[PATCH] scsi: ufs: Adjust ufshcd_hold() during sending attribute requests

[PATCH] scsi: ufs: Adjust ufshcd_hold() during sending attribute requests

[Jintae Jang]

From: jintae jang <jt77.jang@samsung.com>

Invalidation check of arguments should have been checked before
ufshcd_hold(). It can help to prevent ufshcd_hold()/
ufshcd_release() from being invoked unnecessarily.

Signed-off-by: jintae jang <jt77.jang@samsung.com>
---
 drivers/scsi/ufs/ufshcd.c | 15 +++++++--------
 1 file changed, 7 insertions(+), 8 deletions(-)

diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
index 51b4e0a..0b60931 100644
--- a/drivers/scsi/ufs/ufshcd.c
+++ b/drivers/scsi/ufs/ufshcd.c
@@ -2978,14 +2978,14 @@ int ufshcd_query_attr(struct ufs_hba *hba, enum query_opcode opcode,
 
 	BUG_ON(!hba);
 
-	ufshcd_hold(hba, false);
 	if (!attr_val) {
 		dev_err(hba->dev, "%s: attribute value required for opcode 0x%x\n",
 				__func__, opcode);
-		err = -EINVAL;
-		goto out;
+		return -EINVAL;
 	}
 
+	ufshcd_hold(hba, false);
+
 	mutex_lock(&hba->dev_cmd.lock);
 	ufshcd_init_query(hba, &request, &response, opcode, idn, index,
 			selector);
@@ -3069,21 +3069,20 @@ static int __ufshcd_query_descriptor(struct ufs_hba *hba,
 
 	BUG_ON(!hba);
 
-	ufshcd_hold(hba, false);
 	if (!desc_buf) {
 		dev_err(hba->dev, "%s: descriptor buffer required for opcode 0x%x\n",
 				__func__, opcode);
-		err = -EINVAL;
-		goto out;
+		return -EINVAL;
 	}
 
 	if (*buf_len < QUERY_DESC_MIN_SIZE || *buf_len > QUERY_DESC_MAX_SIZE) {
 		dev_err(hba->dev, "%s: descriptor buffer size (%d) is out of range\n",
 				__func__, *buf_len);
-		err = -EINVAL;
-		goto out;
+		return -EINVAL;
 	}
 
+	ufshcd_hold(hba, false);
+
 	mutex_lock(&hba->dev_cmd.lock);
 	ufshcd_init_query(hba, &request, &response, opcode, idn, index,
 			selector);
-- 
1.9.1

Re: [PATCH] scsi: ufs: Adjust ufshcd_hold() during sending attribute requests

[Stanley Chu]

On Thu, 2020-12-03 at 14:25 +0900, Jintae Jang wrote:
> From: jintae jang <jt77.jang@samsung.com>
> 
> Invalidation check of arguments should have been checked before
> ufshcd_hold(). It can help to prevent ufshcd_hold()/
> ufshcd_release() from being invoked unnecessarily.
> 
> Signed-off-by: jintae jang <jt77.jang@samsung.com>

Reviewed-by: Stanley Chu <stanley.chu@mediatek.com>

Re: [PATCH] scsi: ufs: Adjust ufshcd_hold() during sending attribute requests

[Martin K. Petersen]

Jintae,

> Invalidation check of arguments should have been checked before
> ufshcd_hold(). It can help to prevent ufshcd_hold()/ ufshcd_release()
> from being invoked unnecessarily.

Applied to 5.11/scsi-staging, thanks!

-- 
Martin K. Petersen	Oracle Linux Engineering

Re: [PATCH] scsi: ufs: Adjust ufshcd_hold() during sending attribute requests

[Martin K. Petersen]

On Thu, 3 Dec 2020 14:25:32 +0900, Jintae Jang wrote:

> Invalidation check of arguments should have been checked before
> ufshcd_hold(). It can help to prevent ufshcd_hold()/
> ufshcd_release() from being invoked unnecessarily.

Applied to 5.11/scsi-queue, thanks!

[1/1] scsi: ufs: Adjust ufshcd_hold() during sending attribute requests
      https://git.kernel.org/mkp/scsi/c/8ca1a40b9f9d

-- 
Martin K. Petersen	Oracle Linux Engineering