From mboxrd@z Thu Jan  1 00:00:00 1970
From: Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [PATCH] kvm: fix waitqueue_active without memory barrier in
 virt/kvm/async_pf.c
Date: Fri, 9 Oct 2015 12:26:47 +0200
Message-ID: <56179667.8010609@redhat.com>
References: <56177EAC.2070601@redhat.com>
 <17EC94B0A072C34B8DCF0D30AD16044A02874CB2@BPXM09GP.gisp.nec.co.jp>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-2022-jp
Content-Transfer-Encoding: 7bit
Cc: Gleb Natapov <gleb@kernel.org>,
	"kvm@vger.kernel.org" <kvm@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
To: Kosuke Tatsukawa <tatsu@ab.jp.nec.com>
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <17EC94B0A072C34B8DCF0D30AD16044A02874CB2@BPXM09GP.gisp.nec.co.jp>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: kvm.vger.kernel.org



On 09/10/2015 11:04, Kosuke Tatsukawa wrote:
> smp_store_mb() called from set_current_state(), which is called from
> prepare_to_wait() should prevent reordering such as below from
> happening.  wait_event*() also calls set_current_state() inside.

Ah, I missed that set_current_state has a memory barrier in it.  The
patch is okay, but please expand the comment to say that this memory
barrier pairs with prepare_to_wait's set_current_state().

Paolo