Started failing on x86_64-linux only (x86 is o) between
LAST_UPDATED: Tue Aug 16 18:50:47 UTC 2005
LAST_UPDATED: Fri Aug 19 19:26:06 UTC 2005
(gdb) r -O2
-I/home/guerby/work/gcc/build/build-20050819T214115/gcc/testsuite/ada/acats/support
c48009e.adb
Starting program:
--- Additional Comments From bjoern dot m dot haase at web dot de
2005-08-20 13:49 ---
I have re-investigated this bug today and observed that it no longer appears in
mainline. It is also absent in today's cvs state of the 4.0 branch.
Dunno whether the original problem has been
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
13:54 ---
Testcase:
int f(void)
{
int a = 13;
return a[a];
}
This should be converted over to:
int f(void)
{
__builtin_trap ();
}
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23331
--- Additional Comments From jbglaw at lug-owl dot de 2005-08-20 14:04
---
I spotted this bug at some time before while doing VAX development. Taken
from http://www.pergamentum.com/pipermail/linux-vax/2005-July/31.html,
cvs HEAD was already broken (wrt. this bug) at Jul 24, 2005. I
Oh! You have a t complicated way to report bug.
Ok, += operator works only one time in my program later this act as simple =
I'll send you my saved gdb session. I saved the full session, but only last
5-10 lines are important. If you will have a trouble with undersanding fill
free to
--- Additional Comments From hjl at lucon dot org 2005-08-20 15:18 ---
We also need to check divide-by-zero for -minline-int-divide-min-latency and
-minline-int-divide-max-throughput.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23485
--- Additional Comments From roger at eyesopen dot com 2005-08-20 15:27
---
My apologies for adding a comment to an already resolved PR, but I've some
follow-up thoughts on Diego's recent solution to this regression. From a
high-level perspective, it would probably be more efficient to
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-08-20
16:09 ---
Subject: Bug 23239
CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED] 2005-08-20 16:08:52
Modified files:
gcc/testsuite : ChangeLog
--- Additional Comments From danglin at gcc dot gnu dot org 2005-08-20
16:11 ---
Fixed by patch.
--
What|Removed |Added
Status|NEW
--- Additional Comments From danglin at gcc dot gnu dot org 2005-08-20
16:15 ---
Fixed in 4.0.2 and 4.1.0.
--
What|Removed |Added
Status|NEW
--- Additional Comments From falk at debian dot org 2005-08-20 16:19
---
(In reply to comment #0)
Oh! You have a t complicated way to report bug.
We are always eager to improve the process. Maybe you could tell us what you
tried to report this bug and where you had trouble?
If
Whether capacity of strings shrinks depends on whether the strings are shared.
I believe this also true for 4.0.1, but I only have the sources of it, didn't
compile.
To reproduce (note that `s2' is not shared at the time of push):
int
main()
{
string s1;
s1.reserve (1);
string s2 =
I can see a way to improve the speed of java.lang.String.equals. On x86,
sizeof(void *) is 2*sizeof(jchar), whilst on x86_64, sizeof (void *) is
4*sizeof(jchar). So it should be more efficient to compare as many elements as
possible in batches of 2 (on e.g. x86) or 4 (on e.g. x86_64), by casting
--- Additional Comments From greenrd at greenrd dot org 2005-08-20 16:29
---
Created an attachment (id=9549)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=9549action=view)
Benchmark
This is the equality benchmark I used. Uncomment the //ca [0] = 'b'; line and
negate the assertion
following minimal reproduce fails with gcc 4.0.1:
templatetypename T
class TVector
{
public:
T x,y,z,w;
};
templatetypename T
class TQuaternion : public TVectorT
{
public:
T Sum()
{
return x+y+z+w;
}
};
the output produced is:
g++ -rdynamic -c
--- Additional Comments From falk at debian dot org 2005-08-20 16:56
---
Bug 13300 is about a bad assumption with respect to overflow and sign extensions
in loop.c.
--
What|Removed |Added
With today's trunk,
#include complex
void f()
{
std::complexdouble c1(.0),c2(.0);
c1*=c2;
}
g++ -c -O2 -Wall uninit.cpp
Gives:
/home/jan/local/gcc-head/lib/gcc/i686-pc-linux-gnu/4.1.0/../../../../include/c++/4.1.0/complex:1289:
warning: #8216;__t#8217;
After the classpath merge mid-July two undesired info files are being
installed: hacking.info and vmintegration.info.
See http://gcc.gnu.org/ml/gcc/2005-07/msg00779.html for the original
report and Tom Tromey's confirmation that these are undesirable.
--
Summary: [4.1 regression]
After the classpath/libgcj import, installation creates a new directory
$PREFIX/share/classpath/api/ which is empty.
--
Summary: [4.1 regression] libgcj/classpath create empty directory
$PREFIX/share/classpath/api/
Product: gcc
Version: 4.1.0
--- Additional Comments From falk at debian dot org 2005-08-20 17:38
---
I can confirm this on alphaev68-linux, even without -g. C test case:
void f() {
__complex__ double t;
__imag__ t = 0;
}
Was introduced somewhere between 2005-05-17 and 2005-06-30.
--
What
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
18:23 ---
(In reply to comment #0)
i'd be happy if someone could rethink the decisions on this.
If you can get the standard committe to, then we will.
Other than that, we will not.
--
What|Removed
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
18:27 ---
This is a bug in libstdc++ headers. Since complex has been come a scalar and
cannot be loaded
piece wise.
--
What|Removed |Added
--- Additional Comments From hjl at lucon dot org 2005-08-20 18:33 ---
A patch for inlined calls is posted at
http://gcc.gnu.org/ml/gcc-patches/2005-08/msg01223.html
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23485
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
18:34 ---
Why do you think this is a bug.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23494
--- Additional Comments From andreas dot meier_ at gmx dot de 2005-08-20
18:39 ---
Under http://gcc.gnu.org/ml/gcc-patches/2005-03/msg02316.html the patch was
okayed with small changes. Can it be in 4.1.0?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18851
--
What|Removed |Added
CC||pinskia at gcc dot gnu dot
||org
Target Milestone|---
--
What|Removed |Added
CC||pinskia at gcc dot gnu dot
||org
Status|UNCONFIRMED
--
What|Removed |Added
Status|UNCONFIRMED |NEW
Ever Confirmed||1
Last reconfirmed|-00-00 00:00:00
--- Additional Comments From pogonyshev at gmx dot net 2005-08-20 18:48
---
Because it defeats the effect of reserve() call on `s1'. I'm not saying I know
how to avoid it, but I wonder if there is some strict policy behind
`std::basic_string' reallocation behavior in GNU STL. Maybe
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
19:10 ---
Hmm, doing a profiling, I noticed that equals does its own loop instead of
using memcmp which is
most likely more efficient as memcmp is optimized for each target by the OS.
--
--- Additional Comments From rakdver at gcc dot gnu dot org 2005-08-20
19:18 ---
This is not an issue blocking removal of loop.c (if anything, it is in favor of
removal of loop.c).
--
What|Removed |Added
--- Additional Comments From rakdver at gcc dot gnu dot org 2005-08-20
19:21 ---
Sorry, haven't read the instructions (only the title of the metabug :-)
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22366
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
19:29 ---
Actually this still does not work:
From .final_cleanup:
L1:;
_Jv_InitClass (t.class);
;
i = i + 1;
if (i == 1000) goto L4; else goto L1;
--
What|Removed |Added
--
Bug 17574 depends on bug 18399, which changed state.
Bug 18399 Summary: Class initialization optimization does not work with the
inliner
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18399
What|Old Value |New Value
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
19:31 ---
I should copy and paste the full tree dump:
bb 0:
_Jv_InitClass (t.class);
n = 0;
L1:;
_Jv_InitClass (t.class);
n = n + 1;
;
if (n == 1000) goto L4; else goto L1;
The call to _Jv_InitClass is
--
What|Removed |Added
Target Milestone|4.0.1 |---
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18399
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
19:41 ---
Confirmed.
--
What|Removed |Added
Status|UNCONFIRMED |NEW
--- Additional Comments From jan at etpmod dot phys dot tue dot nl
2005-08-20 19:58 ---
(In reply to comment #1)
I can confirm this on alphaev68-linux, even without -g. C test case:
-g should have been -Wall. Sorry about that.
--
Java requires a class initialization check when any public static method is
called. But when we are calling from another method in the _same_ class, we
already know that the class has been initialized, so we can optimize away the
call to _Jv_InitClass, even if we're not inlining. Here's how:
--- Additional Comments From rakdver at gcc dot gnu dot org 2005-08-20
20:05 ---
Caused by http://gcc.gnu.org/ml/gcc-patches/2005-08/msg00283.html, ssa update
needs to be added to final value replacement pass.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23486
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
20:10 ---
Confirmed.
--
What|Removed |Added
Status|UNCONFIRMED |NEW
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-08-20
20:26 ---
Subject: Bug 21692
CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED] 2005-08-20 20:26:26
Modified files:
libjava: ChangeLog configure.host
Added
--- Additional Comments From communa at ua dot fm 2005-08-20 20:30 ---
Sorry I can not reproduce the bug :(
I just recompile binary and it works fine :(
I lose old binary and maybe some changes in source files I made in order to
catch the bug
Sorry,
Dmytro
P.S. if you still need
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
20:32 ---
Closing as invalid then.
--
What|Removed |Added
Status|SUSPENDED
When translating the message
%s ignored with %s and c% %s format
I wanted to switch order of the arguments to get a natural message in Swedish.
So I tried with
%1$s ignorerad med %2$s och %4$s-format 3$c%
That fails, however, with a message
c.c:5: internal compiler error: in
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-08-20
20:37 ---
Subject: Bug 20624
CVSROOT:/cvs/gcc
Module name:gcc
Branch: gcc-4_0-branch
Changes by: [EMAIL PROTECTED] 2005-08-20 20:37:14
Modified files:
gcc:
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
20:37 ---
Fixed in 4.0.2 and above.
--
What|Removed |Added
Status|ASSIGNED
--- Additional Comments From phython at gcc dot gnu dot org 2005-08-20
20:38 ---
int i = (int){1, 2} segfaults as well, and no longer gives an error.
I suspect it was the following patch:
2005-07-20 Giovanni Bajo [EMAIL PROTECTED]
Make CONSTRUCTOR use VEC to store
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
20:40 ---
Already fixed for 4.1.0 by:
2005-06-30 Zack Weinberg [EMAIL PROTECTED]
Jakub Jelinek [EMAIL PROTECTED]
* pretty-print.h (PP_NL_ARGMAX): New.
(text_info): Add locus.
--
What|Removed |Added
GCC target triplet|sh4-unknown-gnu-linux |sh4-*-linux-gnu
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21623
--- Additional Comments From goeran at uddeborg dot se 2005-08-20 20:52
---
Great! (That answer was quick! :-)
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23501
--
What|Removed |Added
GCC build triplet|hppa-unknown-linux-gnu |hppa-*-linux-gnu
GCC host triplet|hppa-unknown-linux-gnu |hppa-*-linux-gnu
GCC target
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
21:09 ---
Hmm, this is not even valid C99, and we reject it in the C front-end with
-std=c99 -pedantic-errors:
t.c:1: error: initializer element is not constant
Though we should not seg fault.
--
--- Additional Comments From falk at debian dot org 2005-08-20 21:10
---
(In reply to comment #0)
If I change line 3043 in alpha.c to load unaligned address of operand[0] into
the
register and pass it to unaligned_storehi() everything works.
That seems like the right thing to do.
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
21:27 ---
_Complex double i = (_Complex double){0.,1.0};
Also ICE but with a differnet error message:
t.cc:1: internal compiler error: in process_init_constructor, at
cp/typeck2.c:1041
Please submit a full bug
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
21:29 ---
I have a fix for this (really I am writting a fix).
--
What|Removed |Added
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
21:47 ---
Another testcase:
int *f = (int){0};
This one is valid C99 also.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23172
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
21:51 ---
I am going to punt as the last example did not work in 2.95.3-4.0.0 also and
that is the only example
in this whole bug which is really valid C99.
--
What|Removed
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
22:08 ---
These also fail at -O0.
--
What|Removed |Added
Component|tree-optimization
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
22:10 ---
Which might mean that libstdc++ is mis-compiled.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23462
During the bootstrap, the GCC build fails at gcc/attribs.c with this error:
attribs.s:514: Error: can't resolve `.text.unlikely' {.text.unlikely section} -
`.LFB96' {.text section}
Line 514 in the .s file is:
.long .LFE96-.LFB96
One label is in a .text section and the other label is in
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
22:18 ---
There are still questions if this is a bug in binutils or gcc.
*** This bug has been marked as a duplicate of 22313 ***
--
What|Removed |Added
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
22:18 ---
*** Bug 23502 has been marked as a duplicate of this bug. ***
--
What|Removed |Added
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
22:20 ---
Actually this effects x86_64 only or if you have a local patch to turn on omit
frame pointer all the time
and -fasynchronous-unwind-tables which is why only x86_64 is effected on a pure
sources.
--
--
What|Removed |Added
CC||ctice at apple dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22313
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
22:21 ---
It sounds like this also effects i686-pc-linux-gnu also with a similar error
and the same backtrace.
--
What|Removed |Added
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-08-20
23:36 ---
Radar 4225347.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23504
--
What|Removed |Added
CC||danglin at gcc dot gnu dot
||org
--
What|Removed |Added
CC||danglin at gcc dot gnu dot
||org
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-08-20
23:55 ---
Subject: Bug 23485
CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED] 2005-08-20 23:55:07
Modified files:
gcc: ChangeLog
gcc/config/ia64:
--- Additional Comments From fitzsim at redhat dot com 2005-08-21 00:00
---
Fixed in Classpath. Closing.
--
What|Removed |Added
Status|NEW
--- Additional Comments From falk at debian dot org 2005-08-21 00:05
---
(In reply to comment #9)
If we really wanted to tackle this better a compile-time, we'd run a
pass to look at all the ARRAY_REFs for those which have an out-of-range
index. It wouldn't be terribly hard to
--- Additional Comments From fitzsim at redhat dot com 2005-08-21 00:12
---
This suggests that a paint event is being delivered to the Canvas before the
canvas's peer has been shown. I don't think this should ever happen though.
We'll re-test this when my latest round of jawt patches
--- Additional Comments From fitzsim at redhat dot com 2005-08-21 00:31
---
Fixed in GNU Classpath. Closing.
--
What|Removed |Added
Status|ASSIGNED
--
Bug 20597 depends on bug 17463, which changed state.
Bug 17463 Summary: new methods introduced in JDK 1.4 missing in java.awt.Window
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17463
What|Old Value |New Value
--- Additional Comments From tromey at gcc dot gnu dot org 2005-08-21
00:31 ---
Thanks, I'll handle this.
--
What|Removed |Added
AssignedTo|unassigned at gcc dot
--- Additional Comments From tromey at gcc dot gnu dot org 2005-08-21
00:31 ---
Thanks, I'll handle this
--
What|Removed |Added
AssignedTo|unassigned at gcc dot
--- Additional Comments From fitzsim at redhat dot com 2005-08-21 00:58
---
I tried to run this under JamVM+Classpath:
$ jamvm -Dpython.home=/home/fitzsim/jython-2.1 -classpath
/home/fitzsim/jython-2.1/jython.jar org.python.util.jython
Jython 2.1 on java1.4.2 (JIT: )
import
--
What|Removed |Added
AssignedTo|graydon at redhat dot com |langel at redhat dot com
Status|NEW |ASSIGNED
--- Additional Comments From fitzsim at redhat dot com 2005-08-21 05:08
---
Fixed in GNU Classpath by Sven de Marothy. Closing.
--
What|Removed |Added
--- Additional Comments From fitzsim at redhat dot com 2005-08-21 05:32
---
Fixed by:
2005-08-21 Thomas Fitzsimmons [EMAIL PROTECTED]
* gnu/java/awt/peer/gtk/GdkGraphics.java (setClip(Shape)): Clear
clip when clip == null.
*
--- Additional Comments From fitzsim at redhat dot com 2005-08-21 05:35
---
Do you have an example program for this? I suspect this was fixed by Sven's
image work.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16824
--- Additional Comments From fitzsim at redhat dot com 2005-08-21 05:38
---
This prints:
Got the image: [EMAIL PROTECTED]
paint 1
1 repaints, probably ok.
Fixed by Sven's image work. Closing.
--
What|Removed |Added
--- Additional Comments From fitzsim at redhat dot com 2005-08-21 05:39
---
Fixed, closing.
--
What|Removed |Added
Status|NEW |RESOLVED
84 matches
Mail list logo
