This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] Fix simd-2.c failure on powerpc
- From: Roger Sayle <roger at www dot eyesopen dot com>
- To: Richard Henderson <rth at redhat dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Thu, 17 Apr 2003 16:51:41 -0600 (MDT)
- Subject: Re: [PATCH] Fix simd-2.c failure on powerpc
On Thu, 17 Apr 2003, Richard Henderson wrote:
> On Thu, Apr 17, 2003 at 04:15:29PM -0600, Roger Sayle wrote:
> > The function reg_overlap_mentioned_p
> > might potentially be called with a zero_extract or a sign_extract
> > from somewhere else in GCC.
>
> Correct. Which is why I think that is the proper fix, and
> that the change in combine is superfluous.
Agreed.
Hopefully, someone will improve can_combine_p to handle SUBREGs and
other SET_DESTs. As currently written, we should never be calling
reg_used_between. If for nothing else, my combine change would be
a performance optimization. Superfluous admittedly, but not incorrect.
Roger
--
p.s. Was that an approval for just the rtlanal.c hunk?