[v1] KVM: x86: Use fast path for Xen timer delivery

[PATCH] KVM: x86: Use fast path for Xen timer delivery

Posted by David Woodhouse 2 years, 4 months ago

From: David Woodhouse <dwmw@amazon.co.uk>

Most of the time there's no need to kick the vCPU and deliver the timer
event through kvm_xen_inject_timer_irqs(). Use kvm_xen_set_evtchn_fast()
directly from the timer callback, and only fall back to the slow path
when it's necessary to do so.

This gives a significant improvement in timer latency testing (using
nanosleep() for various periods and then measuring the actual time
elapsed).

Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
---
 arch/x86/kvm/xen.c | 17 ++++++++++++++---
 1 file changed, 14 insertions(+), 3 deletions(-)

diff --git a/arch/x86/kvm/xen.c b/arch/x86/kvm/xen.c
index 40edf4d1974c..66c4cf93a55c 100644
--- a/arch/x86/kvm/xen.c
+++ b/arch/x86/kvm/xen.c
@@ -134,12 +134,23 @@ static enum hrtimer_restart xen_timer_callback(struct hrtimer *timer)
 {
 	struct kvm_vcpu *vcpu = container_of(timer, struct kvm_vcpu,
 					     arch.xen.timer);
+	struct kvm_xen_evtchn e;
+	int rc;
+
 	if (atomic_read(&vcpu->arch.xen.timer_pending))
 		return HRTIMER_NORESTART;
 
-	atomic_inc(&vcpu->arch.xen.timer_pending);
-	kvm_make_request(KVM_REQ_UNBLOCK, vcpu);
-	kvm_vcpu_kick(vcpu);
+	e.vcpu_id = vcpu->vcpu_id;
+	e.vcpu_idx = vcpu->vcpu_idx;
+	e.port = vcpu->arch.xen.timer_virq;
+	e.priority = KVM_IRQ_ROUTING_XEN_EVTCHN_PRIO_2LEVEL;
+
+	rc = kvm_xen_set_evtchn_fast(&e, vcpu->kvm);
+	if (rc == -EWOULDBLOCK) {
+		atomic_inc(&vcpu->arch.xen.timer_pending);
+		kvm_make_request(KVM_REQ_UNBLOCK, vcpu);
+		kvm_vcpu_kick(vcpu);
+	}
 
 	return HRTIMER_NORESTART;
 }
-- 
2.34.1

Re: [PATCH] KVM: x86: Use fast path for Xen timer delivery

Posted by David Woodhouse 2 years, 4 months ago

On Fri, 2023-09-22 at 10:20 +0100, David Woodhouse wrote:
> From: David Woodhouse <dwmw@amazon.co.uk>
> 
> Most of the time there's no need to kick the vCPU and deliver the timer
> event through kvm_xen_inject_timer_irqs(). Use kvm_xen_set_evtchn_fast()
> directly from the timer callback, and only fall back to the slow path
> when it's necessary to do so.
> 
> This gives a significant improvement in timer latency testing (using
> nanosleep() for various periods and then measuring the actual time
> elapsed).
> 
> Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>

Hm, scratch that. It brings back the race condition Paolo talks about in
https://lore.kernel.org/kvm/0709ac62f664c0f3123fcdeabed3b79038cef3b6.camel@infradead.org/T/

Well, not *quite*... there's no race with clearing timer_expires
because I forgot to clear it at all :)

I think I'll be able to fix it up with an hrtimer_cancel() in the rare
ioctl attr_get path to avoid the race.

Re: [PATCH] KVM: x86: Use fast path for Xen timer delivery

Posted by Paul Durrant 2 years, 4 months ago

On 22/09/2023 10:20, David Woodhouse wrote:
> From: David Woodhouse <dwmw@amazon.co.uk>
> 
> Most of the time there's no need to kick the vCPU and deliver the timer
> event through kvm_xen_inject_timer_irqs(). Use kvm_xen_set_evtchn_fast()
> directly from the timer callback, and only fall back to the slow path
> when it's necessary to do so.
> 
> This gives a significant improvement in timer latency testing (using
> nanosleep() for various periods and then measuring the actual time
> elapsed).
> 
> Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
> ---
>   arch/x86/kvm/xen.c | 17 ++++++++++++++---
>   1 file changed, 14 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kvm/xen.c b/arch/x86/kvm/xen.c
> index 40edf4d1974c..66c4cf93a55c 100644
> --- a/arch/x86/kvm/xen.c
> +++ b/arch/x86/kvm/xen.c
> @@ -134,12 +134,23 @@ static enum hrtimer_restart xen_timer_callback(struct hrtimer *timer)
>   {
>   	struct kvm_vcpu *vcpu = container_of(timer, struct kvm_vcpu,
>   					     arch.xen.timer);
> +	struct kvm_xen_evtchn e;
> +	int rc;
> +
>   	if (atomic_read(&vcpu->arch.xen.timer_pending))
>   		return HRTIMER_NORESTART;
>   
> -	atomic_inc(&vcpu->arch.xen.timer_pending);
> -	kvm_make_request(KVM_REQ_UNBLOCK, vcpu);
> -	kvm_vcpu_kick(vcpu);
> +	e.vcpu_id = vcpu->vcpu_id;
> +	e.vcpu_idx = vcpu->vcpu_idx;
> +	e.port = vcpu->arch.xen.timer_virq;

Don't you need to check the port for validity? The VMM may not have set 
it (and hence be delivering timer VIRQ events itself).

   Paul

> +	e.priority = KVM_IRQ_ROUTING_XEN_EVTCHN_PRIO_2LEVEL;
> +
> +	rc = kvm_xen_set_evtchn_fast(&e, vcpu->kvm);
> +	if (rc == -EWOULDBLOCK) {
> +		atomic_inc(&vcpu->arch.xen.timer_pending);
> +		kvm_make_request(KVM_REQ_UNBLOCK, vcpu);
> +		kvm_vcpu_kick(vcpu);
> +	}
>   
>   	return HRTIMER_NORESTART;
>   }

Re: [PATCH] KVM: x86: Use fast path for Xen timer delivery

Posted by David Woodhouse 2 years, 4 months ago

On Fri, 2023-09-22 at 10:22 +0100, Paul Durrant wrote:
> On 22/09/2023 10:20, David Woodhouse wrote:
> > From: David Woodhouse <dwmw@amazon.co.uk>
> > 
> > Most of the time there's no need to kick the vCPU and deliver the timer
> > event through kvm_xen_inject_timer_irqs(). Use kvm_xen_set_evtchn_fast()
> > directly from the timer callback, and only fall back to the slow path
> > when it's necessary to do so.
> > 
> > This gives a significant improvement in timer latency testing (using
> > nanosleep() for various periods and then measuring the actual time
> > elapsed).
> > 
> > Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
> > ---
> >   arch/x86/kvm/xen.c | 17 ++++++++++++++---
> >   1 file changed, 14 insertions(+), 3 deletions(-)
> > 
> > diff --git a/arch/x86/kvm/xen.c b/arch/x86/kvm/xen.c
> > index 40edf4d1974c..66c4cf93a55c 100644
> > --- a/arch/x86/kvm/xen.c
> > +++ b/arch/x86/kvm/xen.c
> > @@ -134,12 +134,23 @@ static enum hrtimer_restart xen_timer_callback(struct hrtimer *timer)
> >   {
> >         struct kvm_vcpu *vcpu = container_of(timer, struct kvm_vcpu,
> >                                              arch.xen.timer);
> > +       struct kvm_xen_evtchn e;
> > +       int rc;
> > +
> >         if (atomic_read(&vcpu->arch.xen.timer_pending))
> >                 return HRTIMER_NORESTART;
> >   
> > -       atomic_inc(&vcpu->arch.xen.timer_pending);
> > -       kvm_make_request(KVM_REQ_UNBLOCK, vcpu);
> > -       kvm_vcpu_kick(vcpu);
> > +       e.vcpu_id = vcpu->vcpu_id;
> > +       e.vcpu_idx = vcpu->vcpu_idx;
> > +       e.port = vcpu->arch.xen.timer_virq;
> 
> Don't you need to check the port for validity? The VMM may not have set 
> it (and hence be delivering timer VIRQ events itself).

Nah, the kvm_xen_timer_enabled() check which gates all the kernel
interception of timer hypercalls is already testing precisely that.

If the TIMER_VIRQ isn't set, none of this code runs. And when userspace
tears down TIMER_VIRQ, the hrtimer is cancelled. So none of this code
(then) runs.

Re: [PATCH] KVM: x86: Use fast path for Xen timer delivery

Posted by Paul Durrant 2 years, 4 months ago

On 22/09/2023 10:29, David Woodhouse wrote:
> On Fri, 2023-09-22 at 10:22 +0100, Paul Durrant wrote:
>> On 22/09/2023 10:20, David Woodhouse wrote:
>>> From: David Woodhouse <dwmw@amazon.co.uk>
>>>
>>> Most of the time there's no need to kick the vCPU and deliver the timer
>>> event through kvm_xen_inject_timer_irqs(). Use kvm_xen_set_evtchn_fast()
>>> directly from the timer callback, and only fall back to the slow path
>>> when it's necessary to do so.
>>>
>>> This gives a significant improvement in timer latency testing (using
>>> nanosleep() for various periods and then measuring the actual time
>>> elapsed).
>>>
>>> Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
>>> ---
>>>    arch/x86/kvm/xen.c | 17 ++++++++++++++---
>>>    1 file changed, 14 insertions(+), 3 deletions(-)
>>>
>>> diff --git a/arch/x86/kvm/xen.c b/arch/x86/kvm/xen.c
>>> index 40edf4d1974c..66c4cf93a55c 100644
>>> --- a/arch/x86/kvm/xen.c
>>> +++ b/arch/x86/kvm/xen.c
>>> @@ -134,12 +134,23 @@ static enum hrtimer_restart xen_timer_callback(struct hrtimer *timer)
>>>    {
>>>          struct kvm_vcpu *vcpu = container_of(timer, struct kvm_vcpu,
>>>                                               arch.xen.timer);
>>> +       struct kvm_xen_evtchn e;
>>> +       int rc;
>>> +
>>>          if (atomic_read(&vcpu->arch.xen.timer_pending))
>>>                  return HRTIMER_NORESTART;
>>>    
>>> -       atomic_inc(&vcpu->arch.xen.timer_pending);
>>> -       kvm_make_request(KVM_REQ_UNBLOCK, vcpu);
>>> -       kvm_vcpu_kick(vcpu);
>>> +       e.vcpu_id = vcpu->vcpu_id;
>>> +       e.vcpu_idx = vcpu->vcpu_idx;
>>> +       e.port = vcpu->arch.xen.timer_virq;
>>
>> Don't you need to check the port for validity? The VMM may not have set
>> it (and hence be delivering timer VIRQ events itself).
> 
> Nah, the kvm_xen_timer_enabled() check which gates all the kernel
> interception of timer hypercalls is already testing precisely that.
> 
> If the TIMER_VIRQ isn't set, none of this code runs. And when userspace
> tears down TIMER_VIRQ, the hrtimer is cancelled. So none of this code
> (then) runs.

Ok. Good :-)

Reviewed-by: Paul Durrant <paul@xen.org>