--- Comment #5 from domob at gcc dot gnu dot org 2009-01-04 08:40 ---
*** Bug 35943 has been marked as a duplicate of this bug. ***
--
domob at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #1 from domob at gcc dot gnu dot org 2009-01-04 08:40 ---
*** This bug has been marked as a duplicate of 37746 ***
--
domob at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #6 from domob at gcc dot gnu dot org 2009-01-04 08:41 ---
Here's a patch: http://gcc.gnu.org/ml/fortran/2008-12/msg00273.html
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37746
--- Comment #14 from jakub at gcc dot gnu dot org 2009-01-04 09:05 ---
Yes, the http://gcc.gnu.org/ml/gcc-patches/2009-01/msg00108.html
patch fixes the bootstrap (at least --with-tune=pentium-m i686-linux bootstrap
I've tried, haven't tried --with-arch=pentium-m).
--
const _Self
_M_next() const
{
if (_M_node)
return _Fwd_list_iterator(this-_M_node-_M_next);
else
return _Fwd_list_iterator(0);
}
returns a reference to a _Fwd_list_iterator object that is constructed on the
_M_next() stack frame. This causes
--- Comment #1 from paolo at gcc dot gnu dot org 2009-01-04 13:55 ---
Subject: Bug 38719
Author: paolo
Date: Sun Jan 4 13:55:25 2009
New Revision: 143052
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143052
Log:
2009-01-04 Paolo Carlini paolo.carl...@oracle.com
PR
--- Comment #10 from jakub at gcc dot gnu dot org 2009-01-04 14:26 ---
The problem is that if size is smaller than epilogue_size_needed, desired_align
align and size_needed == 1, then ix86_expand_setmem jumps around the
initialization of promoted_val (when val_exp isn't CONST_INT) and
--- Comment #3 from rguenther at suse dot de 2009-01-04 14:35 ---
Subject: Re: _Relative_pointer_impl invokes undefined
behavior
On Sun, 4 Jan 2009, paolo dot carlini at oracle dot com wrote:
--- Comment #2 from paolo dot carlini at oracle dot com 2009-01-04 14:03
---
--- Comment #4 from dfranke at gcc dot gnu dot org 2009-01-04 14:56 ---
simplify.c (gfc_simplify_reshape) has:
3547 if (x[order[i]])
3548{
3549 gfc_error (Invalid permutation in ORDER parameter at %L,
3550 e-where);
3551 gfc_free_expr
--- Comment #5 from jv244 at cam dot ac dot uk 2009-01-04 14:58 ---
(In reply to comment #2)
If the module files cause problems, I can try to get to a source file only
reproducer
http://www.pci.uzh.ch/vandevondele/tmp/CP2K_2008_12_03.tgz can be used as a
source files only testcase,
-unknown-linux-gnu
Configured with: /data03/vondele/gcc_trunk/gcc/configure
--prefix=/data03/vondele/gcc_trunk/build --enable-languages=c,fortran
--disable-multilib
Thread model: posix
gcc version 4.4.0 20090104 (experimental) [trunk revision 143050] (GCC)
COLLECT_GCC_OPTIONS='-c' '-ffree-form
--- Comment #4 from steven at gcc dot gnu dot org 2009-01-04 14:45 ---
Can you fill in the known-to-work field please, if you think this is a
regression?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38722
--- Comment #2 from jv244 at cam dot ac dot uk 2009-01-04 14:19 ---
testcase:
http://www.pci.uzh.ch/vandevondele/tmp/PR38722.tgz
untar, and reproduce with
gfortran -v -c -O1 -cpp bug.f90
If the module files cause problems, I can try to get to a source file only
reproducer.
--
--- Comment #1 from jv244 at cam dot ac dot uk 2009-01-04 14:10 ---
For gcc version 4.4.0 20090104 (experimental) [trunk revision 143050] (GCC)
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38722
--- Comment #2 from dfranke at gcc dot gnu dot org 2009-01-04 16:13 ---
Commit in #1 added a simplifier for MERGE. Having a closer look at
IS_IOSTAT_{END, EOR}, I think they won't make much sense in init expr anyway.
That leaves these to do:
ERFC_SCALED(F2008)
DREAL
--- Comment #1 from dfranke at gcc dot gnu dot org 2009-01-04 16:05 ---
Subject: Bug 38718
Author: dfranke
Date: Sun Jan 4 16:05:42 2009
New Revision: 143053
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143053
Log:
gcc/fortran:
2009-01-04 Daniel Franke
--
jv244 at cam dot ac dot uk changed:
What|Removed |Added
Target Milestone|--- |4.4.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38722
The vectorizer miscompiles gfortran.fortran-torture/execute/elemental.f90 at
-O3
on x86-64-unknown-linux-gnu.
--
Summary: [alias-improvements] vectorizer miscompiles
gfortran.fortran-torture/execute/elemental.f90 at -O3
Product: gcc
Version:
--
paolo dot carlini at oracle dot com changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |paolo dot carlini at oracle
|dot org
--- Comment #8 from hjl dot tools at gmail dot com 2009-01-04 18:21 ---
Revision 143027 is the cause.
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
--- Comment #18 from hjl dot tools at gmail dot com 2009-01-04 18:23
---
Subject: Re: [ira] error in start_allocno_priorities, at ira-color.c:1806
On Fri, Jan 2, 2009 at 4:34 PM, Kenneth Zadeck zad...@naturalbridge.com
wrote:
Kenneth Zadeck wrote:
Steven Bosscher wrote:
On Fri,
--- Comment #2 from paolo dot carlini at oracle dot com 2009-01-04 13:57
---
Fixed.
--
paolo dot carlini at oracle dot com changed:
What|Removed |Added
--- Comment #1 from rguenth at gcc dot gnu dot org 2009-01-04 13:11 ---
The testcases are fixed on the branch.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #2 from rguenth at gcc dot gnu dot org 2009-01-04 13:11 ---
The testcases are fixed on the branch.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--
steven at gcc dot gnu dot org changed:
What|Removed |Added
Severity|normal |critical
Status|UNCONFIRMED |NEW
FRE does not optimize
int i;
int foo (void)
{
int j;
i = j;
return i;
}
because j_1(D) is never in AVAIL_OUT, so we do not find a leader for it.
We can fix this by either pretending used default definitions are defined
in the entry block or by treating default definitions the same as
--- Comment #9 from steven at gcc dot gnu dot org 2009-01-04 18:45 ---
Thanks HJ for looking into this. I think revision 143027 is not the cause,
it's more likely that it uncovers a latent bug. I'm trying to reduce Joost's
code to a small test case. So far, what I'm seeing is that
compiling CP2K at -O1 I get the following ICE:
gfortran -c -O1 -cpp bug.f90
bug.f90: In function qs_ks_build_kohn_sham_matrix:
bug.f90:1392: internal compiler error: Segmentation fault
with a trace
Program received signal SIGSEGV, Segmentation fault.
find_decomposable_subregs (px=value
--- Comment #9 from mikael at gcc dot gnu dot org 2009-01-04 13:01 ---
Subject: Bug 38536
Author: mikael
Date: Sun Jan 4 13:01:12 2009
New Revision: 143050
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143050
Log:
2009-01-04 Mikael Morin mikael.mo...@tele2.fr
PR
--- Comment #2 from paolo dot carlini at oracle dot com 2009-01-04 14:03
---
Richard, this is just an (experimental) extension facility, we can break the
binary interface, return to our decisions, etc, no problem. Thus, just commit
whatever fix you are able to quickly figure out (like
The elemental intrinsic functions listed below lack simplifiers to make them
usable in initialization expressions:
DREAL (GNU extension)
ERFC_SCALED(F2008)
IS_IOSTAT_END (F2003)
IS_IOSTAT_EOR (F2003)
ISNAN (GNU extension)
LSHIFT (GNU extension)
MERGE
--- Comment #5 from mikael at gcc dot gnu dot org 2009-01-04 19:00 ---
(In reply to comment #4)
which detects invalid permutations in the case of constant(!) arguments.
Closing as fixed.
No, it's not. Reopening.
The initial testcase is still not catch.
--
mikael at gcc dot gnu
--- Comment #6 from dfranke at gcc dot gnu dot org 2009-01-04 19:10 ---
(In reply to comment #5)
(In reply to comment #4)
which detects invalid permutations in the case of constant(!) arguments.
Closing as fixed.
No, it's not. Reopening.
The initial testcase is still not
--- Comment #27 from mikael at gcc dot gnu dot org 2009-01-04 19:12 ---
Subject: Bug 35681
Author: mikael
Date: Sun Jan 4 19:12:16 2009
New Revision: 143057
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143057
Log:
2009-01-04 Mikael Morin mikael.mo...@tele2.fr
PR
--- Comment #7 from mikael at gcc dot gnu dot org 2009-01-04 19:12 ---
Subject: Bug 38669
Author: mikael
Date: Sun Jan 4 19:12:16 2009
New Revision: 143057
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143057
Log:
2009-01-04 Mikael Morin mikael.mo...@tele2.fr
PR
--- Comment #7 from mikael at gcc dot gnu dot org 2009-01-04 19:12 ---
Subject: Bug 38487
Author: mikael
Date: Sun Jan 4 19:12:16 2009
New Revision: 143057
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143057
Log:
2009-01-04 Mikael Morin mikael.mo...@tele2.fr
PR
--- Comment #7 from pault at gcc dot gnu dot org 2009-01-04 17:59 ---
(In reply to comment #6)
In the mean time, I know how to apply a fix along the lines of 1.
Ok, then; go for (1). I have nothing to propose.
I have just realised that the matching can be eliminated, although I
--- Comment #6 from steven at gcc dot gnu dot org 2009-01-04 15:13 ---
Does not fail for me on i686-pc-cygwin with gcc version 4.4.0 20090103
(experimental) [trunk revision 143030]. What target are you compiling for?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38722
--- Comment #1 from rguenth at gcc dot gnu dot org 2009-01-04 13:22 ---
Created an attachment (id=17029)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17029action=view)
patch to show the idea
prototype patch (unfortunately there's no std::uintptr_t, so I used unsigned
long for
--- Comment #8 from dominiq at lps dot ens dot fr 2009-01-04 12:27 ---
Withe the patch in comment #5 the problems reported in comment #4 disappear,
the test in comment #6 gives a bus error at compile time:
pr38536_3.f90: In function 'MAIN__':
pr38536_3.f90:1: internal compiler error:
--- Comment #4 from paolo dot carlini at oracle dot com 2009-01-04 14:45
---
Hmm, for now I think you can include ext/type_traits.h and do something like:
typedef __gnu_cxx::__conditional_type(sizeof(unsigned long) =
sizeof(void*)),
unsigned long, unsigned long
The _Relative_pointer_impl (and its const variant) implementation invokes
undefined behavior as it computes differences and compares pointers pointing
to different objects.
For example
int i;
RelPtrint p, q;
p.set(i);
q.set(i);
if (p == q)
will reconstruct pointer values based on the
The attached code compiles but dies with a segmentation fault. If the
omp directive is changed from
!$omp parallel private(k,cell)
to
!$omp parallel private(k)
the code executes.
--
Summary: Segfault with allocatable component in private clause
Product: gcc
--- Comment #1 from kargl at gcc dot gnu dot org 2009-01-04 19:41 ---
Created an attachment (id=17030)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17030action=view)
openmp code showing segfault
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38724
--- Comment #3 from jv244 at cam dot ac dot uk 2009-01-04 14:26 ---
The ICE only happens at -O1 and not at -O0 or -O2
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38722
--- Comment #2 from kargl at gcc dot gnu dot org 2009-01-04 19:43 ---
gfc4x -g -o z -O -fopenmp gib.f90
(gdb) run ./z
Starting program: /usr/home/kargl/tmp/z ./z
[New LWP 100056]
[New LWP 100056]
10 10
1 1
Assertion failed: (arena != NULL),
--- Comment #3 from kargl at gcc dot gnu dot org 2009-01-04 19:46 ---
Using gfortran 4.3.3, a segfault also occurs but the backtrace is different.
(gdb) run
Starting program: /usr/home/kargl/tmp/z
[New LWP 100173]
[New Thread 48301140 (LWP 100173)]
10 10
--- Comment #4 from kargl at gcc dot gnu dot org 2009-01-04 19:47 ---
Make the summary a little clear.
--
kargl at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #5 from rguenth at gcc dot gnu dot org 2009-01-04 19:51 ---
Subject: Bug 38720
Author: rguenth
Date: Sun Jan 4 19:51:18 2009
New Revision: 143058
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143058
Log:
2009-01-04 Richard Guenther rguent...@suse.de
PR
--- Comment #10 from steven at gcc dot gnu dot org 2009-01-04 19:51 ---
Created an attachment (id=17031)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17031action=view)
Failing test case (still needs all the .mod files)
This is as far as I can reduce it with delta.
Joost, could
--- Comment #6 from rguenth at gcc dot gnu dot org 2009-01-04 19:52 ---
Fixed.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Status|NEW
--- Comment #12 from laurent at guerby dot net 2009-01-04 10:42 ---
My second patch testing was successful but it added a lot of time to C only
make bootstrap: 20h = 25h30, so about +25% vs --enable-stage1-checking=release
What should we do?
--
--- Comment #11 from steven at gcc dot gnu dot org 2009-01-04 20:02 ---
Note that this test case ICEs in IRA, but I've checked that it ICEs on the same
insn, and in both cases we're looking at incorrect recog_data.
$ gdb --args ../f951.exe -O t.f90
GNU gdb 6.8.0.20080328-cvs
--- Comment #7 from patriciak784-gccmainling at yahoo dot de 2009-01-04
20:26 ---
(In reply to comment #6)
read_original_filename lexes a token, which hits EOF, which
causes the buffer to be popped.
This is sort of an odd scenario.
Perhaps working around it in preprocess_file is
--- Comment #8 from patriciak784-gccmainling at yahoo dot de 2009-01-04
20:53 ---
(In reply to comment #7)
does indeed fix or work around (with gcc 4.3 branch revision somewhere around
rev143000) the original problem ('gcc -E -dM -fpreprocessed - /dev/null') on
mingw32 and x86_64
--- Comment #12 from steven at gcc dot gnu dot org 2009-01-04 21:07 ---
This fixes it for me. However, I'm no RTL expert, and especially combine I
know nothing about :-)
I'll test/post this and see how the RTL guru's judge it.
* combine.c (try_combine): Adjust INSN_CODE after
--- Comment #7 from dfranke at gcc dot gnu dot org 2009-01-04 21:25 ---
(In reply to comment #6)
Stupidity needs to be punished. Assigning myself to fix it.
Patch: http://gcc.gnu.org/ml/fortran/2009-01/msg00049.html
--
dfranke at gcc dot gnu dot org changed:
What
--- Comment #5 from jakub at gcc dot gnu dot org 2009-01-04 21:31 ---
OpenMP 3.0 doesn't cover types with allocatable components, so this is
undefined. If anything IMHO we should reject any attempts to privatize
types with allocatable components, because it isn't clear what will OpenMP
--- Comment #1 from rguenth at gcc dot gnu dot org 2009-01-04 21:34 ---
Subject: Bug 38723
Author: rguenth
Date: Sun Jan 4 21:33:46 2009
New Revision: 143059
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143059
Log:
2009-01-04 Richard Guenther rguent...@suse.de
PR
--- Comment #4 from reichelt at gcc dot gnu dot org 2009-01-04 21:37
---
Jason, the fix is unfortunately incomplete, because the following testcase
still crashes:
=
union U {};
U u({ 1 });
=
bug.cc:3: internal compiler error: in
--- Comment #13 from steven at gcc dot gnu dot org 2009-01-04 21:38 ---
My patch is wrong. The changes are reverted by undo_all() later on. However,
the bug still is in combine.c. It should not leave insns with the wrong
INSN_CODE.
--
--- Comment #14 from steven at gcc dot gnu dot org 2009-01-04 21:39 ---
Leaving this to an RTL expert.
--
steven at gcc dot gnu dot org changed:
What|Removed |Added
The following code snippet triggers an ICE on the trunk:
=
void foo(int i)
{
goto *i;
}
=
bug.cc: In function 'void foo(int)':
bug.cc:1: error: goto destination is neither a label nor a pointer
goto i;
bug.cc:1: internal compiler error: verify_gimple
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.4.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38725
--- Comment #7 from reichelt at gcc dot gnu dot org 2009-01-04 22:08
---
Maybe the bug boils down to PR38725.
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #2 from dberlin at gcc dot gnu dot org 2009-01-04 22:35 ---
Subject: Re: New: default definitions not in avail_out
At one time we pretended they were defined in the entry block, and
IIRC, it worked out okay.
Dunno what happened to this :)
On Sun, Jan 4, 2009 at 1:40 PM,
--- Comment #2 from sebor at roguewave dot com 2009-01-04 22:37 ---
Some additional background on the problem: it's likely that the gcc binary
used to reproduce the problem on Red Hat Enterprise Linux AS release 4 has
been configured and built on SUSE Linux Enterprise Server 10. Should
--- Comment #1 from rguenth at gcc dot gnu dot org 2009-01-04 23:04 ---
IMHO this is invalid. extend.texi says
@smallexample
goto *ptr;
@end smallexample
@noindent
Any expression of type @code{void *} is allowed.
and integers are not promoted to pointers.
This checking is only
--- Comment #3 from rguenth at gcc dot gnu dot org 2009-01-04 23:05 ---
Mine.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #8 from pault at gcc dot gnu dot org 2009-01-04 23:17 ---
Subject: Bug 38665
Author: pault
Date: Sun Jan 4 23:17:37 2009
New Revision: 143064
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=143064
Log:
2009-01-05 Paul Thomas pa...@gcc.gnu.org
PR
--- Comment #3 from jakub at gcc dot gnu dot org 2009-01-04 23:29 ---
You certainly shouldn't move over fixincluded headers, either you need to rerun
fixinclude on the other system, or don't use any fixincluded headers.
Fixincluded headers are modified headers from the system where
--- Comment #2 from jakub at gcc dot gnu dot org 2009-01-04 23:38 ---
The C FE converts this to void * pointer silently:
7168 tree
7169 c_finish_goto_ptr (tree expr)
7170 {
7171 pedwarn (input_location, OPT_pedantic, ISO C forbids %goto *expr;%);
7172 expr = convert (ptr_type_node,
On Linux/ia64, revision 143058 gave
FAIL: gfortran.dg/elemental_subroutine_7.f90 -O0 execution test
--
Summary: gfortran.dg/elemental_subroutine_7.f90 fail on
Linux/ia64
Product: gcc
Version: 4.4.0
Status: UNCONFIRMED
--- Comment #6 from sgk at troutmask dot apl dot washington dot edu
2009-01-05 02:32 ---
Subject: Re: Segfault caused by derived-type with allocatable component in
private clause
On Sun, Jan 04, 2009 at 09:31:20PM -, jakub at gcc dot gnu dot org wrote:
--- Comment #5 from
--- Comment #3 from joseph at codesourcery dot com 2009-01-05 05:05 ---
Subject: Re: [4.4 regression] ICE with goto
This being accepted for C is bug 32122.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38725
--- Comment #15 from jv244 at cam dot ac dot uk 2009-01-05 05:55 ---
Created an attachment (id=17032)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17032action=view)
testcase without module dependencies
based on Steven's testcase
gfortran -c -O1 reduced.f90
reduced.f90: In
76 matches
Mail list logo
