On Mon, Dec 28, 2020 at 01:27:42PM +0000, wangyunjian wrote:

quoted

-----Original Message-----
From: Michael S. Tsirkin [mailto:mst@redhat.com]
Sent: Sunday, December 27, 2020 7:21 PM
To: wangyunjian <redacted>
Cc: netdev@vger.kernel.org; jasowang@redhat.com;
willemdebruijn.kernel@gmail.com; virtualization@lists.linux-foundation.org;
Lilijun (Jerry) [off-list ref]; chenchanghu
[off-list ref]; xudingke [off-list ref]; huangbin (J)
[off-list ref]
Subject: Re: [PATCH net v5 2/2] vhost_net: fix tx queue stuck when sendmsg
fails

On Fri, Dec 25, 2020 at 03:24:33PM +0800, wangyunjian wrote:

quoted

From: Yunjian Wang <redacted>

Currently the driver doesn't drop a packet which can't be sent by tun
(e.g bad packet). In this case, the driver will always process the
same packet lead to the tx queue stuck.

So not making progress on a bad packet has some advantages, e.g. this is
easier to debug.
When is it important to drop the packet and continue?

quoted

To fix this issue:
1. in the case of persistent failure (e.g bad packet), the driver
   can skip this descriptor by ignoring the error.
2. in the case of transient failure (e.g -ENOBUFS, -EAGAIN and -ENOMEM),
   the driver schedules the worker to try again.

Fixes: 3a4d5c94e959 ("vhost_net: a kernel-level virtio server")

I'd just drop this tag, looks more like a feature than a bug ...

Do these two patches need to be sent separately?

Thanks

Makes sense to me.

quoted

quoted

Signed-off-by: Yunjian Wang <redacted>
Acked-by: Willem de Bruijn <willemb@google.com>
---
 drivers/vhost/net.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c index

c8784dfafdd7..01558fb2c552 100644

--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c

@@ -827,14 +827,13 @@ static void handle_tx_copy(struct vhost_net *net,

struct socket *sock)

quoted

 				msg.msg_flags &= ~MSG_MORE;
 		}

-		/* TODO: Check specific error and bomb out unless ENOBUFS? */
 		err = sock->ops->sendmsg(sock, &msg, len);
-		if (unlikely(err < 0)) {
+		if (unlikely(err == -EAGAIN || err == -ENOMEM || err == -ENOBUFS))
+{
 			vhost_discard_vq_desc(vq, 1);
 			vhost_net_enable_vq(net, vq);
 			break;
 		}
-		if (err != len)
+		if (err >= 0 && err != len)
 			pr_debug("Truncated TX packet: len %d != %zd\n",
 				 err, len);
 done:

@@ -922,7 +921,6 @@ static void handle_tx_zerocopy(struct vhost_net

*net, struct socket *sock)

quoted

 			msg.msg_flags &= ~MSG_MORE;
 		}

-		/* TODO: Check specific error and bomb out unless ENOBUFS? */
 		err = sock->ops->sendmsg(sock, &msg, len);
 		if (unlikely(err < 0)) {
 			if (zcopy_used) {

@@ -931,11 +929,13 @@ static void handle_tx_zerocopy(struct vhost_net

*net, struct socket *sock)

quoted

 				nvq->upend_idx = ((unsigned)nvq->upend_idx - 1)
 					% UIO_MAXIOV;
 			}
-			vhost_discard_vq_desc(vq, 1);
-			vhost_net_enable_vq(net, vq);
-			break;
+			if (err == -EAGAIN || err == -ENOMEM || err == -ENOBUFS) {
+				vhost_discard_vq_desc(vq, 1);
+				vhost_net_enable_vq(net, vq);
+				break;
+			}
 		}
-		if (err != len)
+		if (err >= 0 && err != len)
 			pr_debug("Truncated TX packet: "
 				 " len %d != %zd\n", err, len);
 		if (!zcopy_used)
--
2.23.0