2021-10-15 13:51:11

by Miles Chen

[permalink] [raw]
Subject: [PATCH] scsi: sd: fix crashes in sd_resume_runtime

After merging commit ed4246d37f3b ("scsi: sd: REQUEST SENSE for
BLIST_IGN_MEDIA_CHANGE devices in runtime_resume()"), I hit the
following crash on my device.

static int sd_resume_runtime(struct device *dev)
{
struct scsi_disk *sdkp = dev_get_drvdata(dev);
struct scsi_device *sdp = sdkp->device; // sdkp == NULL and crash

if (sdp->ignore_media_change) {
...
}

I checked sd_resume() and found that sdkp is possbile to be NULL, and
there is a null pointer test in sd_resume() for this case.
To fix this crash, follow sd_resume() to test if sdkp is NULL
before dereferencing it.

Crash:
[ 4.695171][ T151] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000008
[ 4.696591][ T151] Mem abort info:
[ 4.697919][ T151] ESR = 0x96000005
[ 4.699692][ T151] EC = 0x25: DABT (current EL), IL = 32 bits
[ 4.701990][ T151] SET = 0, FnV = 0
[ 4.702513][ T151] EA = 0, S1PTW = 0
[ 4.704431][ T151] FSC = 0x05: level 1 translation fault
[ 4.705254][ T151] Data abort info:
[ 4.705806][ T151] ISV = 0, ISS = 0x00000005
[ 4.706484][ T151] CM = 0, WnR = 0
[ 4.707048][ T151] [0000000000000008] user address but active_mm is swapper
[ 4.710577][ T151] Internal error: Oops: 96000005 [#1] PREEMPT SMP
[ 4.832361][ T151] Kernel Offset: 0x12acc80000 from 0xffffffc010000000
[ 4.833254][ T151] PHYS_OFFSET: 0x40000000
[ 4.833814][ T151] pstate: 80400005 (Nzcv daif +PAN -UAO)
[ 4.834546][ T151] pc : sd_resume_runtime+0x20/0x14c
[ 4.835227][ T151] lr : scsi_runtime_resume+0x84/0xe4
[ 4.835916][ T151] sp : ffffffc0110db8d0
[ 4.836450][ T151] x29: ffffffc0110db8d0 x28: 0000000000000001
[ 4.837258][ T151] x27: ffffff80c0bd1ac0 x26: ffffff80c0bd1ad0
[ 4.838063][ T151] x25: ffffff80cea7e448 x24: ffffffd2bf961000
[ 4.838867][ T151] x23: ffffffd2be69f838 x22: ffffffd2bd9dfb4c
[ 4.839670][ T151] x21: 0000000000000000 x20: ffffff80cea7e000
[ 4.840474][ T151] x19: ffffff80cea7e260 x18: ffffffc0110dd078
[ 4.841277][ T151] x17: 00000000658783d9 x16: 0000000051469dac
[ 4.842081][ T151] x15: 00000000b87f6327 x14: 0000000068fd680d
[ 4.842885][ T151] x13: ffffff80c0bd2470 x12: ffffffd2bfa7f5f0
[ 4.843688][ T151] x11: 0000000000000078 x10: 0000000000000001
[ 4.844492][ T151] x9 : 00000000000000b1 x8 : ffffffd2be69f88c
[ 4.845295][ T151] x7 : ffffffd2bd9e0e5c x6 : 0000000000000000
[ 4.846099][ T151] x5 : 0000000000000080 x4 : 0000000000000001
[ 4.846902][ T151] x3 : 68fd680dfe4ebe5e x2 : 0000000000000003
[ 4.847706][ T151] x1 : ffffffd2bf7f9380 x0 : ffffff80cea7e260
[ 4.856708][ T151] die+0x16c/0x59c
[ 4.857191][ T151] __do_kernel_fault+0x1e8/0x210
[ 4.857833][ T151] do_page_fault+0xa4/0x654
[ 4.858418][ T151] do_translation_fault+0x6c/0x1b0
[ 4.859083][ T151] do_mem_abort+0x68/0x10c
[ 4.859655][ T151] el1_abort+0x40/0x64
[ 4.860182][ T151] el1h_64_sync_handler+0x54/0x88
[ 4.860834][ T151] el1h_64_sync+0x7c/0x80
[ 4.861395][ T151] sd_resume_runtime+0x20/0x14c
[ 4.862025][ T151] scsi_runtime_resume+0x84/0xe4
[ 4.862667][ T151] __rpm_callback+0x1f4/0x8cc
[ 4.863275][ T151] rpm_resume+0x7e8/0xaa4
[ 4.863836][ T151] __pm_runtime_resume+0xa0/0x110
[ 4.864489][ T151] sd_probe+0x30/0x428
[ 4.865016][ T151] really_probe+0x14c/0x500
[ 4.865602][ T151] __driver_probe_device+0xb4/0x18c
[ 4.866278][ T151] driver_probe_device+0x60/0x2c4
[ 4.866931][ T151] __device_attach_driver+0x228/0x2bc
[ 4.867630][ T151] __device_attach_async_helper+0x154/0x21c
[ 4.868398][ T151] async_run_entry_fn+0x5c/0x1c4
[ 4.869038][ T151] process_one_work+0x3ac/0x590
[ 4.869670][ T151] worker_thread+0x320/0x758
[ 4.870265][ T151] kthread+0x2e8/0x35c
[ 4.870792][ T151] ret_from_fork+0x10/0x20

Cc: Stanley Chu <[email protected]>
Fixes: ed4246d37f3b ("scsi: sd: REQUEST SENSE for BLIST_IGN_MEDIA_CHANGE devices in runtime_resume()")
Signed-off-by: Miles Chen <[email protected]>
---
drivers/scsi/sd.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
index 523bf2fdc253..fce63335084e 100644
--- a/drivers/scsi/sd.c
+++ b/drivers/scsi/sd.c
@@ -3683,7 +3683,12 @@ static int sd_resume(struct device *dev)
static int sd_resume_runtime(struct device *dev)
{
struct scsi_disk *sdkp = dev_get_drvdata(dev);
- struct scsi_device *sdp = sdkp->device;
+ struct scsi_device *sdp;
+
+ if (!sdkp) /* E.g.: runtime resume at the start of sd_probe() */
+ return 0;
+
+ sdp = sdkp->device;

if (sdp->ignore_media_change) {
/* clear the device's sense data */
--
2.18.0


2021-10-15 13:53:11

by Martin Kepplinger

[permalink] [raw]
Subject: Re: [PATCH] scsi: sd: fix crashes in sd_resume_runtime

Am Freitag, dem 15.10.2021 um 15:46 +0800 schrieb Miles Chen:
> After merging commit ed4246d37f3b ("scsi: sd: REQUEST SENSE for
> BLIST_IGN_MEDIA_CHANGE devices in runtime_resume()"), I hit the
> following crash on my device.
>
> static int sd_resume_runtime(struct device *dev)
> {
>         struct scsi_disk *sdkp = dev_get_drvdata(dev);
>         struct scsi_device *sdp = sdkp->device; // sdkp == NULL and
> crash
>
>         if (sdp->ignore_media_change) {
>         ...
> }
>
> I checked sd_resume() and found that sdkp is possbile to be NULL, and
> there is a null pointer test in sd_resume() for this case.
> To fix this crash, follow sd_resume() to test if sdkp is NULL
> before dereferencing it.
>
> Crash:
> [    4.695171][  T151] Unable to handle kernel NULL pointer
> dereference at virtual address 0000000000000008
> [    4.696591][  T151] Mem abort info:
> [    4.697919][  T151]   ESR = 0x96000005
> [    4.699692][  T151]   EC = 0x25: DABT (current EL), IL = 32 bits
> [    4.701990][  T151]   SET = 0, FnV = 0
> [    4.702513][  T151]   EA = 0, S1PTW = 0
> [    4.704431][  T151]   FSC = 0x05: level 1 translation fault
> [    4.705254][  T151] Data abort info:
> [    4.705806][  T151]   ISV = 0, ISS = 0x00000005
> [    4.706484][  T151]   CM = 0, WnR = 0
> [    4.707048][  T151] [0000000000000008] user address but active_mm
> is swapper
> [    4.710577][  T151] Internal error: Oops: 96000005 [#1] PREEMPT
> SMP
> [    4.832361][  T151] Kernel Offset: 0x12acc80000 from
> 0xffffffc010000000
> [    4.833254][  T151] PHYS_OFFSET: 0x40000000
> [    4.833814][  T151] pstate: 80400005 (Nzcv daif +PAN -UAO)
> [    4.834546][  T151] pc : sd_resume_runtime+0x20/0x14c
> [    4.835227][  T151] lr : scsi_runtime_resume+0x84/0xe4
> [    4.835916][  T151] sp : ffffffc0110db8d0
> [    4.836450][  T151] x29: ffffffc0110db8d0 x28: 0000000000000001
> [    4.837258][  T151] x27: ffffff80c0bd1ac0 x26: ffffff80c0bd1ad0
> [    4.838063][  T151] x25: ffffff80cea7e448 x24: ffffffd2bf961000
> [    4.838867][  T151] x23: ffffffd2be69f838 x22: ffffffd2bd9dfb4c
> [    4.839670][  T151] x21: 0000000000000000 x20: ffffff80cea7e000
> [    4.840474][  T151] x19: ffffff80cea7e260 x18: ffffffc0110dd078
> [    4.841277][  T151] x17: 00000000658783d9 x16: 0000000051469dac
> [    4.842081][  T151] x15: 00000000b87f6327 x14: 0000000068fd680d
> [    4.842885][  T151] x13: ffffff80c0bd2470 x12: ffffffd2bfa7f5f0
> [    4.843688][  T151] x11: 0000000000000078 x10: 0000000000000001
> [    4.844492][  T151] x9 : 00000000000000b1 x8 : ffffffd2be69f88c
> [    4.845295][  T151] x7 : ffffffd2bd9e0e5c x6 : 0000000000000000
> [    4.846099][  T151] x5 : 0000000000000080 x4 : 0000000000000001
> [    4.846902][  T151] x3 : 68fd680dfe4ebe5e x2 : 0000000000000003
> [    4.847706][  T151] x1 : ffffffd2bf7f9380 x0 : ffffff80cea7e260
> [    4.856708][  T151]  die+0x16c/0x59c
> [    4.857191][  T151]  __do_kernel_fault+0x1e8/0x210
> [    4.857833][  T151]  do_page_fault+0xa4/0x654
> [    4.858418][  T151]  do_translation_fault+0x6c/0x1b0
> [    4.859083][  T151]  do_mem_abort+0x68/0x10c
> [    4.859655][  T151]  el1_abort+0x40/0x64
> [    4.860182][  T151]  el1h_64_sync_handler+0x54/0x88
> [    4.860834][  T151]  el1h_64_sync+0x7c/0x80
> [    4.861395][  T151]  sd_resume_runtime+0x20/0x14c
> [    4.862025][  T151]  scsi_runtime_resume+0x84/0xe4
> [    4.862667][  T151]  __rpm_callback+0x1f4/0x8cc
> [    4.863275][  T151]  rpm_resume+0x7e8/0xaa4
> [    4.863836][  T151]  __pm_runtime_resume+0xa0/0x110
> [    4.864489][  T151]  sd_probe+0x30/0x428
> [    4.865016][  T151]  really_probe+0x14c/0x500
> [    4.865602][  T151]  __driver_probe_device+0xb4/0x18c
> [    4.866278][  T151]  driver_probe_device+0x60/0x2c4
> [    4.866931][  T151]  __device_attach_driver+0x228/0x2bc
> [    4.867630][  T151]  __device_attach_async_helper+0x154/0x21c
> [    4.868398][  T151]  async_run_entry_fn+0x5c/0x1c4
> [    4.869038][  T151]  process_one_work+0x3ac/0x590
> [    4.869670][  T151]  worker_thread+0x320/0x758
> [    4.870265][  T151]  kthread+0x2e8/0x35c
> [    4.870792][  T151]  ret_from_fork+0x10/0x20
>
> Cc: Stanley Chu <[email protected]>
> Fixes: ed4246d37f3b ("scsi: sd: REQUEST SENSE for
> BLIST_IGN_MEDIA_CHANGE devices in runtime_resume()")
> Signed-off-by: Miles Chen <[email protected]>
> ---
>  drivers/scsi/sd.c | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
> index 523bf2fdc253..fce63335084e 100644
> --- a/drivers/scsi/sd.c
> +++ b/drivers/scsi/sd.c
> @@ -3683,7 +3683,12 @@ static int sd_resume(struct device *dev)
>  static int sd_resume_runtime(struct device *dev)
>  {
>         struct scsi_disk *sdkp = dev_get_drvdata(dev);
> -       struct scsi_device *sdp = sdkp->device;
> +       struct scsi_device *sdp;
> +
> +       if (!sdkp)      /* E.g.: runtime resume at the start of
> sd_probe() */
> +               return 0;
> +
> +       sdp = sdkp->device;
>  
>         if (sdp->ignore_media_change) {
>                 /* clear the device's sense data */

I'm fine with this, thank you.

Reviewed-by: Martin Kepplinger <[email protected]>



2021-10-16 07:00:20

by Stanley Chu

[permalink] [raw]
Subject: Re: [PATCH] scsi: sd: fix crashes in sd_resume_runtime

Hi Miles,

On Fri, 2021-10-15 at 15:46 +0800, Miles Chen wrote:
> After merging commit ed4246d37f3b ("scsi: sd: REQUEST SENSE for
> BLIST_IGN_MEDIA_CHANGE devices in runtime_resume()"), I hit the
> following crash on my device.
>
> static int sd_resume_runtime(struct device *dev)
> {
> struct scsi_disk *sdkp = dev_get_drvdata(dev);
> struct scsi_device *sdp = sdkp->device; // sdkp == NULL and
> crash
>
> if (sdp->ignore_media_change) {
> ...
> }
>
> I checked sd_resume() and found that sdkp is possbile to be NULL, and
> there is a null pointer test in sd_resume() for this case.
> To fix this crash, follow sd_resume() to test if sdkp is NULL
> before dereferencing it.


LGTM.

Reviewed-by: Stanley Chu <[email protected]>

2021-10-17 16:37:34

by Bart Van Assche

[permalink] [raw]
Subject: Re: [PATCH] scsi: sd: fix crashes in sd_resume_runtime

On 10/15/21 00:46, Miles Chen wrote:
> Crash:
> [ 4.695171][ T151] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000008
> [ 4.710577][ T151] Internal error: Oops: 96000005 [#1] PREEMPT SMP
> [ 4.856708][ T151] die+0x16c/0x59c
> [ 4.857191][ T151] __do_kernel_fault+0x1e8/0x210
> [ 4.857833][ T151] do_page_fault+0xa4/0x654
> [ 4.858418][ T151] do_translation_fault+0x6c/0x1b0
> [ 4.859083][ T151] do_mem_abort+0x68/0x10c
> [ 4.859655][ T151] el1_abort+0x40/0x64
> [ 4.860182][ T151] el1h_64_sync_handler+0x54/0x88
> [ 4.860834][ T151] el1h_64_sync+0x7c/0x80
> [ 4.861395][ T151] sd_resume_runtime+0x20/0x14c
> [ 4.862025][ T151] scsi_runtime_resume+0x84/0xe4
> [ 4.862667][ T151] __rpm_callback+0x1f4/0x8cc
> [ 4.863275][ T151] rpm_resume+0x7e8/0xaa4
> [ 4.863836][ T151] __pm_runtime_resume+0xa0/0x110
> [ 4.864489][ T151] sd_probe+0x30/0x428
> [ 4.865016][ T151] really_probe+0x14c/0x500
> [ 4.865602][ T151] __driver_probe_device+0xb4/0x18c
> [ 4.866278][ T151] driver_probe_device+0x60/0x2c4
> [ 4.866931][ T151] __device_attach_driver+0x228/0x2bc
> [ 4.867630][ T151] __device_attach_async_helper+0x154/0x21c
> [ 4.868398][ T151] async_run_entry_fn+0x5c/0x1c4
> [ 4.869038][ T151] process_one_work+0x3ac/0x590
> [ 4.869670][ T151] worker_thread+0x320/0x758
> [ 4.870265][ T151] kthread+0x2e8/0x35c
> [ 4.870792][ T151] ret_from_fork+0x10/0x20
>
> Cc: Stanley Chu <[email protected]>
> Fixes: ed4246d37f3b ("scsi: sd: REQUEST SENSE for BLIST_IGN_MEDIA_CHANGE devices in runtime_resume()")
> Signed-off-by: Miles Chen <[email protected]>
> ---
> drivers/scsi/sd.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
> index 523bf2fdc253..fce63335084e 100644
> --- a/drivers/scsi/sd.c
> +++ b/drivers/scsi/sd.c
> @@ -3683,7 +3683,12 @@ static int sd_resume(struct device *dev)
> static int sd_resume_runtime(struct device *dev)
> {
> struct scsi_disk *sdkp = dev_get_drvdata(dev);
> - struct scsi_device *sdp = sdkp->device;
> + struct scsi_device *sdp;
> +
> + if (!sdkp) /* E.g.: runtime resume at the start of sd_probe() */
> + return 0;
> +
> + sdp = sdkp->device;
>
> if (sdp->ignore_media_change) {
> /* clear the device's sense data */

Fixing this crash by adding a check inside sd_resume_runtime() seems
wrong to me. sd_probe() namely calls dev_set_drvdata(dev, sdkp) before
sd_probe() has finished so even with the above patch applied sd_resume()
can be called before sd_probe() has finished.

With which kernel version has this crash been encountered? The
scsi_autopm_get_device() / scsi_autopm_put_device() pair added by commit
6fe8c1dbefd6 ("scsi: balance out autopm get/put calls in
scsi_sysfs_add_sdev()"; kernel v3.18) should be sufficient to prevent
the reported crash.

Thanks,

Bart.


2021-10-18 01:50:41

by Miles Chen

[permalink] [raw]
Subject: Re: [PATCH] scsi: sd: fix crashes in sd_resume_runtime

From: Miles Chen <[email protected]>

> Fixing this crash by adding a check inside sd_resume_runtime() seems wrong to me. sd_probe() namely calls dev_set_drvdata(dev, sdkp) before
> sd_probe() has finished so even with the above patch applied sd_resume() can be called before sd_probe() has finished.
>
> With which kernel version has this crash been encountered? The
> scsi_autopm_get_device() / scsi_autopm_put_device() pair added by commit
> 6fe8c1dbefd6 ("scsi: balance out autopm get/put calls in scsi_sysfs_add_sdev()"; kernel v3.18) should be sufficient to prevent the reported crash.
>
> Thanks,

Thanks for your comment.

I hit this in v5.15-rc1 merge, I can still reproduce this with v5.15-rc5.
I found two ways to avoid the crash:
1) revert commit ed4246d37f3b ("scsi: sd: REQUEST SENSE for
BLIST_IGN_MEDIA_CHANGE devices in runtime_resume()") works for me.
2) adding the NULL point check in this patch.

From the backtrace, dev_set_drvdata() is called after sd_resume_runtime()
is called.

sd_probe()
{
scsi_autopm_get_device()
pm_runtime_get_sync()
__pm_runtime_resume()
rpm_resume()
...
sd_resume_runtime() // crash here

dev_set_drvdata(dev, sdkp); // sdkp is set later
}

[ 4.861395][ T151] sd_resume_runtime+0x20/0x14c
[ 4.862025][ T151] scsi_runtime_resume+0x84/0xe4
[ 4.862667][ T151] __rpm_callback+0x1f4/0x8cc
[ 4.863275][ T151] rpm_resume+0x7e8/0xaa4
[ 4.863836][ T151] __pm_runtime_resume+0xa0/0x110
[ 4.864489][ T151] sd_probe+0x30/0x428
[ 4.865016][ T151] really_probe+0x14c/0x500

2021-10-18 03:05:16

by Bart Van Assche

[permalink] [raw]
Subject: Re: [PATCH] scsi: sd: fix crashes in sd_resume_runtime

On 10/15/21 13:11, [email protected] wrote:
> I hit this in v5.15-rc1 merge, I can still reproduce this with v5.15-rc5.
> I found two ways to avoid the crash:
> 1) revert commit ed4246d37f3b ("scsi: sd: REQUEST SENSE for
> BLIST_IGN_MEDIA_CHANGE devices in runtime_resume()") works for me.
> 2) adding the NULL point check in this patch.
>
>>From the backtrace, dev_set_drvdata() is called after sd_resume_runtime()
> is called.
>
> sd_probe()
> {
> scsi_autopm_get_device()
> pm_runtime_get_sync()
> __pm_runtime_resume()
> rpm_resume()
> ...
> sd_resume_runtime() // crash here
>
> dev_set_drvdata(dev, sdkp); // sdkp is set later
> }
>
> [ 4.861395][ T151] sd_resume_runtime+0x20/0x14c
> [ 4.862025][ T151] scsi_runtime_resume+0x84/0xe4
> [ 4.862667][ T151] __rpm_callback+0x1f4/0x8cc
> [ 4.863275][ T151] rpm_resume+0x7e8/0xaa4
> [ 4.863836][ T151] __pm_runtime_resume+0xa0/0x110
> [ 4.864489][ T151] sd_probe+0x30/0x428
> [ 4.865016][ T151] really_probe+0x14c/0x500

Thanks for the clarification. Given this clarification I'm fine with
your patch.

Bart.


2021-10-18 03:21:54

by Miles Chen

[permalink] [raw]
Subject: Re: [PATCH] scsi: sd: fix crashes in sd_resume_runtime

> Thanks for the clarification. Given this clarification I'm fine with your patch.

> Bart.


thanks for your review

Miles

2021-10-19 03:50:10

by Martin K. Petersen

[permalink] [raw]
Subject: Re: [PATCH] scsi: sd: fix crashes in sd_resume_runtime

On Fri, 15 Oct 2021 15:46:54 +0800, Miles Chen wrote:

> After merging commit ed4246d37f3b ("scsi: sd: REQUEST SENSE for
> BLIST_IGN_MEDIA_CHANGE devices in runtime_resume()"), I hit the
> following crash on my device.
>
> static int sd_resume_runtime(struct device *dev)
> {
> struct scsi_disk *sdkp = dev_get_drvdata(dev);
> struct scsi_device *sdp = sdkp->device; // sdkp == NULL and crash
>
> [...]

Applied to 5.15/scsi-fixes, thanks!

[1/1] scsi: sd: fix crashes in sd_resume_runtime
https://git.kernel.org/mkp/scsi/c/85374b639229

--
Martin K. Petersen Oracle Linux Engineering