On Jan 9 15:25, Ken Brown wrote: > On 1/9/2025 1:57 PM, Corinna Vinschen wrote: > > Given we don't do filler pages because Windows doesn't let us anyway, we > > could switch to 64K allocation_granularity() bookkeeping now, too. > I'll take a look at doing that before I try to implement your suggestion > about maintaining per page protection in an array. I've got a lot going on > at the moment, so this might take some time.
No worries. Lately I had a look (again) into the MapViewOfFile3 function and the placeholder type of allocation, in the hope this might allow us filler pages. There's also the funny (and wrong) description of MapViewOfFile3 on MSDN claiming "[in] ViewSize The number of bytes to map. A value of zero (0) specifies that the entire section is to be mapped. The size must always be a multiple of the page size." ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ This doesn't match with how file-backed mappings are handled, but grasping for straws, it gave me hope for the most part of a minute... Corinna
