On 5/5/12 11:05 AM, Andrew Whitworth wrote:
After some discussion on #parrot last week, we wanted to try and get GC finalization working. I've created the whiteknight/gc_finalize branch to try and do exactly that. This branch makes some non-trivial changes to the interpreter destruction sequence, enables GC finalization unconditionally, and removes the --leak-test commandline option (which was optionally activating a broken version of gc finalization).
Can you provide for the quasi-layperson a definition of "GC finalization."
Right now the branch is passing coretest.
It's passing the basis 'make test' on linux/i386. I will run it through my usual testing routines. Are there any tests that should have been added?
Thank you very much. Jim Keenan _______________________________________________ http://lists.parrot.org/mailman/listinfo/parrot-dev
