#51 0.3.4b Error - Appender::_addAppender()

[ms]

I have been using 0.2.7 for a few months without
problem. I recently discovered a problem regarding the
Log4cppCleanup class, but noticed that this class has
been removed in the latest (non-stable) release. I
downloaded 0.3.4b which compiled fine. When I try and
use my existing code in VC++ .NET (i just called
Category::getInstance( "name" ) infact), I get an
exception Appender::_addAppender.

An attempt to create a ScopedLock object fails,
seemingly because of null pointers.

void Appender::_addAppender(Appender* appender) {

//REQUIRE(_allAppenders.find(appender->getName()) ==
_getAllAppenders().end())
threading::ScopedLock lock(_appenderMapMutex);
_getAllAppenders()[appender->getName()] = appender;
}

Has anyone else experienced this and are there any
solutions?

When is the next stable release scheduled for?

[Marco Villani]

Logged In: YES
user_id=658521

Sorry for my English.

Your problem (Error [ 650940 ] 0.3.4b Error -
Appender::_addAppender() )
could be similar to this problem:
[ 645270 ] May be there is a problem in PThreads.hh

See:
https://sourceforge.net/tracker/index.php?
func=detail&aid=645270&group_id=15190&atid=11519
0
and then:
http://cvs.sourceforge.net/cgi-
bin/viewcvs.cgi/log4cpp/log4cpp/include/log4cpp/thre
ading/PThreads.hh

CVS log for
log4cpp/log4cpp/include/log4cpp/threading/PThreads.
hh

Revision 1.4 / (view) - annotate - [select for diffs] , Thu
Nov 28 18:15:29 2002 UTC (11 days, 20 hours ago) by
bastiaan
Branch: MAIN
CVS Tags: HEAD
Changes since 1.3: +6 -6 lines
Diff to previous 1.3
Fix Mutex copy bug (#645270) as suggested by Robert
Ballarin.

Best regards.

[Nobody/Anonymous]

Logged In: NO

Hello,

i experienced the same problem, some other bug reports
seem to go in same direction...

Well the cause is just an old problem of dependant global
objects: construction order.

// snip begin+++ appender.cpp

namespace log4cpp {
Appender::AppenderMap* Appender::_allAppenders;
threading::Mutex Appender::_appenderMapMutex; <----
this mutex object poses the problem!!!

// snip end---

The order of initialization is undefined across compilation
units. So if someone uses (like me) static log instance
references at global files scope they must be initialized at the
beginning.

example:

static log4cpp::Category& = log4cpp::Category::getInstance
() .... constructs a new Appender, which gets added to map
of appenders.
If the mutex, which is used to protect the map is not yet
constructed (in another compiltion unit) -> *boom*.

Solutions: not elegant ones yet ...

1.) instead of "threading::Mutex
Appender::_appenderMapMutex;"

use "first time access - first init"

threading::Mutex& _getAppenderMapMutex()
{
static threading::Mutex mutex;
return mutex;
}

2.) #pragma init_seg (non portable)

Regards,

Anastasius Focht

[Steve Pole]

Logged In: YES
user_id=455201

Yep, same thing here on W2K/VC6. The reason why I went
from 0.2.7 to 0.3.4b was the garbled output in my multi-
threaded application, so I guess I'm just gonna go back and
try to live with it...