From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1763148AbZD1SIS (ORCPT ); Tue, 28 Apr 2009 14:08:18 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1756325AbZD1SIF (ORCPT ); Tue, 28 Apr 2009 14:08:05 -0400 Received: from gir.skynet.ie ([193.1.99.77]:41653 "EHLO gir.skynet.ie" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755719AbZD1SIE (ORCPT ); Tue, 28 Apr 2009 14:08:04 -0400 Date: Tue, 28 Apr 2009 19:07:59 +0100 From: Mel Gorman To: Andrew Morton Cc: Christoph Lameter , Linux Memory Management List , KOSAKI Motohiro , Nick Piggin , Linux Kernel Mailing List , Lin Ming , Peter Zijlstra , Pekka Enberg , "Zhang, Yanmin" , Hugh Dickins Subject: [PATCH] Properly account for freed pages in free_pages_bulk() and when allocating high-order pages in buffered_rmqueue() V2 Message-ID: <20090428180759.GB18893@csn.ul.ie> References: <1240408407-21848-1-git-send-email-mel@csn.ul.ie> <1240819119.2567.884.camel@ymzhang> <20090427143845.GC912@csn.ul.ie> <1240883957.2567.886.camel@ymzhang> <20090428103159.GB23540@csn.ul.ie> <20090428165129.GA18893@csn.ul.ie> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-15 Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.17+20080114 (2008-01-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org free_pages_bulk() updates the number of free pages in the zone but it is assuming that the pages being freed are order-0. While this is currently always true, it's wrong to assume the order is 0. This patch fixes the problem. buffered_rmqueue() is not updating NR_FREE_PAGES when allocating pages with __rmqueue(). As a result, high-order allocation will appear to increase the number of free pages leading to the situation where the free page count exceeds available RAM. This patch accounts for those allocated pages properly. This is a fix for page-allocator-update-nr_free_pages-only-as-necessary.patch. Changelog since V1 o Change 1UL to 1 as it's unnecessary in this case to be unsigned long Reported-by: Zhang, Yanmin Signed-off-by: Mel Gorman --- mm/page_alloc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 5dd2d59..59eb2e1 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -545,7 +545,7 @@ static void free_pages_bulk(struct zone *zone, int count, zone_clear_flag(zone, ZONE_ALL_UNRECLAIMABLE); zone->pages_scanned = 0; - __mod_zone_page_state(zone, NR_FREE_PAGES, count); + __mod_zone_page_state(zone, NR_FREE_PAGES, count << order); while (count--) { struct page *page; @@ -1151,6 +1151,7 @@ again: } else { spin_lock_irqsave(&zone->lock, flags); page = __rmqueue(zone, order, migratetype); + __mod_zone_page_state(zone, NR_FREE_PAGES, -(1 << order)); spin_unlock(&zone->lock); if (!page) goto failed;