Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!cmcl2!husc6!bu-cs!bzs
From: bzs@bu-cs.BU.EDU (Barry Shein)
Newsgroups: net.arch
Subject: Re: VERY LARGE main memories: crypt
Message-ID: <1276@bu-cs.bu-cs.BU.EDU>
Date: Sun, 14-Sep-86 20:12:02 EDT
Article-I.D.: bu-cs.1276
Posted: Sun Sep 14 20:12:02 1986
Date-Received: Sun, 14-Sep-86 21:51:32 EDT
Organization: Boston U. Comp. Sci.
Lines: 23


Am I missing something? What's the huge problem with this hypothetical
vadvise(CYCLIC)? Isn't it just MRU (most recently used)?

It seems to me that the conjecture was that scientific number crunchers
tend to just

	DO 100 I=1,N
	ARR[I] = F(ARR[I])
100	CONTINUE

or something analagous to that, and when done do it again. Thus, the
most recently used page is the one that will be used again the
furthest in the future and the least recently used is the one that
will be used soon. Thus, they do hand to hand combat with LRU
algorithms. Therefore, it seems they want LRU for the text space
(being as there's no particular reason to believe the text is any
different than other programs, locality of reference applies) and MRU
for their (impure) data space (probably LRU for the rest.) Nothing's
perfect, but might that not be better? Anyone know if anyone's tried
it? I assume so.

	-Barry Shein, Boston University