IO-Uring Archive on lore.kernel.org
 help / color / Atom feed
* [PATCH v2] io_uring: reduced function parameter ctx if possible
@ 2019-11-08  7:29 Jackie Liu
  2019-11-08  9:00 ` Jackie Liu
  2019-11-08 14:11 ` Jens Axboe
  0 siblings, 2 replies; 9+ messages in thread
From: Jackie Liu @ 2019-11-08  7:29 UTC (permalink / raw)
  To: axboe; +Cc: io-uring, liuyun01

Many times, the core of the function is req, and req has already set
req->ctx at initialization time, so there is no need to pass in from
outside.

Cleanup, no function change.

Signed-off-by: Jackie Liu <liuyun01@kylinos.cn>
---
V2:
 - rebase to branch for-5.5/io_uring: 2665abfd757fb35a241c6f0b1ebf620e3ffb36fb

 fs/io_uring.c | 109 +++++++++++++++++++++++++++++++---------------------------
 1 file changed, 58 insertions(+), 51 deletions(-)

diff --git a/fs/io_uring.c b/fs/io_uring.c
index eadd19a..2cc53e3 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -429,20 +429,20 @@ static struct io_ring_ctx *io_ring_ctx_alloc(struct io_uring_params *p)
 	return ctx;
 }
 
-static inline bool __io_sequence_defer(struct io_ring_ctx *ctx,
-				       struct io_kiocb *req)
+static inline bool __io_sequence_defer(struct io_kiocb *req)
 {
+	struct io_ring_ctx *ctx = req->ctx;
+
 	return req->sequence != ctx->cached_cq_tail + ctx->cached_sq_dropped
 					+ atomic_read(&ctx->cached_cq_overflow);
 }
 
-static inline bool io_sequence_defer(struct io_ring_ctx *ctx,
-				     struct io_kiocb *req)
+static inline bool io_sequence_defer(struct io_kiocb *req)
 {
 	if ((req->flags & (REQ_F_IO_DRAIN|REQ_F_IO_DRAINED)) != REQ_F_IO_DRAIN)
 		return false;
 
-	return __io_sequence_defer(ctx, req);
+	return __io_sequence_defer(req);
 }
 
 static struct io_kiocb *io_get_deferred_req(struct io_ring_ctx *ctx)
@@ -450,7 +450,7 @@ static struct io_kiocb *io_get_deferred_req(struct io_ring_ctx *ctx)
 	struct io_kiocb *req;
 
 	req = list_first_entry_or_null(&ctx->defer_list, struct io_kiocb, list);
-	if (req && !io_sequence_defer(ctx, req)) {
+	if (req && !io_sequence_defer(req)) {
 		list_del_init(&req->list);
 		return req;
 	}
@@ -463,7 +463,7 @@ static struct io_kiocb *io_get_timeout_req(struct io_ring_ctx *ctx)
 	struct io_kiocb *req;
 
 	req = list_first_entry_or_null(&ctx->timeout_list, struct io_kiocb, list);
-	if (req && !__io_sequence_defer(ctx, req)) {
+	if (req && !__io_sequence_defer(req)) {
 		list_del_init(&req->list);
 		return req;
 	}
@@ -512,10 +512,10 @@ static inline bool io_prep_async_work(struct io_kiocb *req)
 	return do_hashed;
 }
 
-static inline void io_queue_async_work(struct io_ring_ctx *ctx,
-				       struct io_kiocb *req)
+static inline void io_queue_async_work(struct io_kiocb *req)
 {
 	bool do_hashed = io_prep_async_work(req);
+	struct io_ring_ctx *ctx = req->ctx;
 
 	trace_io_uring_queue_async_work(ctx, do_hashed, req, &req->work,
 					req->flags);
@@ -566,7 +566,7 @@ static void io_commit_cqring(struct io_ring_ctx *ctx)
 			continue;
 		}
 		req->flags |= REQ_F_IO_DRAINED;
-		io_queue_async_work(ctx, req);
+		io_queue_async_work(req);
 	}
 }
 
@@ -714,9 +714,9 @@ static void __io_free_req(struct io_kiocb *req)
 	kmem_cache_free(req_cachep, req);
 }
 
-static bool io_link_cancel_timeout(struct io_ring_ctx *ctx,
-				   struct io_kiocb *req)
+static bool io_link_cancel_timeout(struct io_kiocb *req)
 {
+	struct io_ring_ctx *ctx = req->ctx;
 	int ret;
 
 	ret = hrtimer_try_to_cancel(&req->timeout.timer);
@@ -756,7 +756,7 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr)
 		 * in this context instead of having to queue up new async work.
 		 */
 		if (req->flags & REQ_F_LINK_TIMEOUT) {
-			wake_ev = io_link_cancel_timeout(ctx, nxt);
+			wake_ev = io_link_cancel_timeout(nxt);
 
 			/* we dropped this link, get next */
 			nxt = list_first_entry_or_null(&req->link_list,
@@ -765,7 +765,7 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr)
 			*nxtptr = nxt;
 			break;
 		} else {
-			io_queue_async_work(req->ctx, nxt);
+			io_queue_async_work(nxt);
 			break;
 		}
 	}
@@ -793,7 +793,7 @@ static void io_fail_links(struct io_kiocb *req)
 
 		if ((req->flags & REQ_F_LINK_TIMEOUT) &&
 		    link->submit.sqe->opcode == IORING_OP_LINK_TIMEOUT) {
-			io_link_cancel_timeout(ctx, link);
+			io_link_cancel_timeout(link);
 		} else {
 			io_cqring_fill_event(ctx, link->user_data, -ECANCELED);
 			__io_free_req(link);
@@ -862,7 +862,7 @@ static void io_put_req(struct io_kiocb *req, struct io_kiocb **nxtptr)
 		if (nxtptr)
 			*nxtptr = nxt;
 		else
-			io_queue_async_work(nxt->ctx, nxt);
+			io_queue_async_work(nxt);
 	}
 }
 
@@ -1803,7 +1803,7 @@ static void io_poll_remove_one(struct io_kiocb *req)
 	WRITE_ONCE(poll->canceled, true);
 	if (!list_empty(&poll->wait.entry)) {
 		list_del_init(&poll->wait.entry);
-		io_queue_async_work(req->ctx, req);
+		io_queue_async_work(req);
 	}
 	spin_unlock(&poll->head->lock);
 
@@ -1855,9 +1855,10 @@ static int io_poll_remove(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 	return 0;
 }
 
-static void io_poll_complete(struct io_ring_ctx *ctx, struct io_kiocb *req,
-			     __poll_t mask)
+static void io_poll_complete(struct io_kiocb *req, __poll_t mask)
 {
+	struct io_ring_ctx *ctx = req->ctx;
+
 	req->poll.done = true;
 	io_cqring_fill_event(ctx, req->user_data, mangle_poll(mask));
 	io_commit_cqring(ctx);
@@ -1893,7 +1894,7 @@ static void io_poll_complete_work(struct io_wq_work **workptr)
 		return;
 	}
 	list_del_init(&req->list);
-	io_poll_complete(ctx, req, mask);
+	io_poll_complete(req, mask);
 	spin_unlock_irq(&ctx->completion_lock);
 
 	io_cqring_ev_posted(ctx);
@@ -1921,13 +1922,13 @@ static int io_poll_wake(struct wait_queue_entry *wait, unsigned mode, int sync,
 
 	if (mask && spin_trylock_irqsave(&ctx->completion_lock, flags)) {
 		list_del(&req->list);
-		io_poll_complete(ctx, req, mask);
+		io_poll_complete(req, mask);
 		spin_unlock_irqrestore(&ctx->completion_lock, flags);
 
 		io_cqring_ev_posted(ctx);
 		io_put_req(req, NULL);
 	} else {
-		io_queue_async_work(ctx, req);
+		io_queue_async_work(req);
 	}
 
 	return 1;
@@ -2012,7 +2013,7 @@ static int io_poll_add(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 	}
 	if (mask) { /* no async, we'd stolen it */
 		ipt.error = 0;
-		io_poll_complete(ctx, req, mask);
+		io_poll_complete(req, mask);
 	}
 	spin_unlock_irq(&ctx->completion_lock);
 
@@ -2259,12 +2260,13 @@ static int io_async_cancel(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 	return 0;
 }
 
-static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req)
+static int io_req_defer(struct io_kiocb *req)
 {
 	const struct io_uring_sqe *sqe = req->submit.sqe;
 	struct io_uring_sqe *sqe_copy;
+	struct io_ring_ctx *ctx = req->ctx;
 
-	if (!io_sequence_defer(ctx, req) && list_empty(&ctx->defer_list))
+	if (!io_sequence_defer(req) && list_empty(&ctx->defer_list))
 		return 0;
 
 	sqe_copy = kmalloc(sizeof(*sqe_copy), GFP_KERNEL);
@@ -2272,7 +2274,7 @@ static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req)
 		return -EAGAIN;
 
 	spin_lock_irq(&ctx->completion_lock);
-	if (!io_sequence_defer(ctx, req) && list_empty(&ctx->defer_list)) {
+	if (!io_sequence_defer(req) && list_empty(&ctx->defer_list)) {
 		spin_unlock_irq(&ctx->completion_lock);
 		kfree(sqe_copy);
 		return 0;
@@ -2287,11 +2289,12 @@ static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req)
 	return -EIOCBQUEUED;
 }
 
-static int __io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
-			   struct io_kiocb **nxt, bool force_nonblock)
+static int __io_submit_sqe(struct io_kiocb *req, struct io_kiocb **nxt,
+			   bool force_nonblock)
 {
 	int ret, opcode;
 	struct sqe_submit *s = &req->submit;
+	struct io_ring_ctx *ctx = req->ctx;
 
 	req->user_data = READ_ONCE(s->sqe->user_data);
 
@@ -2389,7 +2392,7 @@ static void io_wq_submit_work(struct io_wq_work **workptr)
 		s->has_user = (work->flags & IO_WQ_WORK_HAS_MM) != 0;
 		s->in_async = true;
 		do {
-			ret = __io_submit_sqe(ctx, req, &nxt, false);
+			ret = __io_submit_sqe(req, &nxt, false);
 			/*
 			 * We can get EAGAIN for polled IO even though we're
 			 * forcing a sync submission from here, since we can't
@@ -2443,10 +2446,10 @@ static inline struct file *io_file_from_index(struct io_ring_ctx *ctx,
 	return table->files[index & IORING_FILE_TABLE_MASK];
 }
 
-static int io_req_set_file(struct io_ring_ctx *ctx,
-			   struct io_submit_state *state, struct io_kiocb *req)
+static int io_req_set_file(struct io_submit_state *state, struct io_kiocb *req)
 {
 	struct sqe_submit *s = &req->submit;
+	struct io_ring_ctx *ctx = req->ctx;
 	unsigned flags;
 	int fd;
 
@@ -2486,9 +2489,10 @@ static int io_req_set_file(struct io_ring_ctx *ctx,
 	return 0;
 }
 
-static int io_grab_files(struct io_ring_ctx *ctx, struct io_kiocb *req)
+static int io_grab_files(struct io_kiocb *req)
 {
 	int ret = -EBADF;
+	struct io_ring_ctx *ctx = req->ctx;
 
 	rcu_read_lock();
 	spin_lock_irq(&ctx->inflight_lock);
@@ -2604,8 +2608,9 @@ static inline struct io_kiocb *io_get_linked_timeout(struct io_kiocb *req)
 	return NULL;
 }
 
-static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
+static int __io_queue_sqe(struct io_kiocb *req)
 {
+	struct io_ring_ctx *ctx = req->ctx;
 	struct io_kiocb *nxt;
 	int ret;
 
@@ -2616,7 +2621,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
 			goto err;
 	}
 
-	ret = __io_submit_sqe(ctx, req, NULL, true);
+	ret = __io_submit_sqe(req, NULL, true);
 
 	/*
 	 * We async punt it if the file wasn't marked NOWAIT, or if the file
@@ -2631,7 +2636,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
 		if (sqe_copy) {
 			s->sqe = sqe_copy;
 			if (req->work.flags & IO_WQ_WORK_NEEDS_FILES) {
-				ret = io_grab_files(ctx, req);
+				ret = io_grab_files(req);
 				if (ret) {
 					kfree(sqe_copy);
 					goto err;
@@ -2642,7 +2647,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
 			 * Queued up for async execution, worker will release
 			 * submit reference when the iocb is actually submitted.
 			 */
-			io_queue_async_work(ctx, req);
+			io_queue_async_work(req);
 			return 0;
 		}
 	}
@@ -2662,11 +2667,12 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
 	return ret;
 }
 
-static int io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
+static int io_queue_sqe(struct io_kiocb *req)
 {
+	struct io_ring_ctx *ctx = req->ctx;
 	int ret;
 
-	ret = io_req_defer(ctx, req);
+	ret = io_req_defer(req);
 	if (ret) {
 		if (ret != -EIOCBQUEUED) {
 			io_cqring_add_event(ctx, req->submit.sqe->user_data, ret);
@@ -2675,17 +2681,17 @@ static int io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
 		return 0;
 	}
 
-	return __io_queue_sqe(ctx, req);
+	return __io_queue_sqe(req);
 }
 
-static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req,
-			      struct io_kiocb *shadow)
+static int io_queue_link_head(struct io_kiocb *req, struct io_kiocb *shadow)
 {
 	int ret;
 	int need_submit = false;
+	struct io_ring_ctx *ctx = req->ctx;
 
 	if (!shadow)
-		return io_queue_sqe(ctx, req);
+		return io_queue_sqe(req);
 
 	/*
 	 * Mark the first IO in link list as DRAIN, let all the following
@@ -2693,7 +2699,7 @@ static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req,
 	 * list.
 	 */
 	req->flags |= REQ_F_IO_DRAIN;
-	ret = io_req_defer(ctx, req);
+	ret = io_req_defer(req);
 	if (ret) {
 		if (ret != -EIOCBQUEUED) {
 			io_cqring_add_event(ctx, req->submit.sqe->user_data, ret);
@@ -2716,18 +2722,19 @@ static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req,
 	spin_unlock_irq(&ctx->completion_lock);
 
 	if (need_submit)
-		return __io_queue_sqe(ctx, req);
+		return __io_queue_sqe(req);
 
 	return 0;
 }
 
 #define SQE_VALID_FLAGS	(IOSQE_FIXED_FILE|IOSQE_IO_DRAIN|IOSQE_IO_LINK)
 
-static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
-			  struct io_submit_state *state, struct io_kiocb **link)
+static void io_submit_sqe(struct io_kiocb *req, struct io_submit_state *state,
+			  struct io_kiocb **link)
 {
 	struct io_uring_sqe *sqe_copy;
 	struct sqe_submit *s = &req->submit;
+	struct io_ring_ctx *ctx = req->ctx;
 	int ret;
 
 	/* enforce forwards compatibility on users */
@@ -2736,7 +2743,7 @@ static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
 		goto err_req;
 	}
 
-	ret = io_req_set_file(ctx, state, req);
+	ret = io_req_set_file(state, req);
 	if (unlikely(ret)) {
 err_req:
 		io_cqring_add_event(ctx, s->sqe->user_data, ret);
@@ -2775,7 +2782,7 @@ static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
 		ret = -EINVAL;
 		goto err_req;
 	} else {
-		io_queue_sqe(ctx, req);
+		io_queue_sqe(req);
 	}
 }
 
@@ -2919,7 +2926,7 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr,
 		req->submit.needs_fixed_file = async;
 		trace_io_uring_submit_sqe(ctx, req->submit.sqe->user_data,
 					  true, async);
-		io_submit_sqe(ctx, req, statep, &link);
+		io_submit_sqe(req, statep, &link);
 		submitted++;
 
 		/*
@@ -2927,14 +2934,14 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr,
 		 * that's the end of the chain. Submit the previous link.
 		 */
 		if (!(sqe_flags & IOSQE_IO_LINK) && link) {
-			io_queue_link_head(ctx, link, shadow_req);
+			io_queue_link_head(link, shadow_req);
 			link = NULL;
 			shadow_req = NULL;
 		}
 	}
 
 	if (link)
-		io_queue_link_head(ctx, link, shadow_req);
+		io_queue_link_head(link, shadow_req);
 	if (statep)
 		io_submit_state_end(&state);
 
-- 
2.7.4




^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: [PATCH v2] io_uring: reduced function parameter ctx if possible
  2019-11-08  7:29 [PATCH v2] io_uring: reduced function parameter ctx if possible Jackie Liu
@ 2019-11-08  9:00 ` Jackie Liu
  2019-11-08  9:10   ` Bob Liu
  2019-11-08 14:11 ` Jens Axboe
  1 sibling, 1 reply; 9+ messages in thread
From: Jackie Liu @ 2019-11-08  9:00 UTC (permalink / raw)
  To: Jens Axboe; +Cc: io-uring

oops, please ignore this, it's crash with commit '2665abfd757fb35a241c6f0b1ebf620e3ffb36fb'
I need test more, and resend later.

> 2019年11月8日 15:29,Jackie Liu <liuyun01@kylinos.cn> 写道:
> 
> Many times, the core of the function is req, and req has already set
> req->ctx at initialization time, so there is no need to pass in from
> outside.
> 
> Cleanup, no function change.
> 
> Signed-off-by: Jackie Liu <liuyun01@kylinos.cn>
> ---
> V2:
> - rebase to branch for-5.5/io_uring: 2665abfd757fb35a241c6f0b1ebf620e3ffb36fb
> 
> fs/io_uring.c | 109 +++++++++++++++++++++++++++++++---------------------------
> 1 file changed, 58 insertions(+), 51 deletions(-)
> 
> diff --git a/fs/io_uring.c b/fs/io_uring.c
> index eadd19a..2cc53e3 100644
> --- a/fs/io_uring.c
> +++ b/fs/io_uring.c
> @@ -429,20 +429,20 @@ static struct io_ring_ctx *io_ring_ctx_alloc(struct io_uring_params *p)
> 	return ctx;
> }
> 
> -static inline bool __io_sequence_defer(struct io_ring_ctx *ctx,
> -				       struct io_kiocb *req)
> +static inline bool __io_sequence_defer(struct io_kiocb *req)
> {
> +	struct io_ring_ctx *ctx = req->ctx;
> +
> 	return req->sequence != ctx->cached_cq_tail + ctx->cached_sq_dropped
> 					+ atomic_read(&ctx->cached_cq_overflow);
> }
> 
> -static inline bool io_sequence_defer(struct io_ring_ctx *ctx,
> -				     struct io_kiocb *req)
> +static inline bool io_sequence_defer(struct io_kiocb *req)
> {
> 	if ((req->flags & (REQ_F_IO_DRAIN|REQ_F_IO_DRAINED)) != REQ_F_IO_DRAIN)
> 		return false;
> 
> -	return __io_sequence_defer(ctx, req);
> +	return __io_sequence_defer(req);
> }
> 
> static struct io_kiocb *io_get_deferred_req(struct io_ring_ctx *ctx)
> @@ -450,7 +450,7 @@ static struct io_kiocb *io_get_deferred_req(struct io_ring_ctx *ctx)
> 	struct io_kiocb *req;
> 
> 	req = list_first_entry_or_null(&ctx->defer_list, struct io_kiocb, list);
> -	if (req && !io_sequence_defer(ctx, req)) {
> +	if (req && !io_sequence_defer(req)) {
> 		list_del_init(&req->list);
> 		return req;
> 	}
> @@ -463,7 +463,7 @@ static struct io_kiocb *io_get_timeout_req(struct io_ring_ctx *ctx)
> 	struct io_kiocb *req;
> 
> 	req = list_first_entry_or_null(&ctx->timeout_list, struct io_kiocb, list);
> -	if (req && !__io_sequence_defer(ctx, req)) {
> +	if (req && !__io_sequence_defer(req)) {
> 		list_del_init(&req->list);
> 		return req;
> 	}
> @@ -512,10 +512,10 @@ static inline bool io_prep_async_work(struct io_kiocb *req)
> 	return do_hashed;
> }
> 
> -static inline void io_queue_async_work(struct io_ring_ctx *ctx,
> -				       struct io_kiocb *req)
> +static inline void io_queue_async_work(struct io_kiocb *req)
> {
> 	bool do_hashed = io_prep_async_work(req);
> +	struct io_ring_ctx *ctx = req->ctx;
> 
> 	trace_io_uring_queue_async_work(ctx, do_hashed, req, &req->work,
> 					req->flags);
> @@ -566,7 +566,7 @@ static void io_commit_cqring(struct io_ring_ctx *ctx)
> 			continue;
> 		}
> 		req->flags |= REQ_F_IO_DRAINED;
> -		io_queue_async_work(ctx, req);
> +		io_queue_async_work(req);
> 	}
> }
> 
> @@ -714,9 +714,9 @@ static void __io_free_req(struct io_kiocb *req)
> 	kmem_cache_free(req_cachep, req);
> }
> 
> -static bool io_link_cancel_timeout(struct io_ring_ctx *ctx,
> -				   struct io_kiocb *req)
> +static bool io_link_cancel_timeout(struct io_kiocb *req)
> {
> +	struct io_ring_ctx *ctx = req->ctx;
> 	int ret;
> 
> 	ret = hrtimer_try_to_cancel(&req->timeout.timer);
> @@ -756,7 +756,7 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr)
> 		 * in this context instead of having to queue up new async work.
> 		 */
> 		if (req->flags & REQ_F_LINK_TIMEOUT) {
> -			wake_ev = io_link_cancel_timeout(ctx, nxt);
> +			wake_ev = io_link_cancel_timeout(nxt);
> 
> 			/* we dropped this link, get next */
> 			nxt = list_first_entry_or_null(&req->link_list,
> @@ -765,7 +765,7 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr)
> 			*nxtptr = nxt;
> 			break;
> 		} else {
> -			io_queue_async_work(req->ctx, nxt);
> +			io_queue_async_work(nxt);
> 			break;
> 		}
> 	}
> @@ -793,7 +793,7 @@ static void io_fail_links(struct io_kiocb *req)
> 
> 		if ((req->flags & REQ_F_LINK_TIMEOUT) &&
> 		    link->submit.sqe->opcode == IORING_OP_LINK_TIMEOUT) {
> -			io_link_cancel_timeout(ctx, link);
> +			io_link_cancel_timeout(link);
> 		} else {
> 			io_cqring_fill_event(ctx, link->user_data, -ECANCELED);
> 			__io_free_req(link);
> @@ -862,7 +862,7 @@ static void io_put_req(struct io_kiocb *req, struct io_kiocb **nxtptr)
> 		if (nxtptr)
> 			*nxtptr = nxt;
> 		else
> -			io_queue_async_work(nxt->ctx, nxt);
> +			io_queue_async_work(nxt);
> 	}
> }
> 
> @@ -1803,7 +1803,7 @@ static void io_poll_remove_one(struct io_kiocb *req)
> 	WRITE_ONCE(poll->canceled, true);
> 	if (!list_empty(&poll->wait.entry)) {
> 		list_del_init(&poll->wait.entry);
> -		io_queue_async_work(req->ctx, req);
> +		io_queue_async_work(req);
> 	}
> 	spin_unlock(&poll->head->lock);
> 
> @@ -1855,9 +1855,10 @@ static int io_poll_remove(struct io_kiocb *req, const struct io_uring_sqe *sqe)
> 	return 0;
> }
> 
> -static void io_poll_complete(struct io_ring_ctx *ctx, struct io_kiocb *req,
> -			     __poll_t mask)
> +static void io_poll_complete(struct io_kiocb *req, __poll_t mask)
> {
> +	struct io_ring_ctx *ctx = req->ctx;
> +
> 	req->poll.done = true;
> 	io_cqring_fill_event(ctx, req->user_data, mangle_poll(mask));
> 	io_commit_cqring(ctx);
> @@ -1893,7 +1894,7 @@ static void io_poll_complete_work(struct io_wq_work **workptr)
> 		return;
> 	}
> 	list_del_init(&req->list);
> -	io_poll_complete(ctx, req, mask);
> +	io_poll_complete(req, mask);
> 	spin_unlock_irq(&ctx->completion_lock);
> 
> 	io_cqring_ev_posted(ctx);
> @@ -1921,13 +1922,13 @@ static int io_poll_wake(struct wait_queue_entry *wait, unsigned mode, int sync,
> 
> 	if (mask && spin_trylock_irqsave(&ctx->completion_lock, flags)) {
> 		list_del(&req->list);
> -		io_poll_complete(ctx, req, mask);
> +		io_poll_complete(req, mask);
> 		spin_unlock_irqrestore(&ctx->completion_lock, flags);
> 
> 		io_cqring_ev_posted(ctx);
> 		io_put_req(req, NULL);
> 	} else {
> -		io_queue_async_work(ctx, req);
> +		io_queue_async_work(req);
> 	}
> 
> 	return 1;
> @@ -2012,7 +2013,7 @@ static int io_poll_add(struct io_kiocb *req, const struct io_uring_sqe *sqe,
> 	}
> 	if (mask) { /* no async, we'd stolen it */
> 		ipt.error = 0;
> -		io_poll_complete(ctx, req, mask);
> +		io_poll_complete(req, mask);
> 	}
> 	spin_unlock_irq(&ctx->completion_lock);
> 
> @@ -2259,12 +2260,13 @@ static int io_async_cancel(struct io_kiocb *req, const struct io_uring_sqe *sqe,
> 	return 0;
> }
> 
> -static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req)
> +static int io_req_defer(struct io_kiocb *req)
> {
> 	const struct io_uring_sqe *sqe = req->submit.sqe;
> 	struct io_uring_sqe *sqe_copy;
> +	struct io_ring_ctx *ctx = req->ctx;
> 
> -	if (!io_sequence_defer(ctx, req) && list_empty(&ctx->defer_list))
> +	if (!io_sequence_defer(req) && list_empty(&ctx->defer_list))
> 		return 0;
> 
> 	sqe_copy = kmalloc(sizeof(*sqe_copy), GFP_KERNEL);
> @@ -2272,7 +2274,7 @@ static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req)
> 		return -EAGAIN;
> 
> 	spin_lock_irq(&ctx->completion_lock);
> -	if (!io_sequence_defer(ctx, req) && list_empty(&ctx->defer_list)) {
> +	if (!io_sequence_defer(req) && list_empty(&ctx->defer_list)) {
> 		spin_unlock_irq(&ctx->completion_lock);
> 		kfree(sqe_copy);
> 		return 0;
> @@ -2287,11 +2289,12 @@ static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req)
> 	return -EIOCBQUEUED;
> }
> 
> -static int __io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
> -			   struct io_kiocb **nxt, bool force_nonblock)
> +static int __io_submit_sqe(struct io_kiocb *req, struct io_kiocb **nxt,
> +			   bool force_nonblock)
> {
> 	int ret, opcode;
> 	struct sqe_submit *s = &req->submit;
> +	struct io_ring_ctx *ctx = req->ctx;
> 
> 	req->user_data = READ_ONCE(s->sqe->user_data);
> 
> @@ -2389,7 +2392,7 @@ static void io_wq_submit_work(struct io_wq_work **workptr)
> 		s->has_user = (work->flags & IO_WQ_WORK_HAS_MM) != 0;
> 		s->in_async = true;
> 		do {
> -			ret = __io_submit_sqe(ctx, req, &nxt, false);
> +			ret = __io_submit_sqe(req, &nxt, false);
> 			/*
> 			 * We can get EAGAIN for polled IO even though we're
> 			 * forcing a sync submission from here, since we can't
> @@ -2443,10 +2446,10 @@ static inline struct file *io_file_from_index(struct io_ring_ctx *ctx,
> 	return table->files[index & IORING_FILE_TABLE_MASK];
> }
> 
> -static int io_req_set_file(struct io_ring_ctx *ctx,
> -			   struct io_submit_state *state, struct io_kiocb *req)
> +static int io_req_set_file(struct io_submit_state *state, struct io_kiocb *req)
> {
> 	struct sqe_submit *s = &req->submit;
> +	struct io_ring_ctx *ctx = req->ctx;
> 	unsigned flags;
> 	int fd;
> 
> @@ -2486,9 +2489,10 @@ static int io_req_set_file(struct io_ring_ctx *ctx,
> 	return 0;
> }
> 
> -static int io_grab_files(struct io_ring_ctx *ctx, struct io_kiocb *req)
> +static int io_grab_files(struct io_kiocb *req)
> {
> 	int ret = -EBADF;
> +	struct io_ring_ctx *ctx = req->ctx;
> 
> 	rcu_read_lock();
> 	spin_lock_irq(&ctx->inflight_lock);
> @@ -2604,8 +2608,9 @@ static inline struct io_kiocb *io_get_linked_timeout(struct io_kiocb *req)
> 	return NULL;
> }
> 
> -static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
> +static int __io_queue_sqe(struct io_kiocb *req)
> {
> +	struct io_ring_ctx *ctx = req->ctx;
> 	struct io_kiocb *nxt;
> 	int ret;
> 
> @@ -2616,7 +2621,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
> 			goto err;
> 	}
> 
> -	ret = __io_submit_sqe(ctx, req, NULL, true);
> +	ret = __io_submit_sqe(req, NULL, true);
> 
> 	/*
> 	 * We async punt it if the file wasn't marked NOWAIT, or if the file
> @@ -2631,7 +2636,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
> 		if (sqe_copy) {
> 			s->sqe = sqe_copy;
> 			if (req->work.flags & IO_WQ_WORK_NEEDS_FILES) {
> -				ret = io_grab_files(ctx, req);
> +				ret = io_grab_files(req);
> 				if (ret) {
> 					kfree(sqe_copy);
> 					goto err;
> @@ -2642,7 +2647,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
> 			 * Queued up for async execution, worker will release
> 			 * submit reference when the iocb is actually submitted.
> 			 */
> -			io_queue_async_work(ctx, req);
> +			io_queue_async_work(req);
> 			return 0;
> 		}
> 	}
> @@ -2662,11 +2667,12 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
> 	return ret;
> }
> 
> -static int io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
> +static int io_queue_sqe(struct io_kiocb *req)
> {
> +	struct io_ring_ctx *ctx = req->ctx;
> 	int ret;
> 
> -	ret = io_req_defer(ctx, req);
> +	ret = io_req_defer(req);
> 	if (ret) {
> 		if (ret != -EIOCBQUEUED) {
> 			io_cqring_add_event(ctx, req->submit.sqe->user_data, ret);
> @@ -2675,17 +2681,17 @@ static int io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
> 		return 0;
> 	}
> 
> -	return __io_queue_sqe(ctx, req);
> +	return __io_queue_sqe(req);
> }
> 
> -static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req,
> -			      struct io_kiocb *shadow)
> +static int io_queue_link_head(struct io_kiocb *req, struct io_kiocb *shadow)
> {
> 	int ret;
> 	int need_submit = false;
> +	struct io_ring_ctx *ctx = req->ctx;
> 
> 	if (!shadow)
> -		return io_queue_sqe(ctx, req);
> +		return io_queue_sqe(req);
> 
> 	/*
> 	 * Mark the first IO in link list as DRAIN, let all the following
> @@ -2693,7 +2699,7 @@ static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req,
> 	 * list.
> 	 */
> 	req->flags |= REQ_F_IO_DRAIN;
> -	ret = io_req_defer(ctx, req);
> +	ret = io_req_defer(req);
> 	if (ret) {
> 		if (ret != -EIOCBQUEUED) {
> 			io_cqring_add_event(ctx, req->submit.sqe->user_data, ret);
> @@ -2716,18 +2722,19 @@ static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req,
> 	spin_unlock_irq(&ctx->completion_lock);
> 
> 	if (need_submit)
> -		return __io_queue_sqe(ctx, req);
> +		return __io_queue_sqe(req);
> 
> 	return 0;
> }
> 
> #define SQE_VALID_FLAGS	(IOSQE_FIXED_FILE|IOSQE_IO_DRAIN|IOSQE_IO_LINK)
> 
> -static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
> -			  struct io_submit_state *state, struct io_kiocb **link)
> +static void io_submit_sqe(struct io_kiocb *req, struct io_submit_state *state,
> +			  struct io_kiocb **link)
> {
> 	struct io_uring_sqe *sqe_copy;
> 	struct sqe_submit *s = &req->submit;
> +	struct io_ring_ctx *ctx = req->ctx;
> 	int ret;
> 
> 	/* enforce forwards compatibility on users */
> @@ -2736,7 +2743,7 @@ static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
> 		goto err_req;
> 	}
> 
> -	ret = io_req_set_file(ctx, state, req);
> +	ret = io_req_set_file(state, req);
> 	if (unlikely(ret)) {
> err_req:
> 		io_cqring_add_event(ctx, s->sqe->user_data, ret);
> @@ -2775,7 +2782,7 @@ static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
> 		ret = -EINVAL;
> 		goto err_req;
> 	} else {
> -		io_queue_sqe(ctx, req);
> +		io_queue_sqe(req);
> 	}
> }
> 
> @@ -2919,7 +2926,7 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr,
> 		req->submit.needs_fixed_file = async;
> 		trace_io_uring_submit_sqe(ctx, req->submit.sqe->user_data,
> 					  true, async);
> -		io_submit_sqe(ctx, req, statep, &link);
> +		io_submit_sqe(req, statep, &link);
> 		submitted++;
> 
> 		/*
> @@ -2927,14 +2934,14 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr,
> 		 * that's the end of the chain. Submit the previous link.
> 		 */
> 		if (!(sqe_flags & IOSQE_IO_LINK) && link) {
> -			io_queue_link_head(ctx, link, shadow_req);
> +			io_queue_link_head(link, shadow_req);
> 			link = NULL;
> 			shadow_req = NULL;
> 		}
> 	}
> 
> 	if (link)
> -		io_queue_link_head(ctx, link, shadow_req);
> +		io_queue_link_head(link, shadow_req);
> 	if (statep)
> 		io_submit_state_end(&state);
> 
> -- 
> 2.7.4
> 
> 
> 


--
BR, Jackie Liu




^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: [PATCH v2] io_uring: reduced function parameter ctx if possible
  2019-11-08  9:00 ` Jackie Liu
@ 2019-11-08  9:10   ` Bob Liu
       [not found]     ` <7B6E98CA-22DE-4AFA-94C3-D65395A1629C@kylinos.cn>
  0 siblings, 1 reply; 9+ messages in thread
From: Bob Liu @ 2019-11-08  9:10 UTC (permalink / raw)
  To: Jackie Liu, Jens Axboe; +Cc: io-uring

On 11/8/19 5:00 PM, Jackie Liu wrote:
> oops, please ignore this, it's crash with commit '2665abfd757fb35a241c6f0b1ebf620e3ffb36fb'
> I need test more, and resend later.
> 

Hmm.. io_link_cancel_timeout() may free req..

>> 2019年11月8日 15:29,Jackie Liu <liuyun01@kylinos.cn> 写道:
>>
>> Many times, the core of the function is req, and req has already set
>> req->ctx at initialization time, so there is no need to pass in from
>> outside.
>>
>> Cleanup, no function change.
>>
>> Signed-off-by: Jackie Liu <liuyun01@kylinos.cn>
>> ---
>> V2:
>> - rebase to branch for-5.5/io_uring: 2665abfd757fb35a241c6f0b1ebf620e3ffb36fb
>>
>> fs/io_uring.c | 109 +++++++++++++++++++++++++++++++---------------------------
>> 1 file changed, 58 insertions(+), 51 deletions(-)
>>
>> diff --git a/fs/io_uring.c b/fs/io_uring.c
>> index eadd19a..2cc53e3 100644
>> --- a/fs/io_uring.c
>> +++ b/fs/io_uring.c
>> @@ -429,20 +429,20 @@ static struct io_ring_ctx *io_ring_ctx_alloc(struct io_uring_params *p)
>> 	return ctx;
>> }
>>
>> -static inline bool __io_sequence_defer(struct io_ring_ctx *ctx,
>> -				       struct io_kiocb *req)
>> +static inline bool __io_sequence_defer(struct io_kiocb *req)
>> {
>> +	struct io_ring_ctx *ctx = req->ctx;
>> +
>> 	return req->sequence != ctx->cached_cq_tail + ctx->cached_sq_dropped
>> 					+ atomic_read(&ctx->cached_cq_overflow);
>> }
>>
>> -static inline bool io_sequence_defer(struct io_ring_ctx *ctx,
>> -				     struct io_kiocb *req)
>> +static inline bool io_sequence_defer(struct io_kiocb *req)
>> {
>> 	if ((req->flags & (REQ_F_IO_DRAIN|REQ_F_IO_DRAINED)) != REQ_F_IO_DRAIN)
>> 		return false;
>>
>> -	return __io_sequence_defer(ctx, req);
>> +	return __io_sequence_defer(req);
>> }
>>
>> static struct io_kiocb *io_get_deferred_req(struct io_ring_ctx *ctx)
>> @@ -450,7 +450,7 @@ static struct io_kiocb *io_get_deferred_req(struct io_ring_ctx *ctx)
>> 	struct io_kiocb *req;
>>
>> 	req = list_first_entry_or_null(&ctx->defer_list, struct io_kiocb, list);
>> -	if (req && !io_sequence_defer(ctx, req)) {
>> +	if (req && !io_sequence_defer(req)) {
>> 		list_del_init(&req->list);
>> 		return req;
>> 	}
>> @@ -463,7 +463,7 @@ static struct io_kiocb *io_get_timeout_req(struct io_ring_ctx *ctx)
>> 	struct io_kiocb *req;
>>
>> 	req = list_first_entry_or_null(&ctx->timeout_list, struct io_kiocb, list);
>> -	if (req && !__io_sequence_defer(ctx, req)) {
>> +	if (req && !__io_sequence_defer(req)) {
>> 		list_del_init(&req->list);
>> 		return req;
>> 	}
>> @@ -512,10 +512,10 @@ static inline bool io_prep_async_work(struct io_kiocb *req)
>> 	return do_hashed;
>> }
>>
>> -static inline void io_queue_async_work(struct io_ring_ctx *ctx,
>> -				       struct io_kiocb *req)
>> +static inline void io_queue_async_work(struct io_kiocb *req)
>> {
>> 	bool do_hashed = io_prep_async_work(req);
>> +	struct io_ring_ctx *ctx = req->ctx;
>>
>> 	trace_io_uring_queue_async_work(ctx, do_hashed, req, &req->work,
>> 					req->flags);
>> @@ -566,7 +566,7 @@ static void io_commit_cqring(struct io_ring_ctx *ctx)
>> 			continue;
>> 		}
>> 		req->flags |= REQ_F_IO_DRAINED;
>> -		io_queue_async_work(ctx, req);
>> +		io_queue_async_work(req);
>> 	}
>> }
>>
>> @@ -714,9 +714,9 @@ static void __io_free_req(struct io_kiocb *req)
>> 	kmem_cache_free(req_cachep, req);
>> }
>>
>> -static bool io_link_cancel_timeout(struct io_ring_ctx *ctx,
>> -				   struct io_kiocb *req)
>> +static bool io_link_cancel_timeout(struct io_kiocb *req)
>> {
>> +	struct io_ring_ctx *ctx = req->ctx;
>> 	int ret;
>>
>> 	ret = hrtimer_try_to_cancel(&req->timeout.timer);
>> @@ -756,7 +756,7 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr)
>> 		 * in this context instead of having to queue up new async work.
>> 		 */
>> 		if (req->flags & REQ_F_LINK_TIMEOUT) {
>> -			wake_ev = io_link_cancel_timeout(ctx, nxt);
>> +			wake_ev = io_link_cancel_timeout(nxt);
>>
>> 			/* we dropped this link, get next */
>> 			nxt = list_first_entry_or_null(&req->link_list,
>> @@ -765,7 +765,7 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr)
>> 			*nxtptr = nxt;
>> 			break;
>> 		} else {
>> -			io_queue_async_work(req->ctx, nxt);
>> +			io_queue_async_work(nxt);
>> 			break;
>> 		}
>> 	}
>> @@ -793,7 +793,7 @@ static void io_fail_links(struct io_kiocb *req)
>>
>> 		if ((req->flags & REQ_F_LINK_TIMEOUT) &&
>> 		    link->submit.sqe->opcode == IORING_OP_LINK_TIMEOUT) {
>> -			io_link_cancel_timeout(ctx, link);
>> +			io_link_cancel_timeout(link);
>> 		} else {
>> 			io_cqring_fill_event(ctx, link->user_data, -ECANCELED);
>> 			__io_free_req(link);
>> @@ -862,7 +862,7 @@ static void io_put_req(struct io_kiocb *req, struct io_kiocb **nxtptr)
>> 		if (nxtptr)
>> 			*nxtptr = nxt;
>> 		else
>> -			io_queue_async_work(nxt->ctx, nxt);
>> +			io_queue_async_work(nxt);
>> 	}
>> }
>>
>> @@ -1803,7 +1803,7 @@ static void io_poll_remove_one(struct io_kiocb *req)
>> 	WRITE_ONCE(poll->canceled, true);
>> 	if (!list_empty(&poll->wait.entry)) {
>> 		list_del_init(&poll->wait.entry);
>> -		io_queue_async_work(req->ctx, req);
>> +		io_queue_async_work(req);
>> 	}
>> 	spin_unlock(&poll->head->lock);
>>
>> @@ -1855,9 +1855,10 @@ static int io_poll_remove(struct io_kiocb *req, const struct io_uring_sqe *sqe)
>> 	return 0;
>> }
>>
>> -static void io_poll_complete(struct io_ring_ctx *ctx, struct io_kiocb *req,
>> -			     __poll_t mask)
>> +static void io_poll_complete(struct io_kiocb *req, __poll_t mask)
>> {
>> +	struct io_ring_ctx *ctx = req->ctx;
>> +
>> 	req->poll.done = true;
>> 	io_cqring_fill_event(ctx, req->user_data, mangle_poll(mask));
>> 	io_commit_cqring(ctx);
>> @@ -1893,7 +1894,7 @@ static void io_poll_complete_work(struct io_wq_work **workptr)
>> 		return;
>> 	}
>> 	list_del_init(&req->list);
>> -	io_poll_complete(ctx, req, mask);
>> +	io_poll_complete(req, mask);
>> 	spin_unlock_irq(&ctx->completion_lock);
>>
>> 	io_cqring_ev_posted(ctx);
>> @@ -1921,13 +1922,13 @@ static int io_poll_wake(struct wait_queue_entry *wait, unsigned mode, int sync,
>>
>> 	if (mask && spin_trylock_irqsave(&ctx->completion_lock, flags)) {
>> 		list_del(&req->list);
>> -		io_poll_complete(ctx, req, mask);
>> +		io_poll_complete(req, mask);
>> 		spin_unlock_irqrestore(&ctx->completion_lock, flags);
>>
>> 		io_cqring_ev_posted(ctx);
>> 		io_put_req(req, NULL);
>> 	} else {
>> -		io_queue_async_work(ctx, req);
>> +		io_queue_async_work(req);
>> 	}
>>
>> 	return 1;
>> @@ -2012,7 +2013,7 @@ static int io_poll_add(struct io_kiocb *req, const struct io_uring_sqe *sqe,
>> 	}
>> 	if (mask) { /* no async, we'd stolen it */
>> 		ipt.error = 0;
>> -		io_poll_complete(ctx, req, mask);
>> +		io_poll_complete(req, mask);
>> 	}
>> 	spin_unlock_irq(&ctx->completion_lock);
>>
>> @@ -2259,12 +2260,13 @@ static int io_async_cancel(struct io_kiocb *req, const struct io_uring_sqe *sqe,
>> 	return 0;
>> }
>>
>> -static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> +static int io_req_defer(struct io_kiocb *req)
>> {
>> 	const struct io_uring_sqe *sqe = req->submit.sqe;
>> 	struct io_uring_sqe *sqe_copy;
>> +	struct io_ring_ctx *ctx = req->ctx;
>>
>> -	if (!io_sequence_defer(ctx, req) && list_empty(&ctx->defer_list))
>> +	if (!io_sequence_defer(req) && list_empty(&ctx->defer_list))
>> 		return 0;
>>
>> 	sqe_copy = kmalloc(sizeof(*sqe_copy), GFP_KERNEL);
>> @@ -2272,7 +2274,7 @@ static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> 		return -EAGAIN;
>>
>> 	spin_lock_irq(&ctx->completion_lock);
>> -	if (!io_sequence_defer(ctx, req) && list_empty(&ctx->defer_list)) {
>> +	if (!io_sequence_defer(req) && list_empty(&ctx->defer_list)) {
>> 		spin_unlock_irq(&ctx->completion_lock);
>> 		kfree(sqe_copy);
>> 		return 0;
>> @@ -2287,11 +2289,12 @@ static int io_req_defer(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> 	return -EIOCBQUEUED;
>> }
>>
>> -static int __io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
>> -			   struct io_kiocb **nxt, bool force_nonblock)
>> +static int __io_submit_sqe(struct io_kiocb *req, struct io_kiocb **nxt,
>> +			   bool force_nonblock)
>> {
>> 	int ret, opcode;
>> 	struct sqe_submit *s = &req->submit;
>> +	struct io_ring_ctx *ctx = req->ctx;
>>
>> 	req->user_data = READ_ONCE(s->sqe->user_data);
>>
>> @@ -2389,7 +2392,7 @@ static void io_wq_submit_work(struct io_wq_work **workptr)
>> 		s->has_user = (work->flags & IO_WQ_WORK_HAS_MM) != 0;
>> 		s->in_async = true;
>> 		do {
>> -			ret = __io_submit_sqe(ctx, req, &nxt, false);
>> +			ret = __io_submit_sqe(req, &nxt, false);
>> 			/*
>> 			 * We can get EAGAIN for polled IO even though we're
>> 			 * forcing a sync submission from here, since we can't
>> @@ -2443,10 +2446,10 @@ static inline struct file *io_file_from_index(struct io_ring_ctx *ctx,
>> 	return table->files[index & IORING_FILE_TABLE_MASK];
>> }
>>
>> -static int io_req_set_file(struct io_ring_ctx *ctx,
>> -			   struct io_submit_state *state, struct io_kiocb *req)
>> +static int io_req_set_file(struct io_submit_state *state, struct io_kiocb *req)
>> {
>> 	struct sqe_submit *s = &req->submit;
>> +	struct io_ring_ctx *ctx = req->ctx;
>> 	unsigned flags;
>> 	int fd;
>>
>> @@ -2486,9 +2489,10 @@ static int io_req_set_file(struct io_ring_ctx *ctx,
>> 	return 0;
>> }
>>
>> -static int io_grab_files(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> +static int io_grab_files(struct io_kiocb *req)
>> {
>> 	int ret = -EBADF;
>> +	struct io_ring_ctx *ctx = req->ctx;
>>
>> 	rcu_read_lock();
>> 	spin_lock_irq(&ctx->inflight_lock);
>> @@ -2604,8 +2608,9 @@ static inline struct io_kiocb *io_get_linked_timeout(struct io_kiocb *req)
>> 	return NULL;
>> }
>>
>> -static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> +static int __io_queue_sqe(struct io_kiocb *req)
>> {
>> +	struct io_ring_ctx *ctx = req->ctx;
>> 	struct io_kiocb *nxt;
>> 	int ret;
>>
>> @@ -2616,7 +2621,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> 			goto err;
>> 	}
>>
>> -	ret = __io_submit_sqe(ctx, req, NULL, true);
>> +	ret = __io_submit_sqe(req, NULL, true);
>>
>> 	/*
>> 	 * We async punt it if the file wasn't marked NOWAIT, or if the file
>> @@ -2631,7 +2636,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> 		if (sqe_copy) {
>> 			s->sqe = sqe_copy;
>> 			if (req->work.flags & IO_WQ_WORK_NEEDS_FILES) {
>> -				ret = io_grab_files(ctx, req);
>> +				ret = io_grab_files(req);
>> 				if (ret) {
>> 					kfree(sqe_copy);
>> 					goto err;
>> @@ -2642,7 +2647,7 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> 			 * Queued up for async execution, worker will release
>> 			 * submit reference when the iocb is actually submitted.
>> 			 */
>> -			io_queue_async_work(ctx, req);
>> +			io_queue_async_work(req);
>> 			return 0;
>> 		}
>> 	}
>> @@ -2662,11 +2667,12 @@ static int __io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> 	return ret;
>> }
>>
>> -static int io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> +static int io_queue_sqe(struct io_kiocb *req)
>> {
>> +	struct io_ring_ctx *ctx = req->ctx;
>> 	int ret;
>>
>> -	ret = io_req_defer(ctx, req);
>> +	ret = io_req_defer(req);
>> 	if (ret) {
>> 		if (ret != -EIOCBQUEUED) {
>> 			io_cqring_add_event(ctx, req->submit.sqe->user_data, ret);
>> @@ -2675,17 +2681,17 @@ static int io_queue_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req)
>> 		return 0;
>> 	}
>>
>> -	return __io_queue_sqe(ctx, req);
>> +	return __io_queue_sqe(req);
>> }
>>
>> -static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req,
>> -			      struct io_kiocb *shadow)
>> +static int io_queue_link_head(struct io_kiocb *req, struct io_kiocb *shadow)
>> {
>> 	int ret;
>> 	int need_submit = false;
>> +	struct io_ring_ctx *ctx = req->ctx;
>>
>> 	if (!shadow)
>> -		return io_queue_sqe(ctx, req);
>> +		return io_queue_sqe(req);
>>
>> 	/*
>> 	 * Mark the first IO in link list as DRAIN, let all the following
>> @@ -2693,7 +2699,7 @@ static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req,
>> 	 * list.
>> 	 */
>> 	req->flags |= REQ_F_IO_DRAIN;
>> -	ret = io_req_defer(ctx, req);
>> +	ret = io_req_defer(req);
>> 	if (ret) {
>> 		if (ret != -EIOCBQUEUED) {
>> 			io_cqring_add_event(ctx, req->submit.sqe->user_data, ret);
>> @@ -2716,18 +2722,19 @@ static int io_queue_link_head(struct io_ring_ctx *ctx, struct io_kiocb *req,
>> 	spin_unlock_irq(&ctx->completion_lock);
>>
>> 	if (need_submit)
>> -		return __io_queue_sqe(ctx, req);
>> +		return __io_queue_sqe(req);
>>
>> 	return 0;
>> }
>>
>> #define SQE_VALID_FLAGS	(IOSQE_FIXED_FILE|IOSQE_IO_DRAIN|IOSQE_IO_LINK)
>>
>> -static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
>> -			  struct io_submit_state *state, struct io_kiocb **link)
>> +static void io_submit_sqe(struct io_kiocb *req, struct io_submit_state *state,
>> +			  struct io_kiocb **link)
>> {
>> 	struct io_uring_sqe *sqe_copy;
>> 	struct sqe_submit *s = &req->submit;
>> +	struct io_ring_ctx *ctx = req->ctx;
>> 	int ret;
>>
>> 	/* enforce forwards compatibility on users */
>> @@ -2736,7 +2743,7 @@ static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
>> 		goto err_req;
>> 	}
>>
>> -	ret = io_req_set_file(ctx, state, req);
>> +	ret = io_req_set_file(state, req);
>> 	if (unlikely(ret)) {
>> err_req:
>> 		io_cqring_add_event(ctx, s->sqe->user_data, ret);
>> @@ -2775,7 +2782,7 @@ static void io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
>> 		ret = -EINVAL;
>> 		goto err_req;
>> 	} else {
>> -		io_queue_sqe(ctx, req);
>> +		io_queue_sqe(req);
>> 	}
>> }
>>
>> @@ -2919,7 +2926,7 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr,
>> 		req->submit.needs_fixed_file = async;
>> 		trace_io_uring_submit_sqe(ctx, req->submit.sqe->user_data,
>> 					  true, async);
>> -		io_submit_sqe(ctx, req, statep, &link);
>> +		io_submit_sqe(req, statep, &link);
>> 		submitted++;
>>
>> 		/*
>> @@ -2927,14 +2934,14 @@ static int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr,
>> 		 * that's the end of the chain. Submit the previous link.
>> 		 */
>> 		if (!(sqe_flags & IOSQE_IO_LINK) && link) {
>> -			io_queue_link_head(ctx, link, shadow_req);
>> +			io_queue_link_head(link, shadow_req);
>> 			link = NULL;
>> 			shadow_req = NULL;
>> 		}
>> 	}
>>
>> 	if (link)
>> -		io_queue_link_head(ctx, link, shadow_req);
>> +		io_queue_link_head(link, shadow_req);
>> 	if (statep)
>> 		io_submit_state_end(&state);
>>
>> -- 
>> 2.7.4
>>
>>
>>
> 
> 
> --
> BR, Jackie Liu
> 
> 
> 


^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: [PATCH v2] io_uring: reduced function parameter ctx if possible
       [not found]     ` <7B6E98CA-22DE-4AFA-94C3-D65395A1629C@kylinos.cn>
@ 2019-11-08  9:24       ` Jackie Liu
  2019-11-08  9:35         ` Bob Liu
  0 siblings, 1 reply; 9+ messages in thread
From: Jackie Liu @ 2019-11-08  9:24 UTC (permalink / raw)
  To: Bob Liu; +Cc: Jens Axboe, io-uring


> 
>> 2019年11月8日 17:10,Bob Liu <bob.liu@oracle.com> 写道:
>> 
>> On 11/8/19 5:00 PM, Jackie Liu wrote:
>>> oops, please ignore this, it's crash with commit '2665abfd757fb35a241c6f0b1ebf620e3ffb36fb'
>>> I need test more, and resend later.
>>> 
>> 
>> Hmm.. io_link_cancel_timeout() may free req..
>> 
> 
> Actually, there is no problem with this patch. It is caused by a patch that I
> have not sent yet, I will send it after doing more tests.
> 

Sad. rejected by mailing list? 

@@ -714,9 +714,9 @@ static void __io_free_req(struct io_kiocb *req)
        kmem_cache_free(req_cachep, req);
 }

-static bool io_link_cancel_timeout(struct io_ring_ctx *ctx,
-                                  struct io_kiocb *req)
+static bool io_link_cancel_timeout(struct io_kiocb *req)
 {
+       struct io_ring_ctx *ctx = req->ctx;
        int ret;

        ret = hrtimer_try_to_cancel(&req->timeout.timer);

I use req directly to convert to ctx, there is no crash problem here.

--
BR, Jackie Liu




^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: [PATCH v2] io_uring: reduced function parameter ctx if possible
  2019-11-08  9:24       ` Jackie Liu
@ 2019-11-08  9:35         ` Bob Liu
  0 siblings, 0 replies; 9+ messages in thread
From: Bob Liu @ 2019-11-08  9:35 UTC (permalink / raw)
  To: Jackie Liu; +Cc: Jens Axboe, io-uring

On 11/8/19 5:24 PM, Jackie Liu wrote:
> 
>>
>>> 2019年11月8日 17:10,Bob Liu <bob.liu@oracle.com> 写道:
>>>
>>> On 11/8/19 5:00 PM, Jackie Liu wrote:
>>>> oops, please ignore this, it's crash with commit '2665abfd757fb35a241c6f0b1ebf620e3ffb36fb'
>>>> I need test more, and resend later.
>>>>
>>>
>>> Hmm.. io_link_cancel_timeout() may free req..
>>>
>>
>> Actually, there is no problem with this patch. It is caused by a patch that I
>> have not sent yet, I will send it after doing more tests.
>>
> 
> Sad. rejected by mailing list? 
> 
> @@ -714,9 +714,9 @@ static void __io_free_req(struct io_kiocb *req)
>         kmem_cache_free(req_cachep, req);
>  }
> 
> -static bool io_link_cancel_timeout(struct io_ring_ctx *ctx,
> -                                  struct io_kiocb *req)
> +static bool io_link_cancel_timeout(struct io_kiocb *req)
>  {
> +       struct io_ring_ctx *ctx = req->ctx;
>         int ret;
> 
>         ret = hrtimer_try_to_cancel(&req->timeout.timer);
> 
> I use req directly to convert to ctx, there is no crash problem here.
> 

Oh, I see.. Sorry for miss that..

> --
> BR, Jackie Liu
> 
> 
> 


^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: [PATCH v2] io_uring: reduced function parameter ctx if possible
  2019-11-08  7:29 [PATCH v2] io_uring: reduced function parameter ctx if possible Jackie Liu
  2019-11-08  9:00 ` Jackie Liu
@ 2019-11-08 14:11 ` Jens Axboe
  2019-11-08 14:16   ` Jens Axboe
  1 sibling, 1 reply; 9+ messages in thread
From: Jens Axboe @ 2019-11-08 14:11 UTC (permalink / raw)
  To: Jackie Liu; +Cc: io-uring

On 11/8/19 12:29 AM, Jackie Liu wrote:
> Many times, the core of the function is req, and req has already set
> req->ctx at initialization time, so there is no need to pass in from
> outside.
> 
> Cleanup, no function change.

I was curious if this patch netted us any improvements as well, but it
actually blows up the text segment a lot on my laptop. Before the
patch:

   text	   data	    bss	    dec	    hex	filename
  87504	  17588	    256	 105348	  19b84	fs/io_uring.o

and after:

   text	   data	    bss	    dec	    hex	filename
  99098	  17876	    256	 117230	  1c9ee	fs/io_uring.o

which seems really odd. I double checked just to be sure!

axboe@x1:~ $ gcc --version
gcc (Ubuntu 9.2.1-17ubuntu1~18.04.1) 9.2.1 20191102

-- 
Jens Axboe


^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: [PATCH v2] io_uring: reduced function parameter ctx if possible
  2019-11-08 14:11 ` Jens Axboe
@ 2019-11-08 14:16   ` Jens Axboe
       [not found]     ` <5dc57d18.1c69fb81.78a44.b83dSMTPIN_ADDED_BROKEN@mx.google.com>
  0 siblings, 1 reply; 9+ messages in thread
From: Jens Axboe @ 2019-11-08 14:16 UTC (permalink / raw)
  To: Jackie Liu; +Cc: io-uring

On 11/8/19 7:11 AM, Jens Axboe wrote:
> On 11/8/19 12:29 AM, Jackie Liu wrote:
>> Many times, the core of the function is req, and req has already set
>> req->ctx at initialization time, so there is no need to pass in from
>> outside.
>>
>> Cleanup, no function change.
> 
> I was curious if this patch netted us any improvements as well, but it
> actually blows up the text segment a lot on my laptop. Before the
> patch:
> 
>     text	   data	    bss	    dec	    hex	filename
>    87504	  17588	    256	 105348	  19b84	fs/io_uring.o
> 
> and after:
> 
>     text	   data	    bss	    dec	    hex	filename
>    99098	  17876	    256	 117230	  1c9ee	fs/io_uring.o
> 
> which seems really odd. I double checked just to be sure!
> 
> axboe@x1:~ $ gcc --version
> gcc (Ubuntu 9.2.1-17ubuntu1~18.04.1) 9.2.1 20191102

I took a look at the generated code, and it looks like it's just
enabling gcc to inline a lot more aggressively.

-- 
Jens Axboe


^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: [PATCH v2] io_uring: reduced function parameter ctx if possible
       [not found]     ` <5dc57d18.1c69fb81.78a44.b83dSMTPIN_ADDED_BROKEN@mx.google.com>
@ 2019-11-08 14:39       ` Jens Axboe
       [not found]         ` <5dc57f27.1c69fb81.27d15.ccf4SMTPIN_ADDED_BROKEN@mx.google.com>
  0 siblings, 1 reply; 9+ messages in thread
From: Jens Axboe @ 2019-11-08 14:39 UTC (permalink / raw)
  To: Jackie Liu; +Cc: io-uring

On 11/8/19 7:33 AM, Jackie Liu wrote:
> 
> 
> 在 2019/11/8 22:16, Jens Axboe 写道:
>> On 11/8/19 7:11 AM, Jens Axboe wrote:
>>> On 11/8/19 12:29 AM, Jackie Liu wrote:
>>>> Many times, the core of the function is req, and req has already set
>>>> req->ctx at initialization time, so there is no need to pass in from
>>>> outside.
>>>>
>>>> Cleanup, no function change.
>>>
>>> I was curious if this patch netted us any improvements as well, but it
>>> actually blows up the text segment a lot on my laptop. Before the
>>> patch:
>>>
>>>       text	   data	    bss	    dec	    hex	filename
>>>      87504	  17588	    256	 105348	  19b84	fs/io_uring.o
>>>
>>> and after:
>>>
>>>       text	   data	    bss	    dec	    hex	filename
>>>      99098	  17876	    256	 117230	  1c9ee	fs/io_uring.o
>>>
>>> which seems really odd. I double checked just to be sure!
>>>
>>> axboe@x1:~ $ gcc --version
>>> gcc (Ubuntu 9.2.1-17ubuntu1~18.04.1) 9.2.1 20191102
>>
>> I took a look at the generated code, and it looks like it's just
>> enabling gcc to inline a lot more aggressively.
>>
> 
> Hum... I don't have a lot of attention to the size of the generated
> code, just to make the code look easier to understand. Will this affect
> performance?

Seems to be an extreme case on the laptop since I have various debug
options enabled for testing in qemu, but it's an increase on my
test box with gcc 7.2 as well. I don't think it's anything to worry
about, it's just inlining what it should inline. I did a quick perf
run and we're not down, so I'm going to apply this one as it cleans
up the code and that's always the most important.

-- 
Jens Axboe


^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: [PATCH v2] io_uring: reduced function parameter ctx if possible
       [not found]         ` <5dc57f27.1c69fb81.27d15.ccf4SMTPIN_ADDED_BROKEN@mx.google.com>
@ 2019-11-08 14:47           ` Jens Axboe
  0 siblings, 0 replies; 9+ messages in thread
From: Jens Axboe @ 2019-11-08 14:47 UTC (permalink / raw)
  To: Jackie Liu; +Cc: io-uring

On 11/8/19 7:43 AM, Jackie Liu wrote:
> 
> 
> 在 2019/11/8 22:39, Jens Axboe 写道:
>> On 11/8/19 7:33 AM, Jackie Liu wrote:
>>>
>>>
>>> 在 2019/11/8 22:16, Jens Axboe 写道:
>>>> On 11/8/19 7:11 AM, Jens Axboe wrote:
>>>>> On 11/8/19 12:29 AM, Jackie Liu wrote:
>>>>>> Many times, the core of the function is req, and req has already set
>>>>>> req->ctx at initialization time, so there is no need to pass in from
>>>>>> outside.
>>>>>>
>>>>>> Cleanup, no function change.
>>>>>
>>>>> I was curious if this patch netted us any improvements as well, but it
>>>>> actually blows up the text segment a lot on my laptop. Before the
>>>>> patch:
>>>>>
>>>>>         text	   data	    bss	    dec	    hex	filename
>>>>>        87504	  17588	    256	 105348	  19b84	fs/io_uring.o
>>>>>
>>>>> and after:
>>>>>
>>>>>         text	   data	    bss	    dec	    hex	filename
>>>>>        99098	  17876	    256	 117230	  1c9ee	fs/io_uring.o
>>>>>
>>>>> which seems really odd. I double checked just to be sure!
>>>>>
>>>>> axboe@x1:~ $ gcc --version
>>>>> gcc (Ubuntu 9.2.1-17ubuntu1~18.04.1) 9.2.1 20191102
>>>>
>>>> I took a look at the generated code, and it looks like it's just
>>>> enabling gcc to inline a lot more aggressively.
>>>>
>>>
>>> Hum... I don't have a lot of attention to the size of the generated
>>> code, just to make the code look easier to understand. Will this affect
>>> performance?
>>
>> Seems to be an extreme case on the laptop since I have various debug
>> options enabled for testing in qemu, but it's an increase on my
>> test box with gcc 7.2 as well. I don't think it's anything to worry
>> about, it's just inlining what it should inline. I did a quick perf
>> run and we're not down, so I'm going to apply this one as it cleans
>> up the code and that's always the most important.
>>
> 
> Please wait a moment, I will send the third version, although this patch
> has not been modified, but there is a separate series of patches sent
> together, please reviewed them for me.

OK, will do.

-- 
Jens Axboe


^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, back to index

Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-11-08  7:29 [PATCH v2] io_uring: reduced function parameter ctx if possible Jackie Liu
2019-11-08  9:00 ` Jackie Liu
2019-11-08  9:10   ` Bob Liu
     [not found]     ` <7B6E98CA-22DE-4AFA-94C3-D65395A1629C@kylinos.cn>
2019-11-08  9:24       ` Jackie Liu
2019-11-08  9:35         ` Bob Liu
2019-11-08 14:11 ` Jens Axboe
2019-11-08 14:16   ` Jens Axboe
     [not found]     ` <5dc57d18.1c69fb81.78a44.b83dSMTPIN_ADDED_BROKEN@mx.google.com>
2019-11-08 14:39       ` Jens Axboe
     [not found]         ` <5dc57f27.1c69fb81.27d15.ccf4SMTPIN_ADDED_BROKEN@mx.google.com>
2019-11-08 14:47           ` Jens Axboe

IO-Uring Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/io-uring/0 io-uring/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 io-uring io-uring/ https://lore.kernel.org/io-uring \
		io-uring@vger.kernel.org
	public-inbox-index io-uring

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.io-uring


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git