-
CVE-2026-99067
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
btrfs: scrub: fix local_root reference leak in scrub_print_warning_inode()
When paths_from_inode() fails, scrub_print_warning_inode() jumps to err
without dropping the reference taken by btrfs_get_fs_root(), leaking a
reference to the root every time path resolution fails while printing
scrub warnings. Every other error and success path of the function
drops the reference.
Drop the reference on the paths_from_inode() failure path too.
-
CVE-2026-99058
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
workqueue: Fix NULL current_pwq deref in chained work check
current_wq_worker() only tells us that %current is a kworker. It does
not guarantee that the worker is currently executing a work item, as
worker-current_pwq is populated only while process_one_work() is running
the work function.
is_chained_work() can be reached while queueing on a draining or
destroying workqueue. If a kworker outside work-item execution reaches
that path, current_wq_worker() returns a worker but worker->current_pwq is
NULL, and the chained-work test can fault before it emits the intended
warning.
Guard the chained-work test with worker->current_pwq. A kworker without
current_pwq is not executing work on the target workqueue, so the helper
should return false and let the draining/destroying warning fire.
-
CVE-2026-99057
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
xen/netfront: don't leak the skb when xennet_fill_frags() fails
When a response chain has more slots than fit in the skb's frags,
xennet_fill_frags() returns an error and xennet_poll() jumps to its
error path. That path moves what's left on tmpq to errq to be freed,
but the skb being filled was already dequeued from tmpq, so it's never
freed. Each chain that overflows leaks the skb and the pages attached
to it as frags, and the backend decides how many slots it sends.
Put the skb back on tmpq before taking the error path, like the
xennet_set_skb_gso() failure just above it does.
-
CVE-2026-99049
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
gpio: xilinx: fix runtime PM leak on request error path
pm_runtime_get_sync() leaves the usage counter incremented even when it
fails, and xgpio_request() returns the error without dropping it.
gpiolib does not call -free() when ->request() fails, so xgpio_free(),
which holds the only matching pm_runtime_put(), never runs. The reference
is leaked and the controller can no longer runtime-suspend, so its clock
stays enabled.
Switch to pm_runtime_resume_and_get(), which only increments the usage
counter on success.
-
CVE-2026-99041
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
ip6_gre: remove obsolete ERSPAN PMTU update
A tc tunnel_key action can attach a METADATA_IP_TUNNEL dst to an skb,
and mirred can redirect it to a native ip6erspan device. The legacy PMTU
update in ip6erspan_tunnel_xmit() treats that metadata dst as a route.
Since it has no output device, dst_dev(dst)-mtu dereferences NULL.
On v7.2 with KASAN, an initial UID/GID 65534 process with CapEff 0 used
self-created user and network namespaces to trigger:
KASAN: null-ptr-deref
RIP: ip6erspan_tunnel_xmit+0x10fc/0x2cc0
Kernel panic - not syncing: Fatal exception in interrupt
The update is obsolete. Commit fe1a4ca0a2b7 ("ip6_gre: process toobig
in a better way") changed IPv6 GRE Too Big handling to update the
underlay route and removed the equivalent block from __gre6_xmit().
ip6_tnl_xmit() then propagates the underlay PMTU to the overlay route.
Remove the stale ERSPAN copy and its now-unused dst variable. This
eliminates the metadata-dst dereference and keeps ERSPAN aligned with
IPv6 GRE PMTU handling.
-
CVE-2026-99031
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
net: bridge: avoid recursive multicast port cleanup
br_multicast_sg_del_exclude_ports() removes automatically added STAR_EXCL
port groups by calling br_multicast_del_pg(). For S,G entries, that
deletion calls back into br_multicast_sg_del_exclude_ports(), so a large
number of STAR_EXCL ports consumes one kernel stack frame per port and can
hit the stack guard page.
Keep the complete port-group deletion path for cleanup, but suppress the
S,G exclude-port cleanup while that helper is already walking the same
entry. Normal deletion callers retain the existing behavior.
-
CVE-2026-99027
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
net: team: stop reusing skb after queue override
A nonzero dev_queue_xmit() result does not return ownership of the skb.
When an override port's qdisc drops it, the queue override helper used to
try another port or the mode transmit op with the freed skb.
Return the handoff decision separately from transmit success. Select only
the first override port under RCU and stop after handing it the skb,
regardless of the lower transmit result. Keep the existing success and
drop accounting.
BUG: KASAN: slab-use-after-free in sk_skb_reason_drop
Write of size 4 by task poc/5040
sk_skb_reason_drop (net/core/skbuff.c:1240)
dev_kfree_skb_any_reason (net/core/dev.c:3470)
ab_transmit (drivers/net/team/team_mode_activebackup.c:49)
team_xmit (drivers/net/team/team_core.c:1849)
__dev_direct_xmit (net/core/dev.c:4934)
Freed by task 5040:
__tcf_kfree_skb_list (net/sched/sch_generic.c:43)
__dev_queue_xmit (net/core/dev.c:4831)
team_xmit (drivers/net/team/team_core.c:1846)
Kernel panic - not syncing: KASAN: panic_on_warn set ...
-
CVE-2026-99025
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
openvswitch: fix soft lockup in the netlink flow dump
A production compute node carrying a few thousand datapath flows hit a
soft lockup inside a single netlink flow dump and panicked.
ovs_flow_cmd_dump() calls ovs_flow_stats_get() for every flow it
emits, and that releases stats-lock with spin_unlock_bh() once per
CPU that has touched the flow. Every release is a local_bh_enable(),
and each one runs the pending softirq backlog in the dumping thread's
own context.
The skb bounds the flows one callback emits, but not the softirq work
it absorbs. On a CPU that carries the box's packet load the backlog
refills as fast as it drains, so the dumping thread becomes that CPU's
softirq engine. It never sleeps and it has no reschedule point, so
under voluntary preemption nothing can take the CPU away from it:
neither the ksoftirqd the kernel woke to take the work over, nor the
stopper thread the softlockup detector dispatches to refresh its
timestamp.
Hold BH off across the whole callback instead, the way
ctnetlink_dump_table() does, so the nested spin_unlock_bh() stop
draining softirqs. The loop already runs under rcu_read_lock() and
cannot sleep. What it gives up is preemption under CONFIG_PREEMPT,
since a BH-off region is not preemptible outside PREEMPT_RT. The
region stays short: the skb caps the flows one callback emits, and
empty buckets cost no skb space but are each visited once per dump, as
the cursor only moves forward. The table grows on insert and shrinks
only on flush, so the walk is bounded by the largest flow count the
datapath has held. The softirq backlog the callback used to absorb has
no bound at all.
-
CVE-2026-99024
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
optee: register TEE devices only once fully initialized
optee_probe() called tee_device_register() on both the client and the
supplicant device before the rest of struct optee had been initialized.
tee_device_register() calls cdev_device_add(), which does two things at
once: it creates /dev/tee0 and /dev/teepriv0, and it links the device
into the tee class so that class_find_device() can find it. From that
moment on the device is reachable both from user space via tee_open()
and from kernel space via tee_client_open_context(). The only gate in
teedev_open() is tee_device_get(), which merely checks that
teedev-desc is non-NULL, which was already set by tee_device_alloc().
Therefore, there is effectively no gate at all.
A context opened in that window can run against a struct optee where
- optee->call_queue.mutex is not initialized by optee_cq_init()
- optee->supp mutex and completions is not initialized by
optee_supp_init()
- optee->rpmb_dev_mutex is not initialized yet,
- the message argument cache is not initialized by
optee_shm_arg_cache_init()
- optee->ctx is still NULL
Any open session or invoke during this window takes uninitialized
mutexes and dereferences a NULL pointer.
Fix it by moving both tee_device_register() calls down to the point
where all of struct optee is set up.
[JW: adding fixes and cc]
-
CVE-2026-99017
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
riscv: errata: select RISCV_ALTERNATIVE_EARLY for THEAD GHOSTWRITE
errata_probe_ghostwrite() only marks the CPU as vulnerable when it is
called at the RISCV_ALTERNATIVES_EARLY_BOOT stage, and the mitigation
in ghostwrite_enable_mitigation() only takes effect if the vulnerable
state was recorded before the boot alternatives are patched.
The early boot stage is only executed when CONFIG_RISCV_ALTERNATIVE_EARLY
is enabled, which currently is only selected by ERRATA_THEAD_MAE. In a
kernel built with ERRATA_THEAD_GHOSTWRITE=y but ERRATA_THEAD_MAE=n, the
early alternatives never run, so ghostwrite_state is never set to
VULNERABLE: ghostwrite_enable_mitigation() returns false, xtheadvector
remains enabled, and sysfs reports "Not affected" on a vulnerable
T-Head C9xx CPU.
Select RISCV_ALTERNATIVE_EARLY from ERRATA_THEAD_GHOSTWRITE so that the
vulnerability is detected in the early boot stage and the default
mitigation is applied.
-
CVE-2026-99013
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
Bluetooth: ISO: Reject concurrent BIS listener setup
iso_sock_listen() drops the socket lock before calling iso_listen_bis()
to preserve the hdev-lock ordering. Two concurrent listen() calls on
the same broadcast socket can therefore both observe BT_BOUND. Caller
A can attach connA and release the locks, then caller B can acquire
them and attach connB before A publishes BT_LISTEN.
__iso_chan_add() checks whether connB already has a socket, but not
whether the socket already has a connection. The second attachment
overwrites iso_pi(sk)->conn, leaking the connA reference and leaving
connA->sk intact. Socket teardown follows connB, so later events on
connA can cause a stale socket access.
Recheck BT_BOUND and require iso_pi(sk)->conn to be NULL after taking
both the HCI device and socket locks in iso_listen_bis(). The connection
check also covers the interval between attachment and publication of
BT_LISTEN. Reject a competing setup with -EBADFD through the existing
unlock path before allocating another connection.
A recorded KASAN report from an instrumented kernel shows:
BUG: KASAN: slab-use-after-free in iso_sock_hold+0x119/0x1f0
Call Trace:
iso_sock_hold+0x119/0x1f0
iso_conn_del+0xe4/0x270
hci_disconn_complete_evt+0x351/0x8c0
hci_event_packet+0x71b/0xb20
hci_rx_work+0x293/0x730
Allocated by task 93:
sk_prot_alloc+0x113/0x220
sk_alloc+0x2b/0x6d0
bt_sock_alloc+0x29/0x370
iso_sock_alloc.constprop.0+0x19/0x300
iso_sock_create+0x94/0x100
Freed by task 93:
kfree+0x121/0x3c0
__sk_destruct+0x42b/0x540
iso_sock_release+0x29d/0x340
__sock_release+0xa1/0x260
sock_close+0x10/0x20
-
CVE-2026-99012
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
Bluetooth: ISO: Serialize concurrent connect calls
iso_sock_connect() checks the socket state before taking the socket lock
and drops the lock again before setting up a connection. Two callers can
both pass the admission check while the socket is open or bound.
Caller A can copy its destination for route selection, then caller B can
replace the socket destination before A binds or connects the CIS. A
then uses B's destination with the route selected for its own request.
B can also proceed after A attaches a connection and sets BT_CONNECT.
For deferred BIS setup, B can bind a second BIS and overwrite
iso_pi(sk)-conn, leaving the first connection's reference and conn->sk
back-pointer stranded. Concurrent CIS connects can likewise replace the
socket's connection. Later teardown only detaches the current connection,
so a callback on the old connection can race with socket release and
access a freed socket.
KASAN reported:
BUG: KASAN: slab-use-after-free in iso_sock_hold+0xf7/0x1b0
Call Trace:
iso_sock_hold+0xf7/0x1b0
iso_conn_del+0x7b/0x1d0
iso_connect_cfm+0x186/0x16a0
hci_conn_failed+0x154/0x280
hci_abort_conn_sync+0x3dc/0x7d0
hci_cmd_sync_work+0x173/0x300
Allocated by task 121:
sk_alloc+0x2b/0x6d0
bt_sock_alloc+0x29/0x370
iso_sock_alloc.constprop.0+0x19/0x300
iso_sock_create+0x94/0x100
Freed by task 121:
kfree+0x121/0x3c0
__sk_destruct+0x42b/0x540
iso_sock_release+0x29d/0x340
__sock_release+0xa1/0x260
Check admission under the socket lock and mark the connect operation in
progress before publishing its destination. Keep that flag set across
route lookup and connection setup, rejecting another connect with
-EBADFD before it can change the destination or attach a connection.
Clear the flag after either helper returns, including on failure, so a
failed setup can be retried.
A separate flag is needed because the socket lock must be released before
acquiring the HCI device lock, while BT_CONNECT requires an attached
connection. Keep the existing state transitions and the deferred CIS
completion through iso_sock_recvmsg().
Reject an existing socket attachment in __iso_chan_add() before taking
a new reference or publishing either pointer, so every caller preserves
the connection association. Keep the same-socket, same-connection success
case for deferred CIS setup. Reject an attached socket before BIS setup
so a reusable BIS is not claimed before the attachment is rejected.
In CIS setup, reject a different HCI connection before iso_conn_add()
and drop the per-attempt HCI hold directly. An unowned iso_conn may still
carry another reference after detachment, so its destruction cannot be
relied on to release that hold.
iso_listen_bis() creates a fresh PA HCI/ISO pair under the device lock,
so the temporary iso_conn_put() frees the ISO candidate and drops its
HCI hold on rejection. iso_conn_ready() uses a freshly allocated,
unpublished child with no connection, making the socket-side rejection
unreachable.
-
CVE-2026-99006
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
drm/i915/gem: Prevent overstepping exec array boundary
It is possible for eb_relocate_parse_slow() to call kvfree() on
entries outside the original exec array during following
scenario inside eb_relocate_parse_slow():
1) eb_copy_relocations() allocates as many as eb-buffer_count
copies for exec entries;
2) eb_parse() appends up to two VMAs, right after incrementing
eb->buffer_count;
3) cleanup under "out" label uses this incremented buffer_count
to release relocation pointers from exec array via kvfree(),
not from VMA array.
To amend this problem, ensure that relocation cleanup uses the
original exec object count, excluding the VMAs appended by the
command parser.
(cherry picked from commit 2a0158925fb943c7ba8f3216e2734952128defc9)
-
CVE-2026-99003
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
drm/panthor: validate userspace queue count against initialized firmware slot count
In panthor_drv.c:panthor_ioctl_group_create(),
the queue count is validated against the hardcoded maximum:
if (!args-queues.count || args->queues.count > MAX_CS_PER_CSG)
return -EINVAL;
However, if the firmware reports fewer slots than MAX_CS_PER_CSG,
panthor_fw_init_ifaces() will only initialize that smaller number of
stream interfaces.
If unprivileged userspace requests more queues than the firmware reported
(but no more than MAX_CS_PER_CSG), cs_slot_prog_locked() fetches an
uninitialized stream interface and may crash when dereferencing
cs_iface->input->ringbuf_base.
-
CVE-2026-99000
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
drm/xe: Fix shrinker accounting double-subtraction on nested external pins
xe_bo_pin_external() and xe_bo_unpin_external() support nested pins,
since the underlying ttm_bo_pin()/ttm_bo_unpin() maintain a pin_count
refcount rather than a boolean. However, the shrinker accounting calls
xe_ttm_tt_account_subtract() and xe_ttm_tt_account_add() are invoked
unconditionally on every pin_external()/unpin_external() call, instead
of only on the transition into and out of the pinned state.
An external BO (dma-buf) can be pinned more than once while already
pinned, for example when it has multiple attachments/importers, each
independently calling xe_bo_pin_external(). Each such nested pin
subtracts the BO's pages from the shrinker's accounting again, even
though the BO's pages were already removed from consideration by the
first pin. Symmetrically, each nested unpin adds them back again
before the BO is actually unpinned. This desynchronizes the
shrinker's page and object counts from reality, and since they are
declared as signed long but interpreted as unsigned long in
xe_shrinker_count(), the underflow can turn into an enormous
shrinkable/purgeable page count, causing the shrinker to be invoked
excessively under memory pressure.
Guard the accounting calls with the same pin-count transition checks
already used to guard the pinned_link list maintenance, so accounting
is only updated on the outermost pin and the final unpin.
While centralizing this logic into xe_bo_account_pin()/
xe_bo_account_unpin(), also close a second, related accounting gap
affecting imported dma-bufs. For an imported bo's ttm_tt, xe's
ttm_tt_populate hook (xe_ttm_tt_populate()) early-returns without
calling xe_ttm_tt_account_add(), since TTM_TT_FLAG_EXTERNAL is set
without TTM_TT_FLAG_EXTERNAL_MAPPABLE. However, TTM core's
ttm_tt_populate() wrapper unconditionally marks the tt as populated
on driver-hook success, so ttm_tt_is_populated() still reports true
for it. xe_bo_account_pin()/xe_bo_account_unpin() trusted
ttm_tt_is_populated() as a proxy for "these pages are accounted for
by the shrinker", and would subtract/add pages for such a tt that
were never added in the first place, again underflowing the
shrinker's counts. This is reachable today via xe_bo_pin_external(),
and would have become newly reachable via the fb-pin path once this
patch routes __xe_pin_fb_vma() through xe_bo_pin_account() below, so
fix it here rather than in a later patch, to avoid a commit in this
series that introduces the regression before fixing it. Skip the
accounting for imported bos, using the same EXTERNAL &&
!EXTERNAL_MAPPABLE test xe_ttm_tt_populate()/xe_ttm_tt_unpopulate()
use to decide whether to act, via a new xe_ttm_bo_is_imported()
helper (hoisted up from its only prior use site).
v2:
- Fix the critical issue caught by Sashiko AI review: the pin-count
transition check introduced here was still bypassed by the fb-pin
path (__xe_pin_fb_vma()/__xe_unpin_fb_vma()), which could leave a
stale pinned_link entry and a permanent shrinker accounting leak
when the fb-unpin performed the true last unpin. Addressed in a
separate follow-up patch in this series rather than folded in here,
since it is a distinct bug with its own Fixes: tag.
- Drop the unneeded local "last_unpin" variable in
xe_bo_unpin_external(), restoring the original inline
bo-ttm.pin_count == 1 check; it was a no-op rename not used by any
of the new accounting helpers, and was going to be deleted again by
the follow-up patch anyway.
v3:
- Exclude imported dma-bufs from the pin/unpin shrinker accounting,
fixing another underflow caught by Sashiko AI review.
- Add a short comment on the backup-object pin site noting it's
already populated/accounted at pin time, addressing (as a
false-positive clarification, no code fix needed) a separate
Medium-severity accounting-leak concern also caught by Sashiko AI
---truncated---
-
CVE-2026-98999
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
drm/xe: Fix stale pinned_link entry when fb-pin performs the final unpin
xe_bo_pin_external() and xe_bo_unpin_external() maintain the bo's
pinned_link list membership in xe-pinned.late.external, based on
whether the current call is the outermost pin or the final unpin.
However, the same external bo's pin_count can also be raised and
lowered directly by __xe_pin_fb_vma()/__xe_unpin_fb_vma(), which pin
the bo as a display scanout buffer without going through
xe_bo_pin_external()/xe_bo_unpin_external() at all, and have no
notion of, or ownership over, pinned_link.
If a bo is pinned both externally (e.g. dma-buf export) and as an fb,
and the external unpin happens first, xe_bo_unpin_external() correctly
observes pin_count > 1 and leaves the bo on pinned_link. When the fb
unpin later performs the true last unpin (pin_count 1 -> 0), it never
touches pinned_link, leaving the bo linked on xe->pinned.late.external
indefinitely. Once the bo is subsequently freed, this stale list entry
points into freed memory, corrupting the list and risking a
use-after-free the next time the list is walked or spliced.
The backup object pin/unpin sites in xe_bo_notifier_prepare_pinned()/
xe_bo_notifier_unprepare_pinned() have the same bypass characteristic,
though they never add their bo to a pinned list, so are not affected
by this particular list-corruption issue.
Move the pinned_link removal into xe_bo_account_unpin(), which already
runs on every unpin path (kernel, external, framebuffer, backup
object) right before the true 1 -> 0 pin_count transition. Since
list_del_init() only operates on the node itself, this removal is
list-agnostic and safe to perform regardless of which list (external
or kernel_bo_present) the bo happens to be linked on, or which code
path is performing the final unpin. Drop the now-redundant explicit
list_del_init() calls in xe_bo_unpin_external() and xe_bo_unpin().
v2:
- New patch
v3:
- Rebased on changes to previous patch.
(cherry picked from commit ca58d36a1f93af586ee93b340c9f9fd81312cc01)
-
CVE-2026-98998
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
drm/bridge: th1520-dw-hdmi: Fix remove() callback
This driver stores struct th1520_hdmi * in platform drvdata, but
th1520_dw_hdmi_remove() was reading it back as struct dw_hdmi *
and passing it to dw_hdmi_remove(), so teardown runs on the wrong
pointer.
Retrieve struct th1520_hdmi * from platform drvdata and pass
hdmi-dw_hdmi to dw_hdmi_remove().
-
CVE-2026-98997
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
drm/amdgpu/userq: only accept doorbell BOs as queue doorbell
amdgpu_userq_get_doorbell_index() pins any BO passed as doorbell_handle
into the doorbell domain. For a regular BO, TTM moves it with
ttm_bo_move_null() and its contents are lost. Since an importer on the
same device gets the exporter's GEM object, a client can do this to a
buffer another client shared with it.
Reject BOs that were not created in the doorbell domain.
v2: drop the error message, use a local abo variable (Christian)
(cherry picked from commit f7497bbc19db4821c62b9d97b31c9086feb2d917)
-
CVE-2026-98996
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
fs: don't silently unmount busy mounts
The propagate_mount_busy() helper exists to decide whether a synchronous
can suceed. It does that by looking at the copies of the victims in the
mounts its parent propagates to. One of those functions from propagation
hell.
It skips a copy that has child mounts except for the case of a single
child mount covering the copy's root. That algorithm used to be exact.
It isn't anymore.
Originally only childless copies were unmounted during propagation.
And 1064f874abc0 ("mnt: Tuck mounts under others instead of creating
shadow/side mounts.") ended up adding covered mounts to both sides in
one go.
Along came 99b19d16471e ("mnt: In propgate_umount handle visiting mounts
in any order") and that started to unmount more than before:
[1]: A copy of a mount is unmounted when each of its children is either
its overmount or a copy of the victim itself. In the former case
the overmount gets reparented. In the latter case the copy of the
victim will end up being unmounted.
So propagate_umount() ended up being rewritten without adjusting
propagate_mount_busy(). And currently propagate_mount_busy() doesn't
look at copies of the victim that have two or more children nor a single
child that is not the overmount.
So any such copy in [1] is not checked for
references. The result is that a synchronous unmount succeeds, pulling
out a mount that is still in use.
A container can run into this with its own mounts. The host shares a
tree with the container and the host has a mount in that tree:
mount -t tmpfs none /mnt
mount --make-shared /mnt
mount -t tmpfs none /mnt/a
The container's /mnt is a slave copy of the host mount. The container
now takes a detached copy of the tree and mounts that copy beneath its
propagated copy of /mnt/a and keeps the file descriptor to the tree
open:
unshare -m --propagation unchanged
mount --make-rslave /
fd = open_tree(AT_FDCWD, "/mnt", OPEN_TREE_CLONE | AT_RECURSIVE)
move_mount(fd, "", AT_FDCWD, "/mnt/a", MOVE_MOUNT_BENEATH)
grep /mnt /proc/self/mountinfo
775 355 ... /mnt ... master:646
777 775 ... /mnt/a ... master:646
776 777 ... /mnt/a ... master:647
778 777 ... /mnt/a/a ... master:647
The now attached tree (777) is now mounted beneath the container's copy
of /mnt/a. That's also where the host mount sits (as seen from the host
mount namespace). So the propagated copy (776) of the host's mount is on
top of its root and the copy of that (778) is inside it. Now the host
runs:
umount /mnt/a
and it succeeds. The moved tree tree and the copy inside of it are gone
from the container. Now only the propagated copy is left and it got
reparented to where it was before:
grep /mnt /proc/self/mountinfo
775 355 ... /mnt ... master:646
776 775 ... /mnt/a ...
But the file descriptor still refers to the moved tree. So a synchronous
umount detached a mount that is still in use.
Teach propagate_mount_busy() the same checks that propagate_umount()
does. It now checks for a single victim mount without children
for which it can assume that MNT_LOCKED is cleared on every copy like
propagate_umount() does. The copies form chains. IOW, when a copy itself
receives propagation from the victim's parent then the mount at the same
mountpoint inside that copy is a copy as well. Whether a mount receives
propagation from the victim's can simply be checked by walking its
masters. Walk each chain once from its top so we are sure that nested
copies stay linear. Check every reference once of every copy that gets
unmounted.
-
CVE-2026-98991
•
published on October 23, 2026
In the Linux kernel, the following vulnerability has been resolved:
namespace: queue a mount only once for mount notifications
mnt_notify_add() puts a mount on notify_list. It doesn't check whether
the mount is on the list already. That's fine as long as every mount is
queued at most once per namespace_sem hold. It isn't.
propagate_umount() moves a surviving overmount off a stack of mounts
that are going away and queues it as moved. shrink_submounts() and
mark_mounts_for_expiry() call umount_tree() for several mounts under a
single namespace_sem hold. So the second umount_tree() can take down
exactly the mount the first one reparented, reparent it once more and so
end up queueing it a second time.
The second list_add_tail() cuts the mounts queued in between out of the
list while the head still points at the last of them. notify_mnt_list()
then only visits that one mount and the others are freed after the grace
period with notify_list still pointing at them. Every later mount
operation on the host walks freed memory:
BUG: KASAN: slab-use-after-free in __list_add_valid_or_report
Read of size 8 at addr ffff8881003d57e0 by task notify_dq/162
__list_add_valid_or_report+0x15c/0x1a0
mnt_add_to_ns+0x322/0x890
attach_recursive_mnt.isra.0+0xf2a/0x1a00
path_mount+0x139c/0x1d40
This needs a mount with MNT_SHRINKABLE to bind from. That's what
finish_automount() creates (submounts of NFS, AFS and CIFS and the
tracefs mount below debugfs). Bind mounts inherit the flag. A mount
namespace with a FAN_MARK_MNTNS mark does the rest and root in a user
namespace can have both.
Initialize to_notify in alloc_vfsmnt() and leave a mount alone that is
queued already. mnt_notify() looks at the state the mount has when the
list is drained so a single entry per mount is enough. The move event
for a mount that is unmounted under the same hold is lost. That's fine.
The detach is what matters.