[PATCH] scsi: ufs: Check workqueue exists before destroying it

[PATCH] scsi: ufs: Check workqueue exists before destroying it

[Matthias Kaehlcke]

In ufshcd_hba_exit() the clock gating workqueue is always destroyed
when clock gating is supported. However the workqueue pointer is NULL
when ufshcd_init() fails before the workqueue is created. Only destroy
the workqueue if it exists.

Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
---
 drivers/scsi/ufs/ufshcd.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
index 397081d320b1..401f7dbe2981 100644
--- a/drivers/scsi/ufs/ufshcd.c
+++ b/drivers/scsi/ufs/ufshcd.c
@@ -7190,7 +7190,8 @@ static void ufshcd_hba_exit(struct ufs_hba *hba)
 		if (ufshcd_is_clkscaling_supported(hba)) {
 			if (hba->devfreq)
 				ufshcd_suspend_clkscaling(hba);
-			destroy_workqueue(hba->clk_scaling.workq);
+			if (hba->clk_scaling.workq)
+				destroy_workqueue(hba->clk_scaling.workq);
 			ufshcd_devfreq_remove(hba);
 		}
 		ufshcd_setup_clocks(hba, false);
-- 
2.18.0.597.ga71716f1ad-goog

Re: [PATCH] scsi: ufs: Check workqueue exists before destroying it

[Evan Green]

On Wed, Aug 1, 2018 at 5:14 PM Matthias Kaehlcke <mka@chromium.org> wrote:
>
> In ufshcd_hba_exit() the clock gating workqueue is always destroyed
> when clock gating is supported. However the workqueue pointer is NULL
> when ufshcd_init() fails before the workqueue is created. Only destroy
> the workqueue if it exists.
>
> Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> ---
>  drivers/scsi/ufs/ufshcd.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
> index 397081d320b1..401f7dbe2981 100644
> --- a/drivers/scsi/ufs/ufshcd.c
> +++ b/drivers/scsi/ufs/ufshcd.c
> @@ -7190,7 +7190,8 @@ static void ufshcd_hba_exit(struct ufs_hba *hba)
>                 if (ufshcd_is_clkscaling_supported(hba)) {
>                         if (hba->devfreq)
>                                 ufshcd_suspend_clkscaling(hba);
> -                       destroy_workqueue(hba->clk_scaling.workq);
> +                       if (hba->clk_scaling.workq)
> +                               destroy_workqueue(hba->clk_scaling.workq);
>                         ufshcd_devfreq_remove(hba);
>                 }
>                 ufshcd_setup_clocks(hba, false);
> --
> 2.18.0.597.ga71716f1ad-goog
>

Reviewed-by: Evan Green <evgreen@chromium.org>

Re: [PATCH] scsi: ufs: Check workqueue exists before destroying it

[Stephen Boyd]

Quoting Matthias Kaehlcke (2018-08-01 17:14:35)
> In ufshcd_hba_exit() the clock gating workqueue is always destroyed
> when clock gating is supported. However the workqueue pointer is NULL
> when ufshcd_init() fails before the workqueue is created. Only destroy
> the workqueue if it exists.
> 
> Signed-off-by: Matthias Kaehlcke <mka@chromium.org>

Any sort of Fixes: tag?

Re: [PATCH] scsi: ufs: Check workqueue exists before destroying it

[Matthias Kaehlcke]

On Tue, Aug 07, 2018 at 11:24:50PM -0700, Stephen Boyd wrote:
> Quoting Matthias Kaehlcke (2018-08-01 17:14:35)
> > In ufshcd_hba_exit() the clock gating workqueue is always destroyed
> > when clock gating is supported. However the workqueue pointer is NULL
> > when ufshcd_init() fails before the workqueue is created. Only destroy
> > the workqueue if it exists.
> > 
> > Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> 
> Any sort of Fixes: tag?

I could add one, though it seems this patch is superseded by
https://patchwork.kernel.org/patch/10559013/