* [PATCH v5 5.10/5.15] ata: libata-scsi: check cdb length for VARIABLE_LENGTH_CMD commands
@ 2024-08-16 19:27 Artem Sadovnikov
0 siblings, 0 replies; 3+ messages in thread
From: Artem Sadovnikov @ 2024-08-16 19:27 UTC (permalink / raw)
To: stable, Greg Kroah-Hartman
Cc: Artem Sadovnikov, Damien Le Moal, Niklas Cassel, linux-ide,
linux-kernel, lvc-project
No upstream commit exists for this commit.
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 ce70fd9a551a ("scsi: core: Remove the
cmd field from struct scsi_request") upstream.
Backporting this commit would require significant changes to the code so
it is better 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.
Acked-by: Damien Le Moal <dlemoal@kernel.org>
Co-developed-by: Mikhail Ivanov <iwanov-23@bk.ru>
Signed-off-by: Mikhail Ivanov <iwanov-23@bk.ru>
Co-developed-by: Mikhail Ukhin <mish.uxin2012@yandex.ru>
Signed-off-by: Mikhail Ukhin <mish.uxin2012@yandex.ru>
Signed-off-by: Artem Sadovnikov <ancowi69@gmail.com>
---
Link: https://lore.kernel.org/lkml/20240711151546.341491-1-ancowi69@gmail.com/T/#u
unfortunately, stable@vger.kernel.org wasn't initially mentioned.
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 36f32fa052df..4397986db053 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -3949,6 +3949,9 @@ static unsigned int ata_scsi_var_len_cdb_xlat(struct ata_queued_cmd *qc)
const u8 *cdb = scmd->cmnd;
const u16 sa = get_unaligned_be16(&cdb[8]);
+ if (scmd->cmd_len != 32)
+ return 1;
+
/*
* if service action represents a ata pass-thru(32) command,
* then pass it to ata_scsi_pass_thru handler.
--
2.34.1
^ permalink raw reply related [flat|nested] 3+ messages in thread* [PATCH v5 5.10/5.15] ata: libata-scsi: check cdb length for VARIABLE_LENGTH_CMD commands
@ 2024-07-11 15:15 Artem Sadovnikov
2024-07-11 22:22 ` Damien Le Moal
0 siblings, 1 reply; 3+ messages in thread
From: Artem Sadovnikov @ 2024-07-11 15:15 UTC (permalink / raw)
To: Damien Le Moal
Cc: Artem Sadovnikov, Niklas Cassel, linux-ide, linux-kernel,
lvc-project
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 ce70fd9a551a ("scsi: core: Remove the
cmd field from struct scsi_request") upstream.
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>
Signed-off-by: Mikhail Ukhin <mish.uxin2012@yandex.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 36f32fa052df..4397986db053 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -3949,6 +3949,9 @@ static unsigned int ata_scsi_var_len_cdb_xlat(struct ata_queued_cmd *qc)
const u8 *cdb = scmd->cmnd;
const u16 sa = get_unaligned_be16(&cdb[8]);
+ if (scmd->cmd_len != 32)
+ return 1;
+
/*
* if service action represents a ata pass-thru(32) command,
* then pass it to ata_scsi_pass_thru handler.
--
2.34.1
^ permalink raw reply related [flat|nested] 3+ messages in thread* Re: [PATCH v5 5.10/5.15] ata: libata-scsi: check cdb length for VARIABLE_LENGTH_CMD commands
2024-07-11 15:15 Artem Sadovnikov
@ 2024-07-11 22:22 ` Damien Le Moal
0 siblings, 0 replies; 3+ messages in thread
From: Damien Le Moal @ 2024-07-11 22:22 UTC (permalink / raw)
To: Artem Sadovnikov; +Cc: Niklas Cassel, linux-ide, linux-kernel, lvc-project
On 7/12/24 00:15, Artem Sadovnikov 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 ce70fd9a551a ("scsi: core: Remove the
> cmd field from struct scsi_request") upstream.
>
> 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>
> Signed-off-by: Mikhail Ukhin <mish.uxin2012@yandex.ru>
Looks good, so please feel free to add:
Acked-by: Damien Le Moal <dlemoal@kernel.org>
However, you did not address this patch to linux-stable and Greg, so it will go
nowhere as is since I cannot apply that to stable. So please check:
Documentation/process/stable-kernel-rules.rst
and see "option 2".
> ---
> 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 36f32fa052df..4397986db053 100644
> --- a/drivers/ata/libata-scsi.c
> +++ b/drivers/ata/libata-scsi.c
> @@ -3949,6 +3949,9 @@ static unsigned int ata_scsi_var_len_cdb_xlat(struct ata_queued_cmd *qc)
> const u8 *cdb = scmd->cmnd;
> const u16 sa = get_unaligned_be16(&cdb[8]);
>
> + if (scmd->cmd_len != 32)
> + return 1;
> +
> /*
> * if service action represents a ata pass-thru(32) command,
> * then pass it to ata_scsi_pass_thru handler.
--
Damien Le Moal
Western Digital Research
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2024-08-16 19:28 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-08-16 19:27 [PATCH v5 5.10/5.15] ata: libata-scsi: check cdb length for VARIABLE_LENGTH_CMD commands Artem Sadovnikov
-- strict thread matches above, loose matches on Subject: below --
2024-07-11 15:15 Artem Sadovnikov
2024-07-11 22:22 ` Damien Le Moal
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox