From: Eugene Leitl <eugene DOT leitl AT lrz DOT uni-muenchen DOT de>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 20 May 1999 10:34:18 -0700 (PDT)
To: pgcc AT delorie DOT com
Subject: Re: Benchmark PGCC vs EGCS on a K6-2
In-Reply-To: <19990520160622.20808@atrey.karlin.mff.cuni.cz>
References: <3743ADE8 DOT C938ADBB AT informatik DOT hu-berlin DOT de>
	<Pine DOT NEB DOT 3 DOT 96 DOT 990520070839 DOT 20661O-100000 AT slavery DOT tainted DOT org>
	<19990520160622 DOT 20808 AT atrey DOT karlin DOT mff DOT cuni DOT cz>
X-Mailer: VM 6.43 under 20.4 "Emerald" XEmacs  Lucid
Message-ID: <14148.18189.460333.721469@lrz.de>
Reply-To: pgcc AT delorie DOT com
X-Mailing-List: pgcc AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

Jan Hubicka writes:

 > Problem with K6 is, that it seems to be sinsitive for memory layout. So
 > cleaning up memory by starting some large program or rebooting is recommended
 > as well...

I don't remember where I heard it, but it seems code and stack
aligment matters a lot for the K6.