Andi Vajda wrote: > On Tue, 18 Jul 2006, David Fraser wrote: >> gcc 4.1.1 with the patch from bug 13212 runs most of the tests fine but: >> - FAILS all 18 tests in test_PyLucene with "Lock obtain timed out" > This is because of the earlier crashes. You probably need to remove > the lock files in /tmp manually. Also, remove the left over index > directory that is created by test_PyLucene.py Indeed, thanks - this version now passes all tests. >> I presume therefore that the Lock obtain timed out is unrelated to the >> threading tests (any ideas Andi), and that this patch solves the >> threading problems on gcc 4.1.1. Yay! > Yeah, this is what I saw on Gentoo with 4.1.1 + the earlier patch by > Bryce > McKinlay. >> The next steps are: >> 1) Run tests for using standard Python threads rather than creating them >> with PyLucene.PythonThread, and see what needs to be added to support >> them (presuming calls to AttachThread methods...) > Yes, this would be the major achievement. If you can get rid of > PyLucene.PythonThread, this would be a big day in PyLucene land ! > > When creating a thread that is not a PyLucene.PythonThread, before > calling any function in libgcj that involves allocating memory, you > need to first call libgcj's JvAttachCurrentThread() API, the API that > was allegedly made to work correctly with that patch to gcj 4.1.1. > There currently is no such API accessible from Python (I did not wrap > it in java.cpp) but it should be pretty simple to add it. Let me know > if you need help with that. Thanks for the pointers. I'll see what I can do, no guarantees, but I'd really like to get this working :-)
Cheers David _______________________________________________ pylucene-dev mailing list [email protected] http://lists.osafoundation.org/mailman/listinfo/pylucene-dev
