From: Pavel Begunkov <[email protected]>
To: [email protected]
Cc: Jens Axboe <[email protected]>,
[email protected], [email protected]
Subject: [PATCH v2 8/8] io_uring: optimise io_req_local_work_add
Date: Thu, 6 Apr 2023 14:20:14 +0100 [thread overview]
Message-ID: <fb11597e9bbcb365901824f8c5c2cf0d6ee100d0.1680782017.git.asml.silence@gmail.com> (raw)
In-Reply-To: <[email protected]>
Chains of memory accesses are never good for performance.
The req->task->io_uring->in_cancel in io_req_local_work_add() is there
so that when a task is exiting via io_uring_try_cancel_requests() and
starts waiting for completions it gets woken up by every new task_work
item queued.
Do a little trick by announcing waiting in io_uring_try_cancel_requests()
making io_req_local_work_add() to wake us up. We also need to check for
deferred tw items after prepare_to_wait(TASK_INTERRUPTIBLE);
Signed-off-by: Pavel Begunkov <[email protected]>
---
io_uring/io_uring.c | 24 ++++++++++++++++--------
1 file changed, 16 insertions(+), 8 deletions(-)
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index 0ea50c46f27f..9bbf58297a0e 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -1335,10 +1335,6 @@ static void io_req_local_work_add(struct io_kiocb *req, unsigned flags)
&req->io_task_work.node));
if (!first) {
- if (unlikely(atomic_read(&req->task->io_uring->in_cancel))) {
- io_move_task_work_from_local(ctx);
- return;
- }
if (ctx->flags & IORING_SETUP_TASKRUN_FLAG)
atomic_or(IORING_SQ_TASKRUN, &ctx->rings->sq_flags);
if (ctx->has_evfd)
@@ -3205,6 +3201,12 @@ static __cold bool io_uring_try_cancel_requests(struct io_ring_ctx *ctx,
enum io_wq_cancel cret;
bool ret = false;
+ /* set it so io_req_local_work_add() would wake us up */
+ if (ctx->flags & IORING_SETUP_DEFER_TASKRUN) {
+ atomic_set(&ctx->cq_wait_nr, 1);
+ smp_mb();
+ }
+
/* failed during ring init, it couldn't have issued any requests */
if (!ctx->rings)
return false;
@@ -3259,6 +3261,8 @@ __cold void io_uring_cancel_generic(bool cancel_all, struct io_sq_data *sqd)
{
struct io_uring_task *tctx = current->io_uring;
struct io_ring_ctx *ctx;
+ struct io_tctx_node *node;
+ unsigned long index;
s64 inflight;
DEFINE_WAIT(wait);
@@ -3280,9 +3284,6 @@ __cold void io_uring_cancel_generic(bool cancel_all, struct io_sq_data *sqd)
break;
if (!sqd) {
- struct io_tctx_node *node;
- unsigned long index;
-
xa_for_each(&tctx->xa, index, node) {
/* sqpoll task will cancel all its requests */
if (node->ctx->sq_data)
@@ -3305,7 +3306,13 @@ __cold void io_uring_cancel_generic(bool cancel_all, struct io_sq_data *sqd)
prepare_to_wait(&tctx->wait, &wait, TASK_INTERRUPTIBLE);
io_run_task_work();
io_uring_drop_tctx_refs(current);
-
+ xa_for_each(&tctx->xa, index, node) {
+ if (!llist_empty(&node->ctx->work_llist)) {
+ WARN_ON_ONCE(node->ctx->submitter_task &&
+ node->ctx->submitter_task != current);
+ goto end_wait;
+ }
+ }
/*
* If we've seen completions, retry without waiting. This
* avoids a race where a completion comes in before we did
@@ -3313,6 +3320,7 @@ __cold void io_uring_cancel_generic(bool cancel_all, struct io_sq_data *sqd)
*/
if (inflight == tctx_inflight(tctx, !cancel_all))
schedule();
+end_wait:
finish_wait(&tctx->wait, &wait);
} while (1);
--
2.40.0
next prev parent reply other threads:[~2023-04-06 13:21 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-04-06 13:20 [PATCH v2 0/8] optimise resheduling due to deferred tw Pavel Begunkov
2023-04-06 13:20 ` [PATCH v2 1/8] io_uring: move pinning out of io_req_local_work_add Pavel Begunkov
2023-04-06 13:20 ` [PATCH v2 2/8] io_uring: optimie local tw add ctx pinning Pavel Begunkov
2023-04-06 13:20 ` [PATCH v2 3/8] io_uring: refactor __io_cq_unlock_post_flush() Pavel Begunkov
2023-04-06 14:23 ` Pavel Begunkov
2023-04-06 13:20 ` [PATCH v2 4/8] io_uring: add tw add flags Pavel Begunkov
2023-04-06 13:20 ` [PATCH v2 5/8] io_uring: inline llist_add() Pavel Begunkov
2023-04-06 13:20 ` [PATCH v2 6/8] io_uring: reduce scheduling due to tw Pavel Begunkov
2023-04-06 13:20 ` [PATCH v2 7/8] io_uring: refactor __io_cq_unlock_post_flush() Pavel Begunkov
2023-04-06 13:20 ` Pavel Begunkov [this message]
2023-04-12 1:53 ` [PATCH v2 0/8] optimise resheduling due to deferred tw 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=fb11597e9bbcb365901824f8c5c2cf0d6ee100d0.1680782017.git.asml.silence@gmail.com \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
/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