From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Mon, 2 Oct 2000 12:48:50 -0300 (BRST) From: Rik van Riel Subject: Re: Ignore my: Re: [PATCH] fix for VM test9-pre7 In-Reply-To: <39D85CBE.28783101@norran.net> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-mm@kvack.org Return-Path: To: Roger Larsson Cc: "linux-mm@kvack.org" , Linus Torvalds List-ID: On Mon, 2 Oct 2000, Roger Larsson wrote: > My report was bogus - I had forgot to do a cp... *grin* > I have now retried. With good results. > The only thing that worries me is that dbench results > has declined a little from earlier test9. > > System is responsive even during mmap002, > and run of mmap002 is much faster then before :-) This is due partly to Ananth' bugfix (we now set the page age right for all pages) and partly due to the dynamic free target we have right now. The dbench thing I have to look into. I haven't figured out yet why dbench performs slightly worse (but maybe it's just because dbench uses an access pattern that benefits from having worse page replacement ;)) > Summary: > + No lookups, kills, good performance... > = I like it. It seems that performance during the misc001 patch is pretty bad at times, but indeed, during mmap001 and mmap002 the system is quite usable (in console mode). regards, Rik -- "What you're running that piece of shit Gnome?!?!" -- Miguel de Icaza, UKUUG 2000 http://www.conectiva.com/ http://www.surriel.com/ -- 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/