From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail6.bemta7.messagelabs.com (mail6.bemta7.messagelabs.com [216.82.255.55]) by kanga.kvack.org (Postfix) with ESMTP id 92F689000BD for ; Tue, 27 Sep 2011 22:19:11 -0400 (EDT) Received: from m1.gw.fujitsu.co.jp (unknown [10.0.50.71]) by fgwmail5.fujitsu.co.jp (Postfix) with ESMTP id 4F8003EE0C2 for ; Wed, 28 Sep 2011 11:19:06 +0900 (JST) Received: from smail (m1 [127.0.0.1]) by outgoing.m1.gw.fujitsu.co.jp (Postfix) with ESMTP id 30E9045DE5D for ; Wed, 28 Sep 2011 11:19:06 +0900 (JST) Received: from s1.gw.fujitsu.co.jp (s1.gw.fujitsu.co.jp [10.0.50.91]) by m1.gw.fujitsu.co.jp (Postfix) with ESMTP id 17B2545DE58 for ; Wed, 28 Sep 2011 11:19:06 +0900 (JST) Received: from s1.gw.fujitsu.co.jp (localhost.localdomain [127.0.0.1]) by s1.gw.fujitsu.co.jp (Postfix) with ESMTP id 0A4EA1DB8054 for ; Wed, 28 Sep 2011 11:19:06 +0900 (JST) Received: from m107.s.css.fujitsu.com (m107.s.css.fujitsu.com [10.240.81.147]) by s1.gw.fujitsu.co.jp (Postfix) with ESMTP id CA2651DB804E for ; Wed, 28 Sep 2011 11:19:05 +0900 (JST) Message-ID: <4E8284C6.1050900@jp.fujitsu.com> Date: Wed, 28 Sep 2011 11:21:58 +0900 From: KOSAKI Motohiro MIME-Version: 1.0 Subject: Re: [PATCH] vmscan: add barrier to prevent evictable page in unevictable list References: <1317174330-2677-1-git-send-email-minchan.kim@gmail.com> In-Reply-To: <1317174330-2677-1-git-send-email-minchan.kim@gmail.com> Content-Type: text/plain; charset=ISO-2022-JP Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: To: minchan.kim@gmail.com Cc: akpm@linux-foundation.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org, jweiner@redhat.com, mel@csn.ul.ie, riel@redhat.com, lee.schermerhorn@hp.com (2011/09/28 10:45), Minchan Kim wrote: > When racing between putback_lru_page and shmem_unlock happens, > progrom execution order is as follows, but clear_bit in processor #1 > could be reordered right before spin_unlock of processor #1. > Then, the page would be stranded on the unevictable list. > > spin_lock > SetPageLRU > spin_unlock > clear_bit(AS_UNEVICTABLE) > spin_lock > if PageLRU() > if !test_bit(AS_UNEVICTABLE) > move evictable list > smp_mb > if !test_bit(AS_UNEVICTABLE) > move evictable list > spin_unlock > > But, pagevec_lookup in scan_mapping_unevictable_pages has rcu_read_[un]lock so > it could protect reordering before reaching test_bit(AS_UNEVICTABLE) on processor #1 > so this problem never happens. But it's a unexpected side effect and we should > solve this problem properly. Do we still need this after Hannes removes scan_mapping_unevictable_pages? > > This patch adds a barrier after mapping_clear_unevictable. > > side-note: I didn't meet this problem but just found during review. -- 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/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: email@kvack.org