Zygo Blaxell [Sun, 18 Oct 2026 19:50:43 +0000 (15:50 -0400)]
btrfs: allocator hints: make the available space estimates honor the hints
The chunk allocator keeps data chunks off METADATA_ONLY and NONE_ONLY
devices and metadata and system chunks off DATA_ONLY and NONE_ONLY
devices (gather_device_info()). The available space estimates do not:
btrfs_update_per_profile_avail() simulates chunk allocation over every
writable device, and statfs adds up every device's unallocated space.
The space of a device a hint excludes is counted as available although
no chunk of that type can ever be made from it.
Metadata overcommit trusts that estimate (calc_available_free_space()),
so once the devices metadata may use are full, reservations are granted
against chunks that cannot be allocated, and the transactions that use
them abort. Reproducer (raid56-wh-harness alloc-hints-space-test.sh):
raid1 data and metadata on three devices of 6, 3 and 6 GiB, devid 3 set
to NONE_ONLY (5) and emptied with a full balance, then data written
until ENOSPC (devid 2 fills while devid 1 still has 3 GiB) and small
files created until ENOSPC:
BTRFS: error (device loop2 state A) in btrfs_finish_one_ordered:3410: errno=-28 No space left
BTRFS error (device loop2 state EA): Transaction 29 aborted (-ENOSPC)
BTRFS info (device loop2 state EA): forced readonly
and, after a remount, a second abort from __btrfs_update_delayed_inode()
while the files were read back. The files written since the last
commit were lost. statfs reported 3103 MiB available at the data ENOSPC.
The same filesystem without devid 3 (6 + 3 GiB, no hints) fills
metadata to a clean ENOSPC and reports 55 MiB.
Give the estimates the allocator's view of the hints:
- btrfs_dev_alloc_hint() decides whether a chunk type may use a device
and returns its tier; btrfs_select_hint_tiers() keeps the tiers a
chunk would use. gather_device_info() now uses both instead of its
open-coded copy, so the rules live in one place.
- The per-profile estimate is kept separately for data and for metadata
and system chunks; alloc_virtual_chunk() skips the devices the hints
exclude for the type and picks hint tiers the way the allocator does.
btrfs_get_per_profile_avail() takes the block group type. Mixed block
groups ignore the hints, as in the allocator.
- statfs leaves out devices that data chunks may not use. Only the
available space changes: f_bavail (df Avail) leaves out the
unallocated space of those devices, while f_blocks (df Size) and
f_bfree (and so df Used) still count every device, including the ones
the hints keep data off. On a filesystem with a METADATA_ONLY device,
for example, Size minus Used is larger than Avail by about that
device's unallocated space.
- A type change recomputes the estimate.
PREFERRED_NONE devices still count: the allocator uses them as the last
tier. Without CONFIG_BTRFS_ALLOCATOR_HINTS the data and metadata
estimates are identical and only one is computed.
The estimates have been hint-blind since "btrfs: add allocator_hint
mode", earlier in this series, taught the allocator the hints.
The fallback to fs_info->free_chunk_space (zoned filesystems, or a
failed estimate) and the reclaim thresholds that read it stay hint-blind.
With this change the reproducer fills metadata to a clean ENOSPC with no
abort and no lost files, and statfs reports the space data can really
use.
6.18 port: needs the per-profile available space estimate backported
first (upstream 52fead5eb8a7, c84053d9f7f7, 2672a26a7551, v7.1); before
it, calc_available_free_space() divided fs_info->free_chunk_space, the
unallocated space of every device, by the profile's copy count, which is
hint-blind and also wrong for unequal devices without hints.
Zygo Blaxell [Sun, 4 Oct 2026 14:55:00 +0000 (10:55 -0400)]
btrfs: allocator hints: don't run a transaction from the type store
btrfs_devinfo_type_store() starts and commits a transaction while kernfs
holds the attribute active. btrfs_rm_device() removes the device's
sysfs directory, draining that attribute, while it holds its own
transaction handle and fs_devices->device_list_mutex. A type write
racing a device remove can deadlock:
rm_device: transaction handle, device_list_mutex -> drain devinfo/<id>/type
type store: devinfo/<id>/type active -> commit waits for rm_device's handle
Lockdep reports the cycle (kn->active -> btrfs_trans_num_writers ->
btrfs_trans_num_extwriters -> device_list_mutex) with a type writer
looping during 'btrfs device remove'. Joining the transaction without
committing does not help: a chunk allocation during the dev item update
makes the end of the transaction take device_list_mutex.
Do what the label and feature stores do and keep transactions out of
sysfs. The store sets device->type under chunk_mutex (which also orders
it against chunk allocation and against the end of a device replace),
marks the device BTRFS_DEV_STATE_TYPE_DIRTY and requests a commit.
btrfs_update_dirty_device_types() writes the dirty dev items at the start
of the next commit, under device_list_mutex, after the transaction
handle, the same order as btrfs_rm_device() and btrfs_run_dev_stats(), with the
chunk tree space reserved beforehand (btrfs_reserve_chunk_metadata(), as
btrfs_grow_device() does; it takes chunk_mutex, so before the mutex). A
dev item already deleted by a device remove is skipped. A dev item that
cannot be written stays dirty, with a rate-limited warning, and is retried
at the next commit: a hint is not worth aborting the transaction for. The
hint takes effect for new chunks at once and reaches the disk at the next
commit, sync or unmount.
Taking chunk_mutex in the store also closes the race at the end of a
device replace: the finish moves the hint and the dirty bit to the
target under the same mutex, so a hint change on the source either
reaches the target or, once the source has the replace devid, returns
-EBUSY. The replace target (devid 0, no dev item) still returns -EBUSY.
Also check sb_rdonly() before BTRFS_FS_OPEN, so a read-only mount
reports -EROFS rather than -EBUSY, and refuse seed devices, which are
read-only and not written by the commit-time update.
The store orders the new type before the dirty bit
(smp_mb__before_atomic(); the commit's test_and_clear_bit() is the other
half), and btrfs_update_device() reads the type with READ_ONCE(), so the
commit cannot write the old type and clear the bit on a weakly ordered
machine.
Zygo Blaxell [Sun, 4 Oct 2026 04:44:58 +0000 (00:44 -0400)]
btrfs: allocator hints: keep the hint across device replace
Device replace drops the allocation hint, and changing a hint during a
replace can abort the transaction.
The replace target takes over the source's dev item when the replace
finishes, but not the in-memory device->type. The on-disk item keeps the
hint only until the next btrfs_update_device() on the new device (any
chunk allocation on it), which writes type 0 over it. Copy the type
together with the devid, under the same locks.
During a replace, the target is visible as devinfo/0 and has no dev item.
After the finish, the source's devid is BTRFS_DEV_REPLACE_DEVID until its
sysfs directory goes away. A write to 'type' on either device makes
btrfs_update_device() miss the dev item (-ENOENT), and the store aborts
the transaction, forcing the filesystem read-only in the middle of the
replace. Device remove has the same window between deleting the dev
item and removing the sysfs directory.
Return -EBUSY for the replace target and for a device with the replace
devid. An -ENOENT that still gets past these checks (remove, or a
replace that finishes after them) ends the transaction without aborting
it and returns -ENODEV.
The source's hint can still be changed during a replace, and the target
inherits it at the finish, so a replace can move to a device of a
different type, e.g. SSD to HDD. A write that races the finish can be
lost: it misses the dev item and returns -ENODEV, or the finish has
already copied the old type. That is tolerable, and the next patch
closes the window.
Zygo Blaxell [Sun, 4 Oct 2026 04:44:02 +0000 (00:44 -0400)]
btrfs: allocator hints: define the hint values unconditionally in the UAPI header
The BTRFS_DEV_ALLOCATION_* values sit in include/uapi/linux/btrfs_tree.h
inside #ifdef CONFIG_BTRFS_ALLOCATOR_HINTS, to keep them away from
userspace until the interface is settled. A CONFIG_ symbol in a UAPI
header is rejected by scripts/headers_install.sh:
error: include/uapi/linux/btrfs_tree.h: leak CONFIG_BTRFS_ALLOCATOR_HINTS to user-space
so 'make headers_install' (and the deb/rpm package targets, which run
it) fail with the patch set applied.
The values are on-disk format: they give the meaning of the low bits of
dev_item.type, and userspace (btrfs-progs) needs them to read and set a
device's hint. Defining them changes no userspace-visible structure.
So drop the guard and always export the values. The code that acts on
the hints stays under CONFIG_BTRFS_ALLOCATOR_HINTS.
KeinNiemand [Sun, 30 Aug 2026 12:50:54 +0000 (14:50 +0200)]
btrfs: honor allocator hints for fixed-width mirrors
gather_device_info() admits complete allocation-hint groups until it
has at least devs_min devices. It then clears the hints and sorts the
admitted devices globally by max_avail. For a fixed-width mirrored
profile, that second sort can move preferred devices past the later
devs_max cutoff.
RAID1, RAID1C3 and RAID1C4 currently use exactly devs_min devices.
Limit the hint-ordered prefix to that fixed count before clearing the
hints. Devices within each hint group are already ordered by max_avail,
so this keeps the best devices needed from the final group.
The existing second sort then restores the max_avail ordering expected
by stripe-size calculation over the selected devices only.
Check that devs_min and devs_max match instead of relying on the
current RAID1 profile definitions implicitly. If a future RAID1 profile
becomes variable-width, it will retain the existing whole-group
behavior.
Leave all other profiles unchanged, in particular the wide striped
profiles that need the complete final candidate group.
A preferred device remains eligible while it has at least
dev_extent_min free. It can therefore constrain stripe_size and produce
a small block group. This behavior already exists when the preferred
tier alone satisfies the profile; this change intentionally extends it
to fixed-width mirrors assembled from multiple hint tiers.
v2:
- Document the possible chunk-size constraint from a nearly full
preferred device.
Kai Krakow [Fri, 10 Jul 2026 21:51:46 +0000 (23:51 +0200)]
btrfs: add allocator_hint to disable allocation completely
This is useful where you want to prevent new allocations of chunks to a
set of multiple disks which are going to be removed from the pool. This
acts as a multiple `btrfs dev remove` on steroids that can remove
multiple disks in parallel without moving data to disks which would be
removed in the next round. In such cases, it will avoid moving the same
data multiple times, and thus avoid placing it on potentially bad
disks.
The restriction applies to non-mixed block groups. Mixed data and
metadata block groups intentionally ignore allocator hints. Device
replace does not copy the hint to the target device, so reapply it
after replacement.
v2:
- Clarify the mixed-block-group scope and device replace behavior.
Thanks to @Zygo for the explanation and suggestion.
Kai Krakow [Fri, 10 Jul 2026 22:09:55 +0000 (00:09 +0200)]
btrfs: add allocator_hint for no allocation preferred
This is useful where you want to avoid new allocations of chunks on a
disk which is going to be removed from the pool anyways, e.g. due to
bad blocks or because it's slow.
The preference applies to non-mixed block groups. Mixed data and
metadata block groups intentionally ignore allocator hints.
v2:
- Clarify that mixed block groups are outside the allocator-hints
scope.
Kai Krakow [Fri, 10 Jul 2026 22:18:02 +0000 (00:18 +0200)]
btrfs: add allocator_hint mode
When this mode is enabled, the chunk allocation policy is modified as
follows:
Each disk may have a different tag:
- BTRFS_DEV_ALLOCATION_PREFERRED_METADATA
- BTRFS_DEV_ALLOCATION_METADATA_ONLY
- BTRFS_DEV_ALLOCATION_DATA_ONLY
- BTRFS_DEV_ALLOCATION_PREFERRED_DATA (default)
Where:
- ALLOCATION_PREFERRED_X means that it is preferred to use this disk
for the X chunk type (the other type may be allowed when the space is
low)
- ALLOCATION_X_ONLY means that it is used *only* for the X chunk type.
This means also that it is a preferred choice.
Each time the allocator allocates a chunk of type X, first it takes the
disks tagged as ALLOCATION_X_ONLY or ALLOCATION_PREFERRED_X.
If the space is not enough, it uses also the disks tagged as
ALLOCATION_METADATA_ONLY.
If the space is not enough, it uses also the other disks, with the
exception of the one marked as ALLOCATION_PREFERRED_Y, where Y is the
other type of chunk (i.e. not X).
Mixed data and metadata block groups intentionally ignore all hints. They
cannot separate the two chunk classes and are outside the tiering use case
targeted by this feature.
The current type-based interface also has two known limitations: device
replace does not copy hints to the target device, and hint updates are not
serialized with concurrent chunk allocation.
v2:
- Document the mixed-block-group scope and the limitations of the
type-based interface.
Co-authored-by: Goffredo Baroncelli <kreijack@inwind.it> Signed-off-by: Kai Krakow <kai@kaishome.de>
Kai Krakow [Fri, 10 Jul 2026 22:25:21 +0000 (00:25 +0200)]
btrfs: change the DEV_ITEM 'type' field via sysfs
Known limitations:
- Device replace does not copy the allocation hint to the target
device. Users must reapply it after replacement.
- The sysfs update is not serialized with chunk allocation through
chunk_mutex, so an allocation concurrent with an update may observe
either value.
A future revision should replace the type field with dedicated
transactional structures. Transactions should not be triggered directly
from sysfs writes, see:
Kai Krakow [Fri, 10 Jul 2026 22:51:17 +0000 (00:51 +0200)]
btrfs: add flags to give a hint to the chunk allocator
Add the following flags to give a hint about which chunk should be
allocated on which a disk.
The following flags are created:
- BTRFS_DEV_ALLOCATION_PREFERRED_DATA
preferred data chunk, but metadata chunk allowed
- BTRFS_DEV_ALLOCATION_PREFERRED_METADATA
preferred metadata chunk, but data chunk allowed
- BTRFS_DEV_ALLOCATION_METADATA_ONLY
only metadata chunk allowed
- BTRFS_DEV_ALLOCATION_DATA_ONLY
only data chunk allowed
Co-authored-by: Goffredo Baroncelli <kreijack@inwind.it> Signed-off-by: Kai Krakow <kai@kaishome.de>
Cheng Yongkang [Mon, 26 Oct 2026 10:51:09 +0000 (03:51 -0700)]
btrfs: remove devid kobject when deleting a missing device
btrfs_rm_device() removes the sysfs device entry only when
device->bdev_file is non-NULL. A missing device (created by
add_missing_dev() for an absent disk) has bdev_file == NULL, but
btrfs_sysfs_add_device() still registers its devid kobject under
/sys/fs/btrfs/<uuid>/devinfo/<devid>.
When such a device is deleted, btrfs_free_device() frees the
btrfs_device while the devid kobject remains registered. The kobject
is embedded in the freed btrfs_device and its show functions use
container_of() to reach it, so reading any devinfo attribute is a
use-after-free:
BUG: KASAN: slab-use-after-free in sysfs_kf_seq_show+0x157/0x490
Read of size 8 at addr ffff88801d8172a8 by task poc/9385
...
Freed by task 9385:
kfree
btrfs_rm_device+0x11ee/0x17b0
btrfs_ioctl_rm_dev+0x265/0x400
Allocated by task 9390:
btrfs_alloc_device+0x7e/0x400
device_list_add
btrfs_scan_one_device
btrfs_control_ioctl
The freed object is kmalloc-1k and the faulting access is 680 bytes
inside it (offsetof(struct btrfs_device, devid_kobj) = 640 plus
offsetof(struct kobject, ktype) = 40), confirming the embedded
devid_kobj.
btrfs_sysfs_remove_device() already skips the block-device link when
device->bdev is NULL and only touches the devid kobject when it is
initialized. Call it for missing devices too.
As can_overcommit() simply uses unallocated space with factor to
calculate the allocatable metadata chunk size, resulting 25.5GiB
available space.
But in reality we can only allocate one 1GiB RAID1 chunk, the remaining
49GiB on devid 2 will never be utilized to fulfill a RAID1 chunk.
This leads to various ENOSPC related transaction abort and flips the fs
read-only.
Now use per-profile available space in calc_available_free_space(), and
only when that failed we fall back to the old factor based estimation.
And for zoned devices or for the very low chance of temporary memory
allocation failure, we will still fallback to factor based estimation.
But I hope in reality it's very rare.
Reviewed-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: Qu Wenruo <wqu@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
(cherry picked from commit 2672a26a75517a2b4409f2a379ee2bb5c38cff06)
Qu Wenruo [Wed, 4 Feb 2026 02:54:07 +0000 (13:24 +1030)]
btrfs: update per-profile available estimation
This involves the following timing:
- Chunk allocation
- Chunk removal
- After Mount
- New device
- Device removal
- Device shrink
- Device enlarge
And since the function btrfs_update_per_profile_avail() will not return
an error, this won't cause new error handling path.
Although when btrfs_update_per_profile_avail() failed (only ENOSPC
possible) it will mark the per-profile available estimation as
unreliable, so later btrfs_get_per_profile_avail() will return false and
require the caller to have a fallback solution.
The function btrfs_update_per_profile_avail() will be executed with
chunk_mutex hold, thus it will slightly slow down those involved
functions, but not a lot.
As all the core workload is just various u64 calculations inside a loop,
without any tree search, the overhead should be acceptable even for all
supported 9 profiles.
For 4 disks (which exercises all 9 profiles), the execution time of that
function will still be less than 10 us.
Reviewed-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: Qu Wenruo <wqu@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
(cherry picked from commit c84053d9f7f758c79320716caa098cafc70a74da)
Qu Wenruo [Wed, 4 Feb 2026 02:54:06 +0000 (13:24 +1030)]
btrfs: introduce the device layout aware per-profile available space
[BUG]
There is a long known bug that if metadata is using RAID1 on two disks
with unbalanced sizes, there is a very high chance to hit ENOSPC related
transaction abort.
[CAUSE]
The root cause is in the available space estimation code:
- Factor based calculation
Just use all unallocated space, divide by the profile factor
One obvious user is can_overcommit().
If using factor based estimation, we can use (1GiB + 50GiB) / 2 = 25.5GiB
free space for metadata.
Thus we can continue allocating metadata (over-commit) way beyond the
1GiB limit.
But this estimation is completely wrong, in reality we can only allocate
one single 1GiB RAID1 block group, thus if we continue over-commit, at
one time we will hit ENOSPC at some critical path and flips the fs
read-only.
[SOLUTION]
This patch will introduce per-profile available space estimation,
which can provide chunk-allocator like behavior to give a (mostly)
accurate result, with under-estimate corner cases.
There are some differences between the estimation and real chunk
allocator:
- No consideration on hole size
It's fine for most cases, as all data/metadata strips are in 1GiB size
thus there should not be any hole wasting much space.
And chunk allocator is able to use smaller stripes when there is
really no other choice.
Although in theory this means it can lead to some over-estimation, it
should not cause too much hassle in the real world.
The other benefit of such behavior is, we avoid dev-extent tree search
completely, thus the overhead is very small.
- No true balance for certain cases
If we have 3 disks RAID1, and each device has 2GiB unallocated space,
we can load balance the chunk allocation so that we can allocate 3GiB
RAID1 chunks, and that's what chunk allocator will do.
But this current estimation code is using the largest available space
to do a single allocation. Meaning the estimation will be 2GiB, thus
under estimate.
Such under estimation is fine and after the first chunk allocation, the
estimation will be updated and still give a correct 2GiB
estimation.
So this only means the estimation will be a little conservative, which
is safer for call sites like metadata over-commit check.
With that facility, for above 1GiB + 50GiB case, it will give a RAID1
estimation of 1GiB, instead of the incorrect 25.5GiB.
Or for a more complex example:
devid 1 unallocated: 1T
devid 2 unallocated: 1T
devid 3 unallocated: 10T
We will get an array of:
RAID10: 2T
RAID1: 2T
RAID1C3: 1T
RAID1C4: 0 (not enough devices)
DUP: 6T
RAID0: 3T
SINGLE: 12T
RAID5: 2T
RAID6: 1T
[IMPLEMENTATION]
And for the each profile , we go chunk allocator level calculation:
The pseudo code looks like:
clear_virtual_used_space_of_all_rw_devices();
do {
/*
* The same as chunk allocator, despite used space,
* we also take virtual used space into consideration.
*/
sort_device_with_virtual_free_space();
/*
* Unlike chunk allocator, we don't need to bother hole/stripe
* size, so we use the smallest device to make sure we can
* allocated as many stripes as regular chunk allocator
*/
stripe_size = device_with_smallest_free->avail_space;
stripe_size = min(stripe_size, to_alloc / ndevs);
/*
* Allocate a virtual chunk, allocated virtual chunk will
* increase virtual used space, allow next iteration to
* properly emulate chunk allocator behavior.
*/
ret = alloc_virtual_chunk(stripe_size, &allocated_size);
if (ret == 0)
avail += allocated_size;
} while (ret == 0)
This minimal available space based calculation is not perfect, but the
important part is, the estimation is never exceeding the real available
space.
This patch just introduces the infrastructure, no hooks are executed
yet.
Reviewed-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: Qu Wenruo <wqu@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
(cherry picked from commit 52fead5eb8a743384bab24ca8c3695257c755f0f)
Reading I/O statistics via sysfs concurrently with tearing down a
frontend device might result in UAF or a NULL dereference, as the
reading of the statistics data is not synchronized with freeing the
data.
Use a rwlock for synchronizing.
This is XSA-528 / CVE-2026-100082
Reported-by: Yehyeong Lee <yhlee@isslab.korea.ac.kr> Signed-off-by: Juergen Gross <jgross@suse.com> Acked-by: Roger Pau Monné <roger@xenproject.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In order for the function to legitimately avoid locking/RCU, it needs to
run after all TX/RX processing has finished/stopped: Lookups happen from
xenvif_start_xmit(), while adds/removes happen from xenvif_tx_action()
(called out of xenvif_poll()).
Move the call out of xenvif_disconnect_data() (the call to the function
from connect()'s error path doesn't require any cleaning up there), and
put it past the synchronization point in backend_disconnect().
The network statistic data in the driver is protected by RCU.
Unfortunately when tearing down the queues after an error during
connecting to the frontend (e.g. due to illegal Xenstore data filled
in by the guest), the statistic data is being freed without
synchronizing the RCU state. In case the data is being read via sysfs
in this phase, access to already freed memory or memory being in new
use is possible.
Fix that by calling synchronize_net() before freeing the queue data.
To avoid hard-coding these values in drivers, define constants for them
that drivers can reference. Also, update all instances in bitmap.rs and
id_pool.rs that use these values to use the new constants.
Twenty years ago I made creating a tuntap device require CAP_NET_ADMIN
in commit ca6bb5d7ab22 ("[NET]: Require CAP_NET_ADMIN to create tuntap
devices.")
The idea was to allow the permissions on /dev/net/tun to be relaxed to
0666 and thus allow owners of tun devices assigned with TUNSETOWNER to
actually open it and get at their devices.
What we missed at the time was that persistent devices with *no* owner
were already essentially world-consumable, and anyone who could open
the device node could also attach to those.
Even then, using persistent devices with no explicitly set owner — on
the (false) presumption that CAP_NET_ADMIN would be required to open
them — wasn't an egregious userspace bug. There are cases where it does
make sense to separate the creation/configuration of the netdevice, from
the process which shovels packets into it.
Since then, TUNSETGROUP was added a year later in commit 8c644623fe7e
("[NET]: Allow group ownership of TUN/TAP devices."). Its commit message
noted that if both uid and gid are set, both have to match (in itself a
bizarre choice) — but still not that with *neither* set, no permission
at all was required.
Then in 2012 commit c8d68e6be1c3 ("tuntap: multiqueue support") allowed
new queues to be attached to a device from the moment of its creation,
with no way for userspace to close the race between creation and a
subsequent TUNSET{OWNER,GROUP}.
In 2024, commit 3ca459eaba1bf ("tun: fix group permission check") tried
to fix things to be saner: only require the user *or* the group to
match, in line with standard POSIX permissions, and require
CAP_NET_ADMIN if neither the user nor group authorised the access
(although it put the CAP_NET_ADMIN check first, and would wrongly have
reported the capability as having been used even when it was not).
That fix was reverted in commit a70c7b3cbc06 ("tun: revert fix group
permission check") due to allegedly causing a regression in an selinux
test suite which *expected* the attachment to succeed when authorised
by none of user, group, nor CAP_NET_ADMIN:
https://lore.kernel.org/netdev/CAFqZXNtkCBT4f+PwyVRmQGoT3p1eVa01fCG_aNtpt6dakXncUg@mail.gmail.com/
We should not break UAPI. But we don't have to consider an selinux test
case which deliberately exercises a security flaw to be UAPI worth
preserving either. To which end, I have fixed it in selinux-testsuite
commit b2f894f0016b ("tests/tun_tap: set an owner on the device for the
relabel tests").
So: fix the lack of CAP_NET_ADMIN check on devices which *haven't* been
assigned an owner or group, while leaving the weird tun ownership model
as it was (user *and* group have to match, if both are set). If someone
still wants to fix the latter (which wasn't actually the reason the
previous attempt was reverted), they can do it under separate cover.
Permit attaching to an existing device only if an owner and/or group was
set and the caller matches all that were set, or if the caller has
CAP_NET_ADMIN. The only behavioural change is that a device with no
owner or group now requires CAP_NET_ADMIN to attach — precisely the
loophole we want to close.
And put the CAP_NET_ADMIN check in the right place this time.
Fixes: ca6bb5d7ab22 ("[NET]: Require CAP_NET_ADMIN to create tuntap devices.") Reported-by: Andrew Lutomirski <luto@kernel.org> Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Cc: stable@vger.kernel.org Reviewed-by: Willem de Bruijn <willemb@google.com> Link: https://patch.msgid.link/3c77f88acff5b3c2617b22d64faad15f6212619c.camel@infradead.org Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The ASUS Vivobook S 14 Q423SA-U5512 (DMI family S5406SA) with the
ALC294 codec has working analog audio when booted directly after a full
shutdown. After a warm reboot from Windows 11, audio works briefly during
Linux startup, then both the internal speakers and headphones become
silent. PCM playback continues without errors.
Coefficient 0x1b is 0x4e4b on a direct cold boot, but 0x4a4b after the
Windows warm reboot. Restoring bit 10 changes it back to 0x4e4b and
immediately enables both analog outputs. Speaker playback, headphone
playback, jack switching, cold boot, and suspend/resume were tested
successfully. The same 0x4a4b state and immediate recovery with 0x4e4b
were independently confirmed on another S5406SA with subsystem 1043:10c4.
The same coefficient write is already used by
ALC294_FIXUP_ASUS_COEF_1B for the ASUS UX431DA. That fixup cannot be
reused here because it chains to ALC289_FIXUP_ASUS_GA401 and applies
unrelated DAC routing and microphone pin configuration.
Add a model-specific fixup for subsystem 1043:10c4 and chain it to the
existing ALC2XX_FIXUP_HEADSET_MIC fixup instead.
dev_pm_opp_put_opp_table() frees the opp_table which is subsquently used
by dev_err_probe(). This causes an Oops during boot on gs101-oriole.
cpu cpu0: error 000000006b6b6b6b: Couldn't find clock
Unable to handle kernel paging request at virtual address 006b6b6b6b6b6cd3
...
Hardware name: Oriole (DT)
pstate: 00400005 (nzcv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : _of_add_table_indexed+0x80/0xbb0
lr : _of_add_table_indexed+0x6c/0xbb0
...
Call trace:
_of_add_table_indexed+0x80/0xbb0 (P)
dev_pm_opp_of_cpumask_add_table+0x70/0x120
dt_cpufreq_probe+0x23c/0x480
platform_probe+0x64/0xb8
Fixes: 84f05af0975c9 ("opp: Use clk_get_optional() to avoid leaving opp_table->clk as an error pointer") Signed-off-by: Peter Griffin <peter.griffin@linaro.org> Reviewed-by: Tudor Ambarus <tudor.ambarus@linaro.org>
[ Viresh: use return value of dev_err_probe() ] Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A new utility metadata_size was added in
commit 261b67f4e347 ("selftests: ublk: add utility to get block device metadata size")
but it was not added to .gitignore. Fix that by adding it there.
While at it sort all entries alphabetically and add a SPDX license header.
Reviewed-by: Caleb Sander Mateos <csander@purestorage.com> Fixes: 261b67f4e347 ("selftests: ublk: add utility to get block device metadata size") Signed-off-by: Alexander Atanasov <alex@zazolabs.com> Reviewed-by: Ming Lei <ming.lei@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
__discard_anon_folio_pmd_locked() clears a lazyfree THP's PMD before it
determines whether the folio can be discarded, and restores it if the
folio is dirty or has extra references. It is done under only the PMD
lock and the anon_vma read lock inside try_to_unmap_one() and no mmap lock
is held. The temporarily none PMD causes UAF.
A parallel munmap() or mremap() is supposed to serialize with it on the
PMD lock, but zap_pmd_range() (called by munmap()) and move_page_tables()
(called by mremap()) do not take the lock when they see the temporarily
none PMD and skip the PMD. There is no more page table walk after the
none PMD is restored:
- munmap(): free_pgtables() waits in unlink_anon_vmas() for the anon_vma
lock, then free_pmd_range() drops the restored PMD as bad without
removing the rmap, so the folio keeps a stale mapcount.
- whole-VMA MREMAP_DONTUNMAP: dontunmap_complete() unlinks the source VMA
from its anon_vma while the restored PMD still maps the folio.
- plain mremap(): the THP is not moved but unmapped, so the THP's data is
lost.
In the first two cases, the folio stays mapped, but no VMA that maps it is
linked to its anon_vma. When the anon_vma is freed, it is still reachable
when one rmap walks the folio. Racing lazyfree reclaim against
MREMAP_DONTUNMAP as an unprivileged user reproduces a KASAN use-after-free
in folio_lock_anon_vma_read() as reported in the Closes link.
One solution is to use an invalidated PMD instead of a cleared one until
the discard is committed, but it needs arch fixes first.
1. GUP-fast still follows an invalidated PMD on riscv, LoongArch and s390.
2. sparc64 miscounts thp_pte_count when an invalidated PMD is set again or
cleared.
To keep the fix simple for backport, stop discarding lazyfree THPs at PMD
level instead. try_to_unmap_one() splits the PMD and discards the folio at
PTE level.
unmap_huge_pmd_locked() and __discard_anon_folio_pmd_locked() are left in
place, unused, to minimize this fix for backport and the invalidated PMD
approach is planned as a follow-up series to use them again.
Link: https://lore.kernel.org/20261012-lazyfree-stable-fix-v2-1-6387d6e30063@nvidia.com Fixes: 735ecdfaf4e8 ("mm/vmscan: avoid split lazyfree THP during shrink_folio_list()") Signed-off-by: Zi Yan <ziy@nvidia.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Reported-by: Kyle Zeng <kylebot@openai.com> Closes: https://lore.kernel.org/r/20261009165214.40212-2-kylebot@openai.com Reviewed-by: Kyle Zeng <kylebot@openai.com> Reviewed-by: Lance Yang <lance.yang@linux.dev> Acked-by: David Hildenbrand (Arm) <david@kernel.org> Reviewed-by: Lorenzo Stoakes (ARM) <ljs@kernel.org> Tested-by: JongHyuk Kim(MalHyuk) <malhyuk97@gmail.com> Link: https://lore.kernel.org/CADQRuqvXQ=YE=7T3oR66vTxE33pi=ygEeZKPTOU=cALZL0Figw@mail.gmail.com Assisted-by: LLM Cc: Rik van Riel <riel@surriel.com> Cc: Liam R. Howlett <liam@infradead.org> Cc: Vlastimil Babka <vbabka@kernel.org> Cc: Harry Yoo <harry@kernel.org> Cc: Jann Horn <jannh@google.com> Cc: <stable@vger.kernel.org> Signed-off-by: Zi Yan <ziy@nvidia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The cdev of a PPS generator is embedded in struct pps_gen_device, but
nothing ties the lifetime of that structure to the cdev: pps_gen is
freed by the release function of its device, and an open file holds a
device reference only until pps_gen_cdev_release() drops it.
When the generator is unregistered while /dev/pps-genN is open, that
put_device() drops the last reference and frees pps_gen, and __fput()
then calls cdev_put() on the freed cdev:
BUG: KASAN: slab-use-after-free in cdev_put+0x53/0x60
Read of size 8 at addr ffff88801383e138 by task ppsgen64/149
Call Trace:
cdev_put+0x53/0x60
__fput+0x745/0xad0
fput_close_sync+0xd9/0x1b0
__x64_sys_close+0x86/0xf0
...
Freed by task 149:
kfree+0x25a/0x6d0
device_release+0xca/0x3c0
kobject_put+0x169/0x320
pps_gen_cdev_release+0x51/0x80
__fput+0x36a/0xad0
pps.c had the same bug, fixed in commit c79a39dc8d06 ("pps: Fix a
use-after-free").
Fix it the usual way: embed the struct device in pps_gen_device and
register both with cdev_device_add(). This makes the device the parent
of the cdev, so the cdev holds a device reference until the last file
is closed.
The class_create() call has been deprecated in favor of class_register()
as the driver core now allows for a struct class to be in read-only
memory. Change pps_gen_class to be a const struct class and drop the
class_create() call.
ipu_bridge_get_ivsc_acpi_dev() returns an ACPI device with a reference
held. If no matching CSI device is found, ipu_bridge_check_ivsc_dev()
drops that reference before passing the device to dev_err(). Concurrent
CSI device and ACPI namespace removal can drop the other references,
making this the final put. dev_err() then reads the freed struct
acpi_device and its kobject name.
This was reproduced without kprobes by racing normal CSI platform-device
unregistration and ACPI configfs table removal against the production
ipu_bridge_init() entry. KASAN reported the final free through
device_release(), followed by invalid reads through _dev_err().
Emit the error message while the lookup reference is still held, then
drop the reference as before.
Source reproducers and vulnerable/fixed serial logs, with only the build
hostname redacted, are available at:
https://github.com/kimaiden1984-boop/linux-ipu-bridge-acpi-uaf-reproducer
The fixed target builds with W=1 on x86-64. This was not tested on
physical IPU hardware. The tests establish runtime reachability but do
not measure occurrence frequency on shipping systems or establish
exploitability.
Fixes: c66821f381ae ("media: pci: intel: Add IVSC support for IPU bridge driver") Cc: stable@vger.kernel.org Assisted-by: Codex:GPT-5 Signed-off-by: Yibo Tan <lhfff@tju.edu.cn> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
[ sashal: Retained the existing "MEI CSI dev" diagnostic because CVS support is absent. ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ksmbd_reopen_durable_fd() republishes a detached durable handle in the
reconnecting session's file table and takes no reference for that entry.
If smb2_open() then fails, cleanup relies on the final ksmbd_fd_put() to
remove the volatile id.
Two reconnects of the same handle can both take a lookup reference before
either attaches it. If the one that reopens first then fails and drops
its references first, the other request makes the final put through
ksmbd_put_durable_fd(). That frees the handle without removing the first
session's table entry, so tearing that session down writes to freed
memory.
Commit 3515503322f4 ("ksmbd: fix durable reconnect error path file
lifetime") handles a final put made by the reconnecting request, but not
one made by another holder.
Remove the volatile id under the file-table lock on this error path,
before the references are dropped, as ksmbd_close_fd() does on the
request_open path just above. Removal then no longer depends on which
holder makes the final put. The stale comment in the reconnect cleanup
is updated to match. A successful reconnect is unchanged.
Fixes: c8efcc786146 ("ksmbd: add support for durable handles v1/v2") Cc: stable@vger.kernel.org Signed-off-by: Ian Bridges <icb@fastmail.org> Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
[ sashal: Adapted the cleanup call to a braced if (fp) block because the replay and request-open cleanup branches are absent. ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When an encrypted request comes in, the decryption key is picked by
the SessionId in the transform header, but the request is then
authorized under whatever session the decrypted inner header names.
Nothing ever compares the two, so on a connection with more than
one session a client can get a request decrypted with one session's
key and run it as another session. Encrypted requests are exempt
from signing too, so a valid AEAD tag is the only proof of session
identity, and it is checked against the wrong session.
MS-SMB2 says the server must check that the transform SessionId
matches the one in the decrypted SMB2 header and treat a mismatch
as a protocol error. Add that check in __handle_ksmbd_work() after
decryption, before dispatching: the message has to be at least one
fixed SMB2 header, the first operation must not set
SMB2_FLAGS_RELATED_OPERATIONS and its SessionId must match the
transform SessionId, and every following non-related operation in
a compound chain must carry the same SessionId. NextCommand
offsets must be 8-byte aligned and stay inside the message; the
bounds check is done by subtraction so it cannot wrap.
When the decrypted payload is a compression transform, the
existing decompression step runs first and the decompressed SMB2
message is checked instead, since a compression transform header
carries no SessionId.
Fixes: e2f34481b24d ("cifsd: add server-side procedures for SMB3") Reported-by: Dairui Zhang <zhangdairui@gmail.com> Assisted-by: LLM Cc: stable@vger.kernel.org Signed-off-by: Dairui Zhang <zhangdairui@gmail.com> Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
[ sashal: Placed session validation directly after decryption because the intervening decompression code is absent. ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
An SMB WRITE to an existing named stream can silently discard data after
its written range. ksmbd stores streams as xattr values, but the write
path passes offset + count as the length of the replacement xattr even
when the old stream is longer. Writing 26 bytes at offset 0 to a 60-byte
stream reports 26 bytes written while discarding the remaining 34 bytes.
Keep the existing xattr length when an in-place write ends before EOF.
Reject writes that cross XATTR_SIZE_MAX rather than storing only the
portion that fits while reporting the full requested count.
Commit 4ea0bb8aaedf ("ksmbd: handle set/get info file for streamed file")
intentionally skipped ordinary inode truncation for stream handles: an
EOF SetInfo on an ADS had truncated the base file. That fixed the base
file corruption but left stream EOF changes reporting success without
resizing the stream. Handle these requests by resizing only the stream
xattr; the base file size remains untouched. Zero-fill stream extensions.
This is needed alongside the write fix: a short write's accidental
truncation previously masked an ignored explicit EOF change. The stream
resize helper returns xattr errors directly rather than using the
base-file truncate path's EBADF conversion.
Compound response handling pads the last response of a compound to an
8-byte boundary by growing the final iov and the RFC1002 length.
smb2_allocate_rsp_buf() allocates exactly response_sz. A command can
fill the response buffer to within 7 bytes of its end, for example a
QUERY_INFO whose output is bounded only by the space left. The padding
can then run past the allocation, and ksmbd_tcp_writev() reads those
bytes when it sends the response.
Allocate room for the padding when the request is a compound, as
smb2_read() and smb2_read_pipe() do for their payloads. response_sz is
unchanged, so commands still fill only up to it, and kvzalloc() zeroes
the padding.
Fixes: e2f34481b24d ("cifsd: add server-side procedures for SMB3") Cc: stable@vger.kernel.org Signed-off-by: Ian Bridges <icb@fastmail.org> Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
[ sashal: Adjusted context to retain work->conn->vals->max_trans_size instead of vals->max_trans_size. ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
SMB stream names are case-insensitive, but ksmbd stores each named
stream in an xattr whose name is case-sensitive. Opening an existing
stream finds its xattr with a case-insensitive match, yet the handle
keeps the name as the client spelled it, and every later operation that
needs an exact name uses that spelling:
- a WRITE or an end-of-file change creates a second xattr, so a stream
created as "Foo" and written through "foo" ends up stored twice, and
which value a later lookup returns depends on xattr list order;
- delete-on-close and delete-pending removal can fail with -ENODATA,
or remove only a case-variant copy and leave the original;
- the share mode check compares stream names with strcmp(), so two
opens of the same stream that differ only in case do not conflict.
Have ksmbd_vfs_casexattr_len() copy the matching xattr's name into the
handle when the stream is opened, so later operations use the name
already on disk. The match must have the same length as the name it
replaces; the stream lookup passes a length that includes the
terminating NUL, which already guarantees that, and the check makes the
copy safe for any caller. A stream that does not exist yet keeps the
client's spelling, as before.
smb2_rename() compares and looks up fp->stream.name case-insensitively,
so it behaves the same with either spelling.
Fixes: e2f34481b24d ("cifsd: add server-side procedures for SMB3") Cc: stable@vger.kernel.org Signed-off-by: DaeMyung Kang <charsyam@gmail.com> Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
[ sashal: adapted stream lookup signature changes to the existing callers and non-const mnt_idmap interface ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When get_inode_acl() fails, it returns an ERR_PTR(), but
ksmbd_acls_fattr() stored that pointer in smb_fattr without
checking for errors. Later callers then passed the error pointer
to security descriptor helpers or posix_acl_release(), resulting
in an invalid pointer dereference.
This was reproduced on an XFS-backed share when xfs_get_acl()
returned ERR_PTR(-ENOMEM). Querying the SMB ACL from a CIFS client
triggered the crash:
BUG: unable to handle page fault for address: fffffffffffffff8
RIP: smb_acl_sec_desc_scratch_len+0x58/0x90 [ksmbd]
Call Trace:
smb2_query_info+0x40d/0x17c0 [ksmbd]
handle_ksmbd_work+0x1ca/0x770 [ksmbd]
Fix this by checking the return values from get_inode_acl() in
ksmbd_acls_fattr(). Treat -EOPNOTSUPP as indicating that no ACL is
available, and propagate all other errors to the callers. Update
smb2_open() and smb2_get_info_sec() to handle these errors.
Cc: stable@vger.kernel.org Fixes: e2f34481b24d ("cifsd: add server-side procedures for SMB3") Assisted-by: Codex:gpt-5.6-sol Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn> Signed-off-by: Henry Hu <huhai@kylinos.cn> Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
[ sashal: retained the non-const mnt_idmap parameter in ksmbd_acls_fattr() ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Initialize the compatibility byte in request buffers without zeroing
the entire allocation. Keep the one-byte compatibility exception for
SMB2 messages so valid zero-length WRITE requests remain compatible,
and validate the data range in the WRITE handlers against the current
compound request before accessing it.
This prevents stale request-buffer data from being written to files or
sent to IPC pipes.
Fixes: 443d61d1fa9f ("ksmbd: allocate one more byte for implied bcc[0]") Cc: stable@vger.kernel.org Reported-by: zdi-disclosures@trendmicro.com # ZDI-CAN-32436 Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
[ sashal: Inlined the unavailable smb2_current_req_len() helper and omitted the decompression change because compression support is absent. ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A FileLinkInformation request can supply "/" as the new link name.
smb2_get_name() strips the trailing slash and passes an empty string to
ksmbd_vfs_kern_path_start_removing().
ksmbd_vfs_path_lookup() resolves an empty pathname as the configured share
path, without LOOKUP_BENEATH. With ReplaceIfExists set, smb2_create_link()
can therefore remove the share root from its parent before link creation
fails. An authenticated client with a writable share and DELETE access to
a file can trigger this if the share root is empty and the mapped user
has write and search permission on its parent directory.
Return -EINVAL for empty pathnames when for_remove is set, before
resolving the share path or acquiring removal resources. Ordinary lookups
continue to accept empty pathnames for opening the share root.
Fixes: 74d7970febf7 ("ksmbd: fix racy issue from using ->d_parent and ->d_name") Cc: stable@vger.kernel.org Reported-by: Zachariah Mithani (AploSec) <zachariah@aplosai.com> Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
[ sashal: Replaced for_remove with the existing do_lock parameter. ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A compound request can place SESSION_SETUP after another command. That
earlier command leaves a reference to its session in work->sess.
smb2_sess_setup() then overwrites work->sess with the session it looks up
or creates, and the dispatcher later puts work->sess only once. The
inherited reference is never dropped. That leaks the session, together
with its setup connection and socket, and every repeat of the compound
leaks another.
Save the inherited session on entry. After the request has been
processed, release it once, but only when smb2_sess_setup() has replaced
work->sess with its own session or cleared it. Do this after
ksmbd_conn_unlock() so the put does not run under srv_mutex. An error path
may instead leave the inherited session in work->sess. Keep it there, so
the dispatcher still puts it and the error response is still signed or
encrypted with it as before.
Fixes: 7aa8804c0b67 ("ksmbd: fix user-after-free from session log off") Cc: stable@vger.kernel.org Signed-off-by: Ian Bridges <icb@fastmail.org> Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
[ sashal: Omitted the installed = true assignment for the absent rejected-binding signature-check block. ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
smb_check_perm_dacl() returns the access mask it computes to the SMB2
CREATE handler, which uses it as the granted access for the open. For a
security descriptor whose DACL is present but contains no ACEs, the
function is supposed to deny access: an empty (non-NULL) DACL grants no
access to anyone except the object owner's implicit READ_CONTROL and
WRITE_DAC (MS-DTYP 2.4.5).
The empty-DACL branch only denies when both
- the DACL has no trailing bytes (pdacl_size == sizeof(struct smb_acl)),
and
- the request contains bits beyond READ_CONTROL / WRITE_DAC.
In every other case it takes the shared "goto err_out" with rc still 0,
which returns success without modifying *pdaccess, so the caller grants
the full requested access. Three problems follow:
- The READ_CONTROL / WRITE_DAC carve-out models the owner's implicit
rights, but the check runs before is_owner is computed, so those
rights are handed to *every* caller, not just the owner. For
READ_CONTROL the SMB2 CREATE path does not fall back to
inode_permission(), so any user can read the security descriptor of
an object whose empty DACL is meant to deny them.
- A request of exactly READ_CONTROL and/or WRITE_DAC is granted to any
caller. WRITE_DAC lets the handle rewrite the DACL.
- A DACL that declares zero ACEs but carries trailing bytes
(pdacl_size > sizeof(struct smb_acl)) skips the deny entirely and is
granted whatever was requested.
On shares where the NT ACL is the authoritative access control -- e.g.
"force user" shares, where every client maps to one POSIX identity so
POSIX cannot distinguish users -- this is an access-control bypass.
Compute is_owner before the empty-DACL branch and deny access from an
empty DACL unless the caller is the owner requesting no more than
READ_CONTROL / WRITE_DAC.
Fixes: e2f34481b24d ("cifsd: add server-side procedures for SMB3") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Aamir Ahmed <elb12345@hotmail.co.uk> Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
[ sashal: added missing ownership declarations and initialization of is_owner using the mount’s ID mapping ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For device types with need_parent_lock set (only USB interfaces),
probe() must run with the parent device locked. A synchronous attach
after device_add() gets this from the caller, e.g.
usb_set_configuration() holds the udev lock while adding interfaces.
__device_attach_async_helper() runs the probe from an async worker
instead, where that lock isn't held, and only takes device_lock(dev).
With async probing enabled for the hub driver (e.g. driver_async_probe=*
or driver_async_probe=hub), this races when a multi-TT hub is
configured. For multi-TT hubs, hub_probe() calls
usb_set_interface(hdev, 0, 1) to select the TT-per-port altsetting, and
usb_set_interface() creates the endpoint devices of the interface a
second time:
create_intf_ep_devs() checks and sets intf->ep_devs_created without a
lock of its own. If usb_set_configuration() has created ep_81 but not
yet set the flag, usb_set_interface() skips removing the endpoint
devices of the old altsetting and creates ep_81 again. With a
synchronous probe, hub_probe() runs inside device_add() with the udev
lock held, and usb_set_configuration() then finds ep_devs_created
already set.
Single-TT hubs don't call usb_set_interface() from hub_probe() and
don't show this warning, but their probe runs without the udev lock as
well.
On an i.MX8MM board with an onboard USB2514 hub this hit 11 of 100
boots:
It also shows on v7.3-rc6 in QEMU, with dummy_hcd and a raw-gadget
program emulating a multi-TT hub: re-authorizing the hub in a loop,
while a SCHED_FIFO task preempts the writing task at random points for
3 ms, hit the warning in 29 of 6000 iterations.
Take the parent lock in __device_attach_async_helper() as well, like
__driver_attach_async_helper() does, and move
__device_driver_lock/unlock() up for that. With this the warning was
gone in 300 boots on the board and in 6000 iterations in QEMU.
Fixes: 765230b5f084 ("driver-core: add asynchronous probing support for drivers") Signed-off-by: Mario Peter <mario.peter@leica-geosystems.com> Cc: stable@vger.kernel.org # 4.2 Cc: Dmitry Torokhov <dmitry.torokhov@gmail.com>
[ johan: switch to async_schedule_dev_nocall() to avoid deadlock if
async_schedule_dev() runs synchronously when parent lock is
held ] Signed-off-by: Johan Hovold <johan@kernel.org> Link: https://patch.msgid.link/20261014133109.1392265-4-johan@kernel.org
[ Fold the nested conditions into a single if statement; correct commit
message to mention "device types" rather than "buses" for
"need_parent_lock". - Danilo ] Signed-off-by: Danilo Krummrich <dakr@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When a device is hot-plugged, the drivers_autoprobe sysfs attribute is
not checked (at least for PCI devices). This means that
drivers_autoprobe is not working as intended, e.g. hot-plugged PCI
devices will still be autoprobed and bound to drivers even with
drivers_autoprobe disabled.
The problem likely started when device_add() was removed from
pci_bus_add_device() in commit 4f535093cf8f ("PCI: Put pci_dev in device
tree as early as possible") which means that the check for
drivers_autoprobe which used to happen in bus_probe_device() is no
longer present (previously bus_add_device() calls bus_probe_device()).
Conveniently, in commit 91703041697c ("PCI: Allow built-in drivers to
use async initial probing") device_attach() was replaced with
device_initial_probe() which faciliates this change to push the check
for drivers_autoprobe into device_initial_probe().
Make sure all devices check drivers_autoprobe by pushing the
drivers_autoprobe check into device_initial_probe(). This will only
affect devices on the PCI bus for now as device_initial_probe() is only
called by pci_bus_add_device() and bus_probe_device(), but
bus_probe_device() already checks for autoprobe, so callers of
bus_probe_device() should not observe changes on autoprobing.
Note also that pushing this check into device_initial_probe() rather
than device_attach() makes it only affect automatic probing of
drivers (e.g. when a device is hot-plugged), userspace can still choose
to manually bind a driver by writing to drivers_probe sysfs attribute,
even with autoprobe disabled.
Any future callers of device_initial_probe() will respect the
drivers_autoprobe sysfs attribute, which is the intended purpose of
drivers_autoprobe.
Signed-off-by: Vincent Liu <vincent.liu@nutanix.com> Link: https://patch.msgid.link/20251022120740.2476482-1-vincent.liu@nutanix.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Stable-dep-of: 187a62f99da2 ("driver core: take parent lock in async device attach") Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, init_cifs() calls cifs_proc_init() near the beginning of
the function, before INIT_LIST_HEAD, leading to a momentarily
uninitialized, but exposed, procfs. If a userspace process reads this
empty procfs, iterating over a zeroed list_head leads to a NULL
pointer dereference and a kernel oops/panic.
For example, if a delay is added between cifs_proc_init() and
INIT_LIST_HEAD, and a userspace process is polling
/proc/fs/cifs/DebugData, the following oops occurs:
Fix this by moving cifs_proc_init() down in init_cifs(), ensuring it runs
only after all internal lists, locks, counters, and workqueues are fully
ready, but before registering the filesystem. Update the error cleanup
ladder and exit_cifs() to maintain proper LIFO order.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Fredric Cover <fredric.cover.lkernel@gmail.com> Cc: stable@vger.kernel.org Signed-off-by: Paulo Alcantara <pc@manguebit.org>
[ sashal: Adjusted init_cifs() context because smb1_init_maperror() and smb2_init_maperror() are absent. ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
receive_encrypted_read() sets *num_mids = 1 and never writes the
caller's bufs[]. Offload returns -1 with *num_mids = 0, so the
demultiplex thread skips the frame. The inline decrypt path
returns 0 even when smb2_find_mid() fails.
cifs_demultiplex_thread() then treats mids[i] == NULL as an
oplock break and calls is_oplock_break(bufs[i], server) with
bufs[i] still NULL. smb2_is_valid_oplock_break() reads
hdr.Command at offset 12 and oopses.
A large encrypted READ whose MessageId does not match is enough
to panic cifsd.
Oops: general protection fault, probably for non-canonical address
0xdffffc0000000001: 0000 [#1] SMP KASAN PTI
KASAN: null-ptr-deref in range
[0x0000000000000008-0x000000000000000f]
CPU: 0 PID: 100 Comm: cifsd
RIP: smb2_is_valid_oplock_break+0xc8/0x2270 [cifs]
cifs_demultiplex_thread
Kernel panic - not syncing: Fatal exception
Return -1 when inline decrypt finds no mid, and skip
is_oplock_break() plus the unknown-frame dump when bufs[i]
is NULL. receive_encrypted_standard() already assigns
bufs[*num_mids].
Fixes: c42a6abe3012 ("CIFS: Add capability to decrypt big read responses") Assisted-by: LLM Signed-off-by: Zikang Gao <2715417602@qq.com> Reviewed-by: David Howells <dhowells@redhat.com> Signed-off-by: Paulo Alcantara <pc@manguebit.org> Cc: stable@vger.kernel.org
[ sashal: Adjusted context to retain the existing one-argument mid_execute_callback() and release_mid() calls. ] Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
io_recv_finish() debits sr->mshot_total_len with min_t(int, sel->val,
sr->mshot_total_len). sel->val is the cumulative byte count of the
request: on a bundle retry the receive path adds sr->done_io back to the
return value, and the bundle branch in io_recv_finish() recovers the new
bytes with this_ret = sel->val - sr->done_io. The debit therefore uses
the wrong value after a bundle retry, and the int cast truncates a limit
above INT_MAX.
For an IORING_OP_RECV multishot request with optlen above INT_MAX, the
limit becomes negative when cast to int, so the first completion debits
the entire limit and ends the multishot with IORING_RECV_MSHOT_DONE
while the socket still holds data that covers the limit. With 120 bytes
queued, a 64-byte buffer and optlen 0x80000000, the first 64-byte CQE
lacks IORING_CQE_F_MORE and 56 bytes remain unread; the same setup with
optlen 0x7fffffff posts 64 and 56 bytes with F_MORE.
After a bundle retry, the bytes already debited for the previous
attempts are debited again, so the limit is exhausted early. With 1400
bytes queued, optlen 1000 and buffers of 600, 100, 100, 50, 50 and 200
bytes, the first attempt delivers 700 bytes and retries; the second
delivers 200 more, but the debit accounts for 900 and finishes the
request with a terminal 900-byte CQE while 500 bytes remain unread. The
limit could have been reached.
Debit only the bytes received by this invocation, sel->val -
sr->done_io, and compare as unsigned so a limit above INT_MAX is
accounted correctly. Non-bundle multishot requests keep sr->done_io at
zero, so their accounting is unchanged.
Fixes: 0ebc9a7ecf6a ("io_uring/net: Support multishot receive len cap") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Cheng Yongkang <teel4res@gmail.com> Link: https://patch.msgid.link/179328314078.648427.10086632105828535293@gmail.com
[axboe: fix-up near-by incorrect comment while in there] Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
io_timeout_cancel() posts the cancelled timeout's -ECANCELED CQE without
advancing ctx->cq_timeouts. io_timeout_fn() and io_kill_timeout()
advance it for every timeout CQE they post, and io_flush_timeouts()
computes the number of non-timeout completions as
seq = cached_cq_tail - cq_timeouts
to compare against each off timeout's target_seq. The cancelled
timeout's CQE is therefore counted as a regular completion, so every
cancellation makes an already armed off timeout fire one completion
early.
A ring with a keep-alive timeout of off=4 completes it after two NOPs
when one off=100 victim is cancelled through IORING_OP_ASYNC_CANCEL, and
before any NOP when two victims are cancelled. The cancel CQE should
count, but the cancelled timeouts' -ECANCELED CQEs must not: the
expected counts are three and two NOPs respectively.
Advance cq_timeouts when io_timeout_extract() succeeds, under
timeout_lock and before the -ECANCELED completion is queued, matching
io_timeout_fn() and io_kill_timeout(). The -ENOENT and -EALREADY paths
post no new timeout CQE and keep their accounting. Both the
ASYNC_CANCEL and the TIMEOUT_REMOVE path go through io_timeout_cancel().
Fixes: bfe68a221905 ("io_uring: off timeouts based only on completions") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Cheng Yongkang <teel4res@gmail.com> Link: https://patch.msgid.link/179328314078.648427.13761781546953594668@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
An IOPOLL ring without DEFER_TASKRUN can be reaped by a task other than
the one that submitted the I/O. If a polled request is marked
REQ_F_REISSUE, io_free_batch_list() calls io_queue_iowq() with the
original tctx. Ordinary task exit can tear down tctx->io_wq after
draining only the tracked requests. The live request pins the task, but
not its io-wq, so the foreign task can race io_wq_put_and_exit() and
enqueue through a freed or NULL queue.
Forward foreign userspace enqueues to the owning task before accessing
tctx->io_wq. Use a dedicated task_work callback so reissues still go
straight to io-wq, and fail them if task_work is running in cancellation
mode. Preserve the existing PF_KTHREAD cancellation path. This also
covers the other callers of io_queue_iowq() without changing the
worker-owned enqueue path.
Signed-off-by: Kyle Zeng <kylebot@openai.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
[axboe: 6.18 lacks 13b0742fd058, which removed io_req_queue_iowq(). Name
the new task_work callback io_queue_iowq_foreign_tw() so it doesn't
clash with the existing io_req_queue_iowq_tw()] Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
io_sq_thread_stop() wakes the SQPOLL thread and waits for it to exit.
The thread belongs to the thread group that created the ring, so a
SIGSTOP or the cgroup freezer holds it in get_signal(), and the wake up
doesn't change that. io_ring_exit_work() then sleeps in that wait until
someone continues or kills the process. iou_exit has 64 slots, so 64
closed SQPOLL rings of a stopped process keep every other ring from
being freed, for everyone.
Don't wait. Set the stop bit, wake the thread, and let it exit when it
gets to run again. The sqd can now outlive its last ring, so give it a
second refcount with one reference for the rings and one for the thread,
and free it with the last one. The thread can't reach a freed ring,
rings leave ctx_list before dropping their reference, and the io-wq hash
is refcounted.
Fixes: 16efa4fce3b7 ("io_uring: allow IO worker threads to be frozen") Signed-off-by: Chris Mason <mason@kernel.org> Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A failed xa_store() can leave nodes without entries in the tree, and
erasing every entry doesn't free those. An XArray needs xa_destroy()
before it goes away, even an empty one. That was dropped for tctx->xa in
commit ca6484cd308a ("io_uring: no need to call xa_destroy() on empty
xarray") over a lockdep report from freeing the tctx in softirq context,
and ctx->personalities never had it.
Now that the nodes are charged, a store can fail at memory.max and the
leaked nodes stay charged. Call xa_destroy() in io_uring_clean_tctx(),
which the task runs itself in process context before the tctx is freed,
in the error path of __io_uring_add_tctx_node(), and for
ctx->personalities in io_ring_ctx_free().
Fixes: ca6484cd308a ("io_uring: no need to call xa_destroy() on empty xarray") Signed-off-by: Chris Mason <mason@kernel.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
[axboe: 6.18 installs the tctx before adding the node and has no error
path that frees it, drop that hunk. There's no ctx->hpage_acct either] Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The io_tctx_node allocated the first time a task touches a ring, and
the XArray nodes behind tctx->xa and ctx->personalities, are plain
GFP_KERNEL. tctx->xa is indexed by the ring pointer, so nearly every
entry needs nodes of its own, and a task gets one entry per ring it has
used. 642 threads on 4000 rings held 2.3GB with 59MB charged against a
64MB memory.max.
Allocate the node with GFP_KERNEL_ACCOUNT and set XA_FLAGS_ACCOUNT on
both XArrays.
Fixes: 0f2122045b94 ("io_uring: don't rely on weak ->files references") Signed-off-by: Chris Mason <mason@kernel.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
[axboe: 6.18 has kmalloc() for the node, inline in
__io_uring_add_tctx_node()] Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The io_mapped_ubuf of a registered buffer has one bio_vec per page,
unless the pages coalesce into huge pages. Huge pages only count against
RLIMIT_MEMLOCK once per ring, so a buffer that maps the same 2MB folio
over and over in a way that defeats coalescing gets a full bvec array
per registration while only the one folio counts as locked. 1GB of that
registered in 100 slots took several times a 64MB memory.max, with 2.6MB
charged.
Allocate the io_mapped_ubuf and the io_rsrc_node with
GFP_KERNEL_ACCOUNT. An object reused from the ring caches stays charged
to whoever allocated it first, same as the io_kiocb cache.
Fixes: de2939388be5 ("io_uring: improve registered buffer accounting for huge pages") Signed-off-by: Chris Mason <mason@kernel.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
[axboe: 6.18 uses kvmalloc() + struct_size_t() rather than
kvmalloc_flex()] Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
io_wq_exit_start() sets IO_WQ_BIT_EXIT, but worker creation task_work
queued before that stays queued until io_wq_exit_workers() cancels it.
In between the task runs its task_work while canceling requests, and a
worker can still get created.
That's a problem for exec. begin_new_exec() calls io_uring_task_cancel()
after de_thread(), when it's supposedly the only thread left, and
io_wq_exit_workers() only waits for ->worker_done, which a worker
completes before do_exit(). So the exiting worker is still in the thread
group, using the mm, files and sighand, while exec replaces them. Easy
to hit with a signal pending when create_io_thread() fails and the retry
gets queued right before execve().
Cancel the queued creation task_work as soon as the exit bit is set.
Pending work that isn't otherwise canceled now ends with -ECANCELED when
the io-wq is destroyed, rather than depending on whether the task
happened to run its task_work first.
Fixes: 685fe7feedb9 ("io-wq: eliminate the need for a manager thread") Signed-off-by: Josef Bacik <josef@toxicpanda.com>
[axboe: the barrier is already upstream, see 0e5663856b69, only the
cancel is new] Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
IORING_OP_CONNECT on an O_NONBLOCK AF_UNIX stream socket retries a
connect that fails with -EAGAIN because the listener's backlog is full.
Polling the client socket doesn't help: an unconnected stream socket
reports EPOLLOUT and EPOLLHUP right away, before the backlog drains. So
io_uring retries the connect until the apoll retry limit is hit, then
punts the request to io-wq. With a full backlog, io_connect() runs 128
times in the submitting task and once more in a new io-wq worker. That
worker stays in the task's thread group, so a later
unshare(CLONE_NEWUSER) fails with -EINVAL.
None of this changes the result. __sys_connect_file() adds the socket's
own O_NONBLOCK to every attempt, so the attempt from io-wq fails with
-EAGAIN as well. That -EAGAIN becomes the CQE result, and connect(2)
returns the same -EAGAIN.
Let's complete the request with -EAGAIN right away when the socket has
O_NONBLOCK set, like connect(2) does. io_accept() already completes
-EAGAIN this way for IORING_ACCEPT_DONTWAIT. -EINPROGRESS still waits
for the connection to complete. A blocking socket still ends up in
io-wq, where the connect sleeps until the listener accepts.
io_recvmsg() takes the transfer length from the ring provided buffer when
sqe->len is 0, and such a buffer can be up to 4G. min_ret is an int, so 2G
or more makes it negative and the (ret < min_ret) block is skipped, and the
oversized length reaches the socket layer unchanged: an empty socket then
completes with 0 instead of arming poll, and a short read is reported as
complete. Clamp the length to MAX_RW_COUNT, as io_recv() does.
Fixes: c7fb19428d67 ("io_uring: add support for ring mapped supplied buffers") Signed-off-by: Yichong Chen <chenyichong@uniontech.com> Link: https://patch.msgid.link/20261023112209.87979-1-chenyichong@uniontech.com Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
io_ring_buffers_peek() flags the request with REQ_F_BL_EMPTY if the
buffers it picked were the last ones available, comparing head to tail.
When the loop breaks because max_len was exhausted the buffer was picked,
but head is left pointing at it, so the flag isn't set if it was the last
buffer. A bundle receive then retries, finds no buffer, and the request,
which has received data, is marked as failed on account of the -ENOBUFS.
That severs a link, and ends a multishot.
When a remote buffer descriptor is larger than the remaining local
buffer, smbdirect_connection_rdma_xmit() clamps its length and sets
buf_len to zero. The common accounting code then subtracts the clamped
length from buf_len, causing it to underflow. If another descriptor
follows, the loop processes it with a wrapped remaining length and
includes memory beyond the local buffer in the RDMA transfer.
A client can trigger an out-of-bounds kernel write with an SMB WRITE
request, or an out-of-bounds kernel read with an SMB READ request, by
supplying multiple RDMA buffer descriptors with an oversized descriptor
before the last one.
Remove the premature assignment to buf_len. Subtracting the clamped length
already consumes the remaining buffer, so the next iteration stops and
only descriptors within the local buffer are used to build work requests.
Fixes: e628bf939aaf ("ksmbd: reduce descriptor size if remaining bytes is less than request size") Cc: stable@vger.kernel.org Reported-by: zdi-disclosures@trendmicro.com # ZDI-CAN-33549 Signed-off-by: Stefan Metzmacher <metze@samba.org> Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
[ metze: in 6.18 the code is smb_direct_rdma_xmit() in
fs/smb/server/transport_rdma.c ] Assisted-by: Claude:claude-opus-5-5 Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When a process drops CAP_SYS_ADMIN and calls fsconfig(FSCONFIG_SET_FLAG,
"override_creds"), overlayfs saves those restricted credentials in
ofs->creator_cred. ovl_fill_super() then wraps the entire mount setup in
with_ovl_creds(), switching to those restricted credentials for the
duration.
Mount setup calls clone_private_mount(), which checks CAP_SYS_ADMIN
against current credentials. With restricted credentials active, this
check fails with -EPERM even though the task issuing
FSCONFIG_CMD_CREATE is fully privileged.
Use the mounter's real credentials only for cloning the private layer
mounts. Keep all other backing filesystem operations during mount setup
under the stashed credentials.
Reported-by: Disha Goel <disgoel@linux.ibm.com> Closes: https://lore.kernel.org/linux-unionfs/20260924112858.51392-1-disgoel@linux.ibm.com Fixes: c28f922c9dcee ("clone_private_mnt(): make sure that caller has CAP_SYS_ADMIN in the right userns") Cc: stable@vger.kernel.org Reviewed-by: Christian Brauner (Amutable) <brauner@kernel.org> Signed-off-by: Amir Goldstein <amir73il@gmail.com>
[backport: 6.18 has no scoped_with_creds(), use scoped_class(override_creds)] Signed-off-by: Amir Goldstein <amir73il@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A change to a removed file through an open descriptor replaces whatever
has its old name in the upper layer.
Copy up must only create a name in the upper directory, never replace
one. After an unlink, an rmdir or a rename over a lower dentry, a
dentry that an open file keeps alive still has its old parent and name,
but in the upper directory that name is now a whiteout or a new object.
ovl_copy_up_workdir() renames the copy over it, which brings back the
removed object and destroys whatever took its name.
An unprivileged user opened a lower FIFO and unlinked it, root created
a file at that name, and fchmod on the old descriptor replaced the file
of root with the FIFO.
Fix this by looking up the target name of the rename in
ovl_copy_up_workdir() with LOOKUP_EXCL, so that copy up fails with
-EEXIST if the name is taken, as copy up with O_TMPFILE already does
when it links the copy to the name.
Fixes: e9be9d5e76e3 ("overlay filesystem") Cc: stable@vger.kernel.org Signed-off-by: Chris Mason <mason@kernel.org> Signed-off-by: Josef Bacik <josef@toxicpanda.com>
[amir: use LOOKUP_EXCL instead of relying on d_unhashed()] Signed-off-by: Amir Goldstein <amir73il@gmail.com>
[backport: no start_renaming_dentry() and LOOKUP_EXCL, fail with -EEXIST if
the target of ovl_lookup_upper() is positive] Signed-off-by: Amir Goldstein <amir73il@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Listing a directory on an overlay uses a freed dentry and crashes the
kernel when a lookup moves the directory at the same time.
Code that can sleep may only use d_parent without a reference while
something stops a move of the dentry. ovl_iterate_real(),
ovl_need_adjust_d_ino() and ovl_cache_update() use the d_parent of the
listed directory and hold only the lock of that directory. When a
lookup finds the directory under another parent after a change to a
layer, d_splice_alias() moves it without that lock and drops its
reference on the old parent.
With a delay in ovl_iterate_real(), getdents64 on an overlay directory
raced a rename in the lower layer and a lookup that moved its dentry,
and after an rmdir of the old parent KASAN reported a use-after-free in
security_inode_getattr().
Fix this by using dget_parent() in the three functions.
Fixes: b5efccbe0a12 ("ovl: constant d_ino across copy up") Cc: stable@vger.kernel.org Signed-off-by: Chris Mason <mason@kernel.org> Signed-off-by: Josef Bacik <josef@toxicpanda.com>
[amir: fix comments in ovl_need_adjust_d_ino()] Signed-off-by: Amir Goldstein <amir73il@gmail.com>
[backport: no ovl_need_adjust_d_ino(), take the parent with dget_parent()
in ovl_iterate() and test a local flag] Signed-off-by: Amir Goldstein <amir73il@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
An unprivileged user crashes the kernel by mounting overlayfs with the
volatile option.
A dentry that serves as the parent of a lookup must be a directory.
ovl_create_volatile_dirty() walks work/incompat/volatile/dirty by name,
and ovl_lookup_or_create() returns an existing entry whatever its type.
A file or a symlink that another task creates during the mount becomes
the parent of the next lookup, which calls a NULL ->lookup.
An unprivileged user mounted overlayfs with "volatile" in a user
namespace while a second process created a symlink named work/incompat,
and the mount task oopsed on a NULL ->lookup call.
Fix this by returning -ENOTDIR from ovl_lookup_or_create() when the
caller asked for a directory and the existing entry is a non-dir.
Fixes: c86243b090bc ("ovl: provide a mount option "volatile"") Cc: stable@vger.kernel.org Signed-off-by: Chris Mason <mason@kernel.org> Signed-off-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Amir Goldstein <amir73il@gmail.com>
[backport: no ovl_start_creating_upper(), check the child of
ovl_lookup_upper() in ovl_lookup_or_create() instead] Signed-off-by: Amir Goldstein <amir73il@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The raw event callback runs before HID core validates and zero-pads the
report. It passes the buffer to a helper which inspects data[0] and, for a
special report, reads the complete struct savu_mouse_report_special. A
truncated report can therefore cause an out-of-bounds read.
Require enough data for the special report before calling the helper.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw-event length validation to hid-asus.
Fixes: 6a2a6390cf09 ("HID: roccat: add support for Roccat Savu") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Benjamin Tissoires <bentiss@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The raw event callback runs before HID core validates and zero-pads the
report. It reads the report number without checking the received length
and passes matching data to roccat_report_event(), which copies a complete
struct ryos_report_special. A truncated report can therefore cause an
out-of-bounds read and expose adjacent data through the Roccat character
device.
Require a complete special report before inspecting or forwarding it.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw-event length validation to hid-asus.
Fixes: 6f3a19360545 ("HID: roccat: add support for Ryos MK keyboards") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Benjamin Tissoires <bentiss@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The raw event callback runs before HID core validates and zero-pads the
report. It passes the buffer to helpers which inspect data[0] and read a
complete struct pyra_mouse_event_button for button reports. A truncated
report can therefore cause an out-of-bounds read.
Require enough data for the button report before calling either helper.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw-event length validation to hid-asus.
Fixes: cb7cf3da0daa ("HID: roccat: add driver for Roccat Pyra mouse") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Benjamin Tissoires <bentiss@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The raw event callback runs before HID core validates and zero-pads the
report. It passes the buffer to helpers which inspect data[0] and read a
complete struct kovaplus_mouse_report_button for button reports. A
truncated report can therefore cause an out-of-bounds read.
Require enough data for the button report before calling either helper.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw-event length validation to hid-asus.
Fixes: 0e70f97f257e ("HID: roccat: Add support for Kova[+] mouse") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Benjamin Tissoires <bentiss@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The raw event callback runs before HID core validates and zero-pads the
report. It reads the report number without checking the received length
and passes matching data to roccat_report_event(), which copies a complete
struct konepure_mouse_report_button. A truncated report can therefore
cause an out-of-bounds read and expose adjacent data through the Roccat
character device.
Require a complete button report before inspecting or forwarding it.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw-event length validation to hid-asus.
Fixes: 8936aa31cd5f ("HID: roccat: add support for Roccat Kone Pure gaming mouse") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Benjamin Tissoires <bentiss@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The raw event callback runs before HID core validates and zero-pads the
report. It passes the buffer to helpers which inspect data[0] and, for a
button report, read the complete struct koneplus_mouse_report_button. A
truncated report can therefore cause an out-of-bounds read.
Require enough data for the button report before calling either helper.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw-event length validation to hid-asus.
Fixes: 47dbdbffe15b ("HID: roccat: Add support for Roccat Kone[+] v2") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Benjamin Tissoires <bentiss@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The raw event callback runs before HID core validates and zero-pads the
report. It passes the buffer to helpers which inspect data[0] and, for a
button report, read the complete struct isku_report_button. A truncated
report can therefore cause an out-of-bounds read.
Require enough data for the button report before calling either helper.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw-event length validation to hid-asus.
Fixes: d41c2a7011df ("HID: roccat: Add support for Isku keyboard") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Benjamin Tissoires <bentiss@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The wired Anne Pro 2 identifying as 04d9:a292 stops delivering keyboard
input after a period of inactivity. Unplugging and reconnecting the
keyboard restores input.
Keeping ObinsKit running prevents the problem. A read-only userspace
reader that keeps USB HID interface 1 open also prevents it, as does
usbhid.quirks=0x04d9:0xa292:0x00000400 without either application running.
Commit ca28aff0e1dc ("HID: quirks: Always poll Obins Anne Pro 2
keyboard") added HID_QUIRK_ALWAYS_POLL for product ID a293.
The variant identifying as 04d9:a292 exhibits the same symptom.
Apply the quirk to this product ID as well.
The HID core invokes raw_event callbacks before validating the report
length. pxrc_raw_event() reads byte seven and writes bytes one and seven
without checking that the report contains those bytes.
A one-byte input report reproduced the first invalid access under KASAN:
BUG: KASAN: slab-out-of-bounds in pxrc_raw_event+0x161/0x260
Read of size 1 by task hidtrigger/90
Call Trace:
pxrc_raw_event+0x161/0x260
kasan_report+0x139/0x170
Require all eight bytes before reading or rewriting the report.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw_event length validation to hid-asus.
The HID core invokes raw_event callbacks before validating the report
length. Prodikeys reports 0x01 and 0x04 are passed to handlers that always
read bytes one through three, regardless of the supplied size. A truncated
report can therefore cause an out-of-bounds read.
Require four bytes for the two fixed-format reports. Keep report 0x03 on
its existing variable-length path, which derives the number of note pairs
from size and only accesses complete pairs.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw_event length validation to hid-asus.
MCP2221 responses are 64-byte reports, but mcp2221_raw_event() currently
accepts any report of at least four bytes. Several response handlers read
far beyond that minimum. In particular, the I2C status response reads
offset 20 and copies ADC samples starting at offset 50. Other GPIO, SRAM,
and flash response handlers also access fixed offsets beyond byte three.
The HID core invokes raw_event callbacks before validating the report
length, so a truncated response can cause an out-of-bounds read. Require a
complete protocol report before dispatching any response handler.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw_event length validation to hid-asus.
Fixes: 67a95c21463d ("HID: mcp2221: add usb to i2c-smbus host bridge") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The HID core invokes raw_event callbacks before validating the report
length. mcp2200_raw_event() casts a READ_ALL response to struct
mcp_read_all_resp and reads fields through offset ten without checking that
the response contains the structure.
A one-byte READ_ALL response reproduced the issue under KASAN:
BUG: KASAN: slab-out-of-bounds in mcp2200_raw_event+0x24b/0x3a0
Read of size 1 by task hidtrigger/89
Call Trace:
mcp2200_raw_event+0x24b/0x3a0
kasan_report+0x139/0x170
Reject an incomplete READ_ALL response with a protocol error before reading
its fields. The common completion path then wakes the command waiter with
the error instead of leaving it to time out.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw_event length validation to hid-asus.
Fixes: 740329d7120f ("HID: mcp2200: added driver for GPIOs of MCP2200") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The HID core invokes raw_event callbacks before validating the report
length. For an unnumbered keyboard report, logi_dj_raw_event() uses the
second byte as temporary storage while prepending a report ID. A one-byte
report therefore makes both the initial write and the later restoration
access data beyond the received report.
Reject unnumbered keyboard reports that do not provide the temporary byte.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw_event length validation to hid-asus.
Fixes: 74808f9115ce ("HID: logitech-dj: add support for non unifying receivers") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The HID core invokes raw_event callbacks before validating the report
length. elo_raw_event() only checks the first-byte packet marker before
elo_process_data() reads coordinates, flags, and pressure through offset
seven. A truncated packet beginning with the expected marker can therefore
cause an out-of-bounds read.
Require the complete eight-byte SmartSet packet before parsing it.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw_event length validation to hid-asus.
Fixes: d23efc19478a ("HID: add driver for ELO 4000/4500") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The HID core invokes raw_event callbacks before validating the report
length. cp2112_raw_event() reads every field of a transfer status response
without checking size. For a data response, it reads the length at offset
two and copies that many bytes from offset three, again without ensuring
that the input report contains them.
Require a complete transfer status structure. For data responses, first
require the header and then ensure the clamped payload length fits in the
received report before copying it.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw_event length validation to hid-asus.
The HID core invokes raw_event callbacks before validating the report
length. cougar_raw_event() reads the key code and action from offsets one
and two without checking that those bytes are present. A short report on
the special interface can therefore cause an out-of-bounds read.
Consume reports that do not contain the action field before accessing the
fixed offsets, consistent with the callback's handling of other reports on
the special interface.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw_event length validation to hid-asus.
Fixes: b8e759b8f6da ("HID: cougar: Add support for the Cougar 500k Gaming Keyboard") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao <yaojiale02@163.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Syzbot reported an out-of-bounds shift (32-bit integer shifted left by
32) in the s32ton() routine in the HID core. This routine converts a
signed 32-bit integer to a signed n-bit integer, and it caps the value
of n at 32. But when n == 32 there's no need to do anything; it can
simply return the input value. This will avoid calculating (1 << n)
when n is 32, which was the cause of the invalid shift.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Fixes: 69c02ffde6ed ("HID: core: clamp report_size in s32ton() to avoid undefined shift") Reported-by: syzbot+25a1d2907885b2182275@syzkaller.appspotmail.com Closes: https://lore.kernel.org/linux-usb/6ac9cbd8.a150a118.27e199.0025.GAE@google.com/ Tested-by: syzbot+25a1d2907885b2182275@syzkaller.appspotmail.com Cc: stable@vger.kernel.org Link: https://patch.msgid.link/dfa0aa1c-e55e-4316-8307-f72a07383634@rowland.harvard.edu Signed-off-by: Benjamin Tissoires <bentiss@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use report type from request node while sending feature and input reports
to HID core. The cli_data report type array is never assigned and
feature reports are sent as input reports.
The accelerometer HID report descriptor declares a trailing motion-state
field (usage 0x0451), but the Sensor Fusion Hub never reports this data,
so each accelerometer input report is one byte shorter than the descriptor
advertises.
This was previously harmless because hid-core zero-padded short reports.
Since commit 0a3fe972a7cb ("HID: core: Mitigate potential OOB by removing
bogus memset()") such reports are rejected instead:
hid-sensor-hub 0020:1022:0001.0003: Event data for report 1 was too short (15 vs 14)
As a result the accelerometer input events are dropped and screen
auto-rotation stops working on AMD SFH systems.
Remove the motion-state field from the accelerometer report descriptor and
drop the unused in_accel_shake_detection member, so the descriptor matches
the report actually provided by the hardware.
Fixes: 1434f9fc0e47 ("SFH: Create HID report to Enable support of AMD sensor fusion Hub (SFH)") Reported-by: Andrej Scek <scek.andrej@gmail.com> Closes: https://lore.kernel.org/linux-input/CADrS85612m-Hh-G4BJmOhy-y6r3Ni3kD8_GM-2QRBeWWEEZWuA@mail.gmail.com/ Cc: stable@vger.kernel.org Signed-off-by: Basavaraj Natikar <Basavaraj.Natikar@amd.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The HID core invokes raw_event callbacks before validating the report
length. zc_raw_event() verifies the report ID in byte zero and then reads
the key code from byte one for report IDs 0x02 and 0x03 without checking
that the second byte is present.
A one-byte report reproduced the issue under KASAN:
BUG: KASAN: slab-out-of-bounds in zc_raw_event+0x54c/0x580
Read of size 1 by task hidtrigger/93
Call Trace:
zc_raw_event+0x54c/0x580
__hid_input_report+0x2ef/0x3a0
inject_store+0x479/0x490 [faketrans]
Require both the report ID and key code before processing the report.
Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init
sequence") added the same kind of raw_event length validation to hid-asus.
get_d_signext() has two compounding bugs since its introduction in 2010:
1. The extraction mask 0x8FF silently drops bits 8-10 of the 12-bit D
field (PPC ISA bits 21-23), corrupting any displacement that has any
of those bits set.
2. The sign-magnitude idiom "-(d & 0x7ff)" is wrong for two's-complement:
for D=0xFFC (encoding of -4) it returns -252 instead of -4.
Together these errors produce a wrong effective address for psq_l, psq_lu,
psq_st, and psq_stu whenever the displacement is negative or is a positive
value >= 0x100 with bits 8-10 set — essentially any real-world paired-
single stack-relative access.
The D field occupies the bottom 12 bits of the instruction word (confirmed
by the adjacent W and I extractions via inst_get_field(inst,16,16) and
inst_get_field(inst,17,19)). Replace the open-coded logic with the standard
sign_extend32(inst & 0xfff, 11), which correctly performs two's-complement
sign extension from 12 bits to 32 bits.
kvmppc_hv_find_lock_hpte() requires virtual-mode callers to run with
preemption disabled, because it can return with HPTE_V_HVLOCK still held
until the caller later unlocks the HPTE. Existing virtual-mode callers
in book3s_64_mmu_hv.c already follow that rule.
Two virtual-mode call paths do not:
1. kvmppc_handle_exit_hv() calls kvmppc_hpte_hv_fault() for hash-mode
data-side and instruction-side faults after guest exit, with
preemption already enabled. kvmppc_hpte_hv_fault() calls
kvmppc_hv_find_lock_hpte() directly.
2. kvmppc_pseries_do_hcall() executes virtual-mode handlers for HPT
hcalls via kvmppc_pseries_do_hpt_hcall() with preemption enabled.
The handlers for H_ENTER, H_REMOVE, H_READ, H_CLEAR_MOD, H_CLEAR_REF,
H_PROTECT, and H_BULK_REMOVE spin on try_lock_hpte() or lock_rmap(),
and H_ENTER also reaches kvmppc_do_h_enter(), which uses
arch_spin_lock() on kvm->mmu_lock. That raw lock choice is
intentional because kvmppc_do_h_enter() is also used by real-mode
callers, so the correct fix is to establish the proper preemption
context at the virtual-mode caller boundary.
If a vCPU thread is preempted while holding HPTE_V_HVLOCK, any other
thread on the same CPU trying to acquire the same bit-lock can spin
indefinitely. The same applies to lock_rmap() users and to the H_ENTER
path while holding the raw mmu_lock.
Fix this by adding preempt_disable()/preempt_enable() pairs around the
two kvmppc_hpte_hv_fault() call sites in kvmppc_handle_exit_hv(), and
around the kvmppc_pseries_do_hpt_hcall() invocation in
kvmppc_pseries_do_hcall(). Targeting only the HPT hcall helper avoids
wrapping non-HPT hcalls (such as H_RTAS, H_CONFER, H_REGISTER_VPA) that
perform page faults, blocking memory allocations, or context switches.
Enable Notify VM exit functionality for TDX guests.
Notify VM exit is an existing feature supported by KVM. Userspace can
enable Notify VM exit through KVM_CAP_X86_NOTIFY_VMEXIT when it's
reported as supported. However, KVM reports the support of this CAP just
based on the hardware capability but doesn't differentiate between VMX
and TDX. This leads to the issue that userspace can enable this cap for
TDX guests without getting an error, but the feature is not actually
enabled because KVM doesn't call the TDX module API to program the
relevant TD VMCS fields.
Enable Notify VM exit for TDX guests by:
- Invoking TDX module API calls to set NOTIFY_VM_EXITING and Notify
Window in TD VMCS. It's done in tdx_vcpu_init() where other TD VMCS
bits are set. Since TDX vCPU cannot be reset, it only needs to be
configured once when initializing the TDX vCPU.
- Adding corresponding exit handler for TDX Notify VM Exit.
Notify VM exit can happen when executing the IRET instruction. If the
IRET unblocks the NMI blocking state, bit 12 of the exit qualification
is set. In this case, the VMM needs to restore the "blocked by NMI" state
when it decides to re-enter the guest. For TDX, KVM cannot manage the
GUEST_INTERRUPTIBILITY_INFO and it's TDX module's responsibility to
handle it. Extract the common part without NMI blocking handling into a
helper in common.h so that it can be shared between VMX and TDX.
Note, KVM uses "pre-production" terminology for the feature formally called
Notify VM-Exit. All public versions of the SDM refer to the feature as
Instruction Timeout. This will be remedied in the near future, for now,
use KVM's terminology for consistency.
Note, #2, there is no enumeration bit for Notify VM exit by TDX module
because all TDX modules support it, and allow to set the corresponding
TD VMCS fields as long as the hardware supports the feature.
Fixes: 161d34609f9b ("KVM: TDX: Make TDX VM type supported") Cc: stable@vger.kernel.org Signed-off-by: Xiaoyao Li <xiaoyao.li@intel.com> Reviewed-by: Rick Edgecombe <rick.p.edgecombe@intel.com> Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com> Link: https://patch.msgid.link/20260819094903.3060020-2-xiaoyao.li@intel.com Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Check if there is valid exit info based on vp_enter_ret instead of relying
on the clobbered Exit Reason, in tdx_get_exit_info().
Current KVM uses "Exit Reason is not equal to the synthesized invalid
Exit Reason, -1u," as the condition to identify there is a real TD Exit
and valid exit infos. However, there is one issue with this approach:
KVM updates the Exit Reason to the synthesized invalid Exit Reason for
real EPT MISCONFIG as well. This is a false positive for real EPT
MISCONFIG, which has valid exit infos.
Though the issue can be addressed by changing the handling for real EPT
MISCONFIG to not update the Exit Reason to the synthesized one, relying
on the clobbered Exit Reason itself is brittle. Instead, check
vp_enter_ret directly to identify if it is a valid Exit Reason.
Ensure all reserved flags are zeroed when setting the msg_pending flag as
part of SynIC message delivery, as setting only msg_pending can leak seven
bits of kernel stack data to the guest. E.g. gcc typically uses a
bitwise-OR to set the msg_pending flag, without initializing the on-stack
variable.
Precisely zero the flags, e.g. as opposed to zeroing the entire structure,
to make it somewhat more obvious that the flags *need* to be zeroed. E.g.
it would be quite easy to misread the initial read of guest memory and
think that zeroing the entire structure is completely superfluous.
kvm_zap_gfn_range() walks memslots and rmaps and may drop mmu_lock to
reschedule. Callers that do not already hold kvm->srcu (or slots_lock)
can race with memslot deletion: synchronize_srcu_expedited() does not
wait, kvm_free_memslot() frees the old slot and its rmap, and the zap
resumes on freed memory.
The VFIO noncoherent-DMA path hits this by zapping the entire GPA
space without SRCU. KASAN reported:
BUG: KASAN: vmalloc-out-of-bounds in slot_rmap_walk_next+0x82/0x1c0
Read of size 8 at addr ffffc900005c1008
Call Trace:
slot_rmap_walk_next+0x82/0x1c0
__kvm_rmap_zap_gfn_range+0x17a/0x280
kvm_zap_gfn_range+0x2a6/0x6a0
kvm_vfio_set_attr+0x576/0x770
kvm_device_ioctl+0x1ff/0x3b0
Take SRCU inside kvm_zap_gfn_range() so every caller is covered.
Nesting with an existing kvm->srcu critical section is fine; the
helper uses a local index. Drop the now-redundant SRCU pair from
__kvm_set_or_clear_apicv_inhibit().
Fixes: 362ff6dca541 ("KVM: x86/mmu: Zap KVM TDP when noncoherent DMA assignment starts/stops") Cc: stable@vger.kernel.org Signed-off-by: Chengfeng Ye <nicoyip.dev@gmail.com> Reviewed-by: Kai Huang <kai.huang@intel.com> Link: https://patch.msgid.link/20260825183924.237727-1-nicoyip.dev@gmail.com Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Request event processing whenever kvm_queue_exception_vmexit() queues an
exception VM-Exit. Most callers reach the helper through
kvm_multiple_exception(), which already sets KVM_REQ_EVENT, but nested
asynchronous page faults call it directly.
Without another event request, vcpu_enter_guest() skips
kvm_check_and_inject_events() and re-enters L2. A subsequent VM-Exit can
clear the pending exception while reconstructing vectoring state, losing
the synthetic #PF VM-Exit. The APF reason remains PAGE_NOT_PRESENT and a
later regular #PF can consume the stale value.
Setting KVM_REQ_EVENT in the common helper ensures that L1 receives the
queued VM-Exit before KVM re-enters L2.
Fixes: 7709aba8f716 ("KVM: x86: Morph pending exceptions to pending VM-Exits at queue time") Cc: stable@vger.kernel.org Signed-off-by: Loc Nguyen <loc@ryzome.com> Link: https://patch.msgid.link/20261002053723.3275202-2-loc@ryzome.com Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Preserve Bus Lock Detect if it isn't present in the payload. Unlike RTM,
which is explicitly modified on every #DB, DR6.BLD is modified if and only
if the #DB is due to a (coincident) Bus Lock Detect.
Per Intel's SDM:
Other debug exceptions do not modify this bit.
And AMD's APM
All other #DB exceptions leave DR6[BLD] unmodified,
Restore the explicit RTM handling that was clobbered by commit 9a3ecd5e2aa1 ("KVM: X86: Rename DR6_INIT to DR6_ACTIVE_LOW"), as the
semantics of any given bit are feature specific, i.e. are obviously not
consistent for active-high versus active-low. In other words, while the
bug was technically introduced when support for Bus Lock Detect was added,
it's really that attempt to genericize handling of active-low bits that set
KVM up to fail.
Fixes: e8ea85fb280e ("KVM: X86: Add support for the emulation of DR6_BUS_LOCK bit") Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260930193634.3415089-2-seanjc@google.com Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Force a vCPU out of its hardware-tracked halted state when userspace puts
the vCPU into any state other than HALTED via KVM_SET_MP_STATE, i.e. clear
VMCS.GUEST_ACTIVITY_STATE if it says the vCPU is halted. Add an optional
kvm_x86_ops hook to do the clearing, as SVM has no equivalent VMCB field.
When HLT-exiting is disabled for a VM (KVM_CAP_X86_DISABLE_EXITS with
KVM_X86_DISABLE_EXITS_HLT, e.g. QEMU's "-overcommit cpu-pm=on"), a guest
HLT halts the physical CPU instead of exiting to KVM, and hardware saves
GUEST_ACTIVITY_STATE=HLT into the VMCS on the next VM-Exit. That field is
sticky: it survives VM-Exit/VM-Enter and is only cleared by vmx_clear_hlt()
on event injection, or by vmx_vcpu_reset() on INIT / vCPU creation.
Nothing clears it on a userspace-driven state change. KVM_SET_REGS only
writes the software register cache and KVM_SET_MP_STATE only writes
vcpu->arch.mp_state; kvm_vcpu_running() likewise consults software state
only. A VMM that emulates a machine reset therefore ends up with a vCPU
that KVM happily VM-Enters while hardware refuses to fetch instructions.
Reproduce with a Linux guest by triggering a panic/kdump on a non-boot
vCPU: nmi_shootdown_cpus() parks the other vCPUs -- including vCPU0 -- in
crash_nmi_callback(), which does local_irq_disable() followed by a bare
HLT. The capture kernel then resets the machine via port 0xCF9. QEMU
rewrites RIP to 0xfff0 and sets mp_state to RUNNABLE, but vCPU0's
GUEST_ACTIVITY_STATE is still HLT, so the BSP never executes the reset
vector, never sends SIPIs, and the entire VM hangs at "reboot: machine
restart" forever. Only destroying and recreating the VM recovers it.
Key off "not HALTED" rather than "is RUNNABLE" as HALTED is the only target
state for which retaining the hardware halted state is meaningful; every
other transition either can't be reached on VMX, e.g. AP_RESET_HOLD, or
goes through vmx_vcpu_reset(), which stuffs GUEST_ACTIVITY_ACTIVE anyways.
This also captures the intent more directly: KVM only ever drops the
hardware halted state, it never installs it.
Hook KVM_SET_MP_STATE rather than the VM-Enter path so that the clearing
is driven by an explicit userspace declaration, and so that no work is
added to vmx_vcpu_run(). Gate the call on kvm_hlt_in_guest() even though
vmx_clear_hlt() checks it as well, as the check is cheap and makes it
obvious in common code when the hardware state can exist at all.
Do nothing for TDX vCPUs as TDX doesn't support disabling HLT-exiting and
KVM cannot access a TD's VMCS, i.e. the VMCS used to run the vCPU can't
in HALTED state, and even if it were, KVM can't change it.
Fixes: caa057a2cad6 ("KVM: X86: Provide a capability to disable HLT intercepts") Cc: stable@vger.kernel.org Tested-by: Qinguang Chen <chenqinguang@kuaishou.com> Signed-off-by: Zhiping Du <duzhiping@kuaishou.com> Signed-off-by: Keqiang Duan <duankeqiangcym@gmail.com> Reviewed-by: Chao Gao <chao.gao@intel.com> Link: https://patch.msgid.link/20260820123512.87236-1-duankeqiangcym@gmail.com
[sean: drop claim that spurious wakeups are ok, massage changelog] Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
kvm_gpc_check() checks the cached memslot generation against the current
one, which dereferences kvm->memslots and therefore requires kvm->srcu to
be held. __kvm_xen_has_interrupt() does not take it.
Most callers do happen to hold kvm->srcu, but not all of them:
- kvm_emulate_halt() on the VM-Exit path, via kvm_vcpu_has_events() and
kvm_cpu_has_extint(). vcpu_enter_guest() drops the vCPU's SRCU lock
before entering the guest, so it is not held on the way back out.
- kvm_vcpu_block() -> kvm_vcpu_check_block() -> kvm_arch_vcpu_runnable(),
which is the case the existing comment in this function describes.
Use guard(srcu) so that the three existing early returns don't each need
an explicit unlock. SRCU read sections nest, so this is harmless on the
paths which already hold it, and srcu_read_lock() does not sleep, so it
is also safe in the atomic case which this function already handles.
Fixes: 7caf9571563e ("KVM: x86/xen: Use gfn_to_pfn_cache for vcpu_info") Cc: stable@vger.kernel.org Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Assisted-by: Claude:claude-mythos-5 Reviewed-by: Paul Durrant <paul@xen.org> Link: https://patch.msgid.link/20260831213632.81023-11-dwmw2@infradead.org Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Xen's map_guest_area() enforces that vcpu_info is aligned to
sizeof(xen_ulong_t). KVM has no such check, allowing a guest to
register vcpu_info at an arbitrary byte alignment.
Enforce unconditional 4-byte alignment regardless of the current
shinfo mode. This is sufficient because subsequent commits ensure
that all locked atomic operations on vcpu_info fields use at most
32-bit accesses. Return -ENXIO on failure, matching Xen's
map_guest_area() behaviour for unaligned requests.
Originally observed in review of an unrelated patch:
https://lore.kernel.org/all/20260604193554.1BA311F00893@smtp.kernel.org/
Bug the VM if KVM attempts to construct a CPU role with the should-be-
impossible combination of long mode being active without PAE paging being
enabled. KVM's MMU construction assumes that EFER.LMA can be set if and
only CR4.PAE is set, and will create a completely invalid MMU if that
assumption fails. FNAME(walk_addr_generic) already has sanity checks to
try and mitigate the fallout, but attempt to catch such bugs earlier, as
this is (at least) the second time KVM has had bugs that escaped into
FNAME(walk_addr_generic), and it's entirely possible the bad state could
cause problems elsewhere.
Extend the "EFER.LMA && !CR4.PAE" check, which exists largely to guard
against KVM configuring a paging32 MMU with more than 2 levels of paging,
with a very explicit check for exactly that: that KVM isn't trying to walk
more levels of paging than the MMU template provides. I.e. harden KVM
against all bugs that would cause KVM to generates accesses beyond the
bounds of guest_walker's arrays, regardless of how KVM ended up with the
misconfigured MMU.
Note, don't use w->cpu_role.base.level directly as the paging64 template
only provides two levels of page tables for PAE paging on 32-bit hosts, and
handles the third level by manually emulating the PDPTR access.
Preserve the negative return value from __vmx_handle_exit() when a bus
lock is detected, instead of always overwriting it with 0.
The purpose of bus_lock_detected handling is to force a userspace exit to
inform userspace that a bus lock happened. The negative return value can
achieve this purpose, and changing the negative value to 0 fails to return
an error to userspace. So, preserve the negative return value.
Fixes: fe6b6bc802b4 ("KVM: VMX: Enable bus lock VM exit") Cc: stable@vger.kernel.org Reported-by: sashiko-bot@kernel.org Closes: https://lore.kernel.org/all/20260805034602.5B2BB1F000E9@smtp.kernel.org/ Signed-off-by: Xiaoyao Li <xiaoyao.li@intel.com> Link: https://patch.msgid.link/20260819094903.3060020-7-xiaoyao.li@intel.com Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move handle_bus_lock_vmexit() from vmx.c to main.c to make it a shared
helper so that TDX can use it for Bus Lock VM exit handling as well.
Rename it to add vt_ prefix to reflect that it is a helper for both VMX
and TDX.
Cc: stable@vger.kernel.org Signed-off-by: Xiaoyao Li <xiaoyao.li@intel.com> Reviewed-by: Nikolay Borisov <nik.borisov@suse.com> Link: https://patch.msgid.link/20260819094903.3060020-8-xiaoyao.li@intel.com Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Poison to_vmx() in main.c and posted_intr.c so that attempting to interpret
the vCPU as a VMX vCPU will fail at compile time, as opposed to failing at
runtime, or worse corrupting state without outright failing.
Note, to_tdx() is buried in tdx.c, i.e. isn't broadly reachable, and so
doesn't need the same treatment as to_vmx().
Reviewed-by: Xiaoyao Li <xiaoyao.li@intel.com> Reviewed-by: Kai Huang <kai.huang@intel.com> Tested-by: Kai Huang <kai.huang@intel.com> Reviewed-by: Binbin Wu <binbin.wu@linux.intel.com> Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260904010353.3175819-2-seanjc@google.com Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>