lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date:	Tue, 18 Dec 2007 17:54:15 +1100
From:	Rusty Russell <rusty@...tcorp.com.au>
To:	Christian Borntraeger <borntraeger@...ibm.com>
Cc:	dor.laor@...ranet.com, kvm-devel <kvm-devel@...ts.sourceforge.net>,
	netdev@...r.kernel.org, virtualization@...ts.linux-foundation.org
Subject: Re: [kvm-devel] [PATCH resent] virtio_net: Fix stalled inbound trafficon early packets

On Friday 14 December 2007 05:30:31 Christian Borntraeger wrote:
> Rusty, if you decide to apply my patch, there is one downside: The
> debugging code in virtio_ring sometimes triggers with a false positive:
>
> try_fill_recv calls vring_kick. Here we do a notify to the host. This might
> cause an immediate interrupt, triggering the poll routine before vring_kick
> can run END_USE. This is no real problem, as we dont touch the vq after
> notify.

Hmm, how about we just do something like this?  (needs previous enable_cb
patch)

> Christian, facing 64 guest cpus unconvering all kind of races

Heroic is the word which comes to mind... :)

Cheers,
Rusty.
---
virtio: flush buffers on open

Fix bug found by Christian Borntraeger: if the other side fills all
the registered network buffers before we enable NAPI, we will never
get an interrupt.  The simplest fix is to process the input queue once
on open.

Signed-off-by: Rusty Russell <rusty@...tcorp.com.au>

diff -r 1dd61213039c drivers/net/virtio_net.c
--- a/drivers/net/virtio_net.c	Tue Dec 18 17:34:18 2007 +1100
+++ b/drivers/net/virtio_net.c	Tue Dec 18 17:47:39 2007 +1100
@@ -326,6 +326,13 @@ static int virtnet_open(struct net_devic
 		return -ENOMEM;
 
 	napi_enable(&vi->napi);
+
+	/* If all buffers were filled by other side before we napi_enabled, we
+	 * won't get another interrupt, so process any outstanding packets
+	 * now.  virtnet_poll wants re-enable the queue, so we disable here. */
+	vi->rvq->vq_ops->disable_cb(vi->rvq);
+	netif_rx_schedule(vi->dev, &vi->napi);
+
 	return 0;
 }
 
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists