All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jens Axboe <axboe@kernel.dk>
To: linux-block@vger.kernel.org, linux-nvme@lists.infradead.org
Cc: Jens Axboe <axboe@kernel.dk>
Subject: [PATCH 2/7] blk-mq: add mq_ops->commit_rqs()
Date: Wed, 28 Nov 2018 06:35:33 -0700	[thread overview]
Message-ID: <20181128133538.20329-3-axboe@kernel.dk> (raw)
In-Reply-To: <20181128133538.20329-1-axboe@kernel.dk>

blk-mq passes information to the hardware about any given request being
the last that we will issue in this sequence. The point is that hardware
can defer costly doorbell type writes to the last request. But if we run
into errors issuing a sequence of requests, we may never send the request
with bd->last == true set. For that case, we need a hook that tells the
hardware that nothing else is coming right now.

For failures returned by the drivers ->queue_rq() hook, the driver is
responsible for flushing pending requests, if it uses bd->last to
optimize that part. This works like before, no changes there.

Reviewed-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-mq.c         | 16 ++++++++++++++++
 include/linux/blk-mq.h | 10 ++++++++++
 2 files changed, 26 insertions(+)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 0e2663048d95..18ff47a85ad3 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -1259,6 +1259,14 @@ bool blk_mq_dispatch_rq_list(struct request_queue *q, struct list_head *list,
 	if (!list_empty(list)) {
 		bool needs_restart;
 
+		/*
+		 * If we didn't flush the entire list, we could have told
+		 * the driver there was more coming, but that turned out to
+		 * be a lie.
+		 */
+		if (q->mq_ops->commit_rqs)
+			q->mq_ops->commit_rqs(hctx);
+
 		spin_lock(&hctx->lock);
 		list_splice_init(list, &hctx->dispatch);
 		spin_unlock(&hctx->lock);
@@ -1865,6 +1873,14 @@ void blk_mq_try_issue_list_directly(struct blk_mq_hw_ctx *hctx,
 			blk_mq_end_request(rq, ret);
 		}
 	}
+
+	/*
+	 * If we didn't flush the entire list, we could have told
+	 * the driver there was more coming, but that turned out to
+	 * be a lie.
+	 */
+	if (!list_empty(list) && hctx->queue->mq_ops->commit_rqs)
+		hctx->queue->mq_ops->commit_rqs(hctx);
 }
 
 static void blk_add_rq_to_plug(struct blk_plug *plug, struct request *rq)
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index b8de11e0603b..467f1dd21ccf 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -117,6 +117,7 @@ struct blk_mq_queue_data {
 
 typedef blk_status_t (queue_rq_fn)(struct blk_mq_hw_ctx *,
 		const struct blk_mq_queue_data *);
+typedef void (commit_rqs_fn)(struct blk_mq_hw_ctx *);
 /* takes rq->cmd_flags as input, returns a hardware type index */
 typedef int (rq_flags_to_type_fn)(struct request_queue *, unsigned int);
 typedef bool (get_budget_fn)(struct blk_mq_hw_ctx *);
@@ -144,6 +145,15 @@ struct blk_mq_ops {
 	 */
 	queue_rq_fn		*queue_rq;
 
+	/*
+	 * If a driver uses bd->last to judge when to submit requests to
+	 * hardware, it must define this function. In case of errors that
+	 * make us stop issuing further requests, this hook serves the
+	 * purpose of kicking the hardware (which the last request otherwise
+	 * would have done).
+	 */
+	commit_rqs_fn		*commit_rqs;
+
 	/*
 	 * Return a queue map type for the given request/bio flags
 	 */
-- 
2.17.1


WARNING: multiple messages have this Message-ID (diff)
From: axboe@kernel.dk (Jens Axboe)
Subject: [PATCH 2/7] blk-mq: add mq_ops->commit_rqs()
Date: Wed, 28 Nov 2018 06:35:33 -0700	[thread overview]
Message-ID: <20181128133538.20329-3-axboe@kernel.dk> (raw)
In-Reply-To: <20181128133538.20329-1-axboe@kernel.dk>

blk-mq passes information to the hardware about any given request being
the last that we will issue in this sequence. The point is that hardware
can defer costly doorbell type writes to the last request. But if we run
into errors issuing a sequence of requests, we may never send the request
with bd->last == true set. For that case, we need a hook that tells the
hardware that nothing else is coming right now.

For failures returned by the drivers ->queue_rq() hook, the driver is
responsible for flushing pending requests, if it uses bd->last to
optimize that part. This works like before, no changes there.

Reviewed-by: Omar Sandoval <osandov at fb.com>
Reviewed-by: Ming Lei <ming.lei at redhat.com>
Reviewed-by: Christoph Hellwig <hch at lst.de>
Signed-off-by: Jens Axboe <axboe at kernel.dk>
---
 block/blk-mq.c         | 16 ++++++++++++++++
 include/linux/blk-mq.h | 10 ++++++++++
 2 files changed, 26 insertions(+)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 0e2663048d95..18ff47a85ad3 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -1259,6 +1259,14 @@ bool blk_mq_dispatch_rq_list(struct request_queue *q, struct list_head *list,
 	if (!list_empty(list)) {
 		bool needs_restart;
 
+		/*
+		 * If we didn't flush the entire list, we could have told
+		 * the driver there was more coming, but that turned out to
+		 * be a lie.
+		 */
+		if (q->mq_ops->commit_rqs)
+			q->mq_ops->commit_rqs(hctx);
+
 		spin_lock(&hctx->lock);
 		list_splice_init(list, &hctx->dispatch);
 		spin_unlock(&hctx->lock);
@@ -1865,6 +1873,14 @@ void blk_mq_try_issue_list_directly(struct blk_mq_hw_ctx *hctx,
 			blk_mq_end_request(rq, ret);
 		}
 	}
+
+	/*
+	 * If we didn't flush the entire list, we could have told
+	 * the driver there was more coming, but that turned out to
+	 * be a lie.
+	 */
+	if (!list_empty(list) && hctx->queue->mq_ops->commit_rqs)
+		hctx->queue->mq_ops->commit_rqs(hctx);
 }
 
 static void blk_add_rq_to_plug(struct blk_plug *plug, struct request *rq)
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index b8de11e0603b..467f1dd21ccf 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -117,6 +117,7 @@ struct blk_mq_queue_data {
 
 typedef blk_status_t (queue_rq_fn)(struct blk_mq_hw_ctx *,
 		const struct blk_mq_queue_data *);
+typedef void (commit_rqs_fn)(struct blk_mq_hw_ctx *);
 /* takes rq->cmd_flags as input, returns a hardware type index */
 typedef int (rq_flags_to_type_fn)(struct request_queue *, unsigned int);
 typedef bool (get_budget_fn)(struct blk_mq_hw_ctx *);
@@ -144,6 +145,15 @@ struct blk_mq_ops {
 	 */
 	queue_rq_fn		*queue_rq;
 
+	/*
+	 * If a driver uses bd->last to judge when to submit requests to
+	 * hardware, it must define this function. In case of errors that
+	 * make us stop issuing further requests, this hook serves the
+	 * purpose of kicking the hardware (which the last request otherwise
+	 * would have done).
+	 */
+	commit_rqs_fn		*commit_rqs;
+
 	/*
 	 * Return a queue map type for the given request/bio flags
 	 */
-- 
2.17.1

  parent reply	other threads:[~2018-11-28 13:35 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-11-28 13:35 [PATCHSET v2 0/8] block plugging improvements Jens Axboe
2018-11-28 13:35 ` Jens Axboe
2018-11-28 13:35 ` [PATCH 1/7] block: improve logic around when to sort a plug list Jens Axboe
2018-11-28 13:35   ` Jens Axboe
2018-11-29 15:44   ` Christoph Hellwig
2018-11-29 15:44     ` Christoph Hellwig
2018-12-04  1:35   ` Sagi Grimberg
2018-12-04  1:35     ` Sagi Grimberg
2018-11-28 13:35 ` Jens Axboe [this message]
2018-11-28 13:35   ` [PATCH 2/7] blk-mq: add mq_ops->commit_rqs() Jens Axboe
2018-11-29 15:45   ` Christoph Hellwig
2018-11-29 15:45     ` Christoph Hellwig
2018-11-29 16:52     ` Jens Axboe
2018-11-29 16:52       ` Jens Axboe
2018-12-04  1:35   ` Sagi Grimberg
2018-12-04  1:35     ` Sagi Grimberg
2018-11-28 13:35 ` [PATCH 3/7] nvme: implement mq_ops->commit_rqs() hook Jens Axboe
2018-11-28 13:35   ` Jens Axboe
2018-11-29 15:47   ` Christoph Hellwig
2018-11-29 15:47     ` Christoph Hellwig
2018-11-29 17:02     ` Jens Axboe
2018-11-29 17:02       ` Jens Axboe
2018-11-29 17:04       ` Christoph Hellwig
2018-11-29 17:04         ` Christoph Hellwig
2018-11-29 17:06         ` Jens Axboe
2018-11-29 17:06           ` Jens Axboe
2018-11-29 17:38           ` Keith Busch
2018-11-29 17:38             ` Keith Busch
2018-11-29 17:42             ` Jens Axboe
2018-11-29 17:42               ` Jens Axboe
2018-11-28 13:35 ` [PATCH 4/7] virtio_blk: " Jens Axboe
2018-11-28 13:35   ` Jens Axboe
2018-12-04  1:36   ` Sagi Grimberg
2018-12-04  1:36     ` Sagi Grimberg
2018-11-28 13:35 ` [PATCH 5/7] ataflop: " Jens Axboe
2018-11-28 13:35   ` Jens Axboe
2018-11-29  7:03   ` Ming Lei
2018-11-29  7:03     ` Ming Lei
2018-11-29 15:47   ` Christoph Hellwig
2018-11-29 15:47     ` Christoph Hellwig
2018-12-04  1:37   ` Sagi Grimberg
2018-12-04  1:37     ` Sagi Grimberg
2018-11-28 13:35 ` [PATCH 6/7] blk-mq: use bd->last == true for list inserts Jens Axboe
2018-11-28 13:35   ` Jens Axboe
2018-11-29  7:01   ` Ming Lei
2018-11-29  7:01     ` Ming Lei
2018-11-29 15:49   ` Christoph Hellwig
2018-11-29 15:49     ` Christoph Hellwig
2018-12-04  1:39   ` Sagi Grimberg
2018-12-04  1:39     ` Sagi Grimberg
2018-11-28 13:35 ` [PATCH 7/7] blk-mq: use plug for devices that implement ->commits_rqs() Jens Axboe
2018-11-28 13:35   ` Jens Axboe
2018-11-29  7:13   ` Ming Lei
2018-11-29  7:13     ` Ming Lei
2018-11-29 15:49   ` Christoph Hellwig
2018-11-29 15:49     ` Christoph Hellwig
2018-11-29 15:50     ` Christoph Hellwig
2018-11-29 15:50       ` Christoph Hellwig
2018-11-29 17:05     ` Jens Axboe
2018-11-29 17:05       ` Jens Axboe
2018-12-04  1:39   ` Sagi Grimberg
2018-12-04  1:39     ` Sagi Grimberg

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20181128133538.20329-3-axboe@kernel.dk \
    --to=axboe@kernel.dk \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-nvme@lists.infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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.