io_uring: fix failed linkchain code logic

Given a linkchain like this:
req0(link_flag)-->req1(link_flag)-->...-->reqn(no link_flag)

There is a problem:
 - if some intermediate linked req like req1 's submittion fails, reqs
   after it won't be cancelled.

   - sqpoll disabled: maybe it's ok since users can get the error info
     of req1 and stop submitting the following sqes.

   - sqpoll enabled: definitely a problem, the following sqes will be
     submitted in the next round.

The solution is to refactor the code logic to:
 - if a linked req's submittion fails, just mark it and the head(if it
   exists) as REQ_F_FAIL. Leverage req->result to indicate whether it
   is failed or cancelled.
 - submit or fail the whole chain when we come to the end of it.

Signed-off-by: Hao Xu <haoxu@linux.alibaba.com>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/20210827094609.36052-3-haoxu@linux.alibaba.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

fs/io_uring.c

@@ -1188,6 +1188,12 @@ static inline void req_set_fail(struct io_kiocb *req)
 	req->flags |= REQ_F_FAIL;
 }
 
+static inline void req_fail_link_node(struct io_kiocb *req, int res)
+{
+	req_set_fail(req);
+	req->result = res;
+}
+
 static void io_ring_ctx_ref_free(struct percpu_ref *ref)
 {
 	struct io_ring_ctx *ctx = container_of(ref, struct io_ring_ctx, refs);
@@ -1957,11 +1963,16 @@ static void io_fail_links(struct io_kiocb *req)
 
 	req->link = NULL;
 	while (link) {
+		long res = -ECANCELED;
+
+		if (link->flags & REQ_F_FAIL)
+			res = link->result;
+
 		nxt = link->link;
 		link->link = NULL;
 
 		trace_io_uring_fail_link(req, link);
-		io_cqring_fill_event(link->ctx, link->user_data, -ECANCELED, 0);
+		io_cqring_fill_event(link->ctx, link->user_data, res, 0);
 		io_put_req_deferred(link);
 		link = nxt;
 	}
@@ -6622,8 +6633,10 @@ static inline void io_queue_sqe(struct io_kiocb *req)
 	if (unlikely(req->ctx->drain_active) && io_drain_req(req))
 		return;
 
-	if (likely(!(req->flags & REQ_F_FORCE_ASYNC))) {
+	if (likely(!(req->flags & (REQ_F_FORCE_ASYNC | REQ_F_FAIL)))) {
 		__io_queue_sqe(req);
+	} else if (req->flags & REQ_F_FAIL) {
+		io_req_complete_failed(req, req->result);
 	} else {
 		int ret = io_req_prep_async(req);
 
@@ -6732,18 +6745,33 @@ static int io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
 	ret = io_init_req(ctx, req, sqe);
 	if (unlikely(ret)) {
 fail_req:
-		if (link->head) {
 		/* fail even hard links since we don't submit */
-			io_req_complete_failed(link->head, -ECANCELED);
+		if (link->head) {
-			link->head = NULL;
+			/*
-		}
+			 * we can judge a link req is failed or cancelled by if
+			 * REQ_F_FAIL is set, but the head is an exception since
+			 * it may be set REQ_F_FAIL because of other req's failure
+			 * so let's leverage req->result to distinguish if a head
+			 * is set REQ_F_FAIL because of its failure or other req's
+			 * failure so that we can set the correct ret code for it.
+			 * init result here to avoid affecting the normal path.
+			 */
+			if (!(link->head->flags & REQ_F_FAIL))
+				req_fail_link_node(link->head, -ECANCELED);
+		} else if (!(req->flags & (REQ_F_LINK | REQ_F_HARDLINK))) {
+			/*
+			 * the current req is a normal req, we should return
+			 * error and thus break the submittion loop.
+			 */
 			io_req_complete_failed(req, ret);
 			return ret;
 		}
-
+		req_fail_link_node(req, ret);
+	} else {
 		ret = io_req_prep(req, sqe);
 		if (unlikely(ret))
 			goto fail_req;
+	}
 
 	/* don't need @sqe from now on */
 	trace_io_uring_submit_sqe(ctx, req, req->opcode, req->user_data,
@@ -6760,9 +6788,14 @@ fail_req:
 	if (link->head) {
 		struct io_kiocb *head = link->head;
 
+		if (!(req->flags & REQ_F_FAIL)) {
 			ret = io_req_prep_async(req);
-		if (unlikely(ret))
+			if (unlikely(ret)) {
-			goto fail_req;
+				req_fail_link_node(req, ret);
+				if (!(head->flags & REQ_F_FAIL))
+					req_fail_link_node(head, -ECANCELED);
+			}
+		}
 		trace_io_uring_link(ctx, req, head);
 		link->last->link = req;
 		link->last = req;