linux-stable.git/drivers/md, branch v6.6.22

dm-integrity, dm-verity: reduce stack usage for recheck

2024-03-01T12:35:01+00:00

commit 66ad2fbcdbeab0edfd40c5d94f32f053b98c2320 upstream.

The newly added integrity_recheck() function has another larger stack
allocation, just like its caller integrity_metadata(). When it gets
inlined, the combination of the two exceeds the warning limit for 32-bit
architectures and possibly risks an overflow when this is called from
a deep call chain through a file system:

drivers/md/dm-integrity.c:1767:13: error: stack frame size (1048) exceeds limit (1024) in 'integrity_metadata' [-Werror,-Wframe-larger-than]
 1767 | static void integrity_metadata(struct work_struct *w)

Since the caller at this point is done using its checksum buffer,
just reuse the same buffer in the new function to avoid the double
allocation.

[Mikulas: add "noinline" to integrity_recheck and verity_recheck.
These functions are only called on error, so they shouldn't bloat the
stack frame or code size of the caller.]

Fixes: c88f5e553fe3 ("dm-integrity: recheck the integrity tag after a failure")
Fixes: 9177f3c0dea6 ("dm-verity: recheck the hash after a failure")
Cc: stable@vger.kernel.org
Signed-off-by: Arnd Bergmann 
Signed-off-by: Mikulas Patocka 
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

md: Fix missing release of 'active_io' for flush

2024-03-01T12:35:00+00:00

commit 855678ed8534518e2b428bcbcec695de9ba248e8 upstream.

submit_flushes
 atomic_set(&mddev->flush_pending, 1);
 rdev_for_each_rcu(rdev, mddev)
  atomic_inc(&mddev->flush_pending);
  bi->bi_end_io = md_end_flush
  submit_bio(bi);
                        /* flush io is done first */
                        md_end_flush
                         if (atomic_dec_and_test(&mddev->flush_pending))
                          percpu_ref_put(&mddev->active_io)
                          -> active_io is not released

 if (atomic_dec_and_test(&mddev->flush_pending))
  -> missing release of active_io

For consequence, mddev_suspend() will wait for 'active_io' to be zero
forever.

Fix this problem by releasing 'active_io' in submit_flushes() if
'flush_pending' is decreased to zero.

Fixes: fa2bbff7b0b4 ("md: synchronize flush io with array reconfiguration")
Cc: stable@vger.kernel.org # v6.1+
Reported-by: Blazej Kucman 
Closes: https://lore.kernel.org/lkml/20240130172524.0000417b@linux.intel.com/
Signed-off-by: Yu Kuai 
Signed-off-by: Song Liu 
Link: https://lore.kernel.org/r/20240201092559.910982-7-yukuai1@huaweicloud.com
Signed-off-by: Greg Kroah-Hartman

dm-verity: recheck the hash after a failure

2024-03-01T12:34:59+00:00

commit 9177f3c0dea6143d05cac1bbd28668fd0e216d11 upstream.

If a userspace process reads (with O_DIRECT) multiple blocks into the same
buffer, dm-verity reports an error [1].

This commit fixes dm-verity, so that if hash verification fails, the data
is read again into a kernel buffer (where userspace can't modify it) and
the hash is rechecked. If the recheck succeeds, the content of the kernel
buffer is copied into the user buffer; if the recheck fails, an error is
reported.

[1] https://people.redhat.com/~mpatocka/testcases/blk-auth-modify/read2.c

Signed-off-by: Mikulas Patocka 
Cc: stable@vger.kernel.org
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm-crypt: don't modify the data when using authenticated encryption

2024-03-01T12:34:59+00:00

commit 50c70240097ce41fe6bce6478b80478281e4d0f7 upstream.

It was said that authenticated encryption could produce invalid tag when
the data that is being encrypted is modified [1]. So, fix this problem by
copying the data into the clone bio first and then encrypt them inside the
clone bio.

This may reduce performance, but it is needed to prevent the user from
corrupting the device by writing data with O_DIRECT and modifying them at
the same time.

[1] https://lore.kernel.org/all/20240207004723.GA35324@sol.localdomain/T/

Signed-off-by: Mikulas Patocka 
Cc: stable@vger.kernel.org
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm-integrity: recheck the integrity tag after a failure

2024-03-01T12:34:59+00:00

commit c88f5e553fe38b2ffc4c33d08654e5281b297677 upstream.

If a userspace process reads (with O_DIRECT) multiple blocks into the same
buffer, dm-integrity reports an error [1]. The error is reported in a log
and it may cause RAID leg being kicked out of the array.

This commit fixes dm-integrity, so that if integrity verification fails,
the data is read again into a kernel buffer (where userspace can't modify
it) and the integrity tag is rechecked. If the recheck succeeds, the
content of the kernel buffer is copied into the user buffer; if the
recheck fails, an integrity error is reported.

[1] https://people.redhat.com/~mpatocka/testcases/blk-auth-modify/read2.c

Signed-off-by: Mikulas Patocka 
Cc: stable@vger.kernel.org
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm-crypt: recheck the integrity tag after a failure

2024-03-01T12:34:59+00:00

commit 42e15d12070b4ff9af2b980f1b65774c2dab0507 upstream.

If a userspace process reads (with O_DIRECT) multiple blocks into the same
buffer, dm-crypt reports an authentication error [1]. The error is
reported in a log and it may cause RAID leg being kicked out of the
array.

This commit fixes dm-crypt, so that if integrity verification fails, the
data is read again into a kernel buffer (where userspace can't modify it)
and the integrity tag is rechecked. If the recheck succeeds, the content
of the kernel buffer is copied into the user buffer; if the recheck fails,
an integrity error is reported.

[1] https://people.redhat.com/~mpatocka/testcases/blk-auth-modify/read2.c

Signed-off-by: Mikulas Patocka 
Cc: stable@vger.kernel.org
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm: limit the number of targets and parameter size area

2024-02-23T08:25:27+00:00

commit bd504bcfec41a503b32054da5472904b404341a4 upstream.

The kvmalloc function fails with a warning if the size is larger than
INT_MAX. The warning was triggered by a syscall testing robot.

In order to avoid the warning, this commit limits the number of targets to
1048576 and the size of the parameter area to 1073741824.

Signed-off-by: Mikulas Patocka 
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

md: bypass block throttle for superblock update

2024-02-23T08:25:18+00:00

[ Upstream commit d6e035aad6c09991da1c667fb83419329a3baed8 ]

commit 5e2cf333b7bd ("md/raid5: Wait for MD_SB_CHANGE_PENDING in raid5d")
introduced a hung bug and will be reverted in next patch, since the issue
that commit is fixing is due to md superblock write is throttled by wbt,
to fix it, we can have superblock write bypass block layer throttle.

Fixes: 5e2cf333b7bd ("md/raid5: Wait for MD_SB_CHANGE_PENDING in raid5d")
Cc: stable@vger.kernel.org # v5.19+
Suggested-by: Yu Kuai 
Signed-off-by: Junxiao Bi 
Reviewed-by: Logan Gunthorpe 
Reviewed-by: Yu Kuai 
Signed-off-by: Song Liu 
Link: https://lore.kernel.org/r/20231108182216.73611-1-junxiao.bi@oracle.com
Signed-off-by: Sasha Levin

dm-crypt, dm-verity: disable tasklets

2024-02-23T08:24:55+00:00

commit 0a9bab391e336489169b95cb0d4553d921302189 upstream.

Tasklets have an inherent problem with memory corruption. The function
tasklet_action_common calls tasklet_trylock, then it calls the tasklet
callback and then it calls tasklet_unlock. If the tasklet callback frees
the structure that contains the tasklet or if it calls some code that may
free it, tasklet_unlock will write into free memory.

The commits 8e14f610159d and d9a02e016aaf try to fix it for dm-crypt, but
it is not a sufficient fix and the data corruption can still happen [1].
There is no fix for dm-verity and dm-verity will write into free memory
with every tasklet-processed bio.

There will be atomic workqueues implemented in the kernel 6.9 [2]. They
will have better interface and they will not suffer from the memory
corruption problem.

But we need something that stops the memory corruption now and that can be
backported to the stable kernels. So, I'm proposing this commit that
disables tasklets in both dm-crypt and dm-verity. This commit doesn't
remove the tasklet support, because the tasklet code will be reused when
atomic workqueues will be implemented.

[1] https://lore.kernel.org/all/d390d7ee-f142-44d3-822a-87949e14608b@suse.de/T/
[2] https://lore.kernel.org/lkml/20240130091300.2968534-1-tj@kernel.org/

Signed-off-by: Mikulas Patocka 
Cc: stable@vger.kernel.org
Fixes: 39d42fa96ba1b ("dm crypt: add flags to optionally bypass kcryptd workqueues")
Fixes: 5721d4e5a9cdb ("dm verity: Add optional "try_verify_in_tasklet" feature")
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

md: Whenassemble the array, consult the superblock of the freshest device

2024-02-05T20:14:24+00:00

[ Upstream commit dc1cc22ed58f11d58d8553c5ec5f11cbfc3e3039 ]

Upon assembling the array, both kernel and mdadm allow the devices to have event
counter difference of 1, and still consider them as up-to-date.
However, a device whose event count is behind by 1, may in fact not be up-to-date,
and array resync with such a device may cause data corruption.
To avoid this, consult the superblock of the freshest device about the status
of a device, whose event counter is behind by 1.

Signed-off-by: Alex Lyakas 
Signed-off-by: Song Liu 
Link: https://lore.kernel.org/r/1702470271-16073-1-git-send-email-alex.lyakas@zadara.com
Signed-off-by: Sasha Levin