[gclist] Allocator algorithms.

[Nick Barnes]

> We've seen one synthetic program for which the Sun allocator is bad
> news, at least locality-wise.  I don't know if that means anything.

Has anyone out there ever got around to writing a pessimal allocator?
That is, a program which hits the worst case of whichever memory
manager it's plugged into. It would seem to be a fairly
straightforward programming exercise, in non-conforming C. Maybe some
academic on the list could get a student to do one....

Such a monster is not really of great utility, since worst-case
programs don't exist in the real world. But it would be mildly
interesting (and would make measuring the worst case of a black-box
allocator trivial).

Nick Barnes