This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Fix PR rtl-optimization/28096
- From: Roger Sayle <roger at eyesopen dot com>
- To: Eric Botcazou <ebotcazou at libertysurf dot fr>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Sat, 30 Sep 2006 10:17:36 -0600 (MDT)
- Subject: Re: Fix PR rtl-optimization/28096
Hi Eric,
Good catch and thanks for the fix for PR 28096. But one hunk of your
change got my grey cells thinking on an unrelated GNU coding style issue.
On Sat, 30 Sep 2006, Eric Botcazou wrote:
>--- ifcvt.c (revision 117300)
> +++ ifcvt.c (working copy)
> @@ -2424,7 +2424,7 @@ check_cond_move_block (basic_block bb, r
> src = SET_SRC (set);
> if (!REG_P (dest)
> || (SMALL_REGISTER_CLASSES && HARD_REGISTER_P (dest)))
>- return false;
>+ return FALSE;
>
> if (!CONSTANT_P (src) && !register_operand (src, VOIDmode))
> return FALSE;
I appreciate that this clean up is the minimal change to be consistent
with the rest of the function, but statistically the lower case "false"
is more prevalent in GCC's source code.
"return false;" 1514 occurences
"return FALSE;" 147 occurences
So whilst I fully agree with your change during stage 3, we should
probably plan a Kazu-style clean-up of the tree once we return to
stage 1, when we have the freedom to enforce global consistency rather
than local (function-wide) consistency. I suspect this is already part
of the ongoing transition from "int" to "bool" for Boolean types.
Roger
--