netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Gregory Haskins <gregory.haskins@gmail.com>
To: Avi Kivity <avi@redhat.com>
Cc: Anthony Liguori <anthony@codemonkey.ws>,
	Ingo Molnar <mingo@elte.hu>,
	kvm@vger.kernel.org, Andrew Morton <akpm@linux-foundation.org>,
	torvalds@linux-foundation.org,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	netdev@vger.kernel.org,
	"alacrityvm-devel@lists.sourceforge.net"
	<alacrityvm-devel@lists.sourceforge.net>
Subject: Re: [GIT PULL] AlacrityVM guest drivers for 2.6.33
Date: Mon, 21 Dec 2009 11:56:52 -0500	[thread overview]
Message-ID: <4B2FA8D4.3050206@gmail.com> (raw)
In-Reply-To: <4B2FA4F2.8000401@redhat.com>

[-- Attachment #1: Type: text/plain, Size: 1886 bytes --]

On 12/21/09 11:40 AM, Avi Kivity wrote:
> On 12/21/2009 06:37 PM, Anthony Liguori wrote:
>> Since virtio-pci supports MSI-X, there should be no IO exits on
>> host->guest notification other than EOI in the virtual APIC.  This is
>> a light weight exit today and will likely disappear entirely with
>> newer hardware.
> 
> I'm working on disappearing EOI exits on older hardware as well.  Same
> idea as the old TPR patching, without most of the magic.
> 

While I applaud any engineering effort that results in more optimal
execution, if you are talking about what we have discussed in the past
its not quite in the same league as my proposal.

You are talking about the ability to optimize the final EOI if there are
no pending interrupts remaining, right?  The problem with this approach
is it addresses the wrong side of the curve: That is, it optimizes the
code as its about to go io-idle.  You still have to take an extra exit
for each injection during the heat of battle, which is when you actually
need it most.

To that front, what I have done is re-used the lockless shared-memory
concept for even "interrupt injection".  Lockless shared-memory rings
have the property that both producer and consumer can simultaneously
manipulate the ring.  So what we do in alacrityvm is deliver shm-signals
(shm-signal == "interrupt" in vbus) over a ring so that the host can
inject a signal to a running vcpu and a vcpu can complete an
ack/re-inject cycle directly from vcpu context.  Therefore, we only need
a physical IDT injection when the vcpu is io-idle transitioning to
io-busy, and remain completely in parallel guest/host context until we
go idle again.

That said, your suggestion would play nicely with the above mentioned
scheme, so I look forward to seeing it in the tree.  Feel free to send
me patches for testing.

Kind Regards,
-Greg


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 267 bytes --]

  reply	other threads:[~2009-12-21 16:56 UTC|newest]

Thread overview: 105+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-12-07 18:53 [GIT PULL] AlacrityVM guest drivers for 2.6.33 Gregory Haskins
2009-12-18 21:51 ` Ingo Molnar
2009-12-21 15:34   ` Gregory Haskins
2009-12-21 15:43     ` Avi Kivity
2009-12-21 16:04       ` Gregory Haskins
2009-12-21 16:37         ` Anthony Liguori
2009-12-21 16:40           ` Avi Kivity
2009-12-21 16:56             ` Gregory Haskins [this message]
2009-12-21 17:05               ` Avi Kivity
2009-12-21 17:24                 ` Gregory Haskins
2009-12-21 16:46           ` Gregory Haskins
2009-12-21 17:20             ` Anthony Liguori
2009-12-21 17:44               ` Gregory Haskins
2009-12-22  0:12                 ` Anthony Liguori
2009-12-22  7:59                   ` Ingo Molnar
2009-12-22 11:49                   ` Andi Kleen
2009-12-22 15:31                     ` Anthony Liguori
2009-12-22 16:01                       ` Bartlomiej Zolnierkiewicz
2009-12-22 16:21                         ` Anthony Liguori
2009-12-23  6:51                           ` Ingo Molnar
2009-12-23 10:13                             ` Andi Kleen
2009-12-23 10:22                               ` Avi Kivity
2009-12-23 12:14                                 ` Andi Kleen
2009-12-23 12:48                                   ` Avi Kivity
2009-12-23 17:00                                     ` Chris Wright
2009-12-23 17:20                                       ` Andi Kleen
2009-12-23 15:03                                   ` Anthony Liguori
2009-12-23 17:52                                   ` [Alacrityvm-devel] " Peter W. Morreale
2009-12-23 18:17                                     ` Gregory Haskins
2009-12-23 18:22                                     ` Chris Wright
2009-12-23 18:15                                 ` Gregory Haskins
2009-12-23 20:20                                   ` Avi Kivity
2009-12-23 21:21                                     ` Gregory Haskins
2009-12-27  9:15                                       ` Avi Kivity
2009-12-27 13:18                                         ` Gregory Haskins
2009-12-27 13:27                                           ` Avi Kivity
2009-12-27 13:39                                             ` Gregory Haskins
2009-12-27 13:49                                               ` Avi Kivity
2009-12-28  1:01                                                 ` Gregory Haskins
2009-12-23 18:51                               ` Ingo Molnar
2009-12-23 19:27                                 ` Andi Kleen
2009-12-23 20:26                                   ` Avi Kivity
2009-12-24  6:58                                 ` Gleb Natapov
2009-12-24 20:41                                   ` Roland Dreier
2009-12-23 13:07                             ` Bartlomiej Zolnierkiewicz
2009-12-23 13:31                               ` Avi Kivity
2009-12-23 14:08                                 ` Bartlomiej Zolnierkiewicz
2009-12-23 14:28                                   ` Avi Kivity
2009-12-23 14:57                               ` Anthony Liguori
2009-12-23 18:31                                 ` Chris Wright
2009-12-23 16:44                             ` Gregory Haskins
2009-12-23 17:10                               ` Andi Kleen
2009-12-23 17:17                                 ` Gregory Haskins
2009-12-23 17:29                                   ` Linus Torvalds
2009-12-23 23:27                                     ` Anthony Liguori
2009-12-23 17:33                                   ` Andi Kleen
2009-12-23 20:36                               ` Avi Kivity
2009-12-23 21:01                                 ` Avi Kivity
2009-12-24  9:36                                   ` Gregory Haskins
2009-12-27  9:33                                     ` Avi Kivity
2009-12-27 13:34                                       ` Gregory Haskins
2009-12-27 13:49                                         ` Avi Kivity
2009-12-27 14:29                                           ` Gregory Haskins
2009-12-24  9:31                                 ` Gregory Haskins
2009-12-27  9:29                                   ` Avi Kivity
2009-12-22 16:21                       ` Andi Kleen
2009-12-22 16:27                         ` Anthony Liguori
2009-12-22 17:33                           ` Andi Kleen
2009-12-22 17:35                             ` Avi Kivity
2009-12-22 21:14                             ` Anthony Liguori
2009-12-23  0:02                               ` Chris Wright
2009-12-23 15:12                                 ` Anthony Liguori
2009-12-23 15:17                                   ` Chris Wright
2009-12-22 17:05                         ` Avi Kivity
2009-12-22 20:41                   ` Gregory Haskins
2009-12-22  7:22                 ` Gleb Natapov
2009-12-22  7:57     ` Ingo Molnar
2009-12-22 17:36       ` Gregory Haskins
2009-12-22 18:53         ` Avi Kivity
2009-12-22 18:56           ` Gregory Haskins
2009-12-22 19:15           ` Gregory Haskins
2009-12-22 19:25             ` Avi Kivity
2009-12-22 19:32               ` Gregory Haskins
2009-12-22 19:37                 ` Gregory Haskins
2009-12-22 19:38                 ` Avi Kivity
2009-12-22 19:41                   ` Gregory Haskins
2009-12-22 19:43                     ` Avi Kivity
2009-12-22 19:47                       ` Gregory Haskins
2009-12-22 19:40                 ` Avi Kivity
2009-12-22 19:39             ` Davide Libenzi
2009-12-22 19:53               ` Gregory Haskins
2009-12-23  1:05                 ` Davide Libenzi
2009-12-23  6:15         ` Kyle Moffett
2009-12-23 15:09           ` Anthony Liguori
2009-12-23 19:54             ` [Alacrityvm-devel] " Ira W. Snyder
2009-12-23 22:58               ` Anthony Liguori
2009-12-23 23:42                 ` Ira W. Snyder
2009-12-24 17:09                   ` Anthony Liguori
2009-12-25  0:38                     ` Ira W. Snyder
2009-12-24  4:52                 ` Kyle Moffett
2009-12-24 16:57                   ` Anthony Liguori
2009-12-23 17:34           ` Gregory Haskins
2009-12-23 19:28             ` Ira W. Snyder
2009-12-23 19:50               ` Andi Kleen
2009-12-24 10:01               ` 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=4B2FA8D4.3050206@gmail.com \
    --to=gregory.haskins@gmail.com \
    --cc=akpm@linux-foundation.org \
    --cc=alacrityvm-devel@lists.sourceforge.net \
    --cc=anthony@codemonkey.ws \
    --cc=avi@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=netdev@vger.kernel.org \
    --cc=torvalds@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 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).