Re: gprof - wrong time results

[Paul Pluzhnikov]

Jan Schäfer <schaefer@neuro.informatik.uni-ulm.de> writes:

> and with -lc_p I get:
>
> --- time output ---
> real    0m1.664s
> user    0m1.562s
> sys     0m0.004s
> --- time output ---
>
> --- flat profile ---
> Jedes Muster zählt als 0.01 seconds.
>   %    kumulativ   Selbst            Selbst   Gesamt
>  Zeit   seconds   seconds  Aufrufe  s/Aufru  s/Aufru  Name
>  99.40     14.90    14.90  1000000     0.00     0.00  memset
>   0.60     14.99     0.09        1     0.09    14.99  foo(int)
>   0.00     14.99     0.00       15     0.00     0.00  strncmp
>   0.00     14.99     0.00       11     0.00     0.00  strlen
...
> mmmmh, strange...?

Strange indeed.

I don't know what the explanation might be. You may want to ask
in c.o.l.d.apps, specifying system details, etc.

Here is my result:

$ time ./a.out

real    0m3.312s
user    0m3.290s
sys     0m0.000s

$ gprof a.out
Flat profile:

Each sample counts as 0.01 seconds.
  %   cumulative   self              self     total
 time   seconds   seconds    calls   s/call   s/call  name
 99.38      3.21     3.21  1000000     0.00     0.00  memset
  0.62      3.23     0.02        1     0.02     3.23  foo
  0.00      3.23     0.00        1     0.00     0.00  __cxa_atexit
  0.00      3.23     0.00        1     0.00     0.00  _mcleanup
...

This appears to be in good agreement with 'time'.

System details: dual PIII/600MHz,
Red Hat Enterprise Linux ES release 3 (Taroon Update 4)
Linux dev1 2.4.21-27.0.2.ELsmp #1 SMP Wed Jan 12 23:35:44 EST 2005 i686 i686 
i386 GNU/Linux
glibc-profile-2.3.2-95.30
binutils-2.14.90.0.4-35    (gprof is part of binutils).

Cheers,
-- 
In order to understand recursion you must first understand recursion.
Remove /-nsp/ for email.