From: Pavel Begunkov <[email protected]>
To: Jens Axboe <[email protected]>, [email protected]
Subject: [PATCH v3 09/17] io_uring: remove useless is_dying check on quiesce
Date: Thu, 25 Mar 2021 13:07:58 +0000 [thread overview]
Message-ID: <11c2a6b222cc3bae50e4643edebd45a4fc9737d3.1616677487.git.asml.silence@gmail.com> (raw)
In-Reply-To: <[email protected]>
rsrc_data refs should always be valid for potential submitters,
io_rsrc_ref_quiesce() restores it before unlocking, so
percpu_ref_is_dying() check in io_sqe_files_unregister() does nothing
and misleading. Concurrent quiesce is prevented with
struct io_rsrc_data::quiesce.
Signed-off-by: Pavel Begunkov <[email protected]>
---
fs/io_uring.c | 8 ++------
1 file changed, 2 insertions(+), 6 deletions(-)
diff --git a/fs/io_uring.c b/fs/io_uring.c
index 9da4a1981560..a57fe91f06d2 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -7005,6 +7005,7 @@ static int io_rsrc_ref_quiesce(struct io_rsrc_data *data, struct io_ring_ctx *ct
{
int ret;
+ /* As we may drop ->uring_lock, other task may have started quiesce */
if (data->quiesce)
return -ENXIO;
@@ -7067,12 +7068,7 @@ static int io_sqe_files_unregister(struct io_ring_ctx *ctx)
unsigned nr_tables, i;
int ret;
- /*
- * percpu_ref_is_dying() is to stop parallel files unregister
- * Since we possibly drop uring lock later in this function to
- * run task work.
- */
- if (!data || percpu_ref_is_dying(&data->refs))
+ if (!data)
return -ENXIO;
ret = io_rsrc_ref_quiesce(data, ctx);
if (ret)
--
2.24.0
next prev parent reply other threads:[~2021-03-25 13:13 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-25 13:07 [PATCH v3 for-5.13 00/17] ctx wide rsrc nodes + other stuff Pavel Begunkov
2021-03-25 13:07 ` [PATCH v3 01/17] io_uring: name rsrc bits consistently Pavel Begunkov
2021-03-25 13:07 ` [PATCH v3 02/17] io_uring: simplify io_rsrc_node_ref_zero Pavel Begunkov
2021-03-25 13:07 ` [PATCH v3 03/17] io_uring: use rsrc prealloc infra for files reg Pavel Begunkov
2021-03-25 13:07 ` [PATCH v3 04/17] io_uring: encapsulate rsrc node manipulations Pavel Begunkov
2021-03-25 13:07 ` [PATCH v3 05/17] io_uring: move rsrc_put callback into io_rsrc_data Pavel Begunkov
2021-03-25 13:07 ` [PATCH v3 06/17] io_uring: refactor io_queue_rsrc_removal() Pavel Begunkov
2021-03-25 13:07 ` [PATCH v3 07/17] io_uring: ctx-wide rsrc nodes Pavel Begunkov
2021-03-25 13:07 ` [PATCH v3 08/17] io_uring: reuse io_rsrc_node_destroy() Pavel Begunkov
2021-03-25 13:07 ` Pavel Begunkov [this message]
2021-03-25 13:07 ` [PATCH v3 10/17] io_uring: refactor rw reissue Pavel Begunkov
2021-03-25 13:08 ` [PATCH v3 11/17] io_uring: combine lock/unlock sections on exit Pavel Begunkov
2021-03-25 13:08 ` [PATCH v3 12/17] io_uring: better ref handling in poll_remove_one Pavel Begunkov
2021-03-25 13:08 ` [PATCH v3 13/17] io_uring: remove unused hash_wait Pavel Begunkov
2021-03-25 13:08 ` [PATCH v3 14/17] io_uring: refactor io_async_cancel() Pavel Begunkov
2021-03-25 13:08 ` [PATCH v3 15/17] io_uring: improve import_fixed overflow checks Pavel Begunkov
2021-03-25 13:08 ` [PATCH v3 16/17] io_uring: store reg buffer end instead of length Pavel Begunkov
2021-03-25 13:08 ` [PATCH v3 17/17] io_uring: kill unused forward decls 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=11c2a6b222cc3bae50e4643edebd45a4fc9737d3.1616677487.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