Re: [PATCH 9/9] scsi: ufs: Properly release resources if a task is aborted successfully

Re: [PATCH 9/9] scsi: ufs: Properly release resources if a task is aborted successfully

[Can Guo]

Hi Markus,

On 2020-08-07 17:33, Markus Elfring wrote:
>> … To fix it, …
> 
> I propose to replace this wording by the tag “Fixes”.
> 
> 
>> … "mannually", …
> 
> Please avoid a typo:
> … "manually", …
> 
> 
> Regards,
> Markus

Thanks, will fix these in next version.

Regards,
Can Guo.

Re: [9/9] scsi: ufs: Properly release resources if a task is aborted successfully

[Can Guo]

Hi Markus,

On 2020-08-10 02:00, Markus Elfring wrote:
>> Thanks, will fix these in next version.
> 
> Thanks for your adjustment of the proposed commit message.
> Should the corresponding patch be marked with an other version number?
> https://lore.kernel.org/linux-arm-kernel/1596975355-39813-10-git-send-email-cang@codeaurora.org/
> https://lore.kernel.org/patchwork/patch/1285629/
> https://lkml.org/lkml/2020/8/9/87
> 
> Should a cover letter be provided for such a patch series?
> 
> Regards,
> Markus

I am not sure if you got my mails, this patch is included in
a patch series with other 8 changes and they do have a cover
letter. Let me re-send them to you.

Thanks,

Can Guo.

[PATCH 9/9] scsi: ufs: Properly release resources if a task is aborted successfully

[Can Guo]

In current UFS task abort hook, namely ufshcd_abort(), if one task is
aborted successfully, clk_gating.active_reqs held by this task is not
decreased, which makes clk_gating.active_reqs stay above zero forever,
thus clock gating would never happen. Instead of releasing resources of
one task "manually", use the existing func __ufshcd_transfer_req_compl().
This change can also eliminate possible racing of scsi_dma_unmap() from
the real completion in IRQ handler path.

Fixes: 5a0b0cb9bee76 ("ufs: Add support for clock gating")
Signed-off-by: Can Guo <cang@codeaurora.org>
CC: Stanley Chu <stanley.chu@mediatek.com>
Reviewed-by: Stanley Chu <stanley.chu@mediatek.com>
Reviewed-by: Asutosh Das <asutoshd@codeaurora.org>

diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
index 9ebb5cd..efb40b1 100644
--- a/drivers/scsi/ufs/ufshcd.c
+++ b/drivers/scsi/ufs/ufshcd.c
@@ -6636,11 +6636,8 @@ static int ufshcd_abort(struct scsi_cmnd *cmd)
 		goto out;
 	}
 
-	scsi_dma_unmap(cmd);
-
 	spin_lock_irqsave(host->host_lock, flags);
-	ufshcd_outstanding_req_clear(hba, tag);
-	hba->lrb[tag].cmd = NULL;
+	__ufshcd_transfer_req_compl(hba, (1UL << tag));
 	spin_unlock_irqrestore(host->host_lock, flags);
 
 out:
-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project.

[PATCH 9/9] scsi: ufs: Properly release resources if a task is aborted successfully

[Can Guo]

In current UFS task abort hook, namely ufshcd_abort(), if a task is
aborted successfully, clock scaling busy time statistics is not updated
and, most important, clk_gating.active_reqs is not decreased, which makes
clk_gating.active_reqs stay above zero forever, thus clock gating would
never happen. To fix it, instead of releasing resources "mannually", use
the existing func __ufshcd_transfer_req_compl(). This can also eliminate
racing of scsi_dma_unmap() from the real completion in IRQ handler path.

Signed-off-by: Can Guo <cang@codeaurora.org>
CC: Stanley Chu <stanley.chu@mediatek.com>
Reviewed-by: Stanley Chu <stanley.chu@mediatek.com>

diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
index b2947ab..9541fc7 100644
--- a/drivers/scsi/ufs/ufshcd.c
+++ b/drivers/scsi/ufs/ufshcd.c
@@ -6636,11 +6636,8 @@ static int ufshcd_abort(struct scsi_cmnd *cmd)
 		goto out;
 	}
 
-	scsi_dma_unmap(cmd);
-
 	spin_lock_irqsave(host->host_lock, flags);
-	ufshcd_outstanding_req_clear(hba, tag);
-	hba->lrb[tag].cmd = NULL;
+	__ufshcd_transfer_req_compl(hba, (1UL << tag));
 	spin_unlock_irqrestore(host->host_lock, flags);
 
 out:
-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project.