[PATCH] KVM: arm64: Convert ICH_HCR_EL2_TDIR cap to EARLY_LOCAL_CPU_FEATURE

[PATCH] KVM: arm64: Convert ICH_HCR_EL2_TDIR cap to EARLY_LOCAL_CPU_FEATURE

[Marc Zyngier]

Suzuki notices that making the ICH_HCR_EL2_TDIR capability a system
one isn't a very good idea, should we end-up with CPUs that have
asymmetric TDIR support (somehow unlikely, but you never know what
level of stupidity vendors are up to). For this hypothetical setup,
making this an "EARLY_LOCAL_CPU_FEATURE" is a much better option.

This is actually consistent with what we already do with GICv5
legacy interface, so flip the capability over.

Reported-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Fixes: 2a28810cbb8b2 ("KVM: arm64: GICv3: Detect and work around the lack of ICV_DIR_EL1 trapping")
Link: https://lore.kerenl.org/r/5df713d4-8b79-4456-8fd1-707ca89a61b6@arm.com
---
 arch/arm64/kernel/cpufeature.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c
index 5de51cb1b8fe2..75fb9a0efcc8e 100644
--- a/arch/arm64/kernel/cpufeature.c
+++ b/arch/arm64/kernel/cpufeature.c
@@ -2325,14 +2325,14 @@ static bool can_trap_icv_dir_el1(const struct arm64_cpu_capabilities *entry,
 
 	BUILD_BUG_ON(ARM64_HAS_ICH_HCR_EL2_TDIR <= ARM64_HAS_GICV3_CPUIF);
 	BUILD_BUG_ON(ARM64_HAS_ICH_HCR_EL2_TDIR <= ARM64_HAS_GICV5_LEGACY);
-	if (!cpus_have_cap(ARM64_HAS_GICV3_CPUIF) &&
+	if (!this_cpu_has_cap(ARM64_HAS_GICV3_CPUIF) &&
 	    !is_midr_in_range_list(has_vgic_v3))
 		return false;
 
 	if (!is_hyp_mode_available())
 		return false;
 
-	if (cpus_have_cap(ARM64_HAS_GICV5_LEGACY))
+	if (this_cpu_has_cap(ARM64_HAS_GICV5_LEGACY))
 		return true;
 
 	if (is_kernel_in_hyp_mode())
@@ -2863,7 +2863,7 @@ static const struct arm64_cpu_capabilities arm64_features[] = {
 		 */
 		.desc = "ICV_DIR_EL1 trapping",
 		.capability = ARM64_HAS_ICH_HCR_EL2_TDIR,
-		.type = ARM64_CPUCAP_SYSTEM_FEATURE,
+		.type = ARM64_CPUCAP_EARLY_LOCAL_CPU_FEATURE,
 		.matches = can_trap_icv_dir_el1,
 	},
 #ifdef CONFIG_ARM64_E0PD
-- 
2.47.3

Re: [PATCH] KVM: arm64: Convert ICH_HCR_EL2_TDIR cap to EARLY_LOCAL_CPU_FEATURE

[Suzuki K Poulose]

On 25/11/2025 16:01, Marc Zyngier wrote:
> Suzuki notices that making the ICH_HCR_EL2_TDIR capability a system
> one isn't a very good idea, should we end-up with CPUs that have
> asymmetric TDIR support (somehow unlikely, but you never know what
> level of stupidity vendors are up to). For this hypothetical setup,
> making this an "EARLY_LOCAL_CPU_FEATURE" is a much better option.
> 
> This is actually consistent with what we already do with GICv5
> legacy interface, so flip the capability over.
> 
> Reported-by: Suzuki K Poulose <suzuki.poulose@arm.com>
> Signed-off-by: Marc Zyngier <maz@kernel.org>
> Fixes: 2a28810cbb8b2 ("KVM: arm64: GICv3: Detect and work around the lack of ICV_DIR_EL1 trapping")
> Link: https://lore.kerenl.org/r/5df713d4-8b79-4456-8fd1-707ca89a61b6@arm.com

Reviewed-by: Suzuki K Poulose <suzuki.poulose@arm.com>

> ---
>   arch/arm64/kernel/cpufeature.c | 6 +++---
>   1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c
> index 5de51cb1b8fe2..75fb9a0efcc8e 100644
> --- a/arch/arm64/kernel/cpufeature.c
> +++ b/arch/arm64/kernel/cpufeature.c
> @@ -2325,14 +2325,14 @@ static bool can_trap_icv_dir_el1(const struct arm64_cpu_capabilities *entry,
>   
>   	BUILD_BUG_ON(ARM64_HAS_ICH_HCR_EL2_TDIR <= ARM64_HAS_GICV3_CPUIF);
>   	BUILD_BUG_ON(ARM64_HAS_ICH_HCR_EL2_TDIR <= ARM64_HAS_GICV5_LEGACY);
> -	if (!cpus_have_cap(ARM64_HAS_GICV3_CPUIF) &&
> +	if (!this_cpu_has_cap(ARM64_HAS_GICV3_CPUIF) &&
>   	    !is_midr_in_range_list(has_vgic_v3))
>   		return false;
>   
>   	if (!is_hyp_mode_available())
>   		return false;
>   
> -	if (cpus_have_cap(ARM64_HAS_GICV5_LEGACY))
> +	if (this_cpu_has_cap(ARM64_HAS_GICV5_LEGACY))
>   		return true;
>   
>   	if (is_kernel_in_hyp_mode())
> @@ -2863,7 +2863,7 @@ static const struct arm64_cpu_capabilities arm64_features[] = {
>   		 */
>   		.desc = "ICV_DIR_EL1 trapping",
>   		.capability = ARM64_HAS_ICH_HCR_EL2_TDIR,
> -		.type = ARM64_CPUCAP_SYSTEM_FEATURE,
> +		.type = ARM64_CPUCAP_EARLY_LOCAL_CPU_FEATURE,
>   		.matches = can_trap_icv_dir_el1,
>   	},
>   #ifdef CONFIG_ARM64_E0PD

Re: [PATCH] KVM: arm64: Convert ICH_HCR_EL2_TDIR cap to EARLY_LOCAL_CPU_FEATURE

[Oliver Upton]

On Tue, 25 Nov 2025 16:01:44 +0000, Marc Zyngier wrote:
> Suzuki notices that making the ICH_HCR_EL2_TDIR capability a system
> one isn't a very good idea, should we end-up with CPUs that have
> asymmetric TDIR support (somehow unlikely, but you never know what
> level of stupidity vendors are up to). For this hypothetical setup,
> making this an "EARLY_LOCAL_CPU_FEATURE" is a much better option.
> 
> This is actually consistent with what we already do with GICv5
> legacy interface, so flip the capability over.
> 
> [...]

Applied to next, thanks!

[1/1] KVM: arm64: Convert ICH_HCR_EL2_TDIR cap to EARLY_LOCAL_CPU_FEATURE
      https://git.kernel.org/kvmarm/kvmarm/c/64d67e7add10

--
Best,
Oliver