public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: Xin Tong <xerox.time.tech@gmail.com>, kvm@vger.kernel.org
Subject: Re: Measuring KVM Performance using Hardware Performance Counters
Date: Sun, 02 Feb 2014 12:37:09 +0100	[thread overview]
Message-ID: <52EE2DE5.6040007@redhat.com> (raw)
In-Reply-To: <CALKntY066W=JCq4J=i5MT-ggiumPypRb7+YOymXSZFhWHOoA_g@mail.gmail.com>

Il 02/02/2014 03:08, Xin Tong ha scritto:
> I am getting very weird profile results by running operf on linux on
> the host and profiling the a kvm virtual machine running dacapo
> eclipse benchmark.  I am expecting a lot of time should be spent in
> the qemu-system-x86_64 as the instructions from the eclipse benchmark
> would be treated as part of the qemu-system-x86_64 process, but the
> results tell different. any suggestions ?

Most of the time should be spent running the guest.  This is in the 
context of process qemu-system-x86_64, but it is not running code from 
the qemu executable.

What likely happens is that when the profiling counter expires, it 
causes the VM to exit before the profiling interrupt is delivered.  The 
sample should then be associated to the kvm_intel module.

31.5% in the KVM module seems like a lot.  Can you detail the results 
further, so that I can see the function names for kvm and kvm_intel?

Paolo

>
> CPU_CLK_UNHALTED %     Module
>          Symbol/Functions
> 1054             31.49 /kvm
>          /kvm
> 645             19.27 /kvm_intel
>         /kvm_intel
> 248             7.41
> /home/xtong/qemu-kvm-1.2.0/install/bin/qemu-system-x86_64
> /home/xtong/qemu-kvm-1.2.0/install/bin/qemu-system-x86_64
> 215             6.42 /usr/lib/debug/boot/vmlinux-3.11.0-13-generic
>         native_write_msr_safe
> 101             3.02 /usr/lib/debug/boot/vmlinux-3.11.0-13-generic
>         native_read_msr_safe
> 93               2.78 /lib/x86_64-linux-gnu/libc-2.15.so
>         /lib/x86_64-linux-gnu/libc-2.15.so
> 77               2.30 /lib/x86_64-linux-gnu/libglib-2.0.so.0.3200.4
>          /lib/x86_64-linux-gnu/libglib-2.0.so.0.3200.4
> 65               1.94 /usr/lib/debug/boot/vmlinux-3.11.0-13-generic
>          __srcu_read_lock
> 48               1.43 /usr/lib/debug/boot/vmlinux-3.11.0-13-generic
>          __srcu_read_unlock
> 40               1.20 /usr/lib/debug/boot/vmlinux-3.11.0-13-generic
>          native_load_tr_desc
> 29               0.87 /usr/lib/debug/boot/vmlinux-3.11.0-13-generic
>          __ticket_spin_lock
> 25               0.75 /usr/lib/debug/boot/vmlinux-3.11.0-13-generic
>          guest_exit
> 21               0.63 /usr/lib/debug/boot/vmlinux-3.11.0-13-generic
>          native_load_gdt
> ...
> Thank you.


  reply	other threads:[~2014-02-02 11:37 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-01-30  5:06 Measuring KVM Performance using Hardware Performance Counters Xin Tong
2014-02-02  2:08 ` Xin Tong
2014-02-02 11:37   ` Paolo Bonzini [this message]
     [not found]     ` <CA+JLOitGuTcX9coon0EKWBuhzuw4_7BxEab3=sk80J_WFakqhA@mail.gmail.com>
     [not found]       ` <52EF4F48.7080609@redhat.com>
2014-02-03 17:06         ` Xin Tong
2014-02-04  3:31           ` Paolo Bonzini
2014-02-04 20:53             ` Xin Tong

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=52EE2DE5.6040007@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=xerox.time.tech@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