2020-11-09 09:12:06

by Qinglang Miao

[permalink] [raw]
Subject: [PATCH] scsi: ufshcd: fix missing destroy_workqueue() on error in ufshcd_init

Add the missing destroy_workqueue() before return from
ufshcd_init in the error handling case. It seems that
exit_gating is an appropriate place.

Fixes: 4db7a2360597 ("scsi: ufs: Fix concurrency of error handler and other error recovery paths")
Signed-off-by: Qinglang Miao <[email protected]>
---
drivers/scsi/ufs/ufshcd.c | 1 +
1 file changed, 1 insertion(+)

diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
index b8f573a02713..9eaa0eaca374 100644
--- a/drivers/scsi/ufs/ufshcd.c
+++ b/drivers/scsi/ufs/ufshcd.c
@@ -9206,6 +9206,7 @@ int ufshcd_init(struct ufs_hba *hba, void __iomem *mmio_base, unsigned int irq)
exit_gating:
ufshcd_exit_clk_scaling(hba);
ufshcd_exit_clk_gating(hba);
+ destroy_workqueue(hba->eh_wq);
out_disable:
hba->is_irq_enabled = false;
ufshcd_hba_exit(hba);
--
2.23.0


2020-11-09 09:42:33

by Avri Altman

[permalink] [raw]
Subject: RE: [PATCH] scsi: ufshcd: fix missing destroy_workqueue() on error in ufshcd_init

>
> Add the missing destroy_workqueue() before return from
> ufshcd_init in the error handling case. It seems that
> exit_gating is an appropriate place.
>
> Fixes: 4db7a2360597 ("scsi: ufs: Fix concurrency of error handler and other
> error recovery paths")
> Signed-off-by: Qinglang Miao <[email protected]>
> ---
> drivers/scsi/ufs/ufshcd.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
> index b8f573a02713..9eaa0eaca374 100644
> --- a/drivers/scsi/ufs/ufshcd.c
> +++ b/drivers/scsi/ufs/ufshcd.c
> @@ -9206,6 +9206,7 @@ int ufshcd_init(struct ufs_hba *hba, void __iomem
> *mmio_base, unsigned int irq)
> exit_gating:
> ufshcd_exit_clk_scaling(hba);
> ufshcd_exit_clk_gating(hba);
> + destroy_workqueue(hba->eh_wq);
Maybe also in ufshcd_remove?

2020-11-10 01:21:15

by Qinglang Miao

[permalink] [raw]
Subject: Re: [PATCH] scsi: ufshcd: fix missing destroy_workqueue() on error in ufshcd_init



?? 2020/11/9 17:40, Avri Altman ะด??:
>>
>> Add the missing destroy_workqueue() before return from
>> ufshcd_init in the error handling case. It seems that
>> exit_gating is an appropriate place.
>>
>> Fixes: 4db7a2360597 ("scsi: ufs: Fix concurrency of error handler and other
>> error recovery paths")
>> Signed-off-by: Qinglang Miao <[email protected]>
>> ---
>> drivers/scsi/ufs/ufshcd.c | 1 +
>> 1 file changed, 1 insertion(+)
>>
>> diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
>> index b8f573a02713..9eaa0eaca374 100644
>> --- a/drivers/scsi/ufs/ufshcd.c
>> +++ b/drivers/scsi/ufs/ufshcd.c
>> @@ -9206,6 +9206,7 @@ int ufshcd_init(struct ufs_hba *hba, void __iomem
>> *mmio_base, unsigned int irq)
>> exit_gating:
>> ufshcd_exit_clk_scaling(hba);
>> ufshcd_exit_clk_gating(hba);
>> + destroy_workqueue(hba->eh_wq);
> Maybe also in ufshcd_remove?
> .
You're right Avri, thanks!

I'm gonna send a v2 on this patch to cover ufshcd_remove.
>