www.delorie.com/archives/browse.cgi   search  
Mail Archives: pgcc/1998/01/24/00:27:17

X-POP3-Rcpt: mlehmann AT universe DOT sgh-net DOT de
24 Jan 1998 00:27:17 +0100 (CET) :
From: Ronald Wahl <Ronald DOT Wahl AT Informatik DOT TU-Chemnitz DOT DE>
X-Sender: rwa AT goliath DOT csn DOT tu-chemnitz DOT de
To: Steve Bergman <steve AT netplus DOT net>
cc: beastium-list AT Desk DOT nl
Subject: Re: Status of AMD K6 Support
In-Reply-To: <34C92380.B2C424BC@netplus.net>
Message-ID: <Pine.LNX.3.96.980124001554.4337A-100000@goliath.csn.tu-chemnitz.de>
MIME-Version: 1.0
Sender: Marc Lehmann <pcg AT goof DOT com>
Status: RO
Lines: 39

On Fri, 23 Jan 1998, Steve Bergman wrote:

> What is the current status of AMD K6 optimization in egcs/pgcc? 
> Information on this subject seems virtually nonexistent.  What I have
> gleaned about it amounts to this recommendation:
> 
> gcc -O6 -mamdk6 -fno-software-pipe -fno-swap-for-agi
> -fno-schedule-stack-reg-insns -fno-exceptions 

The options no-software-pipe, no-swap-for-agi and
no-schedule-stack-reg-insns are now included in -mamdk6.

> I am running redhat 5.0 (glibc2) and am specifically interested in
> optimizing Mesa, which is an OPENGL compatible graphics library.  It is
> fpu intensive and I notice that there are some recommendations for fpu
> intensive programs in the FAQ.  I have compiled the library in the past
> with "-O6 -mamdk6" and found it to be slightly slower than standard gcc
> 2.7.*.  Does anyone have any recommendations?

Try -funroll-loops or -funroll-all-loops. This may give a speedup or a
slowdown. Maybe you should also check the latest snapshots. But note: The
code generated by this two options is still broken in some cases. What you
also could try is to optimize the assembler routines in Mesa. Try to
minimize the fpu instructions (e.g. remove the fxch instructions and
adapt the code) and use integer code in parallel with fpu code. Further
more recommendations you will find in the optimization guide that is
located here: http://www.amd.com/K6/k6docs.

.... but the best optimization you will get with a 3D accelerator by 3dfx
since these chips are supported by Mesa.

ron

-- 
\ Ronald Wahl --- rwa AT informatik DOT tu-chemnitz DOT de   \
 \ WWW: http://www.tu-chemnitz.de/~row             \
  \ Talk: rwa AT goliath DOT csn DOT tu-chemnitz DOT de            \
   \ PGP key available by finger to my email address \


- Raw text -


  webmaster     delorie software   privacy  
  Copyright © 2019   by DJ Delorie     Updated Jul 2019