Subject: Re: pageable kernel pmap entries
To: Jason Thorpe <email@example.com>
From: Chris G. Demetriou <firstname.lastname@example.org>
Date: 05/03/1999 16:37:39
Jason Thorpe <email@example.com> writes:
> On Mon, 03 May 1999 16:12:09 -0700
> Jonathan Stone <jonathan@DSG.Stanford.EDU> wrote:
> > Why, given that it has a significant runtime cost, do we need it in
> > the first place?
> (1) Chuq needs it to work properly for the unified buffer cache
> he's been working on.
Further, i'd expect that, assuming usage stays approximately the same,
it'd have much the same effect as the (evil) buffer cache munging that
currently goes on.
Both cases use TLB entries for dynamic mapping, and if active "buffer
cache" virtual space usage was approximately the same i'd expect that
both would use similar numbers of TLB entries and cause equivalent
the real difference is now there's a way to implement the buffer cache
that doesn't involve the gross hack that's there now. 8-)
Chris Demetriou - firstname.lastname@example.org - http://www.netbsd.org/People/Pages/cgd.html
Disclaimer: Not speaking for NetBSD, just expressing my own opinion.