On Aug 25, 10:33 am, Martin Albrecht <m...@informatik.uni-bremen.de> wrote: > Excuse my ignorance, I am sure this was answered a long time ago but I can't > find it right now: why don't you just steal the memory manager back from GMP, > i.e. why does ECL require the GMP memory manager to use its garbage collected > memory?
Yes, this did come up. ECL doesn't do reference counting. If a bignum becomes unreachable, it gets collected by Boehm-Weiser, but mpz_clear () is not called on the mpz_t part. ECL relies on the limbs being allocated by Boehm-Weiser, so that they are reclaimed once they become unreachable. So if we just change the memory manager back, ECL will be creating all kinds of "mpz_t"s without freeing them and hence create a memory leak. ECL does have a "finalize" option, which allows a finalizer to be connected to an object. If we were to change ECL so that every bignum gets a finalizer that calls mpz_clear(), we could change the memory manager back. I have no idea how to do that, though. --~--~---------~--~----~------------~-------~--~----~ To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
