From mboxrd@z Thu Jan  1 00:00:00 1970
From: Eric Dumazet <eric.dumazet@gmail.com>
Subject: Re: [PATCH] net: disable UDP punt on sockets in RCV_SHUTDWON
Date: Fri, 4 May 2018 16:44:52 -0700
Message-ID: <7ef94995-9c5b-4f53-e1e3-e12e74817e9f@gmail.com>
References: <1525468117-61242-1-git-send-email-chintsha@cisco.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Cc: kamensky@cisco.com, takondra@cisco.com,
        xe-linux-external@cisco.com, enkechen@cisco.com
To: Chintan Shah <chintsha@cisco.com>, davem@davemloft.net,
        kuznet@ms2.inr.ac.ru, jmorris@namei.org, yoshfuji@linux-ipv6.org,
        kaber@trash.net, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <1525468117-61242-1-git-send-email-chintsha@cisco.com>
Content-Language: en-US
Sender: linux-kernel-owner@vger.kernel.org
List-Id: netdev.vger.kernel.org



On 05/04/2018 02:08 PM, Chintan Shah wrote:
> A UDP application which opens multiple sockets with same local
> address/port combination (using SO_REUSEPORT/SO_REUSEADDR socket options);
> and issues connect to a remote socket (using one of these local socket).
> Now if the same socket, which issued connect, issues shutdown (SHUT_RD);
> packets would still be queued to this socket (if sent from same remote
> client, which the local socket connected to), and not delivered to the
> other socket in the normal state.
> 

Confusing changelog.

sk_shutdown is on a different cache line, so this additional fetch would cause
loss of performance if many sockets are scanned in the hash bucket.

If you are trying to add full 4-tuple hash table to UDP, and accept() ability,
this would require a bit more than this hack...