[PATCH] scsi: ufs: Fix referencing invalid rsp field

[PATCH] scsi: ufs: Fix referencing invalid rsp field

[Daejun Park]

This is a patch for fixing referencing sense data when it is invalid.
When the length of the data segment is 0, there is no valid information in
the rsp field, so ufshpb_rsp_upiu() is returned without additional
operation.

Signed-off-by: Daejun Park <daejun7.park@samsung.com>
---
 drivers/scsi/ufs/ufshpb.c | 19 +++++++------------
 1 file changed, 7 insertions(+), 12 deletions(-)

diff --git a/drivers/scsi/ufs/ufshpb.c b/drivers/scsi/ufs/ufshpb.c
index f1f65383e97d..8882b47f76d3 100644
--- a/drivers/scsi/ufs/ufshpb.c
+++ b/drivers/scsi/ufs/ufshpb.c
@@ -1304,6 +1304,13 @@ void ufshpb_rsp_upiu(struct ufs_hba *hba, struct ufshcd_lrb *lrbp)
 	struct utp_hpb_rsp *rsp_field = &lrbp->ucd_rsp_ptr->hr;
 	int data_seg_len;
 
+	data_seg_len = be32_to_cpu(lrbp->ucd_rsp_ptr->header.dword_2)
+		& MASK_RSP_UPIU_DATA_SEG_LEN;
+
+	/* If data segment length is zero, rsp_field is not valid */
+	if (!data_seg_len)
+		return;
+
 	if (unlikely(lrbp->lun != rsp_field->lun)) {
 		struct scsi_device *sdev;
 		bool found = false;
@@ -1338,18 +1345,6 @@ void ufshpb_rsp_upiu(struct ufs_hba *hba, struct ufshcd_lrb *lrbp)
 		return;
 	}
 
-	data_seg_len = be32_to_cpu(lrbp->ucd_rsp_ptr->header.dword_2)
-		& MASK_RSP_UPIU_DATA_SEG_LEN;
-
-	/* To flush remained rsp_list, we queue the map_work task */
-	if (!data_seg_len) {
-		if (!ufshpb_is_general_lun(hpb->lun))
-			return;
-
-		ufshpb_kick_map_work(hpb);
-		return;
-	}
-
 	BUILD_BUG_ON(sizeof(struct utp_hpb_rsp) != UTP_HPB_RSP_SIZE);
 
 	if (!ufshpb_is_hpb_rsp_valid(hba, lrbp, rsp_field))
-- 
2.25.1

RE: [PATCH] scsi: ufs: Fix referencing invalid rsp field

[Avri Altman]

> This is a patch for fixing referencing sense data when it is invalid.
> When the length of the data segment is 0, there is no valid information in
> the rsp field, so ufshpb_rsp_upiu() is returned without additional
> operation.
> 
> Signed-off-by: Daejun Park <daejun7.park@samsung.com>
Acked-by: Avri Altman <avri.altman@wdc.com>

RE: [PATCH] scsi: ufs: Fix referencing invalid rsp field

[Chanho Park]

> This is a patch for fixing referencing sense data when it is invalid.
> When the length of the data segment is 0, there is no valid information in
> the rsp field, so ufshpb_rsp_upiu() is returned without additional
> operation.
> 

You should put a proper fix tag to be applied the corresponding stable tree.
See https://www.kernel.org/doc/html/latest/process/submitting-patches.html

Fixes: 4b5f49079c52 ("scsi: ufs: ufshpb: L2P map management for HPB read")

Best Regards,
Chanho Park

> Signed-off-by: Daejun Park <daejun7.park@samsung.com>
> ---
>  drivers/scsi/ufs/ufshpb.c | 19 +++++++------------
>  1 file changed, 7 insertions(+), 12 deletions(-)
> 
> diff --git a/drivers/scsi/ufs/ufshpb.c b/drivers/scsi/ufs/ufshpb.c index
> f1f65383e97d..8882b47f76d3 100644
> --- a/drivers/scsi/ufs/ufshpb.c
> +++ b/drivers/scsi/ufs/ufshpb.c
> @@ -1304,6 +1304,13 @@ void ufshpb_rsp_upiu(struct ufs_hba *hba, struct
> ufshcd_lrb *lrbp)
>  	struct utp_hpb_rsp *rsp_field = &lrbp->ucd_rsp_ptr->hr;
>  	int data_seg_len;
> 
> +	data_seg_len = be32_to_cpu(lrbp->ucd_rsp_ptr->header.dword_2)
> +		& MASK_RSP_UPIU_DATA_SEG_LEN;
> +
> +	/* If data segment length is zero, rsp_field is not valid */
> +	if (!data_seg_len)
> +		return;
> +
>  	if (unlikely(lrbp->lun != rsp_field->lun)) {
>  		struct scsi_device *sdev;
>  		bool found = false;
> @@ -1338,18 +1345,6 @@ void ufshpb_rsp_upiu(struct ufs_hba *hba, struct
> ufshcd_lrb *lrbp)
>  		return;
>  	}
> 
> -	data_seg_len = be32_to_cpu(lrbp->ucd_rsp_ptr->header.dword_2)
> -		& MASK_RSP_UPIU_DATA_SEG_LEN;
> -
> -	/* To flush remained rsp_list, we queue the map_work task */
> -	if (!data_seg_len) {
> -		if (!ufshpb_is_general_lun(hpb->lun))
> -			return;
> -
> -		ufshpb_kick_map_work(hpb);
> -		return;
> -	}
> -
>  	BUILD_BUG_ON(sizeof(struct utp_hpb_rsp) != UTP_HPB_RSP_SIZE);
> 
>  	if (!ufshpb_is_hpb_rsp_valid(hba, lrbp, rsp_field))
> --
> 2.25.1
> 

Re: [PATCH] scsi: ufs: Fix referencing invalid rsp field

[Martin K. Petersen]

On Thu, 19 May 2022 15:05:29 +0900, Daejun Park wrote:

> This is a patch for fixing referencing sense data when it is invalid.
> When the length of the data segment is 0, there is no valid information in
> the rsp field, so ufshpb_rsp_upiu() is returned without additional
> operation.
> 
> 

Applied to 5.18/scsi-fixes, thanks!

[1/1] scsi: ufs: Fix referencing invalid rsp field
      https://git.kernel.org/mkp/scsi/c/d5d92b644084

-- 
Martin K. Petersen	Oracle Linux Engineering