When we automatically mount snapshot we want to return vnode of the mount point

from the lookup and not covered vnode. This is one of the fixes for using .zfs/
over NFS.

MFC after:	1 week

1 files changed, 19 insertions, 3 deletions

diff --git a/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c b/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c
index 72db1f0..e845a99 100644
--- a/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c
+++ b/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c
@@ -882,9 +882,25 @@ domount:
 	    dvp->v_vfsp->mnt_stat.f_mntonname, nm);
 	err = domount(curthread, *vpp, "zfs", mountpoint, snapname, 0);
 	kmem_free(mountpoint, mountpoint_len);
-	/* FreeBSD: This line was moved from below to avoid a lock recursion. */
-	if (err == 0)
-		vn_lock(*vpp, LK_EXCLUSIVE | LK_RETRY);
+	if (err == 0) {
+		vnode_t *mvp;
+
+		ASSERT((*vpp)->v_mountedhere != NULL);
+		err = VFS_ROOT((*vpp)->v_mountedhere, LK_EXCLUSIVE, &mvp);
+		ASSERT(err == 0);
+		VN_RELE(*vpp);
+		*vpp = mvp;
+
+		/*
+		 * Fix up the root vnode mounted on .zfs/snapshot/<snapname>.
+		 *
+		 * This is where we lie about our v_vfsp in order to
+		 * make .zfs/snapshot/<snapname> accessible over NFS
+		 * without requiring manual mounts of <snapname>.
+		 */
+		ASSERT(VTOZ(*vpp)->z_zfsvfs != zfsvfs);
+		VTOZ(*vpp)->z_zfsvfs->z_parent = zfsvfs;
+	}
 	mutex_exit(&sdp->sd_lock);
 	/*
 	 * If we had an error, drop our hold on the vnode and