Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!nrl-cmf!ames!amdcad!sun!pitstop!sundc!seismo!uunet!tiamat!jim
From: jim@tiamat.fsc.com (Jim O'Connor)
Newsgroups: comp.sources.d
Subject: Re: Perl 2.0 problem
Summary: -O munges perl2 on other machines, too
Message-ID: <382@tiamat.fsc.com>
Date: 8 Feb 89 21:15:02 GMT
References: <744@attdso.att.com> <4349@jpl-devvax.JPL.NASA.GOV> <763@attdso.att.com>
Organization: Filtration Sciences Corp., Chattanooga, TN
Lines: 26

In article <763@attdso.att.com>, tim@attdso.att.com (Tim J Ihde) writes:
> 
> I retrieved all the patches from your server, and applied them in turn.
> However, I still got the core dump when running base.lex.  After trying
> various things, I decided to try compiling again with -g and running under
> sdb.  Low and behold, this fixed the problem.  More specifically, removing
> the -O for debugging solved the problem.

When compiling perl2 on an Altos 2086 with Xenix 3.4b, the -O option also
produced a binary that would not run.  With the -O removed, everything
ran just fine. (Well, actually there were other problems, such as Altos's
yacc not being able to swallow perl.y, and not having the bcopy()-type
functions or the memcpy()-type functions, but I managed to get around most
of this.)

> >(By the way, I'm posting this from the terminal room at USENIX.  Fun!)
> 
> Gee, some guys get all the excitement. :-)

If I would've known they were going to have a terminal room, I would've
gone. 8-).

--jim
-------------
James B. O'Connor			jim@FSC.COM
Filtration Sciences Corporation		615/821-4022 x. 651