Re: [PATCH v3 11/28] KVM: arm64: Expose FEAT_NV2p1 to NV guests

[Marc Zyngier <maz@kernel.org>]

On Wed, 22 Jul 2026 09:32:52 +0100,
sashiko-bot@kernel.org wrote:
> 
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Medium] The ARM64_HAS_NV2P1 CPU feature capability is checked dynamically to enable FEAT_NV2p1 logic in KVM, but the capability is never registered in the kernel's CPU feature framework, rendering the new feature entirely dead code.
> --
> 
> --- Patch [11]: [PATCH v3 11/28] KVM: arm64: Expose FEAT_NV2p1 to NV guests ---
> commit 2bb48d2c7459805b9d877b250f95afb9c6786a43
> Author: Marc Zyngier <maz@kernel.org>
> 
> KVM: arm64: Expose FEAT_NV2p1 to NV guests
> 
> This commit exposes the FEAT_NV2p1 feature to nested virtualization guests by
> updating the ID_AA64MMFR4_EL1 register limits when the hardware capability is
> present. This intends to reduce the number of traps for NV guests.
> 
> > diff --git a/arch/arm64/kvm/nested.c b/arch/arm64/kvm/nested.c
> > index fb54f6dad995c..82be8bcb07035 100644
> > --- a/arch/arm64/kvm/nested.c
> > +++ b/arch/arm64/kvm/nested.c
> [ ... ]
> > @@ -1738,7 +1738,11 @@ u64 limit_nv_id_reg(struct kvm *kvm, u32 reg, u64 val)
> >  		if (test_bit(KVM_ARM_VCPU_HAS_EL2_E2H0, kvm->arch.vcpu_features)) {
> >  			val = 0;
> >  		} else {
> > -			val = SYS_FIELD_PREP_ENUM(ID_AA64MMFR4_EL1, NV_frac, NV2_ONLY);
> > +			val &= ID_AA64MMFR4_EL1_NV_frac;
> > +			if (cpus_have_final_cap(ARM64_HAS_NV2P1))
> 
> [Severity: Medium]
> Is ARM64_HAS_NV2P1 registered in the kernel's CPU feature framework?
> 
> It appears that the ARM64_HAS_NV2P1 capability is missing from the
> arm64_features array in arch/arm64/kernel/cpufeature.c. Because it is not
> mapped to a hardware ID register field, this capability check in
> limit_nv_id_reg() will always evaluate to false.
> 
> Does this render the newly added FEAT_NV2p1 support unreachable?

See follow-up patch.

	M.

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