From f28c240e7152462f0750a8939db28d985ecf7c67 Mon Sep 17 00:00:00 2001
From: Hao Xu <haoxu@linux.alibaba.com>
Date: Wed, 8 Dec 2021 13:21:25 +0800
Subject: [PATCH] io_uring: batch completion in prior_task_list
Git-commit: f28c240e7152462f0750a8939db28d985ecf7c67
Patch-mainline: v5.17-rc1
References: bsc#1205205

In previous patches, we have already gathered some tw with
io_req_task_complete() as callback in prior_task_list, let's complete
them in batch while we cannot grab uring lock. In this way, we batch
the req_complete_post path.

Signed-off-by: Hao Xu <haoxu@linux.alibaba.com>
Link: https://lore.kernel.org/r/20211208052125.351587-1-haoxu@linux.alibaba.com
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Gabriel Krisman Bertazi <krisman@suse.de>
---
 fs/io_uring.c |   71 +++++++++++++++++++++++++++++++++++++++++++++++++---------
 1 file changed, 60 insertions(+), 11 deletions(-)

--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -2230,7 +2230,49 @@ static void ctx_flush_and_put(struct io_
 	percpu_ref_put(&ctx->refs);
 }
 
-static void handle_tw_list(struct io_wq_work_node *node, struct io_ring_ctx **ctx, bool *locked)
+static inline void ctx_commit_and_unlock(struct io_ring_ctx *ctx)
+{
+	io_commit_cqring(ctx);
+	spin_unlock(&ctx->completion_lock);
+	io_cqring_ev_posted(ctx);
+}
+
+static void handle_prev_tw_list(struct io_wq_work_node *node,
+				struct io_ring_ctx **ctx, bool *uring_locked)
+{
+	if (*ctx && !*uring_locked)
+		spin_lock(&(*ctx)->completion_lock);
+
+	do {
+		struct io_wq_work_node *next = node->next;
+		struct io_kiocb *req = container_of(node, struct io_kiocb,
+						    io_task_work.node);
+
+		if (req->ctx != *ctx) {
+			if (unlikely(!*uring_locked && *ctx))
+				ctx_commit_and_unlock(*ctx);
+
+			ctx_flush_and_put(*ctx, uring_locked);
+			*ctx = req->ctx;
+			/* if not contended, grab and improve batching */
+			*uring_locked = mutex_trylock(&(*ctx)->uring_lock);
+			percpu_ref_get(&(*ctx)->refs);
+			if (unlikely(!*uring_locked))
+				spin_lock(&(*ctx)->completion_lock);
+		}
+		if (likely(*uring_locked))
+			req->io_task_work.func(req, uring_locked);
+		else
+			__io_req_complete_post(req, req->result, io_put_kbuf(req));
+		node = next;
+	} while (node);
+
+	if (unlikely(!*uring_locked))
+		ctx_commit_and_unlock(*ctx);
+}
+
+static void handle_tw_list(struct io_wq_work_node *node,
+			   struct io_ring_ctx **ctx, bool *locked)
 {
 	do {
 		struct io_wq_work_node *next = node->next;
@@ -2251,31 +2293,38 @@ static void handle_tw_list(struct io_wq_
 
 static void tctx_task_work(struct callback_head *cb)
 {
-	bool locked = false;
+	bool uring_locked = false;
 	struct io_ring_ctx *ctx = NULL;
 	struct io_uring_task *tctx = container_of(cb, struct io_uring_task,
 						  task_work);
 
 	while (1) {
-		struct io_wq_work_node *node;
+		struct io_wq_work_node *node1, *node2;
 
-		if (!tctx->prior_task_list.first &&
-		    !tctx->task_list.first && locked)
+		if (!tctx->task_list.first &&
+		    !tctx->prior_task_list.first && uring_locked)
 			io_submit_flush_completions(ctx);
 
 		spin_lock_irq(&tctx->task_lock);
-		node= wq_list_merge(&tctx->prior_task_list, &tctx->task_list);
-		if (!node)
+		node1 = tctx->prior_task_list.first;
+		node2 = tctx->task_list.first;
+		INIT_WQ_LIST(&tctx->task_list);
+		INIT_WQ_LIST(&tctx->prior_task_list);
+		if (!node2 && !node1)
 			tctx->task_running = false;
 		spin_unlock_irq(&tctx->task_lock);
-		if (!node)
+		if (!node2 && !node1)
 			break;
 
-		handle_tw_list(node, &ctx, &locked);
+		if (node1)
+			handle_prev_tw_list(node1, &ctx, &uring_locked);
+
+		if (node2)
+			handle_tw_list(node2, &ctx, &uring_locked);
 		cond_resched();
 	}
 
-	ctx_flush_and_put(ctx, &locked);
+	ctx_flush_and_put(ctx, &uring_locked);
 }
 
 static void io_req_task_work_add(struct io_kiocb *req, bool priority)
@@ -2764,7 +2813,7 @@ static void io_complete_rw(struct kiocb
 		return;
 	req->result = res;
 	req->io_task_work.func = io_req_task_complete;
-	io_req_task_work_add(req, true);
+	io_req_task_work_add(req, !!(req->ctx->flags & IORING_SETUP_SQPOLL));
 }
 
 static void io_complete_rw_iopoll(struct kiocb *kiocb, long res)