Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Wanpeng Li <kernellwp@gmail.com>
To:     linux-kernel@vger.kernel.org, kvm@vger.kernel.org
Cc:     Paolo Bonzini <pbonzini@redhat.com>,
        Sean Christopherson <seanjc@google.com>,
        Vitaly Kuznetsov <vkuznets@redhat.com>,
        Wanpeng Li <wanpengli@tencent.com>,
        Jim Mattson <jmattson@google.com>,
        Joerg Roedel <joro@8bytes.org>
Subject: [PATCH v4 5/5] KVM: LAPIC: Narrow the timer latency between wait_lapic_expire and world switch
Date:   Tue, 18 May 2021 05:00:35 -0700
Message-Id: <1621339235-11131-5-git-send-email-wanpengli@tencent.com>
In-Reply-To: <1621339235-11131-1-git-send-email-wanpengli@tencent.com>
References: <1621339235-11131-1-git-send-email-wanpengli@tencent.com>
Precedence: bulk

From: Wanpeng Li <wanpengli@tencent.com>

Let's treat lapic_timer_advance_ns automatic tuning logic as hypervisor
overhead, move it before wait_lapic_expire instead of between wait_lapic_expire 
and the world switch, the wait duration should be calculated by the 
up-to-date guest_tsc after the overhead of automatic tuning logic. This 
patch reduces ~30+ cycles for kvm-unit-tests/tscdeadline-latency when testing 
busy waits.

Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
---
v3 -> v4:
 * consider automatic tuning is disabled and timer did not arrive early
 * add a code comment

 arch/x86/kvm/lapic.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index c0ebef560bd1..5d91f2367c31 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -1598,11 +1598,19 @@ static void __kvm_wait_lapic_expire(struct kvm_vcpu *vcpu)
 	guest_tsc = kvm_read_l1_tsc(vcpu, rdtsc());
 	apic->lapic_timer.advance_expire_delta = guest_tsc - tsc_deadline;
 
+	if (lapic_timer_advance_dynamic) {
+		adjust_lapic_timer_advance(vcpu, apic->lapic_timer.advance_expire_delta);
+		/*
+		 * If the timer fired early, reread the TSC to account for the
+		 * overhead of the above adjustment to avoid waiting longer
+		 * than is necessary.
+		 */
+		if (guest_tsc < tsc_deadline)
+			guest_tsc = kvm_read_l1_tsc(vcpu, rdtsc());
+	}
+
 	if (guest_tsc < tsc_deadline)
 		__wait_lapic_expire(vcpu, tsc_deadline - guest_tsc);
-
-	if (lapic_timer_advance_dynamic)
-		adjust_lapic_timer_advance(vcpu, apic->lapic_timer.advance_expire_delta);
 }
 
 void kvm_wait_lapic_expire(struct kvm_vcpu *vcpu)
-- 
2.25.1