--- Additional Comments From stevenb at suse dot de 2005-03-06 09:30
---
Subject: Re: [meta-bug] optimizations that CSE still catches
On Sunday 06 March 2005 06:59, law at redhat dot com wrote:
Ah. Yes. What did it look like in the tree dumps? Unless
one of the expanders is
Compiling code below (g++ -O5 -c) fails with the error message below.
This is the major source of performance loss since it leaves silly functions
like this called zillions of times w/out inlining.
Any workaround ?
Yuri
code
templatetypename T __attribute__ ((__always_inline__)) T
The attached preprocessed source triggers a segfault in 4.1 (and probably also
4.0) which totally breaks Boost.Serialization and other libraries.
--
Summary: [4.0?/4.1 Regression] Segfualt on valid C++ code (breaks
Boost)
Product: gcc
--- Additional Comments From giovannibajo at libero dot it 2005-03-06
10:05 ---
Created an attachment (id=8341)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8341action=view)
Preprocessed source
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20341
--
/* { dg-do assemble { target x86_64-*-* } } */
/* { dg-options -O2 } */
typedef int vec __attribute__ ((vector_size (8)));
void foo (int j, vec r)
{
unsigned short *dst;
while (j--)
*dst = (unsigned long
--- Additional Comments From belyshev at depni dot sinp dot msu dot ru
2005-03-06 10:15 ---
similar program, but with
-typedef int vec __attribute__ ((vector_size (8)));
+typedef int vec __attribute__ ((mode(DI)));
compiles fine with 3.4.4.
--
What|Removed
--- Additional Comments From yuri at tsoft dot com 2005-03-06 10:45 ---
actually this one is fixed in 4.0, problem though is that 4.0 isn't usable yet
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20340
Another example of failed inlining, now also existent in 4.0 as of 2005-02-13:
When I tried to compile with -O5 -Dinline=__attribute__\(\(__always_inline__\)\)
I get the error message below.
Interestingly there's no inline word here, it must have been taken from some
global declaration ?
---
--- Additional Comments From pcarlini at suse dot de 2005-03-06 11:39
---
I think this can be closed as duplicate of tree-optimization/3713. See also
libstdc++/14078 for explanations.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20343
--
What|Removed |Added
Summary|__always_inline__ fails on |__always_inline__ fails on
|STL templetized function|STL templatized function
$ gnatmake --RTS=/tmp/rts test1.adb
gcc -c --RTS=/tmp/rts test1.adb
gnat1: error: unrecognized command line option -fRTS=/tmp/rts
Note: This works for 4.0.0-20041123 and fails for 4.0.0-20041129 and newer.
--
Summary: [4.0/4.1 regression] gnat1: error: unrecognized command
--- Additional Comments From ebotcazou at gcc dot gnu dot org 2005-03-06
11:54 ---
Confirmed, worked only by accident.
--
What|Removed |Added
--- Additional Comments From ebotcazou at gcc dot gnu dot org 2005-03-06
11:55 ---
It should be fixed by the next push from AdaCore.
--
What|Removed |Added
--- Additional Comments From ebotcazou at gcc dot gnu dot org 2005-03-06
11:56 ---
The change comes from:
2004-11-28 Matt Kraai [EMAIL PROTECTED]
PR diagnostic/17594
* opts.c (find_opt): Require that the input match a switch exactly
or that the switch take a
Forward-declaring a class with a __dllexport__ or __dllimport__
attribute elicits the warning
type attributes are honored only at type definition
even when the same attribute is repeated on the class definition.
Until recently, this usage was accepted with no diagnostic. An
option to inhibit
--- Additional Comments From giovannibajo at libero dot it 2005-03-06
12:13 ---
this might be a fallout of rth's cleanups.
--
What|Removed |Added
CC|
~/gcc4build/bin/g++ -v -c t1.cc
Using built-in specs.
Target: i686-pc-linux-gnu
Configured with: ../gcc-4.0-20050220/configure --prefix=/home/joseph/gcc4build
--enable-languages=c,c++
Thread model: posix
gcc version 4.0.0 20050220 (experimental)
--- Additional Comments From pepster at users dot sourceforge dot net
2005-03-06 12:18 ---
Created an attachment (id=8342)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8342action=view)
source for bug
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20346
--- Additional Comments From pepster at users dot sourceforge dot net
2005-03-06 12:18 ---
Created an attachment (id=8343)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8343action=view)
source for bug
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20346
--- Additional Comments From christian dot joensson at gmail dot com
2005-03-06 12:30 ---
Laurent Guerby suggested I use this in the gcc/ada/Makefile.in:
ifeq ($(strip $(filter-out sparc% linux%,$(arch) $(osys))),)
LIBGNAT_TARGET_PAIRS = \
a-intnam.adsa-intnam-linux.ads \
--- Additional Comments From schwab at suse dot de 2005-03-06 12:30 ---
This is invalid code, uint is not defined as a type.
--
What|Removed |Added
This is a request for simplification of error messages, not a bug report.
Consider the following incorrect program, which generates an error
from inside the STL :
--
#include algorithm
struct A {};
A a[10];
int main() {
std::find(a, a+10, a[1]);
return 0;
--
What|Removed |Added
Version|unknown |4.1.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20342
--- Additional Comments From pepster at users dot sourceforge dot net
2005-03-06 13:05 ---
Sorry. Ignore. Reporting real bug soon ...
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20346
Source submitted as attachment
~/gcc4build/bin/gcc -v -I. -Iinc1 -E t1.cc
Using built-in specs.
Target: i686-pc-linux-gnu
Configured with: ../gcc-4.0-20050220/configure --prefix=/home/joseph/gcc4build
--enable-languages=c,c++
Thread model: posix
gcc version 4.0.0 20050220 (experimental)
--- Additional Comments From pepster at users dot sourceforge dot net
2005-03-06 13:09 ---
Created an attachment (id=8344)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8344action=view)
setup for recreating bug
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20348
--- Additional Comments From jsm28 at gcc dot gnu dot org 2005-03-06 13:22
---
Kazu, you've been fixing a lot of this sort of thing lately, are any
of the problems mentioned in this PR still present?
--
What|Removed |Added
--- Additional Comments From kazu at cs dot umass dot edu 2005-03-06 14:15
---
cfg.c(821): remark #593: variable num_bb_notes was set but never used
Confirmed this in cfghooks.c:verify_flow_info on mainline.
loop-unswitch.c(326): remark #593: variable src was set but never used
--
What|Removed |Added
Component|rtl-optimization|target
Keywords||ssemmx
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
15:10 ---
*** This bug has been marked as a duplicate of 14950 ***
--
What|Removed |Added
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
15:10 ---
*** Bug 20340 has been marked as a duplicate of this bug. ***
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14950
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
15:12 ---
There is no preprocessed source here or even a source?
Could you read http://gcc.gnu.org/bugs.html and attach the needed information?
--
What|Removed |Added
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
15:17 ---
*** Bug 20341 has been marked as a duplicate of this bug. ***
--
What|Removed |Added
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
15:17 ---
Reduces to PR 20186's reduction. Also looks to be the same preprocessed source.
*** This bug has been marked as a duplicate of 20186 ***
--
What|Removed |Added
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
15:26 ---
Just a note that this is actually the include patch which is embeded into gcc
so there is a reason for ../
../ because there could be symbolic links involved.
--
What|Removed
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
15:34 ---
Confirmed. There is some caching going wrong here.
--
What|Removed |Added
--
What|Removed |Added
Last reconfirmed|2004-09-30 17:42:47 |2005-03-06 15:37:17
date||
Target Milestone|---
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
15:40 ---
No feedback in 3 months.
--
What|Removed |Added
Status|WAITING
--
Bug 16651 depends on bug 16371, which changed state.
Bug 16371 Summary: libstdc++ fails for crosses
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16371
What|Old Value |New Value
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
15:40 ---
No feedback in 3 months.
--
What|Removed |Added
Status|WAITING
--- Additional Comments From nick at sqrt dot co dot uk 2005-03-06 15:51
---
This is a longstanding bug that still occurs for the 4.0 branch.
For ARM change:
Integer'Min (2, Standard'Maximum_Alignment);
to:
Integer'Min (4, Standard'Maximum_Alignment);
--
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-03-06
16:02 ---
Subject: Bug 20326
CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED] 2005-03-06 16:02:41
Modified files:
gcc/treelang : ChangeLog Make-lang.in
Added files:
--- Additional Comments From laurent at guerby dot net 2005-03-06 16:21
---
On possible solution for your target is to add in your target ada/Makefile.in
LIBGNAT_TARGET_PAIRS = \
...
s-auxdec.adss-auxdec-arm.ads \
...
And commit s-auxdec-arm.ads with your change.
--
--- Additional Comments From phython at gcc dot gnu dot org 2005-03-06
16:34 ---
Fixed, in mainline, I'll backport this later.
--
What|Removed |Added
Target
--- Additional Comments From tobi at gcc dot gnu dot org 2005-03-06 16:37
---
Does this happen with an up-to-date gfortran? I've just fixed the UBOUND
intrinsic a few days ago. I'm also adding Steve to CC because he confirmed the
bug, and I'm sure he has an up-to-date compiler lying
--- Additional Comments From tobi at gcc dot gnu dot org 2005-03-06 16:39
---
Oh, forgot to say: I'm not seeing any problems with an up-to-date compiler.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20335
--- Additional Comments From phillip dot m dot jordan at gmail dot com
2005-03-06 16:45 ---
I'm downloading gfortran-4.0-20050305 right now, will report back when it's
downloaded and built. (will take a few hours, sorry)
Thanks for looking at the problem.
~phil
--
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-03-06
16:59 ---
Subject: Bug 19311
CVSROOT:/cvs/gcc
Module name:gcc
Branch: gcc-3_4-branch
Changes by: [EMAIL PROTECTED] 2005-03-06 16:59:19
Modified files:
gcc/cp :
--- Additional Comments From sgk at troutmask dot apl dot washington dot
edu 2005-03-06 17:00 ---
Subject: Re: ICE with attached Fortran source
On Sun, Mar 06, 2005 at 04:37:26PM -, tobi at gcc dot gnu dot org wrote:
--- Additional Comments From tobi at gcc dot gnu dot org
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-03-06
17:12 ---
Subject: Bug 19311
CVSROOT:/cvs/gcc
Module name:gcc
Branch: gcc-4_0-branch
Changes by: [EMAIL PROTECTED] 2005-03-06 17:12:12
Modified files:
gcc/cp :
--- Additional Comments From lerdsuwa at gcc dot gnu dot org 2005-03-06
17:16 ---
Fixed in 3.4 and 4.0 branches.
--
What|Removed |Added
Status|ASSIGNED
--- Additional Comments From mark at codesourcery dot com 2005-03-06 18:02
---
Subject: Re: [PR c++/20103] failure to gimplify constructors for addressable
types
Alexandre Oliva wrote:
+case TARGET_EXPR:
+ {
+ tree r = tsubst_copy (t, args, complain, in_decl);
+
+
--- Additional Comments From yuri at tsoft dot com 2005-03-06 18:10 ---
Sorry about this, this appears to be more complex to reporduce than I thought.
Compile the following with the options:
-O5 -Dinline=__attribute__\(\(__always_inline__\)\)
to see the error message.
(gcc 4.0 cvs
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
18:16 ---
Actually this is not a bug.
You need to define inline as inline __attribute__((always_inline)).
--
What|Removed |Added
--- Additional Comments From tobi at gcc dot gnu dot org 2005-03-06 18:10
---
Patch here: http://gcc.gnu.org/ml/fortran/2005-03/msg00091.html
--
What|Removed |Added
--- Additional Comments From pcarlini at suse dot de 2005-03-06 18:18
---
In anyone paying attention to my comment #1 ? This is a *known issue*, basically
we cannot inline currently pointers to functions.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20343
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
18:21 ---
(In reply to comment #5)
In anyone paying attention to my comment #1 ? This is a *known issue*,
basically
we cannot inline currently pointers to functions.
Yes but his definition of inline is incorrect
--- Additional Comments From phillip dot m dot jordan at gmail dot com
2005-03-06 18:33 ---
Just tested with the 20050305 snapshot, and it works perfectly.
Thanks!
~phil
--
What|Removed |Added
--
What|Removed |Added
Target Milestone|--- |4.0.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20335
--- Additional Comments From jsm28 at gcc dot gnu dot org 2005-03-06 18:49
---
The tests have now been changed not to try to use TImode if __hppa__.
However, support for TImode would still make sense as a feature on HPPA64
(in which case titype-1.c and uninit-C.c should be changed
--- Additional Comments From yuri at tsoft dot com 2005-03-06 19:37 ---
You need to define inline as inline __attribute__((always_inline))
Did this, same error message.
In my project I NEED to specify explicitely inlining of the functions.
Including STL since vector::operator[] or
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-03-06
19:48 ---
Subject: Bug 20155
CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED] 2005-03-06 19:48:18
Modified files:
libjava: ChangeLog Makefile.am Makefile.in
--- Additional Comments From pcarlini at suse dot de 2005-03-06 19:49
---
The error message that you are seeing is triggered by 'std::endl', a
manipulator
in standard terms, that cannot currently inlined. I have trouble believing that
the performance bottleneck of your application is
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
19:50 ---
The patch had to be reverted.
--
What|Removed |Added
Status|RESOLVED
--
Bug 17574 depends on bug 20155, which changed state.
Bug 20155 Summary: [4.0/4.1 Regression] libgcj build fails with execvp:
/bin/sh: Argument list too long
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20155
What|Old Value |New Value
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
19:52 ---
Patch for 4.1.0 was posted here:
http://gcc.gnu.org/ml/gcc-patches/2005-03/msg00548.html.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17671
--- Additional Comments From law at redhat dot com 2005-03-06 19:56 ---
Subject: Re: [meta-bug] optimizations that CSE still
catches
On Sun, 2005-03-06 at 09:30 +, stevenb at suse dot de wrote:
--- Additional Comments From stevenb at suse dot de 2005-03-06 09:30
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-03-06
20:15 ---
Subject: Bug 20155
CVSROOT:/cvs/gcc
Module name:gcc
Branch: gcc-4_0-branch
Changes by: [EMAIL PROTECTED] 2005-03-06 20:15:23
Modified files:
libjava:
(...)
make[5]: Entering directory
`/home/users/pluto/rpm/BUILD/gcc-4.0-20050305/obj-amd64-pld-linux/amd64-pld-linux/32/libjava'
depbase=`echo prims.lo | sed 's|[^/]*$|.deps/|;s|\.lo$||'`; \
if /bin/sh ./libtool
--mode=compile
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
20:23 ---
Can you try again? I think this was caused by the patch for PR 20155 which had
to be reverted.
--
What|Removed |Added
--
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |phython at gcc dot gnu dot
|dot org |org
Status|NEW
--- Additional Comments From bangerth at dealii dot org 2005-03-06 21:37
---
I think the whole idea of second-guessing what a user might have meant when
he misspelled an option is debatable. I personally am of the opinion that
this should yield an error or a warning, but that we
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-03-06
21:50 ---
Subject: Bug 20288
CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED] 2005-03-06 21:50:37
Modified files:
gcc: ChangeLog
gcc/config/avr :
--- Additional Comments From steven at gcc dot gnu dot org 2005-03-06
22:14 ---
Just to give people an idea of how close we are to optimizing well enough that
the calls to fold_rtx in CSE are almost all no-ops, here are some numbers
taken over all cc1-i files on amd64:
Number of
-exceptions -fPIC -o threading.s
GNU C++ version 4.1.0 20050306 (experimental) (i686-redhat-linux)
compiled by GNU C version 4.1.0 20050306 (experimental).
GGC heuristics: --param ggc-min-expand=90 --param ggc-min-heapsize=113285
../libcwd/threading.cc:704: internal compiler error: Segmentation
--- Additional Comments From carlo at gcc dot gnu dot org 2005-03-06 22:33
---
Created an attachment (id=8345)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8345action=view)
bzip2-ed preprocessed source code.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20350
This specifies a sub-bug of 18212. The test-case is simple, not the severity!
Given are 2 independet(!) and (almost) empty classes: A1 and A2. A1 has the
main() method. A1 is in jar J1, A2 in jar J2.
The bug: Compiling the jar J1 by giving the output file by -o fails, when
compiled with J2.
--- Additional Comments From bojan at antonovic dot com 2005-03-06 22:43
---
Created an attachment (id=8346)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8346action=view)
jar with class A1
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20351
--- Additional Comments From bojan at antonovic dot com 2005-03-06 22:44
---
Created an attachment (id=8347)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8347action=view)
jar with class A2
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20351
--
What|Removed |Added
OtherBugsDependingO||18212
nThis||
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20351
Executing on host: /xxx/gnu/gcc-3.3/objdir/gcc/g++ -shared-libgcc -B/xxx/gnu/gcc
-3.3/objdir/gcc/ -nostdinc++ -L/xxx/gnu/gcc-3.3/objdir/hppa1.1-hp-hpux10.20/libs
tdc++-v3/src -L/xxx/gnu/gcc-3.3/objdir/hppa1.1-hp-hpux10.20/libstdc++-v3/src/.li
bs -B/opt/gnu/gcc/gcc-4.0.0/hppa1.1-hp-hpux10.20/bin/
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
22:52 ---
Compiling two jars at the same time is known to be broken.
--
What|Removed |Added
--- Additional Comments From danglin at gcc dot gnu dot org 2005-03-06
22:52 ---
This fail also occurs on 4.0.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20352
--- Additional Comments From rakdver at gcc dot gnu dot org 2005-03-06
23:05 ---
This seems to be a cse problem. The code looks like
(set (reg:si 124) (reg:si 5))
...
(set (reg:si 5) (reg:si 124))
...
(set (reg:di 119) (reg:di 5))
...
(set (reg:di 3) (reg:di 119))
cse propagates the
--- Additional Comments From carlo at gcc dot gnu dot org 2005-03-06 23:20
---
Created an attachment (id=8348)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8348action=view)
Same, but without PCH.
Sorry, the previous one was compiled with pch.
This one is without PCH.
--
--- Additional Comments From Thomas dot Koenig at online dot de 2005-03-06
23:34 ---
Updated patch:
http://gcc.gnu.org/ml/gcc-patches/2005-03/msg00566.html
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19568
--- Additional Comments From Thomas dot Koenig at online dot de 2005-03-06
23:34 ---
Patch:
http://gcc.gnu.org/ml/gcc-patches/2005-03/msg00566.html
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20131
When optimizing, GCC converts calls to cos() and sin() (and probably others,
these were the ones I saw) into calls to cosf() and sinf() when cos() and sin()
are called with a float argument. The problem with this is that neither
libgcc nor uClibc provide cosf and sinf! The bug can be worked
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
23:49 ---
Confirmed, reduced testcase:
struct a { int i, j; };
template int i struct mutex_tct { static a S_mutex; };
template int i a mutex_tcti::S_mutex = {0,1};
template extern a mutex_tct0::S_mutex;
template a
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
23:51 ---
It worked with 4.0.0 20041124 so it is semi new.
--
What|Removed |Added
Known to fail|
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
23:54 ---
Note Linux defaults to being glibc which implies that it has C99 math
functions, really uclibc really
should have a different configure file if it does not have C99 functions. Well
technicially it
--- Additional Comments From pcarlini at suse dot de 2005-03-06 23:55
---
The implementation of complex::pow can be definitely improved, and will, as soon
as Roger's work on signbit goes in, still, this issue seems limited to some very
specific, old version of hpux, pre release 11,
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
23:55 ---
Fixed so closing.
--
What|Removed |Added
Status|UNCONFIRMED
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-03-06
23:58 ---
This even works with 4.0.0 20050113 but fails with 4.0.0 20050201 more
prove that this is very
recent regression.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20350
--- Additional Comments From dave at hiauly1 dot hia dot nrc dot ca
2005-03-07 00:06 ---
Subject: Re: FAIL: 26_numerics/complex/pow.cc execution test
The implementation of complex::pow can be definitely improved, and will, as
soon
as Roger's work on signbit goes in, still, this
--- Additional Comments From pcarlini at suse dot de 2005-03-07 00:14
---
Yes, this test doesn't fail on hpux 11. This probably implies a
hpux 10 math library issue. I see the following math routines are
called log, atan2, exp, cos and sin in that order.
Indeed, I was thinking the
--- Additional Comments From kkojima at gcc dot gnu dot org 2005-03-07
00:20 ---
I've traced what's going on:
http://gcc.gnu.org/ml/gcc-patches/2005-03/msg00525.html
It includes a patch for comment.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20331
--- Additional Comments From bojan at antonovic dot com 2005-03-07 00:59
---
The part of the bug that produces the ICE (bus error) can be reproduced by
extracting and compiling all classes from a single jar. Unzip all classes from
SMOOD.jar by
unzip -l SMOOD.jar
and then compile with
--- Additional Comments From dave at hiauly1 dot hia dot nrc dot ca
2005-03-07 01:31 ---
Subject: Re: FAIL: 26_numerics/complex/pow.cc execution test
We have log(x) == -inf for x == 0+. The exp call is returning 0+
but the argument isn't -inf. It's -5.9923104495410517e+307. -inf
--- Additional Comments From pcarlini at suse dot de 2005-03-07 01:32
---
We have log(x) == -inf for x == 0+. The exp call is returning 0+
but the argument isn't -inf. It's -5.9923104495410517e+307. -inf
is -1.7976931348623157e+308.
Interesting. Therefore, 1/3*(-inf) instead of
1 - 100 of 121 matches
Mail list logo