From: Pavel Begunkov <[email protected]>
To: Jens Axboe <[email protected]>, [email protected]
Subject: [PATCH 01/13] io_uring: introduce a new helper for ctx quiesce
Date: Tue, 23 Feb 2021 01:55:36 +0000 [thread overview]
Message-ID: <9f40843d172aa8dfb631755b9a71ced79d310f49.1614045169.git.asml.silence@gmail.com> (raw)
In-Reply-To: <[email protected]>
Extract a helper io_uring_quiesce_ctx(), which takes care of ctx
quiesce, it's complex enough.
Signed-off-by: Pavel Begunkov <[email protected]>
---
fs/io_uring.c | 52 ++++++++++++++++++++++++++++-----------------------
1 file changed, 29 insertions(+), 23 deletions(-)
diff --git a/fs/io_uring.c b/fs/io_uring.c
index 3f764f2f2982..6e781c064e74 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -9683,6 +9683,33 @@ static bool io_register_op_must_quiesce(int op)
}
}
+static int io_uring_quiesce_ctx(struct io_ring_ctx *ctx)
+{
+ int ret;
+
+ percpu_ref_kill(&ctx->refs);
+ /*
+ * Drop uring mutex before waiting for references to exit. If
+ * another thread is currently inside io_uring_enter() it might
+ * need to grab the uring_lock to make progress. If we hold it
+ * here across the drain wait, then we can deadlock. It's safe
+ * to drop the mutex here, since no new references will come in
+ * after we've killed the percpu ref.
+ */
+ mutex_unlock(&ctx->uring_lock);
+ do {
+ ret = wait_for_completion_interruptible(&ctx->ref_comp);
+ if (!ret)
+ break;
+ ret = io_run_task_work_sig();
+ } while (ret >= 0);
+ mutex_lock(&ctx->uring_lock);
+
+ if (ret && io_refs_resurrect(&ctx->refs, &ctx->ref_comp))
+ return ret;
+ return 0;
+}
+
static int __io_uring_register(struct io_ring_ctx *ctx, unsigned opcode,
void __user *arg, unsigned nr_args)
__releases(ctx->uring_lock)
@@ -9699,29 +9726,8 @@ static int __io_uring_register(struct io_ring_ctx *ctx, unsigned opcode,
return -ENXIO;
if (io_register_op_must_quiesce(opcode)) {
- percpu_ref_kill(&ctx->refs);
-
- /*
- * Drop uring mutex before waiting for references to exit. If
- * another thread is currently inside io_uring_enter() it might
- * need to grab the uring_lock to make progress. If we hold it
- * here across the drain wait, then we can deadlock. It's safe
- * to drop the mutex here, since no new references will come in
- * after we've killed the percpu ref.
- */
- mutex_unlock(&ctx->uring_lock);
- do {
- ret = wait_for_completion_interruptible(&ctx->ref_comp);
- if (!ret)
- break;
- ret = io_run_task_work_sig();
- if (ret < 0)
- break;
- } while (1);
-
- mutex_lock(&ctx->uring_lock);
-
- if (ret && io_refs_resurrect(&ctx->refs, &ctx->ref_comp))
+ ret = io_uring_quiesce_ctx(ctx);
+ if (ret)
return ret;
}
--
2.24.0
next prev parent reply other threads:[~2021-02-23 2:00 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-02-23 1:55 [PATCH for-next 00/13] simple 5.13 cleanups Pavel Begunkov
2021-02-23 1:55 ` Pavel Begunkov [this message]
2021-02-23 1:55 ` [PATCH 02/13] io_uring: avoid taking ctx refs for task-cancel Pavel Begunkov
2021-02-23 1:55 ` [PATCH 03/13] io_uring: reuse io_req_task_queue_fail() Pavel Begunkov
2021-02-23 1:55 ` [PATCH 04/13] io_uring: further deduplicate file slot selection Pavel Begunkov
2021-02-23 1:55 ` [PATCH 05/13] io_uring: add a helper failing not issued requests Pavel Begunkov
2021-02-23 1:55 ` [PATCH 06/13] io_uring: refactor provide/remove buffer locking Pavel Begunkov
2021-02-23 1:55 ` [PATCH 07/13] io_uring: don't restirct issue_flags for io_openat Pavel Begunkov
2021-02-23 1:55 ` [PATCH 08/13] io_uring: use better types for cflags Pavel Begunkov
2021-02-23 1:55 ` [PATCH 09/13] io_uring: refactor out send/recv async setup Pavel Begunkov
2021-02-23 1:55 ` [PATCH 10/13] io_uring: untie alloc_async_data and needs_async_data Pavel Begunkov
2021-02-23 1:55 ` [PATCH 11/13] io_uring: rethink def->needs_async_data Pavel Begunkov
2021-02-23 1:55 ` [PATCH 12/13] io_uring: merge defer_prep() and prep_async() Pavel Begunkov
2021-02-23 1:55 ` [PATCH 13/13] io_uring: simplify io_resubmit_prep() Pavel Begunkov
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=9f40843d172aa8dfb631755b9a71ced79d310f49.1614045169.git.asml.silence@gmail.com \
[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