From 7d67af2c013402537385dae343a2d0f6a4cb3bfd Mon Sep 17 00:00:00 2001
From: Pavel Begunkov <asml.silence@gmail.com>
Date: Mon, 24 Feb 2020 11:32:45 +0300
Subject: io_uring: add splice(2) support

Add support for splice(2).

- output file is specified as sqe->fd, so it's handled by generic code
- hash_reg_file handled by generic code as well
- len is 32bit, but should be fine
- the fd_in is registered file, when SPLICE_F_FD_IN_FIXED is set, which
is a splice flag (i.e. sqe->splice_flags).

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/uapi/linux/io_uring.h | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)

(limited to 'include/uapi/linux')

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 3f7961c1c243..08891cc1c1e7 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -23,7 +23,10 @@ struct io_uring_sqe {
 		__u64	off;	/* offset into file */
 		__u64	addr2;
 	};
-	__u64	addr;		/* pointer to buffer or iovecs */
+	union {
+		__u64	addr;	/* pointer to buffer or iovecs */
+		__u64	splice_off_in;
+	};
 	__u32	len;		/* buffer size or number of iovecs */
 	union {
 		__kernel_rwf_t	rw_flags;
@@ -37,6 +40,7 @@ struct io_uring_sqe {
 		__u32		open_flags;
 		__u32		statx_flags;
 		__u32		fadvise_advice;
+		__u32		splice_flags;
 	};
 	__u64	user_data;	/* data to be passed back at completion time */
 	union {
@@ -45,6 +49,7 @@ struct io_uring_sqe {
 			__u16	buf_index;
 			/* personality to use, if used */
 			__u16	personality;
+			__s32	splice_fd_in;
 		};
 		__u64	__pad2[3];
 	};
@@ -113,6 +118,7 @@ enum {
 	IORING_OP_RECV,
 	IORING_OP_OPENAT2,
 	IORING_OP_EPOLL_CTL,
+	IORING_OP_SPLICE,
 
 	/* this goes last, obviously */
 	IORING_OP_LAST,
@@ -128,6 +134,12 @@ enum {
  */
 #define IORING_TIMEOUT_ABS	(1U << 0)
 
+/*
+ * sqe->splice_flags
+ * extends splice(2) flags
+ */
+#define SPLICE_F_FD_IN_FIXED	(1U << 31) /* the last bit of __u32 */
+
 /*
  * IO completion data structure (Completion Queue Entry)
  */
-- 
cgit v1.2.3


From d7718a9d25a61442da8ee8aeeff6a0097f0ccfd6 Mon Sep 17 00:00:00 2001
From: Jens Axboe <axboe@kernel.dk>
Date: Fri, 14 Feb 2020 22:23:12 -0700
Subject: io_uring: use poll driven retry for files that support it

Currently io_uring tries any request in a non-blocking manner, if it can,
and then retries from a worker thread if we get -EAGAIN. Now that we have
a new and fancy poll based retry backend, use that to retry requests if
the file supports it.

This means that, for example, an IORING_OP_RECVMSG on a socket no longer
requires an async thread to complete the IO. If we get -EAGAIN reading
from the socket in a non-blocking manner, we arm a poll handler for
notification on when the socket becomes readable. When it does, the
pending read is executed directly by the task again, through the io_uring
task work handlers. Not only is this faster and more efficient, it also
means we're not generating potentially tons of async threads that just
sit and block, waiting for the IO to complete.

The feature is marked with IORING_FEAT_FAST_POLL, meaning that async
pollable IO is fast, and that poll<link>other_op is fast as well.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 fs/io_uring.c                   | 354 ++++++++++++++++++++++++++++++----------
 include/trace/events/io_uring.h | 103 ++++++++++++
 include/uapi/linux/io_uring.h   |   1 +
 3 files changed, 375 insertions(+), 83 deletions(-)

(limited to 'include/uapi/linux')

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 0d973de75127..8c976fde40bd 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -487,6 +487,7 @@ enum {
 	REQ_F_COMP_LOCKED_BIT,
 	REQ_F_NEED_CLEANUP_BIT,
 	REQ_F_OVERFLOW_BIT,
+	REQ_F_POLLED_BIT,
 };
 
 enum {
@@ -529,6 +530,13 @@ enum {
 	REQ_F_NEED_CLEANUP	= BIT(REQ_F_NEED_CLEANUP_BIT),
 	/* in overflow list */
 	REQ_F_OVERFLOW		= BIT(REQ_F_OVERFLOW_BIT),
+	/* already went through poll handler */
+	REQ_F_POLLED		= BIT(REQ_F_POLLED_BIT),
+};
+
+struct async_poll {
+	struct io_poll_iocb	poll;
+	struct io_wq_work	work;
 };
 
 /*
@@ -562,27 +570,29 @@ struct io_kiocb {
 	u8				opcode;
 
 	struct io_ring_ctx	*ctx;
-	union {
-		struct list_head	list;
-		struct hlist_node	hash_node;
-	};
-	struct list_head	link_list;
+	struct list_head	list;
 	unsigned int		flags;
 	refcount_t		refs;
+	struct task_struct	*task;
 	u64			user_data;
 	u32			result;
 	u32			sequence;
 
+	struct list_head	link_list;
+
 	struct list_head	inflight_entry;
 
 	union {
 		/*
 		 * Only commands that never go async can use the below fields,
-		 * obviously. Right now only IORING_OP_POLL_ADD uses them.
+		 * obviously. Right now only IORING_OP_POLL_ADD uses them, and
+		 * async armed poll handlers for regular commands. The latter
+		 * restore the work, if needed.
 		 */
 		struct {
-			struct task_struct	*task;
 			struct callback_head	task_work;
+			struct hlist_node	hash_node;
+			struct async_poll	*apoll;
 		};
 		struct io_wq_work	work;
 	};
@@ -3563,9 +3573,209 @@ out:
 #endif
 }
 
-static bool io_poll_remove_one(struct io_kiocb *req)
+struct io_poll_table {
+	struct poll_table_struct pt;
+	struct io_kiocb *req;
+	int error;
+};
+
+static void __io_queue_proc(struct io_poll_iocb *poll, struct io_poll_table *pt,
+			    struct wait_queue_head *head)
+{
+	if (unlikely(poll->head)) {
+		pt->error = -EINVAL;
+		return;
+	}
+
+	pt->error = 0;
+	poll->head = head;
+	add_wait_queue(head, &poll->wait);
+}
+
+static void io_async_queue_proc(struct file *file, struct wait_queue_head *head,
+			       struct poll_table_struct *p)
+{
+	struct io_poll_table *pt = container_of(p, struct io_poll_table, pt);
+
+	__io_queue_proc(&pt->req->apoll->poll, pt, head);
+}
+
+static int __io_async_wake(struct io_kiocb *req, struct io_poll_iocb *poll,
+			   __poll_t mask, task_work_func_t func)
+{
+	struct task_struct *tsk;
+
+	/* for instances that support it check for an event match first: */
+	if (mask && !(mask & poll->events))
+		return 0;
+
+	trace_io_uring_task_add(req->ctx, req->opcode, req->user_data, mask);
+
+	list_del_init(&poll->wait.entry);
+
+	tsk = req->task;
+	req->result = mask;
+	init_task_work(&req->task_work, func);
+	/*
+	 * If this fails, then the task is exiting. If that is the case, then
+	 * the exit check will ultimately cancel these work items. Hence we
+	 * don't need to check here and handle it specifically.
+	 */
+	task_work_add(tsk, &req->task_work, true);
+	wake_up_process(tsk);
+	return 1;
+}
+
+static void io_async_task_func(struct callback_head *cb)
+{
+	struct io_kiocb *req = container_of(cb, struct io_kiocb, task_work);
+	struct async_poll *apoll = req->apoll;
+	struct io_ring_ctx *ctx = req->ctx;
+
+	trace_io_uring_task_run(req->ctx, req->opcode, req->user_data);
+
+	WARN_ON_ONCE(!list_empty(&req->apoll->poll.wait.entry));
+
+	if (hash_hashed(&req->hash_node)) {
+		spin_lock_irq(&ctx->completion_lock);
+		hash_del(&req->hash_node);
+		spin_unlock_irq(&ctx->completion_lock);
+	}
+
+	/* restore ->work in case we need to retry again */
+	memcpy(&req->work, &apoll->work, sizeof(req->work));
+
+	__set_current_state(TASK_RUNNING);
+	mutex_lock(&ctx->uring_lock);
+	__io_queue_sqe(req, NULL);
+	mutex_unlock(&ctx->uring_lock);
+
+	kfree(apoll);
+}
+
+static int io_async_wake(struct wait_queue_entry *wait, unsigned mode, int sync,
+			void *key)
+{
+	struct io_kiocb *req = wait->private;
+	struct io_poll_iocb *poll = &req->apoll->poll;
+
+	trace_io_uring_poll_wake(req->ctx, req->opcode, req->user_data,
+					key_to_poll(key));
+
+	return __io_async_wake(req, poll, key_to_poll(key), io_async_task_func);
+}
+
+static void io_poll_req_insert(struct io_kiocb *req)
+{
+	struct io_ring_ctx *ctx = req->ctx;
+	struct hlist_head *list;
+
+	list = &ctx->cancel_hash[hash_long(req->user_data, ctx->cancel_hash_bits)];
+	hlist_add_head(&req->hash_node, list);
+}
+
+static __poll_t __io_arm_poll_handler(struct io_kiocb *req,
+				      struct io_poll_iocb *poll,
+				      struct io_poll_table *ipt, __poll_t mask,
+				      wait_queue_func_t wake_func)
+	__acquires(&ctx->completion_lock)
+{
+	struct io_ring_ctx *ctx = req->ctx;
+	bool cancel = false;
+
+	poll->file = req->file;
+	poll->head = NULL;
+	poll->done = poll->canceled = false;
+	poll->events = mask;
+
+	ipt->pt._key = mask;
+	ipt->req = req;
+	ipt->error = -EINVAL;
+
+	INIT_LIST_HEAD(&poll->wait.entry);
+	init_waitqueue_func_entry(&poll->wait, wake_func);
+	poll->wait.private = req;
+
+	mask = vfs_poll(req->file, &ipt->pt) & poll->events;
+
+	spin_lock_irq(&ctx->completion_lock);
+	if (likely(poll->head)) {
+		spin_lock(&poll->head->lock);
+		if (unlikely(list_empty(&poll->wait.entry))) {
+			if (ipt->error)
+				cancel = true;
+			ipt->error = 0;
+			mask = 0;
+		}
+		if (mask || ipt->error)
+			list_del_init(&poll->wait.entry);
+		else if (cancel)
+			WRITE_ONCE(poll->canceled, true);
+		else if (!poll->done) /* actually waiting for an event */
+			io_poll_req_insert(req);
+		spin_unlock(&poll->head->lock);
+	}
+
+	return mask;
+}
+
+static bool io_arm_poll_handler(struct io_kiocb *req)
+{
+	const struct io_op_def *def = &io_op_defs[req->opcode];
+	struct io_ring_ctx *ctx = req->ctx;
+	struct async_poll *apoll;
+	struct io_poll_table ipt;
+	__poll_t mask, ret;
+
+	if (!req->file || !file_can_poll(req->file))
+		return false;
+	if (req->flags & (REQ_F_MUST_PUNT | REQ_F_POLLED))
+		return false;
+	if (!def->pollin && !def->pollout)
+		return false;
+
+	apoll = kmalloc(sizeof(*apoll), GFP_ATOMIC);
+	if (unlikely(!apoll))
+		return false;
+
+	req->flags |= REQ_F_POLLED;
+	memcpy(&apoll->work, &req->work, sizeof(req->work));
+
+	/*
+	 * Don't need a reference here, as we're adding it to the task
+	 * task_works list. If the task exits, the list is pruned.
+	 */
+	req->task = current;
+	req->apoll = apoll;
+	INIT_HLIST_NODE(&req->hash_node);
+
+	if (def->pollin)
+		mask = POLLIN | POLLRDNORM;
+	if (def->pollout)
+		mask |= POLLOUT | POLLWRNORM;
+	mask |= POLLERR | POLLPRI;
+
+	ipt.pt._qproc = io_async_queue_proc;
+
+	ret = __io_arm_poll_handler(req, &apoll->poll, &ipt, mask,
+					io_async_wake);
+	if (ret) {
+		ipt.error = 0;
+		apoll->poll.done = true;
+		spin_unlock_irq(&ctx->completion_lock);
+		memcpy(&req->work, &apoll->work, sizeof(req->work));
+		kfree(apoll);
+		return false;
+	}
+	spin_unlock_irq(&ctx->completion_lock);
+	trace_io_uring_poll_arm(ctx, req->opcode, req->user_data, mask,
+					apoll->poll.events);
+	return true;
+}
+
+static bool __io_poll_remove_one(struct io_kiocb *req,
+				 struct io_poll_iocb *poll)
 {
-	struct io_poll_iocb *poll = &req->poll;
 	bool do_complete = false;
 
 	spin_lock(&poll->head->lock);
@@ -3575,7 +3785,24 @@ static bool io_poll_remove_one(struct io_kiocb *req)
 		do_complete = true;
 	}
 	spin_unlock(&poll->head->lock);
+	return do_complete;
+}
+
+static bool io_poll_remove_one(struct io_kiocb *req)
+{
+	bool do_complete;
+
+	if (req->opcode == IORING_OP_POLL_ADD) {
+		do_complete = __io_poll_remove_one(req, &req->poll);
+	} else {
+		/* non-poll requests have submit ref still */
+		do_complete = __io_poll_remove_one(req, &req->apoll->poll);
+		if (do_complete)
+			io_put_req(req);
+	}
+
 	hash_del(&req->hash_node);
+
 	if (do_complete) {
 		io_cqring_fill_event(req, -ECANCELED);
 		io_commit_cqring(req->ctx);
@@ -3686,8 +3913,13 @@ static void io_poll_task_func(struct callback_head *cb)
 	struct io_kiocb *nxt = NULL;
 
 	io_poll_task_handler(req, &nxt);
-	if (nxt)
+	if (nxt) {
+		struct io_ring_ctx *ctx = nxt->ctx;
+
+		mutex_lock(&ctx->uring_lock);
 		__io_queue_sqe(nxt, NULL);
+		mutex_unlock(&ctx->uring_lock);
+	}
 }
 
 static int io_poll_wake(struct wait_queue_entry *wait, unsigned mode, int sync,
@@ -3695,51 +3927,16 @@ static int io_poll_wake(struct wait_queue_entry *wait, unsigned mode, int sync,
 {
 	struct io_kiocb *req = wait->private;
 	struct io_poll_iocb *poll = &req->poll;
-	__poll_t mask = key_to_poll(key);
-	struct task_struct *tsk;
 
-	/* for instances that support it check for an event match first: */
-	if (mask && !(mask & poll->events))
-		return 0;
-
-	list_del_init(&poll->wait.entry);
-
-	tsk = req->task;
-	req->result = mask;
-	init_task_work(&req->task_work, io_poll_task_func);
-	task_work_add(tsk, &req->task_work, true);
-	wake_up_process(tsk);
-	return 1;
+	return __io_async_wake(req, poll, key_to_poll(key), io_poll_task_func);
 }
 
-struct io_poll_table {
-	struct poll_table_struct pt;
-	struct io_kiocb *req;
-	int error;
-};
-
 static void io_poll_queue_proc(struct file *file, struct wait_queue_head *head,
 			       struct poll_table_struct *p)
 {
 	struct io_poll_table *pt = container_of(p, struct io_poll_table, pt);
 
-	if (unlikely(pt->req->poll.head)) {
-		pt->error = -EINVAL;
-		return;
-	}
-
-	pt->error = 0;
-	pt->req->poll.head = head;
-	add_wait_queue(head, &pt->req->poll.wait);
-}
-
-static void io_poll_req_insert(struct io_kiocb *req)
-{
-	struct io_ring_ctx *ctx = req->ctx;
-	struct hlist_head *list;
-
-	list = &ctx->cancel_hash[hash_long(req->user_data, ctx->cancel_hash_bits)];
-	hlist_add_head(&req->hash_node, list);
+	__io_queue_proc(&pt->req->poll, pt, head);
 }
 
 static int io_poll_add_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
@@ -3757,7 +3954,10 @@ static int io_poll_add_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe
 	events = READ_ONCE(sqe->poll_events);
 	poll->events = demangle_poll(events) | EPOLLERR | EPOLLHUP;
 
-	/* task will wait for requests on exit, don't need a ref */
+	/*
+	 * Don't need a reference here, as we're adding it to the task
+	 * task_works list. If the task exits, the list is pruned.
+	 */
 	req->task = current;
 	return 0;
 }
@@ -3767,46 +3967,15 @@ static int io_poll_add(struct io_kiocb *req, struct io_kiocb **nxt)
 	struct io_poll_iocb *poll = &req->poll;
 	struct io_ring_ctx *ctx = req->ctx;
 	struct io_poll_table ipt;
-	bool cancel = false;
 	__poll_t mask;
 
 	INIT_HLIST_NODE(&req->hash_node);
-
-	poll->head = NULL;
-	poll->done = false;
-	poll->canceled = false;
-
-	ipt.pt._qproc = io_poll_queue_proc;
-	ipt.pt._key = poll->events;
-	ipt.req = req;
-	ipt.error = -EINVAL; /* same as no support for IOCB_CMD_POLL */
-
-	/* initialized the list so that we can do list_empty checks */
-	INIT_LIST_HEAD(&poll->wait.entry);
-	init_waitqueue_func_entry(&poll->wait, io_poll_wake);
-	poll->wait.private = req;
-
 	INIT_LIST_HEAD(&req->list);
+	ipt.pt._qproc = io_poll_queue_proc;
 
-	mask = vfs_poll(poll->file, &ipt.pt) & poll->events;
+	mask = __io_arm_poll_handler(req, &req->poll, &ipt, poll->events,
+					io_poll_wake);
 
-	spin_lock_irq(&ctx->completion_lock);
-	if (likely(poll->head)) {
-		spin_lock(&poll->head->lock);
-		if (unlikely(list_empty(&poll->wait.entry))) {
-			if (ipt.error)
-				cancel = true;
-			ipt.error = 0;
-			mask = 0;
-		}
-		if (mask || ipt.error)
-			list_del_init(&poll->wait.entry);
-		else if (cancel)
-			WRITE_ONCE(poll->canceled, true);
-		else if (!poll->done) /* actually waiting for an event */
-			io_poll_req_insert(req);
-		spin_unlock(&poll->head->lock);
-	}
 	if (mask) { /* no async, we'd stolen it */
 		ipt.error = 0;
 		io_poll_complete(req, mask, 0);
@@ -4751,6 +4920,9 @@ static struct io_kiocb *io_prep_linked_timeout(struct io_kiocb *req)
 
 	if (!(req->flags & REQ_F_LINK))
 		return NULL;
+	/* for polled retry, if flag is set, we already went through here */
+	if (req->flags & REQ_F_POLLED)
+		return NULL;
 
 	nxt = list_first_entry_or_null(&req->link_list, struct io_kiocb,
 					link_list);
@@ -4788,6 +4960,11 @@ again:
 	 */
 	if (ret == -EAGAIN && (!(req->flags & REQ_F_NOWAIT) ||
 	    (req->flags & REQ_F_MUST_PUNT))) {
+		if (io_arm_poll_handler(req)) {
+			if (linked_timeout)
+				io_queue_linked_timeout(linked_timeout);
+			goto done_req;
+		}
 punt:
 		if (io_op_defs[req->opcode].file_table) {
 			ret = io_grab_files(req);
@@ -6782,6 +6959,17 @@ static void __io_uring_show_fdinfo(struct io_ring_ctx *ctx, struct seq_file *m)
 		seq_printf(m, "Personalities:\n");
 		idr_for_each(&ctx->personality_idr, io_uring_show_cred, m);
 	}
+	seq_printf(m, "PollList:\n");
+	spin_lock_irq(&ctx->completion_lock);
+	for (i = 0; i < (1U << ctx->cancel_hash_bits); i++) {
+		struct hlist_head *list = &ctx->cancel_hash[i];
+		struct io_kiocb *req;
+
+		hlist_for_each_entry(req, list, hash_node)
+			seq_printf(m, "  op=%d, task_works=%d\n", req->opcode,
+					req->task->task_works != NULL);
+	}
+	spin_unlock_irq(&ctx->completion_lock);
 	mutex_unlock(&ctx->uring_lock);
 }
 
@@ -6998,7 +7186,7 @@ static int io_uring_create(unsigned entries, struct io_uring_params *p)
 
 	p->features = IORING_FEAT_SINGLE_MMAP | IORING_FEAT_NODROP |
 			IORING_FEAT_SUBMIT_STABLE | IORING_FEAT_RW_CUR_POS |
-			IORING_FEAT_CUR_PERSONALITY;
+			IORING_FEAT_CUR_PERSONALITY | IORING_FEAT_FAST_POLL;
 	trace_io_uring_create(ret, ctx, p->sq_entries, p->cq_entries, p->flags);
 	return ret;
 err:
diff --git a/include/trace/events/io_uring.h b/include/trace/events/io_uring.h
index 27bd9e4f927b..9f0d3b7d56b0 100644
--- a/include/trace/events/io_uring.h
+++ b/include/trace/events/io_uring.h
@@ -357,6 +357,109 @@ TRACE_EVENT(io_uring_submit_sqe,
 			  __entry->force_nonblock, __entry->sq_thread)
 );
 
+TRACE_EVENT(io_uring_poll_arm,
+
+	TP_PROTO(void *ctx, u8 opcode, u64 user_data, int mask, int events),
+
+	TP_ARGS(ctx, opcode, user_data, mask, events),
+
+	TP_STRUCT__entry (
+		__field(  void *,	ctx		)
+		__field(  u8,		opcode		)
+		__field(  u64,		user_data	)
+		__field(  int,		mask		)
+		__field(  int,		events		)
+	),
+
+	TP_fast_assign(
+		__entry->ctx		= ctx;
+		__entry->opcode		= opcode;
+		__entry->user_data	= user_data;
+		__entry->mask		= mask;
+		__entry->events		= events;
+	),
+
+	TP_printk("ring %p, op %d, data 0x%llx, mask 0x%x, events 0x%x",
+			  __entry->ctx, __entry->opcode,
+			  (unsigned long long) __entry->user_data,
+			  __entry->mask, __entry->events)
+);
+
+TRACE_EVENT(io_uring_poll_wake,
+
+	TP_PROTO(void *ctx, u8 opcode, u64 user_data, int mask),
+
+	TP_ARGS(ctx, opcode, user_data, mask),
+
+	TP_STRUCT__entry (
+		__field(  void *,	ctx		)
+		__field(  u8,		opcode		)
+		__field(  u64,		user_data	)
+		__field(  int,		mask		)
+	),
+
+	TP_fast_assign(
+		__entry->ctx		= ctx;
+		__entry->opcode		= opcode;
+		__entry->user_data	= user_data;
+		__entry->mask		= mask;
+	),
+
+	TP_printk("ring %p, op %d, data 0x%llx, mask 0x%x",
+			  __entry->ctx, __entry->opcode,
+			  (unsigned long long) __entry->user_data,
+			  __entry->mask)
+);
+
+TRACE_EVENT(io_uring_task_add,
+
+	TP_PROTO(void *ctx, u8 opcode, u64 user_data, int mask),
+
+	TP_ARGS(ctx, opcode, user_data, mask),
+
+	TP_STRUCT__entry (
+		__field(  void *,	ctx		)
+		__field(  u8,		opcode		)
+		__field(  u64,		user_data	)
+		__field(  int,		mask		)
+	),
+
+	TP_fast_assign(
+		__entry->ctx		= ctx;
+		__entry->opcode		= opcode;
+		__entry->user_data	= user_data;
+		__entry->mask		= mask;
+	),
+
+	TP_printk("ring %p, op %d, data 0x%llx, mask %x",
+			  __entry->ctx, __entry->opcode,
+			  (unsigned long long) __entry->user_data,
+			  __entry->mask)
+);
+
+TRACE_EVENT(io_uring_task_run,
+
+	TP_PROTO(void *ctx, u8 opcode, u64 user_data),
+
+	TP_ARGS(ctx, opcode, user_data),
+
+	TP_STRUCT__entry (
+		__field(  void *,	ctx		)
+		__field(  u8,		opcode		)
+		__field(  u64,		user_data	)
+	),
+
+	TP_fast_assign(
+		__entry->ctx		= ctx;
+		__entry->opcode		= opcode;
+		__entry->user_data	= user_data;
+	),
+
+	TP_printk("ring %p, op %d, data 0x%llx",
+			  __entry->ctx, __entry->opcode,
+			  (unsigned long long) __entry->user_data)
+);
+
 #endif /* _TRACE_IO_URING_H */
 
 /* This part must be outside protection */
diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 08891cc1c1e7..53b36311cdac 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -216,6 +216,7 @@ struct io_uring_params {
 #define IORING_FEAT_SUBMIT_STABLE	(1U << 2)
 #define IORING_FEAT_RW_CUR_POS		(1U << 3)
 #define IORING_FEAT_CUR_PERSONALITY	(1U << 4)
+#define IORING_FEAT_FAST_POLL		(1U << 5)
 
 /*
  * io_uring_register(2) opcodes and arguments
-- 
cgit v1.2.3


From ddf0322db79c5984dc1a1db890f946dd19b7d6d9 Mon Sep 17 00:00:00 2001
From: Jens Axboe <axboe@kernel.dk>
Date: Sun, 23 Feb 2020 16:41:33 -0700
Subject: io_uring: add IORING_OP_PROVIDE_BUFFERS

IORING_OP_PROVIDE_BUFFERS uses the buffer registration infrastructure to
support passing in an addr/len that is associated with a buffer ID and
buffer group ID. The group ID is used to index and lookup the buffers,
while the buffer ID can be used to notify the application which buffer
in the group was used. The addr passed in is the starting buffer address,
and length is each buffer length. A number of buffers to add with can be
specified, in which case addr is incremented by length for each addition,
and each buffer increments the buffer ID specified.

No validation is done of the buffer ID. If the application provides
buffers within the same group with identical buffer IDs, then it'll have
a hard time telling which buffer ID was used. The only restriction is
that the buffer ID can be a max of 16-bits in size, so USHRT_MAX is the
maximum ID that can be used.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 fs/io_uring.c                 | 138 +++++++++++++++++++++++++++++++++++++++++-
 include/uapi/linux/io_uring.h |  10 ++-
 2 files changed, 145 insertions(+), 3 deletions(-)

(limited to 'include/uapi/linux')

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 1f3ae208f6a6..1a58f2042815 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -447,6 +447,15 @@ struct io_splice {
 	unsigned int			flags;
 };
 
+struct io_provide_buf {
+	struct file			*file;
+	__u64				addr;
+	__s32				len;
+	__u32				bgid;
+	__u16				nbufs;
+	__u16				bid;
+};
+
 struct io_async_connect {
 	struct sockaddr_storage		address;
 };
@@ -572,6 +581,7 @@ struct io_kiocb {
 		struct io_madvise	madvise;
 		struct io_epoll		epoll;
 		struct io_splice	splice;
+		struct io_provide_buf	pbuf;
 	};
 
 	struct io_async_ctx		*io;
@@ -799,7 +809,8 @@ static const struct io_op_def io_op_defs[] = {
 		.needs_file		= 1,
 		.hash_reg_file		= 1,
 		.unbound_nonreg_file	= 1,
-	}
+	},
+	[IORING_OP_PROVIDE_BUFFERS] = {},
 };
 
 static void io_wq_submit_work(struct io_wq_work **workptr);
@@ -2785,6 +2796,120 @@ static int io_openat(struct io_kiocb *req, bool force_nonblock)
 	return io_openat2(req, force_nonblock);
 }
 
+static int io_provide_buffers_prep(struct io_kiocb *req,
+				   const struct io_uring_sqe *sqe)
+{
+	struct io_provide_buf *p = &req->pbuf;
+	u64 tmp;
+
+	if (sqe->ioprio || sqe->rw_flags)
+		return -EINVAL;
+
+	tmp = READ_ONCE(sqe->fd);
+	if (!tmp || tmp > USHRT_MAX)
+		return -E2BIG;
+	p->nbufs = tmp;
+	p->addr = READ_ONCE(sqe->addr);
+	p->len = READ_ONCE(sqe->len);
+
+	if (!access_ok(u64_to_user_ptr(p->addr), p->len))
+		return -EFAULT;
+
+	p->bgid = READ_ONCE(sqe->buf_group);
+	tmp = READ_ONCE(sqe->off);
+	if (tmp > USHRT_MAX)
+		return -E2BIG;
+	p->bid = tmp;
+	return 0;
+}
+
+static int io_add_buffers(struct io_provide_buf *pbuf, struct io_buffer **head)
+{
+	struct io_buffer *buf;
+	u64 addr = pbuf->addr;
+	int i, bid = pbuf->bid;
+
+	for (i = 0; i < pbuf->nbufs; i++) {
+		buf = kmalloc(sizeof(*buf), GFP_KERNEL);
+		if (!buf)
+			break;
+
+		buf->addr = addr;
+		buf->len = pbuf->len;
+		buf->bid = bid;
+		addr += pbuf->len;
+		bid++;
+		if (!*head) {
+			INIT_LIST_HEAD(&buf->list);
+			*head = buf;
+		} else {
+			list_add_tail(&buf->list, &(*head)->list);
+		}
+	}
+
+	return i ? i : -ENOMEM;
+}
+
+static void io_ring_submit_unlock(struct io_ring_ctx *ctx, bool needs_lock)
+{
+	if (needs_lock)
+		mutex_unlock(&ctx->uring_lock);
+}
+
+static void io_ring_submit_lock(struct io_ring_ctx *ctx, bool needs_lock)
+{
+	/*
+	 * "Normal" inline submissions always hold the uring_lock, since we
+	 * grab it from the system call. Same is true for the SQPOLL offload.
+	 * The only exception is when we've detached the request and issue it
+	 * from an async worker thread, grab the lock for that case.
+	 */
+	if (needs_lock)
+		mutex_lock(&ctx->uring_lock);
+}
+
+static int io_provide_buffers(struct io_kiocb *req, bool force_nonblock)
+{
+	struct io_provide_buf *p = &req->pbuf;
+	struct io_ring_ctx *ctx = req->ctx;
+	struct io_buffer *head, *list;
+	int ret = 0;
+
+	io_ring_submit_lock(ctx, !force_nonblock);
+
+	lockdep_assert_held(&ctx->uring_lock);
+
+	list = head = idr_find(&ctx->io_buffer_idr, p->bgid);
+
+	ret = io_add_buffers(p, &head);
+	if (ret < 0)
+		goto out;
+
+	if (!list) {
+		ret = idr_alloc(&ctx->io_buffer_idr, head, p->bgid, p->bgid + 1,
+					GFP_KERNEL);
+		if (ret < 0) {
+			while (!list_empty(&head->list)) {
+				struct io_buffer *buf;
+
+				buf = list_first_entry(&head->list,
+							struct io_buffer, list);
+				list_del(&buf->list);
+				kfree(buf);
+			}
+			kfree(head);
+			goto out;
+		}
+	}
+out:
+	io_ring_submit_unlock(ctx, !force_nonblock);
+	if (ret < 0)
+		req_set_fail_links(req);
+	io_cqring_add_event(req, ret);
+	io_put_req(req);
+	return 0;
+}
+
 static int io_epoll_ctl_prep(struct io_kiocb *req,
 			     const struct io_uring_sqe *sqe)
 {
@@ -4392,6 +4517,9 @@ static int io_req_defer_prep(struct io_kiocb *req,
 	case IORING_OP_SPLICE:
 		ret = io_splice_prep(req, sqe);
 		break;
+	case IORING_OP_PROVIDE_BUFFERS:
+		ret = io_provide_buffers_prep(req, sqe);
+		break;
 	default:
 		printk_once(KERN_WARNING "io_uring: unhandled opcode %d\n",
 				req->opcode);
@@ -4669,6 +4797,14 @@ static int io_issue_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 		}
 		ret = io_splice(req, force_nonblock);
 		break;
+	case IORING_OP_PROVIDE_BUFFERS:
+		if (sqe) {
+			ret = io_provide_buffers_prep(req, sqe);
+			if (ret)
+				break;
+		}
+		ret = io_provide_buffers(req, force_nonblock);
+		break;
 	default:
 		ret = -EINVAL;
 		break;
diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 53b36311cdac..bc34a57a660b 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -45,8 +45,13 @@ struct io_uring_sqe {
 	__u64	user_data;	/* data to be passed back at completion time */
 	union {
 		struct {
-			/* index into fixed buffers, if used */
-			__u16	buf_index;
+			/* pack this to avoid bogus arm OABI complaints */
+			union {
+				/* index into fixed buffers, if used */
+				__u16	buf_index;
+				/* for grouped buffer selection */
+				__u16	buf_group;
+			} __attribute__((packed));
 			/* personality to use, if used */
 			__u16	personality;
 			__s32	splice_fd_in;
@@ -119,6 +124,7 @@ enum {
 	IORING_OP_OPENAT2,
 	IORING_OP_EPOLL_CTL,
 	IORING_OP_SPLICE,
+	IORING_OP_PROVIDE_BUFFERS,
 
 	/* this goes last, obviously */
 	IORING_OP_LAST,
-- 
cgit v1.2.3


From bcda7baaa3f15c7a95db3c024bb046d6e298f76b Mon Sep 17 00:00:00 2001
From: Jens Axboe <axboe@kernel.dk>
Date: Sun, 23 Feb 2020 16:42:51 -0700
Subject: io_uring: support buffer selection for OP_READ and OP_RECV

If a server process has tons of pending socket connections, generally
it uses epoll to wait for activity. When the socket is ready for reading
(or writing), the task can select a buffer and issue a recv/send on the
given fd.

Now that we have fast (non-async thread) support, a task can have tons
of pending reads or writes pending. But that means they need buffers to
back that data, and if the number of connections is high enough, having
them preallocated for all possible connections is unfeasible.

With IORING_OP_PROVIDE_BUFFERS, an application can register buffers to
use for any request. The request then sets IOSQE_BUFFER_SELECT in the
sqe, and a given group ID in sqe->buf_group. When the fd becomes ready,
a free buffer from the specified group is selected. If none are
available, the request is terminated with -ENOBUFS. If successful, the
CQE on completion will contain the buffer ID chosen in the cqe->flags
member, encoded as:

	(buffer_id << IORING_CQE_BUFFER_SHIFT) | IORING_CQE_F_BUFFER;

Once a buffer has been consumed by a request, it is no longer available
and must be registered again with IORING_OP_PROVIDE_BUFFERS.

Requests need to support this feature. For now, IORING_OP_READ and
IORING_OP_RECV support it. This is checked on SQE submission, a CQE with
res == -EOPNOTSUPP will be posted if attempted on unsupported requests.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 fs/io_uring.c                 | 224 ++++++++++++++++++++++++++++++++++--------
 include/uapi/linux/io_uring.h |  14 +++
 2 files changed, 199 insertions(+), 39 deletions(-)

(limited to 'include/uapi/linux')

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 1a58f2042815..a80b5c189c14 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -395,7 +395,9 @@ struct io_sr_msg {
 		void __user		*buf;
 	};
 	int				msg_flags;
+	int				bgid;
 	size_t				len;
+	struct io_buffer		*kbuf;
 };
 
 struct io_open {
@@ -490,6 +492,7 @@ enum {
 	REQ_F_LINK_BIT		= IOSQE_IO_LINK_BIT,
 	REQ_F_HARDLINK_BIT	= IOSQE_IO_HARDLINK_BIT,
 	REQ_F_FORCE_ASYNC_BIT	= IOSQE_ASYNC_BIT,
+	REQ_F_BUFFER_SELECT_BIT	= IOSQE_BUFFER_SELECT_BIT,
 
 	REQ_F_LINK_NEXT_BIT,
 	REQ_F_FAIL_LINK_BIT,
@@ -506,6 +509,7 @@ enum {
 	REQ_F_NEED_CLEANUP_BIT,
 	REQ_F_OVERFLOW_BIT,
 	REQ_F_POLLED_BIT,
+	REQ_F_BUFFER_SELECTED_BIT,
 };
 
 enum {
@@ -519,6 +523,8 @@ enum {
 	REQ_F_HARDLINK		= BIT(REQ_F_HARDLINK_BIT),
 	/* IOSQE_ASYNC */
 	REQ_F_FORCE_ASYNC	= BIT(REQ_F_FORCE_ASYNC_BIT),
+	/* IOSQE_BUFFER_SELECT */
+	REQ_F_BUFFER_SELECT	= BIT(REQ_F_BUFFER_SELECT_BIT),
 
 	/* already grabbed next link */
 	REQ_F_LINK_NEXT		= BIT(REQ_F_LINK_NEXT_BIT),
@@ -550,6 +556,8 @@ enum {
 	REQ_F_OVERFLOW		= BIT(REQ_F_OVERFLOW_BIT),
 	/* already went through poll handler */
 	REQ_F_POLLED		= BIT(REQ_F_POLLED_BIT),
+	/* buffer already selected */
+	REQ_F_BUFFER_SELECTED	= BIT(REQ_F_BUFFER_SELECTED_BIT),
 };
 
 struct async_poll {
@@ -612,6 +620,7 @@ struct io_kiocb {
 			struct callback_head	task_work;
 			struct hlist_node	hash_node;
 			struct async_poll	*apoll;
+			int			cflags;
 		};
 		struct io_wq_work	work;
 	};
@@ -661,6 +670,8 @@ struct io_op_def {
 	/* set if opcode supports polled "wait" */
 	unsigned		pollin : 1;
 	unsigned		pollout : 1;
+	/* op supports buffer selection */
+	unsigned		buffer_select : 1;
 };
 
 static const struct io_op_def io_op_defs[] = {
@@ -770,6 +781,7 @@ static const struct io_op_def io_op_defs[] = {
 		.needs_file		= 1,
 		.unbound_nonreg_file	= 1,
 		.pollin			= 1,
+		.buffer_select		= 1,
 	},
 	[IORING_OP_WRITE] = {
 		.needs_mm		= 1,
@@ -794,6 +806,7 @@ static const struct io_op_def io_op_defs[] = {
 		.needs_file		= 1,
 		.unbound_nonreg_file	= 1,
 		.pollin			= 1,
+		.buffer_select		= 1,
 	},
 	[IORING_OP_OPENAT2] = {
 		.needs_file		= 1,
@@ -1170,7 +1183,7 @@ static bool io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force)
 		if (cqe) {
 			WRITE_ONCE(cqe->user_data, req->user_data);
 			WRITE_ONCE(cqe->res, req->result);
-			WRITE_ONCE(cqe->flags, 0);
+			WRITE_ONCE(cqe->flags, req->cflags);
 		} else {
 			WRITE_ONCE(ctx->rings->cq_overflow,
 				atomic_inc_return(&ctx->cached_cq_overflow));
@@ -1194,7 +1207,7 @@ static bool io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force)
 	return cqe != NULL;
 }
 
-static void io_cqring_fill_event(struct io_kiocb *req, long res)
+static void __io_cqring_fill_event(struct io_kiocb *req, long res, long cflags)
 {
 	struct io_ring_ctx *ctx = req->ctx;
 	struct io_uring_cqe *cqe;
@@ -1210,7 +1223,7 @@ static void io_cqring_fill_event(struct io_kiocb *req, long res)
 	if (likely(cqe)) {
 		WRITE_ONCE(cqe->user_data, req->user_data);
 		WRITE_ONCE(cqe->res, res);
-		WRITE_ONCE(cqe->flags, 0);
+		WRITE_ONCE(cqe->flags, cflags);
 	} else if (ctx->cq_overflow_flushed) {
 		WRITE_ONCE(ctx->rings->cq_overflow,
 				atomic_inc_return(&ctx->cached_cq_overflow));
@@ -1222,23 +1235,34 @@ static void io_cqring_fill_event(struct io_kiocb *req, long res)
 		req->flags |= REQ_F_OVERFLOW;
 		refcount_inc(&req->refs);
 		req->result = res;
+		req->cflags = cflags;
 		list_add_tail(&req->list, &ctx->cq_overflow_list);
 	}
 }
 
-static void io_cqring_add_event(struct io_kiocb *req, long res)
+static void io_cqring_fill_event(struct io_kiocb *req, long res)
+{
+	__io_cqring_fill_event(req, res, 0);
+}
+
+static void __io_cqring_add_event(struct io_kiocb *req, long res, long cflags)
 {
 	struct io_ring_ctx *ctx = req->ctx;
 	unsigned long flags;
 
 	spin_lock_irqsave(&ctx->completion_lock, flags);
-	io_cqring_fill_event(req, res);
+	__io_cqring_fill_event(req, res, cflags);
 	io_commit_cqring(ctx);
 	spin_unlock_irqrestore(&ctx->completion_lock, flags);
 
 	io_cqring_ev_posted(ctx);
 }
 
+static void io_cqring_add_event(struct io_kiocb *req, long res)
+{
+	__io_cqring_add_event(req, res, 0);
+}
+
 static inline bool io_is_fallback_req(struct io_kiocb *req)
 {
 	return req == (struct io_kiocb *)
@@ -1660,6 +1684,18 @@ static inline bool io_req_multi_free(struct req_batch *rb, struct io_kiocb *req)
 	return true;
 }
 
+static int io_put_kbuf(struct io_kiocb *req)
+{
+	struct io_buffer *kbuf = (struct io_buffer *) req->rw.addr;
+	int cflags;
+
+	cflags = kbuf->bid << IORING_CQE_BUFFER_SHIFT;
+	cflags |= IORING_CQE_F_BUFFER;
+	req->rw.addr = 0;
+	kfree(kbuf);
+	return cflags;
+}
+
 /*
  * Find and free completed poll iocbs
  */
@@ -1671,10 +1707,15 @@ static void io_iopoll_complete(struct io_ring_ctx *ctx, unsigned int *nr_events,
 
 	rb.to_free = rb.need_iter = 0;
 	while (!list_empty(done)) {
+		int cflags = 0;
+
 		req = list_first_entry(done, struct io_kiocb, list);
 		list_del(&req->list);
 
-		io_cqring_fill_event(req, req->result);
+		if (req->flags & REQ_F_BUFFER_SELECTED)
+			cflags = io_put_kbuf(req);
+
+		__io_cqring_fill_event(req, req->result, cflags);
 		(*nr_events)++;
 
 		if (refcount_dec_and_test(&req->refs) &&
@@ -1849,13 +1890,16 @@ static inline void req_set_fail_links(struct io_kiocb *req)
 static void io_complete_rw_common(struct kiocb *kiocb, long res)
 {
 	struct io_kiocb *req = container_of(kiocb, struct io_kiocb, rw.kiocb);
+	int cflags = 0;
 
 	if (kiocb->ki_flags & IOCB_WRITE)
 		kiocb_end_write(req);
 
 	if (res != req->result)
 		req_set_fail_links(req);
-	io_cqring_add_event(req, res);
+	if (req->flags & REQ_F_BUFFER_SELECTED)
+		cflags = io_put_kbuf(req);
+	__io_cqring_add_event(req, res, cflags);
 }
 
 static void io_complete_rw(struct kiocb *kiocb, long res, long res2)
@@ -2033,7 +2077,7 @@ static int io_prep_rw(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 
 	req->rw.addr = READ_ONCE(sqe->addr);
 	req->rw.len = READ_ONCE(sqe->len);
-	/* we own ->private, reuse it for the buffer index */
+	/* we own ->private, reuse it for the buffer index  / buffer ID */
 	req->rw.kiocb.private = (void *) (unsigned long)
 					READ_ONCE(sqe->buf_index);
 	return 0;
@@ -2146,8 +2190,61 @@ static ssize_t io_import_fixed(struct io_kiocb *req, int rw,
 	return len;
 }
 
+static void io_ring_submit_unlock(struct io_ring_ctx *ctx, bool needs_lock)
+{
+	if (needs_lock)
+		mutex_unlock(&ctx->uring_lock);
+}
+
+static void io_ring_submit_lock(struct io_ring_ctx *ctx, bool needs_lock)
+{
+	/*
+	 * "Normal" inline submissions always hold the uring_lock, since we
+	 * grab it from the system call. Same is true for the SQPOLL offload.
+	 * The only exception is when we've detached the request and issue it
+	 * from an async worker thread, grab the lock for that case.
+	 */
+	if (needs_lock)
+		mutex_lock(&ctx->uring_lock);
+}
+
+static struct io_buffer *io_buffer_select(struct io_kiocb *req, size_t *len,
+					  int bgid, struct io_buffer *kbuf,
+					  bool needs_lock)
+{
+	struct io_buffer *head;
+
+	if (req->flags & REQ_F_BUFFER_SELECTED)
+		return kbuf;
+
+	io_ring_submit_lock(req->ctx, needs_lock);
+
+	lockdep_assert_held(&req->ctx->uring_lock);
+
+	head = idr_find(&req->ctx->io_buffer_idr, bgid);
+	if (head) {
+		if (!list_empty(&head->list)) {
+			kbuf = list_last_entry(&head->list, struct io_buffer,
+							list);
+			list_del(&kbuf->list);
+		} else {
+			kbuf = head;
+			idr_remove(&req->ctx->io_buffer_idr, bgid);
+		}
+		if (*len > kbuf->len)
+			*len = kbuf->len;
+	} else {
+		kbuf = ERR_PTR(-ENOBUFS);
+	}
+
+	io_ring_submit_unlock(req->ctx, needs_lock);
+
+	return kbuf;
+}
+
 static ssize_t io_import_iovec(int rw, struct io_kiocb *req,
-			       struct iovec **iovec, struct iov_iter *iter)
+			       struct iovec **iovec, struct iov_iter *iter,
+			       bool needs_lock)
 {
 	void __user *buf = u64_to_user_ptr(req->rw.addr);
 	size_t sqe_len = req->rw.len;
@@ -2159,12 +2256,29 @@ static ssize_t io_import_iovec(int rw, struct io_kiocb *req,
 		return io_import_fixed(req, rw, iter);
 	}
 
-	/* buffer index only valid with fixed read/write */
-	if (req->rw.kiocb.private)
+	/* buffer index only valid with fixed read/write, or buffer select  */
+	if (req->rw.kiocb.private && !(req->flags & REQ_F_BUFFER_SELECT))
 		return -EINVAL;
 
 	if (opcode == IORING_OP_READ || opcode == IORING_OP_WRITE) {
 		ssize_t ret;
+
+		if (req->flags & REQ_F_BUFFER_SELECT) {
+			struct io_buffer *kbuf = (struct io_buffer *) req->rw.addr;
+			int bgid;
+
+			bgid = (int) (unsigned long) req->rw.kiocb.private;
+			kbuf = io_buffer_select(req, &sqe_len, bgid, kbuf,
+						needs_lock);
+			if (IS_ERR(kbuf)) {
+				*iovec = NULL;
+				return PTR_ERR(kbuf);
+			}
+			req->rw.addr = (u64) kbuf;
+			req->flags |= REQ_F_BUFFER_SELECTED;
+			buf = u64_to_user_ptr(kbuf->addr);
+		}
+
 		ret = import_single_range(rw, buf, sqe_len, *iovec, iter);
 		*iovec = NULL;
 		return ret < 0 ? ret : sqe_len;
@@ -2307,7 +2421,7 @@ static int io_read_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 	io = req->io;
 	io->rw.iov = io->rw.fast_iov;
 	req->io = NULL;
-	ret = io_import_iovec(READ, req, &io->rw.iov, &iter);
+	ret = io_import_iovec(READ, req, &io->rw.iov, &iter, !force_nonblock);
 	req->io = io;
 	if (ret < 0)
 		return ret;
@@ -2324,7 +2438,7 @@ static int io_read(struct io_kiocb *req, bool force_nonblock)
 	size_t iov_count;
 	ssize_t io_size, ret;
 
-	ret = io_import_iovec(READ, req, &iovec, &iter);
+	ret = io_import_iovec(READ, req, &iovec, &iter, !force_nonblock);
 	if (ret < 0)
 		return ret;
 
@@ -2396,7 +2510,7 @@ static int io_write_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 	io = req->io;
 	io->rw.iov = io->rw.fast_iov;
 	req->io = NULL;
-	ret = io_import_iovec(WRITE, req, &io->rw.iov, &iter);
+	ret = io_import_iovec(WRITE, req, &io->rw.iov, &iter, !force_nonblock);
 	req->io = io;
 	if (ret < 0)
 		return ret;
@@ -2413,7 +2527,7 @@ static int io_write(struct io_kiocb *req, bool force_nonblock)
 	size_t iov_count;
 	ssize_t ret, io_size;
 
-	ret = io_import_iovec(WRITE, req, &iovec, &iter);
+	ret = io_import_iovec(WRITE, req, &iovec, &iter, !force_nonblock);
 	if (ret < 0)
 		return ret;
 
@@ -2850,24 +2964,6 @@ static int io_add_buffers(struct io_provide_buf *pbuf, struct io_buffer **head)
 	return i ? i : -ENOMEM;
 }
 
-static void io_ring_submit_unlock(struct io_ring_ctx *ctx, bool needs_lock)
-{
-	if (needs_lock)
-		mutex_unlock(&ctx->uring_lock);
-}
-
-static void io_ring_submit_lock(struct io_ring_ctx *ctx, bool needs_lock)
-{
-	/*
-	 * "Normal" inline submissions always hold the uring_lock, since we
-	 * grab it from the system call. Same is true for the SQPOLL offload.
-	 * The only exception is when we've detached the request and issue it
-	 * from an async worker thread, grab the lock for that case.
-	 */
-	if (needs_lock)
-		mutex_lock(&ctx->uring_lock);
-}
-
 static int io_provide_buffers(struct io_kiocb *req, bool force_nonblock)
 {
 	struct io_provide_buf *p = &req->pbuf;
@@ -3390,6 +3486,27 @@ static int io_send(struct io_kiocb *req, bool force_nonblock)
 #endif
 }
 
+static struct io_buffer *io_recv_buffer_select(struct io_kiocb *req,
+					       int *cflags, bool needs_lock)
+{
+	struct io_sr_msg *sr = &req->sr_msg;
+	struct io_buffer *kbuf;
+
+	if (!(req->flags & REQ_F_BUFFER_SELECT))
+		return NULL;
+
+	kbuf = io_buffer_select(req, &sr->len, sr->bgid, sr->kbuf, needs_lock);
+	if (IS_ERR(kbuf))
+		return kbuf;
+
+	sr->kbuf = kbuf;
+	req->flags |= REQ_F_BUFFER_SELECTED;
+
+	*cflags = kbuf->bid << IORING_CQE_BUFFER_SHIFT;
+	*cflags |= IORING_CQE_F_BUFFER;
+	return kbuf;
+}
+
 static int io_recvmsg_prep(struct io_kiocb *req,
 			   const struct io_uring_sqe *sqe)
 {
@@ -3401,6 +3518,7 @@ static int io_recvmsg_prep(struct io_kiocb *req,
 	sr->msg_flags = READ_ONCE(sqe->msg_flags);
 	sr->msg = u64_to_user_ptr(READ_ONCE(sqe->addr));
 	sr->len = READ_ONCE(sqe->len);
+	sr->bgid = READ_ONCE(sqe->buf_group);
 
 #ifdef CONFIG_COMPAT
 	if (req->ctx->compat)
@@ -3490,8 +3608,9 @@ static int io_recvmsg(struct io_kiocb *req, bool force_nonblock)
 static int io_recv(struct io_kiocb *req, bool force_nonblock)
 {
 #if defined(CONFIG_NET)
+	struct io_buffer *kbuf = NULL;
 	struct socket *sock;
-	int ret;
+	int ret, cflags = 0;
 
 	if (unlikely(req->ctx->flags & IORING_SETUP_IOPOLL))
 		return -EINVAL;
@@ -3499,15 +3618,25 @@ static int io_recv(struct io_kiocb *req, bool force_nonblock)
 	sock = sock_from_file(req->file, &ret);
 	if (sock) {
 		struct io_sr_msg *sr = &req->sr_msg;
+		void __user *buf = sr->buf;
 		struct msghdr msg;
 		struct iovec iov;
 		unsigned flags;
 
-		ret = import_single_range(READ, sr->buf, sr->len, &iov,
+		kbuf = io_recv_buffer_select(req, &cflags, !force_nonblock);
+		if (IS_ERR(kbuf))
+			return PTR_ERR(kbuf);
+		else if (kbuf)
+			buf = u64_to_user_ptr(kbuf->addr);
+
+		ret = import_single_range(READ, buf, sr->len, &iov,
 						&msg.msg_iter);
-		if (ret)
+		if (ret) {
+			kfree(kbuf);
 			return ret;
+		}
 
+		req->flags |= REQ_F_NEED_CLEANUP;
 		msg.msg_name = NULL;
 		msg.msg_control = NULL;
 		msg.msg_controllen = 0;
@@ -3528,7 +3657,9 @@ static int io_recv(struct io_kiocb *req, bool force_nonblock)
 			ret = -EINTR;
 	}
 
-	io_cqring_add_event(req, ret);
+	kfree(kbuf);
+	req->flags &= ~REQ_F_NEED_CLEANUP;
+	__io_cqring_add_event(req, ret, cflags);
 	if (ret < 0)
 		req_set_fail_links(req);
 	io_put_req(req);
@@ -4566,6 +4697,9 @@ static void io_cleanup_req(struct io_kiocb *req)
 	case IORING_OP_READV:
 	case IORING_OP_READ_FIXED:
 	case IORING_OP_READ:
+		if (req->flags & REQ_F_BUFFER_SELECTED)
+			kfree((void *)(unsigned long)req->rw.addr);
+		/* fallthrough */
 	case IORING_OP_WRITEV:
 	case IORING_OP_WRITE_FIXED:
 	case IORING_OP_WRITE:
@@ -4577,6 +4711,10 @@ static void io_cleanup_req(struct io_kiocb *req)
 		if (io->msg.iov != io->msg.fast_iov)
 			kfree(io->msg.iov);
 		break;
+	case IORING_OP_RECV:
+		if (req->flags & REQ_F_BUFFER_SELECTED)
+			kfree(req->sr_msg.kbuf);
+		break;
 	case IORING_OP_OPENAT:
 	case IORING_OP_OPENAT2:
 	case IORING_OP_STATX:
@@ -5154,7 +5292,8 @@ static inline void io_queue_link_head(struct io_kiocb *req)
 }
 
 #define SQE_VALID_FLAGS	(IOSQE_FIXED_FILE|IOSQE_IO_DRAIN|IOSQE_IO_LINK|	\
-				IOSQE_IO_HARDLINK | IOSQE_ASYNC)
+				IOSQE_IO_HARDLINK | IOSQE_ASYNC | \
+				IOSQE_BUFFER_SELECT)
 
 static bool io_submit_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 			  struct io_submit_state *state, struct io_kiocb **link)
@@ -5171,6 +5310,12 @@ static bool io_submit_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 		goto err_req;
 	}
 
+	if ((sqe_flags & IOSQE_BUFFER_SELECT) &&
+	    !io_op_defs[req->opcode].buffer_select) {
+		ret = -EOPNOTSUPP;
+		goto err_req;
+	}
+
 	id = READ_ONCE(sqe->personality);
 	if (id) {
 		req->work.creds = idr_find(&ctx->personality_idr, id);
@@ -5183,7 +5328,8 @@ static bool io_submit_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 
 	/* same numerical values with corresponding REQ_F_*, safe to copy */
 	req->flags |= sqe_flags & (IOSQE_IO_DRAIN | IOSQE_IO_HARDLINK |
-					IOSQE_ASYNC | IOSQE_FIXED_FILE);
+					IOSQE_ASYNC | IOSQE_FIXED_FILE |
+					IOSQE_BUFFER_SELECT);
 
 	ret = io_req_set_file(state, req, sqe);
 	if (unlikely(ret)) {
diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index bc34a57a660b..9b263d9b24e6 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -66,6 +66,7 @@ enum {
 	IOSQE_IO_LINK_BIT,
 	IOSQE_IO_HARDLINK_BIT,
 	IOSQE_ASYNC_BIT,
+	IOSQE_BUFFER_SELECT_BIT,
 };
 
 /*
@@ -81,6 +82,8 @@ enum {
 #define IOSQE_IO_HARDLINK	(1U << IOSQE_IO_HARDLINK_BIT)
 /* always go async */
 #define IOSQE_ASYNC		(1U << IOSQE_ASYNC_BIT)
+/* select buffer from sqe->buf_group */
+#define IOSQE_BUFFER_SELECT	(1U << IOSQE_BUFFER_SELECT_BIT)
 
 /*
  * io_uring_setup() flags
@@ -155,6 +158,17 @@ struct io_uring_cqe {
 	__u32	flags;
 };
 
+/*
+ * cqe->flags
+ *
+ * IORING_CQE_F_BUFFER	If set, the upper 16 bits are the buffer ID
+ */
+#define IORING_CQE_F_BUFFER		(1U << 0)
+
+enum {
+	IORING_CQE_BUFFER_SHIFT		= 16,
+};
+
 /*
  * Magic offsets for the application to mmap the data it needs
  */
-- 
cgit v1.2.3


From 067524e914cb23e20d59480b318fe2625eaee7c8 Mon Sep 17 00:00:00 2001
From: Jens Axboe <axboe@kernel.dk>
Date: Mon, 2 Mar 2020 16:32:28 -0700
Subject: io_uring: provide means of removing buffers

We have IORING_OP_PROVIDE_BUFFERS, but the only way to remove buffers
is to trigger IO on them. The usual case of shrinking a buffer pool
would be to just not replenish the buffers when IO completes, and
instead just free it. But it may be nice to have a way to manually
remove a number of buffers from a given group, and
IORING_OP_REMOVE_BUFFERS provides that functionality.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 fs/io_uring.c                 | 102 ++++++++++++++++++++++++++++++++++--------
 include/uapi/linux/io_uring.h |   1 +
 2 files changed, 84 insertions(+), 19 deletions(-)

(limited to 'include/uapi/linux')

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 455d53fd840f..f131105fa12a 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -827,6 +827,7 @@ static const struct io_op_def io_op_defs[] = {
 		.unbound_nonreg_file	= 1,
 	},
 	[IORING_OP_PROVIDE_BUFFERS] = {},
+	[IORING_OP_REMOVE_BUFFERS] = {},
 };
 
 static void io_wq_submit_work(struct io_wq_work **workptr);
@@ -2995,6 +2996,75 @@ static int io_openat(struct io_kiocb *req, bool force_nonblock)
 	return io_openat2(req, force_nonblock);
 }
 
+static int io_remove_buffers_prep(struct io_kiocb *req,
+				  const struct io_uring_sqe *sqe)
+{
+	struct io_provide_buf *p = &req->pbuf;
+	u64 tmp;
+
+	if (sqe->ioprio || sqe->rw_flags || sqe->addr || sqe->len || sqe->off)
+		return -EINVAL;
+
+	tmp = READ_ONCE(sqe->fd);
+	if (!tmp || tmp > USHRT_MAX)
+		return -EINVAL;
+
+	memset(p, 0, sizeof(*p));
+	p->nbufs = tmp;
+	p->bgid = READ_ONCE(sqe->buf_group);
+	return 0;
+}
+
+static int __io_remove_buffers(struct io_ring_ctx *ctx, struct io_buffer *buf,
+			       int bgid, unsigned nbufs)
+{
+	unsigned i = 0;
+
+	/* shouldn't happen */
+	if (!nbufs)
+		return 0;
+
+	/* the head kbuf is the list itself */
+	while (!list_empty(&buf->list)) {
+		struct io_buffer *nxt;
+
+		nxt = list_first_entry(&buf->list, struct io_buffer, list);
+		list_del(&nxt->list);
+		kfree(nxt);
+		if (++i == nbufs)
+			return i;
+	}
+	i++;
+	kfree(buf);
+	idr_remove(&ctx->io_buffer_idr, bgid);
+
+	return i;
+}
+
+static int io_remove_buffers(struct io_kiocb *req, bool force_nonblock)
+{
+	struct io_provide_buf *p = &req->pbuf;
+	struct io_ring_ctx *ctx = req->ctx;
+	struct io_buffer *head;
+	int ret = 0;
+
+	io_ring_submit_lock(ctx, !force_nonblock);
+
+	lockdep_assert_held(&ctx->uring_lock);
+
+	ret = -ENOENT;
+	head = idr_find(&ctx->io_buffer_idr, p->bgid);
+	if (head)
+		ret = __io_remove_buffers(ctx, head, p->bgid, p->nbufs);
+
+	io_ring_submit_lock(ctx, !force_nonblock);
+	if (ret < 0)
+		req_set_fail_links(req);
+	io_cqring_add_event(req, ret);
+	io_put_req(req);
+	return 0;
+}
+
 static int io_provide_buffers_prep(struct io_kiocb *req,
 				   const struct io_uring_sqe *sqe)
 {
@@ -3070,15 +3140,7 @@ static int io_provide_buffers(struct io_kiocb *req, bool force_nonblock)
 		ret = idr_alloc(&ctx->io_buffer_idr, head, p->bgid, p->bgid + 1,
 					GFP_KERNEL);
 		if (ret < 0) {
-			while (!list_empty(&head->list)) {
-				struct io_buffer *buf;
-
-				buf = list_first_entry(&head->list,
-							struct io_buffer, list);
-				list_del(&buf->list);
-				kfree(buf);
-			}
-			kfree(head);
+			__io_remove_buffers(ctx, head, p->bgid, -1U);
 			goto out;
 		}
 	}
@@ -4825,6 +4887,9 @@ static int io_req_defer_prep(struct io_kiocb *req,
 	case IORING_OP_PROVIDE_BUFFERS:
 		ret = io_provide_buffers_prep(req, sqe);
 		break;
+	case IORING_OP_REMOVE_BUFFERS:
+		ret = io_remove_buffers_prep(req, sqe);
+		break;
 	default:
 		printk_once(KERN_WARNING "io_uring: unhandled opcode %d\n",
 				req->opcode);
@@ -5120,6 +5185,14 @@ static int io_issue_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe,
 		}
 		ret = io_provide_buffers(req, force_nonblock);
 		break;
+	case IORING_OP_REMOVE_BUFFERS:
+		if (sqe) {
+			ret = io_remove_buffers_prep(req, sqe);
+			if (ret)
+				break;
+		}
+		ret = io_remove_buffers(req, force_nonblock);
+		break;
 	default:
 		ret = -EINVAL;
 		break;
@@ -6998,16 +7071,7 @@ static int __io_destroy_buffers(int id, void *p, void *data)
 	struct io_ring_ctx *ctx = data;
 	struct io_buffer *buf = p;
 
-	/* the head kbuf is the list itself */
-	while (!list_empty(&buf->list)) {
-		struct io_buffer *nxt;
-
-		nxt = list_first_entry(&buf->list, struct io_buffer, list);
-		list_del(&nxt->list);
-		kfree(nxt);
-	}
-	kfree(buf);
-	idr_remove(&ctx->io_buffer_idr, id);
+	__io_remove_buffers(ctx, buf, id, -1U);
 	return 0;
 }
 
diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 9b263d9b24e6..cef4c0c0f26b 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -128,6 +128,7 @@ enum {
 	IORING_OP_EPOLL_CTL,
 	IORING_OP_SPLICE,
 	IORING_OP_PROVIDE_BUFFERS,
+	IORING_OP_REMOVE_BUFFERS,
 
 	/* this goes last, obviously */
 	IORING_OP_LAST,
-- 
cgit v1.2.3


From bbbdeb4720a0759ec90e3bcb20ad28d19e531346 Mon Sep 17 00:00:00 2001
From: Jens Axboe <axboe@kernel.dk>
Date: Wed, 11 Mar 2020 07:45:46 -0600
Subject: io_uring: dual license io_uring.h uapi header

This just syncs the header it with the liburing version, so there's no
confusion on the license of the header parts.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/uapi/linux/io_uring.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/uapi/linux')

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index cef4c0c0f26b..6d9d2b1cc523 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note OR MIT */
 /*
  * Header file for the io_uring interface.
  *
-- 
cgit v1.2.3


From 9f5834c868e901b00f1bfe4d0052b5906b4a2b7f Mon Sep 17 00:00:00 2001
From: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Date: Sat, 21 Mar 2020 12:19:07 +0100
Subject: io_uring: make spdxcheck.py happy

Commit bbbdeb4720a0 ("io_uring: dual license io_uring.h uapi header")
uses a nested SPDX-License-Identifier to dual license the header.

Since then, ./scripts/spdxcheck.py complains:

  include/uapi/linux/io_uring.h: 1:60 Missing parentheses: OR

Add parentheses to make spdxcheck.py happy.

Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/uapi/linux/io_uring.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/uapi/linux')

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 6d9d2b1cc523..e48d746b8e2a 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -1,4 +1,4 @@
-/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note OR MIT */
+/* SPDX-License-Identifier: (GPL-2.0 WITH Linux-syscall-note) OR MIT */
 /*
  * Header file for the io_uring interface.
  *
-- 
cgit v1.2.3