$count = $order->where($con) ->join(left join x_goods on x_goods.goods_id = x_order.order_goodsid) ->order(order_createtime desc) ->group(order_no) ->count();// 查询满足要求的总记录数

无法获取数量
$count = $order->where($con) ->join(left join x_goods on x_goods.goods_id = x_order.order_goodsid) ->order(order_createtime desc) ->group(order_no) ->select();// 查询满足要求的总记录数

可以获取数据
回复内容： $count = $order->where($con) ->join(left join x_goods on x_goods.goods_id = x_order.order_goodsid) ->order(order_createtime desc) ->group(order_no) ->count();// 查询满足要求的总记录数

无法获取数量
$count = $order->where($con) ->join(left join x_goods on x_goods.goods_id = x_order.order_goodsid) ->order(order_createtime desc) ->group(order_no) ->select();// 查询满足要求的总记录数

可以获取数据
那就直接写成
$count = count($order->where($con) ->join(left join x_goods on x_goods.goods_id = x_order.order_goodsid) ->order(order_createtime desc) ->group(order_no) ->select());// 查询满足要求的总记录数
缺陷是大数据量下会影响效率，导致网络io和资源消耗过大。小业务的话就不必多虑。
我瞎猜的哈，你试试
$count = $order->where($con) ->join(left join x_goods on x_goods.goods_id = x_order.order_goodsid) ->order(order_createtime desc) ->group(order_no) ->field(count(*)) ->select();//
最好把执行的sql打印出来看下，比较复杂的逻辑用原生sql最好，没必要拘泥。