From owner-freebsd-current  Wed Nov 29 22: 8:47 2000
Delivered-To: freebsd-current@freebsd.org
Received: from mailman.zeta.org.au (mailman.zeta.org.au [203.26.10.16])
	by hub.freebsd.org (Postfix) with ESMTP id 2C59C37B400
	for <current@FreeBSD.ORG>; Wed, 29 Nov 2000 22:08:43 -0800 (PST)
Received: from bde.zeta.org.au (bde.zeta.org.au [203.2.228.102])
	by mailman.zeta.org.au (8.8.7/8.8.7) with ESMTP id RAA23773;
	Thu, 30 Nov 2000 17:08:29 +1100
Date: Thu, 30 Nov 2000 17:09:01 +1100 (EST)
From: Bruce Evans <bde@zeta.org.au>
X-Sender: bde@besplex.bde.org
To: Julian Elischer <julian@elischer.org>
Cc: current@FreeBSD.ORG
Subject: Re: slight improvement in locore.s?
In-Reply-To: <3A2525DD.3D1E8D5A@elischer.org>
Message-ID: <Pine.BSF.4.21.0011301650450.4214-100000@besplex.bde.org>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-freebsd-current@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG

On Wed, 29 Nov 2000, Julian Elischer wrote:

> Bruce Evans wrote:
> > 
> > On Thu, 23 Nov 2000, Julian Elischer wrote:
> > 
> > > locore.s includes:
> > > #define ALLOCPAGES(foo) \
> > >         movl    R(physfree), %esi ; \
> > >         movl    $((foo)*PAGE_SIZE), %eax ; \
> > >         addl    %esi, %eax ; \
> > >         movl    %eax, R(physfree) ; \
> > >         movl    %esi, %edi ; \
> > >         movl    $((foo)*PAGE_SIZE),%ecx ; \
> > >         xorl    %eax,%eax ; \
> > >         cld ; \
> > >         rep ; \
> > >         stosb
> > >
> > >
> > > might it be a very slight optimisation to change this to:
> > > #define ALLOCPAGES(foo) \
> > >         movl    R(physfree), %esi ; \
> > >         movl    $((foo)*PAGE_SIZE), %eax ; \
> > >         movl    %eax, %ecx ; \
> > >         addl    %esi, %eax ; \
> > >         movl    %eax, R(physfree) ; \
> > >         movl    %esi, %edi ; \
> > >         xorl    %eax,%eax ; \
> > >         cld ; \
> > >         rep ; \
> > >         stosb
> > 
> > This can be improved more (3 instructions) by not loading physfree into
> > the wrong register, and depending on stosb to increment the register,
> 
> so your assembly code would look like? (just curious).

This is completely untested:

#define ALLOCPAGES(foo) \
	movl    R(physfree),%edi ;		\
	movl    $(foo)*PAGE_SIZE,%ecx ;		\
	movl	%edi,%esi;			\
	xorl    %eax,%eax ;			\
	cld ;					\
	rep ;					\
	stosb ;					\
	movl    %edi,R(physfree)

I wouldn't write it like this if either space or time efficiency were
important.

> > but it should be written in C anyway.  A relocation macro like R()
> > should work just as well in C as in asm.  In C, the above is:
> > 
> >         bzero(R(physfree), (foo) * PAGE_SIZE);
> >         R(physfree) += (foo) * PAGE_SIZE);
> >         return (R(physfree));   /* In unusual as well as wrong reg %esi. */

Oops, this should return the original value of physfree.  %esi is used to
hold this value, so it is not "wrong".  Setting it takes one instruction, so
my version is only 2 instructions shorter.

Bruce



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message