[gclist] Finalizers & Reference counting.

[Boehm, Hans]

> -----Original Message-----
> From: Lassi A. Tuura [mailto:lassi.tuura@cern.ch]
> 
> There was also discussion on how GCC's memory allocator was spreading
> objects into different pages (by size buckets I believe).  In 
> some cases
> this was causing closely-related objects to be allocated on different
> pages, which of course also affects memory access patterns.  That is
> obviously an issue with that particular implementation of GC; 
> I have no
> idea how hard-wired a design choice it is.

It's also not at all clear to me how much impact that has, or even whether the impact is positive or negative.  For some reason everyone's inutition seems to be that spreading things out is bad.  But caches contain more than one line, so I'm not sure that there's a real cost to accessing a small number of memory regions instead of one.  And segregating by size does mean that objects of the same type tend to be colocated.

Hans