--- Comment #4 from domob at gcc dot gnu dot org 2008-10-04 06:45 ---
Created an attachment (id=16461)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16461action=view)
Patch addressing most of the specific remarks
This patch (posted to the list some time ago) addresses most of the
--- Comment #31 from nickc at redhat dot com 2008-10-04 08:27 ---
Subject: Re: [cygming] Invalid alignment for SSE store
to .comm data generated with -O3
the patch looks ok but unfortunately does not always solves the problem,
something in the chain misalignes the symbol
This
--- Comment #33 from wbrana at gmail dot com 2008-10-04 09:22 ---
results with -fno-tree-pre
1749 - 4.4.0 20080926 (experimental)
1701 - 4.3.2
2476 - 4.2.4
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21485
--- Comment #4 from domob at gcc dot gnu dot org 2008-10-04 10:16 ---
The problem is clearly that because of the error the type-bound procedure
remains in a somewhat dubious state (its pass_arg_num is invalid) and this
causes the ICE when resolving an actual call to it afterwards. This
--- Comment #5 from sfilippone at uniroma2 dot it 2008-10-04 10:29 ---
(In reply to comment #4)
The problem is clearly that because of the error the type-bound procedure
remains in a somewhat dubious state (its pass_arg_num is invalid) and this
causes the ICE when resolving an actual
--- Comment #6 from domob at gcc dot gnu dot org 2008-10-04 10:40 ---
(In reply to comment #5)
Hmm. I see that in my previous comment #3 I said the wrong thing: the attached
sample code should be correct, once the name in the PASS argument is fixed.
The reasoning behind #3 is that
--- Comment #14 from pault at gcc dot gnu dot org 2008-10-04 11:56 ---
(In reply to comment #13)
In fact the ICE is precisely due to the argument not being a restricted
expression, since the size of x and its kind are unavailable to the caller
and
the interfacing cannot be
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
CC||rguenth at gcc dot gnu dot
|
--- Comment #3 from kazu at gcc dot gnu dot org 2008-10-04 14:44 ---
Added a pach URL above.
--
kazu at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #34 from rguenth at gcc dot gnu dot org 2008-10-04 15:11
---
Fastest result on a Intel Core Duo with
gcc-4.1 -O3 -fomit-frame-pointer -fno-tree-pre -fno-inline -fschedule-insns:
1273
the interesting thing is that with the above we if-convert
if (array[k] array[k
--- Comment #35 from rguenth at gcc dot gnu dot org 2008-10-04 15:58
---
Another missed optimization on the tree level is hoisting of the load of
array[k*4] before the k j condition which is possible after the PRE
insertion:
bb 3:
if (k_4 j_5(D))
goto bb 4;
else
goto bb
--- Comment #36 from rguenth at gcc dot gnu dot org 2008-10-04 16:16
---
Testcase for that:
/* { dg-do compile } */
/* { dg-options -O2 -fdump-tree-pre } */
long
NumSift (long *array, int b, unsigned long k)
{
if (b)
if (array[k] array[k + 1L])
++k;
return array[k];
}
--- Comment #5 from danglin at gcc dot gnu dot org 2008-10-04 18:22 ---
Fixed
--
danglin at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #9 from hjl dot tools at gmail dot com 2008-10-04 18:25 ---
When we multiple a DImode constant in CONST_DOUBLE, if
CONST_DOUBLE_HIGH (op1) == 0, the constant must be
positive. In this case, we can't just check coeff != 0.
Does this patch make senses?
--- ./expmed.c.ll
--- Comment #4 from danglin at gcc dot gnu dot org 2008-10-04 18:08 ---
Subject: Bug 37603
Author: danglin
Date: Sat Oct 4 18:06:46 2008
New Revision: 140871
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=140871
Log:
PR target/37603
* pa.c
--- Comment #3 from danglin at gcc dot gnu dot org 2008-10-04 18:04 ---
Subject: Bug 37603
Author: danglin
Date: Sat Oct 4 18:03:07 2008
New Revision: 140870
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=140870
Log:
PR target/37603
* pa.c
--- Comment #5 from pault at gcc dot gnu dot org 2008-10-04 20:26 ---
Created an attachment (id=16462)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16462action=view)
The fix
I just tried to commit this as obvious but authentication failed for some
reason. One way or another, it
--- Comment #10 from hjl dot tools at gmail dot com 2008-10-04 20:43
---
A patch is posted at
http://gcc.gnu.org/ml/gcc-patches/2008-10/msg00128.html
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
ffmpeg_g.exe |grep ff_cos_16
00e62d14 B _ff_cos_16
00e25654 B _ff_cos_16384
$ nm libavcodec/fft.o |grep ff_cos_16
0020 C _ff_cos_16
8000 C _ff_cos_16384
gcc version 4.3.3 20081004 (prerelease) (Sherpya) (plain svn with the patch)
GNU ld (GNU Binutils) 2.18.91.20080917
I can also provide all
--- Comment #12 from ubizjak at gmail dot com 2008-10-04 21:47 ---
(In reply to comment #10)
Must be darwin specific then, can't reproduce on x86_64-linux and from quick
skim of gcc-testresults nobody else that supplied test summary recently
managed
to reproduced it.
$ gfortran-4.exe -v -save-temps -c PrettyPix.f95
Using built-in specs.
Target: i686-pc-cygwin
Configured with: /gnu/gcc/release/gcc4-4.3.2-1/src/gcc-4.3.2/configure
--srcdir=/gnu/gcc/release/gcc4-4.3.2-1/src/gcc-4.3.2 --prefix=/usr
--exec-prefix=/usr --bindir=/usr/bin --sbindir=/usr/sbin
--- Comment #2 from fang at csl dot cornell dot edu 2008-10-05 01:16
---
dupe PR 36460 (fixed in 4.4)?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35497
--- Comment #3 from 3dw4rd at verizon dot net 2008-10-05 01:33 ---
Subject: Re: Weirdness with numeric_limits in special
functions
paolo dot carlini at oracle dot com wrote:
--- Comment #2 from paolo dot carlini at oracle dot com 2008-10-03 10:54
---
Any news Ed?
--- Comment #6 from pault at gcc dot gnu dot org 2008-10-05 05:51 ---
Subject: Bug 37706
Author: pault
Date: Sun Oct 5 05:50:00 2008
New Revision: 140879
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=140879
Log:
2008-10-04 Paul Thomas [EMAIL PROTECTED]
PR
24 matches
Mail list logo