From: "Michael S. Tsirkin" <mst@redhat.com>
To: Rusty Russell <rusty@rustcorp.com.au>
Cc: linux-kernel@vger.kernel.org,
virtualization@lists.linux-foundation.org, kvm@vger.kernel.org,
qemu-devel@nongnu.org
Subject: Re: [PATCHv2-RFC 0/2] virtio: put last seen used index into ring itself
Date: Mon, 31 May 2010 15:20:56 +0300 [thread overview]
Message-ID: <20100531122055.GB1704@redhat.com> (raw)
In-Reply-To: <201005311716.43573.rusty@rustcorp.com.au>
On Mon, May 31, 2010 at 05:16:42PM +0930, Rusty Russell wrote:
> On Thu, 27 May 2010 05:20:35 am Michael S. Tsirkin wrote:
> > Here's a rewrite of the original patch with a new layout.
> > I haven't tested it yet so no idea how this performs, but
> > I think this addresses the cache bounce issue raised by Avi.
> > Posting for early flames/comments.
>
> Sorry, not without some evidence that it'll actually reduce cacheline
> bouncing. I *think* it will, but it's not obvious: the host may keep
> looking at avail_idx as we're updating last_seen. Or does qemu always
> look at both together anyway?
> Can someone convince me this is a win?
> Rusty.
What really happens is host looks at flags and last_seen together.
And flags happens to be in the same cache line with avail idx.
So to get an obvious win, we should put flags and last_seen
in a separate cache line from avail, which us easy - just add some padding.
And I'll relax the requirement from guest to only require it to update
last_seen when interrupts are enabled. This way flags and
last_seen are written together and read together.
Makes sense?
--
MST
WARNING: multiple messages have this Message-ID (diff)
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Rusty Russell <rusty@rustcorp.com.au>
Cc: qemu-devel@nongnu.org, linux-kernel@vger.kernel.org,
kvm@vger.kernel.org, virtualization@lists.linux-foundation.org
Subject: [Qemu-devel] Re: [PATCHv2-RFC 0/2] virtio: put last seen used index into ring itself
Date: Mon, 31 May 2010 15:20:56 +0300 [thread overview]
Message-ID: <20100531122055.GB1704@redhat.com> (raw)
In-Reply-To: <201005311716.43573.rusty@rustcorp.com.au>
On Mon, May 31, 2010 at 05:16:42PM +0930, Rusty Russell wrote:
> On Thu, 27 May 2010 05:20:35 am Michael S. Tsirkin wrote:
> > Here's a rewrite of the original patch with a new layout.
> > I haven't tested it yet so no idea how this performs, but
> > I think this addresses the cache bounce issue raised by Avi.
> > Posting for early flames/comments.
>
> Sorry, not without some evidence that it'll actually reduce cacheline
> bouncing. I *think* it will, but it's not obvious: the host may keep
> looking at avail_idx as we're updating last_seen. Or does qemu always
> look at both together anyway?
> Can someone convince me this is a win?
> Rusty.
What really happens is host looks at flags and last_seen together.
And flags happens to be in the same cache line with avail idx.
So to get an obvious win, we should put flags and last_seen
in a separate cache line from avail, which us easy - just add some padding.
And I'll relax the requirement from guest to only require it to update
last_seen when interrupts are enabled. This way flags and
last_seen are written together and read together.
Makes sense?
--
MST
next prev parent reply other threads:[~2010-05-31 12:25 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-05-26 19:50 [PATCHv2-RFC 0/2] virtio: put last seen used index into ring itself Michael S. Tsirkin
2010-05-26 19:50 ` [Qemu-devel] " Michael S. Tsirkin
2010-05-26 19:50 ` [PATCHv2-RFC 1/2] virtio: support layout with avail ring before idx Michael S. Tsirkin
2010-05-26 19:50 ` Michael S. Tsirkin
2010-05-26 19:50 ` [Qemu-devel] " Michael S. Tsirkin
2010-05-26 19:50 ` [PATCHv2-RFC 2/2] virtio: publish used idx Michael S. Tsirkin
2010-05-26 19:50 ` [Qemu-devel] " Michael S. Tsirkin
2010-05-26 19:50 ` Michael S. Tsirkin
2010-05-27 12:07 ` [PATCHv2-RFC 0/2] virtio: put last seen used index into ring itself Avi Kivity
2010-05-27 12:07 ` [Qemu-devel] " Avi Kivity
2010-05-27 13:59 ` Michael S. Tsirkin
2010-05-27 13:59 ` [Qemu-devel] " Michael S. Tsirkin
2010-05-27 13:59 ` Michael S. Tsirkin
2010-05-27 12:07 ` Avi Kivity
2010-05-28 9:56 ` Jes Sorensen
2010-05-28 9:56 ` Jes Sorensen
2010-05-28 9:56 ` [Qemu-devel] " Jes Sorensen
2010-05-30 11:22 ` Michael S. Tsirkin
2010-05-30 11:22 ` Michael S. Tsirkin
2010-05-30 11:22 ` [Qemu-devel] " Michael S. Tsirkin
2010-05-31 7:36 ` Jes Sorensen
2010-05-31 7:36 ` [Qemu-devel] " Jes Sorensen
2010-05-31 13:29 ` Michael S. Tsirkin
2010-05-31 13:29 ` Michael S. Tsirkin
2010-05-31 13:29 ` [Qemu-devel] " Michael S. Tsirkin
2010-05-31 7:36 ` Jes Sorensen
2010-05-31 7:46 ` Rusty Russell
2010-05-31 7:46 ` [Qemu-devel] " Rusty Russell
2010-05-31 12:20 ` Michael S. Tsirkin
2010-05-31 12:20 ` Michael S. Tsirkin [this message]
2010-05-31 12:20 ` [Qemu-devel] " Michael S. Tsirkin
2010-06-02 20:39 ` Jes Sorensen
2010-06-02 20:39 ` Jes Sorensen
2010-06-02 20:39 ` Jes Sorensen
2010-05-31 7:46 ` Rusty Russell
-- strict thread matches above, loose matches on Subject: below --
2010-05-26 19:50 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=20100531122055.GB1704@redhat.com \
--to=mst@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=qemu-devel@nongnu.org \
--cc=rusty@rustcorp.com.au \
--cc=virtualization@lists.linux-foundation.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.