--- Comment #2 from ubizjak at gmail dot com 2009-01-23 08:47 ---
Works for me with:
--cut here--
#define _GNU_SOURCE
#include fenv.h
#include signal.h
#include stdlib.h
static void foo (int sig)
{
exit (0);
}
float __attribute__((noinline)) test (float x) { return 2.0f / x; };
Following test:
--cut here--
#define _GNU_SOURCE
#include fenv.h
#include signal.h
#include stdlib.h
static void foo (int sig)
{
exit (0);
}
float __attribute__((noinline)) test (float x) { return 2.0f / x; };
int main()
{
signal (SIGFPE, foo);
feclearexcept (FE_ALL_EXCEPT);
--- Comment #1 from rguenth at gcc dot gnu dot org 2009-01-23 09:21 ---
FENV access is not implemented.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #1 from rguenth at gcc dot gnu dot org 2009-01-23 10:33 ---
Seems to work for me.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38942
--- Comment #46 from r dot emrich at de dot tecosim dot com 2009-01-23
10:45 ---
Created an attachment (id=17164)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17164action=view)
Verbose output of static link step
As you can see for all libraries except libdld archives are used.
--- Comment #47 from r dot emrich at de dot tecosim dot com 2009-01-23
10:50 ---
Created an attachment (id=17165)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17165action=view)
chatr output for the resulting object
As you can see there is one dependency to libdl.1 what is ok
comiling common.c of conky-1.6.1
CFLAGS=-O2 -march=native -mtune=native -pipe -fomit-frame-pointer -ffast-math
i486-slackware-linux-gcc -DHAVE_CONFIG_H -I.
-DSYSTEM_CONFIG_FILE=\/etc/conky/c
onky.conf\-O2 -march=native -mtune=native -pipe -fomit-frame-pointer
-ffast
-math -pthread
--- Comment #3 from rob1weld at aol dot com 2009-01-23 12:25 ---
From: Ben Elliston b...@air.net.au
To: rob1w...@aol.com
Sent: Thu, 22 Jan 2009 8:12 pm
Subject: Re: Patch for config.guess
Hi Rob,
Thanks for your report.
The script config.guess lumps together many of Sun's
The 4.x compilers does not warn when using unset variables.
The 3.x compilers did warn on this:
Example:
int use_unset_variable(int y)
{
int x;
switch(y) {
case 0:
x = 0;
break;
default:
break;
}
x++;
return x;
}
Variable x could be unset when returning.
I use all
I built gcc and submitted a Test Report here:
http://gcc.gnu.org/ml/gcc-testresults/2009-01/msg02341.html
http://gcc.gnu.org/ml/gcc-testresults/2009-01/msg02197.html
...
=== gfortran tests ===
FAIL: gfortran.dg/array_constructor_23.f -O3 -fomit-frame-pointer execution
test
FAIL:
--- Comment #4 from rguenth at gcc dot gnu dot org 2009-01-23 15:33 ---
The testcase should be
subroutine to_product_of(self,a,b,a1,a2)
complex(kind=8) :: self (:)
complex(kind=8), intent(in) :: a(:,:)
complex(kind=8), intent(in) :: b(:)
integer a1,a2
do i = 1,a1
do j =
--- Comment #5 from rguenth at gcc dot gnu dot org 2009-01-23 15:36 ---
So,
4) The vectorized version sucks because we have to use peeling for niters
because we need to unroll the loop once and cannot apply SLP here.
Q1: does SLP work with reductions at all?
Q2: does SLP do
--- Comment #12 from bonzini at gnu dot org 2009-01-23 15:57 ---
Subject: Bug 38932
Author: bonzini
Date: Fri Jan 23 15:57:19 2009
New Revision: 143588
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143588
Log:
2008-01-23 Paolo Bonzini bonz...@gnu.org
PR
--- Comment #7 from vmakarov at redhat dot com 2009-01-23 15:58 ---
If it is a problem, it is not a problem of IRA. The usage of the old RA
results in the same abort. Neither the old RA, nor IRA assigns CX to pseudos.
CX is started to be used by reload, so I think it is reload
--- Comment #13 from rguenth at gcc dot gnu dot org 2009-01-23 16:18
---
Fixed on the trunk.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Known
--- Comment #10 from rguenth at gcc dot gnu dot org 2009-01-23 16:18
---
Fixed.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Status|NEW
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Keywords||missed-optimization
Summary|[4.4 regression]
--- Comment #5 from rguenth at gcc dot gnu dot org 2009-01-23 16:22 ---
GCC 3.4.4 is no longer maintained.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.4.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38740
Legacy f77 code often uses equivalence statements between array elements and
common members extensively, e.g.:
COMMON/GCPARM/IPARAM,PCUTGA,PCUTEL,PCUTNE,PCUTHA,PCUTMU
+ ,NSPARA,MPSTAK,NPGENE
REAL PACUTS(5)
EQUIVALENCE (PACUTS(1),PCUTGA)
When parallelizing such
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.3.3
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38471
Noticed with gcc-4_3-branch revision 143533 and -march=v32 -O2 -fno-tree-sra.
The attached test-case (from ghostscript-5.50) goes:
/h/hp/gcc/cris/bugs/20090122/gs42.i: In function 'append_outline':
/h/hp/gcc/cris/bugs/20090122/gs42.i:1387: error: insn does not satisfy its
constraints:
(insn 1860
--
hp at gcc dot gnu dot org changed:
What|Removed |Added
Status|UNCONFIRMED |NEW
Ever Confirmed|0 |1
Last
--
hp at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |hp at gcc dot gnu dot org
|dot org
--- Comment #1 from hp at gcc dot gnu dot org 2009-01-23 16:33 ---
Created an attachment (id=17166)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17166action=view)
test-case
Compile with -O2 -march=v32 -fno-tree-sra for cris-axis-elf (or
crisv32-axis-elf).
--
--
burnus at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|4.3.3 |4.5.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38471
--- Comment #12 from rguenth at gcc dot gnu dot org 2009-01-23 16:52
---
this is placement-new related.
D.20941_8 = ox
ox = { } no-tbaa-pruning
D.20941_8 = { ox }
alias warning for D.20941_8
After alias:
ox.D.20531.m_initialized = 0;
D.20941_8 = (struct X *)
--- Comment #1 from ubizjak at gmail dot com 2009-01-23 17:02 ---
1) The source as inlined in bugreport is useless, please see [1] for a detailed
bug report instructions and _attach_ generated preprocessed source (unless it
is a couple of tens of lines long...).
2) -march=native
--- Comment #2 from hp at gcc dot gnu dot org 2009-01-23 17:08 ---
Created an attachment (id=17167)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17167action=view)
Reduced test-case, suggested for gcc.dg/torture but untested in framework.
Also -fstrict-aliasing (the default) is
--- Comment #48 from dave at hiauly1 dot hia dot nrc dot ca 2009-01-23
17:41 ---
Subject: Re: shared link/execute fails for cross gcc from linux to target
hppa64-hp-hpux11.00
As you can see there is one dependency to libdl.1 what is ok AFAIK. There is
no
libdl archive.
Yes,
I'm seeing all these test failures in regression runs on HEAD:
FAIL: g++.dg/torture/stackalign/unwind-0.C (test for excess errors)
WARNING: g++.dg/torture/stackalign/unwind-0.C compilation failed to produce
executable
FAIL: g++.dg/torture/stackalign/unwind-0.C (test for excess errors)
WARNING:
--- Comment #1 from dave dot korn dot cygwin at gmail dot com 2009-01-23
18:10 ---
Created an attachment (id=17168)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17168action=view)
Force assembler labels to match.
Now testing this fairly straightforward approach to making the
--- Comment #2 from dannysmith at users dot sourceforge dot net 2009-01-23
18:53 ---
There is an alternative patch at
http://gcc.gnu.org/ml/gcc-patches/2008-10/msg00044.html
which i had forgotten about. It has been tested on i686-pc-mingw32 and
i686-pc-linux
--
--- Comment #3 from dave dot korn dot cygwin at gmail dot com 2009-01-23
19:02 ---
Right you are. Either one should work, but I don't have any more spare time
than you for testing things on Linux right now. It's non-critical, so I'll
keep a patch in my local tree and maybe we should
--- Comment #3 from bangerth at dealii dot org 2009-01-23 19:26 ---
I see this as well. It triggers a ton of time in boost::graph code.
I think it should have higher priority than P3.
W.
--
bangerth at dealii dot org changed:
What|Removed |Added
--- Comment #7 from bangerth at dealii dot org 2009-01-23 19:31 ---
I see this as well. It triggers a lot when using boost::graph which
uses empty classes as tags all over the place. A simple case with
boost::graph would be this:
--
#include
--- Comment #1 from mikael at gcc dot gnu dot org 2009-01-23 20:04 ---
(In reply to comment #0)
=== gfortran tests ===
FAIL: gfortran.dg/array_constructor_23.f -O3 -fomit-frame-pointer execution
test
FAIL: gfortran.dg/array_constructor_23.f -O3 -fomit-frame-pointer
-funroll-loops
--- Comment #2 from nemet at gcc dot gnu dot org 2009-01-23 21:19 ---
The fixed-point tests fail to compile when configure with
--disable-fixed-point.
I think the easiest would be to extend Richard's new infrastructure with a new
keyword and then skip these tests completely. I take a
The following snippet makes GCC 4.3.2 (and SVN versions as of 09/2008 still)
cause an ICE:
template typename T, T N void f(T()[N]);
int main() {
int x[2];
unsigned int y[2];
f(x); // works
f(y); // ICE
}
Getting this output:
test.cpp: In function 'int main()':
test.cpp:7:
--- Comment #1 from rob1weld at aol dot com 2009-01-23 22:31 ---
The adding of these assertions to replace an error message might have been
first discussed here: http://gcc.gnu.org/ml/java-patches/2007-q1/msg00478.html
Rob
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38892
Hi,
this is just to mention that support for MINVAL/MAXVAL/MINLOC/MAXLOC
for character arrays is required in Fortran 2003.
Example:
print *, minval ( [ A ] )
print *, minloc ( [ A ] )
should both work.
--
Summary: Fortran 2003: MINVAL/MAXVAL/MINLOC/MAXLOC for character
SjLj EH (the default) is completely broken on Cygwin (and MinGW, confirmed by
Danny in private email).
The failures have been occurring since at least November: see, e.g.
http://gcc.gnu.org/ml/gcc-testresults/2008-11/msg02144.html
(The distro maintainers have all been so focussed recently on
--- Comment #1 from dave dot korn dot cygwin at gmail dot com 2009-01-23
23:31 ---
Created an attachment (id=17169)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17169action=view)
Simple throw-catch testcase
Test cases don't come much simpler than this.
--
--- Comment #2 from dave dot korn dot cygwin at gmail dot com 2009-01-23
23:31 ---
Created an attachment (id=17170)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17170action=view)
Pre-processed source of testcase.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38952
--- Comment #3 from dave dot korn dot cygwin at gmail dot com 2009-01-23
23:32 ---
Created an attachment (id=17171)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17171action=view)
Generated assembler for testcase
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38952
--- Comment #4 from dave dot korn dot cygwin at gmail dot com 2009-01-23
23:44 ---
The bug manifests itself as a crash on exit from main(); $eip is set to zero
and we get a SEGV.
On entry to main(), the registers show:
esp0x22cc40 0x22cc40
ebp0x22cca8
--- Comment #8 from hjl dot tools at gmail dot com 2009-01-24 00:08 ---
Reload can't deal with:
[...@gnu-6 reg-1]$ cat f.i
int
bar (int g)
{
register int x __asm__(ecx);
x = 8;
if ((1 g) != x)
return 0;
return x;
}
[...@gnu-6 reg-1]$
--- Comment #5 from dave dot korn dot cygwin at gmail dot com 2009-01-24
00:10 ---
Here is a disassembly of the start of the main function:
(gdb) disass main
Dump of assembler code for function main:
0x00401070 main+0:push %ebp
0x00401071 main+1:mov%esp,%ebp
0x00401073
This problem happened in the graphite branch where after the graphite
pass we scheduled a pass of loop invariant motion (LIM). LIM does
expect to see the loop closed SSA form, but that property is not
maintained by the code generator of graphite. The testcase is
extracted from the polyhedron
--- Comment #1 from spop at gcc dot gnu dot org 2009-01-24 00:47 ---
Created an attachment (id=17172)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17172action=view)
patch
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38953
--- Comment #2 from spop at gcc dot gnu dot org 2009-01-24 01:00 ---
Subject: Bug 38953
Author: spop
Date: Sat Jan 24 00:59:51 2009
New Revision: 143599
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143599
Log:
2009-01-24 Sebastian Pop sebastian@amd.com
PR
--- Comment #3 from spop at gcc dot gnu dot org 2009-01-24 01:00 ---
Fixed.
--
spop at gcc dot gnu dot org changed:
What|Removed |Added
Status|UNCONFIRMED
--- Comment #6 from dave dot korn dot cygwin at gmail dot com 2009-01-24
01:08 ---
Here is the RTL that is created by the .130r.eh pass: everything between note 2
and call_insn 3, was added after expand.
try_optimize_cfg iteration 2
(note 1 0 4 NOTE_INSN_DELETED)
(note 4 1 2 2 [bb
--- Comment #9 from hjl dot tools at gmail dot com 2009-01-24 01:32 ---
From gcc doc:
---
Sometimes you need to make an `asm' operand be a specific register,
but there's no matching constraint letter for that register _by
itself_. To force the operand into that register, use a local
--- Comment #1 from manu at gcc dot gnu dot org 2009-01-24 01:40 ---
Quoting the FAQ at http://gcc.gnu.org/wiki/NewWconversion#faq
Why Wconversion is not enabled by -Wall or at least by -Wextra?
Implicit conversions are very common in C. This tied with the fact that there
is no
--- Comment #6 from bkoz at gcc dot gnu dot org 2009-01-24 01:52 ---
Mine.
--
bkoz at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at
I was instructed by avr-gcc 4.3.2 to send a bug report. Here's the command
line:
$ avr-gcc -v -save-temps -Os -g -Wall -I. -I../usbtiny -mmcu=attiny2313 -c -o
main.o main.c
Using built-in specs.
Target: avr
Configured with: ../configure --disable-libssp --disable-nls
--enable-languages=c,c++
With recent trunk (4.4.0 20090107), test gfortran.dg/array_constructor_24.f
fails on i686-pc-cygwin for the four -O3 execution tests. The failure is in
subroutine FA6013.
program try_fa6013
call fa6013 (10, 1, -1)
call fa6077 (10, 1, -1, (/1,2,3,4,5,6,7,8,9,10/))
call
--
billingd at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |billingd at gcc dot gnu dot
|dot org
--- Comment #7 from bkoz at gcc dot gnu dot org 2009-01-24 02:30 ---
Please note there are no ABI baseline files checked in to the FSF GCC
repository for darwin for this (or any) release on this (or any) architecture.
Without these, there can be no automated ABI testing by the GCC
--- Comment #2 from bkoz at gcc dot gnu dot org 2009-01-24 02:33 ---
Mine.
--
bkoz at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at
--- Comment #1 from billingd at gcc dot gnu dot org 2009-01-24 02:44
---
Patch http://gcc.gnu.org/ml/gcc-patches/2009-01/msg01181.html
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38955
--- Comment #3 from bkoz at gcc dot gnu dot org 2009-01-24 02:45 ---
Please note there are no ABI baseline files checked in to the FSF GCC
repository for darwin for this (or any) release on this (or any) architecture.
Without these, there can be no automated ABI testing by the GCC
--- Comment #3 from nemet at gcc dot gnu dot org 2009-01-24 02:47 ---
It was actually Richard's mips.exp rewrite that removed { target { fixed_point
} } from these tests.
Richard, was this intentional? It seems to me that since fixed-point is not
dependent on a command-line flag
--- Comment #2 from bkoz at gcc dot gnu dot org 2009-01-24 02:57 ---
Mine.
--
bkoz at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at
Test gfortran.dg/chmod_1.f90 fails on i686-pc-cygwin. The part of the test
that fails is:
call chmod (n, a-w, i
if (i == 0 .and. getuid() /= 0) then
if (access(n,w) == 0 .or. access(n,W) == 0) call abort
end if
n = foobar_file, and after the test we have
$ ls -l foobar_file
--- Comment #1 from billingd at gcc dot gnu dot org 2009-01-24 03:06
---
This is with cygwin-1.7.
--
billingd at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #3 from bkoz at gcc dot gnu dot org 2009-01-24 03:20 ---
Please note there are no ABI baseline files checked in to the FSF GCC
repository for darwin for this (or any) release on this (or any) architecture.
Without these, there can be no automated ABI testing by the GCC
--- Comment #21 from bkoz at gcc dot gnu dot org 2009-01-24 03:42 ---
This is fixed for hpux on trunk, and I believe it should be fixed for solaris
as well. Can some solaris tester confirm?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28125
--- Comment #8 from mrs at apple dot com 2009-01-24 04:18 ---
First, you didn't test 4.0.0, so all your reasoning is invalid. Second,
claiming no ABI breakage when there is abi breakage is silly. I guess if you
have no customers with no software, with do previously deployed software
--- Comment #4 from mrs at apple dot com 2009-01-24 04:22 ---
First, you didn't test 4.0.0, so all your reasoning is invalid. Second,
claiming no ABI breakage when there is abi breakage is silly. I guess if you
have no customers with no software, with do previously deployed software
--- Comment #8 from xuepeng dot guo at intel dot com 2009-01-24 05:12
---
Created an attachment (id=17173)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17173action=view)
An extracted test case for this bug.
Hi tim, I extracted this test case from your website. But I can't
--- Comment #3 from abramobagnara at tin dot it 2009-01-24 05:43 ---
The test case you have used it's different from the original that showed the
bug.
Nevertheless it's useful to understand the possible nature of the bug.
It seems that if feenableexcept(FE_INEXACT) is not called
--- Comment #5 from dirtyepic at gentoo dot org 2009-01-24 05:53 ---
Just tested again w/ r143592 with no issues.
--
dirtyepic at gentoo dot org changed:
What|Removed |Added
--- Comment #2 from lisp2d at lisp2d dot net 2009-01-24 06:00 ---
The niche is selected incorrectly. Basically the considerable quantity of
errors of programming is linked to implicit type conversion.
void insert(char*,char,unsigned int){..}
insert(..,..,size_t);
Example
double d =
--- Comment #7 from dave dot korn dot cygwin at gmail dot com 2009-01-24
06:23 ---
Not IRA-related, it seems, but reload-backend interaction. -fno-ira doesn't
make any difference to the generated code to calculate the frame pointer for
the jmp_buf. In a non-IRA build, pass 172r.lreg
--- Comment #2 from billingd at gcc dot gnu dot org 2009-01-24 07:41
---
This looks like a problem with the access function on cygwin-1.7
$ touch foobar_file
$ chmod a-w foobar_file
$ cat test_access.c
#include stdio.h
#include unistd.h
int main (void) {
char *file=foobar_file;
77 matches
Mail list logo
