_handle to trace_dma_alloc() without
checking whether the allocation succeeded. No backend writes it on
failure: dma_direct_alloc(), iommu_dma_alloc() and the dma_map_ops
instances assign it only on the path that returns a buffer. Callers
usually pass an uninitialized automatic variable, so a failed allocation
records whatever the stack held, next to the virt_addr=(null) that marks
the record as an error:
dma_alloc: dmatrace dir=BIDIRECTIONAL dma_addr=deadbeefdeadbeef
size=1099511627776 virt_addr=0000000000000000
The device coherent pool path reaches the same call: a non-zero return
from dma_alloc_from_dev_coherent() means the request was handled, not
that it succeeded, so cpu_addr is NULL and dma_handle is untouched once
the pool runs out.
For an allocation event a NULL virt_addr already means the request
failed, so the address field carries nothing. Report 0 for it in the
event class rather than at each call site, which covers dma_alloc_pages()
and dma_alloc_sgt_err() as well.
Fixes: 038eb433dc14 ("dma-mapping: add tracing for dma-mapping API calls")
Fixes: 68b6dbf1f441 ("dma-mapping: trace more error paths")
Suggested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Donggeun Yoo <donggeunyoo.kernel@gmail.com>
Link: https://lore.kernel.org/r/20260907120124.603373-1-donggeunyoo.kernel@gmail.com
Reviewed-by: Sean Anderson <sean.anderson@linux.dev>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Shivank Garg <shivankg@amd.com>
Date: Sat Aug 22 19:22:04 2026 +0000
dmaengine: Fix device kref underflow in dma_chan_put()
[ Upstream commit 44dab659064eb5c10adb0306510eebe848ed592d ]
dma_chan_get() takes chan->device->ref only on the slow path:
/* no kref on fast path */
if (chan->client_count) {
__module_get(owner);
chan->client_count++;
return 0;
}
if (!try_module_get(owner))
return -ENODEV;
if (!dma_device_get(chan->device)) { // calls kref_get_unless_zero()
dma_chan_put() drops the ref unconditionally, so every fast-path
get/put pair drops one extra device reference.
The bug fires when two conditions hold together: a non-private
provider has a persistent client holding chan->client_count > 0
and another client cycles dmaengine_get()/dmaengine_put().
When the kref hits zero, the subsequent dma_find_channel() returns
NULL even though the provider module is still loaded.
Fix this by dropping device->ref only on the last put, matching the
single slow-path get.
Fixes: 8ad342a86359 ("dmaengine: Add reference counting to dma_device struct")
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Shivank Garg <shivankg@amd.com>
Link: https://patch.msgid.link/20260822-dmaengine-kref-fix-v5-2-d4a4ee47d927@amd.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Shivank Garg <shivankg@amd.com>
Date: Sat Aug 22 19:22:05 2026 +0000
dmaengine: fix use-after-free in dma_chan_put() and dma_release_channel()
[ Upstream commit e873c74132f0c5f1452816cd9bb26208f0bba1e1 ]
When dma_device_put() drops the last reference on chan->device->ref,
dma_device_release() runs and may free the dma_device along with its
channels.
dma_chan_put() then still reads chan->device->owner via
dma_chan_to_owner() for the trailing module_put(). KASAN catches it:
slab-use-after-free in dma_chan_put+0x3e6/0x4c0
Read of size 8 by task insmod/6319
Freed by task 6319:
kfree+0x225/0x470
dma_chan_put+0x395/0x4c0
dmaengine_put+0xf8/0x160
Cache the module owner in dma_chan_put() before the put so the trailing
module_put() does not need chan->device.
Fixes: 8ad342a86359 ("dmaengine: Add reference counting to dma_device struct")
Suggested-by: Sashiko <sashiko-bot@kernel.org>
Link: https://sashiko.dev/#/patchset/20260518-dmaengine-kref-fix-v1-1-4d6125048fb7@amd.com
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Shivank Garg <shivankg@amd.com>
Link: https://patch.msgid.link/20260822-dmaengine-kref-fix-v5-3-d4a4ee47d927@amd.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Troy Mitchell <troy.mitchell@linux.spacemit.com>
Date: Mon Jul 27 01:26:15 2026 -0700
dmaengine: mmp_pdma: fix wrong extended DRCMR base for SpacemiT K3
[ Upstream commit d1fc569fcbc7be6de06b034cf954a95e30ca1fb2 ]
The extended DRCMR window on SpacemiT K3 starts at 0x1100. Commit
6587b8661a0b ("dmaengine: mmp_pdma: add SpacemiT K3 support") incorrectly
set it to 0x1000, causing DRCMR accesses for request IDs >= 64 to target
offsets 0x100 too low.
The 0x1100 base has been verified on K3 silicon using real SPI and QSPI
DMA transactions. The K3 DMA documentation [1] was updated on June 24,
2026, to reflect the corrected register addresses.
Drop the bogus DRCMR_EXT_BASE_K3 macro and reuse
DRCMR_EXT_BASE_DEFAULT for the K3 ops.
Fixes: 6587b8661a0b ("dmaengine: mmp_pdma: add SpacemiT K3 support")
Link: https://www.spacemit.com/community/document/info?nodepath=hardware/key_stone/k3/k3_docs/k3_usermanual/16_peripherals/dma.md&lang=en [1]
Signed-off-by: Troy Mitchell <troy.mitchell@linux.spacemit.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Link: https://patch.msgid.link/20260727-k3-pdma-fix-drcmr-base-v2-1-afba55cba1f3@linux.spacemit.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Baineng Shou <shoubaineng@gmail.com>
Date: Thu Sep 10 10:16:52 2026 +0800
dmaengine: mmp_pdma: fix wrong sg length in mmp_pdma_prep_slave_sg()
[ Upstream commit 075bc7b1d3dde5ed43fbaabbc1a69f09b7fc3a47 ]
In mmp_pdma_prep_slave_sg(), for_each_sg() iterates the scatterlist
putting each entry into 'sg', but the entry length is read from 'sgl'
(the list head) instead of 'sg' (the current entry):
for_each_sg(sgl, sg, sg_len, i) {
addr = sg_dma_address(sg);
avail = sg_dma_len(sgl); /* should be 'sg' */
Consequently 'avail' is always the length of the first entry. For
multi-sg lists this causes out-of-bounds reads when a later entry is
shorter than the first, and silent data loss when it is longer.
Single-sg or uniformly-sized lists happen to mask the issue.
Fixes: c8acd6aa6bed3 ("dmaengine: mmp-pdma support")
Signed-off-by: Baineng Shou <shoubaineng@gmail.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Link: https://patch.msgid.link/20260910021652.1296640-1-shoubaineng@gmail.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Sascha Hauer <s.hauer@pengutronix.de>
Date: Mon Aug 17 22:44:33 2026 +0200
dmaengine: pxa: fix double counting of the hw descriptors
[ Upstream commit f6504be006aa4bb4bd26285f410a885c17920d65 ]
pxad_alloc_desc() was converted from
kzalloc(struct_size(sw_desc, hw_desc, nb_hw_desc), GFP_NOWAIT)
to kzalloc_flex(), which sets the __counted_by() counter sw_desc->nb_desc
itself - but only where the compiler has __builtin_counted_by_ref(), so
from gcc 15.1 or clang 22.1 on. The loop below it still increments
nb_desc, which makes it come out doubled there and correct elsewhere.
nb_desc is what pxad_free_desc() iterates over and what
set_updater_desc() indexes from, so set it explicitly and drop the
increment. The error path has to lower it to the number of descriptors
allocated so far, otherwise pxad_free_desc() would free entries that were
never allocated.
Fixes: 69050f8d6d075 ("treewide: Replace kmalloc with kmalloc_obj for non-scalar types")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Link: https://lore.kernel.org/r/20260817-dmaengine-pxa-v1-1-850c215c1196@pengutronix.de
Link: https://patch.msgid.link/20260817-dmaengine-pxa-v2-1-f42ab0569a48@pengutronix.de
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ruoyu Wang <ruoyuw560@gmail.com>
Date: Thu Aug 13 23:31:49 2026 +0800
dmaengine: sprd: Fix runtime PM reference leak in probe
[ Upstream commit a7df136ec529ee49a789c5029bc37b98b0d4bedd ]
pm_runtime_get_sync() increments a device's usage counter even when it
fails. sprd_dma_probe() currently jumps directly to controller clock
cleanup on that error, bypassing both pm_runtime_put_noidle() and
pm_runtime_disable(). This can happen if the preceding unchecked
pm_runtime_set_active() fails and the following runtime-resume attempt
also returns an error.
Enter the existing runtime-PM unwind path instead. This drops the
reference without idling the partially initialized device, disables
runtime PM, and then releases the controller clocks. The success path
and propagated error code are unchanged.
This issue was found by a static analysis checker and confirmed by manual
source review.
Fixes: 9b3b8171f7f4 ("dmaengine: sprd: Add Spreadtrum DMA driver")
Signed-off-by: Ruoyu Wang <ruoyuw560@gmail.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Link: https://patch.msgid.link/20260813153149.3953497-1-ruoyuw560@gmail.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Christian Lugnberg <christian.lugnberg@soundtrack.io>
Date: Mon Aug 17 15:51:22 2026 +0200
dmaengine: sun6i: fix non-atomic read of DMA position registers
commit c90b6973daa37f4c283342dff881ae001dea4fe6 upstream.
sun6i_get_chan_size() reads DMA_CHAN_LLI_ADDR and DMA_CHAN_CUR_CNT in two
separate readl() calls with no synchronisation between them:
pos = readl(pchan->base + DMA_CHAN_LLI_ADDR);
bytes = readl(pchan->base + DMA_CHAN_CUR_CNT);
DMA_CHAN_LLI_ADDR holds the physical address of the *next* descriptor the
engine will load once the current one completes. DMA_CHAN_CUR_CNT holds the
remaining byte count for the *current* descriptor. If the DMA engine
advances to the next LLI entry between the two reads, pos becomes stale: it
still points to what was the next descriptor at the time of the first read,
but that descriptor is now the current one and CUR_CNT reflects its initial
(full) byte count. The subsequent virtual-chain walk starts one entry too
early and accumulates an extra full period's worth of bytes into the
residue estimate.
Fix this by re-reading DMA_CHAN_LLI_ADDR after DMA_CHAN_CUR_CNT and
retrying if the value changed. This double-read pattern guarantees that
both registers were sampled during the same descriptor interval. The cost
is at most one extra readl() pair per call in the racy case, which occurs
only at descriptor boundaries (~every 2 ms) and is negligible.
Fixes: a90e173f3faf ("dmaengine: sun6i: Add cyclic capability")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-sonnet-4-6
Signed-off-by: Christian Lugnberg <christian.lugnberg@soundtrack.io>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Link: https://patch.msgid.link/20260817135723.12807-2-christian.lugnberg@soundtrack.io
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Christian Lugnberg <christian.lugnberg@soundtrack.io>
Date: Mon Aug 17 15:51:23 2026 +0200
dmaengine: sun6i: fix undefined behaviour in sun6i_dma_tx_status
commit 9096bdc8d930147f7c39a493a859acbd3a8485d8 upstream.
sun6i_dma_tx_status() calls vchan_find_desc() to look up the virtual
descriptor for a given cookie, before checking whether the pointer
vd is NULL:
vd = vchan_find_desc(&vchan->vc, cookie);
txd = to_sun6i_desc(&vd->tx); /* vd may be NULL here */
if (vd) {
for (lli = txd->v_lli; ...)
vchan_find_desc() returns NULL when the descriptor has already been
completed or is in-flight on a physical channel and no longer present
in the virtual channel's descriptor list. When vd is NULL,
to_sun6i_desc() is called unconditionally on &vd->tx before the NULL
check, which is undefined behaviour. Move the call inside the if (vd)
guard to ensure it is only reached with a valid pointer.
vd = vchan_find_desc(&vchan->vc, cookie);
if (vd) {
struct sun6i_desc *txd = to_sun6i_desc(&vd->tx);
for (lli = txd->v_lli; ...)
Fixes: 555859308723 ("dmaengine: sun6i: Add driver for the Allwinner A31 DMA controller")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-sonnet-4-6
Signed-off-by: Christian Lugnberg <christian.lugnberg@soundtrack.io>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Link: https://patch.msgid.link/20260817135723.12807-3-christian.lugnberg@soundtrack.io
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Alexander Chesnokov <Alexander.Chesnokov@kaspersky.com>
Date: Wed Aug 12 08:34:26 2026 +0300
dmaengine: ti: k3-udma-glue: fix NULL dereference in k3_udma_glue_release_rx_chn()
commit 0294b6dd515256c03ea2dbf508ddd3826d788579 upstream.
If devm_kcalloc() for rx_chn->flows fails in a channel request function,
the error path calls k3_udma_glue_release_rx_chn(), which dereferences
the NULL rx_chn->flows pointer in k3_udma_glue_release_rx_flow().
Skip the flow release loop in k3_udma_glue_release_rx_chn() when
rx_chn->flows is not allocated.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Fixes: d70241913413 ("dmaengine: ti: k3-udma: Add glue layer for non DMAengine users")
Cc: stable@vger.kernel.org
Reported-by: Pavel Zhigulin <Pavel.Zhigulin@kaspersky.com>
Signed-off-by: Alexander Chesnokov <Alexander.Chesnokov@kaspersky.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Link: https://patch.msgid.link/20260812053426.3521589-1-Alexander.Chesnokov@kaspersky.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shivank Garg <shivankg@amd.com>
Date: Sat Aug 22 19:22:06 2026 +0000
dmaengine: wait for RCU readers before releasing dma_device
[ Upstream commit dc750422170a563c7a81f6e49d36bb02c62ae37f ]
dma_issue_pending_all() walks the dma_device_list with
list_for_each_entry_rcu() under rcu_read_lock(). dma_device_release()
unlinks the device with list_del_rcu() and then calls
device->device_release() (which in many drivers, such as plx_dma.c,
directly calls kfree()).
Because there is no grace period between unlinking the device and
freeing it, concurrent RCU readers in dma_issue_pending_all() can
access the device after it has been freed.
The lockless walk originally relied on clients holding a dmaengine
reference to pin the provider module, and therefore the device, for as
long as they might traverse the list. Commit 8ad342a86359 ("dmaengine:
Add reference counting to dma_device struct") decoupled the dma_device
lifetime from the module reference, so the device can now be released
while a reader is still walking the list.
Add synchronize_rcu() before the device is freed, so RCU readers are
guaranteed to have finished. Keep it unconditional: providers that do
not implement device_release() free the device themselves once
dma_async_device_unregister() returns. This call will delay for a grace
period with dma_list_mutex held, which is safe and only teardown path is
delayed.
Fixes: 2ba05622b8b1 ("dmaengine: provide a common 'issue_pending_all' implementation")
Suggested-by: Sashiko <sashiko-bot@kernel.org>
Link: https://sashiko.dev/#/patchset/20260526-dmaengine-kref-fix-v2-0-3df60afac01d@amd.com
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Shivank Garg <shivankg@amd.com>
Link: https://patch.msgid.link/20260822-dmaengine-kref-fix-v5-4-d4a4ee47d927@amd.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Alex Bereza <alex@bereza.email>
Date: Tue Aug 18 09:36:29 2026 +0200
dmaengine: xilinx_dma: Fix hardware buffer descriptor chain after cyclic DMA
[ Upstream commit 7ed1e3070c9b4bbd67d5519e14711038dd53ab13 ]
Using the DMA in cyclic mode modifies the hardware buffer descriptor
chain in xilinx_dma_prep_dma_cyclic so that the last descriptor used by
the cyclic transfer points back to the first descriptor, but it never
restores the original descriptor ring. This breaks using non-cyclic mode
after cyclic mode with an error like:
xilinx-vdma 86000000.dma: Channel 00000000354d5c8d has errors 100, cdr 6de40000 tdr 6de40400
The only way to get out of this error state is to rebuild the hardware
buffer descriptor ring by releasing and re-acquiring the channel.
Fix using non-cyclic mode after cyclic mode by always restoring the
original buffer descriptor ring in the same manner as it is set up by
xilinx_dma_alloc_chan_resources().
Fixes: 23059408b6a3 ("dmaengine: xilinx_dma: Fix race condition in the driver for multiple descriptor scenario")
Signed-off-by: Alex Bereza <alex@bereza.email>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Suraj Gupta <suraj.gupta2@amd.com>
Link: https://patch.msgid.link/20260817-fix-hw-buf-desc-after-cyclic-mode-v1-1-1fe47e701d6c@bereza.email
Link: https://patch.msgid.link/20260818-fix-hw-buf-desc-after-cyclic-mode-v2-1-530ff44c6a81@bereza.email
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Alex Bereza <alex@bereza.email>
Date: Mon Aug 17 11:23:55 2026 +0200
dmaengine: xilinx_dma: Fix hardware buffer descriptor reuse order
[ Upstream commit cee9c863ee68cb27d66745eb03f60e357f4f8ad2 ]
xilinx_dma_alloc_chan_resources() builds a static ring of hardware
buffer descriptors once and the driver uses this ring throughout the
lifetime of a channel. This requires the allocation order of hardware
buffer descriptors from chan->free_seg_list to stay in sync with the
hardware buffer descriptor ring built at channel allocation time by
returning oldest descriptors to chan->free_seg_list first.
When chan->pending_list is not empty e.g. during
xilinx_dma_terminate_all() the chan->free_seg_list and the order of the
static hardware buffer descriptor ring get out of sync. Descriptors age
in this order: pending -> active -> done. So freeing pending_list first
returns the newest buffer descriptors to the chan->free_seg_list first
and thus breaks the order required by the static hardware buffer
descriptor ring. Then when the channel is reused, after a wrap around of
the free_seg_list the DMA will find a hardware buffer descriptor with a
length field that is still zeroed and stop with something like this:
xilinx-vdma 86000000.dma: Channel 000000003a21d7b8 has errors 10, cdr 6de4c000 tdr 6de4c000
After this no more descriptors are completed and a consumer potentially
blocks and waits forever. The only way to get out of this error state is
to rebuild the static hardware buffer descriptor ring and the
free_seg_list by releasing and re-acquiring the channel.
Fix the order in which hardware buffer descriptors are returned to
free_seg_list to ensure the mentioned requirement holds.
Fixes: 23059408b6a3 ("dmaengine: xilinx_dma: Fix race condition in the driver for multiple descriptor scenario")
Signed-off-by: Alex Bereza <alex@bereza.email>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Suraj Gupta <suraj.gupta2@amd.com>
Link: https://patch.msgid.link/20260817-fix-hw-buf-desc-reuse-v1-1-d79827a844c7@bereza.email
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Francis Marlou Pacaro <pacaro.francis.marlou.n@gmail.com>
Date: Fri Sep 4 08:20:21 2026 +0800
drm/amd/display: fix MALL hysteresis timer underflow at high refresh rates
[ Upstream commit 2ac2fe765ef475f409616ac0b57c4a3922749b0f ]
dcn30_apply_idle_power_optimizations() derives the MALL frame cache
hysteresis timer with
tmr_delay = (uint32_t)(div_u64(..., denom) - 64LL);
div_u64() returns a u64, so when the quotient is smaller than 64 the
subtraction wraps instead of going negative and tmr_delay ends up huge.
The loop that follows tries to squeeze it into the 6 bit register field
by doubling denom, but that only makes the quotient smaller, so tmr_delay
can never converge. tmr_scale is bumped past 3 and the function gives up
with
/* Delay exceeds range of hysteresis timer */
ASSERT(false);
even though the requested delay is too *short* to encode, not too long.
With mall_additional_timer_percent left at its default of 0, the quotient
drops below 64 once the refresh rate used for the calculation goes above
~243 Hz. Every DCN 3.0 display above that loses MALL static screen
entirely and splats a WARN once per boot. Reproduced on Navi 23
(RX 6600) driving 1920x1080, resetting /sys/kernel/debug/clear_warn_once
between modes:
refresh MALL ASSERT
144 Hz enabled no
240 Hz enabled no
280 Hz skipped yes
360 Hz skipped yes
Commit 3bb68cec4db8 ("drm/amd/display: Add Overflow check to skip MALL")
already covered the other end of the range, where a large stutter period
makes the delay too long to encode. Cover the short end by clamping to
0, which selects the shortest hysteresis the register can express,
65.28us * 64 = ~4.18ms. That is marginally longer than what the formula
asks for at these refresh rates, and erring long is the safe direction:
it only delays MALL entry, it can never enter early.
The numerator does not change between iterations, only denom does, so
compute it once and keep both call sites inside 100 columns.
The genuinely out of range case at very low refresh rates still reaches
the ASSERT, which is where it belongs.
Fixes: 52f2e83e2fe5 ("drm/amdgpu/display: add MALL support (v2)")
Signed-off-by: Francis Marlou Pacaro <pacaro.francis.marlou.n@gmail.com>
Reviewed-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 387550e53e1405f1f960b62b22f8783db17c8e1d)
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Dmitriy Chumachenko <Dmitry.Chumachenko@cyberprotect.ru>
Date: Mon Sep 14 17:33:03 2026 +0300
drm/amdgpu: check ras and obj before dereference
[ Upstream commit 723d4dc628d764b19cf9efca14b82cca5ff020c9 ]
nbio_v7_9_handle_ras_controller_intr_no_bifring() dereferences ras and obj
without checking either for NULL. Both amdgpu_ras_get_context() and
amdgpu_ras_find_obj() can return NULL, e.g. during the window between
adev->nbio.ras being set (early in amdgpu_ras_init(), by design, to
enable the fatal-error interrupt as soon as possible) and the PCIE_BIF
ras object actually being created in RAS late_init. Any interrupt in that
window crashes in hard-IRQ context.
This is analogous to commit d190b459b2a4 ("drm/amdgpu: the warning
dereferencing obj for nbio_v7_4"), which fixed the same issue in the
nbio_v7_4 handler.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Fixes: 7692e1ee2446 ("drm/amdgpu: add RAS fatal error handler for NBIO v7.9")
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Dmitriy Chumachenko <Dmitry.Chumachenko@cyberprotect.ru>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit c7071767a50a32ed727cf800ac84372429e3b4b3)
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Mario Limonciello <mario.limonciello@amd.com>
Date: Tue Sep 22 15:34:00 2026 -0400
drm/amdgpu: Fix GPU PCIe link capability reporting
[ Upstream commit 04de4007d32385b8b6a5dd72bff3146dfdc592c3 ]
Commit eb53125a7ad9 ("drm/amd: Add dedicated helper for
amdgpu_device_find_parent()") made amdgpu_device_gpu_bandwidth() query
the first device outside the dGPU. That is the host side of the
physical link, not the GPU side.
As a result, the ASIC and platform capability masks can both be based
on the host port. drm_amdgpu_info_device then exposes the host
capabilities to userspace, such as Gen5 x16 for a Gen4 x8 GPU.
Cache both ends of the physical link during device initialization.
Use link_dev for the GPU capability and link_partner for the platform
capability and _PR3 detection.
Reported-by: "Marek Olšák" <maraeo@gmail.com>
Closes: https://lore.kernel.org/amd-gfx/CAAxE2A4VhsAzzO1QjBjUg+NgnbD04ZzMyN6xsUJxjKJHH6hxiw@mail.gmail.com/
Suggested-by: Lijo Lazar <lijo.lazar@amd.com>
Fixes: eb53125a7ad9 ("drm/amd: Add dedicated helper for amdgpu_device_find_parent()")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 7ea6a47224e2c6e89a3a682d7fbaace4817a55aa)
Cc: stable@vger.kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Chengjun Yao <Chengjun.Yao@amd.com>
Date: Tue Sep 8 10:15:43 2026 +0800
drm/amdgpu: fix rmmio iounmap skipped on device removal
commit 5155002b03b24ba3ef91c5c313b8cf0171b24904 upstream.
amdgpu_pci_remove() calls drm_dev_unplug() before fini_sw(), so
drm_dev_enter() is already false there and the iounmap() guarded by it
is skipped. This .remove path runs on both hot-unplug and plain rmmod,
so the register BAR ioremap mapping leaks one instance per unload.
Unmap rmmio unconditionally (guard only on non-NULL) and drop the now
unused idx.
Fixes: 62d5f9f7110a ("drm/amdgpu: Unmap MMIO mappings when device is not unplugged")
Signed-off-by: Chengjun Yao <Chengjun.Yao@amd.com>
Reviewed-by: Asad Kamal <asad.kamal@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit dd6f86a97260e5207d3329ad03aa89fdad61b1e6)
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Mike Lothian <mike@fireburn.co.uk>
Date: Sat Sep 12 00:29:08 2026 +0100
drm/amdgpu: hold a runtime PM reference for P2P dma-buf attachments
commit 636139603b99d2e3a18a46cf3f8d39313ce8042e upstream.
amdgpu_dma_buf_map() adds VRAM to the allowed domains for a peer2peer
attachment. GTT is only a fallback placement when VRAM is preferred, so
ttm_bo_validate() migrates the buffer from GTT into VRAM. While the
exporting device is runtime suspended its SDMA rings are down and the
move fails:
amdgpu: Move buffer fallback to memcpy unavailable
An importer on a second GPU reaches this holding no runtime PM
reference on the exporter, e.g. a compositor on the APU submitting a
frame that references a buffer exported by an idle dGPU:
amdgpu_cs_ioctl -> amdgpu_cs_parser_bos -> amdgpu_cs_bo_validate
-> ttm_bo_validate -> amdgpu_bo_move -> dma_buf_map_attachment
-> amdgpu_dma_buf_map -> ttm_bo_validate -> amdgpu_bo_move
Pinning a dma-buf into VRAM has the same requirement, which
commit 030631e97b20 ("drm/amdgpu: revert "take runtime pm reference
when we attach a buffer" v2") called out as the one case that would
need the reference back.
Take it in attach and drop it in detach. pm_runtime_get_if_active()
never resumes the device, so it cannot deadlock against the reservation
taken during resume, which is why the old pm_runtime_get_sync() had to
go. If the device is not active, clear peer2peer instead: the buffer
then stays in GTT, which remains accessible while the GPU is powered
down. If runtime PM is disabled, take a plain reference so the put in
detach stays balanced.
Fixes: 030631e97b20 ("drm/amdgpu: revert "take runtime pm reference when we attach a buffer" v2")
Suggested-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Mike Lothian <mike@fireburn.co.uk>
Assisted-by: Claude:Opus-5 [Claude Code]
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 062ff15e30a48d14fb7d7558eba84f8dc97197f0)
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: chong li <chongli2@amd.com>
Date: Tue Sep 22 15:33:59 2026 -0400
drm/amdgpu: reduce early full GPU access during SR-IOV init
[ Upstream commit cba4928cdffaa0f9012acff0ec4f896320107077 ]
Allow early FB reads to fall back to BAR0 when the VRAM aperture is not ready.
This lets SR-IOV VFs consume host-provided init data before requesting full GPU access.
For ASICs that support request_init_data,
defer full GPU access until after non-GPU early init to shorten the full-access window.
Legacy ASICs(before NV12) do not send request_init_data;
the host dumps init data only during full GPU access,
so keep the original early full-access request path for them.
Signed-off-by: chong li <chongli2@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Stable-dep-of: 04de4007d323 ("drm/amdgpu: Fix GPU PCIe link capability reporting")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Mario Limonciello <mario.limonciello@amd.com>
Date: Wed Aug 26 13:51:02 2026 -0500
drm/amdgpu: restrict BAR0 fallback read to SR-IOV VFs only
commit bd1f08246b8a2564d8ac61df715b6bcd5f994729 upstream.
The BAR0 fallback read path was introduced as a workaround for SR-IOV VFs
where the VRAM aperture is not available during early init. Restrict this
workaround to only SR-IOV VFs where it's needed.
Reported-by: gloveless@jqluv.com
Fixes: cba4928cdffa ("drm/amdgpu: reduce early full GPU access during SR-IOV init")
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://patch.msgid.link/20260826185102.2269511-1-mario.limonciello@amd.com
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit d8a0affd207c813bd063fa2c27786f449eaf92b8)
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Mario Limonciello <mario.limonciello@amd.com>
Date: Tue Sep 15 12:51:24 2026 -0500
drm/amdgpu: Skip KFD mapping clear before initialization
commit 7f9caa70aef0950e06d395ca0035831214d88187 upstream.
amdgpu_amdkfd_clear_kfd_mapping() assumes that a non-NULL kfd_dev
has a fully populated node array. This is not true when KFD device
initialization fails after probe.
For example, kgd2kfd_device_init() sets num_nodes before checking
PCIe atomics support. On Polaris systems without the required atomics,
it returns before allocating nodes[0], but the kfd_dev remains attached
to the amdgpu device. A later GPU reset then dereferences nodes[0]->id.
Require the authoritative KFD initialization flag before walking the
node array, matching the existing KFD reset and teardown paths.
Fixes: 70cadefcc616 ("drm/amdgpu: unmap all user mappings of framebuffer and doorbell before mode1 reset")
Closes: https://gitlab.freedesktop.org/drm/amd/-/work_items/5833
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 4ac1835823c47903fbb278bbf474773c46f59edc)
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: David Francis <David.Francis@amd.com>
Date: Wed Aug 5 09:16:51 2026 -0400
drm/amdkfd: Avoid integer underflow in EOP ring size calculation.
commit 8ee521b8b189799e361d4233c5180ba56656d4d4 upstream.
The low 6 bits of cp_hqd_eop_control store the base-2 logarithm
of the EOP ring size. This was calculated as
order_base_2(q->eop_ring_buffer_size / 4) - 1
But order_base_2 can in theory return 0, so this could underflow
(although in practice the ring buffer size cannot be less than 4096).
Change this to
order_base_2(q->eop_ring_buffer_size / 8)
using properties of logarithms.
Also add to the above comment to make the mathematics more clear.
Reviewed-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: David Francis <David.Francis@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit f0f43fcf8b2b3a924cad9444340921c96ed5f634)
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: David Francis <David.Francis@amd.com>
Date: Wed Aug 5 09:51:35 2026 -0400
drm/amdkfd: Avoid integer underflow with ffs in EOP ring size calc
commit c883d0a132d430ef7ebb23fd94323be94d0fbdb8 upstream.
The low 6 bits of cp_hqd_eop_control store the base-2 logarithm
of the EOP ring size. This was calculated as
ffs(q->eop_ring_buffer_size / sizeof(unsigned int)) - 1 - 1
But ffs can in theory return 1 or 0, so this could underflow
(although in practice the ring buffer size cannot be less than 4096).
Change this to
ffs(q->eop_ring_buffer_size / sizeof(unsigned int) / 4)
using properties of logarithms.
Reviewed-by: Kent Russell <kent.russell@amd.com>
Signed-off-by: David Francis <David.Francis@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 4f18c56630383c14bfc6b2d65f88f2f895d2121a)
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Vladimir Marioukhine <Vladimir.Marioukhine@amd.com>
Date: Wed Aug 12 13:19:46 2026 -0400
drm/amdkfd: implement restore_mqd callbacks for GFX12/12.1
commit 5f28bb1c2cd9dcdb76a20d61b3ea069b85893c59 upstream.
kfd_mqd_manager_v12.c (GFX 12.0) and kfd_mqd_manager_v12_1.c (GFX 12.1)
do not implement restore_mqd callbacks, leaving the function pointers
NULL and causing CRIU restore to return -EOPNOTSUPP on GFX12.
Implement restore_mqd for both compute and SDMA queues in
kfd_mqd_manager_v12.c and kfd_mqd_manager_v12_1.c, modeled after the
GFX 11 implementation with the following improvements:
- update cp_mqd_base_addr_lo/hi to the newly allocated MQD address,
fixing a pre-existing gap shared with v11 where the in-MQD copy
still pointed at the old checkpoint-time address after restore
- memset the full allocation before memcpy for compute queues to avoid
stale data in the GTT sub-allocator tail; SDMA MQDs use sizeof(*m)
since they are packed at mqd_size stride in a shared BO
checkpoint_mqd registration is deferred to a follow-up patch that also
implements get_checkpoint_info, so that checkpoint and restore are
enabled together as a complete and testable unit.
Note: GFX12.1 restore handles XCC0 only. Multi-XCC CRIU restore is
currently unreachable due to a separate validation issue in
kfd_criu_restore_queue(). A pr_warn_once() is emitted if a multi-XCC
device is encountered.
Signed-off-by: Vladimir Marioukhine <Vladimir.Marioukhine@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit b1f9601237d050f5df478464cf51bf1fff29a256)
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Sajal Gupta <sajal2005gupta@gmail.com>
Date: Wed Sep 2 18:00:57 2026 +0530
drm/gud: fix out-of-bounds write in gud_plane_atomic_check()
commit 59ced288fcba9e91bd38e61a972ad782c4edb7d0 upstream.
The plane property loop uses req->properties[num_properties + i] as write
index while simultaneously incrementing `num_properties` inside the loop.
At iteration i, num_properties has also incremented by i, so the write
is done at `initial_num_properties + 2*i`, skipping every other index and
advancing by 2 per iteration.
With just 2 connector and 32 plane properties the last write happens at
index 64, one slot past the end of the 64-slot (indices 0–63)
allocation. A USB device can trigger OOB by advertising the maximum
number of properties.
Fix by dropping the redundant `+ i`; num_properties is already the correct
running index, as gud_connector_fill_properties() fills the preceding
slots.
Fixes: 40e1a70b4aed ("drm: Add GUD USB Display driver")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Link: https://sashiko.dev/#/patchset/20260821071812.16500-1-sajal2005gupta%40gmail.com?part=1
Signed-off-by: Sajal Gupta <sajal2005gupta@gmail.com>
Cc: <stable@vger.kernel.org>
Acked-by: Ruben Wauters <rubenru09@aol.com>
Signed-off-by: Ruben Wauters <rubenru09@aol.com>
Link: https://patch.msgid.link/20260902123254.36987-1-sajal2005gupta@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Sophie D <patches@scd31.com>
Date: Wed Sep 9 21:49:10 2026 -0400
drm/gud: Ignore damage clips in full update mode
commit effce1cb87ee0d8b3a8cbe7722968f4ea7efd360 upstream.
When running in full update mode, previously small updates (such as
moving the mouse across the screen) would cause many full frames to be
generated. This would bog down the bus and lower the effective framerate
significantly - I was seeing a drop from 60 FPS to 2 FPS.
Set ignore_damage_clips in full update mode so the damage iterator
yields a single full-plane rectangle instead of one per clip.
Fixes: 73cfd166e045 ("drm/gud: Replace simple display pipe with DRM atomic helpers")
Cc: <stable@vger.kernel.org> # 6.18.x
Signed-off-by: Sophie D <patches@scd31.com>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Ruben Wauters <rubenru09@aol.com>
Signed-off-by: Ruben Wauters <rubenru09@aol.com>
Link: https://patch.msgid.link/20260910014910.8564-1-patches@scd31.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Guangshuo Li <lgs201920130244@gmail.com>
Date: Sat Aug 8 21:16:24 2026 +0800
drm/msm/adreno: fix autosuspend cleanup during teardown
commit 6fbbf1e152f34ad3913e4a6476680aba672c5068 upstream.
adreno_gpu_init() calls pm_runtime_use_autosuspend(), but
adreno_gpu_cleanup() does not call the matching
pm_runtime_dont_use_autosuspend() during teardown.
If the autosuspend delay is set to a negative value while autosuspend
is enabled, the runtime PM core increments usage_count to prevent
runtime suspend. Without calling pm_runtime_dont_use_autosuspend()
during teardown, this reference is not dropped and usage_count remains
unbalanced.
The documentation for pm_runtime_use_autosuspend() also notes that it
is important to undo it with pm_runtime_dont_use_autosuspend() at
driver exit time, unless runtime PM was initially enabled with
devm_pm_runtime_enable().
Add the missing pm_runtime_dont_use_autosuspend() call to
adreno_gpu_cleanup().
This issue was found by manual code inspection.
Fixes: eeb754746b14 ("drm/msm/gpu: use pm-runtime")
Cc: stable@vger.kernel.org
Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Patchwork: https://patchwork.freedesktop.org/patch/745110/
Link: https://lore.kernel.org/r/20260808131624.2854412-1-lgs201920130244@gmail.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Karl Mehltretter <kmehltretter@gmail.com>
Date: Sun Sep 6 19:03:01 2026 +0200
drm/msm/adreno: Fix the skip_gpu parameter description
[ Upstream commit 8061ee61b9426fe38350fa9eead2d9c50b03deb6 ]
The module parameter is skip_gpu, but its MODULE_PARM_DESC() names
no_gpu, so modinfo describes a parameter that does not exist and shows
no description for the real one.
Use the parameter name in the description.
Fixes: 3f17991488af ("drm/msm/adreno: Add a modparam to skip GPU")
Assisted-by: LLM
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/751406/
Message-ID: <20260906170301.2393-1-kmehltretter@gmail.com>
Signed-off-by: Rob Clark <robin.clark@oss.qualcomm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: William Bright <william.bright@imd-tec.com>
Date: Wed Aug 12 11:05:52 2026 +0100
drm/msm/dp: fix link bandwidth check when wide bus is enabled
[ Upstream commit 58995b11dfb7dda095d23f22fa4dc79b923b5adf ]
msm_dp_display_mode_valid() halves the pixel clock when either YUV420 or
wide bus is in use, then uses that halved value both for the controller
pixel clock limit and for the DP link bandwidth check.
Only YUV420 halves the data crossing the link. Wide bus widens the
internal DPU to DP interface to two pixels per clock, halving the
controller clock. Every pixel is still transmitted, so the link
bandwidth requirement remains.
As a result, modes needing up to twice the available link bandwidth pass
validation. On the IMDT QCS8550 SBC (rev5 with CYPD6125), where DP runs
over USB-C alt mode where only two lanes are available, 3840x2160@60 was
accepted despite needing 9.6 Gbps against the 8.64 Gbps the link can
carry.
Use a separate link pixel clock that is only halved for YUV420 for the
bandwidth calculation, leaving the wide bus halving to apply solely to
the controller pixel clock limit. With this, 4k@60 is correctly rejected
and 4k@30 selected instead.
Fixes: df9cf852ca30 ("drm/msm/dp: account for widebus and yuv420 during mode validation")
Assisted-by: Claude:claude-opus-5
Signed-off-by: William Bright <william.bright@imd-tec.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Patchwork: https://patchwork.freedesktop.org/patch/746145/
Link: https://lore.kernel.org/r/20260812-msm-dp-link-bw-v1-1-b0e3ce1190be@imd-tec.com
[DB: dropped useless comment]
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jesse Casco <jesse.casco@gmail.com>
Date: Sat Aug 8 13:13:25 2026 -0400
drm/msm/dp: skip PUSH_IDLE when the link was never enabled
[ Upstream commit e249a6e2a130c08bb4d8b0a55cbe29754307e5c9 ]
msm_dp_display_atomic_enable() returns early when link training fails,
leaving ->power_on false and the main link down.
msm_dp_display_atomic_disable() nevertheless writes DP_STATE_CTRL_PUSH_IDLE
and waits for an idle-pattern completion that cannot arrive, so every failed
enable is followed by "PUSH_IDLE pattern timedout".
Every other step of the teardown is already gated on that flag:
msm_dp_display_disable(), called from .atomic_post_disable(), returns early
on !power_on. The PUSH_IDLE write is the only one that is not, so the
controller's runtime-PM reference is then dropped without the link having
been taken down.
On glymur (Snapdragon X2 Elite) the consequence is not a warning. The SoC
does not survive it: TrustZone force-stops the SOCCP and ADSP remote
processors and the machine resets silently about 50 ms later, with no oops
and no panic. On an ASUS Zenbook A16 (UX3607OA), whose eDP panel does not
currently train, this reproduces without any compositor or GPU involvement:
# eDP enable has already failed with "Failed link training (rc=-104)"
echo 1 > /sys/class/graphics/fb0/blank
[535.645455] === marker ===
[535.694833] qcom_q6v5_pas d00000.remoteproc: fatal error received: \
sys_m_smsm.c:512:TZ force stop
[535.694875] remoteproc remoteproc0: crash detected in soccp: type fatal error
[535.728857] qcom_q6v5_pas 6800000.remoteproc: fatal error received: \
sys_m_smsm.c:783:err fatal notification received from TZ
<SoC reset>
Gate the PUSH_IDLE write on ->power_on so the disable path is consistent
with the rest of the teardown. With this applied the same sequence is
harmless and the machine stays up; without it, it resets every time.
The unconditional write dates back to the original DP driver
(c943b4948b58 ("drm/msm/dp: add displayPort driver support")), but the
surrounding code has been restructured several times since, so no Fixes:
tag is offered.
Note that the eDP link-training failure that exposes this on the A16 is a
separate problem in the glymur eDP PHY and is reported separately; this
change is about not damaging the machine when training fails, for whatever
reason.
Tested on ASUS Zenbook A16 (UX3607OA), Snapdragon X2 Elite Extreme, on
linux-next next-20260803 and next-20260807. The machine has since been
running next-20260807 with this patch as its daily driver.
Assisted-by: Anthropic:Claude-Opus-5
Signed-off-by: Jesse Casco <jesse.casco@gmail.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Patchwork: https://patchwork.freedesktop.org/patch/745167/
Link: https://lore.kernel.org/r/20260808171325.133041-1-jesse.casco@gmail.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Saim Shujah <saimzst@gmail.com>
Date: Fri Aug 28 11:54:40 2026 +0500
drm/msm/dpu: clear pending peripheral flush state
commit a5b5cc909931572aec446e129c035b76b3f0c1fa upstream.
dpu_hw_ctl_clear_pending_flush() resets the cached per-block state after a
flush transaction, but misses pending_periph_flush_mask.
The peripheral flush updater accumulates interface bits in this mask. A
later transaction which sets the top-level peripheral flush bit can write
stale interface bits to CTL_PERIPH_FLUSH together with the current state.
Peripheral flush support was added after the helper started clearing every
individual pending flush mask. Clear the peripheral mask together with the
other cached child masks.
Fixes: 64f7b81f0358 ("drm/msm/dpu: add support of new peripheral flush mechanism")
Cc: stable@vger.kernel.org
Signed-off-by: Saim Shujah <saimzst@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/748968/
Link: https://lore.kernel.org/r/20260828065440.140410-1-saimzst@gmail.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Date: Thu Sep 3 15:19:10 2026 +0300
drm/msm/dsi: round the byte clock rate after reparenting to the PHY PLL
[ Upstream commit 2028280686f4fa78e2f1f6dede4b6c1fd782b9e3 ]
DSI 6G v2.9 hosts (SM8650, SM8750, Kaanapali, etc.) reparent the byte and
pixel RCGs to the DSI PHY PLL at runtime from
dsi_link_clk_set_rate_6g_v2_9(), after the PHY has been enabled. However
dsi_calc_clk_rate_6g() runs earlier, in order to compute the bit clock
request for the PHY. At that point the byte RCG still has its reset
parent (XO), so clk_round_rate() returns a bogus rate, which then ends up
in the PHY bit clock request and the PLL gets programmed to a wrong
frequency, breaking the panel.
Move the rounding to dsi_link_clk_set_rate_6g(), which is called after
the RCGs have been reparented to the PLL. Storing the rounded rate at
this point still makes later link_clk_set_rate() calls no-ops in the
CCF. Derive the byte interface clock rate from the rounded byte clock
rate, otherwise it would keep requesting the idealized rate and
retrigger the PLL on every transfer.
Reported-by: Abel Vesa <abel.vesa@oss.qualcomm.com>
Reported-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com>
Fixes: 6cd33b6f4155 ("drm/msm/dsi: round 6G byte clock rate to the PLL-achievable value")
Assisted-by: LLM
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Tested-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> # SM6115P J606F
Tested-by: Abel Vesa <abel.vesa@oss.qualcomm.com>
Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com>
Patchwork: https://patchwork.freedesktop.org/patch/750496/
Link: https://lore.kernel.org/r/20260903-fix-eliza-dsi-v1-1-3474a6c9f2e0@oss.qualcomm.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Guangshuo Li <lgs201920130244@gmail.com>
Date: Sun Sep 13 16:58:14 2026 +0800
drm/msm/hdmi_phy: fix runtime PM cleanup on probe failure
commit f4fae975db08a9aeec0b15e145c7d4d0fe02a0ec upstream.
msm_hdmi_phy_probe() enables runtime PM before enabling the PHY
resources and initializing the PLL, but failures from either operation
return without calling the matching pm_runtime_disable().
The remove path disables runtime PM, but it is not called when probe
fails. As a result, runtime PM remains enabled after an unsuccessful
probe.
Route failures after pm_runtime_enable() through a common error path
and disable runtime PM before returning.
This issue was found by manual code inspection.
Fixes: 15b4a4523859 ("drm/msm/hdmi: Create a separate HDMI PHY driver")
Cc: stable@vger.kernel.org
Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Patchwork: https://patchwork.freedesktop.org/patch/753043/
Link: https://lore.kernel.org/r/20260913085814.1509352-1-lgs201920130244@gmail.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Karl Mehltretter <kmehltretter@gmail.com>
Date: Sun Sep 6 19:03:47 2026 +0200
drm/msm: Fix the separate_gpu_kms parameter description
[ Upstream commit adf5967331318bcb436fc80069915231ec039352 ]
The module parameter is separate_gpu_kms, but its MODULE_PARM_DESC()
names separate_gpu_drm, so modinfo describes a parameter that does not
exist and shows no description for the real one.
Use the parameter name in the description.
Fixes: 217ed15bd399 ("drm/msm: enable separate binding of GPU and display devices")
Assisted-by: LLM
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/751407/
Message-ID: <20260906170347.2427-1-kmehltretter@gmail.com>
Signed-off-by: Rob Clark <robin.clark@oss.qualcomm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jonghyuk Kim(MalHyuk) <malhyuk97@gmail.com>
Date: Wed Sep 2 10:27:20 2026 +0900
drm/msm: RCU-free the scheduler-containing ring and VM objects
commit 01c8d1f385f788f1bbbbb7687c4386d614281218 upstream.
Both struct msm_ringbuffer and struct msm_gem_vm embed a struct
drm_gpu_scheduler. msm_ringbuffer_destroy() and the VM free callback
msm_gem_vm_free() call drm_sched_fini() on the embedded scheduler and then
free the containing object with plain kfree().
drm_sched_fence_get_timeline_name() returns fence->sched->name, and the
scheduler fence keeps a .release callback so it is not ops-detached on
signalling. A finished fence exported to userspace (the submit out-fence, or
a VM_BIND fence, via sync_file / drm_syncobj) keeps pointing at the embedded
scheduler after the ring/VM is freed, so a later get_timeline_name() --
reachable unprivileged through SYNC_IOC_FILE_INFO -- dereferences freed slab
memory (KASAN slab-use-after-free read).
Per the dma-fence lifetime contract the exporter must keep the data backing a
signalled fence alive for an RCU grace period. Free the scheduler-containing
objects with kfree_rcu() instead of kfree().
Fixes: 1d8a5ca436ee ("drm/msm: Conversion to drm scheduler")
Fixes: 92395af63a99 ("drm/msm: Add VM_BIND submitqueue")
Cc: stable@vger.kernel.org
Signed-off-by: Jonghyuk Kim(MalHyuk) <malhyuk97@gmail.com>
Patchwork: https://patchwork.freedesktop.org/patch/750234/
Message-ID: <20260902012720.880783-1-malhyuk97@gmail.com>
Signed-off-by: Rob Clark <robin.clark@oss.qualcomm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Lyude Paul <lyude@redhat.com>
Date: Fri Aug 14 15:43:50 2026 -0400
drm/nouveau/gsp/r570: Enable S/R Display workaround in GSP
commit 24fbd6d4bcf3363ef13ebe0d36dea93f30396c6d upstream.
There's two flags that we've never been setting when asking GSP to suspend
the GPU, which OpenRM does set:
GPU_STATE_FLAGS_PRESERVING
GPU_STATE_FLAGS_PM_TRANSITION
These flags aren't -supposed- to do much in GSP, they're mostly used by
OpenRM itself for state tracking. The only thing they do from GSP's side is
control whether or not a single display related workaround is applied
during suspend.
But as it turns out, that single workaround is actually quite crucial for
getting runtime PM working with nouveau - and without it set we end up
seeing a lot more failures with runtime PM resume. So, let's start setting
it.
Signed-off-by: Lyude Paul <lyude@redhat.com>
Fixes: 53dac0623853 ("drm/nouveau/gsp: add support for 570.144")
Cc: <stable@vger.kernel.org> # v6.16+
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patch.msgid.link/20260814194542.781955-4-lyude@redhat.com
(cherry picked from commit ca57629b3eb912c77bc4357178a2130ea6c2d6df)
Signed-off-by: Lyude Paul <lyude@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Lyude Paul <lyude@redhat.com>
Date: Fri Aug 14 15:43:49 2026 -0400
drm/nouveau/gsp/r570: Set GcOff = 0 in fbsr
commit 12f6eff11ccf9cad3b2dfcdd94184fdb9ffface2 upstream.
Previously, it looked as if we were able to fix suspend/resume on some
desktops by setting Gcoff based on whether or not we were entering runtime
PM. This was a mistake though - the only time suspend/resume would end up
actually working was if Gcoff = 0.
It seems like it's likely the main reason for this is the FBSR GcOff
argument actually controls GSP's behavior with regards to which buffers it
decides to save across suspend/resume. When GcOff = 1, RM reserved regions
are saved unless they are marked as LOST_ON_SUSPEND, and RM channel-context
and kernel-client buffers are also saved -including- when they are
LOST_ON_SUSPEND. This means with GcOff = 1, we end up having GSP save and
restore buffers that actually need to be reinitialized on resume - causing
the failures we're setting.
Thanks to John Hubbard from Nvidia for providing some background on what
these options do in the GSP firmware do!
Signed-off-by: Lyude Paul <lyude@redhat.com>
Fixes: 53dac0623853 ("drm/nouveau/gsp: add support for 570.144")
Cc: <stable@vger.kernel.org> # v6.16+
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patch.msgid.link/20260814194542.781955-3-lyude@redhat.com
(cherry picked from commit c7abe771e013848970421e5ca29c6b2f05c31965)
Signed-off-by: Lyude Paul <lyude@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Lyude Paul <lyude@redhat.com>
Date: Fri Aug 14 15:43:51 2026 -0400
drm/nouveau/gsp: Increase delay for magic sleep in r535_gsp_fini()
commit bbb9293c9bb792f3f16c842f223b8c97bdbaf227 upstream.
As it turns out, Turing isn't the only architecture that needs this. On
this Dell Precision 7780 with an AD103 GPU, along with pretty much every
other laptop I tested, runtime PM is still somewhat unreliable. At first
glance it seems as if it's fixed, but lowering the autosuspend delay to
500ms and then doing a stress test of suspend/resume cycles on the GPU ends
up causing everything to start timing out.
After quite a lot of digging, I eventually landed back on this magic
timeout in r535_gsp_fini(). As it turns out, increasing the timeout ends up
fixing the runtime PM issues as far as I can tell, even during intense
stress testing.
Unfortunately after spending quite a bit of time trying to dig through
OpenRM to figure out what this magic sleep is actually doing, I've also
come up short with any reasonable explanation. In lieu of that, I'm going
to include the observations I did make while trying to figure this out in
hopes someone eventually does figure this out:
* The magic sleep has to occur after fbsr is initialized. Performing it at
any time before that doesn't appear to work.
* In situations where runtime PM starts getting flaky, some rather
interesting visual effects end up happening on occasion before the GPU
fully falls over. In particular, squares that look like the result of an
incomplete blitting operation to a tiled buffer end up showing up on
applications like vkcube. Interestingly enough, they remain in precisely
the same place between runtime PM cycles until the GPU falls over - even
when restarting vkcube multiple times, and even when vkcube is actively
updating the screen. Even more interestingly, they're not limited to a
specific framebuffer - you can see the squares changing as the cube
rotates around.
We cannot however, say that this is likely to be a incomplete fbsr
operation. The magic sleep happens before fbsr is actually saved (which
happens on the GSP unload), so it's something else.
* During a short bit of testing with a desktop that I have, the magic sleep
seemed to make no difference to whether or not suspend/resume works. It
seems to generally work almost always. So we can assume this is likely
exclusive to runtime PM, not S3.
As well, here's a list of the things I tried before settling on the magic
sleep:
* Hooking up NV2080_CTRL_CMD_INTERNAL_GCX_ENTRY_PREREQUISITE and then
blocking runtime PM until OpenRM signals that GC6/GCOFF is ready appears
to make no difference.
* Hooking up some (maybe not all, unsure about that part) bits of comptag
saving including:
* Fetching static memsys information from GSP
* Adding the size of the comptag storage to the fbsr data
* Adding a GA103+ workaround for disabling raw compression mode during
fbsr (it doesn't seem like it applies for any systems I tried it on
anyhow)
* Setting bPreserveVideoMemoryAllocations=1 in GspSystemInfo
So, until we can figure this out properly - just sleep for longer.
Signed-off-by: Lyude Paul <lyude@redhat.com>
Fixes: 53dac0623853 ("drm/nouveau/gsp: add support for 570.144")
Cc: <stable@vger.kernel.org> # v6.16+
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patch.msgid.link/20260814194542.781955-5-lyude@redhat.com
(cherry picked from commit 09b47186a4164f3aaa3591313f80794443117342)
Signed-off-by: Lyude Paul <lyude@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Date: Tue Sep 15 16:05:57 2026 +0100
drm/sched: Fix virtual runtime race
commit 2ab510e63197360945f915dd5631a77c63ac6b27 upstream.
Prevent pushing a new job to an entity seeing it being the first in the
queue, and hence entering the drm_sched_rq_add_entity() path, if the pop
side in drm_sched_entity_pop_job() has just de-queued the job but not yet
updated the saved virtual time.
Restoring the unsaved virtual time, which is at this point not a delta but
still an absolute value, pushes the said entity to the rear of the run queue
for a potentially very long time.
We close this race by pulling the locked sections out to encompass both
the queue push/pop and corresponding rbtree management.
This is aligned with the future direction to replace the current lockless
job queue with one of the fully locked standard list primitives.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Fixes: 2fa4d8e2c109 ("drm/sched: Add fair scheduling policy")
Suggested-by: Luke.Wildhardt@proton.me # via Claude Opus
Tested-by: Luke.Wildhardt@proton.me
Cc: Christian König <christian.koenig@amd.com>
Cc: Danilo Krummrich <dakr@kernel.org>
Cc: Philipp Stanner <phasta@kernel.org>
Cc: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Vitaly Prosyak <vitaly.prosyak@amd.com>
Cc: stable@vger.kernel.org # v7.2+
[phasta: commit title]
Signed-off-by: Philipp Stanner <phasta@kernel.org>
Link: https://patch.msgid.link/20260915150557.62847-1-tvrtko.ursulin@igalia.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Vadim Nikitushkin <bub4z0r@gmail.com>
Date: Wed Sep 9 23:50:28 2026 +0300
drm/ttm: fix swapped-out resources never leaving their bulk_move range
commit 3db7d7d583419f7b1f2e141e36418802dbb25cf8 upstream.
ttm_tt_swapout() returns the number of pages swapped out on success and
a negative error code on failure; for a populated ttm it never returns
zero. Commit b2ed01e7ad3d ("drm/ttm: Fix ttm_bo_swapout() infinite LRU
walk on swapout failure") moved the bulk_move bookkeeping in
ttm_bo_swapout_cb() under "if (!ret)", so the
ttm_resource_del_bulk_move_unevictable() / ttm_resource_move_to_lru_tail()
pair is now skipped on every successful swapout. The equivalent change
for the shrinker in commit 1d59f36e95f7 ("drm/ttm: Fix ttm_bo_shrink()
infinite LRU walk on backup failure") tests "lret > 0", which is what
was intended here as well.
Before b2ed01e7ad3d the resource was taken off the bulk_move before the
swapout; since then a swapped-out resource stays inside its BO's
bulk_move range (and on the manager LRU) although it is unevictable.
When it is later freed or the BO leaves the bulk_move
(ttm_resource_free(), ttm_bo_set_bulk_move() via amdgpu_vm_bo_del()),
ttm_resource_del_bulk_move() skips it because of its
!ttm_resource_unevictable() guard, so a range endpoint in pos->first /
pos->last is left pointing at freed memory. The next
ttm_lru_bulk_move_tail() or ttm_resource_add_bulk_move() on that cursor
is a use-after-free, seen as the resv WARN in ttm_lru_bulk_move_add(),
"list_del corruption" in ttm_resource_move_to_lru_tail() or a NULL
dereference in ttm_resource_manager_next() -- minutes to hours after a
hibernation, or at process exit / reboot following one. Samuel
Ainsworth's analysis of drm/amd issue 5387 (see Link) identified the
dangling cursor; the missing removal at swapout time is the reason it
dangles.
Testing the condition for success restores the removal. On an AMD
Phoenix APU (ASUS UM3406GA, gfx1103) running suspend-then-hibernate on
a 7.0.y stable kernel carrying the backport (Ubuntu 7.0.0-31) the bug
crashed 5 of 18 hibernation cycles; a function profile of one
hibernation showed 336 ttm_tt_swapout() calls and zero
ttm_resource_del_bulk_move_unevictable() calls. With this change the
removal happens for every swapped-out resource and 12 further cycles
were clean.
Fixes: b2ed01e7ad3d ("drm/ttm: Fix ttm_bo_swapout() infinite LRU walk on swapout failure")
Cc: stable@vger.kernel.org # v7.1+
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/5387
Link: https://lore.kernel.org/dri-devel/CAHYiNPa6aVacJoLOje-qZ1GyYx-9p0tN4NuP8D_eSL+UJeevXw@mail.gmail.com/
Signed-off-by: Vadim Nikitushkin <bub4z0r@gmail.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Link: https://lore.kernel.org/r/20260909205028.13799-1-bub4z0r@gmail.com
[ Squashed with commit fcfe64715b425262af1b36f498f9197f3537ceed
("drm/ttm: apply the swapout bulk_move fix to the intended condition"):
upstream 3db7d7d58341 was applied to the "if (ret)" after
ttm_resource_try_charge() in ttm_bo_alloc_at_place() instead of the
"if (!ret)" after ttm_tt_swapout() in ttm_bo_swapout_cb(), and
fcfe64715b42 restored the former and changed the latter. 7.2.y has no
ttm_resource_try_charge() (dmem cgroup charging is 7.3 material), so
neither commit applies on its own; the net effect of the two is the
single hunk below, which is the change the changelog describes. ]
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Karl Mehltretter <kmehltretter@gmail.com>
Date: Sat Aug 22 16:31:10 2026 +0200
drm/vc4: Use managed KMS polling to fix UAF on unbind
[ Upstream commit 073a30d75f309812ed61af134f24ffef4107b13a ]
vc4_kms_load() calls drm_kms_helper_poll_init() but the driver provides
no matching drm_kms_helper_poll_fini(). The output poll work stays
scheduled after unbind and runs on the freed drm_device:
# modprobe vc4; rmmod vc4; sleep 10
BUG: KASAN: slab-use-after-free in delayed_work_timer_fn
BUG: KASAN: slab-use-after-free in drm_client_dev_hotplug [drm]
Workqueue: events output_poll_execute [drm_kms_helper]
Allocated by task 171: __devm_drm_dev_alloc
Freed by task 262 (rmmod): drm_dev_put / component_del
Use drmm_kms_helper_poll_init() so polling is finalized with the device,
as other drivers do.
Fixes: c8b75bca92cb ("drm/vc4: Add KMS support for Raspberry Pi.")
Assisted-by: Claude:claude-fable-5
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Link: https://patch.msgid.link/20260822143110.68594-1-kmehltretter@gmail.com
Reviewed-by: Maíra Canal <mcanal@igalia.com>
Signed-off-by: Maíra Canal <mcanal@igalia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Raag Jadav <raag.jadav@intel.com>
Date: Fri Sep 11 17:45:47 2026 +0530
drm/xe/i2c: Disable IRQ on unbind
[ Upstream commit f0e9f963a3d209d7dc7ddd61116118ab5da2797d ]
Currently, struct xe_i2c is freed before SGUnit IRQ is disabled in unbind
path, leaving a potential UAF in case I2C IRQ is hit during this small
window. Explicitly disable I2C IRQ in xe_i2c_remove() and fix this.
Fixes: 0bb78ce09926 ("drm/xe/i2c: Wire up reset/postinstall for I2C IRQ")
Signed-off-by: Raag Jadav <raag.jadav@intel.com>
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Link: https://patch.msgid.link/20260911121547.2407261-1-raag.jadav@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
(cherry picked from commit 8ba5c8b8ab3fd362267c11df2cd5a90ee46f6e24)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ilia Levi <ilia.levi@intel.com>
Date: Tue Sep 8 17:50:48 2026 +0100
drm/xe/mmio_gem: forbid VMA split
[ Upstream commit 247a82da6f563dcfd9074a68f99a0c0997d0679c ]
The fault handler assumes it always operates on a VMA spanning the entire
GEM object. This does not hold when the VMA has been split, e.g. by a
partial munmap or mprotect. In that case the handler may map wrong
physical pages or cause SIGBUS.
Handle this by forbidding VMA split, as partial unmaps are not deemed
useful for MMIO GEMs.
Suggested-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Ilia Levi <ilia.levi@intel.com>
Fixes: 1ffcf8b8ae8a ("drm/xe: Support for mmap-ing mmio regions")
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20260908165046.1393557-11-matthew.auld@intel.com
(cherry picked from commit f3391a0b12d7bf826a0b21600d2f294f3dce4c14)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Shuicheng Lin <shuicheng.lin@intel.com>
Date: Tue Sep 8 17:50:51 2026 +0100
drm/xe/mmio_gem: Revoke drm_vma_node on xe_mmio_gem destroy
[ Upstream commit 37fcbd7b2f8996d783932dab11bc668e169b0de6 ]
xe_mmio_gem_create() calls drm_vma_node_allow() but nothing ever calls
drm_vma_node_revoke(). The drm_vma_offset_file rb-tree entry allocated
by drm_vma_node_allow() is not freed by drm_gem_object_release(), so
it is leaked on every create/destroy cycle.
Add a struct drm_file * parameter to xe_mmio_gem_destroy() and call
drm_vma_node_revoke() from there, mirroring the drm_vma_node_allow()
call in xe_mmio_gem_create().
Fixes: 1ffcf8b8ae8a ("drm/xe: Support for mmap-ing mmio regions")
Suggested-by: Ilia Levi <ilia.levi@intel.com>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Shuicheng Lin <shuicheng.lin@intel.com>
Reviewed-by: Ilia Levi <ilia.levi@intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20260908165046.1393557-14-matthew.auld@intel.com
(cherry picked from commit 32f0cb250598456d812fb7ca57a040282858323d)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ilia Levi <ilia.levi@intel.com>
Date: Tue Sep 8 17:50:49 2026 +0100
drm/xe/mmio_gem: use write-back mapping for dummy page
[ Upstream commit 819f189265a5955da743e78e20a713a038982e89 ]
Currently vmf_insert_pfn() maps the dummy page as UC, inheriting the
VMA's page protection which was set for the real MMIO region. This
conflicts with the direct map's WB mapping of the same page, creating a
cache type alias which is architecturally undefined on some platforms.
Use vmf_insert_pfn_prot() with a WB pgprot instead. Also simplify to
fault in the requested page instead of the whole VMA.
Fixes: 1ffcf8b8ae8a ("drm/xe: Support for mmap-ing mmio regions")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Closes: https://sashiko.dev/#/patchset/20260525125801.975038-6-ilia.levi%40intel.com
Assisted-by: GitHub-Copilot:claude-opus-4.6
Signed-off-by: Ilia Levi <ilia.levi@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20260908165046.1393557-12-matthew.auld@intel.com
(cherry picked from commit 1e8e28e35df0e77ae1b22fc091c1f422f62fa5e9)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Shuicheng Lin <shuicheng.lin@intel.com>
Date: Wed Sep 9 16:21:01 2026 +0000
drm/xe/shrinker: Return the freed page count through a parameter
[ Upstream commit 3c90e42a01426262f0cd166bc01b45c05562640d ]
__xe_shrinker_walk() and xe_shrinker_walk() return either the number of
pages freed or a negative error, so the two cannot be reported at once.
On error the pages already freed are dropped, and since xe_shrinker_scan()
only accumulates non-negative returns while *scanned is updated by
pointer, the shrinker tells mm that it scanned without freeing.
Accumulate the count into a caller-provided counter and return only the
status, so an error no longer discards what the walk had freed.
Fixes: 00c8efc3180f ("drm/xe: Add a shrinker for xe bos")
Assisted-by: Claude:claude-opus-5
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260909162102.1097006-2-shuicheng.lin@intel.com
Signed-off-by: Shuicheng Lin <shuicheng.lin@intel.com>
(cherry picked from commit d7aac1a0235a6ce41e30cec385e2db8c33dad12d)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Shuicheng Lin <shuicheng.lin@intel.com>
Date: Wed Sep 9 16:21:02 2026 +0000
drm/xe/shrinker: Take a runtime PM ref before shrinking non-system memory
[ Upstream commit 985862be16c7e4da808c51f393d631fb60c0be5c ]
__xe_shrinker_walk() walks the SYSTEM and TT LRUs without a runtime PM
reference. Shrinking a bo outside system memory invalidates its GPU
mappings, which needs the device resumed, so while it is runtime
suspended the page table zap trips an assert and the TLB invalidation
returns -ENODEV:
WARNING: drivers/gpu/drm/xe/xe_bo.c:770 at xe_bo_move_notify+0x1fc/0x450 [xe]
xe_bo_shrink+0x20f/0x2b0 [xe]
__xe_shrinker_walk+0x174/0x410 [xe]
xe_shrinker_scan+0x10c/0x1e0 [xe]
do_shrink_slab+0x176/0x7e0
drop_caches_sysctl_handler+0x9c/0xf0
Take a reference before walking a memory type other than XE_PL_SYSTEM
and stop there if it cannot be acquired. Reuse the shrinker's existing
acquire path, which resumes the device directly where reclaim allows
that and otherwise queues the PM worker for a later scan. Stop the walk
once the scan target is met, so a satisfied scan does not wake the
device. System memory is still reclaimed while the device is suspended.
Gate this on xe_device_is_l2_flush_optimized(), the same condition under
which xe_bo_trigger_rebind() issues the invalidation for a non-fault-mode
vm, so reclaim is unaffected elsewhere. The System CCS copy already has
its own reference in xe_bo_shrink().
Only a non-fault-mode vm can reach this, since a fault-mode vm requires
LR mode and that holds a runtime PM reference for the vm's lifetime.
Reproduced with igt@xe_madvise@dontneed-before-exec while the GPU is
runtime suspended.
v2: simplify needs_rpm check. (Matt)
retarget Fixes tag since the issue occurs with the non-fault-mode
path added by 4e7ebff69aed.
v3: handle this in xe_shrinker.c instead of xe_bo.c (Thomas)
v4: stop the walk once the scan target is met. (Sashiko)
v5: rebase on the freed page accounting fix. (Sashiko)
v6: reuse the shrinker acquire path so runtime pm can be resumed
directly instead of always queueing a worker. (Thomas)
v7: replace xe_pm_runtime_put() with xe_shrinker_runtime_pm_put(). (Thomas)
Fixes: 4e7ebff69aed ("drm/xe/xe3p_lpg: flush shrinker bo cachelines manually")
Assisted-by: Claude:claude-opus-5
Cc: Tejas Upadhyay <tejas.upadhyay@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://patch.msgid.link/20260909162102.1097006-3-shuicheng.lin@intel.com
Signed-off-by: Shuicheng Lin <shuicheng.lin@intel.com>
(cherry picked from commit 628f92b28bf4c371c10207daf6fc4caee0c0db2e)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Thadeu Lima de Souza Cascardo <cascardo@igalia.com>
Date: Wed Aug 26 08:46:58 2026 -0300
drm: Fix drm_pending_vblank_event leak in error path for out_fence_ptr
[ Upstream commit 9eb1a393c89a79c4210230d23e7d88d239c61d7b ]
When an out_fence_ptr is provided but DRM_MODE_PAGE_FLIP_EVENT is not
set, a drm_pending_vblank_event will be allocated. If later, there is an
allocation failure or another failure at setup_out_fence(), that event
will not have base.fence set and it will not be released at
complete_signaling().
Release the event and set crtc_state->event to NULL just like in the
DRM_MODE_PAGE_FLIP_EVENT case when there is a failure at
drm_event_reserve_init(). That is, prepare_signaling() releases the
event and there is nothing to be done at complete_signaling(). Use
drm_event_cancel_free() as that will also undo drm_event_reserve_init()
in case it has been called.
Reported-by: sashiko-bot@kernel.org
Closes: https://sashiko.dev/#/patchset/20260727-drm_crtc_atomic_commit_leak-v1-1-23d9948a9d7c@igalia.com?part=1
Fixes: 92c715fca907 ("drm/atomic: Fix double free in drm_atomic_state_default_clear")
Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@igalia.com>
Reviewed-by: Melissa Wen <mwen@igalia.com>
Signed-off-by: Melissa Wen <mwen@igalia.com>
Link: https://patch.msgid.link/20260826-drm_pending_vblank_event_leak-v4-1-f8de8b996b9d@igalia.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Thu Sep 10 20:46:12 2026 +0000
drop_monitor: fix out-of-bounds write in reset_per_cpu_data()
[ Upstream commit 439f392084f8f7f59ab9d47a9579185accefe1d8 ]
In reset_per_cpu_data(), al is computed as:
al = sizeof(struct net_dm_alert_msg);
al += dm_hit_limit * sizeof(struct net_dm_drop_point);
al += sizeof(struct nlattr);
skb = genlmsg_new(al, GFP_KERNEL);
...
nla = nla_reserve(skb, NLA_UNSPEC, sizeof(struct net_dm_alert_msg));
...
msg = nla_data(nla);
memset(msg, 0, al);
Because al includes sizeof(struct nlattr) (the 4-byte attribute header),
genlmsg_new() allocates al bytes of tailroom starting at nla.
However, msg points to nla_data(nla), which is located
sizeof(struct nlattr) bytes past nla. Calling memset(msg, 0, al)
therefore writes al bytes starting from msg, exceeding the allocated
buffer by sizeof(struct nlattr) (4 bytes) and corrupting
skb_shared_info.
Fix this by letting al represent only the payload length, allocating
the skb with genlmsg_new(nla_total_size(al), GFP_KERNEL), and zeroing
al bytes from msg.
Fixes: 683703a26e46 ("drop_monitor: Update netlink protocol to include netlink attribute header in alert message")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Link: https://patch.msgid.link/20260910204612.3762015-5-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Thu Sep 10 20:46:09 2026 +0000
drop_monitor: synchronize tracepoint unregistration on error path
[ Upstream commit 6a038ef2b57922b6d9ca98ddac0df0681849b704 ]
If register_trace_napi_poll() fails in net_dm_trace_on_set(),
unregister_trace_kfree_skb() is called to roll back the kfree_skb
tracepoint registration.
However, tracepoint_synchronize_unregister() is omitted before calling
cancel_work_sync() and module_put(). An in-flight probe executing
concurrently on another CPU could call schedule_work() after
cancel_work_sync() has already returned, leaving a pending work item
scheduled after the module reference is dropped. If the module is then
unloaded, executing the work item triggers a kernel panic.
Add tracepoint_synchronize_unregister() after unregister_trace_kfree_skb()
in the error path, matching net_dm_trace_off_set() and
net_dm_hw_probe_unregister().
Fixes: 7c747838a558 ("drop_monitor: Split tracing enable / disable to different functions")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Link: https://patch.msgid.link/20260910204612.3762015-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Thu Sep 10 20:46:11 2026 +0000
drop_monitor: use raw_cpu_ptr() in tracepoint probes
[ Upstream commit c19b7d35086b7d240f1ca3088b0079d2bd39ffb9 ]
syzbot reported a preemption warning in sk_skb_reason_drop():
BUG: using smp_processor_id() in preemptible [00000000] code: syz.0.17/5917
caller is net_dm_packet_trace_kfree_skb_hit+0x119/0x350 net/core/drop_monitor.c:519
In net_dm_packet_trace_kfree_skb_hit(), data = this_cpu_ptr(&dm_cpu_data)
is evaluated before spin_lock_irqsave(&data->drop_queue.lock, flags).
When kfree_skb() is called from preemptible context (e.g. process context
during close() on /dev/net/tun), preemption is enabled, triggering the
CONFIG_DEBUG_PREEMPT warning in smp_processor_id().
The same pattern exists in net_dm_hw_trap_summary_probe() and
net_dm_hw_trap_packet_probe() for dm_hw_cpu_data.
This is a false positive because each per-cpu structure is protected
by its own spinlock. If the task migrates to another CPU right after
reading the per-cpu pointer, the lock still safely synchronizes
access to that queue.
Use raw_cpu_ptr() instead of this_cpu_ptr() to silence
CONFIG_DEBUG_PREEMPT without disturbing interrupt state or breaking
PREEMPT_RT locking semantics.
Fixes: ca30707dee2b ("drop_monitor: Add packet alert mode")
Fixes: 5855357cd40e ("drop_monitor: Prepare probe functions for devlink tracepoint")
Reported-by: syzbot+dc57fd6722deb17e92af@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/netdev/6aa316b2.f81106d8.2ab401.0014.GAE@google.com/
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260910204612.3762015-4-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Thu Sep 10 20:46:10 2026 +0000
drop_monitor: use timer_shutdown_sync() to prevent timer rearming during teardown
[ Upstream commit c391a40f71886b28c082b47270f0e856fa3e1150 ]
In drop_monitor teardown paths (net_dm_trace_off_set(),
net_dm_hw_monitor_stop(), and error unwind paths in net_dm_trace_on_set()
and net_dm_hw_monitor_start()), per-CPU timers are stopped using
timer_delete_sync() followed by cancel_work_sync().
However, there is a circular dependency between send_timer and
dm_alert_work:
1) sched_send_work() (timer callback) schedules dm_alert_work.
2) send_dm_alert() / net_dm_hw_summary_work() calls reset_per_cpu_data()
or net_dm_hw_reset_per_cpu_data().
3) If memory allocation fails under memory pressure in the reset
function, it re-arms the timer via mod_timer(&data->send_timer, ...).
If dm_alert_work is running concurrently while timer_delete_sync()
executes on another CPU, an allocation failure in the worker will
re-arm the timer after timer_delete_sync() has already returned.
Once cancel_work_sync() completes and module_put() is called, the timer
remains active in the timer wheel. If the module is then unloaded, the
timer will fire and execute sched_send_work() in freed memory,
triggering a kernel panic / use-after-free.
Switch from timer_delete_sync() to timer_shutdown_sync(). This guarantees
that any in-flight timer handler has finished and prevents subsequent
re-arming attempts from running workers from succeeding. When monitoring
is restarted later, timer_setup() is invoked, which cleanly
re-initializes the timer.
Fixes: 9398e9c0b1d4 ("drop_monitor: Perform cleanup upon probe registration failure")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260910204612.3762015-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Maher Azzouzi <maherazz04@gmail.com>
Date: Mon Aug 17 14:37:52 2026 +0100
esp: downgrade zerocopy managed frags before mutating skb frags
[ Upstream commit f89416eb3db151170a6f3c6dfc5239d26cdce4d2 ]
On the out-of-place output path (esp->inplace == false) ESP rewrites the
skb frag array: esp_output_head() appends a trailer frag and
esp_output_tail() replaces the frags with a destination page, both
referenced with get_page().
When the skb carries zerocopy managed frags (SKBFL_MANAGED_FRAG_REFS) the
payload frags are owned by the ubuf and must not be referenced or
unreferenced individually, but ESP mutates the frag array without ever
downgrading the skb. This breaks the managed-frag invariant two ways:
- esp_ssg_unref() walks the source scatterlist and drops a page
reference for every frag, including the ubuf-owned payload frags,
pushing their refcount below the GUP pin bias while the pages are
still pinned, i.e. a use-after-free of the zerocopy pages;
- esp_output_tail() installs its destination page as frag 0 with
get_page() but leaves SKBFL_MANAGED_FRAG_REFS set, so
skb_release_data() takes the skip_unref branch and never drops that
reference, leaking the x->xfrag page at packet rate.
Fix this the way every other frag-mutating site does (__ip_append_data(),
__ip6_append_data(), tcp_sendmsg_locked()) and call
skb_zcopy_downgrade_managed() before ESP touches the frag array: it takes
a real reference on each existing frag and clears SKBFL_MANAGED_FRAG_REFS,
so the per-frag unref in esp_ssg_unref() and the frag release in
skb_release_data() are both balanced and no mixed-ownership frag array is
left behind.
Fixes: 753f1ca4e1e5 ("net: introduce managed frags infrastructure")
Signed-off-by: Maher Azzouzi <maherazz04@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jakub Kicinski <kuba@kernel.org>
Date: Mon Sep 14 19:23:27 2026 -0700
eth: fbnic: ring the doorbell if a burst ends in a drop
[ Upstream commit 490599ab23134962a6d18a024e84541d77bdb999 ]
fbnic_tx_map() skips the doorbell write, and the completion request,
for every packet handed to it with xmit_more set, counting on the
packet which ends the burst to publish them all. When that packet is
dropped instead - skb_put_padto(), skb_cow_head() or a DMA mapping
failure - nothing rings. The descriptors of the preceding packets stay
invisible to the HW until the next transmit on that queue, which for a
burst-then-idle workload may never come.
Remember the meta descriptor of the last packet left without a doorbell
and flush it from the error paths. The completion request has to be set
on that descriptor rather than simply writing the tail, otherwise the HW
would transmit the packets but never report a head, and the ring would
fill up and stall for good.
This is very similar to Joe's recent series of fixes for bnxt.
Not seen in real life, reproduced under QEMU with failure injection.
Fixes: 9a57bacd574b ("eth: fbnic: Add basic Tx handling")
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260915022327.913218-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Hyunwoo Kim <imv4bel@gmail.com>
Date: Fri Sep 11 11:09:17 2026 +0200
exec: Cleanup POSIX timers right after de_thread()
commit acb03d3881818581052924a9bbbe92b8741ed448 upstream.
A per-thread CPU timer holds a reference to the PID of the thread it is
attached to and, while it is armed, its node is queued in that thread's
posix_cputimers. The task is looked up by that PID.
When a non-leader thread exec()s, de_thread() changes which task owns
that PID. pid_task(timer->it.cpu.pid, PIDTYPE_PID) then returns NULL,
but the node is still queued on tsk, which is alive. timer_lock_sighand()
takes a failed lookup to mean that the node is already dequeued, so it
has nothing to undo.
begin_new_exec() calls posix_cpu_timers_exit(me) right after
exec_task_namespaces() and that removes the leftover node, so the state
normally stays invisible. But bprm->point_of_no_return is set before
de_thread(), so if unshare_files(), set_mm_exe_file(), exec_mmap() or
exec_task_namespaces() fails, the task dies before it gets there.
exit_itimers() then frees the k_itimer while its node is still queued,
and reaping tsk later erases that freed node from the rbtree.
In short:
the non-leader thread B the parent
timer_create(CLOCK_THREAD_CPUTIME_ID)
timer_settime()
arm_timer() // the node is queued on B
execve()
de_thread(B)
exchange_tids(B, leader) // B's PID now belongs to the leader
release_task(leader)
__exit_signal(leader)
posix_cpu_timers_exit(leader) // cleans leader's queue, not B's
__unhash_process(leader) // that PID has no task anymore
exec_mmap()
mmap_read_lock_killable(old_mm)
kill(B, SIGKILL)
// -EINTR
get_signal()
do_exit()
exit_itimers()
posix_timer_delete()
posix_cpu_timer_del()
posix_timer_unhash_and_free() // freed while still queued
wait4()
release_task(B)
posix_cpu_timers_exit(B)
cleanup_timerqueue()
timerqueue_del() // use-after-free
Move the POSIX timer cleanup right after de_thread() before any of the
later failure conditions brings the task into do_exit().
[ tglx: Move the cleanup right after de_thread() ]
Fixes: 55e8c8eb2c7b ("posix-cpu-timers: Store a reference to a pid not a task")
Signed-off-by: Hyunwoo Kim <imv4bel@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Kijo Park <red993688@gmail.com>
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/ao7Q8miiuLAPVnWv@v4bel
Link: https://patch.msgid.link/20260911090541.627712075@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Xixin Liu <liuxixin@kylinos.cn>
Date: Tue Jul 28 08:50:00 2026 +0800
firmware: arm_scpi: reject DVFS OPP count above MAX_DVFS_OPPS
[ Upstream commit 32471d84a487c7fd74532bc96be56f8028cf4a3f ]
scpi_dvfs_get_info() already rejected a zero opp_count, but still trusted
any larger value from the SCP firmware. The shared-memory reply only holds
MAX_DVFS_OPPS entries in buf.opps[]; a bigger count over-reads that array
and then sizes the allocated OPP table incorrectly (garbage OPPs / OOB).
The missing upper bound dates back to the original SCPI DVFS support.
Reject zero and out-of-range counts in one check and return -EINVAL.
Fixes: 8cb7cf56c9fe ("firmware: add support for ARM System Control and Power Interface(SCPI) protocol")
Signed-off-by: Xixin Liu <liuxixin@kylinos.cn>
Link: https://patch.msgid.link/022802f0b38f.v2.1785200642.git.liuxixin@kylinos.cn
Signed-off-by: Sudeep Holla <sudeep.holla@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Seunguk Shin <seunguk.shin@arm.com>
Date: Mon Aug 3 13:34:55 2026 +0100
fs/dax: check zero or empty entry before converting xarray entry
commit 8e2b8614039853e68d5338e37821e8bcee9fc05f upstream.
Calling dax_to_folio() with empty entry causes kernel panic below when
booting a VM with DAX enabled storage.
This patch checks empty entry before calling dax_to_folio() on
dax_associate_entry(), dax_disassociate_entry(), and dax_busy_page().
Commit 98c183a4fccf ("fs/dax: don't disassociate zero page entries") added
guards in the associate and disassociate paths, but the guards still come
after dax_to_folio(), and dax_busy_page() still has the same problem.
[ 0.737679] EXT4-fs (pmem0p1): mounted filesystem 79676804-7c8b-491a-b2a6-9bae3c72af70 ro with ordered data mode. Quota mode: disabled.
[ 0.737891] VFS: Mounted root (ext4 filesystem) readonly on device 259:1.
[ 0.739119] devtmpfs: mounted
[ 0.739476] Freeing unused kernel memory: 1920K
[ 0.740156] Run /sbin/init as init process
[ 0.740229] with arguments:
[ 0.740286] /sbin/init
[ 0.740321] with environment:
[ 0.740369] HOME=/
[ 0.740400] TERM=linux
[ 0.743162] Unable to handle kernel paging request at virtual address fffffdffbf000008
[ 0.743285] Mem abort info:
[ 0.743316] ESR = 0x0000000096000006
[ 0.743371] EC = 0x25: DABT (current EL), IL = 32 bits
[ 0.743444] SET = 0, FnV = 0
[ 0.743489] EA = 0, S1PTW = 0
[ 0.743545] FSC = 0x06: level 2 translation fault
[ 0.743610] Data abort info:
[ 0.743656] ISV = 0, ISS = 0x00000006, ISS2 = 0x00000000
[ 0.743720] CM = 0, WnR = 0, TnD = 0, TagAccess = 0
[ 0.743785] GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0
[ 0.743848] swapper pgtable: 4k pages, 48-bit VAs, pgdp=00000000b9d17000
[ 0.743931] [fffffdffbf000008] pgd=10000000bfa3d403, p4d=10000000bfa3d403, pud=1000000040bfe403, pmd=0000000000000000
[ 0.744070] Internal error: Oops: 0000000096000006 [#1] SMP
[ 0.748888] CPU: 0 UID: 0 PID: 1 Comm: init Not tainted 6.18.4 #1 NONE
[ 0.749421] pstate: 004000c5 (nzcv daIF +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[ 0.749969] pc : dax_disassociate_entry.constprop.0+0x20/0x50
[ 0.750444] lr : dax_insert_entry+0xcc/0x408
[ 0.750802] sp : ffff80008000b9e0
[ 0.751083] x29: ffff80008000b9e0 x28: 0000000000000000 x27: 0000000000000000
[ 0.751682] x26: 0000000001963d01 x25: ffff0000004f7d90 x24: 0000000000000000
[ 0.752264] x23: 0000000000000000 x22: ffff80008000bcc8 x21: 0000000000000011
[ 0.752836] x20: ffff80008000ba90 x19: 0000000001963d01 x18: 0000000000000000
[ 0.753407] x17: 0000000000000000 x16: 0000000000000000 x15: 0000000000000000
[ 0.753970] x14: ffffbf3154b9ae70 x13: 0000000000000000 x12: ffffbf3154b9ae70
[ 0.754548] x11: ffffffffffffffff x10: 0000000000000000 x9 : 0000000000000000
[ 0.755122] x8 : 000000000000000d x7 : 000000000000001f x6 : 0000000000000000
[ 0.755707] x5 : 0000000000000000 x4 : 0000000000000000 x3 : fffffdffc0000000
[ 0.756287] x2 : 0000000000000008 x1 : 0000000040000000 x0 : fffffdffbf000000
[ 0.756871] Call trace:
[ 0.757107] dax_disassociate_entry.constprop.0+0x20/0x50 (P)
[ 0.757592] dax_iomap_pte_fault+0x4fc/0x808
[ 0.757951] dax_iomap_fault+0x28/0x30
[ 0.758258] ext4_dax_huge_fault+0x80/0x2dc
[ 0.758594] ext4_dax_fault+0x10/0x3c
[ 0.758892] __do_fault+0x38/0x12c
[ 0.759175] __handle_mm_fault+0x530/0xcf0
[ 0.759518] handle_mm_fault+0xe4/0x230
[ 0.759833] do_page_fault+0x17c/0x4dc
[ 0.760144] do_translation_fault+0x30/0x38
[ 0.760483] do_mem_abort+0x40/0x8c
[ 0.760771] el0_ia+0x4c/0x170
[ 0.761032] el0t_64_sync_handler+0xd8/0xdc
[ 0.761371] el0t_64_sync+0x168/0x16c
[ 0.761677] Code: f9453021 f2dfbfe3 cb813080 8b001860 (f9400401)
[ 0.762168] ---[ end trace 0000000000000000 ]---
[ 0.762550] note: init[1] exited with irqs disabled
[ 0.762631] Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b
Link: https://lore.kernel.org/m2y0enxtzk.fsf@arm.com
Fixes: 38607c62b34b ("fs/dax: properly refcount fs dax pages")
Signed-off-by: Seunguk Shin <seunguk.shin@arm.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Reported-by: Kiara Grouwstra <cinereal@riseup.net>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Usama Arif <usama.arif@linux.dev>
Date: Tue Sep 22 08:19:24 2026 -0400
fs/super: skip non-memcg-aware nr_cached_objects in memcg slab shrink
[ Upstream commit 0baad6f9b9970c6e3f1d33dbfd17d1a77702771d ]
The super_block shrinker is registered with SHRINKER_MEMCG_AWARE because its
dentry and inode LRUs are memcg-aware (via list_lru). But the optional
->nr_cached_objects() hooks that the shrinker also drives are not memcg-aware:
btrfs extent maps and xfs inode reclaim operate on filesystem-global
state, and shmem's unused-huge shrinker walks a per-superblock shrinklist.
None of them filter by sc->memcg.
The mismatch shows up under memcg-heavy slab reclaim. shrink_slab_memcg()
calls do_shrink_slab() once per (memcg, NUMA node) pair for every memcg
whose bit is set in the per-superblock shrinker bitmap, which on a busy
host means hundreds of calls per reclaim pass. Each scan queues the same
global shrinker work item that's already kicked from the root path.
Because btrfs/xfs global count is typically non-zero on any in-use filesystem,
the returned total stays positive even if a memcg's own dentry/inode LRUs
are empty. shrink_slab_memcg() therefore never clears the SB shrinker bit
in the memcg bitmap, so subsequent reclaim passes from the same memcg
re-enter super_cache_count() and pay for the global counter walk again.
Restrict ->nr_cached_objects() to the global shrink path (sc->memcg NULL
or root). The memcg-aware dentry/inode LRUs keep being counted and
scanned per memcg as before; only the global fs-specific hooks are skipped.
The root/global shrink path still drives those hooks; only their
invocation from non-root memcg slab reclaim is removed.
Signed-off-by: Usama Arif <usama.arif@linux.dev>
Link: https://patch.msgid.link/20260609123047.1948242-1-usama.arif@linux.dev
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
Stable-dep-of: 641aade99f06 ("fs: fix missed removal of super_fs_objects_eligible()")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Qi Zheng <qi.zheng@linux.dev>
Date: Tue Sep 22 08:19:25 2026 -0400
fs: fix missed removal of super_fs_objects_eligible()
[ Upstream commit 641aade99f06df0037e52b5c81645461b7132947 ]
Commit 0ef8faff490be ("fs: push nr_cached_objects memcg gating into
individual filesystems") was meant to drop the blanket memcg gate in
fs/super.c and let each ->nr_cached_objects() implementation decide for
itself whether it is meaningful in per-memcg reclaim. However, when
that patch was applied the removal of super_fs_objects_eligible() and
its two call sites in super_cache_scan() / super_cache_count() was
lost, so the helper is still gating every ->nr_cached_objects() hook
and 0ef8faff490be is effectively a no-op.
Consequences of the leftover gate:
- XFS's inode-reclaim hook, which is intentionally driven from
per-memcg contexts to free memcg-charged slab, is still
short-circuited in fs/super.c exactly the regression from
commit 0baad6f9b997 ("fs/super: skip non-memcg-aware
nr_cached_objects in memcg slab shrink") that 0ef8faff490be was
written to undo. Memcg-charged XFS inode slab therefore keeps
piling up under per-memcg pressure until global reclaim kicks in.
- Any future ->nr_cached_objects()/->free_cached_objects() that
grows memcg awareness is likewise blocked before it can run, so
filesystems cannot opt in to per-memcg reclaim on their own
defeating the whole point of pushing the gating decision down
into the callbacks.
Drop the leftover helper and its call sites so the intent of
0ef8faff490be actually takes effect.
Link: https://lore.kernel.org/cover.1786955972.git.zhengqi.arch@bytedance.com
Link: https://lore.kernel.org/3b038d373c70ebac7cdabfb0035bb91d1d6e6cfe.1786955972.git.zhengqi.arch@bytedance.com
Link: https://lore.kernel.org/all/20260715103516.2410175-1-usama.arif@linux.dev/ [0]
Fixes: 0ef8faff490b ("fs: push nr_cached_objects memcg gating into individual filesystems")
Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
Acked-by: Usama Arif <usama.arif@linux.dev>
Cc: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: Christian Brauner <brauner@kernel.org>
Cc: David Hildenbrand <david@kernel.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeel.butt@linux.dev>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Peter Zijlstra <peterz@infradead.org>
Date: Fri Sep 11 11:04:47 2026 +0200
futex: Also allocate private hash on vfork()
[ Upstream commit b61b6f95d6722ddbbbd09e689fa41b55fd36f9a5 ]
As Jann demonstrated, it is entirely feasible to access the mm through vfork().
Therefore we need to allocate a private hash on vfork() as well as any other
CLONE_VM user.
Specifically, it must be avoided to have (private) futex waiters before
allocating the private hash.
Fixes: ee9dce44362b ("futex: Drop CLONE_THREAD requirement for private default hash alloc")
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://patch.msgid.link/20260911090447.GT788244@noisy.programming.kicks-ass.net
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Runyu Xiao <runyu.xiao@seu.edu.cn>
Date: Mon Sep 14 13:15:37 2026 +0800
gpio: virtuser: skip free_irq when no IRQ is installed
[ Upstream commit 50fd0ada8d37587223001600933270b59cb30e19 ]
Disabling interrupt monitoring uses atomic_xchg() to clear the stored IRQ.
When monitoring is already disabled, atomic_xchg() returns 0. It must not
be passed to free_irq().
The bug is reproducible on an x86_64 QEMU guest with
CONFIG_GPIO_VIRTUSER=y and CONFIG_GPIO_SIM=y. Configure a live
gpio-virtuser device through configfs. Its input lookup must refer to a
live gpio-sim bank, such as key gpio-sim-test with offset 0. The
consumer's dev_name attribute is shown as <dev> below; then run:
echo 0 > /sys/kernel/debug/gpio-virtuser/<dev>/gpiod:input:0/interrupts
On an unpatched kernel, this reaches gpio_virtuser_interrupts_set() with
ld->irq still at its initial value 0, and free_irq() reports:
Trying to free already-free IRQ 0
The same reproducer completes without the warning on the patched kernel.
Fixes: 91581c4b3f29 ("gpio: virtuser: new virtual testing driver for the GPIO API")
Assisted-by: LLM
Signed-off-by: Runyu Xiao <runyu.xiao@seu.edu.cn>
Reviewed-by: Linus Walleij <linusw@kernel.org>
Link: https://patch.msgid.link/20260914051537.15320-1-runyu.xiao@seu.edu.cn
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Abdurrahman Hussain <abdurrahman@nexthop.ai>
Date: Sat Aug 15 13:44:19 2026 -0700
gpiolib: of: don't mark hog nodes OF_POPULATED before a chip is found
commit 1f1d0812f6a8ab8e6f709c599f137c99646512cc upstream.
When a gpio-hog node is attached by a device-tree overlay before its
parent GPIO chip has been registered, of_gpio_notify() sets
OF_POPULATED on the node via of_node_test_and_set_flag() and only then
discovers that there is no gpio_device for the parent, returning
NOTIFY_DONE without clearing the flag.
Since gpiochip_hog_lines() skips any hog child whose of_node carries
OF_POPULATED, the leaked flag makes the hog silently ignored when the
chip is registered later. Applying an overlay containing both a GPIO
controller node and its hog children - and populating devices only
after the overlay apply completes - hits this on every boot; the hog
is only applied if the chip driver is unbound (which clears the flag
in the remove path) and rebound.
Look up the parent gpio_device before claiming the node so that a hog
attached ahead of its chip stays unclaimed and is picked up normally
by gpiochip_hog_lines() at registration time.
Signed-off-by: Abdurrahman Hussain <abdurrahman@nexthop.ai>
Fixes: a23226b7c1f6 ("gpiolib: handle gpio-hogs only once")
Cc: stable@vger.kernel.org
Reviewed-by: Daniel Drake <dan@reactivated.net>
Link: https://patch.msgid.link/20260815-gpiolib-of-hog-flag-leak-v1-1-6126aac5f6f3@nexthop.ai
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Michail Tatas <michail.tatas@gmail.com>
Date: Thu Aug 13 20:36:50 2026 +0300
gpiolib: Put fwnode reference on failure
commit 16b10f64c63f78220c3b4035f1ed6cd3bdcb0b02 upstream.
We get a reference to the fwnode handle which we pass to
gpio_shared_make_ref. In case it fails we do not put the reference.
Fix by putting the reference in the failure case
Fixes: 49416483a953 ("gpio: shared: allow sharing a reset-gpios pin between reset-gpio and gpiolib")
Cc: stable@vger.kernel.org
Signed-off-by: Michail Tatas <michail.tatas@gmail.com>
Link: https://patch.msgid.link/an4Asr4tx3D2QvLD@michalis-linux
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Thomas Richard (congatec GmbH) <thomas.richard@bootlin.com>
Date: Fri Sep 11 19:31:59 2026 +0200
hwmon: (cgbc-hwmon) Add missing sensors
commit 3550d1dbbcb9f51b77e077e8958423ee2c401c6c upstream.
Add the following sensors:
- Alternate Board Temperature (temp11_input)
- Top DIMM 1-7 Temperature (temp12_input to temp18_input)
- Bottom DIMM 1 Temperature (temp19_input)
- 12V Standby Voltage (in14_input)
This fixes the following warning on conga-SA7:
Board Controller returned an unknown sensor (bc_type=1, bc_id=11), ignore it
Also update existing labels to match Congatec documentation.
Cc: stable@kernel.org
Fixes: 08ebc9def79f ("hwmon: Add Congatec Board Controller monitoring driver")
Signed-off-by: Thomas Richard (congatec GmbH) <thomas.richard@bootlin.com>
Link: https://patch.msgid.link/20260911-cgbc-hwmon-fix-and-new-sensors-v2-2-0c6bf078d173@bootlin.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Thomas Richard (congatec GmbH) <thomas.richard@bootlin.com>
Date: Fri Sep 11 19:31:58 2026 +0200
hwmon: (cgbc-hwmon) Fix current sensors ID lookup
commit 7bae83ffb133bc373d098fa6828cef2ef4da49fe upstream.
Current sensors on the Congatec Board Controller don't use consecutive IDs,
unlike other sensor types (voltage, temperature, fan). The driver assumed
consecutive IDs and performed a simple lookup, which caused an unknown
sensor warning. Define current sensor IDs explicitly.
Changes the warning on conga-SA7 (type and channel are correct now).
Before:
Board Controller returned an unknown sensor (type=2, channel=17), ignore it
After:
Board Controller returned an unknown sensor (bc_type=1, bc_id=11), ignore it
Cc: stable@kernel.org
Fixes: 08ebc9def79f ("hwmon: Add Congatec Board Controller monitoring driver")
Signed-off-by: Thomas Richard (congatec GmbH) <thomas.richard@bootlin.com>
Link: https://patch.msgid.link/20260911-cgbc-hwmon-fix-and-new-sensors-v2-1-0c6bf078d173@bootlin.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Cong Nguyen <congnt264@gmail.com>
Date: Mon Sep 14 17:41:36 2026 +0700
hwmon: (gpio-fan) return IRQ_HANDLED from the shared alarm IRQ handler
[ Upstream commit bdf5f731957de48acada392f28e82bc019713adb ]
fan_alarm_irq_handler() always schedules alarm_work but returns IRQ_NONE,
so the kernel treats every alarm interrupt as unhandled. On a shared
line that risks the whole line being disabled as spurious.
v1 just fixed that, but it was still IRQF_SHARED, and always returning
IRQ_HANDLED there defeats spurious-interrupt detection for the line --
if the interrupt ever fires without a real event, nothing catches it,
and a fault could spin the CPU in the handler.
Sashiko flagged this in v1, and Guenter confirmed: this interrupt must
not be shared. So v2 drops IRQF_SHARED too.
Fixes: d6fe1360f42e ("hwmon: add generic GPIO fan driver")
Reported-by: Sashiko AI review <sashiko-bot@kernel.org>
Link: https://lore.kernel.org/r/20260901160931.DD3811F00A3D@smtp.kernel.org
Assisted-by: Claude:claude-opus-4
Signed-off-by: Cong Nguyen <congnt264@gmail.com>
Link: https://patch.msgid.link/20260914104136.1797979-1-congnt264@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Muhammad Bilal <meatuni001@gmail.com>
Date: Wed Sep 16 05:29:26 2026 +0500
hwmon: (hp-wmi-sensors) Fix use-after-free in fungible_show()
commit e6cb0b4d4ecb8e71fd2200d907ab2e9663356f69 upstream.
nsensor->current_state is dynamically replaced as the sensor's state
changes. update_numeric_sensor_from_wobj() does this by freeing the
old string and installing a new one:
if (strcmp(trimmed, nsensor->current_state)) {
new_string = hp_wmi_strdup(dev, trimmed);
if (new_string) {
devm_kfree(dev, nsensor->current_state);
nsensor->current_state = new_string;
}
}
This function is only ever called from hp_wmi_update_info() while
state->lock is held, so the free-and-replace itself is properly
serialized against concurrent updates.
fungible_show(), however, reads the same pointer after the lock has
already been dropped:
err = hp_wmi_update_info(state, info);
if (err)
return err;
switch (prop) {
...
case HP_WMI_PROPERTY_CURRENT_STATE:
seq_printf(seqf, "%s\n", nsensor->current_state);
break;
hp_wmi_update_info() takes state->lock internally and releases it
before returning, so by the time fungible_show() dereferences
nsensor->current_state in seq_printf(), no lock is held. Two
processes reading a sensor's current_state debugfs entry at
overlapping times (or one reading it while another read of the same
sensor triggers a refresh) can race: one thread's seq_printf() can
be part-way through printing the string at the moment another
thread's call into update_numeric_sensor_from_wobj() frees it with
devm_kfree() and installs a new pointer, causing a use-after-free
read.
Take state->lock around the read in fungible_show() as well, so it
can never run concurrently with the free-and-replace in
update_numeric_sensor_from_wobj().
Fixes: 23902f98f8d4 ("hwmon: add HP WMI Sensors driver")
Cc: stable@vger.kernel.org
Signed-off-by: Muhammad Bilal <meatuni001@gmail.com>
Acked-by: James Seo <james@equiv.tech>
Link: https://patch.msgid.link/20260916002926.161595-1-meatuni001@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: James Seo <james@equiv.tech>
Date: Wed Sep 16 15:19:15 2026 -0700
hwmon: (hp-wmi-sensors) Improve raw WMI string handling
[ Upstream commit 92b68492eae701e5b0e9d142ffe229921af7b1fa ]
Commit c9ba59258094 ("hwmon: (hp-wmi-sensors) Fix failure to load on
EliteDesk 800 G6") left out some logic for recognizing raw WMI
strings in check_numeric_sensor_wobj(). This issue was reported by a
user along with an incomplete and unsuitable proposed solution [1].
Add the missing logic and properly remedy the issue. Also slightly
refactor how raw WMI strings are recognized elsewhere to make the
intent that they should be treated as regular ACPI strings clearer.
Reported-by: Muhammad Bilal <meatuni001@gmail.com>
Link: https://lore.kernel.org/linux-hwmon/20260916002907.161210-1-meatuni001@gmail.com/ [1]
Fixes: c9ba59258094 ("hwmon: (hp-wmi-sensors) Fix failure to load on EliteDesk 800 G6")
Signed-off-by: James Seo <james@equiv.tech>
Link: https://patch.msgid.link/20260916221912.434119-5-james@equiv.tech
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Nuno Sá <nuno.sa@analog.com>
Date: Fri Sep 11 14:53:37 2026 +0100
hwmon: (pmbus/core) increase number of phases and add new mask
commit 06bd6794b5fd2163880ac3bfe973d4cc61f359f3 upstream.
Increase the number of phases to 16 as a new upcoming device supports
such a number.
While at it, add a new mask for controlling the source of the output
voltage.
Note (groeck):
This patch was meant to prepare for support of MAX20826 and compatible
devices, which support more than 10 phases per page. However, Sashiko
reports that the mp2975 driver already supports up to 14 phases, and the
mp2856 driver supports up to 12 phases. This already has the potential for
out-of-bounds writes when probing the affected chips, making this patch a
bug fix.
Fixes: 2c6fcbb21149 ("hwmon: (pmbus) Add support for MPS Multi-phase mp2975 controller")
Fixes: f9e5f289b686 ("hwmon: (pmbus) Add support for MPS Multi-phase mp2856/mp2857 controller")
Signed-off-by: Nuno Sá <nuno.sa@analog.com>
Link: https://patch.msgid.link/20260911-hwmon-max20826-support-v2-1-5e30cbd97d84@analog.com
Cc: stable@vger.kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Sanman Pradhan <psanman@juniper.net>
Date: Tue Sep 15 16:48:35 2026 +0000
hwmon: (pmbus/tps53679) Fix TPS53676 phase page decoding
commit 1d12fb94ac0975566545871dda100df34df5f845 upstream.
tps53676_identify() reads the USER_DATA_03 phase configuration to count
the phases assigned to each channel and derive the number of PMBus pages.
In each 16-bit phase descriptor the channel (PAGE) is encoded in bit 4 and
the firing order in bits 3:0, but the code tested bit 3 (0x08), which is
part of the firing-order field.
TPS53676 supports up to seven phases, so firing-order bit 3 is never set.
As a result the existing test classifies every enabled phase as channel A.
On a dual-channel configuration the phases assigned to channel B are
therefore miscounted as channel A and page 1 is not exposed.
Test the PAGE field (bit 4) instead.
Fixes: cb3d37b59012 ("hwmon: (pmbus/tps53679) Add support for TI TPS53676")
Cc: stable@vger.kernel.org
Signed-off-by: Sanman Pradhan <psanman@juniper.net>
Link: https://patch.msgid.link/20260915164823.160977-2-sanman.pradhan@hpe.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Sanman Pradhan <psanman@juniper.net>
Date: Wed Sep 16 23:54:17 2026 +0000
hwmon: (pmbus/tps53679) Select page 0 for single-page TPS53676
commit 089070b51ccbac411462a30a454690274c6e4270 upstream.
tps53676_identify() derives the number of PMBus pages but does not
ensure that page 0 is selected for single-page configurations.
pmbus_set_page() does not update the PAGE register when info->pages is
1, so if boot firmware leaves PAGE set to another value subsequent
register accesses may target the wrong page.
For single-page devices, select page 0 explicitly.
Fixes: cb3d37b59012 ("hwmon: (pmbus/tps53679) Add support for TI TPS53676")
Cc: stable@vger.kernel.org
Signed-off-by: Sanman Pradhan <psanman@juniper.net>
Link: https://patch.msgid.link/20260916235406.681131-2-sanman.pradhan@hpe.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Yibo Tan <lhfff@tju.edu.cn>
Date: Fri Sep 11 15:18:09 2026 +0800
hwmon: (pwm-fan) Stop RPM timer before freeing tach data
commit 26d5ff79768548efb1e604bb6e8697c101e06269 upstream.
sample_timer() rearms the RPM timer and accesses the devm-managed
ctx->tachs and ctx->pulses_per_revolution arrays. The cleanup action
which stops the timer is registered before those arrays are allocated.
Since devres releases entries in reverse order, driver detach can free
the arrays before pwm_fan_cleanup() shuts down the timer. A timer expiry
in that window accesses the freed tach data.
With a KASAN kernel, a test-only kprobe delayed entry to
pwm_fan_cleanup() while normal sysfs unbind ran. Each of three runs
reported three four-byte reads and two four-byte writes in sample_timer()
after its backing devm allocations had been freed. The helper did not
invoke the timer callback, cleanup actions or free functions.
With the fix, three matching unbind runs completed without KASAN, BUG,
WARNING, Oops or panic. Instrumentation confirmed that timer retirement
completed before the first timer backing allocation was released.
Split timer retirement from the power cleanup and register its devres
action after the timer backing data and IRQ actions are installed. This
preserves the early power rollback action while ensuring the timer is
retired before its backing data is released. Use timer_shutdown_sync()
because the callback can rearm itself.
Fixes: 01695410d452 ("hwmon: (pwm-fan) Store tach data separately")
Cc: stable@vger.kernel.org
Assisted-by: Codex:GPT-5
Signed-off-by: Yibo Tan <lhfff@tju.edu.cn>
Link: https://patch.msgid.link/20260911071809.130151-1-lhfff@tju.edu.cn
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Guangshuo Li <lgs201920130244@gmail.com>
Date: Mon Sep 14 14:28:09 2026 +0800
hwmon: (w83791d) remove fan/pwm 4-5 sysfs group on remove
commit 0ff9c7775e51ac6d47b1bb5c46f06b1434fe58a8 upstream.
When the fan/pwm 4-5 pins are not used as GPIO, w83791d_probe()
creates the w83791d_group_fanpwm45 sysfs group on the I2C client
device.
The probe error path removes this group when a later initialization
step fails, but the normal remove path only removes w83791d_group.
As a result, the optional fan/pwm 4-5 sysfs files can remain after the
driver is unbound.
The callbacks associated with these files access the driver data,
which is devm allocated and released after driver unbind. Leaving the
sysfs files behind can therefore result in accesses to stale driver
data.
Remove w83791d_group_fanpwm45 during normal teardown as well.
This issue was found by manual code inspection.
Fixes: 6e1ecd9b8f13 ("hwmon: (w83791d) fan 4/5 pins can also be used for gpio")
Cc: stable@vger.kernel.org
Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com>
Link: https://patch.msgid.link/20260914062809.1650538-1-lgs201920130244@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Guangshuo Li <lgs201920130244@gmail.com>
Date: Mon Sep 14 15:36:38 2026 +0800
hwmon: (w83793) release probe data through kref
commit c702a5f18b780e477eccbbab558e590e9673e4cb upstream.
w83793_probe() initializes data->kref to manage the lifetime of the
driver data. The normal remove path drops the driver-owned reference
with kref_put(), while watchdog users take and release additional
references through the same kref.
However, the probe error path still frees data directly with kfree().
This bypasses the kref-managed lifetime and discards the initial
reference without a matching kref_put(), leaving the reference
accounting unbalanced.
Drop the probe-owned reference with kref_put() instead and let
w83793_release_resources() perform the final free, matching the normal
remove path.
This issue was found by manual code inspection.
Fixes: 5852f9609d21 ("hwmon: (w83793) Add watchdog functionality")
Cc: stable@vger.kernel.org
Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com>
Link: https://patch.msgid.link/20260914073638.1662500-1-lgs201920130244@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shengzhuo Wei <me@cherr.cc>
Date: Thu Aug 27 23:43:01 2026 +0800
i2c: at91: release DMA channels on remove and probe error
commit f7eeb1af8537b05953fb1c88ab8b59d94059a381 upstream.
at91_twi_configure_dma() requests exclusive tx/rx DMA channels, but
nothing ever releases them on driver detach, and the probe error path
after the channels are acquired (i2c_add_numbered_adapter() failure)
returns without releasing them either, because the remove callback is
not invoked after a failed probe.
Move the release into a helper, call it from the existing
configure-failure path, the adapter-registration failure path, and
at91_twi_remove().
Fixes: 60937b2cdbf9 ("i2c: at91: add dma support")
Assisted-by: GLM:5.3
Signed-off-by: Shengzhuo Wei <me@cherr.cc>
Cc: <stable@vger.kernel.org> # v3.8+
Acked-by: Mukesh Kumar Savaliya <mukesh.savaliya@oss.qualcomm.com>
Signed-off-by: Andi Shyti <andi.shyti@kernel.org>
Link: https://patch.msgid.link/20260827-i2c-dma-channel-leak-v1-1-271d4adc03a0@cherr.cc
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Linkai Gong <gonglinkai@kylinos.cn>
Date: Mon Sep 7 15:11:02 2026 +0800
i2c: atr: fix dangling adapter pointer on add failure
commit ad34235808b63a70ca4989b7a2852923193d06ef upstream.
i2c_atr_add_adapter() stores atr->adapter[chan_id] before
i2c_add_adapter() so that the I2C bus notifier can match child clients
during registration. On failure the channel is freed but the slot was
left pointing at freed memory, which can lead to use-after-free in
i2c_atr_del_adapter() / cleanup and also block reuse with -EEXIST.
Clear the slot on the i2c_add_adapter() error path before freeing chan.
Fixes: a076a860acae ("media: i2c: add I2C Address Translator (ATR) support")
Signed-off-by: Linkai Gong <gonglinkai@kylinos.cn>
Cc: <stable@vger.kernel.org> # v6.6+
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@kernel.org>
Link: https://patch.msgid.link/20260907071102.1080840-1-gonglinkai@kylinos.cn
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Guangshuo Li <lgs201920130244@gmail.com>
Date: Mon Sep 14 17:15:44 2026 +0800
i2c: imx: disable autosuspend on remove
commit e0c3e9d76adbe522dd420a766ce42d03ce887c29 upstream.
i2c_imx_probe() enables runtime PM autosuspend with
pm_runtime_use_autosuspend(). The probe error path correctly undoes
this setting with pm_runtime_dont_use_autosuspend(), but the normal
remove path only disables runtime PM.
The runtime PM API requires pm_runtime_use_autosuspend() to be undone
with pm_runtime_dont_use_autosuspend() at driver exit unless runtime PM
was enabled with devm_pm_runtime_enable(). Leaving the autosuspend flag
set therefore leaves the runtime PM state incompletely cleaned up after
the driver is unbound.
Add the missing pm_runtime_dont_use_autosuspend() call to the remove
path.
This issue was found by manual code inspection.
Fixes: 588eb93ea49f ("i2c: imx: add runtime pm support to improve the performance")
Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com>
Cc: <stable@vger.kernel.org> # v4.5+
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Andi Shyti <andi.shyti@kernel.org>
Link: https://patch.msgid.link/20260914091544.1667137-1-lgs201920130244@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shengzhuo Wei <me@cherr.cc>
Date: Thu Aug 27 23:43:02 2026 +0800
i2c: imx: release DMA channels on probe error
commit e9f03b9625e2eeaca357b065c92d5b14064a1583 upstream.
i2c_imx_dma_request() acquires exclusive tx/rx DMA channels and is
optional: on errors other than -EPROBE_DEFER the driver falls back to
PIO mode and probe continues. If i2c_add_numbered_adapter() then fails,
probe returns through clk_notifier_unregister without releasing the
channels, because the remove callback is not invoked after a failed
probe.
Release the channels on the probe error path, mirroring
i2c_imx_remove().
Fixes: ce1a78840ff7 ("i2c: imx: add DMA support for freescale i2c driver")
Assisted-by: GLM:5.3
Signed-off-by: Shengzhuo Wei <me@cherr.cc>
Cc: <stable@vger.kernel.org> # v3.19+
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Andi Shyti <andi.shyti@kernel.org>
Link: https://patch.msgid.link/20260827-i2c-dma-channel-leak-v1-2-271d4adc03a0@cherr.cc
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Liu Zhenlong <dragonliu2018@gmail.com>
Date: Wed Aug 19 01:57:50 2026 +0800
i2c: qcom-cci: fix device_node refcount leak in cci_probe()/cci_remove()
commit 7362a1553eb09a8cdf8be7e509bd5309a8342486 upstream.
The of_node_put() matching of_node_get() runs after i2c_del_adapter(),
whose trailing memset() zeroes adap->dev and thus adap->dev.of_node,
making the put a no-op and leaking the node on every adapter removal
and error cleanup.
Use a devm action: the pointer is captured at registration, out of
reach of that memset(), and devres runs the put once on probe failure
and detach, replacing the three manual of_node_put() calls. The
setup loop uses the scoped iterator form so the child node is released
automatically if devm_add_action_or_reset() fails mid-loop.
Suggested-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Fixes: 02a4a69667a2 ("i2c: qcom-cci: don't put a device tree node before i2c_add_adapter()")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Liu Zhenlong <dragonliu2018@gmail.com>
Cc: <stable@vger.kernel.org> # v5.17+
Reviewed-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Signed-off-by: Andi Shyti <andi.shyti@kernel.org>
Link: https://patch.msgid.link/20260818175750.4205-1-dragonliu2018@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shuhei Takeshita <jyohuku.alterego@gmail.com>
Date: Sun Aug 9 12:27:43 2026 +0900
IB/hfi1: Fix the PIO_CRED credit-return mmap
commit 62f0f34fbd2b2d5653d33d3b9d42fdcabb1c0101 upstream.
hfi1_file_mmap()'s PIO_CRED case must hand user space the single
credit-return page that holds this context's entry. That page is the
second or third page of the per-node credit-return allocation once the
hardware send context index reaches 64 or 128, so the failure below is
intermittent: when the entry lands on the first page the offset is zero
and everything works.
Two things are wrong.
First, cr_page_offset is a byte offset but .va is a struct
credit_return *, so adding it is pointer arithmetic and scales the offset
by sizeof(struct credit_return) == 64. memvirt then lands 256 KiB or
512 KiB past a 10240-byte allocation. With an IOMMU translating, that
address is inside the vmalloc range but in no vm_area, so
dma_mmap_coherent() -> iommu_dma_mmap() finds no pages, vmalloc_to_pfn()
returns page_to_pfn(NULL), and remap_pfn_range() installs a frame above
MAXPHYADDR. The first user read then takes:
psm2_ep_open_pr: Corrupted page table at address 7a14d007e000
PGD 800000013886a067 P4D 800000013886a067 PUD 13886b067 PMD 13886c067
PTE 800049168e911235
Oops: Bad pagetable: 000d [#1] SMP PTI
Second, and still wrong once the arithmetic is corrected,
dma_mmap_coherent() describes a whole coherent buffer and selects the
page within it with vma->vm_pgoff. Offsetting cpu_addr has no effect:
for a vmap'd allocation iommu_dma_mmap() uses cpu_addr only to locate the
vm_area and then maps pages[vm_pgoff], which hfi1_file_mmap() has just
set to 0. User space therefore always receives the first credit-return
page, every credit read is for the wrong context, and send PIO stalls
forever.
Use the DMA API as intended: pass the base of the allocation with its
full length and select the page with vm_pgoff. A separate length is
needed because memlen must keep describing the VMA for the existing size
check. The dma-direct path stays correct as well, since dma_direct_mmap()
adds the same vm_pgoff to the base pfn.
Tested on a Dell T7610 (Xeon E5-2650 v2, Intel IOMMU in DMA-FQ mode)
against a Threadripper PRO 3995WX peer, both Omni-Path 100. Before this
change psm2_ep_open() Oopses the kernel; with only the arithmetic
corrected psm2_ep_open() succeeds but any transfer that uses send PIO
hangs, PSM2_SDMA=2 (send PIO disabled) completing normally while
PSM2_SDMA=0 (send PIO only) hangs every time. With this change send PIO,
send DMA and the default mixed mode all work.
Fixes: 1ec82317a1da ("IB/hfi1: Use dma_mmap_coherent for matching buffers")
Cc: stable@vger.kernel.org
Signed-off-by: Shuhei Takeshita <jyohuku.alterego@gmail.com>
Link: https://patch.msgid.link/20260809032743.2671579-3-jyohuku.alterego@gmail.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shuhei Takeshita <jyohuku.alterego@gmail.com>
Date: Sun Aug 9 12:27:42 2026 +0900
IB/hfi1: Resolve the credit-return buffer through the send context's node
commit 975396b9e5a4028e649f4b9a6a5ca5dfb76a824b upstream.
hfi1_file_mmap()'s PIO_CRED case derives this context's credit-return
page offset, and the DMA handle for it, from dd->cr_base[uctxt->numa_id].
uctxt->numa_id is the node of whichever CPU the process happened to be
running on, but the entry itself lives in the credit-return allocation of
the send context's own node:
sc->hw_free = &sc->dd->cr_base[sc->node].va[gc].cr[index];
and user send contexts are allocated with sc_alloc(dd, SC_USER, ...,
dd->node), the HFI-local node. On a multi-socket host with the process
running off that node the two allocations differ, so the subtraction
produces an offset into an unrelated buffer and the DMA handle belongs to
the wrong allocation.
Use the send context's own node for all three references. The
continuation lines are reindented at the same time; they mixed spaces and
tabs.
Fixes: 7724105686e7 ("IB/hfi1: add driver files")
Cc: stable@vger.kernel.org
Signed-off-by: Shuhei Takeshita <jyohuku.alterego@gmail.com>
Link: https://patch.msgid.link/20260809032743.2671579-2-jyohuku.alterego@gmail.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Carolina Jubran <cjubran@nvidia.com>
Date: Wed Sep 2 17:06:32 2026 +0300
IB/IPoIB: Avoid restoring OPER_UP after multicast flush
[ Upstream commit 9a141d3dc869d18b2eab35e999f4790a9b84e40f ]
ipoib_ib_dev_flush_light() temporarily clears IPOIB_FLAG_OPER_UP to
prevent multicast joins while ipoib_mcast_dev_flush() is running, and
restores the flag afterwards if it was previously set.
This restore races with ipoib_ib_dev_down(). If the interface is brought
down while the flush is in progress, ipoib_ib_dev_down() clears
IPOIB_FLAG_OPER_UP, but the flush path may set it again after the device
has already gone down.
Since commit 894021a75291 ("IB/ipoib: Make the carrier_on_task race
aware"), ipoib_mcast_carrier_on_task() relies on IPOIB_FLAG_OPER_UP
being cleared to terminate its rtnl_trylock() retry loop. If the flag is
left set after shutdown, the workqueue retries forever, causing teardown
to deadlock when ipoib_ndo_uninit() waits in destroy_workqueue() while
holding RTNL.
Instead of overloading IPOIB_FLAG_OPER_UP to block multicast joins
during a light flush, introduce a dedicated IPOIB_FLAG_MCAST_FLUSH flag.
Use it together with IPOIB_FLAG_OPER_UP to determine whether multicast
joins are allowed, avoiding the race with device shutdown.
Fixes: 344bacca8cd8 ("IB/ipoib: Don't allow MC joins during light MC flush")
Reported-by: Ben Davies <ben.davies@gresearch.co.uk>
Signed-off-by: Carolina Jubran <cjubran@nvidia.com>
Reviewed-by: Cosmin Ratiu <cratiu@nvidia.com>
Signed-off-by: Edward Srouji <edwards@nvidia.com>
Link: https://patch.msgid.link/20260902-avoid-rest-oper-up-v1-1-04fcd4916cae@nvidia.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Yehyeong Lee <yhlee@isslab.korea.ac.kr>
Date: Wed Aug 19 10:08:04 2026 +0900
IB/iser: reject a remote invalidation of an unregistered direction
[ Upstream commit d85f0f0a7c85756fc992c70d869706f19dac9259 ]
A write command whose data is sent entirely as immediate data is not
registered. iser_reg_mem_fastreg() takes the DMA key path and leaves
rdma_reg[ISER_DIR_OUT].desc at NULL, while iser_dma_map_task_data() has
already set dir[ISER_DIR_OUT].
iser_check_remote_inv() looks at dir[] alone and hands the descriptor to
iser_inv_desc(), which reads desc->sig_protected. A target that answers
such a command with IB_WR_SEND_WITH_INV faults the initiator.
Leaving those commands unregistered is deliberate.
The same function already terminates the connection when a target sends
a remote invalidation the initiator did not ask for. A target that
invalidates a direction that was never registered is in the same class,
so give it the same answer.
Oops: general protection fault, probably for non-canonical address 0xdffffc0000000004: 0000 [#1] SMP KASAN NOPTI
KASAN: null-ptr-deref in range [0x0000000000000020-0x0000000000000027]
CPU: 0 UID: 0 PID: 40 Comm: kworker/u8:2 Not tainted 7.2.0-rc5-ISERHOST-gf5098b6bae76-dirty #3 PREEMPT(lazy)
Hardware name: QEMU Ubuntu 24.04 PC v2 (i440FX + PIIX, arch_caps fix, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014
Workqueue: rxe_wq do_work
RIP: 0010:iser_task_rsp+0x6d6/0xec0
Code: 48 c1 ea 03 80 3c 02 00 0f 85 ba 06 00 00 48 8b 9b 78 01 00 00 48 b8 00 00 00 00 00 fc ff df 48 8d 7b 20 48 89 fa 48 c1 ea 03 <0f> b6 04 02 84 c0 74 06 0f 8e 76 06 00 00 80 7b 20 00 0f 84 3d 04
RSP: 0018:ffff88811b008db8 EFLAGS: 00010202
RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 0000000000001848
RDX: 0000000000000004 RSI: 1ffff11021587b12 RDI: 0000000000000020
RBP: ffff88810adc1ae4 R08: ffff888109b7f860 R09: ffffffff90a922c0
R10: ffff88810adc1a1c R11: 000000000000003c R12: ffff888109b7f800
R13: ffff88810adc1acc R14: ffff888109b7f820 R15: 0000000000000000
FS: 0000000000000000(0000) GS:ffff88818a676000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000005afe2b CR3: 000000010af23005 CR4: 0000000000770ef0
PKRU: 55555554
Call Trace:
<IRQ>
__ib_process_cq+0xe1/0x390
ib_poll_handler+0x6e/0x200
irq_poll_softirq+0x1df/0x480
? clockevents_program_event+0x2ba/0x860
? __pfx_irq_poll_softirq+0x10/0x10
handle_softirqs+0x18e/0x590
? __pfx_handle_softirqs+0x10/0x10
? __hrtimer_rearm_deferred+0x156/0x450
do_softirq+0x3b/0x60
</IRQ>
<TASK>
__local_bh_enable_ip+0x61/0x70
__alloc_skb+0x732/0x890
? _raw_spin_lock_irqsave+0x85/0xe0
? __pfx___alloc_skb+0x10/0x10
? _raw_read_unlock_irqrestore+0x16/0x50
rxe_init_packet+0x16b/0x4f0
prepare_ack_packet+0xb8/0x830
rxe_receiver+0x499/0x9980
? __pfx_rxe_receiver+0x10/0x10
? rxe_completer+0x29e5/0x38c0
? hrtimer_start_range_ns_common+0x75f/0x1730
? hrtimer_start_range_ns+0xa6/0x2c0
? __pfx__raw_spin_lock_irqsave+0x10/0x10
? __pfx_rxe_receiver+0x10/0x10
do_work+0x144/0x470
process_one_work+0x633/0x1030
? assign_work+0x11d/0x370
worker_thread+0x45b/0xd10
? __pfx_worker_thread+0x10/0x10
kthread+0x2c6/0x3b0
? recalc_sigpending+0x15c/0x1e0
? __pfx_kthread+0x10/0x10
ret_from_fork+0x36e/0x5a0
? __pfx_ret_from_fork+0x10/0x10
? __switch_to+0x572/0xdd0
? __pfx_kthread+0x10/0x10
ret_from_fork_asm+0x1a/0x30
</TASK>
Modules linked in:
---[ end trace 0000000000000000 ]---
Fixes: 59caaed7a72a ("IB/iser: Support the remote invalidation exception")
Signed-off-by: Yehyeong Lee <yhlee@isslab.korea.ac.kr>
Link: https://patch.msgid.link/20260819010804.641772-1-yhlee@isslab.korea.ac.kr
Reviewed-by: Max Gurtovoy <mgurtovoy@nvidia.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Yehyeong Lee <yhlee@isslab.korea.ac.kr>
Date: Fri Aug 21 17:06:20 2026 +0900
IB/isert: wait for deferred control PDU completions before releasing the connection
[ Upstream commit a8fe3dfce8c0d8a76dc3d8486a5bff5feebe156f ]
isert_send_done() hands ISTATE_SEND_TASKMGTRSP, ISTATE_SEND_REJECT and
ISTATE_SEND_TEXTRSP completions off to isert_comp_wq and returns. The work
item then runs isert_completion_put() -> isert_put_cmd(), which reads
isert_conn->conn and takes conn->cmd_lock.
Nothing orders that work item against teardown. isert_wait_conn() queues
isert_release_work, which frees isert_conn, and iscsit_close_connection()
frees the iscsit_conn right after it returns, so the queued work can run
against freed memory.
Count the deferred control PDU completions per connection and let
isert_wait_conn() wait for them before the release work is queued.
ISTATE_SEND_LOGOUTRSP is deliberately not counted: that branch runs
iscsit_logout_post_handler(), which ends up waiting for
conn->conn_wait_comp, and that completion is only sent by
iscsit_close_connection() after it has called iscsit_wait_conn().
Waiting for it here would deadlock. Its wait stays the existing
isert_wait4logout().
The splat below is from a kernel with tracing printk()s and an msleep(200)
injected into isert_do_control_comp() to widen the window:
BUG: KASAN: slab-use-after-free in isert_put_cmd+0x53d/0x620
Read of size 8 at addr ffff8881054f1038 by task kworker/u17:1/182
CPU: 0 UID: 0 PID: 182 Comm: kworker/u17:1 Tainted: G B 7.2.0-rc5-TWIDE-gb8babf08acc7 #1 PREEMPT(lazy)
Tainted: [B]=BAD_PAGE
Hardware name: QEMU Ubuntu 24.04 PC v2 (i440FX + PIIX, arch_caps fix, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014
Workqueue: isert_comp_wq isert_do_control_comp
Call Trace:
<TASK>
dump_stack_lvl+0x53/0x70
print_report+0xd0/0x630
? __pfx__raw_spin_lock_irqsave+0x10/0x10
? _raw_spin_unlock_irqrestore+0x3e/0x70
? isert_put_cmd+0x53d/0x620
kasan_report+0xce/0x100
? isert_put_cmd+0x53d/0x620
isert_put_cmd+0x53d/0x620
? isert_completion_put+0x305/0x330
? isert_do_control_comp+0x2ef/0x310
process_one_work+0x633/0x1030
? assign_work+0x11d/0x370
worker_thread+0x45b/0xd10
? __pfx_worker_thread+0x10/0x10
? __pfx_worker_thread+0x10/0x10
kthread+0x2c6/0x3b0
? recalc_sigpending+0x15c/0x1e0
? __pfx_kthread+0x10/0x10
ret_from_fork+0x36e/0x5a0
? __pfx_ret_from_fork+0x10/0x10
? __switch_to+0x572/0xdd0
? __pfx_kthread+0x10/0x10
ret_from_fork_asm+0x1a/0x30
</TASK>
Allocated by task 48:
kasan_save_stack+0x33/0x60
kasan_save_track+0x14/0x30
__kasan_kmalloc+0x8f/0xa0
__kmalloc_cache_noprof+0x158/0x370
isert_cma_handler+0x1e3/0x2ae0
cma_cm_event_handler+0x3e/0x240
cma_ib_req_handler+0x17d9/0x4490
cm_process_work+0x41/0x330
cm_work_handler+0x5727/0xc160
process_one_work+0x633/0x1030
worker_thread+0x45b/0xd10
kthread+0x2c6/0x3b0
ret_from_fork+0x36e/0x5a0
ret_from_fork_asm+0x1a/0x30
Freed by task 184:
kasan_save_stack+0x33/0x60
kasan_save_track+0x14/0x30
kasan_save_free_info+0x3b/0x60
__kasan_slab_free+0x43/0x70
kfree+0x121/0x380
iscsit_close_connection+0x7cf/0x1e60
iscsit_take_action_for_connection_exit+0x1b6/0x360
iscsi_target_tx_thread+0x472/0x690
kthread+0x2c6/0x3b0
ret_from_fork+0x36e/0x5a0
ret_from_fork_asm+0x1a/0x30
Fixes: b8d26b3be8b3 ("iser-target: Add iSCSI Extensions for RDMA (iSER) target driver")
Signed-off-by: Yehyeong Lee <yhlee@isslab.korea.ac.kr>
Link: https://patch.msgid.link/20260821080620.1694119-1-yhlee@isslab.korea.ac.kr
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Shuangpeng Bai <shuangpeng.kernel@gmail.com>
Date: Sun Aug 16 00:45:10 2026 -0400
IB/mlx4: Fix use-after-free on pkey sysfs registration failure
commit 1af874e9f4ce22ccf8b10ab5462f32c70d3be21a upstream.
register_pkey_tree() ignores errors from register_one_pkey_tree() and
continues registering the remaining slaves. The per-slave error path has
already released the pkey parent kobjects, but their pointers remain
stored in the device. A later device cleanup therefore passes the stale
pointers to kobject_put(), causing a use-after-free.
Clear the parent pointers after releasing a failed slave tree and skip
unregistered trees during device cleanup. This preserves the existing
best-effort registration behavior while preventing a second cleanup of
the failed tree.
Fixes: c1e7e466120b ("IB/mlx4: Add iov directory in sysfs under the ib device")
Cc: stable@vger.kernel.org
Signed-off-by: Shuangpeng Bai <shuangpeng.kernel@gmail.com>
Link: https://patch.msgid.link/20260816044510.3848996-1-shuangpeng.kernel@gmail.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Alvin Šipraga <alvin.sipraga@analog.com>
Date: Tue Aug 18 18:00:02 2026 +0200
Input: adp5588-keys - cache GPIO state before registering the gpiochip
commit 21efadc62272cabee9bec27777ae75d84a9ca8a8 upstream.
So as not to clobber any pre-programmed GPIO state in the execution
of its gpiochip ops, the driver caches things during probe time.
However, since those ops can be called both during and immediately after
the call to devm_gpiochip_add_data(), it is imperative that things are
cached before that. That's not the case right now, so reorder the two
steps to prevent any clobbering.
In the concrete example which motivated this change, a bootloader was
preconfiguring an important GPIO output to HIGH before booting the
kernel. Linux would then inadvertently set that output to LOW while
configuring a GPIO hog on a discrete GPIO line within the same 8-bit
bank (because the cached value was 0=LOW).
Fixes: ba9f507a1bea ("Input: adp5588-keys - export unused GPIO pins")
Signed-off-by: Alvin Šipraga <alvin.sipraga@analog.com>
Reviewed-by: Nuno Sá <nuno.sa@analog.com>
Link: https://patch.msgid.link/20260818-adp5588-gpio-cache-v1-1-650a2674fc0d@analog.com
Cc: stable@vger.kernel.org
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Alexei Turtanov <9alexei9@gmail.com>
Date: Fri Aug 28 14:22:39 2026 +0300
Input: atkbd - skip deactivate for Xiaomi Redmi Book Pro 16 2026
commit aefbda23eeba234c3ff0f135dc5be6e294bd25a6 upstream.
The internal keyboard of the Xiaomi Redmi Book Pro 16 2026 (board TM2425)
does not work: atkbd_probe() succeeds and every command is ACKed, but no
scancodes ever arrive afterwards.
Testing on the hardware through serio_raw shows that ATKBD_CMD_RESET_DIS
(0xF5) is the culprit. After 0xF5 the embedded controller keeps ACKing
commands but stops delivering scancodes, and neither ATKBD_CMD_ENABLE
(0xF4) nor ATKBD_CMD_RESET_BAT (0xFF) bring them back. Only re-enabling
the keyboard interface at the controller level (i8042 command 0xAE, or
rewriting the command byte as i8042_port_close() does) revives it.
Running the init sequence without 0xF5 (0xED 0x00, 0xF3 0x00, 0xF4)
keeps the keyboard working.
'i8042.dumbkbd=1' also works around this, but then the driver never
writes to the keyboard and the LEDs cannot be controlled. Use the
existing atkbd_deactivate_fixup quirk instead, as done for the sibling
TM2424 by commit 3a046db33bb9 ("Input: atkbd - skip deactivate for
Xiaomi Book Pro 14's internal keyboard"). Tested on v7.2: keyboard,
Caps Lock LED and s2idle suspend/resume all work.
DMI: XIAOMI REDMI Book Pro 16 2026/TM2425, BIOS RMAPT6B0P0909 05/22/2026
Fixes: 9cf6e24c9fbf ("Input: atkbd - do not skip atkbd_deactivate() when skipping ATKBD_CMD_GETID")
Cc: stable@vger.kernel.org
Signed-off-by: Alexei Turtanov <9alexei9@gmail.com>
Link: https://patch.msgid.link/20260828112239.18081-1-9alexei9@gmail.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Linkai Gong <gonglinkai@kylinos.cn>
Date: Tue Sep 1 20:26:49 2026 +0800
Input: cyttsp5 - clamp the HID report size before memcpy
commit 85f080fb87ed5cd3e46121be677f52c82f26a0ab upstream.
The size field comes from the device and is used as the memcpy()
length into response_buf, which is CY_MAX_INPUT bytes.
Fixes: 5b0c03e24a06 ("Input: Add driver for Cypress Generation 5 touchscreen")
Signed-off-by: Linkai Gong <gonglinkai@kylinos.cn>
Link: https://patch.msgid.link/20260901122649.1173066-1-gonglinkai@kylinos.cn
Cc: stable@vger.kernel.org
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: hpp.iscas <hppiscas@163.com>
Date: Sat Sep 5 21:40:04 2026 +0800
Input: eeti_ts - publish the OF module alias
[ Upstream commit a52ae68a937efc353251aec27fc995ff66cbe1ca ]
The EETI driver matches eeti,exc3000-i2c Device Tree clients, but only
publishes the legacy eeti_ts I2C ID. The I2C core emits an OF modalias
for a Device Tree client.
Publish the existing OF match table within its CONFIG_OF guard.
Fixes: e32d7f1b246c ("Input: eeti - add device tree matching table")
Signed-off-by: hpp.iscas <hppiscas@163.com>
Link: https://patch.msgid.link/20260905134004.66336-1-hppiscas@163.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Iván Ezequiel Rodriguez <ivanrwcm25@gmail.com>
Date: Tue Sep 1 10:06:27 2026 -0300
Input: evdev - zero absinfo before partial copy in EVIOCSABS
commit 8b852965b8eaf910c314dc346967ed82c8d4f235 upstream.
The EVIOCSABS handler copies at most the user supplied ioctl size into
an uninitialized on-stack struct input_absinfo:
if (copy_from_user(&abs, p, min_t(size_t,
size, sizeof(struct input_absinfo))))
The size comes from _IOC_SIZE() of the ioctl command and is therefore
fully controlled by userspace. A short size leaves the trailing part of
the structure holding whatever was on the kernel stack, and the whole
structure is then stored into the device:
dev->absinfo[t] = abs;
EVIOCGABS hands that back to userspace, disclosing the stale stack
bytes. Only the resolution field is currently cleared, which covers the
legacy struct layout but not an arbitrarily short size.
Zero the structure before the copy so any part not supplied by the
caller reads back as zero. The existing resolution fixup is kept, since
it also handles a size that partially overlaps that field.
Fixes: 448cd1664a57 ("Input: evdev - rearrange ioctl handling")
Cc: stable@vger.kernel.org
Signed-off-by: Iván Ezequiel Rodriguez <ivanrwcm25@gmail.com>
Link: https://patch.msgid.link/20260901130629.24078-2-ivanrwcm25@gmail.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Runyu Xiao <runyu.xiao@seu.edu.cn>
Date: Wed Sep 2 23:40:04 2026 +0800
Input: hp_sdc - shut down kicker timer on module exit
commit 309731e95917125bbd13626a7a5600490a5bf44f upstream.
hp_sdc_kicker() rearms hp_sdc.kicker with mod_timer() after scheduling the
tasklet. The module exit path uses timer_delete_sync(). That waits for a
callback already running but can still leave the timer rearmed.
A callback can therefore leave the timer pending while hp_sdc_exit() tears
down the driver, allowing timer activity to access dismantled driver state.
Use timer_shutdown_sync() for final teardown. It waits for a running
callback and prevents rearming after module exit begins.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Assisted-by: Codex:GPT-5
Signed-off-by: Runyu Xiao <runyu.xiao@seu.edu.cn>
Acked-by: Helge Deller <deller@gmx.de>
Link: https://patch.msgid.link/20260902154004.3595416-1-runyu.xiao@seu.edu.cn
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Chris Sommers <chris.sommers@icloud.com>
Date: Mon Sep 7 11:27:23 2026 -0700
Input: i8042 - add quirk for Acer Aspire Go 15 AG15-42P
commit 25e424eb4ae1a662d9c3573218d06ac32f797fc5 upstream.
On the Acer Aspire Go 15 (AG15-42P), the internal keyboard drops out
~5 seconds after boot on both Linux and Linux-LTS kernels. Keystrokes on
the built-in keyboard stop registering while the trackpad and external
keyboards remain functional.
Testing confirms that booting with the i8042.reset kernel parameter
resolves the issue and keeps the internal keyboard responsive.
Add SERIO_QUIRK_RESET_ALWAYS to i8042_dmi_quirk_table for the Acer
Aspire AG15-42P to automatically apply this quirk on boot.
Signed-off-by: Chris Sommers <chris.sommers@icloud.com>
Link: https://patch.msgid.link/20260907182723.2709981-1-chris.sommers@icloud.com
Cc: stable@vger.kernel.org
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Date: Tue Aug 4 22:08:54 2026 -0700
Input: rmi_smbus - fix out-of-bounds read in rmi_smb_write_block()
commit 51cfe54f815ae175c7d1126b983d4d7c89715004 upstream.
When chunking writes into SMBus blocks in rmi_smb_write_block(), the
loop calculates block_len using the original total length (len) instead
of the remaining length (cur_len).
If len is greater than 32 bytes (SMB_MAX_COUNT), block_len remains 32
for every iteration, even on the final partial chunk where fewer than 32
bytes remain. This causes smb_block_write() to read 32 bytes from the
advanced data buffer pointer, reading past the end of the input buffer.
Fix this by calculating block_len using cur_len and advancing the buffer
and address pointers by block_len.
Fixes: 82264d0cf7ae ("Input: synaptics-rmi4 - add SMBus support")
Cc: stable@vger.kernel.org
Reported-by: sashiko-bot@kernel.org
Assisted-by: LLM
Link: https://patch.msgid.link/anLFSMKSoKyyZ272@google.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Hans de Goede <johannes.goede@oss.qualcomm.com>
Date: Wed Sep 9 11:39:34 2026 +0200
Input: soc_button_array - check btns_desc->package.count
commit fb5022278b6ea7f1838e3ef78028d5d5e3375f65 upstream.
Check that btns_desc->package.count is not 0 before accessing
btns_desc->package.elements[0].
Fixes: 4c3362f44980 ("Input: soc_button_array - add support for ACPI 6.0 Generic Button Device")
Cc: stable@vger.kernel.org
Reported-by: Shashiko <sashiko-bot@kernel.org>
Closes: https://lore.kernel.org/linux-input/20260909091440.3384C1F00A3A@smtp.kernel.org/
Signed-off-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
Link: https://patch.msgid.link/20260909093934.29411-2-johannes.goede@oss.qualcomm.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Hans de Goede <johannes.goede@oss.qualcomm.com>
Date: Wed Sep 9 11:39:33 2026 +0200
Input: soc_button_array - fix MS Surface Pro 11 probe failure
commit ed22ad5fdbdbf9b4cb4ad3003f60314b5a5eb89d upstream.
On the MS Surface Pro 11 soc_button_array probing races with the GPIO
driver probing. If soc_button_array wins the race then gpiod_get() returns
EPROBE_DEFER, which should normally take care of retrying later, but
the soc_button_array code deliberately ignores EPROBE_DEFER causing it
to fail its probe() which causes the volume and power buttons to now work.
The ignoring of EPROBE_DEFER is there to deal with a problem specific to
older Bay Trail (BYT) and Cherry Trail (CHT) tablets which often use this
driver. Modify the error handling to only ignore EPROBE_DEFER on BYT and
CHT platforms and propagate EPROBE_DEFER normally on other platforms.
Fixes: bcf059578980 ("Input: soc_button_array - partial revert of support for newer surface devices")
Cc: stable@vger.kernel.org
Reported-by: Sergey Lebedev <lsa.uz@pm.me>
Closes: https://lore.kernel.org/lkml/20260830141355.55898-1-lsa.uz@pm.me/
Signed-off-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
Tested-by: Sergey Lebedev <lsa.uz@pm.me>
Link: https://patch.msgid.link/20260909093934.29411-1-johannes.goede@oss.qualcomm.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Raphaël Larocque <rlarocque@disroot.org>
Date: Thu Sep 10 12:44:25 2026 -0400
Input: synaptics - disable InterTouch on ThinkPad T440p (board id 2722)
commit 26eb3d92c7a4d7adb1ae1740ca6e8e100b11d1ec upstream.
The Lenovo ThinkPad T440p (PNP ID LEN0036, board id 2722) has a
Synaptics touchpad whose SMBus companion is not ready at boot and
takes roughly 200 seconds to appear. During this window the touchpad
and TrackPoint are completely unresponsive on approximately 50% of
boots, making the machine unusable until the companion finally
registers.
The device is in the topbuttonpad_pnp_ids[] SMBus allowlist, so the
kernel attempts to use SMBus/RMI4 mode by default. When the companion
is not ready, psmouse_smbus_init() leaves breadcrumbs and returns
-EAGAIN, the PS/2 fallback path is taken, but the device does not
function properly until the companion appears and RMI4 takes over.
Disable SMBus InterTouch for board id 2722 so the touchpad and
TrackPoint work immediately via PS/2 from boot. Users can still force
SMBus with psmouse.synaptics_intertouch=1 if needed.
Tested-by: Raphaël Larocque <rlarocque@disroot.org>
Signed-off-by: Raphaël Larocque <rlarocque@disroot.org>
Link: https://patch.msgid.link/20260910164425.12832-1-rlarocque@disroot.org
Cc: stable@vger.kernel.org
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Date: Wed Aug 5 22:44:23 2026 -0700
Input: synaptics-rmi4 - fix GPF in suspend and resume when unbound
commit fe10579b6dc3f0dac61e51e1797cacbba5039ac2 upstream.
Transport drivers (such as rmi_i2c and rmi_spi) invoke
rmi_driver_suspend() and rmi_driver_resume() on their child rmi_dev
device during system power management events. However, transport drivers
are fully registered and operational even if the physical RMI driver
failed to bind or probe the rmi_dev device.
When rmi_driver_suspend() or rmi_driver_resume() is called on an unbound
rmi_dev, dev_get_drvdata() returns NULL. Calling rmi_disable_irq() or
rmi_enable_irq() without driver data attached causes a NULL pointer
dereference and General Protection Fault when attempting to lock
data->enabled_mutex.
Fix this by checking if driver data is attached to rmi_dev in
rmi_driver_suspend() and rmi_driver_resume(), exiting early if
no driver data is present.
Fixes: 2b6a321da9a2 ("Input: synaptics-rmi4 - add support for Synaptics RMI4 devices")
Reported-by: syzbot+09103639e39c989e3ed3@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=09103639e39c989e3ed3
Cc: stable@vger.kernel.org
Assisted-by: LLM
Link: https://patch.msgid.link/anQe8UiyUR4x0flD@google.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Karl Mehltretter <kmehltretter@gmail.com>
Date: Sat Sep 5 12:20:38 2026 +0200
Input: trackpoint - fix the inertia attribute name in the ABI document
[ Upstream commit 45b0037899704caf9078be2be4de69361ca7d933 ]
The attribute is created as "inertia" (TRACKPOINT_INT_ATTR(inertia, ...)
in drivers/input/mouse/trackpoint.c); the ABI file spells the path
"intertia". The description below it already says inertia.
Fix the spelling.
Fixes: aebb47d4e7a9 ("Input: trackpoint: document sysfs interface")
Assisted-by: LLM
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Link: https://patch.msgid.link/20260905102038.42882-1-kmehltretter@gmail.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Roberts Kursitis <roberts.kursitis@azeron.eu>
Date: Sun Sep 6 17:30:40 2026 +0300
Input: xpad - add support for Azeron devices
commit cba76c0f47af1a389d718c5bb69e75cbd67bba98 upstream.
Azeron controllers (Cyro, Cyborg, Classic/Compact, Cyro Lefty,
Cyborg II and Keyzen) present a standard Xbox 360 controller
interface, so they work with the existing xpad driver once their
USB IDs are added.
The 0x16d0 vendor ID is a shared block, but this is safe because
xpad only binds interfaces that match the Xbox 360 signature.
Tested with an Azeron Keyzen.
Signed-off-by: Roberts Kursitis <roberts.kursitis@azeron.eu>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260906143040.162418-1-roberts.kursitis@azeron.eu
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Erich Sartison <byt.es@mailbox.org>
Date: Thu Sep 3 12:31:37 2026 +0200
Input: xpad - add support for Victrix Pro BFG Controller
commit 971fa7ea8621e123feb9c8d7dc61be1c656bd945 upstream.
The controller doesn't currently work via USB-cable.
Signed-off-by: Erich Sartison <byt.es@mailbox.org>
Link: https://patch.msgid.link/20260903103137.630170-1-byt.es@mailbox.org
Cc: stable@vger.kernel.org
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Jeremy Nyberg <slickstretch3.0@gmail.com>
Date: Sun Sep 13 17:43:02 2026 -0700
Input: xpad - fix PDP Marvel Xbox 360 controller
commit 7bc369cb3d3f3656eb77285628ee264264d28ad4 upstream.
The PDP Marvel Xbox 360 controller with USB ID 0e6f:0147 is
incorrectly classified as an Xbox One controller.
With the current XTYPE_XBOXONE classification, the controller is
detected but produces no input, while its four player LEDs continue
blinking indefinitely.
Classify USB ID 0e6f:0147 as an Xbox 360 controller instead.
Tested on a PDP Marvel Xbox 360 controller with USB ID 0e6f:0147.
All inputs register correctly and the player LED indicates the
current player.
Fixes: c225370e01b8 ("Input: xpad - sync supported devices with 360Controller")
Cc: stable@vger.kernel.org
Signed-off-by: Jeremy Nyberg <SlickStretch3.0@gmail.com>
Link: https://patch.msgid.link/20260910071627.236014-1-SlickStretch3.0@gmail.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Iván Ezequiel Rodriguez <ivanrwcm25@gmail.com>
Date: Tue Sep 1 10:06:28 2026 -0300
Input: zero ff_effect before compat copy in input_ff_effect_from_user
commit f84819ef8d66931ee3998fee3c4f03230f4cb6cc upstream.
In the compat path input_ff_effect_from_user() aliases the caller's
native struct ff_effect with the smaller struct ff_effect_compat and
copies only the compat sized prefix:
compat_effect = (struct ff_effect_compat *)effect;
if (copy_from_user(compat_effect, buffer,
sizeof(struct ff_effect_compat)))
The tail of the native structure is never written. Callers pass an
uninitialized on-stack object, for example evdev_do_ioctl() for
EVIOCSFF, so those bytes keep their previous stack contents.
input_ff_upload() then stores the full native structure in
ff->effects[id], from where a uinput based force feedback daemon can
read it back via UI_BEGIN_FF_UPLOAD, disclosing kernel stack memory to
userspace.
Zero the effect before the compat copy.
Fixes: 2d56f3a32c0e ("Input: refactor evdev 32bit compat to be shareable with uinput")
Cc: stable@vger.kernel.org
Signed-off-by: Iván Ezequiel Rodriguez <ivanrwcm25@gmail.com>
Link: https://patch.msgid.link/20260901130629.24078-3-ivanrwcm25@gmail.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Dong Chenchen <dongchenchen2@huawei.com>
Date: Thu Sep 10 22:00:42 2026 +0800
ipv4: icmp: reject RTN_UNREACHABLE input routes in icmp_route_lookup
[ Upstream commit 2998147b59c9df0a51477c7a6b3d1f0ba3127dd4 ]
When the forward output route cannot be used in icmp_route_lookup(),
it enters the "reverse path" and calls ip_route_input() on fl4_dec.daddr,
the original packet's source address.
ip_route_input() only returns an error for truly invalid packets. For
unreachable addresses it will succeed and return an input route whose
dst.output is set to ip_rt_bug(). The existing check only rejects
RTN_LOCAL routes, so the RTN_UNREACHABLE route types can still be returned
and later used for output, syzkaller triggering a WARN_ON_ONCE()
in ip_rt_bug() as bellow:
------------[ cut here ]------------
WARNING: net/ipv4/route.c:1273 at ip_rt_bug+0x14/0x20
RIP: 0010:ip_rt_bug+0x14/0x20
Call Trace:
ip_push_pending_frames+0xfa/0x100
__icmp_send+0x905/0xf10
ip_options_compile+0xc0/0xd0
ip_rcv_finish_core+0x321/0xae0
ip_rcv+0x1de/0x260
__netif_receive_skb_one_core+0x11a/0x130
netif_receive_skb+0x7b/0x260
tun_get_user+0x11bf/0x1c10
------------[ cut here ]------------
Reject input route that is RTN_UNREACHABLE to fix it. The net warning
is only printed for RTN_LOCAL, as RTN_UNREACHABLE is not the result of
a race condition.
Fixes: 8b7817f3a959 ("[IPSEC]: Add ICMP host relookup support")
Suggested-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Dong Chenchen <dongchenchen2@huawei.com>
Link: https://patch.msgid.link/20260910140042.1880242-1-dongchenchen2@huawei.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Zhiling Zou <zhilinz@nebusec.ai>
Date: Mon Aug 3 21:28:58 2026 +0800
ipv6: xfrm: use full sockets in local error paths
commit 6973a21ee73c5567f883813c8ef414774b45892f upstream.
xfrm6_local_rxpmtu() and xfrm6_local_error() dereference skb->sk as if it
always pointed at a full IPv6 socket.
That is not guaranteed. TCP SYN-ACK skbs can be owned by a
TCP_NEW_SYN_RECV request_sock while the output path itself is driven by the
full listener. If rerouting selects an IPv6 XFRM tunnel route with a lower
MTU, the local PMTU/error handling path can reach these callbacks with that
mini-socket still attached to the skb.
The callbacks then miscast the request socket as a full inet/IPv6 socket and
can read beyond the request_sock allocation when they access inet_sock or
ipv6_pinfo state.
Resolve the owner with skb_to_full_sk() in both callbacks and bail out when
no full socket is attached. This matches the surrounding XFRM IPv6 PMTU/error
logic, which already reasons about full sockets with skb_to_full_sk().
Fixes: dd767856a36e ("xfrm6: Don't call icmpv6_send on local error")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Cen Zhang <cenzhang@linux.microsoft.com>
Date: Sat Sep 12 00:31:16 2026 +0300
KEYS: encrypted: fix integer overflow of datablob_len
commit 8697c431e297eb0d0ab13dda6bc172b48a34f05c upstream.
encrypted_key_alloc() stores datablob_len in a u16. It is computed from
multiple string and payload lengths. If the result exceeds U16_MAX, the
assignment truncates the allocation size. KASAN reports a 32760-byte
slab-out-of-bounds write when __ekey_init() copies the master key
description into the undersized buffer.
The total payload length stored in key->datalen is also a u16. Use
check_add_overflow() to reject values that do not fit either destination,
and use kzalloc_flex() for the flexible-array allocation.
Fixes: 7e70cb497850 ("keys: add new key-type encrypted")
Cc: stable@vger.kernel.org
Assisted-by: GitHub-Copilot:claude-opus-4.6
Signed-off-by: Cen Zhang <cenzhang@linux.microsoft.com>
Signed-off-by: Francis Perron <francis@akrites.dev>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Tested-by: R Nageswara Sastry <rnsastry@linux.ibm.com>
Link: https://lore.kernel.org/r/20260909153433.83117-1-cenzhang@linux.microsoft.com
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Karl Mehltretter <kmehltretter@gmail.com>
Date: Fri Aug 21 04:53:27 2026 +0200
keys: fix lost wakeup when reaping a dead key type
[ Upstream commit 2725ab3f5ad1c5f375c7c9fee4af02a9b138f701 ]
clear_bit() is atomic with respect to the word it modifies, but it is
an unordered operation: it implies no memory barrier on either side
(Documentation/atomic_bitops.txt).
key_garbage_collector() clears KEY_GC_REAPING_KEYTYPE with clear_bit()
and calls wake_up_bit() after reaping a dead key type. wake_up_bit()
uses a lockless waitqueue check and requires a full barrier after the
clear.
The existing smp_mb() is before clear_bit(), so nothing orders the clear
against that check. The GC can see an empty waitqueue while
unregister_key_type() still sees the bit set. The final wakeup is then
lost, leaving module unload stuck in wait_on_bit().
Use clear_and_wake_up_bit(). Its clear_bit_unlock() has RELEASE
semantics, so the completed GC work stays ordered before the clear, and
its smp_mb__after_atomic() orders the clear before the waitqueue check.
Fixes: 0c061b5707ab ("KEYS: Correctly destroy key payloads when their keytype is removed")
Assisted-by: Claude:claude-fable-5
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Link: https://lore.kernel.org/r/20260821025327.61488-1-kmehltretter@gmail.com
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Maoyi Xie <maoyixie.tju@gmail.com>
Date: Fri Aug 21 17:59:35 2026 +0800
keys: translate request_key_auth pid for the reading procfs instance
commit 0d6a4268b06084baafd8ee5d66955c7e1c2e053b upstream.
request_key_auth_describe() prints rka->pid into /proc/keys as a raw
pid_t in the initial pid namespace. A reader can open /proc/keys through
a mount in another pid namespace. That reader sees a number with no
meaning there. The number can even name an unrelated task. The line
needs VIEW on the key. So the reader either shares the key owner's uid
or possesses the key.
The fix keeps a struct pid. Commit 4f82f45730c6 ("net ip6 flowlabel:
Make owner a union of struct pid * and kuid_t") gave
/proc/net/ip6_flowlabel the same storage. The print goes through
pid_nr_ns(). It renders against the pid namespace of the procfs instance
the line is read through. Commit ad08978ab41c ("ipv6/flowlabel: simplify
pid namespace lookup") moved that print to the same anchor. Output
through an initial namespace /proc does not change. The line shows 0 for
a requestor with no number in that namespace.
Translating at read time was the alternative. find_pid_ns() can resolve
a recycled number. The line would then name a live task with no
connection to the key. A stored struct pid gives 0 instead when the
requestor has no number there.
Link: https://lore.kernel.org/keyrings/20260809110202.2180410-1-maoyixie.tju@gmail.com/
Fixes: 78b7280cce23 ("KEYS: Improve /proc/keys")
Cc: stable@vger.kernel.org # v5.10+
Assisted-by: Claude:claude-opus-5 codeql
Signed-off-by: Maoyi Xie <maoyixie.tju@gmail.com>
Link: https://lore.kernel.org/r/20260821095935.1864998-1-maoyixie.tju@gmail.com
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Jarkko Sakkinen <jarkko@kernel.org>
Date: Tue Sep 1 23:58:06 2026 +0300
KEYS: trusted: Fix tpm2_load_cmd() boundary check
commit 114f00d738f15dd8c7318369edcdc53dd6d08763 upstream.
tpm2_load_cmd() does boundary checks against the ASN.1 size i.e.,
payload->blob_len. Address this by passing the decoded blob size to
tpm2_load_cmd(), and use it for the boundary checks.
Cc: stable@vger.kernel.org # v5.13+
Fixes: f2219745250f ("security: keys: trusted: use ASN.1 TPM2 key format for the blobs")
Reported-by: co+6a581c4284f721d4@bugs.sh
Closes: https://bugs.sh/b/6a581c4284f721d4/
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Tested-by: Srish Srinivasan <ssrish@linux.ibm.com>
Link: https://lore.kernel.org/r/20260901205809.2028454-1-jarkko@kernel.org
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Thomas Huth <thuth@redhat.com>
Date: Wed Sep 9 17:57:07 2026 +0200
kselftest/arm64: Fix size of thread_data values for pthread_join()
commit 3d1ba5cbfb622025690c218d8f20da92a9ecb383 upstream.
pthread_join() stores the thread's return value (a "void *", i.e.
8 bytes on 64 bit computers) into the address that is passed as second
parameter. However, the entries of thread_data are only normal "int"s,
i.e. only 4 bytes. The additional 4 bytes of the return value clobber
whatever is adjacent on the stack, i.e. other members of the thread_data
array (which will be re-written in the next iteration of the for-loop,
so that nobody noticed this problem), or another other local variable
on the stack for the last iteration. Use "intptr_t" to declare the
thread_data array entries with the correct size.
Fixes: 29f080881601c ("kselftest/arm64: check GCR_EL1 after context switch")
Cc: stable@vger.kernel.org
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Sun Jun 28 15:43:58 2026 +0900
ksmbd: extend procfs server statistics
[ Upstream commit 4c670ccd5790816fc0f5714d5ee3c67dd5a9c67a ]
The server proc entry does not expose configured limits or enough outcome
data to distinguish protocol errors from transport stalls.
Report the server state, listener and signing configuration, connection
limits, timeout values, current client and open-file totals, IPC activity,
and durable scavenger state. Classify processed SMB2 response statuses by
NTSTATUS severity and provide counters for common error groups while
retaining the per-command counters.
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Stable-dep-of: d7fd1f98607f ("ksmbd: follow SMB2 session expiration semantics")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Sun Jun 28 15:35:29 2026 +0900
ksmbd: fix malformed procfs status output
[ Upstream commit 1f7dd03a88a8405143aab195f6fa9ed2243494b1 ]
The ksmbd procfs monitoring files produce misleading or malformed output.
The constant-name helper uses a bitwise test for enum values. This omits
zero-valued constants and can print multiple names for one lease state. It
also unconditionally emits a newline, splitting entries in the open-file
table across two lines. Session capabilities are printed as numeric flag
values even though a table of descriptive names is available.
Use exact matching for enum values. Print flag names as a comma-separated
list, preserving unknown bits as hexadecimal values. Let callers control
line termination so each open-file entry remains on one line. Print common
session properties once, and report signing and encryption independently.
Adjust client and open-file column widths for IPv6 addresses and 64-bit
file IDs, and fix the misspelled OPLOCK_EXCLUSIVE name. Also expose and
maintain the total request count alongside the per-command counters.
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Stable-dep-of: d7fd1f98607f ("ksmbd: follow SMB2 session expiration semantics")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Tue Jul 7 00:07:09 2026 +0900
ksmbd: fix partial file information responses
[ Upstream commit 6b8b79226bc3e0ac3fdd4e91836241af712e8cd1 ]
Variable-length file information handlers use the client output length
while constructing the response. FILE_ALL_INFORMATION can consequently
return -EINVAL before the common buffer check, while stream information
can stop building the complete result too early.
Build the complete response within the available server response buffer
and apply the client output length only when selecting the final status
and transmitted length. Use the protocol-defined fixed sizes for all,
alternate-name, and stream information to distinguish
STATUS_INFO_LENGTH_MISMATCH from STATUS_BUFFER_OVERFLOW.
This fixes smb2.getinfo.qfile_buffercheck.
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Stable-dep-of: 9fa26285ae70 ("ksmbd: keep compound responses on query info errors")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Wed Sep 9 09:58:22 2026 +0900
ksmbd: fix partial normalized name responses
commit f4fafaf02174c32bce2f9bb4196fadf13f1fd96e upstream.
Windows may request FILE_NORMALIZED_NAME_INFORMATION with an output
buffer that only fits the fixed portion of the variable-length response.
Treat the fixed portion as FILE_NORMALIZED_NAME_INFORMATION_SIZE so ksmbd
returns STATUS_BUFFER_OVERFLOW instead of STATUS_INFO_LENGTH_MISMATCH.
This avoids rejecting valid partial normalized-name responses.
Fixes: 6b8b79226bc3 ("ksmbd: fix partial file information responses")
Reported-by: Mobin Aydinfar <mobin@mobintestserver.ir>
Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Thu Sep 3 14:06:02 2026 +0900
ksmbd: follow SMB2 session expiration semantics
[ Upstream commit d7fd1f98607f2cd358e583f9548bdf0173090a86 ]
ksmbd_session_register() destroys valid sessions after ten seconds of
inactivity whenever a client starts another SessionSetup exchange. This
confuses Session.IdleTime with Session.ExpirationTime. Windows can create
additional authenticated sessions on an existing connection, so deleting
the older session invalidates its tree connects and makes mapped drives
fail with STATUS_NETWORK_NAME_DELETED.
The session expiration rules require the server to change a valid session
to expired only after its credential expiration time passes. A valid or
expired session otherwise keeps its connection from being scavenged.
Connections that have not negotiated a dialect, have no sessions, or have
only InProgress sessions are disconnected after an implementation-specific
timeout. Use the Windows-compatible 45 second value and run the expiration
check periodically for both TCP and SMB Direct.
Keep zero as an infinite credential expiration time, and count each
Valid-to-Expired transition. Set expired sessions to InProgress when they
reauthenticate. If authentication fails, remove the session from the
global and per-connection tables immediately, including SMB3 multichannel
connections.
Retain protection against abandoned SessionId-zero exchanges by allowing
only one InProgress authentication per connection. Additional exchanges
fail with STATUS_INSUFFICIENT_RESOURCES, and stale InProgress sessions are
reaped after the same 45 second setup timeout. This bounds the original
unauthenticated memory-exhaustion path without evicting established
sessions.
Fixes: ea174a918939 ("ksmbd: destroy expired sessions")
Reported-by: Mobin Aydinfar <mobin@mobintestserver.ir>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Wed Sep 9 09:58:48 2026 +0900
ksmbd: keep compound responses on query info errors
[ Upstream commit 9fa26285ae70ac2d3d1b47459a6b4463ab053e1c ]
Do not reset the RFC1002 length of the complete response when a query
info buffer is too small. The current command will add its error response
through ksmbd_iov_pin_rsp(), while resetting the base length can truncate
earlier responses in a compound request.
This lets ksmbd return the earlier responses and the query-info error
response together. Remove the now-unused rsp_org parameter from the pipe
query-info helpers.
Fixes: e2b76ab8b5c9 ("ksmbd: add support for read compound")
Reported-by: Mobin Aydinfar <mobin@mobintestserver.ir>
Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Sun Jul 5 22:43:46 2026 +0900
ksmbd: return buffer overflow for partial filesystem info
[ Upstream commit 0ecd35fac4b4f2828490689b46039744d201dcb0 ]
The query-info buffer check returns STATUS_INFO_LENGTH_MISMATCH for
every output buffer smaller than the complete response. Variable-length
filesystem information instead requires STATUS_BUFFER_OVERFLOW when the
fixed portion fits but the complete data does not.
Pass the fixed size for each filesystem information class to the buffer
checker. Keep INFO_LENGTH_MISMATCH for buffers below that size, and
return BUFFER_OVERFLOW with a response truncated to the requested length
for larger partial buffers.
This fixes smb2.getinfo.qfs_buffercheck.
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Stable-dep-of: 9fa26285ae70 ("ksmbd: keep compound responses on query info errors")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Amit Machhiwal <amachhiw@linux.ibm.com>
Date: Tue Sep 15 22:04:16 2026 +0530
KVM: PPC: Book3S HV: fix secure device page leak on uv_page_in() failure
[ Upstream commit 0a416ee20bcccddf91ca5b63696a23b9d11d73aa ]
In kvmppc_svm_page_in(), if uv_page_in() fails after
kvmppc_uvmem_get_page() has succeeded, the secure device page is never
released. kvmppc_uvmem_get_page() sets a bit in kvmppc_uvmem_bitmap,
allocates a kvmppc_uvmem_page_pvt struct, marks the GFN as
KVMPPC_GFN_UVMEM_PFN, and calls zone_device_page_init() which sets
refcount=1 and locks the page. The subsequent goto out_finalize skips
the *mig.dst assignment, so migrate_vma_finalize() is a no-op for the
page, and none of those resources are ever reclaimed.
Each occurrence permanently consumes one entry from the firmware-bounded
secure memory pool (kvmppc_uvmem_bitmap), leaks pvt, and leaves the GFN
marked as secure — making it unusable for the lifetime of the VM.
The twin __kvmppc_svm_page_out() already handles the analogous uv_page_out()
failure correctly with unlock_page(dpage); __free_page(dpage). Apply
the same pattern here: unlock_page() followed by put_page(), which
chains through free_zone_device_folio() into kvmppc_uvmem_folio_free()
to clear the bitmap bit, free pvt, and reset the GFN state.
Reachable whenever uv_page_in() returns an error (e.g. UV pool
exhaustion) on any POWER9/10 + Ultravisor/PEF system.
Fixes: ca9f4942670c ("KVM: PPC: Book3S HV: Support for running secure guests")
Reviewed-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Tested-by: R Nageswara Sastry <rnsastry@linux.ibm.com>
Signed-off-by: Amit Machhiwal <amachhiw@linux.ibm.com>
Signed-off-by: Gautam Menghani <gautam@linux.ibm.com>
Signed-off-by: Madhavan Srinivasan <maddy@linux.ibm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Amit Machhiwal <amachhiw@linux.ibm.com>
Date: Tue Sep 15 22:04:15 2026 +0530
KVM: PPC: Book3S HV: fix use-after-free in kvmhv_emulate_tlbie_all_lpid()
[ Upstream commit 51938dfa8a51a4f85328413fca9b6e21f9d2d088 ]
kvmhv_emulate_tlbie_all_lpid() iterates the nested-guest IDR and drops
mmu_lock before calling kvmhv_emulate_tlbie_lpid(), but does not hold a
reference on the kvm_nested_guest pointer obtained from the IDR. A
concurrent vCPU issuing a single-LPID tlbie (is=2, ric=2) can race
through kvmhv_flush_nested() -> kvmhv_remove_nested() -> idr_remove /
--refcnt -> kvmhv_release_nested() -> kfree(gp) in that window, leaving
the iterating vCPU with a dangling pointer. The subsequent
mutex_lock(&gp->tlb_lock) and accesses to gp->shadow_pgtable,
gp->shadow_lpid and gp->l1_host all touch freed memory. The free path
is fully L1-controlled.
Fix this by incrementing gp->refcnt inside the loop before dropping
mmu_lock, mirroring what kvmhv_get_nested() does, and releasing the
reference with kvmhv_put_nested() after the per-guest work completes.
This is the same get/put discipline already used at every other
call site that drops mmu_lock while holding a nested-guest pointer.
Fixes: e3b6b4661527 ("KVM: PPC: Book3S HV: Implement H_TLB_INVALIDATE hcall")
Reviewed-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Tested-by: R Nageswara Sastry <rnsastry@linux.ibm.com>
Signed-off-by: Amit Machhiwal <amachhiw@linux.ibm.com>
Signed-off-by: Gautam Menghani <gautam@linux.ibm.com>
Signed-off-by: Madhavan Srinivasan <maddy@linux.ibm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Date: Fri Sep 25 16:37:14 2026 +0200
Linux 7.2.8
Link: https://lore.kernel.org/r/20260923140644.756254324@linuxfoundation.org
Tested-by: Brett A C Sheffield <bacs@librecast.net>
Tested-by: Ronald Warsow <rwarsow@gmx.de>
Tested-by: Florian Fainelli <florian.fainelli@broadcom.com>
Tested-by: Peter Schneider <pschneider1968@googlemail.com>
Tested-by: Salvatore Bonaccorso <carnil@debian.org>
Tested-by: Ron Economos <re@w6rz.net>
Tested-by: Pavel Machek (CIP) <pavel@nabladev.com>
Tested-by: Takeshi Ogasawara <takeshi.ogasawara@futuring-girl.com>
Tested-by: Markus Reichelt <lkt+2023@mareichelt.com>
Tested-by: Justin M. Forbes <jforbes@fedoraproject.org>
Tested-by: Wentao Guan <guanwentao@uniontech.com>
Tested-by: Benjamin Boortz <bennib@mailbox.org>
Tested-by: Barry K. Nathan <barryn@pobox.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shakeel Butt <shakeel.butt@linux.dev>
Date: Fri Aug 28 19:32:51 2026 -0700
memcg: avoid charging the root memcg from obj_cgroup_charge_pages()
commit 6e673d0879ef78c395cfe0d3ba316690a60055d8 upstream.
obj_cgroup_charge_pages() resolves the objcg to its memcg and calls
try_charge_memcg(), which does not short circuit the root memcg. That
memcg can be the root memcg: obj_cgroup_is_root() reflects the memcg the
objcg was created for and is never updated, while memcg_reparent_objcgs()
does redirect objcg->memcg to the parent on rmdir. An objcg of a dying
child of root therefore passes every obj_cgroup_is_root() filter but
resolves to the root memcg.
Folios keep the objcg they were charged with, so this is easy to reach
through zswap: allocate anon memory in a cgroup, move the task out, remove
the cgroup, then write to the root cgroup's memory.reclaim. The reclaimed
folios are charged through the reparented objcg and end up in
refill_stock() with the root memcg:
WARNING: mm/memcontrol.c:2198 at refill_stock+0x644/0x940
refill_stock+0x644/0x940
try_charge_memcg+0x12d6/0x1570
__obj_cgroup_charge+0x35/0xf0
obj_cgroup_charge+0x1de/0x210
obj_cgroup_charge_zswap+0x83/0x270
zswap_store+0x1620/0x2000
swap_writeout+0x94c/0x14c0
shrink_folio_list+0x3388/0x52b0
[...]
try_to_free_mem_cgroup_pages+0x30d/0x830
user_proactive_reclaim+0x504/0x840
memory_reclaim+0x1f/0x30
Beyond the warning, the charge is asymmetric: obj_cgroup_uncharge_pages()
skips refill_stock() for the root memcg, so the root's page counter grows
and is never uncharged. It is not user visible, since memory.current is
not exposed on the root, but it is a leak.
Use try_charge(), which returns early for the root memcg, restoring the
symmetry with obj_cgroup_uncharge_pages().
The above sequence was scripted into a standalone reproducer (zswap on,
swap on a virtio disk, 512MB of anon memory faulted in inside a child of
the root cgroup, the task then migrated to the root cgroup, the child
removed, followed by "echo 600M swappiness=max > memory.reclaim" on the
root) and run in a CONFIG_DEBUG_VM=y VM. It reproduces the splat on the
first zswap store of a reparented folio, with the same call chain as the
report. With this patch applied the splat is gone while the zswap store
count over the run is unchanged, so the same path is still exercised.
cgroup selftests test_zswap, test_kmem and test_memcontrol show no new
failures.
Link: https://lore.kernel.org/20260829023251.474083-1-shakeel.butt@linux.dev
Fixes: 20d6c1725228 ("memcg: avoid refill_stock for root memcg")
Signed-off-by: Shakeel Butt <shakeel.butt@linux.dev>
Reported-by: Farhad Alemi <farhad.alemi@berkeley.edu>
Closes: https://lore.kernel.org/all/CA+0ovCgWzUMK+nNbbtH7eV65Ca=fDN4Ozu7iASgryjvv8Tk8zQ@mail.gmail.com/
Reviewed-by: Muchun Song <muchun.song@linux.dev>
Reviewed-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Yifei Gao <gyf161023@gmail.com>
Date: Tue Aug 4 21:34:56 2026 +0000
memstick: ms_block: destroy io_queue workqueue on removal
commit 90af7fde083e1b22c349c3a8b1626728e44e474c upstream.
msb_init_disk() creates the per-card ordered workqueue msb->io_queue with
alloc_ordered_workqueue(). It is torn down with destroy_workqueue() only
on the init error path; msb_remove() never destroys it. msb_stop() merely
flushes the queue, and neither msb_data_clear() nor put_disk() free it. As
a result every card insert/remove cycle leaks the workqueue and its
kworker, exhausting kernel memory over repeated cycles.
Destroy the workqueue in msb_remove() after the disk has been removed and
the queue drained.
Fixes: 0ab30494bc4f ("memstick: add support for legacy memorysticks")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Yifei Gao <gyf161023@gmail.com>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Julian Braha <julianbraha@gmail.com>
Date: Sun Aug 16 00:51:09 2026 +0100
mips: econet: fix unmet dependencies for ECONET
[ Upstream commit fb5a08026a68be0ec5f660588311fb1091bf5d58 ]
ECONET selects EARLY_PRINTK_8250, SERIAL_OF_PLATFORM, and SERIAL_8250
without ensuring their dependencies, EARLY_PRINTK and TTY are met. This
causes unmet dependencies:
WARNING: unmet direct dependencies detected for SERIAL_8250
Depends on [n]: TTY [=n] && HAS_IOMEM [=y] && !S390
Selected by [y]:
- ECONET [=y]
WARNING: unmet direct dependencies detected for EARLY_PRINTK_8250
Depends on [n]: EARLY_PRINTK [=n] && USE_GENERIC_EARLY_PRINTK_8250 [=y]
Selected by [y]:
- ECONET [=y]
WARNING: unmet direct dependencies detected for SERIAL_OF_PLATFORM
Depends on [n]: TTY [=n] && HAS_IOMEM [=y] && SERIAL_8250 [=y] && OF [=y]
Selected by [y]:
- ECONET [=y]
However, in the discussion of v1 of this patch, Arnd pointed out that these
selects don't belong here in the first place. [1]
Let's remove them to resolve the unmet dependencies.
These unmet dependencies were found by kconfirm, a static analysis tool for
Kconfig.
Fixes: 79ee1d20e37c ("mips: econet: Fix incorrect Kconfig dependencies")
Fixes: 35fb26f94dfa ("mips: Add EcoNet MIPS platform support")
Suggested-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Julian Braha <julianbraha@gmail.com>
Tested-by: Caleb James DeLisle <cjd@cjdns.fr>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Orgad Shaneh <orgads@gmail.com>
Date: Tue Sep 1 19:33:55 2026 +0000
MIPS: Octeon: apply USB FDT fixups also when USB is modular
[ Upstream commit 126f16e0a1b353c2ba5c7e2c8626cfa865934f9f ]
The uctl/usbn device-tree fixups in octeon_prune_device_tree() - which
set the board's USB reference-clock frequency and type from
__cvmx_helper_board_usb_get_clock_type() - are guarded by
"#ifdef CONFIG_USB", which is false when USB is built as a module. The
fixups then silently disappear and octeon-hcd sees whatever default the
DTS carries (12MHz crystal in octeon_3xxx.dts), leaving the PHY dead or
the bus erroring on boards with a different reference clock.
Use IS_ENABLED() so USB=m gets the same fixups as USB=y.
Fixes: 7fd57ab9d9cf ("MIPS: Octeon: Fix compile error when USB is not enabled.")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Orgad Shaneh <orgads@gmail.com>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Théo Lebrun <theo.lebrun@bootlin.com>
Date: Fri Aug 28 16:07:33 2026 +0200
mips: select CONFIG_WEAK_REORDERING_BEYOND_LLSC from CONFIG_EYEQ
commit bbc448c541ed90d52f9ec22f17394304e3bf170e upstream.
On I6500 CPU cores, lld and scd give no ordering guarantees (same as all
other instructions). To respect the assumption that arch_cmpxchg() is
fully ordered, we must inject sync instructions above and below our
lld/scd loops using the already in place WEAK_REORDERING_BEYOND_LLSC
infrastructure.
Otherwise, bad things can happen:
[ 34.054496] CPU 3 Unable to handle kernel paging request at virtual address 0000000000000000, epc == a80000080838e01c, ra == a80000080838dfc4
[ 34.054559] Oops[#1]:
[ 34.069561] CPU: 3 UID: 0 PID: 170 Comm: pipe_race Not tainted 7.2.0-rc6-01553-gb73c35220968-dirty #103 VOLUNTARY
[ 34.079932] Hardware name: Mobile EyeQ5 MP5 Evaluation board
[ 34.085592] $ 0 : 0000000000000000 0000000000000001 0000000000000000 0000000000000000
[ 34.093616] $ 4 : a800000808ee2618 000000000b7a879d 0000000000001000 0000000000000000
[ 34.101638] $ 8 : 0000000000e3f2c9 0000000000000000 a800000808a2a9f8 0000000000000000
[ 34.109660] $12 : a8000008139ffcd8 ffffffff84080018 a80000080837fae0 7878787878787878
[ 34.117682] $16 : a800000807e82940 0000000000001000 0000000000000000 0000000000000000
[ 34.125704] $20 : a800000802920e00 a8000008139ffdf8 a800000802649400 0000000000e3f2c9
[ 34.133726] $24 : 0000000000000006 00000001200406e0
[ 34.141783] $28 : a8000008139fc000 a8000008139ffd10 0000000000e3f2c8 a80000080838dfc4
[ 34.149837] epc : a80000080838e01c anon_pipe_read+0xd4/0x428
[ 34.155697] ra : a80000080838dfc4 anon_pipe_read+0x7c/0x428
[ 34.161549] Status: 140000e3 KX SX UX KERNEL EXL IE
[ 34.166551] Cause : 40800408 (ExcCode 02)
[ 34.170574] BadVA : 0000000000000000
[ 34.174161] PrId : 0001b028 (MIPS I6500)
[ 34.178183] Process pipe_race (pid: 170, threadinfo=000000005ca35720, task=00000000e1013890, tls=000000014ebbb780)
[ 34.188568] Stack : a800000802649400 0000000000000000 0000000000000000 a8000008139ffdd0
[ 34.196623] 0000000000000fba a800000808ee0000 0000000000000001 a8000008130c3e80
[ 34.204676] a8000008080d1280 a8000008139ffd58 a8000008139ffd58 1dbd2b22ea1dd500
[ 34.212729] a800000802649400 a800000808ee0000 ffffffffffffffea 0000000000000001
[ 34.220783] 0000000000001000 0000000000000000 00000001200ae518 ffffffffffffffff
[ 34.228836] 000000fffbe0e530 a80000080837edf4 000000fffbe0e530 0000000000000000
[ 34.236890] 0000000000000000 0000000000000000 000000014ebb55a0 0000000000001000
[ 34.244943] 0000000000000001 a800000802649400 0000000000000000 0000000000000000
[ 34.252996] 0000000000000000 0000400400000000 0000000000000000 1dbd2b22ea1dd500
[ 34.261049] 00000000140000e3 a800000802649400 a800000802649400 a800000808ee0000
[ 34.269103] ...
[ 34.271568] Call Trace:
[ 34.274026] [<a80000080838e01c>] anon_pipe_read+0xd4/0x428
[ 34.279533] [<a80000080837edf4>] vfs_read+0x25c/0x318
[ 34.284607] [<a80000080837faac>] ksys_read+0x104/0x138
[ 34.289763] [<a80000080802b9cc>] syscall_common+0x44/0x68
[ 34.295187]
[ 34.296689] Code: f84000cf 02209825 de020010 <dc420000> d8400004 02002825 0040f809 02802025 f84000c3
[ 34.306504]
[ 34.308099] ---[ end trace 0000000000000000 ]---
My initial reproducer was the xdp-tools test suite. A standalone
reproducer would be an lld/scd loop that, when the read is reordered by
the CPU, triggers a fault. We can achieve this from userspace by
stressing an anonymous pipe, which uses a mutex. Program used:
// SPDX-License-Identifier: GPL-2.0
// pipe_race.c - reproducer for MIPS LL/SC reordering vs fs/pipe.c
//
// Two userspace processes on an anonymous pipe:
// parent = writer: tight write() loop
// child = reader: tight read() loop
#define _GNU_SOURCE
#include <assert.h>
#include <errno.h>
#include <sched.h>
#include <signal.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/types.h>
#include <sys/wait.h>
#include <time.h>
#include <unistd.h>
int main(void)
{
long wrsize = 70; // bytes per write()
long rdsize = 4096; // bytes per read()
int pfd[2];
char *buf;
pid_t pid;
int ret;
ret = pipe(pfd);
assert(!ret);
pid = fork();
assert(pid >= 0);
if (pid == 0) { /* reader */
close(pfd[1]);
buf = malloc(rdsize);
assert(buf);
for (;;) {
ssize_t n = read(pfd[0], buf, rdsize);
if (n < 0 && errno == EINTR)
continue;
if (n <= 0)
_exit(n < 0 ? 1 : 0);
}
}
close(pfd[0]); /* writer */
buf = malloc(wrsize);
assert(buf);
memset(buf, 'x', wrsize);
for (;;) {
ssize_t n = write(pfd[1], buf, wrsize);
if (n < 0 && errno == EINTR)
continue;
if (n != wrsize)
break;
}
kill(pid, SIGKILL);
wait(NULL);
return 0;
}
Fixes: fbe0fae601b7 ("MIPS: mobileye: Add EyeQ6H support")
Cc: stable@vger.kernel.org
Signed-off-by: Théo Lebrun <theo.lebrun@bootlin.com>
Reviewed-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Nhat Pham <nphamcs@gmail.com>
Date: Fri Aug 28 12:14:33 2026 -0700
mm, swap: fix SWAP_USAGE_OFFLIST_BIT collision with real usage count
commit 12e9ac7bc5b254048f886bf421e3a15491106c1f upstream.
SWAP_USAGE_OFFLIST_BIT is embedded in the si->inuse_pages usage counter,
and is meant to sit above any value that counter can reach. However, it
is defined from BITS_PER_TYPE(atomic_t), so it is bit 30. On a system
with 4 KiB pages the flag collides with the usage count once that count
reaches 4 TiB.
swap_usage_in_pages() masks bit 30 out, so whenever the real count has
that bit set, every caller of it reads 4 TiB low:
* /proc/swaps understates Used by 4 TiB.
* A raw count of exactly 2^30 masks to zero, so try_to_unuse() takes its
"if (!swap_usage_in_pages(si)) goto success;" early exit and swapoff
tears the device down while pages are still swapped out. Nothing in
the rest of swapoff aborts the teardown, so those pages are lost.
Independently of swapoff, the collision also corrupts the counter and the
plist. On a device in normal use, a free that leaves bit 30 set in the
count makes swap_usage_sub() see the flag where there is only count, and
call add_to_avail_list(). It clears the bit with
fetch_and(~SWAP_USAGE_OFFLIST_BIT), leaving the stored count 4 TiB below
the real one, and calls plist_add() on a device that is already listed,
tripping the WARN_ON(!plist_node_empty(node)) in plist_add() and linking
the node a second time.
Change the definition of SWAP_USAGE_OFFLIST_BIT to be based on
atomic_long_t instead. Note that the usage counter field itself is of
this same type, so it is still a valid bit.
Link: https://lore.kernel.org/20260828191433.3304458-1-nphamcs@gmail.com
Fixes: b228386cf237 ("mm, swap: clean up plist removal and adding")
Signed-off-by: Nhat Pham <nphamcs@gmail.com>
Reported-by: Sashiko <sashiko-bot@kernel.org>
Closes: https://sashiko.dev/#/patchset/20260825153238.2695446-1-nphamcs%40gmail.com
Suggested-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Kairui Song <kasong@tencent.com>
Cc: Baoquan He <baoquan.he@linux.dev>
Cc: Barry Song <baohua@kernel.org>
Cc: Chris Li <chrisl@kernel.org>
Cc: Gregory Price <gourry@gourry.net>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Joshua Hahn <joshua.hahnjy@gmail.com>
Cc: Kemeng Shi <shikemeng@huaweicloud.com>
Cc: Shakeel Butt <shakeel.butt@linux.dev>
Cc: Youngjun Park <youngjun.park@lge.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Lorenzo Stoakes (ARM) <ljs@kernel.org>
Date: Thu Aug 27 20:55:57 2026 +0100
mm/huge_memory: bypass THP tuneables for huge pfnmap mappings
commit e384abeb559d10d6505aec053ede9368d81d4c71 upstream.
The sysfs THP tuneables at /sys/kernel/mm/transparent_huge_pages/ rather
confusingly only control the behaviour of THP in some instances.
They are not applicable to MADV_COLLAPSE operations, nor to DAX mappings.
Long-term, THP is predicated upon compaction being able to obtain large
folios to populate THP ranges.
However, vm_normal_folio() returns NULL for PFN map mappings, thus their
reference count is maintained by the driver, not core mm.
As a consequence, the folios are not subject to reclaim nor compaction, so
are not truly part of the THP mechanism at all.
However, since commit 5dd40721f147 ("mm: allow THP orders for PFNMAPs")
introduced the ability to establish huge PFN maps, they have been subject
to THP tuneables.
This is incorrect - if a huge PFN map is available (defined by
vma->vm_ops->huge_fault being non-NULL for a VMA_PFNMAP_BIT VMA), then it
should be mapped huge upon fault-in.
Correct this by explicitly checking for this while ensuring that smaps
continues to accurately report THPeligible statistics.
While here, abstract the entire file-backed THP check in
vma_can_map_huge_file(), with sensible separation of logic into helper
functions.
Note that drm_gem_shmem_mmap() and panthor_gem_mmap() establish huge PFN
maps of shmem folios, however they are marked unevictable in
drm_gem_get_pages(), and in any case would fail the reference check in
__remove_mapping() even if they weren't.
Failing to map huge PFN maps has resulted in significant real-world
performance degradation, see links for details.
[ziy@nvidia.com: rename some functions]
Link: https://lore.kernel.org/DL1HIHWYJ7TB.1CY76SJS0V03L@nvidia.com
Link: https://lore.kernel.org/20260827-hugepfn-allowable-orders-v1-1-94819c8807c8@kernel.org
Fixes: 5dd40721f147 ("mm: allow THP orders for PFNMAPs")
Signed-off-by: Lorenzo Stoakes (ARM) <ljs@kernel.org>
Signed-off-by: Zi Yan <ziy@nvidia.com>
Reported-by: Cedric Le Goater <clg@redhat.com>
Closes: https://lore.kernel.org/linux-mm/20260805055544.1568534-1-clg@redhat.com/
Reported-by: Saravanan D <saravanand@crusoe.ai>
Closes: https://lore.kernel.org/linux-mm/20260821070520.25759-1-saravanand@crusoe.ai/
Reviewed-by: Zi Yan <ziy@nvidia.com>
Tested-by: Saravanan D <saravanand@crusoe.ai>
Tested-by: Lance Yang <lance.yang@linux.dev>
Reviewed-by: SJ Park <sj@kernel.org>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: Barry Song <baohua@kernel.org>
Cc: David Hildenbrand <david@kernel.org>
Cc: Dev Jain <dev.jain@arm.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Liam R. Howlett <liam@infradead.org>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shakeel Butt <shakeel.butt@linux.dev>
Date: Tue Sep 1 11:01:09 2026 -0700
mm/mlock: use the IRQ-safe accessor for NR_MLOCK in __munlock_folio()
commit e14a3454806468b086fe2e4ca2e1bff95b528531 upstream.
NR_MLOCK is updated from interrupt context. __free_pages_prepare() clears
a stray PG_mlocked and adjusts NR_MLOCK, and a folio can reach it with the
flag still set from a bio completion handler:
__free_pages_ok+0x6af/0x7a0
<IRQ>
__bio_release_pages+0xde/0x260
__iomap_dio_bio_end_io+0x16e/0x1a0
blk_update_request+0x14b/0x3d0
blk_mq_end_request+0x18/0x30
blk_done_softirq+0x49/0x60
The folio gets there like this. A MAP_SHARED file mapping is mlocked, so
its page cache folios carry PG_mlocked, and an O_DIRECT write sourced from
that mapping GUP-pins those same folios. munlock() then runs
mlock_vma_pages_range(), which clears VM_LOCKED before walking the page
tables to munlock each folio. A concurrent hole punch reaches the folio
through the rmap (i_mmap_rwsem, not mmap_lock) and can land inside that
window: __folio_remove_rmap() -> munlock_vma_folio() sees VM_LOCKED
already clear, so it neither queues the folio on the mlock batch nor takes
a reference, and the pte it clears makes the pending mlock_pte_range()
walk skip the folio at its !pte_present() check. filemap_remove_folio()
then drops the page cache reference, leaving the bio's pin as the last
one, released from the completion handler above.
So __zone_stat_mod_folio() here needs interrupts disabled, not merely
preemption, and __munlock_folio() has a path where they are not: when the
folio has already been taken off the LRU by somebody else the function
jumps straight to the counter update without taking the lruvec lock. The
read-modify-write of the per-CPU NR_MLOCK diff can then be interrupted by
the softirq above, and one of the two decrements is lost, leaving Mlocked
in /proc/meminfo permanently overstated.
Use zone_stat_mod_folio(). mod_zone_state()'s this_cpu_try_cmpxchg() is
atomic against a same-CPU interrupt and retries, and on the path where the
lruvec lock is held its cost is negligible next to the lock itself.
The UNEVICTABLE_PG* events are deliberately left on the __ accessors: they
occupy different vm_event_states slots from the UNEVICTABLE_PGCLEARED that
__free_pages_prepare() bumps, and nothing updates those two from interrupt
context.
Link: https://lore.kernel.org/20260901180109.3797944-1-shakeel.butt@linux.dev
Fixes: 2fbb0c10d1e8 ("mm/munlock: mlock_page() munlock_page() batch by pagevec")
Signed-off-by: Shakeel Butt <shakeel.butt@linux.dev>
Reported-by: syzbot+cd2073ee6d958a8d0fcd@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/linux-mm/6a931c5a.08e933ee.dbf97.0093.GAE@google.com/
Acked-by: Hugh Dickins <hughd@google.com>
Cc: Jann Horn <jannh@google.com>
Cc: Liam R. Howlett <liam@infradead.org>
Cc: Lorenzo Stoakes <ljs@kernel.org>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Pedro Falcato <pfalcato@suse.de>
Cc: Vlastimil Babka <vbabka@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Lorenzo Stoakes (ARM) <ljs@kernel.org>
Date: Fri Aug 28 12:20:37 2026 +0100
mm/mremap: account mm->locked_vm correctly for MREMAP_DONTUNMAP
commit 397432cab17bccb600fd6c16ed593f1149042268 upstream.
When a VMA is mremap()'d with MREMAP_DONTUNMAP set, that results in the
VMA being copied, but the source VMA not being unmapped.
If the VMA is mlock()'d this is a legal operation, though the source VMA
has its VMA_LOCKED_BIT cleared.
However this is done in dontunmap_complete(), after mm->locked_vm was
incremented via vrm_stat_account(), resulting in double-counting.
Worse, this is not even corrected when source VMA is unmapped, due to the
VMA_LOCKED_BIT flag having been cleared.
This all works fine in the usual mremap() case (without MREMAP_DONTUNMAP),
as the source VMA is unmapped with VMA_LOCKED_BIT intact, at which time
mm->locked_vm is decremented accordingly.
Resolve the issue by invoking vrm_stat_account() only after
dontunmap_complete() has run.
Note that MREMAP_DONTUNMAP requires old_len == new_len, so no need to
account for a delta in size in this case.
The bug was introduced by commit b714ccb02a76 ("mm/mremap: complete
refactor of move_vma()") which incorrectly reordered the accounting and
the clearing of the VMA_LOCKED_BIT flag.
Link: https://lore.kernel.org/20260828-mremap-fix-locked-vm-v1-1-c80be7505d1e@kernel.org
Fixes: b714ccb02a76 ("mm/mremap: complete refactor of move_vma()")
Signed-off-by: Lorenzo Stoakes (ARM) <ljs@kernel.org>
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Closes: https://sashiko.dev/#/patchset/20260825-fix-mremap-dontunmap-pgoff-v1-1-39a40b2c98b3@kernel.org
Reported-by: Kunwu Chan <kunwu.chan@gmail.com>
Closes: https://lore.kernel.org/all/20260828094823.594279-1-kunwu.chan@linux.dev/
Acked-by: Vlastimil Babka (SUSE) <vbabka@kernel.org>
Tested-by: Kunwu Chan <kunwu.chan@gmail.com>
Reviewed-by: Kunwu Chan <kunwu.chan@gmail.com>
Cc: Jann Horn <jannh@google.com>
Cc: Liam R. Howlett <liam@infradead.org>
Cc: Pedro Falcato <pfalcato@suse.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Jiayuan Chen <jiayuan.chen@linux.dev>
Date: Wed Sep 2 15:37:59 2026 +0800
mm/shrinker: fix bogus set_shrinker_bit() with cgroup.memory=nokmem
commit 932cfb25e7ce98d1f93895671ec186a3087e4f80 upstream.
With cgroup.memory=nokmem, shrinker_memcg_alloc() bails out early and
never allocates an id, so shrinker->id keeps the 0 it got from the
kzalloc() in shrinker_alloc(). __list_lru_init() then copies that 0 into
lru->shrinker_id, where it looks like a valid bit index.
Nothing calls expand_shrinker_info() on nokmem either, so shrinker_nr_max
stays 0 and every memcg ends up with an empty map (map_nr_max == 0).
deferred_split_folio() hands a real memcg to __list_lru_add() regardless
of whether the lru is memcg aware, so the first THP queued in a cgroup
does set_shrinker_bit(memcg, nid, 0) and trips the bounds check:
WARNING: mm/shrinker.c:212 at set_shrinker_bit+0x7d/0x90, CPU#126
Call Trace:
<TASK>
deferred_split_folio+0x18c/0x220
map_anon_folio_pmd_nopf+0xdd/0x130
map_anon_folio_pmd_pf+0x14/0xb0
do_huge_pmd_anonymous_page+0x1a1/0x620
__handle_mm_fault+0xea9/0x10d0
handle_mm_fault+0xe5/0x320
do_user_addr_fault+0x1cc/0x870
exc_page_fault+0x81/0x1b0
asm_exc_page_fault+0x27/0x30
</TASK>
Harmless, the WARN_ON_ONCE() is what keeps the out of bounds unit[] read
from happening, but the id should not look valid in the first place.
Clear it before returning.
Two other spots could paper over this: drop the id in __list_lru_init()
when nokmem turns memcg_aware off, or make deferred_split_folio() pass
NULL like list_lru_add_obj() does. Both leave shrinker->id lying around
for the next caller, so fix it where the id is handed out.
Link: https://lore.kernel.org/20260902073800.305481-1-jiayuan.chen@linux.dev
Fixes: fafaeceb89a5 ("mm: switch deferred split shrinker to list_lru")
Signed-off-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Acked-by: Shakeel Butt <shakeel.butt@linux.dev>
Cc: Usama Arif <usama.arif@linux.dev>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Kairui Song <kasong@tencent.com>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Lorenzo Stoakes (ARM) <ljs@kernel.org>
Date: Wed Sep 2 19:08:08 2026 +0100
mm/vma: correctly unaccount on mmap_prepare() failure
commit 6cc27d82196385fe06853319f74312a7d8019726 upstream.
__mmap_setup() accounts memory for relevant mappings via:
security_vm_enough_memory_mm()
-> __vm_enough_memory()
-> vm_acct_memory()
If __mmap_setup() fails, this indicates that this accounting did not take
place, and thus it's appropriate for __mmap_region() to jump to
abort_munmap.
However if call_mmap_prepare() fails, it also jumps there and any accounted
memory is not correctly unaccounted.
Fix this by handling each error separately.
Link: https://lore.kernel.org/20260902-fix-unaccount-mmap_prepare-v1-1-ea070189fdfb@kernel.org
Fixes: c84bf6dd2b83 ("mm: introduce new .mmap_prepare() file callback")
Signed-off-by: Lorenzo Stoakes (ARM) <ljs@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Liam R. Howlett <liam@infradead.org>
Cc: Pedro Falcato <pfalcato@suse.de>
Cc: Vlastimil Babka <vbabka@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Wenjie Qi <qiwenjie@xiaomi.com>
Date: Sun Aug 30 01:36:12 2026 +0800
mm: filemap: retain mapped dropbehind folios
commit 848d2ce2fce15fbdc083fbf9691bfa72911033c4 upstream.
Fault-around can map ready dropbehind folios without going through the
normal page-cache lookup that clears dropbehind. A mapping represents a
competing cached user, so retain the folio instead of forcibly unmapping
it when writeback completes.
For a mapped folio, folio_unmap_invalidate() can call
unmap_mapping_folio(), which takes i_mmap_rwsem and may sleep. Retaining
mapped folios avoids this path when folio_end_dropbehind() runs in
non-preemptible task context.
Tal was able to trigger a sleeping-in-atomic warning due to this [1].
Unmapped dropbehind folios continue through the existing invalidation path.
Link: https://lore.kernel.org/4aba05e1a2c3b61cb337d373eb9b7a8db4ddd822.1788024049.git.qiwenjie@xiaomi.com
Link: https://lore.kernel.org/076bb01b-6fcf-4691-be8c-0e8507c9fe64@columbia.edu [1]
Fixes: fb7d3bc41493 ("mm/filemap: drop streaming/uncached pages when writeback completes")
Signed-off-by: Wenjie Qi <qiwenjie@xiaomi.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Tal Zussman <tz2294@columbia.edu>
Tested-by: Tal Zussman <tz2294@columbia.edu>
Cc: Barry Song <baohua@kernel.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Trond Myklebust <trond.myklebust@hammerspace.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Meijing Zhao <zhaomeijing@lixiang.com>
Date: Wed Sep 2 15:59:44 2026 +0800
mm: memblock: show all region flags in debugfs
[ Upstream commit e2d5b01f878d76bd1142e512a0b979a1d3cd0abf ]
Commit 493f349e38d0 ("memblock: Add flags and nid info in memblock
debugfs") made memblock_debug_show() stop after finding the first set
flag. A memblock region can carry multiple flags, so the remaining flags
are hidden from debugfs.
Walk all bits in the region flags and print every set flag separated by
"|". Keep walking beyond flagname[] so that a set flag without a known
name is reported as UNKNOWN rather than silently ignored.
Fixes: 493f349e38d0 ("memblock: Add flags and nid info in memblock debugfs")
Signed-off-by: Meijing Zhao <zhaomeijing@lixiang.com>
Link: https://patch.msgid.link/20260902075944.3742866-1-zhaomeijing100@gmail.com
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Fan Wu <fanwu01@zju.edu.cn>
Date: Thu Aug 6 13:02:33 2026 +0000
mmc: core: Cancel SDIO IRQ work before freeing host
commit 6feadbecdae60a6324c967f3b1493741083793a3 upstream.
A host controller that uses sdio_signal_irq() schedules host->sdio_irq_work
from its interrupt handler. That work is only cancelled on the suspend
path (mmc_sdio_suspend()), not on the remove/free path, so a worker armed
just before the controller freed its IRQ can run after
mmc_host_classdev_release() has freed the host and dereference it through
container_of().
Cancel host->sdio_irq_work in mmc_free_host(), like the existing
host->detect drain added by commit 1036f69e2513 ("mmc: core: Cancel
delayed work before releasing host").
This issue was found by an in-house static analysis tool.
Fixes: 682696605c70 ("mmc: sdio: Add API to manage SDIO IRQs from a workqueue")
Cc: stable@vger.kernel.org
Assisted-by: Codex:gpt-5.6
Signed-off-by: Fan Wu <fanwu01@zju.edu.cn>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Zhu Ling <zhuling0805@qq.com>
Date: Fri Sep 4 17:07:46 2026 +0800
mmc: core: Fix OF node reference leak on card add failure
commit 08b54e16d547d5c1aa61bf7a3595bb1620975eeb upstream.
mmc_of_find_child_device() returns a device node with its reference count
incremented. mmc_add_card() stores the reference before calling
device_add(), while the card is marked present only after device_add()
succeeds.
If device_add() fails, the callers release the card through
mmc_remove_card(). However, mmc_remove_card() only drops the OF node
reference for a present card, leaking the reference on this error path.
Move of_node_put() outside the present-card conditional so the reference
is released for both registered cards and card-add failures.
Fixes: 25185f3f31c9 ("mmc: Add SDIO function devicetree subnode parsing")
Cc: stable@vger.kernel.org
Signed-off-by: Zhu Ling <zhuling0805@qq.com>
Reviewed-by: Shawn Lin <shawn.lin@linux.dev>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Fan Wu <fanwu01@zju.edu.cn>
Date: Fri Aug 14 08:23:54 2026 +0000
mmc: hsq: Fix use-after-free in retry work
commit 5d132990475f02cfa1debe03d50b479432864ebd upstream.
mmc_hsq_pump_requests() queues retry_work when request_atomic() returns
-EBUSY; today sdhci-sprd is the only consumer that implements
request_atomic(). The work is embedded in a devm-allocated mmc_hsq, but
is never cancelled during driver removal. Work still pending at unbind
can therefore run after the devm allocation has been released and
dereference hsq->mmc and hsq->mrq.
Use devm_work_autocancel() to cancel and drain retry_work before the devm
allocation is released. By the time devres cleanup begins,
mmc_remove_host() has already stopped the host, so no new requests can
arm the work.
This issue was found by an in-house static analysis tool.
Fixes: 6db96e5810e0 ("mmc: host: Introduce the request_atomic() for the host")
Cc: stable@vger.kernel.org
Assisted-by: Codex:gpt-5.6
Signed-off-by: Fan Wu <fanwu01@zju.edu.cn>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Fan Wu <fanwu01@zju.edu.cn>
Date: Fri Aug 14 08:25:50 2026 +0000
mmc: mmci: Fix use-after-free in busy-timeout work
commit 2b19cf3e50cddaff07b657dae1a8f30f06032852 upstream.
ux500_busy_complete() can queue ux500_busy_timeout_work for an R1b
command, but mmci_remove() never cancels it. The work can subsequently
dereference the devm-allocated mmci_host after it has been released.
Mask the controller interrupts and disable the delayed work during
removal. This drains any queued instance and stops an IRQ handler that
is still in progress from queueing the work again once it has been
disabled.
This issue was found by an in-house static analysis tool.
Fixes: b1a665932dc2 ("mmc: mmci: Add support for SW busy-end timeouts")
Cc: stable@vger.kernel.org # v6.10+
Assisted-by: Codex:gpt-5.6
Signed-off-by: Fan Wu <fanwu01@zju.edu.cn>
Reviewed-by: Linus Walleij <linusw@kernel.org>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Fan Wu <fanwu01@zju.edu.cn>
Date: Fri Aug 7 03:26:54 2026 +0000
mmc: mxcmmc: cancel data work and watchdog on remove
commit d3a421c82412344022982d5b91ba23194a0a6f29 upstream.
mxcmci_remove() frees the host through the devm tail, but neither it nor
mmc_remove_host() drains the driver's own asynchronous state.
host->watchdog, a 10 s timer armed on the DMA path in mxcmci_setup_data(),
is deleted only by the DMA- and IRQ-complete paths, which the remove path
does not explicitly drain; it can therefore fire after the host is freed
and dereference it in mxcmci_watchdog(). host->datawork, armed from the
IRQ handler on the PIO path, is not cancelled by the remove path either.
Free the devm-registered IRQ, then cancel datawork and delete the watchdog
in mxcmci_remove(), before dma_release_channel(). Freeing the IRQ first
keeps a trailing handler from re-arming datawork between the cancel and
the host free. Both callbacks are non-self-rearming.
This issue was found by an in-house static analysis tool.
Fixes: f6ad0a481342 ("mmc: mxcmmc: fix bug that may block a data transfer forever")
Cc: stable@vger.kernel.org
Assisted-by: Codex:gpt-5.6
Signed-off-by: Fan Wu <fanwu01@zju.edu.cn>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Florian Maillard <florian.maillard@mailoo.org>
Date: Mon Aug 24 08:57:55 2026 +0200
mmc: rtsx_pci_sdmmc: ignore broken write-protect on ThinkPad X260
commit 9c182bc5d7817437a7d04ab96133f9191846d93d upstream.
The Realtek RTS522A card reader in the Lenovo ThinkPad X260
(subsystem 17aa:504a) incorrectly reports inserted SD cards as
write-protected.
This causes the MMC core to expose the card as read-only:
mmcblk0: mmc0:aaaa SN256 238 GiB (ro)
and /sys/block/mmcblk0/ro reports 1.
Setting MMC_CAP2_NO_WRITE_PROTECT makes the card writable again.
Limit the quirk to the affected Lenovo subsystem.
Assisted-by: ChatGPT:GPT-5.6 Sol
Signed-off-by: Florian Maillard <florian.maillard@mailoo.org>
Cc: stable@vger.kernel.org
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Myeonghun Pak <mhun512@gmail.com>
Date: Sun Sep 13 18:35:38 2026 -0400
mmc: sdhci-of-aspeed: Remove children before releasing SDC resources
commit 4396d70bb7fec531bcf934fed016b2f3300c670b upstream.
Probe failure and removal leave SDHCI child devices registered after the
parent clock and managed resources are released.
Unregister the OF children in reverse order before disabling the parent
clock on both paths. Use of_platform_device_destroy() because manual
child creation does not set the flag required by of_platform_depopulate().
This issue was identified during our ongoing static-analysis research
while reviewing kernel code.
Fixes: bb7b8ec62dfb ("mmc: sdhci-of-aspeed: Add support for the ASPEED SD controller")
Co-developed-by: Ijae Kim <ae878000@gmail.com>
Signed-off-by: Ijae Kim <ae878000@gmail.com>
Signed-off-by: Myeonghun Pak <mhun512@gmail.com>
Assisted-by: OpenAI:GPT-5.6
Cc: stable@vger.kernel.org
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Diogo Ivo (Schneider Electric) <diogo.ivo@bootlin.com>
Date: Fri Aug 7 13:06:59 2026 +0200
mmc: sdhci_am654: Clear ITAPDLY on tuning failure
commit c9f47cc8c37f7659897142ffe216c250fbc1d4ed upstream.
When tuning fails, stale ITAPDLY values can persist and interfere with
subsequent I/O accesses, for example in DDR50 mode in cards with no tuning
support. Move the ITAPDLY enable setting out of the tuning loop to after
successful tuning, and explicitly clear ITAPDLY (delay and enable) when
tuning fails so that we are sure only working values are actually left in
hardware.
Fixes: 901d16e46296 ("mmc: sdhci_am654: Add retry tuning")
Cc: stable@vger.kernel.org
Signed-off-by: Diogo Ivo (Schneider Electric) <diogo.ivo@bootlin.com>
Reviewed-by: Judith Mendez <jm@ti.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Diogo Ivo (Schneider Electric) <diogo.ivo@bootlin.com>
Date: Fri Aug 7 13:07:00 2026 +0200
mmc: sdhci_am654: Fallback to DT-provided itap delay on DDR50 tuning failure
commit 308d05225281d86150d88141990d6caf8c902349 upstream.
DDR50 mode is not required to support the tuning command CMD19, meaning
that calibration may fail on cards that do not implement it, in which
case a known-good itap delay value should be programmed into the host
controller.
Do this by reading the (already defined) itap delay DT property for DDR50
and, if tuning fails for this mode, fall back to the DT-provided itap delay
value. If the DT does not provide a value for DDR50 fallback then this
simply disables using itapdly.
Fixes: 901d16e46296 ("mmc: sdhci_am654: Add retry tuning")
Cc: stable@vger.kernel.org
Signed-off-by: Diogo Ivo (Schneider Electric) <diogo.ivo@bootlin.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Reviewed-by: Judith Mendez <jm@ti.com>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Diogo Ivo (Schneider Electric) <diogo.ivo@bootlin.com>
Date: Fri Aug 7 13:06:57 2026 +0200
mmc: sdhci_am654: Move tuning_loop to local variable
commit ff894dced1a7ad7523f9c65dbdb53d02474cca0f upstream.
The tuning_loop field in struct sdhci_am654_data is only used within
sdhci_am654_platform_execute_tuning() as a loop counter that is
initialized to 0 in sdhci_am654_init(). Since it shouldn't persist across
function calls, otherwise every failure expends its "budget", move it to a
local variable and remove the struct field along with the now-unnecessary
initialization.
Signed-off-by: Diogo Ivo (Schneider Electric) <diogo.ivo@bootlin.com>
Reviewed-by: Judith Mendez <jm@ti.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Fixes: de31f6ab68a3 ("mmc: sdhci_am654: Reset Command and Data line after tuning")
Cc: stable@vger.kernel.org
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Diogo Ivo (Schneider Electric) <diogo.ivo@bootlin.com>
Date: Fri Aug 7 13:06:58 2026 +0200
mmc: sdhci_am654: Reset command and data lines on failed tuning
commit 7197d9107d9545730153b82ea5a411c5208b443f upstream.
The CMD/DATA reset after tuning should be performed regardless of
whether tuning succeeded or failed, since tuning data may remain in
the buffer in either case. Move the error return after the reset so
that the controller is always cleaned up.
Fixes: de31f6ab68a3 ("mmc: sdhci_am654: Reset Command and Data line after tuning")
Cc: stable@vger.kernel.org
Signed-off-by: Diogo Ivo (Schneider Electric) <diogo.ivo@bootlin.com>
Reviewed-by: Judith Mendez <jm@ti.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Felix Gu <ustc.gu@gmail.com>
Date: Sat Aug 22 02:58:47 2026 +0800
mmc: sdio_uart: fix xmit_fifo leak when the port table is full
commit 53823e25793a97d07e6e98e0904bbf74cac8bc76 upstream.
sdio_uart_add_port() allocates the transmit fifo before claiming a
slot in sdio_uart_table[]. When all UART_NR slots are taken, it
returns -EBUSY with the fifo still allocated, but the probe error
path only kfree()s the port, leaking the transmit fifo.
Free the fifo in the failure path of sdio_uart_add_port() itself so
the function retains nothing on error.
Fixes: 8b197a5ce7a7 ("sdio_uart: Use kfifo instead of the messy circ stuff")
Signed-off-by: Felix Gu <ustc.gu@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Runyu Xiao <runyu.xiao@seu.edu.cn>
Date: Wed Sep 2 22:09:26 2026 +0800
mmc: sh_mmcif: initialize IRQ-thread mutex before requesting interrupt
commit d5ea0d226e8f0801d78702142a124d78c317d822 upstream.
The threaded IRQ handler can run before devm_request_threaded_irq()
returns, but thread_lock was initialized afterwards. Initialize it before
requesting either interrupt.
Fixes: 8047310ee984 ("mmc: sh_mmcif: fix a race, causing an Oops on SMP")
Cc: stable@vger.kernel.org
Assisted-by: Codex:GPT-5
Signed-off-by: Runyu Xiao <runyu.xiao@seu.edu.cn>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Xu Rao <raoxu@uniontech.com>
Date: Tue Aug 18 19:31:53 2026 +0800
mmc: spi: reset bytes_xfered before retrying CRC failures
commit 8b0cc8707f65e0f51912e764e1b309b2559db1ec upstream.
mmc_spi_data_do() updates data->bytes_xfered after each block has been
transferred successfully. If a later block in the same data request
fails with a CRC error, data->bytes_xfered may therefore contain the
number of bytes completed before the failing block.
mmc_spi_request() has a private recovery path for such CRC failures. It
sends STOP_TRANSMISSION, clears data->error and jumps back to
crc_recover to issue the same command and data request again. However,
it does not clear data->bytes_xfered before the retry.
If the retry succeeds, the request is completed with the bytes from the
failed attempt still included in data->bytes_xfered. For a multi-block
request this can make the completed request report more bytes than were
transferred by the successful retry, and can even exceed the request size
when most blocks completed before the CRC error.
This is most likely to be observed on MMC-over-SPI systems where long
multi-block transfers occasionally hit a data CRC error but the
mmc_spi-internal retry succeeds. The data itself is retried, but the
completion accounting is not.
Clear data->bytes_xfered together with data->error before repeating the
request so the final completion reports only the bytes transferred by the
successful attempt.
Fixes: 061c6c847eeb ("mmc_spi: Recover from CRC errors for r/w operation over SPI.")
Cc: stable@vger.kernel.org
Signed-off-by: Xu Rao <raoxu@uniontech.com>
Signed-off-by: Ulf Hansson <ulfh@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Kuniyuki Iwashima <kuniyu@google.com>
Date: Wed Sep 9 23:31:23 2026 +0000
neighbour: Add missing RCU annotation for neightbl_dump_info().
[ Upstream commit 764dcebb033764633700a036c7351a7c6350eec6 ]
neightbl_dump_info() fetches the first non-default neigh_parms
with list_next_entry(&tbl->parms, ...) and iterates through the
list with list_for_each_entry_from_rcu().
However, list_next_entry() does not use RCU helper.
Let's use list_for_each_entry_rcu() and skip the default parms.
Fixes: 4ae34be50064 ("neighbour: Convert RTM_GETNEIGHTBL to RCU.")
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260909233143.2401847-2-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Kuniyuki Iwashima <kuniyu@google.com>
Date: Wed Sep 9 23:31:25 2026 +0000
neighbour: Don't render blackhole_netdev via RTM_GETNEIGHTBL.
[ Upstream commit 7b430fcfc972f61b09cc19ca95997586af4a147d ]
The cited commits started to initialise blackhole_netdev with
neigh_parms_alloc().
This is visible in init_net as the ifindex==0 entries via
RTM_GETNEIGHTBL:
# ynl --family rt-neigh --dump getneightbl --output-json \
| jq '.[] | select(.parms.ifindex == 0)
| {name: .name, ifindex: .parms.ifindex}'
{
"name": "arp_cache",
"ifindex": 0
}
{
"name": "ndisc_cache",
"ifindex": 0
}
For RTM_SETNEIGHTBL, ifindex being 0 means wildcard.
Let's skip blackhole_netdev's parms in neightbl_dump_info().
Note that lookup_neigh_parms() does not need the same change
because the default parms is always the first entry and matches
with ifindex == 0.
Fixes: e5f80fcf869a ("ipv6: give an IPv6 dev to blackhole_netdev")
Fixes: 22600596b675 ("ipv4: give an IPv4 dev to blackhole_netdev")
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260909233143.2401847-4-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Kuniyuki Iwashima <kuniyu@google.com>
Date: Wed Sep 9 23:31:24 2026 +0000
neighbour: Enforce min/max to NDTPA_INTERVAL_PROBE_TIME_MS.
[ Upstream commit 6d79b223ec44ada58ad37db42f539b60985a7722 ]
NDTPA_INTERVAL_PROBE_TIME_MS sets .type and .min but misses
.validation_type, so no validation is applied:
# ynl --family rt-neigh --do setneightbl \
--json '{"name": "arp_cache", "parms": {"interval-probe-time-ms": 0}}'
# ynl --family rt-neigh --dump getneightbl --output-json | \
jq '.[] | select(.name == "arp_cache" and has("config"))
| .parms["interval-probe-time-ms"]'
0
Moreover, nla_get_msecs() uses msecs_to_jiffies(), and u64 is
silently cast to u32, so a larger value can bypass the min check:
e.g. 4294967296 == 0x100000000
# ynl --family rt-neigh --do setneightbl \
--json '{"name": "arp_cache", "parms": {"interval-probe-time-ms": 4294967296}}'
# ynl --family rt-neigh --dump getneightbl --output-json | \
jq '.[] | select(.name == "arp_cache" and has("config"))
| .parms["interval-probe-time-ms"]'
0
msecs_to_jiffies() returns MAX_JIFFY_OFFSET if the value is
larger than INT_MAX. Also, INT_MAX ms overflows int NEIGH_VAR()
when HZ > 1000 (Alpha, MIPS), and passing a negative integer to
queue_delayed_work(unsigned long delay) causes sign extension,
which wraps around the expiry time to the past, resulting in it
being handled as 0 delay in the timer wheel.
Let's use NLA_POLICY_FULL_RANGE() and limit the max to 1 day.
The same max check is applied to sysctl as well.
Note that this controls the probe interval for NTF_MANAGED
entries, so the max of 1 day is unlikely to break any
deployments.
Fixes: 211da42eaa45 ("net, neigh: introduce interval_probe_time_ms for periodic probe")
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260909233143.2401847-3-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Kuniyuki Iwashima <kuniyu@google.com>
Date: Wed Sep 9 23:31:26 2026 +0000
neighbour: Skip default parms when resumed in neightbl_dump_info().
[ Upstream commit 979aabdad8dd03394467ee484a1a70f3d40b19ba ]
neightbl_dump_info() calls neightbl_fill_info() in each loop
to render the default parms.
If there are many devices and neightbl_fill_param_info() failed,
neightbl_fill_info() is called again when the dump resumes:
# ynl --family rt-neigh --dump getneightbl --output-json |
jq '.[] | {name: .name, ifindex: .parms.ifindex}'
...
{
"name": "ndisc_cache",
"ifindex": null
}
...
{
"name": "ndisc_cache",
"ifindex": 6
}
{
"name": "ndisc_cache",
"ifindex": null
}
{
"name": "ndisc_cache",
"ifindex": 5
}
Let's skip neightbl_fill_info() if it is already called in
neightbl_dump_info().
Note that we cannot use !neigh_skip instead of !default_skip
because default_skip == 1 && neigh_skip == 0 could be true
if the first neightbl_fill_param_info() fails.
Also, nidx must be cleared at the end of each table loop;
otherwise, if neightbl_fill_info() for a subsequent table
fails, the leftover nidx from the previous table would be
saved in cb->args[1], resulting in erroneously skipping parms
of the subsequent table in the next dump.
Fixes: c7fb64db001f ("[NETLINK]: Neighbour table configuration and statistics via rtnetlink")
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260909233143.2401847-5-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Mark Amirkan <markdamirkan@gmail.com>
Date: Sun Sep 13 10:31:08 2026 +0000
net/packet: avoid truncating TPACKET_V3 private size
commit 37213e61120297920ae4c937fcb326a360da5084 upstream.
tpacket_req3.tp_sizeof_priv is an unsigned int, and packet_set_ring()
validates the full value against the block size. init_prb_bdqc() then
stores it in the unsigned short blk_sizeof_priv field.
Commit 2b6867c2ce76 ("net/packet: fix overflow in check for priv area
size") fixed the validation arithmetic, but an accepted value above
USHRT_MAX still narrows when it is stored.
For a 131072-byte block, tp_sizeof_priv=65536 is valid. The narrowing
makes offset_to_first_pkt 48 instead of 65584, so packet records can be
placed in the private area that userspace asked the kernel to preserve.
blk_sizeof_priv is internal state, so widen it to hold the validated
UAPI value.
Fixes: f6fb8f100b80 ("af-packet: TPACKET_V3 flexible buffer implementation.")
Cc: stable@vger.kernel.org
Signed-off-by: Mark Amirkan <markdamirkan@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260913-b4-send-packet-private-v1-1-925eab2cd388@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Mark Amirkan <markdamirkan@gmail.com>
Date: Sun Sep 13 10:28:08 2026 +0000
net/packet: clear RX owner on VNET header error
commit 33ff111d7ba3beb86e28938d6382bb5beabd865a upstream.
Commit 61fad6816fc1 ("net/packet: tpacket_rcv: avoid a producer race
condition") added rx_owner_map and made tpacket_rcv() claim a V1 or V2
ring slot before converting the virtio-net header. If the conversion
fails, the drop path leaves the slot claimed.
With a one-frame TPACKET_V2 ring, an unsupported UDP GSO packet leaves
the only slot unavailable, so the ring also drops the next valid packet.
Clear the ownership bit on this error path. TPACKET_V3 already clears
its block state here.
Fixes: 61fad6816fc1 ("net/packet: tpacket_rcv: avoid a producer race condition")
Cc: stable@vger.kernel.org
Signed-off-by: Mark Amirkan <markdamirkan@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260913-b4-send-packet-vnet-v1-1-5545ffb528ae@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Date: Thu Sep 10 17:34:12 2026 +0800
net/sched: act_api: release tail references on DELACTION failure
commit 6e05e46fa821a5c1b281355f1f622ac76cb6080a upstream.
A batched RTM_DELACTION request takes a temporary reference on each
action before attempting any deletion. tcf_action_delete() clears
each processed slot and drops its temporary reference before attempting
the deletion. If deletion fails, tca_action_gd() calls
tcf_action_put_many() to release the remaining references, but its
tcf_act_for_each_action() iterator stops at the first NULL slot.
When a batch stops at an action bound to a filter, this leaks a
reference on each subsequent action. A later delete of an unbound
action can then return success without removing it from the IDR.
Walk the full array in tcf_action_put_many() and skip NULL slots to
release the references held on the unprocessed actions.
Fixes: a0e947c9ccff ("net/sched: act_api: avoid non-contiguous action array")
Cc: stable@vger.kernel.org
Signed-off-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Link: https://patch.msgid.link/20260910093413.34509-2-xuanqiang.luo@linux.dev
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Jamal Hadi Salim <jhs@mojatatu.com>
Date: Sat Sep 12 14:08:30 2026 -0400
net/sched: codel: bound the dropping loop per dequeue call
[ Upstream commit 7f4a5ec6258fd7c92633ec4b0493fc51166d9398 ]
The CoDel control law schedules the next drop one interval/sqrt(count)
after the previous drop, using the configured interval
(codel_params.interval). For very small intervals the scheduled step
rounds down to zero, so the dropping loop in codel_dequeue() never
advances and drains the entire backlog under the qdisc lock in one
call - an unprivileged user can trigger a soft lockup this way.
Fix in the shared codel code used by both codel and fq_codel:
1. Make the control-law step at least 1 tick so the dropping loop
always moves forward.
2. Cap the dropping loop at CODEL_MAX_DROPS_PER_DEQUEUE (256) drops
per codel_dequeue() call, resyncing drop_next to now when the cap
is hit: the catch-up owed to the loop grows with the idle gap and
the backlog, which no interval threshold can bound. This is a
deliberate behaviour change after long idle gaps.
The cap applies to fq_codel (4b549a2ef4be) and the mac80211 TXQ path
(fixed interval, cap only).
The target sojourn delay (codel_params.target) is not validated: it
does not feed the control law, so a sub-tick value is aggressive
rather than deadlock-prone.
Conditions to recreate the bug:
- tc qdisc add dev lo root handle 1: tbf rate 1kbit burst 2kb limit 1000000
- tc qdisc add dev lo parent 1:1 handle 10: codel interval 2us target 1ms noecn limit 1000000 (same for fq_codel)
- unpatched kernel: tc accepts it; a UDP flood under the 1kbit tbf
soft-lockups (watchdog: BUG: soft lockup) while one
codel_dequeue() call drops the backlog under the qdisc lock
- patched kernel: same setup, at most 256 drops per dequeue call,
no soft lockup
Testing: claim reproducer and interval 2us/3us variants run clean;
tdc qdisc category passes (see the selftests patch).
Fixes: 76e3cc126bb2 ("codel: Controlled Delay AQM")
Reported-by: Vega <vega@nebusec.ai>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Tested-by: Victor Nogueira <victor@mojatatu.com>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Reviewed-by: Toke Høiland-Jørgensen <toke@toke.dk>
Link: https://patch.msgid.link/QDISC-1L5H.v1.20260912080102@mojatatu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jamal Hadi Salim <jhs@mojatatu.com>
Date: Sat Sep 12 14:09:19 2026 -0400
net/sched: hhf: cap hh_flows_limit at change time
commit 2cef2588c995722a901368def30befeef9ae55c6 upstream.
hhf_change() stores TCA_HHF_HH_FLOWS_LIMIT with no upper bound. A huge
hh_flows_limit lets each new heavy-hitter flow pass the
hh_flows_current_cnt check in alloc_new_hh() and forces a fixed-size
kzalloc(GFP_ATOMIC) per flow under spoofed traffic, for unbounded memory
growth.
Bound the attribute with NLA_POLICY_MAX() at 2*HH_FLOWS_CNT (the
hhf_init() default) and report the rejected value via extack. The
deprecated nested parse is kept: legacy tc does not set NLA_F_NESTED on
TCA_OPTIONS. Configs relying on hh_limit above the default were relying
on unbounded, unsafe behaviour and are not supported going forward.
hhf_init() also ran hhf_change() before setting the default
hh_flows_limit, so a user-supplied hh_limit at add time was clobbered
back to 2048. Set the default before hhf_change() so the configured
value sticks.
This is a follow-up to commit eb56a495f59b ("net/sched: hhf: clamp
quantum in change and init paths"), which bounded the quantum of the
same qdisc; the hh_flows_limit bound is the remaining unbounded knob of
that series' scope.
Conditions to recreate the bug: CAP_NET_ADMIN in a user namespace;
tc qdisc change dev X root hhf hh_limit 4294967295 succeeds and the
value is echoed by tc qdisc show, unbounding heavy-hitter flow
allocations; also tc qdisc add dev X root hhf hh_limit 500 stores 2048
instead of 500.
Fixes: 10239edf86f1 ("net-qdisc-hhf: Heavy-Hitter Filter (HHF) qdisc")
Cc: stable@vger.kernel.org
Reported-by: Sashiko (gemini) <sashiko-bot@kernel.org>
Closes: https://sashiko.dev/#/patchset/20260822195509.112717-1-jhs@mojatatu.com
Reviewed-by: Victor Nogueira <victor@mojatatu.com>
Tested-by: hybris <hybris@mojatatu.ai>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/QDISC-B855.v1.20260911153152@mojatatu.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Nicolai Buchwitz <nb@tipi-net.de>
Date: Sun Sep 13 21:00:52 2026 +0200
net: bcmgenet: restore the hardware filters on open
[ Upstream commit 23ca4ddc4fce2c233a49e9fd34d4b5b02bd7324e ]
bcmgenet_hfb_init() runs INIT_LIST_HEAD() on priv->rxnfc_list, which drops
every rule off the list, and bcmgenet_open() calls it on each ifup. Every
rule the user configured is silently lost:
# ethtool -N eth0 flow-type ether dst $MAC action 0
Added rule with ID 0
# ethtool -n eth0 | grep -c Filter:
1
# ip link set eth0 down && ip link set eth0 up
# ethtool -n eth0 | grep -c Filter:
0
Initialise the lists once at probe and restore the rules on open, as
bcmgenet_resume() already does.
Fixes: 3e370952287c ("net: bcmgenet: add support for ethtool rxnfc flows")
Signed-off-by: Nicolai Buchwitz <nb@tipi-net.de>
Reviewed-by: Justin Chen <justin.chen@broadcom.com>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://patch.msgid.link/20260913190052.939955-1-nb@tipi-net.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Nikolay Aleksandrov <razor@blackwall.org>
Date: Fri Sep 11 13:50:21 2026 +0300
net: bridge: mst: move switchdev call outside rcu
[ Upstream commit 18a6fe05fb6e18de29fa90d388bb34044114b3d8 ]
This is a follow-up of one of sashiko's pre-existing bug reports.
br_mst_set_state() calls switchdev_port_attr_set() for nonzero MSTIs
while holding rcu_read_lock() which invokes the blocking switchdev
notifier chain and may sleep. Nonzero MSTI changes come from netlink
with rtnl held. Move the switchdev call before entering the rcu section and
assert that rtnl is held.
The call cannot be deferred because netlink needs its error and extack.
Also DSA reads the old bridge MST state during the callback and checks it.
A deferred callback will be late and will see the updated state.
Fixes: 3a7c1661ae13 ("net: bridge: mst: fix vlan use-after-free")
Signed-off-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260911105021.1385934-1-razor@blackwall.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Nikolay Aleksandrov <razor@blackwall.org>
Date: Mon Sep 14 13:52:58 2026 +0300
net: bridge: vlan: fix bugs caused by switchdev deletion errors
[ Upstream commit 2842ce397dd09882530b42f7fdb0c855767eb24e ]
Allowing switchdev to prevent vlan deletion and error out in __vlan_del
could cause multiple different issues - inconsistent state, memory leaks
when flushing, NULL pointer dereference on bridge error when flushing.
It doesn't make sense to allow it to stop __vlan_del, so log the error
and continue with software vlan deletion. This is also consistent with
8021q behaviour.
Suggested-by: Ido Schimmel <idosch@nvidia.com>
Fixes: bf361ad38165 ("net: bridge: check __vlan_vid_del for error")
Fixes: 5454f5c28eca ("net: bridge: vlan: check for errors from __vlan_del in __vlan_flush")
Fixes: 2594e9064a57 ("bridge: vlan: add per-vlan struct and move to rhashtables")
Fixes: 9c86ce2c1ae3 ("net: bridge: Notify about bridge VLANs")
Signed-off-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260914105258.3436918-1-razor@blackwall.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Pablo Neira Ayuso <pablo@netfilter.org>
Date: Tue Jul 28 18:30:46 2026 +0200
net: do not advance stack index from dev_fwd_path()
[ Upstream commit 5deda60c56eeeab25beb10cf4d48e07587076b11 ]
Update stack index from dev_fill_forward_path() instead, once the
forward path slot has been populated.
Acked-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Stable-dep-of: 150dba2c69e9 ("net: remove WARN_ON_ONCE() from the dev_fill_forward_path() loop check")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Daniel Golle <daniel@makrotopia.org>
Date: Thu Sep 10 14:13:15 2026 +0100
net: dsa: mxl862xx: disable the stats poll on teardown
[ Upstream commit 9e92ad4630f5dd1838ce6bbe6b1bd2c73d34de36 ]
mxl862xx_setup() arms the stats poll before mxl862xx_setup_mdio(), and
nothing stops it until dsa_register_switch() has returned an error to
mxl862xx_probe(). DSA frees the dsa_port list before it returns, so a
poll that fires once .setup or a later step of dsa_tree_setup() has
failed walks freed ports. On shutdown the user ports stay registered,
and the WORK_STOPPED flag test in mxl862xx_get_stats64() is not atomic
with the cancel in mxl862xx_shutdown(), so a re-arm that read the flag
before it was set queues the poll after cancel_delayed_work_sync() has
returned.
Arm the poll once .setup has succeeded and stop it from a .teardown op,
which DSA calls on unregister and after a failed registration, in both
cases before it frees the ports. Use disable_delayed_work_sync() there
and in shutdown(): it drains a running poll as the cancel did and turns
every later attempt to queue the work into a no-op, so the re-arm
cannot bring the poll back. remove() and the probe error path only set
WORK_STOPPED, which crc_err_work tests before it walks the ports.
Fixes: a21d33a5265f ("net: dsa: mxl862xx: implement .get_stats64")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Link: https://patch.msgid.link/1eb6f7fc1789b67e4b11e3f4d5ff080d0b6f7cbb.1789045590.git.daniel@makrotopia.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Linus Walleij <linusw@kernel.org>
Date: Mon Sep 14 23:26:41 2026 +0200
net: ethernet: cortina: Ack RX overrun interrupt correctly
[ Upstream commit 1dd85662fee6e2ac580b1c4f9a0c0a7ae6e31f0e ]
The RX overrun interrupt is reported in interrupt status register 4, but
gmac_irq() acknowledges it using the RX descriptor error bit from status
register 0. For GMAC0 this writes the GMAC1 overrun bit, while for GMAC1
the shift leaves no bit in the 32-bit register.
Acknowledge the same per-port RX overrun bit that was detected.
Fixes: 4d5ae32f5e1e ("net: ethernet: Add a driver for Gemini gigabit ethernet")
Signed-off-by: Linus Walleij <linusw@kernel.org>
Link: https://patch.msgid.link/20260914-b4-gemini-ethernet-fixes-2-v2-1-5ab39a047b90@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Hohyun Sim <tlaghgus0425@korea.ac.kr>
Date: Thu Sep 10 15:37:43 2026 +0900
net: fddi: skfp: fix NULL deref when setting the MAC address while down
[ Upstream commit 7c8810c2e69c3d9ca6df870b40ae9218e50b4fb1 ]
skfp_ctl_set_mac_address() calls ResetAdapter() unconditionally, without
checking netif_running(). ResetAdapter() first calls card_stop(), which
sets smc->hw.hw_state to STOPPED, and then mac_drv_clear_tx_queue(),
which walks the two transmit queues:
for (i = QUEUE_S; i <= QUEUE_A0; i++) {
queue = smc->hw.fp.tx[i] ;
...
t = queue->tx_curr_get ;
smc->hw.fp.tx[] is only populated by init_tx(), which is reached from
skfp_open() through init_smt() -> init_fddi_driver() -> init_fplus() ->
init_mac() -> init_tx(). The private area is allocated and zeroed by
alloc_fddidev(), so on an interface that has never been brought up both
queue pointers are still NULL. The hw_state test at the top of
mac_drv_clear_tx_queue() does not catch this, because card_stop() has
just set STOPPED; the function proceeds into the loop and dereferences
NULL. ResetAdapter() does call init_smt() itself, but only after the
queues have been cleared.
Setting the MAC address on a down interface therefore oopses:
ip link set dev fddi0 address 02:00:00:00:00:01
BUG: KASAN: null-ptr-deref in mac_drv_clear_tx_queue+0x68/0x2c0 [skfp]
Read of size 8 at addr 0000000000000010 by task ip/302
Call Trace:
<TASK>
mac_drv_clear_tx_queue+0x68/0x2c0 [skfp 6c01d4bab63c36978bd0a7d7e90837adb44cc37b]
ResetAdapter+0x29/0x100 [skfp 6c01d4bab63c36978bd0a7d7e90837adb44cc37b]
skfp_ctl_set_mac_address+0x57/0x80 [skfp 6c01d4bab63c36978bd0a7d7e90837adb44cc37b]
netif_set_mac_address+0x1e4/0x2c0
do_setlink+0x684/0x2680
</TASK>
Address 0x10 is the offset of tx_curr_get, the third pointer in
struct s_smt_tx_queue, on 64-bit. mac_drv_clear_rx_queue(), which
ResetAdapter() calls immediately afterwards, dereferences
smc->hw.fp.rx[QUEUE_R1] in the same way behind the same ineffective
hw_state test; the transmit queue merely crashes first. Both are
covered by the guard below.
Skip the adapter reset when the interface is down. dev_addr_set() is
left unconditional, so the new address is still recorded in
dev->dev_addr. Nothing is lost by not resetting the adapter here:
skfp_open() deliberately re-reads the factory address on every open,
read_address(smc, NULL);
eth_hw_addr_set(dev, smc->hw.fddi_canon_addr.a);
and the comment above it states this is done to discard exactly such an
address override across a close/open cycle. An address set while the
interface is down could not have survived the following open even
before this change, so the guard removes no working behaviour. Guarding
the hardware side of ndo_set_mac_address() with netif_running() is
established practice; skge_set_mac_address() has done so since commit
2eb3e621c4e0 ("skge: set mac address bonding fix").
Guarding the reset as a whole, rather than NULL-checking the queues, is
also what the rest of the driver expects. After a previous open/close
the queue pointers are stale but non-NULL, so there is no crash, yet
ResetAdapter() goes on to call smt_online() and STI_FBI() ("Enable
Board Interrupts") while skfp_close() has already called free_irq() -
the adapter would be brought back online with no handler installed. The
only other ResetAdapter() caller is skfp_interrupt(), which by
construction runs only while the device is open.
Found by automated driver testing against an emulated SysKonnect FDDI
adapter under a KASAN-enabled 7.0.0 kernel. Triggering it requires
CAP_NET_ADMIN.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Assisted-by: LLM KASAN
Signed-off-by: Hohyun Sim <tlaghgus0425@korea.ac.kr>
Link: https://patch.msgid.link/20260910063743.110747-1-tlaghgus0425@korea.ac.kr
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Gris Ge <cnfourt@gmail.com>
Date: Sun Sep 13 17:08:50 2026 +0800
net: ip_tunnel: initialize `options_len` before referencing options
commit 455ebeadf714f51e1dbbd6a022c74c9215b1cd76 upstream.
The following command triggers a kernel panic:
ip link add d0 type dummy; ip link set d0 up
ip route add 10.30.0.0/16 \
encap ip id 300 geneve_opts 4660:66:11223344 dev d0
memcpy: detected buffer overflow: 4 byte write of buffer size 0
kernel BUG at lib/string_helpers.c:1044!
...
ip_tun_parse_opts.part.0.cold+0x10/0x10
ip_tun_build_state+0x116/0x2a0
On kernels built with GCC 15+ and `CONFIG_FORTIFY_SOURCE`, the fortified
`memcpy()` got 0 sized destination with request of 4 bytes length:
static int ip_tun_parse_opts_geneve(...)
{
...
attr = tb[LWTUNNEL_IP_OPT_GENEVE_DATA];
data_len = nla_len(attr); /* == 4 */
struct geneve_opt *opt = ip_tunnel_info_opts(info) + opts_len;
memcpy(opt->opt_data, nla_data(attr), data_len);
/* ^^^^^^^^^^^^^ 0 since options_len is assigned afterwards */
Fixed by initializing the counter before the options are referenced.
Matching what `tunnel_key_opts_set()` already does.
Fixes: bb5e62f2d547 ("net: Add options as a flexible array to struct ip_tunnel_info")
Cc: stable@vger.kernel.org
Signed-off-by: Gris Ge <cnfourt@gmail.com>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Link: https://patch.msgid.link/20260913090851.468216-1-cnfourt@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Mark Amirkan <markdamirkan@gmail.com>
Date: Sun Sep 13 17:14:09 2026 -0700
net: lan743x: fix RX checksum use-after-free
commit a9ce4053dc945c5372dedba5017ee675b30dc0c5 upstream.
lan743x_rx_process_buffer() adds each non-first receive buffer to the
head skb's frag_list. On the last descriptor, lan743x_rx_trim_skb()
linearizes the head and frees the fragment skb metadata.
The checksum-success path then writes ip_summed through the local skb
pointer, which still points to the final fragment. This causes a
use-after-free write when a packet spans more than one receive buffer.
Set ip_summed on the surviving head skb instead. Multi-buffer receive
can occur after a live MTU increase because existing ring entries keep
their old buffer size until they are replenished.
A KUnit test invoking lan743x_rx_process_buffer() with a two-buffer
packet produced a one-byte KASAN use-after-free write before this change.
The same test passed after the change. The driver object also builds
with W=1. This was not tested on physical LAN743x hardware.
Fixes: cd6910501cfd ("net: lan743x: Add support for Rx IP & TCP checksum offload")
Cc: stable@vger.kernel.org
Signed-off-by: Mark Amirkan <markdamirkan@gmail.com>
Reviewed-by: Chenguang Zhao <zhaochenguang@kylinos.cn>
Link: https://patch.msgid.link/20260913-b4-send-lan743x-uaf-v1-1-73d563d08ba9@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Tue Sep 15 04:30:54 2026 +0000
net: lock the socket in sock_gettstamp()
[ Upstream commit 9ed55f3dbef4f4adfe65eb03b0c35c53229a8490 ]
sk->sk_flags must only be changed while holding the socket lock,
because sock_set_flag() and sock_reset_flag() use non atomic
operations (__set_bit() and __clear_bit()).
sock_gettstamp() is one of the last places where a bit of sk->sk_flags
is changed from a syscall without owning the socket lock, through
sock_enable_timestamp(sk, SOCK_TIMESTAMP).
sk_set_memalloc() and sk_clear_memalloc() also change sk->sk_flags
without the socket lock, but their callers (nbd, iscsi_tcp, nvme-tcp,
sunrpc, wireguard) need a careful audit, this will be addressed in a
separate patch.
Jungwoo Lee and Wongi Lee reported an UDP socket use-after-free
caused by this bug: a SIOCGSTAMPNS_NEW ioctl racing with bind()
can cancel the SOCK_RCU_FREE bit that udp_lib_get_port() just set,
because both threads perform a read-modify-write on the same word.
CPU 0 (bind) CPU 1 (SIOCGSTAMPNS_NEW)
-------------------------------- ----------------------------
read sk_flags = F read sk_flags = F
compute F | BIT(SOCK_RCU_FREE) compute F | BIT(SOCK_TIMESTAMP)
store F | BIT(SOCK_RCU_FREE)
sk_add_node_rcu(sk, ...)
store F | BIT(SOCK_TIMESTAMP)
After the lost update, SOCK_RCU_FREE is clear while the socket is
visible to lockless UDP receive lookups. sk_destruct() then frees
the socket immediately instead of waiting for a RCU grace period,
while the receive path still holds a reference-less pointer to it:
BUG: KASAN: slab-use-after-free in ipv4_pktinfo_prepare+0x30/0x410
Read of size 8 at addr ffff888008806610 by task exploit/207
CPU: 0 UID: 1000 PID: 207 Comm: exploit Not tainted 6.12.95+ #1
ipv4_pktinfo_prepare+0x30/0x410
udp_queue_rcv_one_skb+0x51c/0x1180
udp_unicast_rcv_skb+0x109/0x350
ip_protocol_deliver_rcu+0x14b/0x310
ip_local_deliver_finish+0x29d/0x390
ip_local_deliver+0x24d/0x2a0
Only grab the socket lock when SOCK_TIMESTAMP has to be set,
to keep the common case lockless.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Reported-by: Jungwoo Lee <jwlee2217@gmail.com>
Reported-by: Wongi Lee <qw3rtyp0@gmail.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260915043055.3441600-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: James Clark <jjc@jclark.com>
Date: Tue Sep 15 11:58:17 2026 +0700
net: macb: fix ordering around PTP timestamp read
[ Upstream commit 9ca4ba24259183ce15665be86b2956cd896c4687 ]
PTP_SYS_OFFSET_EXTENDED returns system timestamps that do not correctly
bracket the PHC register read on MACB/GEM. On a Raspberry Pi 5, the
returned interval can be as short as 37 ns, while an ordered register
read takes approximately 1 us. This biases the midpoint used by phc2sys,
causing CLOCK_REALTIME to run approximately 0.5 us ahead when synchronized
to the PHC.
gem_tsu_get_time() reads the nanoseconds register using the driver's
relaxed MMIO accessor. On weakly ordered systems, the subsequent system
timestamp can be taken before the register read completes. The internal
smp_rmb() in the pre-timestamp path also does not guarantee ordering
against the subsequent MMIO read.
Add rmb() before and after the bracketed nanoseconds read in both the
normal and seconds rollover paths so the system timestamps bracket the
PHC read. Adding the post-read barrier increases the minimum interval on
the same Raspberry Pi 5 to approximately 1 us.
Fixes: e51bb5c2784c ("net: macb: ptp: Switch to gettimex64() interface")
Tested-by: Nicolai Buchwitz <nb@tipi-net.de> # Raspberry Pi CM5, min bracket 37 ns -> 981 ns
Reviewed-by: Nicolai Buchwitz <nb@tipi-net.de>
Reviewed-by: Théo Lebrun <theo.lebrun@bootlin.com>
Assisted-by: LLM
Signed-off-by: James Clark <jjc@jclark.com>
Link: https://patch.msgid.link/20260915045823.76100-1-jjc@jclark.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Dmitriy Okunev <dokunevdmitriy@gmail.com>
Date: Mon Sep 14 12:15:57 2026 +0300
net: mvpp2: prevent buffer overflow in page_pool allocation
[ Upstream commit 14cb1e7702e5cb3c58888f6aed498381a73927d2 ]
The per‑processor buffering scheme is supported only if the
number of pools (nrxqs * 2) does not exceed MVPP2_BM_MAX_POOLS (8).
This is already checked in mvpp2_probe() during the initial
activation of percpu_pools.
However, mvpp2_change_mtu() may later call
mvpp2_bm_switch_buffers(priv, true) without this check, which can
lead to an out-of-bounds access in the priv->page_pool array in
mvpp2_bm_init(). The array is sized to hold MVPP2_PORT_MAX_RXQ
entries, and mvpp2_get_nrxqs() may return exactly that value. The
per-CPU scheme then doubles it to nrxqs * 2, exceeding the array
bounds.
Check that the hardware version is MVPP22 or newer and that the
number of pools (nrxqs * 2) does not exceed MVPP2_BM_MAX_POOLS
before switching to per-CPU mode.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Fixes: 7d04b0b13b11 ("mvpp2: percpu buffers")
Signed-off-by: Dmitriy Okunev <dokunevdmitriy@gmail.com>
Link: https://patch.msgid.link/20260914091557.71769-1-dokunevdmitriy@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Yige Jiang <yigejiang86@gmail.com>
Date: Sun Sep 13 14:41:02 2026 +0800
net: netsec: fix device_node reference leak on phy_np
[ Upstream commit 5ae916fabca141b79b32e2e57f3c915c0f1e1b2e ]
netsec_of_probe() takes a reference on the PHY device_node with
of_parse_phandle() and stores it in priv->phy_np, but the driver never
drops it. One device_node reference is leaked per probe, on the success
path as well as on every error path reached after netsec_of_probe().
Neither consumer takes ownership. of_mdio_parse_addr() is a static
inline taking a const struct device_node * that only reads the "reg"
property. of_phy_connect() borrows as well: of_phy_get_and_connect() in
drivers/net/mdio/of_mdio.c brackets its own call with of_node_get() at
:364 and of_node_put() at :373, which would be a double put if
of_phy_connect() consumed the reference.
The node is still in use at netsec_netdev_open() time, where it is
passed to of_phy_connect(), so it has device lifetime. Release it at
the probe error label, which every failure path after the acquire
funnels through, and in netsec_remove(). Both releases precede
free_netdev(), since priv is netdev_priv(ndev). The ACPI probe path
leaves priv->phy_np NULL and of_node_put(NULL) is a no-op.
There is no end-user visible symptom on currently supported platforms:
a device_node is only freed once OF_DYNAMIC is enabled and the node has
been detached, so on a static device tree the imbalance is inert. It is
observable as a refcount that grows across bind/unbind cycles, and would
matter under device tree overlays.
Found by static analysis of reference acquire/release pairing rather
than from a runtime report. No reproducer was produced and the change
has not been runtime tested; it is compile-tested only (arm64,
CONFIG_SNI_NETSEC=m via COMPILE_TEST).
Fixes: 533dd11a12f6 ("net: socionext: Add Synquacer NetSec driver")
Signed-off-by: Yige Jiang <yigejiang86@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260913064102.37452-1-yigejiang86@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Pablo Neira Ayuso <pablo@netfilter.org>
Date: Tue Jul 28 18:30:48 2026 +0200
net: pass dst via net_device_path in dev_fill_forward_path()
[ Upstream commit 0ad8404e776698de4dac5cc0df3be68d344e741c ]
Add dst_entry to tunnel device path, this will allow us to remove
a duplicated route lookup.
This is a preparation patch to retrieve the tunnel route directly
from the .fill_forward_path. This new dst_entry in the tunnel will be
used by a follow up patch.
Since dst_release() works fine on NULL interface, this is still
noop until the flowtable starts using this.
Add a new dev_fill_forward_path_release() function to drop the refcount
on the tunnel device route and use it in case of error out. Export it so
to drop the refcount on the tunnel route at a later stage.
Adjust existing drivers that recycle dev_fill_forward_path() to call
dev_fill_forward_path_release() for safety reasons.
Acked-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Stable-dep-of: 150dba2c69e9 ("net: remove WARN_ON_ONCE() from the dev_fill_forward_path() loop check")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com>
Date: Thu Aug 6 23:09:03 2026 +0200
net: pass net_device_path_ctx to dev_fill_forward_path()
[ Upstream commit 95133a416809c7e822da4023b7f4193ef2620796 ]
Refactor dev_fill_forward_path() to take a struct net_device_path_ctx
pointer instead of a (dev, daddr) pair, so the caller can build and
populate the context up front and keep it after the forward path walk.
This allows additional fields (e.g. vlan and ether_type) to be carried
in the context and shared with ndo_fill_forward_path implementations,
instead of being reconstructed on the stack inside the core helper.
Update the mtk_ppe_offload, airoha_ppe and nf_flow_table_path callers to
allocate and fill the context before invoking dev_fill_forward_path().
The network topology resolution behaviour is unchanged.
This is a preliminary patch to enable HW flowtable offload for IPv4
over IPv6 tunnels.
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Stable-dep-of: 150dba2c69e9 ("net: remove WARN_ON_ONCE() from the dev_fill_forward_path() loop check")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ahmed Naseef <naseefkm@gmail.com>
Date: Sat Sep 12 17:43:06 2026 +0400
net: phy: mediatek: do not report link and per-speed LED rules together
commit bde5212360bd44506edec073ebbd6d0c72f75820 upstream.
mtk_phy_led_hw_ctrl_get() reports TRIGGER_NETDEV_LINK whenever any of the
speed bits in on_set is on, and in addition reports every individual
TRIGGER_NETDEV_LINK_* bit that is set. The netdev trigger refuses that
combination: netdev_led_attr_store() rejects TRIGGER_NETDEV_LINK together
with any per-speed rule, and it validates the whole resulting mode rather
than just the bit being written. Once the hardware has any link bit
programmed, every write to the trigger attributes of that LED therefore
fails with -EINVAL and the LED can no longer be configured.
The rules are also fed back into the hardware: the trigger stores what is
read back, and a later write of device_name programs it again, expanding
TRIGGER_NETDEV_LINK to every speed in on_set. An LED configured for a
single speed is thereby silently widened to "on at any link speed".
Both are easy to see on the EcoNet EN7528, whose four PHYs share one LED
block. The first LED programs the block correctly, the second reads those
rules back and rewrites them widened, and the remaining two then read the
widened value, so an LED configured for "link_10 link_100" ends up lit on a
1000 Mbps link.
on_set holds every speed the LED can indicate and is exactly what
mtk_phy_led_hw_ctrl_set() programs for TRIGGER_NETDEV_LINK, so report the
speed independent rule only when all of them are on, and the individual
speeds otherwise. The mapping is then the inverse of the one used when
programming the LED and round trips without changing the register.
Fixes: c66937b0f8db ("net: phy: mediatek-ge-soc: support PHY LEDs")
Cc: stable@vger.kernel.org
Signed-off-by: Ahmed Naseef <naseefkm@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://patch.msgid.link/20260912134306.3544329-1-naseefkm@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Wed Aug 12 08:54:38 2026 +0000
net: prevent torn reads in netdev_tc_txq
[ Upstream commit 21ef2d065ad3f0cfbf2ae51260bf962a9fa2c643 ]
netdev_set_tc_queue() (and related helpers/drivers such as
netdev_bind_sb_channel_queue(), netdev_reset_tc(), and
netdev_unbind_sb_channel()) perform separate 16-bit writes to
dev->tc_to_txq[tc].count and dev->tc_to_txq[tc].offset.
Furthermore, memset() in netdev_reset_tc() and
netdev_unbind_sb_channel() provides no guarantee of performing
full 32-bit word stores.
Concurrent lockless readers (e.g. skb_tx_hash(), netdev_txq_to_tc(),
ixgbe_select_queue(), taprio, mqprio, FPE drivers) can observe torn
values where offset and count belong to inconsistent configurations.
Redefine struct netdev_tc_txq to embed count and offset inside a union
with a u32 combined field, allowing atomic manipulation via
READ_ONCE() and WRITE_ONCE().
Update all lockless readers and writers across the kernel to use
READ_ONCE() and WRITE_ONCE() on the combined field.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260812085440.3917924-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Stable-dep-of: 02fffd1939f6 ("net: stmmac: preserve real_num_tx_queues on mqprio setup failure")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Daniel Zahka <daniel.zahka@gmail.com>
Date: Tue Sep 15 16:11:37 2026 -0700
net: psp: avoid conflicts with skb->decrypted and sk_validate_xmit_skb()
[ Upstream commit a41f24c612c3f5139a3143307eb85bbcf1bd4d07 ]
PSP conflicts with TLS ULP in its usage of both skb->decrypted and
sk->sk_validate_xmit_skb().
Make PSP mutually exclusive with TLS ULP, the only other user of either
of these. As other users of skb->decrypted come along, they can be added
to sk_has_decrypt_user(). It would make sense to also assert that
sk->sk_validate_xmit_skb() is also NULL in both of these setup paths for
similar future proofing, but the PSP listener/sk_clone() path is still
broken and it could be seen as a regression to not allow rx assoc to run
on a child of a listener socket with PSP tx assoc state.
Include all TCP ULPs in the sk_has_decrypt_user() check, even though TLS
is the only one that conflicts with PSP via the decrypted bit. This is
intentional because PSP was not designed to be used with ULPs. It is
best to close off surface area that may make bugs reachable, until
someone wishes to design and test an actual user of PSP with ULPs.
Fixes: 6b46ca260e22 ("net: psp: add socket security association code")
Signed-off-by: Daniel Zahka <daniel.zahka@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260915-psp-ktls-fix-v2-1-0eedc3b148ec@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Farhad Alemi <farhad.alemi@berkeley.edu>
Date: Sat Sep 12 07:40:09 2026 +0000
net: remove WARN_ON_ONCE() from the dev_fill_forward_path() loop check
[ Upstream commit 150dba2c69e93302af24a0c868eebe4871e2e107 ]
ipip_fill_forward_path() and ip6_tnl_fill_forward_path() look up the
route to the tunnel's remote endpoint and set ctx->dev to its device,
which is the tunnel itself when that route resolves back to the tunnel.
dev_fill_forward_path() then makes no progress and trips
WARN_ON_ONCE(last_dev == ctx->dev) as soon as a flowtable tries to
offload a flow through the tunnel. That routing loop is a configuration
any CAP_NET_ADMIN user can set up, and ip_tunnel_xmit() and
ip6_tnl_xmit() already treat it as a tx error, so remove the warning and
just fail the walk, as commit 008e7a7c293b ("net: remove WARN_ON_ONCE
when accessing forward path array") did for the path stack overflow.
Fixes: ab427db17885 ("netfilter: flowtable: Add IPIP rx sw acceleration")
Fixes: d98103575dcd ("netfilter: flowtable: Add IP6IP6 rx sw acceleration")
Closes: https://lore.kernel.org/all/CA+0ovCgaRvbd0Udj70b2xxG8Cx3CaCpNhnf1V4RWQuDveZYZhA@mail.gmail.com/
Suggested-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Farhad Alemi <farhad.alemi@berkeley.edu>
Reviewed-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Link: https://patch.msgid.link/CA+0ovCgKDOk+Bg6Gh5Lwx94u_jJjQ30-vY1JcY2BYfhnWJJbPA@mail.gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Tue Sep 15 13:04:23 2026 +0000
net: skbuff: do not leave stale header offsets after pskb_carve()
[ Upstream commit a5117e1eccac6ee3bd4aed7cacf8ebcb6b3eb309 ]
pskb_carve_inside_header() and pskb_carve_inside_nonlinear() remove
the first bytes of a packet and reallocate skb->head.
All the headers that were present before the operation are gone,
but both functions call skb_headers_offset_update(skb, 0), which
is a no-op : skb->mac_header, skb->network_header,
skb->transport_header and skb->csum_start keep their old values and
now describe bytes which are no longer there.
Both helpers size the new head from the old skb_end_offset(), so the
stale offsets still land inside the new allocation. They point past
skb_tail_pointer() though, to bytes that were never initialized.
pskb_carve_inside_nonlinear() is the worst case, because it leaves a
zombie skb with an empty linear part (skb->data ==
skb_tail_pointer(skb), skb_headlen(skb) == 0), while
skb_mac_header_was_set() is still true and skb->mac_header is way
ahead of skb->data.
The only user of pskb_extract() is rds_tcp_data_recv(), and the
carved skb is queued on tinc->ti_skb_list. When the RDS incoming
message is released, rds_tcp_inc_free() calls skb_queue_purge(),
which frees the skbs with SKB_DROP_REASON_QUEUE_PURGE. This is
visible from drop_monitor, which then tries to pull back to the
(bogus) mac header :
skbuff: __skb_pull(len=234)
skb len=6968 data_len=6968 headroom=0 headlen=0 tailroom=0
end-tail=384 mac=(234,14) mac_len=14 net=(248,40) trans=288
shinfo(txflags=0 nr_frags=1 gso(size=1428 type=16 segs=5))
csum(0x100120 start=288 offset=16 ip_summed=3 complete_sw=0 valid=1 level=0)
hash(0x7b446c6c sw=0 l4=1) proto=0x86dd pkttype=0 iif=60
kernel BUG at ./include/linux/skbuff.h:2847!
Add skb_carve_reset_headers() to mark the mac and transport headers
as not set, reset the network header, clear skb->mac_len, and drop
a now meaningless CHECKSUM_PARTIAL (csum_start no longer describes
anything).
Invalidate the inner offsets as well. Unlike mac_header and
transport_header they have no "unset" sentinel, so a leftover
non-zero value still looks like a real header. Zero
skb->inner_mac_header, skb->inner_network_header,
skb->inner_transport_header, skb->inner_protocol and
skb->encapsulation, so that all the header state is invalidated in
one place.
v2: fixed an inaccurate changelog. The stale offsets stay inside the
new skb->head, which is never smaller than the old one, they
simply point past skb_tail_pointer() to bytes that are gone.
Thanks to Xuanqiang Luo for insisting on this.
Also invalidate the inner header state, as suggested by the
netdev AI review :
https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260911114922.621937-1-edumazet%40google.com
Fixes: 6fa01ccd8830 ("skbuff: Add pskb_extract() helper function")
Reported-by: syzbot+586af68eb819833c2d91@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/netdev/6aa3e9d3.f2639fcc.29487d.0028.GAE@google.com/
Cc: Xuanqiang Luo <xuanqiang.luo@linux.dev>
Cc: Allison Henderson <achender@kernel.org>
Cc: rds-devel@oss.oracle.com
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Link: https://patch.msgid.link/20260915130423.3956471-1-edumazet@google.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com>
Date: Mon Sep 14 09:41:07 2026 +0200
net: stmmac: do not overwrite phc_index when no PTP clock is registered
[ Upstream commit f0ef4b1eaed000a304726a43091588e8426ba08a ]
stmmac_get_ts_info() reports phc_index as 0 when hardware timestamping
is supported but no PTP clock has been registered yet (e.g. while the
interface is down). Zero is a valid PHC index and would make userspace
resolve the wrong clock; the absence of a clock should be reported as
-1.
The ethtool core already initializes phc_index to -1 before invoking
the get_ts_info callback (ethtool_init_tsinfo()), so just drop the
erroneous assignment.
Fixes: 9364fa7fcf12 ("net: stmmac: Remove setting of RX software timestamp")
Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Reviewed-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com>
Reviewed-by: Gal Pressman <gal@nvidia.com>
Link: https://patch.msgid.link/20260914-stmmac-fix-phc_index-v2-1-bf3d90373fe4@oss.qualcomm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com>
Date: Fri Sep 11 11:20:15 2026 +0200
net: stmmac: fix TSO header length truncation
[ Upstream commit 15989abd74f16f44bf953d056b95f1d2fda9b0cd ]
stmmac_tso_xmit() stores the protocol header length returned by
stmmac_tso_header_size() in a u8. stmmac_tso_valid_packet() admits
headers up to 1023 bytes, so a header longer than 255 bytes wraps modulo
256 (486 becomes 230, 256 becomes 0).
A TCP over IPv6 socket carrying a few hundred bytes of sticky
destination/hop-by-hop options makes skb_tcp_all_headers() exceed 255
while staying below the 1023-byte limit, so such an skb reaches
stmmac_tso_xmit().
Widen proto_hdr_len to unsigned int, which is sufficient since the value
is bounded by the hardware limit, and adjust the debug print specifier
accordingly.
Fixes: 9edfa7dab811 ("net: stmmac: enable TSO for IPv6")
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com>
Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Link: https://patch.msgid.link/20260911-stmmac-fix-header-length-v1-1-8fc103334327@oss.qualcomm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com>
Date: Fri Sep 11 10:58:30 2026 +0200
net: stmmac: preserve real_num_tx_queues on mqprio setup failure
[ Upstream commit 02fffd1939f6b45892f61822459953ce95e42948 ]
With the FPE preemption-class mapping error now propagated from
stmmac_fpe_map_preemption_class(), tc_setup_dwmac510_mqprio() can fail
on the mapping step. The error path used to call stmmac_reset_tc_mqprio(),
which resets the number of real TX queues to priv->plat->tx_queues_to_use
(the platform maximum), overwriting the value that was active before the
offload was attempted (for example a lower count left over from a previous
mqprio configuration).
The issue can be triggered using the following configuration:
# First mqprio config lowers the hw queue count below the platform
# default (e.g. 8 TX queues).
$tc qdisc add dev eth0 root handle 1: mqprio queues 2@0 2@2
# Replace mqprio configuration with a second one that fails FPE
# preemption-class mapping. stmmac driver resets the real_num_tx_queues
# to the platform maximum, losing the previous configuration.
$tc qdisc replace dev eth0 root handle 2: mqprio queues 2@0 2@2 fp E P
Save ndev->real_num_tx_queues before lowering it and restore it,
together with the TC-to-queue and priority-to-TC mappings, when the FPE
preemption-class mapping fails, instead of resetting the queue count to
the platform maximum.
Note that a failed setup makes the qdisc layer run mqprio_destroy() on
the new qdisc. Because priv->hw_offload is only assigned after
ndo_setup_tc() succeeds, mqprio_destroy() calls netdev_set_num_tc(dev, 0),
so dev->num_tc ends up 0 regardless of the driver-side restore and the
previous qdisc is not reactivated. The restore is still needed to keep
real_num_tx_queues and to avoid leaving the failed configuration's
TC-to-queue and priority-to-TC mappings in place.
Fixes: 195e4f409a40 ("net: stmmac: support fp parameter of tc-mqprio")
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com>
Link: https://patch.msgid.link/20260911-stmmac-tc_setup_dwmac510_mqprio-error-path-v3-2-a76b1e2547c1@oss.qualcomm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com>
Date: Fri Sep 11 10:58:29 2026 +0200
net: stmmac: propagate FPE preemption-class mapping errors
[ Upstream commit 90e4b849dfa6fc8e6c050bcfe1b331b69c015d28 ]
stmmac_fpe_map_preemption_class() dispatches through the
stmmac_do_void_callback() helper, which forces the callback's return
value to 0 whenever the op pointer is populated. As a result the
-EINVAL returned by dwmac5_fpe_map_preemption_class() (e.g. when a
preemptible TC owns more than one TXQ under SP scheduling) is silently
swallowed by every caller.
Switch the dispatch macro to stmmac_do_callback() so the callback's real
result is propagated, and honour it in the taprio and mqprio qdisc
offload.
Note that the taprio "if (ret)" check in tc_taprio_configure() used to
be dead code and now becomes live: a preemptible TC spanning more than
one TXQ under SP scheduling cannot be programmed in hardware, so a
taprio or mqprio configuration that previously returned success while
leaving the preemption-class register unprogrammed now fails with
-EINVAL. For taprio, the failure also runs the disable path, tearing
down the schedule that was just installed; this is the intended
behaviour.
Fixes: 195e4f409a40 ("net: stmmac: support fp parameter of tc-mqprio")
Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com>
Link: https://patch.msgid.link/20260911-stmmac-tc_setup_dwmac510_mqprio-error-path-v3-1-a76b1e2547c1@oss.qualcomm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Guanglei Zhu <zhugl3@xiaopeng.com>
Date: Fri Sep 11 10:17:33 2026 +0800
net: wwan: mhi_wwan_mbim: check skb_copy_bits() return value
commit 31550d585589fde1ae95bf7f7a8188b2d2fdf1c7 upstream.
mhi_mbim_rx() ignores the return value of skb_copy_bits() when it
copies each datagram out of the NTB. The datagram offset and length
come from the DPE, which is only checked to lie within the NTB
itself, so a modem can point a datagram outside the received skb.
The copy then fails and the freshly allocated skbn is passed to
netif_rx() with its uninitialized contents still in place, leaking
kernel heap memory into the network stack.
Free the skb and account an error when the copy fails.
Fixes: aa730a9905b7 ("net: wwan: Add MHI MBIM network driver")
Cc: stable@vger.kernel.org
Suggested-by: Loic Poulain <loic.poulain@oss.qualcomm.com>
Signed-off-by: Guanglei Zhu <zhugl3@xiaopeng.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Verified in a QEMU guest with a fault injector pointing a DPE
outside the received NTB: the copy fails, and the unpatched driver
hands the uninitialized skbn to the network stack (observed as
"unknown protocol" on bytes that were never written). With this
check the failed datagram is dropped and counted as an rx error.
Changes in v2: factor the free-and-count sequence out into
mhi_mbim_rx_drop(), shared with the unknown-protocol path, as
suggested by Loic Poulain.
Link: https://patch.msgid.link/20260911021734.1396599-2-zhugl3@xiaopeng.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Author: Guanglei Zhu <zhugl3@xiaopeng.com>
Date: Fri Sep 11 10:17:32 2026 +0800
net: wwan: mhi_wwan_mbim: guard against a cyclic NDP chain
commit 5d063822ac5184939c1ed377a339a01d8ae814e8 upstream.
The NDP traversal in mhi_mbim_rx() only stops when wNextNdpIndex is
zero. Nothing requires the offsets to advance, so a modem that
points an NDP at itself, or at an earlier NDP, keeps the loop
spinning forever on one CPU.
Break out when the next NDP offset is not larger than the current
one.
Fixes: aa730a9905b7 ("net: wwan: Add MHI MBIM network driver")
Cc: stable@vger.kernel.org
Suggested-by: Loic Poulain <loic.poulain@oss.qualcomm.com>
Signed-off-by: Guanglei Zhu <zhugl3@xiaopeng.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Verified in a QEMU guest with a fault injector feeding the driver's
receive callback an NTB whose single NDP points at itself: the
unpatched driver spins in mhi_mbim_rx() with one CPU pinned at 100%
and the thread never returns. With this check the loop terminates
within one iteration.
Changes in v2: move the non-increasing check to the wNextNdpIndex
retrieval site, as suggested by Loic Poulain, instead of tracking
the previous offset in a separate variable.
Link: https://patch.msgid.link/20260911021734.1396599-1-zhugl3@xiaopeng.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Author: Guanglei Zhu <zhugl3@xiaopeng.com>
Date: Fri Sep 11 10:17:34 2026 +0800
net: wwan: t7xx: validate the netif index in t7xx_ccmni_recv_skb()
commit c7ead9704249d57d4693a04697e3bbd285138fa9 upstream.
The netif index carried in the DPMAIF PIT header is five bits wide,
but ccmni_inst[] only has room for NIC_DEV_MAX (21) entries.
t7xx_ccmni_recv_skb() indexes the array without a bounds check, so
indexes 21 to 31 read past it. The out-of-bounds value lands in the
callback table that follows the array, which is never NULL, so the
existing !ccmni check does not catch it and the driver dereferences
whatever sits there as a struct t7xx_ccmni.
Drop the skb when the index is out of range.
Fixes: 05d19bf500f8 ("net: wwan: t7xx: Add WWAN network interface")
Cc: stable@vger.kernel.org
Signed-off-by: Guanglei Zhu <zhugl3@xiaopeng.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Verified in a QEMU guest with a fault injector setting the netif
index to 25: the unpatched driver reads a value past ccmni_inst[],
which lands in the callback table, and dereferences it far enough to
queue the skb. With this check the packet is dropped. Well-formed
traffic on index 0 is unaffected.
Changes in v2: none.
Link: https://patch.msgid.link/20260911021734.1396599-3-zhugl3@xiaopeng.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Author: Wyatt Feng <wf.kernel.dev@gmail.com>
Date: Sat Aug 29 23:44:32 2026 +0800
net: xfrm: reject unrepresentable espintcp transport headers
commit 96f01b53c2d05e003b040892256de54a586e8529 upstream.
ESP-in-TCP can hand xfrm packets whose transport header offset no longer
fits after the stream parser trims the TCP envelope. The plain transport
header reset truncates that offset and triggers the skb warning path.
Use the careful transport-header helper and drop the skb through the
existing XFRM error path when the offset cannot be represented.
Fixes: e27cca96cd68 ("xfrm: add espintcp (RFC 8229)")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Assisted-by: Codex:GPT-5.4
Signed-off-by: Wyatt Feng <wf.kernel.dev@gmail.com>
Signed-off-by: Ren Wei <weir@nebusec.ai>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Pablo Neira Ayuso <pablo@netfilter.org>
Date: Mon Sep 7 21:04:05 2026 +0200
netfilter: flowtable: hold reference on ct until flow is released
[ Upstream commit e75a9fa1d44bcbd66ea02e8781bcca6ea4076e0d ]
nf_ct_put() releases the ct->ext area inmediately, the rcu typesafe
semantics also allow to refer to the wrong conntrack from the flowtable
datapath. Hold reference on ct until flow is released after rcu grace
period.
Add rcu_barrier() on module exit path, to ensure pending flow entries
are release before module goes away.
Fixes: 0ff90b6c2034 ("netfilter: nf_flow_offload: fix use-after-free and a resource leak")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Pablo Neira Ayuso <pablo@netfilter.org>
Date: Thu Sep 3 01:28:56 2026 +0200
netfilter: nf_nat: unregister and release hooks on error
[ Upstream commit cbdd39ce42530a193c56beb206a3356cb6d01016 ]
If nf_hook_entries_insert_raw() fails, the NAT hooks get never released,
resulting in a memleak.
Postpone setting nat_proto_net->nat_hook_ops when the hooks are
registered to simplify the error path to decide whether the nat hooks
need unwinding.
Fixes: 1cd472bf036c ("netfilter: nf_nat: add nat hook register functions to nf_nat")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Fernando Fernandez Mancera <fmancera@suse.de>
Date: Thu Aug 27 12:32:56 2026 +0200
netfilter: nf_tables: fix device name and prefix match in hook lookup
[ Upstream commit 444e4c88c9c62a3d823069006563513fe7d5aa66 ]
Currently, a netdev chain or flowtable hooked to a device prefix can be
unintentionally deleted by a control-plane request targeting an exact
device name or even a shorter one due to the usage of min() to calculate
the length to match.
Fix this by making sure an exact device match never matches a prefix and
that both the target and the candidate have the same length during
delete operation. The add and update paths retain the existing overlap
matching to prevent a single device from matching multiple hooks.
Reported-by: Wei Fang <void0red@gmail.com>
Closes: https://lore.kernel.org/netfilter-devel/CANE+tVrDeNCHQVmsqkV2ozeBqyE3GtRDMhZgsg1bhw10yGNTRQ@mail.gmail.com/
Fixes: 6d07a289504a ("netfilter: nf_tables: Support wildcard netdev hook specs")
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Theodor Arsenij Larionov Trichkine <theodorlarionov@gmail.com>
Date: Tue Aug 25 12:10:56 2026 +0300
netfilter: nft_nat: fully initialise new_addr in netmap setup
[ Upstream commit d313499df66159b4b7971d760d16729598ab7e5a ]
nft_nat_setup_netmap() builds the mapped address in an on-stack
union nf_inet_addr. For an IPv4 mapping it writes only the 4-byte .ip
member and the loop runs a single 32-bit iteration, but it then copies
the whole 16-byte union into range->min_addr and range->max_addr, so the
upper 12 bytes reach nf_nat_setup_info() uninitialised.
KMSAN reports an uninit-value in nf_nat_setup_info() reached from
nft_nat_eval(). The IPv6 path fills all 16 bytes and is not affected.
Zero-initialise new_addr.
Fixes: 3ff7ddb1353d ("netfilter: nft_nat: add netmap support")
Signed-off-by: Theodor Arsenij Larionov Trichkine <theodorlarionov@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Fri Sep 11 16:08:04 2026 +0000
netlink: do not free nlk->groups while lockless readers can use it
[ Upstream commit ceac0de741bfb47ca255eee075257b3bb31f0651 ]
netlink_realloc_groups() uses krealloc() under netlink_table_grab().
Whenever NLGRPSZ(groups) lands in a different kmalloc bucket, the old
bitmap is freed immediately.
Two readers of nlk->groups / nlk->ngroups do not hold the netlink
table lock:
1) sk_diag_dump_groups(). Hashed (bound) sockets are dumped from the
rhashtable walk in __netlink_diag_dump(), which only holds RCU.
Only the mc_list part of the dump takes nl_table_lock.
2) netlink_native_seq_show() (/proc/net/netlink), whose walk has been
lockless since commit 21e4902aea80 ("netlink: Lockless lookup with
RCU grace period in socket release").
Both can read a freed buffer, and sk_diag_dump_groups() can also read
past the end of the old (smaller) buffer if it happens to load the old
@groups pointer together with the new @ngroups value, copying the
result into a NETLINK_DIAG_GROUPS attribute.
This is the same class of bug that commit f773608026ee ("netlink:
access nlk groups safely in netlink bind and getname") fixed for bind()
and getname(); these two readers were missed. Simply grabbing the table
lock in sk_diag_dump_groups() is not an option, because it is also
called with nl_table_lock already held from the mc_list section of the
dump.
Make the lockless readers safe instead:
- Allocate a new bitmap and free the old one after an RCU grace period,
instead of relying on the implicit kfree() done by krealloc().
- Publish @groups before @ngroups, both with release semantics, and have
the lockless readers load @ngroups first. A reader can then never pair
the new (bigger) size with the old (smaller) buffer, and a reader
picking up the new pointer while still seeing the old size is
guaranteed to see the initialized bitmap.
netlink_realloc_groups() is called from process context (bind() and
setsockopt()), so kfree_rcu_mightsleep() can be used, once the table
has been released.
Fixes: 21e4902aea80 ("netlink: Lockless lookup with RCU grace period in socket release")
Fixes: ad202074320c ("netlink: Use rhashtable walk interface in diag dump")
Reported-by: James Burton <jamesburton@meta.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260911160804.917099-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jeff Layton <jlayton@kernel.org>
Date: Tue Apr 14 07:18:05 2026 -0700
nfsd: fix handling of NFSEXP_PNFS in the netlink codepath
commit f76017a7663c4ce5e379f8a8d39f032bdb1fd865 upstream.
The rework of how block layouts were checked moved the check for
NFSEXP_PNFS out of nfsd4_setup_layout_type() and into the callers. That
patch didn't account for the new call in nfsd4_setup_layout_type().
Cc: Christoph Hellwig <hch@lst.de>
Fixes: da9baa5470dc ("exportfs,nfsd: rework checking for layout-based block device access support")
Cc: stable@vger.kernel.org
Reported-by: Olga Kornievskaia <okorniev@redhat.com>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://patch.msgid.link/20260414-pnfs-exp-fix-v1-1-9face14c16c2@kernel.org
Signed-off-by: Chuck Lever <cel@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Fri Sep 4 14:46:38 2026 +0900
ntfs: account for MFT records added during allocation
[ Upstream commit 631946431ddc66a472c5cc629cd654e62dfa1f88 ]
When no free MFT record is available in the initialized $MFT/$BITMAP,
ntfs_mft_record_alloc() extends $MFT/$DATA and formats the requested record
together with a dynamically sized tail reserve. Those records become
visible through the $MFT file size before charging the requested record to
the free-record counter.
Account for all newly visible records before releasing the MFT allocation
lock, then subtract the one record being allocated. Keep MFT counter
updates independent of the asynchronous free-cluster scan and update the
counter when a record is successfully cleared in the MFT bitmap.
Store the clamped result of the MFT bitmap scan and keep statfs from
exposing an invalid cached count if an accounting error occurs.
Fixes: 115380f9a2f9 ("ntfs: update mft operations")
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Mon Sep 7 13:34:25 2026 +0900
ntfs: ignore interrupted inode reads as corruption
[ Upstream commit 8c5dc7587fdd45f957af81a9adc1e16863f300fc ]
Do not mark the volume in error or report an inode as corrupt when
reading it was interrupted by a signal. -EINTR and -ERESTARTSYS indicate
a transient read failure rather than on-disk NTFS corruption.
Fixes: 115380f9a2f9 ("ntfs: update mft operations")
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Mon Sep 7 13:34:01 2026 +0900
ntfs: propagate folio errors
[ Upstream commit 1923eeffa63edeff427d76fc302bc5eb835771ce ]
Return the error from __filemap_get_folio() instead of replacing it
with -ENOMEM.
Fixes: af0db57d4293 ("ntfs: update inode operations")
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Sun Sep 6 21:37:28 2026 +0900
ntfs: protect runlist updates with the runlist lock
[ Upstream commit 91709ba5d6d709b2b663287b7e871e2c6b480502 ]
ntfs_non_resident_attr_shrink() calls runlist helpers that require the
runlist write lock, but did not hold it while freeing clusters and
truncating the runlist. Serialize those operations and the resident
conversion with the runlist lock.
ntfs_attr_map_cluster() can merge a newly allocated run before updating
mapping pairs. If the update fails, free the clusters and restore both
the in-memory runlist and on-disk mapping pairs from a saved runlist.
Mark the volume in error if either rollback step fails.
Fixes: 495e90fa3348 ("ntfs: update attrib operations")
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Sun Sep 6 11:12:47 2026 +0900
ntfs: repack $MFT/$ATTRIBUTE LIST
[ Upstream commit b1d732e62a5b3942546e4edaab8976258e779287 ]
Repack the non-resident $MFT/$ATTRIBUTE_LIST into a contiguous run
when its mapping pairs no longer fit in the base MFT record. Propagate
allocation and writeback errors, and check synchronous replacement writes.
Fixes: 495e90fa3348 ("ntfs: update attrib operations")
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Namjae Jeon <linkinjeon@kernel.org>
Date: Tue Sep 8 15:52:46 2026 +0900
ntfs: use dynamic MFT tail reservation
[ Upstream commit 6d8c197c9992659a65525a07de4368c8401fdda7 ]
The ntfs MFT allocator historically treated records below 64 as a
permanent extension area and stopped searching for $MFT extent records
after record 400. Windows and ntfs3 do not maintain that on-disk
layout, so an NTFS volume can have free MFT records while ntfs returns
-ENOSPC when $MFT:$DATA needs another mapping-pairs extent.
Use record 24 as the first normal record and maintain an in-memory tail
reserve of up to four initialized records. Normal allocations skip the
reserve, while $MFT metadata extent allocations consume it. When a new
tail is initialized, allocate at least two records and reserve the
following records to avoid recursive allocation during MFT extension.
Existing free runs can seed the reserve on volumes mounted without one.
For $MFT/$DATA, constrain an extent record to a record whose byte offset
is below the new extent lowest VCN byte offset. This preserves bootstrap
reachability without an arbitrary record 400 limit. If no safe record is
available, validate and use reserved records 15, 12, 13, and 14 as
bootstrap candidates while keeping their MFT bitmap entries in use.
This allows existing Windows volumes to extend $MFT using their actual
free records and prevents normal file allocation from consuming
the metadata reserve.
Fixes: 115380f9a2f9 ("ntfs: update mft operations")
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Iván Ezequiel Rodriguez <ivanrwcm25@gmail.com>
Date: Thu Jul 23 15:12:59 2026 -0500
ntsync: reject wait ioctls with zero owner
commit 61611481f7b599e0526a3782c626b1a9deeb48bd upstream.
setup_wait() already validates pad and flags but not owner, while
Documentation/userspace-api/ntsync.rst requires EINVAL when owner is
zero. Reject early before queueing waiters.
Signed-off-by: Iván Ezequiel Rodriguez <ivanrwcm25@gmail.com>
Signed-off-by: Griffin Kroah-Hartman <griffin@kroah.com>
Reviewed-by: Elizabeth Figura <zfigura@codeweavers.com>
Signed-off-by: Elizabeth Figura <zfigura@codeweavers.com>
Link: https://patch.msgid.link/20260723201301.11826-4-zfigura@codeweavers.com
Cc: Alice Ryhl <aliceryhl@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Ulises Mendez Martinez <umendez@google.com>
Date: Fri Sep 4 15:07:09 2026 +0000
objtool: Validate disassembler headers in libopcodes probe
[ Upstream commit 7e61560628d17ea6b1d8ee370f6d42694cff8758 ]
commit 3f2de814c059 ("objtool: Fix libopcodes linking with static libraries")
tested for libopcodes availability by linking a test snippet with a forward
declaration of disassemble_init_for_target().
However, testing symbol linkage with an extern declaration only verifies
the presence of the library (.so/.a) and bypasses checking for development
headers (binutils-dev). On systems where libopcodes is present without
development headers installed, the probe succeeds, enabling BUILD_DISAS.
Subsequent compilation of objtool then fails:
fatal error: 'bfd.h' file not found
113 | #include <bfd.h>
Additionally, the probe invokes $(HOSTCC) without $(HOSTCFLAGS), ignoring
any sysroot or include flags specified for the host compiler.
Fix this by including <bfd.h> and <dis-asm.h> directly in the test snippet,
passing $(HOSTCFLAGS) so host compiler options are respected, and defining
PACKAGE="objtool" to satisfy the configuration check in <bfd.h>.
Fixes: 3f2de814c059 ("objtool: Fix libopcodes linking with static libraries")
Fixes: 436326bc525d ("objtool: fix build failure due to missing libopcodes check")
Reported-by: Alice Ryhl <aliceryhl@google.com>
Assisted-by: Antigravity:Gemini-Next
Signed-off-by: Ulises Mendez Martinez <umendez@google.com>
Link: https://patch.msgid.link/20260904150710.2997558-1-umendez@google.com
Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Zhiling Zou <zhilinz@nebusec.ai>
Date: Sat Sep 12 21:22:43 2026 +0800
openvswitch: avoid reallocating confirmed conntrack labels
commit 3f118c8217c109fd13ca61caa301d72c483897ef upstream.
ovs_ct_get_conn_labels() adds the labels extension when a conntrack
entry does not have one. Confirmed conntracks can be read locklessly,
so adding an extension may reallocate and free the extension block
while another CPU accesses it.
Only add the extension for unconfirmed conntracks. A confirmed
conntrack without labels now fails the caller's label operation instead
of reallocating its extension storage.
Fixes: c2ac66735870 ("openvswitch: Allow matching on conntrack label")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai>
Reviewed-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Aaron Conole <aconole@redhat.com>
Link: https://patch.msgid.link/372fbb062b40ae6723684f55484be86ff0064f8e.1789218015.git.zhilinz@nebusec.ai
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Richard Zhu <hongxing.zhu@nxp.com>
Date: Thu Aug 13 17:50:03 2026 +0800
PCI: imx6: Move clock enable after core reset assertion
commit c9dc7d730319ad64b51570c5387f1fee7b07b510 upstream.
Commit 610fa91d9863 ("PCI: imx6: Assert PERST# before enabling regulators")
inadvertently moved clock enablement before core reset assertion, breaking
PCI device initialization on i.MX6Q Apalis platforms with
ASM1061/ASM1062 SATA controllers connected:
imx6q-pcie 1ffc000.pcie: host bridge /soc/pcie@1ffc000 ranges:
imx6q-pcie 1ffc000.pcie: IO 0x0001f80000..0x0001f8ffff -> 0x0000000000
imx6q-pcie 1ffc000.pcie: MEM 0x0001000000..0x0001efffff -> 0x0001000000
imx6q-pcie 1ffc000.pcie: config reg[1] 0x01f00000 == cpu 0x01f00000
imx6q-pcie 1ffc000.pcie: iATU: unroll F, 4 ob, 4 ib, align 64K, limit 4G
imx6q-pcie 1ffc000.pcie: Link: Only Gen1 is enabled
imx6q-pcie 1ffc000.pcie: Link failed to come up. LTSSM: POLL_CONFIG
imx6q-pcie 1ffc000.pcie: probe with driver imx6q-pcie failed with error -110
NOTE: It is not 100% clear if the issue is specific to the ASM1061/ASM1062
device or on the specific power-up sequence (reset vs cold-power-on).
To fix this regression, restore the original sequence where clocks are
enabled after asserting core reset and configuring the controller type.
Fixes: 610fa91d9863 ("PCI: imx6: Assert PERST# before enabling regulators")
Reported-by: Leonardo Costa <leoreis.costa@gmail.com>
Closes: https://lore.kernel.org/all/bl7i3obu2clzsgeoct2a4mtfhv6typcjdqmgneropf3hpgwve6@n2m5uhlduw57/T/#u
Reported-by: Franz Schnyder <fra.schnyder@gmail.com>
Closes: https://lore.kernel.org/all/t65y5d54axtksbfs7r4olcefqhwm6m4dz3njgnrnf7fcotj74i@o7avoznlafbj/
Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@oss.qualcomm.com>
[bhelgaas: move to pci/for-linus for v7.3]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: stable@vger.kernel.org # 7.2+
Link: https://patch.msgid.link/20260813095003.356062-1-hongxing.zhu@oss.nxp.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shouping Wang <allen.wang@hj-micro.com>
Date: Thu Sep 10 19:46:01 2026 +0800
perf/arm-cmn: Fix wp_dev_sel2 setting for multi-DTM configurations
[ Upstream commit 49daa3d668b69a5454b5aba0078848a479f79f1c ]
When MXP_MULTIPLE_DTM_EN is TRUE, each DTM will monitor at most
two device ports. In this case, {wp_dev_sel2, wp_dev_sel} will
only use values 2'b00 and 2'b01 per DTM.
Previously the setting allowed values beyond the supported range
per DTM, which could cause each DTM to select invalid ports when
MXP_MULTIPLE_DTM_EN is TRUE.
Fix this by only setting CMN_DTM_WPn_CONFIG_WP_DEV_SEL2 when
!multi_dtm.
Fixes: 60d1504070c2 ("perf/arm-cmn: Support new IP features")
Signed-off-by: Shouping Wang <allen.wang@hj-micro.com>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Vinay Belgaumkar <vinay.belgaumkar@intel.com>
Date: Fri Sep 4 11:16:24 2026 -0700
perf: Fix null pointer access in is_include_guest_event()
[ Upstream commit 88aed0422f39b22406f35f1e758cea25e7bbcfb5 ]
A typical module unload occurring event when there is an active perf
connection leads to freeing of the pmu pointer. The call log is something
like:
..
__pmu_detach_event
pmu_detach_event
pmu_detach_events
perf_pmu_unregister
..
__pmu_detach_event() sets event->pmu to null. When the perf connection
finally is closed, the following stack trace is observed:
Oops: general protection fault, kernel NULL pointer dereference
...
RIP: 0010:_free_event+0x3e/0x370
...
Call Trace:
...
perf_event_release_kernel+0x260/0x2d0
perf_release+0x12/0x20
A call to mediated_pmu_unaccount_event() inside _free_event() is the root
cause of this crash. Adding a check inside is_include_guest_event() ensures
we don't accidentally access a null pmu ptr. In addition to this, we will
now call mediated_pmu_unaccount_event() before clearing the pmu ptr so that
nr_include_guest_events counts are maintained correctly.
Fixes: eff95e170275 ("perf: Add APIs to create/release mediated guest vPMUs")
Assisted-by: Claude:Claude-Sonnet-5
Signed-off-by: Vinay Belgaumkar <vinay.belgaumkar@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Link: https://patch.msgid.link/20260904181625.1394082-1-vinay.belgaumkar@intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Date: Fri Sep 11 09:40:14 2026 +0200
phy: mediatek: phy-mtk-hdmi-mt8195: Fix PLL calc divisor overflow
[ Upstream commit de7f29a1fe1dc2864d8a47f8c39d508442cae167 ]
When trying to calculate a PLL rate for target display resolutions
above 2560x1440, 24bpp, 30Hz, the pixel clock value will be more
than 32-bits long but the division to finally calculate the digital
clock divider is being done with div_u64(), which expects a 32bit
unsigned divisor.
Fix the overflow by using div64_u64() instead.
Fixes: 9d9ff3d2a4a5 ("phy: mediatek: hdmi: mt8195: fix wrong pll calculus")
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@oss.qualcomm.com>
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://patch.msgid.link/20260911074015.9994-2-angelogioacchino.delregno@collabora.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Date: Fri Sep 11 09:40:15 2026 +0200
phy: mediatek: phy-mtk-hdmi-mt8195: Fix TMDS clk bit ratio setting
[ Upstream commit 486a70ef848264dcf9a57f0bb0452848db9537de ]
The comment in the mtk_phy_tmds_clk_ratio() function clearly and
correctly explains that the TMDS ratio has to be 1/10 for data
rates under 3.4Gbps, and 1/40 over that.
Unfortunately though, the TXC_DIV register setting was wrong, as
in value 3 means to divide by 8 and, in order to achieve the in
spec 1/40 (tmds) data rate, this has to divide by 4 instead!
Add definitions for the TXC_DIV register values clearly explaining
the meanings (DIV2, DIV4, DIV8), and program the correct, DIV 4,
value to the register in mtk_phy_tmds_clk_ratio().
This fixes out of spec clocking and, with this change, SoCs using
the MT8195 class HDMI PHYs can now successfully be configured to
output 3840x2160@60Hz over HDMI.
Fixes: 45810d486bb4 ("phy: mediatek: add support for phy-mtk-hdmi-mt8195")
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@oss.qualcomm.com>
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://patch.msgid.link/20260911074015.9994-3-angelogioacchino.delregno@collabora.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
Date: Thu Jul 16 21:32:46 2026 +0300
phy: renesas: rcar-gen3-usb2: Avoid long delay in atomic context
commit 48e97c59a49c9e90270f5e3d221db253b76de5da upstream.
The OTG PHY initialization sequence needs to wait for 20 ms at a specific
step, as described in commit 72c0339c115b ("phy: renesas:
rcar-gen3-usb2: follow the hardware manual procedure").
Commit 55a387ebb921 ("phy: renesas: rcar-gen3-usb2: Lock around hardware
registers and driver data") tried to address various problems in the
rcar-gen3-usb2 driver and converted the mutex protecting HW register
accesses to a spin lock, leaving, however, a long delay in the critical
section protected by the spin lock. This may become a problem,
especially on RT kernels.
To address this, release the spin lock before sleeping for 20 ms as
required by the HW manual and reacquire it afterwards. To avoid other
threads entering the critical section and configuring the HW while the
software is waiting for the OTG initialization to complete, introduce the
otg_initializing variable alongside the otg_init_done wait queue. Any
other thread trying to configure the HW while the OTG PHY initialization
is in progress waits for the wait queue instead of immediately returning
errors to PHY users. The IRQs were also disabled while waiting for the OTG
PHY initialization to complete, as the interrupt handler may also apply HW
settings.
The OTG can only be initialized once. It is initialized by the first PHY
that calls struct phy_ops::rcar_gen3_phy_usb2_init().
To avoid failures when multiple PHYs call struct
phy_ops::rcar_gen3_phy_usb2_init() simultaneously, and the PHY responsible
for initializing the OTG either fails or deinit quiqly and another PHY
takes over the PHY init role), the code waiting for the
channel->otg_init_done wait queue retries up to NUM_OF_PHYS times.
Fixes: 55a387ebb921 ("phy: renesas: rcar-gen3-usb2: Lock around hardware registers and driver data")
Cc: stable@vger.kernel.org
Reported-by: Pavel Machek <pavel@nabladev.com>
Closes: https://lore.kernel.org/all/afhkX2Ys2BG1gnqy@duo.ucw.cz
Reported-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org>
Closes: https://lore.kernel.org/all/afhkX2Ys2BG1gnqy@duo.ucw.cz
Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@oss.qualcomm.com>
Link: https://lore.kernel.org/all/afhkX2Ys2BG1gnqy@duo.ucw.cz
Link: https://patch.msgid.link/20260716183246.3183877-1-claudiu.beznea+renesas@tuxon.dev
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Thomas Gleixner <tglx@kernel.org>
Date: Wed Sep 16 20:48:30 2026 +0200
posix-cpu-timers: Prevent freeing a timer which is queued on the expiry list
commit c21eaa72f02fc6e85621cbe09d303d8fb8bd39cd upstream.
Kijo analyzed another race in the POSIX CPU timer code:
Commit bf635681c906 converted cpu_timer::firing from a tristate value to a
boolean. This lost the distinction between "not owned by the firing list"
and "still owned, but delivery was canceled". The resulting race is:
expiry handler timer_settime() timer_delete()
-------------- --------------- --------------
collect timer onto
private firing list
firing = true
observes firing = true
firing = false
return TIMER_RETRY
wait for handler
observes firing = false
finish deletion
unhash and free timer
resume list traversal
read freed elist.next
-> UAF
The firing bit is clearly the wrong indicator since that commit.
Check whether the timer is queued on the expiry list or not instead. If it
is queued clear the firing bit to prevent signal delivery as before and
return TIMER_RETRY so the caller unlocks the timer which allows the expiry
code to make progress and remove it from the list.
Fixes: bf635681c906 ("posix-cpu-timers: Cleanup the firing logic")
Reported-by: Kijo Park <red993688@gmail.com>
Debugged-by: Kijo Park <red993688@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Kijo Park <red993688@gmail.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Date: Wed Sep 9 14:37:07 2026 +0200
power: sequencing: don't call .post_enable() if pwrseq_unit_enable() failed
commit 5f90f85eae4e9d2e9628b2019870994ba830b533 upstream.
If the call to pwrseq_unit_enable() failed in pwrseq_enable(), bail out
instead of calling target->post_enable() which assumes the target was
successfully enabled.
Fixes: 249ebf3f65f8 ("power: sequencing: implement the pwrseq core")
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260909-pwrseq-kunit-v2-1-ef496afc89d2@oss.qualcomm.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Biju Das <biju.das.jz@bp.renesas.com>
Date: Wed Aug 26 13:27:24 2026 +0100
power: sequencing: Fix build issue with COMPILE_TEST
[ Upstream commit 3b54dbd119805361695cb50ca6a875f4c7518b74 ]
The POWER_SEQUENCING_TH1520_GPU driver depends on
(ARCH_THEAD && AUXILIARY_BUS) || COMPILE_TEST. This means when
COMPILE_TEST=y and ARCH_THEAD is not set, the driver can still be
built even though it requires AUXILIARY_BUS, which may not be
selected in that configuration, leading to a build failure.
Fix this by dropping AUXILIARY_BUS from the dependency and instead
selecting it directly, so the dependency is satisfied regardless of
whether COMPILE_TEST or ARCH_THEAD is enabled.
Fixes: 1a7312b93ab0 ("power: sequencing: extend build coverage with COMPILE_TEST=y")
Signed-off-by: Biju Das <biju.das.jz@bp.renesas.com>
Link: https://patch.msgid.link/20260826122742.153643-3-biju.das.jz@bp.renesas.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Date: Wed Sep 9 14:37:09 2026 +0200
power: sequencing: fix NULL-pointer dereference in pwrseq_device_register()
commit 242da4318d97380741516b595af3920207b2f0f1 upstream.
If dev_set_name() fails in pwrseq_device_register(), we jump to the
err_put_pwrseq label before initializing pwrseq->targets.
pwrseq_release() will try to iterate over targets unconditionally and
subsequently dereference an invalid pointer. Move the call to
dev_set_name() after the list head is initialized.
Fixes: 249ebf3f65f8 ("power: sequencing: implement the pwrseq core")
Cc: stable@vger.kernel.org
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Closes: https://sashiko.dev/#/patchset/20260903-pwrseq-kunit-v1-0-1f893d2cabc2%40oss.qualcomm.com?part=2
Link: https://patch.msgid.link/20260909-pwrseq-kunit-v2-3-ef496afc89d2@oss.qualcomm.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Date: Wed Sep 9 14:37:08 2026 +0200
power: sequencing: fix NULL-pointer dereference in pwrseq_unit_new()
commit 115b303e8e093d964089ec6f3c40d984d77b33d0 upstream.
If memory allocation fails in pwrseq_unit_setup_deps(), pwrseq_unit_put()
is called to release the partially initialized unit. However, we've
never initialized unit->list and pwrseq_unit_release() will
unconditionally call list_del() on it. Initialize unit->list right after
allocating the unit struct.
Fixes: 249ebf3f65f8 ("power: sequencing: implement the pwrseq core")
Cc: stable@vger.kernel.org
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Closes: https://sashiko.dev/#/patchset/20260903-pwrseq-kunit-v1-0-1f893d2cabc2%40oss.qualcomm.com?part=1
Link: https://patch.msgid.link/20260909-pwrseq-kunit-v2-2-ef496afc89d2@oss.qualcomm.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shivaprasad G Bhat <sbhat@linux.ibm.com>
Date: Tue Sep 15 22:04:17 2026 +0530
powerpc/iommu: Fix the overflow validation in iommu_tce_check_ioba
[ Upstream commit 0b271f7d7f5ed45bc498a03ce0aa9cfd8402fc71 ]
The commit b1af23d836f8 ("KVM: PPC: iommu: Unify TCE checking") unified
IOBA parameter checking across KVM and VFIO into iommu_tce_check_ioba().
While doing so, the passed in argument npages is ignored and constant
value '1' is used leaving out a possible overflow as the callers can
legitimately be using npages > 1 for H_STUFF_TCE or H_PUT_TCE_INDIRECT
cases.
Fix this by accounting for 'npages', checking for arithmetic overflow,
and verifying that the entire requested range (ioba - offset + npages)
does not exceed the table capacity 'size'.
Fixes: b1af23d836f8 ("KVM: PPC: iommu: Unify TCE checking")
Reviewed-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Tested-by: R Nageswara Sastry <rnsastry@linux.ibm.com>
Signed-off-by: Shivaprasad G Bhat <sbhat@linux.ibm.com>
Signed-off-by: Gautam Menghani <gautam@linux.ibm.com>
Signed-off-by: Madhavan Srinivasan <maddy@linux.ibm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Sat Sep 12 23:30:48 2026 +0000
pppoatm: ensure a writable skb header and linear data
[ Upstream commit ecc7253683a3c55caa868ce0ee530fcb0044bd3c ]
In pppoatm_send(), LLC encapsulation checks whether there is sufficient
headroom for the 4-byte LLC header, but does not ensure that the skb header
is writable.
Normal transmit packets passing through ppp_start_xmit() have their header
unshared via skb_cow_head(). However, packets can also reach pppoatm_send()
via PPP channel bridging (PPPIOCBRIDGECHAN) without going through
ppp_start_xmit().
Use skb_cow_head() to ensure both sufficient headroom and a writable
header before pushing the LLC header.
While at it:
- Call pskb_may_pull(skb, 1) before inspecting skb->data[0] to prevent
out-of-bounds reads on zero-length or non-linear frames (e.g. from
bridging).
- Defer SC_COMP_PROT protocol compression until after pppoatm_may_send()
succeeds. This eliminates the temporary skb allocation on admission failure
and completely removes the fragile "undo" heuristic at the nospace label,
avoiding any risk of reading uninitialized headroom or performing an
unbalanced skb_push().
Fixes: 4cf476ced45d ("ppp: add PPPIOCBRIDGECHAN and PPPIOCUNBRIDGECHAN ioctls")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260912233048.3977192-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Date: Mon Aug 24 22:54:43 2026 +0530
RDMA/bnxt_re: Avoid exposing umdbr to userspace
[ Upstream commit 23d7e03a52ece66b992620aa9b8fa5f164077c0f ]
The umdbr field in struct bnxt_re_db_region returns the raw
unmapped PCI BAR address of the doorbell region. Avoid sharing
this field to the userspace. Change this to a reserved field
and stop populating it, keeping the ABI layout and size
unchanged for existing binaries.
Fixes: 1234a9d8aebb ("RDMA/bnxt_re: Support doorbell extensions")
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Link: https://patch.msgid.link/20260824172443.33943-1-sriharsha.basavapatna@broadcom.com
Reviewed-by: Selvin Xavier <selvin.xavier@broadcom.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Linkai Gong <gonglinkai@kylinos.cn>
Date: Fri Aug 21 17:09:08 2026 +0800
RDMA/bnxt_re: check create_singlethread_workqueue() in DCB setup
[ Upstream commit 6c368f7baaea63c1c7c28c6df271511f2a1562c9 ]
bnxt_re_init_dcb_wq() ignores a failed allocation. The async DCB
handler later calls queue_work() on the NULL pointer.
Fixes: 51dc5312dcd9 ("RDMA/bnxt_re: Add support to handle DCB_CONFIG_CHANGE event")
Signed-off-by: Linkai Gong <gonglinkai@kylinos.cn>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jeffin Philip <jeffinphilip14@gmail.com>
Date: Fri Sep 4 18:44:37 2026 +0530
RDMA/core: fix refcount bug in iwpm_get_nlmsg_request()
commit 33fb59da49c4c3f5c2ec9f9d4447a56857a02c02 upstream.
iwpm_get_nlmsg_request() initializes refcount _after_ list_add_tail()
making it accessible to global list where another CPU can kref_get()
on nlmsg_request causing a refcount "addition on 0" bug. Fix this
by initializing kref _before_ list_add_tail() so refcount for
nlmsg_request can be incremented/decremented normally. In addition,
also initialize every field before list_add_tail().
Reported-by: syzbot+bd317784d628820741b5@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=bd317784d628820741b5
Fixes: 30dc5e63d6a5 ("RDMA/core: Add support for iWARP Port Mapper user space service")
Cc: stable@vger.kernel.org
Signed-off-by: Jeffin Philip <jeffinphilip14@gmail.com>
Link: https://patch.msgid.link/20260904131437.12917-1-jeffinphilip14@gmail.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Krystian Kaniewski <krystianmkaniewski@gmail.com>
Date: Wed Aug 12 10:16:41 2026 +0200
RDMA/core: Reject unregistering netdevs in ib_get_eth_speed
[ Upstream commit ef9fbe1b93f3b617b96e86d5cd76b3fa44514cb5 ]
ib_device_get_netdev() intentionally returns a referenced net_device even
when it is unregistering, so matching and cleanup callers can still find
the association. The reference keeps struct net_device allocated, but does
not guarantee that the device remains operational.
ib_get_eth_speed() uses the returned device operationally by invoking its
ethtool callback. Although that call is made under RTNL, the function does
not verify the registration state first. An asynchronous RDMA port query
can therefore call into a netdev after NETDEV_UNREGISTER and ndo_uninit
have completed.
Check for NETREG_REGISTERED while holding RTNL and return -ENODEV for a
device which is being unregistered. Keeping RTNL across the check and the
ethtool operation prevents unregister from starting between them.
Keep the speed fallback and warning under RTNL as well, so the warning can
safely read netdev->name. Drop the netdev reference before releasing RTNL
once all accesses to the device are complete.
Fixes: d41861942fc5 ("IB/core: Add generic function to extract IB speed from netdev")
Reported-by: syzbot+5fe14f2ff4ccbace9a26@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=5fe14f2ff4ccbace9a26
Signed-off-by: Krystian Kaniewski <krystianmkaniewski@gmail.com>
Link: https://patch.msgid.link/20260812081708.32468-1-krystianmkaniewski@gmail.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Leon Romanovsky <leon@kernel.org>
Date: Thu Sep 10 09:35:13 2026 -0400
RDMA/efa: Keep admin queues alive while IRQ is registered
[ Upstream commit e08aca85c02ff290f785f07acae758f0daf5f49e ]
The management IRQ handler accesses both the admin completion queue and the
async event queue. The driver registered the IRQ before constructing these
queues and destroyed them before freeing the IRQ, so the handler's lifetime
was not contained by the resources it accesses.
Initialize the queues with interrupts masked, request the IRQ, and then
switch to interrupt mode. On removal, reset the device and free the IRQ
before destroying the queues. Also reset the device before destroying the
queues if IRQ registration fails, because the device already has their DMA
addresses.
Fixes: b7f5e880f377 ("RDMA/efa: Add the efa module")
Link: https://patch.msgid.link/20260907-use-after-free-of-admin-queue-struct-v1-1-dd9d9267fbf4@nvidia.com
Reviewed-by: Michael Margolin <mrgolin@amazon.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Leon Romanovsky <leon@kernel.org>
Date: Thu Sep 10 09:35:13 2026 -0400
RDMA/efa: Keep EQ resources alive while IRQ is registered
[ Upstream commit e22a3627b7151754f07f90ea3d1ab6e85f5d93f4 ]
The completion IRQ handler accesses the EQ state and DMA buffer. Its IRQ was
registered before that state was initialized, while teardown released the
buffer before free_irq() synchronized the handler.
Initialize the EQ without arming it, register the IRQ, and then arm it.
Reverse the resource order during teardown by freeing the IRQ before
destroying the EQ.
Fixes: 2a152512a155 ("RDMA/efa: CQ notifications")
Link: https://patch.msgid.link/20260907-use-after-free-of-admin-queue-struct-v1-2-dd9d9267fbf4@nvidia.com
Reviewed-by: Michael Margolin <mrgolin@amazon.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Cheng Xu <chengyou@linux.alibaba.com>
Date: Fri Aug 28 11:03:44 2026 +0800
RDMA/erdma: Use IRQ-safe XArray helpers for QP and CQ tables
[ Upstream commit 00baeade709fb66da647e8327e8398bb532e30f7 ]
Locked QP and CQ lookups from EQ interrupts can deadlock with
create-path XArray updates. If an interrupt arrives while the create
path holds the plain xa_lock, the lookup spins forever trying to
acquire the same lock.
Use IRQ-safe XArray helpers for all QP and CQ create-path updates,
including the GSI QP store and error paths. Initialize both arrays with
XA_FLAGS_LOCK_IRQ so sleeping allocations preserve interrupt state.
Fixes: 98df2aee1459 ("RDMA/erdma: Hold CQ references when processing EQ events")
Fixes: a52eeff32024 ("RDMA/erdma: Hold QP references for AE and CM processing")
Signed-off-by: Cheng Xu <chengyou@linux.alibaba.com>
Link: https://patch.msgid.link/20260828030344.88021-1-chengyou@linux.alibaba.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jacob Moroni <jmoroni@google.com>
Date: Tue Sep 1 16:00:14 2026 +0000
RDMA/irdma: Enforce local fence for IB_WR_REG_MR
[ Upstream commit 3fb905f07ea45b31c8f67ba6e4668de46f527e65 ]
Enforce local fence for IB_WR_REG_MR to avoid spurious
FASTREG_VALID_MKEY async events during heavy invalidation
and registration activity.
Commit 69e8e429bca2 ("RDMA/irdma: Enforce local fence for LOCAL_INV WRs")
was very similar, but was not sufficient to prevent all occurrences
of these async events.
Fixes: b48c24c2d710 ("RDMA/irdma: Implement device supported verb APIs")
Signed-off-by: Jacob Moroni <jmoroni@google.com>
Link: https://patch.msgid.link/20260901160014.2026285-1-jmoroni@google.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Li RongQing <lirongqing@baidu.com>
Date: Wed Aug 26 15:32:16 2026 +0800
RDMA/mad: Fix receive buffer leak when PKey enforcement fails
[ Upstream commit 3476c28c9addfa253f505e6bd87f1f5598b961d0 ]
ib_mad_complete_recv() initializes mad_recv_wc->rmpp_list and then runs
ib_mad_enforce_security() before linking recv_buf onto that list. On
failure it calls ib_free_recv_mad(), which only walks rmpp_list and frees
the ib_mad_private of every buffer found there. As the list is still
empty at that point, nothing is freed at all.
The caller cannot clean up either: ib_mad_recv_done() sets recv to NULL
right after ib_mad_complete_recv() returns, assuming the MAD layer took
ownership of the buffer. Every MAD that fails the PKey check therefore
leaks one ib_mad_private (about 300 bytes per IB port MAD, ~2K for OPA),
and a remote node can trigger this repeatedly by sending MADs with a
wrong PKey.
Link recv_buf onto rmpp_list right after the list is initialized, so the
error path has something to free.
Fixes: 47a2b338fe63 ("IB/core: Enforce security on management datagrams")
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Link: https://patch.msgid.link/20260826073216.2367-1-lirongqing@baidu.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Or Har-Toov <ohartoov@nvidia.com>
Date: Tue Aug 11 19:25:57 2026 +0300
RDMA/mlx5: Remove warn on missing representor in query_port_speed
[ Upstream commit 2be77295316c2dff0a33c0dc3a65abdab4ccf796 ]
The representor ib_device's phys_port_cnt is set to the total vport
count when the uplink vport rep loads. Individual port[i].rep entries
are populated only as each VF/SF vport rep registers. A NULL .rep for
a given port index is therefore expected while VF reps are still
loading or haven't been enabled yet.
Tools like ibstat and ibv_devinfo iterate over all ports of all RDMA
devices. Some ports may not have an eswitch representor, causing
repeated dmesg warnings when these tools run without a device argument.
This causes dmesg to be flooded with this message on every ibstat
invocation.
Remove the warning and return -ENODEV when no representor exists for
the queried port.
Fixes: aaecff5e13cd ("RDMA/mlx5: Implement query_port_speed callback")
Signed-off-by: Or Har-Toov <ohartoov@nvidia.com>
Reviewed-by: Shay Drory <shayd@nvidia.com>
Signed-off-by: Edward Srouji <edwards@nvidia.com>
Link: https://patch.msgid.link/20260811-remove-warn-on-miss-rep-v1-1-eccf399bc6af@nvidia.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Quanye Yang <quanyeyang@proton.me>
Date: Sun Aug 30 15:09:55 2026 +0800
RDMA/rtrs-clt: Fix CQ pool leak when connect is interrupted
[ Upstream commit 2ae16aaa78b5edc6e6d0904c84fd9cdfb762bcda ]
The client borrows shared CQ credits in the ADDR_RESOLVED handler via
ib_cq_pool_get(), before the peer is connected. create_cm() can return
-ERESTARTSYS from wait_event_interruptible_timeout() without destroying
the CM ID. The init_conns() and stop-and-destroy paths then call
destroy_con_cq_qp() while cq is still NULL (no PUT) and only afterwards
rdma_destroy_id().
CMA serializes the handler against rdma_destroy_id() with handler_mutex,
but that does not order the GET against destroy_con_cq_qp(). If
ADDR_RESOLVED has already passed the DESTROYING check, it can take
con_mutex, GET credits, and then lose the con to kfree. Device
unregister later hits WARN_ON(cq->cqe_used) in ib_cq_pool_cleanup().
Set a per-connection flag under con_mutex before CQ/QP teardown so a
racing ADDR_RESOLVED cannot borrow credits after teardown has begun.
Reported-by: syzbot+d396918a29afb8543e1c@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=d396918a29afb8543e1c
Fixes: 3b89e92c2a95 ("RDMA/rtrs: Use new shared CQ mechanism")
Signed-off-by: Quanye Yang <quanyeyang@proton.me>
Link: https://patch.msgid.link/20260830-rdma-rtrs-clt-cq-pool-leak-v1-1-b169434fd3df@proton.me
Reviewed-by: Jack Wang <jinpu.wang@cloud.ionos.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ryan Mehri <ryan.mehri1@gmail.com>
Date: Sat Aug 22 23:33:43 2026 -0400
RDMA/rtrs: guard against null kobj name
[ Upstream commit 99c24a8968ebef0573825b5cb89d5985d51635b9 ]
In the client, if `init_path()` errors, the callee tries to clean up
with `rtrs_clt_close_conns()`. However, this can lead to calling the
event tracing code with `clt_path->kobj->name` being `NULL` and thus
causing a null pointer dereference when trying to copy from it.
This just adds a guard to check that the name is not `NULL` before
copying from it. The server appears to have a similar pattern.
Fixes: 5a93929d9f9a1 ("RDMA/rtrs-clt: Add event tracing support")
Fixes: c16762b7bf54d ("RDMA/rtrs-srv: Add event tracing support")
Reported-by: syzbot+1695193198994f4e7fed@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=1695193198994f4e7fed
Signed-off-by: Ryan Mehri <ryan.mehri1@gmail.com>
Link: https://patch.msgid.link/20260823034303.163403-1-ryan.mehri1@gmail.com
Reviewed-by: Jack Wang <jinpu.wang@cloud.ionos.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Gang Yan <yangang@kylinos.cn>
Date: Fri Aug 14 17:37:40 2026 +0800
RDMA/rxe: Fix integer overflow in mr_check_range() leading to OOB access
[ Upstream commit d10e2a08799e858d3e71ea4169bcd018f216d444 ]
mr_check_range() validates that [iova, iova+length) falls within the
registered MR range using wraparound-prone arithmetic:
if (iova < mr->ibmr.iova ||
iova + length > mr->ibmr.iova + mr->ibmr.length)
A remote peer can craft an RDMA-Write/Read RETH so that iova + length
wraps to 0 (e.g. iova=0xfffffffffffffff8, length=8), bypassing the
check. rxe_mr_iova_to_index() then computes a huge index (int idx, only
guarded by WARN_ON) and rxe_mr_copy_xarray() dereferences
mr->page_info[huge], causing an out-of-bounds read/write and a kernel
oops that is triggerable by an unauthenticated remote peer.
Rewrite the check in overflow-safe form; the first two clauses guarantee
that the subsequent subtractions do not underflow:
if (iova < mr->ibmr.iova ||
length > mr->ibmr.length ||
iova - mr->ibmr.iova > mr->ibmr.length - length)
With the fix, mr_check_range() returns -EINVAL for the crafted iova and
the responder reports REMOTE_ACCESS_ERROR instead of triggering the OOB.
Fixes: 8700e3e7c485 ("Soft RoCE driver")
Signed-off-by: Gang Yan <yangang@kylinos.cn>
Link: https://patch.msgid.link/20260814093740.292954-1-gang.yan@linux.dev
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Reviewed-by: Shukai Ni <shukai.ni@kuleuven.be>
Tested-by: Shukai Ni <shukai.ni@kuleuven.be>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Michael Bommarito <michael.bommarito@gmail.com>
Date: Tue Jun 16 22:27:28 2026 -0400
RDMA/rxe: insert mcg into mcg_tree only after rxe_mcast_add() succeeds
[ Upstream commit 1caceeb2d74bbe88223aea55eb8626b4c5f076fd ]
rxe_get_mcg() publishes a newly allocated multicast group in
rxe->mcg_tree before programming the backing Ethernet multicast address
with rxe_mcast_add(), which runs outside mcg_lock. A local userspace
RDMA client reaches this path with ATTACH_MCAST on a UD QP; if
rxe_mcast_add() then returns an error (for example -ENODEV when the
backing netdev has been removed, or a propagated dev_mc_add() error),
the unwind frees the published group without removing it from the tree.
A later lookup of the same MGID dereferences the freed struct rxe_mcg
from __rxe_lookup_mcg().
Fix this by keeping the new mcg private until rxe_mcast_add() succeeds.
Split the tree publication into __rxe_publish_mcg(), call rxe_mcast_add()
before taking the tree reference, and free the still-private mcg on
failure. Because the group is never visible in mcg_tree until the
multicast address is programmed, no concurrent caller can look it up or
attach a QP to a group that is about to be torn down, so the error path
needs no conditional unwind. If another caller publishes the same MGID
while the address is being programmed, the post-add re-check under
mcg_lock finds the winner; this caller then drops its private object and
balances its own rxe_mcast_add() with rxe_mcast_del() before returning
the winner.
Reproduced by forcing the rxe_mcast_add() error return under KASAN:
without the change the next attach to the same MGID reports a
slab-use-after-free in __rxe_lookup_mcg(); with it the forced failure
returns cleanly. A no-injection attach/detach regression, including a
two-QP shared join/leave and re-attach, stays KASAN- and leak-clean.
Fixes: a926a903b7dc ("RDMA/rxe: Do not call dev_mc_add/del() under a spinlock")
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Link: https://patch.msgid.link/20260617022728.2770116-1-michael.bommarito@gmail.com
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Weiming Shi <bestswngs@gmail.com>
Date: Sun Jul 26 19:15:30 2026 +0800
RDMA/rxe: Restore HMM_PFN_WRITE check in ODP write paths
[ Upstream commit 769001ce838d907ecaa95f1d0a4e8fc86f761f9f ]
Commit 0b261d7c1cd3 ("RDMA/rxe: Break endless pagefault loop for RO
pages") dropped the access permission test from rxe_check_pagefault()
and left only HMM_PFN_VALID. A page faulted in read-only, for example
a page-cache folio behind a PROT_READ file mapping, then satisfies the
check and ODP write operations (RDMA WRITE, RDMA READ response, SEND
payload, atomics) modify it through kmap without ever breaking CoW.
An unprivileged user can register an ODP MR over such a mapping and
have incoming RDMA traffic overwrite the page cache of a file it only
holds O_RDONLY, including /etc/passwd or setuid binaries. This is the
same primitive class as Dirty COW and CVE-2022-2590.
mlx5 has the missing invariant: its ODP path sets the device write bit
only for pfns that carry HMM_PFN_WRITE. Restore it in rxe by requiring
HMM_PFN_WRITE in rxe_check_pagefault() for every operation except
RXE_PAGEFAULT_RDONLY. A write to a non-writable VMA now fails the one
fault attempt with -EPERM from hmm_vma_fault() instead of re-faulting
forever. For a writable VMA the fault breaks CoW and the write lands
in the private page.
Keep pmem flushes on the read-only check. arch_wb_cache_pmem() never
modifies memory, and the FLUSH access bits do not make the umem
writable, so classifying flushes as writes would make every flush
against a flush-only MR fail.
Fixes: 0b261d7c1cd3 ("RDMA/rxe: Break endless pagefault loop for RO pages")
Signed-off-by: Weiming Shi <bestswngs@gmail.com>
Link: https://patch.msgid.link/20260726111533.1037819-1-bestswngs@gmail.com
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Tested-by: Hongqiang Luo <wanbafv@gmail.com>
Tested-by: Xinyu Ma <mmmxny@gmail.com>
Tested-by: Zhanbo Ye <cainyzb@gmail.com>
Reported-by: Weiming Shi <bestswngs@gmail.com>
Reported-by: Shaomin Chen <eeesssooo020@gmail.com>
Reported-by: Rui Ding <threonine42@gmail.com>
Reported-by: Miao Zhao <muel@nova.gal>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Norbert Szetei <norbert@doyensec.com>
Date: Thu Aug 27 19:18:07 2026 +0200
RDMA/rxe: validate access flags before swapping the MR's PD
[ Upstream commit ae36a5b609ae79f4de966328b78d2584be9719a4 ]
rxe_rereg_user_mr() reassigns mr->ibmr.pd first and only then
validates the IB_MR_REREG_ACCESS argument:
if (flags & IB_MR_REREG_PD) {
rxe_put(old_pd);
rxe_get(pd);
mr->ibmr.pd = ibpd;
}
if (flags & IB_MR_REREG_ACCESS) {
if (access & ~RXE_ACCESS_SUPPORTED_MR)
return ERR_PTR(-EOPNOTSUPP);
mr->access = access;
}
Both flags pass the entry check because RXE_MR_REREG_SUPPORTED is
IB_MR_REREG_PD | IB_MR_REREG_ACCESS, so a caller can reach the access
check with mr->ibmr.pd already reassigned.
mr->ibmr.pd is owned by the core, which adjusts pd->usecnt only on the
success path: ib_uverbs_rereg_mr() jumps to put_new_uobj on a driver error
without undoing the reassignment, so mr->pd == new_pd while the usecnts
still charge the MR to orig_pd. ib_dereg_mr_user() then decrements
new_pd, whose count can reach zero while a memory window still references
it; uverbs_free_pd() frees the PD on that count alone and rxe_mw_cleanup()
writes to freed memory:
BUG: KASAN: slab-use-after-free in __rxe_put+0x31/0xa0
Write of size 4 at addr ffff8881301dd690 by task rxe_poc/591
__rxe_put+0x31/0xa0
rxe_mw_cleanup+0x42/0x200
__rxe_cleanup+0x115/0x370
rxe_dealloc_mw+0x4c/0x80
Allocated by task 591:
ib_uverbs_alloc_pd+0x258/0x540
Freed by task 591:
ib_dealloc_pd_user+0x174/0x210
uverbs_free_pd+0x8d/0xc0
ib_uverbs_dealloc_pd+0x18e/0x1d0
Validate the access flags before mutating any state so the callback either
applies every requested change or none.
Fixes: 544c7f62cf32 ("RDMA/rxe: Implement rereg_user_mr")
Signed-off-by: Norbert Szetei <norbert@doyensec.com>
Link: https://patch.msgid.link/46E1D5C0-24BE-4D01-BDB3-634FE09B22C5@doyensec.com
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jérémy Jean <Jeremy.Jean@oss.cyber.gouv.fr>
Date: Tue Sep 8 08:55:20 2026 +0000
RDMA/siw: Bound fragmented header copies by the remaining length
[ Upstream commit 9ff797e516dbc1ecb73701ec4c24055712d44411 ]
siw_get_hdr() can receive an extended DDP/RDMAP header across more than
one TCP callback. The first callback may receive most of the header,
while the next one still limits the copy to hdrlen - MIN_DDP_HDR instead
of the number of missing bytes. This makes the destination move past the
end of the header and overwrite the receive state, including
fpdu_part_rcvd. A later callback can then use a negative fpdu_part_rcvd
value as a copy offset, which creates an OOB write.
Use the number of header bytes already received when calculating the
next copy length.
Fixes: 754209850df8 ("RDMA/siw: Always consume all skbuf data in sk_data_ready() upcall.")
Signed-off-by: Jérémy Jean <Jeremy.Jean@oss.cyber.gouv.fr>
Link: https://patch.msgid.link/20260908085520.1746329-1-Jeremy.Jean@oss.cyber.gouv.fr
Assisted-by: Codex:gpt-6
Acked-by: Bernard Metzler <bernard.metzler@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Guoqing Jiang <guoqing.jiang@linux.dev>
Date: Thu Aug 27 20:55:53 2026 +0800
RDMA/siw: Clear association under lock if siw_qp_modify fails in siw_accept
[ Upstream commit 32cd87f54dd1070020e664ccb0312a9f0fea79b4 ]
We need to clear cep before release state_lock as siw_qp_llp_close and
siw_qp_modify->siw_qp_llp_close did.
Otherwise if siw_qp_modify() fails in siw_accept(), the QP's state_lock
is released before the error path cleanup. A concurrent ibv_modify_qp()
transitioning the QP to ERROR can race in this window:
siw_accept() ibv_modify_qp(ERROR)
---------------------- ----------------------
siw_qp_modify() fails
up_write(&qp->state_lock)
down_write(&qp->state_lock)
nextstate_from_idle():
if (qp->cep)
siw_cep_put(qp->cep) <- frees cep
qp->cep = NULL
goto error
cep->qp = NULL <- UAF
Clear qp->cep and drop the association reference taken by siw_cep_get(),
all under the write lock held from the initial down_write(&qp->state_lock).
Thread B therefore sees qp->cep == NULL, skips its own put, and cannot free
the cep before siw_accept() is done with it.
Fixes: 6c52fdc244b5 ("rdma/siw: connection management")
Reported-by: Shuangpeng Bai <shuangpeng.kernel@gmail.com>
Link: https://lore.kernel.org/linux-rdma/d6fbe475-a5c2-f975-99b0-a0bd6b6d10e8@linux.dev/T/#m5876c1ff2de8686a9a1173b8f1aa0ff5363a785c
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://patch.msgid.link/20260827125553.12831-1-guoqing.jiang@linux.dev
Acked-by: Bernard Metzler <bernard.metzler@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Quanye Yang <quanyeyang@proton.me>
Date: Mon Aug 31 20:30:58 2026 +0800
RDMA/ucma: Serialize join and leave on copy_to_user failure
commit 662ade4de9ff5eceb0820a9f8e9fac70ba6a815b upstream.
rdma_join_multicast() queues RoCE work that later reads the ucma_multicast
through event->param.ud.private_data, then list_add()s the CMA multicast
at the head of id_priv->mc_list. rdma_leave_multicast() matches only by
sockaddr and destroys the first hit.
ucma_process_join() used to drop ctx->mutex after a successful join and
retake it only if copy_to_user() failed. Two concurrent JOIN_MCAST calls
with the same address can therefore insert a second CMA entry before the
first thread's leave. leave then cancels the newer work and the older
worker still dereferences the ucma_multicast that the first thread frees.
Keep ctx->mutex held from rdma_join_multicast() through copy_to_user() and,
on -EFAULT, through rdma_leave_multicast() so leave cannot miss this join.
Do not leave if join itself failed: that path never published this address
on mc_list, and a leave-by-addr would destroy an earlier successful join.
Reported-by: syzbot+a6ffe86390c8a6afc818@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=a6ffe86390c8a6afc818
Fixes: fe454dc31e84 ("RDMA/ucma: Fix use-after-free bug in ucma_create_uevent")
Cc: stable@vger.kernel.org
Signed-off-by: Quanye Yang <quanyeyang@proton.me>
Link: https://patch.msgid.link/20260831-rdma-ucma-mc-uaf-v1-1-b8eeb7046aff@proton.me
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Or Har-Toov <ohartoov@nvidia.com>
Date: Tue Aug 11 19:19:16 2026 +0300
RDMA/uverbs: Fix mmap_lock/disassociation_lock circular dependency
[ Upstream commit a44a3f175eaee7e5aeb6a8fed381c4a0d5f49236 ]
Commit 51976c6cd786 ("RDMA/core: Provide rdma_user_mmap_disassociate()
to disassociate mmap pages") introduced disassociation_lock to protect
new mmap registrations against uverbs_user_mmap_disassociate(), but
created an ABBA deadlock:
Thread A (mmap / fork):
mmap_lock -> disassociation_lock
Thread B (disassociate):
disassociation_lock -> mmap_lock
Fix by removing disassociation_lock entirely and using the pre-existing
hw_destroy_rwsem instead. hw_destroy_rwsem already provides the same
protection: rdma_umap_open() and ib_uverbs_mmap() both use
down_read_trylock() before registering a new VMA, so holding hw_destroy_rwsem
in uverbs_user_mmap_disassociate() is sufficient to block new registrations.
trylock is used in both mmap paths (not blocking down_read) because
mmap_lock is already held on entry, and uverbs_user_mmap_disassociate()
acquires mmap_lock internally — a blocking read would recreate the same
deadlock.
The only caller that was not taking hw_destroy_rwsem for write was
rdma_user_mmap_disassociate(). Fix it to take the rwsem per-ufile while
iterating under lists_mutex. This is safe because ib_uverbs_close()
releases hw_destroy_rwsem entirely before acquiring lists_mutex, so the
two locks are never held simultaneously.
lockdep warning:
[ 776.654252] ======================================================
[ 776.655214] WARNING: possible circular locking dependency detected
[ 776.656167] 6.18.0for-upstream_debug_94e244d9ccab #1 Not tainted
[ 776.657114] ------------------------------------------------------
[ 776.658087] devlink/14824 is trying to acquire lock:
[ 776.658879] ffff88811170c800 (&mm->mmap_lock){++++}-{4:4}, at: uverbs_user_mmap_disassociate+0x168/0x780 [ib_uverbs]
[ 776.660479]
[ 776.660479] but task is already holding lock:
[ 776.661460] ffff888142d92b08 (&file->disassociation_lock){+.+.}-{4:4}, at: uverbs_user_mmap_disassociate+0x39/0x780 [ib_uverbs]
[ 776.663177]
[ 776.663177] which lock already depends on the new lock.
[ 776.663177]
[ 776.664525]
[ 776.664525] the existing dependency chain (in reverse order) is:
[ 776.665724]
[ 776.665724] -> #2 (&file->disassociation_lock){+.+.}-{4:4}:
[ 776.666887] __mutex_lock+0x16d/0x2330
[ 776.667633] rdma_umap_open+0x129/0x280 [ib_uverbs]
[ 776.668489] dup_mmap+0xa40/0x1790
[ 776.669170] copy_process+0x5dd2/0x6170
[ 776.669933] kernel_clone+0xb6/0x610
[ 776.670636] __do_sys_clone+0xb5/0xf0
[ 776.671354] do_syscall_64+0x70/0x12e0
[ 776.672083] entry_SYSCALL_64_after_hwframe+0x4b/0x53
[ 776.672940]
[ 776.672940] -> #1 (&mm->mmap_lock/1){+.+.}-{4:4}:
[ 776.673985] down_write_nested+0x90/0x1e0
[ 776.674751] dup_mmap+0x201/0x1790
[ 776.675448] copy_process+0x5dd2/0x6170
[ 776.676180] kernel_clone+0xb6/0x610
[ 776.676904] __do_sys_clone+0xb5/0xf0
[ 776.677615] do_syscall_64+0x70/0x12e0
[ 776.678351] entry_SYSCALL_64_after_hwframe+0x4b/0x53
[ 776.679239]
[ 776.679239] -> #0 (&mm->mmap_lock){++++}-{4:4}:
[ 776.680253] __lock_acquire+0x18c6/0x2ec0
[ 776.681018] lock_acquire+0x10e/0x2e0
[ 776.681742] down_read+0x95/0x430
[ 776.682395] uverbs_user_mmap_disassociate+0x168/0x780 [ib_uverbs]
[ 776.683436] uverbs_destroy_ufile_hw+0x1ae/0x270 [ib_uverbs]
[ 776.684416] ib_uverbs_remove_one+0x22b/0x420 [ib_uverbs]
[ 776.685371] remove_client_context+0xa6/0xf0 [ib_core]
[ 776.686342] disable_device+0x12b/0x240 [ib_core]
[ 776.687249] __ib_unregister_device+0x269/0x460 [ib_core]
[ 776.688233] ib_unregister_device+0x21/0x30 [ib_core]
[ 776.689140] mlx5r_remove+0xd0/0x170 [mlx5_ib]
[ 776.689999] device_release_driver_internal+0x3b2/0x560
[ 776.694876] bus_remove_device+0x1f5/0x3e0
[ 776.695638] device_del+0x3b9/0x990
[ 776.696329] mlx5_detach_device+0x17e/0x350 [mlx5_core]
[ 776.697429] mlx5_unload_one_devl_locked+0x3f/0xb0 [mlx5_core]
[ 776.698578] mlx5_devlink_reload_down+0x1f9/0x550 [mlx5_core]
[ 776.699712] devlink_reload+0x13e/0x680
[ 776.700456] devlink_nl_reload_doit+0xc29/0x1160
[ 776.701293] genl_family_rcv_msg_doit+0x1c9/0x2a0
[ 776.702135] genl_rcv_msg+0x3f0/0x6b0
[ 776.702854] netlink_rcv_skb+0x11d/0x370
[ 776.703605] genl_rcv+0x24/0x40
[ 776.704236] netlink_unicast+0x5b4/0x970
[ 776.704984] netlink_sendmsg+0x730/0xbf0
[ 776.705748] __sock_sendmsg+0xc5/0x190
[ 776.706461] __sys_sendto+0x201/0x2f0
[ 776.707188] __x64_sys_sendto+0xdc/0x1b0
[ 776.707931] do_syscall_64+0x70/0x12e0
[ 776.708643] entry_SYSCALL_64_after_hwframe+0x4b/0x53
[ 776.709546]
[ 776.709546] other info that might help us debug this:
[ 776.709546]
[ 776.710910] Chain exists of:
[ 776.710910] &mm->mmap_lock --> &mm->mmap_lock/1 --> &file->disassociation_lock
[ 776.710910]
[ 776.712805] Possible unsafe locking scenario:
[ 776.712805]
[ 776.713828] CPU0 CPU1
[ 776.714589] ---- ----
[ 776.715347] lock(&file->disassociation_lock);
[ 776.716097] lock(&mm->mmap_lock/1);
[ 776.717067] lock(&file->disassociation_lock);
[ 776.718199] rlock(&mm->mmap_lock);
[ 776.718857]
[ 776.718857] *** DEADLOCK ***
Fixes: 51976c6cd786 ("RDMA/core: Provide rdma_user_mmap_disassociate() to disassociate mmap pages")
Signed-off-by: Or Har-Toov <ohartoov@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Edward Srouji <edwards@nvidia.com>
Link: https://patch.msgid.link/20260811-fix-mmap-lockdep-v1-1-1151b41063b4@nvidia.com
Acked-by: Junxian Huang <huangjunxian6@hisilicon.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Li RongQing <lirongqing@baidu.com>
Date: Wed Aug 26 15:31:46 2026 +0800
RDMA/uverbs: Fix potential leak of resources->collection in flow_resources_alloc()
[ Upstream commit 08d4d9802d58bf032099091e6acf719f3298f28e ]
The two array allocations are done unconditionally and only checked
afterwards, so if the counters allocation fails while the collection
allocation succeeds, the error path frees counters and the containing
struct but never frees resources->collection, losing the only pointer
to it.
Fixes: de7498147d00 ("RDMA/uverbs: Refactor flow_resources_alloc() function")
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Link: https://patch.msgid.link/20260826073146.2203-1-lirongqing@baidu.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Aohan Mei <henrymei@tencent.com>
Date: Fri Sep 11 15:34:32 2026 +0800
rds: ib: use rds_conn_drop() on protocol version mismatch
commit f97d8c7bab7843631206a114986c9059da03efeb upstream.
rds_ib_cm_connect_complete() runs from the RDMA-CM event handler with
conn->c_cm_lock held. When the peer negotiates a protocol version
older than RDS_PROTOCOL_COMPAT_VERSION, the handler calls
rds_conn_destroy(), which is only safe in the rmmod path: it
synchronously tears the connection down and flush_work()es the
shutdown work cp_down_w.
That shutdown work (rds_conn_shutdown()) needs cp_cm_lock, which is
the very lock the event handler still holds, so the flush never
completes: the two workers wait on each other and the RDS connection
workqueues stall for good.
All other RDMA-CM failure paths (REJECTED, CONNECT_ERROR,
DISCONNECTED) use rds_conn_drop(), which marks the connection
RDS_CONN_ERROR and schedules the shutdown work asynchronously. Use
it here as well.
Fixes: f147dd9ecabf ("RDS/IB: Disallow connections less than RDS 3.1")
Reported-by: TencentOS Corvus AI <corvus@tencent.com>
Cc: stable@vger.kernel.org
Reviewed-by: Allison Henderson <achender@kernel.org>
Signed-off-by: Aohan Mei <henrymei@tencent.com>
Link: https://patch.msgid.link/20260911073436.3542080-1-ljp1205831794@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Nemesa Garg <nemesa.garg@intel.com>
Date: Wed Sep 9 16:33:31 2026 +0530
Revert "drm/i915/display: Clear SEL_FETCH_PLANE_CTL on plane disable"
[ Upstream commit a26204be587c57bd5c54fa513be26c4fd7bf252d ]
This reverts commit 7f1172a2ac0d7e50850785e2e65789c8aac8411a.
This commit replaced the crtc_state->enable_psr2_sel_fetch guard in
icl_plane_disable_sel_fetch_arm() and i9xx_cursor_disable_sel_fetch_arm()
with HAS_PSR2_SEL_FETCH(). This is a display version check and
says nothing about the pipe, so every plane and cursor disable on a
display 12+ platform started writing SEL_FETCH_PLANE_CTL() /
SEL_FETCH_CUR_CTL(), including on pipes that do not implement them.
It shows up as an unclaimed register access on pipes driving HDMI where
selective fetch was never enabled.
The stale selective fetch enable bit that commit addressed is handled
in the next patch.
Fixes: 7f1172a2ac0d ("drm/i915/display: Clear SEL_FETCH_PLANE_CTL on plane disable")
Closes: https://gitlab.freedesktop.org/drm/i915/kernel/-/work_items/16876
Signed-off-by: Nemesa Garg <nemesa.garg@intel.com>
Reviewed-by: Jouni Högander <jouni.hogander@intel.com>
Signed-off-by: Suraj Kandpal <suraj.kandpal@intel.com>
Link: https://patch.msgid.link/20260909110332.3528029-2-nemesa.garg@intel.com
(cherry picked from commit d393529394167e0f5f706657eebe84d8529ce4fc)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lyude Paul <lyude@redhat.com>
Date: Fri Aug 14 15:43:48 2026 -0400
Revert "nouveau/gsp: fix suspend/resume regression on r570 firmware"
commit a5c41fa7f925fda2db394329fa0b26243fa63a81 upstream.
This reverts commit 8302d0afeaec0bc57d951dd085e0cffe997d4d18.
It turns out this looked like the right fix on some systems, but it's not -
as this causes runtime PM to actually fail on many a laptop.
Fixes: 8302d0afeaec ("nouveau/gsp: fix suspend/resume regression on r570 firmware")
Cc: <stable@vger.kernel.org> # v6.19+
Signed-off-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patch.msgid.link/20260814194542.781955-2-lyude@redhat.com
(cherry picked from commit 94097122bfd701976bc1a62ccd434c13f3f67cde)
Signed-off-by: Lyude Paul <lyude@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Chunfeng Song <springbreeze@stu.pku.edu.cn>
Date: Thu Sep 10 05:51:10 2026 +0000
rust: net: phy: fix off-by-one bit positions in device status accessors
commit 6fb0a9d9071f1ff0cc5cfc0782302d9c90d642cb upstream.
The hand-written bitfield offsets in is_link_up(), is_autoneg_enabled()
and is_autoneg_completed() were correct when the abstraction was
merged: at that time autoneg, link, and autoneg_complete were at bits
13, 14, and 15 of struct phy_device's first bitfield unit. Commit
2796ff1e3dca ("net: phy: add flag is_genphy_driven to struct phy_device")
later inserted is_genphy_driven just before autoneg, shifting the three
fields up by one, so the accessors now read:
is_link_up() reads bit 14 = autoneg
is_autoneg_enabled() reads bit 13 = is_genphy_driven
is_autoneg_completed() reads bit 15 = link
The official ax88796b Rust driver uses all three accessors in its
read_status() implementation, so it inherits the bug.
phy_attach_direct() sets is_genphy_driven only when it falls back to
the generic driver, and ax88796b has a real driver, so
is_genphy_driven stays 0. The broken is_autoneg_enabled() therefore
reads bit 13 as 0, compares it against AUTONEG_ENABLE (1), and always
returns false, so read_status() never reaches the
resolve_aneg_linkmode() call.
The ordinary bindgen accessors take &self. Calling them through
(*phydev).link() would create a shared reference to the complete
bindings::phy_device, which is not appropriate for an object wrapped in
Opaque.
Use the bindgen-generated raw accessors (link_raw(), autoneg_raw(),
and autoneg_complete_raw()) instead. They retain the bit positions and
endianness handling generated from the C layout without creating a Rust
reference to the complete phy_device. Drop the hand-written numbers
together with the TODO comment that marked them as a stopgap.
The raw accessors are only emitted by bindgen 0.71 and later, and were
added at the Rust-for-Linux project's request, so this fix can only be
backported to stable branches whose minimum bindgen version is at least
that, hence the scope on the Cc: stable line below.
Found by a static equivalence audit (C2RustDrv, a C-to-Rust driver
migration tool) that compares hand-written bitfield offsets against
the bindgen layout of struct phy_device. Verified by building the
bindings and checking the generated accessors; no runtime testing was
possible without PHY hardware.
Fixes: 2796ff1e3dca ("net: phy: add flag is_genphy_driven to struct phy_device")
Cc: stable@vger.kernel.org # Only 7.1.y and later (requires bindgen's raw pointer accessors).
Link: https://github.com/rust-lang/rust-bindgen/issues/2674
Signed-off-by: Chunfeng Song <springbreeze@stu.pku.edu.cn>
Reviewed-by: FUJITA Tomonori <fujita.tomonori@gmail.com>
Link: https://patch.msgid.link/20260910055110.167110-1-springbreeze@stu.pku.edu.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Andrea Righi <arighi@nvidia.com>
Date: Tue Sep 15 20:41:01 2026 +0200
sched/core: Avoid false migration warning for proxy donors
[ Upstream commit fe3c73d7bc769e7afc252f867a3421fe168b898d ]
Proxy execution can move a blocked donor's scheduling context to the
lock owner's CPU even when the donor is migration-disabled. The donor
does not execute there, and its original execution CPU remains recorded
in wake_cpu.
set_task_cpu() warns unconditionally for migration-disabled tasks, so a
subsequent proxy migration or the wakeup path returning the donor home
triggers a false positive: moving a blocked scheduling context does not
violate the migration-disabled execution context.
For example, creating a mutex owner on CPU1 and a migration-disabled
waiter on CPU0 can trigger the following warning:
proxy_migrate_repro: donor blocking on CPU0 with migration disabled
proxy_migrate_repro: donor moved from CPU0 to CPU1
WARNING: kernel/sched/core.c:3389 at set_task_cpu+0x1d3/0x280
...
Call Trace:
try_to_wake_up+0x43f/0x780
__mutex_unlock_slowpath+0x330/0x540
owner_fn+0x9f/0xc0 [proxy_migrate_repro]
...
proxy_migrate_repro: donor woke on CPU0, task_cpu=0
proxy_migrate_repro: completed
Exclude blocked proxy donors from the warning. The proxy wakeup path
restores an executable placement before clearing the blocked state.
Fixes: b049b81bdff6 ("sched: Handle blocked-waiter migration (and return migration)")
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: John Stultz <jstultz@google.com>
Link: https://patch.msgid.link/20260915184101.2621252-1-arighi@nvidia.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: fangqiurong <fangqiurong@kylinos.cn>
Date: Sat Sep 12 21:15:18 2026 +0800
sched_ext: Close the pre-enable ops error claim window
commit c7a1c6e8004ab12a9c9bfdcb603f60f9bf4a3cee upstream.
scx_alloc_and_add_sched() publishes ops->priv before
scx_root_enable_workfn() switches the state to SCX_ENABLING. An error
claimed via scx_bpf_error_bstr() from an associated BPF program in that
window is consumed by scx_disable_workfn(), which takes the pre-enable
shortcut in scx_root_disable(). The shortcut returns without any teardown
and restores SCX_DISABLED with an unconditional scx_set_enable_state() xchg
racing the enable workfn's own transition. The enable then completes with
the claim consumed: the scheduler stays up but can never be disabled again,
and bpf_scx_unreg() frees it while still in use, resulting in a
use-after-free. Both WARN_ON_ONCE()s fire back to back:
WARNING: kernel/sched/ext/ext.c:7522 at
scx_root_enable_workfn+0xeec/0x1be0, CPU#3: scx_enable_help/276
WARNING: kernel/sched/ext/ext.c:6398 at scx_root_disable+0xb50/0xdb8,
CPU#0: sched_ext_helpe/664
scx_root_enable_workfn() switches to SCX_ENABLING before the scheduler
allocation, so ops->priv is never visible while SCX_DISABLED. The allocation
failure path restores SCX_DISABLED.
Fixes: 105dcd005be2 ("sched_ext: Introduce scx_prog_sched()")
Cc: stable@vger.kernel.org
Signed-off-by: fangqiurong <fangqiurong@kylinos.cn>
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Wanwu Li <liwanwu@kylinos.cn>
Date: Thu Sep 3 14:06:26 2026 +0800
sched_ext: Fix NULL sched deref in kfunc sub-sched error paths
commit 0a85182723b65ad8bee8131bc38fcf0347d6679b upstream.
When the root scheduler has sub-scheds attached, the COMPAT kfunc
wrappers scx_bpf_select_cpu_and() and scx_bpf_dsq_insert_vtime() refuse
the call and report to @p's scheduler:
scx_error(scx_task_sched(p), "... must be used");
The wrappers are reachable with tasks that have no scheduler.
scx_bpf_select_cpu_and() is in the select_cpu kfunc group, which
scx_kfunc_context_filter() opens to BPF_PROG_TYPE_SYSCALL programs;
scx_bpf_dsq_insert_vtime() is in the enqueue_dispatch group, which
ops.enqueue() and ops.dispatch() may call with any KF_RCU task -- the
group has no kf_tasks validation, and scx_dsq_insert_preamble() checks
task ownership with scx_task_on_sched() precisely because @p may be an
arbitrary task.
scx_task_sched(p) is p->scx.sched, which is NULL for tasks past
sched_ext_dead() -- which clears it via scx_disable_and_exit_task() on
exit -- and for idle tasks, which the enable paths skip as they are
never scheduled through SCX. It is also an rcu_dereference_protected()
that expects @p's pi_lock or rq lock, which neither wrapper holds.
Passing NULL to scx_error() reaches scx_vexit(), which dereferences
sch->exit_info, oopsing the kernel.
One concrete trigger exercised while developing the fix: a
BPF_PROG_TYPE_SYSCALL program calling the select_cpu_and wrapper on an
exited-but-not-reaped task while a sub-scheduler was attached (its pid
stays findable while the zombie is unreaped; faulting instruction is
the scx_vexit() prologue "mov r15,[rdi+0x398]" with RDI=NULL and 0x398
the offset of sch->exit_info):
sched_ext: BPF scheduler "kfunc_subsched_null" enabled
sched_ext: BPF sub-scheduler "kfunc_subsched_null" enabled
sched_ext: Unassociated program run_select_cpu_ (id 76)
BUG: kernel NULL pointer dereference, address: 0000000000000398
#PF: supervisor read access in kernel mode
#PF: error_code(0x0000) - not-present page
Oops: Oops: 0000 [#1] SMP NOPTI
CPU: 7 UID: 0 PID: 8201 Comm: kfunc_test_runn Tainted: G W
RIP: 0010:scx_vexit+0x25/0xa0
Code: ... <4c> 8b bf 98 03 00 00 ...
CR2: 0000000000000398
Call Trace:
<TASK>
__scx_exit+0x4f/0x70
scx_bpf_select_cpu_and+0xab/0xb0
bpf_prog_430ed61a7b66e03a_run_select_cpu_and+0x9c/0xe7
? __x64_sys_bpf+0x2c/0x40
bpf_prog_test_run_syscall+0x130/0x2f0
__sys_bpf+0x930/0x10d0
? __x64_sys_bpf+0x2c/0x40
__x64_sys_bpf+0x2c/0x40
do_syscall_64+0xbc/0x460
entry_SYSCALL_64_after_hwframe+0x76/0x7e
</TASK>
Read @p's scheduler under RCU instead, which the wrappers can do from
their guard(rcu)(): fault it when it can be determined, and when it
can't be determined -- @p is a task past sched_ext_dead() or an idle
task -- there is nothing obviously wrong to report, so just refuse the
call as before without faulting any scheduler.
These COMPAT wrappers are scheduled for eventual removal once the
deprecation grace period elapses, but until then -- and regardless of
their removal timeline -- they must not oops the kernel on a task they
are handed.
Cc: stable@vger.kernel.org # v7.1+
Fixes: a5fa0708cbfd ("sched_ext: Enforce scheduling authority in dispatch and select_cpu operations")
Suggested-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Wanwu Li <liwanwu@kylinos.cn>
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Alberto Carboneri <acarboneri@drivesec.com>
Date: Fri Sep 4 13:54:37 2026 +0000
scsi: core: Validate MODE SENSE lengths in scsi_cdl_enable()
commit 3d676e458fe0c566f5a62753dc696b6a862fc412 upstream.
scsi_cdl_enable() uses length fields returned by MODE SENSE to locate
the ATA feature mode page in a 64-byte stack buffer. A target can report
a total length shorter than its mode header and block descriptors. The
unsigned subtraction used for the MODE SELECT length can wrap, and the
separately computed buf_data can point beyond buf.
During automatic scan, enable is false, so the read-modify-write of
buf_data[4] can clear the low two bits of a target-selected
out-of-bounds stack byte. scsi_mode_select() can then copy up to 64
bytes from outside the buffer into the outgoing MODE SELECT payload,
disclosing stack contents to the target.
This is reachable while scanning a USB storage device that identifies as
an ATA device and advertises CDL support. No filesystem mount or
userspace access to the block device is required.
On upstream commit cee9395acd80 ("Linux 7.3-rc1"), a build-specific,
one-vCPU QEMU/Raw Gadget proof using QEMU-only multi-UDC allocator
sampling executed a fixed proof command inside the guest and created a
UID-0-owned marker during automatic enumeration, with KASLR and NX
enabled.
The issue was independently found during security research at Drivesec
S.r.l.
Cap the available length to the buffer size. Validate and consume the
mode header and block descriptor lengths before using the page, and
require the five bytes needed to access the CDL field.
Fixes: 1b22cfb14142 ("scsi: core: Allow enabling and disabling command duration limits")
Reported-by: Sashiko AI Review <sashiko-bot@kernel.org>
Closes: https://lore.kernel.org/linux-scsi/20260717192313.93D791F000E9@smtp.kernel.org/
Link: https://lore.kernel.org/linux-scsi/20260717222931.AC4EE1F000E9@smtp.kernel.org/
Link: https://lore.kernel.org/linux-scsi/df13ec87ac9b28e3b0a2d9eb26477e276ff0278a.camel@HansenPartnership.com/
Cc: stable@vger.kernel.org
Assisted-by: LLM
Co-developed-by: Pimen Flavian Dei (Drivesec S.r.l.) <fdei@drivesec.com>
Signed-off-by: Pimen Flavian Dei (Drivesec S.r.l.) <fdei@drivesec.com>
Signed-off-by: Alberto Carboneri (Drivesec S.r.l.) <acarboneri@drivesec.com>
Link: https://lore.kernel.org/linux-scsi/20260717192313.93D791F000E9@smtp.kernel.org/
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://patch.msgid.link/20260904135410.360314-1-acarboneri@drivesec.com
Signed-off-by: Martin K. Petersen (Oracle) <mkp@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Karan Tilak Kumar <kartilak@cisco.com>
Date: Tue Sep 22 11:23:11 2026 -0400
scsi: fnic: Bump up version number
[ Upstream commit 2265541d221dc550dc89b52e49e9d9eb2f824996 ]
Bump up version number to 1.9.0.0.
Reviewed-by: Sesidhar Baddela <sebaddel@cisco.com>
Reviewed-by: Arulprabhu Ponnusamy <arulponn@cisco.com>
Reviewed-by: Gian Carlo Boffa <gcboffa@cisco.com>
Reviewed-by: Arun Easi <aeasi@cisco.com>
Reviewed-by: Hannes Reinecke <hare@kernel.org>
Reviewed-by: Lee Duncan <lduncan@suse.com>
Tested-by: Karan Tilak Kumar <kartilak@cisco.com>
Signed-off-by: Karan Tilak Kumar <kartilak@cisco.com>
Co-developed-by: Hannes Reinecke <hare@kernel.org>
Link: https://patch.msgid.link/20260724174811.5118-14-kartilak@cisco.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Stable-dep-of: 0cb1fd924126 ("scsi: fnic: Fix missed link-up when critical IRQ targets offline CPU")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Arun Easi <aeasi@cisco.com>
Date: Tue Sep 22 11:23:12 2026 -0400
scsi: fnic: Fix missed link-up when critical IRQ targets offline CPU
[ Upstream commit 0cb1fd924126f1f581621a5e804df98a02be9dff ]
When CPU Hyper Threading is disabled, sibling CPUs remain present but
are reported offline. Managed MSI-X IRQs can still receive affinity
masks that include those offline CPUs. If a driver-critical vector is
managed, it can be parked on an offline CPU and the driver may miss
critical events such as link-up.
Keep driver-critical vectors unmanaged so they can be migrated by the
IRQ core when their target CPU is offlined.
Since HWQ-0 is unmanaged now, in some queue combinations there can be no
mappings to it in mq_map. So without the blk-mq fix mentioned below,
system may crash during cpu offline/online tests.
Fixes: 8a8449ca5e33 ("scsi: fnic: Modify ISRs to support multiqueue (MQ)")
Cc: stable@vger.kernel.org
Depends-on: commit 10845a105bbc ("blk-mq: skip CPU offline notify on unmapped hctx")
Reviewed-by: Sesidhar Baddela <sebaddel@cisco.com>
Reviewed-by: Arulprabhu Ponnusamy <arulponn@cisco.com>
Reviewed-by: Gian Carlo Boffa <gcboffa@cisco.com>
Reviewed-by: Karan Tilak Kumar <kartilak@cisco.com>
Signed-off-by: Arun Easi <aeasi@cisco.com>
Reviewed-by: Laurence Oberman <loberman@redhat.com>
Link: https://patch.msgid.link/20260903175547.57971-1-aeasi@cisco.com
Signed-off-by: Martin K. Petersen (Oracle) <mkp@kernel.org>
[ Inlined fnic_mq_init_queue_map() into fnic_mq_map_queues_cpus() using the existing cpu variable. ]
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Karan Tilak Kumar <kartilak@cisco.com>
Date: Tue Sep 22 11:23:10 2026 -0400
scsi: fnic: Make debug logging protocol independent
[ Upstream commit b93c38a9f2ce5441c90de55776f8df97679cf8a2 ]
Make the fnic debug macros take struct fnic instead of struct Scsi_Host so
FCP and NVMe initiator roles can share the same logging interface.
Add fnic_printk() to route FCP initiator messages through shost_printk()
and non-SCSI role messages through printk(). Add role and non-SCSI role
messages through printk(). Add role predicates and separate FDLS, FIP, and
NVMe logging masks.
Convert FCS, FIP, SCSI, ISR, and main debug call sites to pass the fnic
instance directly, and keep FIP VLAN MAC descriptors skipped while
reporting unexpected descriptor types.
Reviewed-by: Sesidhar Baddela <sebaddel@cisco.com>
Reviewed-by: Arulprabhu Ponnusamy <arulponn@cisco.com>
Reviewed-by: Gian Carlo Boffa <gcboffa@cisco.com>
Reviewed-by: Arun Easi <aeasi@cisco.com>
Reviewed-by: Hannes Reinecke <hare@kernel.org>
Reviewed-by: Lee Duncan <lduncan@suse.com>
Signed-off-by: Karan Tilak Kumar <kartilak@cisco.com>
Co-developed-by: Hannes Reinecke <hare@kernel.org>
Link: https://patch.msgid.link/20260724174811.5118-2-kartilak@cisco.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Stable-dep-of: 0cb1fd924126 ("scsi: fnic: Fix missed link-up when critical IRQ targets offline CPU")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Karl Mehltretter <kmehltretter@gmail.com>
Date: Sun Sep 6 19:09:25 2026 +0200
scsi: pm80xx: Fix the use_msix, use_tasklet and read_wwn parameter descriptions
[ Upstream commit 264bf9655c3d067d775a46f05eb8c871c488a864 ]
The MODULE_PARM_DESC() lines of use_msix, use_tasklet and read_wwn all
name a parameter zoned, which does not exist, and the use_tasklet one
repeats the use_msix text. modinfo shows three "zoned" entries and no
description for the real parameters.
Name the right parameters and describe use_tasklet.
Fixes: efa1fca45082 ("scsi: pm8001: Remove PM8001_USE_MSIX")
Fixes: 205430290ad0 ("scsi: pm8001: Remove PM8001_USE_TASKLET")
Fixes: 80975adc79dd ("scsi: pm8001: Remove PM8001_READ_VPD")
Assisted-by: LLM
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://patch.msgid.link/20260906170925.2524-1-kmehltretter@gmail.com
Signed-off-by: Martin K. Petersen (Oracle) <mkp@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Karl Mehltretter <kmehltretter@gmail.com>
Date: Sun Sep 6 19:10:09 2026 +0200
scsi: qla2xxx: Fix the ql2xfc2target parameter description
[ Upstream commit 779f202a92ef10a426efc07d0f4267918cb07ca3 ]
The module parameter is ql2xfc2target, but its MODULE_PARM_DESC() names
qla2xfc2target, so modinfo describes a parameter that does not exist and
shows no description for the real one.
Use the parameter name in the description.
Fixes: 877b03795fcf ("scsi: qla2xxx: Add option to disable FC2 Target support")
Assisted-by: LLM
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Link: https://patch.msgid.link/20260906171009.2560-1-kmehltretter@gmail.com
Signed-off-by: Martin K. Petersen (Oracle) <mkp@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Andrea Mayer <andrea.mayer@uniroma2.it>
Date: Sun Sep 13 21:44:21 2026 +0200
seg6: set IPSKB_L3SLAVE from IP6SKB_L3SLAVE on IPIP decapsulation
[ Upstream commit 7616242a2b37883f7322aaa1d2bd6cd0fed28315 ]
When an SRv6 packet arrives on an interface enslaved to a VRF,
vrf_ip6_rcv() sets IP6SKB_L3SLAVE in IP6CB, but decap_and_validate()
has never set IPSKB_L3SLAVE in IPCB. The bit stayed clear in the
common case, and with CONFIG_IPV6_MIP6 the leftover frag_max_size of
a reassembled outer packet could even set it, with no VRF involved.
Commit 44930446dde4 ("ipv6: seg6: clear IPv4 control block on IPIP
decapsulation") then made the unreliable bit reliably clear.
The effect of the missing flag is visible with End.DX4 when a
delivery to a local address of the node reaches the socket lookup.
For example, a UDP socket bound to the enslaved ingress interface
does not receive any of the decapsulated packets, while an unbound
socket outside the VRF does.
This contradicts Documentation/networking/vrf.rst: by default the
scope of an unbound UDP or TCP socket is limited to the default VRF.
Set IPSKB_L3SLAVE for IPv4 in decap_and_validate(), which already does
the same for IPv6. The socket lookup then matches the decapsulated
packet like any other packet received on that enslaved interface. Such
a packet matches an unbound UDP or TCP socket only when
udp_l3mdev_accept or tcp_l3mdev_accept is set.
Fixes: 891ef8dd2a8d ("ipv6: sr: implement additional seg6local actions")
Signed-off-by: Andrea Mayer <andrea.mayer@uniroma2.it>
Reviewed-by: David Ahern <dsahern@kernel.org>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Link: https://patch.msgid.link/20260913194421.31-1-andrea.mayer@uniroma2.it
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Günther Noack <gnoack@google.com>
Date: Thu Sep 17 17:42:29 2026 +0200
selftests/landlock: Add audit test for whiteout object creation
[ Upstream commit 8c46c6acbebe0d8544fd1b55e5ddf36828d7b9ea ]
Add audit_layout1.make_whiteout: This test looks similar to
audit_layout1.make_char, but creates a whiteout object through mknod().
Since whiteout object creation is now guarded with
LANDLOCK_ACCESS_FS_MAKE_REG rather than LANDLOCK_ACCESS_FS_MAKE_CHAR, it
also needs to log the matching denial to audit.
Signed-off-by: Günther Noack <gnoack@google.com>
Link: https://patch.msgid.link/20260813093157.1436894-5-gnoack@google.com
Signed-off-by: Mickaël Salaün <mic@digikod.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Günther Noack <gnoack@google.com>
Date: Thu Sep 17 17:42:28 2026 +0200
selftests/landlock: Add tests for whiteout object creation
[ Upstream commit ee890889b30b22f9a21636061def7a04e4f89380 ]
Add tests to check that whiteout object creation is guarded by
LANDLOCK_ACCESS_FS_MAKE_REG, in the cases where these are created from
userspace:
* Conventional creation with mknod()
* Linking or renaming an existing whiteout object
* renameat2() with RENAME_WHITEOUT,
which creates a new whiteout object in the source location
* renameat2() with RENAME_EXCHANGE,
with one of the renamed objects being a whiteout object
Signed-off-by: Günther Noack <gnoack@google.com>
Link: https://patch.msgid.link/20260813093157.1436894-4-gnoack@google.com
[mic: Update commit message as requested]
Signed-off-by: Mickaël Salaün <mic@digikod.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Günther Noack <gnoack@google.com>
Date: Thu Sep 17 17:42:27 2026 +0200
selftests/landlock: Use an actual chardev for MAKE_CHAR audit test
[ Upstream commit 173b1bd8730825e1f6862dbd07856e7d68447a41 ]
By passing a (0, 0) device number, the audit test for
LANDLOCK_ACCESS_FS_MAKE_CHAR was accidentally creating a whiteout object
rather than a char device. In preparation to treating whiteout objects
differently, use an actual character device instead.
Signed-off-by: Günther Noack <gnoack@google.com>
Link: https://patch.msgid.link/20260813093157.1436894-2-gnoack@google.com
Signed-off-by: Mickaël Salaün <mic@digikod.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Christian Göttsche <cgzones@googlemail.com>
Date: Thu Sep 3 13:43:38 2026 +0200
selinux: always fill AVC decision in avc_has_perm_noaudit()
commit 8861db305103107199b1426f25fde1fb6d465583 upstream.
avc_has_perm_noaudit() is documented to return a copy of the access
decision in @avd, but its early return for an empty requested permission
set leaves the buffer untouched. All callers pass an uninitialized
stack variable and afterwards feed it to avc_audit(), and the inode hook
even stores it in the per-task decision cache.
Fill in a deny-all, audit-all decision, similar to avd_init(), so every
caller receives a defined value at no cost on the hot path.
Cc: stable@vger.kernel.org
Fixes: e6f2f381e4015386 ("selinux: replace BUG_ONs with WARN_ONs in avc.c")
Signed-off-by: Christian Göttsche <cgzones@googlemail.com>
Reviewed-by: Stephen Smalley <stephen.smalley.work@gmail.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Karl Mehltretter <kmehltretter@gmail.com>
Date: Sat Aug 29 23:32:55 2026 +0200
selinux: preserve user SID across nested backing files
commit 8c0c602202b9a4909b00bc3354e3c0355bc69e65 upstream.
SELinux saves the user file SID in a backing-file security blob so it
remains available after mmap() replaces vma->vm_file with a backing file.
For nested backing files (overlayfs over overlayfs, or FUSE passthrough
backed by overlayfs), user_file may itself be a backing file. Its
fsec->sid is the SID of the mounter that opened it, rather than the user
that opened the top-level file. mprotect() then checks fd { use } against
the mounter SID. This can incorrectly deny access without a domain
transition, or check the wrong target SID after one.
Copy the saved user SID when user_file is a backing file. Keep using the
regular file SID for the first backing layer.
With two nested overlayfs mounts and SELinux enforcing,
mprotect(PROT_READ) returns EACCES with an fd { use } denial against the
mounter SID. With this change, mprotect() succeeds.
Tested on arm64 QEMU with a small BusyBox initramfs and a purpose-built
SELinux policy. The original test was also repeated with Fedora Cloud
Base 44 userspace and gave the same result.
Cc: stable@vger.kernel.org
Fixes: 82544d36b172 ("selinux: fix overlayfs mmap() and mprotect() access checks")
Assisted-by: LLM
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Stephen Smalley <stephen.smalley.work@gmail.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Karl Mehltretter <kmehltretter@gmail.com>
Date: Sat Aug 29 23:32:56 2026 +0200
selinux: recheck intermediate backing files on mprotect()
commit 78fc54b934bfb2c18aad8154c7302067146946f9 upstream.
mprotect() can be used to bypass the SELinux checks that mmap() performs
against the intermediate layers of a stacked filesystem.
mmap() checks every backing layer as the request descends through the
stack. mprotect() only has the lowest backing file in vma->vm_file, so it
rechecks the top-level user and the lowest mounter, but skips the mounters
of every layer in between. With two nested overlayfs mounts and a policy
denying mounter_t -> middle_file_t:file { execute }, a direct
mmap(PROT_EXEC) is denied:
avc: denied { execute } for pid=71 comm="nested_exec"
path="/payload" dev="overlay" ino=9
scontext=user_u:base_r:mounter_t
tcontext=user_u:object_r:middle_file_t tclass=file permissive=0
while mmap(PROT_NONE) followed by mprotect(PROT_EXEC) succeeds.
Preserve each intermediate path, mounter SID and file-description SID in
the backing-file security blob, copying the saved entries when another
backing layer is opened. Allocate the array only for nested backing files,
and release it and the path references in the backing_file_free hook.
During mprotect(), recheck fd { use } and the requested inode permissions
for every saved mounter, and include the intermediate layers in the execmod
checks. Policy for nested stacking may then need to grant intermediate
mounters what a direct mmap() already requires, and execmod on intermediate
labels for binaries using text relocations.
Tested on arm64 QEMU with a small BusyBox initramfs and a purpose-built
SELinux policy, on a mainline tree containing
commit f2381b546e7e ("fs: fix user path of nested backing files").
Cc: stable@vger.kernel.org
Fixes: 82544d36b172 ("selinux: fix overlayfs mmap() and mprotect() access checks")
Assisted-by: LLM
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Reviewed-by: Stephen Smalley <stephen.smalley.work@gmail.com>
[PM: subject tweak]
Signed-off-by: Paul Moore <paul@paul-moore.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Thomas Gleixner <tglx@kernel.org>
Date: Fri Sep 11 11:09:12 2026 +0200
signal: Prevent exec() race
commit d2710c8d938ae6a825a6463158e6e6f31eac792a upstream.
Hyunwoo debugged the following KASAN UAF splat:
BUG: KASAN: slab-use-after-free in __send_signal_locked+0xb27/0xba0
Write of size 8 at addr ffff888007ed80c8 by task poc/79
...
Call Trace:
__send_signal_locked+0xb27/0xba0
do_send_sig_info+0xa7/0x160
do_send_specific+0x76/0xa0
__x64_sys_tgkill+0x193/0x270
...
Allocated by task 80:
do_timer_create+0x1a4/0x1030
__x64_sys_timer_create+0x145/0x190
...
Freed by task 12:
kmem_cache_free_bulk+0x1f8/0x4a0
kvfree_rcu_bulk+0x14f/0x1c0
kfree_rcu_work+0x128/0x1a0
...
Last potentially related work creation:
kvfree_call_rcu+0x39/0x390
__flush_itimer_signals+0x211/0x320
flush_itimer_signals+0x47/0x90
begin_new_exec+0xa6b/0x28c0
It turned out that this happens with a non-leader exec() as Hyunwoo
explained:
de_thread() calls exchange_tids() before release_task(leader), so the
struct pid held by a SIGEV_THREAD_ID timer created against the leader's tid
now points to the thread which called execve(). pid_task() returns that
thread and lock_task_sighand() on it succeeds.
If the timer signal is blocked, its sigqueue stays queued on the leader's
task::pending. The next expiry of that timer can then run while
release_task() flushes the queue.
posixtimer_send_sigqueue() checks whether the sigqueue is already queued
with a plain list_empty(), which only reads list_head::next.
list_del_init() is not atomic and INIT_LIST_HEAD() stores list_head::next
before list_head::prev, so the check can pass in between. list_add_tail()
queues the entry on the task::pending of the live thread, and the
list_head::prev store from the flush then overwrites the list_head::prev
link that list_add_tail() has just set.
__flush_itimer_signals() does not undo that either. With list_head::prev
pointing at the entry itself, its list_del_init() only stores the same
values again, so the entry is not removed from the list. It is still there
after the last reference is dropped and the timer is freed by RCU, and the
list_add_tail() of a later tgkill() follows that list_head::prev into the
freed timer.
This problem surfaced with the recent commit which moved the sigqueue flush
out of the sighand lock held region.
Hyonwoo proposed to fix this by using list_del_init_careful(), but that
just papers over the problem. After some disucssions and various attempts
to solve it, Eric pointed out that there is no reason to flush
task::pending late in release_task() and it should be done in
exit_signals() already.
As nothing can collect and deliver signals which are queued in a dying
task's pending queue, there is no reason to delay it further.
But it has to be ensured that no signals can be queued into it after that
point. exit_signals() sets PF_EXITING in task::flags, which can be used as
an indicator for this.
Cure it by:
- Preventing signal queueing for task private signals (PIDTYPE_PID) when
the task has PF_EXITING set in __send_signal_locked() and in
posixtimer_send_sigqueue().
- Protecting the unlocked setting of PF_EXITING in exit_signals() for the
task group empty and the group exit case with sighand lock
- Flushing task::pending signals right there.
Optimize that by moving the whole pending list to an on-stack list head
under sighand lock and free the signals without the lock held.
There has been quite some discussion about the lockless flush and the
non-leader exec case on weakly ordered systems. The problem is that a third
party which tries to send a posix timer signal relies on the PID lookup to
find the target task and that lookup might result in the new leader when
the signal was originaly directed to the old leader. In case that the
signal was queued on the old leader then the lockless flush raised a
concern over the following situation:
old_leader new_leader third party
A: flush_list() // list_del_init() stores to sigqueue
LOCK (tasklist)
old_leader->exit_state = EXIT_ZOMBIE;
B: UNLOCK (tasklist)
C: LOCK (tasklist)
if (old_leader->exit_state)
transfer_tids()
D: store PID
posix_timer_send_sigqueue()
// Observes #D so t = new_leader
E: t = get_target()
F: LOCK (sighand)
G: if (list_empty(sigqueue))
list_add(sigqueue)
The concern was that the third party might observe #D but not observe #A
and therefore would proceed to #G while the list_del() stores (#A) in
flush_list() are not visible yet, which could result in list corruption.
That would be possible if looking at it solely from a RELEASE+ACQUIRE
ordering point of view, but B-C is a UNLOCK+LOCK hand-over, which is not
the same as RELEASE+ACQUIRE:
RELEASE+ACQUIRE: RCpc, only the CPUs involved agree on the ordering
UNLOCK+LOCK: RCtso, the hand-over is store-ordering
As B-C is UNLOCK+LOCK, which is RCtso and that does impose store order,
A stores must happen before the D store.
Combine with E-F, which has a data dependency from the LOAD to the LOCK and
thereby constraints later LOADs, those sigqueue loads in G that come after
F must in fact observe the A stores.
Fixes: fb3bbcfe344e ("exit: change the release_task() paths to call flush_sigqueue() lockless")
Reported-by: Hyunwoo Kim <imv4bel@gmail.com>
Debugged-by: Hyunwoo Kim <imv4bel@gmail.com>
Suggested-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Thomas Gleixner <tglx@kernel.org>
Tested-by: Kijo Park <red993688@gmail.com>
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260911090541.572536604@kernel.org
Closes: https://patch.msgid.link/aok1rdkBgZsynHZB@v4bel
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: April Cardenas <april.cardenas@canonical.com>
Date: Thu Sep 10 00:16:49 2026 -0500
smb/client: send lease break ACKs thru correct session for multiuser mounts
commit ebc5660132ddd244b57f03ed324922013a3d7363 upstream.
Currently, when cifs_oplock_break handles a break request from the server
it searches for the appropriate tlink to handle the request
but incorrectly uses the current fsuid as the search key, eventually
causing read errors for users with multiuser mounts on NetApp.
Fix this by using the tlink from the cfile struct instead to respond
through the correct session.
As breaks are handled in a worker thread, the current fsuid
isn't guaranteed to match the session that the break is intended for.
This means that cifs_sb_tlink may search the rbtree using the wrong fsuid,
and return a tlink with an incorrect session than
the lease break was intended for. As a result, the breaks
may be ACKed through an incorrect session.
While it seems that Samba/Windows Servers 2016-2025 ignore this as long as
the lease key is correct, we ran into a case where if you're using
NetApp ONTAP or Azure NetApp Files they will reject the ACK
and return `STATUS_LOCK_NOT_GRANTED` errors on any future read requests
a user may initiate through their still held open file handle,
and the server will eventually close the file.
In the dmesg logs, the user may see errors like these:
CIFS: Status code returned 0xc0000128 STATUS_FILE_CLOSED
CIFS: VFS: Send error in read = -9
With a multiuser mount using NetApp, this issue is really easy
for users to hit on a wide variety of kernel versions
by attempting to copy a file from the share
to the local machine through GNOME Files/Nautilus.
This copy will always result in Nautilus throwing
a `Bad File Descriptor` error to the user and fail.
With this fix, you can copy files through Nautilus without issue.
>From looking at the traces, it seems that glib will
open the file first, and call listxattr before actually attempting
to copy the file data. The listxattr call always triggers a break,
causing the copy to fail.
The proposed fix returns to the way the client grabbed the tlink before
commit e8f5f849ffce2 ("cifs: fix potential oops in cifs_oplock_break").
The bulk of that commit (checking for list empty) remains untouched, and
I think the change to using cifs_sb_tlink was intended to avoid a
NULL/ERR deference on the tlink as well as update the reference count.
I believe this fix should preserve those safety properties, but of course
I'd appreciate any corrections here.
Fixes: e8f5f849ffce2 ("cifs: fix potential oops in cifs_oplock_break")
Cc: stable@vger.kernel.org
Signed-off-by: April Cardenas <april.cardenas@canonical.com>
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Reviewed-by: Bharath S M <bharathsm@microsoft.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Paulo Alcantara <pc@manguebit.org>
Date: Fri Sep 11 22:38:04 2026 -0300
smb: client: cancel reconnect work in clean_demultiplex_info()
commit c65eae6f61d1778ff7a82e4aae4080e26f486af1 upstream.
clean_demultiplex_info() cancels server->echo delayed work but not
server->reconnect, which can cause a use-after-free when the
demultiplex thread exits while a reconnect work is still queued:
cifs_demultiplex_thread()
cifs_readv_from_socket()
cifs_reconnect()
__cifs_reconnect()
cifs_queue_server_reconn()
mod_delayed_work(cifsiod_wq, &server->reconnect, 0)
clean_demultiplex_info()
cancel_delayed_work_sync(&server->echo) // echo canceled
// reconnect NOT canceled
kfree_sensitive(server) // server freed
...later, on cifsiod_wq:
smb2_reconnect_server()
server->srv_count // UAF read of freed server
Fix this by canceling server->reconnect delayed work in
clean_demultiplex_info() before the server is freed, the same way
cifs_put_tcp_session() already does.
Reported-by: syzbot+5003556314abc915a71f@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/r/6aa4a12d.f81106d8.2ab401.0023.GAE@google.com
Fixes: 53e0e11efe92 ("CIFS: Fix a possible memory corruption during reconnect")
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Cc: David Howells <dhowells@redhat.com>
Cc: Shyam Prasad N <sprasad@microsoft.com>
Cc: Ronnie Sahlberg <ronniesahlberg@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Bharath SM <bharathsm@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Paulo Alcantara <pc@manguebit.org>
Date: Sun Sep 13 21:15:41 2026 -0300
smb: client: fix fattr leaking on wsl_to_fattr() failure
commit e1253a82bb4c0fed6706a5839fc8b6e01be1abe2 upstream.
wsl_to_fattr() mutates fattr fields as it parses each WSL EA. If
validation later fails, the function returns false with partially
mutated fattr fields that callers do not reset.
Fix this by parsing into local variables and only committing them to
fattr on success.
Closes: https://sashiko.dev/#/patchset/20260906200517.725015-1-pc%40manguebit.org
Fixes: 78e26bec4d6d ("smb: client: parse uid, gid, mode and dev from WSL reparse points")
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Cc: David Howells <dhowells@redhat.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Shyam Prasad N <sprasad@microsoft.com>
Cc: Ronnie Sahlberg <ronniesahlberg@gmail.com>
Cc: Bharath SM <bharathsm@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Frank Sorenson <sorenson@redhat.com>
Date: Wed Sep 16 16:33:58 2026 -0500
smb: client: fix missing iov bounds check in parse_posix_sids()
commit b09d092eb24ad0110f16a9b7c1ed5d2a0c1733dc upstream.
In parse_posix_sids(), sidsbuf_end is calculated using the server-supplied
out_len without being validated against the actual length of the received
iov (iov_len).
If a server provides an inflated out_len, sidsbuf_end will point past the
end of the iov. This defeats the bounds guards in posix_info_sid_size(),
allowing out-of-bounds reads into adjacent kernel memory.
Fix this by rejecting responses where the calculated sidsbuf_end would
exceed the received iov boundaries or cause pointer wraparound.
Fixes: a90f37e3d7ac ("smb: client: parse owner/group when creating reparse points")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Frank Sorenson <sorenson@redhat.com>
Date: Wed Sep 16 16:33:55 2026 -0500
smb: client: fix missing lower-bound check on DFS referral string offsets
commit e83330c55edc0c3ac08aa6c95e49e4694c65523b upstream.
parse_dfs_referrals() checks that DfsPathOffset and NetworkAddressOffset
do not exceed the buffer end, but fails to check that they don't point
inside the referral header itself.
If a server provides an offset smaller than
sizeof(struct dfs_referral_level_3), the derived string pointer overlaps
with the struct fields, causing cifs_strndup_from_utf16() to interpret
header data as UTF-16 strings.
Fix this by enforcing that string offsets are at least sizeof(*ref).
Fixes: 4ecce920e13a ("CIFS: move DFS response parsing out of SMB1 code")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Frank Sorenson <sorenson@redhat.com>
Date: Wed Sep 16 16:33:52 2026 -0500
smb: client: fix next_buffer UAF and NextCommand bounds in compound PDUs
commit 05762c5bc1cfdcac36747994fde2c04387a457f1 upstream.
Fix several related bounds checking and pointer lifecycle issues in
receive_encrypted_standard()'s handling of compound encrypted frames:
- Clear next_buffer after assigning it to server->bigbuf. A stale
next_buffer pointer can lead to a use-after-free on subsequent
error paths.
- Update pdu_length to the decrypted plaintext size (buf_size). Using
the pre-decryption length allows NextCommand to point into stale
ciphertext residue.
- Reject next_cmd values smaller than MID_HEADER_SIZE(server).
- Fix an integer overflow in the upper bound check by verifying
pdu_length - next_cmd < MID_HEADER_SIZE(server), ensuring the
trailing slice is large enough for a header.
Fixes: b24df3e30cbf ("cifs: update receive_encrypted_standard to handle compounded responses")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Frank Sorenson <sorenson@redhat.com>
Date: Wed Sep 16 16:33:57 2026 -0500
smb: client: fix OOB struct field reads in move_smb2_ea_to_cifs()
commit eeb5ef6083e1cefa2ef75041b5597ff228b8d7bb upstream.
In move_smb2_ea_to_cifs(), the while (src_size > 0) loop condition is
insufficient. It allows iteration to continue even if the remaining
src_size is too small to contain a complete smb2_ea_info structure.
Consequently, reads of ea_name_length and ea_value_length can occur
out-of-bounds.
Fix this by ensuring src_size >= sizeof(*src) before attempting to read
any structure fields. Additionally, reject any next_entry_offset that is
smaller than sizeof(*src) or that would advance the pointer beyond the
available buffer.
Note that for calls where the server returns a malformed EA list, the
error returned to userspace changes from -ENODATA (getxattr) or
-ERANGE (listxattr) to -EIO. This correctly signals a server protocol
error rather than misleadingly indicating "attribute not present" or
"output buffer too small".
Fixes: 95907fea4fd8 ("cifs: Add support for reading attributes on SMB2+")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Frank Sorenson <sorenson@redhat.com>
Date: Wed Sep 16 16:33:59 2026 -0500
smb: client: fix potential OOB read in smb3_enum_snapshots()
commit 4775c3b7a597907e0b97556c7986fda238a377ae upstream.
If snapshot_array_size is smaller than GMT_TOKEN_SIZE,
smb3_enum_snapshots() sets ret_data_len to
sizeof(struct smb_snapshot_array) without verifying the actual length
of the server's reply.
Because SMB2_ioctl() places no lower bound on the server-supplied
OutputCount and allocates retbuf to exactly that length, a short reply
results in ret_data_len exceeding the size of retbuf. The subsequent
copy_to_user() then reads past the end of retbuf, leaking adjacent slab
memory to userspace. The subsequent clamp check is ineffective as it
only reduces ret_data_len.
Fix this by rejecting replies shorter than
sizeof(struct smb_snapshot_array) with -EIO. Note that the bound is set
to the 12-byte struct size rather than the 16-byte
MIN_SNAPSHOT_ARRAY_SIZE defined in MS-SMB2 3.3.5.15.1, because 12 bytes
is exactly what copy_to_user() attempts to read.
Fixes: e02789a53d71 ("smb3: enumerating snapshots was leaving part of the data off end")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Frank Sorenson <sorenson@redhat.com>
Date: Wed Sep 16 16:34:00 2026 -0500
smb: client: fix reparse buffer bounds in cifs_query_reparse_point()
commit 5f0306e731e2f46e91419eae57eee3a241c055e0 upstream.
In cifs_query_reparse_point(), the start >= end check before casting to
struct reparse_data_buffer * only ensures the start pointer is within the
response. It fails to verify that there is enough space remaining for the
fixed 8-byte header of the structure.
If a server provides a DataOffset that leaves less than 8 bytes remaining,
the check passes, but subsequent reads of ReparseTag and ReparseDataLength
will occur out-of-bounds.
Fix this by ensuring the remaining space is at least the size of the
reparse_data_buffer structure before accessing its fields.
Fixes: 56e84c64fc25 ("cifs: Fix validation of SMB1 query reparse point response")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Paulo Alcantara <pc@manguebit.org>
Date: Wed Sep 9 13:52:14 2026 -0300
smb: client: fix rlist race and missing initialization
commit 5f270f091256da1338c3631083e15d7f83cc05e1 upstream.
TCP_Server_Info.rlist is allocated via kzalloc which zeros both ->next
and ->prev to NULL instead of pointing to itself, making list_empty()
always return false and list_add() dereference a NULL ->prev pointer.
Also, cifs_signal_cifsd_for_reconnect() can be called concurrently
from multiple cifsd threads, allowing the same server's rlist node to
be added twice into the local list, corrupting it.
Closes: https://sashiko.dev/#/patchset/20260911204446.1719356-1-pc%40manguebit.org
Fixes: df0e03a4fb94 ("smb: client: fix potential deadlock when reconnecting channels")
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Cc: David Howells <dhowells@redhat.com>
Cc: Shyam Prasad N <sprasad@microsoft.com>
Cc: Ronnie Sahlberg <ronniesahlberg@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Bharath SM <bharathsm@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Frank Sorenson <sorenson@redhat.com>
Date: Wed Sep 16 16:33:54 2026 -0500
smb: client: fix server->total_read for compound encrypted PDUs
commit f73726b83e4756fdaa099e1bc1143293bd57ad79 upstream.
In receive_encrypted_standard(), server->total_read is left at the
full decrypted frame size when walking sub-PDUs of a compound encrypted
frame. As a result, cifs_handle_standard() passes this full size
to smb2_check_message(), causing the PDU length guards to incorrectly
validate the entire compound frame instead of the current sub-PDU.
This allows truncated non-last sub-PDUs to bypass length validation,
leading to out-of-bounds reads in smb2_get_data_area_len().
Fix this by setting server->total_read to the true length of the
current sub-PDU: next_cmd for non-last sub-PDUs, and the remaining
pdu_length for the last one.
Fixes: b24df3e30cbf ("cifs: update receive_encrypted_standard to handle compounded responses")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Paulo Alcantara <pc@manguebit.org>
Date: Sat Sep 12 14:20:08 2026 -0300
smb: client: fix smbd_connection leak on cifs_get_tcp_session() error
commit e75c96157d45e498970158c8f7373d90102e33b9 upstream.
When an RDMA connection is successfully established via
smbd_get_connection() but cifs_get_tcp_session() later fails (e.g.
kthread_create() returns an error), the error path frees tcp_ses
without first destroying the smbd_connection.
Fix this by calling smbd_destroy() in the out_err cleanup path before
kfree(tcp_ses). smbd_destroy() safely handles the case where
smbd_conn is NULL, so it can be called unconditionally.
Closes: https://sashiko.dev/#/patchset/20260912165503.521597-1-pc%40manguebit.org
Fixes: 2f8946464b11 ("CIFS: SMBD: Upper layer connects to SMBDirect session")
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Cc: Tom Talpey <tom@talpey.com>
Cc: Stefan Metzmacher <metze@samba.org>
Cc: Shyam Prasad N <sprasad@microsoft.com>
Cc: Ronnie Sahlberg <ronniesahlberg@gmail.com>
Cc: Bharath SM <bharathsm@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Paulo Alcantara <pc@manguebit.org>
Date: Sun Sep 13 21:09:15 2026 -0300
smb: client: fix unaligned access in WSL reparse point parser
commit e1aeaf79dea51e6065da56924bc07e22d59012ac upstream.
When wsl_to_fattr() parses WSL extended attributes, it computes a
payload pointer from ea->ea_data + ea_name_length + 1. Since the
smb2_file_full_ea_info struct is __packed and all WSL xattr names are
6 bytes long, the value pointer always lands at an odd byte offset,
never satisfying __le32 or __le64 alignment requirements.
The code then casts this pointer to __le32 * or __le64 * and
dereferences it directly, which may cause alignment faults on some
architectures.
Replace all such casts with get_unaligned_le32() and
get_unaligned_le64() in reparse_mkdev(), wsl_make_kuid(),
wsl_make_kgid() and wsl_to_fattr().
Closes: https://sashiko.dev/#/patchset/20260906200517.725015-1-pc%40manguebit.org
Fixes: 78e26bec4d6d ("smb: client: parse uid, gid, mode and dev from WSL reparse points")
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Cc: David Howells <dhowells@redhat.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Shyam Prasad N <sprasad@microsoft.com>
Cc: Ronnie Sahlberg <ronniesahlberg@gmail.com>
Cc: Bharath SM <bharathsm@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Joseph Qi <joseph.qi@linux.alibaba.com>
Date: Tue Sep 1 09:04:13 2026 +0800
smb: client: fix use-after-free of iface in cifs_try_adding_channels()
commit d034e836eefd7ce75e588f7031cffbeec594f5ac upstream.
cifs_try_adding_channels() iterates ses->iface_list with
list_for_each_entry_safe_from(), which captures the next entry
(niface) under iface_lock. The loop body then drops iface_lock for
the whole duration of cifs_ses_add_channel().
A concurrent interface refresh (SMB3_request_interfaces() ->
parse_server_interfaces()) marks all ifaces inactive and removes and
frees any that are not re-advertised via list_del() + kref_put(),
where release_iface() is a bare kfree(). Since niface typically has
no channel holding a reference, the list reference is its last and it
can be freed inside the unlocked window. On continue, the iterator
advance step then dereferences niface->iface_head.next, and the loop
body reads iface->rdma_capable/is_active, both on freed memory.
Fix this by never keeping an unreferenced list pointer across the
unlocked window. Each channel attempt now re-scans the list from the
head under iface_lock, takes a kref on the selected candidate, and
passes only that referenced candidate to cifs_ses_add_channel().
weight_fulfilled still tracks selection progress, so restarting the
scan preserves the original weighted distribution and the
weight_fulfilled-before-kref_put ordering on the failure path.
Add a per-pass attempts cap so a flapping interface refresh cannot
keep the inner loop spinning within a single tries increment.
Fixes: aa45dadd34e4 ("cifs: change iface_list from array to sorted linked list")
Cc: stable@vger.kernel.org
Assisted-by: Qoder:Qwen3.8-Max
Signed-off-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Acked-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Frank Sorenson <sorenson@redhat.com>
Date: Wed Sep 16 16:33:56 2026 -0500
smb: client: reject short Next offsets in parse_server_interfaces()
commit 1b3221bb121079ad79a1f3c3aa360ba649832e7a upstream.
In parse_server_interfaces(), the server-supplied Next offset is
validated against bytes_left, but not against the size of the interface
structure itself.
A small, non-zero Next value can pass the bounds check but advance the
pointer by less than sizeof(*p). This causes the next iteration of the
loop to read misaligned, overlapping structure fields.
Fix this by ensuring the Next offset is at least sizeof(*p).
Fixes: 7d34ec36abb8 ("smb3: fix for slab out of bounds on mount to ksmbd")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Jérémy Jean <Jeremy.Jean@oss.cyber.gouv.fr>
Date: Thu Sep 10 18:37:43 2026 +0000
smb: client: validate absolute native symlink targets before NT fixups
[ Upstream commit 23c240d9509e15f72e4112fc95f0160ab32ec430 ]
With symlinkroot unset, an absolute target is copied without conversion
to an NT drive path. Later code still assumes an NT prefix is present
when modifying the target and calculating the print name length.
For "/ab", this causes two failures: sym[5] and path[5] are written
past their allocations, and plen -= 2 * poff subtracts an assumed
8-byte prefix from a 6-byte UTF-16 target, wrapping u16 plen to 65534.
That underflow causes another overflow: memcpy() copies 65534 bytes
into a 24-byte buffer. A user with write access to a mounted share
can trigger these bugs with default settings.
Validate the NT drive prefix, including an ASCII drive letter, before
accessing fixed offsets or subtracting the prefix length.
Fixes: 3363da82e02f ("smb: client: fix native SMB symlink traversal")
Assisted-by: Codex:gpt-5
Signed-off-by: Jérémy Jean <Jeremy.Jean@oss.cyber.gouv.fr>
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Alexey Klimov <alexey.klimov@linaro.org>
Date: Thu Sep 17 10:16:42 2026 +0200
soc: samsung: exynos-pmu: fix use-after-free of interrupt generator node
commit 4dd1999783d7d12434006289338373e49492dc96 upstream.
The setup_cpuhp_and_cpuidle() parses the device tree node for the
interrupt generation block via of_parse_phandle() and decrements its
reference count using of_node_put() immediately after fetching the resource
address. However, later the intr_gen_node pointer is passed into
of_syscon_register_regmap().
Fix this by declaring intr_gen_node with __free() and removing
of_node_put().
Reported-by: Sashiko <sashiko-bot@kernel.org>
Closes: https://sashiko.dev/#/patchset/20260513-exynos850-cpuhotplug-v4-0-54fec5f65362@linaro.org?part=3
Fixes: 78b72897a5c8 ("soc: samsung: exynos-pmu: Enable CPU Idle for gs101")
Cc: stable@vger.kernel.org
Signed-off-by: Alexey Klimov <alexey.klimov@linaro.org>
Link: https://patch.msgid.link/20260828-exynos-pmu-cpuhp-idle-fixes-v2-1-06bce6107bd6@linaro.org
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20260917081641.72291-2-krzk@kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Bard Liao <yung-chuan.liao@linux.intel.com>
Date: Tue Sep 1 11:10:19 2026 +0800
soundwire: cadence_master: wait and cancel cdns->work before clock stop
[ Upstream commit aba7b41faeecb7692458095ce6fafc341fe0b80e ]
A peripheral event could happen during the clock stop process. We need
to wait for the event be handled before stopping the bus clock.
Otherwise, we will get the IO transfer timed out issue.
Fixes: af4cc917826f ("soundwire: cadence: mask Slave interrupt before stopping clock")
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: David Lin <david.lin@intel.com>
Reviewed-by: Shuming Fan <shumingf@realtek.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.dev>
Link: https://patch.msgid.link/20260901031019.233254-1-yung-chuan.liao@linux.intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ian Luites <ian@luites.com>
Date: Mon Aug 31 08:25:47 2026 +0000
soundwire: dmi-quirks: Disable ghost Realtek on Asus GX651AX
commit 6d49beec658f61801e79019fd73691b17252dee3 upstream.
The Asus ROG Zephyrus Duo GX651AX exposes a Realtek RT722 device in
ACPI which does not exist in the physical hardware. The device remains
unattached while the CS42L43 and both CS35L56 devices attach
successfully.
This confuses the function topology machine driver into creating
duplicate DAI links named SDW3-Playback-SimpleJack, and the sof_sdw
probe fails with error -12. Add a model-specific quirk to remove the
ghost RT722 device.
Fixes: 45cf24da0a10 ("ASoC: Intel: soc-acpi-intel-ptl-match: Remove unnecessary cs42l43 match")
Cc: stable@vger.kernel.org # 7.2.x
Assisted-by: LLM
Signed-off-by: Ian Luites <ian@luites.com>
Reviewed-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://patch.msgid.link/20260831082534.224716-1-ian@luites.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Frieder Schrempf <frieder.schrempf@kontron.de>
Date: Thu Sep 17 16:10:15 2026 +0200
spi: fsl-qspi: Reprogram the clock rate when the operation frequency changes
commit 3d743adf090cd4c9a2120c1e02b0482e88aa0d2d upstream.
fsl_qspi_select_mem() returns early when the chip select has not changed,
which happens before it reaches clk_set_rate(). Since the rate is now
taken from the spi-mem operation rather than from the SPI device, the
controller honours op->max_freq exactly once per chip select and ignores
it for every operation after that.
q->selected is only reset to -1 in fsl_qspi_default_setup(), i.e. at probe
and on resume, so on the common single chip select board the very first
operation latches a rate that all subsequent operations inherit, whatever
frequency they asked for.
This results in operations being issued with the wrong frequency.
Cache the operation frequency the clock was programmed for next to the
selected chip select, and redo the clock setup when either changes.
Fixes: 2438db5253eb ("spi: fsl-qspi: Support per spi-mem operation frequency switches")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-5
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Acked-by: Han Xu <han.xu@nxp.com>
Link: https://patch.msgid.link/20260917-fsl-qspi-freq-op-fix-v1-1-5fbe6b02f738@kontron.de
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Gabor Juhos <j4g8y7@gmail.com>
Date: Wed Sep 9 16:08:25 2026 +0200
spi: spi-qpic-snand: avoid writing QPIC_EBI2_ECC_BUF_CFG register
[ Upstream commit 930a7312c946bf4731721cadd82bb9a2ada496ca ]
The description of commit bfb34eced559 ("mtd: rawnand: qcom: avoid writing
to obsolete register") says this:
"QPIC_EBI2_ECC_BUF_CFG register got obsolete from QPIC V2.0 onwards.
Avoid writing this register if QPIC version is V2.0 or newer."
Although the referenced commit is related to the 'qcom-nandc' driver,
however the hardware supported by the current driver is also based on
QPIC v2.0 so we should avoid writing that register here as well.
Remove the register writing code to avoid undefined behaviour.
Fixes: 7304d1909080 ("spi: spi-qpic: add driver for QCOM SPI NAND flash Interface")
Signed-off-by: Gabor Juhos <j4g8y7@gmail.com>
Reviewed-by: Md Sadre Alam <md.alam@oss.qualcomm.com>
Link: https://patch.msgid.link/20260909-qpic-snand-avoid-ebi2-reg-write-v1-1-9b1b1466cc75@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Itai Handler <itai.handler@gmail.com>
Date: Thu Sep 10 20:48:32 2026 +0300
spi: spi-zynqmp-gqspi: stop the controller on shutdown
commit e922bad8b2d5028c51a096d083fea41cd0987154 upstream.
The driver has no ->shutdown, and platform_drv_shutdown() has no
fallback of its own. Unlike pci_device_shutdown(), which clears bus
mastering when kexec_in_progress, nothing on the platform bus disarms a
device that can still write to memory. The normal kexec path never
calls ->suspend either, so the quiesce in zynqmp_qspi_suspend() is not
reached.
A controller that is still executing a DMA read may therefore keep
writing to memory across a kexec. QSPIDMA_DST_ADDR still points at
memory owned by the kernel that called kexec, DST_SIZE is non-zero and
the flash is still clocked, so data can keep landing in RAM while the
new kernel is being relocated, and after it has started executing.
That destination is a physical address which means nothing to the new
kernel, so the writes can corrupt whatever now occupies it: kernel text
or data, page tables, or the initrd. Nothing reports an error and the
resulting behaviour is undefined.
This can be observed by reading GQSPI_EN (offset 0x114) and
QSPIDMA_DST_ADDR/SIZE/STS/CTRL (offsets 0x800 to 0x80c) early in the new
kernel, before the driver probes: without this patch GQSPI_EN reads 1
and QSPIDMA_DST_ADDR still points into the previous kernel's memory.
Add a ->shutdown that stops the controller the way zynqmp_qspi_suspend()
already does. spi_controller_suspend() stops the queue, waits for a
message that is already executing and makes any later transfer fail with
-ESHUTDOWN, so nothing can be cut short by the register write that
follows. It may sleep, which is fine here: device_shutdown() runs in
process context. Unlike ->suspend this cannot abort on error, because a
controller left mastering the bus is worse than a truncated transfer, so
a failure to drain is only logged.
GQSPI_EN_OFST is then cleared, as zynqmp_qspi_remove() and
zynqmp_qspi_suspend() already do. Skip that write only when
pm_runtime_get_if_in_use() returns 0, i.e. runtime suspended: the clocks
are gated, so the registers are unreachable and the controller cannot be
mastering the bus. A negative return is not the same thing - it is what
the CONFIG_PM=n stub always returns, and there probe() has enabled pclk
and refclk for good, so the controller is running and must be stopped.
Fixes: dfe11a11d523 ("spi: Add support for Zynq Ultrascale+ MPSoC GQSPI controller")
Cc: stable@vger.kernel.org
Signed-off-by: Itai Handler <itai.handler@gmail.com>
Link: https://patch.msgid.link/20260910174832.873352-1-itai.handler@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Hao-Qun Huang <alvinhuang0603@gmail.com>
Date: Sun Sep 13 03:20:49 2026 +0800
spi: virtio: Use the per-transfer bits per word
commit 095858324f063dba830041f067872f0a08765d2f upstream.
virtio_spi_transfer_one() puts spi->bits_per_word into the request
header, so a transfer that sets its own word size reaches the backend
with the device default instead. The SPI core has already copied that
default into xfer->bits_per_word when the transfer leaves it at zero,
the same way it does for xfer->speed_hz, which this function already
uses.
Per-transfer word sizes are ordinary SPI usage. mipi_dbi, for one, sends
a 9-bit command and reads the reply as 8-bit data in the same message.
With a 16-bit device default, a one-byte transfer asking for 8 bits goes
out as a partial 16-bit word, which the backend may reject.
Fixes: f98cabe3f6cf ("SPI: Add virtio SPI driver")
Cc: stable@vger.kernel.org
Assisted-by: LLM
Signed-off-by: Hao-Qun Huang <alvinhuang0603@gmail.com>
Link: https://patch.msgid.link/20260913032049.11209.alvinhuang0603@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Jens Axboe <axboe@kernel.dk>
Date: Tue Sep 1 19:39:46 2026 +0200
sunvdc: fix -EIO issue due to lack of retries
[ Upstream commit 5067d4ba713961d8ccea1e06cd4c453793f3121e ]
John reports that since commit:
a11f6ca9aef9 ("sunvdc: Do not spin in an infinite loop when vio_ldc_send() returns EAGAIN")
users of Linux inside Solaris ldom see occasional -EIO errors because
the request send loop now times out. The current loop does 10 retries,
and inside vio_ldc_send() a further 1000 1usec retries are done as well.
Even with 10.5 msec of busy loop retries that's apparently not enough to
always succeed.
Rather than introduce continued busy looping, requeue the request and
have the delayed queue kicking retry the request after another 10ms.
This obviously isn't ideal, but there's seemingly no way to wait for
this type of event. And if 10ms of busy looping was not enough to make
progress, then presumably this is an edge condition and we just need to
guarantee to make forward progress at some later point in time. That's
more suitably done through letting the CPU tend to other work, rather
than sitting in a tight loop retrying.
[stian: rebased on top of the cookie-unmap fix, without which every
requeued attempt leaks LDC map table entries; tested on an
UltraSPARC T4 LDOM where the vdc_tx_trigger failure condition was
reproduced and absorbed by the requeue with no I/O error]
Reported-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Link: https://lore.kernel.org/all/20251006100226.4246-2-glaubitz@physik.fu-berlin.de/
Link: https://lore.kernel.org/all/418310b3-2b77-4534-b2fd-27dcc11e333c@kernel.dk/
Signed-off-by: Stian Halseth <stian@itx.no>
Link: https://patch.msgid.link/20260901173947.3292110-3-stian@itx.no
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Donggeun Yoo <donggeunyoo.kernel@gmail.com>
Date: Sat Sep 5 17:42:10 2026 +0900
swiotlb: use the adjusted address for the highmem page lookup
commit b7d7914a9ae3097e63d113007e4fb44d33d515b1 upstream.
swiotlb_bounce() reads the page frame number from the slot's recorded
orig_addr, then advances orig_addr by tlb_offset to reach the address
the caller asked about. The highmem branch mixes the two: the offset
within the page comes from the adjusted address, the page from the value
before it.
Once the adjustment crosses a page boundary the pair no longer describes
one location, and the whole copy lands one page below the intended one
for a positive tlb_offset, one above for a negative one. DMA_FROM_DEVICE
writes the device data over the wrong page and leaves the intended one
stale, DMA_TO_DEVICE feeds the device from a page the mapping may not
cover. Partial syncs through dma_sync_single_range_for_*() are what make
tlb_offset non-zero.
The branch test is picked the same way, so a slot recorded in lowmem can
be adjusted into highmem and the lowmem path then hands a highmem
address to phys_to_virt().
Take both from orig_addr once it is final and keep pfn in the branch
that uses it. PhysHighMem() asks the question straight from the address,
as dma-debug already does.
Fixes: 5f89468e2f06 ("swiotlb: manipulate orig_addr when tlb_addr has offset")
Cc: stable@vger.kernel.org
Signed-off-by: Donggeun Yoo <donggeunyoo.kernel@gmail.com>
Reviewed-by: Michael Kelley <mhklinux@outlook.com>
Link: https://lore.kernel.org/r/20260905084210.148255-1-donggeunyoo.kernel@gmail.com
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Kuniyuki Iwashima <kuniyu@google.com>
Date: Sat Sep 5 23:36:30 2026 +0000
sysctl: Check range in proc_dointvec_ms_jiffies_minmax
[ Upstream commit 82431877d837a6c2593efd8e66ba28b35a220ca4 ]
Add the range check to do_proc_int_conv_ms_jiffies_minmax that commit
d174174c6776 ("sysctl: replace SYSCTL_INT_CONV_CUSTOM macro with
functions") incorrectly removed.
Fixes: d174174c6776 ("sysctl: replace SYSCTL_INT_CONV_CUSTOM macro with functions")
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Signed-off-by: Joel Granados <joel.granados@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Sat Sep 12 14:48:48 2026 +0000
tcp: do not let tcp_rmem be set below 4096
[ Upstream commit 83a945a529d6e002dd7339c532288a931f463dba ]
We can hit a division by zero crash in tcp_rcvbuf_grow()
and tcp_rcv_space_adjust():
divide error: 0000 [#1] PREEMPT SMP
RIP: 0010:tcp_rcvbuf_grow+0x187/0x450 net/ipv4/tcp_input.c:939
...
grow = div_u64(((u64)rcvwin << 1) * (newval - oldval), oldval);
The division uses oldval = tp->rcvq_space.space as divisor.
When tp->rcvq_space.space is zero, this leads to a divide-by-zero
exception.
tp->rcvq_space.space is initialized in tcp_init_buffer_space():
tp->rcvq_space.space = min3(tp->rcv_ssthresh, tp->rcv_wnd,
(u32)TCP_INIT_CWND * tp->advmss);
If tcp_rmem[1] is configured to very small values (such as 1),
sk->sk_rcvbuf is initialized to 1. Then tcp_full_space(sk), which
computes (sk->sk_rcvbuf * scaling_ratio) >> 8, truncates to 0.
This sets tp->window_clamp = 0, tp->rcv_ssthresh = 0, and
tp->rcvq_space.space = 0. Later, when data arrives and DRS is invoked,
tcp_rcvbuf_grow() divides by oldval == 0.
Back in 2015, commit b1cb59cf2efe ("net: sysctl_net_core: check SNDBUF
and RCVBUF for min length") ensured that net.core.rmem_default and
net.core.rmem_max cannot be set below SOCK_MIN_RCVBUF. Similarly,
SO_RCVBUF setsockopt enforces max_t(int, val * 2, SOCK_MIN_RCVBUF).
However, net.ipv4.tcp_rmem still had .extra1 = SYSCTL_ONE, allowing
arbitrarily small values.
Because SOCK_MIN_RCVBUF depends on sizeof(struct sk_buff) and cacheline
alignment, its value varies across architectures and configuration options.
Using a fixed constant of 4096 ensures a predictable, architecture-
independent lower bound that is safely above SOCK_MIN_RCVBUF everywhere
and matches the documented 4K default.
Fix this by setting tcp_rmem.extra1 to 4096 and updating the documentation.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260912144848.3448026-1-edumazet@google.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Kuniyuki Iwashima <kuniyu@google.com>
Date: Mon Sep 14 01:14:01 2026 +0000
tcp: Don't call skb_clone_and_charge_r() for close()d listener in tcp_v6_do_rcv().
[ Upstream commit 8e759cd1f6444a946bd1fd2b2b29eea582eea1d5 ]
tcp_v6_do_rcv() no longer calls skb_clone_and_charge_r() for
TCP_LISTEN since commit 073d89808c06 ("net: fix data-races around
sk->sk_forward_alloc").
However, there is still a small race window between tcp_v6_rcv()
and tcp_v6_do_rcv(), where concurrent close() changes TCP_LISTEN
to TCP_CLOSE, causing skb_clone_and_charge_r() to be called
locklessly and resulting in the splat below. [0]
Let's avoid calling skb_clone_and_charge_r() for TCP_CLOSE as well.
This is fine for non-listeners because tcp_rcv_state_process()
drops skb for TCP_CLOSE and opt_skb was freed immediately anyway.
[0]:
sk->sk_forward_alloc
WARNING: net/ipv4/af_inet.c:162 at inet_sock_destruct+0x64d/0x810 net/ipv4/af_inet.c:162, CPU#1: ksoftirqd/1/28
Modules linked in:
CPU: 1 UID: 0 PID: 28 Comm: ksoftirqd/1 Not tainted 7.2.0 #17 PREEMPT(full)
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.17.0-debian-1.17.0-1 04/01/2014
RIP: 0010:inet_sock_destruct+0x64d/0x810 net/ipv4/af_inet.c:162
Code: 3d 49 ff e9 06 fd ff ff e8 d0 5b 83 f8 90 0f 0b 90 e9 35 fe ff ff e8 c2 5b 83 f8 90 0f 0b 90 e9 c5 fe ff ff e8 b4 5b 83 f8 90 <0f> 0b 90 e9 04 ff ff ff e8 a6 5b 83 f8 90 0f 0b 90 e9 65 fe ff ff
RSP: 0018:ffffc90000677bb8 EFLAGS: 00010246
RAX: 0000000000000000 RBX: ffff8880117bde80 RCX: ffffffff8957eb41
RDX: ffff88801dad5d00 RSI: ffffffff8957ec3c RDI: 0000000000000005
RBP: 00000000fffff000 R08: ffffffff8957eb41 R09: 00000000fffff000
R10: 0000000000000005 R11: 0000000000000000 R12: dffffc0000000000
R13: ffff8880117bdf10 R14: ffffffff81c08eb7 R15: 0000000000000003
FS: 0000000000000000(0000) GS:ffff8880d7ae5000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f93a1021138 CR3: 00000000207a9000 CR4: 0000000000350ef0
Call Trace:
<TASK>
__sk_destruct+0x82/0xae0 net/core/sock.c:2356
rcu_do_batch kernel/rcu/tree.c:2645 [inline]
rcu_core+0x59c/0x1100 kernel/rcu/tree.c:2897
handle_softirqs+0x1e4/0x9b0 kernel/softirq.c:622
run_ksoftirqd kernel/softirq.c:1076 [inline]
run_ksoftirqd+0x38/0x60 kernel/softirq.c:1068
smpboot_thread_fn+0x458/0xc80 kernel/smpboot.c:160
kthread+0x396/0x4a0 kernel/kthread.c:436
ret_from_fork+0x8e0/0xe40 arch/x86/kernel/process.c:158
ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245
</TASK>
Fixes: e994b2f0fb92 ("tcp: do not lock listener to process SYN packets")
Reported-by: Taras Madan <tarasmadan@google.com>
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260914011420.115556-1-kuniyu@google.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Inbal Schussheim <inbal.lipshtat@mail.huji.ac.il>
Date: Mon Sep 14 12:04:07 2026 +0300
tcp: exclude old ACKs from tcp fast path
commit f81e6c3fb06327bc49cdd6e559845293ba06a704 upstream.
Exclude old ACKs before SND.UNA from the tcp fast path
as well as ACKs after SND.NXT.
Such ACKs will fall through to the slow path, where tcp_ack()
performs the appropriate validation and challenge ACK handling
according to RFC5961 and Commit 3d501dd326fb1c7 ("tcp: do not
accept ACK of bytes we never sent").
This prevents old ACKs from being accepted
or modifying connection state as part of the fast path before
appropriate ACK validation is applied.
In particular, this prevents payload carried by a segment with
an excessively old ACK from advancing RCV.NXT before the ACK
is rejected.
Fixes: 31770e34e43d ("tcp: Revert "tcp: remove header prediction"")
Reported-by: Amit Klein <amit.klein@mail.huji.ac.il>
Reported-by: Tamir Shahar <tamir.shahar1@mail.huji.ac.il>
Reported-by: Inbal Schussheim <inbal.lipshtat@mail.huji.ac.il>
Suggested-by: Eric Dumazet <edumazet@google.com>
Cc: stable@vger.kernel.org
Signed-off-by: Inbal Schussheim <inbal.lipshtat@mail.huji.ac.il>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260914090408.1435080-2-inbal.lipshtat@mail.huji.ac.il
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Li Jun <lijun01@kylinos.cn>
Date: Mon Sep 14 14:23:53 2026 +0800
watchdog: da9062: fix suspend/resume handling of HW_RUNNING watchdog
[ Upstream commit 5071122bf5a628494db16d98d253f622a5aab074 ]
da9062_wdt_suspend() and da9062_wdt_resume() only check watchdog_active(),
when the watchdog is left running by the driver sets
WDOG_HW_RUNNING in da9062_wdt_probe() but userspace never opens the
device, so WDOG_ACTIVE remains cleared, the wdt_disable() will not be
executed in da9062_wdt_suspend. In this case, the suspend callback is
a no-op and the watchdog keeps counting during system suspend,
leading to an unexpected system reset.
Check WDOG_HW_RUNNING and wdt->wdd，can fix this issue.
Fixes: f6c98b08381c7 ("watchdog: da9062: add power management ops")
Cs: stable@vger.kernel.org
Signed-off-by: Li Jun <lijun01@kylinos.cn>
Link: https://patch.msgid.link/20260914062353.582205-1-lijun01@kylinos.cn
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Li Jun <lijun01@kylinos.cn>
Date: Thu Sep 17 09:37:10 2026 +0800
watchdog: da9063: fix suspend/resume handling of HW_RUNNING watchdog
commit 7cb575b71ab98194d2e040bded3a7281e089c5ed upstream.
da9063_wdt_suspend() and da9063_wdt_resume() only check watchdog_active(),
when the watchdog is left running by the driver sets
WDOG_HW_RUNNING in da9063_wdt_probe() but userspace never opens the
device, so WDOG_ACTIVE remains cleared, the wdt_disable() will not be
executed in da9063_wdt_suspend. In this case, the suspend callback is
a no-op and the watchdog keeps counting during system suspend,
leading to an unexpected system reset.
Check WDOG_HW_RUNNING and wdd，can fix this issue.
Fixes: a7ceca4398bc8 ("watchdog: da9063: optionally disable watchdog during suspend")
Cc: stable@vger.kernel.org
Signed-off-by: Li Jun <lijun01@kylinos.cn>
Link: https://patch.msgid.link/20260917013710.2754679-1-lijun01@kylinos.cn
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Tzung-Bi Shih <tzungbi@kernel.org>
Date: Sun Sep 13 14:48:49 2026 +0800
watchdog: digicolor: Avoid division by zero
commit 400cb663ca019bae6eb878f06f1094ddf7c0b0df upstream.
clk_get_rate() could return 0. Avoid a division by zero panic.
Since get_timeleft() cannot propagate errors, check the clock rate early
in probe() and cache the rate in the driver data as it is unlikely to
change at runtime.
Fixes: 336694a01dae ("watchdog: digicolor: driver for Conexant Digicolor CX92755 SoC")
Cc: stable@vger.kernel.org
Signed-off-by: Tzung-Bi Shih <tzungbi@kernel.org>
Acked-by: Baruch Siach <baruch@tkos.co.il>
Link: https://patch.msgid.link/20260913064851.8239-2-tzungbi@kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Tzung-Bi Shih <tzungbi@kernel.org>
Date: Sun Sep 13 14:51:26 2026 +0800
watchdog: msc313e: Fix premature reset during timeout update
commit 22737cfced627ffcb4b5c36d63bb3d4476f63213 upstream.
Updating the 32-bit hardware timeout requires writing to two 16-bit
registers sequentially. If the watchdog is actively running, this
non-atomic update might trigger a premature system reset.
Clear the watchdog counter before updating the registers to prevent the
timer from timing out prematurely against an intermediate threshold.
Fixes: e9800b799464 ("watchdog: Add Mstar MSC313e WDT driver")
Cc: stable@vger.kernel.org
Signed-off-by: Tzung-Bi Shih <tzungbi@kernel.org>
Link: https://patch.msgid.link/20260913065126.8350-1-tzungbi@kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Tzung-Bi Shih <tzungbi@kernel.org>
Date: Sun Sep 13 00:33:34 2026 +0800
watchdog: msc313e: Propagate error code in resume()
commit 1d9763f34a85680db1e8233d654fdb85e5f897cc upstream.
If msc313e_wdt_start() fails during system resume, the error is
currently ignored. Consequently, the watchdog isn't running without the
user's knowledge.
Propagate the error code and print a message if msc313e_wdt_start()
fails.
Signed-off-by: Tzung-Bi Shih <tzungbi@kernel.org>
Fixes: e9800b7994642 ("watchdog: Add Mstar MSC313e WDT driver")
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260912163334.28636-1-tzungbi@kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Tzung-Bi Shih <tzungbi@kernel.org>
Date: Sun Sep 13 14:48:50 2026 +0800
watchdog: rtd119x: Avoid division by zero
commit 5af7d2cbd20f893def03c8310a460ade66a5d822 upstream.
clk_get_rate() could return 0. Avoid a division by zero panic.
Fixes: 2bdf6acbfead ("watchdog: Add Realtek RTD1295")
Cc: stable@vger.kernel.org
Signed-off-by: Tzung-Bi Shih <tzungbi@kernel.org>
Link: https://patch.msgid.link/20260913064851.8239-3-tzungbi@kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Tzung-Bi Shih <tzungbi@kernel.org>
Date: Sun Sep 13 14:48:51 2026 +0800
watchdog: rzv2h: Avoid division by zero
commit 6274281c41efa8dd1aa5234c59ad904ff89d7af4 upstream.
clk_get_rate() could return 0. Avoid a division by zero panic.
Fixes: f6febd0a30b6 ("watchdog: Add Watchdog Timer driver for RZ/V2H(P)")
Cc: stable@vger.kernel.org
Signed-off-by: Tzung-Bi Shih <tzungbi@kernel.org>
Link: https://patch.msgid.link/20260913064851.8239-4-tzungbi@kernel.org
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Wentao Liang <vulab@iscas.ac.cn>
Date: Wed Sep 16 17:05:11 2026 +0000
watchdog: sp5100_tco: Fix pci_dev reference leak in sp5100_tco_init()
commit 88f113634028ca90a857031837d8061d1a9e1a7b upstream.
sp5100_tco_init() stores the PCI device matched by for_each_pci_dev()
in the global sp5100_tco_pci and keeps its reference for the lifetime
of the driver, but neither sp5100_tco_exit() nor the error paths of
sp5100_tco_init() call pci_dev_put(), leaking the reference on driver
registration failure and on every module load/unload cycle.
Drop the reference when the platform driver or device registration
fails and when the module is unloaded.
Fixes: 15e28bf13008 ("watchdog: Add support for sp5100 chipset TCO")
Cc: stable@vger.kernel.org
Signed-off-by: Wentao Liang <vulab@iscas.ac.cn>
Link: https://patch.msgid.link/20260916170511.2086199-1-vulab@iscas.ac.cn
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Wentao Liang <vulab@iscas.ac.cn>
Date: Wed Sep 16 17:07:04 2026 +0000
watchdog: starfive-wdt: Fix runtime PM leak in starfive_wdt_pm_start()
commit 8f0ca55016a7647109ae2bc91bcb346fc8b13785 upstream.
starfive_wdt_pm_start() takes a runtime PM reference with
pm_runtime_get_sync(), which increments the usage counter even when it
fails, and returns the error without dropping it again. The watchdog
core does not invoke the stop callback when start fails, so the
reference taken on the error path is leaked.
Use pm_runtime_resume_and_get() instead, which keeps the usage counter
balanced when the resume fails.
Fixes: db728ea9c7be ("drivers: watchdog: Add StarFive Watchdog driver")
Cc: stable@vger.kernel.org
Signed-off-by: Wentao Liang <vulab@iscas.ac.cn>
Link: https://patch.msgid.link/20260916170704.2086331-1-vulab@iscas.ac.cn
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Nicolas Escande <nico.escande@gmail.com>
Date: Fri Jul 31 16:58:30 2026 +0200
wifi: ath11k: cleanup arsta in ath11k_mac_peer_cleanup_all()
[ Upstream commit 820b8cff81c796ba20573e04722ab62500713f97 ]
When mac80211 removes a sta, it calls .sta_state() which in turn calls
ath11k_mac_station_remove(). In that function we clean up both peers &
arsta related resources.
But when the firmware crashes, ath11k calls ieee80211_restart_hw(), which
assumes that all driver related resources are cleaned up beforehand. This
cleanup is supposedly done by ath11k_mac_peer_cleanup_all() but does not
in fact free arsta->rx_stats / tx_stats.
Extract the arsta cleanup from ath11k_mac_station_remove() into a
new ath11k_mac_station_cleanup() and call it from both there and
ath11k_mac_peer_cleanup_all().
This should handle kmemleaks reports like:
unreferenced object 0xffffff801ae66400 (size 1024):
comm "hostapd", pid 1306, jiffies 4295011565
hex dump (first 32 bytes):
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
backtrace (crc d61c08ec):
kmemleak_alloc+0x3c/0x50
__kmalloc_cache_noprof+0x2b0/0x3e0
ath11k_mac_op_sta_state+0x1dc/0xb10
drv_sta_state+0xac/0x6f8
sta_info_insert_rcu+0x314/0x5e0
sta_info_insert+0x14/0x38
ieee80211_add_station+0x10c/0x1a0
nl80211_new_station+0x3e8/0x680
genl_family_rcv_msg_doit+0xc0/0x120
genl_rcv_msg+0x1b4/0x258
netlink_rcv_skb+0x4c/0x108
genl_rcv+0x38/0x60
netlink_unicast+0x190/0x278
netlink_sendmsg+0x15c/0x370
____sys_sendmsg+0x120/0x290
___sys_sendmsg+0x70/0xa0
Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.9.0.1-01977-QCAHKSWPL_SILICONZ-1
Fixes: d5c65159f289 ("ath11k: driver for Qualcomm IEEE 802.11ax devices")
Signed-off-by: Nicolas Escande <nico.escande@gmail.com>
Reviewed-by: Rameshkumar Sundaram <rameshkumar.sundaram@oss.qualcomm.com>
Reviewed-by: Baochen Qiang <baochen.qiang@oss.qualcomm.com>
Link: https://patch.msgid.link/20260731145830.769811-1-nico.escande@gmail.com
Signed-off-by: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Bogdan Nicolae <bogdan.nicolae@gmail.com>
Date: Fri Aug 7 11:34:18 2026 -0500
wifi: brcmfmac: cyw: pass PMKID to firmware if present
[ Upstream commit e2de8d5eb2984416affdd9559e55f37c7f1bbf47 ]
Zero out auth_status on initialization. Otherwise, garbage will
leak from the stack to the firmware (when ssid is less than 32 bytes
and/or when params->pmkid is set). Then, pass the params->pmkid to the
firmware (without it, the firmware caches a garbage PMKID on successful
authentication and denies a subsequent association request that includes
the PMKID).
Fixes: 66f909308a7c ("wifi: brcmfmac: cyw: support external SAE authentication in station mode")
Signed-off-by: Bogdan Nicolae <bogdan.nicolae@gmail.com>
Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Link: https://patch.msgid.link/20260807163418.487508-1-bogdan.nicolae@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Karl Mehltretter <kmehltretter@gmail.com>
Date: Tue Aug 11 10:27:02 2026 +0200
wifi: brcmfmac: fix lost 802.1x TX completion wakeup
[ Upstream commit 621d90169cef6c8da5b6134db5c0c4e23cdd09ce ]
brcmf_txfinalize() decrements pend_8021x_cnt before a lockless
waitqueue_active() check. atomic_dec() does not order the decrement
against the check.
The waiter can therefore observe a nonzero count while the waker observes
an empty queue, losing the final wakeup and delaying key installation
until the 950 ms timeout.
Add smp_mb__after_atomic() to order the decrement before the queue
check. wait_event_timeout() provides the matching barrier. LKMM confirms
that this forbids the lost-wakeup outcome.
Fixes: 21fff75d2fb6 ("brcmfmac: use wait_event_timeout for 8021x pending count")
Assisted-by: Claude:claude-fable-5
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Link: https://patch.msgid.link/20260811082702.44521-1-kmehltretter@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jiangshan Yi <yijiangshan@kylinos.cn>
Date: Sat Aug 15 20:10:43 2026 +0800
wifi: brcmsmac: fix UAF in brcms_free_timer()
commit 1eeca1d5e0920fbdad6449768fd2d4364e714180 upstream.
brcms_free_timer() calls brcms_del_timer() which uses the non-synchronous
cancel_delayed_work() to cancel the timer's underlying delayed work. If
the work callback (_brcms_timer) is already running, cancel_delayed_work()
returns false without waiting, and brcms_free_timer() proceeds to kfree(t)
while the callback still accesses t through container_of().
Add an explicit cancel_delayed_work_sync() after brcms_del_timer() to
guarantee that any in-flight callback has completed before the timer
structure is freed.
Fixes: 5b435de0d786 ("net: wireless: add brcm80211 drivers")
Cc: stable@vger.kernel.org
Signed-off-by: Jiangshan Yi <yijiangshan@kylinos.cn>
Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Link: https://patch.msgid.link/20260815121043.938414-1-yijiangshan@kylinos.cn
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:55:02 2026 +0200
wifi: cfg80211: check IP header size in cfg80211_classify8021d()
[ Upstream commit 48b2c5c628b09cf36cbeca53e0432fc2a7518be7 ]
A frame that looks like IP can be transmitted, but be too short, so
the DS field is read incorrectly:
BUG: KMSAN: uninit-value in cfg80211_classify8021d+0x99d/0x12b0 net/wireless/util.c:1027
cfg80211_classify8021d+0x99d/0x12b0 net/wireless/util.c:1027
ieee80211_select_queue+0x37a/0x9e0 net/mac80211/wme.c:180
__ieee80211_subif_start_xmit+0x60f/0x1d90 net/mac80211/tx.c:4304
ieee80211_subif_start_xmit+0xa8/0x6d0 net/mac80211/tx.c:4538
...
packet_sendmsg+0x9173/0xa2a0 net/packet/af_packet.c:3108
Use skb_header_pointer() like the MPLS case.
Assisted-by: LLM
Fixes: e31a16d6f64e ("wireless: move some utility functions from mac80211 to cfg80211")
Reported-by: syzbot+878ddc3962f792e9af59@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=878ddc3962f792e9af59
Link: https://patch.msgid.link/20260904165614.5e61a4c80b92.I37d68d3f406cb3b90b32e6943418d66070b65197@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:55:05 2026 +0200
wifi: cfg80211: don't filter by BSS type when removing stale entries
[ Upstream commit b377e1000d963e7182a987082b4b06580bd7ac84 ]
When an assoc AP switches to a channel that already has a BSS entry,
cfg80211_update_assoc_bss_entry() removes that entry before rehashing
the real one, since the two would otherwise collide in the BSS rbtree.
The lookup for that entry also required it to match the connection's BSS
type, so an entry advertising e.g. the IBSS capability bit was left in
place, and the following cfg80211_rehash_bss() then ran into it:
WARN_ON(!cmp)
Changing the type shouldn't really happen, but can be triggered by a
rogue AP/device, so drop the check and remove any entries matching
the comparison.
Assisted-by: LLM
Fixes: 0afd425b1b64 ("cfg80211: fix duplicated scan entries after channel switch")
Reported-by: syzbot+dc6f4dce0d707900cdea@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=dc6f4dce0d707900cdea
Link: https://patch.msgid.link/20260904165614.1f05dae1c546.Ib52d57b57caa912efee020f9d4a033a5160617ce@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:55:03 2026 +0200
wifi: cfg80211: don't free driver-owned scan requests
[ Upstream commit dab68a74e90b8e07f08ed9deaa5884857a3cfe89 ]
When an interface goes down while a scan is running, cfg80211 completes
the scan towards userspace and frees the scan request. However, the
driver can be convinced that it owns the request, since the cancellation
is (intended to be) asynchronous.
The WARN_ON() in the netdev notifier was meant to catch this, but it's
not actually avoidable, so it triggers and we get a UAF in scan_done().
There doesn't seem to be a great way around it, so just track that the
driver is still convinced it owns the request, and then just free it on
completion if it was already cancelled. Also remove the warnings since
they can trigger in the intended architecture.
Assisted-by: LLM
Fixes: 4a58e7c38443 ("cfg80211: don't "leak" uncompleted scans")
Reported-by: syzbot+189dcafc06865d38178d@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=189dcafc06865d38178d
Link: https://patch.msgid.link/20260904165614.375e543228b1.I03cbb5a54cb02d6bba5034286af1ed73aba134d1@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:55:01 2026 +0200
wifi: cfg80211: don't get the radio mask for netdev-less wdevs
[ Upstream commit a7783e585360ee05dfe21d3173dbbe985c94f29e ]
cfg80211_calculate_bi_data() calls rdev_get_radio_mask() with
wdev->netdev, which can be NULL and then crashes in mac80211.
To avoid that, invert the order of checks since wdev->netdev
is always valid for beaconing interfaces.
Assisted-by: LLM
Fixes: abb4cfe3661a ("wifi: cfg80211: extend interface combination check for multi-radio")
Reported-by: syzbot+abff43d2d045e37c0bb2@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=abff43d2d045e37c0bb2
Link: https://patch.msgid.link/20260904165614.2056a8b7dc91.I7412c5062d8166ad6c81ee7252cec49dea19a60f@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:55:07 2026 +0200
wifi: cfg80211: fix NAN regulatory enforcement
[ Upstream commit 17a5f8571d1d40c88b78cfc154a7da0d60f13f37 ]
reg_wdev_chan_valid() returns early for any wdev that has no netdev,
which is fine for P2P originally (and later PD still), but NAN has
no netdev and yet enforcement code was added and is needed, but is
dead code right now.
Use wdev_running() instead so that netdev-less wdevs aren't skipped.
P2P/PD don't do anything in the later switch, but NAN code can now
be reached.
Assisted-by: LLM
Fixes: 0e8ec738a71e ("wifi: cfg80211: add support for NAN data interface")
Link: https://patch.msgid.link/20260904165614.6abc075b5401.Ib90696e3fa49b1698c27d64db5360d51f6f187a9@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 17:02:02 2026 +0200
wifi: cfg80211: get the wiphy out of a dying network namespace
[ Upstream commit 4635b1a1c1d693178a537446a6e09963f0fdae52 ]
When a network namespace is destroyed, cfg80211_pernet_exit() moves any
wiphy back to the initial namespace, and just warns if that fails. But
moving an interface can fail (due to allocation failures), and then the
wiphy is left behind with a garbage netns pointer:
Kernel mode fault at addr 0x30
genlmsg_multicast_netns.constprop.0+0x46/0xcf [cfg80211]
nl80211_notify_wiphy+0xcd/0xe8 [cfg80211]
wiphy_unregister+0x169/0x3fc [cfg80211]
Note that commit debac3a20dec ("net: Remove conflicting altnames for
dying netns in __dev_change_net_namespace().") fixed another path
that could reach it without allocation failures.
Remove interfaces that cannot be moved instead of failing the switch,
so that the wiphy always ends up in the initial namespace. In this
case the netdev core will unregister the interfaces anyway.
Assisted-by: LLM
Reported-by: syzbot+c5f8a81e794d4a4f2014@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=c5f8a81e794d4a4f2014
Fixes: 463d018323851 ("cfg80211: make aware of net namespaces")
Link: https://patch.msgid.link/20260904170220.7f3edc6d9992.I5e57921011244d3d8ef14d89e738aa19a5d972a0@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:55:06 2026 +0200
wifi: cfg80211: ibss: ref BSS entry for joined event
[ Upstream commit 708f9d43d6a2eb9c6b83fe62af628de9dffd9314 ]
When the IBSS is joined, we only record the BSSID/channel in the event
and look up the BSS entry when processing it. However, that's racy,
e.g. a new scan with NL80211_SCAN_FLAG_FLUSH can remove it, causing a
warning in the event work:
!bss
WARNING: net/wireless/ibss.c:37 at __cfg80211_ibss_joined+0x3d3/0x440
Workqueue: cfg80211 cfg80211_event_work
cfg80211_process_wdev_events+0x39f/0x5b0 net/wireless/util.c:1144
cfg80211_process_rdev_events+0xa1/0x110 net/wireless/util.c:1179
cfg80211_event_work+0x2f/0x40 net/wireless/core.c:393
Do the lookup early (the driver is expected to only join an IBSS that
has a BSS entry) and keep a reference to it.
Assisted-by: LLM
Fixes: 667503ddcb96 ("cfg80211: fix locking")
Reported-by: syzbot+7f064ba1704c2466e36d@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=7f064ba1704c2466e36d
Link: https://patch.msgid.link/20260904165614.f49a213f0e49.I192bfe738750ebb5f2c4faa3019a428da64cd3ec@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:55:04 2026 +0200
wifi: cfg80211: only group hidden BSSes with beacon entries
[ Upstream commit 068843ed0902c552a13860c5ec6b2ca65b57a065 ]
When a probe response for an unknown BSS comes in, __cfg80211_bss_update()
looks for an existing entry with the same BSSID and a hidden (zero-length
or NUL-filled) SSID, and if it finds one it groups them, using the beacon
IEs from the existing entry.
But that could find another entry without a beacon, if it was also from a
probe response (with SSID), so there's a group without beacon elements.
If a beacon with a hidden SSID for that BSSID arrives later,
cfg80211_combine_bsses() goes looking for the probe response entries that
belong to it - i.e. entries with the same BSSID and channel that have no
beacon IEs - and finds those two. They are already grouped with each
other, so it hits its
WARN_ON_ONCE(bss->pub.hidden_beacon_bss)
WARN_ON_ONCE(!list_empty(&bss->hidden_list))
which are there because an entry without beacon elements is not supposed
to be part of a group yet.
Only combine entries when a beacon was already received, ones that are
kept separate will be combined when a beacon arrives.
Assisted-by: LLM
Fixes: 4593c4cbe1c9 ("cfg80211: fix BSS list hidden SSID lookup")
Reported-by: syzbot+1a797e1c81be78a2ace7@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=1a797e1c81be78a2ace7
Link: https://patch.msgid.link/20260904165614.bcfa64715745.Iad740347c86de56d4ff4f96a95f3c3afc47c42de@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 17:01:59 2026 +0200
wifi: cfg80211: restore netns_immutable on failures
[ Upstream commit eeee52cfd1d639774c9812e8890631404a057dd2 ]
Switching a wiphy's netns has to clear netns_immutable before moving
interfaces, but then if any of the interfaces fails to move, it gets
netns_immutable cleared forever. Then userspace can move it by itself,
breaking the assumption that they all move together.
Fix the order here and always reset netns_immutable after attempting
the move.
Assisted-by: LLM
Fixes: 463d018323851 ("cfg80211: make aware of net namespaces")
Link: https://patch.msgid.link/20260904170220.7ea88157dcbc.Id868585a790be8b9ece9b39b0db464a5963faaf3@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 17:02:00 2026 +0200
wifi: cfg80211: undo netns switch if renaming the wiphy fails
[ Upstream commit a41bd1938a9bfe226d444172a7e20e4bd5097960 ]
Once all the interfaces have been moved, cfg80211_switch_netns()
moves the wiphy itself by setting its network namespace and then
renaming it, which makes sysfs move it. The rename can fail (but
only on allocation failures), leaving things mixed up and hitting
the warning there.
Ignoring it isn't great, undo the move and let the change fail
in this case. If undo fails then WARN, then things would again
be stuck in two different network namespaces.
Assisted-by: LLM
Reported-by: syzbot+3515319a302224e081b4@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=3515319a302224e081b4
Fixes: 463d018323851 ("cfg80211: make aware of net namespaces")
Link: https://patch.msgid.link/20260904170220.7966cc705e33.Ib398351113bbd3cab85302467060cab378564421@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Stanislaw Gruszka <stf_xl@wp.pl>
Date: Thu Aug 20 11:30:59 2026 +0200
wifi: iwlegacy: fix broadcast stations deallocation
commit b5526b780f8b297a76030410b96ba29153afb98f upstream.
On the error path of __il4965_up(), il_dealloc_bcast_stations() clears
only IL_STA_UCODE_ACTIVE, leaving IL_STA_BCAST set. This causes the
same broadcast stations to be deallocated again by __il4965_down().
This can occur when RF_KILL is toggled during driver startup.
To fix clear the entire 'used' field, since we will not do any
other operations on the station.
Reported-and-tested-by: Martin-Éric Racine <martin-eric.racine+kernel-bugzilla@iki.fi>
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=221733
Fixes: c2fd34469d16 ("iwl4965: Fix a memory leak in error handling code of __il4965_up")
Cc: <stable@vger.kernel.org> # 7.1.x: 57aa1718d595 wifi: iwlegacy: replace BUG_ON() with WARN_ON() on num_stations check
Cc: <stable@vger.kernel.org> # 6.x.x: 57aa1718d595 wifi: iwlegacy: replace BUG_ON() with WARN_ON() on num_stations check
Cc: <stable@vger.kernel.org> # 5.x.x: 57aa1718d595 wifi: iwlegacy: replace BUG_ON() with WARN_ON() on num_stations check
Signed-off-by: Stanislaw Gruszka <stf_xl@wp.pl>
Link: https://patch.msgid.link/20260820093059.18779-1-stf_xl@wp.pl
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Jiangshan Yi <yijiangshan@kylinos.cn>
Date: Sat Aug 15 19:57:24 2026 +0800
wifi: libertas_tf: fix UAF in lbtf_free_adapter()
commit bbb9a0ab96d44a64529aafc7a16de460a1712f6a upstream.
lbtf_free_adapter() calls lbtf_free_cmd_buffer() to free the command
buffers before calling timer_delete_sync() to wait for the command
timer callback. If the timer callback (command_timer_fn) is already
running when lbtf_free_cmd_buffer() frees the command array, the
callback dereferences priv->cur_cmd->cmdbuf which points to freed
memory.
Swap the order so that timer_delete_sync() runs first, ensuring any
in-flight callback has completed before the command buffers are freed.
Fixes: 06b16ae53192 ("libertas_tf: main.c, data paths and mac80211 handlers")
Cc: stable@vger.kernel.org
Signed-off-by: Jiangshan Yi <yijiangshan@kylinos.cn>
Link: https://patch.msgid.link/20260815115724.920628-1-yijiangshan@kylinos.cn
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Daehyeon Ko <4ncienth@gmail.com>
Date: Wed Sep 9 15:11:24 2026 +0900
wifi: libipw: reject TKIP frames without a full MIC
commit 06f42accaf3c6aecab1dcc57f68dde6c06c8b380 upstream.
libipw_michael_mic_verify() assumes that an skb contains an eight-byte
Michael MIC. A short TKIP frame makes the unsigned payload length wrap,
causing michael_mic() to read past the skb.
Check that the MIC is present before verifying it, and use the existing
MICHAEL_MIC_LEN constant for all MIC lengths in the verifier.
Fixes: b453872c35cf ("[NET] ieee80211 subsystem")
Cc: stable@vger.kernel.org
Assisted-by: LLM
Signed-off-by: Daehyeon Ko <4ncienth@gmail.com>
Link: https://patch.msgid.link/20260909061124.3802517-1-4ncienth@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shmulik Cohen <anuk909@gmail.com>
Date: Wed Aug 12 22:04:11 2026 +0300
wifi: libipw: reject too-short association responses
[ Upstream commit adb7118b7d2cfd7e8213c17d7d2829f353017754 ]
libipw_handle_assoc_resp() reads the capability, status and aid fields
of the 30-byte association response prefix and then computes the
information element length as
stats->len - sizeof(*frame)
stats->len is a u16 and sizeof() has type size_t, so the subtraction is
evaluated as size_t and wraps instead of going negative. Truncating
that to the u16 length parameter of libipw_parse_info_param() turns a
frame shorter than the fixed fields into a length near 64 KiB, and the
parser then reads past the receive buffer.
Both the ipw2100 and ipw2200 management receive paths reach this
function having established only that the frame carries the generic
24-byte three-address header.
Reject the frame before any fixed field is touched.
Found by an AI-assisted review of length arithmetic in management frame
parsers. Verified with a KUnit case under Generic KASAN on arm64 under
QEMU; I do not have the hardware, so it is not tested on a real device.
Fixes: 9e8571affd1c ("[PATCH] ieee80211: Add QoS (WME) support to the ieee80211 subsystem")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Shmulik Cohen <anuk909@gmail.com>
Link: https://patch.msgid.link/20260812190412.18333-3-anuk909@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Shmulik Cohen <anuk909@gmail.com>
Date: Wed Aug 12 22:04:10 2026 +0300
wifi: libipw: reject too-short beacon and probe responses
[ Upstream commit 5ce5721e8cbe3e80db8f43851cc2a2a92485ef4b ]
libipw_process_probe_response() and the libipw_network_init() call it
makes assume the frame contains the full 36-byte beacon and probe
response prefix, but the ipw2100 and ipw2200 receive paths only
establish that a management frame carries the generic 24-byte
three-address header.
libipw_network_init() then computes the information element length as
stats->len - sizeof(*beacon)
stats->len is a u16 and sizeof() has type size_t, so the subtraction is
evaluated as size_t and wraps instead of going negative. Truncating
that to the u16 length parameter of libipw_parse_info_param() yields
65524 for a 24-byte beacon, and the parser then walks the receive
buffer as if it held almost 64 KiB of information elements, reading
past the allocation.
Reject the frame before any fixed field is touched.
Found by an AI-assisted review of length arithmetic in management frame
parsers. Verified with a KUnit case under Generic KASAN on arm64 under
QEMU; I do not have the hardware, so it is not tested on a real device.
Fixes: b453872c35cf ("[NET] ieee80211 subsystem")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Shmulik Cohen <anuk909@gmail.com>
Link: https://patch.msgid.link/20260812190412.18333-2-anuk909@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:57:11 2026 +0200
wifi: mac80211: abort chanswitch when leaving a mesh
[ Upstream commit ac7472a24bd433b81c06582835dd1d5547c10da9 ]
The code in ieee80211_stop_mesh() leaves CSA active, but leaving
the mesh released the channel context, so the CSA finalize work
crashes:
Oops: general protection fault, probably for non-canonical address
0xdffffc0000000003
KASAN: null-ptr-deref in range [0x0000000000000018-0x000000000000001f]
RIP: 0010:ieee80211_put_srates_elem+0x42/0x640 net/mac80211/util.c:3272
Call Trace:
ieee80211_mesh_build_beacon+0xa83/0x1b50 net/mac80211/mesh.c:1093
ieee80211_mesh_rebuild_beacon+0xc7/0x170 net/mac80211/mesh.c:1147
ieee80211_mesh_finish_csa+0x131/0x210 net/mac80211/mesh.c:1542
ieee80211_set_after_csa_beacon net/mac80211/cfg.c:4085 [inline]
__ieee80211_csa_finalize net/mac80211/cfg.c:4133 [inline]
ieee80211_csa_finalize+0x633/0x1150 net/mac80211/cfg.c:4155
cfg80211_wiphy_work+0x2ab/0x450 net/wireless/core.c:438
Abort the channel switch properly.
Assisted-by: LLM
Fixes: b8456a14e9d2 ("{nl,cfg,mac}80211: implement mesh channel switch userspace API")
Reported-by: syzbot+81cd9dc1596563141d19@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=81cd9dc1596563141d19
Link: https://patch.msgid.link/20260904165722.d0b87eee08aa.I80550d6127e0bb26efb49a5fbe95be1aef1cd0cb@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Tue Sep 8 14:28:18 2026 +0200
wifi: mac80211: add HE 6 GHz capability in the scan elems len
[ Upstream commit cd54bf333f5631d3630bab0a832e9ae648f73515 ]
The HE 6 GHz Band Capability element is in the probe request for
every band if 6 GHz is supported, so add the size to scan_ies_len.
Otherwise, building probe request elements can fail, triggering the
WARN_ON in __ieee80211_start_scan().
Assisted-by: LLM
Fixes: 2ad2274c58ee ("mac80211: Add HE 6GHz capabilities element to probe request")
Reported-by: syzbot+f961b9f94edbc266f1f8@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=f961b9f94edbc266f1f8
Link: https://patch.msgid.link/20260908122838.201719-19-johannes@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Ivan Pustogarov <ivan@ipust.net>
Date: Thu Sep 3 17:26:16 2026 +0200
wifi: mac80211: avoid out-of-bounds read for empty PREQ elements
commit e6031f02269c0f51cf67886d177f02bc300b47cd upstream.
ieee80211_mesh_preq_size_ok() derives the location of the PREQ bottom
fields before checking whether the element contains even the fixed
header. ieee80211_mesh_hwmp_preq_get_bottom() reads the flags byte to
account for the optional Address Extension field. Consequently, an
empty PREQ element causes a one-byte read beyond its declared payload.
Move the helper call after both size checks, so the bottom fields are
only accessed when they are present.
Fixes: 8b40b1d24a60 ("wifi: mac80211: Fix overread in PREQ frame processing")
Cc: stable@vger.kernel.org
Signed-off-by: Ivan Pustogarov <ivan@ipust.net>
Link: https://patch.msgid.link/20260903152616.1646637-1-ivan@ipust.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Rik van Riel <riel@surriel.com>
Date: Sat Aug 8 10:47:55 2026 -0400
wifi: mac80211: avoid WARN in set_bitrate_mask when sdata not in driver
commit da2ca406f45a6e21760243152ed8d2e8e72915c2 upstream.
ieee80211_set_bitrate_mask() checks if the interface is running via
ieee80211_sdata_running(), but it does not check if the interface is
still present in the driver.
When sdata is running but IEEE80211_SDATA_IN_DRIVER is not set, the
call reaches drv_set_bitrate_mask() in driver-ops.h which hits
wlan1: Failed check-sdata-in-driver check, flags: 0x0
WARNING: net/mac80211/driver-ops.h:884 at drv_set_bitrate_mask
Syzkaller triggers this via wext SIOCSIWRATE ioctl. The Call Trace shows
wext_ioctl_dispatch() in wext-core.c dispatching the ioctl, calling
ioctl_standard_call() for SIOCSIWRATE, which calls cfg80211_wext_siwrate()
in wext-compat.c. That builds a bitrate mask and calls
rdev_set_bitrate_mask() which ends up in ieee80211_set_bitrate_mask() in
cfg.c. The interface is marked running via SDATA_STATE_RUNNING but
flags is 0, so check_sdata_in_driver() fails.
When the interface is being torn down, or when wext ioctl is issued
during interface bringup before drv_add_interface() sets IN_DRIVER, the
running check passes while IN_DRIVER is clear.
Check IEEE80211_SDATA_IN_DRIVER in ieee80211_set_bitrate_mask() before
calling the driver, returning -ENETDOWN. This avoids the WARN_ONCE in
driver-ops.h and matches other cfg.c operations that bail early when not
in driver.
This change should be safe because wiphy mutex is held in
cfg80211_wext_siwrate() via guard(wiphy), and IN_DRIVER is set/cleared
under RTNL and wiphy paths in drv_add_interface() and
drv_remove_interface() in driver-ops.c, so the check is race-free
against driver add/remove. Returning -ENETDOWN is the same error other
not-running paths use and does not introduce new locking.
Reported-by: syzbot+af177aa139efdd13a9da@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=af177aa139efdd13a9da
Link: https://lore.kernel.org/all/6a75205c.59b6c763.2bba34.00c3.GAE@google.com/
Fixes: 554a43d5e77e ("mac80211: check sdata_running on ieee80211_set_bitrate_mask")
Cc: stable@vger.kernel.org
Assisted-by: Hermes:muse-spark-1.2 syzkaller
Signed-off-by: Rik van Riel <riel@surriel.com>
Link: https://patch.msgid.link/20260808104755.319c686e@fangorn
Reported-by: syzbot+dcaca020ca8377e7ced0@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=dcaca020ca8377e7ced0
[also add second syzbot report]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Tue Sep 8 14:28:17 2026 +0200
wifi: mac80211: don't access the TSF of a down interface
[ Upstream commit 0b1de9feeb8651f7a3bb53ed7c9006e3b5298c01 ]
The tsf debugfs files call the driver even if the interface
isn't up, tgriggering check-sdata-in-driver warnings.
Reject the access in that case.
Assisted-by: LLM
Fixes: 37a41b4affa3 ("mac80211: add ieee80211_vif param to tsf functions")
Reported-by: syzbot+1c8c45017f784e646b47@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=1c8c45017f784e646b47
Link: https://patch.msgid.link/20260908122838.201719-18-johannes@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Tue Sep 8 14:28:12 2026 +0200
wifi: mac80211: don't allow injecting frames wider than the chanctx
[ Upstream commit e14bf37bb2b3853012ff160131d1c6233f7a9cc9 ]
Frames injected on a monitor interface can carry a radiotap
field requesting a bandwidth, which mac80211 passes down to
the driver regardless of the the actual operational bandwidth.
If the bandwidth requested is too wide, that triggers a warning
in hwsim:
WARN_ON(hwsim_get_chanwidth(bw) > hwsim_get_chanwidth(confbw))
Drop such frames entirely instead since they cannot be sent.
Assisted-by: LLM
Fixes: 646e76bb5daf ("mac80211: parse VHT info in injected frames")
Reported-by: syzbot+435fdb053cf98bfa5778@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=435fdb053cf98bfa5778
Link: https://patch.msgid.link/20260908122838.201719-13-johannes@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Tue Sep 8 14:28:15 2026 +0200
wifi: mac80211: don't allow link changes when iface is down
[ Upstream commit 370872d30349d81dec519e15ea2949fd63511cf7 ]
ieee80211_set_active_links() only checks that the interface is running in
the inner __ieee80211_set_active_links(), after drv_can_activate_links()
was already called, so using active_links on an interface that's down
triggers the check-sdata-in-driver warning.
Add the missing check in the debugfs file.
Assisted-by: LLM
Fixes: 3d9011029227 ("wifi: mac80211: implement link switching")
Reported-by: syzbot+582469b3a9ef5f13606b@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=582469b3a9ef5f13606b
Link: https://patch.msgid.link/20260908122838.201719-16-johannes@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:57:09 2026 +0200
wifi: mac80211: don't offload TC setup on AP_VLAN interfaces
[ Upstream commit 362bd5bce29ed0f6fd3d39a7065567777d70606e ]
AP_VLAN interfaces are purely virtual, so don't try to offload
TC setup to drivers. We can't really use the AP interface either
since we may not know it all the time, and it could technically
even change.
Just reject the TC offload so things get done in software.
Assisted-by: LLM
Fixes: 61587f1556fe ("wifi: mac80211: add support for letting drivers register tc offload support")
Reported-by: syzbot+f1ba58d6b55abd13239e@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=f1ba58d6b55abd13239e
Link: https://patch.msgid.link/20260904165722.726cc076cecb.Iccfd88b13635425e850ce031376eb60a4ce5f4f8@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Tue Sep 8 14:28:16 2026 +0200
wifi: mac80211: don't RCU-dereference the mesh CSA settings we just set
[ Upstream commit b481e64e4498e2c053d5954f546ee02338f6ab63 ]
In the error path of ieee80211_mesh_csa_beacon() the settings that were
just assigned are read back with rcu_dereference(), which lockdep then
complains about.
There's no need to read the pointer at all, tmp_csa_settings still is
the right value anyway.
Assisted-by: LLM
Fixes: b8456a14e9d2 ("{nl,cfg,mac}80211: implement mesh channel switch userspace API")
Reported-by: syzbot+b59873f5699e941717ca@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=b59873f5699e941717ca
Link: https://patch.msgid.link/20260908122838.201719-17-johannes@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:57:07 2026 +0200
wifi: mac80211: don't start a ROC while scanning
[ Upstream commit 733f0fde95392ed5f61a4e36aee661ea8d0e8581 ]
The ROC work can be pending when a scan starts (which requires
ROC list to be empty, but that's possible), and then a new ROC
can be added to the list and the work will pick it up.
Avoid starting that ROC if a scan made it between things, as
otherwise we'll hit a warning later:
WARNING: net/mac80211/offchannel.c:404 at ieee80211_start_next_roc+0x256/0x2d0
Workqueue: events_unbound cfg80211_wiphy_work
Call Trace:
__ieee80211_scan_completed+0x4fd/0xe40 net/mac80211/scan.c:537
ieee80211_scan_work+0x472/0x1ff0 net/mac80211/scan.c:1193
cfg80211_wiphy_work+0x410/0x570 net/wireless/core.c:513
Assisted-by: LLM
Fixes: aaa016ccd5df ("mac80211: rewrite remain-on-channel logic")
Reported-by: syzbot+c3a167b5615df4ccd7fb@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=c3a167b5615df4ccd7fb
Link: https://patch.msgid.link/20260904165722.f9d5b150edd8.I61bc9de8c8d089096ad695213b9c85c7df38c3bd@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:57:08 2026 +0200
wifi: mac80211: don't warn when an IBSS has no channel to scan
[ Upstream commit a7491b7efbd9136b120a12ed72af9c12121dd134 ]
ieee80211_request_ibss_scan() warns when regulatory leaves no
allowed channel, but that can happen as the regdomain can change
while IBSS is operating, and it can continue to operate briefly
during the 60s grace period until it's shut down.
Just remove the warning in this case.
Assisted-by: LLM
Fixes: 34bcf7150241 ("mac80211: fix ibss scanning")
Reported-by: syzbot+1634c5399e29d8b66789@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=1634c5399e29d8b66789
Link: https://patch.msgid.link/20260904165722.fe380c27fef4.I0e8bee2e12a40d240851a4bc724d47753af46159@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Lachlan Hodges <lachlan.hodges@morsemicro.com>
Date: Thu Aug 27 15:43:02 2026 +1000
wifi: mac80211: include TIM bitmap control for buffered S1G mcast traffic
[ Upstream commit af72b5946d493cecced27d0951ea37c1d178601e ]
Currently when building the S1G TIM element, we only build the bitmap
control if we have buffered unicast traffic. Since AID 0 sits within
the bitmap control if we have buffered multicast traffic with no
buffered unicast traffic the bitmap control won't be emitted and
dozing stations will be unaware of buffered multicast.
To fix, only exclude the bitmap control byte when we don't have
both buffered unicast and multicast traffic.
Fixes: ee6360945483 ("wifi: mac80211: support block bitmap S1G TIM encoding")
Signed-off-by: Lachlan Hodges <lachlan.hodges@morsemicro.com>
Link: https://patch.msgid.link/20260827054302.254124-1-lachlan.hodges@morsemicro.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Tue Sep 8 14:28:20 2026 +0200
wifi: mac80211: mesh: release the channel if start fails
[ Upstream commit ae97fff6495a8764bc0ef281cfe5444f701e527f ]
ieee80211_join_mesh() acquires a channel context and then calls
ieee80211_start_mesh(), which can fail. In that case, the chanctx
isn't released then interface removal will attempt to unassign it
after it's removed from the driver, hitting:
wlan0: Failed check-sdata-in-driver check, flags: 0x0
WARNING: net/mac80211/driver-ops.c:366 at drv_unassign_vif_chanctx
ieee80211_assign_link_chanctx
__ieee80211_link_release_channel
ieee80211_link_release_channel
ieee80211_teardown_sdata
unregister_netdevice_many_notify
_cfg80211_unregister_wdev
ieee80211_remove_interfaces
ieee80211_unregister_hw
mac80211_hwsim_del_radio
hwsim_exit_net
Correctly release the channel on start failures.
Assisted-by: LLM
Reported-by: syzbot+63a84ea9c0f57d6133fa@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=63a84ea9c0f57d6133fa
Fixes: 2b5e19677592 ("mac80211: cache mesh beacon")
Link: https://patch.msgid.link/20260908122838.201719-21-johannes@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Tue Sep 8 14:28:19 2026 +0200
wifi: mac80211: mesh: reset the CSA state when leaving
[ Upstream commit 860134b3af77970e006feab7e5decb8c84771c7f ]
ifmsh->csa is allocated in ieee80211_mesh_csa_beacon() and only freed
in ieee80211_mesh_finish_csa(), i.e. when the channel switch completes.
Leaving the mesh while a switch is still pending therefore leaks it.
Additionally, ifmsh->csa_role and ifmsh->chsw_ttl have their state leak
in this case, so things can get mixed up in addition to the memory
leak.
Refactor the reset and call it in ieee80211_stop_mesh() to fix it all.
Assisted-by: LLM
Reported-by: syzbot+f5752cd6b94fe38be666@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=f5752cd6b94fe38be666
Fixes: b8456a14e9d2 ("{nl,cfg,mac}80211: implement mesh channel switch userspace API")
Link: https://patch.msgid.link/20260908122838.201719-20-johannes@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:57:14 2026 +0200
wifi: mac80211: only operate on TDLS peers in the TDLS code
[ Upstream commit 6f0a100df8539ce90f37c14e1945f396ca2410bc ]
ieee80211_tdls_oper() can operate on the AP station, which then
yields various warnings when the AP station is removed then or
at a later point in time after being confused for a TDLS peer.
Always check that the station is a TDLS peer.
Assisted-by: LLM
Fixes: dfe018bf9953 ("mac80211: handle TDLS high-level commands and frames")
Fixes: 17e6a59a365a ("mac80211: cleanup TDLS state during failed setup")
Reported-by: syzbot+a59b5291776979816910@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=a59b5291776979816910
Link: https://patch.msgid.link/20260904165722.3bad8b79679b.I99618745e83cbe9b9804179387be15fcd3505ae3@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Devin Wittmayer <lucid_duck@justthetip.ca>
Date: Fri Sep 4 13:03:38 2026 -0700
wifi: mac80211: refuse to make a monitor active when it has no queue
commit 2b04d6556964ae9f89819b86a0a7801e39c3aae5 upstream.
A monitor interface only gets a TXQ if it's created active, and one can't
be added later. Setting the flag on a down interface is still allowed, so
the driver is handed a monitor with no queue. ath9k dereferences it:
BUG: kernel NULL pointer dereference, address: 0000000000000066
RIP: 0010:ath_tx_node_init+0x49/0x170 [ath9k]
ath9k_add_interface+0x10c/0x140 [ath9k]
drv_add_interface+0x54/0x250 [mac80211]
ieee80211_do_open+0x32f/0x800 [mac80211]
Reached with CAP_NET_ADMIN by "iw dev X set monitor active" followed by
"ip link set X up". RTNL is held, so netlink operations block behind it.
Refuse the flag when there is no queue to give.
Fixes: 79af1f866193 ("mac80211: avoid allocating TXQs that won't be used")
Cc: stable@vger.kernel.org
Signed-off-by: Devin Wittmayer <lucid_duck@justthetip.ca>
Link: https://patch.msgid.link/20260904200338.10829-1-lucid_duck@justthetip.ca
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Tue Sep 8 14:28:14 2026 +0200
wifi: mac80211: require a peer station for TDLS setup confirm
[ Upstream commit 038e1d126304fd25d507fd4e671232df57bd1799 ]
It's nonsense for the setup confirm to go to station that
doesn't even exist, and it hits a warning when building
the frame:
WARN_ON_ONCE(!sta || !ap_sta)
Only accept WLAN_TDLS_SETUP_CONFIRM when the station is
already there as a TDLS station. Need to copy the call
to ieee80211_tdls_prep_mgmt_packet() since the existing
WLAN_TDLS_DISCOVERY_REQUEST already falls through to it.
Assisted-by: LLM
Fixes: 6f7eaa47e1de ("mac80211: add TDLS QoS param IE on setup-confirm")
Reported-by: syzbot+e55106f8389651870be0@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=e55106f8389651870be0
Link: https://patch.msgid.link/20260908122838.201719-15-johannes@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:57:12 2026 +0200
wifi: mac80211: reset state when starting AP fails
[ Upstream commit 3f28551d0241254a75626d868041c6340285088b ]
ieee80211_start_ap() can set enable_beacon (and beacon_int) and fail
later, leaving it set forever. Scanning can then attempt to restore
beaconing on such an interface, leading to:
Oops: divide error: 0000 [#1] SMP KASAN NOPTI
RIP: 0010:mac80211_hwsim_link_info_changed+0xca7/0xf00
Call Trace:
drv_link_info_changed+0x413/0x860 net/mac80211/driver-ops.c:495
ieee80211_link_info_change_notify+0x24b/0x3c0 net/mac80211/main.c:427
ieee80211_offchannel_return+0x381/0x580 net/mac80211/offchannel.c:160
__ieee80211_scan_completed+0x993/0xe30 net/mac80211/scan.c:519
ieee80211_scan_work+0x472/0x2010 net/mac80211/scan.c:1193
cfg80211_wiphy_work+0x2b7/0x550 net/wireless/core.c:538
in hwsim. Also, cfg80211 then allows changing the interface type,
and the off-channel path getgs confused about beaconing as well,
leading to another warning:
WARNING: net/mac80211/driver-ops.c:468 at drv_link_info_changed+0x583/0x880
ieee80211_link_info_change_notify+0x24b/0x3c0 net/mac80211/main.c:427
ieee80211_offchannel_stop_vifs+0x328/0x5c0 net/mac80211/offchannel.c:122
ieee80211_start_sw_scan net/mac80211/scan.c:583 [inline]
__ieee80211_start_scan+0xfb6/0x1af0 net/mac80211/scan.c:882
Reset the state on failures to always have it correct.
Assisted-by: LLM
Fixes: d6a83228823f ("mac80211: track enable_beacon explicitly")
Reported-by: syzbot+ca7a2759caaa6cd4e3db@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=ca7a2759caaa6cd4e3db
Reported-by: syzbot+c4686c3eb8b64032618f@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=c4686c3eb8b64032618f
Link: https://patch.msgid.link/20260904165722.9629429a5221.I7f599412bfe12a09d41ea4901be9ad165d07d133@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Tue Sep 8 14:28:13 2026 +0200
wifi: mac80211: reset the AP_VLAN tailroom counter on ifdown
[ Upstream commit 4504f3960dc4501c73be9f99eabda2e26e9db41e ]
On ifup, AP_VLAN interfaces get crypto_tx_tailroom_needed_cnt from
the AP interface, but it's never decremented again unless the AP is
also brought down. Thus, bringing the same AP_VLAN up again will
increment the counter again and eventually hit the sanity check:
WARN_ON_ONCE(sdata->crypto_tx_tailroom_needed_cnt !=
master->crypto_tx_tailroom_needed_cnt);
Reset it on ifdown to avoid that.
Assisted-by: LLM
Fixes: f9dca80b98ca ("mac80211: fix AP_VLAN crypto tailroom calculation")
Reported-by: syzbot+de3ee5362db09487ea37@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=de3ee5362db09487ea37
Link: https://patch.msgid.link/20260908122838.201719-14-johannes@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:57:13 2026 +0200
wifi: mac80211: reset the LED state when ifup fails
[ Upstream commit 78183e8331958fda11cd2b6850bb424a9747c4b2 ]
When the first interface comes up, the radio LED is turned on. This can
start the TPT trigger timer, which continues running.
But if bringing up the interface fails then the timer keeps running and
won't be stopped by anything, eventually it can be freed:
ODEBUG: free active (active state 0) object: ffff888127e12130 object type: timer_list hint: tpt_trig_timer+0x0/0x300 net/mac80211/led.c:145
WARNING: CPU: 0 PID: 5923 at lib/debugobjects.c:612 debug_print_object+0x1a2/0x2b0
debug_check_no_obj_freed+0x4b7/0x600 lib/debugobjects.c:1129
kfree+0x436/0x670 mm/slub.c:6818
ieee80211_led_exit+0x162/0x1c0 net/mac80211/led.c:210
ieee80211_unregister_hw+0x27e/0x3a0 net/mac80211/main.c:1706
rt2x00lib_remove_dev+0x55b/0x670
Undo the LED state in the error path.
Assisted-by: LLM
Fixes: 67408c8c7b9d ("mac80211: selective throughput LED trigger active")
Reported-by: syzbot+e84ecca6d1fa09a9b3d9@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=e84ecca6d1fa09a9b3d9
Link: https://patch.msgid.link/20260904165722.044aa432f873.I601a67a2cd558b8ef8416a07554ae7efe896e9d8@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Tue Sep 8 14:28:21 2026 +0200
wifi: mac80211: set up the TX info early to fix failure paths
[ Upstream commit 50d3d79dc0743b616afb00d01a626c76758721f7 ]
The previous commit 2c51457d930f ("wifi: mac80211: free ack status
frame on TX header build failure") cleaned up the leak, but still
left the code a bit messy and the failed SKB didn't get reported
to userspace.
Fix this up by initialising skb->cb[] earlier, which allows using
ieee80211_free_txskb() and therefore reports it for the failure
in ieee80211_build_hdr(), and unifies the ieee80211_skb_resize()
failure path with it.
Assisted-by: LLM
Fixes: c3e7724b6bc2 ("mac80211: use ieee80211_free_txskb to fix possible skb leaks")
Link: https://patch.msgid.link/20260908122838.201719-22-johannes@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 16:57:10 2026 +0200
wifi: mac80211: suppress chanctx warning for debugfs reset
[ Upstream commit bf29d085e0eba92388518719d044f4702a8c6644 ]
Before suspend all the channel contexts should removed, so the
warning makes sense and should be there, but during reset the
same code is called without first removing. Limit the check to
the real suspend case.
Assisted-by: LLM
Fixes: 12e7f517029d ("mac80211: cleanup generic suspend/resume procedures")
Reported-by: syzbot+56a1a45a9a2c04d425ff@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=56a1a45a9a2c04d425ff
Link: https://patch.msgid.link/20260904165722.fe46395e310b.Ic4aaa95bd9d0ceb6a3cd7d84c425afee7d7d3dd7@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 17:02:01 2026 +0200
wifi: mac80211: unlist vifs when their netdev is unregistered
[ Upstream commit eee2efd82867b623982ac51925b5a1812a74c50d ]
mac80211 only removes vifs from the local->interfaces list when
an interface is removed via ieee80211_if_remove(), before it
unregisters the netdev. However, it's possible for a netdev to
be unregistered without going through that: When the netns that
holds the wiphy is destroyed, the wiphy is supposed to move to
the init_ns, but that can run into allocation failures.
Then, mac80211 has an interface listed that doesn't exist, and
will eventually hit
BUG: failure at net/wireless/core.h:141/wiphy_to_rdev()!
...
_cfg80211_unregister_wdev+0x24/0x36a [cfg80211]
cfg80211_unregister_wdev+0x15/0x1d [cfg80211]
ieee80211_remove_interfaces+0x1ff/0x257 [mac80211]
ieee80211_unregister_hw+0x73/0x1d1 [mac80211]
mac80211_hwsim_del_radio+0x114/0x166 [mac80211_hwsim]
Remove the interface from the list in ->ndo_uninit if it's still
around to avoid this.
Assisted-by: LLM
Fixes: 463d018323851 ("cfg80211: make aware of net namespaces")
Link: https://patch.msgid.link/20260904170220.038ad73e6c04.I990abca78483e058746b6f42b4796717c3028164@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Johannes Berg <johannes.berg@intel.com>
Date: Fri Sep 4 17:01:35 2026 +0200
wifi: mac80211_hwsim: don't hand frames to mac80211 while stopping
[ Upstream commit 87840d4a3a21b1c19b867a80e16ba69dff284de2 ]
The code checks ->started for frames coming from wmediumd, but the
radio can be stopped after the check and before frame delivery,
causing mac80211 to hit the WARN_ON(!local->started).
Expand the mutex for this case and synchronise against it when the
radio is stopped to avoid being able to hit the warning with hwsim.
Drop the error print that would've complicated the error path, it
only triggers for allocation failures (already noisy) and malformed
frames anyway.
Assisted-by: LLM
Fixes: 7882513bacb1 ("mac80211_hwsim driver support userspace frame tx/rx")
Reported-by: syzbot+b4aa2b672b18f1d4dc5f@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=b4aa2b672b18f1d4dc5f
Link: https://patch.msgid.link/20260904170140.5f69a10d606b.I4a7921d00643f69e439c7a3b221d104f66a3dcdc@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Doruk Tan Ozturk <doruk@0sec.ai>
Date: Fri Aug 14 15:47:04 2026 +0200
wifi: mwifiex: bound the pairwise-cipher OUI walk to the IE length
commit e667aee1c192d67d27c803007bfa9c6e0873e959 upstream.
mwifiex_search_oui_in_ie() reads a pairwise-cipher (PTK) count from a
beacon/probe-response RSN or WPA information element and then walks that
many 4-byte OUIs, comparing each with memcmp(). The count comes straight
from the (attacker-supplied) IE and is never checked against the
element's own length, and the callers admit the element on element_id
alone (has_ieee_hdr() / has_vendor_hdr(), no length check). A crafted
RSN/WPA IE with a large pairwise count therefore makes the walk read up
to 255 * 4 bytes past the element -- an out-of-bounds read of the
kmemdup()'d beacon buffer, reachable from any AP whose beacon/probe
response is processed during scan-result parsing.
Pass the number of IE bytes available at the OUI list and bound the walk
to the element. Keep the length signed and reject a negative value
before any unsigned arithmetic, so a small or zero IE length cannot
underflow to a large size_t and defeat the bound.
Found by 0sec automated security-research tooling (https://0sec.ai).
Fixes: 5e6e3a92b9a4 ("wireless: mwifiex: initial commit for Marvell mwifiex driver")
Cc: stable@vger.kernel.org
Assisted-by: 0sec:multi-model
Signed-off-by: Doruk Tan Ozturk <doruk@0sec.ai>
Link: https://patch.msgid.link/20260814134704.85902-1-doruk@0sec.ai
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Peng Hao <flyingpenghao@gmail.com>
Date: Fri Aug 28 19:15:31 2026 +0800
wifi: mwifiex: fix IRQ leak using wrong index in MSI-X error path
[ Upstream commit a3d722190cdef18da4878b5efc27c3c386dda248 ]
mwifiex_pcie_request_irq() registers each MSI-X vector with a per-index
dev_id (&card->msix_ctx[i]). On a request_irq() failure the cleanup loop
"for (j = 0; j < i; j++)" frees msix_entries[j].vector but passes the
failed index's &card->msix_ctx[i] as the dev_id. free_irq() matches on
(irq, dev_id), so it fails to find the action registered with
&card->msix_ctx[j]: the already-requested IRQ j is not freed (leaked) and
free_irq() warns about freeing a non-existent IRQ. Use &card->msix_ctx[j].
Fixes: 99074fc1e67b ("mwifiex: enable pcie MSIx interrupt mode support")
Signed-off-by: Peng Hao <flyingpeng@tencent.com>
Link: https://patch.msgid.link/20260828111531.56723-1-flyingpeng@tencent.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Linmao Li <lilinmao@kylinos.cn>
Date: Thu Aug 20 14:21:55 2026 +0800
wifi: mwifiex: prevent authentication frame length truncation
commit fa00193eb991f92b007aefe7afb6a7566976dacf upstream.
mwifiex_cfg80211_authenticate() derives the authentication frame length
from req->ie_len and req->auth_data_len, both of type size_t, but stores
it in a u16.
NL80211_ATTR_AUTH_DATA only has a minimum length policy. Since nla_len is
a u16, a single attribute can carry up to 65531 bytes of payload, so the
sum can exceed U16_MAX before it is assigned to pkt_len. The truncated
pkt_len determines the skb frame area, while the copy length remains
req->auth_data_len - 4, resulting in a heap buffer overflow.
For example, with auth_data_len equal to 65510 and no IEs, the sum is
65546. It is truncated to 10 and then reduced by four to 6. The driver
appends only six bytes to the skb with skb_put(), but then copies 65506
user-provided bytes into the authentication body.
Reaching this path requires CAP_NET_ADMIN in the user namespace owning
the network namespace, an up station netdev, and a suitable BSS/SAE
authentication request.
Compute the length in size_t, reject values that cannot be represented by
the firmware's u16 frame length field, and only then assign it to pkt_len.
Fixes: 36995892c271 ("wifi: mwifiex: add host mlme for client mode")
Cc: stable@vger.kernel.org # 6.12+
Signed-off-by: Linmao Li <lilinmao@kylinos.cn>
Link: https://patch.msgid.link/20260820062155.3981976-1-lilinmao@kylinos.cn
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Zhao Li <enderaoelyther@gmail.com>
Date: Tue Aug 25 19:25:23 2026 +0800
wifi: mwifiex: validate action frame fixed fields
commit 1c25bfad93e69ce13f744a2fb919f02ea396a985 upstream.
mwifiex_process_mgmt_packet() accepts an rx_pkt_length as small as a
four-address struct ieee80211_hdr plus the two-byte firmware length prefix.
After stripping the prefix, mwifiex_parse_mgmt_packet() can receive a
frame equal to sizeof(struct ieee80211_hdr).
For action frames, the parser reads the category byte immediately after
that header and, for a public action frame, reads the following action
code byte without verifying that either field is present. A truncated frame
can therefore make the parser consume up to two bytes past the
firmware-declared frame length. If those bytes look like a TDLS discovery
response, the malformed frame can spuriously update peer signal state.
Require the category and public action-code fields before reading them.
Use sizeof(*ieee_hdr) so the checks and field accesses directly match the
firmware four-address layout being parsed before address4 is removed.
Suggested-by: Johannes Berg <johannes@sipsolutions.net>
Suggested-by: Brian Norris <briannorris@chromium.org>
Fixes: 72e5aa8d2a6d ("mwifiex: support for parsing TDLS discovery frames")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/all/66f148d83eb9f0970b9abbccc85d1b61244e54ad.camel@sipsolutions.net/
Link: https://lore.kernel.org/all/20260708195911.84365-8-enderaoelyther@gmail.com/
Link: https://lore.kernel.org/all/20260723011013.76968-1-enderaoelyther@gmail.com/
Link: https://lore.kernel.org/all/20260723202257.688-1-enderaoelyther@gmail.com/
Link: https://lore.kernel.org/all/anuWyiPQja6_5vly@google.com/
Assisted-by: Codex:gpt-5
Assisted-by: Kimi:K3
Signed-off-by: Zhao Li <enderaoelyther@gmail.com>
Link: https://patch.msgid.link/20260825112523.95774-1-enderaoelyther@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Pengpeng Hou <pengpeng@iscas.ac.cn>
Date: Sat Aug 15 21:52:27 2026 +0800
wifi: mwifiex: validate scan response extents
commit 3687d7d48070838cc2953431b3a27717cab0aaf6 upstream.
mwifiex_ret_802_11_scan() subtracts the fixed response fields and the
firmware-provided BSS length from resp->size without first proving that
either extent fits. A short response or oversized BSS length can
therefore underflow tlv_buf_size and make the TLV parser walk beyond the
command response.
Compute the fixed extent from the selected normal or background scan
response. Validate that the fixed fields and BSS data fit before deriving
the TLV extent and entering the parser.
Fixes: 5e6e3a92b9a4 ("wireless: mwifiex: initial commit for Marvell mwifiex driver")
Cc: stable@vger.kernel.org
Assisted-by: Codex:gpt-5
Signed-off-by: Pengpeng Hou <pengpeng@iscas.ac.cn>
Link: https://patch.msgid.link/20260815135227.50392-1-pengpeng@iscas.ac.cn
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shengzhuo Wei <me@cherr.cc>
Date: Mon Aug 31 02:42:13 2026 +0800
wifi: p54: require a full exp_if record in PDR_INTERFACE_LIST
commit d8efd84f49379ed28624098821f80e992657d935 upstream.
The PDR_INTERFACE_LIST loop only checks that the record start is within
the entry before reading an entire struct exp_if from it. A truncated
trailing record makes the if_id/variant reads cross the entry boundary
into the heap beyond the EEPROM buffer (verified with a KASAN
reproducer of the loop). The variant also feeds the synth front-end
selection, so this is not only a leak.
Advance only while a full record still fits in the entry.
Fixes: eff1a59c48e3 ("[P54]: add mac80211-based driver for prism54 softmac hardware")
Cc: stable@vger.kernel.org
Acked-by: Christian Lamparter <chunkeey@gmail.com>
Assisted-by: GLM:5.3
Signed-off-by: Shengzhuo Wei <me@cherr.cc>
Link: https://patch.msgid.link/20260831-p54-pda-validation-v2-2-dae566b388c8@cherr.cc
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Shengzhuo Wei <me@cherr.cc>
Date: Mon Aug 31 02:42:12 2026 +0800
wifi: p54: validate curve data length in the calibration curve converters
commit ce858fa6b8a214dee5adb82358885fa024cdd887 upstream.
p54_convert_rev0() and p54_convert_rev1() read calibration curve
data from the device-supplied EEPROM entry using channel and
points-per-channel counts taken verbatim from that same entry, so
an entry that declares more data than it carries drives an
out-of-bounds read past the EEPROM buffer (verified with a KASAN
reproducer of the conversion loop). The sibling converters
p54_convert_output_limits() and p54_convert_db() already validate
their counts against the entry length; this path was missed.
Reject the entry when the counts do not fit in the entry data.
Fixes: eff1a59c48e3 ("[P54]: add mac80211-based driver for prism54 softmac hardware")
Cc: stable@vger.kernel.org
Assisted-by: GLM:5.3
Signed-off-by: Shengzhuo Wei <me@cherr.cc>
Link: https://patch.msgid.link/20260831-p54-pda-validation-v2-1-dae566b388c8@cherr.cc
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Tianchu Chen <flynnnchen@tencent.com>
Date: Fri Sep 4 14:24:45 2026 +0000
wifi: rsi: fix heap OOB write on key removal
commit e6c5ed7a98d7bc8b0f7918246f1c90ddb3f79dfa upstream.
When a key is removed (data == NULL), rsi_hal_load_key() runs:
memset(&set_key[FRAME_DESC_SZ], 0, frame_len - FRAME_DESC_SZ);
set_key is a struct rsi_set_key *, so the subscript is scaled by
sizeof(struct rsi_set_key) (160 bytes): &set_key[FRAME_DESC_SZ] is
skb->data + 2560, and the memset writes 144 zero bytes starting
2.4KB past the end of the 160-byte skb data buffer, corrupting
unrelated heap objects. The intended byte offset would have been
(u8 *)set_key + FRAME_DESC_SZ.
The write fires on every DISABLE_KEY callback, so plain disconnects,
roams and interface teardowns trigger it on real networks.
The memset is redundant: the whole buffer is zeroed right after
allocation, so the frame sent to the device is byte-identical
without it. Drop the else branch; normal operation is unaffected.
Discovered by Atuin - Automated Vulnerability Discovery Engine.
Fixes: dad0d04fa7ba ("rsi: Add RS9113 wireless driver")
Cc: stable@vger.kernel.org
Assisted-by: LLM
Signed-off-by: Tianchu Chen <flynnnchen@tencent.com>
Link: https://patch.msgid.link/90bb2b07007942064c04aa3729cedd9eb1e930b1@linux.dev
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Zihan Xi <zihanx@nebusec.ai>
Date: Wed Sep 9 12:37:18 2026 +0000
wifi: virt_wifi: don't transfer operstate before register
[ Upstream commit e5c8d7acd31b27057ea42cd405d0b3ece097bc89 ]
virt_wifi_newlink() calls netif_stacked_transfer_operstate() before
register_netdevice(). If the lower device is dormant, that queues the
new netdev on lweventlist while it is still uninitialized. If
registration fails after that, for example because of an invalid name
such as "bad/name", free_netdev() immediately frees the object. A
later linkwatch_fire_event() then use-after-frees the list entry.
Move the transfer to after netdev_upper_dev_link(), as macvlan and
ipvlan already do.
Fixes: c7cdba31ed8b ("mac80211-next: rtnetlink wifi simulation device")
Reported-by: Vega <vega@nebusec.ai>
Assisted-by: LLM
Co-developed-by: Luxing Yin <root@tr0jan.top>
Signed-off-by: Luxing Yin <root@tr0jan.top>
Signed-off-by: Zihan Xi <zihanx@nebusec.ai>
Link: https://patch.msgid.link/f5a832fb0ab228ce6e2b5a91fba4ca8b79198a2f.1788948455.git.zihanx@nebusec.ai
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Mariano Baragiola <mbaragiola@linux.com>
Date: Sun Aug 9 09:49:47 2026 -0300
wifi: virt_wifi: free skb when disconnected
[ Upstream commit f9edf7cf63b96d2b776fca8d258d3c5256e40c8e ]
When the simulated link is disconnected, virt_wifi_start_xmit() returns
NET_XMIT_DROP without freeing the skb. dev_hard_start_xmit() treats this
return value as consumed, so every packet sent while disconnected leaks its
skb.
Free the skb before returning the drop status.
Fixes: c7cdba31ed8b ("mac80211-next: rtnetlink wifi simulation device")
Signed-off-by: Mariano Baragiola <mbaragiola@linux.com>
Link: https://patch.msgid.link/20260809124947.3590270-1-mbaragiola@linux.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Fan Wu <fanwu01@zju.edu.cn>
Date: Thu Sep 10 02:09:07 2026 +0000
wifi: wcn36xx: Fix potential use-after-free in TX ack timer teardown
commit d9be5e75530772fc31637070d51e5717d6aeaa2a upstream.
wcn36xx_dxe_deinit() tears down the TX ack timer with timer_delete(),
which only dequeues the timer and does not wait for a callback that is
already executing; the preceding free_irq() calls synchronize the
interrupt handlers only. The callback, wcn36xx_dxe_tx_timer(), can
therefore be running past the teardown and use the wcn freed along
with the ieee80211_hw in wcn36xx_remove(): it takes wcn->dxe_lock,
reads wcn->tx_ack_skb and passes wcn->hw to
ieee80211_tx_status_irqsafe().
Fix this by using timer_shutdown_sync(), which waits for a running
callback and also prevents the timer from being rearmed again. The
timer is set up again by wcn36xx_dxe_init() on the next start, so the
start/stop cycle is unaffected.
This issue was found by an in-house static analysis tool.
Fixes: fdf21cc37149 ("wcn36xx: Add TX ack support")
Cc: stable@vger.kernel.org
Assisted-by: LLM
Co-developed-by: Song Li <songl@zju.edu.cn>
Signed-off-by: Song Li <songl@zju.edu.cn>
Signed-off-by: Fan Wu <fanwu01@zju.edu.cn>
Reviewed-by: Loic Poulain <loic.poulain@oss.qualcomm.com>
Link: https://patch.msgid.link/20260910020907.3353-1-fanwu01@zju.edu.cn
Signed-off-by: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Ali Ahmet Memis <ali@iusegentoo.com>
Date: Fri Aug 7 11:52:30 2026 +0000
wifi: wilc1000: fix out-of-bounds read in P2P public action frames
commit ba6cb7c0868a412c2eb68e8efd5aa38bfb258a14 upstream.
wilc_wfi_p2p_rx() and mgmt_tx() start parsing a frame once
ieee80211_is_public_action() returns true. That helper only verifies the
frame is long enough for the action category field, that is
offsetofend(struct ieee80211_mgmt, u.action.category), 25 bytes. Both
functions then read the P2P public action header up to oui_subtype at
offset 30 and pass "size - ie_offset" to cfg80211_find_vendor_ie(), where
ie_offset is offsetof(struct ieee80211_mgmt, u) + sizeof(*d), i.e. 32.
A public action frame of 25 to 31 bytes passes the check but is shorter
than that 32 byte header, so oui_subtype can be read out of bounds, and
because the length is unsigned, "size - ie_offset" underflows to a value
close to 4 GiB. cfg80211_find_vendor_ie() takes an unsigned int length,
so even the size_t subtraction in mgmt_tx() is truncated to the same
value. It then walks far past the buffer searching for a vendor element
until it reaches unmapped memory.
In the receive path the frame arrives over the air and needs no
association, so a nearby unauthenticated device can crash the host while
it is in P2P listen. Reject frames shorter than the P2P public action
header in both paths before dereferencing it.
Fixes: 4fb8b5aa2a11 ("staging: wilc1000: refactor p2p action frames handling API's")
Cc: stable@vger.kernel.org
Signed-off-by: Ali Ahmet Memis <ali@iusegentoo.com>
Link: https://patch.msgid.link/20260807115230.136767-1-ali@iusegentoo.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Tianchu Chen <flynnnchen@tencent.com>
Date: Fri Sep 4 13:39:34 2026 +0000
wifi: wilc1000: fix RX buffer OOB-write in wilc_wlan_handle_isr_ext()
commit c1ba7f7f18465e259cf1b4d9c73fc73853d7f790 upstream.
wilc_wlan_handle_isr_ext() takes the RX transfer size from the
device-reported interrupt status register (a 15-bit field shifted left by 2,
up to 131068 bytes) and reads that many bytes from the device into
rx_buffer, which is only WILC_RX_BUFF_SIZE (96K) large. The wrap
check only handles the current offset; the size itself is never
compared against the buffer, so a bogus SDIO device can make the driver
OOB-write rx_buffer by up to ~32K with data it controls.
The oversized transfer also leaves rx_buffer_offset past the end of
the buffer, after which the unsigned wrap check stops working and
the overflow can repeat.
Drop any transfer whose size exceeds the RX buffer, acknowledging
the data interrupt and re-arming the RX engine so the bogus frame is
discarded and reception can continue. This also restores the
rx_buffer_offset <= WILC_RX_BUFF_SIZE invariant the wrap check
relies on.
This is not expected to change driver behavior in most cases:
without this check, an oversized transfer would most likely
corrupt neighboring kernel memory instead of completing anyway, and
the drop path performs the same interrupt acknowledgment and RX
engine re-arming as the normal path, so subsequent transfers are
received unaffected.
Discovered by Atuin - Automated Vulnerability Discovery Engine.
Fixes: c5c77ba18ea6 ("staging: wilc1000: Add SDIO/SPI 802.11 driver")
Cc: stable@vger.kernel.org
Assisted-by: LLM
Signed-off-by: Tianchu Chen <flynnnchen@tencent.com>
Link: https://patch.msgid.link/7c971924c6bdccf6c2f75704a5a746e9303aaf64@linux.dev
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Runyu Xiao <runyu.xiao@seu.edu.cn>
Date: Thu Aug 20 20:51:26 2026 +0800
wifi: wlcore: release runtime PM ref on regdomain config failure
commit 8a1f3cf89ddcc700e25afe42cfad333059adcc94 upstream.
wlcore_regdomain_config() gets a runtime PM reference before sending
the regulatory-domain command. When
wlcore_cmd_regdomain_config_locked() fails, the function queues recovery
and returns without dropping that reference.
Release the reference after handling the command result so both success
and failure paths balance the preceding
pm_runtime_resume_and_get(). The recovery worker takes a separate
runtime PM reference and cannot release the reference held here.
Fixes: fa2648a34e73 ("wlcore: Add support for runtime PM")
Cc: stable@vger.kernel.org
Signed-off-by: Runyu Xiao <runyu.xiao@seu.edu.cn>
Link: https://patch.msgid.link/20260820125126.12757-1-runyu.xiao@seu.edu.cn
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Chang S. Bae <chang.seok.bae@intel.com>
Date: Wed Sep 16 23:00:03 2026 +0000
x86/build/64: Prevent native builds from generating EGPR use
commit 63edf5a009ae366369a1b484cd9ae4ee7c51946c upstream.
Omar reports that CONFIG_X86_NATIVE_CPU=y allows builds to opportunistically
emit instructions using %r16-%r31 (EGPRs) when the build host supports APX
since the commit:
ea1dcca1de12 ("x86/kbuild/64: Add the CONFIG_X86_NATIVE_CPU option to locally optimize the kernel with '-march=native'")
But the kernel is not yet prepared to use new registers internally. For
example, there is no context-switch support for general in-kernel use.
Explicitly disable EGPR use when building with -march=native.
For C, since GCC 14 and Clang 18, both compilers support suppressing EGPR
use with -mno-apx-features=egpr, whose availability can be detected via
cc-option.
For Rust, pass features=-apxf through the generated JSON to avoid
unstable-feature warnings, see
https://github.com/rust-lang/rust/issues/139284
Note Rust only accepts the option to disable APX instructions entirely or not.
Support for this gating also depends on the Rust/LLVM combination. Rust
1.88 introduced the `apxf` feature option, but versions prior to 1.93 may
emit an `apxf` attribute to the backend that only LLVM 23 or later can
interpret. Restrict native Rust builds accordingly.
Fixes: ea1dcca1de12 ("x86/kbuild/64: Add the CONFIG_X86_NATIVE_CPU option to locally optimize the kernel with '-march=native'")
Reported-by: Omar Avelar <omar.avelar@intel.com>
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Miguel Ojeda <ojeda@kernel.org>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260916230003.1144622-1-chang.seok.bae@intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: David Laight <david.laight.linux@gmail.com>
Date: Mon Aug 3 10:47:01 2026 +0100
x86/div64: Fix addition of large constants in mul_u64_add_u64_div_u64()
[ Upstream commit f65d38155aef069c897a64643a03db237dd1e0c8 ]
Adding constants over 2^31 fails to compile because the ADD instruction only
supports 32bit signed immediates.
Replace the "irm" constraint with "erm" so that the compiler loads large
constants into a register.
Found by a patch to drivers/iio/frequency/ad9910.c
[ bp: Massage commit message. ]
Fixes: 6480241f31f5 ("lib: add mul_u64_add_u64_div_u64() and mul_u64_u64_div_u64_roundup()")
Signed-off-by: David Laight <david.laight.linux@gmail.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: H. Peter Anvin <hpa@zytor.com>
Link: https://patch.msgid.link/20260803094702.3852-2-david.laight.linux@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Matthew Schwartz <matthew.schwartz@linux.dev>
Date: Thu Sep 17 16:09:06 2026 -0700
x86/fred: Reconstruct the #GP context for rejected INT instructions
[ Upstream commit 93f53499d0b945e8ae447f497faf743d60069f61 ]
FRED event delivery does not use the IDT, so the gate DPL check that
rejects a user INT n falls to software (Intel FRED specification [1],
section 8.3). fred_intx() rejects the same vectors as IDT delivery, but
reports a zero error code and the IP after the INT. This breaks the
signal ABI. Wine uses the error code to recognize INT 0x2d, so the
changed context turns a handled breakpoint into an access violation in
Elden Ring.
Rewind IP using the instruction length in the augmented SS and
synthesize the IDT selector error code, (vector << 3) | 2. Set RF in the
saved flags, as the CPU does for a #GP fault. Section 5.2.1 defines the
saved vector, instruction length and RF state. The supplied length
handles prefixes without reading user memory. Limit the changes to
already-rejected software interrupts, preserving the accepted INT3, INT4
and enabled INT80 paths and hardware exceptions. With IA32 emulation
disabled, INT 0x80 now reports the same #GP as the DPL 0 gate IDT
installs there. The rewound IP also stops fixup_iopl_exception() from
inspecting the byte after the INT.
Also clear the software event flag. Section 6.2.3 specifies that ERETU
with this flag and TF set traps before executing any user instruction. A
tracer that suppresses SIGSEGV and resumes with TF set expects the next
instruction to run first, as after IRET. The sigreturn path clears the
same flag for this reason in prevent_single_step_upon_eretu().
[1] Intel Flexible Return and Event Delivery (FRED) Specification,
revision 9.0 (346446-009US), sections 5.2.1, 6.2.3 and 8.3.
Fixes: 14619d912b65 ("x86/fred: FRED entry/exit and dispatch code")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/work_items/15745
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/work_items/16132
Reported-by: Paul Gofman <pgofman@codeweavers.com>
Signed-off-by: Matthew Schwartz <matthew.schwartz@linux.dev>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: H. Peter Anvin <hpa@zytor.com>
Link: https://cdrdv2.intel.com/v1/dl/getContent/678938 # [1]
Link: https://patch.msgid.link/20260917230907.2080792-2-matthew.schwartz@linux.dev
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Jinke Han <jinkehan@didiglobal.com>
Date: Tue Sep 8 15:37:42 2026 +0800
x86/kprobes: Fix crash when probing CS CALL instructions
[ Upstream commit a5f7a5bb3b7f28ba7e4fa246775b29a0e5537255 ]
When using eBPF to probe CS CALL instructions within a function,
a crash can be triggered.
The eBPF tool probes offset 257 of the __hrtimer_run_queues()
function:
<__hrtimer_run_queues+249>: nopl 0x0(%rax,%rax,1)
<__hrtimer_run_queues+254>: mov %r14,%rdi
<__hrtimer_run_queues+257>: cs call <__x86_indirect_thunk_r12>
<__hrtimer_run_queues+263>: mov %eax,%r12d
<__hrtimer_run_queues+266>: xchg %ax,%ax
<__hrtimer_run_queues+268>: mov %r13,%rdi
Which triggers this crash:
BUG: unable to handle page fault for address: 00000000000f41c9
#PF: supervisor write access in kernel mode
#PF: error_code(0x0002) - not-present page
PGD 0 P4D 0
Oops: 0002 [#1] SMP NOPTI
CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P
RIP: 0010:__hrtimer_run_queues+0x106/0x230
Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is
at the 6th byte of the above CS CALL instruction. Since the CS CALL
instruction occupies 6 bytes, the exception occurred in the middle of that
call instruction.
The root cause is that when using eBPF tools to probe in the middle of a
function, a kprobe with INT3 is used as the underlying implementation.
During single-step emulation of the original CALL instruction,
int3_emulate_call() assumes that the probed CALL instruction is 5 bytes
long. However, the actual CS-prefixed CALL instruction occupies 6 bytes,
so it constructs an incorrect exception return address. When the CPU
returns from the kprobe handler, the next instruction to be executed is at
the address of the last byte of that CS CALL instruction. Coincidentally,
starting from that address, the CPU fetches and decodes a completely
different instruction, which ultimately triggers a kernel crash.
Fix the issue by using the actual instruction length obtained from
the instruction decoder when constructing the exception return
address, rather than relying on the hardcoded CALL_INSN_SIZE macro.
[ mingo: Refined the changelog ]
Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step")
Suggested-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Chang S. Bae <chang.seok.bae@intel.com>
Date: Wed Sep 16 22:59:39 2026 +0000
x86/microcode/intel: Reject problematic loading on Granite Rapids systems
commit e7d3e2f46dd5a69046e6d95a0f189155a5516b93 upstream.
Microcode updates can usually jump revisions. However, there is an erratum on
Granite Rapids systems. If they "jump over" revision 0x1000405, they result in
an #MC. Avoid it.
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260916225939.1144524-1-chang.seok.bae@intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Aleksandr Nogikh <nogikh@google.com>
Date: Fri Jul 31 10:06:20 2026 +0000
xfrm: add missing RCU read lock in xfrm_send_migrate_state()
[ Upstream commit 42d100f5232f39b8ea7b00a7c2482325c7f032a4 ]
xfrm_nlmsg_multicast() requires the RCU read lock to be held because it
safely dereferences the net->xfrm.nlsk pointer using rcu_dereference().
When it is called from xfrm_send_migrate_state(), the RCU read lock is not
held, which triggers a suspicious RCU usage warning:
WARNING: suspicious RCU usage
net/xfrm/xfrm_user.c:1630 suspicious rcu_dereference_check() usage!
Call Trace:
lockdep_rcu_suspicious+0x13f/0x1d0 kernel/locking/lockdep.c:6876
xfrm_nlmsg_multicast+0x1d8/0x1f0 net/xfrm/xfrm_user.c:1630
xfrm_send_migrate_state+0x870/0xae0 net/xfrm/xfrm_user.c:3340
xfrm_do_migrate_state+0x1749/0x1e90 net/xfrm/xfrm_user.c:3507
xfrm_user_rcv_msg+0x7a8/0xf30 net/xfrm/xfrm_user.c:3907
Fix this by wrapping the xfrm_nlmsg_multicast() call in
xfrm_send_migrate_state() with rcu_read_lock() and rcu_read_unlock().
Fixes: a9d155ea9b44 ("xfrm: add XFRM_MSG_MIGRATE_STATE for single SA migration")
Assisted-by: Gemini:gemini-3.5-flash Gemini:gemini-3.1-pro-preview syzbot
Reported-by: syzbot+c0e99a1aa85a286d7a3b@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=c0e99a1aa85a286d7a3b
Link: https://syzkaller.appspot.com/ai_job?id=8977f559-3a7e-4bb5-b4d6-1196956260b6
Signed-off-by: Aleksandr Nogikh <nogikh@google.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Eric Dumazet <edumazet@google.com>
Date: Fri Aug 7 17:15:33 2026 +0000
xfrm: add missing rcu_read_lock(), skb_dst_force() and dev_hold() for xfrm_trans_reinject()
[ Upstream commit d2f5082f9e84653fa1a9e8aebaaff23e688f5e19 ]
syzbot reported a suspicious RCU usage warning in ip6_pkt_drop():
WARNING: suspicious RCU usage in ip6_pkt_drop
include/net/addrconf.h:389 suspicious rcu_dereference_check() usage!
Call Trace:
__in6_dev_get_safely include/net/addrconf.h:389 [inline]
ip6_pkt_drop+0x596/0x610 net/ipv6/route.c:4620
ip6_pkt_discard+0x1c/0x30 net/ipv6/route.c:4651
xfrm_trans_reinject+0x324/0x630 net/xfrm/xfrm_input.c:806
process_one_work kernel/workqueue.c:3322 [inline]
process_scheduled_works+0xa8e/0x14e0 kernel/workqueue.c:3405
worker_thread+0xa47/0xfb0 kernel/workqueue.c:3486
When commit 4f4920669d21 ("xfrm: Reinject transport-mode packets through
workqueue") converted xfrm_trans_reinject from a tasklet to a workqueue,
the reinjection loop ceased running in softirq context. Workqueue workers
run in process context where local_bh_disable() does not enter an RCU
read-side critical section under CONFIG_PREEMPT_RCU.
Because finish callbacks (such as ip6_rcv_finish) expect to run under an
RCU read lock (performing route lookups, l3mdev lookups, and accessing
RCU-protected data structures), invoking them in workqueue context without
rcu_read_lock() triggers RCU lockdep warnings.
Furthermore, packets queued to the workqueue via xfrm_trans_queue_net()
may carry non-refcounted (noref) dst entries (e.g. from ip_route_input_noref).
Additionally, on netdevice unregistration, dst_dev_put() replaces dst->dev
with blackhole_netdev, so dst entries do not keep skb->dev alive while
queued in the workqueue.
Fix these issues by:
1. Calling skb_dst_force(skb) in xfrm_trans_queue_net() while still in the
caller's RCU section to ensure dst is reference-counted before queuing.
2. Holding a reference on skb->dev via dev_hold()/dev_put() across workqueue
deferral so skb->dev remains valid during finish() callback processing.
3. Acquiring rcu_read_lock() around the finish callback invocation loop in
xfrm_trans_reinject().
Fixes: 4f4920669d21 ("xfrm: Reinject transport-mode packets through workqueue")
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Steffen Klassert <steffen.klassert@secunet.com>
Cc: Liu Jian <liujian56@huawei.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Kyle Zeng <kylebot@openai.com>
Date: Tue Aug 4 06:10:37 2026 +0000
xfrm: fix compat ALLOCSPI request use-after-free
[ Upstream commit d1ebd9081879fd9ae9c8fb7e8928f19cc88ae320 ]
xfrm_state_netlink() builds the ALLOCSPI response with
dump_one_state(), which already calls alloc_compat() with the response
skb and header.
xfrm_alloc_userspi() then calls alloc_compat() again, but passes the
original request skb and its header. For a compat request, the
translator therefore interprets the 228-byte compat xfrm_userspi_info
as the 232-byte native layout and reads four bytes past the declared
payload. It also publishes the translated child through the request's
frag_list.
A multicast clone of the request shares skb_shared_info and can observe
that child. xfrm_user_rcv_msg() frees it after the request handler
returns, racing a compat receiver which may still be copying from it and
resulting in a use-after-free.
Remove the redundant conversion. The response keeps its correct compat
translation from dump_one_state(), and no child is attached to the
inbound request.
Fixes: 5f3eea6b7e8f ("xfrm/compat: Attach xfrm dumps to 64=>32 bit translator")
Assisted-by: Codex:gpt-5.6-sol Codex:gpt-5.5-cyber
Signed-off-by: Kyle Zeng <kylebot@openai.com>
Co-developed-by: David Lee <david.lee@trailofbits.com>
Signed-off-by: David Lee <david.lee@trailofbits.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Cen Zhang (Microsoft Security FORGE Labs) <blbllhy@gmail.com>
Date: Wed Aug 26 16:17:45 2026 -0400
xfrm: hold net_device reference under RCU in bundle creation
[ Upstream commit 9fa903b24b1f46b4ff5443bcd4aca23e5c57f9c1 ]
xfrm_bundle_create() and xfrm_create_dummy_bundle() read dst->dev into
a local pointer without taking a device reference, then pass it to
xfrm_fill_dst(). A concurrent RTM_DELLINK replaces dst->dev via
dst_dev_put() and frees the old net_device, causing a use-after-free
when xfrm6_fill_dst() later dereferences the stale dev pointer.
BUG: KASAN: slab-use-after-free in xfrm6_fill_dst+0x82c/0x860
(net/ipv6/xfrm6_policy.c:86 netdev_hold())
Read of size 8 at addr ffff8880142fe588 by task exploit/153
Call Trace:
xfrm6_fill_dst+0x82c/0x860
xfrm_resolve_and_create_bundle+0x21d4/0x2bd0
xfrm_lookup_with_ifid+0x485/0x1640
ip6_dst_lookup_flow+0x19b/0x1e0
udpv6_sendmsg+0x1443/0x2dd0
Fix this by reading dst->dev via dst_dev_rcu() and keeping the RCU
read-side critical section active until xfrm_fill_dst() has taken the
required device references.
Fixes: 25ee3286dcbc ("[IPSEC]: Merge common code into xfrm_bundle_create")
Fixes: a0073fe18e71 ("xfrm: Add a state resolution packet queue")
Suggested-by: Steffen Klassert <steffen.klassert@secunet.com>
Reported-by: Xiang Mei (Microsoft) <xmei5@asu.edu>
Link: https://lore.kernel.org/all/20260820200245.44312-1-blbllhy@gmail.com/
Cc: AutonomousCodeSecurity@microsoft.com
Assisted-by: GitHub-Copilot:claude-opus-4.6
Signed-off-by: Cen Zhang (Microsoft Security FORGE Labs) <blbllhy@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Henry Martin <bsdhenrymartin@gmail.com>
Date: Mon Aug 3 12:01:54 2026 +0800
xfrm: iptfs: fix runt reassembly panic from short inner tot_len
[ Upstream commit dc33262be1fe43d0eb0b84fb58c6ed42e2f64a8c ]
When the start of an inner packet is split across two outer packets
such that fewer than 4 bytes land at the end of the first one,
__input_process_payload() saves those bytes as a runt and skips the
iplen/iphlen validation performed for in-place packets. When the
continuation packet arrives, iptfs_reassem_cont() only requires the
declared inner length to be >= sizeof(ra_runt) (6) before allocating
the reassembly skb with that attacker-controlled length.
However, __iptfs_iphlen() always returns the fixed minimum IP header
size (20 for IPv4, 40 for IPv6), so for an inner IPv4 tot_len in
[6, 19] the header-completion copy writes past the declared packet
length, and the subsequent "ipremain -= copylen" underflows to ~4GB,
leaving the payload copy length bounded only by blkoff (up to 64KB).
At runtime the skb_put() tailroom check turns this into
skb_over_panic(), i.e. an unprivileged kernel panic (DoS), reachable
locally via userns+netns IPTFS SAs and remotely against IPTFS VPN
gateways when the decrypted outer skb is linear (e.g. AF_PACKET taps,
tun/tap delivery).
Align the runt path with the normal path by requiring the declared
inner length to cover at least the IP header size. This also subsumes
the previous >= sizeof(ra_runt) check, since the minimum IP header
is always larger than the runt buffer.
This issue was found by the autokbug dynamic kernel fuzzer at
Tencent Yunding Lab.
Fixes: 075694765446 ("xfrm: iptfs: handle received fragmented inner packets")
Reported-by: Henry Martin <bsdhenrymartin@gmail.com>
Signed-off-by: Henry Martin <bsdhenrymartin@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Roshan Kumar <roshaen09@gmail.com>
Date: Tue Jul 28 10:56:08 2026 +0530
xfrm: iptfs: fix stack OOB read in iptfs_skb_reset_frag_walk()
[ Upstream commit d042487dc118e494db2e2c1382310255c90ff544 ]
iptfs_skb_reset_frag_walk() advances to the fragment containing @offset
with an unbounded loop:
while (offset >= walk->past + walk->frags[walk->fragi].len)
walk->past += walk->frags[walk->fragi++].len;
walk->fragi is advanced and walk->frags[walk->fragi] is dereferenced
without ever checking fragi against walk->nr_frags. When the requested
offset is at or beyond the total length spanned by the walk's fragments,
fragi runs past nr_frags and off the end of the fixed-size on-stack
frags[MAX_SKB_FRAGS + 1] array, reading out-of-bounds stack memory.
The two callers behave differently: iptfs_skb_add_frags() already guards
against this with
if (!walk->nr_frags ||
offset >= walk->total + walk->initial_offset)
return len;
but iptfs_skb_can_add_frags() has no such guard and calls
iptfs_skb_reset_frag_walk() unconditionally, so it performs the
out-of-range walk. Its own "fragi < walk->nr_frags" bound check runs only
afterwards, too late to prevent the read.
This is reachable from the receive path: a crafted IP-TFS (AGGFRAG)
payload delivered to an IPTFS SA drives iptfs_reassem_cont() ->
iptfs_skb_can_add_frags() with an offset past the fragment total, e.g.:
BUG: KASAN: stack-out-of-bounds in iptfs_skb_reset_frag_walk+0x235/0x250
Read of size 4 at addr ffff888008ad7210 by task repro/345
iptfs_skb_reset_frag_walk+0x235/0x250 net/xfrm/xfrm_iptfs.c:392
iptfs_skb_can_add_frags+0x155/0x310 net/xfrm/xfrm_iptfs.c:420
iptfs_reassem_cont+0xcf8/0x1140 net/xfrm/xfrm_iptfs.c:902
iptfs_input_ordered+0x552/0x670 net/xfrm/xfrm_iptfs.c:1280
iptfs_input+0x3d6/0xde0 net/xfrm/xfrm_iptfs.c:1741
xfrm_input+0x282f/0x6140 net/xfrm/xfrm_input.c:700
xfrm4_esp_rcv+0x93/0x120 net/ipv4/xfrm4_protocol.c:104
ip_rcv+0x278/0x2d0 net/ipv4/ip_input.c:612
Give iptfs_skb_can_add_frags() the same up-front guard that
iptfs_skb_add_frags() already has, so the walk is never entered with an
out-of-range offset. When it triggers, the caller falls back to the
existing linearize-and-copy path, which is safe.
Fixes: 5f2b6a909574 ("xfrm: iptfs: add skb-fragment sharing code")
Reported-by: Roshan Kumar <roshaen09@gmail.com>
Signed-off-by: Roshan Kumar <roshaen09@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Author: Zhiling Zou <zhilinz@nebusec.ai>
Date: Sat Aug 29 17:24:24 2026 +0800
xfrm: save input state data before secpath resets
commit 3cf5cdecd99c9c186a5ea518d93bbf3045b6e3aa upstream.
xfrm_input() stores the current xfrm_state in the skb secpath while it
continues receive-side processing. Some input paths can reset that secpath
before xfrm_input() has finished dereferencing the state.
Receive callback users such as VTI and XFRM interfaces can reset the
secpath. The VTI receive path does so before checking whether the packet
crosses network namespaces, while the XFRM interface path does so only for
cross-network-namespace packets. The XFRM_MAX_DEPTH error path can also
reset the secpath before the final drop callback reports the current
state's protocol.
If secpath_reset() drops the last state reference while the state is
concurrently deleted, xfrm_input() can still dereference the freed state
when selecting transport_finish() or reporting the drop callback protocol.
Save the state protocol on the stack while the state is still valid,
and use the already saved address family for transport_finish(). A larval
XFRM_STATE_ACQ state has no type, so retain nexthdr as its protocol. This
preserves the existing drop-path fallback while avoiding the post-reset
state dereferences without adding an extra state reference to every
received packet.
Fixes: df3893c176e9 ("vti: Update the ipv4 side to use it's own receive hook.")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Chengfeng Ye <nicoyip.dev@gmail.com>
Date: Thu Jul 30 18:35:43 2026 +0800
xfrm: serialize state GC with device state flush
commit 89fefad9f971bc637fb22373078144f2563c4be9 upstream.
The deferred-device pass in xfrm_dev_state_flush() finds states under
xfrm_state_dev_gc_lock, but drops the lock before calling
xfrm_dev_state_free() because the driver callback may sleep. The device
GC list does not hold an xfrm_state reference, so the state GC worker can
destroy the same state concurrently.
The race can proceed as follows:
CPU 0 CPU 1
find x on the device GC list
drop xfrm_state_dev_gc_lock
read x->xso.dev
xfrm_state_gc_destroy(x)
xfrm_dev_state_free(x)
xfrm_state_free(x)
continue xfrm_dev_state_free(x)
Both paths can invoke the driver callback and drop the device reference.
CPU 0 can also access the xfrm_state after CPU 1 has freed it.
KASAN reported:
BUG: KASAN: slab-use-after-free in xfrm_dev_state_free+0x24c/0x2a0
Read of size 8 at addr ffff88810bbaa960 by task poc/102
Call Trace:
xfrm_dev_state_free+0x24c/0x2a0
xfrm_dev_state_flush+0x353/0x400
xfrm_dev_event+0x26d/0x3a0
notifier_call_chain+0xc0/0x280
__dev_notify_flags+0x169/0x250
netif_change_flags+0xe7/0x160
dev_change_flags+0x96/0x220
devinet_ioctl+0x7f4/0x1880
Allocated by task 87:
xfrm_state_alloc+0x1e/0x5c0
xfrm_add_sa+0xe7f/0x5820
xfrm_user_rcv_msg+0x4f3/0x940
Freed by task 57:
kmem_cache_free+0xcb/0x3d0
xfrm_state_gc_task+0x4a8/0x650
process_one_work+0x63a/0x1070
Serialize xfrm_state destruction against the deferred-device pass with a
mutex. Keep xfrm_state_dev_gc_lock limited to list operations and retain
the existing callback and device-reference release ordering.
Fixes: 07b87f9eea0c ("xfrm: Fix unregister netdevice hang on hardware offload.")
Cc: stable@vger.kernel.org
Signed-off-by: Chengfeng Ye <nicoyip.dev@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Author: Siwei Zhang <fourdizhang@tencent.com>
Date: Thu Jul 30 19:40:08 2026 +0800
xfrm: use hlist_del_init_rcu for state_cache and state_cache_input
commit 2afb8dc1f4390f164db8352f8e685e126e9db566 upstream.
Commit 14acf9652e56 ("xfrm: defensively unhash xfrm_state lists in
__xfrm_state_delete") converted bydst/bysrc/byseq/byspi from
hlist_del_rcu() to hlist_del_init_rcu() so that a second
__xfrm_state_delete() on the same object becomes a no-op rather than a
write through LIST_POISON pprev. It missed state_cache and
state_cache_input, which kept hlist_del_rcu():
- hlist_del_rcu() leaves pprev = LIST_POISON2 (non-NULL), so
hlist_unhashed() returns false.
- hlist_del_init_rcu() leaves pprev = NULL, so hlist_unhashed()
returns true.
A second __xfrm_state_delete() therefore enters __hlist_del() on the
already-deleted state_cache/state_cache_input nodes and does
WRITE_ONCE(*pprev, next) through LIST_POISON2 — a write use-after-free
once the slab is reused. The corruption can in turn cause a subsequent
hlist_for_each_entry_rcu traversal to follow a dangling next pointer,
producing the read use-after-free reported in xfrm_input_state_lookup().
Switch state_cache and state_cache_input to hlist_del_init_rcu() to
match the other four lists, closing the write use-after-free and, with
it, the read use-after-free it spawns.
Assisted-by: CodeBuddy:GLM-5.2
Fixes: 0045e3d80613 ("xfrm: Cache used outbound xfrm states at the policy.")
Fixes: 81a331a0e72d ("xfrm: Add an inbound percpu state cache.")
Cc: stable@vger.kernel.org
Signed-off-by: Siwei Zhang <fourdizhang@tencent.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>