From: Aaron Conole <aconole@redhat.com>
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Jason Wang <jasowang@redhat.com>,
Maxime Coquelin <maxime.coquelin@redhat.com>,
yuanhan.liu@linux.intel.com, pbonzini@redhat.com,
qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [RFC v2 0/3] virtio-net: Add support to MTU feature
Date: Wed, 23 Nov 2016 09:02:53 -0500 [thread overview]
Message-ID: <f7tlgwal2eq.fsf@redhat.com> (raw)
In-Reply-To: <20161123062418-mutt-send-email-mst@kernel.org> (Michael S. Tsirkin's message of "Wed, 23 Nov 2016 06:26:41 +0200")
"Michael S. Tsirkin" <mst@redhat.com> writes:
> On Wed, Nov 23, 2016 at 11:42:52AM +0800, Jason Wang wrote:
>> > > > > > > > Seems to me like an easy way to get out of sync.
>> > > > > >
>> > > > > >If we send it to the backend, that has a chance to check
>> > > > > >mtu and disconnect on error.
>> > > >
>> > > >For vhost-user backend, we can send it the MTU value with a
>> > > >vhost-user protocol feature.
>> > > >
>> > > >For tun/macvtap, how do you do without adding a new ioctl ?
>> > Have management configure same mtu on the backend and in qemu.
>> >
>> >
>>
>> Then why not do same for vhost-user (instead of using two different
>> methods)?
>
> That's what I'm saying. If backend supports that, we can also
> check the mtu in some way to make sure it matches.
I'm not sure why we need a new ioctl (or an ioctl at all - netlink
supports all of this)?
ex:
08:58:34 aconole {fast-datapath-beta-rhel-7} ~/rhpkg/openvswitch$ sudo ip tuntap add dev tap0 mode tap
[sudo] password for aconole:
08:58:40 aconole {fast-datapath-beta-rhel-7} ~/rhpkg/openvswitch$ ip l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
...
7: tap0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 46:e0:fc:83:54:1c brd ff:ff:ff:ff:ff:ff
08:58:51 aconole {fast-datapath-beta-rhel-7} ~/rhpkg/openvswitch$ sudo ip l set tap0 mtu 8000
08:58:54 aconole {fast-datapath-beta-rhel-7} ~/rhpkg/openvswitch$ ip l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
...
7: tap0: <BROADCAST,MULTICAST> mtu 8000 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether 46:e0:fc:83:54:1c brd ff:ff:ff:ff:ff:ff
So, at least with iproute2, we can already read and write using the netlink
interface for tuntap devices. I haven't played with macvtap, but I
think it's similar support - just do a netlink query, get the configured
MTU, and advertise it. I might be missing something though - I'm a
simple guy with simple ideas. Maybe there's a cross-platform issue or
something?
-Aaron
next prev parent reply other threads:[~2016-11-23 14:03 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-11-17 21:58 [Qemu-devel] [RFC v2 0/3] virtio-net: Add support to MTU feature Maxime Coquelin
2016-11-17 21:58 ` [Qemu-devel] [RFC v2 1/3] vhost-user: Add new protocol feature MTU Maxime Coquelin
2016-11-18 14:26 ` Aaron Conole
2016-11-21 12:50 ` Maxime Coquelin
2016-11-17 21:58 ` [Qemu-devel] [RFC v2 2/3] vhost-net: Add new MTU feature support Maxime Coquelin
2016-11-17 22:39 ` Michael S. Tsirkin
2016-11-21 12:51 ` Maxime Coquelin
2016-11-18 18:13 ` Aaron Conole
2016-11-17 21:58 ` [Qemu-devel] [RFC v2 3/3] virtio-net: Add " Maxime Coquelin
2016-11-17 22:38 ` Michael S. Tsirkin
2016-11-21 12:34 ` Maxime Coquelin
2016-11-21 16:48 ` Michael S. Tsirkin
2016-11-22 12:11 ` Maxime Coquelin
2016-11-22 14:18 ` Michael S. Tsirkin
2016-11-22 15:33 ` Maxime Coquelin
2016-11-17 22:34 ` [Qemu-devel] [RFC v2 0/3] virtio-net: Add support to MTU feature Michael S. Tsirkin
2016-11-18 6:42 ` John Fastabend
2016-11-18 18:15 ` Aaron Conole
2016-11-18 18:52 ` Maxime Coquelin
2016-11-18 19:21 ` Aaron Conole
2016-11-21 16:23 ` Michael S. Tsirkin
2016-11-22 4:07 ` Jason Wang
2016-11-22 7:40 ` Maxime Coquelin
2016-11-22 14:32 ` Aaron Conole
2016-11-22 14:41 ` Michael S. Tsirkin
2016-11-22 17:56 ` Maxime Coquelin
2016-11-22 20:38 ` Michael S. Tsirkin
2016-11-23 3:42 ` Jason Wang
2016-11-23 4:26 ` Michael S. Tsirkin
2016-11-23 14:02 ` Aaron Conole [this message]
2016-11-23 17:42 ` Michael S. Tsirkin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=f7tlgwal2eq.fsf@redhat.com \
--to=aconole@redhat.com \
--cc=jasowang@redhat.com \
--cc=maxime.coquelin@redhat.com \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=yuanhan.liu@linux.intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).