From mboxrd@z Thu Jan  1 00:00:00 1970
From: David Miller <davem@davemloft.net>
Subject: Re: [PATCH][net-next] net: slightly optimize eth_type_trans
Date: Sat, 17 Nov 2018 16:51:38 -0800 (PST)
Message-ID: <20181117.165138.801331179496411533.davem@davemloft.net>
References: <1542072871-21208-1-git-send-email-lirongqing@baidu.com>
        <20181115.151131.596213640354724307.davem@davemloft.net>
        <d5201107-08b3-8681-993d-2ba3480707e7@gmail.com>
Mime-Version: 1.0
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Cc: lirongqing@baidu.com, netdev@vger.kernel.org
To: eric.dumazet@gmail.com
Return-path: <netdev-owner@vger.kernel.org>
Received: from shards.monkeyblade.net ([23.128.96.9]:38448 "EHLO
        shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S1725734AbeKRLKK (ORCPT
        <rfc822;netdev@vger.kernel.org>); Sun, 18 Nov 2018 06:10:10 -0500
In-Reply-To: <d5201107-08b3-8681-993d-2ba3480707e7@gmail.com>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Sat, 17 Nov 2018 14:37:12 -0800

> 
> 
> On 11/15/2018 03:11 PM, David Miller wrote:
> 
>> 
>> Applied.
>> 
> 
> While reviewing this stuff, I found we have a bug.
> 
> If napi_reuse_skb() is called, we might inherit from prior skb->pkt_type value.
> 
> It seems that GRO could aggregate packets with pkt_type != PACKET_HOST, right ?
> 
> David, any objection if I submit the following fix ?

Oh weird, so we do GRO frags accumulation using SKB which never goes through
eth_type_trans()?

I don't understand how we can, in this circumstance, assume PACKET_HOST?

Because that is what your suggested patch does.

Frame could be UDP multicast, and we could legitimately GRO accumulate it.
In that situations setting PACKET_HOST doesn't seem correct.