6.12-stable review patch. If anyone has any objections, please let me know.
------------------
From: Pavel Begunkov asml.silence@gmail.com
[ Upstream commit 5e16f1a68d28965c12b6fa227a306fef8a680f84 ]
io_req_post_cqe() sets submit_state.cq_flush so that *flush_completions() can take care of batch commiting CQEs. Don't commit it twice by using __io_cq_unlock_post().
Signed-off-by: Pavel Begunkov asml.silence@gmail.com Link: https://lore.kernel.org/r/41c416660c509cee676b6cad96081274bcb459f3.174549386... Signed-off-by: Jens Axboe axboe@kernel.dk Signed-off-by: Sasha Levin sashal@kernel.org --- io_uring/io_uring.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-)
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c index 8ef0603c07f11..985c87ea09a90 100644 --- a/io_uring/io_uring.c +++ b/io_uring/io_uring.c @@ -874,10 +874,15 @@ bool io_req_post_cqe(struct io_kiocb *req, s32 res, u32 cflags) lockdep_assert(!io_wq_current_is_worker()); lockdep_assert_held(&ctx->uring_lock);
- __io_cq_lock(ctx); - posted = io_fill_cqe_aux(ctx, req->cqe.user_data, res, cflags); + if (!ctx->lockless_cq) { + spin_lock(&ctx->completion_lock); + posted = io_fill_cqe_aux(ctx, req->cqe.user_data, res, cflags); + spin_unlock(&ctx->completion_lock); + } else { + posted = io_fill_cqe_aux(ctx, req->cqe.user_data, res, cflags); + } + ctx->submit_state.cq_flush = true; - __io_cq_unlock_post(ctx); return posted; }