*BSD News Article 14383


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!agate!howland.reston.ans.net!bogus.sura.net!darwin.sura.net!haven.umd.edu!uunet!pipex!bnr.co.uk!demon!centrix.demon.co.uk!damian
Newsgroups: comp.os.386bsd.questions
From: damian@centrix.demon.co.uk (damian)
Subject: Re: gcc eating up my machine
References: <9304120856.aa13605@gate.demon.co.uk> <1993Apr12.140748.343@nhqvax.hq.nasa.gov>
Distribution: world
Organization: Centrix
Date: Tue, 13 Apr 1993 00:13:34 +0000
Message-ID: <9304130133.aa09013@gate.demon.co.uk>
Sender: usenet@demon.co.uk
Lines: 23

Hmm, I don't the <test>?<value1>:<value2> is the problem here because certainly
in this X source (I think it was Xmailtool) the offending code was in a
header file, it had nothing else in it but this static array. I cut down the
size of the array (Who wants the logo anyway? :-)), it then went through
fine. One thing that did confuse the hell out of me and makes me think it
is a gcc bug is that one a few occasions it compiled fine (that really helped
me to track it down!). Sounds like a variable not initialised or even something
wierd going on with the memory management in the OS).

Oh I agree with you that if the nice value is lower it can and should hog
the cpu, but these were at the same level (possibly the cc was lower priority,
because I ran it in background). I think SCO allows two progs at the same
level to grab the CPU equally, but I have a feeling that SCO doesn't let
anyone hog the CPU regardless of priority, certainly nice doesn't seem to
work as well as it used to.

Damian
-- 
+----------------------------+------------------------------------------------+
| Damian Ivereigh            | If you can't suss out what this is replying to |
| damian@centrix.demon.co.uk | get a threaded news reader, like trn. :-)      |
| Twickenham, U.K.           | This is the best way to cut wasted traffic     |
+----------------------------+------------------------------------------------+