--- Comment #12 from ubizjak at gmail dot com 2010-09-02 06:05 ---
Created an attachment (id=21652)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=21652action=view)
preprocessed source, fails with valgrind on linux
A non-sse4 sse2 capable host (core2 class) is needed to trigger
--- Comment #13 from ubizjak at gmail dot com 2010-09-02 11:24 ---
(In reply to comment #12)
A non-sse4 sse2 capable host (core2 class) is needed to trigger valgrind
failure.
This is actually PR45386. Sorry for the noise...
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45484
--- Comment #1 from howarth at nitro dot med dot uc dot edu 2010-09-01
16:21 ---
Original patch submitted with
http://gcc.gnu.org/ml/gcc-patches/2010-08/msg00440.html.
Test cases added with http://gcc.gnu.org/ml/gcc-patches/2010-08/msg00503.html.
--
howarth at nitro dot med dot uc
--- Comment #2 from howarth at nitro dot med dot uc dot edu 2010-09-01
19:26 ---
I am finding in current gcc trunk that these tests randomly pass and fail if
you repeat them enough times. This correlates with Richard Henderson
observation that the failures at r163660 looks like random
--- Comment #3 from ebotcazou at gcc dot gnu dot org 2010-09-01 20:57
---
Original patch submitted with
http://gcc.gnu.org/ml/gcc-patches/2010-08/msg00440.html.
I think I can prove that my patch doesn't affect code generation, except
possibly on the SPARC, so I'm a little skeptical
--- Comment #4 from howarth at nitro dot med dot uc dot edu 2010-09-01
23:11 ---
(In reply to comment #3)
I think I can prove that my patch doesn't affect code generation, except
possibly on the SPARC, so I'm a little skeptical about your diagnosis. Are
you
sure it's not
--- Comment #5 from howarth at nitro dot med dot uc dot edu 2010-09-01
23:17 ---
The code generate IS random. I repeatedly execute...
/Users/howarth/darwin_objdir/gcc/xgcc -B/Users/howarth/darwin_objdir/gcc/
/Users/howarth/gcc/gcc/testsuite/gcc.c-torture/execute/builtins/sprintf-chk.c
--- Comment #6 from howarth at nitro dot med dot uc dot edu 2010-09-01
23:21 ---
I should also note that sometimes the test in comment 5 does produce identical
code between two instances of executing the compile command whereas other times
I get a crash...
--- Comment #7 from howarth at nitro dot med dot uc dot edu 2010-09-02
04:59 ---
Created an attachment (id=21651)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=21651action=view)
preprocessed source of reduced failing test case
--
--- Comment #8 from howarth at nitro dot med dot uc dot edu 2010-09-02
05:05 ---
At r163660, the reduced test case (of just the test3 subroutine of
sprint-chk.c) triggers the compiler ICE at high frequency on
x86-apple-darwin10...
/Users/howarth/darwin_objdir/gcc/xgcc
--- Comment #9 from howarth at nitro dot med dot uc dot edu 2010-09-02
05:10 ---
Note that in current gcc trunk (r163747) with...
http://gcc.gnu.org/ml/gcc-patches/2010-08/msg01916.html
and
http://gcc.gnu.org/bugzilla/attachment.cgi?id=21650
the frequency of the test case ICEing
--- Comment #10 from ubizjak at gmail dot com 2010-09-02 05:47 ---
(In reply to comment #8)
Since this doesn't backtrace in gdb, I recompiled dwarf2out.c with the
patch...
You should use bigger hammer.
Try valgrind using following procedure:
a) Create a preprocessed source
On Sep 1, 2010, at 10:47 PM, ubizjak at gmail dot com gcc-bugzi...@gcc.gnu.org
wrote:
--- Comment #10 from ubizjak at gmail dot com 2010-09-02 05:47
---
(In reply to comment #8)
Since this doesn't backtrace in gdb, I recompiled dwarf2out.c with
the patch...
You should
--- Comment #11 from pinskia at gmail dot com 2010-09-02 05:55 ---
Subject: Re: r163660 ICEs gcc.c-torture/execute/builtins/sprintf-chk.c
compilation, -Os at -m32
On Sep 1, 2010, at 10:47 PM, ubizjak at gmail dot com
gcc-bugzi...@gcc.gnu.org
wrote:
--- Comment #10 from
14 matches
Mail list logo
