On Wed, 2008-01-30 at 19:52 +0100, Mark Kettenis wrote:
> > From: Thiago Jung Bauermann <bauerman@br.ibm.com>
> > ppc_dl0_upper_regnum is used ppc_pseudo_register_{read,write}. Are you
> > saying that given it always has the value of ppc_fp0_regnum, I should
> > use that instead? Could be, but then there's the lack of symmetry in
> > using an fp variable in dl-related code. But that's minor. 
> 
> Yes, I think that using ppc_fp0_regnum is better, since it makes it
> explicit in the code that the decimal float (pseudo) registers are
> just a different representation of the floating point registers.
> 
> > ppcdl15_regnum is indeed unused. I kept it also for symmetry reasons,
> > since there's ppc_ev31_regnum. But it can be removed (specially if the
> > later is removed as well as you suggested).
> 
> Yes, that was what I was thinking.  The less uneeded clutter, the better.

Ok, removed those. I also refreshed it against the reworked SPE macro
cleanup patch. What about this?
-- 
[]'s
Thiago Jung Bauermann
Software Engineer
IBM Linux Technology Center