* [PATCH V2] scsi: avoid to run synchronize_rcu for each device in scsi_host_block
@ 2020-04-23 2:07 Ming Lei
2020-04-23 5:41 ` Hannes Reinecke
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: Ming Lei @ 2020-04-23 2:07 UTC (permalink / raw)
To: James Bottomley, linux-scsi, Martin K . Petersen
Cc: Ming Lei, Steffen Maier, Bart Van Assche, Christoph Hellwig,
Dexuan Cui, Hannes Reinecke
scsi_host_block() calls scsi_internal_device_block() for each
scsi_device, and scsi_internal_device_block() calls
blk_mq_quiesce_queue() for each LUN. However synchronize_rcu is
run from blk_mq_quiesce_queue().
This way may become unnecessary slow in case of lots of LUNs.
Use scsi_internal_device_block_nowait() to implement scsi_host_block(),
so it is enough to run synchronize_rcu() once since scsi never
supports blk-mq's BLK_MQ_F_BLOCKING.
Cc: Steffen Maier <maier@linux.ibm.com>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dexuan Cui <decui@microsoft.com>
Cc: Hannes Reinecke <hare@suse.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
---
V2:
- fix commit log as pointed by Steffen
- add comment on BLK_MQ_F_BLOCKING as suggested by Bart
drivers/scsi/scsi_lib.c | 17 ++++++++++++++++-
1 file changed, 16 insertions(+), 1 deletion(-)
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 47835c4b4ee0..86007a523145 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -2841,11 +2841,26 @@ scsi_host_block(struct Scsi_Host *shost)
struct scsi_device *sdev;
int ret = 0;
+ /*
+ * Call scsi_internal_device_block_nowait then we can avoid to
+ * run synchronize_rcu() for each LUN
+ */
shost_for_each_device(sdev, shost) {
- ret = scsi_internal_device_block(sdev);
+ mutex_lock(&sdev->state_mutex);
+ ret = scsi_internal_device_block_nowait(sdev);
+ mutex_unlock(&sdev->state_mutex);
if (ret)
break;
}
+
+ /*
+ * SCSI never enables blk-mq's BLK_MQ_F_BLOCKING, so run
+ * synchronize_rcu() once is enough
+ */
+ WARN_ON_ONCE(shost->tag_set.flags & BLK_MQ_F_BLOCKING);
+
+ if (!ret)
+ synchronize_rcu();
return ret;
}
EXPORT_SYMBOL_GPL(scsi_host_block);
--
2.25.2
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH V2] scsi: avoid to run synchronize_rcu for each device in scsi_host_block
2020-04-23 2:07 [PATCH V2] scsi: avoid to run synchronize_rcu for each device in scsi_host_block Ming Lei
@ 2020-04-23 5:41 ` Hannes Reinecke
2020-04-24 6:47 ` Christoph Hellwig
2020-04-24 18:12 ` Martin K. Petersen
2 siblings, 0 replies; 4+ messages in thread
From: Hannes Reinecke @ 2020-04-23 5:41 UTC (permalink / raw)
To: Ming Lei, James Bottomley, linux-scsi, Martin K . Petersen
Cc: Steffen Maier, Bart Van Assche, Christoph Hellwig, Dexuan Cui
On 4/23/20 4:07 AM, Ming Lei wrote:
> scsi_host_block() calls scsi_internal_device_block() for each
> scsi_device, and scsi_internal_device_block() calls
> blk_mq_quiesce_queue() for each LUN. However synchronize_rcu is
> run from blk_mq_quiesce_queue().
>
> This way may become unnecessary slow in case of lots of LUNs.
>
> Use scsi_internal_device_block_nowait() to implement scsi_host_block(),
> so it is enough to run synchronize_rcu() once since scsi never
> supports blk-mq's BLK_MQ_F_BLOCKING.
>
> Cc: Steffen Maier <maier@linux.ibm.com>
> Cc: Bart Van Assche <bvanassche@acm.org>
> Cc: Christoph Hellwig <hch@lst.de>
> Cc: Dexuan Cui <decui@microsoft.com>
> Cc: Hannes Reinecke <hare@suse.de>
> Signed-off-by: Ming Lei <ming.lei@redhat.com>
> ---
> V2:
> - fix commit log as pointed by Steffen
> - add comment on BLK_MQ_F_BLOCKING as suggested by Bart
>
> drivers/scsi/scsi_lib.c | 17 ++++++++++++++++-
> 1 file changed, 16 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
> index 47835c4b4ee0..86007a523145 100644
> --- a/drivers/scsi/scsi_lib.c
> +++ b/drivers/scsi/scsi_lib.c
> @@ -2841,11 +2841,26 @@ scsi_host_block(struct Scsi_Host *shost)
> struct scsi_device *sdev;
> int ret = 0;
>
> + /*
> + * Call scsi_internal_device_block_nowait then we can avoid to
> + * run synchronize_rcu() for each LUN
> + */
> shost_for_each_device(sdev, shost) {
> - ret = scsi_internal_device_block(sdev);
> + mutex_lock(&sdev->state_mutex);
> + ret = scsi_internal_device_block_nowait(sdev);
> + mutex_unlock(&sdev->state_mutex);
> if (ret)
> break;
> }
> +
> + /*
> + * SCSI never enables blk-mq's BLK_MQ_F_BLOCKING, so run
> + * synchronize_rcu() once is enough
> + */
> + WARN_ON_ONCE(shost->tag_set.flags & BLK_MQ_F_BLOCKING);
> +
> + if (!ret)
> + synchronize_rcu();
> return ret;
> }
> EXPORT_SYMBOL_GPL(scsi_host_block);
>
Looks good.
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cheers,
Hannes
--
Dr. Hannes Reinecke Teamlead Storage & Networking
hare@suse.de +49 911 74053 688
SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), Geschäftsführer: Felix Imendörffer
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH V2] scsi: avoid to run synchronize_rcu for each device in scsi_host_block
2020-04-23 2:07 [PATCH V2] scsi: avoid to run synchronize_rcu for each device in scsi_host_block Ming Lei
2020-04-23 5:41 ` Hannes Reinecke
@ 2020-04-24 6:47 ` Christoph Hellwig
2020-04-24 18:12 ` Martin K. Petersen
2 siblings, 0 replies; 4+ messages in thread
From: Christoph Hellwig @ 2020-04-24 6:47 UTC (permalink / raw)
To: Ming Lei
Cc: James Bottomley, linux-scsi, Martin K . Petersen, Steffen Maier,
Bart Van Assche, Christoph Hellwig, Dexuan Cui, Hannes Reinecke
Looks good,
Reviewed-by: Christoph Hellwig <hch@lst.de>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH V2] scsi: avoid to run synchronize_rcu for each device in scsi_host_block
2020-04-23 2:07 [PATCH V2] scsi: avoid to run synchronize_rcu for each device in scsi_host_block Ming Lei
2020-04-23 5:41 ` Hannes Reinecke
2020-04-24 6:47 ` Christoph Hellwig
@ 2020-04-24 18:12 ` Martin K. Petersen
2 siblings, 0 replies; 4+ messages in thread
From: Martin K. Petersen @ 2020-04-24 18:12 UTC (permalink / raw)
To: Ming Lei
Cc: James Bottomley, linux-scsi, Martin K . Petersen, Steffen Maier,
Bart Van Assche, Christoph Hellwig, Dexuan Cui, Hannes Reinecke
Ming,
> scsi_host_block() calls scsi_internal_device_block() for each
> scsi_device, and scsi_internal_device_block() calls
> blk_mq_quiesce_queue() for each LUN. However synchronize_rcu is
> run from blk_mq_quiesce_queue().
Applied to 5.8/scsi-queue, thank you!
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2020-04-24 18:12 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-04-23 2:07 [PATCH V2] scsi: avoid to run synchronize_rcu for each device in scsi_host_block Ming Lei
2020-04-23 5:41 ` Hannes Reinecke
2020-04-24 6:47 ` Christoph Hellwig
2020-04-24 18:12 ` Martin K. Petersen
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.