syzbot reports that KMSAN complains that 'nr_tw' is an uninit-value
with the following report:

BUG: KMSAN: uninit-value in io_req_local_work_add io_uring/io_uring.c:1192 [inline]
BUG: KMSAN: uninit-value in io_req_task_work_add_remote+0x588/0x5d0 io_uring/io_uring.c:1240
 io_req_local_work_add io_uring/io_uring.c:1192 [inline]
 io_req_task_work_add_remote+0x588/0x5d0 io_uring/io_uring.c:1240
 io_msg_remote_post io_uring/msg_ring.c:102 [inline]
 io_msg_data_remote io_uring/msg_ring.c:133 [inline]
 io_msg_ring_data io_uring/msg_ring.c:152 [inline]
 io_msg_ring+0x1c38/0x1ef0 io_uring/msg_ring.c:305
 io_issue_sqe+0x383/0x22c0 io_uring/io_uring.c:1710
 io_queue_sqe io_uring/io_uring.c:1924 [inline]
 io_submit_sqe io_uring/io_uring.c:2180 [inline]
 io_submit_sqes+0x1259/0x2f20 io_uring/io_uring.c:2295
 __do_sys_io_uring_enter io_uring/io_uring.c:3205 [inline]
 __se_sys_io_uring_enter+0x40c/0x3ca0 io_uring/io_uring.c:3142
 __x64_sys_io_uring_enter+0x11f/0x1a0 io_uring/io_uring.c:3142
 x64_sys_call+0x2d82/0x3c10 arch/x86/include/generated/asm/syscalls_64.h:427
 do_syscall_x64 arch/x86/entry/common.c:52 [inline]
 do_syscall_64+0xcd/0x1e0 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

which is the following check:

if (nr_tw < nr_wait)
	return;

in io_req_local_work_add(). While nr_tw itself cannot be uninitialized,
it does depend on req->flags, which off the msg ring issue path can
indeed be uninitialized.

Fix this by always clearing the allocated 'req' fully if we can't grab
one from the cache itself.

Fixes: 50cf5f3842af ("io_uring/msg_ring: add an alloc cache for io_kiocb entries")
Reported-by: syzbot+82609b8937a4458106ca@syzkaller.appspotmail.com
Link: https://lore.kernel.org/io-uring/000000000000fd3d8d061dfc0e4a@google.com/
Signed-off-by: Jens Axboe <axboe@kernel.dk>

The change adding caching around the request allocated and freed for
data messages changed a kmem_cache_free() to a kfree(), which isn't
correct as the request came from slab in the first place. Fix that up
and use the right freeing function if the cache is already at its limit.

Note that the current mixing of kmem_cache_alloc and kfree is fine, but
consistent alloc/free functions should be used as it's otherwise somewhat
confusing.

Fixes: 50cf5f3842af ("io_uring/msg_ring: add an alloc cache for io_kiocb entries")
Signed-off-by: Jens Axboe <axboe@kernel.dk>

The change for improving the handling of the target CQE posting
inadvertently dropped the NULL check for the submitter task on the target
ring, reinstate that.

Fixes: 0617bb500bfa ("io_uring/msg_ring: improve handling of target CQE posting")
Reported-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

With slab accounting, allocating and freeing memory has considerable
overhead. Add a basic alloc cache for the io_kiocb allocations that
msg_ring needs to do. Unlike other caches, this one is used by the
sender, grabbing it from the remote ring. When the remote ring gets
the posted completion, it'll free it locally. Hence it is separately
locked, using ctx->msg_lock.

Signed-off-by: Jens Axboe <axboe@kernel.dk>

Use the exported helper for queueing task_work for message passing,
rather than rolling our own.

Note that this is only done for strict data messages for now, file
descriptor passing messages still rely on the kernel task_work. It could
get converted at some point if it's performance critical.

This improves peak performance of message passing by about 5x in some
basic testing, with 2 threads just sending messages to each other.
Before this change, it was capped at around 700K/sec, with the change
it's at over 4M/sec.

Signed-off-by: Jens Axboe <axboe@kernel.dk>

Currently this is gated on whether or not the target ring needs a local
completion - and if so, whether or not we're running on the right task.
The use case for same thread cross posting is probably a lot less
relevant than remote posting. And since we're going to improve this
situation anyway, just gate it on local posting and ignore what task
we're currently running on.

Signed-off-by: Jens Axboe <axboe@kernel.dk>

Move the posting outside the checking and locking, it's cleaner that
way.

Signed-off-by: Jens Axboe <axboe@kernel.dk>

In io_msg_exec_remote(), ctx->submitter_task is read using READ_ONCE at
the beginning of the function, checked, and then re-read from
ctx->submitter_task, voiding all guarantees of the checks. Reuse the value
that was read by READ_ONCE to ensure the consistency of the task struct
throughout the function.

Signed-off-by: linke li <lilinke99@qq.com>
Link: https://lore.kernel.org/r/tencent_F9B2296C93928D6F68FF0C95C33475C68209@qq.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

Use io_file_from_index instead of open coding it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20230620113235.920399-8-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

msg_ring requests transferring files support auto index selection via
IORING_FILE_INDEX_ALLOC, however they don't return the selected index
to the target ring and there is no other good way for the userspace to
know where is the receieved file.

Return the index for allocated slots and 0 otherwise, which is
consistent with other fixed file installing requests.

Cc: stable@vger.kernel.org # v6.0+
Fixes: e6130eba8a848 ("io_uring: add support for passing fixed file descriptors")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://github.com/axboe/liburing/issues/809
Signed-off-by: Jens Axboe <axboe@kernel.dk>