Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932247AbcCMD1b (ORCPT ); Sat, 12 Mar 2016 22:27:31 -0500 Received: from mga14.intel.com ([192.55.52.115]:44961 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753882AbcCMD0w (ORCPT ); Sat, 12 Mar 2016 22:26:52 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.24,328,1455004800"; d="scan'208";a="763027724" From: Lan Tianyu To: gleb@kernel.org, pbonzini@redhat.com, tglx@linutronix.de, mingo@redhat.com, hpa@zytor.com, x86@kernel.org, guangrong.xiao@linux.intel.com Cc: Lan Tianyu , kvm@vger.kernel.org, linux-kernel@vger.kernel.org Subject: [PATCH 1/7] KVM: Remove redundant smp_mb() in the kvm_mmu_commit_zap_page() Date: Sun, 13 Mar 2016 11:10:24 +0800 Message-Id: <1457838631-26775-2-git-send-email-tianyu.lan@intel.com> X-Mailer: git-send-email 1.7.9.5 In-Reply-To: <1457838631-26775-1-git-send-email-tianyu.lan@intel.com> References: <1457838631-26775-1-git-send-email-tianyu.lan@intel.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1303 Lines: 37 There is already a barrier inside of kvm_flush_remote_tlbs() which can help to make sure everyone sees our modifications to the page tables and see changes to vcpu->mode here. So remove the smp_mb in the kvm_mmu_commit_zap_page() and update the comment. Signed-off-by: Lan Tianyu --- arch/x86/kvm/mmu.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 2463de0..5e795af 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -2390,14 +2390,12 @@ static void kvm_mmu_commit_zap_page(struct kvm *kvm, return; /* - * wmb: make sure everyone sees our modifications to the page tables - * rmb: make sure we see changes to vcpu->mode - */ - smp_mb(); - - /* - * Wait for all vcpus to exit guest mode and/or lockless shadow - * page table walks. + * We need to make sure everyone sees our modifications to + * the page tables and see changes to vcpu->mode here. The barrier + * in the kvm_flush_remote_tlbs() helps us to achieve these. This pairs + * with vcpu_enter_guest and walk_shadow_page_lockless_begin/end. + * In addition, wait for all vcpus to exit guest mode and/or lockless + * shadow page table walks. */ kvm_flush_remote_tlbs(kvm); -- 1.8.4.rc0.1.g8f6a3e5.dirty