[PATCH] KVM: VMX: Refactor to not compare set PI control bits directly to 1

[PATCH] KVM: VMX: Refactor to not compare set PI control bits directly to 1

[Liran Alon]

This is a pure code refactoring.
No semantic change is expected.

Reviewed-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
Signed-off-by: Liran Alon <liran.alon@oracle.com>
---
 arch/x86/kvm/vmx/vmx.c | 18 ++++++++----------
 1 file changed, 8 insertions(+), 10 deletions(-)

diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index e31317fc8c95..92eb4910fe9f 100644
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -5302,6 +5302,11 @@ static void shrink_ple_window(struct kvm_vcpu *vcpu)
 	}
 }
 
+static bool vmx_dy_apicv_has_pending_interrupt(struct kvm_vcpu *vcpu)
+{
+	return pi_test_on(vcpu_to_pi_desc(vcpu));
+}
+
 /*
  * Handler for POSTED_INTERRUPT_WAKEUP_VECTOR.
  */
@@ -5313,9 +5318,7 @@ static void wakeup_handler(void)
 	spin_lock(&per_cpu(blocked_vcpu_on_cpu_lock, cpu));
 	list_for_each_entry(vcpu, &per_cpu(blocked_vcpu_on_cpu, cpu),
 			blocked_vcpu_list) {
-		struct pi_desc *pi_desc = vcpu_to_pi_desc(vcpu);
-
-		if (pi_test_on(pi_desc) == 1)
+		if (vmx_dy_apicv_has_pending_interrupt(vcpu))
 			kvm_vcpu_kick(vcpu);
 	}
 	spin_unlock(&per_cpu(blocked_vcpu_on_cpu_lock, cpu));
@@ -6168,11 +6171,6 @@ static int vmx_sync_pir_to_irr(struct kvm_vcpu *vcpu)
 	return max_irr;
 }
 
-static bool vmx_dy_apicv_has_pending_interrupt(struct kvm_vcpu *vcpu)
-{
-	return pi_test_on(vcpu_to_pi_desc(vcpu));
-}
-
 static void vmx_load_eoi_exitmap(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap)
 {
 	if (!kvm_vcpu_apicv_active(vcpu))
@@ -7336,7 +7334,7 @@ static int pi_pre_block(struct kvm_vcpu *vcpu)
 	do {
 		old.control = new.control = pi_desc->control;
 
-		WARN((pi_desc->sn == 1),
+		WARN(pi_desc->sn,
 		     "Warning: SN field of posted-interrupts "
 		     "is set before blocking\n");
 
@@ -7361,7 +7359,7 @@ static int pi_pre_block(struct kvm_vcpu *vcpu)
 			   new.control) != old.control);
 
 	/* We should not block the vCPU if an interrupt is posted for it.  */
-	if (pi_test_on(pi_desc) == 1)
+	if (pi_test_on(pi_desc))
 		__pi_post_block(vcpu);
 
 	local_irq_enable();
-- 
2.20.1

Re: [PATCH] KVM: VMX: Refactor to not compare set PI control bits directly to 1

[Jim Mattson]

On Mon, Sep 30, 2019 at 5:54 PM Liran Alon <liran.alon@oracle.com> wrote:
>
> This is a pure code refactoring.
> No semantic change is expected.
>
> Reviewed-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
> Signed-off-by: Liran Alon <liran.alon@oracle.com>
Reviewed-by: Jim Mattson <jmattson@google.com>

Re: [PATCH] KVM: VMX: Refactor to not compare set PI control bits directly to 1

[Paolo Bonzini]

On 01/10/19 02:54, Liran Alon wrote:
> This is a pure code refactoring.
> No semantic change is expected.
> 
> Reviewed-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
> Signed-off-by: Liran Alon <liran.alon@oracle.com>
> ---
>  arch/x86/kvm/vmx/vmx.c | 18 ++++++++----------
>  1 file changed, 8 insertions(+), 10 deletions(-)

I'm not sure this is an improvement, for two reasons:

1) the "dy" in vmx_dy_apicv_has_pending_interrupt callback is meant for
directed yield, so it's a bit ugly to call it from wakeup_handler.

2) the wakeup_handler is an interrupt handler specific to posted
interrupts, so it makes sense to check the posted interrupts descriptor.

I wouldn't say ON is a control bit in fact (unlike for example SN or
NV), since it is written by the processor or IOMMU rather than the
hypervisor.

Paolo

> diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
> index e31317fc8c95..92eb4910fe9f 100644
> --- a/arch/x86/kvm/vmx/vmx.c
> +++ b/arch/x86/kvm/vmx/vmx.c
> @@ -5302,6 +5302,11 @@ static void shrink_ple_window(struct kvm_vcpu *vcpu)
>  	}
>  }
>  
> +static bool vmx_dy_apicv_has_pending_interrupt(struct kvm_vcpu *vcpu)
> +{
> +	return pi_test_on(vcpu_to_pi_desc(vcpu));
> +}
> +
>  /*
>   * Handler for POSTED_INTERRUPT_WAKEUP_VECTOR.
>   */
> @@ -5313,9 +5318,7 @@ static void wakeup_handler(void)
>  	spin_lock(&per_cpu(blocked_vcpu_on_cpu_lock, cpu));
>  	list_for_each_entry(vcpu, &per_cpu(blocked_vcpu_on_cpu, cpu),
>  			blocked_vcpu_list) {
> -		struct pi_desc *pi_desc = vcpu_to_pi_desc(vcpu);
> -
> -		if (pi_test_on(pi_desc) == 1)
> +		if (vmx_dy_apicv_has_pending_interrupt(vcpu))
>  			kvm_vcpu_kick(vcpu);
>  	}
>  	spin_unlock(&per_cpu(blocked_vcpu_on_cpu_lock, cpu));
> @@ -6168,11 +6171,6 @@ static int vmx_sync_pir_to_irr(struct kvm_vcpu *vcpu)
>  	return max_irr;
>  }
>  
> -static bool vmx_dy_apicv_has_pending_interrupt(struct kvm_vcpu *vcpu)
> -{
> -	return pi_test_on(vcpu_to_pi_desc(vcpu));
> -}
> -
>  static void vmx_load_eoi_exitmap(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap)
>  {
>  	if (!kvm_vcpu_apicv_active(vcpu))
> @@ -7336,7 +7334,7 @@ static int pi_pre_block(struct kvm_vcpu *vcpu)
>  	do {
>  		old.control = new.control = pi_desc->control;
>  
> -		WARN((pi_desc->sn == 1),
> +		WARN(pi_desc->sn,
>  		     "Warning: SN field of posted-interrupts "
>  		     "is set before blocking\n");
>  
> @@ -7361,7 +7359,7 @@ static int pi_pre_block(struct kvm_vcpu *vcpu)
>  			   new.control) != old.control);
>  
>  	/* We should not block the vCPU if an interrupt is posted for it.  */
> -	if (pi_test_on(pi_desc) == 1)
> +	if (pi_test_on(pi_desc))
>  		__pi_post_block(vcpu);
>  
>  	local_irq_enable();
>