> In article <V65t61KlN2yj.csx0zJNL@smtp.free.fr> you write:
> >This is quite strange. I've been on the RiscBSD
> >mailing-list (this is NetBSD/arm32) and it was
> >said that NetBSD required *signed* char (I don't
> >remember at which level), and this was annoying
> >because the ARM processors don't have a signed
> >byte-load or it is slower than the unsigned one.

arm32 has always used unsigned chars.

Cheers,
			Mark