Andi Vajda wrote:
Hm ... didn't realise this. Are there any details on how the deadlocks can occur and what Python 2.4 does differently?
e.g. would it only be in multiprocess situation?
This bug is fixed in python 2.4: http://mail.python.org/pipermail/patches/2004-September/015924.html
Great, thanks for the reference...
Even when not creating threads, with PyLucene there is always one extra thread, libgcj's garbage collector finalizer thread that may need to call into python to release a wrapped python object. It needs to acquire the GIL when it does that.
OK I see, thanks for the info
David _______________________________________________ pylucene-dev mailing list [email protected] http://lists.osafoundation.org/mailman/listinfo/pylucene-dev
