You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@stdcxx.apache.org by Liviu Nicoara <ni...@roguewave.com> on 2006/03/30 19:06:02 UTC
std::locale mt test failing
The following test case fails intermittently with SIGSEGV:
$ uname -a
SunOS doc 5.10 Generic_118822-19 sun4u sparc SUNW,Ultra-Enterprise
$ gcc --version
gcc (GCC) 3.4.4
Copyright (C) 2004 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.
$ cat t.cpp
#include <locale>
#include <pthread.h>
extern "C" void* thread_func (void* pv)
{
for (int i = 0; i < 1000; ++i)
std::locale loc ("de");
return 0;
}
int main ()
{
pthread_t tid [4];
for (int i = 0; i < 4; ++i)
if (pthread_create (tid + i, 0, thread_func, 0))
return 1;
for (int i = 0; i < 4; ++i)
if (pthread_join (tid [i], 0))
return 2;
return 0;
}
$ make t
gcc -D_RWCONFIG=12d
-I/build/nicoara/hal9u1/tests/stdlib/locale/../include
-I./../../../../include -I/build/nicoara/hal9u1/include
-I/build/nicoara/hal9u1/include/ansi -I./../../../..
-I/build/nicoara/hal9u1 -I/build/nicoara/hal9u1/tests/stdlib/locale -I.
-O2 -pthreads --pedantic -Wall -W -Wno-long-long -c
/build/nicoara/hal9u1/tests/stdlib/locale/t.cpp
/build/nicoara/hal9u1/tests/stdlib/locale/t.cpp:5: warning: unused
parameter 'pv'
gcc -pthreads -L./../../../../lib -o t t.o -lm -ltestx1012d
-ltlt1012d -lstd4112d -lsupc++
Segmentation Fault (core dumped)
I do not have a stack trace yet. I will file a JIRA incident soon.
Thanks,
Liviu
Re: std::locale mt test failing
Posted by Liviu Nicoara <ni...@roguewave.com>.
Yep, with SunPro it is getting a shorter one but it shows too t_cancel
and thr_exit_common (in 4.1.4 though).
Martin Sebor wrote:
> Liviu Nicoara wrote:
>> I got it. I will test against dev and file a bug report only if that
>> fails too.
>
> It does with gcc but I haven't been able to reproduce it with Sun
> C++ yet. The stack trace I got with my optimized build isn't very
> helpful:
>
> (gdb) where
> #0 0xff3c26d8 in ?? ()
> #1 0xff3c3330 in ?? ()
> #2 0xff3c5dc0 in ?? ()
> #3 0xff3c5ec4 in ?? ()
> #4 0xff13fd88 in _t_cancel () from /usr/lib/libc.so.1
> #5 0xff13c81c in _thr_exit_common () from /usr/lib/libc.so.1
>
> Martin
>
>> Thanks,
>> Liviu
>>
>> Martin Sebor wrote:
>>
>>> Liviu Nicoara wrote:
>>>
>>>> Liviu Nicoara wrote:
>>>>
>>>>
>>>>> The following test case fails intermittently with SIGSEGV:
>>>>> ...........
>>>>> I will file a JIRA incident soon.
>>>> Will 4.1.4 be revisited and the bug addressed if dev does not show the
>>>> problem? What is the policy in this case?
>>> I don't plan to fix anything but showstoppers in the upcoming (Rogue
>>> Wave SourcePro 9 Update 1) release. I'm also not planning to release
>>> stdcxx 4.1.4. The next stdcxx release will be either 4.2 or 5.0
>>> depending on if we decide to break binary compatibility.
>>>
>>> Martin
>>>
>
Re: std::locale mt test failing
Posted by Martin Sebor <se...@roguewave.com>.
Liviu Nicoara wrote:
> I got it. I will test against dev and file a bug report only if that
> fails too.
It does with gcc but I haven't been able to reproduce it with Sun
C++ yet. The stack trace I got with my optimized build isn't very
helpful:
(gdb) where
#0 0xff3c26d8 in ?? ()
#1 0xff3c3330 in ?? ()
#2 0xff3c5dc0 in ?? ()
#3 0xff3c5ec4 in ?? ()
#4 0xff13fd88 in _t_cancel () from /usr/lib/libc.so.1
#5 0xff13c81c in _thr_exit_common () from /usr/lib/libc.so.1
Martin
>
> Thanks,
> Liviu
>
> Martin Sebor wrote:
>
>>Liviu Nicoara wrote:
>>
>>>Liviu Nicoara wrote:
>>>
>>>
>>>>The following test case fails intermittently with SIGSEGV:
>>>>...........
>>>>I will file a JIRA incident soon.
>>>
>>>Will 4.1.4 be revisited and the bug addressed if dev does not show the
>>>problem? What is the policy in this case?
>>
>>I don't plan to fix anything but showstoppers in the upcoming (Rogue
>>Wave SourcePro 9 Update 1) release. I'm also not planning to release
>>stdcxx 4.1.4. The next stdcxx release will be either 4.2 or 5.0
>>depending on if we decide to break binary compatibility.
>>
>>Martin
>>
Re: std::locale mt test failing
Posted by Liviu Nicoara <ni...@roguewave.com>.
I got it. I will test against dev and file a bug report only if that
fails too.
Thanks,
Liviu
Martin Sebor wrote:
> Liviu Nicoara wrote:
>> Liviu Nicoara wrote:
>>
>>> The following test case fails intermittently with SIGSEGV:
>>> ...........
>>> I will file a JIRA incident soon.
>>
>> Will 4.1.4 be revisited and the bug addressed if dev does not show the
>> problem? What is the policy in this case?
>
> I don't plan to fix anything but showstoppers in the upcoming (Rogue
> Wave SourcePro 9 Update 1) release. I'm also not planning to release
> stdcxx 4.1.4. The next stdcxx release will be either 4.2 or 5.0
> depending on if we decide to break binary compatibility.
>
> Martin
>
Re: std::locale mt test failing
Posted by Martin Sebor <se...@roguewave.com>.
Liviu Nicoara wrote:
> Liviu Nicoara wrote:
>
>>The following test case fails intermittently with SIGSEGV:
>>...........
>>I will file a JIRA incident soon.
>
>
> Will 4.1.4 be revisited and the bug addressed if dev does not show the
> problem? What is the policy in this case?
I don't plan to fix anything but showstoppers in the upcoming (Rogue
Wave SourcePro 9 Update 1) release. I'm also not planning to release
stdcxx 4.1.4. The next stdcxx release will be either 4.2 or 5.0
depending on if we decide to break binary compatibility.
Martin
Re: std::locale mt test failing
Posted by Liviu Nicoara <ni...@roguewave.com>.
Liviu Nicoara wrote:
> The following test case fails intermittently with SIGSEGV:
> ...........
> I will file a JIRA incident soon.
Will 4.1.4 be revisited and the bug addressed if dev does not show the
problem? What is the policy in this case?
Thanks,
Liviu