This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug target/67732] [SH] Strange LRA addsi3 usage
- From: "olegendo at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Tue, 29 Sep 2015 14:22:38 +0000
- Subject: [Bug target/67732] [SH] Strange LRA addsi3 usage
- Auto-submitted: auto-generated
- References: <bug-67732-4 at http dot gcc dot gnu dot org/bugzilla/>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=67732
--- Comment #3 from Oleg Endo <olegendo at gcc dot gnu.org> ---
Hm, those addressing mode issues are supposed to be improved by the AMS pass.
But it's not ready yet.
Compiling CSiBE with the current AMS branch, without LRA:
sum: 3333879 -> 3318319 -15560 / -0.466724 %
and teem is again the bigger chunk:
total: 1106660 -> 1098296 -8364 / -0.755788 %
It seems that the big improvement there comes from better utilization of
post-inc/pre-dec and index-mode FP loads/stores.
Unfortunately, running AMS + LRA doesn't work, so I can't get any numbers for
that ( https://gcc.gnu.org/bugzilla/show_bug.cgi?id=55212#c117 )