From: Pavel Begunkov <[email protected]>
To: [email protected]
Cc: Jens Axboe <[email protected]>,
[email protected], Li Shi <[email protected]>
Subject: [PATCH] io_uring: fix cancellation overwriting req->flags
Date: Fri, 14 Jun 2024 01:04:29 +0100 [thread overview]
Message-ID: <6827b129f8f0ad76fa9d1f0a773de938b240ffab.1718323430.git.asml.silence@gmail.com> (raw)
Only the current owner of a request is allowed to write into req->flags.
Hence, the cancellation path should never touch it. Add a new field
instead of the flag, move it into the 3rd cache line because it should
always be initialised. poll_refs can move further as polling is an
involved process anyway.
It's a minimal patch, in the future we can and should find a better
place for it and remove now unused REQ_F_CANCEL_SEQ.
Fixes: 521223d7c229f ("io_uring/cancel: don't default to setting req->work.cancel_seq")
Cc: <[email protected]>
Reported-by: Li Shi <[email protected]>
Signed-off-by: Pavel Begunkov <[email protected]>
---
include/linux/io_uring_types.h | 3 ++-
io_uring/cancel.h | 4 ++--
io_uring/io_uring.c | 1 +
3 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/include/linux/io_uring_types.h b/include/linux/io_uring_types.h
index ac333ea81d31..327d7f43c1fb 100644
--- a/include/linux/io_uring_types.h
+++ b/include/linux/io_uring_types.h
@@ -653,7 +653,7 @@ struct io_kiocb {
struct io_rsrc_node *rsrc_node;
atomic_t refs;
- atomic_t poll_refs;
+ bool cancel_seq_set;
struct io_task_work io_task_work;
/* for polled requests, i.e. IORING_OP_POLL_ADD and async armed poll */
struct hlist_node hash_node;
@@ -662,6 +662,7 @@ struct io_kiocb {
/* opcode allocated if it needs to store data for async defer */
void *async_data;
/* linked requests, IFF REQ_F_HARDLINK or REQ_F_LINK are set */
+ atomic_t poll_refs;
struct io_kiocb *link;
/* custom credentials, valid IFF REQ_F_CREDS is set */
const struct cred *creds;
diff --git a/io_uring/cancel.h b/io_uring/cancel.h
index 76b32e65c03c..b33995e00ba9 100644
--- a/io_uring/cancel.h
+++ b/io_uring/cancel.h
@@ -27,10 +27,10 @@ bool io_cancel_req_match(struct io_kiocb *req, struct io_cancel_data *cd);
static inline bool io_cancel_match_sequence(struct io_kiocb *req, int sequence)
{
- if ((req->flags & REQ_F_CANCEL_SEQ) && sequence == req->work.cancel_seq)
+ if (req->cancel_seq_set && sequence == req->work.cancel_seq)
return true;
- req->flags |= REQ_F_CANCEL_SEQ;
+ req->cancel_seq_set = true;
req->work.cancel_seq = sequence;
return false;
}
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index c170a2b8d2cf..8a216b1d6dc1 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -2211,6 +2211,7 @@ static int io_init_req(struct io_ring_ctx *ctx, struct io_kiocb *req,
req->file = NULL;
req->rsrc_node = NULL;
req->task = current;
+ req->cancel_seq_set = false;
if (unlikely(opcode >= IORING_OP_LAST)) {
req->opcode = 0;
--
2.44.0
next reply other threads:[~2024-06-14 0:04 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-14 0:04 Pavel Begunkov [this message]
2024-06-14 12:14 ` [PATCH] io_uring: fix cancellation overwriting req->flags 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=6827b129f8f0ad76fa9d1f0a773de938b240ffab.1718323430.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