[PATCH 09/11] io_uring: remove struct io_tw_state::locked

[Pavel Begunkov <asml.silence@gmail.com>]

ctx is always locked for task_work now, so get rid of struct
io_tw_state::locked. Note I'm stopping one step before removing
io_tw_state altogether, which is not empty, because it still serves the
purpose of indicating which function is a tw callback and forcing users
not to invoke them carelessly out of a wrong context. The removal can
always be done later.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/linux/io_uring_types.h |  2 --
 io_uring/io_uring.c            | 31 ++++++++-----------------------
 io_uring/io_uring.h            |  5 +----
 io_uring/poll.c                |  2 +-
 io_uring/rw.c                  |  6 ++----
 io_uring/timeout.c             |  8 ++------
 io_uring/uring_cmd.c           |  8 ++------
 io_uring/waitid.c              |  2 +-
 8 files changed, 17 insertions(+), 47 deletions(-)

diff --git a/include/linux/io_uring_types.h b/include/linux/io_uring_types.h
index e24893625085..5a2afbc93887 100644
--- a/include/linux/io_uring_types.h
+++ b/include/linux/io_uring_types.h
@@ -439,8 +439,6 @@ struct io_ring_ctx {
 };
 
 struct io_tw_state {
-	/* ->uring_lock is taken, callbacks can use io_tw_lock to lock it */
-	bool locked;
 };
 
 enum {
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index 0cef5c4ddc98..1c4bbfc411d1 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -252,14 +252,12 @@ static __cold void io_fallback_req_func(struct work_struct *work)
 						fallback_work.work);
 	struct llist_node *node = llist_del_all(&ctx->fallback_llist);
 	struct io_kiocb *req, *tmp;
-	struct io_tw_state ts = { .locked = true, };
+	struct io_tw_state ts = {};
 
 	percpu_ref_get(&ctx->refs);
 	mutex_lock(&ctx->uring_lock);
 	llist_for_each_entry_safe(req, tmp, node, io_task_work.node)
 		req->io_task_work.func(req, &ts);
-	if (WARN_ON_ONCE(!ts.locked))
-		return;
 	io_submit_flush_completions(ctx);
 	mutex_unlock(&ctx->uring_lock);
 	percpu_ref_put(&ctx->refs);
@@ -1163,11 +1161,9 @@ static void ctx_flush_and_put(struct io_ring_ctx *ctx, struct io_tw_state *ts)
 		return;
 	if (ctx->flags & IORING_SETUP_TASKRUN_FLAG)
 		atomic_andnot(IORING_SQ_TASKRUN, &ctx->rings->sq_flags);
-	if (ts->locked) {
-		io_submit_flush_completions(ctx);
-		mutex_unlock(&ctx->uring_lock);
-		ts->locked = false;
-	}
+
+	io_submit_flush_completions(ctx);
+	mutex_unlock(&ctx->uring_lock);
 	percpu_ref_put(&ctx->refs);
 }
 
@@ -1191,8 +1187,6 @@ struct llist_node *io_handle_tw_list(struct llist_node *node,
 		if (req->ctx != ctx) {
 			ctx_flush_and_put(ctx, &ts);
 			ctx = req->ctx;
-
-			ts.locked = true;
 			mutex_lock(&ctx->uring_lock);
 			percpu_ref_get(&ctx->refs);
 		}
@@ -1465,22 +1459,16 @@ static int __io_run_local_work(struct io_ring_ctx *ctx, struct io_tw_state *ts,
 static inline int io_run_local_work_locked(struct io_ring_ctx *ctx,
 					   int min_events)
 {
-	struct io_tw_state ts = { .locked = true, };
-	int ret;
+	struct io_tw_state ts = {};
 
 	if (llist_empty(&ctx->work_llist))
 		return 0;
-
-	ret = __io_run_local_work(ctx, &ts, min_events);
-	/* shouldn't happen! */
-	if (WARN_ON_ONCE(!ts.locked))
-		mutex_lock(&ctx->uring_lock);
-	return ret;
+	return __io_run_local_work(ctx, &ts, min_events);
 }
 
 static int io_run_local_work(struct io_ring_ctx *ctx, int min_events)
 {
-	struct io_tw_state ts = { .locked = true };
+	struct io_tw_state ts = {};
 	int ret;
 
 	mutex_lock(&ctx->uring_lock);
@@ -1708,10 +1696,7 @@ static int io_iopoll_check(struct io_ring_ctx *ctx, long min)
 
 void io_req_task_complete(struct io_kiocb *req, struct io_tw_state *ts)
 {
-	if (ts->locked)
-		io_req_complete_defer(req);
-	else
-		io_req_complete_post(req, IO_URING_F_UNLOCKED);
+	io_req_complete_defer(req);
 }
 
 /*
diff --git a/io_uring/io_uring.h b/io_uring/io_uring.h
index 472ba5692ba8..6cad3ef3408b 100644
--- a/io_uring/io_uring.h
+++ b/io_uring/io_uring.h
@@ -344,10 +344,7 @@ static inline bool io_task_work_pending(struct io_ring_ctx *ctx)
 
 static inline void io_tw_lock(struct io_ring_ctx *ctx, struct io_tw_state *ts)
 {
-	if (!ts->locked) {
-		mutex_lock(&ctx->uring_lock);
-		ts->locked = true;
-	}
+	lockdep_assert_held(&ctx->uring_lock);
 }
 
 /*
diff --git a/io_uring/poll.c b/io_uring/poll.c
index 6db1dcb2c797..8901dd118e50 100644
--- a/io_uring/poll.c
+++ b/io_uring/poll.c
@@ -322,7 +322,7 @@ static int io_poll_check_events(struct io_kiocb *req, struct io_tw_state *ts)
 			__poll_t mask = mangle_poll(req->cqe.res &
 						    req->apoll_events);
 
-			if (!io_fill_cqe_req_aux(req, ts->locked, mask,
+			if (!io_fill_cqe_req_aux(req, true, mask,
 						 IORING_CQE_F_MORE)) {
 				io_req_set_res(req, mask, 0);
 				return IOU_POLL_REMOVE_POLL_USE_RES;
diff --git a/io_uring/rw.c b/io_uring/rw.c
index 576934dbf833..c7f9246ff508 100644
--- a/io_uring/rw.c
+++ b/io_uring/rw.c
@@ -305,11 +305,9 @@ void io_req_rw_complete(struct io_kiocb *req, struct io_tw_state *ts)
 
 	io_req_io_end(req);
 
-	if (req->flags & (REQ_F_BUFFER_SELECTED|REQ_F_BUFFER_RING)) {
-		unsigned issue_flags = ts->locked ? 0 : IO_URING_F_UNLOCKED;
+	if (req->flags & (REQ_F_BUFFER_SELECTED|REQ_F_BUFFER_RING))
+		req->cqe.flags |= io_put_kbuf(req, 0);
 
-		req->cqe.flags |= io_put_kbuf(req, issue_flags);
-	}
 	io_req_task_complete(req, ts);
 }
 
diff --git a/io_uring/timeout.c b/io_uring/timeout.c
index 7fd7dbb211d6..0a48e6acd0b2 100644
--- a/io_uring/timeout.c
+++ b/io_uring/timeout.c
@@ -72,10 +72,7 @@ static void io_timeout_complete(struct io_kiocb *req, struct io_tw_state *ts)
 	struct io_ring_ctx *ctx = req->ctx;
 
 	if (!io_timeout_finish(timeout, data)) {
-		bool filled;
-		filled = io_fill_cqe_req_aux(req, ts->locked, -ETIME,
-					     IORING_CQE_F_MORE);
-		if (filled) {
+		if (io_fill_cqe_req_aux(req, true, -ETIME, IORING_CQE_F_MORE)) {
 			/* re-arm timer */
 			spin_lock_irq(&ctx->timeout_lock);
 			list_add(&timeout->list, ctx->timeout_list.prev);
@@ -301,7 +298,6 @@ int io_timeout_cancel(struct io_ring_ctx *ctx, struct io_cancel_data *cd)
 
 static void io_req_task_link_timeout(struct io_kiocb *req, struct io_tw_state *ts)
 {
-	unsigned issue_flags = ts->locked ? 0 : IO_URING_F_UNLOCKED;
 	struct io_timeout *timeout = io_kiocb_to_cmd(req, struct io_timeout);
 	struct io_kiocb *prev = timeout->prev;
 	int ret = -ENOENT;
@@ -313,7 +309,7 @@ static void io_req_task_link_timeout(struct io_kiocb *req, struct io_tw_state *t
 				.data		= prev->cqe.user_data,
 			};
 
-			ret = io_try_cancel(req->task->io_uring, &cd, issue_flags);
+			ret = io_try_cancel(req->task->io_uring, &cd, 0);
 		}
 		io_req_set_res(req, ret ?: -ETIME, 0);
 		io_req_task_complete(req, ts);
diff --git a/io_uring/uring_cmd.c b/io_uring/uring_cmd.c
index ec38a8d4836d..e45d4cd5ef82 100644
--- a/io_uring/uring_cmd.c
+++ b/io_uring/uring_cmd.c
@@ -56,13 +56,9 @@ EXPORT_SYMBOL_GPL(io_uring_cmd_mark_cancelable);
 static void io_uring_cmd_work(struct io_kiocb *req, struct io_tw_state *ts)
 {
 	struct io_uring_cmd *ioucmd = io_kiocb_to_cmd(req, struct io_uring_cmd);
-	unsigned issue_flags = IO_URING_F_UNLOCKED;
 
-	/* locked task_work executor checks the deffered list completion */
-	if (ts->locked)
-		issue_flags = IO_URING_F_COMPLETE_DEFER;
-
-	ioucmd->task_work_cb(ioucmd, issue_flags);
+	/* task_work executor checks the deffered list completion */
+	ioucmd->task_work_cb(ioucmd, IO_URING_F_COMPLETE_DEFER);
 }
 
 void __io_uring_cmd_do_in_task(struct io_uring_cmd *ioucmd,
diff --git a/io_uring/waitid.c b/io_uring/waitid.c
index 6f851978606d..791ea752b0b1 100644
--- a/io_uring/waitid.c
+++ b/io_uring/waitid.c
@@ -118,7 +118,7 @@ static int io_waitid_finish(struct io_kiocb *req, int ret)
 static void io_waitid_complete(struct io_kiocb *req, int ret)
 {
 	struct io_waitid *iw = io_kiocb_to_cmd(req, struct io_waitid);
-	struct io_tw_state ts = { .locked = true };
+	struct io_tw_state ts = {};
 
 	/* anyone completing better be holding a reference */
 	WARN_ON_ONCE(!(atomic_read(&iw->refs) & IO_WAITID_REF_MASK));
-- 
2.43.0