--- Comment #20 from spop at gcc dot gnu dot org 2007-07-04 07:08 ---
Fixed.
--
spop at gcc dot gnu dot org changed:
What|Removed |Added
Status|NEW
--- Comment #19 from spop at gcc dot gnu dot org 2007-07-04 07:04 ---
Subject: Bug 32457
Author: spop
Date: Wed Jul 4 07:04:31 2007
New Revision: 126305
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=126305
Log:
PR middle-end/32457
* tree-data-ref.c (analyze_siv
--
mmitchel at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P1
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32457
--- Comment #18 from sebpop at gmail dot com 2007-06-26 23:54 ---
Subject: Re: [4.3 Regression] Complete program optimized away (i686,
-ftree-vectorize)
Hi,
> I'm bootstrapping this patch on both amd64-linux and i686-linux.
The patch passed bootstrap and test on both machines with no
--- Comment #16 from spop at gcc dot gnu dot org 2007-06-26 18:02 ---
Subject: Re: [4.3 Regression] Complete program optimized away (i686,
-ftree-vectorize)
Hi,
The problem comes from the fact that estimated_loop_iterations_int
returns HWI whereas the code in analyze_subscript_affine_
--- Comment #15 from spop at gcc dot gnu dot org 2007-06-26 16:33 ---
Subject: Re: [4.3 Regression] Complete program optimized away (i686,
-ftree-vectorize)
Aha, thanks for the clarifications, and sorry for the noise.
I should read again some fortran tutorial or some more code
written
--- Comment #14 from burnus at gcc dot gnu dot org 2007-06-26 16:25 ---
> REAL, DIMENSION(0:100) :: RBOUND
> I thought that the array indices start from 1, and not from 0.
Well, dimension(5) means from 1 to 5, but e.g. dimension(-2:4:2) means the
indices {-2, 0, 2, 4}. And in the
--- Comment #13 from spop at gcc dot gnu dot org 2007-06-26 16:03 ---
Subject: Re: [4.3 Regression] Complete program optimized away (i686,
-ftree-vectorize)
I just reduced the problem to this snippet: the problem seems to be
that the data dependence analysis is wrong by answering that
--- Comment #12 from spop at gcc dot gnu dot org 2007-06-22 20:20 ---
Subject: Re: [4.3 Regression] Complete program optimized away (i686,
-ftree-vectorize)
> Do you still need the stuff mentioned in comment #7 or not? I would guess not.
>
No, I'm debugging that on an amd64. Thanks.
--- Comment #11 from burnus at gcc dot gnu dot org 2007-06-22 18:51 ---
> Grrr, I mean, I will test the patch now on amd64 ;-)
Do you still need the stuff mentioned in comment #7 or not? I would guess not.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32457
--- Comment #10 from sebpop at gmail dot com 2007-06-22 10:12 ---
Subject: Re: [4.3 Regression] Complete program optimized away (i686,
-ftree-vectorize)
Grrr, I mean, I will test the patch now on amd64 ;-)
On 6/22/07, Sebastian Pop <[EMAIL PROTECTED]> wrote:
> Okay, I'm testing that o
--- Comment #9 from spop at gcc dot gnu dot org 2007-06-22 10:11 ---
Subject: Re: [4.3 Regression] Complete program optimized away (i686,
-ftree-vectorize)
Okay, I'm testing that on an amd64 machine. Sorry for the confusion.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32457
--- Comment #8 from ubizjak at gmail dot com 2007-06-22 09:36 ---
(In reply to comment #7)
> I fail to see what changes. Both before and after applying my patch
> the assembly diff is the same. I cannot reproduce the bug on
> i686-linux.
This bug can be reproduced only on x86_64 host
--- Comment #7 from spop at gcc dot gnu dot org 2007-06-22 08:09 ---
Subject: Re: [4.3 Regression] Complete program optimized away (i686,
-ftree-vectorize)
I fail to see what changes. Both before and after applying my patch
the assembly diff is the same. I cannot reproduce the bug on
--- Comment #6 from spop at gcc dot gnu dot org 2007-06-22 07:52 ---
Subject: Re: [4.3 Regression] Complete program optimized away (i686,
-ftree-vectorize)
It is not the vectorizer that is in fault here: applying my patch does
not impact the output of -fdump-tree-vect-all-all except fo
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
CC||pinskia at gcc dot gnu dot
|
--- Comment #5 from spop at gcc dot gnu dot org 2007-06-21 21:38 ---
Subject: Re: [4.3 Regression] Complete program optimized away (i686,
-ftree-vectorize)
Thanks for pointing me to this bug. I'll have a look.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32457
--- Comment #4 from ubizjak at gmail dot com 2007-06-21 20:58 ---
Bisection points to:
Author: spop
Date: Wed Jun 20 23:42:28 2007
New Revision: 125900
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=125900
Log:
PR tree-optimization/32075
* tree-data-ref.c (subscr
--- Comment #3 from ubizjak at gmail dot com 2007-06-21 19:52 ---
Confirmed, for some reason we hit STOP in line 544 (subroutine KEEL). I'm not
familiar with Fortran, so perhaps someone who knows it should step through KEEL
to find what went wrong.
--
ubizjak at gmail dot com changed
--- Comment #2 from burnus at gcc dot gnu dot org 2007-06-21 17:22 ---
Further note:
-march=pentium2 works
-march=pentium3 fails
-march=pentium2 -msse fails
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32457
--- Comment #1 from burnus at gcc dot gnu dot org 2007-06-21 15:58 ---
Missed to say that yesterday's version (r125874) was ok.
Occurs here with 125909 (bootstrapped) and 125922 (only build). svn status
shows no changes in my tree.
However, it could not be reproduced by Uros with 125920
21 matches
Mail list logo
