From: Christoph Hellwig <hch@lst.de>
To: Jens Axboe <axboe@kernel.dk>
Cc: Bart Van Assche <bvanassche@acm.org>,
Damien Le Moal <dlemoal@kernel.org>,
linux-block@vger.kernel.org
Subject: [PATCH 4/7] blk-mq: use the I/O scheduler for writes from the flush state machine
Date: Fri, 19 May 2023 06:40:47 +0200 [thread overview]
Message-ID: <20230519044050.107790-5-hch@lst.de> (raw)
In-Reply-To: <20230519044050.107790-1-hch@lst.de>
From: Bart Van Assche <bvanassche@acm.org>
Send write requests issued by the flush state machine through the normal
I/O submission path including the I/O scheduler (if present) so that I/O
scheduler policies are applied to writes with the FUA flag set.
Separate the I/O scheduler members from the flush members in struct
request since now a request may pass through both an I/O scheduler
and the flush machinery.
Note that the actual flush requests, which have no bio attached to the
request still bypass the I/O schedulers.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
[hch: rebased]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
---
block/blk-mq.c | 4 ++--
include/linux/blk-mq.h | 27 +++++++++++----------------
2 files changed, 13 insertions(+), 18 deletions(-)
diff --git a/block/blk-mq.c b/block/blk-mq.c
index c0b394096b6b6b..aac67bc3d3680c 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -458,7 +458,7 @@ static struct request *__blk_mq_alloc_requests(struct blk_mq_alloc_data *data)
* Flush/passthrough requests are special and go directly to the
* dispatch list.
*/
- if (!op_is_flush(data->cmd_flags) &&
+ if ((data->cmd_flags & REQ_OP_MASK) != REQ_OP_FLUSH &&
!blk_op_is_passthrough(data->cmd_flags)) {
struct elevator_mq_ops *ops = &q->elevator->type->ops;
@@ -2497,7 +2497,7 @@ static void blk_mq_insert_request(struct request *rq, blk_insert_t flags)
* dispatch it given we prioritize requests in hctx->dispatch.
*/
blk_mq_request_bypass_insert(rq, flags);
- } else if (rq->rq_flags & RQF_FLUSH_SEQ) {
+ } else if (req_op(rq) == REQ_OP_FLUSH) {
/*
* Firstly normal IO request is inserted to scheduler queue or
* sw queue, meantime we add flush request to dispatch queue(
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index 49d14b1acfa5df..935201c8974371 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -169,25 +169,20 @@ struct request {
void *completion_data;
};
-
/*
* Three pointers are available for the IO schedulers, if they need
- * more they have to dynamically allocate it. Flush requests are
- * never put on the IO scheduler. So let the flush fields share
- * space with the elevator data.
+ * more they have to dynamically allocate it.
*/
- union {
- struct {
- struct io_cq *icq;
- void *priv[2];
- } elv;
-
- struct {
- unsigned int seq;
- struct list_head list;
- rq_end_io_fn *saved_end_io;
- } flush;
- };
+ struct {
+ struct io_cq *icq;
+ void *priv[2];
+ } elv;
+
+ struct {
+ unsigned int seq;
+ struct list_head list;
+ rq_end_io_fn *saved_end_io;
+ } flush;
union {
struct __call_single_data csd;
--
2.39.2
next prev parent reply other threads:[~2023-05-19 4:41 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-19 4:40 less special casing for flush requests v2 Christoph Hellwig
2023-05-19 4:40 ` [PATCH 1/7] blk-mq: factor out a blk_rq_init_flush helper Christoph Hellwig
2023-05-19 18:54 ` Bart Van Assche
2023-05-19 4:40 ` [PATCH 2/7] blk-mq: reflow blk_insert_flush Christoph Hellwig
2023-05-19 19:38 ` Bart Van Assche
2023-05-19 4:40 ` [PATCH 3/7] blk-mq: defer to the normal submission path for non-flush flush commands Christoph Hellwig
2023-05-19 4:40 ` Christoph Hellwig [this message]
2023-05-19 19:44 ` [PATCH 4/7] blk-mq: use the I/O scheduler for writes from the flush state machine Bart Van Assche
2023-05-24 5:53 ` Christoph Hellwig
2023-05-24 18:07 ` Bart Van Assche
2023-05-25 8:14 ` Christoph Hellwig
2023-05-28 20:50 ` Bart Van Assche
2023-05-30 14:55 ` Christoph Hellwig
2023-05-31 13:34 ` Bart Van Assche
2023-05-31 15:00 ` Christoph Hellwig
2023-05-19 4:40 ` [PATCH 5/7] blk-mq: defer to the normal submission path for post-flush requests Christoph Hellwig
2023-05-19 19:42 ` Bart Van Assche
2023-05-19 4:40 ` [PATCH 6/7] blk-mq: do not do head insertions post-pre-flush commands Christoph Hellwig
2023-05-19 4:40 ` [PATCH 7/7] blk-mq: don't use the requeue list to queue flush commands Christoph Hellwig
2023-05-19 19:55 ` Bart Van Assche
2023-05-20 4:56 ` Christoph Hellwig
2023-05-21 14:06 ` Bart Van Assche
2023-05-20 1:53 ` less special casing for flush requests v2 Jens Axboe
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=20230519044050.107790-5-hch@lst.de \
--to=hch@lst.de \
--cc=axboe@kernel.dk \
--cc=bvanassche@acm.org \
--cc=dlemoal@kernel.org \
--cc=linux-block@vger.kernel.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).