From:	jtgalkowski AT alum DOT mit DOT edu
Newsgroups:	comp.os.msdos.djgpp
Subject:	Re: Recently installed version of DJGPP's gcc hangs
Date:	Wed, 16 Sep 1998 18:23:33 GMT
Organization:	Deja News - The Leader in Internet Discussion
Lines:	92
Message-ID:	<6tovn5$h7k$1@nnrp1.dejanews.com>
References:	<Pine DOT SUN DOT 3 DOT 91 DOT 980916121747 DOT 5874P-100000 AT is>
NNTP-Posting-Host:	128.253.114.234
To:	djgpp AT delorie DOT com
DJ-Gateway:	from newsgroup comp.os.msdos.djgpp

In article <Pine DOT SUN DOT 3 DOT 91 DOT 980916121747 DOT 5874P-100000 AT is>,
  Eli Zaretskii <eliz AT is DOT elta DOT co DOT il> wrote:
>


[snip]


>
> Please do let it run longer, because it's quite possible that the
> compiler was paging due to insufficient memory (6MB free is not always
> enough, especially for large sources and some pathological cases).
> The disk activity seems to indicate that paging is indeed happening,
> and this would usually considerably slow down the compiler, so be
> patient.

Ran for over an hour on 25 MHz (I know, slow) 486.  Result was
the Page Fault in RMDB reported earlier today.  Based upon Charles
Sandmann's (sandmann AT clio DOT rice DOT edu) suggestion, I am double-checking
that I have the R4 versus R3 version of cwsdpmi.  The source in question
is one piece some 99,900 bytes C source that compiles and runs fine for
Win95 using Cygnus.  Doing this DOS version to provide users an
alternative on older Win 3.1 systems.

After that, Mr Sandmann indicates a long struggle with debugging
images of the execution or, my suggestion, perturbing the source
being compiled so to try to move the error out of the execution
space.  I s'ppose I could try breaking up the source unit in
question....

>
> If the compiler doesn't seem to finish even if you leave it to run for
> hours, please tell more:
>
>        1) what package are you trying to compile;
>        2) do you compile it using Make or with some batch file/manual
> 	  procedure;
>        3) details about the source file where it gets stuck: size,
> 	  source language, complexity;
>        4) the exact output of go2-v2.exe when run without arguments.
>
> If you don't have a disk cache installed, I'd suggest to install one
> (e.g., SmartDrv that comes with DOS); it can speed out things
> tremendously.  See section 3.9 of the DJGPP FAQ list (v2/faq210b.zip
> from the same place you get DJGPP) for other tips on system
> configuration.

I can provide the source itself.  I'm using a modified makefile--
simply one which compiles all the pieces to .o using "-c -O".
I'll build 'em after they all get done.  The culprit is a file
called "toam.c" which I am sending a zipped copy of to you and
Mr Sandmann just for completeness.  That is, I am grateful for
your suggestions and advice but don't expect anything more towards
what is, after all, basically my problem.

I am Using a Norton NCACHE2 with 1 MB cache size.  The package is
Professor Neng-Fa Zhou's BPROLOG from
http://www.cad.mse.kyutech.ac.jp/people/zhou/bprolog.html.

>
> > Also, while I've run gcc and gdb and bash many times before, it
> > was always on a Solaris or Ultra or Alpha.  Not quite sure what
> > to expect in this DOS environment.
>
> Expect them to work almost 100% like they would on Unix.  The
> differences in GDB functionality are described in section 12.1 of the
> FAQ.  Bash has a README.dos file that explains some DJGPP-specific
> features.

Thanks.  That's great.

[snip]

 go2-v2 is a diagnostic tool, and is also used
> to run raw COFF executables in some rare circumstances.  Section 22.12
> in the FAQ has more details.
>


Thanks very much.

 (:-)}

--
 Jan Theodore Galkowski

   500-446-2770

   http://www.hotwired.com/members/profile/algebraist/

-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/rg_mkgrp.xp   Create Your Own Free Member Forum

- Raw text -