Newsgroups: comp.parallel.pvm
From: sm1hk@tuhhco.rz.tu-harburg.de (Heiko Koenig)
Subject: What makes pvm_serror with 13%run-time??
Organization: Technische Universtaet Hamburg-Harburg
Date: 28 Jun 1994 11:12:27 GMT
Message-ID: <2up0ir$duh@rztsun.tu-harburg.de>


Hallo pvm-user,

after I have test my programm with 'prof' I saw, that pvm_serror has 13.1%
run-time (the top mark in my Protokoll :-( ). 
I have never called that routine! - And after a look in the pvm-3.2.6-Sources
I have found out, that this routine isn't called of any other routine of pvm! -
So, whats happen with pvm_serror() ??



prof Protokoll:

 %time  cumsecs  #call  ms/call  name
  13.5    31.14                  _vfc_yuvread_pal
  13.1    61.39                  _pvm_serror
  12.9    91.14                  mcount
  11.5   117.6330192265     0.00  _bcopy
   9.6   139.7960382301     0.00  .mul
   7.0   156.00                  _vfc_yuv2y_pal
   6.2   170.40                  _midtobuf
   5.6   183.26    100   128.60  _picshift2
   5.1   194.92  24438     0.48  _write
   4.9   206.13    100   112.10  _CompressPic
   4.4   216.18                  _pvm_pkbyte
   3.4   224.08    413    19.13  _ioctl
   1.4   227.38  42295     0.08  _select
   0.5   228.47  17919     0.06  _read
   0.2   228.84                  _mxfer
   0.1   229.10      1   260.00  _sethard
   0.1   229.29  26198     0.01  _cfree
   0.1   229.44  26173     0.01  _malloc
   0.1   229.58  44950     0.00  _bzero
   0.0   229.69  15065     0.01  _readbyte
   0.0   229.79  15482     0.01  _waitwrite




Heiko


Koenig@tu-harburg.d400.de

