mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Caleb Sander Mateos <csander@purestorage.com>
To: Jens Axboe <axboe@kernel.dk>
Cc: io-uring@vger.kernel.org, linux-kernel@vger.kernel.org,
	Caleb Sander Mateos <csander@purestorage.com>
Subject: [PATCH 4/4] io_uring/futex: avoid ifd NULL check in io_futex_wait()
Date: Tue,  3 Nov 2026 14:10:31 -0700	[thread overview]
Message-ID: <20261103211031.592745-5-csander@purestorage.com> (raw)
In-Reply-To: <20261103211031.592745-1-csander@purestorage.com>

There are two paths that hit the done_unlock label in io_futex_wait():
the ENOMEM error path allocating ifd (via goto done_unlock) and the
futex_wait_setup() error path (via a fallthrough). In the ENOMEM path,
ifd is always NULL; in the futex_wait_setup() error path, it's always
non-NULL. Move the done_unlock label so the goto skips the ifd freeing
and the success path can free it unconditionally. This also avoids the
need to initialize ifd before the io_cache_alloc() call, as it's only
used if io_cache_alloc() succeeds.

Signed-off-by: Caleb Sander Mateos <csander@purestorage.com>
---
 io_uring/futex.c | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)

diff --git a/io_uring/futex.c b/io_uring/futex.c
index 9c7d384bab9b..e0cb815fd37a 100644
--- a/io_uring/futex.c
+++ b/io_uring/futex.c
@@ -317,11 +317,11 @@ int io_futexv_wait(struct io_kiocb *req, unsigned int issue_flags)
 
 int io_futex_wait(struct io_kiocb *req, unsigned int issue_flags)
 {
 	struct io_futex *iof = io_kiocb_to_cmd(req, struct io_futex);
 	struct io_ring_ctx *ctx = req->ctx;
-	struct io_futex_data *ifd = NULL;
+	struct io_futex_data *ifd;
 	int ret;
 
 	if (!iof->futex_mask) {
 		ret = -EINVAL;
 		goto done;
@@ -349,16 +349,14 @@ int io_futex_wait(struct io_kiocb *req, unsigned int issue_flags)
 		io_ring_submit_unlock(ctx, issue_flags);
 
 		return IOU_ISSUE_SKIP_COMPLETE;
 	}
 
+	/* As in io_futex_complete(), or failed arms would drain it. */
+	io_cache_free(&ctx->futex_cache, ifd);
+	io_req_async_data_clear(req, 0);
 done_unlock:
-	if (ifd) {
-		/* As in io_futex_complete(), or failed arms would drain it. */
-		io_cache_free(&ctx->futex_cache, ifd);
-		io_req_async_data_clear(req, 0);
-	}
 	io_ring_submit_unlock(ctx, issue_flags);
 done:
 	if (ret < 0)
 		req_set_fail(req);
 	io_req_set_res(req, ret, 0);
-- 
2.56.0


      parent reply	other threads:[~2026-11-03 21:10 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-03 21:10 [PATCH 0/4] io_uring: improve alloc_cache branch prediction Caleb Sander Mateos
2026-11-03 21:10 ` [PATCH 1/4] io_uring: split io_uring_alloc_async_data() Caleb Sander Mateos
2026-11-03 21:10 ` [PATCH 2/4] io_uring/alloc_cache: mark cache success paths as likely Caleb Sander Mateos
2026-11-03 21:10 ` [PATCH 3/4] io_uring/alloc_cache: assume io_alloc_cache_get() doesn't return NULL Caleb Sander Mateos
2026-11-03 21:10 ` Caleb Sander Mateos [this message]

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=20261103211031.592745-5-csander@purestorage.com \
    --to=csander@purestorage.com \
    --cc=axboe@kernel.dk \
    --cc=io-uring@vger.kernel.org \
    --cc=linux-kernel@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

all inboxes | Powered by JetHome®