From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755532Ab0DORWi (ORCPT ); Thu, 15 Apr 2010 13:22:38 -0400 Received: from gir.skynet.ie ([193.1.99.77]:49796 "EHLO gir.skynet.ie" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755271Ab0DORVo (ORCPT ); Thu, 15 Apr 2010 13:21:44 -0400 From: Mel Gorman To: linux-mm@kvack.org, linux-fsdevel@vger.kernel.org Cc: linux-kernel@vger.kernel.org, Chris Mason , Dave Chinner , KOSAKI Motohiro , Andi Kleen , Johannes Weiner , Mel Gorman Subject: [PATCH 09/10] vmscan: Setup pagevec as late as possible in shrink_page_list() Date: Thu, 15 Apr 2010 18:21:42 +0100 Message-Id: <1271352103-2280-10-git-send-email-mel@csn.ul.ie> X-Mailer: git-send-email 1.6.5 In-Reply-To: <1271352103-2280-1-git-send-email-mel@csn.ul.ie> References: <1271352103-2280-1-git-send-email-mel@csn.ul.ie> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org shrink_page_list() sets up a pagevec to release pages as according as they are free. It uses significant amounts of stack on the pagevec. This patch adds pages to be freed via pagevec to a linked list which is then freed en-masse at the end. This avoids using stack in the main path that potentially calls writepage(). Signed-off-by: Mel Gorman --- mm/vmscan.c | 34 ++++++++++++++++++++++++++-------- 1 files changed, 26 insertions(+), 8 deletions(-) diff --git a/mm/vmscan.c b/mm/vmscan.c index 9bc1ede..2c22c83 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -619,6 +619,22 @@ static enum page_references page_check_references(struct page *page, return PAGEREF_RECLAIM; } +static void free_page_list(struct list_head *free_list) +{ + struct pagevec freed_pvec; + struct page *page, *tmp; + + pagevec_init(&freed_pvec, 1); + + list_for_each_entry_safe(page, tmp, free_list, lru) { + list_del(&page->lru); + if (!pagevec_add(&freed_pvec, page)) { + __pagevec_free(&freed_pvec); + pagevec_reinit(&freed_pvec); + } + } +} + /* * shrink_page_list() returns the number of reclaimed pages */ @@ -627,13 +643,12 @@ static unsigned long shrink_page_list(struct list_head *page_list, enum pageout_io sync_writeback) { LIST_HEAD(ret_pages); - struct pagevec freed_pvec; + LIST_HEAD(free_list); int pgactivate = 0; unsigned long nr_reclaimed = 0; cond_resched(); - pagevec_init(&freed_pvec, 1); while (!list_empty(page_list)) { enum page_references references; struct address_space *mapping; @@ -808,10 +823,12 @@ static unsigned long shrink_page_list(struct list_head *page_list, __clear_page_locked(page); free_it: nr_reclaimed++; - if (!pagevec_add(&freed_pvec, page)) { - __pagevec_free(&freed_pvec); - pagevec_reinit(&freed_pvec); - } + + /* + * Is there need to periodically free_page_list? It would + * appear not as the counts should be low + */ + list_add(&page->lru, &free_list); continue; cull_mlocked: @@ -834,9 +851,10 @@ keep: list_add(&page->lru, &ret_pages); VM_BUG_ON(PageLRU(page) || PageUnevictable(page)); } + + free_page_list(&free_list); + list_splice(&ret_pages, page_list); - if (pagevec_count(&freed_pvec)) - __pagevec_free(&freed_pvec); count_vm_events(PGACTIVATE, pgactivate); return nr_reclaimed; } -- 1.6.5