Re: FAILED: patch "[PATCH] io_uring: use right issue_flags for splice/tee" failed to apply to 5.17-stable tree

[Jens Axboe <axboe@kernel.dk>]

[-- Attachment #1: Type: text/plain, Size: 386 bytes --]

On 4/18/22 3:54 AM, gregkh@linuxfoundation.org wrote:
> 
> The patch below does not apply to the 5.17-stable tree.
> If someone wants it applied there, or to any other stable or longterm
> tree, then please email the backport, including the original git commit
> id to <stable@vger.kernel.org>.

Here's this one, and the two others that failed to apply for
5.17-stable.

-- 
Jens Axboe

[-- Attachment #2: 0003-io_uring-fix-poll-error-reporting.patch --]
[-- Type: text/x-patch, Size: 1400 bytes --]

From 8f8cabb77826b786c2f5a1854abad465f9fbfef2 Mon Sep 17 00:00:00 2001
From: Pavel Begunkov <asml.silence@gmail.com>
Date: Mon, 18 Apr 2022 06:41:20 -0600
Subject: [PATCH 3/3] io_uring: fix poll error reporting

commit 7179c3ce3dbff646c55f7cd664a895f462f049e5 upstream.

We should not return an error code in req->result in
io_poll_check_events(), because it may get mangled and returned as
success. Just return the error code directly, the callers will fail the
request or proceed accordingly.

Fixes: 6bf9c47a3989 ("io_uring: defer file assignment")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/5f03514ee33324dc811fb93df84aee0f695fb044.1649862516.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 fs/io_uring.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 293b04c7edda..7b4d48cfe44c 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -5509,9 +5509,8 @@ static int io_poll_check_events(struct io_kiocb *req, bool locked)
 			unsigned flags = locked ? 0 : IO_URING_F_UNLOCKED;
 
 			if (unlikely(!io_assign_file(req, flags)))
-				req->result = -EBADF;
-			else
-				req->result = vfs_poll(req->file, &pt) & poll->events;
+				return -EBADF;
+			req->result = vfs_poll(req->file, &pt) & poll->events;
 		}
 
 		/* multishot, just fill an CQE and proceed */
-- 
2.35.1


[-- Attachment #3: 0002-io_uring-fix-poll-file-assign-deadlock.patch --]
[-- Type: text/x-patch, Size: 1291 bytes --]

From cbb3f5938cdf975f2970f9b5674a56e6fa1a59ff Mon Sep 17 00:00:00 2001
From: Pavel Begunkov <asml.silence@gmail.com>
Date: Mon, 18 Apr 2022 06:40:30 -0600
Subject: [PATCH 2/3] io_uring: fix poll file assign deadlock

commit cce64ef01308b677a687d90927fc2b2e0e1cba67 upstream.

We pass "unlocked" into io_assign_file() in io_poll_check_events(),
which can lead to double locking.

Fixes: 6bf9c47a3989 ("io_uring: defer file assignment")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/2476d4ae46554324b599ee4055447b105f20a75a.1649862516.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 fs/io_uring.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 26832d166ca2..293b04c7edda 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -5506,8 +5506,9 @@ static int io_poll_check_events(struct io_kiocb *req, bool locked)
 
 		if (!req->result) {
 			struct poll_table_struct pt = { ._key = poll->events };
+			unsigned flags = locked ? 0 : IO_URING_F_UNLOCKED;
 
-			if (unlikely(!io_assign_file(req, IO_URING_F_UNLOCKED)))
+			if (unlikely(!io_assign_file(req, flags)))
 				req->result = -EBADF;
 			else
 				req->result = vfs_poll(req->file, &pt) & poll->events;
-- 
2.35.1


[-- Attachment #4: 0001-io_uring-use-right-issue_flags-for-splice-tee.patch --]
[-- Type: text/x-patch, Size: 1624 bytes --]

From d5f62d8199e00d5af1a57f83ca1fcd49fecd3a36 Mon Sep 17 00:00:00 2001
From: Pavel Begunkov <asml.silence@gmail.com>
Date: Wed, 13 Apr 2022 16:10:33 +0100
Subject: [PATCH 1/3] io_uring: use right issue_flags for splice/tee

commit e941976659f1f6834077a1596bf53e6bdb10e90b upstream.

Pass right issue_flags into into io_file_get_fixed() instead of
IO_URING_F_UNLOCKED. It's probably not a problem at the moment but let's
do it safer.

Fixes: 6bf9c47a3989 ("io_uring: defer file assignment")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/7d242daa9df5d776907686977cd29fbceb4a2d8d.1649862516.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 fs/io_uring.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 1a9630dc5361..26832d166ca2 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -4136,7 +4136,7 @@ static int io_tee(struct io_kiocb *req, unsigned int issue_flags)
 		return -EAGAIN;
 
 	if (sp->flags & SPLICE_F_FD_IN_FIXED)
-		in = io_file_get_fixed(req, sp->splice_fd_in, IO_URING_F_UNLOCKED);
+		in = io_file_get_fixed(req, sp->splice_fd_in, issue_flags);
 	else
 		in = io_file_get_normal(req, sp->splice_fd_in);
 	if (!in) {
@@ -4178,7 +4178,7 @@ static int io_splice(struct io_kiocb *req, unsigned int issue_flags)
 		return -EAGAIN;
 
 	if (sp->flags & SPLICE_F_FD_IN_FIXED)
-		in = io_file_get_fixed(req, sp->splice_fd_in, IO_URING_F_UNLOCKED);
+		in = io_file_get_fixed(req, sp->splice_fd_in, issue_flags);
 	else
 		in = io_file_get_normal(req, sp->splice_fd_in);
 	if (!in) {
-- 
2.35.1