Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
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] KVM: LAPIC: Fix advance timer expiration delta trace
Date:   Tue, 26 Apr 2022 01:25:51 -0700
Message-Id: <1650961551-38390-1-git-send-email-wanpengli@tencent.com>
Precedence: bulk

From: Wanpeng Li <wanpengli@tencent.com>

The advance timer expiration delta trace is moved after guest_exit_irqoff()
since it can avoid to violate RCU extended quiescent state when moving 
wait_lapic_expire() just before vmentry. Now kvm can enter a guest multiple 
times since IPI/Timer fastpath, snapshot the delta multiple times before 
vmentry and trace it one time after vmexit is not accurately this time. 
Commit 87fa7f3e98a1 (x86/kvm: Move context tracking where it belongs) moves 
rcu state updates more closer around vmentry/vmexit, we will not violate RCU 
any more. Let's fix it by keeping the advance timer expiration delta trace 
in __kvm_wait_lapic_epire() and dropping advance_expire_delta.

Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
---
 arch/x86/kvm/lapic.c | 4 ++--
 arch/x86/kvm/lapic.h | 1 -
 arch/x86/kvm/x86.c   | 8 --------
 3 files changed, 2 insertions(+), 11 deletions(-)

diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index 66b0eb0bda94..a4e9eb329e42 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -1648,10 +1648,10 @@ static void __kvm_wait_lapic_expire(struct kvm_vcpu *vcpu)
 	tsc_deadline = apic->lapic_timer.expired_tscdeadline;
 	apic->lapic_timer.expired_tscdeadline = 0;
 	guest_tsc = kvm_read_l1_tsc(vcpu, rdtsc());
-	apic->lapic_timer.advance_expire_delta = guest_tsc - tsc_deadline;
+	trace_kvm_wait_lapic_expire(vcpu->vcpu_id, guest_tsc - tsc_deadline);
 
 	if (lapic_timer_advance_dynamic) {
-		adjust_lapic_timer_advance(vcpu, apic->lapic_timer.advance_expire_delta);
+		adjust_lapic_timer_advance(vcpu, guest_tsc - tsc_deadline);
 		/*
 		 * If the timer fired early, reread the TSC to account for the
 		 * overhead of the above adjustment to avoid waiting longer
diff --git a/arch/x86/kvm/lapic.h b/arch/x86/kvm/lapic.h
index 4e4f8a22754f..65bb2a8cf145 100644
--- a/arch/x86/kvm/lapic.h
+++ b/arch/x86/kvm/lapic.h
@@ -38,7 +38,6 @@ struct kvm_timer {
 	u64 tscdeadline;
 	u64 expired_tscdeadline;
 	u32 timer_advance_ns;
-	s64 advance_expire_delta;
 	atomic_t pending;			/* accumulated triggered timers */
 	bool hv_timer_in_use;
 };
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index a6ab19afc638..5fa465b268d2 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -10248,14 +10248,6 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 	 */
 	guest_timing_exit_irqoff();
 
-	if (lapic_in_kernel(vcpu)) {
-		s64 delta = vcpu->arch.apic->lapic_timer.advance_expire_delta;
-		if (delta != S64_MIN) {
-			trace_kvm_wait_lapic_expire(vcpu->vcpu_id, delta);
-			vcpu->arch.apic->lapic_timer.advance_expire_delta = S64_MIN;
-		}
-	}
-
 	local_irq_enable();
 	preempt_enable();
 
-- 
2.25.1