[PATCHSET v2 0/3] Support for provided buffers for send

[PATCHSET v2 0/3] Support for provided buffers for send

[Jens Axboe]

Hi,

We never supported provided buffers for sends, because it didn't seem
to make a lot of sense. But it actually does make a lot of sense! If
an app is receiving data, doing something with it, and then sending
either the same or another buffer out based on that, then if we use
provided buffers for sends we can guarantee that the sends are
serialized. This is because provided buffer rings are FIFO ordered,
as it's a ring buffer, and hence it doesn't really matter if you
have more than one send inflight.

This provides a nice efficiency win, but more importantly, it reduces
the complexity in the application as it no longer needs to track a
potential backlog of sends. The app just sets up a send based buffer
ring, exactly like it does for incoming data. And that's it, no more
dealing with serialized sends.

In some testing with proxy [1], in basic shuffling of packets I see
a 36% improvement with this over manually dealing with sends. That's
a pretty big win on top of making the app simpler.

This also opens the door for multishot send requests, which is an
interesting future venue to pursue!

You can also find the patches here:

https://git.kernel.dk/cgit/linux/log/?h=io_uring-send-queue

[1] https://git.kernel.dk/cgit/liburing/tree/examples/proxy.c

Changes since v1:

- Forgot that the sendmsg side uses the generic helper for this, whereas
  recvmsg rolls its own to deal with provided buffers. Unify them in patch
  1 instead.

 include/uapi/linux/io_uring.h |   1 +
 io_uring/io_uring.c           |   3 +-
 io_uring/net.c                | 309 +++++++++++++++++++---------------
 io_uring/opdef.c              |   2 +
 4 files changed, 175 insertions(+), 140 deletions(-)

-- 
Jens Axboe

[PATCH 1/3] io_uring/net: unify how recvmsg and sendmsg copy in the msghdr

[Jens Axboe]

For recvmsg, we roll our own since we support buffer selections. This
isn't the case for sendmsg right now, but in preparation for doing so,
make the recvmsg copy helpers generic so we can call them from the
sendmsg side as well.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 io_uring/net.c | 275 +++++++++++++++++++++++++------------------------
 1 file changed, 140 insertions(+), 135 deletions(-)

diff --git a/io_uring/net.c b/io_uring/net.c
index 161622029147..fcbaeb7cc045 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -204,16 +204,148 @@ static int io_setup_async_msg(struct io_kiocb *req,
 	return -EAGAIN;
 }
 
+static bool io_recvmsg_multishot_overflow(struct io_async_msghdr *iomsg)
+{
+	int hdr;
+
+	if (iomsg->namelen < 0)
+		return true;
+	if (check_add_overflow((int)sizeof(struct io_uring_recvmsg_out),
+			       iomsg->namelen, &hdr))
+		return true;
+	if (check_add_overflow(hdr, (int)iomsg->controllen, &hdr))
+		return true;
+
+	return false;
+}
+
+#ifdef CONFIG_COMPAT
+static int __io_compat_msg_copy_hdr(struct io_kiocb *req,
+				    struct io_async_msghdr *iomsg,
+				    struct sockaddr __user **addr, int ddir)
+{
+	struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg);
+	struct compat_msghdr msg;
+	struct compat_iovec __user *uiov;
+	int ret;
+
+	if (copy_from_user(&msg, sr->umsg_compat, sizeof(msg)))
+		return -EFAULT;
+
+	ret = __get_compat_msghdr(&iomsg->msg, &msg, addr);
+	if (ret)
+		return ret;
+
+	uiov = compat_ptr(msg.msg_iov);
+	if (req->flags & REQ_F_BUFFER_SELECT) {
+		compat_ssize_t clen;
+
+		iomsg->free_iov = NULL;
+		if (msg.msg_iovlen == 0) {
+			sr->len = 0;
+		} else if (msg.msg_iovlen > 1) {
+			return -EINVAL;
+		} else {
+			if (!access_ok(uiov, sizeof(*uiov)))
+				return -EFAULT;
+			if (__get_user(clen, &uiov->iov_len))
+				return -EFAULT;
+			if (clen < 0)
+				return -EINVAL;
+			sr->len = clen;
+		}
+
+		if (ddir == ITER_DEST && req->flags & REQ_F_APOLL_MULTISHOT) {
+			iomsg->namelen = msg.msg_namelen;
+			iomsg->controllen = msg.msg_controllen;
+			if (io_recvmsg_multishot_overflow(iomsg))
+				return -EOVERFLOW;
+		}
+	} else {
+		iomsg->free_iov = iomsg->fast_iov;
+		ret = __import_iovec(ddir, (struct iovec __user *)uiov,
+				     msg.msg_iovlen, UIO_FASTIOV,
+				     &iomsg->free_iov, &iomsg->msg.msg_iter,
+				     true);
+		if (ret < 0)
+			return ret;
+	}
+
+	return 0;
+}
+#endif
+
+static int __io_msg_copy_hdr(struct io_kiocb *req, struct io_async_msghdr *iomsg,
+			     struct sockaddr __user **addr, int ddir)
+{
+	struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg);
+	struct user_msghdr msg;
+	int ret;
+
+	if (copy_from_user(&msg, sr->umsg, sizeof(*sr->umsg)))
+		return -EFAULT;
+
+	ret = __copy_msghdr(&iomsg->msg, &msg, addr);
+	if (ret)
+		return ret;
+
+	if (req->flags & REQ_F_BUFFER_SELECT) {
+		if (msg.msg_iovlen == 0) {
+			sr->len = iomsg->fast_iov[0].iov_len = 0;
+			iomsg->fast_iov[0].iov_base = NULL;
+			iomsg->free_iov = NULL;
+		} else if (msg.msg_iovlen > 1) {
+			return -EINVAL;
+		} else {
+			if (copy_from_user(iomsg->fast_iov, msg.msg_iov,
+					   sizeof(*msg.msg_iov)))
+				return -EFAULT;
+			sr->len = iomsg->fast_iov[0].iov_len;
+			iomsg->free_iov = NULL;
+		}
+
+		if (ddir == ITER_DEST && req->flags & REQ_F_APOLL_MULTISHOT) {
+			iomsg->namelen = msg.msg_namelen;
+			iomsg->controllen = msg.msg_controllen;
+			if (io_recvmsg_multishot_overflow(iomsg))
+				return -EOVERFLOW;
+		}
+	} else {
+		iomsg->free_iov = iomsg->fast_iov;
+		ret = __import_iovec(ddir, msg.msg_iov, msg.msg_iovlen,
+				     UIO_FASTIOV, &iomsg->free_iov,
+				     &iomsg->msg.msg_iter, false);
+		if (ret > 0)
+			ret = 0;
+	}
+
+	return ret;
+}
+
+static int io_msg_copy_hdr(struct io_kiocb *req, struct io_async_msghdr *iomsg,
+			   struct sockaddr __user **addr, int ddir)
+{
+	iomsg->msg.msg_name = &iomsg->addr;
+	iomsg->msg.msg_iter.nr_segs = 0;
+
+#ifdef CONFIG_COMPAT
+	if (req->ctx->compat)
+		return __io_compat_msg_copy_hdr(req, iomsg, addr, ddir);
+#endif
+
+	return __io_msg_copy_hdr(req, iomsg, addr, ddir);
+}
+
 static int io_sendmsg_copy_hdr(struct io_kiocb *req,
 			       struct io_async_msghdr *iomsg)
 {
 	struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg);
 	int ret;
 
-	iomsg->msg.msg_name = &iomsg->addr;
-	iomsg->free_iov = iomsg->fast_iov;
-	ret = sendmsg_copy_msghdr(&iomsg->msg, sr->umsg, sr->msg_flags,
-					&iomsg->free_iov);
+	ret = io_msg_copy_hdr(req, iomsg, NULL, ITER_SOURCE);
+	if (ret)
+		return ret;
+
 	/* save msg_control as sys_sendmsg() overwrites it */
 	sr->msg_control = iomsg->msg.msg_control_user;
 	return ret;
@@ -435,142 +567,15 @@ int io_send(struct io_kiocb *req, unsigned int issue_flags)
 	return IOU_OK;
 }
 
-static bool io_recvmsg_multishot_overflow(struct io_async_msghdr *iomsg)
-{
-	int hdr;
-
-	if (iomsg->namelen < 0)
-		return true;
-	if (check_add_overflow((int)sizeof(struct io_uring_recvmsg_out),
-			       iomsg->namelen, &hdr))
-		return true;
-	if (check_add_overflow(hdr, (int)iomsg->controllen, &hdr))
-		return true;
-
-	return false;
-}
-
-static int __io_recvmsg_copy_hdr(struct io_kiocb *req,
-				 struct io_async_msghdr *iomsg)
-{
-	struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg);
-	struct user_msghdr msg;
-	int ret;
-
-	if (copy_from_user(&msg, sr->umsg, sizeof(*sr->umsg)))
-		return -EFAULT;
-
-	ret = __copy_msghdr(&iomsg->msg, &msg, &iomsg->uaddr);
-	if (ret)
-		return ret;
-
-	if (req->flags & REQ_F_BUFFER_SELECT) {
-		if (msg.msg_iovlen == 0) {
-			sr->len = iomsg->fast_iov[0].iov_len = 0;
-			iomsg->fast_iov[0].iov_base = NULL;
-			iomsg->free_iov = NULL;
-		} else if (msg.msg_iovlen > 1) {
-			return -EINVAL;
-		} else {
-			if (copy_from_user(iomsg->fast_iov, msg.msg_iov, sizeof(*msg.msg_iov)))
-				return -EFAULT;
-			sr->len = iomsg->fast_iov[0].iov_len;
-			iomsg->free_iov = NULL;
-		}
-
-		if (req->flags & REQ_F_APOLL_MULTISHOT) {
-			iomsg->namelen = msg.msg_namelen;
-			iomsg->controllen = msg.msg_controllen;
-			if (io_recvmsg_multishot_overflow(iomsg))
-				return -EOVERFLOW;
-		}
-	} else {
-		iomsg->free_iov = iomsg->fast_iov;
-		ret = __import_iovec(ITER_DEST, msg.msg_iov, msg.msg_iovlen, UIO_FASTIOV,
-				     &iomsg->free_iov, &iomsg->msg.msg_iter,
-				     false);
-		if (ret > 0)
-			ret = 0;
-	}
-
-	return ret;
-}
-
-#ifdef CONFIG_COMPAT
-static int __io_compat_recvmsg_copy_hdr(struct io_kiocb *req,
-					struct io_async_msghdr *iomsg)
-{
-	struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg);
-	struct compat_msghdr msg;
-	struct compat_iovec __user *uiov;
-	int ret;
-
-	if (copy_from_user(&msg, sr->umsg_compat, sizeof(msg)))
-		return -EFAULT;
-
-	ret = __get_compat_msghdr(&iomsg->msg, &msg, &iomsg->uaddr);
-	if (ret)
-		return ret;
-
-	uiov = compat_ptr(msg.msg_iov);
-	if (req->flags & REQ_F_BUFFER_SELECT) {
-		compat_ssize_t clen;
-
-		iomsg->free_iov = NULL;
-		if (msg.msg_iovlen == 0) {
-			sr->len = 0;
-		} else if (msg.msg_iovlen > 1) {
-			return -EINVAL;
-		} else {
-			if (!access_ok(uiov, sizeof(*uiov)))
-				return -EFAULT;
-			if (__get_user(clen, &uiov->iov_len))
-				return -EFAULT;
-			if (clen < 0)
-				return -EINVAL;
-			sr->len = clen;
-		}
-
-		if (req->flags & REQ_F_APOLL_MULTISHOT) {
-			iomsg->namelen = msg.msg_namelen;
-			iomsg->controllen = msg.msg_controllen;
-			if (io_recvmsg_multishot_overflow(iomsg))
-				return -EOVERFLOW;
-		}
-	} else {
-		iomsg->free_iov = iomsg->fast_iov;
-		ret = __import_iovec(ITER_DEST, (struct iovec __user *)uiov, msg.msg_iovlen,
-				   UIO_FASTIOV, &iomsg->free_iov,
-				   &iomsg->msg.msg_iter, true);
-		if (ret < 0)
-			return ret;
-	}
-
-	return 0;
-}
-#endif
-
-static int io_recvmsg_copy_hdr(struct io_kiocb *req,
-			       struct io_async_msghdr *iomsg)
-{
-	iomsg->msg.msg_name = &iomsg->addr;
-	iomsg->msg.msg_iter.nr_segs = 0;
-
-#ifdef CONFIG_COMPAT
-	if (req->ctx->compat)
-		return __io_compat_recvmsg_copy_hdr(req, iomsg);
-#endif
-
-	return __io_recvmsg_copy_hdr(req, iomsg);
-}
-
 int io_recvmsg_prep_async(struct io_kiocb *req)
 {
+	struct io_async_msghdr *iomsg;
 	int ret;
 
 	if (!io_msg_alloc_async_prep(req))
 		return -ENOMEM;
-	ret = io_recvmsg_copy_hdr(req, req->async_data);
+	iomsg = req->async_data;
+	ret = io_msg_copy_hdr(req, iomsg, &iomsg->uaddr, ITER_DEST);
 	if (!ret)
 		req->flags |= REQ_F_NEED_CLEANUP;
 	return ret;
@@ -793,7 +798,7 @@ int io_recvmsg(struct io_kiocb *req, unsigned int issue_flags)
 	if (req_has_async_data(req)) {
 		kmsg = req->async_data;
 	} else {
-		ret = io_recvmsg_copy_hdr(req, &iomsg);
+		ret = io_msg_copy_hdr(req, &iomsg, &iomsg.uaddr, ITER_DEST);
 		if (ret)
 			return ret;
 		kmsg = &iomsg;
-- 
2.43.0

[PATCH 2/3] io_uring/net: add provided buffer support for IORING_OP_SEND

[Jens Axboe]

It's pretty trivial to wire up provided buffer support for the send
side, just like we do on the receive side. This enables setting up
a buffer ring that an application can use to push pending sends to,
and then have a send pick a buffer from that ring.

One of the challenges with async IO and networking sends is that you
can get into reordering conditions if you have more than one inflight
at the same time. Consider the following scenario where everything is
fine:

1) App queues sendA for socket1
2) App queues sendB for socket1
3) App does io_uring_submit()
4) sendA is issued, completes successfully, posts CQE
5) sendB is issued, completes successfully, posts CQE

All is fine. Requests are always issued in-order, and both complete
inline as most sends do.

However, if we're flooding socket1 with sends, the following could
also result from the same sequence:

1) App queues sendA for socket1
2) App queues sendB for socket1
3) App does io_uring_submit()
4) sendA is issued, socket1 is full, poll is armed for retry
5) Space frees up in socket1, this triggers sendA retry via task_work
6) sendB is issued, completes successfully, posts CQE
7) sendA is retried, completes successfully, posts CQE

Now we've sent sendB before sendA, which can make things unhappy. If
both sendA and sendB had been using provided buffers, then it would look
as follows instead:

1) App queues dataA for sendA, queues sendA for socket1
2) App queues dataB for sendB queues sendB for socket1
3) App does io_uring_submit()
4) sendA is issued, socket1 is full, poll is armed for retry
5) Space frees up in socket1, this triggers sendA retry via task_work
6) sendB is issued, picks first buffer (dataA), completes successfully,
   posts CQE (which says "I sent dataA")
7) sendA is retried, picks first buffer (dataB), completes successfully,
   posts CQE (which says "I sent dataB")

Now we've sent the data in order, and everybody is happy.

It's worth noting that this also opens the door for supporting multishot
sends, as provided buffers would be a prerequisite for that. Those can
trigger either when new buffers are added to the outgoing ring, or (if
stalled due to lack of space) when space frees up in the socket.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/uapi/linux/io_uring.h |  1 +
 io_uring/io_uring.c           |  3 ++-
 io_uring/net.c                | 19 ++++++++++++++++---
 io_uring/opdef.c              |  1 +
 4 files changed, 20 insertions(+), 4 deletions(-)

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 7bd10201a02b..74c3afac9c63 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -522,6 +522,7 @@ struct io_uring_params {
 #define IORING_FEAT_CQE_SKIP		(1U << 11)
 #define IORING_FEAT_LINKED_FILE		(1U << 12)
 #define IORING_FEAT_REG_REG_RING	(1U << 13)
+#define IORING_FEAT_SEND_BUFS		(1U << 14)
 
 /*
  * io_uring_register(2) opcodes and arguments
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index cf2f514b7cc0..f6332fc56bed 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -3962,7 +3962,8 @@ static __cold int io_uring_create(unsigned entries, struct io_uring_params *p,
 			IORING_FEAT_POLL_32BITS | IORING_FEAT_SQPOLL_NONFIXED |
 			IORING_FEAT_EXT_ARG | IORING_FEAT_NATIVE_WORKERS |
 			IORING_FEAT_RSRC_TAGS | IORING_FEAT_CQE_SKIP |
-			IORING_FEAT_LINKED_FILE | IORING_FEAT_REG_REG_RING;
+			IORING_FEAT_LINKED_FILE | IORING_FEAT_REG_REG_RING |
+			IORING_FEAT_SEND_BUFS;
 
 	if (copy_to_user(params, p, sizeof(*p))) {
 		ret = -EFAULT;
diff --git a/io_uring/net.c b/io_uring/net.c
index fcbaeb7cc045..10b6d8caf4da 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -489,7 +489,8 @@ int io_sendmsg(struct io_kiocb *req, unsigned int issue_flags)
 		ret += sr->done_io;
 	else if (sr->done_io)
 		ret = sr->done_io;
-	io_req_set_res(req, ret, 0);
+	cflags = io_put_kbuf(req, issue_flags);
+	io_req_set_res(req, ret, cflags);
 	return IOU_OK;
 }
 
@@ -497,8 +498,10 @@ int io_send(struct io_kiocb *req, unsigned int issue_flags)
 {
 	struct sockaddr_storage __address;
 	struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg);
-	struct msghdr msg;
+	size_t len = sr->len;
+	unsigned int cflags;
 	struct socket *sock;
+	struct msghdr msg;
 	unsigned flags;
 	int min_ret = 0;
 	int ret;
@@ -531,7 +534,17 @@ int io_send(struct io_kiocb *req, unsigned int issue_flags)
 	if (unlikely(!sock))
 		return -ENOTSOCK;
 
-	ret = import_ubuf(ITER_SOURCE, sr->buf, sr->len, &msg.msg_iter);
+	if (io_do_buffer_select(req)) {
+		void __user *buf;
+
+		buf = io_buffer_select(req, &len, issue_flags);
+		if (!buf)
+			return -ENOBUFS;
+		sr->buf = buf;
+		sr->len = len;
+	}
+
+	ret = import_ubuf(ITER_SOURCE, sr->buf, len, &msg.msg_iter);
 	if (unlikely(ret))
 		return ret;
 
diff --git a/io_uring/opdef.c b/io_uring/opdef.c
index 9c080aadc5a6..88fbe5cfd379 100644
--- a/io_uring/opdef.c
+++ b/io_uring/opdef.c
@@ -273,6 +273,7 @@ const struct io_issue_def io_issue_defs[] = {
 		.audit_skip		= 1,
 		.ioprio			= 1,
 		.manual_alloc		= 1,
+		.buffer_select		= 1,
 #if defined(CONFIG_NET)
 		.prep			= io_sendmsg_prep,
 		.issue			= io_send,
-- 
2.43.0

[PATCH 3/3] io_uring/net: add provided buffer support for IORING_OP_SENDMSG

[Jens Axboe]

Adds provided buffer support for sendmsg as well, see the previous commit
that added it to IORING_OP_SEND for a longer explanation of why this
makes sense.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 io_uring/net.c   | 15 ++++++++++++++-
 io_uring/opdef.c |  1 +
 2 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/io_uring/net.c b/io_uring/net.c
index 10b6d8caf4da..30afb394efd7 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -436,6 +436,7 @@ int io_sendmsg(struct io_kiocb *req, unsigned int issue_flags)
 	struct io_sr_msg *sr = io_kiocb_to_cmd(req, struct io_sr_msg);
 	struct io_async_msghdr iomsg, *kmsg;
 	struct socket *sock;
+	unsigned int cflags;
 	unsigned flags;
 	int min_ret = 0;
 	int ret;
@@ -458,6 +459,17 @@ int io_sendmsg(struct io_kiocb *req, unsigned int issue_flags)
 	    (sr->flags & IORING_RECVSEND_POLL_FIRST))
 		return io_setup_async_msg(req, kmsg, issue_flags);
 
+	if (io_do_buffer_select(req)) {
+		void __user *buf;
+		size_t len = sr->len;
+
+		buf = io_buffer_select(req, &len, issue_flags);
+		if (!buf)
+			return -ENOBUFS;
+
+		iov_iter_ubuf(&kmsg->msg.msg_iter, ITER_SOURCE, buf, len);
+	}
+
 	flags = sr->msg_flags;
 	if (issue_flags & IO_URING_F_NONBLOCK)
 		flags |= MSG_DONTWAIT;
@@ -576,7 +588,8 @@ int io_send(struct io_kiocb *req, unsigned int issue_flags)
 		ret += sr->done_io;
 	else if (sr->done_io)
 		ret = sr->done_io;
-	io_req_set_res(req, ret, 0);
+	cflags = io_put_kbuf(req, issue_flags);
+	io_req_set_res(req, ret, cflags);
 	return IOU_OK;
 }
 
diff --git a/io_uring/opdef.c b/io_uring/opdef.c
index 88fbe5cfd379..1f6b09e61ef8 100644
--- a/io_uring/opdef.c
+++ b/io_uring/opdef.c
@@ -139,6 +139,7 @@ const struct io_issue_def io_issue_defs[] = {
 		.pollout		= 1,
 		.ioprio			= 1,
 		.manual_alloc		= 1,
+		.buffer_select		= 1,
 #if defined(CONFIG_NET)
 		.prep			= io_sendmsg_prep,
 		.issue			= io_sendmsg,
-- 
2.43.0