From: Avi Kivity <avi@redhat.com>
To: Rik van Riel <riel@redhat.com>
Cc: kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>,
Peter Zijlstra <a.p.zijlstra@chello.nl>,
Mike Galbraith <efault@gmx.de>,
Chris Wright <chrisw@sous-sol.org>
Subject: Re: [RFC -v3 PATCH 0/3] directed yield for Pause Loop Exiting
Date: Tue, 04 Jan 2011 11:14:33 +0200 [thread overview]
Message-ID: <4D22E4F9.4060909@redhat.com> (raw)
In-Reply-To: <20110103162637.29f23c40@annuminas.surriel.com>
On 01/03/2011 11:26 PM, Rik van Riel wrote:
> When running SMP virtual machines, it is possible for one VCPU to be
> spinning on a spinlock, while the VCPU that holds the spinlock is not
> currently running, because the host scheduler preempted it to run
> something else.
>
> Both Intel and AMD CPUs have a feature that detects when a virtual
> CPU is spinning on a lock and will trap to the host.
>
> The current KVM code sleeps for a bit whenever that happens, which
> results in eg. a 64 VCPU Windows guest taking forever and a bit to
> boot up. This is because the VCPU holding the lock is actually
> running and not sleeping, so the pause is counter-productive.
>
> In other workloads a pause can also be counter-productive, with
> spinlock detection resulting in one guest giving up its CPU time
> to the others. Instead of spinning, it ends up simply not running
> much at all.
>
> This patch series aims to fix that, by having a VCPU that spins
> give the remainder of its timeslice to another VCPU in the same
> guest before yielding the CPU - one that is runnable but got
> preempted, hopefully the lock holder.
>
Looks great.
Assuming there are no objections, Mike, can you get 2/3 into a
fast-forward-only branch of tip.git? I'll then merge it into kvm.git
and apply 1/3 and 2/3.
Since the merge window is about to open I'd like to merge this for
2.6.39 so it gets to stew a while in tip.git, kvm.git, and linux-next.
--
error compiling committee.c: too many arguments to function
next prev parent reply other threads:[~2011-01-04 9:14 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-01-03 21:26 [RFC -v3 PATCH 0/3] directed yield for Pause Loop Exiting Rik van Riel
2011-01-03 21:27 ` [RFC -v3 PATCH 1/3] kvm: keep track of which task is running a KVM vcpu Rik van Riel
2011-01-03 21:29 ` [RFC -v3 PATCH 2/3] sched: add yield_to function Rik van Riel
2011-01-04 1:51 ` Mike Galbraith
2011-01-04 6:14 ` KOSAKI Motohiro
2011-01-04 12:03 ` Avi Kivity
2011-01-05 2:39 ` KOSAKI Motohiro
2011-01-05 8:35 ` Avi Kivity
2011-01-05 8:40 ` KOSAKI Motohiro
2011-01-05 9:08 ` Avi Kivity
2011-01-05 9:30 ` KOSAKI Motohiro
2011-01-05 9:34 ` Avi Kivity
2011-01-05 10:24 ` Peter Zijlstra
2011-01-05 10:04 ` Peter Zijlstra
2011-01-04 17:16 ` Peter Zijlstra
2011-01-05 3:17 ` KOSAKI Motohiro
2011-01-04 14:28 ` Hillf Danton
2011-01-04 16:41 ` Hillf Danton
2011-01-04 16:44 ` Rik van Riel
2011-01-04 16:51 ` Hillf Danton
2011-01-04 16:54 ` Rik van Riel
2011-01-04 17:02 ` Hillf Danton
2011-01-04 17:08 ` Peter Zijlstra
2011-01-04 17:12 ` Hillf Danton
2011-01-04 17:22 ` Peter Zijlstra
2011-01-04 17:53 ` Rik van Riel
2011-01-04 18:05 ` Peter Zijlstra
2011-01-04 18:04 ` Peter Zijlstra
2011-01-04 18:53 ` Mike Galbraith
2011-01-05 16:57 ` Mike Galbraith
2011-01-05 17:04 ` Peter Zijlstra
2011-01-05 17:23 ` Mike Galbraith
2011-01-07 5:29 ` Mike Galbraith
2011-01-13 3:02 ` Rik van Riel
2011-01-13 3:26 ` Mike Galbraith
2011-01-13 5:08 ` Rik van Riel
2011-01-06 14:33 ` Hillf Danton
2011-01-05 17:10 ` Avi Kivity
2011-01-05 17:15 ` Peter Zijlstra
2011-01-05 17:19 ` Avi Kivity
2011-01-05 17:28 ` Peter Zijlstra
2011-01-05 17:35 ` Avi Kivity
2011-01-05 17:39 ` Peter Zijlstra
2011-01-06 3:49 ` Tejun Heo
2011-01-03 21:30 ` [RFC -v3 PATCH 3/3] Subject: kvm: use yield_to instead of sleep in kvm_vcpu_on_spin Rik van Riel
2011-01-04 6:42 ` [RFC -v3 PATCH 0/3] directed yield for Pause Loop Exiting Mike Galbraith
2011-01-04 9:09 ` Avi Kivity
2011-01-04 10:32 ` Mike Galbraith
2011-01-04 10:35 ` Avi Kivity
2011-01-04 9:14 ` Avi Kivity [this message]
2011-01-04 10:26 ` Mike Galbraith
2011-01-04 17:20 ` Peter Zijlstra
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=4D22E4F9.4060909@redhat.com \
--to=avi@redhat.com \
--cc=a.p.zijlstra@chello.nl \
--cc=chrisw@sous-sol.org \
--cc=efault@gmx.de \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=riel@redhat.com \
--cc=vatsa@linux.vnet.ibm.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