On Sunday, March 23, 2025, David G. Johnston <david(dot)g(dot)johnston(at)gmail(dot)com>
wrote:
>
> Maybe IOS helps though I do wonder whether a sequential scan skips over
> known all-dead pages making that relative benefit go away.
>
Well, no, since it tracks known visible, not known non-visible, though for
something like a fast exists test with no where clause if the VM had any 1
bits the answer to the exists would be yes. No idea if we take that
shortcut.
David J.