And the reason there's no impact is that even after growing Lua_State by sizeof(void *), the structure is well below the margin of a certain multiplier.
Which was my point all along .. though it has NOTHING to do with virtual memory or VM page sizes. I have failed to follow your argument, except that it seems that you somehow are connecting the fine-grained malloc() allocations with the behind the scenes coarse grained virtual memory allocation that supports the heap algorithm. The "certain multiplier" for malloc() is not directly coupled to the page allocation size of the VM system.
--Tim
|
