2020-11-19 06:33:05

by Stanley Chu

Subject: [PATCH v1] scsi: ufs: Fix race between shutdown and runtime resume flow

If UFS host device is in runtime-suspended state while
UFS shutdown callback is invoked, UFS device shall be
resumed for register accesses. Currently only UFS local
runtime resume function will be invoked to wake up the host.
This is not enough because if someone triggers runtime
resume from block layer, then race may happen between
shutdown and runtime resume flow, and finally lead to
unlocked register access.

To fix this kind of issues, in ufshcd_shutdown(), use
pm_runtime_get_sync() instead of resuming UFS device by
ufshcd_runtime_resume() "internally" to let runtime PM
framework manage the whole resume flow.

Fixes: 57d104c153d3 ("ufs: add UFS power management support")
Signed-off-by: Stanley Chu <[email protected]>
---
drivers/scsi/ufs/ufshcd.c | 6 +-----
1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
index 80cbce414678..bb16cc04f106 100644
--- a/drivers/scsi/ufs/ufshcd.c
+++ b/drivers/scsi/ufs/ufshcd.c
@@ -8941,11 +8941,7 @@ int ufshcd_shutdown(struct ufs_hba *hba)
if (ufshcd_is_ufs_dev_poweroff(hba) && ufshcd_is_link_off(hba))
goto out;

- if (pm_runtime_suspended(hba->dev)) {
- ret = ufshcd_runtime_resume(hba);
- if (ret)
- goto out;
- }
+ pm_runtime_get_sync(hba->dev);

ret = ufshcd_suspend(hba, UFS_SHUTDOWN_PM);
out:
--
2.18.0

2020-11-19 07:16:38

by Can Guo

Subject: Re: [PATCH v1] scsi: ufs: Fix race between shutdown and runtime resume flow

On 2020-11-19 14:29, Stanley Chu wrote:
> If UFS host device is in runtime-suspended state while
> UFS shutdown callback is invoked, UFS device shall be
> resumed for register accesses. Currently only UFS local
> runtime resume function will be invoked to wake up the host.
> This is not enough because if someone triggers runtime
> resume from block layer, then race may happen between
> shutdown and runtime resume flow, and finally lead to
> unlocked register access.
>
> To fix this kind of issues, in ufshcd_shutdown(), use
> pm_runtime_get_sync() instead of resuming UFS device by
> ufshcd_runtime_resume() "internally" to let runtime PM
> framework manage the whole resume flow.
>
> Fixes: 57d104c153d3 ("ufs: add UFS power management support")
> Signed-off-by: Stanley Chu <[email protected]>

Reviewed-by: Can Guo <[email protected]>

> ---
> drivers/scsi/ufs/ufshcd.c | 6 +-----
> 1 file changed, 1 insertion(+), 5 deletions(-)
>
> diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
> index 80cbce414678..bb16cc04f106 100644
> --- a/drivers/scsi/ufs/ufshcd.c
> +++ b/drivers/scsi/ufs/ufshcd.c
> @@ -8941,11 +8941,7 @@ int ufshcd_shutdown(struct ufs_hba *hba)
> if (ufshcd_is_ufs_dev_poweroff(hba) && ufshcd_is_link_off(hba))
> goto out;
>
> - if (pm_runtime_suspended(hba->dev)) {
> - ret = ufshcd_runtime_resume(hba);
> - if (ret)
> - goto out;
> - }
> + pm_runtime_get_sync(hba->dev);
>
> ret = ufshcd_suspend(hba, UFS_SHUTDOWN_PM);
> out:

2020-11-20 03:35:03

by Martin K. Petersen

Subject: Re: [PATCH v1] scsi: ufs: Fix race between shutdown and runtime resume flow

On Thu, 19 Nov 2020 14:29:16 +0800, Stanley Chu wrote:

> If UFS host device is in runtime-suspended state while
> UFS shutdown callback is invoked, UFS device shall be
> resumed for register accesses. Currently only UFS local
> runtime resume function will be invoked to wake up the host.
> This is not enough because if someone triggers runtime
> resume from block layer, then race may happen between
> shutdown and runtime resume flow, and finally lead to
> unlocked register access.
>
> [...]

Applied to 5.10/scsi-fixes, thanks!

[1/1] scsi: ufs: Fix race between shutdown and runtime resume flow
https://git.kernel.org/mkp/scsi/c/e92643db5148

--
Martin K. Petersen Oracle Linux Engineering