aboutsummaryrefslogtreecommitdiffstats
path: root/fs
diff options
context:
space:
mode:
authorIan Kent <raven@themaw.net>2006-06-25 08:48:47 -0400
committerLinus Torvalds <torvalds@g5.osdl.org>2006-06-25 13:01:18 -0400
commitf9022f66336b8ad680884d4810f1c421fff899e3 (patch)
tree7b2607f0d726c3e14976564fde0f6a570f150625 /fs
parenta2f20c7c274b6241767330c902b3111b9326d901 (diff)
[PATCH] autofs4: need to invalidate children on tree mount expire
I've found a case where invalid dentrys in a mount tree, waiting to be cleaned up by d_invalidate, prevent the expected expire. In this case dentrys created during a lookup for which a mount fails or has no entry in the mount map contribute to the d_count of the parent dentry. These dentrys may not be invalidated prior to comparing the interanl usage count of valid autofs dentrys against the dentry d_count which makes a mount tree appear busy so it doesn't expire. Signed-off-by: Ian Kent <raven@themaw.net> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Diffstat (limited to 'fs')
-rw-r--r--fs/autofs4/expire.c6
1 files changed, 6 insertions, 0 deletions
diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c
index b8ce02607d66..4456d1daa40f 100644
--- a/fs/autofs4/expire.c
+++ b/fs/autofs4/expire.c
@@ -174,6 +174,12 @@ static int autofs4_tree_busy(struct vfsmount *mnt,
174 struct autofs_info *ino = autofs4_dentry_ino(p); 174 struct autofs_info *ino = autofs4_dentry_ino(p);
175 unsigned int ino_count = atomic_read(&ino->count); 175 unsigned int ino_count = atomic_read(&ino->count);
176 176
177 /*
178 * Clean stale dentries below that have not been
179 * invalidated after a mount fail during lookup
180 */
181 d_invalidate(p);
182
177 /* allow for dget above and top is already dgot */ 183 /* allow for dget above and top is already dgot */
178 if (p == top) 184 if (p == top)
179 ino_count += 2; 185 ino_count += 2;