| Age | Commit message (Collapse) | Author |
|---|
|
virtio_pmem_host_ack() publishes the device response by setting done and
waking the submitter. The submitter reads resp.ret after wait_event()
observes done.
Use smp_store_release() on done and smp_load_acquire() in the wait
condition so the response read is ordered after completion.
Signed-off-by: Li Chen <me@linux.beauty>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260630092338.2094628-10-me@linux.beauty>
|
|
KASAN reports slab-use-after-free in __wake_up_common():
BUG: KASAN: slab-use-after-free in __wake_up_common+0x114/0x160
Read of size 8 at addr ffff88810fdcb710 by task swapper/0/0
CPU: 0 UID: 0 PID: 0 Comm: swapper/0 Not tainted
6.19.0-next-20260220-00006-g1eae5f204ec3 #4 PREEMPT(full)
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Arch Linux
1.17.0-2-2 04/01/2014
Call Trace:
<IRQ>
dump_stack_lvl+0x6d/0xb0
print_report+0x170/0x4e2
? __pfx__raw_spin_lock_irqsave+0x10/0x10
? __virt_addr_valid+0x1dc/0x380
kasan_report+0xbc/0xf0
? __wake_up_common+0x114/0x160
? __wake_up_common+0x114/0x160
__wake_up_common+0x114/0x160
? __pfx__raw_spin_lock_irqsave+0x10/0x10
__wake_up+0x36/0x60
virtio_pmem_host_ack+0x11d/0x3b0
? sched_balance_domains+0x29f/0xb00
? __pfx_virtio_pmem_host_ack+0x10/0x10
? _raw_spin_lock_irqsave+0x98/0x100
? __pfx__raw_spin_lock_irqsave+0x10/0x10
vring_interrupt+0x1c9/0x5e0
? __pfx_vp_interrupt+0x10/0x10
vp_vring_interrupt+0x87/0x100
? __pfx_vp_interrupt+0x10/0x10
__handle_irq_event_percpu+0x17f/0x550
? __pfx__raw_spin_lock+0x10/0x10
handle_irq_event+0xab/0x1c0
handle_fasteoi_irq+0x276/0xae0
__common_interrupt+0x65/0x130
common_interrupt+0x78/0xa0
</IRQ>
virtio_pmem_host_ack() wakes a request that has already been freed by the
submitter.
This happens when the request token is still reachable via the virtqueue,
but virtio_pmem_flush() returns and frees it.
Fix the token lifetime by refcounting struct virtio_pmem_request.
virtio_pmem_flush() holds a submitter reference, and the virtqueue holds an
extra reference once the request is queued. The completion path drops the
virtqueue reference, and the submitter drops its reference before
returning.
Fixes: 6e84200c0a29 ("virtio-pmem: Add virtio pmem driver")
Cc: stable@vger.kernel.org
Signed-off-by: Li Chen <me@linux.beauty>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260630092338.2094628-9-me@linux.beauty>
|
|
Use READ_ONCE()/WRITE_ONCE() for the wait_event() flags (done and
wq_buf_avail). They are observed by waiters without pmem_lock, so make
the accesses explicit single loads/stores and avoid compiler
reordering/caching across the wait/wake paths.
Acked-by: Pankaj Gupta <pankaj.gupta.linux@gmail.com>
Signed-off-by: Li Chen <me@linux.beauty>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260630092338.2094628-8-me@linux.beauty>
|
|
virtio_pmem_host_ack() reclaims virtqueue descriptors with
virtqueue_get_buf(). The -ENOSPC waiter wakeup is tied to completing the
returned token. If token completion is skipped for any reason, reclaimed
descriptors may not wake a waiter and the submitter may sleep forever
waiting for a free slot. Always wake one -ENOSPC waiter for each virtqueue
completion before touching the returned token.
Signed-off-by: Li Chen <me@linux.beauty>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260630092338.2094628-7-me@linux.beauty>
|
|
virtio_pmem_flush() can run from pmem_submit_bio() while filesystem IO
is waiting on the flush completion. The request object allocation can
sleep, but it should not enter filesystem or block IO reclaim from this
flush path.
Use GFP_NOIO for the request allocation. The virtqueue descriptor
allocation still uses GFP_ATOMIC because it runs under pmem_lock.
Signed-off-by: Li Chen <me@linux.beauty>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260630092338.2094628-6-me@linux.beauty>
|
|
pmem_submit_bio() passes the parent bio to nvdimm_flush() for
REQ_FUA. For virtio-pmem this makes async_pmem_flush() allocate
and submit a child PREFLUSH bio chained to the parent.
That child allocation is in the block submit path. Making it
blocking with GFP_NOIO can consume the same global bio mempool that
submit_bio() uses, while making it GFP_ATOMIC can fail under
pressure. A forced failure of the child allocation produced:
virtio_pmem: forcing child bio allocation failure for test
Buffer I/O error on dev pmem0, logical block 0, lost sync page write
EXT4-fs (pmem0): I/O error while writing superblock
EXT4-fs (pmem0): mount failed
Avoid the child bio without turning REQ_FUA into a synchronous
submit-path wait. Let provider flush callbacks return
NVDIMM_FLUSH_ASYNC after taking ownership of parent bio completion.
pmem_submit_bio() returns in that case, and virtio-pmem queues an
ordered WQ_MEM_RECLAIM work item that runs the existing host flush
path and completes the parent bio.
This keeps the asynchronous completion model of the child-bio path
while removing the child bio allocation from the submit path.
Signed-off-by: Li Chen <me@linux.beauty>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260630092338.2094628-5-me@linux.beauty>
|
|
pmem_submit_bio() handles flush-only bios before and after the data
loop. Keep dataless bios out of bio_for_each_segment() so the data path
only walks bios that actually carry bvec data.
Signed-off-by: Li Chen <me@linux.beauty>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260630092338.2094628-4-me@linux.beauty>
|
|
pmem_submit_bio() records a REQ_PREFLUSH error, but continues to copy the
bio data and can later overwrite the error with a successful REQ_FUA flush.
That lets data writes run after a failed preflush and can complete the bio
successfully despite the failed ordering barrier.
Run the REQ_PREFLUSH flush synchronously before touching the bio data and
complete the bio with the flush error if it fails. Keep asynchronous flush
chaining for REQ_FUA. At that point, data copy has completed and the parent
bio can wait for the chained flush bio.
Signed-off-by: Li Chen <me@linux.beauty>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260630092338.2094628-3-me@linux.beauty>
|
|
nvdimm_flush() maps provider flush failures to -EIO. Keep that default
because provider callbacks can report host-side or backend failures that
should remain generic I/O errors to the guest.
Guest-side allocation failures should not be reported as I/O errors. In the
virtio-pmem path, the flush request allocation can fail with -ENOMEM before
any request is submitted to the host. Mapping that to -EIO makes resource
pressure look like media failure.
Preserve -ENOMEM from provider callbacks and continue to map other non-zero
provider failures to -EIO. The generic flush path still returns 0, and
pmem_submit_bio() already converts errno values to block status for bio
completion.
Suggested-by: Pankaj Gupta <pankaj.gupta.linux@gmail.com>
Signed-off-by: Li Chen <me@linux.beauty>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260630092338.2094628-2-me@linux.beauty>
|
|
In vp_find_vqs_intx(), the admin vq was set up using the local
queue_idx counter instead of avq->vq_index (the actual queue index
obtained from the device). This differs from vp_find_vqs_msix() which
correctly uses avq->vq_index. Using the wrong index causes the admin
virtqueue to be mapped to an incorrect hardware queue.
Fix it by using avq->vq_index consistent with the msix path.
Fixes: af22bbe1f4a5 ("virtio: create admin queues alongside other virtqueues")
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Message-ID: <20260629033538.2476-1-lirongqing@baidu.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
virtio_mem_bbm_for_each_bb() and virtio_mem_bbm_for_each_bb_rev()
accept a '_vm' parameter to allow callers to pass any variable name
referring to the virtio_mem instance. However, the 'for' loop
initializer and part of the loop condition use the bare name 'vm'
instead of the macro parameter '_vm'.
Fix by replacing all bare 'vm->' references inside the macros with
the '_vm' parameter, and wrap in parentheses following kernel macro
conventions.
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Acked-by: David Hildenbrand (Arm) <david@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260630045952.2188-1-lirongqing@baidu.com>
|
|
The @get_uid tag in the virtio_dma_buf_ops kdoc comment is a typo;
the actual field name is get_uuid.
Fixes: a0308938ec81 ("virtio: add dma-buf support for exported objects")
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260629033146.2209-1-lirongqing@baidu.com>
|
|
vdpasim_dma_map() updates the IOTLB and the passthrough (iommu_pt)
state under iommu_lock. vdpasim_dma_unmap() clears iommu_pt and
resets the IOTLB before taking iommu_lock, then deletes the mapping
while holding the lock.
A concurrent dma_map(), dma_unmap(), or reset path that also touches
the same address space can therefore observe or modify the IOTLB and
iommu_pt state without consistent locking.
Perform the passthrough transition and range deletion under the same
iommu_lock scope, matching dma_map().
Tested-on: openEuler VM (6.16.8, /usr/src/linux-6.16.8)
Tested-by: Xiong Weimin <xiongweimin@kylinos.cn>
Signed-off-by: Xiong Weimin <xiongweimin@kylinos.cn>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260626020545.607600-3-15927021679@163.com>
|
|
vdpasim_kick_vq() sets pending_kick when a virtqueue is kicked while
the device is suspended (!running but DRIVER_OK). vdpasim_resume()
later replays kicks for all virtqueues when pending_kick is set.
vdpasim_do_reset() clears running and status but leaves pending_kick
unchanged. If a kick is deferred during suspend and the device is
reset before resume, a later resume can spuriously kick every
virtqueue even though no new work was queued after reset.
Clear pending_kick in vdpasim_do_reset() together with the other
device state that must not survive a reset.
Tested-on: openEuler VM (6.16.8, /usr/src/linux-6.16.8)
Tested-by: Xiong Weimin <xiongweimin@kylinos.cn>
Signed-off-by: Xiong Weimin <xiongweimin@kylinos.cn>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260626020545.607600-2-15927021679@163.com>
|
|
vhost_net_test has --batch in longopts, but not in help.
The parser never handles 'b', so --batch hits assert(0).
Remove the unsupported option.
Signed-off-by: Yichong Chen <chenyichong@uniontech.com>
Acked-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <E091F15D8FBE8F16+20260618100254.513806-1-chenyichong@uniontech.com>
|
|
The clear_user() call in VHOST_GET_FEATURES_ARRAY incorrectly starts
at argp, which is the beginning of the features array, overwriting the
data just written by copy_to_user(). It should start after the copied
elements at argp + copied * sizeof(u64) to only zero the trailing
unused space.
Use size_mul() for both the offset and length calculations so the
arithmetic stays consistent with the surrounding code and remains
overflow-safe.
Fixes: 333c515d1896 ("vhost-net: allow configuring extended features")
Signed-off-by: Yufeng Wang <wangyufeng@kylinos.cn>
Acked-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260626070438.59149-1-r4o5m6e8o@163.com>
|
|
Like tell_host(), stats_handle_request() ignores the return value of
virtqueue_add_outbuf() and kicks the queue regardless. The same "we
should always be able to add one buffer to an empty queue" assumption
does not hold once the virtqueue has been broken (e.g. on device
shutdown), where the add fails with -EIO. Unlike tell_host() it does
not wait_event() afterwards so it cannot hang, but it still kicks a
queue with nothing queued.
Warn and bail out on failure, mirroring tell_host() and
virtballoon_free_page_report().
Suggested-by: David Hildenbrand <david@kernel.org>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: David Hildenbrand (Arm) <david@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260624154001.2733242-1-den@openvz.org>
|
|
tell_host() ignores the return value of virtqueue_add_outbuf() and goes
on to kick the queue and wait_event() for the host's ack. The comment
claims "We should always be able to add one buffer to an empty queue",
but that does not hold once the virtqueue has been broken (e.g. on
device shutdown): the add then fails with -EIO and the following
wait_event() would block forever on a buffer the host can never return.
Warn and bail out on failure, mirroring virtballoon_free_page_report().
Suggested-by: David Hildenbrand <david@kernel.org>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260624140846.2616797-5-den@openvz.org>
|
|
Commit 8bd2fa086a04 ("virtio: break and reset virtio devices on
device_shutdown()") added a generic virtio bus .shutdown handler that
breaks and resets every virtio device during device_shutdown(), i.e. on
reboot and kexec.
virtio_balloon provides no .shutdown of its own, so that generic path
runs while the balloon's asynchronous work is still armed. Once the
device has been broken, virtqueue_add_inbuf() in
virtballoon_free_page_report() returns -EIO and trips its
WARN_ON_ONCE(). On a kernel booted with panic_on_warn that turns an
ordinary reboot, for example a kexec based upgrade, into a fatal panic
in the middle of device_shutdown(), so the machine never reaches the
new kernel.
Relaxing that single WARN_ON_ONCE() would only hide the symptom: the
inflate/deflate and OOM paths do not warn, they call
wait_event(vb->acked, ...) and would instead block forever on a broken
queue that can no longer complete. The device has to be quiesced, not
just kept quiet.
Add a .shutdown handler that quiesces the balloon via the shared
virtballoon_quiesce() helper while the device is still alive, and only
then breaks and resets it via virtio_device_shutdown(). Unlike
virtballoon_remove() the balloon workqueue is not destroyed, as shutdown
does not free the device and cancel_work_sync() together with stop_update
already prevent any further work from being queued.
Fixes: 8bd2fa086a04 ("virtio: break and reset virtio devices on device_shutdown()")
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: David Hildenbrand (Arm) <david@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260624140846.2616797-4-den@openvz.org>
|
|
virtballoon_remove() stops all of the balloon's asynchronous work (the
free page reporting worker, the inflate/deflate and stats workers, the
OOM notifier and the free page shrinker) before tearing the device
down. A following change needs the same teardown from a .shutdown
handler, so move it into a virtballoon_quiesce() helper.
No functional change.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: David Hildenbrand (Arm) <david@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260624140846.2616797-3-den@openvz.org>
|
|
The generic virtio bus .shutdown handler, virtio_dev_shutdown(), breaks
and resets a device once it has established that the driver has no
.shutdown of its own. A driver that does implement .shutdown, to quiesce
its own activity first, still needs the same break and reset afterwards
and would otherwise have to open code it.
Factor the break + synchronize_cbs + reset sequence out of
virtio_dev_shutdown() into an exported virtio_device_shutdown() helper so
such drivers can reuse it instead of duplicating the core logic.
No functional change.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: David Hildenbrand (Arm) <david@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260624140846.2616797-2-den@openvz.org>
|
|
Add VIRTIO_ID_MEDIA definition for virtio-media.
Signed-off-by: Albert Esteve <aesteve@redhat.com>
Message-ID: <20260310-virtio-media-id-v1-1-be211bcf682b@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
There are two cases of sending a __le64 type to a print function
so fix this by adding le64_to_cpu() which fixes the following
(prototype) sparse warnings:
drivers/crypto/virtio/virtio_crypto_skcipher_algs.c:234:17: warning: incorrect type in argument 3 (different base types)
drivers/crypto/virtio/virtio_crypto_skcipher_algs.c:234:17: expected unsigned long long
drivers/crypto/virtio/virtio_crypto_skcipher_algs.c:234:17: got restricted __le64 [usertype] session_id
drivers/crypto/virtio/virtio_crypto_akcipher_algs.c:196:17: warning: incorrect type in argument 3 (different base types)
drivers/crypto/virtio/virtio_crypto_akcipher_algs.c:196:17: expected unsigned long long
drivers/crypto/virtio/virtio_crypto_akcipher_algs.c:196:17: got restricted __le64 [usertype] session_id
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Message-ID: <20260622150322.526375-1-ben.dooks@codethink.co.uk>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
virtio_crypto_dataq_akcipher_callback() sets the result length from the
device-reported response length without bounding it to the destination
buffer, which was allocated for the original request length.
sg_copy_from_buffer() then reads that many bytes from the destination
buffer; a backend reporting a larger length over-reads adjacent kernel
heap into the caller's scatterlist (an out-of-bounds read).
Clamp the reported length to the originally requested destination length.
A conforming device reports no more than that, so valid results are
unaffected.
Fixes: a36bd0ad9fbf ("virtio-crypto: adjust dst_len at ops callback")
Cc: stable@vger.kernel.org
Signed-off-by: Bryam Vargas <hexlabsecurity@proton.me>
Message-ID: <20260622-b4-disp-3a2c09a8-v2-1-d1a809281db4@proton.me>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
When transmission fails in virtio_transport_send_pkt_info, the msg_iter
might have been partially advanced. If we don't restore it, the next
attempt to send data will use an incorrect iterator state, leading to
desync and warnings like "send_pkt() returns 0, but X expected".
Specifically, this can happen in the following scenario, triggered by
the syzkaller repro:
1. A write-only VMA (PROT_WRITE only) is partially populated by a
prior TUN write that failed with -EIO but still faulted in some
pages).
2. A vsock sendmmsg call with MSG_ZEROCOPY requests transmission of a
buffer from this VMA.
3. The first packet (64KB) is sent successfully because the pages are
populated.
4. The second packet allocation fails because GUP fast pins the first page
but GUP slow fails on the next unpopulated page due to PROT_WRITE-only
permissions.
5. The iterator is advanced by the partially successful GUP (68KB total
advanced: 64KB from first packet + 4KB from second), but the send loop
breaks and only reports 64KB sent. This creates a 4KB desync.
6. The next retry starts with a non-zero iov_offset, disabling zerocopy
and falling back to copy mode.
7. In copy mode, the transmission succeeds for the next packets but
exhausts the iterator early because of the desync.
8. The final retry sees an empty iterator but zerocopy is re-enabled
(offset resets). It attempts to send the remaining bytes with zerocopy
but pins 0 pages, creating an empty packet.
9. The transport sends the empty packet, triggering the warning because
the returned bytes (header only) do not match the expected payload size.
10. The loop continues to spin, allocating ubuf_info each time, eventually
exhausting sysctl_optmem_max and returning -ENOMEM to userspace.
Restore msg_iter to its original state before the packet allocation
and transmission attempt if they fail.
Fixes: e0718bd82e27 ("vsock: enable setting SO_ZEROCOPY")
Reported-by: syzbot+28e5f3d207b14bae122a@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=28e5f3d207b14bae122a
Assisted-by: gemini:gemini-3.1-pro
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Octavian Purdila <tavip@google.com>
Message-ID: <20260622222757.2130402-3-tavip@google.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
Export iov_iter_restore so that it can be used by modules.
This is needed by the virtio vsock transport (which can be built as a
module) to restore the msg_iter state when transmission fails.
Acked-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Octavian Purdila <tavip@google.com>
Message-ID: <20260622222757.2130402-2-tavip@google.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
vdpasim_create() leaves vdpasim->worker as an ERR_PTR when
kthread_run_worker() fails. The error path then drops the device
reference, which releases the partially initialized simulator.
vdpasim_free() unconditionally passes the worker pointer to
kthread_destroy_worker(), so the ERR_PTR is dereferenced and can trigger
a general protection fault.
Store the worker error, clear the pointer, and only clean up the worker
when it was successfully initialized. Also make the release path tolerate
partially initialized objects by guarding virtqueue and IOTLB cleanup,
since the same release path can be reached from other initialization
failures.
I found this bug myself, though the patch was written with AI assistance.
Fixes: 76acfa7bc54f ("vdpa_sim: use kthread worker")
Assisted-by: OpenAI-Codex:GPT-5
Reviewed-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Linfeng Sun <linfeng.sun.dev@gamil.com>
Message-ID: <20260620100959.2070316-1-slf@hdu.edu.cn>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
The page reporting callback submits an sg list to the reporting
virtqueue. With VIRTIO_RING_F_INDIRECT_DESC negotiated and
total_sg > 1 (which it typically is), virtqueue_add reports it to the
host by allocating an indirect descriptor via kmalloc(GFP_KERNEL).
This is not pretty: the reporting worker isolates potentially hundreds
of MB of free pages from the buddy allocator (reported pages are at
least pageblock_order, and the sg can contain up to
PAGE_REPORTING_CAPACITY entries of varying orders). As the result,
very theoretically, the kmalloc might trigger OOM when we have in fact a
ton of free memory.
Clear VIRTIO_RING_F_INDIRECT_DESC, to avoid using indirect descriptors.
Fixes: b0c504f15471 ("virtio-balloon: add support for providing free page reports to host")
Assisted-by: Claude:claude-opus-4-6
Acked-by: David Hildenbrand (Arm) <david@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <73fac8a629fd9aca7bb3265ac243a769c28af25d.1783232420.git.mst@redhat.com>
|
|
Virtio MMIO transport version 3 allows device reset to complete
asynchronously. Unlike version 2, where writing zero to Status must
complete the reset before the write returns, version 3 requires the
driver to poll Status until it reads back zero before considering reset
complete.
Update virtio-mmio accordingly: accept transport version 3 and, during
reset, wait for Status to become zero. Keep the polling loop unbounded,
consistent with virtio-pci, since the reset callback does not return an
error code.
Signed-off-by: Peter Hilber <peter.hilber@oss.qualcomm.com>
Link: https://github.com/oasis-tcs/virtio-spec/commit/bb1dd2e1fe89b862f38f15873d835a698b196f89
Message-ID: <20260605142921.2824-1-peter.hilber@oss.qualcomm.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
ntfs_setxattr() updates ctime and marks the inode dirty even when
the operation fails. A failed setxattr(2) must not change file
metadata.
Update ctime only on success.
Fixes: fc053f05ca28 ("ntfs: add reparse and ea operations")
Signed-off-by: Baolin Liu <liubaolin@kylinos.cn>
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
The NTFS boot sector stores the MFT and MFTMirr locations as unsigned
64-bit LCNs, but parse_ntfs_boot_sector() decoded them into an s64.
A crafted high-bit value could therefore become negative and pass
the existing upper-bound check. The invalid value then propagated into
the MFT zone allocator and could result in an out-of-bounds access to
lcn_empty_bits_per_page.
Fixes: 11ccc9107dc4 ("ntfs: update runlist handling and cluster allocator")
Reported-by: Robert Morris <rtm@csail.mit.edu>
Closes: https://lore.kernel.org/all/57514.1787000602@localhost
Cc: stable@vger.kernel.org
Signed-off-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
ntfs_read_iomap_begin_resident() walks the MFT record through
ntfs_attr_lookup() -> ntfs_attr_find() without taking ni->mrec_lock,
while ntfs_attr_record_resize(), ntfs_make_room_for_attr() and
ntfs_resident_attr_record_add() memmove() the same base_ni->mrec buffer
under that lock. map_mft_record() only takes a reference and does not
serialize, so the reader can observe torn attribute length and offset
fields while a writer is relocating the records.
KCSAN reports the race between the mmap read fault path and both link()
and unlink():
BUG: KCSAN: data-race in ntfs_attr_find / ntfs_attr_record_resize
write to 0xffff888100af1018 of 4 bytes by task 96 on cpu 1:
ntfs_attr_record_resize+0xd2/0x130
ntfs_attr_record_rm+0xad/0x530
ntfs_delete+0x224/0x640
ntfs_unlink+0x14d/0x280
vfs_unlink+0x157/0x520
read to 0xffff888100af1018 of 4 bytes by task 95 on cpu 0:
ntfs_attr_find+0x104/0x5b0
ntfs_attr_lookup+0x39c/0x10c0
ntfs_read_iomap_begin_resident+0xc6/0x230
ntfs_read_iomap_begin+0x5d/0xa0
iomap_iter+0x2e2/0x6e0
iomap_read_folio+0x147/0x2a0
ntfs_read_folio+0x108/0x170
filemap_read_folio+0x35/0x100
filemap_fault+0x993/0x1000
value changed: 0x00000250 -> 0x000001f0
The address is mrec + 0x18, i.e. mft_record.bytes_in_use, and the change
is the 96 bytes of one $FILE_NAME attribute being removed.
Keep base_ni->mrec_lock from the resident read iomap lookup through
iomap_end(). This protects both the attribute walk and the subsequent copy
from iomap->inline_data, which points into the MFT record. The non-resident
path is left alone: ntfs_lookup() already holds the directory inode's
mrec_lock when it reads an index folio through read_mapping_folio(), and
taking the lock in the shared wrapper deadlocks there with recursive locking
on mrec_lock. The comment above the read_mapping_folio() call in
fs/ntfs/dir.c notes the same hazard.
The seek path uses the same lookup helper but does not dereference
iomap->inline_data. Release the lock before returning from that path,
whereas the regular read path records base_ni in iomap->private and releases
the lock from its iomap_end() callback.
Tested with a reproducer that faults in a 16-byte resident file while
another thread runs link()/unlink() on it. Before: 40 KCSAN reports in
about one second. After: no reports in 180 seconds over 206,090 read
iterations and 423,540 link/unlink cycles. A PROVE_LOCKING build shows no
lockdep splat with the same reproducer running for 60 seconds.
Fixes: b041ca562526 ("ntfs: update iomap and address space operations")
Link: https://lore.kernel.org/all/20260725042421.109599-1-wonju345@naver.com/
Signed-off-by: Hyeontae Lee <wonju345@naver.com>
Co-developed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
The mapping pairs decoder validates that the starting LCN is within the
volume but does not check if the run extends beyond the volume boundary.
A malformed NTFS image with a crafted mapping pairs array could cause
the kernel to access memory beyond the volume boundary, potentially leading
to memory corruption and privilege escalation.
Add validation to ensure lcn + length stays within nr_clusters.
Cc: stable@vger.kernel.org
Fixes: b4be3a47f8ba4 ("ntfs: bound the free-cluster bitmap scan to the volume")
Signed-off-by: Hongling Zeng <zenghongling@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
Allow a resident attribute record to move to an extent MFT record when the
base record needs room for an attribute list. Retry the root conversion
after creating the list, but do not relocate a root that is already
external.
Roll the root back to the base record if persisting the attribute list
fails, and free extent MFT records left empty by relocation or rollback.
Also preserve bitmap allocation errors in index operations.
Fixes: af0db57d4293 ("ntfs: update inode operations")
Reported-by: yi <691464208@qq.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
ntfs_attr_update_meta() shifts the attribute name when converting between
non-sparse and sparse attributes. Converting to sparse also adds the
compressed_size field before the name and mapping pairs, requiring eight
additional bytes in the attribute record.
However, the validator does not check that name_offset is within safe
boundaries for these operations or that the additional space is available.
A malicious MFT record could set name_offset such that:
1. The name is positioned at the very end of a non-sparse attribute.
Converting to sparse would shift the name forward by 8 bytes,
writing beyond the attribute boundary.
2. The name overlaps with the mapping pairs, causing corruption during
conversion.
Add validation to ensure:
- For named attributes, name_offset is within valid bounds
- Name does not extend beyond the attribute or overlap with mapping pairs
- For non-sparse, non-compressed attributes, eight bytes are available
after mapping_pairs_offset for the compressed_size field
The space check also covers unnamed attributes, for which name_offset = 0
is valid and no name range needs to be checked.
Fixes: 7e2a1c554bc4 ("ntfs: Fix min_len for compressed/sparse attributes in ntfs_non_resident_attr_value_is_valid()")
Cc: stable@vger.kernel.org
Signed-off-by: Hongling Zeng <zenghongling@kylinos.cn>
Co-developed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
Avoid redundant 'strlen()' and use the convenient 'strreplace()'
to simplify 'ntfs_reparse_set_native_symlink()'.
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
Reject setxattr of the reserved $LXUID, $LXGID, $LXMOD and $LXDEV names
from userspace unless the caller has CAP_SYS_ADMIN.
Fixes: fc053f05ca28 ("ntfs: add reparse and ea operations")
Signed-off-by: Pisit Preechapramoth <kml.delusion501@slmail.me>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
vol->lcn_empty_bits_per_page is sized from vol->nr_clusters at mount, but
ntfs_cluster_alloc() bounds its scan of that array by the size of $Bitmap.
Those are independent on-disk quantities and the mount-time check only
rejects a $Bitmap that is too small, so an image whose $Bitmap covers more
clusters than the volume has lets the scan index past the array. A run
whose LCN lies in that gap takes the allocator straight there, since the
caller passes the file's own last LCN as its locality hint. KASAN reports
a slab out-of-bounds read when a file on such a volume is extended.
Clamp the scan to what that array covers, mirroring the max_index
calculation the mount-time scan already uses, and reject a decoded LCN
at or beyond nr_clusters in the mapping pairs decoder. Conforming
volumes are unaffected.
Fixes: 11ccc9107dc4 ("ntfs: update runlist handling and cluster allocator")
Cc: stable@vger.kernel.org
Signed-off-by: Bryam Vargas <hexlabsecurity@proton.me>
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
fmask and dmask provide the default permissions for files without WSL
metadata. Once chmod stores a mode in $LXMOD, however, that per-file mode
must take precedence so selected files can retain permissions such as
execute across remounts.
Record whether $LXMOD was found while loading an inode and apply the mount
masks only when it is absent. Do not remask the in-memory mode after
setattr persists it. Continue loading $LXMOD even when optional $LXUID or
$LXGID metadata is missing, since chmod may create only $LXMOD.
Fixes: fc053f05ca28 ("ntfs: add reparse and ea operations")
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
The windows_names mount option is documented to reject names containing
characters forbidden by Windows. However, ntfs_check_bad_windows_name()
unconditionally rejects those characters before checking the mount option.
Move the character validation after the option check so a default NTFS
mount accepts POSIX names such as names containing ':'. Mounts using
windows_names retain the existing Windows-compatible validation, including
reserved device names and trailing spaces or dots.
Fixes: af0db57d4293 ("ntfs: update inode operations")
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
ntfs_ir_to_ib copies all entries from index_root into a freshly allocated
index_block_size-byte buffer without verifying that the entries fit in the
available space. The entries in index_root may be larger than the usable
entry space in the index block.
This can cause OOB writes past the end of the allocation.
The validator ntfs_index_root_inconsistent() checks that entries are
self-consistent within the IR value, but never cross-checks them against
index_block_size. There is no bounds check in ntfs_ir_to_ib() before the
memcpy.
Fixing this at the sink in ntfs_ir_to_ib() since
ntfs_index_root_inconsistent() validates the logical consistency of
index_root as a structure and a root with large entries is a structurally
valid root. The bug is a size conflict of ntfs_ir_to_ib().
Also, the validator is called once per inode load in
ntfs_read_locked_inode() while ntfs_ir_to_ib() is only called during a
reparent, a check there adds no overhead to the common path.
Moreover, even a future call path that bypasses the validator would still
be protected.
With NULL as first parameter of ntfs_error(), the volume error flag is
never set by this call, so the device name will be absent from the error
message. In any case, that the caller, ntfs_ir_reparent(), prints an error
message that includes the device name on NULL returns.
I think this is the best solution available without adding
'struct super_block *sb' as a parameter to ntfs_ir_to_ib().
This heap out-of-bounds write is triggered by a crafted filesystem image,
which is not in the kernel threat model, anyway, fixing memory errors would
be nice to keep things secure.
Fixes: 0a8ac0c1fa0b ("ntfs: update directory operations")
Signed-off-by: Alexandro Calo <alexandro.calo@nozominetworks.com>
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
ntfs_new_attr_flags() resizes the non-resident attribute record before
moving its name and mapping pairs to their shorter-header offsets when
compression or sparse state is cleared.
Shrinking the record first moves the following attribute over the tail of
the old record. The subsequent memmove() therefore copies bytes from that
following attribute instead of the old mapping pairs. Re-enabling
compression on an empty file persists those bytes as a malformed mapping
pairs array, which ntfsck reports as a missing or invalid run length.
Move the payload before shrinking the record, while retaining the existing
resize-before-move ordering when growing it.
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
When setting sparse/compressed flags on a resident attribute, the
function skipped the resident-to-non-resident conversion and
terminated.
Signed-off-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
ntfs_compress_write() reads every page in a compression unit before copying
new data into it. The read is unnecessary when an aligned write replaces
every byte covered by the page-cache folios.
Detect full page-aligned compression unit overwrites and grab locked cache
folios without reading them. Keep the read-modify-write path for partial
units and units that cover only part of a large page.
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
ntfs_write_cb() allocates a single-vector bio and synchronously submits it
whenever another output page cannot be added. A 64 KiB uncompressed unit
therefore requires up to sixteen separate bio submissions.
Allocate enough vectors for the complete unit, add all output pages, and
perform one synchronous submission.
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
ntfs_write_cb() allocates output pages and creates input and output
vmaps for every compression unit. Sequential writes repeatedly pay
those allocation and page-table costs even though each unit has the
same maximum output size.
Allocate and map the output workspace once per write request. Access
input sub-blocks with kmap_local_page(), and reuse the output pages and
mapping for every compression unit in the request.
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
ntfs_compress_block() allocates and initializes a roughly 40 KiB match
finder context for every 4 KiB sub-block. A 64 KiB compression unit thus
performs sixteen large allocations even though the calls are serialized.
Allocate one context for the complete write request and reset its hash
chains for each sub-block as before.
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
The write iterator now expands attributes before calling
ntfs_compress_write(), so compressed writes must not expand the
attribute themselves. However, the compressed path still needs to reject
zero-byte iterator copies, advance initialized_size after successful I/O,
and invalidate modified folios after a failed compression-unit write.
Reject no-progress copies, persist the new initialized size on success,
and clear folio uptodate state when the synchronous write fails.
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
ntfs_write_cb() punches the old compression unit and publishes the new
mapping before submitting the replacement data. An allocation or I/O
failure after the punch loses the previous contents and can leave the
mapping pointing at unwritten clusters.
Allocate and write the replacement clusters first. Replace the runlist only
after the synchronous write succeeds, and free new clusters on failure.
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
|
When a rewritten compression block consists entirely of zeroes,
ntfs_write_cb() returns without replacing its existing runlist mapping.
The old on-disk contents therefore remain visible after cache eviction.
Punch the compression unit so that reads resolve it as a sparse block and
release any clusters that held the previous contents.
Reviewed-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
|
