Re: [PATCH] Fix PR 29715, fold produces &a - 4

["Andrew Pinski" <pinskia at gmail dot com>]

On 3/1/07, Richard Guenther <richard.guenther@gmail.com> wrote:
> This optimization doesn't look too useful at all now that we gimplify all this
> post/preinc/decrement anyway.  This is actually the only place where we look
> at this operators in fold, so I suggest removing the whole
> transformation instead.

Like this?  I agree that this is no longer needed as we do
gimplification and then optimize it later if we can (though sometimes
we don't because we don't recombined the PLUS with the comparision but
that is not actually tested :) ).

This is not the first time this code has bought troubles:
http://gcc.gnu.org/ml/gcc-patches/2004-03/msg00582.html
http://gcc.gnu.org/ml/gcc-patches/2000-07/msg00575.html
http://gcc.gnu.org/ml/gcc-patches/2004-03/msg02369.html


This was bootstrapped and tested many times on powerpc64-linux-gnu
with no regressions and built and tested for spu-elf with no
regressions.

Thanks,
Andrew Pinski

ChangeLog:
 * fold-const.c (fold_comparision): Remove the "foo++ == CONST" transformation.

Attachment: fixfold.diff.txt
Description: Text document