[PATCH] KVM: arm64: nv: Properly check ESR_EL2.VNCR on taking a VNCR_EL2 related fault

[PATCH] KVM: arm64: nv: Properly check ESR_EL2.VNCR on taking a VNCR_EL2 related fault

[Marc Zyngier]

Instead of checking for the ESR_EL2.VNCR bit being set (the only case
we should be here), we are actually testing random bits in ESR_EL2.DFSC.

13 obviously being a lucky number, it matches both permission and
translation fault status codes, which explains why we never saw it
failing. This was found by inspection, while reviewing a vaguely
related patch.

Whilst we're at it, turn the BUG_ON() into a WARN_ON_ONCE(), as
exploding here is just silly.

Fixes: 069a05e535496 ("KVM: arm64: nv: Handle VNCR_EL2-triggered faults")
Signed-off-by: Marc Zyngier <maz@kernel.org>
---
 arch/arm64/kvm/nested.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm64/kvm/nested.c b/arch/arm64/kvm/nested.c
index c6a4e8f384ac6..046dcfc8bf76b 100644
--- a/arch/arm64/kvm/nested.c
+++ b/arch/arm64/kvm/nested.c
@@ -1287,7 +1287,7 @@ int kvm_handle_vncr_abort(struct kvm_vcpu *vcpu)
 	struct vncr_tlb *vt = vcpu->arch.vncr_tlb;
 	u64 esr = kvm_vcpu_get_esr(vcpu);
 
-	BUG_ON(!(esr & ESR_ELx_VNCR_SHIFT));
+	WARN_ON_ONCE(!(esr & ESR_ELx_VNCR));
 
 	if (esr_fsc_is_permission_fault(esr)) {
 		inject_vncr_perm(vcpu);
-- 
2.39.2

Re: [PATCH] KVM: arm64: nv: Properly check ESR_EL2.VNCR on taking a VNCR_EL2 related fault

[Joey Gouly]

On Wed, Jul 30, 2025 at 11:18:28AM +0100, Marc Zyngier wrote:
> Instead of checking for the ESR_EL2.VNCR bit being set (the only case
> we should be here), we are actually testing random bits in ESR_EL2.DFSC.
> 
> 13 obviously being a lucky number, it matches both permission and
> translation fault status codes, which explains why we never saw it
> failing. This was found by inspection, while reviewing a vaguely
> related patch.
> 
> Whilst we're at it, turn the BUG_ON() into a WARN_ON_ONCE(), as
> exploding here is just silly.
> 
> Fixes: 069a05e535496 ("KVM: arm64: nv: Handle VNCR_EL2-triggered faults")
> Signed-off-by: Marc Zyngier <maz@kernel.org>

Good spot!

Reviewed-by: Joey Gouly <joey.gouly@arm.com>

> ---
>  arch/arm64/kvm/nested.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/arch/arm64/kvm/nested.c b/arch/arm64/kvm/nested.c
> index c6a4e8f384ac6..046dcfc8bf76b 100644
> --- a/arch/arm64/kvm/nested.c
> +++ b/arch/arm64/kvm/nested.c
> @@ -1287,7 +1287,7 @@ int kvm_handle_vncr_abort(struct kvm_vcpu *vcpu)
>  	struct vncr_tlb *vt = vcpu->arch.vncr_tlb;
>  	u64 esr = kvm_vcpu_get_esr(vcpu);
>  
> -	BUG_ON(!(esr & ESR_ELx_VNCR_SHIFT));
> +	WARN_ON_ONCE(!(esr & ESR_ELx_VNCR));
>  
>  	if (esr_fsc_is_permission_fault(esr)) {
>  		inject_vncr_perm(vcpu);
> -- 
> 2.39.2
> 

Re: [PATCH] KVM: arm64: nv: Properly check ESR_EL2.VNCR on taking a VNCR_EL2 related fault

[Oliver Upton]

On Wed, 30 Jul 2025 11:18:28 +0100, Marc Zyngier wrote:
> Instead of checking for the ESR_EL2.VNCR bit being set (the only case
> we should be here), we are actually testing random bits in ESR_EL2.DFSC.
> 
> 13 obviously being a lucky number, it matches both permission and
> translation fault status codes, which explains why we never saw it
> failing. This was found by inspection, while reviewing a vaguely
> related patch.
> 
> [...]

Applied to fixes, thanks!

[1/1] KVM: arm64: nv: Properly check ESR_EL2.VNCR on taking a VNCR_EL2 related fault
      https://git.kernel.org/kvmarm/kvmarm/c/07f557f60a9a

--
Best,
Oliver