+ /* The createtime can be changed by hfs_setattr or hfs_setattrlist.
+ * For NFS, we are assuming that only if the createtime was moved
+ * forward would it mean the fileID got reused in that session by
+ * wrapping. We don't have a volume ID or other unique identifier to
+ * to use here for a generation ID across reboots, crashes where
+ * metadata noting lastFileID didn't make it to disk but client has
+ * it, or volume erasures where fileIDs start over again. Lastly,
+ * with HFS allowing "wraps" of fileIDs now, this becomes more
+ * error prone. Future, would be change the "wrap bit" to a unique
+ * wrap number and use that for generation number. For now do this.
+ */
+ if ((hfsfhp->hfsfid_gen < VTOH(nvp)->h_meta->h_crtime)) {