From: Jianchao Wang <jianchao.w.wang@oracle.com>
To: axboe@kernel.dk
Cc: ming.lei@redhat.com, linux-block@vger.kernel.org,
linux-kernel@vger.kernel.org
Subject: [PATCH V3 2/4] blk-mq: insert request without involving any io scheduler
Date: Mon, 29 Oct 2018 16:33:31 +0800 [thread overview]
Message-ID: <1540802013-2921-3-git-send-email-jianchao.w.wang@oracle.com> (raw)
In-Reply-To: <1540802013-2921-1-git-send-email-jianchao.w.wang@oracle.com>
When try to issue request directly, if the queue is stopped or
quiesced, 'bypass_insert' will be ignored and return BLK_STS_OK to
caller to avoid it dispatch request again. Then the request will
be inserted with blk_mq_sched_insert_request. This is not correct
for dm-rq case where we should avoid to pass through the underlying
paths' io scheduler.
To fix it, use blk_mq_request_bypass_insert to insert the request
into hctx->dispatch directly.
Signed-off-by: Jianchao Wang <jianchao.w.wang@oracle.com>
---
block/blk-mq.c | 17 ++++++++++++-----
1 file changed, 12 insertions(+), 5 deletions(-)
diff --git a/block/blk-mq.c b/block/blk-mq.c
index 0f6328b..e0aa068 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -1729,13 +1729,14 @@ static blk_status_t __blk_mq_issue_directly(struct blk_mq_hw_ctx *hctx,
}
enum mq_decision {
+ MQ_INSERT_DISPATCH,
MQ_INSERT_QUEUE,
MQ_END_REQUEST,
MQ_DO_NOTHING,
};
static inline enum mq_decision
- blk_mq_make_dicision(blk_status_t ret, bool bypass_insert)
+ blk_mq_make_dicision(blk_status_t ret, bool bypass_insert, bool force)
{
enum mq_decision dec;
@@ -1745,7 +1746,10 @@ static inline enum mq_decision
break;
case BLK_STS_DEV_RESOURCE:
case BLK_STS_RESOURCE:
- dec = bypass_insert ? MQ_DO_NOTHING : MQ_INSERT_QUEUE;
+ if (force)
+ dec = bypass_insert ? MQ_INSERT_DISPATCH : MQ_INSERT_QUEUE;
+ else
+ dec = bypass_insert ? MQ_DO_NOTHING : MQ_INSERT_QUEUE;
break;
default:
dec = bypass_insert ? MQ_DO_NOTHING : MQ_END_REQUEST;
@@ -1761,7 +1765,7 @@ static blk_status_t blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx,
bool bypass_insert)
{
struct request_queue *q = rq->q;
- bool run_queue = true;
+ bool run_queue = true, force = false;
enum mq_decision dec;
blk_status_t ret = BLK_STS_RESOURCE;
int srcu_idx;
@@ -1776,7 +1780,7 @@ static blk_status_t blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx,
*/
if (blk_mq_hctx_stopped(hctx) || blk_queue_quiesced(q)) {
run_queue = false;
- bypass_insert = false;
+ force = true;
goto out_unlock;
}
@@ -1795,8 +1799,11 @@ static blk_status_t blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx,
out_unlock:
hctx_unlock(hctx, srcu_idx);
- dec = blk_mq_make_dicision(ret, bypass_insert);
+ dec = blk_mq_make_dicision(ret, bypass_insert, force);
switch(dec) {
+ case MQ_INSERT_DISPATCH:
+ blk_mq_request_bypass_insert(rq, run_queue);
+ break;
case MQ_INSERT_QUEUE:
blk_mq_sched_insert_request(rq, false, run_queue, false);
break;
--
2.7.4
next prev parent reply other threads:[~2018-10-29 8:34 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-10-29 8:33 [PATCH V3 0/4] blk-mq: refactor and fix on issue request directly Jianchao Wang
2018-10-29 8:33 ` [PATCH V3 1/4] blk-mq: refactor the code of " Jianchao Wang
2018-10-29 8:42 ` jianchao.wang
2018-10-29 8:33 ` Jianchao Wang [this message]
2018-10-29 8:33 ` [PATCH V3 3/4] blk-mq: issue directly with bypass_insert 'false' in blk_mq_sched_insert_requests Jianchao Wang
2018-10-29 8:33 ` [PATCH V3 4/4] blk-mq: ensure hctx to be ran on mapped cpu when issue directly Jianchao Wang
2018-10-29 8:49 [PATCH RESENT V3 0/4] blk-mq: refactor and fix on issue request directly Jianchao Wang
2018-10-29 8:49 ` [PATCH V3 2/4] blk-mq: insert request without involving any io scheduler Jianchao Wang
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=1540802013-2921-3-git-send-email-jianchao.w.wang@oracle.com \
--to=jianchao.w.wang@oracle.com \
--cc=axboe@kernel.dk \
--cc=linux-block@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=ming.lei@redhat.com \
/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).