From: Weiqi Gao <weiqigao AT a DOT crl DOT com>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: GCC bugs
Date: Mon, 17 Jan 2000 19:07:14 -0600
Organization: CRL Network Services
Lines: 44
Message-ID: <3883BCC2.2CC1DCD9@a.crl.com>
References: <01bf611a$9c6e9500$LocalHost AT alex>
NNTP-Posting-Host: a116013.stl1.as.crl.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 4.61 [en] (X11; I; Linux 2.2.12-20 i586)
X-Accept-Language: en
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

"Alexei A. Frounze" wrote:
> 
> Hi guys!
> 
> I've found a very strange thing.
> When I run GCC from a command line:
>     c:\kernel\!flat\>gcc -c kernel.c
> and I run a batch file (make-c.bat):
>     @echo off
>     gcc -c kernel.c
> I get different output coff files (kernel.o).
> 
> For example, the first method generates a kernel.o that contains 4 sections
> (.text, .data, .bss and .eh_fram), but the second generates only 3 standard
> sections (.text, .data and .bss).
> 
> What does it mean?

It means your first method and second method didn't do the same thing. 
Are you using aliases, doekey macros? Do you have duplicated commands in
your $PATH?  Try the two methods with absolute paths in all places.

Do you have another brand of GCC installed?  An earlier version?

> The second thing that I wonder is how the GCC works itself.
> Sometimes it terminates with internal error (GPF I think), but if you run
> it again immediately after the crash, it may work normally.
> 
> What does it mean?

I had this happen to me a few times with an earlier version of DJGPP.

> Both these bugs are present on different computers where I use DJGPP.
> 
> Your comments...

The GCC documentation says clearly that a GCC crash under any
circumstance, with legal or illegal source code, is considered a bug. 
If you can reproduce the crash, report a bug with ways to reproduce the
behavior.

-- 
Weiqi Gao
weiqigao AT a DOT crl DOT com