[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Bug-gnubg] Windows build 030203 crashes
From: |
Jim Segrave |
Subject: |
Re: [Bug-gnubg] Windows build 030203 crashes |
Date: |
Sun, 16 Feb 2003 00:49:38 +0100 |
User-agent: |
Mutt/1.2.5.1i |
On Sun 16 Feb 2003 (00:00 +0100), Øystein Johansen wrote:
> Albert Silver wrote:
>
> >Here are 3 URLs. AFTER installing the new Installation archive from
> >Oystein's site, download the new build as it is apparently a tad faster
> >and less prone to crash on some machines.
> >
> >The tutorial is in case you haven't seen it yet, as it should help you
> >get to know the program better.
> >
> >Oystein's site - http://home.online.no/%7Eoeysteij/
> >New build - http://home.online.no/~oeysteij/gnubg-O3.exe
> >Tutorial - http://www.bkgm.com/gnu/AllAboutGNU.html
> >
>
> The setup.exe file in the first URL, contains the gnubg-O3.exe file
> linked in the second URL. (Except that it's called gnubg.exe in the
> archive). I will therefore remove the gnubg-O3.exe file soon! gnubg-O3
> was just ment as a test for finding the problem of the crash of some
> systems.
>
> I'm starting to think of the idea of distributing builds optimised for
> Athlon and Intel processors seperatly. How does that sound?
Assuming that the O3 refers to gcc -O3, I begin to believe the FreeBSD
kernel folk who are adamant that there are serious bugs in -O3
optimisations. When it does run, does -O3 actually produce any
significant improvment over -O2?
I think per-CPU bilds is likely to be a major headache in the long run
- there's Intels, AMDs, Cyrix and they all have various
sub-flavours. Either build to the lowest common denominator or insist
people build locally. For Windows, I fear the latter is not an option.
--
Jim Segrave address@hidden