[PATCH v7 0/2] KVM: LAPIC: Implement Exitless Timer

[PATCH v7 0/2] KVM: LAPIC: Implement Exitless Timer

[Wanpeng Li]

Dedicated instances are currently disturbed by unnecessary jitter due 
to the emulated lapic timers fire on the same pCPUs which vCPUs resident.
There is no hardware virtual timer on Intel for guest like ARM. Both 
programming timer in guest and the emulated timer fires incur vmexits.
This patchset tries to avoid vmexit which is incurred by the emulated 
timer fires in dedicated instance scenario. 

When nohz_full is enabled in dedicated instances scenario, the unpinned 
timer will be moved to the nearest busy housekeepers after commit
9642d18eee2cd (nohz: Affine unpinned timers to housekeepers) and commit 
444969223c8 ("sched/nohz: Fix affine unpinned timers mess"). However, 
KVM always makes lapic timer pinned to the pCPU which vCPU residents, the 
reason is explained by commit 61abdbe0 (kvm: x86: make lapic hrtimer 
pinned). Actually, these emulated timers can be offload to the housekeeping 
cpus since APICv is really common in recent years. The guest timer interrupt 
is injected by posted-interrupt which is delivered by housekeeping cpu 
once the emulated timer fires. 

The host admin should fine tuned, e.g. dedicated instances scenario w/ 
nohz_full cover the pCPUs which vCPUs resident, several pCPUs surplus 
for busy housekeeping, disable mwait/hlt/pause vmexits to keep in non-root  
mode, ~3% redis performance benefit can be observed on Skylake server.

w/o patchset:

            VM-EXIT  Samples  Samples%  Time%   Min Time  Max Time   Avg time

EXTERNAL_INTERRUPT    42916    49.43%   39.30%   0.47us   106.09us   0.71us ( +-   1.09% )

w/ patchset:

            VM-EXIT  Samples  Samples%  Time%   Min Time  Max Time         Avg time

EXTERNAL_INTERRUPT    6871     9.29%     2.96%   0.44us    57.88us   0.72us ( +-   4.02% )

Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>

v6 -> v7:
 * remove bool argument

v5 -> v6:
 * don't overwrites whatever the user specified
 * introduce kvm_can_post_timer_interrupt and kvm_use_posted_timer_interrupt
 * remove kvm_hlt_in_guest() condition
 * squash all of 2/3/4 together

v4 -> v5:
 * update patch description in patch 1/4
 * feed latest apic->lapic_timer.expired_tscdeadline to kvm_wait_lapic_expire()
 * squash advance timer handling to patch 2/4

v3 -> v4:
 * drop the HRTIMER_MODE_ABS_PINNED, add kick after set pending timer
 * don't posted inject already-expired timer

v2 -> v3:
 * disarming the vmx preemption timer when posted_interrupt_inject_timer_enabled()
 * check kvm_hlt_in_guest instead

v1 -> v2:
 * check vcpu_halt_in_guest
 * move module parameter from kvm-intel to kvm
 * add housekeeping_enabled
 * rename apic_timer_expired_pi to kvm_apic_inject_pending_timer_irqs


Wanpeng Li (2):
  KVM: LAPIC: Make lapic timer unpinned
  KVM: LAPIC: Inject timer interrupt via posted interrupt

 arch/x86/kvm/lapic.c            | 109 ++++++++++++++++++++++++++--------------
 arch/x86/kvm/lapic.h            |   1 +
 arch/x86/kvm/vmx/vmx.c          |   3 +-
 arch/x86/kvm/x86.c              |  12 +++--
 arch/x86/kvm/x86.h              |   2 +
 include/linux/sched/isolation.h |   2 +
 kernel/sched/isolation.c        |   6 +++
 7 files changed, 90 insertions(+), 45 deletions(-)

-- 
1.8.3.1

[PATCH v7 1/2] KVM: LAPIC: Make lapic timer unpinned

[Wanpeng Li]

From: Wanpeng Li <wanpengli@tencent.com>

Commit 61abdbe0bcc2 ("kvm: x86: make lapic hrtimer pinned") pinned the
lapic timer to avoid to wait until the next kvm exit for the guest to
see KVM_REQ_PENDING_TIMER set. There is another solution to give a kick
after setting the KVM_REQ_PENDING_TIMER bit, make lapic timer unpinned
will be used in follow up patches.

Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
---
 arch/x86/kvm/lapic.c | 8 ++++----
 arch/x86/kvm/x86.c   | 6 +-----
 2 files changed, 5 insertions(+), 9 deletions(-)

diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index 459d1ee..707ca9c 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -1582,7 +1582,7 @@ static void start_sw_tscdeadline(struct kvm_lapic *apic)
 	    likely(ns > apic->lapic_timer.timer_advance_ns)) {
 		expire = ktime_add_ns(now, ns);
 		expire = ktime_sub_ns(expire, ktimer->timer_advance_ns);
-		hrtimer_start(&ktimer->timer, expire, HRTIMER_MODE_ABS_PINNED);
+		hrtimer_start(&ktimer->timer, expire, HRTIMER_MODE_ABS);
 	} else
 		apic_timer_expired(apic);
 
@@ -1684,7 +1684,7 @@ static void start_sw_period(struct kvm_lapic *apic)
 
 	hrtimer_start(&apic->lapic_timer.timer,
 		apic->lapic_timer.target_expiration,
-		HRTIMER_MODE_ABS_PINNED);
+		HRTIMER_MODE_ABS);
 }
 
 bool kvm_lapic_hv_timer_in_use(struct kvm_vcpu *vcpu)
@@ -2321,7 +2321,7 @@ int kvm_create_lapic(struct kvm_vcpu *vcpu, int timer_advance_ns)
 	apic->vcpu = vcpu;
 
 	hrtimer_init(&apic->lapic_timer.timer, CLOCK_MONOTONIC,
-		     HRTIMER_MODE_ABS_PINNED);
+		     HRTIMER_MODE_ABS);
 	apic->lapic_timer.timer.function = apic_timer_fn;
 	if (timer_advance_ns == -1) {
 		apic->lapic_timer.timer_advance_ns = LAPIC_TIMER_ADVANCE_ADJUST_INIT;
@@ -2510,7 +2510,7 @@ void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu)
 
 	timer = &vcpu->arch.apic->lapic_timer.timer;
 	if (hrtimer_cancel(timer))
-		hrtimer_start_expires(timer, HRTIMER_MODE_ABS_PINNED);
+		hrtimer_start_expires(timer, HRTIMER_MODE_ABS);
 }
 
 /*
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 3a7cd935..e199ac7 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -1437,12 +1437,8 @@ static void update_pvclock_gtod(struct timekeeper *tk)
 
 void kvm_set_pending_timer(struct kvm_vcpu *vcpu)
 {
-	/*
-	 * Note: KVM_REQ_PENDING_TIMER is implicitly checked in
-	 * vcpu_enter_guest.  This function is only called from
-	 * the physical CPU that is running vcpu.
-	 */
 	kvm_make_request(KVM_REQ_PENDING_TIMER, vcpu);
+	kvm_vcpu_kick(vcpu);
 }
 
 static void kvm_write_wall_clock(struct kvm *kvm, gpa_t wall_clock)
-- 
1.8.3.1

Re: [PATCH v7 1/2] KVM: LAPIC: Make lapic timer unpinned

[David Woodhouse]

[-- Attachment #1: Type: text/plain, Size: 1324 bytes --]

On Sat, 2019-07-06 at 09:26 +0800, Wanpeng Li wrote:
> From: Wanpeng Li <wanpengli@tencent.com>
> 
> Commit 61abdbe0bcc2 ("kvm: x86: make lapic hrtimer pinned") pinned the
> lapic timer to avoid to wait until the next kvm exit for the guest to
> see KVM_REQ_PENDING_TIMER set. There is another solution to give a kick
> after setting the KVM_REQ_PENDING_TIMER bit, make lapic timer unpinned
> will be used in follow up patches.
> 
> Cc: Paolo Bonzini <pbonzini@redhat.com>
> Cc: Radim Krčmář <rkrcmar@redhat.com>
> Cc: Marcelo Tosatti <mtosatti@redhat.com>
> Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
> ---
>  arch/x86/kvm/lapic.c | 8 ++++----
>  arch/x86/kvm/x86.c   | 6 +-----
>  2 files changed, 5 insertions(+), 9 deletions(-)

...


> @@ -2510,7 +2510,7 @@ void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu)
>  
>  	timer = &vcpu->arch.apic->lapic_timer.timer;
>  	if (hrtimer_cancel(timer))
> -		hrtimer_start_expires(timer, HRTIMER_MODE_ABS_PINNED);
> +		hrtimer_start_expires(timer, HRTIMER_MODE_ABS);
>  }
>  
>  /*

Wait, in that case why are we even bothering to cancel and restart the
timer? I thought the whole point of that was to pin it to the *new* CPU
that this vCPU is running on.

If not, can't we just kill __kvm_migrate_apic_timer() off completely
not?

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 5965 bytes --]

Re: [PATCH v7 1/2] KVM: LAPIC: Make lapic timer unpinned

[Marcelo Tosatti]

On Wed, Mar 09, 2022 at 10:26:51AM +0100, David Woodhouse wrote:
> On Sat, 2019-07-06 at 09:26 +0800, Wanpeng Li wrote:
> > From: Wanpeng Li <wanpengli@tencent.com>
> > 
> > Commit 61abdbe0bcc2 ("kvm: x86: make lapic hrtimer pinned") pinned the
> > lapic timer to avoid to wait until the next kvm exit for the guest to
> > see KVM_REQ_PENDING_TIMER set. There is another solution to give a kick
> > after setting the KVM_REQ_PENDING_TIMER bit, make lapic timer unpinned
> > will be used in follow up patches.
> > 
> > Cc: Paolo Bonzini <pbonzini@redhat.com>
> > Cc: Radim Krčmář <rkrcmar@redhat.com>
> > Cc: Marcelo Tosatti <mtosatti@redhat.com>
> > Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
> > ---
> >  arch/x86/kvm/lapic.c | 8 ++++----
> >  arch/x86/kvm/x86.c   | 6 +-----
> >  2 files changed, 5 insertions(+), 9 deletions(-)
> 
> ...
> 
> 
> > @@ -2510,7 +2510,7 @@ void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu)
> >  
> >  	timer = &vcpu->arch.apic->lapic_timer.timer;
> >  	if (hrtimer_cancel(timer))
> > -		hrtimer_start_expires(timer, HRTIMER_MODE_ABS_PINNED);
> > +		hrtimer_start_expires(timer, HRTIMER_MODE_ABS);
> >  }
> >  
> >  /*
> 
> Wait, in that case why are we even bothering to cancel and restart the
> timer? I thought the whole point of that was to pin it to the *new* CPU
> that this vCPU is running on.
> 
> If not, can't we just kill __kvm_migrate_apic_timer() off completely
> not?

Current code looks like this:

void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu)
{
        struct hrtimer *timer;

        if (!lapic_in_kernel(vcpu) ||
                kvm_can_post_timer_interrupt(vcpu)) <----------
                return;

        timer = &vcpu->arch.apic->lapic_timer.timer;
        if (hrtimer_cancel(timer))
                hrtimer_start_expires(timer, HRTIMER_MODE_ABS_HARD);
}

Yeah, should be HRTIMER_MODE_ABS_PINNED AFAICS.

Re: [PATCH v7 1/2] KVM: LAPIC: Make lapic timer unpinned

[David Woodhouse]

[-- Attachment #1: Type: text/plain, Size: 2624 bytes --]

On Wed, 2022-03-09 at 08:24 -0300, Marcelo Tosatti wrote:
> On Wed, Mar 09, 2022 at 10:26:51AM +0100, David Woodhouse wrote:
> > On Sat, 2019-07-06 at 09:26 +0800, Wanpeng Li wrote:
> > > From: Wanpeng Li <wanpengli@tencent.com>
> > > 
> > > Commit 61abdbe0bcc2 ("kvm: x86: make lapic hrtimer pinned") pinned the
> > > lapic timer to avoid to wait until the next kvm exit for the guest to
> > > see KVM_REQ_PENDING_TIMER set. There is another solution to give a kick
> > > after setting the KVM_REQ_PENDING_TIMER bit, make lapic timer unpinned
> > > will be used in follow up patches.
> > > 
> > > Cc: Paolo Bonzini <pbonzini@redhat.com>
> > > Cc: Radim Krčmář <rkrcmar@redhat.com>
> > > Cc: Marcelo Tosatti <mtosatti@redhat.com>
> > > Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
> > > ---
> > >  arch/x86/kvm/lapic.c | 8 ++++----
> > >  arch/x86/kvm/x86.c   | 6 +-----
> > >  2 files changed, 5 insertions(+), 9 deletions(-)
> > 
> > ...
> > 
> > 
> > > @@ -2510,7 +2510,7 @@ void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu)
> > >  
> > >  	timer = &vcpu->arch.apic->lapic_timer.timer;
> > >  	if (hrtimer_cancel(timer))
> > > -		hrtimer_start_expires(timer, HRTIMER_MODE_ABS_PINNED);
> > > +		hrtimer_start_expires(timer, HRTIMER_MODE_ABS);
> > >  }
> > >  
> > >  /*
> > 
> > Wait, in that case why are we even bothering to cancel and restart the
> > timer? I thought the whole point of that was to pin it to the *new* CPU
> > that this vCPU is running on.
> > 
> > If not, can't we just kill __kvm_migrate_apic_timer() off completely
> > not?
> 
> Current code looks like this:
> 
> void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu)
> {
>         struct hrtimer *timer;
> 
>         if (!lapic_in_kernel(vcpu) ||
>                 kvm_can_post_timer_interrupt(vcpu)) <----------
>                 return;
> 
>         timer = &vcpu->arch.apic->lapic_timer.timer;
>         if (hrtimer_cancel(timer))
>                 hrtimer_start_expires(timer, HRTIMER_MODE_ABS_HARD);
> }
> 
> Yeah, should be HRTIMER_MODE_ABS_PINNED AFAICS.

No, it's *intentionally* not pinned any more, since this patch that I'm
replying to, which became commit 4d151bf3b89. It doesn't *have* to run
on the same physical CPU, because of the epiphany that it can just call
kvm_vcpu_kick() after making the request.

But if it's a recurring timer it's still *best* for it to run on the
same physical CPU, just for cache locality reasons. So I think I was
wrong; the migration *isn't* pointless. It's still a valid
optimisation; it's just not *mandatory* any more.



[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 5965 bytes --]

[PATCH v7 2/2] KVM: LAPIC: Inject timer interrupt via posted interrupt

[Wanpeng Li]

From: Wanpeng Li <wanpengli@tencent.com>

Dedicated instances are currently disturbed by unnecessary jitter due 
to the emulated lapic timers fire on the same pCPUs which vCPUs resident.
There is no hardware virtual timer on Intel for guest like ARM. Both 
programming timer in guest and the emulated timer fires incur vmexits.
This patch tries to avoid vmexit which is incurred by the emulated 
timer fires in dedicated instance scenario. 

When nohz_full is enabled in dedicated instances scenario, the emulated 
timers can be offload to the nearest busy housekeeping cpus since APICv 
is really common in recent years. The guest timer interrupt is injected 
by posted-interrupt which is delivered by housekeeping cpu once the emulated 
timer fires. 

The host admin should fine tuned, e.g. dedicated instances scenario w/ 
nohz_full cover the pCPUs which vCPUs resident, several pCPUs surplus 
for busy housekeeping, disable mwait/hlt/pause vmexits to keep in non-root  
mode, ~3% redis performance benefit can be observed on Skylake server.

w/o patch:

            VM-EXIT  Samples  Samples%  Time%   Min Time  Max Time   Avg time

EXTERNAL_INTERRUPT    42916    49.43%   39.30%   0.47us   106.09us   0.71us ( +-   1.09% )

w/ patch:

            VM-EXIT  Samples  Samples%  Time%   Min Time  Max Time         Avg time

EXTERNAL_INTERRUPT    6871     9.29%     2.96%   0.44us    57.88us   0.72us ( +-   4.02% )

Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>

Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
---
 arch/x86/kvm/lapic.c            | 101 ++++++++++++++++++++++++++--------------
 arch/x86/kvm/lapic.h            |   1 +
 arch/x86/kvm/vmx/vmx.c          |   3 +-
 arch/x86/kvm/x86.c              |   6 +++
 arch/x86/kvm/x86.h              |   2 +
 include/linux/sched/isolation.h |   2 +
 kernel/sched/isolation.c        |   6 +++
 7 files changed, 85 insertions(+), 36 deletions(-)

diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index 707ca9c..4869691 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -128,6 +128,17 @@ static inline u32 kvm_x2apic_id(struct kvm_lapic *apic)
 	return apic->vcpu->vcpu_id;
 }
 
+bool kvm_can_post_timer_interrupt(struct kvm_vcpu *vcpu)
+{
+	return pi_inject_timer && kvm_vcpu_apicv_active(vcpu);
+}
+EXPORT_SYMBOL_GPL(kvm_can_post_timer_interrupt);
+
+static bool kvm_use_posted_timer_interrupt(struct kvm_vcpu *vcpu)
+{
+	return kvm_can_post_timer_interrupt(vcpu) && vcpu->mode == IN_GUEST_MODE;
+}
+
 static inline bool kvm_apic_map_get_logical_dest(struct kvm_apic_map *map,
 		u32 dest_id, struct kvm_lapic ***cluster, u16 *mask) {
 	switch (map->mode) {
@@ -1436,29 +1447,6 @@ static void apic_update_lvtt(struct kvm_lapic *apic)
 	}
 }
 
-static void apic_timer_expired(struct kvm_lapic *apic)
-{
-	struct kvm_vcpu *vcpu = apic->vcpu;
-	struct swait_queue_head *q = &vcpu->wq;
-	struct kvm_timer *ktimer = &apic->lapic_timer;
-
-	if (atomic_read(&apic->lapic_timer.pending))
-		return;
-
-	atomic_inc(&apic->lapic_timer.pending);
-	kvm_set_pending_timer(vcpu);
-
-	/*
-	 * For x86, the atomic_inc() is serialized, thus
-	 * using swait_active() is safe.
-	 */
-	if (swait_active(q))
-		swake_up_one(q);
-
-	if (apic_lvtt_tscdeadline(apic) || ktimer->hv_timer_in_use)
-		ktimer->expired_tscdeadline = ktimer->tscdeadline;
-}
-
 /*
  * On APICv, this test will cause a busy wait
  * during a higher-priority task.
@@ -1532,7 +1520,7 @@ static inline void adjust_lapic_timer_advance(struct kvm_vcpu *vcpu,
 	apic->lapic_timer.timer_advance_ns = timer_advance_ns;
 }
 
-void kvm_wait_lapic_expire(struct kvm_vcpu *vcpu)
+static void __kvm_wait_lapic_expire(struct kvm_vcpu *vcpu)
 {
 	struct kvm_lapic *apic = vcpu->arch.apic;
 	u64 guest_tsc, tsc_deadline;
@@ -1540,9 +1528,6 @@ void kvm_wait_lapic_expire(struct kvm_vcpu *vcpu)
 	if (apic->lapic_timer.expired_tscdeadline == 0)
 		return;
 
-	if (!lapic_timer_int_injected(vcpu))
-		return;
-
 	tsc_deadline = apic->lapic_timer.expired_tscdeadline;
 	apic->lapic_timer.expired_tscdeadline = 0;
 	guest_tsc = kvm_read_l1_tsc(vcpu, rdtsc());
@@ -1554,8 +1539,59 @@ void kvm_wait_lapic_expire(struct kvm_vcpu *vcpu)
 	if (unlikely(!apic->lapic_timer.timer_advance_adjust_done))
 		adjust_lapic_timer_advance(vcpu, apic->lapic_timer.advance_expire_delta);
 }
+
+void kvm_wait_lapic_expire(struct kvm_vcpu *vcpu)
+{
+	if (!lapic_timer_int_injected(vcpu))
+		return;
+
+	__kvm_wait_lapic_expire(vcpu);
+}
 EXPORT_SYMBOL_GPL(kvm_wait_lapic_expire);
 
+static void kvm_apic_inject_pending_timer_irqs(struct kvm_lapic *apic)
+{
+	struct kvm_timer *ktimer = &apic->lapic_timer;
+
+	kvm_apic_local_deliver(apic, APIC_LVTT);
+	if (apic_lvtt_tscdeadline(apic))
+		ktimer->tscdeadline = 0;
+	if (apic_lvtt_oneshot(apic)) {
+		ktimer->tscdeadline = 0;
+		ktimer->target_expiration = 0;
+	}
+}
+
+static void apic_timer_expired(struct kvm_lapic *apic)
+{
+	struct kvm_vcpu *vcpu = apic->vcpu;
+	struct swait_queue_head *q = &vcpu->wq;
+	struct kvm_timer *ktimer = &apic->lapic_timer;
+
+	if (atomic_read(&apic->lapic_timer.pending))
+		return;
+
+	if (apic_lvtt_tscdeadline(apic) || ktimer->hv_timer_in_use)
+		ktimer->expired_tscdeadline = ktimer->tscdeadline;
+
+	if (kvm_use_posted_timer_interrupt(apic->vcpu)) {
+		if (apic->lapic_timer.timer_advance_ns)
+			__kvm_wait_lapic_expire(vcpu);
+		kvm_apic_inject_pending_timer_irqs(apic);
+		return;
+	}
+
+	atomic_inc(&apic->lapic_timer.pending);
+	kvm_set_pending_timer(vcpu);
+
+	/*
+	 * For x86, the atomic_inc() is serialized, thus
+	 * using swait_active() is safe.
+	 */
+	if (swait_active(q))
+		swake_up_one(q);
+}
+
 static void start_sw_tscdeadline(struct kvm_lapic *apic)
 {
 	struct kvm_timer *ktimer = &apic->lapic_timer;
@@ -2377,13 +2413,7 @@ void kvm_inject_apic_timer_irqs(struct kvm_vcpu *vcpu)
 	struct kvm_lapic *apic = vcpu->arch.apic;
 
 	if (atomic_read(&apic->lapic_timer.pending) > 0) {
-		kvm_apic_local_deliver(apic, APIC_LVTT);
-		if (apic_lvtt_tscdeadline(apic))
-			apic->lapic_timer.tscdeadline = 0;
-		if (apic_lvtt_oneshot(apic)) {
-			apic->lapic_timer.tscdeadline = 0;
-			apic->lapic_timer.target_expiration = 0;
-		}
+		kvm_apic_inject_pending_timer_irqs(apic);
 		atomic_set(&apic->lapic_timer.pending, 0);
 	}
 }
@@ -2505,7 +2535,8 @@ void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu)
 {
 	struct hrtimer *timer;
 
-	if (!lapic_in_kernel(vcpu))
+	if (!lapic_in_kernel(vcpu) ||
+		kvm_can_post_timer_interrupt(vcpu))
 		return;
 
 	timer = &vcpu->arch.apic->lapic_timer.timer;
diff --git a/arch/x86/kvm/lapic.h b/arch/x86/kvm/lapic.h
index f974a3d..e32dd7d 100644
--- a/arch/x86/kvm/lapic.h
+++ b/arch/x86/kvm/lapic.h
@@ -231,6 +231,7 @@ int kvm_vector_to_index(u32 vector, u32 dest_vcpus,
 void kvm_lapic_expired_hv_timer(struct kvm_vcpu *vcpu);
 bool kvm_lapic_hv_timer_in_use(struct kvm_vcpu *vcpu);
 void kvm_lapic_restart_hv_timer(struct kvm_vcpu *vcpu);
+bool kvm_can_post_timer_interrupt(struct kvm_vcpu *vcpu);
 
 static inline enum lapic_mode kvm_apic_mode(u64 apic_base)
 {
diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index b35b3800..d152552a 100644
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -7036,7 +7036,8 @@ static int vmx_set_hv_timer(struct kvm_vcpu *vcpu, u64 guest_deadline_tsc,
 	u64 tscl, guest_tscl, delta_tsc, lapic_timer_advance_cycles;
 	struct kvm_timer *ktimer = &vcpu->arch.apic->lapic_timer;
 
-	if (kvm_mwait_in_guest(vcpu->kvm))
+	if (kvm_mwait_in_guest(vcpu->kvm) ||
+		kvm_can_post_timer_interrupt(vcpu))
 		return -EOPNOTSUPP;
 
 	vmx = to_vmx(vcpu);
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index e199ac7..ed63103 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -54,6 +54,7 @@
 #include <linux/kvm_irqfd.h>
 #include <linux/irqbypass.h>
 #include <linux/sched/stat.h>
+#include <linux/sched/isolation.h>
 #include <linux/mem_encrypt.h>
 
 #include <trace/events/kvm.h>
@@ -155,6 +156,9 @@
 static bool __read_mostly force_emulation_prefix = false;
 module_param(force_emulation_prefix, bool, S_IRUGO);
 
+int __read_mostly pi_inject_timer = -1;
+module_param(pi_inject_timer, bint, S_IRUGO | S_IWUSR);
+
 #define KVM_NR_SHARED_MSRS 16
 
 struct kvm_shared_msrs_global {
@@ -7043,6 +7047,8 @@ int kvm_arch_init(void *opaque)
 		host_xcr0 = xgetbv(XCR_XFEATURE_ENABLED_MASK);
 
 	kvm_lapic_init();
+	if (pi_inject_timer == -1)
+		pi_inject_timer = housekeeping_enabled(HK_FLAG_TIMER);
 #ifdef CONFIG_X86_64
 	pvclock_gtod_register_notifier(&pvclock_gtod_notifier);
 
diff --git a/arch/x86/kvm/x86.h b/arch/x86/kvm/x86.h
index a470ff0..bb1e99b 100644
--- a/arch/x86/kvm/x86.h
+++ b/arch/x86/kvm/x86.h
@@ -296,6 +296,8 @@ int x86_emulate_instruction(struct kvm_vcpu *vcpu, unsigned long cr2,
 
 extern bool enable_vmware_backdoor;
 
+extern int pi_inject_timer;
+
 extern struct static_key kvm_no_apic_vcpu;
 
 static inline u64 nsec_to_cycles(struct kvm_vcpu *vcpu, u64 nsec)
diff --git a/include/linux/sched/isolation.h b/include/linux/sched/isolation.h
index b0fb144..6fc5407 100644
--- a/include/linux/sched/isolation.h
+++ b/include/linux/sched/isolation.h
@@ -19,6 +19,7 @@ enum hk_flags {
 DECLARE_STATIC_KEY_FALSE(housekeeping_overridden);
 extern int housekeeping_any_cpu(enum hk_flags flags);
 extern const struct cpumask *housekeeping_cpumask(enum hk_flags flags);
+extern bool housekeeping_enabled(enum hk_flags flags);
 extern void housekeeping_affine(struct task_struct *t, enum hk_flags flags);
 extern bool housekeeping_test_cpu(int cpu, enum hk_flags flags);
 extern void __init housekeeping_init(void);
@@ -38,6 +39,7 @@ static inline const struct cpumask *housekeeping_cpumask(enum hk_flags flags)
 static inline void housekeeping_affine(struct task_struct *t,
 				       enum hk_flags flags) { }
 static inline void housekeeping_init(void) { }
+static inline bool housekeeping_enabled(enum hk_flags flags) { }
 #endif /* CONFIG_CPU_ISOLATION */
 
 static inline bool housekeeping_cpu(int cpu, enum hk_flags flags)
diff --git a/kernel/sched/isolation.c b/kernel/sched/isolation.c
index 6873020..4d4bae8 100644
--- a/kernel/sched/isolation.c
+++ b/kernel/sched/isolation.c
@@ -13,6 +13,12 @@
 static cpumask_var_t housekeeping_mask;
 static unsigned int housekeeping_flags;
 
+bool housekeeping_enabled(enum hk_flags flags)
+{
+	return !!(housekeeping_flags & flags);
+}
+EXPORT_SYMBOL_GPL(housekeeping_enabled);
+
 int housekeeping_any_cpu(enum hk_flags flags)
 {
 	if (static_branch_unlikely(&housekeeping_overridden))
-- 
1.8.3.1

Re: [PATCH v7 2/2] KVM: LAPIC: Inject timer interrupt via posted interrupt

[Paolo Bonzini]

On 06/07/19 03:26, Wanpeng Li wrote:
> +
> +void kvm_wait_lapic_expire(struct kvm_vcpu *vcpu)
> +{
> +	if (!lapic_timer_int_injected(vcpu))
> +		return;
> +
> +	__kvm_wait_lapic_expire(vcpu);
> +}
>  EXPORT_SYMBOL_GPL(kvm_wait_lapic_expire);

I changed this to

	if (lapic_timer_int_injected(vcpu))
		__kvm_wait_lapic_expire(vcpu);

and queued the two patches.

Paolo

Re: [PATCH v7 0/2] KVM: LAPIC: Implement Exitless Timer

[Paolo Bonzini]

On 11/07/19 15:50, Wanpeng Li wrote:
> kindly ping，

Sorry, I need more time to review this.  It's basically the only
remaining item for the 5.3 merge window, even though it won't be part of
the first pull request to Linus.

Paolo

>> Dedicated instances are currently disturbed by unnecessary jitter due 
>> to the emulated lapic timers fire on the same pCPUs which vCPUs resident.
>> There is no hardware virtual timer on Intel for guest like ARM. Both 
>> programming timer in guest and the emulated timer fires incur vmexits.
>> This patchset tries to avoid vmexit which is incurred by the emulated 
>> timer fires in dedicated instance scenario. 
>>
>> When nohz_full is enabled in dedicated instances scenario, the unpinned 
>> timer will be moved to the nearest busy housekeepers after commit
>> 9642d18eee2cd (nohz: Affine unpinned timers to housekeepers) and commit 
>> 444969223c8 ("sched/nohz: Fix affine unpinned timers mess"). However, 
>> KVM always makes lapic timer pinned to the pCPU which vCPU residents, the 
>> reason is explained by commit 61abdbe0 (kvm: x86: make lapic hrtimer 
>> pinned). Actually, these emulated timers can be offload to the housekeeping 
>> cpus since APICv is really common in recent years. The guest timer interrupt 
>> is injected by posted-interrupt which is delivered by housekeeping cpu 
>> once the emulated timer fires. 
>>
>> The host admin should fine tuned, e.g. dedicated instances scenario w/ 
>> nohz_full cover the pCPUs which vCPUs resident, several pCPUs surplus 
>> for busy housekeeping, disable mwait/hlt/pause vmexits to keep in non-root  
>> mode, ~3% redis performance benefit can be observed on Skylake server.
>>
>> w/o patchset:
>>
>>            VM-EXIT  Samples  Samples%  Time%   Min Time  Max Time   Avg time
>>
>> EXTERNAL_INTERRUPT    42916    49.43%   39.30%   0.47us   106.09us   0.71us ( +-   1.09% )
>>
>> w/ patchset:
>>
>>            VM-EXIT  Samples  Samples%  Time%   Min Time  Max Time         Avg time
>>
>> EXTERNAL_INTERRUPT    6871     9.29%     2.96%   0.44us    57.88us   0.72us ( +-   4.02% )
>>
>> Cc: Paolo Bonzini <pbonzini@redhat.com>
>> Cc: Radim Krčmář <rkrcmar@redhat.com>
>> Cc: Marcelo Tosatti <mtosatti@redhat.com>
>>
>> v6 -> v7:
>> * remove bool argument
>>
>> v5 -> v6:
>> * don't overwrites whatever the user specified
>> * introduce kvm_can_post_timer_interrupt and kvm_use_posted_timer_interrupt
>> * remove kvm_hlt_in_guest() condition
>> * squash all of 2/3/4 together
>>
>> v4 -> v5:
>> * update patch description in patch 1/4
>> * feed latest apic->lapic_timer.expired_tscdeadline to kvm_wait_lapic_expire()
>> * squash advance timer handling to patch 2/4
>>
>> v3 -> v4:
>> * drop the HRTIMER_MODE_ABS_PINNED, add kick after set pending timer
>> * don't posted inject already-expired timer
>>
>> v2 -> v3:
>> * disarming the vmx preemption timer when posted_interrupt_inject_timer_enabled()
>> * check kvm_hlt_in_guest instead
>>
>> v1 -> v2:
>> * check vcpu_halt_in_guest
>> * move module parameter from kvm-intel to kvm
>> * add housekeeping_enabled
>> * rename apic_timer_expired_pi to kvm_apic_inject_pending_timer_irqs
>>
>>
>> Wanpeng Li (2):
>>  KVM: LAPIC: Make lapic timer unpinned
>>  KVM: LAPIC: Inject timer interrupt via posted interrupt
>>
>> arch/x86/kvm/lapic.c            | 109 ++++++++++++++++++++++++++--------------
>> arch/x86/kvm/lapic.h            |   1 +
>> arch/x86/kvm/vmx/vmx.c          |   3 +-
>> arch/x86/kvm/x86.c              |  12 +++--
>> arch/x86/kvm/x86.h              |   2 +
>> include/linux/sched/isolation.h |   2 +
>> kernel/sched/isolation.c        |   6 +++
>> 7 files changed, 90 insertions(+), 45 deletions(-)
>>
>> -- 
>> 1.8.3.1
>>

Re: [PATCH v7 0/2] KVM: LAPIC: Implement Exitless Timer

[Wanpeng Li]

On Thu, 11 Jul 2019 at 22:05, Paolo Bonzini <pbonzini@redhat.com> wrote:
>
> On 11/07/19 15:50, Wanpeng Li wrote:
> > kindly ping，
>
> Sorry, I need more time to review this.  It's basically the only
> remaining item for the 5.3 merge window, even though it won't be part of
> the first pull request to Linus.

Thank you! Hope finally we will not miss it. :)

Regards,
Wanpeng Li

Re: [PATCH v7 0/2] KVM: LAPIC: Implement Exitless Timer

[Wanpeng Li]

On Sat, 6 Jul 2019 at 09:26, Wanpeng Li <kernellwp@gmail.com> wrote:
>
> Dedicated instances are currently disturbed by unnecessary jitter due
> to the emulated lapic timers fire on the same pCPUs which vCPUs resident.
> There is no hardware virtual timer on Intel for guest like ARM. Both
> programming timer in guest and the emulated timer fires incur vmexits.
> This patchset tries to avoid vmexit which is incurred by the emulated
> timer fires in dedicated instance scenario.
>
> When nohz_full is enabled in dedicated instances scenario, the unpinned
> timer will be moved to the nearest busy housekeepers after commit
> 9642d18eee2cd (nohz: Affine unpinned timers to housekeepers) and commit
> 444969223c8 ("sched/nohz: Fix affine unpinned timers mess"). However,
> KVM always makes lapic timer pinned to the pCPU which vCPU residents, the
> reason is explained by commit 61abdbe0 (kvm: x86: make lapic hrtimer
> pinned). Actually, these emulated timers can be offload to the housekeeping
> cpus since APICv is really common in recent years. The guest timer interrupt
> is injected by posted-interrupt which is delivered by housekeeping cpu
> once the emulated timer fires.
>
> The host admin should fine tuned, e.g. dedicated instances scenario w/
> nohz_full cover the pCPUs which vCPUs resident, several pCPUs surplus
> for busy housekeeping, disable mwait/hlt/pause vmexits to keep in non-root
> mode, ~3% redis performance benefit can be observed on Skylake server.
>
> w/o patchset:
>
>             VM-EXIT  Samples  Samples%  Time%   Min Time  Max Time   Avg time
>
> EXTERNAL_INTERRUPT    42916    49.43%   39.30%   0.47us   106.09us   0.71us ( +-   1.09% )
>
> w/ patchset:
>
>             VM-EXIT  Samples  Samples%  Time%   Min Time  Max Time         Avg time
>
> EXTERNAL_INTERRUPT    6871     9.29%     2.96%   0.44us    57.88us   0.72us ( +-   4.02% )
>
> Cc: Paolo Bonzini <pbonzini@redhat.com>
> Cc: Radim Krčmář <rkrcmar@redhat.com>
> Cc: Marcelo Tosatti <mtosatti@redhat.com>
>
> v6 -> v7:
>  * remove bool argument
>
> v5 -> v6:
>  * don't overwrites whatever the user specified
>  * introduce kvm_can_post_timer_interrupt and kvm_use_posted_timer_interrupt
>  * remove kvm_hlt_in_guest() condition
>  * squash all of 2/3/4 together
>
> v4 -> v5:
>  * update patch description in patch 1/4
>  * feed latest apic->lapic_timer.expired_tscdeadline to kvm_wait_lapic_expire()
>  * squash advance timer handling to patch 2/4
>
> v3 -> v4:
>  * drop the HRTIMER_MODE_ABS_PINNED, add kick after set pending timer
>  * don't posted inject already-expired timer
>
> v2 -> v3:
>  * disarming the vmx preemption timer when posted_interrupt_inject_timer_enabled()
>  * check kvm_hlt_in_guest instead
>
> v1 -> v2:
>  * check vcpu_halt_in_guest
>  * move module parameter from kvm-intel to kvm
>  * add housekeeping_enabled
>  * rename apic_timer_expired_pi to kvm_apic_inject_pending_timer_irqs
>
>
> Wanpeng Li (2):
>   KVM: LAPIC: Make lapic timer unpinned
>   KVM: LAPIC: Inject timer interrupt via posted interrupt

There is a further optimization for this feature in houseeking/hrtimer
subsystem.

[1] https://lkml.org/lkml/2019/7/25/963
[2] https://lkml.org/lkml/2019/6/28/231

The [2] patch tries to optimize the worst case, however, it will not
be merged by maintainers and get offline confirm, Thomas will refactor
this to avoid to predict the future on every timer enqueue. Anyway, it
still should be considered to be backported to product environment as
long as get_nohz_timer_target() is using.

Regards,
Wanpeng Li