This is the mail archive of the cygwin@cygwin.com mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Percent of CPU this job got > 100%


"Shankar Unni" <shankar@cotagesoft.com> wrote in message bmmjhd$mmf$1@sea.gmane.org">news:bmmjhd$mmf$1@sea.gmane.org...
> Christopher Faylor wrote:
>
> > On Thu, Oct 16, 2003 at 08:38:33AM +0200, Alex Vinokur wrote:
>
> >>$ /bin/time -v a 1000
> >>[...]
> >> Percent of CPU this job got: 133%
>
> > Automatic, built-in meanness at no extra charge.
>
> :-).
>
> Alex: is it possible your program is multi-threaded, and you have one of
> those shiny new hyperthreading P4s? Remote, but that could give you this
> behavior..

No, it isn't.


Here is program t.c :

------------------
#include <stdio.h>
#include <assert.h>
int main (int argc, char** argv)
{
int i;
  assert (argc > 1);
  for (i = 0; i < atoi(argv[1]); i++);
  return 0;
}
------------------

>
> But even otherwise, small discrepancies between wall clock and computed
> CPU clock are not uncommon, and given that your running time is pretty
> close to one CPU "HZ" tick, you can get weird anomalies like this.
>
> Try timing "a 1000000".
> --
> Shankar.
>

===========================================
Windows 2000 Professional
CYGWIN_NT-5.0 1.5.4(0.94/3/2)
gcc version 3.3.1 (cygming special)
GNU time 1.7
===========================================

$ gcc t.c

$ command time -v a 1000

        Command being timed: "a 1000"
        User time (seconds): 0.03
        System time (seconds): 0.03
        Percent of CPU this job got: 127%
        Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.04
        [---omitted---]

$ command time -v a 100000000

        Command being timed: "a 100000000"
        User time (seconds): 18.82
        System time (seconds): 0.04
        Percent of CPU this job got: 99%
        Elapsed (wall clock) time (h:mm:ss or m:ss): 0:19.03
        [---omitted---]


 =====================================
   Alex Vinokur
     mailto:alexvn@connect.to
     http://mathforum.org/library/view/10978.html
     news://news.gmane.org/gmane.comp.lang.c++.perfometer
   =====================================






--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]