Subject: + dnotify-replace-dnotify_mark_mutex-with-mark-mutex-of-dnotify_group.patch added to -mm tree
To: LinoSanfilippo@xxxxxx,eparis@xxxxxxxxxx,viro@xxxxxxxxxxxxxxxxxx
From: akpm@xxxxxxxxxxxxxxxxxxxx
Date: Tue, 11 Jun 2013 13:58:52 -0700
The patch titled
Subject: dnotify: replace dnotify_mark_mutex with mark mutex of dnotify_group
has been added to the -mm tree. Its filename is
dnotify-replace-dnotify_mark_mutex-with-mark-mutex-of-dnotify_group.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Lino Sanfilippo <LinoSanfilippo@xxxxxx>
Subject: dnotify: replace dnotify_mark_mutex with mark mutex of dnotify_group
There is no need to use a special mutex to protect against the fcntl/close
race (see dnotify.c for a description of this race). Instead the
dnotify_groups mark mutex can be used.
Signed-off-by: Lino Sanfilippo <LinoSanfilippo@xxxxxx>
Cc: Eric Paris <eparis@xxxxxxxxxx>
Cc: Al Viro <viro@xxxxxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
fs/notify/dnotify/dnotify.c | 25 +++++++++++++------------
1 file changed, 13 insertions(+), 12 deletions(-)
diff -puN fs/notify/dnotify/dnotify.c~dnotify-replace-dnotify_mark_mutex-with-mark-mutex-of-dnotify_group fs/notify/dnotify/dnotify.c
--- a/fs/notify/dnotify/dnotify.c~dnotify-replace-dnotify_mark_mutex-with-mark-mutex-of-dnotify_group
+++ a/fs/notify/dnotify/dnotify.c
@@ -31,7 +31,6 @@ int dir_notify_enable __read_mostly = 1;
static struct kmem_cache *dnotify_struct_cache __read_mostly;
static struct kmem_cache *dnotify_mark_cache __read_mostly;
static struct fsnotify_group *dnotify_group __read_mostly;
-static DEFINE_MUTEX(dnotify_mark_mutex);
/*
* dnotify will attach one of these to each inode (i_fsnotify_marks) which
@@ -183,7 +182,7 @@ void dnotify_flush(struct file *filp, fl
return;
dn_mark = container_of(fsn_mark, struct dnotify_mark, fsn_mark);
- mutex_lock(&dnotify_mark_mutex);
+ mutex_lock(&dnotify_group->mark_mutex);
spin_lock(&fsn_mark->lock);
prev = &dn_mark->dn;
@@ -199,11 +198,12 @@ void dnotify_flush(struct file *filp, fl
spin_unlock(&fsn_mark->lock);
- /* nothing else could have found us thanks to the dnotify_mark_mutex */
+ /* nothing else could have found us thanks to the dnotify_groups
+ mark_mutex */
if (dn_mark->dn == NULL)
- fsnotify_destroy_mark(fsn_mark, dnotify_group);
+ fsnotify_destroy_mark_locked(fsn_mark, dnotify_group);
- mutex_unlock(&dnotify_mark_mutex);
+ mutex_unlock(&dnotify_group->mark_mutex);
fsnotify_put_mark(fsn_mark);
}
@@ -326,7 +326,7 @@ int fcntl_dirnotify(int fd, struct file
new_dn_mark->dn = NULL;
/* this is needed to prevent the fcntl/close race described below */
- mutex_lock(&dnotify_mark_mutex);
+ mutex_lock(&dnotify_group->mark_mutex);
/* add the new_fsn_mark or find an old one. */
fsn_mark = fsnotify_find_inode_mark(dnotify_group, inode);
@@ -334,7 +334,8 @@ int fcntl_dirnotify(int fd, struct file
dn_mark = container_of(fsn_mark, struct dnotify_mark, fsn_mark);
spin_lock(&fsn_mark->lock);
} else {
- fsnotify_add_mark(new_fsn_mark, dnotify_group, inode, NULL, 0);
+ fsnotify_add_mark_locked(new_fsn_mark, dnotify_group, inode,
+ NULL, 0);
spin_lock(&new_fsn_mark->lock);
fsn_mark = new_fsn_mark;
dn_mark = new_dn_mark;
@@ -348,9 +349,9 @@ int fcntl_dirnotify(int fd, struct file
/* if (f != filp) means that we lost a race and another task/thread
* actually closed the fd we are still playing with before we grabbed
- * the dnotify_mark_mutex and fsn_mark->lock. Since closing the fd is the
- * only time we clean up the marks we need to get our mark off
- * the list. */
+ * the dnotify_groups mark_mutex and fsn_mark->lock. Since closing the
+ * fd is the only time we clean up the marks we need to get our mark
+ * off the list. */
if (f != filp) {
/* if we added ourselves, shoot ourselves, it's possible that
* the flush actually did shoot this fsn_mark. That's fine too
@@ -385,9 +386,9 @@ out:
spin_unlock(&fsn_mark->lock);
if (destroy)
- fsnotify_destroy_mark(fsn_mark, dnotify_group);
+ fsnotify_destroy_mark_locked(fsn_mark, dnotify_group);
- mutex_unlock(&dnotify_mark_mutex);
+ mutex_unlock(&dnotify_group->mark_mutex);
fsnotify_put_mark(fsn_mark);
out_err:
if (new_fsn_mark)
_
Patches currently in -mm which might be from LinoSanfilippo@xxxxxx are
fanotify-fix-races-when-adding-removing-marks.patch
fanotify-put-duplicate-code-for-adding-vfsmount-inode-marks-into-an-own-function.patch
dnotify-replace-dnotify_mark_mutex-with-mark-mutex-of-dnotify_group.patch
inotify-fix-race-when-adding-a-new-watch.patch
fsnotify-update-comments-concerning-locking-scheme.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
