Subject: Re: .s vs .S
To: Jason R Thorpe <thorpej@zembu.com>
From: Chris Gilbert <chris@paradox.demon.co.uk>
List: tech-kern
Date: 01/26/2001 23:04:55
On Monday 22 January 2001  4:53 pm, Jason R Thorpe wrote:
> On Mon, Jan 22, 2001 at 01:37:58PM +0000, Ben Harris wrote:
>  > At the moment, sys/arch contains many assembler files, some of them with
>  > names ending in ".s" and some in ".S".  As far as I can see,
>  > emitrules() config(8) emits the same rules for assembling both of these
>  > (being invocations of ${NORMAL_S}, but emitsfiles() only puts ".s" files
>  > into ${SFILES}, which means that "make depend" doesn't check
>  > dependencies for ".S" files.
>
> Ooops, yes.  emitsfiles() should include .S files.  .S is what GCC uses
> for "assembler with cpp".  If we expect to pass an asm file through
> CPP, we should be using .S.  (This basically means "always use .S in
> NetBSD".)

I've submitted a PR on this (12051) with a patch that Richard Earnshaw 
provided to the arm32 list.  I've tested it on arm32 and i386 (and sparc, but 
it doesn't have any .S files...) seems to work as desired :)

Cheers,
Chris