From: pavenis AT lanet DOT lv
To: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>, Gero Timann <gtimann AT nexgo DOT de>,
        djgpp AT delorie DOT com
Date: Mon, 9 Jul 2001 16:55:44 +0300
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
Subject: Re: gcc 3.0 problem ?
Message-ID: <3B49E210.16484.20A115@localhost>
References: <Pine DOT A41 DOT 4 DOT 05 DOT 10107091501200 DOT 25696-100000 AT ieva06 DOT lanet DOT lv>
In-reply-to: <Pine.SUN.3.91.1010709164315.8678C-100000@is>
X-mailer: Pegasus Mail for Win32 (v3.12c)
Reply-To: djgpp AT delorie DOT com

On 9 Jul 2001, at 16:44, Eli Zaretskii wrote:

> 
> On Mon, 9 Jul 2001, Andris Pavenis wrote:
> 
> > > Compile run's fine, but the resulting library doesn't work, if it's compiled 
> > > with optimation-flag -O2 or -O3. Ok, allegro is very a complicate near
> > > hardware programmed library, but, strange, -O1 and -O2/-O3 with 
> > > generating debugging code (=-g) works (strange !?). Never seen
> > > this with gcc 2.95.3, therefore it's maybe a gcc3.0-problem.
> > 
> > Known problem
> > 
> > Remove -fomit-frame-pointer from compiler options
> 
> Is this a bug? 

Perhaps so. I tried both removing -fomit-frame-pointer and 
-fstrict-aliasing (default with gcc-3.0) and only presence of
-fomit-frame-pointer seems to cause crashes when trying to use
Allegro compiled with gcc-3.0. I haven't studied it more detailed
though.

> I know that -fomit-frame-pointer makes debugging hard or 
> impossible, but the OP seems to say that the library doesn't work, not 
> just cannot be debug.

Andris