[PATCH V3] arm64/cpufeature: Validate hypervisor capabilities during CPU hotplug

Subsystems: arm64 port (aarch64 architecture), kernel virtual machine for arm64 (kvm/arm64), the rest

STALE2283d LANDED: 5 (5M)

1 review trailer (1 from subsystem maintainers); landed in mainline as c73433fc630c on 2020-05-20.

3 messages, 2 authors, 2020-05-21 · open the first message on its own page

[PATCH V3] arm64/cpufeature: Validate hypervisor capabilities during CPU hotplug

From: Anshuman Khandual <hidden>
Date: 2020-05-12 01:58:08

This validates hypervisor capabilities like VMID width, IPA range for any
hot plug CPU against system finalized values. KVM's view of the IPA space
is used while allowing a given CPU to come up. While here, it factors out
get_vmid_bits() for general use.

Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: kvmarm@lists.cs.columbia.edu
Cc: linux-kernel@vger.kernel.org

Reviewed-by: Marc Zyngier <maz@kernel.org>
Suggested-by: Suzuki Poulose <suzuki.poulose@arm.com>
Signed-off-by: Anshuman Khandual <redacted>
---
Changes in V3:

- Dropped "#ifdef CONFIG_KVM_ARM_HOST" per Marc
- Dropped ID_AA64MMFR0_PARANGE_MASK (for now)
- Updated the commit message per Marc

Changes in V2: (https://patchwork.kernel.org/patch/11535359/)

- Added is_hyp_mode_available() check per Marc
- Moved verify_kvm_capabilities() into cpufeature.c per Marc
- Added helper get_kvm_ipa_limit() to fetch kvm_ipa_limit per Marc
- Renamed kvm as hyp including the commit message per Marc

Changes in V1: (https://patchwork.kernel.org/patch/11532565/)

 arch/arm64/include/asm/cpufeature.h | 18 ++++++++++++++++
 arch/arm64/include/asm/kvm_mmu.h    |  2 +-
 arch/arm64/kernel/cpufeature.c      | 33 +++++++++++++++++++++++++++++
 arch/arm64/kvm/reset.c              |  5 +++++
 4 files changed, 57 insertions(+), 1 deletion(-)
diff --git a/arch/arm64/include/asm/cpufeature.h b/arch/arm64/include/asm/cpufeature.h
index afe08251ff95..1291ad5a9ccb 100644
--- a/arch/arm64/include/asm/cpufeature.h
+++ b/arch/arm64/include/asm/cpufeature.h
@@ -745,6 +745,24 @@ static inline bool cpu_has_hw_af(void)
 extern bool cpu_has_amu_feat(int cpu);
 #endif
 
+static inline unsigned int get_vmid_bits(u64 mmfr1)
+{
+	int vmid_bits;
+
+	vmid_bits = cpuid_feature_extract_unsigned_field(mmfr1,
+						ID_AA64MMFR1_VMIDBITS_SHIFT);
+	if (vmid_bits == ID_AA64MMFR1_VMIDBITS_16)
+		return 16;
+
+	/*
+	 * Return the default here even if any reserved
+	 * value is fetched from the system register.
+	 */
+	return 8;
+}
+
+u32 get_kvm_ipa_limit(void);
+
 #endif /* __ASSEMBLY__ */
 
 #endif
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index 30b0e8d6b895..a7137e144b97 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -416,7 +416,7 @@ static inline unsigned int kvm_get_vmid_bits(void)
 {
 	int reg = read_sanitised_ftr_reg(SYS_ID_AA64MMFR1_EL1);
 
-	return (cpuid_feature_extract_unsigned_field(reg, ID_AA64MMFR1_VMIDBITS_SHIFT) == 2) ? 16 : 8;
+	return get_vmid_bits(reg);
 }
 
 /*
diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c
index 9fac745aa7bb..30917fe7942a 100644
--- a/arch/arm64/kernel/cpufeature.c
+++ b/arch/arm64/kernel/cpufeature.c
@@ -2181,6 +2181,36 @@ static void verify_sve_features(void)
 	/* Add checks on other ZCR bits here if necessary */
 }
 
+#ifdef CONFIG_KVM_ARM_HOST
+void verify_hyp_capabilities(void)
+{
+	u64 safe_mmfr1, mmfr0, mmfr1;
+	int parange, ipa_max;
+	unsigned int safe_vmid_bits, vmid_bits;
+
+	safe_mmfr1 = read_sanitised_ftr_reg(SYS_ID_AA64MMFR1_EL1);
+	mmfr0 = read_cpuid(ID_AA64MMFR0_EL1);
+	mmfr1 = read_cpuid(ID_AA64MMFR1_EL1);
+
+	/* Verify VMID bits */
+	safe_vmid_bits = get_vmid_bits(safe_mmfr1);
+	vmid_bits = get_vmid_bits(mmfr1);
+	if (vmid_bits < safe_vmid_bits) {
+		pr_crit("CPU%d: VMID width mismatch\n", smp_processor_id());
+		cpu_die_early();
+	}
+
+	/* Verify IPA range */
+	parange = mmfr0 & 0x7;
+	ipa_max = id_aa64mmfr0_parange_to_phys_shift(parange);
+	if (ipa_max < get_kvm_ipa_limit()) {
+		pr_crit("CPU%d: IPA range mismatch\n", smp_processor_id());
+		cpu_die_early();
+	}
+}
+#else	/* !CONFIG_KVM_ARM_HOST */
+static inline void verify_hyp_capabilities(void) { }
+#endif	/* CONFIG_KVM_ARM_HOST */
 
 /*
  * Run through the enabled system capabilities and enable() it on this CPU.
@@ -2206,6 +2236,9 @@ static void verify_local_cpu_capabilities(void)
 
 	if (system_supports_sve())
 		verify_sve_features();
+
+	if (is_hyp_mode_available())
+		verify_hyp_capabilities();
 }
 
 void check_local_cpu_capabilities(void)
diff --git a/arch/arm64/kvm/reset.c b/arch/arm64/kvm/reset.c
index 30b7ea680f66..841b492ff334 100644
--- a/arch/arm64/kvm/reset.c
+++ b/arch/arm64/kvm/reset.c
@@ -340,6 +340,11 @@ int kvm_reset_vcpu(struct kvm_vcpu *vcpu)
 	return ret;
 }
 
+u32 get_kvm_ipa_limit(void)
+{
+	return kvm_ipa_limit;
+}
+
 void kvm_set_ipa_limit(void)
 {
 	unsigned int ipa_max, pa_max, va_max, parange;
-- 
2.20.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V3] arm64/cpufeature: Validate hypervisor capabilities during CPU hotplug

From: Will Deacon <will@kernel.org>
Date: 2020-05-20 17:54:49

On Tue, 12 May 2020 07:27:27 +0530, Anshuman Khandual wrote:
This validates hypervisor capabilities like VMID width, IPA range for any
hot plug CPU against system finalized values. KVM's view of the IPA space
is used while allowing a given CPU to come up. While here, it factors out
get_vmid_bits() for general use.

Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: kvmarm@lists.cs.columbia.edu
Cc: linux-kernel@vger.kernel.org
Applied to arm64 (for-next/cpufeature), thanks!

[1/1] arm64/cpufeature: Validate hypervisor capabilities during CPU hotplug
      https://git.kernel.org/arm64/c/c73433fc630c

But please note that I made some changes to verify_hyp_capabilities() so
that it's (a) static and (b) uses IS_ENABLED to avoid the dummy function
definition. I also extended the IS_ENABLED_check so that it doesn't
conflict with the KVM kconfig changes from Fuad. Please shout if you think
I'm still missing something.

Cheers,
-- 
Will

https://fixes.arm64.dev
https://next.arm64.dev
https://will.arm64.dev

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH V3] arm64/cpufeature: Validate hypervisor capabilities during CPU hotplug

From: Anshuman Khandual <hidden>
Date: 2020-05-21 00:48:49


On 05/20/2020 11:24 PM, Will Deacon wrote:
On Tue, 12 May 2020 07:27:27 +0530, Anshuman Khandual wrote:
quoted
This validates hypervisor capabilities like VMID width, IPA range for any
hot plug CPU against system finalized values. KVM's view of the IPA space
is used while allowing a given CPU to come up. While here, it factors out
get_vmid_bits() for general use.

Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: kvmarm@lists.cs.columbia.edu
Cc: linux-kernel@vger.kernel.org
Applied to arm64 (for-next/cpufeature), thanks!

[1/1] arm64/cpufeature: Validate hypervisor capabilities during CPU hotplug
      https://git.kernel.org/arm64/c/c73433fc630c

But please note that I made some changes to verify_hyp_capabilities() so
that it's (a) static and (b) uses IS_ENABLED to avoid the dummy function
definition. I also extended the IS_ENABLED_check so that it doesn't
conflict with the KVM kconfig changes from Fuad. Please shout if you think
I'm still missing something.
Normally I would have expected get_kvm_ipa_limit() to be unavailable without
CONFIG_KVM_ARM_HOST, causing a build failure. But it does compile correctly.
Probably the compiler does not really look for get_kvm_ipa_limit() definition
after an early check and bailout on !IS_ENABLED(CONFIG_KVM_ARM_HOST).

Nonetheless, changes looks good. Thank you.

- Anshuman

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help