From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Michael S. Tsirkin" Subject: [PATCH 2/2] kvm: optimize out smp_mb after srcu_read_unlock Date: Mon, 4 Nov 2013 22:36:25 +0200 Message-ID: <1383592884-2535-2-git-send-email-mst@redhat.com> References: <1383592884-2535-1-git-send-email-mst@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Gleb Natapov , Paolo Bonzini , kvm@vger.kernel.org To: linux-kernel@vger.kernel.org Return-path: Content-Disposition: inline In-Reply-To: <1383592884-2535-1-git-send-email-mst@redhat.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: kvm.vger.kernel.org I noticed that srcu_read_lock/unlock both have a memory barrier, so just by moving srcu_read_unlock earlier we can get rid of one call to smp_mb() using smp_mb__after_srcu_read_unlock instead. Unsurprisingly, the gain is small but measureable using the unit test microbenchmark: before vmcall in the ballpark of 1410 cycles after vmcall in the ballpark of 1360 cycles Signed-off-by: Michael S. Tsirkin --- arch/x86/kvm/x86.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index e5ca72a..d609dce 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -5945,10 +5945,12 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) vcpu->mode = IN_GUEST_MODE; + srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx); + /* We should set ->mode before check ->requests, * see the comment in make_all_cpus_request. */ - smp_mb(); + smp_mb__after_srcu_read_unlock(); local_irq_disable(); @@ -5958,12 +5960,11 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) smp_wmb(); local_irq_enable(); preempt_enable(); + vcpu->srcu_idx = srcu_read_lock(&vcpu->kvm->srcu); r = 1; goto cancel_injection; } - srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx); - if (req_immediate_exit) smp_send_reschedule(vcpu->cpu); -- MST