Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!usc!apple!bionet!ames!sgi!shinobu!odin!flobb4!micah
From: micah@flobb4.csd.sgi.com (Micah Altman)
Newsgroups: comp.sys.sgi
Subject: Re: profiler
Keywords: profiler
Message-ID: <1990Dec11.162333.27363@odin.corp.sgi.com>
Date: 11 Dec 90 16:23:33 GMT
References: <1990Dec5.005445.21368@jarvis.csri.toronto.edu> <1990Dec10.182523.9827@odin.corp.sgi.com>
Sender: news@odin.corp.sgi.com (Net News)
Organization: Silicon Graphics, Inc.  Mountain View, CA
Lines: 26

In <1990Dec10.182523.9827@odin.corp.sgi.com> bennett@sgi.com (Jim Bennett) writes:

>In article <1990Dec5.005445.21368@jarvis.csri.toronto.edu> corkum@csri.toronto.edu (Brent Thomas Corkum) writes:
>>A quick question, is there a profiler available for a C program
>>running on a 4D/25. The profiler being a transparent process, meaning
>>I don't want to place function calls within my program, that will tell me
>>the amount of cpu time spent in each function within a program.
>>
>>If there isn't, what do programmers use to tackle such a question?
>>
>>Brent Corkum
>>corkum@boulder.civ.toronto.edu

>See pixie(1).  It fixes up your binary with basic block counters.

>Ther other profiler is profil(2).  It samples your PC at 10 millisecond
>intervals, but you have to put the call to profil explicitly in your
>program to enable the PC sampling.

You can get automatic pc-sampling of your code by compiling with 
-p ( see man page on prof(1), subject heading "Using pc-sampling" ).
This gives you a routine by routine ( not a basic block count ) pc-sampled
profile without having to insert any profil(2) calls.

- Micah Altman
"Computational Juggler"