linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Etem <etem_deniz@hotmail.com>
To: linux-perf-users@vger.kernel.org
Subject: perf record + perf report
Date: Mon, 29 Jul 2013 08:52:20 +0000 (UTC)	[thread overview]
Message-ID: <loom.20130729T104749-103@post.gmane.org> (raw)

I am running Ubuntu 11.10 64bit on x86_64 machine with 4 core and 
hyperthreading enabled. I am using the following commands in order to get 
per-thread performance characteristics of a multi-threaded application. 
"perf record -s -e cycles,instructions,cache-references,cache-
misses,branches,branch-misses /PATH/TO/APPLICATION". "perf report -T --
stdio".

I get invalid results randomly. In the result, I get invalid (negative) 
process id or very big counter values. I am giving an example output:

# (For a higher level overview, try: perf report --sort comm,dso)
#
#  PID   TID   cycles           cycles  instructions  instructions         
instructions          instructions          instructions  instructions  
instructions  instructions  cache-references  cache-references      
cache-references     cache-references      cache-references  
cache-references  cache-references     cache-references  cache-misses  
cache-misses          cache-misses          cache-misses      cache-misses  
cache-misses  cache-misses      cache-misses  branches  branches              
branches              branches              branches  branches  
branches              branches  branch-misses  branch-misses         
branch-misses      branch-misses         branch-misses  branch-misses  
branch-misses  branch-misses    cycles    cycles                
cycles                cycles            cycles    cycles
  4487  4493        0                0             0             0  
  7494102348578251111   3488601301391995001                 41541         
  37571             0             0                 0                 0   
  7092150592960751218  7810209763779608678                  1611              
  2111                 0  7381246349212319846             0             0   
  7018136542324422001     28838465189274226                36            
  28             0                 0         0         0   
  8026367814273625705   7453295080785733747   7380943991784800358      
  7228         0                     0              0              0     
  28825271049347427  28825292623798627     28825270915195235            
  434              0              0         0         0   
  8462108034846451316                 85567             94053     98435
...

You can see invalid values such as 8026367814273625705. How can I fix this 
problem and get correct per-thread performance counters?

Thanks.

             reply	other threads:[~2013-07-29  8:55 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-07-29  8:52 Etem [this message]
2013-08-02 21:07 ` perf record + perf report Andi Kleen

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=loom.20130729T104749-103@post.gmane.org \
    --to=etem_deniz@hotmail.com \
    --cc=linux-perf-users@vger.kernel.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).