From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
        id S1751108AbdALUje (ORCPT <rfc822;w@1wt.eu>);
        Thu, 12 Jan 2017 15:39:34 -0500
Received: from shards.monkeyblade.net ([184.105.139.130]:56980 "EHLO
        shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S1750922AbdALUjc (ORCPT
        <rfc822;linux-kernel@vger.kernel.org>);
        Thu, 12 Jan 2017 15:39:32 -0500
Date: Thu, 12 Jan 2017 15:39:29 -0500 (EST)
Message-Id: <20170112.153929.270105303415628819.davem@davemloft.net>
To: eric.dumazet@gmail.com
Cc: rob.gardner@oracle.com, shannon.nelson@oracle.com,
        netdev@vger.kernel.org, sparclinux@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH] tcp: fix tcp_fastopen unaligned access complaints on
 sparc
From: David Miller <davem@davemloft.net>
In-Reply-To: <1484252733.13165.1.camel@edumazet-glaptop3.roam.corp.google.com>
References: <1484252011.13165.0.camel@edumazet-glaptop3.roam.corp.google.com>
        <cd152a9f-4855-a552-2b45-bd619436a628@oracle.com>
        <1484252733.13165.1.camel@edumazet-glaptop3.roam.corp.google.com>
X-Mailer: Mew version 6.7 on Emacs 25.1 / Mule 6.0 (HANACHIRUSATO)
Mime-Version: 1.0
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Greylist: Sender succeeded SMTP AUTH, not delayed by milter-greylist-4.5.12 (shards.monkeyblade.net [149.20.54.216]); Thu, 12 Jan 2017 11:40:23 -0800 (PST)
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Thu, 12 Jan 2017 12:25:33 -0800

> On Thu, 2017-01-12 at 13:15 -0700, Rob Gardner wrote:
> 
>> 
>> I suspect that someplace, somebody is casting val to an int * or 
>> something like that.
> 
> Then that would be the bug. Can we root cause this please ?

The three accesses to foc->val are via function calls, at least when I
try to build it, one via memcmp(), one via memcpy() (for the structure
assignment at the end of the function) and one via a call into the
crypto layer when we do tcp_fastopen_cookie_gen).

So if the PC is inside of tcp_try_fastopen() it has to be something
else, or something specific to your gcc and build.