From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Tue, 8 Apr 2008 14:11:35 -0700 From: Andrew Morton Subject: Re: [patch 05/18] SLUB: Slab defrag core Message-Id: <20080408141135.de5a6350.akpm@linux-foundation.org> In-Reply-To: References: <20080404230158.365359425@sgi.com> <20080404230226.847485429@sgi.com> <20080407231129.3c044ba1.akpm@linux-foundation.org> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org Return-Path: To: Christoph Lameter Cc: linux-mm@kvack.org, Mel Gorman , andi@firstfloor.org, Nick Piggin , Rik van Riel , Pekka Enberg List-ID: On Tue, 8 Apr 2008 14:02:46 -0700 (PDT) Christoph Lameter wrote: > On Mon, 7 Apr 2008, Andrew Morton wrote: > > > > Fragmentation is skipped if it was less than a tenth of a second since we > > > last checked a slab cache. An unsuccessful defrag attempt pauses attempts > > > for at least one second. > > > > Can we not do this? It's a really nasty hack. Wall time has almost no > > correlation with reclaim and allocation activity. > > > > If we really cannot think of anything smarter than just throttling then the > > decision regarding when to throttle and for how long should at least be > > driven by something which is vaguely correlated with the present/recent > > allocation/reclaim activity. > > The reclaim interval increases to 1 second if slab reclaim was not > succcessful. I know that. It's still an arbitrary-to-the-point-of-uselessness hack. Reclaim is clocked by scanning rates, allocation rates and disk speed. Not wall time. -- 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