cdevienne2004/08/18 14:17:24
Modified:src ndc.cpp
Log:
Cleaned commented code (made obsolete by the use of ThreadSpecificData_ptr)
Revision ChangesPath
1.6 +2 -19 logging-log4cxx/src/ndc.cpp
Index: ndc.cpp
cdevienne2004/08/18 14:16:08
Modified:include/log4cxx/helpers threadspecificdata.h
Log:
Added default parameter 0 to ThreadSpecificData.reset, so a call without
parameters will just clean the data.
Revision ChangesPath
1.11 +1 -1
logging-log4cxx/include/log4c
[EMAIL PROTECTED] wrote:
cdevienne2004/08/18 14:04:57
Modified:.ChangeLog
include/log4cxx ndc.h
src ndc.cpp
I forgot to specify : I changed NDC::threadSpecificData from
ThreadSpecificData to ThreadSpecificData_ptr. This will fix the
memory leak re
[EMAIL PROTECTED] wrote:
* Hidden thread related classes members (Christophe de Vienne)
* Added TreadSpecificData_ptr (Christophe de Vienne)
* Added CriticalSection::Type so one can use fast mutexes, not only recursive (Christophe de Vienne).
I tested this on linux. Tests on other platforms a
cdevienne2004/08/18 14:04:57
Modified:.ChangeLog
include/log4cxx ndc.h
include/log4cxx/helpers criticalsection.h thread.h
threadspecificdata.h
src criticalsection.cpp ndc.cpp thread.cpp
Christophe de Vienne wrote:
CriticalSection initialise a recursive mutex. For most use it's OK,
but some people may need a simple one for performances issue, and
there is a third one which can report errors.
What I propose is to give the type of mutex we need to the
constructor, with a default t
Michaël CATANZARITI wrote:
Hi,
Christophe de Vienne wrote:
Hi,
I'm currently working on the threading wrappers, mainly hidding the
implementation details dependant from the thread system we use.
I have a few questions :
1) In Thread::~Thread(), the join is done directly instead of using
Thread::
