From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail143.messagelabs.com (mail143.messagelabs.com [216.82.254.35]) by kanga.kvack.org (Postfix) with SMTP id 850066B00F3 for ; Wed, 25 Feb 2009 13:42:55 -0500 (EST) Received: from localhost (smtp.ultrahosting.com [127.0.0.1]) by smtp.ultrahosting.com (Postfix) with ESMTP id E8A7E82C6CF for ; Wed, 25 Feb 2009 13:47:39 -0500 (EST) Received: from smtp.ultrahosting.com ([74.213.175.254]) by localhost (smtp.ultrahosting.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ho21NR8Akb-j for ; Wed, 25 Feb 2009 13:47:39 -0500 (EST) Received: from qirst.com (unknown [74.213.171.31]) by smtp.ultrahosting.com (Postfix) with ESMTP id 7BB7582C6D2 for ; Wed, 25 Feb 2009 13:47:25 -0500 (EST) Date: Wed, 25 Feb 2009 13:33:07 -0500 (EST) From: Christoph Lameter Subject: Re: [PATCH 20/20] Get rid of the concept of hot/cold page freeing In-Reply-To: <20090225160124.GA31915@csn.ul.ie> Message-ID: References: <1235344649-18265-1-git-send-email-mel@csn.ul.ie> <1235344649-18265-21-git-send-email-mel@csn.ul.ie> <20090223013723.1d8f11c1.akpm@linux-foundation.org> <20090223233030.GA26562@csn.ul.ie> <20090223155313.abd41881.akpm@linux-foundation.org> <20090224115126.GB25151@csn.ul.ie> <20090224160103.df238662.akpm@linux-foundation.org> <20090225160124.GA31915@csn.ul.ie> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-mm@kvack.org To: Mel Gorman Cc: Andrew Morton , linux-mm@kvack.org, penberg@cs.helsinki.fi, riel@redhat.com, kosaki.motohiro@jp.fujitsu.com, hannes@cmpxchg.org, npiggin@suse.de, linux-kernel@vger.kernel.org, ming.m.lin@intel.com, yanmin_zhang@linux.intel.com List-ID: On Wed, 25 Feb 2009, Mel Gorman wrote: > It'd impact it for sure. Due to the non-temporal stores, I'm surprised > there is any measurable impact from the patch. This has likely been the > case since commit 0812a579c92fefa57506821fa08e90f47cb6dbdd. My reading of > this (someone correct/enlighten) is that even if the data was cache hot, > it is pushed out as a result of the non-temporal access. A nontemporal store simply does not set the used flag for the cacheline. So the cpu cache LRU will evict the cacheline sooner. Thats at least how it works on I64. -- 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