Re: 2.6.33.3-rt19 kernel BUG's

[john stultz <johnstul@us.ibm.com>]

On Fri, 2010-05-07 at 23:58 +0000, Fernando Lopez-Lezcano wrote:
> On Fri, 2010-05-07 at 16:43 -0700, john stultz wrote:
> > On Fri, 2010-05-07 at 12:41 +0200, Thomas Gleixner wrote:
> > > On Thu, 6 May 2010, Fernando Lopez-Lezcano wrote:
> > > 
> > > CC'ed John and Nick
> > > 
> > > > See below... just trying to boot (quad core desktop machine).
> > 
> > Ok. Got this to reproduce. Will hopefully have a fix for your soon.
> > 
> > Thanks for the bug report!
> 
> Thanks for looking at it! It had been a while since I had problems with
> rt so I could not understand why I could not login :-) 

Bah! I should have caught this faster. I kept thinking it had to be some
missed unlock of the sbi->fs_lock, but it ends up its just the
autofs4_lock.

Apparently the conversion from using the dcache_lock -> autofs4_lock
forgot that this function already grabs the autofs_lock for a small
moment, so we end up grabbing the lock, then a moment later grab it
again. Splat.

This patch should resolve it.

Signed-off-by: John Stultz <johnstul@us.ibm.com>

diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c
index 5b26088..a7eaa32 100644
--- a/fs/autofs4/root.c
+++ b/fs/autofs4/root.c
@@ -438,12 +438,10 @@ static int autofs4_revalidate(struct dentry *dentry, struct nameidata *nd)
 	}
 
 	/* Check for a non-mountpoint directory with no contents */
-	spin_lock(&autofs4_lock);
 	if (S_ISDIR(dentry->d_inode->i_mode) &&
 	    !d_mountpoint(dentry) && list_empty(&dentry->d_subdirs)) {
 		DPRINTK("dentry=%p %.*s, emptydir",
 			 dentry, dentry->d_name.len, dentry->d_name.name);
-		spin_unlock(&autofs4_lock);
 
 		if (autofs4_need_mount(flags) || current->link_count) {
 			int status;