--- Comment #2 from ubizjak at gmail dot com 2007-06-21 06:33 ---
Use fabs/fabsf for double/float operands. Especially with %f.
--
ubizjak at gmail dot com changed:
What|Removed |Added
---
--- Comment #7 from burnus at gcc dot gnu dot org 2007-06-21 06:16 ---
> Host triplet is: processor - platform - os
> So depending on your processor it would be something like,
> i686-pc-linux-gnu
> x86-64-unknown-linux
> You can get this from the command uname
> Try uname -p -i -o
E
--- Comment #1 from rob1weld at aol dot com 2007-06-21 06:01 ---
Created an attachment (id=13753)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=13753&action=view)
Demo for abs / printf bug
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32448
I tried this program on many versions of GCC and it seems to trigger errors.
The curious can edit the program and un-comment various combinations of lines
to chase the error. It compiles as-is and demonstrates the error.
# gcc-3.3 -v
Reading specs from /usr/lib/gcc-lib/i486-linux/3.3.5/specs
Conf
Recent quip: "I do not care that its legal; not in my back-yard" NYMBY.
If you want to complain about not disabling the internal part for dfp, please
submit a new bug but note it will be marked as an enchement because it is
actually hard to do (I already looked into doing it after you and someone
--- Comment #4 from patchapp at dberlin dot org 2007-06-21 05:05 ---
Subject: Bug number PR32298
A patch for this bug has been added to the patch tracker.
The mailing list url for the patch is
http://gcc.gnu.org/ml/gcc-patches/2007-06/msg01521.html
--
http://gcc.gnu.org/bugzilla/sh
--- Comment #21 from kargl at gcc dot gnu dot org 2007-06-21 04:54 ---
(In reply to comment #20)
> $ gfortran -O1 -o TMalign TMalign.f
> In file TMalign.f:2005
>
> VOLATILE D
> 1
> Error: Unclassifiable statement at (1
--- Comment #20 from sunjoong at gmail dot com 2007-06-21 04:35 ---
$ gfortran -O1 -o TMalign TMalign.f
In file TMalign.f:2005
VOLATILE D
1
Error: Unclassifiable statement at (1)
(In reply to comment #19)
> You can ad
--- Comment #6 from jvdelisle at gcc dot gnu dot org 2007-06-21 04:22
---
Host triplet is: processor - platform - os
So depending on your processor it would be something like,
i686-pc-linux-gnu
or
x86-64-unknown-linux
You can get this from the command uname
Try uname -p -i -o
--- Comment #3 from jv244 at cam dot ac dot uk 2007-06-21 04:16 ---
(In reply to comment #2)
> I see a smaller difference, but a difference nonetheless.
yes, looks like better code is now generated, current timings are down to a
200% difference
ifort: 1.988124
gfortran: 3.900243
--
--- Comment #19 from kargl at gcc dot gnu dot org 2007-06-21 04:08 ---
(In reply to comment #18)
>
> I had ONLY HOPEd VOLATILE statement in fortran 77 EXTENSION of gfortran.
> I thought that would be convenient
> on small modification of legacy fortran 77 program.
You've completed miss
--- Comment #4 from jv244 at cam dot ac dot uk 2007-06-21 04:08 ---
> If igmin is -huge(0), (1-ig) can wrap, so ig2 might eventually be equal to ig,
> no? (although I agree for wrapping arithmetic, it will never be the case)
in Fortran 1-ig can't overflow (unless the program is undefine
--- Comment #5 from John dot Harper at mcs dot vuw dot ac dot nz
2007-06-21 03:52 ---
Subject: Re: F0.n output format fails with large numbers
On Thu, 21 Jun 2007, jvdelisle at gcc dot gnu dot org wrote:
> jvdelisle at gcc dot gnu dot org changed:
>
> What|Removed
--- Comment #114 from jv244 at cam dot ac dot uk 2007-06-21 03:41 ---
(In reply to comment #113)
> Great. I hope we can get it working with MPI (should probably already work)
I suspect that will be no real problem, but I do not have an MPI/gfortran setup
to check.
> > this seems quite
--- Comment #18 from sunjoong at gmail dot com 2007-06-21 03:27 ---
I appreciate kargl's comments; they were helpful.
I had known there is VOLATILE attribute in new Fortran standard
but I had worked with "LEGACY" fortran77 program!
I'll write C code if I shuld write one; that is more co
--- Comment #4 from jvdelisle at gcc dot gnu dot org 2007-06-21 03:15
---
Minimal test case:
! printing big numbers in F0.1 format. HP = high precision
PRINT '(F0.1)',huge(1.0)
END
Disregard my previous comment about KIND=10, that nor relevent. I just need
someone to test the #if def
--- Comment #3 from jvdelisle at gcc dot gnu dot org 2007-06-21 03:12
---
Here is a patch. Could someone test for a system that supports KIND=16. I can
only test for KIND=10. The value 45 is the minimum size that will work with
the test case. I am only guessing at the 63.
Index: wr
--- Comment #19 from dave at hiauly1 dot hia dot nrc dot ca 2007-06-21
02:56 ---
Subject: Re: [4.3 Regression] Bootstrap failure in stage1 on hppa*-*-*
> Created an attachment (id=13738)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=13738&action=view)
> --> (http://gcc.gnu.org/
--- Comment #5 from jvdelisle at gcc dot gnu dot org 2007-06-21 02:27
---
Fixed on Trunk, Closing
--
jvdelisle at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #4 from jvdelisle at gcc dot gnu dot org 2007-06-21 02:26
---
Subject: Bug 32361
Author: jvdelisle
Date: Thu Jun 21 02:25:58 2007
New Revision: 125909
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=125909
Log:
2007-06-20 Jerry DeLisle <[EMAIL PROTECTED]>
--- Comment #17 from kargl at gcc dot gnu dot org 2007-06-21 02:24 ---
(In reply to comment #16)
> Thank all of you.
> I could understand what make it different.
>
> There is no 'volatile' statement in fortran77 syntax of gfortran.
> Of course, volatile is not fortran77 standard, I thin
--- Comment #3 from jvdelisle at gcc dot gnu dot org 2007-06-21 02:23
---
Subject: Bug 32361
Author: jvdelisle
Date: Thu Jun 21 02:23:45 2007
New Revision: 125908
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=125908
Log:
2007-06-20 Jerry DeLisle <[EMAIL PROTECTED]>
--- Comment #6 from jvdelisle at gcc dot gnu dot org 2007-06-21 02:11
---
Fixed on 4.3, closing
--
jvdelisle at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #2 from jvdelisle at gcc dot gnu dot org 2007-06-21 01:56
---
Program received signal SIGSEGV, Segmentation fault.
0x0035e2a76ed0 in memcpy () from /lib64/libc.so.6
(gdb) bt
#0 0x0035e2a76ed0 in memcpy () from /lib64/libc.so.6
#1 0x2ab47a8d in output_float
--- Comment #1 from jvdelisle at gcc dot gnu dot org 2007-06-21 01:54
---
Confirmed. I get segfault on x86-64. I will investigate.
--
jvdelisle at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #5 from jvdelisle at gcc dot gnu dot org 2007-06-21 01:48
---
Subject: Bug 25061
Author: jvdelisle
Date: Thu Jun 21 01:48:21 2007
New Revision: 125907
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=125907
Log:
2007-06-20 Jerry DeLisle <[EMAIL PROTECTED]>
Run-time segmentation fault when attempting to print huge(x) with F0.1 format
even when x was default real, let alone kind 8 or 10. Evidence:
[EMAIL PROTECTED] test system: ~/Jfh/Test 52 >cat testf0fmt.f90
! printing big numbers in F0.1 format. HP = high precision
INTEGER,PARAMETER:: &
DP=kin
--- Comment #3 from geoffk at gcc dot gnu dot org 2007-06-21 01:24 ---
I split out the problem with 'i' to PR 32445.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32444
In this short program, on i386-darwin (probably all i386),
struct pt
{
int x, y;
};
struct pt sum(struct pt *p)
{
struct pt tmp = { 0, 0 };
int i;
for (i = 0; i < 100; i++)
{
tmp.x += p[i].x;
tmp.y += p[i].y;
}
return tmp;
}
when compiled with -O1 -g, there is no lo
--- Comment #4 from jvdelisle at gcc dot gnu dot org 2007-06-21 01:18
---
Subject: Bug 25061
Author: jvdelisle
Date: Thu Jun 21 01:18:02 2007
New Revision: 125906
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=125906
Log:
2007-06-20 Jerry DeLisle <[EMAIL PROTECTED]>
--- Comment #2 from geoffk at gcc dot gnu dot org 2007-06-21 01:13 ---
The problem with 'i' is ivopts.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32444
--- Comment #1 from geoffk at gcc dot gnu dot org 2007-06-21 01:08 ---
Happens even with -O1. One of the causes is SRA, but there's more...
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32444
--- Comment #15 from jvdelisle at gcc dot gnu dot org 2007-06-21 01:06
---
I have not patched secnds-1.f yet
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32057
--- Comment #14 from malitzke at metronets dot com 2007-06-21 00:53 ---
This whole series of postings from myself had one aim:
to _shame_ Messrs David Edelsohn an Geoff Keating to step up to their
resposibilities. You, Pinski, are not listed as a maintainer for RS600; they
are.
Mr Ke
Given the following short program,
struct pt
{
int x, y;
};
struct pt sum(struct pt *p)
{
struct pt tmp = { 0, 0 };
int i;
for (i = 0; i < 100; i++)
{
tmp.x += p[i].x;
tmp.y += p[i].y;
}
return tmp;
}
on powerpc-darwin, built with -O2, I see that the main loop is st
--- Comment #2 from jvdelisle at gcc dot gnu dot org 2007-06-21 00:32
---
I am getting the segfault on x86-64. I noticed it this morning but was not
sure is was not my own patches in local trunk. I just sent a note to Paul and
Steve to give heads up that I was seeing this failure, but
--- Comment #5 from spop at gcc dot gnu dot org 2007-06-20 23:45 ---
Subject: Bug 32075
Author: spop
Date: Wed Jun 20 23:44:56 2007
New Revision: 125901
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=125901
Log:
PR tree-optimization/32075
* tree-data-ref.c (subsc
--- Comment #4 from spop at gcc dot gnu dot org 2007-06-20 23:42 ---
Subject: Bug 32075
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 (subsc
--- Comment #16 from sunjoong at gmail dot com 2007-06-20 23:34 ---
Thank all of you.
I could understand what make it different.
There is no 'volatile' statement in fortran77 syntax of gfortran.
Of course, volatile is not fortran77 standard, I think,
but a certian implimentation support
--- Comment #7 from dannysmith at users dot sourceforge dot net 2007-06-20
23:32 ---
(In reply to comment #6)
> Anybody have enough __asm__ foo to write a inline assembly macro taking a long
> double operand and returning one, which I can use to call fsqrt directly in
> inline assembly?
--- Comment #18 from spop at gcc dot gnu dot org 2007-06-20 23:31 ---
Subject: Re: ICE: fold check: original tree changed by fold with
--enable-checking=fold
Hi, here is the final fix for the remaining cases, closing this PR.
Okay for trunk after bootntest on i686-linux with default la
--- Comment #6 from whaley at cs dot utsa dot edu 2007-06-20 23:17 ---
Anybody have enough __asm__ foo to write a inline assembly macro taking a long
double operand and returning one, which I can use to call fsqrt directly in
inline assembly? I'm scoping the docs, but have never found i
--- Comment #17 from spop at gcc dot gnu dot org 2007-06-20 22:58 ---
Subject: Re: ICE: fold check: original tree changed by fold with
--enable-checking=fold
Hi,
The following patch fixes a part of the reported testsuite fails.
With this patch we avoid folding the TREE_CHAIN of an SSA
--- Comment #16 from rob1weld at aol dot com 2007-06-20 22:46 ---
Uros Bizjak seems to have tracked this problems down for 4.3.0 on target
i686-pc-cygwin and i686-pc-linux-gnu in the bug report that Richard just marked
as a dupe.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20623
$ /home/tiberius/local/gcc/bin/gcc -v
Using built-in specs.
Target: i686-pc-linux-gnu
Configured with: ../gcc-4.2.0/configure --prefix=/home/tiberius/local/gcc-4.2.0
--libexecdir=/home/tiberius/local/gcc-4.2.0/lib --enable-shared
--enable-threads=posix --enable-__cxa_atexit --enable-clocale=gnu
--e
--- Comment #2 from danp57 at optonline dot net 2007-06-20 22:41 ---
This has been fixed... gcc-4.2.0 has a build that works on my macbook (x86)
with no problems.
Dan
--
danp57 at optonline dot net changed:
What|Removed |Added
--- Comment #14 from rob1weld at aol dot com 2007-06-20 22:33 ---
GCC version 4.3.0 20070620
http://gcc.gnu.org/ml/gcc-testresults/2007-06/msg00942.html
FAIL: gfortran.dg/secnds-1.f -O1 execution test
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32057
--- Comment #4 from rob1weld at aol dot com 2007-06-20 22:26 ---
My complaint is that collect2 hitlers-in a couple of "-lm"s when I try to link
without it. Since I modifed __my__ copy of GCC source to not use "-lm" I wish
that collect2 did not add the extra library.
I should have to ad
--- Comment #5 from whaley at cs dot utsa dot edu 2007-06-20 22:17 ---
It may be C99, but since it doesn't work on 90% of the machines in the world,
it is a bit of a stretch to call it portable. My point is no standard mandates
you round down a long double (where you don't round down a
--- Comment #2 from pinskia at gcc dot gnu dot org 2007-06-20 22:05 ---
This is more likely related to pointer_plus than dataflow.
Looking into it.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #4 from pinskia at gcc dot gnu dot org 2007-06-20 22:04 ---
sqrtl is C99 so it is portable, just not to cygwin. That is not our fault that
cygwin does not provide a C99 math library.
--
pinskia at gcc dot gnu dot org changed:
What|Removed
--- Comment #3 from fxcoudert at gcc dot gnu dot org 2007-06-20 21:53
---
(In reply to comment #0)
> DO ig=igmin,0
>ig2=1-ig
>if (ig.EQ.ig2) CALL link_error()
> ENDDO
> END
If igmin is -huge(0), (1-ig) can wrap, so ig2 might eventually be equal to ig,
no? (although I agree for
--- Comment #3 from whaley at cs dot utsa dot edu 2007-06-20 21:52 ---
Turns out the proposed solution of using sqrtl is not portable. In particular,
all code using it fails to link on Windows using cygwin. Any idea how to make
this work portably? I still don't understand why it is OK
--- Comment #8 from dfranke at gcc dot gnu dot org 2007-06-20 21:37 ---
Fixed in 4.2 and trunk. Closing.
--
dfranke at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #1 from rask at sygehus dot dk 2007-06-20 21:36 ---
Created an attachment (id=13750)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=13750&action=view)
Preprocessed testcase
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32441
--- Comment #7 from dfranke at gcc dot gnu dot org 2007-06-20 21:35 ---
Subject: Bug 32002
Author: dfranke
Date: Wed Jun 20 21:35:04 2007
New Revision: 125898
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=125898
Log:
gcc/fortran:
2007-07-20 Daniel Franke <[EMAIL PROTECTED]>
Here's how I configure GCC:
$ /n/08/rask/src/gcc/configure --target m32c-unknown-elf
--with-newlib--enable-sim --disable-gdb --disable-nls
--enable-languages=c,c++,java
This is how to reproduce the failure:
$ ./xgcc -B./ ~/__dprintf.c -S -dp -o /dev/null -mcpu=m32cm
/n/08/rask/src/gcc/newlib/libc
--- Comment #2 from ubizjak at gmail dot com 2007-06-20 21:33 ---
Patch at http://gcc.gnu.org/ml/gcc-patches/2007-06/msg01506.html.
--
ubizjak at gmail dot com changed:
What|Removed |Added
---
--- Comment #1 from ubizjak at gmail dot com 2007-06-20 21:30 ---
Confirmed, I have a patch to fix this issue.
--
ubizjak at gmail dot com changed:
What|Removed |Added
--- Comment #28 from pinskia at gcc dot gnu dot org 2007-06-20 21:17
---
*** Bug 32440 has been marked as a duplicate of this bug. ***
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #1 from pinskia at gcc dot gnu dot org 2007-06-20 21:17 ---
This is the way C++ works, yes before GCC 4.1.0, GCC did not work this way but
GCC was incorrect.
*** This bug has been marked as a duplicate of 2922 ***
--
pinskia at gcc dot gnu dot org changed:
Wh
The following:
$ cat t.cpp
template< typename T >
int
baz (T const*)
{
return 0;
}
template
void
bar(T* ptr)
{
baz (*ptr);
}
int
baz (int const&)
{
return 0;
}
void
foo ()
{
int const n = 0;
bar (&n);
}
fails to compile with gcc 4.2.0:
$ g++ -v; g++ -c t.cpp
Using bu
--- Comment #2 from fxcoudert at gcc dot gnu dot org 2007-06-20 20:59
---
I see a smaller difference, but a difference nonetheless.
--
fxcoudert at gcc dot gnu dot org changed:
What|Removed |Added
--
--- Comment #4 from pinskia at gcc dot gnu dot org 2007-06-20 20:53 ---
Related to PR 20297.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
BugsThisDepen
--- Comment #113 from fxcoudert at gcc dot gnu dot org 2007-06-20 20:41
---
(In reply to comment #112)
> after the fix for PR 32140 gfortran compiles CP2K correctly on x86_64 using
> '-O3 -ffast-math -ftree-vectorize -funroll-loops -march=native' . Thanks !
Great. I hope we can get it
--- Comment #10 from dfranke at gcc dot gnu dot org 2007-06-20 20:38
---
*** Bug 32438 has been marked as a duplicate of this bug. ***
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32436
--- Comment #1 from dfranke at gcc dot gnu dot org 2007-06-20 20:38 ---
*** This bug has been marked as a duplicate of 32436 ***
--
dfranke at gcc dot gnu dot org changed:
What|Removed |Added
--
--- Comment #9 from fxcoudert at gcc dot gnu dot org 2007-06-20 20:34
---
(In reply to comment #6)
> 2>&1
If you want to investigate this further, you need to run it directly from
command line, not within make, so that output is not redirected.
--
http://gcc.gnu.org/bugzilla/show_
--- Comment #8 from pinskia at gcc dot gnu dot org 2007-06-20 20:31 ---
> Huh -- that's very sad... I was hoping to use this on my new laptop :-(...
You can use 4.2.0 which is better (and really) supported for x86 Darwin.
I fixed a couple of bugs relating to x86 darwin support only fo
As mentioned in the CP2K PR 29975 comment 112
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29975#c112
on the source made available as
http://www.pci.unizh.ch/vandevondele/tmp/CP2K_gcc_2007_06.tgz
gfortran -O1 -fbounds-check all.f90
fails with
f951: out of memory allocating 2408 bytes after a
--- Comment #1 from burnus at gcc dot gnu dot org 2007-06-20 20:30 ---
It already fails with the smallest possible example:
real :: wlbaer(2)=0.
data wlbaer/0.0,0.0/
end
And it also fails with: gfortran.dg/data_initialized_2.f90
That test used to work. I'm pretty sur
--- Comment #7 from danp57 at optonline dot net 2007-06-20 20:29 ---
Subject: Re: Compile fails on compilation of
compile_options.o
Huh -- that's very sad... I was hoping to use this on my new laptop :-(...
Perhaps this can become a testcase to fix this up?
Dan
dfranke at gcc dot
--- Comment #6 from danp57 at optonline dot net 2007-06-20 20:28 ---
Subject: Re: Compile fails on compilation of
compile_options.o
/bin/sh ./libtool --mode=compile
/Users/platt/platt/install/GccSources/gcc-4.1.2.build/./gcc/xgcc
-B/Users/platt/platt/install/GccSources/gcc-4.1.2.bui
--- Comment #112 from jv244 at cam dot ac dot uk 2007-06-20 20:25 ---
after the fix for PR 32140 gfortran compiles CP2K correctly on x86_64 using
'-O3 -ffast-math -ftree-vectorize -funroll-loops -march=native' . Thanks !
I've made a new tar file available that contains a more recent ver
--- Comment #8 from fxcoudert at gcc dot gnu dot org 2007-06-20 20:21
---
I can reproduce this 4.3 regression with the following C testcase:
$ cat w.c
void foo (void)
{
int i, m;
float xa[21];
m = 0;
while (1)
{
i = 0;
while (1)
{
if (xa[(long int)i] == xa[(
--- Comment #8 from zadeck at naturalbridge dot com 2007-06-20 20:12
---
Subject: Re: ICE in global_alloc, at global.c:514
rask at sygehus dot dk wrote:
> --- Comment #7 from rask at sygehus dot dk 2007-06-20 20:08 ---
> This is something I have no clue about.
>
>
>
try it
--- Comment #3 from pinskia at gcc dot gnu dot org 2007-06-20 20:11 ---
This bug report needs to goto to the translation team as they are not
officially part of the GCC team. But since I always forget where the bug
reports go, I am just going to keep this bug open for now.
--
http:
--- Comment #13 from dberlin at gcc dot gnu dot org 2007-06-20 20:09
---
(In reply to comment #10)
> trunk has the same problem, but different constraints:
>
> Constraints:
>
> ANYTHING = &ANYTHING
> READONLY = &ANYTHING
> INTEGER = &ANYTHING
> barptr = &bar1
> barptr.0_1 = barptr
>
--- Comment #7 from rask at sygehus dot dk 2007-06-20 20:08 ---
This is something I have no clue about.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32418
Configuration invoked with:
../gcc-4.1.2/configure --enable-languages=c,c++,fortran
make
Failing lines show:
../gcc -I../../../gcc-4.1.2/libgfortran/../gcc/config -I../.././gcc
-D_GNU_SOURCE -std=gnu99 -Wall -Wstrict-prototypes -Wmissing-prototypes
-Wold-style-definition -Wextra -Wwrite-strings
Revision: 125825 with the patch 32406-v1.patch from PR 32406 applied.
Configured: ../trunk/configure --target=mipsel-linux
--with-sysroot=/usr/local/mipsel-linux-test
--prefix=/usr/local/mipsel-linux-test --with-arch=mips32 --with-float=soft
--disable-java-awt --without-x --disable-tls --enable-__
--- Comment #3 from pinskia at gcc dot gnu dot org 2007-06-20 20:04 ---
Did you miss "constant arguments", see constant. Nothing is done with non
constant arguments. So we are not replacing libm with mpfr.
--
pinskia at gcc dot gnu dot org changed:
What|Removed
--- Comment #12 from dberlin at gcc dot gnu dot org 2007-06-20 20:03
---
Subject: Re: [4.2 Regression] -fstrict-aliasing causes skipped code
On 20 Jun 2007 15:12:53 -, rguenth at gcc dot gnu dot org
<[EMAIL PROTECTED]> wrote:
>
>
> --- Comment #9 from rguenth at gcc dot gnu do
--- Comment #10 from burnus at gcc dot gnu dot org 2007-06-20 20:01 ---
> have tried your suggestion (commenting out the data
> statement), but the code does not appear to work correctly with this
> modification.
Try the option -fno-automatic.
--
http://gcc.gnu.org/bugzilla/sho
--- Comment #3 from burnus at gcc dot gnu dot org 2007-06-20 20:00 ---
Close as INVALID.
Paul had the tip to try "-fno-automatic"; with this option ifort and gfortran
produce identical output.
-fno-automatic
Treat each program unit as if the "SAVE" statement was specified for every
lo
--- Comment #14 from rask at gcc dot gnu dot org 2007-06-20 19:59 ---
Subject: Bug 32335
Author: rask
Date: Wed Jun 20 19:58:57 2007
New Revision: 125893
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=125893
Log:
2007-06-20 Rask Ingemann Lambertsen <[EMAIL PROTECTED]>
--- Comment #13 from rask at gcc dot gnu dot org 2007-06-20 19:57 ---
Subject: Bug 32335
Author: rask
Date: Wed Jun 20 19:57:32 2007
New Revision: 125892
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=125892
Log:
2007-06-20 Rask Ingemann Lambertsen <[EMAIL PROTECTED]>
--- Comment #5 from dfranke at gcc dot gnu dot org 2007-06-20 19:44 ---
> does the FSF 4.1 branch officially support i386-darwin?
The testresults-ml does no show to many entries for 4.1.x on i386-darwin.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32436
--- Comment #4 from fxcoudert at gcc dot gnu dot org 2007-06-20 19:40
---
Doh, I have another question: does the FSF 4.1 branch officially support
i386-darwin? I'm not sure it does...
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32436
--- Comment #3 from fxcoudert at gcc dot gnu dot org 2007-06-20 19:37
---
> ../gcc -I../../../gcc-4.1.2/libgfortran/../gcc/config -I../.././gcc
> -D_GNU_SOURCE -std=gnu99 -Wall -Wstrict-prototypes -Wmissing-prototypes
> -Wold-style-definition -Wextra -Wwrite-strings -O2 -g -O2 -c
> ../.
--- Comment #29 from fche at redhat dot com 2007-06-20 19:37 ---
> This is the patch mentioned in my explanation. It is against the 4.1.1
> release
> source.
Thanks!
This patch applies fine to CVS head, but it does not appear to help
significantly
with the C++ test cases like the ones
--
burnus at gcc dot gnu dot org changed:
What|Removed |Added
Severity|major |normal
Keywords|ice-on-invalid-code |error-recovery
--- Comment #12 from patchapp at dberlin dot org 2007-06-20 19:35 ---
Subject: Bug number PR target/32335
A patch for this bug has been added to the patch tracker.
The mailing list url for the patch is
http://gcc.gnu.org/ml/gcc-patches/2007-06/msg01501.html
--
http://gcc.gnu.org/bu
--- Comment #2 from danp57 at optonline dot net 2007-06-20 19:34 ---
Subject: Re: Compile fails on compilation of
compile_options.o
I started in a freshly created empty directory!
Dan
On Jun 20, 2007, at 3:30 PM, dfranke at gcc dot gnu dot org wrote:
>
>
> --- Comment #1 from d
--
dfranke at gcc dot gnu dot org changed:
What|Removed |Added
Severity|blocker |normal
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32436
--- Comment #1 from dfranke at gcc dot gnu dot org 2007-06-20 19:30 ---
Since GCC gets compiled over and over again, it is highly unlikely that a
release branch simply fails like that. There is no actual problem shown?!
You started the compilation in a clean directory?
--
dfranke a
Configuration invoked with:
../gcc-4.1.2/configure --enable-languages=c,c++,fortran
make
Failing lines show:
../gcc -I../../../gcc-4.1.2/libgfortran/../gcc/config -I../.././gcc
-D_GNU_SOURCE -std=gnu99 -Wall -Wstrict-prototypes -Wmissing-prototypes
-Wold-style-definition -Wextra -Wwrite-strings
--- Comment #15 from spop at gcc dot gnu dot org 2007-06-20 18:50 ---
(In reply to comment #5)
> This depends on PR 22455 now as --enable-checking=fold is broken.
>
With
--
spop at gcc dot gnu dot org changed:
What|Removed |Added
---
Two tests in SPEC CPU2000 ICE in build build2_stat when compiled for
powerpc-linux with "-O2 -mtree-vectorize -maltivec". Here's a minimized
testcase:
extern char *reg_names[];
extern int target_flags;
void foo ()
{
int i;
if (target_flags)
for (i = 0; i < 64; i++)
reg_names[i]--
--- Comment #3 from scovich at gmail dot com 2007-06-20 18:22 ---
(In reply to comment #1)
Sorry for the double post, but I just tried creating a wrapper_foo() that
copies the values out of the struct, then passes them on to foo() as scalars.
The problem only appears if foo() gets inlin
1 - 100 of 218 matches
Mail list logo
