* [PATCH 5.10/5.15] scsi: add a length check for VARIABLE_LENGTH_CMD commands
@ 2024-03-06 13:50 Mikhail Ukhin
0 siblings, 0 replies; 3+ messages in thread
From: Mikhail Ukhin @ 2024-03-06 13:50 UTC (permalink / raw
To: Dave Kleikamp, Christian Brauner, Jens Axboe, Jan Kara,
Greg Kroah-Hartman
Cc: Mikhail Ukhin, jfs-discussion, stable, lvc-project, linux-kernel,
Pavel Koshutin, Artem Sadovnikov, Mikhail Ivanov
Fuzzing of 5.10 stable branch reports a slab-out-of-bounds error in
ata_scsi_pass_thru.
The error is fixed in 5.18 by commit
ce70fd9a551af7424a7dace2a1ba05a7de8eae27.
Backporting this commit would require significant changes to the code so
it is bettter to use a simple fix for that particular error.
The problem is that the length of the received SCSI command is not
validated if scsi_op == VARIABLE_LENGTH_CMD. It can lead to out-of-bounds
reading if the user sends a request with SCSI command of length less than
32.
Found by Linux Verification Center (linuxtesting.org) with Syzkaller.
Signed-off-by: Artem Sadovnikov <ancowi69@gmail.com>
Signed-off-by: Mikhail Ivanov <iwanov-23@bk.ru>
---
drivers/ata/libata-scsi.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index dfa090ccd21c..77589e911d3d 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -4065,6 +4065,9 @@ int __ata_scsi_queuecmd(struct scsi_cmnd *scmd, struct ata_device *dev)
if (unlikely(!scmd->cmd_len))
goto bad_cdb_len;
+
+ if (scsi_op == VARIABLE_LENGTH_CMD && scmd->cmd_len < 32)
+ goto bad_cdb_len;
if (dev->class == ATA_DEV_ATA || dev->class == ATA_DEV_ZAC) {
if (unlikely(scmd->cmd_len > dev->cdb_len))
--
2.25.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* [PATCH 5.10/5.15] scsi: add a length check for VARIABLE_LENGTH_CMD commands
@ 2024-03-21 11:02 Mikhail Ukhin
2024-03-21 23:53 ` Damien Le Moal
0 siblings, 1 reply; 3+ messages in thread
From: Mikhail Ukhin @ 2024-03-21 11:02 UTC (permalink / raw
To: Jens Axboe
Cc: Mikhail Ukhin, linux-ide, linux-kernel, Damien Le Moal,
Niklas Cassel, Pavel Koshutin, Artem Sadovnikov, Mikhail Ivanov
Fuzzing of 5.10 stable branch reports a slab-out-of-bounds error in
ata_scsi_pass_thru.
The error is fixed in 5.18 by commit
ce70fd9a551af7424a7dace2a1ba05a7de8eae27.
Backporting this commit would require significant changes to the code so
it is bettter to use a simple fix for that particular error.
The problem is that the length of the received SCSI command is not
validated if scsi_op == VARIABLE_LENGTH_CMD. It can lead to out-of-bounds
reading if the user sends a request with SCSI command of length less than
32.
Found by Linux Verification Center (linuxtesting.org) with Syzkaller.
Signed-off-by: Artem Sadovnikov <ancowi69@gmail.com>
Signed-off-by: Mikhail Ivanov <iwanov-23@bk.ru>
---
drivers/ata/libata-scsi.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index dfa090ccd21c..77589e911d3d 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -4065,6 +4065,9 @@ int __ata_scsi_queuecmd(struct scsi_cmnd *scmd, struct ata_device *dev)
if (unlikely(!scmd->cmd_len))
goto bad_cdb_len;
+
+ if (scsi_op == VARIABLE_LENGTH_CMD && scmd->cmd_len < 32)
+ goto bad_cdb_len;
if (dev->class == ATA_DEV_ATA || dev->class == ATA_DEV_ZAC) {
if (unlikely(scmd->cmd_len > dev->cdb_len))
--
2.25.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH 5.10/5.15] scsi: add a length check for VARIABLE_LENGTH_CMD commands
2024-03-21 11:02 [PATCH 5.10/5.15] scsi: add a length check for VARIABLE_LENGTH_CMD commands Mikhail Ukhin
@ 2024-03-21 23:53 ` Damien Le Moal
0 siblings, 0 replies; 3+ messages in thread
From: Damien Le Moal @ 2024-03-21 23:53 UTC (permalink / raw
To: Mikhail Ukhin, Jens Axboe
Cc: linux-ide, linux-kernel, Niklas Cassel, Pavel Koshutin,
Artem Sadovnikov, Mikhail Ivanov
On 3/21/24 20:02, Mikhail Ukhin wrote:
> Fuzzing of 5.10 stable branch reports a slab-out-of-bounds error in
> ata_scsi_pass_thru.
>
> The error is fixed in 5.18 by commit
> ce70fd9a551af7424a7dace2a1ba05a7de8eae27.
> Backporting this commit would require significant changes to the code so
> it is bettter to use a simple fix for that particular error.
>
> The problem is that the length of the received SCSI command is not
> validated if scsi_op == VARIABLE_LENGTH_CMD. It can lead to out-of-bounds
> reading if the user sends a request with SCSI command of length less than
> 32.
>
> Found by Linux Verification Center (linuxtesting.org) with Syzkaller.
>
> Signed-off-by: Artem Sadovnikov <ancowi69@gmail.com>
> Signed-off-by: Mikhail Ivanov <iwanov-23@bk.ru>
Looks OK to me, but the patch title prefix is incorrect since this is not a scsi
patch. Please make that:
ata: libata-scsi: check cdb length for VARIABLE_LENGTH_CMD commands
> ---
> drivers/ata/libata-scsi.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
> index dfa090ccd21c..77589e911d3d 100644
> --- a/drivers/ata/libata-scsi.c
> +++ b/drivers/ata/libata-scsi.c
> @@ -4065,6 +4065,9 @@ int __ata_scsi_queuecmd(struct scsi_cmnd *scmd, struct ata_device *dev)
>
> if (unlikely(!scmd->cmd_len))
> goto bad_cdb_len;
> +
> + if (scsi_op == VARIABLE_LENGTH_CMD && scmd->cmd_len < 32)
> + goto bad_cdb_len;
>
> if (dev->class == ATA_DEV_ATA || dev->class == ATA_DEV_ZAC) {
> if (unlikely(scmd->cmd_len > dev->cdb_len))
--
Damien Le Moal
Western Digital Research
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2024-03-21 23:53 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-03-21 11:02 [PATCH 5.10/5.15] scsi: add a length check for VARIABLE_LENGTH_CMD commands Mikhail Ukhin
2024-03-21 23:53 ` Damien Le Moal
-- strict thread matches above, loose matches on Subject: below --
2024-03-06 13:50 Mikhail Ukhin
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).