Date:	Fri, 21 May 1999 15:06:46 +0200
To:	pgcc AT delorie DOT com
Subject:	Re: Benchmark PGCC vs EGCS on a K6-2
Message-ID:	<19990521150646.H2804@cerebro.laendle>
Mail-Followup-To:	pgcc AT delorie DOT com
References:	<373F3AA2 DOT A446D611 AT informatik DOT hu-berlin DOT de> <Pine DOT LNX DOT 4 DOT 10 DOT 9905181826020 DOT 1284-100000 AT data DOT mandrakesoft DOT com> <19990519105631 DOT 40676 AT atrey DOT karlin DOT mff DOT cuni DOT cz> <19990519151204 DOT E1552 AT cerebro DOT laendle> <19990520103558 DOT 44750 AT atrey DOT karlin DOT mff DOT cuni DOT cz> <19990520233330 DOT J248 AT cerebro DOT laendle> <19990521105104 DOT 58741 AT atrey DOT karlin DOT mff DOT cuni DOT cz>
Mime-Version:	1.0
In-Reply-To:	<19990521105104.58741@atrey.karlin.mff.cuni.cz>; from Jan Hubicka on Fri, May 21, 1999 at 10:51:04AM +0200
X-Operating-System:	Linux version 2.2.9 (root AT cerebro) (gcc driver version pgcc-2.91.66 19990314 (egcs-1.1.2 release) executing gcc version 2.7.2.3)
From:	Marc Lehmann <pcg AT goof DOT com>
Reply-To:	pgcc AT delorie DOT com
X-Mailing-List:	pgcc AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

On Fri, May 21, 1999 at 10:51:04AM +0200, Jan Hubicka wrote:
> OK... :) So I have to start.
> Just a simple question. As I remember, the latest pgcc patch is relative
> to quite old version of egcs (1.1.x times).

No, there is a cvs repository that is mostly up-to-date, so a patch
relative to the one is fine (a patch relative to egcs-snapshot might even
work)

> while updating patch for new egcs snapshots. Do you have somewhere pgcc patches
> updated for this version, so I can do it there?

At the moment only via cvs (http://www.goof.com/pcg/cvs.html). It contains
all pgcc versions (except 1.1.2 and 1.1.3) and most egcs releases (vendor
branch EGCS).

cvs log gcc/version.c is a nice way to see the tags.

> BTW there are also some of my patches pending and especially the scheduling
> one and one that adds new predicates for i387 operands seems to be quite worthwhile.
> Maybe you might consider adding them into pgcc as well, so they should be
> included in pgcc version corresponding to 2.95.0.

Sure, I have no problems with breaking copyrights or stability issues,
much of the code in pgcc is questionable ;-)

> BTW I was also thinking about using integer unit for fp comparsions. This
> technique seems to help _a lot_ for Pentium (whats situation about ppro?)
> (4.5 MLoops p/s to 8 MLoops for FP loops)

I think its not overwhelming on the ppro, and needs very careful tuning
(as not to create worse code). The ppro has extra instructions to
alliviate this case, however, many people still want to create code that
runs on i586.

> The technique I use for XaoS is macro doing just integer comparsion of two
> float values. This works for nonnegative numbers only, so it is hardly useable
> for gcc (does gcc have any support for determining whether number if nonnegative?)

ieee brainies aside, there is some fast way to do this (I believe), I
have to search through my docs to see wether I can find it and wether its
suitable. ("ha, it only costs two jumps" ;)

if you want range analysis in gcc, yes, that would be cool, but no, gcc
does not have it.

> thinking about adding some new flag (-minteger-fp-comparisons) that can
> use simple scheme, where comparing of two floats is done by substracting them,
> storing result to memory as float and testing the sign bits for comparison
> with 0.
> 
> What do you think about this stuff?

Now thats an interesting idea.. Is the speed increase worth the memory
cost?

--  
      -----==-                                             |
      ----==-- _                                           |
      ---==---(_)__  __ ____  __       Marc Lehmann      +--
      --==---/ / _ \/ // /\ \/ /       pcg AT goof DOT com      |e|
      -=====/_/_//_/\_,_/ /_/\_\       XX11-RIPE         --+
    The choice of a GNU generation                       |
                                                         |

- Raw text -