[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: John Fremlin's swap patch



Benjamin Redelings I <bredelin@ucla.edu> writes:

> Your analysis of the problems with the current use of swap_cnt seems
> accurate - though i don't know much about cache miss speeds and such.  I
> guess the problem won't fully be solved until the swapping routine is
> based on pages instead of processes, and can scan pages on the inactive
> list only.

I agree. My patch was just a quick hack, the root of the problem is to
get a good way of finding pages to swap out.

I agree that going through the set of pages that can be discarded is
more reasonable (and logically less complex though I haven't got a
clear picture of how difficult it is to implement).

-- 

	http://altern.org/vii
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux.eu.org/Linux-MM/