--
jv244 at cam dot ac dot uk changed:
What|Removed |Added
OtherBugsDependingO||32834
nThis||
Statu
--
dgregor at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |dgregor at gcc dot gnu dot
|dot org
--- Comment #3 from dgregor at gcc dot gnu dot org 2007-12-15 08:31 ---
This is valid code, so it's more serious than we thought.
--
dgregor at gcc dot gnu dot org changed:
What|Removed |Added
---
The following valid program is rejected with
integer(1), parameter :: ARR1(len(HEX1)) = [( MSKa1(i), i=1,len(HEX1) )]
1
Error: index in dimension 1 is out of bounds at (1)
module abuse_mod
implicit none
integer i
character(8), para
--
jv244 at cam dot ac dot uk changed:
What|Removed |Added
OtherBugsDependingO||32834
nThis||
Statu
--- Comment #1 from burnus at gcc dot gnu dot org 2007-12-15 10:08 ---
Created an attachment (id=14765)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14765&action=view)
Draft patch, does not work yet, looks clumsy
Attachment: Very rough draft patch, which does not work, but tries
ith-system-zlib
--with-pkgversion='Pardus Linux' --with-bugurl=http://bugs.pardus.org.tr
Thread model: posix
gcc version 4.3.0 20071215 [trunk revision 130952] (Pardus Linux)
--
Summary: [4.3 regression] Standard autoconf mktime test is too
sl
--- Comment #1 from ismail at pardus dot org dot tr 2007-12-15 10:27
---
Created an attachment (id=14766)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14766&action=view)
testcase
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34477
--- Comment #2 from ismail at pardus dot org dot tr 2007-12-15 10:30
---
Timings with gcc 3.4.6 :
[~]> gcc -o mktime mktime.c
[~]> time ./mktime
real0m0.052s
user0m0.021s
[~]> gcc -O2 -o mktime mktime.c
[~]> time ./mktime
real0m0.032s
user0m0.021s
sys 0m0.011s
S
--- Comment #4 from mmokrejs at ribosome dot natur dot cuni dot cz
2007-12-15 10:59 ---
None of these works:
find . -type l -or -type d -print
find . -type l OR -type d -print
find . -type l or -type d -print
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34474
--- Comment #5 from mmokrejs at ribosome dot natur dot cuni dot cz
2007-12-15 11:00 ---
Created an attachment (id=14767)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14767&action=view)
fixinc.in.patch
There was one more left ...
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?i
--- Comment #3 from pinskia at gcc dot gnu dot org 2007-12-15 11:43 ---
Hmm, I think time_t is signed so this is just a case of signed overflow being
undefined.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34477
--- Comment #4 from pinskia at gcc dot gnu dot org 2007-12-15 11:46 ---
Yes this is invalid, time_t is signed so the following loop is empty:
for (time_t_max = 1; 0 < time_t_max; time_t_max *= 2)
continue;
--
pinskia at gcc dot gnu dot org changed:
What|Removed
in genconditions source, can see following code:
183int main(int argc, char **argv)
184{
...
194 if (init_md_reader (argv[1]) != SUCCESS_EXIT_CODE)
195 return (FATAL_EXIT_CODE);
196
197 condition_table = htab_create (1000, hash_c_test, cmp_c_test, NULL);
--- Comment #5 from ismail at pardus dot org dot tr 2007-12-15 11:52
---
I don't believe this is invalid changing time_t_max and time_t_min to unsigned
long doesn't fix the problem.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34477
--- Comment #6 from ismail at pardus dot org dot tr 2007-12-15 11:52
---
Created an attachment (id=14768)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14768&action=view)
Testcase with unsigned long values
--
ismail at pardus dot org dot tr changed:
What|Remove
--- Comment #1 from pinskia at gcc dot gnu dot org 2007-12-15 12:04 ---
Fixed by:
r110119 | zack | 2006-01-23 07:15:05 -0800 (Mon, 23 Jan 2006) | 37 lines
[EMAIL PROTECTED]: zack | 2006-01-22 14:44:34 -0800
* genconditions.c (condition_table, add_condition): Delete.
(w
--- Comment #4 from pinskia at gcc dot gnu dot org 2007-12-15 12:13 ---
The code is still invalid C++98 so the regression is still just an error
recovery issue :).
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #6 from aldot at gcc dot gnu dot org 2007-12-15 12:49 ---
(In reply to comment #4)
> None of these works:
>
> find . -type l -or -type d -print
> find . -type l OR -type d -print
> find . -type l or -type d -print
>
In current busybox (IIRC >= 1.6) these are available if yo
--- Comment #7 from steven at gcc dot gnu dot org 2007-12-15 13:25 ---
Three questions:
1. According to the 'find(1)' man pages, -o is specified in POSIX. Should GCC
support non-POSIX find commands? IMHO, no.
2. Why would you build in a busybox environment in the first place? This see
--- Comment #1 from jakub at gcc dot gnu dot org 2007-12-15 13:40 ---
--- gcc/varasm.c.jj101 2007-12-06 11:53:42.0 +0100
+++ gcc/varasm.c2007-12-15 14:33:36.0 +0100
@@ -5519,7 +5519,7 @@ decl_default_tls_model (const_tree decl)
is_local = targetm.binds_local_p
--- Comment #8 from mmokrejs at ribosome dot natur dot cuni dot cz
2007-12-15 14:01 ---
The hardware is a Western Digital MyBookWorldII small NAS system, and it is
bundled with the linux and just these tools. Everybody is glad that at least a
working old gcc is available. We just can't
It should produce files -.i and -.s, or better gccdump.i and gccdump.s.
However, it fails with this error:
cc1: error: unrecognized command line option "-.i"
--
Summary: "gcc -x c - -save-temps" does not work
Product: gcc
Version: 4.2.0
Status: UNCO
The following program fails to compile in c++0x mode:
template void* f(A... a) { return new (char*)(a...); }
int main() {
f(__null);
}
Now, this doesn't sound that bad until one notice that this is functionally
equivalent with
#include
int main() {
std::vector v;
v.push_back(NULL
Tring to make gcc with reinstalled cygwin, which already have patch from:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31039
Still getting the same error, this happend then i catching another error from
here: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34414
host-i686-pc-cygwin\intl\config.log:
//
--- Comment #9 from aldot at gcc dot gnu dot org 2007-12-15 14:44 ---
Is suggest you get a current busybox-1.8.2 and just turn on the applets you
need, including DESKTOP support for find (which is a required feature according
to SUSv3, as steven mentioned).
With everything turned on thi
--- Comment #2 from jakub at gcc dot gnu dot org 2007-12-15 14:59 ---
Actually, thinking about it again, this might be a linker bug instead.
Because while code/data in PIEs is relocatable, the relative addresses of
TLS vars within TLS block should be constant, so ld should be able to ins
--- Comment #10 from rguenth at gcc dot gnu dot org 2007-12-15 16:31
---
Why do you need to bootstrap gcc on a small NAS box? Use a cross-compiler.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34474
Executing on host:
/home/dave/gnu/gcc-4.3/objdir/gcc/testsuite/gfortran/../../gf
ortran -B/home/dave/gnu/gcc-4.3/objdir/gcc/testsuite/gfortran/../../
/home/dave/gnu/gcc-4.3/gcc/gcc/testsuite/gfortran.dg/nan_4.f90 -O -std=gnu
-S -o nan_4.s(timeout = 300)
FAIL: gfortran.dg/nan_4.f90 -O (te
--- Comment #2 from danglin at gcc dot gnu dot org 2007-12-15 16:51 ---
I'm seeing this on hppa64-hp-hpux11.11.
--
danglin at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #3 from dominiq at lps dot ens dot fr 2007-12-15 17:00 ---
Same thing on Darwin at -m64. It comes from
http://gcc.gnu.org/ml/gcc-patches/2007-12/msg00277.html, but I don't know why
it has been commited and why it fails with -m64 but not with -m32.
[ibook-dhum] f90/bug% gfc -
Executing on host: /test/gnu/gcc/objdir/gcc/xgcc -B/test/gnu/gcc/objdir/gcc/
/te
st/gnu/gcc/gcc/gcc/testsuite/gcc.dg/struct/wo_prof_two_strs.c -O3
-fipa-struct
-reorg -fdump-ipa-all -fwhole-program -combine -fipa-type-escape
-fno-show-colum
n -lm -o ./wo_prof_two_strs.exe(timeout = 300)
/t
--- Comment #8 from aldot at gcc dot gnu dot org 2007-12-15 17:23 ---
Two disfunctional testcases:
pr31529_1a.i:
extern getline(){}
pr31529_1b.i:
extern getline(){}
pr31529_2a.i:
extern getline(){}
pr31529_2a.i:
extern __inline getline(){}
Trying to fix this in cfg led to verify_curr_p
My libc is configured to omit any FP support (UCLIBC_HAS_FLOATS is not set)
but the recent libbid updates seems to unconditionally pull in floatingpoint
accessor functions thus breaking bootstrap -- since feclearexcept et al are
thus unavailable. My notes on this read:
8<
Follows:
--
aldot at gcc dot gnu dot org changed:
What|Removed |Added
Known to work||4.2.2
Summary|pulls in allegedly unneeded |[4.3 Regression]
--- Comment #11 from aldot at gcc dot gnu dot org 2007-12-15 17:55 ---
Using programs that are required by SUSv3 in their complete feature set is
fine.
--
aldot at gcc dot gnu dot org changed:
What|Removed |Added
---
--- Comment #1 from burnus at gcc dot gnu dot org 2007-12-15 19:01 ---
Hmm, looking at the test suite results, it works on
i686-pc-linux-gnu, i386-pc-linux-gnu,
ia64-suse-linux-gnu, x86_64-unknown-linux-gnu,
arm-unknown-elf, v850-unknown-elf, i686-pc-cygwin,
hppa2.0w-hp-hpux11.11
--- Comment #2 from dave at hiauly1 dot hia dot nrc dot ca 2007-12-15
19:24 ---
Subject: Re: FAIL: gfortran.dg/nan_4.f90 -O tests for errors
> Hmm, looking at the test suite results, it works on
> hppa2.0w-hp-hpux11.11
My last build on hppa2.0w-hp-hpux11.11 was 130926 (i.e., befor
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
CC||pinskia at gcc dot gnu dot
|
--- Comment #1 from pinskia at gcc dot gnu dot org 2007-12-15 20:07 ---
Note it is unusual to disable fp support for x86. So really I think it is a
bug that you can disable it. Actually I think we should declare this as
invalid.
--
pinskia at gcc dot gnu dot org changed:
--- Comment #1 from pinskia at gcc dot gnu dot org 2007-12-15 20:10 ---
Well what type does NULL would have when passed?
And passing 0 also fails.
This might be a defect in the C++0x part of the standard.
Also I think we really do have an ODR violation in libstdc++ if the methods
chan
--- Comment #2 from aldot at gcc dot gnu dot org 2007-12-15 20:15 ---
IIRC i initially experienced this not on i386 but with mips or arm.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34484
--- Comment #11 from aoliva at gcc dot gnu dot org 2007-12-15 20:19 ---
Subject: Bug 7081
Author: aoliva
Date: Sat Dec 15 20:19:23 2007
New Revision: 130960
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=130960
Log:
gcc/ChangeLog:
PR debug/7081
* dwarf2out.c (dwarf_tag_name): Sy
--- Comment #3 from pinskia at gcc dot gnu dot org 2007-12-15 20:22 ---
Even then you have to have some kind of fp support for GCC to work correctly.
Even if it is soft fp.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
---
--- Comment #12 from aoliva at gcc dot gnu dot org 2007-12-15 20:22 ---
1) is now fixed. I wish I'd had time to get to 2), but my time slot for this
is over for the time being, so I'll leave this up for grabs.
--
aoliva at gcc dot gnu dot org changed:
What|Removed
--- Comment #5 from spop at gcc dot gnu dot org 2007-12-15 20:37 ---
Subject: Re: [4.3 Regression] verify_ssa failed with -ftree-loop-linear
I have a fix for this one, cleaning up the complex logic in
can_convert_to_perfect_nest.
Sebastian
--
http://gcc.gnu.org/bugzilla/show_bug.
--- Comment #2 from pinskia at gcc dot gnu dot org 2007-12-15 20:46 ---
There are too many pointers here for a recursive decent parser to handle. This
is common and really just hard to do correctly.
*** This bug has been marked as a duplicate of 31804 ***
--
pinskia at gcc dot gnu
--- Comment #5 from pinskia at gcc dot gnu dot org 2007-12-15 20:46 ---
*** Bug 34308 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 2007-12-15 20:48 ---
This is a bug in config/bfin/bfin.opt . We should not have mutliple lines
descriptions.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
--
dorit at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |dorit at gcc dot gnu dot org
|dot org
--- Comment #3 from burnus at gcc dot gnu dot org 2007-12-15 20:52 ---
(In reply to comment #2)
> > Which version of the MPFR library are you using?
> 4.3.0
That's the GCC version, the MPFR version should be something like 2.2.x or
2.3.0.
> > print *, real(z'')
>
> On hppa-unkn
--- Comment #2 from pinskia at gcc dot gnu dot org 2007-12-15 21:02 ---
This is not the correct location for asking for help on porting. Ask the gcc@
list. But you might as well run cc1 under the debugger and see why there is an
assert happening.
--
pinskia at gcc dot gnu dot org c
--- Comment #4 from burnus at gcc dot gnu dot org 2007-12-15 21:03 ---
As jerry pointed out on IRC, the default integer size could be 8, which
explains the result. Try instead:
print *, real(z'', 4)
print *, transfer(4294967295_8, 0.0_4)
print '(z12)', transfer(real(z'
--- Comment #7 from howarth at nitro dot med dot uc dot edu 2007-12-15
21:35 ---
I find the following works here...
cd /sw/src/fink.build/gcc43-4.2.999-20071214/darwin_objdir/gcc
./xgcc -m64 -B./ -B/sw/lib/gcc4.3/i686-apple-darwin9/bin/ -isystem
/sw/lib/gcc4.3/i686-apple-darwin9/includ
--- Comment #5 from dave at hiauly1 dot hia dot nrc dot ca 2007-12-15
21:51 ---
Subject: Re: FAIL: gfortran.dg/nan_4.f90 -O tests for errors
> --- Comment #4 from burnus at gcc dot gnu dot org 2007-12-15 21:03
> ---
> As jerry pointed out on IRC, the default integer size co
--- Comment #8 from howarth at nitro dot med dot uc dot edu 2007-12-15
21:58 ---
Hmmm... I see...
file *.o
crt3.o:Mach-O 64-bit object x86_64
crtfastmath.o: Mach-O object i386
crtprec32.o: Mach-O object i386
crtprec64.o: Mach-O object i386
crtprec80.o: Mach-O object i386
--- Comment #4 from hjl at lucon dot org 2007-12-15 22:05 ---
Where is __GCC_FLOAT_NOT_NEEDED documented? Is this a supported gcc macro?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34484
--- Comment #14 from ismail at pardus dot org dot tr 2007-12-15 22:05
---
Created an attachment (id=14769)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14769&action=view)
Better patch
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26061
--- Comment #15 from ismail at pardus dot org dot tr 2007-12-15 22:06
---
Attached is a better patch which adds -f{no}-show-error-count and uses it in
regression tests so regtests works now. IDE's also can use this option.
Is it possible to get this in for gcc 4.3 or gcc 4.4?
--
h
The following invalid testcase triggers an ICE since GCC 4.1.0:
==
struct A
{
template struct X;
};
==
bug.cc:3: error: 'T' has not been declared
bug.cc:3: error: declaration of 'template struct A::X'
bug.cc:3: error: shadows template par
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.1.3
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34485
The following invalid testcase triggers an ICE since GCC 4.1.0:
==
template struct A
{
typedef A* X;
};
template struct B
{
using A::X::Y;
};
B b;
==
bug.cc: In instantiation of 'B':
bug.cc:11: instantiated from here
bug.cc:8: intern
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.1.3
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34486
--- Comment #6 from dave at hiauly1 dot hia dot nrc dot ca 2007-12-15
22:29 ---
Subject: Re: FAIL: gfortran.dg/nan_4.f90 -O tests for errors
> That's the GCC version, the MPFR version should be something like 2.2.x or
> 2.3.0.
2.3.0.
Dave
--
http://gcc.gnu.org/bugzilla/show_bu
--- Comment #7 from dave at hiauly1 dot hia dot nrc dot ca 2007-12-15
22:29 ---
Subject: Re: FAIL: gfortran.dg/nan_4.f90 -O tests for errors
> With 4.2.3, the results for the above are:
> 4.2949673E+09 0.00 4F80 0.00 4EFF8000 2.1432893E+09
>
> The first version gave:
>
--- Comment #8 from jvdelisle at gcc dot gnu dot org 2007-12-15 22:34
---
On ppc64 I get on example of #4:
0.000
0.000
0
0.000
0
0.000
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34482
--- Comment #9 from jvdelisle at gcc dot gnu dot org 2007-12-15 22:37
---
The dump for nan_4.f90 has:
test ()
{
static real(kind=4) r = 0.0;
static integer(kind=4) options.0[7] = {68, 127, 0, 0, 0, 1, 0};
_gfortran_set_options (7, (void *) &options.0);
r = 0.0;
... snip ...
The following invalid testcase triggers an ICE since GCC 4.1.2:
template struct A
{
template struct B { typedef A X; };
};
template struct C
{
class A::template B<0>::X x;
};
C c;
bug.cc: In in
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.1.3
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34487
The following invalid testcase triggers an ICE on mainline:
struct A
{
friend A A() {}
};
bug.cc:3: error: return type specification for constructor invalid
bug.cc: In static member function 'stati
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.3.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34488
+===GNAT BUG DETECTED==+
| 4.3.0 20071103 (experimental) [trunk revision 129867] (i686-apple-darwin8)
GCC error:|
| tree check: expected identifier_node, have type_decl in |
| dbxout_symbol, at dbxout.c:2668
Following testcase distilled from x86_64 glibc clock_gettime.c is miscompiled
on the trunk since r128833. CSE1 removes all the c's return value checks
and assumes it wasn't small negative value:
extern void abort (void);
int (*fnp) (int, void *);
unsigned long bar (int, void *);
int
foo (int x,
--
jakub at gcc dot gnu dot org changed:
What|Removed |Added
Status|UNCONFIRMED |NEW
Ever Confirmed|0 |1
Last reconfirm
--- Comment #34 from jakub at gcc dot gnu dot org 2007-12-15 23:09 ---
Subject: Bug 34003
Author: jakub
Date: Sat Dec 15 23:08:47 2007
New Revision: 130979
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=130979
Log:
PR bootstrap/34003
* c-decl.c (merge_decls): Cop
The following invalid code snippet triggers an ICE since GCC 4.0.3:
template struct A;
template<0> struct A
{
static const int i = 0;
};
int n = A::i;
bug.cc:3: error: expected identifier before numeric constant
bug.cc:
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.1.3
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34491
--- Comment #35 from jakub at gcc dot gnu dot org 2007-12-15 23:10 ---
Fixed on the trunk.
--
jakub at gcc dot gnu dot org changed:
What|Removed |Added
Summary
--- Comment #9 from howarth at nitro dot med dot uc dot edu 2007-12-15
23:14 ---
This should probably be added as...
-- gcc-4.3-20071214/gcc/config/i386/t-darwin.org 2007-12-15
18:12:06.0 -0500
+++ gcc-4.3-20071214/gcc/config/i386/t-darwin 2007-12-15 18:12:43.0
--- Comment #9 from dave at hiauly1 dot hia dot nrc dot ca 2007-12-15
23:17 ---
Subject: Re: [4.1/4.2/4.3 Regression] Endless
loop while building a 64-bit 2.6.20 kernel
> What does the full cse1 dump look like at that point (don't forget to call
> fflush(dump_file) from gdb ;-
--- Comment #1 from rogermc at iinet dot net dot au 2007-12-15 23:24
---
Created an attachment (id=14772)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14772&action=view)
concatenated Framework files (compressed)
command generating error (from mac intel, leopard, Xcode version 3.
--- Comment #2 from rogermc at iinet dot net dot au 2007-12-15 23:29
---
Created an attachment (id=14773)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14773&action=view)
concatenated framework files (compressed)
command causing bug generated by Xcode 3.0, Intel Mac Leopard:
gcc
Mac Intel Leopard Xtools 3.0
Compilation of corefoundation-cfbase.ads failed at line 317.
Confused by Ptr parameter name and return type Ptr!
function CFAllocatorReallocate (
Allocator : Cfallocatorref;
Ptr : Void_Ptr;
Newsize : CFIndex;
--- Comment #1 from pinskia at gcc dot gnu dot org 2007-12-15 23:45 ---
And where did you get corefoundation-cfbase.ads from? And why do you think
this is a GCC bug?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34492
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
CC||pinskia at gcc dot gnu dot
|
--- Comment #2 from rogermc at iinet dot net dot au 2007-12-16 00:24
---
The bindings were installed from Carbon_Bindings_xcode.pkg which I think I got
via the www.adapower.net site.
I previously submitted bug 34489 in response to a compilation message
requesting me to do so, so that's
--- Comment #10 from howarth at nitro dot med dot uc dot edu 2007-12-16
00:35 ---
The last formulation doesn't fix the problem.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34025
--- Comment #11 from howarth at nitro dot med dot uc dot edu 2007-12-16
01:05 ---
For some reason the last change isn't forcing the MULTILIB_CFLAGS to actually
be used on the build lines for the objects added to EXTRA_MULTILIB_PARTS.
# Recursively invoke make in the GCC directory to bu
--- Comment #12 from howarth at nitro dot med dot uc dot edu 2007-12-16
01:13 ---
I can't help but wonder if this issue is due to the fact that darwin is the
only architecture that uses a single value for MULTILIB_OPTIONS. We only
have...
MULTILIB_OPTIONS = m64
whereas all the other a
[ Forwarded from http://bugs.debian.org/452811 ]
We're seeing the following ICE with 4.2/4.3 and -funroll-loops:
/home/tbm/14406.c: In function 'fp_mul_comba_small':
/home/tbm/14406.c:2073: error: insn does not satisfy its constraints:
(insn 30288 1486 1488 11 (set (mem/c:DI (plus:SI (reg/f:SI 1
--- Comment #1 from tbm at cyrius dot com 2007-12-16 01:21 ---
Created an attachment (id=14774)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14774&action=view)
preprocessed source
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34493
--- Comment #13 from howarth at nitro dot med dot uc dot edu 2007-12-16
01:38 ---
The last change just creates an endless loop in the build.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34025
--- Comment #26 from steven at gcc dot gnu dot org 2007-12-16 01:50 ---
For Eric's second test case, pr24400-2.c, I have on cygwin in tree-ssa 1006
basic blocks, 2044 edges and "only" 840 DFS back edges. After expanding to RTL
we have 1046 basic blocks, 78005 edges, and 38760 DFS back e
--- Comment #27 from steven at gcc dot gnu dot org 2007-12-16 01:54 ---
Typos. Wanted to say:
"The test case has no cyclic paths, so A is
38760 and the optimal number of passes in a worklist solver is A + 2 = 38762
if the blocks are presented in reverse post order."
--
http://gcc.
--- Comment #3 from rogermc at iinet dot net dot au 2007-12-16 02:08
---
Error occurs for both MacOSX10.4u.sdk and MacOSX10.5.sdk
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34489
--- Comment #2 from tbm at cyrius dot com 2007-12-16 02:45 ---
Apparently it works with 4.3 now, but 4.2 definitely still has the ICE.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34493
--- Comment #4 from jvdelisle at gcc dot gnu dot org 2007-12-16 03:23
---
I think Sebastian committed this patch with the intent to fix the bug. Usually
we don't commit test cases until after we have it fixed and use the PR to track
the issue. So I think we just wait for Sebastian to
--- Comment #2 from pinskia at gcc dot gnu dot org 2007-12-16 04:15 ---
Hmm, for some reason x86-darwin has a different ABI than x86-linux and I
thought they had the same ABIs here.
What version of GCC are you using?
--
pinskia at gcc dot gnu dot org changed:
What|Re
--- Comment #10 from pinskia at gcc dot gnu dot org 2007-12-16 04:53
---
Aftter
2007-11-06 Jakub Jelinek <[EMAIL PROTECTED]>
PR tree-optimization/33993
* tree-vect-transform.c (vect_get_constant_vectors): Use build_vector
rather than build_constructor_from_li
--- Comment #14 from howarth at nitro dot med dot uc dot edu 2007-12-16
05:43 ---
Created an attachment (id=14775)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14775&action=view)
add missing MULTILIB_CFLAGS
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34025
1 - 100 of 102 matches
Mail list logo
