KVM: arm64: vgic: Update some comments to improve the code readability

[PATCH] KVM: arm64: vgic: Update some comments to improve the code readability

Posted by Zhiyuan Dai 11 months, 2 weeks ago

This patch updates the comments in the vgic_v3_queue_sgi function
to improve the readability of the code.

Signed-off-by: Zhiyuan Dai <daizhiyuan@phytium.com.cn>
---
 arch/arm64/kvm/vgic/vgic-mmio-v3.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/kvm/vgic/vgic-mmio-v3.c b/arch/arm64/kvm/vgic/vgic-mmio-v3.c
index ae4c0593d114..f3b328373869 100644
--- a/arch/arm64/kvm/vgic/vgic-mmio-v3.c
+++ b/arch/arm64/kvm/vgic/vgic-mmio-v3.c
@@ -1040,8 +1040,13 @@ static void vgic_v3_queue_sgi(struct kvm_vcpu *vcpu, u32 sgi, bool allow_group1)
 			irq->pending_latch = true;
 			vgic_queue_irq_unlock(vcpu->kvm, irq, flags);
 		} else {
-			/* HW SGI? Ask the GIC to inject it */
 			int err;
+
+			/*
+                        * With GICv4.1, vSGI can be directly injected.
+                        * So let's pretend that they are HW interrupts,
+                        * tied to a host IRQ.
+                        */
 			err = irq_set_irqchip_state(irq->host_irq,
 						    IRQCHIP_STATE_PENDING,
 						    true);
-- 
2.25.1

Re: [PATCH] KVM: arm64: vgic: Update some comments to improve the code readability

Posted by Marc Zyngier 11 months, 2 weeks ago

[Fixing Suzuki's address]
On Wed, 08 Jan 2025 10:39:19 +0000,
Zhiyuan Dai <daizhiyuan@phytium.com.cn> wrote:
> 
> This patch updates the comments in the vgic_v3_queue_sgi function
> to improve the readability of the code.
> 
> Signed-off-by: Zhiyuan Dai <daizhiyuan@phytium.com.cn>
> ---
>  arch/arm64/kvm/vgic/vgic-mmio-v3.c | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/arm64/kvm/vgic/vgic-mmio-v3.c b/arch/arm64/kvm/vgic/vgic-mmio-v3.c
> index ae4c0593d114..f3b328373869 100644
> --- a/arch/arm64/kvm/vgic/vgic-mmio-v3.c
> +++ b/arch/arm64/kvm/vgic/vgic-mmio-v3.c
> @@ -1040,8 +1040,13 @@ static void vgic_v3_queue_sgi(struct kvm_vcpu *vcpu, u32 sgi, bool allow_group1)
>  			irq->pending_latch = true;
>  			vgic_queue_irq_unlock(vcpu->kvm, irq, flags);
>  		} else {
> -			/* HW SGI? Ask the GIC to inject it */
>  			int err;
> +
> +			/*
> +                        * With GICv4.1, vSGI can be directly injected.
> +                        * So let's pretend that they are HW interrupts,
> +                        * tied to a host IRQ.
> +                        */
>  			err = irq_set_irqchip_state(irq->host_irq,
>  						    IRQCHIP_STATE_PENDING,
>  						    true);

Do you really see this as an improvement? I don't think we need to
paraphrase the code, which is pretty obvious. If anything, I'd happily
remove the *existing* comment.

	M.

-- 
Without deviation from the norm, progress is not possible.