On Wed 05 Feb 2014 16:56, Mark H Weaver <m...@netris.org> writes: > Also: this is on Solaris 10u10. Cole mentioned that on IRC.
On the one hand I don't understand how this can happen. On the other, I unfortunately don't have faith in the buckets-and-chains weak table implementation in 2.0; it usually works but it's gnarly and is a threading nightmare, and accesses disappearing links outside the alloc lock. See for example: http://article.gmane.org/gmane.comp.programming.garbage-collection.boehmgc/4789 In 2.2 the weak table is rewritten to not require the "vacuum" phase and to be threadsafe. That was my best effort at fixing this bug but I couldn't do it with some of the constraints in 2.0. For that reason I am closing this one but feel free to re-open it if you find available hackpower to work towards a solution for 2.0. Andy