*
  * this lock must be held if modifying the page group list
  *
- * returns result from wait_on_bit_lock: 0 on success, < 0 on error
+ * return 0 on success, < 0 on error: -EDELAY if nonblocking or the
+ * result from wait_on_bit_lock
+ *
+ * NOTE: calling with nonblock=false should always have set the
+ *       lock bit (see fs/buffer.c and other uses of wait_on_bit_lock
+ *       with TASK_UNINTERRUPTIBLE), so there is no need to check the result.
  */
 int
 nfs_page_group_lock(struct nfs_page *req, bool nonblock)
 {
        struct nfs_page *head = req->wb_head;
-       int ret;
 
        WARN_ON_ONCE(head != head->wb_head);
 
-       do {
-               ret = wait_on_bit_lock(&head->wb_flags, PG_HEADLOCK,
-                       TASK_UNINTERRUPTIBLE);
-       } while (!nonblock && ret != 0);
+       if (!test_and_set_bit(PG_HEADLOCK, &head->wb_flags))
+               return 0;
 
-       WARN_ON_ONCE(ret > 0);
-       return ret;
+       if (!nonblock)
+               return wait_on_bit_lock(&head->wb_flags, PG_HEADLOCK,
+                               TASK_UNINTERRUPTIBLE);
+
+       return -EAGAIN;
 }
 
 /*