The patch below does not apply to the 4.14-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to stable@vger.kernel.org.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From ecf08dad723d3e000aecff6c396f54772d124733 Mon Sep 17 00:00:00 2001
From: Anthoine Bourgeois anthoine.bourgeois@blade-group.com Date: Sun, 29 Apr 2018 22:05:58 +0000 Subject: [PATCH] KVM: x86: remove APIC Timer periodic/oneshot spikes MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit
Since the commit "8003c9ae204e: add APIC Timer periodic/oneshot mode VMX preemption timer support", a Windows 10 guest has some erratic timer spikes.
Here the results on a 150000 times 1ms timer without any load: Before 8003c9ae204e | After 8003c9ae204e Max 1834us | 86000us Mean 1100us | 1021us Deviation 59us | 149us Here the results on a 150000 times 1ms timer with a cpu-z stress test: Before 8003c9ae204e | After 8003c9ae204e Max 32000us | 140000us Mean 1006us | 1997us Deviation 140us | 11095us
The root cause of the problem is starting hrtimer with an expiry time already in the past can take more than 20 milliseconds to trigger the timer function. It can be solved by forward such past timers immediately, rather than submitting them to hrtimer_start(). In case the timer is periodic, update the target expiration and call hrtimer_start with it.
v2: Check if the tsc deadline is already expired. Thank you Mika. v3: Execute the past timers immediately rather than submitting them to hrtimer_start(). v4: Rearm the periodic timer with advance_periodic_target_expiration() a simpler version of set_target_expiration(). Thank you Paolo.
Cc: Mika Penttilä mika.penttila@nextfour.com Cc: Wanpeng Li kernellwp@gmail.com Cc: Paolo Bonzini pbonzini@redhat.com Cc: stable@vger.kernel.org Signed-off-by: Anthoine Bourgeois anthoine.bourgeois@blade-group.com 8003c9ae204e ("KVM: LAPIC: add APIC Timer periodic/oneshot mode VMX preemption timer support") Signed-off-by: Radim Krčmář rkrcmar@redhat.com
diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c index 70dcb5548022..b74c9c1405b9 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -1463,23 +1463,6 @@ static void start_sw_tscdeadline(struct kvm_lapic *apic) local_irq_restore(flags); }
-static void start_sw_period(struct kvm_lapic *apic) -{ - if (!apic->lapic_timer.period) - return; - - if (apic_lvtt_oneshot(apic) && - ktime_after(ktime_get(), - apic->lapic_timer.target_expiration)) { - apic_timer_expired(apic); - return; - } - - hrtimer_start(&apic->lapic_timer.timer, - apic->lapic_timer.target_expiration, - HRTIMER_MODE_ABS_PINNED); -} - static void update_target_expiration(struct kvm_lapic *apic, uint32_t old_divisor) { ktime_t now, remaining; @@ -1546,6 +1529,26 @@ static void advance_periodic_target_expiration(struct kvm_lapic *apic) apic->lapic_timer.period); }
+static void start_sw_period(struct kvm_lapic *apic) +{ + if (!apic->lapic_timer.period) + return; + + if (ktime_after(ktime_get(), + apic->lapic_timer.target_expiration)) { + apic_timer_expired(apic); + + if (apic_lvtt_oneshot(apic)) + return; + + advance_periodic_target_expiration(apic); + } + + hrtimer_start(&apic->lapic_timer.timer, + apic->lapic_timer.target_expiration, + HRTIMER_MODE_ABS_PINNED); +} + bool kvm_lapic_hv_timer_in_use(struct kvm_vcpu *vcpu) { if (!lapic_in_kernel(vcpu))
commit ecf08dad723d3e000aecff6c396f54772d124733 upstream.
Since the commit "8003c9ae204e: add APIC Timer periodic/oneshot mode VMX preemption timer support", a Windows 10 guest has some erratic timer spikes.
Here the results on a 150000 times 1ms timer without any load: Before 8003c9ae204e | After 8003c9ae204e Max 1834us | 86000us Mean 1100us | 1021us Deviation 59us | 149us Here the results on a 150000 times 1ms timer with a cpu-z stress test: Before 8003c9ae204e | After 8003c9ae204e Max 32000us | 140000us Mean 1006us | 1997us Deviation 140us | 11095us
The root cause of the problem is starting hrtimer with an expiry time already in the past can take more than 20 milliseconds to trigger the timer function. It can be solved by forward such past timers immediately, rather than submitting them to hrtimer_start(). In case the timer is periodic, update the target expiration and call hrtimer_start with it.
v2: Check if the tsc deadline is already expired. Thank you Mika. v3: Execute the past timers immediately rather than submitting them to hrtimer_start(). v4: Rearm the periodic timer with advance_periodic_target_expiration() a simpler version of set_target_expiration(). Thank you Paolo.
Cc: Mika Penttilä mika.penttila@nextfour.com Cc: Wanpeng Li kernellwp@gmail.com Cc: Paolo Bonzini pbonzini@redhat.com Cc: stable@vger.kernel.org # 4.14.x Signed-off-by: Anthoine Bourgeois anthoine.bourgeois@blade-group.com 8003c9ae204e ("KVM: LAPIC: add APIC Timer periodic/oneshot mode VMX preemption timer support") Signed-off-by: Radim Krčmář rkrcmar@redhat.com --- arch/x86/kvm/lapic.c | 37 ++++++++++++++++++++----------------- 1 file changed, 20 insertions(+), 17 deletions(-)
diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c index 8cfdb6484fd0..ab8993fe58cc 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -1418,23 +1418,6 @@ static void start_sw_tscdeadline(struct kvm_lapic *apic) local_irq_restore(flags); }
-static void start_sw_period(struct kvm_lapic *apic) -{ - if (!apic->lapic_timer.period) - return; - - if (apic_lvtt_oneshot(apic) && - ktime_after(ktime_get(), - apic->lapic_timer.target_expiration)) { - apic_timer_expired(apic); - return; - } - - hrtimer_start(&apic->lapic_timer.timer, - apic->lapic_timer.target_expiration, - HRTIMER_MODE_ABS_PINNED); -} - static bool set_target_expiration(struct kvm_lapic *apic) { ktime_t now; @@ -1491,6 +1474,26 @@ static void advance_periodic_target_expiration(struct kvm_lapic *apic) apic->lapic_timer.period); }
+static void start_sw_period(struct kvm_lapic *apic) +{ + if (!apic->lapic_timer.period) + return; + + if (ktime_after(ktime_get(), + apic->lapic_timer.target_expiration)) { + apic_timer_expired(apic); + + if (apic_lvtt_oneshot(apic)) + return; + + advance_periodic_target_expiration(apic); + } + + hrtimer_start(&apic->lapic_timer.timer, + apic->lapic_timer.target_expiration, + HRTIMER_MODE_ABS_PINNED); +} + bool kvm_lapic_hv_timer_in_use(struct kvm_vcpu *vcpu) { if (!lapic_in_kernel(vcpu))
On Tue, May 08, 2018 at 08:17:14AM +0000, Anthoine Bourgeois wrote:
commit ecf08dad723d3e000aecff6c396f54772d124733 upstream.
Since the commit "8003c9ae204e: add APIC Timer periodic/oneshot mode VMX preemption timer support", a Windows 10 guest has some erratic timer spikes.
Here the results on a 150000 times 1ms timer without any load: Before 8003c9ae204e | After 8003c9ae204e Max 1834us | 86000us Mean 1100us | 1021us Deviation 59us | 149us Here the results on a 150000 times 1ms timer with a cpu-z stress test: Before 8003c9ae204e | After 8003c9ae204e Max 32000us | 140000us Mean 1006us | 1997us Deviation 140us | 11095us
The root cause of the problem is starting hrtimer with an expiry time already in the past can take more than 20 milliseconds to trigger the timer function. It can be solved by forward such past timers immediately, rather than submitting them to hrtimer_start(). In case the timer is periodic, update the target expiration and call hrtimer_start with it.
v2: Check if the tsc deadline is already expired. Thank you Mika. v3: Execute the past timers immediately rather than submitting them to hrtimer_start(). v4: Rearm the periodic timer with advance_periodic_target_expiration() a simpler version of set_target_expiration(). Thank you Paolo.
Cc: Mika Penttilä mika.penttila@nextfour.com Cc: Wanpeng Li kernellwp@gmail.com Cc: Paolo Bonzini pbonzini@redhat.com Cc: stable@vger.kernel.org # 4.14.x Signed-off-by: Anthoine Bourgeois anthoine.bourgeois@blade-group.com 8003c9ae204e ("KVM: LAPIC: add APIC Timer periodic/oneshot mode VMX preemption timer support") Signed-off-by: Radim Krčmář rkrcmar@redhat.com
arch/x86/kvm/lapic.c | 37 ++++++++++++++++++++----------------- 1 file changed, 20 insertions(+), 17 deletions(-)
Now applied, thanks.
greg k-h
linux-stable-mirror@lists.linaro.org