Re: How can we make page replacement smarter

[Rik van Riel <riel@redhat.com>]

Alan Cox wrote:
>> Files are different.  File content tends to be grouped
>> in large related chunks, both logically in the file and
>> on disk.  Generally there is a lot more file data on a
>> system than what fits in memory.
> 
> Binary paging patterns don't always look like that unfortunately although
> I suspect we might want to be weighing towards paging out R/O file mapped
> pages earlier simply because they are bigger linear chunks

A properly implemented use-once algorithm should be able
to filter out the spatial locality of reference pages from
the temporal locality of reference ones, though...

-- 
Politics is the struggle between those who want to make their country
the best in the world, and those who believe it already is.  Each group
calls the other unpatriotic.