--- Comment #13 from mmitchel at gcc dot gnu dot org 2006-07-12 07:22
---
Subject: Bug 28016
Author: mmitchel
Date: Wed Jul 12 07:22:47 2006
New Revision: 115365
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115365
Log:
PR c++/28016
* decl.c (cp_finsh_decl): Do
--- Comment #14 from mmitchel at gcc dot gnu dot org 2006-07-12 07:23
---
Fixed in 4.1.2.
--
mmitchel at gcc dot gnu dot org changed:
What|Removed |Added
Sta
Compiling the following snippet
write(*,100) 0.9
write(*,100) 0.99
write(*,100) 0.999
write(*,100) 0.
write(*,100) 0.9
write(*,100) 0.99
100 format((f3.0))
end
with gcc-gfortran-4.1.1-1.fc5, I get as output
--- Comment #1 from rguenth at gcc dot gnu dot org 2006-07-12 08:16 ---
Fixed in 3.4.4.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Status|
Program1:
with Ada.Text_IO; use Ada.Text_IO;
procedure Main is
task Task1 is
entry Stop;
end Task1;
task body Task1 is
begin
accept Stop;
end Task1;
begin
Put_Line ( "Please wait..." );
Delay 3.0;
abort Task1;
Put_Line ( "Task Task1 Aborted!" );
end main
--- Comment #1 from pinskia at gcc dot gnu dot org 2006-07-12 08:56 ---
0.9 in Fortran is 0.9f in C.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28354
--- Comment #11 from rguenth at gcc dot gnu dot org 2006-07-12 09:25
---
Subject: Bug 28187
Author: rguenth
Date: Wed Jul 12 09:25:40 2006
New Revision: 115366
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115366
Log:
2006-07-12 Richard Guenther <[EMAIL PROTECTED]>
--- Comment #12 from rguenth at gcc dot gnu dot org 2006-07-12 09:26
---
Fixed.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Status|ASSIGNE
/usr/bin/gcc-3.4 -mmmx -msse -Winline --param inline-unit-growth=1
--param large-function-growth=1 -O3 -mtune=pentium4 -march=i686
-fomit-frame-pointer -floop-optimize2 -funroll-loops -ftracer -MT
libfbmmx_la-fbmmx.lo -MD -MP -MF .deps/libfbmmx_la-fbmmx.Tpo -c fbmmx.i -fPIC
-DPIC -o /tm
--- Comment #1 from rainer-gccbugzilla at 7val dot com 2006-07-12 10:09
---
Created an attachment (id=11864)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11864&action=view)
preprocessed source file that causes the bug
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28356
--- Comment #2 from rguenth at gcc dot gnu dot org 2006-07-12 10:53 ---
Works for me with 3.4.0 to 3.4.6.
/space/rguenther/install/gcc-3.4.0/bin/gcc -mmmx -msse -Winline --param
inline-unit-growth=1 --param large-function-growth=1 -O3
-mtune=pentium4 -march=i686 -fomit-frame-po
--- Comment #7 from mnylund at emtrion dot de 2006-07-12 11:17 ---
Here's a workaround for the bug (gcc-3.4.4 qt-3.3.6):
- build with -O1 is ok
- build with -O2 leads to ICE
- build with -O2 -fno-schedule-insns2 is OK
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22250
--- Comment #3 from rainer-gccbugzilla at 7val dot com 2006-07-12 11:26
---
Yes, it fails with the preprocessed file that I attached. I also tried a
"apt-get install --reinstall gcc-3.4 gcc-3.4-base" to make sure that the gcc
installation isn't corrupt. Omitting "-ftracer" stops the err
--- Comment #4 from paul dot richard dot thomas at cea dot fr 2006-07-12
11:54 ---
Tobias,
I am so glad to see that somebody is passing components of derived type arrays;
I bust myself to make that work! Anyway, I partially broke it with a recent
patch - sorry.
Reduced testcase:
SUB
--- Comment #4 from rguenth at gcc dot gnu dot org 2006-07-12 12:25 ---
please report the output from the compiler if you specify -v as additional
argument.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28356
--- Comment #26 from multix at gmail dot com 2006-07-12 12:27 ---
I have 6.5.17 too.
I would propose to either improve the autodetect features or point out that
flag in the atchitecture specific page so that other users won't ask.
I'm available for testing on my box. I Think I have acces
--- Comment #27 from pcarlini at suse dot de 2006-07-12 12:29 ---
(In reply to comment #26)
> I have 6.5.17 too.
> ... or point out that
> flag in the atchitecture specific page so that other users won't ask.
That makes sense. Can you prepare a patch? Thanks in advance.
--
http://g
--- Comment #4 from multix at gmail dot com 2006-07-12 12:32 ---
I already configured using --with-gnu-as but not with --with-gnu-ld.
Possibly the two options need to be specified together, that is the gnu
assembler needs the gnu linker?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?
--- Comment #5 from rainer-gccbugzilla at 7val dot com 2006-07-12 12:44
---
debianSarge tiffany:/tmp$ cat ./x
/usr/bin/gcc-3.4 -v -mmmx -msse -Winline --param inline-unit-growth=1
--param large-function-growth=1 -O3 -mtune=pentium4 -march=i686
-fomit-frame-pointer -floop-optim
i.e. the attached test case with
> gcc (GCC) 4.1.1 20060629 (Red Hat 4.1.1-6) and
> gcc (GCC) 4.1.1 20060525 (Red Hat 4.1.1-1)
gives...
New-Style Correct: const 0xbfdb3d2c, unconst 0xbfdb3d2c
Old-Style Correct: const 0xbfdb3d2c, unconst 0xbfdb3d2c
Old-Style Incorrect: const 0xbfdb3d2c, uncon
--- Comment #1 from caolanm at redhat dot com 2006-07-12 12:51 ---
Created an attachment (id=11865)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11865&action=view)
demo source
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28357
Full version string:
g++-4.1 (GCC) 4.1.2 20060708 (prerelease) (Ubuntu 4.1.1-8ubuntu1)
Same problem exists in Debian 'etch'.
The error message after the location of the crash is:
/usr/include/boost/spirit/core/composite/actions.hpp:106: internal compiler
error: in expand_expr_addr_expr_1, at expr.
--- Comment #2 from jakub at gcc dot gnu dot org 2006-07-12 13:02 ---
Works with 3.4.6 and 3.2.3, reproduced on mainline.
--
jakub at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #1 from peter dot soetens at fmtc dot be 2006-07-12 13:05
---
Created an attachment (id=11866)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11866&action=view)
Compressed (bzip2) preprocessed file
After unzipping, compile this file using:
$ g++-4.1 -c attachment.ii -
--- Comment #5 from mrs at gcc dot gnu dot org 2006-07-12 13:12 ---
Subject: Bug 28280
Author: mrs
Date: Wed Jul 12 13:12:08 2006
New Revision: 115369
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115369
Log:
PR c/28280
* c-parser.c (c_parser_typeof_specifier):
--- Comment #6 from mrs at apple dot com 2006-07-12 13:13 ---
I have checked in a fix for this.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28280
--- Comment #6 from rguenth at gcc dot gnu dot org 2006-07-12 13:36 ---
I cannot reproduce this with any FSF release of the 3.4 series. This must be a
Debian specific bug - please report with them.
--
rguenth at gcc dot gnu dot org changed:
What|Removed
--- Comment #3 from bryce at gcc dot gnu dot org 2006-07-12 13:56 ---
Subject: Bug 28352
Author: bryce
Date: Wed Jul 12 13:56:25 2006
New Revision: 115370
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115370
Log:
2006-07-12 Bryce McKinlay <[EMAIL PROTECTED]>
PR libgc
--- Comment #4 from mckinlay at redhat dot com 2006-07-12 13:59 ---
I've committed a patch which fixes the "off by one" error seen with
ecj-produced bytecode. The missing line number issue is a different bug. It
seems that "gcj -C" simply does not generate any debug info for the "snafu"
--- Comment #7 from reichelt at gcc dot gnu dot org 2006-07-12 14:08
---
Even shorter (valid) testcase:
=
template struct A
{
static const bool i = true;
template struct B {};
B<> b;
};
=
This
--- Comment #5 from patchapp at dberlin dot org 2006-07-12 14:15 ---
Subject: Bug number PR other/28251
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/2006-07/msg00517.html
--
http://gcc.gnu.org/bugz
--- Comment #9 from paul dot richard dot thomas at cea dot fr 2006-07-12
14:20 ---
Created an attachment (id=11867)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11867&action=view)
For discussion, perusal and testing: a beta-release of the TR15581 patch
This patch represents some
--- Comment #10 from paul dot richard dot thomas at cea dot fr 2006-07-12
14:23 ---
Bother! Forget the last test - it is an old, incorrect version. I will post
the right one tonight.
Paul
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20541
fastjar contains the following security problem:
When a JAR archive is extracted with filenames with "../" inside, it can
extract files outside of the current directory (a so called directory
traversal).
Unconspicious users unpacking such files could overwrite their own files,
or even system file
--- Comment #1 from marcus at jet dot franken dot de 2006-07-12 14:50
---
Created an attachment (id=11868)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11868&action=view)
cups.jar
sample file with ../ paths inside.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28359
--- Comment #2 from marcus at jet dot franken dot de 2006-07-12 14:54
---
Created an attachment (id=11869)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11869&action=view)
fix.patch
i would propose this patch to fix this problem.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?i
--- Comment #12 from reichelt at gcc dot gnu dot org 2006-07-12 15:03
---
Reduced (from PR 28219) testcase:
=
template struct A
{
enum { a, b = a };
void foo(A);
};
=
bug.h:4: internal compiler error: tree check: expected int
--- Comment #3 from rguenth at gcc dot gnu dot org 2006-07-12 15:08 ---
Confirmed.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
CC|
--- Comment #3 from fxcoudert at gcc dot gnu dot org 2006-07-12 15:11
---
Subject: Bug 28163
Author: fxcoudert
Date: Wed Jul 12 15:11:11 2006
New Revision: 115372
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115372
Log:
PR fortran/28163
* trans-expr.c (gfc_tr
--- Comment #4 from rguenth at gcc dot gnu dot org 2006-07-12 15:14 ---
Created an attachment (id=11870)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11870&action=view)
different fix
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28359
--- Comment #4 from fxcoudert at gcc dot gnu dot org 2006-07-12 15:17
---
Fixed on 4.2
--
fxcoudert at gcc dot gnu dot org changed:
What|Removed |Added
Statu
--- Comment #5 from fxcoudert at gcc dot gnu dot org 2006-07-12 15:19
---
Confirmed, then, if there's a patch already :)
--
fxcoudert at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #5 from rguenth at gcc dot gnu dot org 2006-07-12 15:19 ---
I have applied the 2nd fix to the upstream (savannah) fastjar CVS.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28359
--- Comment #2 from fxcoudert at gcc dot gnu dot org 2006-07-12 15:21
---
Confirmed, and not target-specific. I'll look into it.
--
fxcoudert at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #11 from reichelt at gcc dot gnu dot org 2006-07-12 15:45
---
Fails since GCC 4.1.0 for me.
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
I get an "undefined reference" error with a recent gcc 4.2 when linking to a
class whose header contains an additional namespace that is references from
that class. This is a very recent error. It started to occur between 20060626
and 20060709.
(sid)1690:[EMAIL PROTECTED]: ~/src/t] make
rm -f *.
--- Comment #28 from paolo at gcc dot gnu dot org 2006-07-12 16:03 ---
Subject: Bug 27878
Author: paolo
Date: Wed Jul 12 16:02:46 2006
New Revision: 115374
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115374
Log:
2006-07-12 Paolo Carlini <[EMAIL PROTECTED]>
PR libst
--- Comment #4 from drow at gcc dot gnu dot org 2006-07-12 16:08 ---
(In reply to comment #3)
> Is it OK to add (set_attr "can_delay" "no") for "tls_get_tp_"
> definition?
I think so, with suitable comment. Some future MIPS architecture may provide a
register for this, in which case th
--- Comment #29 from paolo at gcc dot gnu dot org 2006-07-12 16:10 ---
Subject: Bug 27878
Author: paolo
Date: Wed Jul 12 16:10:12 2006
New Revision: 115375
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115375
Log:
2006-07-12 Paolo Carlini <[EMAIL PROTECTED]>
PR libst
--- Comment #30 from ro at techfak dot uni-bielefeld dot de 2006-07-12
16:11 ---
Subject: Re: GCC 4.1.1 build fails on mips-sgi-irix6.5 (libstdc++)/GCC 4.1.0
worked.
multix at gmail dot com writes:
> I'm available for testing on my box. I Think I have access to a 6.5.4 box
> too... b
--- Comment #31 from pcarlini at suse dot de 2006-07-12 16:11 ---
Doc patch committed to mainline and 4_1-branch.
--
pcarlini at suse dot de changed:
What|Removed |Added
--- Comment #2 from reichelt at gcc dot gnu dot org 2006-07-12 16:14
---
With little modifications one can trigger all sorts of behaviour:
ICEs in different places and accepts-invalid bugs.
*** This bug has been marked as a duplicate of 24791 ***
--
reichelt at gcc dot gnu dot org
--- Comment #2 from reichelt at gcc dot gnu dot org 2006-07-12 16:14
---
*** Bug 28316 has been marked as a duplicate of this bug. ***
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #6 from pault at gcc dot gnu dot org 2006-07-12 16:39 ---
FX
I couldn't do the confirmation from the lab!
Thanks
Paul
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28353
--- Comment #5 from reichelt at gcc dot gnu dot org 2006-07-12 16:39
---
Here's a testcase that crashes even at -O0 since GCC 4.1.0:
===
template struct A
{
A() {}
};
A<0> a;
template<> A<0>::A() {}
===
As Wolfgang already mentione
if you do something like this:
char *buffer;
int i;
buffer = (char *) malloc (sizeof (struct something) * 1024);
for (i=0; i < 1024;i++)
memcpy (buffer + i * sizeof (struct something),&i;sizeof (int));
then when i == 255 there is no data stored in buffer. I mean i can read out the
correct va
--- Comment #1 from tbm at cyrius dot com 2006-07-12 16:51 ---
This was caused by:
r115086 | jason | 2006-06-30 03:15:56 +0200 (Fri, 30 Jun 2006) | 58 lines
--
tbm at cyrius dot com changed:
What|Removed |Added
--
--- Comment #2 from reichelt at gcc dot gnu dot org 2006-07-12 16:52
---
This has been fixed on mainline, the 4.1 branch, and the 4.0 branch.
Probably by Lee Millward's patch for PR 27821.
--
reichelt at gcc dot gnu dot org changed:
What|Removed |
gcc creates invalid code for
--
struct snd_mask {
unsigned int bits[6]; // ** (1) ***
};
static int snd_mask_refine(struct snd_mask *mask, const struct snd_mask *v)
{
struct snd_mask old;
old = *mask;
if (mask->bits[0]==0 && mask->bits[1]==0)
*(char *)0 = 0;
re
--- Comment #6 from pinskia at gcc dot gnu dot org 2006-07-12 16:57 ---
Fastjar no longer is included with GCC so closing as will not fix.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
--
--- Comment #3 from aph at gcc dot gnu dot org 2006-07-12 16:59 ---
I can't duplicate this one.
(gdb) s
gnu.java.lang.MainThread.MainThread(java.lang.Class, java.lang.String[])
([EMAIL PROTECTED], [EMAIL PROTECTED],
[EMAIL PROTECTED])
at /home/aph/gcc/trunk/libjava/gnu/java/lan
--- Comment #5 from tromey at gcc dot gnu dot org 2006-07-12 17:00 ---
Subject: Bug 27271
Author: tromey
Date: Wed Jul 12 17:00:49 2006
New Revision: 115378
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115378
Log:
PR libgcj/27271:
* java/util/zip/ZipFile.java (
--- Comment #6 from cvs-commit at developer dot classpath dot org
2006-07-12 17:01 ---
Subject: Bug 27271
CVSROOT:/cvsroot/classpath
Module name:classpath
Changes by: Tom Tromey 06/07/12 17:00:34
Modified files:
. : ChangeLog
java/uti
--- Comment #1 from pinskia at gcc dot gnu dot org 2006-07-12 17:01 ---
memcpy (buffer + i * sizeof (struct something),&i;sizeof (int));
Well that code should be not compile as there is a syntax error? Also what you
gave is not enough to reproduce the bug, do you have a self containe
--- Comment #7 from tromey at gcc dot gnu dot org 2006-07-12 17:02 ---
Fix checked in.
It works with the test case from the PR.
--
tromey at gcc dot gnu dot org changed:
What|Removed |Added
--
--- Comment #2 from tromey at gcc dot gnu dot org 2006-07-12 17:03 ---
This is no longer a problem.
I think the PR 27271 fix cleaned up the lingering problem.
(It is hard to be sure since this report is so vague...)
--
tromey at gcc dot gnu dot org changed:
What|Remov
--- Comment #2 from pinskia at gcc dot gnu dot org 2006-07-12 17:05 ---
The code is invalid as the anonymous namespace causes an ODR violation.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #3 from pinskia at gcc dot gnu dot org 2006-07-12 17:09 ---
To better explain:
test::MonitorHandle references a class in an anonymous namespace which causes
that class to be different in each TU. Since ODR rules say the class needs to
be the same across TUs so we declare tes
--- Comment #13 from bkoz at gcc dot gnu dot org 2006-07-12 17:15 ---
Volker, great job on the reduction for this. I am once again amazed.
-benjamin
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28217
--- Comment #7 from pinskia at gcc dot gnu dot org 2006-07-12 17:23 ---
Fixed.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
Status|NEW
--- Comment #3 from pinskia at gcc dot gnu dot org 2006-07-12 17:39 ---
Hmm:
SdOptionsPrint &rOldUnconst = (SdOptionsPrint&)(rOther);
Looks like that is we are saying this is a reinterpret_cast.
Related to PR 22132.
I bet the fix for PR 22132, just needs to be expanded for reference t
--- Comment #7 from rguenth at gcc dot gnu dot org 2006-07-12 17:47 ---
Well... maybe the RM has an opinion on this.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #5 from ro at gcc dot gnu dot org 2006-07-12 17:48 ---
No, not at all. On the contrary, I usually avoid GNU ld on non-Linux
platforms.
You just need to make sure that GNU ld isn't installed in
/usr/local/mips-sgi-irix6.5/bin/ld where gcc looks for ld by default.
Rainer
--- Comment #8 from marcus at jet dot franken dot de 2006-07-12 18:30
---
yes, this could still be fixed for release branches ;)
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28359
--- Comment #3 from tromey at gcc dot gnu dot org 2006-07-12 19:29 ---
testing a patch
--
tromey at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|un
--- Comment #11 from pault at gcc dot gnu dot org 2006-07-12 19:41 ---
Created an attachment (id=11871)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11871&action=view)
An unpolluted version of the beta-release patch
I do apologise; the previous diff was relative to a development
--- Comment #4 from tromey at gcc dot gnu dot org 2006-07-12 20:01 ---
Subject: Bug 28329
Author: tromey
Date: Wed Jul 12 20:01:47 2006
New Revision: 115396
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115396
Log:
PR java/28329:
* lang-specs.h: Pass '%U'-based
--- Comment #5 from tromey at gcc dot gnu dot org 2006-07-12 20:06 ---
Fixed on branch.
--
tromey at gcc dot gnu dot org changed:
What|Removed |Added
Status|A
[EMAIL PROTECTED] tests]$ gcc --version
gcc (GCC) 4.1.0
Copyright (C) 2006 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
[EMAIL PROTECTED] tests]$ cat test01.cpp
--- Comment #10 from sje at gcc dot gnu dot org 2006-07-12 21:46 ---
Subject: Bug 28114
Author: sje
Date: Wed Jul 12 21:46:43 2006
New Revision: 115397
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115397
Log:
PR c++/28114
* name-lookup.c (pushtag): Return if we
--- Comment #11 from sje at gcc dot gnu dot org 2006-07-12 21:48 ---
Subject: Bug 28114
Author: sje
Date: Wed Jul 12 21:47:58 2006
New Revision: 115398
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115398
Log:
PR c++/28114
* g++.dg/other/pr28114.C: New.
Added:
--- Comment #14 from jason at gcc dot gnu dot org 2006-07-12 21:58 ---
Subject: Bug 28217
Author: jason
Date: Wed Jul 12 21:57:58 2006
New Revision: 115399
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=115399
Log:
PR c++/28217
* semantics.c (note_decl_for_pch):
--- Comment #4 from craigwd2000 at gmail dot com 2006-07-12 22:08 ---
Created an attachment (id=11872)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11872&action=view)
This is my config.log file that was generated when attempting to compile GCC
4.1.1.
--
http://gcc.gnu.org/bug
--- Comment #5 from craigwd2000 at gmail dot com 2006-07-12 22:10 ---
I have also just updated my iMac G5 to 10.4.7. The error in 7/07/06 comment
occurred just shortly after updating to 10.4.7 too.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28297
--- Comment #6 from pinskia at gcc dot gnu dot org 2006-07-12 22:11 ---
The config.log in the libstdc++ directory under the powerpc-apple-darwin8.7.0
directory would be more useful.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28297
--- Comment #7 from craigwd2000 at gmail dot com 2006-07-12 22:25 ---
Created an attachment (id=11873)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11873&action=view)
My config.log file that's related to Libstdc++.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28297
Consider the following test program:
#include
bool has_bad_chars(std::string const & path)
{
for (std::string::const_iterator c = path.begin(); c != path.end(); c++)
{
unsigned char x = static_cast(*c);
if (x <= 0x1f || x == 0x5c || x == 0x7f)
return true;
}
return
--- Comment #1 from zackw at panix dot com 2006-07-12 22:33 ---
Created an attachment (id=11874)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11874&action=view)
assembly output (bad on top, hand-corrected below)
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28364
--- Comment #2 from pinskia at gcc dot gnu dot org 2006-07-12 22:41 ---
Loop-Copy header is doing it Which means there is a confusion at what is the
real header of the loop here.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28364
--- Comment #3 from zackw at panix dot com 2006-07-12 22:42 ---
I should mention that the exact command line flags were -O2
-fomit-frame-pointer -march=pentium4, and that I hand-tweaked the label numbers
for ease of reading.
Also, -fno-tree-ch does suppress this bad optimization, but in
--- Comment #4 from zackw at panix dot com 2006-07-12 22:48 ---
I remembered that I had a build of 4.2 from last week lying around. It
generates even worse code - still with the duplication of most of the loop,
plus a bunch of unnecessary register fiddling and bad addressing mode choice
--- Comment #5 from pinskia at gcc dot gnu dot org 2006-07-12 22:52 ---
For me on PPC-darwin, it generates pretty good code at just -O2 even though
there is a duplicated "header". The loop is pretty good at scheduling the code
too:
L9:
lbz r0,0(r3)
cmplwi cr7,r0,31
class MyClass
{
public:
MyClass() {}
MyClass(int i) { m_i = i;}
private:
MyClass(const MyClass& x) {};
int m_i;
};
class MyInserter
{
public:
MyInserter() {};
MyInserter& operator<<(const MyClass& m) {m_c = m; return *this;};
private:
MyClass m_c;
};
int main()
{
MyClass a(1);
MyInserter
--- Comment #1 from pinskia at gcc dot gnu dot org 2006-07-12 23:11 ---
The C++ standard changed to allow this after we (GCC) decided that we should
follow the C++ standard.
Anyways this is a dup of bug 25950.
*** This bug has been marked as a duplicate of 25950 ***
--
pinskia at gc
--- Comment #23 from pinskia at gcc dot gnu dot org 2006-07-12 23:11
---
*** Bug 28365 has been marked as a duplicate of this bug. ***
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #6 from rakdver at gcc dot gnu dot org 2006-07-12 23:13 ---
Loop header copying is OK; the result is the one I would expect, it certainly
does not make the code worse (unless you are optimizing for code size), and in
many cases may make it better.
I will have a look at the a
--- Comment #7 from zackw at panix dot com 2006-07-12 23:19 ---
Created an attachment (id=11875)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=11875&action=view)
C test case (with interesting implications)
I've found a plain C test case. In the process, I've found that the way
li
--- Comment #8 from zackw at panix dot com 2006-07-12 23:21 ---
Zdenek: I don't see how you can say that what we get is optimal code "unless
optimizing for size". The code generated *will* be slower than the
alternative.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28364
--- Comment #9 from rakdver at atrey dot karlin dot mff dot cuni dot cz
2006-07-12 23:30 ---
Subject: Re: poor optimization choices when iterating over a std::string
(probably not c++-specific)
> Zdenek: I don't see how you can say that what we get is optimal code "unless
> optimizing
--- Comment #10 from zackw at panix dot com 2006-07-12 23:33 ---
I-cache. Also, more iterations before the branch predictors figure out what's
going on.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28364
1 - 100 of 118 matches
Mail list logo
