From:	pavenis AT lanet DOT lv
Message-ID:	<B0000091877@stargate.astr.lu.lv>
To:	salvador <salvador AT inti DOT gov DOT ar>, djgpp-workers AT delorie DOT com
Date:	Fri, 25 Jun 1999 17:00:16 +0300
MIME-Version:	1.0
Subject:	Re: gcc-2.95 benchmark
In-reply-to:	<199906231532.LAA17753@delorie.com>
X-mailer:	Pegasus Mail for Win32 (v3.11)
Reply-To:	djgpp-workers AT delorie DOT com
X-Mailing-List:	djgpp-workers AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

On 23 Jun 99, at 10:24, salvador wrote:

> Hi All:
> 
>   I tried the BYTE benchmarks with gcc 2.95 (june snapshot), even when it
> isn't much better than other versions is overall the best for both Pentium
> and K6 processors.
> 
>   The results are available in:
> 
>   http://www.geocities.com/SiliconValley/Vista/6552/compila.html

Some more thoughts:

When building earlier versions (egcs-1.1.X) of egcs I enabled 
HAIFA scheduler (--enable-haifa when running configure). For 
building current snapshots of gcc-2.95 prerelease for DJGPP I 
didn't enable it.
This can give small changes in performance at least it was so with
egcs-1.1.X but I don't know how it will be with gcc-2.95 as I haven't 
tried. 

Perhaps it would be worth to try also snapshot built with configured 
with --enable-haifa
 
Andris

- Raw text -