* [PATCH v2] scsi: core: fix uninit-value access of variable sshdr
@ 2019-10-12 1:26 zhengbin
2019-10-12 1:58 ` James Bottomley
0 siblings, 1 reply; 5+ messages in thread
From: zhengbin @ 2019-10-12 1:26 UTC (permalink / raw)
To: bvanassche, jejb, martin.petersen, linux-scsi; +Cc: zhengbin13
kmsan report a warning in 5.1-rc4:
BUG: KMSAN: uninit-value in sr_get_events drivers/scsi/sr.c:207 [inline]
BUG: KMSAN: uninit-value in sr_check_events+0x2cf/0x1090 drivers/scsi/sr.c:243
CPU: 1 PID: 13858 Comm: syz-executor.0 Tainted: G B 5.1.0-rc4+ #8
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014
Call Trace:
__dump_stack lib/dump_stack.c:77 [inline]
dump_stack+0x173/0x1d0 lib/dump_stack.c:113
kmsan_report+0x131/0x2a0 mm/kmsan/kmsan.c:619
__msan_warning+0x7a/0xf0 mm/kmsan/kmsan_instr.c:310
sr_get_events drivers/scsi/sr.c:207 [inline]
sr_check_events+0x2cf/0x1090 drivers/scsi/sr.c:243
The reason is as follows:
sr_get_events
struct scsi_sense_hdr sshdr; -->uninit
scsi_execute_req -->If fail, will not set sshdr
scsi_sense_valid(&sshdr) -->access sshdr
We can init sshdr in sr_get_events, but there have many callers of
scsi_execute, scsi_execute_req, we have to troubleshoot all callers,
the simpler way is init sshdr in __scsi_execute.
BTW: we can't just init sshdr->response_code, sr_do_ioctl use
sshdr->sense_key
Signed-off-by: zhengbin <zhengbin13@huawei.com>
---
v1->v2: modify the comments suggested by Bart
drivers/scsi/scsi_lib.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 5447738..d5e29c5 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -255,6 +255,13 @@ int __scsi_execute(struct scsi_device *sdev, const unsigned char *cmd,
struct scsi_request *rq;
int ret = DRIVER_ERROR << 24;
+ /*
+ * Zero-initialize sshdr for those callers that check the *sshdr
+ * contents even if no sense data is available.
+ */
+ if (sshdr)
+ memset(sshdr, 0, sizeof(struct scsi_sense_hdr));
+
req = blk_get_request(sdev->request_queue,
data_direction == DMA_TO_DEVICE ?
REQ_OP_SCSI_OUT : REQ_OP_SCSI_IN, BLK_MQ_REQ_PREEMPT);
--
2.7.4
^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH v2] scsi: core: fix uninit-value access of variable sshdr
2019-10-12 1:26 [PATCH v2] scsi: core: fix uninit-value access of variable sshdr zhengbin
@ 2019-10-12 1:58 ` James Bottomley
2019-10-12 2:03 ` zhengbin (A)
0 siblings, 1 reply; 5+ messages in thread
From: James Bottomley @ 2019-10-12 1:58 UTC (permalink / raw)
To: zhengbin, bvanassche, martin.petersen, linux-scsi
On Sat, 2019-10-12 at 09:26 +0800, zhengbin wrote:
> BTW: we can't just init sshdr->response_code, sr_do_ioctl use
> sshdr->sense_key
That's an actual bug, isn't it?
James
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v2] scsi: core: fix uninit-value access of variable sshdr
2019-10-12 1:58 ` James Bottomley
@ 2019-10-12 2:03 ` zhengbin (A)
2019-10-12 2:06 ` James Bottomley
0 siblings, 1 reply; 5+ messages in thread
From: zhengbin (A) @ 2019-10-12 2:03 UTC (permalink / raw)
To: James Bottomley, bvanassche, martin.petersen, linux-scsi
On 2019/10/12 9:58, James Bottomley wrote:
> On Sat, 2019-10-12 at 09:26 +0800, zhengbin wrote:
>> BTW: we can't just init sshdr->response_code, sr_do_ioctl use
>> sshdr->sense_key
> That's an actual bug, isn't it?
If we init sshdr in __scsi_execute, this will be ok
>
> James
>
>
> .
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v2] scsi: core: fix uninit-value access of variable sshdr
2019-10-12 2:03 ` zhengbin (A)
@ 2019-10-12 2:06 ` James Bottomley
2019-10-12 2:10 ` zhengbin (A)
0 siblings, 1 reply; 5+ messages in thread
From: James Bottomley @ 2019-10-12 2:06 UTC (permalink / raw)
To: zhengbin (A), bvanassche, martin.petersen, linux-scsi
On Sat, 2019-10-12 at 10:03 +0800, zhengbin (A) wrote:
> On 2019/10/12 9:58, James Bottomley wrote:
> > On Sat, 2019-10-12 at 09:26 +0800, zhengbin wrote:
> > > BTW: we can't just init sshdr->response_code, sr_do_ioctl use
> > > sshdr->sense_key
> >
> > That's an actual bug, isn't it?
>
> If we init sshdr in __scsi_execute, this will be ok
No I mean it's a bug because sr_do_ioctl shouldn't be acting on sense
that isn't valid. So all uses of sshdr should be gated on a validity
check.
James
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v2] scsi: core: fix uninit-value access of variable sshdr
2019-10-12 2:06 ` James Bottomley
@ 2019-10-12 2:10 ` zhengbin (A)
0 siblings, 0 replies; 5+ messages in thread
From: zhengbin (A) @ 2019-10-12 2:10 UTC (permalink / raw)
To: James Bottomley, bvanassche, martin.petersen, linux-scsi
On 2019/10/12 10:06, James Bottomley wrote:
> On Sat, 2019-10-12 at 10:03 +0800, zhengbin (A) wrote:
>> On 2019/10/12 9:58, James Bottomley wrote:
>>> On Sat, 2019-10-12 at 09:26 +0800, zhengbin wrote:
>>>> BTW: we can't just init sshdr->response_code, sr_do_ioctl use
>>>> sshdr->sense_key
>>> That's an actual bug, isn't it?
>> If we init sshdr in __scsi_execute, this will be ok
> No I mean it's a bug because sr_do_ioctl shouldn't be acting on sense
> that isn't valid. So all uses of sshdr should be gated on a validity
> check.
Yes you are right, the right way is use scsi_sense_valid(&sshdr), I have troubleshoot callers,
this is the only wrong use(Maybe I miss some...).
>
> James
>
>
> .
>
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2019-10-12 2:10 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-10-12 1:26 [PATCH v2] scsi: core: fix uninit-value access of variable sshdr zhengbin
2019-10-12 1:58 ` James Bottomley
2019-10-12 2:03 ` zhengbin (A)
2019-10-12 2:06 ` James Bottomley
2019-10-12 2:10 ` zhengbin (A)
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 NNTP newsgroup(s).