linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: David Ahern <dsahern@gmail.com>
To: Stefan Hajnoczi <stefanha@gmail.com>
Cc: Mohamad Gebai <mohamad.gebai@polymtl.ca>,
	kvm@vger.kernel.org, linux-perf-users@vger.kernel.org
Subject: Re: Tracing kvm: kvm_entry and kvm_exit
Date: Wed, 27 Feb 2013 21:49:24 -0700	[thread overview]
Message-ID: <512EE1D4.6080808@gmail.com> (raw)
In-Reply-To: <512E36B2.5090100@gmail.com>

On 2/27/13 9:39 AM, David Ahern wrote:
> I have been playing with the live mode a bit lately. I'll add a debug to
> note 2 consecutive entry events without an exit -- see if it sheds some
> light on it.

If you feel game take this for a spin:
   https://github.com/dsahern/linux/commits/perf-kvm-live-3.8

I pushed a few more commits that add debugs to show consecutive events 
(2 entries or 2 exits), max and min times for various exit reasons and 
an option to dump events that take longer than some threshold. If you 
dump the output to a file or run screen with logging you should be able 
to find when the 'long' event happens and which VMEXIT reason was the cause.

Example screen dump:

21:47:37.935302

Analyze events for all VMs, all VCPUs:

              VM-EXIT    Samples  Samples%     Time%   Min Time   Max 
Time         Avg time

          APIC_ACCESS        978    61.51%     0.02%        1us 
113us      7.81us ( +-   3.67% )
       IO_INSTRUCTION        212    13.33%     0.00%        2us 
85us      9.20us ( +-   7.57% )
            CR_ACCESS         93     5.85%     0.00%        2us 
66us      3.49us ( +-  19.73% )
   EXTERNAL_INTERRUPT         53     3.33%     0.01%        3us 
135us     57.02us ( +-   6.59% )
        EXCEPTION_NMI         26     1.64%     0.00%        2us 
3us      2.57us ( +-   1.74% )
    PENDING_INTERRUPT          3     0.19%     0.00%        1us 
1us      1.85us ( +-   4.18% )

Total Samples:1590, Total events handled time:39942741.62us.

295170818723592 VM 11854, vcpu 3: APIC_ACCESS event took 113usec
295170819686668 VM 11854, vcpu 0: EXTERNAL_INTERRUPT event took 112usec
295172155051688 VM 11854, vcpu 0: EXTERNAL_INTERRUPT event took 119usec
295172841724899 VM 11854, vcpu 0: APIC_ACCESS event took 129usec
295172999633325 VM 11854, vcpu 0: EXTERNAL_INTERRUPT event took 111usec

David

  reply	other threads:[~2013-02-28  4:49 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <1361550867.51279e13338a7@www.imp.polymtl.ca>
2013-02-25 10:18 ` Tracing kvm: kvm_entry and kvm_exit Stefan Hajnoczi
2013-02-26 16:24   ` Mohamad Gebai
2013-02-26 16:32   ` Gleb Natapov
2013-02-27 16:39   ` David Ahern
2013-02-28  4:49     ` David Ahern [this message]
2013-02-28  8:44       ` Stefan Hajnoczi
2013-03-04  5:36       ` Mohamad Gebai
2013-03-04  5:40         ` David Ahern
2013-03-04  6:15           ` Mohamad Gebai
2013-03-05 17:23             ` David Ahern

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=512EE1D4.6080808@gmail.com \
    --to=dsahern@gmail.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=mohamad.gebai@polymtl.ca \
    --cc=stefanha@gmail.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).