Re: repeatable ZFS panic: share->excl

Attilio Rao wrote:
> 2009/3/13, John Baldwin <john_at_baldwin.cx>:
>> This is similar to the patch I've asked lulf_at_ to test except that 
>> it is longer and I fix a bug where zfs_lookup() can leak a vnode 
>> lock if the access check fails. :-)  The last one I sent to lulf_at_ 
>> is at www.FreeBSD.org/~jhb/patches/zfs_ea.patch.
> 
> I really thought zfs_lookup() was returning the lock held as a 
> feature assming no LOCKLEAF. Is that a bug instead?

Yes, I think that is the real bug.  Looking at this further I think
zfs_get_xattrdir() will return the vnode locked if it has to create a
new node via zfs_make_attrdir() but only returns it held and unlocked if
it finds an existing one.  So my new patch is to just fix
zfs_get_xattrdir() to unlock the vnode if it creates a new one like so:

(Sorry, TBird is probably going to butcher all the whitespace):

---
//depot/user/jhb/lock/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_dir.c
+++
/Users/jhb/work/p4/lock/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_dir.c
_at__at_ -940,6 +940,7 _at__at_
  		/* NB: we already did dmu_tx_wait() if necessary */
  		goto top;
  	}
+	VOP_UNLOCK(*xvpp, 0);

  	return (error);
  }

A non-butchered version is at www.FreeBSD.org/~jhb/patches/zfs_ea.patch.

-- 
John Baldwin