Added documentation comment describing thread race condition.
git-svn-id: svn://svn.code.sf.net/p/loki-lib/code/trunk@821 7ec92016-0320-0410-acc4-a06ded1c099a
This commit is contained in:
parent
b4f940e59b
commit
5767923ea5
1 changed files with 6 additions and 0 deletions
|
@ -485,6 +485,12 @@ namespace Loki
|
|||
/// \par Warning
|
||||
/// There could be a race condition, see bug "Race condition in RefCountedMTAdj::Release"
|
||||
/// http://sourceforge.net/tracker/index.php?func=detail&aid=1408845&group_id=29557&atid=396644
|
||||
/// As stated in bug 1408845, the Release function is not thread safe if a
|
||||
/// SmartPtr copy-constructor tries to copy the last pointer to an object in
|
||||
/// one thread, while the destructor is acting on the last pointer in another
|
||||
/// thread. The existence of a race between a copy-constructor and destructor
|
||||
/// implies a design flaw at a higher level. That race condition must be
|
||||
/// fixed at a higher design level, and no change to this class could fix it.
|
||||
////////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
template <template <class, class> class ThreadingModel,
|
||||
|
|
Loading…
Reference in a new issue