[PATCH] autofs4 - post expire race fix

[PATCH] autofs4 - post expire race fix

[raven]

At the tail end of an expire it's possible for a process to enter
autofs4_wait, with a waitq type of NFY_NONE but find that the expire
is finished. In this cause autofs4_wait will try to create a new wait but 
not notify the daemon leading to a hang. As the wait type is meant to 
delay mount requests from revalidate or lookup during an expire and the 
expire is done all we need to do is check if the dentry is a mountpoint. 
If it's not then we're done.


diff -Nur linux-2.6.12-rc2-mm3.orig/fs/autofs4/waitq.c linux-2.6.12-rc2-mm3/fs/autofs4/waitq.c
--- linux-2.6.12-rc2-mm3.orig/fs/autofs4/waitq.c	2005-05-01 17:04:49.000000000 +0800
+++ linux-2.6.12-rc2-mm3/fs/autofs4/waitq.c	2005-05-01 17:07:18.000000000 +0800
@@ -191,6 +191,13 @@
 	}
 
 	if ( !wq ) {
+		/* Can't wait for an expire if there's no mount */
+		if (notify == NFY_NONE && !d_mountpoint(dentry)) {
+			kfree(name);
+			up(&sbi->wq_sem);
+			return -ENOENT;
+		}
+
 		/* Create a new wait queue */
 		wq = kmalloc(sizeof(struct autofs_wait_queue),GFP_KERNEL);
 		if ( !wq ) {

Re: [PATCH] autofs4 - post expire race fix

[Jeff Moyer]

==> Regarding [PATCH] autofs4 - post expire race fix; raven@themaw.net adds:

raven> At the tail end of an expire it's possible for a process to enter
raven> autofs4_wait, with a waitq type of NFY_NONE but find that the expire
raven> is finished. In this cause autofs4_wait will try to create a new
raven> wait but not notify the daemon leading to a hang. As the wait type
raven> is meant to delay mount requests from revalidate or lookup during an
raven> expire and the expire is done all we need to do is check if the
raven> dentry is a mountpoint. If it's not then we're done.

FWIW, this looks good to me.  (sorry for the late reply)

Thanks,

Jeff