Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!cs.utexas.edu!usc!henry.jpl.nasa.gov!elroy.jpl.nasa.gov!jato!mars!kaleb
From: kaleb@mars.jpl.nasa.gov (Kaleb Keithley)
Newsgroups: comp.windows.x
Subject: Re: conflicting types for size_t
Message-ID: <2539@jato.Jpl.Nasa.Gov>
Date: 9 Jan 90 00:00:21 GMT
References: <2534@jato.Jpl.Nasa.Gov> <9001082225.AA01547@expire.lcs.mit.edu>
Sender: news@jato.Jpl.Nasa.Gov
Reply-To: kaleb@mars.UUCP (Kaleb Keithley)
Organization: Jet Propulsion Laboratory, Pasadena, CA.
Lines: 25

In article <9001082225.AA01547@expire.lcs.mit.edu> rws@EXPO.LCS.MIT.EDU (Bob Scheifler) writes:
>Does anyone really have
>evidence that gcc on the sparc gives a significant performance increase?

On our Sun 4/260 (SunOS 4.0.3, gcc 1.36) I have a prototype that under R3
would keep the server at 6% cpu pretty consistantly.  With R4 the same
prototype (unchanged, compiled and linked with R3 Xlib before building R4)
runs the cc compiled server at 0.00%, ditto for gcc compiled server.

These are the figures reported by the `top' program.

I know this question belongs in another group, but does anyone know how to
get even finer granularity or accuracy on the amount of cpu a process uses?

(Note to Bob Schleifler: It was -DNOSTDHDRS; boy do I feel stupid, somehow
I got -DNOSTHDRS and I must have looked at it ten times and didn't see it.
thanks)




Chewey, get us outta here!
                 
kaleb@mars.jpl.nasa.gov             (818)354-8771
Kaleb Keithley