[PATCH v2 0/4] KVM: arm/arm64: guest synchronous halt/resume

[PATCH v2 0/4] KVM: arm/arm64: guest synchronous halt/resume

[Eric Auger]

This series introduces the capability to synchronously exit the guest
and prevent it from being re-entered. This modality will be used by
IRQ forwarding series when changing the state of the IRQ.

Former pause flag used when starting the vcpu in KVM_ARM_VCPU_POWER_OFF
state, in PSCI calls and in KVM_SET_MP_STATE ioctl is renamed into
power_off. A new pause flag is introduced. Both now are checked in
kvm_arch_vcpu_runnable and in the VCPU_RUN critical section, before
entering the vcpu.

Best Regards

Eric

History:

v1 -> v2:
- check pause, power_off and in kvm_arch_vcpu_runnable
- check power_off in vcpu_run critcal section before guest entry
- correct compil issue in first patch reported by Andrew
- rename vcpu_pause into vcpu_sleep

RFC -> PATCH v1:
- originally part of [RFC 00/17] ARM IRQ forward control based on IRQ
  bypass manager (https://lkml.org/lkml/2015/7/2/268) and isolated in
  this series.
- added __maybe_unused following Marc's advice


Eric Auger (4):
  KVM: arm/arm64: rename pause into power_off
  KVM: arm/arm64: check power_off in kvm_arch_vcpu_runnable
  KVM: arm/arm64: check power_off in critical section before VCPU run
  KVM: arm/arm64: implement kvm_arm_[halt,resume]_guest

 arch/arm/include/asm/kvm_host.h   |  5 +++-
 arch/arm/kvm/arm.c                | 53 ++++++++++++++++++++++++++++++---------
 arch/arm/kvm/psci.c               | 10 ++++----
 arch/arm64/include/asm/kvm_host.h |  3 +++
 4 files changed, 53 insertions(+), 18 deletions(-)

-- 
1.9.1

[PATCH v2 1/4] KVM: arm/arm64: rename pause into power_off

[Eric Auger]

The kvm_vcpu_arch pause field is renamed into power_off to prepare
for the introduction of a new pause field. Also vcpu_pause is renamed
into vcpu_sleep since we will sleep until both power_off and pause are
false.

Signed-off-by: Eric Auger <eric.auger@linaro.org>

---

v1 -> v2:
- rename pause in kvm_arch_vcpu_ioctl_[set,get]_mpstate
- rename vcpu_pause into vcpu_sleep
---
 arch/arm/include/asm/kvm_host.h   |  4 ++--
 arch/arm/kvm/arm.c                | 20 ++++++++++----------
 arch/arm/kvm/psci.c               | 10 +++++-----
 arch/arm64/include/asm/kvm_host.h |  4 ++--
 4 files changed, 19 insertions(+), 19 deletions(-)

diff --git a/arch/arm/include/asm/kvm_host.h b/arch/arm/include/asm/kvm_host.h
index e896d2c..304004d 100644
--- a/arch/arm/include/asm/kvm_host.h
+++ b/arch/arm/include/asm/kvm_host.h
@@ -129,8 +129,8 @@ struct kvm_vcpu_arch {
 	 * here.
 	 */
 
-	/* Don't run the guest on this vcpu */
-	bool pause;
+	/* vcpu power-off state */
+	bool power_off;
 
 	/* IO related fields */
 	struct kvm_decode mmio_decode;
diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index b598aa4..3ac6b4c 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -313,7 +313,7 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
-	if (vcpu->arch.pause)
+	if (vcpu->arch.power_off)
 		mp_state->mp_state = KVM_MP_STATE_STOPPED;
 	else
 		mp_state->mp_state = KVM_MP_STATE_RUNNABLE;
@@ -326,10 +326,10 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 {
 	switch (mp_state->mp_state) {
 	case KVM_MP_STATE_RUNNABLE:
-		vcpu->arch.pause = false;
+		vcpu->arch.power_off = false;
 		break;
 	case KVM_MP_STATE_STOPPED:
-		vcpu->arch.pause = true;
+		vcpu->arch.power_off = true;
 		break;
 	default:
 		return -EINVAL;
@@ -473,11 +473,11 @@ bool kvm_arch_intc_initialized(struct kvm *kvm)
 	return vgic_initialized(kvm);
 }
 
-static void vcpu_pause(struct kvm_vcpu *vcpu)
+static void vcpu_sleep(struct kvm_vcpu *vcpu)
 {
 	wait_queue_head_t *wq = kvm_arch_vcpu_wq(vcpu);
 
-	wait_event_interruptible(*wq, !vcpu->arch.pause);
+	wait_event_interruptible(*wq, !vcpu->arch.power_off);
 }
 
 static int kvm_vcpu_initialized(struct kvm_vcpu *vcpu)
@@ -527,8 +527,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 
 		update_vttbr(vcpu->kvm);
 
-		if (vcpu->arch.pause)
-			vcpu_pause(vcpu);
+		if (vcpu->arch.power_off)
+			vcpu_sleep(vcpu);
 
 		/*
 		 * Disarming the background timer must be done in a
@@ -766,12 +766,12 @@ static int kvm_arch_vcpu_ioctl_vcpu_init(struct kvm_vcpu *vcpu,
 	vcpu_reset_hcr(vcpu);
 
 	/*
-	 * Handle the "start in power-off" case by marking the VCPU as paused.
+	 * Handle the "start in power-off" case.
 	 */
 	if (test_bit(KVM_ARM_VCPU_POWER_OFF, vcpu->arch.features))
-		vcpu->arch.pause = true;
+		vcpu->arch.power_off = true;
 	else
-		vcpu->arch.pause = false;
+		vcpu->arch.power_off = false;
 
 	return 0;
 }
diff --git a/arch/arm/kvm/psci.c b/arch/arm/kvm/psci.c
index 4b94b51..134971a 100644
--- a/arch/arm/kvm/psci.c
+++ b/arch/arm/kvm/psci.c
@@ -63,7 +63,7 @@ static unsigned long kvm_psci_vcpu_suspend(struct kvm_vcpu *vcpu)
 
 static void kvm_psci_vcpu_off(struct kvm_vcpu *vcpu)
 {
-	vcpu->arch.pause = true;
+	vcpu->arch.power_off = true;
 }
 
 static unsigned long kvm_psci_vcpu_on(struct kvm_vcpu *source_vcpu)
@@ -87,7 +87,7 @@ static unsigned long kvm_psci_vcpu_on(struct kvm_vcpu *source_vcpu)
 	 */
 	if (!vcpu)
 		return PSCI_RET_INVALID_PARAMS;
-	if (!vcpu->arch.pause) {
+	if (!vcpu->arch.power_off) {
 		if (kvm_psci_version(source_vcpu) != KVM_ARM_PSCI_0_1)
 			return PSCI_RET_ALREADY_ON;
 		else
@@ -115,7 +115,7 @@ static unsigned long kvm_psci_vcpu_on(struct kvm_vcpu *source_vcpu)
 	 * the general puspose registers are undefined upon CPU_ON.
 	 */
 	*vcpu_reg(vcpu, 0) = context_id;
-	vcpu->arch.pause = false;
+	vcpu->arch.power_off = false;
 	smp_mb();		/* Make sure the above is visible */
 
 	wq = kvm_arch_vcpu_wq(vcpu);
@@ -152,7 +152,7 @@ static unsigned long kvm_psci_vcpu_affinity_info(struct kvm_vcpu *vcpu)
 	kvm_for_each_vcpu(i, tmp, kvm) {
 		mpidr = kvm_vcpu_get_mpidr_aff(tmp);
 		if (((mpidr & target_affinity_mask) == target_affinity) &&
-		    !tmp->arch.pause) {
+		    !tmp->arch.power_off) {
 			return PSCI_0_2_AFFINITY_LEVEL_ON;
 		}
 	}
@@ -175,7 +175,7 @@ static void kvm_prepare_system_event(struct kvm_vcpu *vcpu, u32 type)
 	 * re-initialized.
 	 */
 	kvm_for_each_vcpu(i, tmp, vcpu->kvm) {
-		tmp->arch.pause = true;
+		tmp->arch.power_off = true;
 		kvm_vcpu_kick(tmp);
 	}
 
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 2709db2..009da6b 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -122,8 +122,8 @@ struct kvm_vcpu_arch {
 	 * here.
 	 */
 
-	/* Don't run the guest */
-	bool pause;
+	/* vcpu power-off state */
+	bool power_off;
 
 	/* IO related fields */
 	struct kvm_decode mmio_decode;
-- 
1.9.1

[PATCH v2 2/4] KVM: arm/arm64: check power_off in kvm_arch_vcpu_runnable

[Eric Auger]

kvm_arch_vcpu_runnable now also checks whether the power_off
flag is set.

Signed-off-by: Eric Auger <eric.auger@linaro.org>
---
 arch/arm/kvm/arm.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index 3ac6b4c..4f50be3 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -347,7 +347,8 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
  */
 int kvm_arch_vcpu_runnable(struct kvm_vcpu *v)
 {
-	return !!v->arch.irq_lines || kvm_vgic_vcpu_pending_irq(v);
+	return ((!!v->arch.irq_lines || kvm_vgic_vcpu_pending_irq(v))
+		&& !v->arch.power_off);
 }
 
 /* Just ensure a guest exit from a particular CPU */
-- 
1.9.1

[PATCH v2 3/4] KVM: arm/arm64: check power_off in critical section before VCPU run

[Eric Auger]

In case KVM_SET_MP_STATE ioctl is called just after we executed the
vcpu_sleep check, we can enter the guest although KVM_MP_STATE_STOPPED
is set. Let's check the power_off state in the critical section,
just before entering the guest.

Signed-off-by: Eric Auger <eric.auger@linaro.org>
Reported-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/arm/kvm/arm.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index 4f50be3..cc404a8 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -555,7 +555,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 			run->exit_reason = KVM_EXIT_INTR;
 		}
 
-		if (ret <= 0 || need_new_vmid_gen(vcpu->kvm)) {
+		if (ret <= 0 || need_new_vmid_gen(vcpu->kvm) ||
+			vcpu->arch.power_off) {
 			local_irq_enable();
 			kvm_vgic_sync_hwstate(vcpu);
 			preempt_enable();
-- 
1.9.1

[PATCH v2 3/4] KVM: arm/arm64: check power_off in critical section before VCPU run

[Christoffer Dall]

On Fri, Aug 07, 2015 at 06:08:32PM +0200, Eric Auger wrote:
> In case KVM_SET_MP_STATE ioctl is called just after we executed the
> vcpu_sleep check, we can enter the guest although KVM_MP_STATE_STOPPED
> is set. Let's check the power_off state in the critical section,
> just before entering the guest.

I don't think this can happen from the KVM_SET_MP_STATE, because that is
a vcpu ioctl and would therefore require userspace sets this before even
entering the run loop in the kernel, but it could happen at the PSCI
system event preparation.

> 
> Signed-off-by: Eric Auger <eric.auger@linaro.org>
> Reported-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/arm/kvm/arm.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
> index 4f50be3..cc404a8 100644
> --- a/arch/arm/kvm/arm.c
> +++ b/arch/arm/kvm/arm.c
> @@ -555,7 +555,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
>  			run->exit_reason = KVM_EXIT_INTR;
>  		}
>  
> -		if (ret <= 0 || need_new_vmid_gen(vcpu->kvm)) {
> +		if (ret <= 0 || need_new_vmid_gen(vcpu->kvm) ||
> +			vcpu->arch.power_off) {
>  			local_irq_enable();
>  			kvm_vgic_sync_hwstate(vcpu);
>  			preempt_enable();
> -- 
> 1.9.1
> 

[PATCH v2 4/4] KVM: arm/arm64: implement kvm_arm_[halt,resume]_guest

[Eric Auger]

We introduce kvm_arm_halt_guest and resume functions. They
will be used for IRQ forward state change.

Halt is synchronous and prevents the guest from being re-entered.
We use the same mechanism put in place for PSCI former pause,
now renamed power_off. A new flag is introduced in arch vcpu state,
pause, only meant to be used by those functions.

Signed-off-by: Eric Auger <eric.auger@linaro.org>

---
v1 -> v2:
- check pause in kvm_arch_vcpu_runnable
- we cannot use kvm_vcpu_block since this latter would exit on
  IRQ/FIQ and this is not what we want
---
 arch/arm/include/asm/kvm_host.h   |  3 +++
 arch/arm/kvm/arm.c                | 35 +++++++++++++++++++++++++++++++----
 arch/arm64/include/asm/kvm_host.h |  3 +++
 3 files changed, 37 insertions(+), 4 deletions(-)

diff --git a/arch/arm/include/asm/kvm_host.h b/arch/arm/include/asm/kvm_host.h
index 304004d..dac85f6 100644
--- a/arch/arm/include/asm/kvm_host.h
+++ b/arch/arm/include/asm/kvm_host.h
@@ -132,6 +132,9 @@ struct kvm_vcpu_arch {
 	/* vcpu power-off state */
 	bool power_off;
 
+	 /* Exit and don't run the guest (internal implementation need) */
+	bool pause;
+
 	/* IO related fields */
 	struct kvm_decode mmio_decode;
 
diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
index cc404a8..0529b38 100644
--- a/arch/arm/kvm/arm.c
+++ b/arch/arm/kvm/arm.c
@@ -348,7 +348,7 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_runnable(struct kvm_vcpu *v)
 {
 	return ((!!v->arch.irq_lines || kvm_vgic_vcpu_pending_irq(v))
-		&& !v->arch.power_off);
+		&& !v->arch.power_off && !v->arch.pause);
 }
 
 /* Just ensure a guest exit from a particular CPU */
@@ -474,11 +474,38 @@ bool kvm_arch_intc_initialized(struct kvm *kvm)
 	return vgic_initialized(kvm);
 }
 
+static void kvm_arm_halt_guest(struct kvm *kvm) __maybe_unused;
+static void kvm_arm_resume_guest(struct kvm *kvm) __maybe_unused;
+
+static void kvm_arm_halt_guest(struct kvm *kvm)
+{
+	int i;
+	struct kvm_vcpu *vcpu;
+
+	kvm_for_each_vcpu(i, vcpu, kvm)
+		vcpu->arch.pause = true;
+	force_vm_exit(cpu_all_mask);
+}
+
+static void kvm_arm_resume_guest(struct kvm *kvm)
+{
+	int i;
+	struct kvm_vcpu *vcpu;
+
+	kvm_for_each_vcpu(i, vcpu, kvm) {
+		wait_queue_head_t *wq = kvm_arch_vcpu_wq(vcpu);
+
+		vcpu->arch.pause = false;
+		wake_up_interruptible(wq);
+	}
+}
+
 static void vcpu_sleep(struct kvm_vcpu *vcpu)
 {
 	wait_queue_head_t *wq = kvm_arch_vcpu_wq(vcpu);
 
-	wait_event_interruptible(*wq, !vcpu->arch.power_off);
+	wait_event_interruptible(*wq, ((!vcpu->arch.power_off) &&
+				       (!vcpu->arch.pause)));
 }
 
 static int kvm_vcpu_initialized(struct kvm_vcpu *vcpu)
@@ -528,7 +555,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 
 		update_vttbr(vcpu->kvm);
 
-		if (vcpu->arch.power_off)
+		if (vcpu->arch.power_off || vcpu->arch.pause)
 			vcpu_sleep(vcpu);
 
 		/*
@@ -556,7 +583,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 		}
 
 		if (ret <= 0 || need_new_vmid_gen(vcpu->kvm) ||
-			vcpu->arch.power_off) {
+			vcpu->arch.power_off || vcpu->arch.pause) {
 			local_irq_enable();
 			kvm_vgic_sync_hwstate(vcpu);
 			preempt_enable();
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 009da6b..69e3785 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -125,6 +125,9 @@ struct kvm_vcpu_arch {
 	/* vcpu power-off state */
 	bool power_off;
 
+	/* Don't run the guest */
+	bool pause;
+
 	/* IO related fields */
 	struct kvm_decode mmio_decode;
 
-- 
1.9.1

[PATCH v2 4/4] KVM: arm/arm64: implement kvm_arm_[halt,resume]_guest

[Christoffer Dall]

On Fri, Aug 07, 2015 at 06:08:33PM +0200, Eric Auger wrote:
> We introduce kvm_arm_halt_guest and resume functions. They
> will be used for IRQ forward state change.
> 
> Halt is synchronous and prevents the guest from being re-entered.
> We use the same mechanism put in place for PSCI former pause,
> now renamed power_off. A new flag is introduced in arch vcpu state,
> pause, only meant to be used by those functions.
> 
> Signed-off-by: Eric Auger <eric.auger@linaro.org>
> 
> ---
> v1 -> v2:
> - check pause in kvm_arch_vcpu_runnable
> - we cannot use kvm_vcpu_block since this latter would exit on
>   IRQ/FIQ and this is not what we want
> ---
>  arch/arm/include/asm/kvm_host.h   |  3 +++
>  arch/arm/kvm/arm.c                | 35 +++++++++++++++++++++++++++++++----
>  arch/arm64/include/asm/kvm_host.h |  3 +++
>  3 files changed, 37 insertions(+), 4 deletions(-)
> 
> diff --git a/arch/arm/include/asm/kvm_host.h b/arch/arm/include/asm/kvm_host.h
> index 304004d..dac85f6 100644
> --- a/arch/arm/include/asm/kvm_host.h
> +++ b/arch/arm/include/asm/kvm_host.h
> @@ -132,6 +132,9 @@ struct kvm_vcpu_arch {
>  	/* vcpu power-off state */
>  	bool power_off;
>  
> +	 /* Exit and don't run the guest (internal implementation need) */

Why exit?  I think it's slightly more correct to just say.
"Don't run the guest (internal implementation need)"

> +	bool pause;
> +
>  	/* IO related fields */
>  	struct kvm_decode mmio_decode;
>  
> diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c
> index cc404a8..0529b38 100644
> --- a/arch/arm/kvm/arm.c
> +++ b/arch/arm/kvm/arm.c
> @@ -348,7 +348,7 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
>  int kvm_arch_vcpu_runnable(struct kvm_vcpu *v)
>  {
>  	return ((!!v->arch.irq_lines || kvm_vgic_vcpu_pending_irq(v))
> -		&& !v->arch.power_off);
> +		&& !v->arch.power_off && !v->arch.pause);
>  }
>  
>  /* Just ensure a guest exit from a particular CPU */
> @@ -474,11 +474,38 @@ bool kvm_arch_intc_initialized(struct kvm *kvm)
>  	return vgic_initialized(kvm);
>  }
>  
> +static void kvm_arm_halt_guest(struct kvm *kvm) __maybe_unused;
> +static void kvm_arm_resume_guest(struct kvm *kvm) __maybe_unused;
> +
> +static void kvm_arm_halt_guest(struct kvm *kvm)
> +{
> +	int i;
> +	struct kvm_vcpu *vcpu;
> +
> +	kvm_for_each_vcpu(i, vcpu, kvm)
> +		vcpu->arch.pause = true;
> +	force_vm_exit(cpu_all_mask);
> +}
> +
> +static void kvm_arm_resume_guest(struct kvm *kvm)
> +{
> +	int i;
> +	struct kvm_vcpu *vcpu;
> +
> +	kvm_for_each_vcpu(i, vcpu, kvm) {
> +		wait_queue_head_t *wq = kvm_arch_vcpu_wq(vcpu);
> +
> +		vcpu->arch.pause = false;
> +		wake_up_interruptible(wq);
> +	}
> +}
> +
>  static void vcpu_sleep(struct kvm_vcpu *vcpu)
>  {
>  	wait_queue_head_t *wq = kvm_arch_vcpu_wq(vcpu);
>  
> -	wait_event_interruptible(*wq, !vcpu->arch.power_off);
> +	wait_event_interruptible(*wq, ((!vcpu->arch.power_off) &&
> +				       (!vcpu->arch.pause)));
>  }
>  
>  static int kvm_vcpu_initialized(struct kvm_vcpu *vcpu)
> @@ -528,7 +555,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
>  
>  		update_vttbr(vcpu->kvm);
>  
> -		if (vcpu->arch.power_off)
> +		if (vcpu->arch.power_off || vcpu->arch.pause)
>  			vcpu_sleep(vcpu);
>  
>  		/*
> @@ -556,7 +583,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
>  		}
>  
>  		if (ret <= 0 || need_new_vmid_gen(vcpu->kvm) ||
> -			vcpu->arch.power_off) {
> +			vcpu->arch.power_off || vcpu->arch.pause) {
>  			local_irq_enable();
>  			kvm_vgic_sync_hwstate(vcpu);
>  			preempt_enable();
> diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
> index 009da6b..69e3785 100644
> --- a/arch/arm64/include/asm/kvm_host.h
> +++ b/arch/arm64/include/asm/kvm_host.h
> @@ -125,6 +125,9 @@ struct kvm_vcpu_arch {
>  	/* vcpu power-off state */
>  	bool power_off;
>  
> +	/* Don't run the guest */

Can we have the same comment on the arm and arm64 version?

> +	bool pause;
> +
>  	/* IO related fields */
>  	struct kvm_decode mmio_decode;
>  
> -- 
> 1.9.1
> 

Besides these commenting nits, I think this looks reasonable overall.

Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>

On the series.