Radford Neal: > > there's a danger of getting carried away and essentially rewriting > > malloc. To avoid this, one might try just calling "free" on the > > no-longer-needed object, letting "malloc" then figure out when it can > > be re-used.
Nathan Kurz: > Yes, I think that's what I was anticipating: add a free() equivalent... Radford Neal: > > Unfortunately, that seems not to be safe, because it's > > possible that there is a reference to the no-longer-needed object on > > the PROTECT stack, even though no one should actually be looking at > > it any more. Nathan Kurz: > Can you explain this case? I don't think I understand it. My comment about it not being safe was referring to actually calling the "free" function in the standard C library, not a "free equivalent". Except for small objects, R calls "free" when it concludes that it no longer needs an object, having allocated space for it earlier with "malloc". After "free" is called, R had better not do anything like try to mark it in the garbage collection phase... Keeping a free list apart from the one maintained by malloc/free would I think have to be how it is done, hence my comment about ending up rewriting malloc/free. But it may not be too hard to restrain oneself and only do the simplest things. Radford Neal ______________________________________________ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
