All of lore.kernel.org
 help / color / mirror / Atom feed
From: Chris Wright <chrisw@sous-sol.org>
To: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Cc: Chris Wright <chrisw@sous-sol.org>,
	Anthony Liguori <aliguori@us.ibm.com>,
	kvm@vger.kernel.org, Avi Kivity <avi@redhat.com>,
	Marcelo Tosatti <mtosatti@redhat.com>
Subject: Re: [PATCH] kvm-vmx: add module parameter to avoid trapping HLT instructions (v2)
Date: Fri, 3 Dec 2010 09:29:06 -0800	[thread overview]
Message-ID: <20101203172906.GD10050@sequoia.sous-sol.org> (raw)
In-Reply-To: <20101203162731.GA11725@linux.vnet.ibm.com>

* Srivatsa Vaddagiri (vatsa@linux.vnet.ibm.com) wrote:
> On Fri, Dec 03, 2010 at 05:27:52PM +0530, Srivatsa Vaddagiri wrote:
> > On Thu, Dec 02, 2010 at 11:14:16AM -0800, Chris Wright wrote:
> > > Perhaps it should be a VM level option.  And then invert the notion.
> > > Create one idle domain w/out hlt trap.  Give that VM a vcpu per pcpu
> > > (pin in place probably).  And have that VM do nothing other than hlt.
> > > Then it's always runnable according to scheduler, and can "consume" the
> > > extra work that CFS wants to give away.
> > 
> > That's not sufficient. Lets we have 3 guests A, B, C that need to be rate
> > limited to 25% on a single cpu system. We create this idle guest D that is 100%
> > cpu hog as per above definition. Now when one of the guest is idle, what ensures
> > that the idle cycles of A is given only to D and not partly to B/C?
> 
> To tackle this problem, I was thinking of having a fill-thread associated with 
> each vcpu (i.e both belong to same cgroup). Fill-thread consumes idle cycles 
> left by vcpu, but otherwise doesn't compete with it for cycles.

That's what Marcelo's suggestion does w/out a fill thread.

  reply	other threads:[~2010-12-03 17:29 UTC|newest]

Thread overview: 69+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-12-02 13:59 [PATCH] kvm-vmx: add module parameter to avoid trapping HLT instructions (v2) Anthony Liguori
2010-12-02 14:39 ` lidong chen
2010-12-02 15:23   ` Anthony Liguori
2010-12-02 15:23   ` Anthony Liguori
2010-12-03  9:38     ` Avi Kivity
2010-12-03 11:12       ` Srivatsa Vaddagiri
2010-12-03 23:28       ` Anthony Liguori
2010-12-02 17:37 ` Marcelo Tosatti
2010-12-02 19:07   ` Anthony Liguori
2010-12-02 20:12     ` Marcelo Tosatti
2010-12-02 20:51       ` Anthony Liguori
2010-12-03  9:36         ` Avi Kivity
2010-12-03 22:45           ` Anthony Liguori
2010-12-04  8:13             ` Avi Kivity
2010-12-04 13:30               ` Anthony Liguori
2010-12-06  8:28                 ` Avi Kivity
2010-12-06  8:35                   ` Avi Kivity
2010-12-06 13:58                     ` Anthony Liguori
2010-12-06 14:01                       ` Avi Kivity
2010-12-06 14:02                         ` Avi Kivity
2010-12-06 14:08                           ` Anthony Liguori
2010-12-06 14:14                             ` Gleb Natapov
2010-12-06 14:03                         ` Anthony Liguori
2010-12-06 14:33                           ` Avi Kivity
2010-12-06 15:07                             ` Anthony Liguori
2010-12-06 15:16                               ` Avi Kivity
2010-12-06 16:21                                 ` Anthony Liguori
2010-12-06 16:30                                   ` Avi Kivity
2010-12-06 16:33                                     ` Anthony Liguori
2010-12-03 12:40         ` Gleb Natapov
2010-12-03 23:31       ` Anthony Liguori
2010-12-03 22:42   ` Anthony Liguori
2010-12-04  8:16     ` Avi Kivity
2010-12-04 13:48       ` Anthony Liguori
2010-12-06  8:32         ` Avi Kivity
2010-12-02 19:14 ` Chris Wright
2010-12-02 20:25   ` Anthony Liguori
2010-12-02 20:40     ` Chris Wright
2010-12-02 20:40   ` Marcelo Tosatti
2010-12-02 21:07     ` Chris Wright
2010-12-02 22:37       ` Anthony Liguori
2010-12-03  2:42         ` Chris Wright
2010-12-03  3:21           ` Anthony Liguori
2010-12-03  3:44             ` Chris Wright
2010-12-03 14:25               ` Anthony Liguori
2010-12-02 22:27     ` Anthony Liguori
2010-12-03 22:49     ` Anthony Liguori
2010-12-04  5:43       ` Srivatsa Vaddagiri
2010-12-03  9:40   ` Avi Kivity
2010-12-03 11:21     ` Srivatsa Vaddagiri
2010-12-03 11:57   ` Srivatsa Vaddagiri
2010-12-03 16:27     ` Srivatsa Vaddagiri
2010-12-03 17:29       ` Chris Wright [this message]
2010-12-03 17:33         ` Srivatsa Vaddagiri
2010-12-04  8:18           ` Avi Kivity
2010-12-03 17:57         ` Srivatsa Vaddagiri
2010-12-03 17:58           ` Chris Wright
2010-12-03 18:07             ` Anthony Liguori
2010-12-03 18:12               ` Srivatsa Vaddagiri
2010-12-04  8:19                 ` Avi Kivity
2010-12-03 18:20               ` Chris Wright
2010-12-03 18:55                 ` Anthony Liguori
2010-12-03 18:10             ` Marcelo Tosatti
2010-12-03 18:24               ` Marcelo Tosatti
2010-12-03 17:28     ` Chris Wright
2010-12-03 17:36       ` Srivatsa Vaddagiri
2010-12-03 17:38         ` Chris Wright
2010-12-03 17:43           ` Srivatsa Vaddagiri
2010-12-03 17:47           ` Anthony Liguori

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=20101203172906.GD10050@sequoia.sous-sol.org \
    --to=chrisw@sous-sol.org \
    --cc=aliguori@us.ibm.com \
    --cc=avi@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=mtosatti@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 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.