* [PATCH] scsi: core: clear driver private data when retry request
@ 2025-02-10 14:03 Ye Bin
2025-02-11 17:24 ` Bart Van Assche
0 siblings, 1 reply; 3+ messages in thread
From: Ye Bin @ 2025-02-10 14:03 UTC (permalink / raw)
To: jejb, martin.petersen, linux-scsi
From: Ye Bin <yebin10@huawei.com>
After commit 1bad6c4a57ef
("scsi: zero per-cmd private driver data for each MQ I/O"),
xen-scsifront/virtio_scsi/snic driver remove code that zeroes
driver-private command data. If request do retry will lead to
driver-private command data remains. Before commit 464a00c9e0ad
("scsi: core: Kill DRIVER_SENSE") if virtio_scsi do capacity
expansion, first request may return UA then request will do retry,
as driver-private command data remains, request will return UA
again. As a result, the request keeps retrying, and the request
times out and fails.
So zeroes driver-private command data when request do retry.
Fixes: f7de50da1479 ("scsi: xen-scsifront: Remove code that zeroes driver-private command data")
Fixes: c2bb87318baa ("scsi: virtio_scsi: Remove code that zeroes driver-private command data")
Fixes: c3006a926468 ("scsi: snic: Remove code that zeroes driver-private command data")
Signed-off-by: Ye Bin <yebin10@huawei.com>
---
drivers/scsi/scsi_lib.c | 19 +++++++++++++------
1 file changed, 13 insertions(+), 6 deletions(-)
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index be0890e4e706..5b0c109c89bb 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -1645,6 +1645,17 @@ static unsigned int scsi_mq_inline_sgl_size(struct Scsi_Host *shost)
sizeof(struct scatterlist);
}
+static inline void scsi_clear_lld_private_data(struct scsi_cmnd *cmd,
+ struct Scsi_Host *shost)
+{
+ /*
+ * Only clear the driver-private command data if the LLD does not supply
+ * a function to initialize that data.
+ */
+ if (!shost->hostt->init_cmd_priv && shost->hostt->cmd_size)
+ memset(cmd + 1, 0, shost->hostt->cmd_size);
+}
+
static blk_status_t scsi_prepare_cmd(struct request *req)
{
struct scsi_cmnd *cmd = blk_mq_rq_to_pdu(req);
@@ -1669,12 +1680,7 @@ static blk_status_t scsi_prepare_cmd(struct request *req)
if (in_flight)
__set_bit(SCMD_STATE_INFLIGHT, &cmd->state);
- /*
- * Only clear the driver-private command data if the LLD does not supply
- * a function to initialize that data.
- */
- if (!shost->hostt->init_cmd_priv)
- memset(cmd + 1, 0, shost->hostt->cmd_size);
+ scsi_clear_lld_private_data(cmd, shost);
cmd->prot_op = SCSI_PROT_NORMAL;
if (blk_rq_bytes(req))
@@ -1848,6 +1854,7 @@ static blk_status_t scsi_queue_rq(struct blk_mq_hw_ctx *hctx,
goto out_dec_host_busy;
req->rq_flags |= RQF_DONTPREP;
} else {
+ scsi_clear_lld_private_data(cmd, shost);
clear_bit(SCMD_STATE_COMPLETE, &cmd->state);
}
--
2.34.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] scsi: core: clear driver private data when retry request
2025-02-10 14:03 [PATCH] scsi: core: clear driver private data when retry request Ye Bin
@ 2025-02-11 17:24 ` Bart Van Assche
2025-02-12 1:33 ` yebin
0 siblings, 1 reply; 3+ messages in thread
From: Bart Van Assche @ 2025-02-11 17:24 UTC (permalink / raw)
To: Ye Bin, jejb, martin.petersen, linux-scsi
On 2/10/25 6:03 AM, Ye Bin wrote:
> From: Ye Bin <yebin10@huawei.com>
>
> After commit 1bad6c4a57ef
> ("scsi: zero per-cmd private driver data for each MQ I/O"),
> xen-scsifront/virtio_scsi/snic driver remove code that zeroes
> driver-private command data. If request do retry will lead to
> driver-private command data remains. Before commit 464a00c9e0ad
> ("scsi: core: Kill DRIVER_SENSE") if virtio_scsi do capacity
> expansion, first request may return UA then request will do retry,
> as driver-private command data remains, request will return UA
> again. As a result, the request keeps retrying, and the request
> times out and fails.
> So zeroes driver-private command data when request do retry.
>
> Fixes: f7de50da1479 ("scsi: xen-scsifront: Remove code that zeroes driver-private command data")
> Fixes: c2bb87318baa ("scsi: virtio_scsi: Remove code that zeroes driver-private command data")
> Fixes: c3006a926468 ("scsi: snic: Remove code that zeroes driver-private command data")
> Signed-off-by: Ye Bin <yebin10@huawei.com>
> ---
> drivers/scsi/scsi_lib.c | 19 +++++++++++++------
> 1 file changed, 13 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
> index be0890e4e706..5b0c109c89bb 100644
> --- a/drivers/scsi/scsi_lib.c
> +++ b/drivers/scsi/scsi_lib.c
> @@ -1645,6 +1645,17 @@ static unsigned int scsi_mq_inline_sgl_size(struct Scsi_Host *shost)
> sizeof(struct scatterlist);
> }
>
> +static inline void scsi_clear_lld_private_data(struct scsi_cmnd *cmd,
> + struct Scsi_Host *shost)
> +{
> + /*
> + * Only clear the driver-private command data if the LLD does not supply
> + * a function to initialize that data.
> + */
> + if (!shost->hostt->init_cmd_priv && shost->hostt->cmd_size)
> + memset(cmd + 1, 0, shost->hostt->cmd_size);
> +}
> +
> static blk_status_t scsi_prepare_cmd(struct request *req)
> {
> struct scsi_cmnd *cmd = blk_mq_rq_to_pdu(req);
> @@ -1669,12 +1680,7 @@ static blk_status_t scsi_prepare_cmd(struct request *req)
> if (in_flight)
> __set_bit(SCMD_STATE_INFLIGHT, &cmd->state);
>
> - /*
> - * Only clear the driver-private command data if the LLD does not supply
> - * a function to initialize that data.
> - */
> - if (!shost->hostt->init_cmd_priv)
> - memset(cmd + 1, 0, shost->hostt->cmd_size);
> + scsi_clear_lld_private_data(cmd, shost);
>
> cmd->prot_op = SCSI_PROT_NORMAL;
> if (blk_rq_bytes(req))
> @@ -1848,6 +1854,7 @@ static blk_status_t scsi_queue_rq(struct blk_mq_hw_ctx *hctx,
> goto out_dec_host_busy;
> req->rq_flags |= RQF_DONTPREP;
> } else {
> + scsi_clear_lld_private_data(cmd, shost);
> clear_bit(SCMD_STATE_COMPLETE, &cmd->state);
> }
Thanks for the detailed analysis. Has the following (untested) simpler
alternative been considered?
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index d776f13cd160..6ee2903b4adb 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -1664,13 +1664,6 @@ static blk_status_t scsi_prepare_cmd(struct
request *req)
if (in_flight)
__set_bit(SCMD_STATE_INFLIGHT, &cmd->state);
- /*
- * Only clear the driver-private command data if the LLD does not supply
- * a function to initialize that data.
- */
- if (!shost->hostt->init_cmd_priv)
- memset(cmd + 1, 0, shost->hostt->cmd_size);
-
cmd->prot_op = SCSI_PROT_NORMAL;
if (blk_rq_bytes(req))
cmd->sc_data_direction = rq_dma_dir(req);
@@ -1837,6 +1830,13 @@ static blk_status_t scsi_queue_rq(struct
blk_mq_hw_ctx *hctx,
if (!scsi_host_queue_ready(q, shost, sdev, cmd))
goto out_dec_target_busy;
+ /*
+ * Only clear the driver-private command data if the LLD does not supply
+ * a function to initialize that data.
+ */
+ if (!shost->hostt->init_cmd_priv && shost->hostt->cmd_size)
+ memset(scsi_cmd_priv(cmd), 0, shost->hostt->cmd_size);
+
if (!(req->rq_flags & RQF_DONTPREP)) {
ret = scsi_prepare_cmd(req);
if (ret != BLK_STS_OK)
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] scsi: core: clear driver private data when retry request
2025-02-11 17:24 ` Bart Van Assche
@ 2025-02-12 1:33 ` yebin
0 siblings, 0 replies; 3+ messages in thread
From: yebin @ 2025-02-12 1:33 UTC (permalink / raw)
To: Bart Van Assche, jejb, martin.petersen, linux-scsi
On 2025/2/12 1:24, Bart Van Assche wrote:
> On 2/10/25 6:03 AM, Ye Bin wrote:
>> From: Ye Bin <yebin10@huawei.com>
>>
>> After commit 1bad6c4a57ef
>> ("scsi: zero per-cmd private driver data for each MQ I/O"),
>> xen-scsifront/virtio_scsi/snic driver remove code that zeroes
>> driver-private command data. If request do retry will lead to
>> driver-private command data remains. Before commit 464a00c9e0ad
>> ("scsi: core: Kill DRIVER_SENSE") if virtio_scsi do capacity
>> expansion, first request may return UA then request will do retry,
>> as driver-private command data remains, request will return UA
>> again. As a result, the request keeps retrying, and the request
>> times out and fails.
>> So zeroes driver-private command data when request do retry.
>>
>> Fixes: f7de50da1479 ("scsi: xen-scsifront: Remove code that zeroes
>> driver-private command data")
>> Fixes: c2bb87318baa ("scsi: virtio_scsi: Remove code that zeroes
>> driver-private command data")
>> Fixes: c3006a926468 ("scsi: snic: Remove code that zeroes
>> driver-private command data")
>> Signed-off-by: Ye Bin <yebin10@huawei.com>
>> ---
>> drivers/scsi/scsi_lib.c | 19 +++++++++++++------
>> 1 file changed, 13 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
>> index be0890e4e706..5b0c109c89bb 100644
>> --- a/drivers/scsi/scsi_lib.c
>> +++ b/drivers/scsi/scsi_lib.c
>> @@ -1645,6 +1645,17 @@ static unsigned int
>> scsi_mq_inline_sgl_size(struct Scsi_Host *shost)
>> sizeof(struct scatterlist);
>> }
>> +static inline void scsi_clear_lld_private_data(struct scsi_cmnd *cmd,
>> + struct Scsi_Host *shost)
>> +{
>> + /*
>> + * Only clear the driver-private command data if the LLD does not
>> supply
>> + * a function to initialize that data.
>> + */
>> + if (!shost->hostt->init_cmd_priv && shost->hostt->cmd_size)
>> + memset(cmd + 1, 0, shost->hostt->cmd_size);
>> +}
>> +
>> static blk_status_t scsi_prepare_cmd(struct request *req)
>> {
>> struct scsi_cmnd *cmd = blk_mq_rq_to_pdu(req);
>> @@ -1669,12 +1680,7 @@ static blk_status_t scsi_prepare_cmd(struct
>> request *req)
>> if (in_flight)
>> __set_bit(SCMD_STATE_INFLIGHT, &cmd->state);
>> - /*
>> - * Only clear the driver-private command data if the LLD does not
>> supply
>> - * a function to initialize that data.
>> - */
>> - if (!shost->hostt->init_cmd_priv)
>> - memset(cmd + 1, 0, shost->hostt->cmd_size);
>> + scsi_clear_lld_private_data(cmd, shost);
>> cmd->prot_op = SCSI_PROT_NORMAL;
>> if (blk_rq_bytes(req))
>> @@ -1848,6 +1854,7 @@ static blk_status_t scsi_queue_rq(struct
>> blk_mq_hw_ctx *hctx,
>> goto out_dec_host_busy;
>> req->rq_flags |= RQF_DONTPREP;
>> } else {
>> + scsi_clear_lld_private_data(cmd, shost);
>> clear_bit(SCMD_STATE_COMPLETE, &cmd->state);
>> }
>
> Thanks for the detailed analysis. Has the following (untested) simpler
> alternative been considered?
>
Thank you for your reply.
I've considered your modification plan, but I think it's a little hard
to understand and the code feels a little loose. Of course, it's just my
own idea.
> diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
> index d776f13cd160..6ee2903b4adb 100644
> --- a/drivers/scsi/scsi_lib.c
> +++ b/drivers/scsi/scsi_lib.c
> @@ -1664,13 +1664,6 @@ static blk_status_t scsi_prepare_cmd(struct
> request *req)
> if (in_flight)
> __set_bit(SCMD_STATE_INFLIGHT, &cmd->state);
>
> - /*
> - * Only clear the driver-private command data if the LLD does not
> supply
> - * a function to initialize that data.
> - */
> - if (!shost->hostt->init_cmd_priv)
> - memset(cmd + 1, 0, shost->hostt->cmd_size);
> -
> cmd->prot_op = SCSI_PROT_NORMAL;
> if (blk_rq_bytes(req))
> cmd->sc_data_direction = rq_dma_dir(req);
> @@ -1837,6 +1830,13 @@ static blk_status_t scsi_queue_rq(struct
> blk_mq_hw_ctx *hctx,
> if (!scsi_host_queue_ready(q, shost, sdev, cmd))
> goto out_dec_target_busy;
>
> + /*
> + * Only clear the driver-private command data if the LLD does not
> supply
> + * a function to initialize that data.
> + */
> + if (!shost->hostt->init_cmd_priv && shost->hostt->cmd_size)
> + memset(scsi_cmd_priv(cmd), 0, shost->hostt->cmd_size);
> +
> if (!(req->rq_flags & RQF_DONTPREP)) {
> ret = scsi_prepare_cmd(req);
> if (ret != BLK_STS_OK)
>
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2025-02-12 1:33 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-02-10 14:03 [PATCH] scsi: core: clear driver private data when retry request Ye Bin
2025-02-11 17:24 ` Bart Van Assche
2025-02-12 1:33 ` yebin
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox