| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| In the Linux kernel, the following vulnerability has been resolved:
X.509: Fix out-of-bounds access when parsing extensions
Leo reports an out-of-bounds access when parsing a certificate with
empty Basic Constraints or Key Usage extension because the first byte of
the extension is read before checking its length. Fix it.
The bug can be triggered by an unprivileged user by submitting a
specially crafted certificate to the kernel through the keyrings(7) API.
Leo has demonstrated this with a proof-of-concept program responsibly
disclosed off-list. |
| In the Linux kernel, the following vulnerability has been resolved:
net/mlx5: Fix deadlock between devlink lock and esw->wq
esw->work_queue executes esw_functions_changed_event_handler ->
esw_vfs_changed_event_handler and acquires the devlink lock.
.eswitch_mode_set (acquires devlink lock in devlink_nl_pre_doit) ->
mlx5_devlink_eswitch_mode_set -> mlx5_eswitch_disable_locked ->
mlx5_eswitch_event_handler_unregister -> flush_workqueue deadlocks
when esw_vfs_changed_event_handler executes.
Fix that by no longer flushing the work to avoid the deadlock, and using
a generation counter to keep track of work relevance. This avoids an old
handler manipulating an esw that has undergone one or more mode changes:
- the counter is incremented in mlx5_eswitch_event_handler_unregister.
- the counter is read and passed to the ephemeral mlx5_host_work struct.
- the work handler takes the devlink lock and bails out if the current
generation is different than the one it was scheduled to operate on.
- mlx5_eswitch_cleanup does the final draining before destroying the wq.
No longer flushing the workqueue has the side effect of maybe no longer
cancelling pending vport_change_handler work items, but that's ok since
those are disabled elsewhere:
- mlx5_eswitch_disable_locked disables the vport eq notifier.
- mlx5_esw_vport_disable disarms the HW EQ notification and marks
vport->enabled under state_lock to false to prevent pending vport
handler from doing anything.
- mlx5_eswitch_cleanup destroys the workqueue and makes sure all events
are disabled/finished. |
| In the Linux kernel, the following vulnerability has been resolved:
unshare: fix unshare_fs() handling
There's an unpleasant corner case in unshare(2), when we have a
CLONE_NEWNS in flags and current->fs hadn't been shared at all; in that
case copy_mnt_ns() gets passed current->fs instead of a private copy,
which causes interesting warts in proof of correctness]
> I guess if private means fs->users == 1, the condition could still be true.
Unfortunately, it's worse than just a convoluted proof of correctness.
Consider the case when we have CLONE_NEWCGROUP in addition to CLONE_NEWNS
(and current->fs->users == 1).
We pass current->fs to copy_mnt_ns(), all right. Suppose it succeeds and
flips current->fs->{pwd,root} to corresponding locations in the new namespace.
Now we proceed to copy_cgroup_ns(), which fails (e.g. with -ENOMEM).
We call put_mnt_ns() on the namespace created by copy_mnt_ns(), it's
destroyed and its mount tree is dissolved, but... current->fs->root and
current->fs->pwd are both left pointing to now detached mounts.
They are pinning those, so it's not a UAF, but it leaves the calling
process with unshare(2) failing with -ENOMEM _and_ leaving it with
pwd and root on detached isolated mounts. The last part is clearly a bug.
There is other fun related to that mess (races with pivot_root(), including
the one between pivot_root() and fork(), of all things), but this one
is easy to isolate and fix - treat CLONE_NEWNS as "allocate a new
fs_struct even if it hadn't been shared in the first place". Sure, we could
go for something like "if both CLONE_NEWNS *and* one of the things that might
end up failing after copy_mnt_ns() call in create_new_namespaces() are set,
force allocation of new fs_struct", but let's keep it simple - the cost
of copy_fs_struct() is trivial.
Another benefit is that copy_mnt_ns() with CLONE_NEWNS *always* gets
a freshly allocated fs_struct, yet to be attached to anything. That
seriously simplifies the analysis...
FWIW, that bug had been there since the introduction of unshare(2) ;-/ |
| In the Linux kernel, the following vulnerability has been resolved:
fs: init flags_valid before calling vfs_fileattr_get
syzbot reported a uninit-value bug in [1].
Similar to the "*get" context where the kernel's internal file_kattr
structure is initialized before calling vfs_fileattr_get(), we should
use the same mechanism when using fa.
[1]
BUG: KMSAN: uninit-value in fuse_fileattr_get+0xeb4/0x1450 fs/fuse/ioctl.c:517
fuse_fileattr_get+0xeb4/0x1450 fs/fuse/ioctl.c:517
vfs_fileattr_get fs/file_attr.c:94 [inline]
__do_sys_file_getattr fs/file_attr.c:416 [inline]
Local variable fa.i created at:
__do_sys_file_getattr fs/file_attr.c:380 [inline]
__se_sys_file_getattr+0x8c/0xbd0 fs/file_attr.c:372 |
| In the Linux kernel, the following vulnerability has been resolved:
scsi: storvsc: Fix scheduling while atomic on PREEMPT_RT
This resolves the follow splat and lock-up when running with PREEMPT_RT
enabled on Hyper-V:
[ 415.140818] BUG: scheduling while atomic: stress-ng-iomix/1048/0x00000002
[ 415.140822] INFO: lockdep is turned off.
[ 415.140823] Modules linked in: intel_rapl_msr intel_rapl_common intel_uncore_frequency_common intel_pmc_core pmt_telemetry pmt_discovery pmt_class intel_pmc_ssram_telemetry intel_vsec ghash_clmulni_intel aesni_intel rapl binfmt_misc nls_ascii nls_cp437 vfat fat snd_pcm hyperv_drm snd_timer drm_client_lib drm_shmem_helper snd sg soundcore drm_kms_helper pcspkr hv_balloon hv_utils evdev joydev drm configfs efi_pstore nfnetlink vsock_loopback vmw_vsock_virtio_transport_common hv_sock vmw_vsock_vmci_transport vsock vmw_vmci efivarfs autofs4 ext4 crc16 mbcache jbd2 sr_mod sd_mod cdrom hv_storvsc serio_raw hid_generic scsi_transport_fc hid_hyperv scsi_mod hid hv_netvsc hyperv_keyboard scsi_common
[ 415.140846] Preemption disabled at:
[ 415.140847] [<ffffffffc0656171>] storvsc_queuecommand+0x2e1/0xbe0 [hv_storvsc]
[ 415.140854] CPU: 8 UID: 0 PID: 1048 Comm: stress-ng-iomix Not tainted 6.19.0-rc7 #30 PREEMPT_{RT,(full)}
[ 415.140856] Hardware name: Microsoft Corporation Virtual Machine/Virtual Machine, BIOS Hyper-V UEFI Release v4.1 09/04/2024
[ 415.140857] Call Trace:
[ 415.140861] <TASK>
[ 415.140861] ? storvsc_queuecommand+0x2e1/0xbe0 [hv_storvsc]
[ 415.140863] dump_stack_lvl+0x91/0xb0
[ 415.140870] __schedule_bug+0x9c/0xc0
[ 415.140875] __schedule+0xdf6/0x1300
[ 415.140877] ? rtlock_slowlock_locked+0x56c/0x1980
[ 415.140879] ? rcu_is_watching+0x12/0x60
[ 415.140883] schedule_rtlock+0x21/0x40
[ 415.140885] rtlock_slowlock_locked+0x502/0x1980
[ 415.140891] rt_spin_lock+0x89/0x1e0
[ 415.140893] hv_ringbuffer_write+0x87/0x2a0
[ 415.140899] vmbus_sendpacket_mpb_desc+0xb6/0xe0
[ 415.140900] ? rcu_is_watching+0x12/0x60
[ 415.140902] storvsc_queuecommand+0x669/0xbe0 [hv_storvsc]
[ 415.140904] ? HARDIRQ_verbose+0x10/0x10
[ 415.140908] ? __rq_qos_issue+0x28/0x40
[ 415.140911] scsi_queue_rq+0x760/0xd80 [scsi_mod]
[ 415.140926] __blk_mq_issue_directly+0x4a/0xc0
[ 415.140928] blk_mq_issue_direct+0x87/0x2b0
[ 415.140931] blk_mq_dispatch_queue_requests+0x120/0x440
[ 415.140933] blk_mq_flush_plug_list+0x7a/0x1a0
[ 415.140935] __blk_flush_plug+0xf4/0x150
[ 415.140940] __submit_bio+0x2b2/0x5c0
[ 415.140944] ? submit_bio_noacct_nocheck+0x272/0x360
[ 415.140946] submit_bio_noacct_nocheck+0x272/0x360
[ 415.140951] ext4_read_bh_lock+0x3e/0x60 [ext4]
[ 415.140995] ext4_block_write_begin+0x396/0x650 [ext4]
[ 415.141018] ? __pfx_ext4_da_get_block_prep+0x10/0x10 [ext4]
[ 415.141038] ext4_da_write_begin+0x1c4/0x350 [ext4]
[ 415.141060] generic_perform_write+0x14e/0x2c0
[ 415.141065] ext4_buffered_write_iter+0x6b/0x120 [ext4]
[ 415.141083] vfs_write+0x2ca/0x570
[ 415.141087] ksys_write+0x76/0xf0
[ 415.141089] do_syscall_64+0x99/0x1490
[ 415.141093] ? rcu_is_watching+0x12/0x60
[ 415.141095] ? finish_task_switch.isra.0+0xdf/0x3d0
[ 415.141097] ? rcu_is_watching+0x12/0x60
[ 415.141098] ? lock_release+0x1f0/0x2a0
[ 415.141100] ? rcu_is_watching+0x12/0x60
[ 415.141101] ? finish_task_switch.isra.0+0xe4/0x3d0
[ 415.141103] ? rcu_is_watching+0x12/0x60
[ 415.141104] ? __schedule+0xb34/0x1300
[ 415.141106] ? hrtimer_try_to_cancel+0x1d/0x170
[ 415.141109] ? do_nanosleep+0x8b/0x160
[ 415.141111] ? hrtimer_nanosleep+0x89/0x100
[ 415.141114] ? __pfx_hrtimer_wakeup+0x10/0x10
[ 415.141116] ? xfd_validate_state+0x26/0x90
[ 415.141118] ? rcu_is_watching+0x12/0x60
[ 415.141120] ? do_syscall_64+0x1e0/0x1490
[ 415.141121] ? do_syscall_64+0x1e0/0x1490
[ 415.141123] ? rcu_is_watching+0x12/0x60
[ 415.141124] ? do_syscall_64+0x1e0/0x1490
[ 415.141125] ? do_syscall_64+0x1e0/0x1490
[ 415.141127] ? irqentry_exit+0x140/0
---truncated--- |
| In the Linux kernel, the following vulnerability has been resolved:
nfs: return EISDIR on nfs3_proc_create if d_alias is a dir
If we found an alias through nfs3_do_create/nfs_add_or_obtain
/d_splice_alias which happens to be a dir dentry, we don't return
any error, and simply forget about this alias, but the original
dentry we were adding and passed as parameter remains negative.
This later causes an oops on nfs_atomic_open_v23/finish_open since we
supply a negative dentry to do_dentry_open.
This has been observed running lustre-racer, where dirs and files are
created/removed concurrently with the same name and O_EXCL is not
used to open files (frequent file redirection).
While d_splice_alias typically returns a directory alias or NULL, we
explicitly check d_is_dir() to ensure that we don't attempt to perform
file operations (like finish_open) on a directory inode, which triggers
the observed oops. |
| In the Linux kernel, the following vulnerability has been resolved:
net: liquidio: Fix off-by-one error in PF setup_nic_devices() cleanup
In setup_nic_devices(), the initialization loop jumps to the label
setup_nic_dev_free on failure. The current cleanup loop while(i--)
skip the failing index i, causing a memory leak.
Fix this by changing the loop to iterate from the current index i
down to 0.
Also, decrement i in the devlink_alloc failure path to point to the
last successfully allocated index.
Compile tested only. Issue found using code review. |
| In the Linux kernel, the following vulnerability has been resolved:
audit: add missing syscalls to read class
The "at" variant of getxattr() and listxattr() are missing from the
audit read class. Calling getxattrat() or listxattrat() on a file to
read its extended attributes will bypass audit rules such as:
-w /tmp/test -p rwa -k test_rwa
The current patch adds missing syscalls to the audit read class. |
| In the Linux kernel, the following vulnerability has been resolved:
net: liquidio: Fix off-by-one error in VF setup_nic_devices() cleanup
In setup_nic_devices(), the initialization loop jumps to the label
setup_nic_dev_free on failure. The current cleanup loop while(i--)
skip the failing index i, causing a memory leak.
Fix this by changing the loop to iterate from the current index i
down to 0.
Compile tested only. Issue found using code review. |
| In the Linux kernel, the following vulnerability has been resolved:
crypto: algif_aead - Fix minimum RX size check for decryption
The check for the minimum receive buffer size did not take the
tag size into account during decryption. Fix this by adding the
required extra length. |
| In the Linux kernel, the following vulnerability has been resolved:
ext4: avoid allocate block from corrupted group in ext4_mb_find_by_goal()
There's issue as follows:
...
EXT4-fs (mmcblk0p1): Delayed block allocation failed for inode 206 at logical offset 0 with max blocks 1 with error 117
EXT4-fs (mmcblk0p1): This should not happen!! Data will be lost
EXT4-fs (mmcblk0p1): Delayed block allocation failed for inode 206 at logical offset 0 with max blocks 1 with error 117
EXT4-fs (mmcblk0p1): This should not happen!! Data will be lost
EXT4-fs (mmcblk0p1): Delayed block allocation failed for inode 206 at logical offset 0 with max blocks 1 with error 117
EXT4-fs (mmcblk0p1): This should not happen!! Data will be lost
EXT4-fs (mmcblk0p1): Delayed block allocation failed for inode 206 at logical offset 0 with max blocks 1 with error 117
EXT4-fs (mmcblk0p1): This should not happen!! Data will be lost
EXT4-fs (mmcblk0p1): Delayed block allocation failed for inode 2243 at logical offset 0 with max blocks 1 with error 117
EXT4-fs (mmcblk0p1): This should not happen!! Data will be lost
EXT4-fs (mmcblk0p1): Delayed block allocation failed for inode 2239 at logical offset 0 with max blocks 1 with error 117
EXT4-fs (mmcblk0p1): This should not happen!! Data will be lost
EXT4-fs (mmcblk0p1): error count since last fsck: 1
EXT4-fs (mmcblk0p1): initial error at time 1765597433: ext4_mb_generate_buddy:760
EXT4-fs (mmcblk0p1): last error at time 1765597433: ext4_mb_generate_buddy:760
...
According to the log analysis, blocks are always requested from the
corrupted block group. This may happen as follows:
ext4_mb_find_by_goal
ext4_mb_load_buddy
ext4_mb_load_buddy_gfp
ext4_mb_init_cache
ext4_read_block_bitmap_nowait
ext4_wait_block_bitmap
ext4_validate_block_bitmap
if (!grp || EXT4_MB_GRP_BBITMAP_CORRUPT(grp))
return -EFSCORRUPTED; // There's no logs.
if (err)
return err; // Will return error
ext4_lock_group(ac->ac_sb, group);
if (unlikely(EXT4_MB_GRP_BBITMAP_CORRUPT(e4b->bd_info))) // Unreachable
goto out;
After commit 9008a58e5dce ("ext4: make the bitmap read routines return
real error codes") merged, Commit 163a203ddb36 ("ext4: mark block group
as corrupt on block bitmap error") is no real solution for allocating
blocks from corrupted block groups. This is because if
'EXT4_MB_GRP_BBITMAP_CORRUPT(e4b->bd_info)' is true, then
'ext4_mb_load_buddy()' may return an error. This means that the block
allocation will fail.
Therefore, check block group if corrupted when ext4_mb_load_buddy()
returns error. |
| In the Linux kernel, the following vulnerability has been resolved:
audit: add fchmodat2() to change attributes class
fchmodat2(), introduced in version 6.6 is currently not in the change
attribute class of audit. Calling fchmodat2() to change a file
attribute in the same fashion than chmod() or fchmodat() will bypass
audit rules such as:
-w /tmp/test -p rwa -k test_rwa
The current patch adds fchmodat2() to the change attributes class. |
| In the Linux kernel, the following vulnerability has been resolved:
usb: gadget: f_tcm: Fix NULL pointer dereferences in nexus handling
The `tpg->tpg_nexus` pointer in the USB Target driver is dynamically
managed and tied to userspace configuration via ConfigFS. It can be
NULL if the USB host sends requests before the nexus is fully
established or immediately after it is dropped.
Currently, functions like `bot_submit_command()` and the data
transfer paths retrieve `tv_nexus = tpg->tpg_nexus` and immediately
dereference `tv_nexus->tvn_se_sess` without any validation. If a
malicious or misconfigured USB host sends a BOT (Bulk-Only Transport)
command during this race window, it triggers a NULL pointer
dereference, leading to a kernel panic (local DoS).
This exposes an inconsistent API usage within the module, as peer
functions like `usbg_submit_command()` and `bot_send_bad_response()`
correctly implement a NULL check for `tv_nexus` before proceeding.
Fix this by bringing consistency to the nexus handling. Add the
missing `if (!tv_nexus)` checks to the vulnerable BOT command and
request processing paths, aborting the command gracefully with an
error instead of crashing the system. |
| In the Linux kernel, the following vulnerability has been resolved:
net: skb: fix cross-cache free of KFENCE-allocated skb head
SKB_SMALL_HEAD_CACHE_SIZE is intentionally set to a non-power-of-2
value (e.g. 704 on x86_64) to avoid collisions with generic kmalloc
bucket sizes. This ensures that skb_kfree_head() can reliably use
skb_end_offset to distinguish skb heads allocated from
skb_small_head_cache vs. generic kmalloc caches.
However, when KFENCE is enabled, kfence_ksize() returns the exact
requested allocation size instead of the slab bucket size. If a caller
(e.g. bpf_test_init) allocates skb head data via kzalloc() and the
requested size happens to equal SKB_SMALL_HEAD_CACHE_SIZE, then
slab_build_skb() -> ksize() returns that exact value. After subtracting
skb_shared_info overhead, skb_end_offset ends up matching
SKB_SMALL_HEAD_HEADROOM, causing skb_kfree_head() to incorrectly free
the object to skb_small_head_cache instead of back to the original
kmalloc cache, resulting in a slab cross-cache free:
kmem_cache_free(skbuff_small_head): Wrong slab cache. Expected
skbuff_small_head but got kmalloc-1k
Fix this by always calling kfree(head) in skb_kfree_head(). This keeps
the free path generic and avoids allocator-specific misclassification
for KFENCE objects. |
| In the Linux kernel, the following vulnerability has been resolved:
spi: rockchip-sfc: Fix double-free in remove() callback
The driver uses devm_spi_register_controller() for registration, which
automatically unregisters the controller via devm cleanup when the
device is removed. The manual call to spi_unregister_controller() in
the remove() callback can lead to a double-free.
And to make sure controller is unregistered before DMA buffer is
unmapped, switch to use spi_register_controller() in probe(). |
| In the Linux kernel, the following vulnerability has been resolved:
spi: amlogic: spifc-a4: Fix DMA mapping error handling
Fix three bugs in aml_sfc_dma_buffer_setup() error paths:
1. Unnecessary goto: When the first DMA mapping (sfc->daddr) fails,
nothing needs cleanup. Use direct return instead of goto.
2. Double-unmap bug: When info DMA mapping failed, the code would
unmap sfc->daddr inline, then fall through to out_map_data which
would unmap it again, causing a double-unmap.
3. Wrong unmap size: The out_map_info label used datalen instead of
infolen when unmapping sfc->iaddr, which could lead to incorrect
DMA sync behavior. |
| In the Linux kernel, the following vulnerability has been resolved:
net: spacemit: Fix error handling in emac_tx_mem_map()
The DMA mappings were leaked on mapping error. Free them with the
existing emac_free_tx_buf() function. |
| In the Linux kernel, the following vulnerability has been resolved:
rxrpc, afs: Fix missing error pointer check after rxrpc_kernel_lookup_peer()
rxrpc_kernel_lookup_peer() can also return error pointers in addition to
NULL, so just checking for NULL is not sufficient.
Fix this by:
(1) Changing rxrpc_kernel_lookup_peer() to return -ENOMEM rather than NULL
on allocation failure.
(2) Making the callers in afs use IS_ERR() and PTR_ERR() to pass on the
error code returned. |
| In the Linux kernel, the following vulnerability has been resolved:
net/mlx5e: RX, Fix XDP multi-buf frag counting for legacy RQ
XDP multi-buf programs can modify the layout of the XDP buffer when the
program calls bpf_xdp_pull_data() or bpf_xdp_adjust_tail(). The
referenced commit in the fixes tag corrected the assumption in the mlx5
driver that the XDP buffer layout doesn't change during a program
execution. However, this fix introduced another issue: the dropped
fragments still need to be counted on the driver side to avoid page
fragment reference counting issues.
Such issue can be observed with the
test_xdp_native_adjst_tail_shrnk_data selftest when using a payload of
3600 and shrinking by 256 bytes (an upcoming selftest patch): the last
fragment gets released by the XDP code but doesn't get tracked by the
driver. This results in a negative pp_ref_count during page release and
the following splat:
WARNING: include/net/page_pool/helpers.h:297 at mlx5e_page_release_fragmented.isra.0+0x4a/0x50 [mlx5_core], CPU#12: ip/3137
Modules linked in: [...]
CPU: 12 UID: 0 PID: 3137 Comm: ip Not tainted 6.19.0-rc3+ #12 NONE
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014
RIP: 0010:mlx5e_page_release_fragmented.isra.0+0x4a/0x50 [mlx5_core]
[...]
Call Trace:
<TASK>
mlx5e_dealloc_rx_wqe+0xcb/0x1a0 [mlx5_core]
mlx5e_free_rx_descs+0x7f/0x110 [mlx5_core]
mlx5e_close_rq+0x50/0x60 [mlx5_core]
mlx5e_close_queues+0x36/0x2c0 [mlx5_core]
mlx5e_close_channel+0x1c/0x50 [mlx5_core]
mlx5e_close_channels+0x45/0x80 [mlx5_core]
mlx5e_safe_switch_params+0x1a5/0x230 [mlx5_core]
mlx5e_change_mtu+0xf3/0x2f0 [mlx5_core]
netif_set_mtu_ext+0xf1/0x230
do_setlink.isra.0+0x219/0x1180
rtnl_newlink+0x79f/0xb60
rtnetlink_rcv_msg+0x213/0x3a0
netlink_rcv_skb+0x48/0xf0
netlink_unicast+0x24a/0x350
netlink_sendmsg+0x1ee/0x410
__sock_sendmsg+0x38/0x60
____sys_sendmsg+0x232/0x280
___sys_sendmsg+0x78/0xb0
__sys_sendmsg+0x5f/0xb0
[...]
do_syscall_64+0x57/0xc50
This patch fixes the issue by doing page frag counting on all the
original XDP buffer fragments for all relevant XDP actions (XDP_TX ,
XDP_REDIRECT and XDP_PASS). This is basically reverting to the original
counting before the commit in the fixes tag.
As frag_page is still pointing to the original tail, the nr_frags
parameter to xdp_update_skb_frags_info() needs to be calculated
in a different way to reflect the new nr_frags. |
| In the Linux kernel, the following vulnerability has been resolved:
net/mlx5e: RX, Fix XDP multi-buf frag counting for striding RQ
XDP multi-buf programs can modify the layout of the XDP buffer when the
program calls bpf_xdp_pull_data() or bpf_xdp_adjust_tail(). The
referenced commit in the fixes tag corrected the assumption in the mlx5
driver that the XDP buffer layout doesn't change during a program
execution. However, this fix introduced another issue: the dropped
fragments still need to be counted on the driver side to avoid page
fragment reference counting issues.
The issue was discovered by the drivers/net/xdp.py selftest,
more specifically the test_xdp_native_tx_mb:
- The mlx5 driver allocates a page_pool page and initializes it with
a frag counter of 64 (pp_ref_count=64) and the internal frag counter
to 0.
- The test sends one packet with no payload.
- On RX (mlx5e_skb_from_cqe_mpwrq_nonlinear()), mlx5 configures the XDP
buffer with the packet data starting in the first fragment which is the
page mentioned above.
- The XDP program runs and calls bpf_xdp_pull_data() which moves the
header into the linear part of the XDP buffer. As the packet doesn't
contain more data, the program drops the tail fragment since it no
longer contains any payload (pp_ref_count=63).
- mlx5 device skips counting this fragment. Internal frag counter
remains 0.
- mlx5 releases all 64 fragments of the page but page pp_ref_count is
63 => negative reference counting error.
Resulting splat during the test:
WARNING: CPU: 0 PID: 188225 at ./include/net/page_pool/helpers.h:297 mlx5e_page_release_fragmented.isra.0+0xbd/0xe0 [mlx5_core]
Modules linked in: [...]
CPU: 0 UID: 0 PID: 188225 Comm: ip Not tainted 6.18.0-rc7_for_upstream_min_debug_2025_12_08_11_44 #1 NONE
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
RIP: 0010:mlx5e_page_release_fragmented.isra.0+0xbd/0xe0 [mlx5_core]
[...]
Call Trace:
<TASK>
mlx5e_free_rx_mpwqe+0x20a/0x250 [mlx5_core]
mlx5e_dealloc_rx_mpwqe+0x37/0xb0 [mlx5_core]
mlx5e_free_rx_descs+0x11a/0x170 [mlx5_core]
mlx5e_close_rq+0x78/0xa0 [mlx5_core]
mlx5e_close_queues+0x46/0x2a0 [mlx5_core]
mlx5e_close_channel+0x24/0x90 [mlx5_core]
mlx5e_close_channels+0x5d/0xf0 [mlx5_core]
mlx5e_safe_switch_params+0x2ec/0x380 [mlx5_core]
mlx5e_change_mtu+0x11d/0x490 [mlx5_core]
mlx5e_change_nic_mtu+0x19/0x30 [mlx5_core]
netif_set_mtu_ext+0xfc/0x240
do_setlink.isra.0+0x226/0x1100
rtnl_newlink+0x7a9/0xba0
rtnetlink_rcv_msg+0x220/0x3c0
netlink_rcv_skb+0x4b/0xf0
netlink_unicast+0x255/0x380
netlink_sendmsg+0x1f3/0x420
__sock_sendmsg+0x38/0x60
____sys_sendmsg+0x1e8/0x240
___sys_sendmsg+0x7c/0xb0
[...]
__sys_sendmsg+0x5f/0xb0
do_syscall_64+0x55/0xc70
The problem applies for XDP_PASS as well which is handled in a different
code path in the driver.
This patch fixes the issue by doing page frag counting on all the
original XDP buffer fragments for all relevant XDP actions (XDP_TX ,
XDP_REDIRECT and XDP_PASS). This is basically reverting to the original
counting before the commit in the fixes tag.
As frag_page is still pointing to the original tail, the nr_frags
parameter to xdp_update_skb_frags_info() needs to be calculated
in a different way to reflect the new nr_frags. |
