Re: [PATCH v3 02/11] KVM: arm64: CCA: Introduce Realms

From: Steven Price

Date: Wed Nov 18 2026 - 11:13:32 EST


On 17/11/2026 14:07, Fuad Tabba wrote:
> On Mon, 16 Nov 2026 16:43:48 +0000, Steven Price <steven.price@xxxxxxx> wrote:
> [...]
>> diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
> [...]
>> @@ -1571,31 +1576,39 @@ struct kvm *kvm_arch_alloc_vm(void);
>>
>> #define __KVM_HAVE_ARCH_FLUSH_REMOTE_TLBS_RANGE
>>
>> -#ifdef __KVM_NVHE_HYPERVISOR__
>> +#define kvm_vm_is_protected(kvm) ((kvm)->arch.vm_flavor >= __VM_PROTECTED)
>>
>> -#define kvm_vm_is_protected(kvm) \
>> - (is_protected_kvm_enabled() && ((kvm)->arch.vm_flavor == VM_PROTECTED_PKVM))
>> +#ifdef __KVM_NVHE_HYPERVISOR__
>> /*
>> * Accessing vcpu->kvm from nVHE hyp stub is tricky, as we need to convert the
>> * pointer to the hyp VA. With pKVM, the nVHE code runs with the hyp_vcpu,
>> - * which is populated correctly and is gated on is_protected_kvm_enabled().
>> + * which is populated correctly. Always vcpu_is_protected_pkvm(), which is
>> + * gated on is_protected_kvm_enabled().
>> */
>> -#define vcpu_is_protected(vcpu) \
>> +#define vcpu_is_protected(vcpu) BUILD_BUG_ON(1)
>> +#else
>> +#define vcpu_is_protected(vcpu) kvm_vm_is_protected((vcpu)->kvm)
>> +#endif
>> +
>> +#define kvm_vm_is_protected_pkvm(kvm) \
>> + (is_protected_kvm_enabled() && ((kvm)->arch.vm_flavor == VM_PROTECTED_PKVM))
>> +
>> +/*
>> + * Rely on is_protected_kvm_enabled() check in kvm_vm_is_protected_pkvm() to
>> + * make sure the vcpu->kvm is always valid VA in the context
>> + */
>> +#define vcpu_is_protected_pkvm(vcpu) \
>> ({ \
>> struct kvm *__kvm = READ_ONCE((vcpu)->kvm); \
>> \
>> - (__kvm && kvm_vm_is_protected(__kvm)); \
>> + (__kvm && kvm_vm_is_protected_pkvm(__kvm)); \
>> })
>>
>> -#else
>> -
>> -#define kvm_vm_is_protected(kvm) ((kvm)->arch.vm_flavor >= __VM_PROTECTED)
>> -#define vcpu_is_protected(vcpu) kvm_vm_is_protected((vcpu)->kvm)
>
> This brings back the _pkvm variants that Marc asked to drop in the
> plumbing series [1]. In the nVHE hyp code, kvm_vm_is_protected() and
> vcpu_is_protected() already mean pKVM-protected, so I think the hyp
> side can stay as it is.
>
> At EL1, the pKVM-only checks that a Realm vCPU also reaches are in
> kvm_arm_vcpu_power_off() and kvm_psci_vcpu_on(). Could they check
> vm_flavor == VM_PROTECTED_PKVM directly?

Ah, yes that should work better. Thanks for the suggestion.

>> +#define kvm_vm_is_realm(kvm) ((kvm)->arch.vm_flavor == VM_REALM)
>> +#define vcpu_is_rec(vcpu) kvm_vm_is_realm((vcpu)->kvm)
>
> vcpu_is_rec() isn't used anywhere in this series. Could it be dropped?

Yep, I can move it after this series.

> [...]
>> diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c
> [...]
>> @@ -906,7 +907,8 @@ static bool kvm_arm_vcpu_suspended(struct kvm_vcpu *vcpu)
>> /* Only the guest's CPU_ON may start a vCPU EL2 holds powered off. */
>> bool kvm_pkvm_vcpu_is_powered_off(struct kvm_vcpu *vcpu)
>> {
>> - return READ_ONCE(vcpu->arch.mp_state.mp_state) == KVM_MP_STATE_UNINITIALIZED;
>> + return vcpu_is_protected_pkvm(vcpu) &&
>> + READ_ONCE(vcpu->arch.mp_state.mp_state) == KVM_MP_STATE_UNINITIALIZED;
>
> Only a pVM's vCPU can be in KVM_MP_STATE_UNINITIALIZED, so is this needed?

I guess not. I'll drop this change.

Thanks,
Steve

> Cheers,
> /fuad
>
> [1] https://lore.kernel.org/all/86pkxs2npx.wl-maz@xxxxxxxxxx/