--- Comment #23 from ebotcazou at gcc dot gnu dot org 2006-09-30 13:35
---
Should work now.
--
ebotcazou at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #22 from ebotcazou at gcc dot gnu dot org 2006-09-30 13:31
---
Subject: Bug 28096
Author: ebotcazou
Date: Sat Sep 30 13:31:29 2006
New Revision: 117331
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=117331
Log:
PR rtl-optimization/28096
* ifcvt.c (ch
--- Comment #21 from ebotcazou at gcc dot gnu dot org 2006-09-30 08:50
---
> This smells like an aliasing issue.
But it's not, it's if-conversion.
--
ebotcazou at gcc dot gnu dot org changed:
What|Removed |Added
--
--- Comment #20 from pinskia at gcc dot gnu dot org 2006-09-30 08:29
---
(In reply to comment #19)
> Confirmed:
>
> andl-144(%ebp), %ebx <--- bogus value in %ebx
> movl%eax, -144(%ebp)
This smells like an aliasing issue.
--
pinskia at gcc dot gnu dot org c
--- Comment #19 from ebotcazou at gcc dot gnu dot org 2006-09-29 20:40
---
> I don't think this bug has anything to do with excess precision. Valgrind
> shows that, when compiled at -O2, we are reading from uninitialized memory.
Confirmed:
andl-144(%ebp), %ebx <--- bogus
--
mmitchel at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P1
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28096
--- Comment #18 from ebotcazou at gcc dot gnu dot org 2006-09-20 17:39
---
Investigating.
--
ebotcazou at gcc dot gnu dot org changed:
What|Removed |Added
Assign
--- Comment #17 from ebotcazou at gcc dot gnu dot org 2006-09-20 17:38
---
> Yes, this is a regression.
> It works fine with -O2 with my system compiler (FC5 gcc, based on gcc 4.1).
> It also works fine with -O2 using my gcc 4.1 build.
> It fails with svn head.
Thanks for the info.
-
