[RFC PATCH v1 0/7] virtio/vsock: introduce MSG_EOR flag for SEQPACKET

	This patchset implements support of MSG_EOR bit for SEQPACKET
AF_VSOCK sockets over virtio transport.
	Idea is to distinguish concepts of 'messages' and 'records'.
Message is result of sending calls: 'write()', 'send()', 'sendmsg()'
etc. It has fixed maximum length, and it bounds are visible using
return from receive calls: 'read()', 'recv()', 'recvmsg()' etc.
Current implementation based on message definition above.
	Record has unlimited length, it consists of multiple message,
and bounds of record are visible via MSG_EOR flag returned from
'recvmsg()' call. Sender passes MSG_EOR to sending system call and
receiver will see MSG_EOR when corresponding message will be processed.
	To support MSG_EOR new bit was added along with existing
'VIRTIO_VSOCK_SEQ_EOR': 'VIRTIO_VSOCK_SEQ_EOM'(end-of-message) - now it
works in the same way as 'VIRTIO_VSOCK_SEQ_EOR'. But 'VIRTIO_VSOCK_SEQ_EOR'
is used to mark 'MSG_EOR' bit passed from userspace.
	This patchset includes simple test for MSG_EOR.
 	Also i've added new vsock test for '-EAGAIN' receive result.

 Arseny Krasnov(7):
  virtio/vsock: add 'VIRTIO_VSOCK_SEQ_EOM' bit
  vsock: rename implementation from 'record' to 'message'
  vhost/vsock: support MSG_EOR bit processing
  virito/vsock: support MSG_EOR bit processing
  af_vsock: rename variables in receive loop
  vsock_test: update message bounds test for MSG_EOR
  vsock_test: 'SO_RCVTIMEO' test for SEQPACKET

 drivers/vhost/vsock.c                   | 28 +++++++----
 include/uapi/linux/virtio_vsock.h       |  1 +
 net/vmw_vsock/af_vsock.c                | 10 ++--
 net/vmw_vsock/virtio_transport_common.c | 23 +++++----
 tools/testing/vsock/vsock_test.c        | 57 ++++++++++++++++++++++-
 5 files changed, 96 insertions(+), 23 deletions(-)

Signed-off-by: Arseny Krasnov <redacted>

-- 
2.25.1

This bit is used to mark end of messages('EOM' - end of message), while
'VIRIO_VSOCK_SEQ_EOR' is used to pass MSG_EOR.

Signed-off-by: Arseny Krasnov <redacted>
---
 include/uapi/linux/virtio_vsock.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/include/uapi/linux/virtio_vsock.h b/include/uapi/linux/virtio_vsock.h
index 3dd3555b2740..1de3211a2988 100644
--- a/include/uapi/linux/virtio_vsock.h
+++ b/include/uapi/linux/virtio_vsock.h

@@ -98,6 +98,7 @@ enum virtio_vsock_shutdown {
 /* VIRTIO_VSOCK_OP_RW flags values */
 enum virtio_vsock_rw {
 	VIRTIO_VSOCK_SEQ_EOR = 1,
+	VIRTIO_VSOCK_SEQ_EOM = 2,
 };
 
 #endif /* _UAPI_LINUX_VIRTIO_VSOCK_H */

-- 
2.25.1

As 'record' is not same as 'message', rename current variables,
comments and defines from 'record' concept to 'message'.

Signed-off-by: Arseny Krasnov <redacted>
---
 drivers/vhost/vsock.c                   | 18 +++++++++---------
 net/vmw_vsock/virtio_transport_common.c | 14 +++++++-------
 2 files changed, 16 insertions(+), 16 deletions(-)

diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index f249622ef11b..3b55de70ac77 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c

@@ -114,7 +114,7 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
 		size_t nbytes;
 		size_t iov_len, payload_len;
 		int head;
-		bool restore_flag = false;
+		bool restore_msg_eom_flag = false;
 
 		spin_lock_bh(&vsock->send_pkt_list_lock);
 		if (list_empty(&vsock->send_pkt_list)) {

@@ -178,16 +178,16 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
 			 * small rx buffers, headers of packets in rx queue are
 			 * created dynamically and are initialized with header
 			 * of current packet(except length). But in case of
-			 * SOCK_SEQPACKET, we also must clear record delimeter
-			 * bit(VIRTIO_VSOCK_SEQ_EOR). Otherwise, instead of one
-			 * packet with delimeter(which marks end of record),
+			 * SOCK_SEQPACKET, we also must clear message delimeter
+			 * bit(VIRTIO_VSOCK_SEQ_EOM). Otherwise, instead of one
+			 * packet with delimeter(which marks end of message),
 			 * there will be sequence of packets with delimeter
 			 * bit set. After initialized header will be copied to
 			 * rx buffer, this bit will be restored.
 			 */
-			if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOR) {
-				pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
-				restore_flag = true;
+			if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOM) {
+				pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
+				restore_msg_eom_flag = true;
 			}
 		}
 

@@ -224,8 +224,8 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
 		 * to send it with the next available buffer.
 		 */
 		if (pkt->off < pkt->len) {
-			if (restore_flag)
-				pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
+			if (restore_msg_eom_flag)
+				pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
 
 			/* We are queueing the same virtio_vsock_pkt to handle
 			 * the remaining bytes, and we want to deliver it

diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
index 169ba8b72a63..dea564e4b482 100644
--- a/net/vmw_vsock/virtio_transport_common.c
+++ b/net/vmw_vsock/virtio_transport_common.c

@@ -77,7 +77,7 @@ virtio_transport_alloc_pkt(struct virtio_vsock_pkt_info *info,
 
 		if (msg_data_left(info->msg) == 0 &&
 		    info->type == VIRTIO_VSOCK_TYPE_SEQPACKET)
-			pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
+			pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
 	}
 
 	trace_virtio_transport_alloc_pkt(src_cid, src_port,

@@ -457,7 +457,7 @@ static int virtio_transport_seqpacket_do_dequeue(struct vsock_sock *vsk,
 				dequeued_len += pkt_len;
 		}
 
-		if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOR) {
+		if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOM) {
 			msg_ready = true;
 			vvs->msg_count--;
 		}

@@ -1029,7 +1029,7 @@ virtio_transport_recv_enqueue(struct vsock_sock *vsk,
 		goto out;
 	}
 
-	if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOR)
+	if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOM)
 		vvs->msg_count++;
 
 	/* Try to copy small packets into the buffer of last packet queued,

@@ -1044,12 +1044,12 @@ virtio_transport_recv_enqueue(struct vsock_sock *vsk,
 
 		/* If there is space in the last packet queued, we copy the
 		 * new packet in its buffer. We avoid this if the last packet
-		 * queued has VIRTIO_VSOCK_SEQ_EOR set, because this is
-		 * delimiter of SEQPACKET record, so 'pkt' is the first packet
-		 * of a new record.
+		 * queued has VIRTIO_VSOCK_SEQ_EOM set, because this is
+		 * delimiter of SEQPACKET message, so 'pkt' is the first packet
+		 * of a new message.
 		 */
 		if ((pkt->len <= last_pkt->buf_len - last_pkt->len) &&
-		    !(le32_to_cpu(last_pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOR)) {
+		    !(le32_to_cpu(last_pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOM)) {
 			memcpy(last_pkt->buf + last_pkt->len, pkt->buf,
 			       pkt->len);
 			last_pkt->len += pkt->len;

-- 
2.25.1

It works in the same way as 'end-of-message' bit: if packet has
'EOM' bit, also check for 'EOR' bit.

Signed-off-by: Arseny Krasnov <redacted>
---
 drivers/vhost/vsock.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index 3b55de70ac77..3e2b150f9c6f 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c

@@ -115,6 +115,7 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
 		size_t iov_len, payload_len;
 		int head;
 		bool restore_msg_eom_flag = false;
+		bool restore_msg_eor_flag = false;
 
 		spin_lock_bh(&vsock->send_pkt_list_lock);
 		if (list_empty(&vsock->send_pkt_list)) {

@@ -188,6 +189,11 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
 			if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOM) {
 				pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
 				restore_msg_eom_flag = true;
+
+				if (le32_to_cpu(pkt->hdr.flags & VIRTIO_VSOCK_SEQ_EOR)) {
+					pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
+					restore_msg_eor_flag = true;
+				}
 			}
 		}
 

@@ -224,9 +230,13 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
 		 * to send it with the next available buffer.
 		 */
 		if (pkt->off < pkt->len) {
-			if (restore_msg_eom_flag)
+			if (restore_msg_eom_flag) {
 				pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
 
+				if (restore_msg_eor_flag)
+					pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
+			}
+
 			/* We are queueing the same virtio_vsock_pkt to handle
 			 * the remaining bytes, and we want to deliver it
 			 * to monitoring devices in the next iteration.

-- 
2.25.1

Record is supported via MSG_EOR flag, while current logic operates
with message, so rename variables from 'record' to 'message'.

Signed-off-by: Arseny Krasnov <redacted>
---
 net/vmw_vsock/af_vsock.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index 3e02cc3b24f8..e2c0cfb334d2 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c

@@ -2014,7 +2014,7 @@ static int __vsock_seqpacket_recvmsg(struct sock *sk, struct msghdr *msg,
 {
 	const struct vsock_transport *transport;
 	struct vsock_sock *vsk;
-	ssize_t record_len;
+	ssize_t msg_len;
 	long timeout;
 	int err = 0;
 	DEFINE_WAIT(wait);

@@ -2028,9 +2028,9 @@ static int __vsock_seqpacket_recvmsg(struct sock *sk, struct msghdr *msg,
 	if (err <= 0)
 		goto out;
 
-	record_len = transport->seqpacket_dequeue(vsk, msg, flags);
+	msg_len = transport->seqpacket_dequeue(vsk, msg, flags);
 
-	if (record_len < 0) {
+	if (msg_len < 0) {
 		err = -ENOMEM;
 		goto out;
 	}

@@ -2044,14 +2044,14 @@ static int __vsock_seqpacket_recvmsg(struct sock *sk, struct msghdr *msg,
 		 * packet.
 		 */
 		if (flags & MSG_TRUNC)
-			err = record_len;
+			err = msg_len;
 		else
 			err = len - msg_data_left(msg);
 
 		/* Always set MSG_TRUNC if real length of packet is
 		 * bigger than user's buffer.
 		 */
-		if (record_len > len)
+		if (msg_len > len)
 			msg->msg_flags |= MSG_TRUNC;
 	}
 

-- 
2.25.1

Set 'MSG_EOR' in one of message sent, check that 'MSG_EOR'
is visible in corresponding message at receiver.

Signed-off-by: Arseny Krasnov <redacted>
---
 tools/testing/vsock/vsock_test.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/tools/testing/vsock/vsock_test.c b/tools/testing/vsock/vsock_test.c
index 67766bfe176f..2a3638c0a008 100644
--- a/tools/testing/vsock/vsock_test.c
+++ b/tools/testing/vsock/vsock_test.c

@@ -282,6 +282,7 @@ static void test_stream_msg_peek_server(const struct test_opts *opts)
 }
 
 #define MESSAGES_CNT 7
+#define MSG_EOR_IDX (MESSAGES_CNT / 2)
 static void test_seqpacket_msg_bounds_client(const struct test_opts *opts)
 {
 	int fd;

@@ -294,7 +295,7 @@ static void test_seqpacket_msg_bounds_client(const struct test_opts *opts)
 
 	/* Send several messages, one with MSG_EOR flag */
 	for (int i = 0; i < MESSAGES_CNT; i++)
-		send_byte(fd, 1, 0);
+		send_byte(fd, 1, (i == MSG_EOR_IDX) ? MSG_EOR : 0);
 
 	control_writeln("SENDDONE");
 	close(fd);

@@ -324,6 +325,11 @@ static void test_seqpacket_msg_bounds_server(const struct test_opts *opts)
 			perror("message bound violated");
 			exit(EXIT_FAILURE);
 		}
+
+		if ((i == MSG_EOR_IDX) ^ !!(msg.msg_flags & MSG_EOR)) {
+			perror("MSG_EOR");
+			exit(EXIT_FAILURE);
+		}
 	}
 
 	close(fd);

-- 
2.25.1

If packet has 'EOR' bit - set MSG_EOR in 'recvmsg()' flags.

Signed-off-by: Arseny Krasnov <redacted>
---
 net/vmw_vsock/virtio_transport_common.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
index dea564e4b482..ee80f31b1368 100644
--- a/net/vmw_vsock/virtio_transport_common.c
+++ b/net/vmw_vsock/virtio_transport_common.c

@@ -76,8 +76,12 @@ virtio_transport_alloc_pkt(struct virtio_vsock_pkt_info *info,
 			goto out;
 
 		if (msg_data_left(info->msg) == 0 &&
-		    info->type == VIRTIO_VSOCK_TYPE_SEQPACKET)
+		    info->type == VIRTIO_VSOCK_TYPE_SEQPACKET) {
 			pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
+
+			if (info->msg->msg_flags & MSG_EOR)
+				pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
+		}
 	}
 
 	trace_virtio_transport_alloc_pkt(src_cid, src_port,

@@ -460,6 +464,9 @@ static int virtio_transport_seqpacket_do_dequeue(struct vsock_sock *vsk,
 		if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOM) {
 			msg_ready = true;
 			vvs->msg_count--;
+
+			if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOR)
+				msg->msg_flags |= MSG_EOR;
 		}
 
 		virtio_transport_dec_rx_pkt(vvs, pkt);

-- 
2.25.1

Test for receive timeout check: connection is established,
receiver sets timeout, but sender does nothing. Receiver's
'read()' call must return EAGAIN.

Signed-off-by: Arseny Krasnov <redacted>
---
 tools/testing/vsock/vsock_test.c | 49 ++++++++++++++++++++++++++++++++
 1 file changed, 49 insertions(+)

diff --git a/tools/testing/vsock/vsock_test.c b/tools/testing/vsock/vsock_test.c
index 2a3638c0a008..aa2de27d0f77 100644
--- a/tools/testing/vsock/vsock_test.c
+++ b/tools/testing/vsock/vsock_test.c

@@ -391,6 +391,50 @@ static void test_seqpacket_msg_trunc_server(const struct test_opts *opts)
 	close(fd);
 }
 
+static void test_seqpacket_timeout_client(const struct test_opts *opts)
+{
+	int fd;
+	struct timeval tv;
+	char dummy;
+
+	fd = vsock_seqpacket_connect(opts->peer_cid, 1234);
+	if (fd < 0) {
+		perror("connect");
+		exit(EXIT_FAILURE);
+	}
+
+	tv.tv_sec = 1;
+	tv.tv_usec = 0;
+
+	if (setsockopt(fd, SOL_SOCKET, SO_RCVTIMEO, (void *)&tv, sizeof(tv)) == -1) {
+		perror("setsockopt 'SO_RCVTIMEO'");
+		exit(EXIT_FAILURE);
+	}
+
+	if ((read(fd, &dummy, sizeof(dummy)) != -1) ||
+	    (errno != EAGAIN)) {
+		perror("EAGAIN expected");
+		exit(EXIT_FAILURE);
+	}
+
+	control_writeln("WAITDONE");
+	close(fd);
+}
+
+static void test_seqpacket_timeout_server(const struct test_opts *opts)
+{
+	int fd;
+
+	fd = vsock_seqpacket_accept(VMADDR_CID_ANY, 1234, NULL);
+	if (fd < 0) {
+		perror("accept");
+		exit(EXIT_FAILURE);
+	}
+
+	control_expectln("WAITDONE");
+	close(fd);
+}
+
 static struct test_case test_cases[] = {
 	{
 		.name = "SOCK_STREAM connection reset",

@@ -431,6 +475,11 @@ static struct test_case test_cases[] = {
 		.run_client = test_seqpacket_msg_trunc_client,
 		.run_server = test_seqpacket_msg_trunc_server,
 	},
+	{
+		.name = "SOCK_SEQPACKET timeout",
+		.run_client = test_seqpacket_timeout_client,
+		.run_server = test_seqpacket_timeout_server,
+	},
 	{},
 };
 

-- 
2.25.1

On Mon, Jul 26, 2021 at 07:31:33PM +0300, Arseny Krasnov wrote:

This patchset implements support of MSG_EOR bit for SEQPACKET
AF_VSOCK sockets over virtio transport.
Idea is to distinguish concepts of 'messages' and 'records'.
Message is result of sending calls: 'write()', 'send()', 'sendmsg()'
etc. It has fixed maximum length, and it bounds are visible using
return from receive calls: 'read()', 'recv()', 'recvmsg()' etc.
Current implementation based on message definition above.
Record has unlimited length, it consists of multiple message,
and bounds of record are visible via MSG_EOR flag returned from
'recvmsg()' call. Sender passes MSG_EOR to sending system call and
receiver will see MSG_EOR when corresponding message will be processed.
To support MSG_EOR new bit was added along with existing
'VIRTIO_VSOCK_SEQ_EOR': 'VIRTIO_VSOCK_SEQ_EOM'(end-of-message) - now it
works in the same way as 'VIRTIO_VSOCK_SEQ_EOR'. But 'VIRTIO_VSOCK_SEQ_EOR'
is used to mark 'MSG_EOR' bit passed from userspace.

At this point it's probably better to rename the old flag, so we stay 
compatible.

What happens if one of the two peers does not support MSG_EOR handling, 
while the other does?

I'll do a closer review in the next few days.

Thanks,
Stefano

On 27.07.2021 10:59, Stefano Garzarella wrote:

Caution: This is an external email. Be cautious while opening links or attachments.



On Mon, Jul 26, 2021 at 07:31:33PM +0300, Arseny Krasnov wrote:

quoted

      This patchset implements support of MSG_EOR bit for SEQPACKET
AF_VSOCK sockets over virtio transport.
      Idea is to distinguish concepts of 'messages' and 'records'.
Message is result of sending calls: 'write()', 'send()', 'sendmsg()'
etc. It has fixed maximum length, and it bounds are visible using
return from receive calls: 'read()', 'recv()', 'recvmsg()' etc.
Current implementation based on message definition above.
      Record has unlimited length, it consists of multiple message,
and bounds of record are visible via MSG_EOR flag returned from
'recvmsg()' call. Sender passes MSG_EOR to sending system call and
receiver will see MSG_EOR when corresponding message will be processed.
      To support MSG_EOR new bit was added along with existing
'VIRTIO_VSOCK_SEQ_EOR': 'VIRTIO_VSOCK_SEQ_EOM'(end-of-message) - now it
works in the same way as 'VIRTIO_VSOCK_SEQ_EOR'. But 'VIRTIO_VSOCK_SEQ_EOR'
is used to mark 'MSG_EOR' bit passed from userspace.

At this point it's probably better to rename the old flag, so we stay
compatible.

What happens if one of the two peers does not support MSG_EOR handling,
while the other does?

I'll do a closer review in the next few days.

Thank You, also i think MSG_EOR support must be described in spec

Thanks,
Stefano

On Tue, Jul 27, 2021 at 12:34:36PM +0300, Arseny Krasnov wrote:

On 27.07.2021 10:59, Stefano Garzarella wrote:

quoted

Caution: This is an external email. Be cautious while opening links or attachments.



On Mon, Jul 26, 2021 at 07:31:33PM +0300, Arseny Krasnov wrote:

quoted

      This patchset implements support of MSG_EOR bit for SEQPACKET
AF_VSOCK sockets over virtio transport.
      Idea is to distinguish concepts of 'messages' and 'records'.
Message is result of sending calls: 'write()', 'send()', 'sendmsg()'
etc. It has fixed maximum length, and it bounds are visible using
return from receive calls: 'read()', 'recv()', 'recvmsg()' etc.
Current implementation based on message definition above.
      Record has unlimited length, it consists of multiple message,
and bounds of record are visible via MSG_EOR flag returned from
'recvmsg()' call. Sender passes MSG_EOR to sending system call and
receiver will see MSG_EOR when corresponding message will be processed.
      To support MSG_EOR new bit was added along with existing
'VIRTIO_VSOCK_SEQ_EOR': 'VIRTIO_VSOCK_SEQ_EOM'(end-of-message) - now it
works in the same way as 'VIRTIO_VSOCK_SEQ_EOR'. But 'VIRTIO_VSOCK_SEQ_EOR'
is used to mark 'MSG_EOR' bit passed from userspace.

At this point it's probably better to rename the old flag, so we stay
compatible.

What happens if one of the two peers does not support MSG_EOR handling,
while the other does?

I'll do a closer review in the next few days.

Thank You, also i think MSG_EOR support must be described in spec

Yep, sure!

What do you think about the concerns above?

Stefano

On 27.07.2021 12:58, Stefano Garzarella wrote:

Caution: This is an external email. Be cautious while opening links or attachments.



On Tue, Jul 27, 2021 at 12:34:36PM +0300, Arseny Krasnov wrote:

quoted

On 27.07.2021 10:59, Stefano Garzarella wrote:

quoted

Caution: This is an external email. Be cautious while opening links or attachments.



On Mon, Jul 26, 2021 at 07:31:33PM +0300, Arseny Krasnov wrote:

quoted

      This patchset implements support of MSG_EOR bit for SEQPACKET
AF_VSOCK sockets over virtio transport.
      Idea is to distinguish concepts of 'messages' and 'records'.
Message is result of sending calls: 'write()', 'send()', 'sendmsg()'
etc. It has fixed maximum length, and it bounds are visible using
return from receive calls: 'read()', 'recv()', 'recvmsg()' etc.
Current implementation based on message definition above.
      Record has unlimited length, it consists of multiple message,
and bounds of record are visible via MSG_EOR flag returned from
'recvmsg()' call. Sender passes MSG_EOR to sending system call and
receiver will see MSG_EOR when corresponding message will be processed.
      To support MSG_EOR new bit was added along with existing
'VIRTIO_VSOCK_SEQ_EOR': 'VIRTIO_VSOCK_SEQ_EOM'(end-of-message) - now it
works in the same way as 'VIRTIO_VSOCK_SEQ_EOR'. But 'VIRTIO_VSOCK_SEQ_EOR'
is used to mark 'MSG_EOR' bit passed from userspace.

At this point it's probably better to rename the old flag, so we stay
compatible.

What happens if one of the two peers does not support MSG_EOR handling,
while the other does?

I'll do a closer review in the next few days.

Thank You, also i think MSG_EOR support must be described in spec

Yep, sure!

What do you think about the concerns above?

I think you are right, i'll rename EOR -> EOM, and EOR will be added by patch

Stefano

Hi Arseny,

On Mon, Jul 26, 2021 at 07:31:33PM +0300, Arseny Krasnov wrote:

This patchset implements support of MSG_EOR bit for SEQPACKET
AF_VSOCK sockets over virtio transport.
Idea is to distinguish concepts of 'messages' and 'records'.
Message is result of sending calls: 'write()', 'send()', 'sendmsg()'
etc. It has fixed maximum length, and it bounds are visible using
return from receive calls: 'read()', 'recv()', 'recvmsg()' etc.
Current implementation based on message definition above.

Okay, so the implementation we merged is wrong right?
Should we disable the feature bit in stable kernels that contain it? Or 
maybe we can backport the fixes...

Record has unlimited length, it consists of multiple message,
and bounds of record are visible via MSG_EOR flag returned from
'recvmsg()' call. Sender passes MSG_EOR to sending system call and
receiver will see MSG_EOR when corresponding message will be processed.
To support MSG_EOR new bit was added along with existing
'VIRTIO_VSOCK_SEQ_EOR': 'VIRTIO_VSOCK_SEQ_EOM'(end-of-message) - now it
works in the same way as 'VIRTIO_VSOCK_SEQ_EOR'. But 'VIRTIO_VSOCK_SEQ_EOR'
is used to mark 'MSG_EOR' bit passed from userspace.

I understand that it makes sense to remap VIRTIO_VSOCK_SEQ_EOR to 
MSG_EOR to make the user understand the boundaries, but why do we need 
EOM as well?

Why do we care about the boundaries of a message within a record?
I mean, if the sender makes 3 calls:
     send(A1,0)
     send(A2,0)
     send(A3, MSG_EOR);

IIUC it should be fine if the receiver for example receives all in one 
single recv() calll with MSG_EOR set, so why do we need EOM?

Thanks,
Stefano

On 04.08.2021 15:57, Stefano Garzarella wrote:

Caution: This is an external email. Be cautious while opening links or attachments.



Hi Arseny,

On Mon, Jul 26, 2021 at 07:31:33PM +0300, Arseny Krasnov wrote:

quoted

      This patchset implements support of MSG_EOR bit for SEQPACKET
AF_VSOCK sockets over virtio transport.
      Idea is to distinguish concepts of 'messages' and 'records'.
Message is result of sending calls: 'write()', 'send()', 'sendmsg()'
etc. It has fixed maximum length, and it bounds are visible using
return from receive calls: 'read()', 'recv()', 'recvmsg()' etc.
Current implementation based on message definition above.

Okay, so the implementation we merged is wrong right?
Should we disable the feature bit in stable kernels that contain it? Or
maybe we can backport the fixes...

Hi,

No, this is correct and it is message boundary based. Idea of this

patchset is to add extra boundaries marker which i think could be

useful when we want to send data in seqpacket mode which length

is bigger than maximum message length(this is limited by transport).

Of course we can fragment big piece of data too small messages, but this

requires to carry fragmentation info in data protocol. So In this case

when we want to maintain boundaries receiver calls recvmsg() until MSG_EOR found.

But when receiver knows, that data is fit in maximum datagram length,

it doesn't care about checking MSG_EOR just calling recv() or read()(e.g.

message based mode).


Thank You

quoted

      Record has unlimited length, it consists of multiple message,
and bounds of record are visible via MSG_EOR flag returned from
'recvmsg()' call. Sender passes MSG_EOR to sending system call and
receiver will see MSG_EOR when corresponding message will be processed.
      To support MSG_EOR new bit was added along with existing
'VIRTIO_VSOCK_SEQ_EOR': 'VIRTIO_VSOCK_SEQ_EOM'(end-of-message) - now it
works in the same way as 'VIRTIO_VSOCK_SEQ_EOR'. But 'VIRTIO_VSOCK_SEQ_EOR'
is used to mark 'MSG_EOR' bit passed from userspace.

I understand that it makes sense to remap VIRTIO_VSOCK_SEQ_EOR to
MSG_EOR to make the user understand the boundaries, but why do we need
EOM as well?

Why do we care about the boundaries of a message within a record?
I mean, if the sender makes 3 calls:
     send(A1,0)
     send(A2,0)
     send(A3, MSG_EOR);

IIUC it should be fine if the receiver for example receives all in one
single recv() calll with MSG_EOR set, so why do we need EOM?

Thanks,
Stefano

On Thu, Aug 05, 2021 at 11:33:12AM +0300, Arseny Krasnov wrote:

On 04.08.2021 15:57, Stefano Garzarella wrote:

quoted

Caution: This is an external email. Be cautious while opening links or attachments.



Hi Arseny,

On Mon, Jul 26, 2021 at 07:31:33PM +0300, Arseny Krasnov wrote:

quoted

      This patchset implements support of MSG_EOR bit for SEQPACKET
AF_VSOCK sockets over virtio transport.
      Idea is to distinguish concepts of 'messages' and 'records'.
Message is result of sending calls: 'write()', 'send()', 'sendmsg()'
etc. It has fixed maximum length, and it bounds are visible using
return from receive calls: 'read()', 'recv()', 'recvmsg()' etc.
Current implementation based on message definition above.

Okay, so the implementation we merged is wrong right?
Should we disable the feature bit in stable kernels that contain it? Or
maybe we can backport the fixes...

Hi,

No, this is correct and it is message boundary based. Idea of this
patchset is to add extra boundaries marker which i think could be
useful when we want to send data in seqpacket mode which length
is bigger than maximum message length(this is limited by transport).
Of course we can fragment big piece of data too small messages, but 
this
requires to carry fragmentation info in data protocol. So In this case
when we want to maintain boundaries receiver calls recvmsg() until 
MSG_EOR found.
But when receiver knows, that data is fit in maximum datagram length,
it doesn't care about checking MSG_EOR just calling recv() or 
read()(e.g.
message based mode).

I'm not sure we should maintain boundaries of multiple send(), from 
POSIX standard [1]:

   SOCK_SEQPACKET
     Provides sequenced, reliable, bidirectional, connection-mode 
     transmission paths for records. A record can be sent using one or 
     more output operations and received using one or more input 
     operations, but a single operation never transfers part of more than 
     one record. Record boundaries are visible to the receiver via the 
     MSG_EOR flag.

 From my understanding a record could be sent with multiple send() and 
received, for example, with a single recvmsg().
The only boundary should be the MSG_EOR flag set by the user on the last 
send() of a record.

 From send() description [2]:

   MSG_EOR
     Terminates a record (if supported by the protocol).

 From recvmsg() description [3]:

   MSG_EOR
     End-of-record was received (if supported by the protocol).

Thanks,
Stefano

[1] 
https://pubs.opengroup.org/onlinepubs/9699919799/functions/socket.html
[2] https://pubs.opengroup.org/onlinepubs/9699919799/functions/send.html
[3] 
https://pubs.opengroup.org/onlinepubs/9699919799/functions/recvmsg.html

On 05.08.2021 12:06, Stefano Garzarella wrote:

Caution: This is an external email. Be cautious while opening links or attachments.



On Thu, Aug 05, 2021 at 11:33:12AM +0300, Arseny Krasnov wrote:

quoted

On 04.08.2021 15:57, Stefano Garzarella wrote:

quoted

Caution: This is an external email. Be cautious while opening links or attachments.



Hi Arseny,

On Mon, Jul 26, 2021 at 07:31:33PM +0300, Arseny Krasnov wrote:

quoted

      This patchset implements support of MSG_EOR bit for SEQPACKET
AF_VSOCK sockets over virtio transport.
      Idea is to distinguish concepts of 'messages' and 'records'.
Message is result of sending calls: 'write()', 'send()', 'sendmsg()'
etc. It has fixed maximum length, and it bounds are visible using
return from receive calls: 'read()', 'recv()', 'recvmsg()' etc.
Current implementation based on message definition above.

Okay, so the implementation we merged is wrong right?
Should we disable the feature bit in stable kernels that contain it? Or
maybe we can backport the fixes...

Hi,

No, this is correct and it is message boundary based. Idea of this
patchset is to add extra boundaries marker which i think could be
useful when we want to send data in seqpacket mode which length
is bigger than maximum message length(this is limited by transport).
Of course we can fragment big piece of data too small messages, but
this
requires to carry fragmentation info in data protocol. So In this case
when we want to maintain boundaries receiver calls recvmsg() until
MSG_EOR found.
But when receiver knows, that data is fit in maximum datagram length,
it doesn't care about checking MSG_EOR just calling recv() or
read()(e.g.
message based mode).

I'm not sure we should maintain boundaries of multiple send(), from
POSIX standard [1]:

Yes, but also from POSIX: such calls like send() and sendmsg()

operates with "message" and if we check recvmsg() we will

find the following thing:


For message-based sockets, such as SOCK_DGRAM and SOCK_SEQPACKET, the entire

message shall be read in a single operation. If a message is too long to fit in the supplied

buffers, and MSG_PEEK is not set in the flags argument, the excess bytes shall be discarded.


I understand this, that send() boundaries also must be maintained.

I've checked SEQPACKET in AF_UNIX and AX_25 - both doesn't support

MSG_EOR, so send() boundaries must be supported.

   SOCK_SEQPACKET
     Provides sequenced, reliable, bidirectional, connection-mode
     transmission paths for records. A record can be sent using one or
     more output operations and received using one or more input
     operations, but a single operation never transfers part of more than
     one record. Record boundaries are visible to the receiver via the
     MSG_EOR flag.

 From my understanding a record could be sent with multiple send() and
received, for example, with a single recvmsg().
The only boundary should be the MSG_EOR flag set by the user on the last
send() of a record.

You are right, if we talking about "record".

 From send() description [2]:

   MSG_EOR
     Terminates a record (if supported by the protocol).

 From recvmsg() description [3]:

   MSG_EOR
     End-of-record was received (if supported by the protocol).

Thanks,
Stefano

[1]
https://pubs.opengroup.org/onlinepubs/9699919799/functions/socket.html
[2] https://pubs.opengroup.org/onlinepubs/9699919799/functions/send.html
[3]
https://pubs.opengroup.org/onlinepubs/9699919799/functions/recvmsg.html

P.S.: seems SEQPACKET is too exotic thing that everyone implements it in

own manner, because i've tested SCTP seqpacket implementation, and found

that:

1) It doesn't support MSG_EOR bit at send side, but uses MSG_EOR at receiver

side to mark MESSAGE boundary.

2) According POSIX any extra bytes that didn't fit in user's buffer must be dropped,

but SCTP doesn't drop it - you can read rest of datagram in next calls.

On Thu, Aug 05, 2021 at 12:21:57PM +0300, Arseny Krasnov wrote:

On 05.08.2021 12:06, Stefano Garzarella wrote:

quoted

Caution: This is an external email. Be cautious while opening links or attachments.



On Thu, Aug 05, 2021 at 11:33:12AM +0300, Arseny Krasnov wrote:

quoted

On 04.08.2021 15:57, Stefano Garzarella wrote:

quoted

Caution: This is an external email. Be cautious while opening links or attachments.



Hi Arseny,

On Mon, Jul 26, 2021 at 07:31:33PM +0300, Arseny Krasnov wrote:

quoted

      This patchset implements support of MSG_EOR bit for SEQPACKET
AF_VSOCK sockets over virtio transport.
      Idea is to distinguish concepts of 'messages' and 'records'.
Message is result of sending calls: 'write()', 'send()', 'sendmsg()'
etc. It has fixed maximum length, and it bounds are visible using
return from receive calls: 'read()', 'recv()', 'recvmsg()' etc.
Current implementation based on message definition above.

Okay, so the implementation we merged is wrong right?
Should we disable the feature bit in stable kernels that contain it? Or
maybe we can backport the fixes...

Hi,

No, this is correct and it is message boundary based. Idea of this
patchset is to add extra boundaries marker which i think could be
useful when we want to send data in seqpacket mode which length
is bigger than maximum message length(this is limited by transport).
Of course we can fragment big piece of data too small messages, but
this
requires to carry fragmentation info in data protocol. So In this case
when we want to maintain boundaries receiver calls recvmsg() until
MSG_EOR found.
But when receiver knows, that data is fit in maximum datagram length,
it doesn't care about checking MSG_EOR just calling recv() or
read()(e.g.
message based mode).

I'm not sure we should maintain boundaries of multiple send(), from
POSIX standard [1]:

Yes, but also from POSIX: such calls like send() and sendmsg()

operates with "message" and if we check recvmsg() we will

find the following thing:


For message-based sockets, such as SOCK_DGRAM and SOCK_SEQPACKET, the entire

message shall be read in a single operation. If a message is too long to fit in the supplied

buffers, and MSG_PEEK is not set in the flags argument, the excess bytes shall be discarded.


I understand this, that send() boundaries also must be maintained.

I've checked SEQPACKET in AF_UNIX and AX_25 - both doesn't support

MSG_EOR, so send() boundaries must be supported.

quoted

   SOCK_SEQPACKET
     Provides sequenced, reliable, bidirectional, connection-mode
     transmission paths for records. A record can be sent using one or
     more output operations and received using one or more input
     operations, but a single operation never transfers part of more than
     one record. Record boundaries are visible to the receiver via the
     MSG_EOR flag.

 From my understanding a record could be sent with multiple send() 
 and
received, for example, with a single recvmsg().
The only boundary should be the MSG_EOR flag set by the user on the 
last
send() of a record.

You are right, if we talking about "record".

quoted

 From send() description [2]:

   MSG_EOR
     Terminates a record (if supported by the protocol).

 From recvmsg() description [3]:

   MSG_EOR
     End-of-record was received (if supported by the protocol).

Thanks,
Stefano

[1]
https://pubs.opengroup.org/onlinepubs/9699919799/functions/socket.html
[2] 
https://pubs.opengroup.org/onlinepubs/9699919799/functions/send.html
[3]
https://pubs.opengroup.org/onlinepubs/9699919799/functions/recvmsg.html

P.S.: seems SEQPACKET is too exotic thing that everyone implements it 
in

own manner, because i've tested SCTP seqpacket implementation, and 
found

that:

1) It doesn't support MSG_EOR bit at send side, but uses MSG_EOR at 
receiver

side to mark MESSAGE boundary.

2) According POSIX any extra bytes that didn't fit in user's buffer 
must be dropped,

but SCTP doesn't drop it - you can read rest of datagram in next calls.

Thanks for this useful information, now I see the differences and why we 
should support both.

I think is better to include them in the cover letter.

I'm going to review the paches right now :-)

Stefano

On Mon, Jul 26, 2021 at 07:33:04PM +0300, Arseny Krasnov wrote:

Already said, but I'll repeat it for completeness.

It's better to rename the flag 1 and use it in the same way we did 
before, so it's backward compatible.

Obviously we have to update the specifications too, explaining the 
difference between the two :-)

Thanks,
Stefano

On Mon, Jul 26, 2021 at 07:33:25PM +0300, Arseny Krasnov wrote:

As 'record' is not same as 'message', rename current variables,
comments and defines from 'record' concept to 'message'.

Signed-off-by: Arseny Krasnov <redacted>
---
drivers/vhost/vsock.c                   | 18 +++++++++---------
net/vmw_vsock/virtio_transport_common.c | 14 +++++++-------
2 files changed, 16 insertions(+), 16 deletions(-)

This patch is fine, I think you can move here the renaming of the flag 
too.

Stefano

On Mon, Jul 26, 2021 at 07:33:38PM +0300, Arseny Krasnov wrote:

Since we now have 2 flags to potentially restore, we could use a single
variable (e.g. uint32_t flags_to_restore), initialized to 0.

We can set all the flags we need to restore and then simply put it
in or with the `pkt->hdr.flags` field.

                                                                ^
Here it should be `le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOR`

If we use a single variable, here we can simply do:

			pkt->hdr.flags |= cpu_to_le32(flags_to_restore);

Stefano

On 06.08.2021 10:28, Stefano Garzarella wrote:

Caution: This is an external email. Be cautious while opening links or attachments.



On Mon, Jul 26, 2021 at 07:33:38PM +0300, Arseny Krasnov wrote:

quoted

It works in the same way as 'end-of-message' bit: if packet has
'EOM' bit, also check for 'EOR' bit.

Signed-off-by: Arseny Krasnov <redacted>
---
drivers/vhost/vsock.c | 12 +++++++++++-
1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index 3b55de70ac77..3e2b150f9c6f 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c

@@ -115,6 +115,7 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
              size_t iov_len, payload_len;
              int head;
              bool restore_msg_eom_flag = false;
+              bool restore_msg_eor_flag = false;

Since we now have 2 flags to potentially restore, we could use a single
variable (e.g. uint32_t flags_to_restore), initialized to 0.

We can set all the flags we need to restore and then simply put it
in or with the `pkt->hdr.flags` field.

quoted

              spin_lock_bh(&vsock->send_pkt_list_lock);
              if (list_empty(&vsock->send_pkt_list)) {

@@ -188,6 +189,11 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
                      if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOM) {
                              pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
                              restore_msg_eom_flag = true;
+
+                              if (le32_to_cpu(pkt->hdr.flags & VIRTIO_VSOCK_SEQ_EOR)) {

                                                                ^
Here it should be `le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOR`

quoted

+                                      pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
+                                      restore_msg_eor_flag = true;
+                              }
                      }
              }

@@ -224,9 +230,13 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
               * to send it with the next available buffer.
               */
              if (pkt->off < pkt->len) {
-                      if (restore_msg_eom_flag)
+                      if (restore_msg_eom_flag) {
                              pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);

+                              if (restore_msg_eor_flag)
+                                      pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
+                      }
+

If we use a single variable, here we can simply do:

                        pkt->hdr.flags |= cpu_to_le32(flags_to_restore);

Stefano

Thanks, i'll prepare v2 both with spec patch. About spec: i've already sent

patch for SEQPACKET, can i prepare spec patch updating current reviewed

SEQPACKET? E.g. i'll include both EOM and EOR in one patch.


Thank You

On Fri, Aug 06, 2021 at 11:40:38AM +0300, Arseny Krasnov wrote:

On 06.08.2021 10:28, Stefano Garzarella wrote:

quoted

Caution: This is an external email. Be cautious while opening links or attachments.



On Mon, Jul 26, 2021 at 07:33:38PM +0300, Arseny Krasnov wrote:

quoted

It works in the same way as 'end-of-message' bit: if packet has
'EOM' bit, also check for 'EOR' bit.

Signed-off-by: Arseny Krasnov <redacted>
---
drivers/vhost/vsock.c | 12 +++++++++++-
1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
index 3b55de70ac77..3e2b150f9c6f 100644
--- a/drivers/vhost/vsock.c
+++ b/drivers/vhost/vsock.c

@@ -115,6 +115,7 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
              size_t iov_len, payload_len;
              int head;
              bool restore_msg_eom_flag = false;
+              bool restore_msg_eor_flag = false;

Since we now have 2 flags to potentially restore, we could use a single
variable (e.g. uint32_t flags_to_restore), initialized to 0.

We can set all the flags we need to restore and then simply put it
in or with the `pkt->hdr.flags` field.

quoted

              spin_lock_bh(&vsock->send_pkt_list_lock);
              if (list_empty(&vsock->send_pkt_list)) {

@@ -188,6 +189,11 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
                      if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOM) {
                              pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
                              restore_msg_eom_flag = true;
+
+                              if (le32_to_cpu(pkt->hdr.flags & VIRTIO_VSOCK_SEQ_EOR)) {

                                                                ^
Here it should be `le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOR`

quoted

+                                      pkt->hdr.flags &= ~cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
+                                      restore_msg_eor_flag = true;
+                              }
                      }
              }

@@ -224,9 +230,13 @@ vhost_transport_do_send_pkt(struct vhost_vsock *vsock,
               * to send it with the next available buffer.
               */
              if (pkt->off < pkt->len) {
-                      if (restore_msg_eom_flag)
+                      if (restore_msg_eom_flag) {
                              pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);

+                              if (restore_msg_eor_flag)
+                                      pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
+                      }
+

If we use a single variable, here we can simply do:

                        pkt->hdr.flags |= cpu_to_le32(flags_to_restore);

Stefano

Thanks, i'll prepare v2 both with spec patch. About spec: i've already sent

patch for SEQPACKET, can i prepare spec patch updating current reviewed

SEQPACKET? E.g. i'll include both EOM and EOR in one patch.

Yep, since spec is not yet merged, I think make sense to have all 
seqpacket stuff in a single patch.

Thanks,
Stefano