after ufs UFS_ABORT_TASK process successfully , host will generate
mcq irq for abort tag with response OCS_ABORTED
ufshcd_compl_one_cqe ->
    ufshcd_release_scsi_cmd

But in ufshcd_mcq_abort already do ufshcd_release_scsi_cmd, this means
 __ufshcd_release will be done twice.

This means hba->clk_gating.active_reqs also will be decrease twice, it
will be negtive, so delete ufshcd_release_scsi_cmd in ufshcd_mcq_abort
function.

static void __ufshcd_release(struct ufs_hba *hba)
{
    if (!ufshcd_is_clkgating_allowed(hba))
        return;

    hba->clk_gating.active_reqs--;

    if (hba->clk_gating.active_reqs < 0) {
        panic("ufs abnormal active_reqs!!!!!!");
    }

    ...
}

Fixes: f1304d442077 ("scsi: ufs: mcq: Added ufshcd_mcq_abort()")
Signed-off-by: ping.gao <ping.gao@samsung.com>
---
 drivers/ufs/core/ufs-mcq.c | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/drivers/ufs/core/ufs-mcq.c b/drivers/ufs/core/ufs-mcq.c
index f1294c29f484..1e50675772fe 100644
--- a/drivers/ufs/core/ufs-mcq.c
+++ b/drivers/ufs/core/ufs-mcq.c
@@ -674,7 +674,6 @@ int ufshcd_mcq_abort(struct scsi_cmnd *cmd)
 	int tag = scsi_cmd_to_rq(cmd)->tag;
 	struct ufshcd_lrb *lrbp = &hba->lrb[tag];
 	struct ufs_hw_queue *hwq;
-	unsigned long flags;
 	int err;
 
 	/* Skip task abort in case previous aborts failed and report failure */
@@ -713,10 +712,5 @@ int ufshcd_mcq_abort(struct scsi_cmnd *cmd)
 		return FAILED;
 	}
 
-	spin_lock_irqsave(&hwq->cq_lock, flags);
-	if (ufshcd_cmd_inflight(lrbp->cmd))
-		ufshcd_release_scsi_cmd(hba, lrbp);
-	spin_unlock_irqrestore(&hwq->cq_lock, flags);
-
 	return SUCCESS;
 }
-- 
2.25.1

> after ufs UFS_ABORT_TASK process successfully , host will generate
> mcq irq for abort tag with response OCS_ABORTED
> ufshcd_compl_one_cqe ->
>     ufshcd_release_scsi_cmd
>
> But in ufshcd_mcq_abort already do ufshcd_release_scsi_cmd, this means
>  __ufshcd_release will be done twice.
>
> This means hba->clk_gating.active_reqs also will be decrease twice, it
> will be negtive, so delete ufshcd_release_scsi_cmd in ufshcd_mcq_abort
> function.

Applied to 6.16/scsi-staging, thanks!

-- 
Martin K. Petersen

On Mon, 2025-05-19 at 10:55 +0800, ping.gao wrote:
> after ufs UFS_ABORT_TASK process successfully , host will generate
> mcq irq for abort tag with response OCS_ABORTED
> ufshcd_compl_one_cqe ->
>     ufshcd_release_scsi_cmd
> 
> But in ufshcd_mcq_abort already do ufshcd_release_scsi_cmd, this
> means
>  __ufshcd_release will be done twice.
> 
> This means hba->clk_gating.active_reqs also will be decrease twice,
> it
> will be negtive, so delete ufshcd_release_scsi_cmd in
> ufshcd_mcq_abort
> function.
> 
> static void __ufshcd_release(struct ufs_hba *hba)
> {
>     if (!ufshcd_is_clkgating_allowed(hba))
>         return;
> 
>     hba->clk_gating.active_reqs--;
> 
>     if (hba->clk_gating.active_reqs < 0) {
>         panic("ufs abnormal active_reqs!!!!!!");
>     }
> 
>     ...
> }
> 
> Fixes: f1304d442077 ("scsi: ufs: mcq: Added ufshcd_mcq_abort()")
> Signed-off-by: ping.gao <ping.gao@samsung.com>

Reviewed-by: Peter Wang <peter.wang@mediatek.com>

On 5/18/25 7:55 PM, ping.gao wrote:
> after ufs UFS_ABORT_TASK process successfully , host will generate
> mcq irq for abort tag with response OCS_ABORTED
> ufshcd_compl_one_cqe ->
>      ufshcd_release_scsi_cmd

Reviewed-by: Bart Van Assche <bvanassche@acm.org>