From: fl0p+@andrew.cmu.edu (Frank T Lofaro) Subject: Re: 0.97p2 Date: 30 Aug 1992 10:52:42 GMT
Excerpts from netnews.comp.os.linux: 30-Aug-92 Re: 0.97p2
bhenning@wimsey.bc.ca (Bill Henning) (195)
>If some of the speed problems are due to the page tables being swapped in
>and out it may be worthwhile to allow a compile time option for the >kernal
>to keep page tables locked in memory...
>Bill
In a multi-user environment, users could (and quite possibly will)
abuse this. A better idea would be to have the sticky bit on a file do
this in addition to its normal effect. This seems like it would also be
easier to implement. (And it wouldn't require yet another complier
release! :).