Subject: Re: stmw/lmw [Was: CVS commit: src/sys/arch/powerpc/powerpc]
To: Matt Thomas <matt@3am-software.com>
From: Eduardo Horvath <eeh@NetBSD.ORG>
List: source-changes
Date: 08/05/2003 20:27:01
On Sun, Aug 03, 2003 at 06:41:30PM -0700, Matt Thomas wrote:
> Actually, when I was doing my internet checksum tests, I found
> individual loads to be *faster* (by 20-25%) than lmw/stmw. This may be
> why gcc doesn't emit them for saving/restoring callframes.
What processors did you run your tests on?
Eduardo