From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Tue, 25 Jul 2006 17:05:54 -0700 (PDT) From: Christoph Lameter Subject: Re: [PATCH] mm: inactive-clean list In-Reply-To: <44C6B111.9010502@redhat.com> Message-ID: References: <1153167857.31891.78.camel@lappy> <44C30E33.2090402@redhat.com> <44C518D6.3090606@redhat.com> <44C68F0E.2050100@redhat.com> <44C6B111.9010502@redhat.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-mm@kvack.org Return-Path: To: Rik van Riel Cc: Peter Zijlstra , linux-mm , linux-kernel List-ID: On Tue, 25 Jul 2006, Rik van Riel wrote: > > Well, I read the whole thing before I replied and I could not figure this > > one out. Maybe I am too dumb to understand. Could you please explain > > yourself in more detail > > Page state transitions can be very expensive in a virtualized > environment, so it would be good if we had fewer transitions. So the hypervisor indeed tracks each individual page state? Note that I do not propose to change the page state but a counter for page states. I am bit confused about how not touching a page can cause page state transitions. But then I do not know much about hypervisors. What magic is going on in the background that could enable the hypervisor to track counter increments? -- 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-mm.org/ . Don't email: email@kvack.org