Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Thu, 6 Sep 2018 21:28:52 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     Dave Hansen <dave.hansen@intel.com>
Cc:     Nadav Amit <namit@vmware.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        linux-kernel@vger.kernel.org, Ingo Molnar <mingo@redhat.com>,
        x86@kernel.org, Dave Hansen <dave.hansen@linux.intel.com>,
        Nadav Amit <nadav.amit@gmail.com>,
        Andi Kleen <ak@linux.intel.com>,
        Josh Poimboeuf <jpoimboe@redhat.com>,
        Michal Hocko <mhocko@suse.com>,
        Vlastimil Babka <vbabka@suse.cz>,
        Sean Christopherson <sean.j.christopherson@intel.com>,
        Andy Lutomirski <luto@kernel.org>
Subject: Re: [PATCH] x86: use WRITE_ONCE() when setting PTEs
Message-ID: <20180906192852.GC4816@worktop.programming.kicks-ass.net>
References: <20180902181451.80520-1-namit@vmware.com>
 <8aef58c7-c27c-5df8-69b3-b188dd5f64e8@intel.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <8aef58c7-c27c-5df8-69b3-b188dd5f64e8@intel.com>
User-Agent: Mutt/1.5.22.1 (2013-10-16)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, Sep 06, 2018 at 10:21:38AM -0700, Dave Hansen wrote:
> There's probably a massive number of things that would break if we
> assumed sane 64-bit writes can be observed piecemeal.

Do assume, it has been observed. WRITE_ONCE()/READ_ONCE() are _required_
if you cannot have load/store tearing.