Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756401Ab1BKKmJ (ORCPT ); Fri, 11 Feb 2011 05:42:09 -0500 Received: from mailout-de.gmx.net ([213.165.64.23]:60911 "HELO mailout-de.gmx.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1751453Ab1BKKkh (ORCPT ); Fri, 11 Feb 2011 05:40:37 -0500 X-Authenticated: #4630777 X-Provags-ID: V01U2FsdGVkX18UsLgQ6fEUqDBQ9kZjAMMw2CC13jFRVY1/E3M8N4 3igE/TmIW+e4Fw From: Lino Sanfilippo To: eparis@redhat.com Cc: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, Lino Sanfilippo Subject: [PATCH 1/7] fsnotify: take groups mark_lock before mark lock Date: Fri, 11 Feb 2011 11:36:37 +0100 Message-Id: <1297420603-11715-2-git-send-email-LinoSanfilippo@gmx.de> X-Mailer: git-send-email 1.7.1 In-Reply-To: <1297420603-11715-1-git-send-email-LinoSanfilippo@gmx.de> References: <1297420603-11715-1-git-send-email-LinoSanfilippo@gmx.de> X-Y-GMX-Trusted: 0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2841 Lines: 107 Race-free addition and removal of a mark to a groups mark list would be easier if we could lock the mark list of group before we lock the specific mark. This patch changes the order used to add/remove marks to/from mark lists from 1. mark->lock 2. group->mark_lock 3. inode->i_lock to 1. group->mark_lock 2. mark->lock 3. inode->i_lock Signed-off-by: Lino Sanfilippo --- fs/notify/mark.c | 21 +++++++++------------ 1 files changed, 9 insertions(+), 12 deletions(-) diff --git a/fs/notify/mark.c b/fs/notify/mark.c index 28b64eb..1ac0447 100644 --- a/fs/notify/mark.c +++ b/fs/notify/mark.c @@ -114,22 +114,21 @@ void fsnotify_destroy_mark(struct fsnotify_mark *mark) struct inode *inode = NULL; spin_lock(&mark->lock); - - group = mark->group; - /* something else already called this function on this mark */ if (!(mark->flags & FSNOTIFY_MARK_FLAG_ALIVE)) { spin_unlock(&mark->lock); return; } - mark->flags &= ~FSNOTIFY_MARK_FLAG_ALIVE; + group = mark->group; + spin_unlock(&mark->lock); + + spin_lock(&group->mark_lock); + spin_lock(&mark->lock); /* 1 from caller and 1 for being on i_list/g_list */ BUG_ON(atomic_read(&mark->refcnt) < 2); - spin_lock(&group->mark_lock); - if (mark->flags & FSNOTIFY_MARK_FLAG_INODE) { inode = mark->i.inode; fsnotify_destroy_inode_mark(mark); @@ -137,11 +136,11 @@ void fsnotify_destroy_mark(struct fsnotify_mark *mark) fsnotify_destroy_vfsmount_mark(mark); else BUG(); + spin_unlock(&mark->lock); list_del_init(&mark->g_list); spin_unlock(&group->mark_lock); - spin_unlock(&mark->lock); spin_lock(&destroy_lock); list_add(&mark->destroy_list, &destroy_list); @@ -217,8 +216,8 @@ int fsnotify_add_mark(struct fsnotify_mark *mark, * group->mark_lock * inode->i_lock */ - spin_lock(&mark->lock); spin_lock(&group->mark_lock); + spin_lock(&mark->lock); mark->flags |= FSNOTIFY_MARK_FLAG_ALIVE; mark->group = group; @@ -235,13 +234,11 @@ int fsnotify_add_mark(struct fsnotify_mark *mark, BUG(); if (ret) goto err; - - spin_unlock(&group->mark_lock); - /* this will pin the object if appropriate */ fsnotify_set_mark_mask_locked(mark, mark->mask); spin_unlock(&mark->lock); + spin_unlock(&group->mark_lock); if (inode) __fsnotify_update_child_dentry_flags(inode); @@ -253,8 +250,8 @@ err: mark->group = NULL; atomic_dec(&group->num_marks); - spin_unlock(&group->mark_lock); spin_unlock(&mark->lock); + spin_unlock(&group->mark_lock); spin_lock(&destroy_lock); list_add(&mark->destroy_list, &destroy_list); -- 1.5.6.5 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/