From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dustin Kirkland Subject: Re: [Qemu-devel] qemu-kvm-0.11 regression, crashes on older guests with virtio network Date: Thu, 29 Oct 2009 09:34:34 -0500 Message-ID: <1256826874.25064.114.camel@x200> References: <1256807803.10825.39.camel@blaa> <1256815818-sup-7805@xpc65.scottt> <1256818566.10825.58.camel@blaa> <4AE9A299.5060003@codemonkey.ws> <1256826351.10825.69.camel@blaa> Reply-To: kirkland@canonical.com Mime-Version: 1.0 Content-Type: multipart/signed; micalg="pgp-sha1"; protocol="application/pgp-signature"; boundary="=-TuqC7zxkIBM7296lTKkB" Cc: Anthony Liguori , Scott Tsai , qemu-devel , kvm To: Mark McLoughlin Return-path: Received: from adelie.canonical.com ([91.189.90.139]:32929 "EHLO adelie.canonical.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752966AbZJ2Oep (ORCPT ); Thu, 29 Oct 2009 10:34:45 -0400 In-Reply-To: <1256826351.10825.69.camel@blaa> Sender: kvm-owner@vger.kernel.org List-ID: --=-TuqC7zxkIBM7296lTKkB Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Thu, 2009-10-29 at 14:25 +0000, Mark McLoughlin wrote: > On Thu, 2009-10-29 at 09:11 -0500, Anthony Liguori wrote: > > Mark McLoughlin wrote: > > > > > >> tap_set_offload(csum: 1, tso4: 1, tso6: 1, ecn: 1) > > >> being called and get an mtu of 1500 on virbr0 using his birdge.sh sc= ript. > > >> > > >> virtio_net_receive2 was trying to transfer a 1534 byte packet (1524 = 'size' + 10 'virtio_net_hdr') > > >> and the guest only had 1524 bytes of space in its input descriptors. > > >> =20 > > > > > > Okay, that sounds like a bug in Dustin's version of the guest virtio-= net > > > driver - if it is only supplying 1524 byte buffers, it should not be > > > saying it supports the VIRTIO_NET_F_GUEST_TSO4 feature > > > =20 > >=20 > > See: > >=20 > > commit 8eca6b1bc770982595db2f7207c65051572436cb > > Author: aliguori > > Date: Sun Apr 5 17:40:08 2009 +0000 > >=20 > > Fix oops on 2.6.25 guest (Rusty Russell) > > =20 > > I believe this is behind the following: > > https://bugs.edge.launchpad.net/ubuntu/jaunty/+source/linux/+bug/33= 1128 > > =20 > > virtio_pci in 2.6.25 didn't do feature negotiation correctly: it=20 > > acked every > > bit. Fortunately, we can detect this. > > =20 > > Signed-off-by: Rusty Russell > > Signed-off-by: Anthony Liguori > >=20 > > It looks like Rusty's fix wasn't enough. If I change virtio-net to onl= y=20 > > advertise F_MAC, we don't run into this problem. >=20 > If it's not acking VBAD_FEATURE, then it doesn't sound like the same > issue >=20 > It's also not acking e.g. MRG_RXBUF, which suggests that it is > selectively acking features, and choosing to ack TSO4 >=20 > A quick look through the guest driver code should clear up the > confusion. Dustion, got a pointer? Hi Mark, I'm currently testing Scott's patch above. In the mean time, Hardy's kernel is in git here: http://kernel.ubuntu.com/git?p=3Dubuntu/ubuntu-hardy.git;a=3Dsummary Thanks, :-Dustin --=-TuqC7zxkIBM7296lTKkB Content-Type: application/pgp-signature; name="signature.asc" Content-Description: This is a digitally signed message part -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) iEYEABECAAYFAkrpp/oACgkQs7pNXIOmEZQazwCg1f9ydSNc9BoXO1BDLbMAF1Qa /ssAoKLDCrhE0gNJeeIG+Byyi3ZFNQ3C =5gfu -----END PGP SIGNATURE----- --=-TuqC7zxkIBM7296lTKkB--