http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
--- Comment #12 from Jakub Jelinek jakub at gcc dot gnu.org 2013-01-03
09:05:49 UTC ---
Author: jakub
Date: Thu Jan 3 09:05:43 2013
New Revision: 194837
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=194837
Log:
PR
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
Jakub Jelinek jakub at gcc dot gnu.org changed:
What|Removed |Added
CC||jakub at
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
Andrew Pinski pinskia at gcc dot gnu.org changed:
What|Removed |Added
Known to work||3.4.6
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
--- Comment #3 from H.J. Lu hjl.tools at gmail dot com 2013-01-01 18:06:37
UTC ---
(const_int 129 [0x81]) isn't considered as a valid
const int for QImode.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
--- Comment #4 from H.J. Lu hjl.tools at gmail dot com 2013-01-01 18:53:32
UTC ---
The original regression was introduced by
http://gcc.gnu.org/ml/gcc-cvs/2004-05/msg00653.html
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
Steven Bosscher steven at gcc dot gnu.org changed:
What|Removed |Added
Status|NEW
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
--- Comment #6 from Steven Bosscher steven at gcc dot gnu.org 2013-01-01
19:51:12 UTC ---
The nonsense set comes from force_operand:
Breakpoint 7, unroll_loop_runtime_iterations (loop=0x76b46f68) at
../../trunk/gcc/loop-unroll.c:1041
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
--- Comment #7 from Steven Bosscher steven at gcc dot gnu.org 2013-01-01
19:52:48 UTC ---
... but actually it looks like it exists even before that:
(gdb) p debug_rtx(desc-niter_expr)
(udiv:QI (minus:QI (const_int -126
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
--- Comment #8 from Steven Bosscher steven at gcc dot gnu.org 2013-01-01
20:44:12 UTC ---
I would expect the insn to match the movqi_internal insn:
(define_insn *movqi_internal
[(set (match_operand:QI 0 nonimmediate_operand =q,q ,q
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
Steven Bosscher steven at gcc dot gnu.org changed:
What|Removed |Added
Status|ASSIGNED|NEW
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55838
--- Comment #10 from H.J. Lu hjl.tools at gmail dot com 2013-01-01 23:02:02
UTC ---
This patch avoids using (const_int 129 [0x81]) as step in QImode:
diff --git a/gcc/loop-iv.c b/gcc/loop-iv.c
index 50b7536..aafaae4 100644
---
11 matches
Mail list logo