>Why does the internal malloc of XFree make the list grow to 157 entries in
>one day when the other allocation method behaves a lot better?  

the xfree internal malloc uses mmap/mumap i believe.   this allows
the X server to free memory back to the system when it is no longer 
in use.

if you look at the libc malloc's free, you'll note that it doesn't 
actually return the memory to the system (it keeps it for reuse).
[see src/lib/libc/stdlib/malloc.c]   thus if you cause X to briefly 
malloc a large chunk of RAM it will never free it back to the system
(although if memory becomes scarce it should get paged out if not used).


>And why do things become so slow when the list length grows to 150+ entries?

i don't have a clear understanding of the slowdown, so it is hard to say.
you could try a uvm kernel and see if that makes any difference.

chuck