[PATCH v3 00/22] ARM64 PMU Partitioning

[PATCH v3 00/22] ARM64 PMU Partitioning

[Colton Lewis]

This series creates a new PMU scheme on ARM, a partitioned PMU that
allows reserving a subset of counters for more direct guest access,
significantly reducing overhead. More details, including performance
benchmarks, can be read in the v1 cover letter linked below.

v3:

* Return to using one kernel command line parameter for reserved host
  counters, but set the default value meaning no partitioning to -1 so
  0 isn't ambiguous. Adjust checks for partitioning accordingly.

* Move the function kvm_pmu_partition_supported() back out of the
  driver to avoid subbing has_vhe in 32-bit ARM code.

* Fix the kernel test robot build failures, one with KVM and no PMU,
  and one with PMU and no KVM.

* Scan entire series to remove vestigial changes and update comments.

v2:
https://lore.kernel.org/kvm/20250620221326.1261128-1-coltonlewis@google.com/

v1:
https://lore.kernel.org/kvm/20250602192702.2125115-1-coltonlewis@google.com/

Colton Lewis (21):
  arm64: cpufeature: Add cpucap for HPMN0
  arm64: Generate sign macro for sysreg Enums
  KVM: arm64: Define PMI{CNTR,FILTR}_EL0 as undef_access
  KVM: arm64: Reorganize PMU functions
  perf: arm_pmuv3: Introduce method to partition the PMU
  perf: arm_pmuv3: Generalize counter bitmasks
  perf: arm_pmuv3: Keep out of guest counter partition
  KVM: arm64: Correct kvm_arm_pmu_get_max_counters()
  KVM: arm64: Set up FGT for Partitioned PMU
  KVM: arm64: Writethrough trapped PMEVTYPER register
  KVM: arm64: Use physical PMSELR for PMXEVTYPER if partitioned
  KVM: arm64: Writethrough trapped PMOVS register
  KVM: arm64: Write fast path PMU register handlers
  KVM: arm64: Setup MDCR_EL2 to handle a partitioned PMU
  KVM: arm64: Account for partitioning in PMCR_EL0 access
  KVM: arm64: Context swap Partitioned PMU guest registers
  KVM: arm64: Enforce PMU event filter at vcpu_load()
  perf: arm_pmuv3: Handle IRQs for Partitioned PMU guest counters
  KVM: arm64: Inject recorded guest interrupts
  KVM: arm64: Add ioctl to partition the PMU when supported
  KVM: arm64: selftests: Add test case for partitioned PMU

Marc Zyngier (1):
  KVM: arm64: Cleanup PMU includes

 Documentation/virt/kvm/api.rst                |  21 +
 arch/arm/include/asm/arm_pmuv3.h              |  38 +
 arch/arm64/include/asm/arm_pmuv3.h            |  61 +-
 arch/arm64/include/asm/kvm_host.h             |  18 +-
 arch/arm64/include/asm/kvm_pmu.h              | 100 +++
 arch/arm64/kernel/cpufeature.c                |   8 +
 arch/arm64/kvm/Makefile                       |   2 +-
 arch/arm64/kvm/arm.c                          |  22 +
 arch/arm64/kvm/debug.c                        |  24 +-
 arch/arm64/kvm/hyp/include/hyp/switch.h       | 233 ++++++
 arch/arm64/kvm/pmu-emul.c                     | 674 +----------------
 arch/arm64/kvm/pmu-part.c                     | 378 ++++++++++
 arch/arm64/kvm/pmu.c                          | 702 ++++++++++++++++++
 arch/arm64/kvm/sys_regs.c                     |  79 +-
 arch/arm64/tools/cpucaps                      |   1 +
 arch/arm64/tools/gen-sysreg.awk               |   1 +
 arch/arm64/tools/sysreg                       |   6 +-
 drivers/perf/arm_pmuv3.c                      | 123 ++-
 include/linux/perf/arm_pmu.h                  |   6 +
 include/linux/perf/arm_pmuv3.h                |  14 +-
 include/uapi/linux/kvm.h                      |   4 +
 tools/include/uapi/linux/kvm.h                |   2 +
 .../selftests/kvm/arm64/vpmu_counter_access.c |  62 +-
 virt/kvm/kvm_main.c                           |   1 +
 24 files changed, 1828 insertions(+), 752 deletions(-)
 create mode 100644 arch/arm64/kvm/pmu-part.c


base-commit: 79150772457f4d45e38b842d786240c36bb1f97f
--
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 01/22] arm64: cpufeature: Add cpucap for HPMN0

[Colton Lewis]

Add a capability for FEAT_HPMN0, whether MDCR_EL2.HPMN can specify 0
counters reserved for the guest.

This required changing HPMN0 to an UnsignedEnum in tools/sysreg
because otherwise not all the appropriate macros are generated to add
it to arm64_cpu_capabilities_arm64_features.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/kernel/cpufeature.c | 8 ++++++++
 arch/arm64/tools/cpucaps       | 1 +
 arch/arm64/tools/sysreg        | 6 +++---
 3 files changed, 12 insertions(+), 3 deletions(-)

diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c
index b34044e20128..73a7dac4b6f6 100644
--- a/arch/arm64/kernel/cpufeature.c
+++ b/arch/arm64/kernel/cpufeature.c
@@ -548,6 +548,7 @@ static const struct arm64_ftr_bits ftr_id_mmfr0[] = {
 };
 
 static const struct arm64_ftr_bits ftr_id_aa64dfr0[] = {
+	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64DFR0_EL1_HPMN0_SHIFT, 4, 0),
 	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64DFR0_EL1_DoubleLock_SHIFT, 4, 0),
 	ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64DFR0_EL1_PMSVer_SHIFT, 4, 0),
 	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64DFR0_EL1_CTX_CMPs_SHIFT, 4, 0),
@@ -2896,6 +2897,13 @@ static const struct arm64_cpu_capabilities arm64_features[] = {
 		.matches = has_cpuid_feature,
 		ARM64_CPUID_FIELDS(ID_AA64MMFR0_EL1, FGT, FGT2)
 	},
+	{
+		.desc = "FEAT_HPMN0",
+		.type = ARM64_CPUCAP_SYSTEM_FEATURE,
+		.capability = ARM64_HAS_HPMN0,
+		.matches = has_cpuid_feature,
+		ARM64_CPUID_FIELDS(ID_AA64DFR0_EL1, HPMN0, IMP)
+	},
 #ifdef CONFIG_ARM64_SME
 	{
 		.desc = "Scalable Matrix Extension",
diff --git a/arch/arm64/tools/cpucaps b/arch/arm64/tools/cpucaps
index 10effd4cff6b..5b196ba21629 100644
--- a/arch/arm64/tools/cpucaps
+++ b/arch/arm64/tools/cpucaps
@@ -39,6 +39,7 @@ HAS_GIC_CPUIF_SYSREGS
 HAS_GIC_PRIO_MASKING
 HAS_GIC_PRIO_RELAXED_SYNC
 HAS_HCR_NV1
+HAS_HPMN0
 HAS_HCX
 HAS_LDAPR
 HAS_LPA2
diff --git a/arch/arm64/tools/sysreg b/arch/arm64/tools/sysreg
index 8a8cf6874298..d29742481754 100644
--- a/arch/arm64/tools/sysreg
+++ b/arch/arm64/tools/sysreg
@@ -1531,9 +1531,9 @@ EndEnum
 EndSysreg
 
 Sysreg	ID_AA64DFR0_EL1	3	0	0	5	0
-Enum	63:60	HPMN0
-	0b0000	UNPREDICTABLE
-	0b0001	DEF
+UnsignedEnum	63:60	HPMN0
+	0b0000	NI
+	0b0001	IMP
 EndEnum
 UnsignedEnum	59:56	ExtTrcBuff
 	0b0000	NI
-- 
2.50.0.727.gbf7dc18ff4-goog

Re: [PATCH v3 01/22] arm64: cpufeature: Add cpucap for HPMN0

[Mark Rutland]

On Thu, Jun 26, 2025 at 08:04:37PM +0000, Colton Lewis wrote:
> Add a capability for FEAT_HPMN0, whether MDCR_EL2.HPMN can specify 0
> counters reserved for the guest.
> 
> This required changing HPMN0 to an UnsignedEnum in tools/sysreg
> because otherwise not all the appropriate macros are generated to add
> it to arm64_cpu_capabilities_arm64_features.

I agree it's appropriate to mark ID_AA64DFR0_EL1.HPMN0 as an
UnsignedEnum. It follows the usual ID scheme per ARM DDI 0487 L.a
section D24.1.3, and zero means not present, so it must be unsigned.

Likewise, the value renames (UNPREDICTABLE => NI and DEF => IMP) look
fine to me.

> Signed-off-by: Colton Lewis <coltonlewis@google.com>

I have one minor nit below, but either way:

Acked-by: Mark Rutland <mark.rutland@arm.com>

> ---
>  arch/arm64/kernel/cpufeature.c | 8 ++++++++
>  arch/arm64/tools/cpucaps       | 1 +
>  arch/arm64/tools/sysreg        | 6 +++---
>  3 files changed, 12 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c
> index b34044e20128..73a7dac4b6f6 100644
> --- a/arch/arm64/kernel/cpufeature.c
> +++ b/arch/arm64/kernel/cpufeature.c
> @@ -548,6 +548,7 @@ static const struct arm64_ftr_bits ftr_id_mmfr0[] = {
>  };
>  
>  static const struct arm64_ftr_bits ftr_id_aa64dfr0[] = {
> +	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64DFR0_EL1_HPMN0_SHIFT, 4, 0),
>  	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64DFR0_EL1_DoubleLock_SHIFT, 4, 0),
>  	ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64DFR0_EL1_PMSVer_SHIFT, 4, 0),
>  	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64DFR0_EL1_CTX_CMPs_SHIFT, 4, 0),
> @@ -2896,6 +2897,13 @@ static const struct arm64_cpu_capabilities arm64_features[] = {
>  		.matches = has_cpuid_feature,
>  		ARM64_CPUID_FIELDS(ID_AA64MMFR0_EL1, FGT, FGT2)
>  	},
> +	{
> +		.desc = "FEAT_HPMN0",

Minor nit, but we can drop the "FEAT_" prefix here, for consistency with
other features (e.g. E0PD, FPMR).

Mark.

> +		.type = ARM64_CPUCAP_SYSTEM_FEATURE,
> +		.capability = ARM64_HAS_HPMN0,
> +		.matches = has_cpuid_feature,
> +		ARM64_CPUID_FIELDS(ID_AA64DFR0_EL1, HPMN0, IMP)
> +	},
>  #ifdef CONFIG_ARM64_SME
>  	{
>  		.desc = "Scalable Matrix Extension",
> diff --git a/arch/arm64/tools/cpucaps b/arch/arm64/tools/cpucaps
> index 10effd4cff6b..5b196ba21629 100644
> --- a/arch/arm64/tools/cpucaps
> +++ b/arch/arm64/tools/cpucaps
> @@ -39,6 +39,7 @@ HAS_GIC_CPUIF_SYSREGS
>  HAS_GIC_PRIO_MASKING
>  HAS_GIC_PRIO_RELAXED_SYNC
>  HAS_HCR_NV1
> +HAS_HPMN0
>  HAS_HCX
>  HAS_LDAPR
>  HAS_LPA2
> diff --git a/arch/arm64/tools/sysreg b/arch/arm64/tools/sysreg
> index 8a8cf6874298..d29742481754 100644
> --- a/arch/arm64/tools/sysreg
> +++ b/arch/arm64/tools/sysreg
> @@ -1531,9 +1531,9 @@ EndEnum
>  EndSysreg
>  
>  Sysreg	ID_AA64DFR0_EL1	3	0	0	5	0
> -Enum	63:60	HPMN0
> -	0b0000	UNPREDICTABLE
> -	0b0001	DEF
> +UnsignedEnum	63:60	HPMN0
> +	0b0000	NI
> +	0b0001	IMP
>  EndEnum
>  UnsignedEnum	59:56	ExtTrcBuff
>  	0b0000	NI
> -- 
> 2.50.0.727.gbf7dc18ff4-goog
> 

Re: [PATCH v3 01/22] arm64: cpufeature: Add cpucap for HPMN0

[Colton Lewis]

Mark Rutland <mark.rutland@arm.com> writes:

> On Thu, Jun 26, 2025 at 08:04:37PM +0000, Colton Lewis wrote:
>> Add a capability for FEAT_HPMN0, whether MDCR_EL2.HPMN can specify 0
>> counters reserved for the guest.

>> This required changing HPMN0 to an UnsignedEnum in tools/sysreg
>> because otherwise not all the appropriate macros are generated to add
>> it to arm64_cpu_capabilities_arm64_features.

> I agree it's appropriate to mark ID_AA64DFR0_EL1.HPMN0 as an
> UnsignedEnum. It follows the usual ID scheme per ARM DDI 0487 L.a
> section D24.1.3, and zero means not present, so it must be unsigned.

> Likewise, the value renames (UNPREDICTABLE => NI and DEF => IMP) look
> fine to me.

>> Signed-off-by: Colton Lewis <coltonlewis@google.com>

> I have one minor nit below, but either way:

> Acked-by: Mark Rutland <mark.rutland@arm.com>

Thank you Mark

>> ---
>>   arch/arm64/kernel/cpufeature.c | 8 ++++++++
>>   arch/arm64/tools/cpucaps       | 1 +
>>   arch/arm64/tools/sysreg        | 6 +++---
>>   3 files changed, 12 insertions(+), 3 deletions(-)

>> diff --git a/arch/arm64/kernel/cpufeature.c  
>> b/arch/arm64/kernel/cpufeature.c
>> index b34044e20128..73a7dac4b6f6 100644
>> --- a/arch/arm64/kernel/cpufeature.c
>> +++ b/arch/arm64/kernel/cpufeature.c
>> @@ -548,6 +548,7 @@ static const struct arm64_ftr_bits ftr_id_mmfr0[] = {
>>   };

>>   static const struct arm64_ftr_bits ftr_id_aa64dfr0[] = {
>> +	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE,  
>> ID_AA64DFR0_EL1_HPMN0_SHIFT, 4, 0),
>>   	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE,  
>> ID_AA64DFR0_EL1_DoubleLock_SHIFT, 4, 0),
>>   	ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE,  
>> ID_AA64DFR0_EL1_PMSVer_SHIFT, 4, 0),
>>   	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE,  
>> ID_AA64DFR0_EL1_CTX_CMPs_SHIFT, 4, 0),
>> @@ -2896,6 +2897,13 @@ static const struct arm64_cpu_capabilities  
>> arm64_features[] = {
>>   		.matches = has_cpuid_feature,
>>   		ARM64_CPUID_FIELDS(ID_AA64MMFR0_EL1, FGT, FGT2)
>>   	},
>> +	{
>> +		.desc = "FEAT_HPMN0",

> Minor nit, but we can drop the "FEAT_" prefix here, for consistency with
> other features (e.g. E0PD, FPMR).

> Mark.

Will do.

>> +		.type = ARM64_CPUCAP_SYSTEM_FEATURE,
>> +		.capability = ARM64_HAS_HPMN0,
>> +		.matches = has_cpuid_feature,
>> +		ARM64_CPUID_FIELDS(ID_AA64DFR0_EL1, HPMN0, IMP)
>> +	},
>>   #ifdef CONFIG_ARM64_SME
>>   	{
>>   		.desc = "Scalable Matrix Extension",
>> diff --git a/arch/arm64/tools/cpucaps b/arch/arm64/tools/cpucaps
>> index 10effd4cff6b..5b196ba21629 100644
>> --- a/arch/arm64/tools/cpucaps
>> +++ b/arch/arm64/tools/cpucaps
>> @@ -39,6 +39,7 @@ HAS_GIC_CPUIF_SYSREGS
>>   HAS_GIC_PRIO_MASKING
>>   HAS_GIC_PRIO_RELAXED_SYNC
>>   HAS_HCR_NV1
>> +HAS_HPMN0
>>   HAS_HCX
>>   HAS_LDAPR
>>   HAS_LPA2
>> diff --git a/arch/arm64/tools/sysreg b/arch/arm64/tools/sysreg
>> index 8a8cf6874298..d29742481754 100644
>> --- a/arch/arm64/tools/sysreg
>> +++ b/arch/arm64/tools/sysreg
>> @@ -1531,9 +1531,9 @@ EndEnum
>>   EndSysreg

>>   Sysreg	ID_AA64DFR0_EL1	3	0	0	5	0
>> -Enum	63:60	HPMN0
>> -	0b0000	UNPREDICTABLE
>> -	0b0001	DEF
>> +UnsignedEnum	63:60	HPMN0
>> +	0b0000	NI
>> +	0b0001	IMP
>>   EndEnum
>>   UnsignedEnum	59:56	ExtTrcBuff
>>   	0b0000	NI
>> --
>> 2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 02/22] arm64: Generate sign macro for sysreg Enums

[Colton Lewis]

There's no reason Enums shouldn't be equivalent to UnsignedEnums and
explicitly specify they are unsigned. This will avoid the annoyance I
had with HPMN0.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/tools/gen-sysreg.awk | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm64/tools/gen-sysreg.awk b/arch/arm64/tools/gen-sysreg.awk
index f2a1732cb1f6..fa21a632d9b7 100755
--- a/arch/arm64/tools/gen-sysreg.awk
+++ b/arch/arm64/tools/gen-sysreg.awk
@@ -308,6 +308,7 @@ $1 == "Enum" && (block_current() == "Sysreg" || block_current() == "SysregFields
 	parse_bitdef(reg, field, $2)
 
 	define_field(reg, field, msb, lsb)
+	define_field_sign(reg, field, "false")
 
 	next
 }
-- 
2.50.0.727.gbf7dc18ff4-goog

Re: [PATCH v3 02/22] arm64: Generate sign macro for sysreg Enums

[Ben Horgan]

Hi Colton,

On 6/26/25 21:04, Colton Lewis wrote:
> There's no reason Enums shouldn't be equivalent to UnsignedEnums and
> explicitly specify they are unsigned. This will avoid the annoyance I
> had with HPMN0.
An Enum can be annotated with the field's sign by updating it to 
UnsignedEnum or SignedEnum. This is explained in [1].

With this change ID_AA64PFR1_EL1.MTE_frac would be marked as unsigned 
when it should really be considered signed.

Enum	43:40	MTE_frac
	0b0000	ASYNC
	0b1111	NI
EndEnum


> 
> Signed-off-by: Colton Lewis <coltonlewis@google.com>
> ---
>   arch/arm64/tools/gen-sysreg.awk | 1 +
>   1 file changed, 1 insertion(+)
> 
> diff --git a/arch/arm64/tools/gen-sysreg.awk b/arch/arm64/tools/gen-sysreg.awk
> index f2a1732cb1f6..fa21a632d9b7 100755
> --- a/arch/arm64/tools/gen-sysreg.awk
> +++ b/arch/arm64/tools/gen-sysreg.awk
> @@ -308,6 +308,7 @@ $1 == "Enum" && (block_current() == "Sysreg" || block_current() == "SysregFields
>   	parse_bitdef(reg, field, $2)
>   
>   	define_field(reg, field, msb, lsb)
> +	define_field_sign(reg, field, "false")
>   
>   	next
>   }

Thanks,

Ben

[1] 
https://lore.kernel.org/all/20221207-arm64-sysreg-helpers-v4-1-25b6b3fb9d18@kernel.org/

Re: [PATCH v3 02/22] arm64: Generate sign macro for sysreg Enums

[Colton Lewis]

Hi Ben. Thanks for the review.

Ben Horgan <ben.horgan@arm.com> writes:

> Hi Colton,

> On 6/26/25 21:04, Colton Lewis wrote:
>> There's no reason Enums shouldn't be equivalent to UnsignedEnums and
>> explicitly specify they are unsigned. This will avoid the annoyance I
>> had with HPMN0.
> An Enum can be annotated with the field's sign by updating it to
> UnsignedEnum or SignedEnum. This is explained in [1].

> With this change ID_AA64PFR1_EL1.MTE_frac would be marked as unsigned
> when it should really be considered signed.

> Enum	43:40	MTE_frac
> 	0b0000	ASYNC
> 	0b1111	NI
> EndEnum

Thanks for the explanation. I made this a separate commit because I
considered people might object and HPMN0 is already an UnsignedEnum in
my previous commit.

Do you think it would be a good idea to make plain Enums signed by
default or should I just remove this commit from the series?


>> Signed-off-by: Colton Lewis <coltonlewis@google.com>
>> ---
>>    arch/arm64/tools/gen-sysreg.awk | 1 +
>>    1 file changed, 1 insertion(+)

>> diff --git a/arch/arm64/tools/gen-sysreg.awk  
>> b/arch/arm64/tools/gen-sysreg.awk
>> index f2a1732cb1f6..fa21a632d9b7 100755
>> --- a/arch/arm64/tools/gen-sysreg.awk
>> +++ b/arch/arm64/tools/gen-sysreg.awk
>> @@ -308,6 +308,7 @@ $1 == "Enum" && (block_current() == "Sysreg" ||  
>> block_current() == "SysregFields
>>    	parse_bitdef(reg, field, $2)

>>    	define_field(reg, field, msb, lsb)
>> +	define_field_sign(reg, field, "false")

>>    	next
>>    }

> Thanks,

> Ben

> [1]
> https://lore.kernel.org/all/20221207-arm64-sysreg-helpers-v4-1-25b6b3fb9d18@kernel.org/

Re: [PATCH v3 02/22] arm64: Generate sign macro for sysreg Enums

[Oliver Upton]

On Fri, Jun 27, 2025 at 08:45:38PM +0000, Colton Lewis wrote:
> Hi Ben. Thanks for the review.
> 
> Ben Horgan <ben.horgan@arm.com> writes:
> 
> > Hi Colton,
> 
> > On 6/26/25 21:04, Colton Lewis wrote:
> > > There's no reason Enums shouldn't be equivalent to UnsignedEnums and
> > > explicitly specify they are unsigned. This will avoid the annoyance I
> > > had with HPMN0.
> > An Enum can be annotated with the field's sign by updating it to
> > UnsignedEnum or SignedEnum. This is explained in [1].
> 
> > With this change ID_AA64PFR1_EL1.MTE_frac would be marked as unsigned
> > when it should really be considered signed.
> 
> > Enum	43:40	MTE_frac
> > 	0b0000	ASYNC
> > 	0b1111	NI
> > EndEnum
> 
> Thanks for the explanation. I made this a separate commit because I
> considered people might object and HPMN0 is already an UnsignedEnum in
> my previous commit.
> 
> Do you think it would be a good idea to make plain Enums signed by
> default or should I just remove this commit from the series?

It is presumptive to associate a sign with an enumeration. Generally
speaking, the only fields that can do signed / unsigned comparisons are
the Feature ID register fields.

So please drop this and only keep the change for HPMN0.

Thanks,
Oliver

> > > Signed-off-by: Colton Lewis <coltonlewis@google.com>
> > > ---
> > >    arch/arm64/tools/gen-sysreg.awk | 1 +
> > >    1 file changed, 1 insertion(+)
> 
> > > diff --git a/arch/arm64/tools/gen-sysreg.awk
> > > b/arch/arm64/tools/gen-sysreg.awk
> > > index f2a1732cb1f6..fa21a632d9b7 100755
> > > --- a/arch/arm64/tools/gen-sysreg.awk
> > > +++ b/arch/arm64/tools/gen-sysreg.awk
> > > @@ -308,6 +308,7 @@ $1 == "Enum" && (block_current() == "Sysreg" ||
> > > block_current() == "SysregFields
> > >    	parse_bitdef(reg, field, $2)
> 
> > >    	define_field(reg, field, msb, lsb)
> > > +	define_field_sign(reg, field, "false")
> 
> > >    	next
> > >    }
> 
> > Thanks,
> 
> > Ben
> 
> > [1]
> > https://lore.kernel.org/all/20221207-arm64-sysreg-helpers-v4-1-25b6b3fb9d18@kernel.org/

Re: [PATCH v3 02/22] arm64: Generate sign macro for sysreg Enums

[Colton Lewis]

Oliver Upton <oliver.upton@linux.dev> writes:

> On Fri, Jun 27, 2025 at 08:45:38PM +0000, Colton Lewis wrote:
>> Hi Ben. Thanks for the review.

>> Ben Horgan <ben.horgan@arm.com> writes:

>> > Hi Colton,

>> > On 6/26/25 21:04, Colton Lewis wrote:
>> > > There's no reason Enums shouldn't be equivalent to UnsignedEnums and
>> > > explicitly specify they are unsigned. This will avoid the annoyance I
>> > > had with HPMN0.
>> > An Enum can be annotated with the field's sign by updating it to
>> > UnsignedEnum or SignedEnum. This is explained in [1].

>> > With this change ID_AA64PFR1_EL1.MTE_frac would be marked as unsigned
>> > when it should really be considered signed.

>> > Enum	43:40	MTE_frac
>> > 	0b0000	ASYNC
>> > 	0b1111	NI
>> > EndEnum

>> Thanks for the explanation. I made this a separate commit because I
>> considered people might object and HPMN0 is already an UnsignedEnum in
>> my previous commit.

>> Do you think it would be a good idea to make plain Enums signed by
>> default or should I just remove this commit from the series?

> It is presumptive to associate a sign with an enumeration. Generally
> speaking, the only fields that can do signed / unsigned comparisons are
> the Feature ID register fields.

> So please drop this and only keep the change for HPMN0.

Done.


> Thanks,
> Oliver

>> > > Signed-off-by: Colton Lewis <coltonlewis@google.com>
>> > > ---
>> > >    arch/arm64/tools/gen-sysreg.awk | 1 +
>> > >    1 file changed, 1 insertion(+)

>> > > diff --git a/arch/arm64/tools/gen-sysreg.awk
>> > > b/arch/arm64/tools/gen-sysreg.awk
>> > > index f2a1732cb1f6..fa21a632d9b7 100755
>> > > --- a/arch/arm64/tools/gen-sysreg.awk
>> > > +++ b/arch/arm64/tools/gen-sysreg.awk
>> > > @@ -308,6 +308,7 @@ $1 == "Enum" && (block_current() == "Sysreg" ||
>> > > block_current() == "SysregFields
>> > >    	parse_bitdef(reg, field, $2)

>> > >    	define_field(reg, field, msb, lsb)
>> > > +	define_field_sign(reg, field, "false")

>> > >    	next
>> > >    }

>> > Thanks,

>> > Ben

>> > [1]
>> >  
>> https://lore.kernel.org/all/20221207-arm64-sysreg-helpers-v4-1-25b6b3fb9d18@kernel.org/

Re: [PATCH v3 02/22] arm64: Generate sign macro for sysreg Enums

[Marc Zyngier]

On Thu, 26 Jun 2025 21:04:38 +0100,
Colton Lewis <coltonlewis@google.com> wrote:
> 
> There's no reason Enums shouldn't be equivalent to UnsignedEnums and
> explicitly specify they are unsigned. This will avoid the annoyance I
> had with HPMN0.

And randomly break unsuspecting cases for which a value is neither
signed not unsigned, but an actual *enumeration*.

If you have issues with HPMN0, start by explaining those, because
ranting in a commit message doesn't help much.

	M.

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

Re: [PATCH v3 02/22] arm64: Generate sign macro for sysreg Enums

[Mark Rutland]

On Thu, Jun 26, 2025 at 08:04:38PM +0000, Colton Lewis wrote:
> There's no reason Enums shouldn't be equivalent to UnsignedEnums and
> explicitly specify they are unsigned. This will avoid the annoyance I
> had with HPMN0.
> 
> Signed-off-by: Colton Lewis <coltonlewis@google.com>

I realise you've already agreed to drop this, so the below is mostly for
the sake of the archive.

We deliberately kept Enum separate from SignedEnum and UnsignedEnum as
there are enumerated ID fields which have no ordering, and are neither
signed nor unsigned (e.g. CTR_EL0.L1Ip).

That was discussed previously at:

  https://lore.kernel.org/all/Y5MrVC3d8MPhvshE@FVFF77S0Q05N/

We could consider adding a comment to that effect in gen-sysreg.awk, but
that needn't be part of this series.

Mark.

> ---
>  arch/arm64/tools/gen-sysreg.awk | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/arch/arm64/tools/gen-sysreg.awk b/arch/arm64/tools/gen-sysreg.awk
> index f2a1732cb1f6..fa21a632d9b7 100755
> --- a/arch/arm64/tools/gen-sysreg.awk
> +++ b/arch/arm64/tools/gen-sysreg.awk
> @@ -308,6 +308,7 @@ $1 == "Enum" && (block_current() == "Sysreg" || block_current() == "SysregFields
>  	parse_bitdef(reg, field, $2)
>  
>  	define_field(reg, field, msb, lsb)
> +	define_field_sign(reg, field, "false")
>  
>  	next
>  }
> -- 
> 2.50.0.727.gbf7dc18ff4-goog
> 

[PATCH v3 03/22] KVM: arm64: Define PMI{CNTR,FILTR}_EL0 as undef_access

[Colton Lewis]

Because KVM isn't fully prepared to support these yet even though the
host PMUv3 driver does, define them as undef_access for now.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/kvm/sys_regs.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
index 76c2f0da821f..99fdbe174202 100644
--- a/arch/arm64/kvm/sys_regs.c
+++ b/arch/arm64/kvm/sys_regs.c
@@ -3092,6 +3092,9 @@ static const struct sys_reg_desc sys_reg_descs[] = {
 	{ SYS_DESC(SYS_SVCR), undef_access, reset_val, SVCR, 0, .visibility = sme_visibility  },
 	{ SYS_DESC(SYS_FPMR), undef_access, reset_val, FPMR, 0, .visibility = fp8_visibility },
 
+	{ SYS_DESC(SYS_PMICNTR_EL0), undef_access },
+	{ SYS_DESC(SYS_PMICFILTR_EL0), undef_access },
+
 	{ PMU_SYS_REG(PMCR_EL0), .access = access_pmcr, .reset = reset_pmcr,
 	  .reg = PMCR_EL0, .get_user = get_pmcr, .set_user = set_pmcr },
 	{ PMU_SYS_REG(PMCNTENSET_EL0),
-- 
2.50.0.727.gbf7dc18ff4-goog

Re: [PATCH v3 03/22] KVM: arm64: Define PMI{CNTR,FILTR}_EL0 as undef_access

[Marc Zyngier]

On Thu, 26 Jun 2025 21:04:39 +0100,
Colton Lewis <coltonlewis@google.com> wrote:
> 
> Because KVM isn't fully prepared to support these yet even though the
> host PMUv3 driver does, define them as undef_access for now.
> 
> Signed-off-by: Colton Lewis <coltonlewis@google.com>
> ---
>  arch/arm64/kvm/sys_regs.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
> index 76c2f0da821f..99fdbe174202 100644
> --- a/arch/arm64/kvm/sys_regs.c
> +++ b/arch/arm64/kvm/sys_regs.c
> @@ -3092,6 +3092,9 @@ static const struct sys_reg_desc sys_reg_descs[] = {
>  	{ SYS_DESC(SYS_SVCR), undef_access, reset_val, SVCR, 0, .visibility = sme_visibility  },
>  	{ SYS_DESC(SYS_FPMR), undef_access, reset_val, FPMR, 0, .visibility = fp8_visibility },
>  
> +	{ SYS_DESC(SYS_PMICNTR_EL0), undef_access },

$ jq -r --arg FEAT "FEAT_PMUv3_ICNTR" -f ./dumpfeat.jq Features.json
(FEAT_PMUv3_ICNTR --> v8Ap8)
(FEAT_PMUv3_ICNTR --> FEAT_PMUv3p9)
((FEAT_PMUv3_ICNTR && FEAT_AA64EL2) --> FEAT_FGT2)

If you have FEAT_PMUv3_ICNTR, then you have FEAT_FGT2. If you have
FEAT_FGT2, then we already trap and UNDEF PMICNTR_EL0 without any
further handling since 4bc0fe0898406 ("KVM: arm64: Add sanitisation
for FEAT_FGT2 registers").

> +	{ SYS_DESC(SYS_PMICFILTR_EL0), undef_access },

Same thing.
> +
>  	{ PMU_SYS_REG(PMCR_EL0), .access = access_pmcr, .reset = reset_pmcr,
>  	  .reg = PMCR_EL0, .get_user = get_pmcr, .set_user = set_pmcr },
>  	{ PMU_SYS_REG(PMCNTENSET_EL0),

So none of this is actually required.

	M.

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

Re: [PATCH v3 03/22] KVM: arm64: Define PMI{CNTR,FILTR}_EL0 as undef_access

[Colton Lewis]

Hi Marc. Thanks for the review.

Marc Zyngier <maz@kernel.org> writes:

> On Thu, 26 Jun 2025 21:04:39 +0100,
> Colton Lewis <coltonlewis@google.com> wrote:

>> Because KVM isn't fully prepared to support these yet even though the
>> host PMUv3 driver does, define them as undef_access for now.

>> Signed-off-by: Colton Lewis <coltonlewis@google.com>
>> ---
>>   arch/arm64/kvm/sys_regs.c | 3 +++
>>   1 file changed, 3 insertions(+)

>> diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
>> index 76c2f0da821f..99fdbe174202 100644
>> --- a/arch/arm64/kvm/sys_regs.c
>> +++ b/arch/arm64/kvm/sys_regs.c
>> @@ -3092,6 +3092,9 @@ static const struct sys_reg_desc sys_reg_descs[] =  
>> {
>>   	{ SYS_DESC(SYS_SVCR), undef_access, reset_val, SVCR, 0, .visibility =  
>> sme_visibility  },
>>   	{ SYS_DESC(SYS_FPMR), undef_access, reset_val, FPMR, 0, .visibility =  
>> fp8_visibility },

>> +	{ SYS_DESC(SYS_PMICNTR_EL0), undef_access },

> $ jq -r --arg FEAT "FEAT_PMUv3_ICNTR" -f ./dumpfeat.jq Features.json
> (FEAT_PMUv3_ICNTR --> v8Ap8)
> (FEAT_PMUv3_ICNTR --> FEAT_PMUv3p9)
> ((FEAT_PMUv3_ICNTR && FEAT_AA64EL2) --> FEAT_FGT2)

> If you have FEAT_PMUv3_ICNTR, then you have FEAT_FGT2. If you have
> FEAT_FGT2, then we already trap and UNDEF PMICNTR_EL0 without any
> further handling since 4bc0fe0898406 ("KVM: arm64: Add sanitisation
> for FEAT_FGT2 registers").

>> +	{ SYS_DESC(SYS_PMICFILTR_EL0), undef_access },

> Same thing.
>> +
>>   	{ PMU_SYS_REG(PMCR_EL0), .access = access_pmcr, .reset = reset_pmcr,
>>   	  .reg = PMCR_EL0, .get_user = get_pmcr, .set_user = set_pmcr },
>>   	{ PMU_SYS_REG(PMCNTENSET_EL0),

> So none of this is actually required.

Thanks for the context. I'll take this patch out.

[PATCH v3 04/22] KVM: arm64: Cleanup PMU includes

[Colton Lewis]

From: Marc Zyngier <maz@kernel.org>

Reorganize these tangled headers.

* Respect the move defining the interface between KVM and PMU in its
  own header asm/kvm_pmu.h

* Define an empty struct arm_pmu so it is defined for those interface
  functions when compiling with CONFIG_KVM but not CONFIG_ARM_PMU

Signed-off-by: Marc Zyngier <maz@kernel.org>
Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/include/asm/arm_pmuv3.h      |  2 +-
 arch/arm64/include/asm/kvm_host.h       | 15 +--------------
 arch/arm64/include/asm/kvm_pmu.h        | 15 +++++++++++++++
 arch/arm64/kvm/debug.c                  |  1 +
 arch/arm64/kvm/hyp/include/hyp/switch.h |  1 +
 arch/arm64/kvm/pmu.c                    |  2 ++
 arch/arm64/kvm/sys_regs.c               |  1 +
 include/linux/perf/arm_pmu.h            |  5 +++++
 virt/kvm/kvm_main.c                     |  1 +
 9 files changed, 28 insertions(+), 15 deletions(-)

diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h
index 8a777dec8d88..32c003a7b810 100644
--- a/arch/arm64/include/asm/arm_pmuv3.h
+++ b/arch/arm64/include/asm/arm_pmuv3.h
@@ -6,7 +6,7 @@
 #ifndef __ASM_PMUV3_H
 #define __ASM_PMUV3_H
 
-#include <asm/kvm_host.h>
+#include <asm/kvm_pmu.h>
 
 #include <asm/cpufeature.h>
 #include <asm/sysreg.h>
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 27ed26bd4381..2df76689381a 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -14,6 +14,7 @@
 #include <linux/arm-smccc.h>
 #include <linux/bitmap.h>
 #include <linux/types.h>
+#include <linux/irq_work.h>
 #include <linux/jump_label.h>
 #include <linux/kvm_types.h>
 #include <linux/maple_tree.h>
@@ -1487,25 +1488,11 @@ void kvm_arch_vcpu_ctxflush_fp(struct kvm_vcpu *vcpu);
 void kvm_arch_vcpu_ctxsync_fp(struct kvm_vcpu *vcpu);
 void kvm_arch_vcpu_put_fp(struct kvm_vcpu *vcpu);
 
-static inline bool kvm_pmu_counter_deferred(struct perf_event_attr *attr)
-{
-	return (!has_vhe() && attr->exclude_host);
-}
-
 #ifdef CONFIG_KVM
-void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr);
-void kvm_clr_pmu_events(u64 clr);
-bool kvm_set_pmuserenr(u64 val);
 void kvm_enable_trbe(void);
 void kvm_disable_trbe(void);
 void kvm_tracing_set_el1_configuration(u64 trfcr_while_in_guest);
 #else
-static inline void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr) {}
-static inline void kvm_clr_pmu_events(u64 clr) {}
-static inline bool kvm_set_pmuserenr(u64 val)
-{
-	return false;
-}
 static inline void kvm_enable_trbe(void) {}
 static inline void kvm_disable_trbe(void) {}
 static inline void kvm_tracing_set_el1_configuration(u64 trfcr_while_in_guest) {}
diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
index baf028d19dfc..ad3247b46838 100644
--- a/arch/arm64/include/asm/kvm_pmu.h
+++ b/arch/arm64/include/asm/kvm_pmu.h
@@ -11,9 +11,15 @@
 #include <linux/kvm_types.h>
 #include <linux/perf_event.h>
 #include <linux/perf/arm_pmuv3.h>
+#include <linux/perf/arm_pmu.h>
 
 #define KVM_ARMV8_PMU_MAX_COUNTERS	32
 
+#define kvm_pmu_counter_deferred(attr)			\
+	({						\
+		!has_vhe() && (attr)->exclude_host;	\
+	})
+
 #if IS_ENABLED(CONFIG_HW_PERF_EVENTS) && IS_ENABLED(CONFIG_KVM)
 struct kvm_pmc {
 	u8 idx;	/* index into the pmu->pmc array */
@@ -68,6 +74,9 @@ int kvm_arm_pmu_v3_has_attr(struct kvm_vcpu *vcpu,
 int kvm_arm_pmu_v3_enable(struct kvm_vcpu *vcpu);
 
 struct kvm_pmu_events *kvm_get_pmu_events(void);
+void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr);
+void kvm_clr_pmu_events(u64 clr);
+bool kvm_set_pmuserenr(u64 val);
 void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu);
 void kvm_vcpu_pmu_restore_host(struct kvm_vcpu *vcpu);
 void kvm_vcpu_pmu_resync_el0(void);
@@ -161,6 +170,12 @@ static inline u64 kvm_pmu_get_pmceid(struct kvm_vcpu *vcpu, bool pmceid1)
 
 #define kvm_vcpu_has_pmu(vcpu)		({ false; })
 static inline void kvm_pmu_update_vcpu_events(struct kvm_vcpu *vcpu) {}
+static inline void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr) {}
+static inline void kvm_clr_pmu_events(u64 clr) {}
+static inline bool kvm_set_pmuserenr(u64 val)
+{
+	return false;
+}
 static inline void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu) {}
 static inline void kvm_vcpu_pmu_restore_host(struct kvm_vcpu *vcpu) {}
 static inline void kvm_vcpu_reload_pmu(struct kvm_vcpu *vcpu) {}
diff --git a/arch/arm64/kvm/debug.c b/arch/arm64/kvm/debug.c
index 1a7dab333f55..a554c3e368dc 100644
--- a/arch/arm64/kvm/debug.c
+++ b/arch/arm64/kvm/debug.c
@@ -9,6 +9,7 @@
 
 #include <linux/kvm_host.h>
 #include <linux/hw_breakpoint.h>
+#include <linux/perf/arm_pmuv3.h>
 
 #include <asm/debug-monitors.h>
 #include <asm/kvm_asm.h>
diff --git a/arch/arm64/kvm/hyp/include/hyp/switch.h b/arch/arm64/kvm/hyp/include/hyp/switch.h
index 7599844908c0..825b81749972 100644
--- a/arch/arm64/kvm/hyp/include/hyp/switch.h
+++ b/arch/arm64/kvm/hyp/include/hyp/switch.h
@@ -14,6 +14,7 @@
 #include <linux/kvm_host.h>
 #include <linux/types.h>
 #include <linux/jump_label.h>
+#include <linux/perf/arm_pmuv3.h>
 #include <uapi/linux/psci.h>
 
 #include <asm/barrier.h>
diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c
index 6b48a3d16d0d..8bfc6b0a85f6 100644
--- a/arch/arm64/kvm/pmu.c
+++ b/arch/arm64/kvm/pmu.c
@@ -8,6 +8,8 @@
 #include <linux/perf/arm_pmu.h>
 #include <linux/perf/arm_pmuv3.h>
 
+#include <asm/kvm_pmu.h>
+
 static DEFINE_PER_CPU(struct kvm_pmu_events, kvm_pmu_events);
 
 /*
diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
index 99fdbe174202..eaff6d63ef77 100644
--- a/arch/arm64/kvm/sys_regs.c
+++ b/arch/arm64/kvm/sys_regs.c
@@ -18,6 +18,7 @@
 #include <linux/printk.h>
 #include <linux/uaccess.h>
 #include <linux/irqchip/arm-gic-v3.h>
+#include <linux/perf/arm_pmuv3.h>
 
 #include <asm/arm_pmuv3.h>
 #include <asm/cacheflush.h>
diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h
index 6dc5e0cd76ca..fb382bcd4f4b 100644
--- a/include/linux/perf/arm_pmu.h
+++ b/include/linux/perf/arm_pmu.h
@@ -187,6 +187,11 @@ void armpmu_free_irq(int irq, int cpu);
 
 #define ARMV8_PMU_PDEV_NAME "armv8-pmu"
 
+#else
+
+struct arm_pmu {
+};
+
 #endif /* CONFIG_ARM_PMU */
 
 #define ARMV8_SPE_PDEV_NAME "arm,spe-v1"
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index e2f6344256ce..25259fcf3115 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -48,6 +48,7 @@
 #include <linux/lockdep.h>
 #include <linux/kthread.h>
 #include <linux/suspend.h>
+#include <linux/perf_event.h>
 
 #include <asm/processor.h>
 #include <asm/ioctl.h>
-- 
2.50.0.727.gbf7dc18ff4-goog

Re: [PATCH v3 04/22] KVM: arm64: Cleanup PMU includes

[Mark Rutland]

On Thu, Jun 26, 2025 at 08:04:40PM +0000, Colton Lewis wrote:
> From: Marc Zyngier <maz@kernel.org>
> 
> Reorganize these tangled headers.

In prior versions of this series, there was some earlier text explaining that
there was a circular header dependency. In v2:

  https://lore.kernel.org/linux-arm-kernel/20250620221326.1261128-6-coltonlewis@google.com/

... it said:

| asm/kvm_host.h includes asm/arm_pmu.h which includes perf/arm_pmuv3.h
| which includes asm/arm_pmuv3.h which includes asm/kvm_host.h This
| causes compilation problems why trying to use anything defined in any
| of the headers in any other headers.

Was there some reason for dropping that? It's a bit odd to start without a
problem statement.

> 
> * Respect the move defining the interface between KVM and PMU in its
>   own header asm/kvm_pmu.h
> 
> * Define an empty struct arm_pmu so it is defined for those interface
>   functions when compiling with CONFIG_KVM but not CONFIG_ARM_PMU

Which functions in particular are those? What prevents them from depending on
CONFIG_ARM_PMU?

Mark.

> Signed-off-by: Marc Zyngier <maz@kernel.org>
> Signed-off-by: Colton Lewis <coltonlewis@google.com>
> ---
>  arch/arm64/include/asm/arm_pmuv3.h      |  2 +-
>  arch/arm64/include/asm/kvm_host.h       | 15 +--------------
>  arch/arm64/include/asm/kvm_pmu.h        | 15 +++++++++++++++
>  arch/arm64/kvm/debug.c                  |  1 +
>  arch/arm64/kvm/hyp/include/hyp/switch.h |  1 +
>  arch/arm64/kvm/pmu.c                    |  2 ++
>  arch/arm64/kvm/sys_regs.c               |  1 +
>  include/linux/perf/arm_pmu.h            |  5 +++++
>  virt/kvm/kvm_main.c                     |  1 +
>  9 files changed, 28 insertions(+), 15 deletions(-)
> 
> diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h
> index 8a777dec8d88..32c003a7b810 100644
> --- a/arch/arm64/include/asm/arm_pmuv3.h
> +++ b/arch/arm64/include/asm/arm_pmuv3.h
> @@ -6,7 +6,7 @@
>  #ifndef __ASM_PMUV3_H
>  #define __ASM_PMUV3_H
>  
> -#include <asm/kvm_host.h>
> +#include <asm/kvm_pmu.h>
>  
>  #include <asm/cpufeature.h>
>  #include <asm/sysreg.h>
> diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
> index 27ed26bd4381..2df76689381a 100644
> --- a/arch/arm64/include/asm/kvm_host.h
> +++ b/arch/arm64/include/asm/kvm_host.h
> @@ -14,6 +14,7 @@
>  #include <linux/arm-smccc.h>
>  #include <linux/bitmap.h>
>  #include <linux/types.h>
> +#include <linux/irq_work.h>
>  #include <linux/jump_label.h>
>  #include <linux/kvm_types.h>
>  #include <linux/maple_tree.h>
> @@ -1487,25 +1488,11 @@ void kvm_arch_vcpu_ctxflush_fp(struct kvm_vcpu *vcpu);
>  void kvm_arch_vcpu_ctxsync_fp(struct kvm_vcpu *vcpu);
>  void kvm_arch_vcpu_put_fp(struct kvm_vcpu *vcpu);
>  
> -static inline bool kvm_pmu_counter_deferred(struct perf_event_attr *attr)
> -{
> -	return (!has_vhe() && attr->exclude_host);
> -}
> -
>  #ifdef CONFIG_KVM
> -void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr);
> -void kvm_clr_pmu_events(u64 clr);
> -bool kvm_set_pmuserenr(u64 val);
>  void kvm_enable_trbe(void);
>  void kvm_disable_trbe(void);
>  void kvm_tracing_set_el1_configuration(u64 trfcr_while_in_guest);
>  #else
> -static inline void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr) {}
> -static inline void kvm_clr_pmu_events(u64 clr) {}
> -static inline bool kvm_set_pmuserenr(u64 val)
> -{
> -	return false;
> -}
>  static inline void kvm_enable_trbe(void) {}
>  static inline void kvm_disable_trbe(void) {}
>  static inline void kvm_tracing_set_el1_configuration(u64 trfcr_while_in_guest) {}
> diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
> index baf028d19dfc..ad3247b46838 100644
> --- a/arch/arm64/include/asm/kvm_pmu.h
> +++ b/arch/arm64/include/asm/kvm_pmu.h
> @@ -11,9 +11,15 @@
>  #include <linux/kvm_types.h>
>  #include <linux/perf_event.h>
>  #include <linux/perf/arm_pmuv3.h>
> +#include <linux/perf/arm_pmu.h>
>  
>  #define KVM_ARMV8_PMU_MAX_COUNTERS	32
>  
> +#define kvm_pmu_counter_deferred(attr)			\
> +	({						\
> +		!has_vhe() && (attr)->exclude_host;	\
> +	})
> +
>  #if IS_ENABLED(CONFIG_HW_PERF_EVENTS) && IS_ENABLED(CONFIG_KVM)
>  struct kvm_pmc {
>  	u8 idx;	/* index into the pmu->pmc array */
> @@ -68,6 +74,9 @@ int kvm_arm_pmu_v3_has_attr(struct kvm_vcpu *vcpu,
>  int kvm_arm_pmu_v3_enable(struct kvm_vcpu *vcpu);
>  
>  struct kvm_pmu_events *kvm_get_pmu_events(void);
> +void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr);
> +void kvm_clr_pmu_events(u64 clr);
> +bool kvm_set_pmuserenr(u64 val);
>  void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu);
>  void kvm_vcpu_pmu_restore_host(struct kvm_vcpu *vcpu);
>  void kvm_vcpu_pmu_resync_el0(void);
> @@ -161,6 +170,12 @@ static inline u64 kvm_pmu_get_pmceid(struct kvm_vcpu *vcpu, bool pmceid1)
>  
>  #define kvm_vcpu_has_pmu(vcpu)		({ false; })
>  static inline void kvm_pmu_update_vcpu_events(struct kvm_vcpu *vcpu) {}
> +static inline void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr) {}
> +static inline void kvm_clr_pmu_events(u64 clr) {}
> +static inline bool kvm_set_pmuserenr(u64 val)
> +{
> +	return false;
> +}
>  static inline void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu) {}
>  static inline void kvm_vcpu_pmu_restore_host(struct kvm_vcpu *vcpu) {}
>  static inline void kvm_vcpu_reload_pmu(struct kvm_vcpu *vcpu) {}
> diff --git a/arch/arm64/kvm/debug.c b/arch/arm64/kvm/debug.c
> index 1a7dab333f55..a554c3e368dc 100644
> --- a/arch/arm64/kvm/debug.c
> +++ b/arch/arm64/kvm/debug.c
> @@ -9,6 +9,7 @@
>  
>  #include <linux/kvm_host.h>
>  #include <linux/hw_breakpoint.h>
> +#include <linux/perf/arm_pmuv3.h>
>  
>  #include <asm/debug-monitors.h>
>  #include <asm/kvm_asm.h>
> diff --git a/arch/arm64/kvm/hyp/include/hyp/switch.h b/arch/arm64/kvm/hyp/include/hyp/switch.h
> index 7599844908c0..825b81749972 100644
> --- a/arch/arm64/kvm/hyp/include/hyp/switch.h
> +++ b/arch/arm64/kvm/hyp/include/hyp/switch.h
> @@ -14,6 +14,7 @@
>  #include <linux/kvm_host.h>
>  #include <linux/types.h>
>  #include <linux/jump_label.h>
> +#include <linux/perf/arm_pmuv3.h>
>  #include <uapi/linux/psci.h>
>  
>  #include <asm/barrier.h>
> diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c
> index 6b48a3d16d0d..8bfc6b0a85f6 100644
> --- a/arch/arm64/kvm/pmu.c
> +++ b/arch/arm64/kvm/pmu.c
> @@ -8,6 +8,8 @@
>  #include <linux/perf/arm_pmu.h>
>  #include <linux/perf/arm_pmuv3.h>
>  
> +#include <asm/kvm_pmu.h>
> +
>  static DEFINE_PER_CPU(struct kvm_pmu_events, kvm_pmu_events);
>  
>  /*
> diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
> index 99fdbe174202..eaff6d63ef77 100644
> --- a/arch/arm64/kvm/sys_regs.c
> +++ b/arch/arm64/kvm/sys_regs.c
> @@ -18,6 +18,7 @@
>  #include <linux/printk.h>
>  #include <linux/uaccess.h>
>  #include <linux/irqchip/arm-gic-v3.h>
> +#include <linux/perf/arm_pmuv3.h>
>  
>  #include <asm/arm_pmuv3.h>
>  #include <asm/cacheflush.h>
> diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h
> index 6dc5e0cd76ca..fb382bcd4f4b 100644
> --- a/include/linux/perf/arm_pmu.h
> +++ b/include/linux/perf/arm_pmu.h
> @@ -187,6 +187,11 @@ void armpmu_free_irq(int irq, int cpu);
>  
>  #define ARMV8_PMU_PDEV_NAME "armv8-pmu"
>  
> +#else
> +
> +struct arm_pmu {
> +};
> +
>  #endif /* CONFIG_ARM_PMU */
>  
>  #define ARMV8_SPE_PDEV_NAME "arm,spe-v1"
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index e2f6344256ce..25259fcf3115 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -48,6 +48,7 @@
>  #include <linux/lockdep.h>
>  #include <linux/kthread.h>
>  #include <linux/suspend.h>
> +#include <linux/perf_event.h>
>  
>  #include <asm/processor.h>
>  #include <asm/ioctl.h>
> -- 
> 2.50.0.727.gbf7dc18ff4-goog
> 

Re: [PATCH v3 04/22] KVM: arm64: Cleanup PMU includes

[Colton Lewis]

Mark Rutland <mark.rutland@arm.com> writes:

> On Thu, Jun 26, 2025 at 08:04:40PM +0000, Colton Lewis wrote:
>> From: Marc Zyngier <maz@kernel.org>

>> Reorganize these tangled headers.

> In prior versions of this series, there was some earlier text explaining  
> that
> there was a circular header dependency. In v2:

>     
> https://lore.kernel.org/linux-arm-kernel/20250620221326.1261128-6-coltonlewis@google.com/

> ... it said:

> | asm/kvm_host.h includes asm/arm_pmu.h which includes perf/arm_pmuv3.h
> | which includes asm/arm_pmuv3.h which includes asm/kvm_host.h This
> | causes compilation problems why trying to use anything defined in any
> | of the headers in any other headers.

> Was there some reason for dropping that? It's a bit odd to start without a
> problem statement.

When I rebased on top of kvm/queue, that included a patch from Sean that
changed the the problem. I thought I wrote a new description but I guess
not. I'll include in my next reroll.

The core issue is still that including *all* of kvm_host.h in
arm_pmuv3.h is a circular dependency.

asm/kvm_host.h includes asm/kvm_pmu.h includes perf/arm_pmuv3.h includes
asm/arm_pmuv3.h includes asm/kvm_host.h

Some other things in this patch are vestigial or needed later, so I
should put them in the patches where they are used.


>> * Respect the move defining the interface between KVM and PMU in its
>>    own header asm/kvm_pmu.h

>> * Define an empty struct arm_pmu so it is defined for those interface
>>    functions when compiling with CONFIG_KVM but not CONFIG_ARM_PMU

> Which functions in particular are those? What prevents them from  
> depending on
> CONFIG_ARM_PMU?

> Mark.

Functions I introduce later. So this hunk should be in a later patch if
I need to keep it.

bool kvm_pmu_is_partitioned(struct arm_pmu *pmu);
u64 kvm_pmu_host_counter_mask(struct arm_pmu *pmu);
u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu);

The first, at least, has to be called in non-PMU KVM code to check if we
can use my ioctl later.

I suppose in that case it's only calling a dummy definition that returns
false, so I don't need struct arm_pmu * in the argument and can just use
void *

>> Signed-off-by: Marc Zyngier <maz@kernel.org>
>> Signed-off-by: Colton Lewis <coltonlewis@google.com>
>> ---
>>   arch/arm64/include/asm/arm_pmuv3.h      |  2 +-
>>   arch/arm64/include/asm/kvm_host.h       | 15 +--------------
>>   arch/arm64/include/asm/kvm_pmu.h        | 15 +++++++++++++++
>>   arch/arm64/kvm/debug.c                  |  1 +
>>   arch/arm64/kvm/hyp/include/hyp/switch.h |  1 +
>>   arch/arm64/kvm/pmu.c                    |  2 ++
>>   arch/arm64/kvm/sys_regs.c               |  1 +
>>   include/linux/perf/arm_pmu.h            |  5 +++++
>>   virt/kvm/kvm_main.c                     |  1 +
>>   9 files changed, 28 insertions(+), 15 deletions(-)

>> diff --git a/arch/arm64/include/asm/arm_pmuv3.h  
>> b/arch/arm64/include/asm/arm_pmuv3.h
>> index 8a777dec8d88..32c003a7b810 100644
>> --- a/arch/arm64/include/asm/arm_pmuv3.h
>> +++ b/arch/arm64/include/asm/arm_pmuv3.h
>> @@ -6,7 +6,7 @@
>>   #ifndef __ASM_PMUV3_H
>>   #define __ASM_PMUV3_H

>> -#include <asm/kvm_host.h>
>> +#include <asm/kvm_pmu.h>

>>   #include <asm/cpufeature.h>
>>   #include <asm/sysreg.h>
>> diff --git a/arch/arm64/include/asm/kvm_host.h  
>> b/arch/arm64/include/asm/kvm_host.h
>> index 27ed26bd4381..2df76689381a 100644
>> --- a/arch/arm64/include/asm/kvm_host.h
>> +++ b/arch/arm64/include/asm/kvm_host.h
>> @@ -14,6 +14,7 @@
>>   #include <linux/arm-smccc.h>
>>   #include <linux/bitmap.h>
>>   #include <linux/types.h>
>> +#include <linux/irq_work.h>
>>   #include <linux/jump_label.h>
>>   #include <linux/kvm_types.h>
>>   #include <linux/maple_tree.h>
>> @@ -1487,25 +1488,11 @@ void kvm_arch_vcpu_ctxflush_fp(struct kvm_vcpu  
>> *vcpu);
>>   void kvm_arch_vcpu_ctxsync_fp(struct kvm_vcpu *vcpu);
>>   void kvm_arch_vcpu_put_fp(struct kvm_vcpu *vcpu);

>> -static inline bool kvm_pmu_counter_deferred(struct perf_event_attr  
>> *attr)
>> -{
>> -	return (!has_vhe() && attr->exclude_host);
>> -}
>> -
>>   #ifdef CONFIG_KVM
>> -void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr);
>> -void kvm_clr_pmu_events(u64 clr);
>> -bool kvm_set_pmuserenr(u64 val);
>>   void kvm_enable_trbe(void);
>>   void kvm_disable_trbe(void);
>>   void kvm_tracing_set_el1_configuration(u64 trfcr_while_in_guest);
>>   #else
>> -static inline void kvm_set_pmu_events(u64 set, struct perf_event_attr  
>> *attr) {}
>> -static inline void kvm_clr_pmu_events(u64 clr) {}
>> -static inline bool kvm_set_pmuserenr(u64 val)
>> -{
>> -	return false;
>> -}
>>   static inline void kvm_enable_trbe(void) {}
>>   static inline void kvm_disable_trbe(void) {}
>>   static inline void kvm_tracing_set_el1_configuration(u64  
>> trfcr_while_in_guest) {}
>> diff --git a/arch/arm64/include/asm/kvm_pmu.h  
>> b/arch/arm64/include/asm/kvm_pmu.h
>> index baf028d19dfc..ad3247b46838 100644
>> --- a/arch/arm64/include/asm/kvm_pmu.h
>> +++ b/arch/arm64/include/asm/kvm_pmu.h
>> @@ -11,9 +11,15 @@
>>   #include <linux/kvm_types.h>
>>   #include <linux/perf_event.h>
>>   #include <linux/perf/arm_pmuv3.h>
>> +#include <linux/perf/arm_pmu.h>

>>   #define KVM_ARMV8_PMU_MAX_COUNTERS	32

>> +#define kvm_pmu_counter_deferred(attr)			\
>> +	({						\
>> +		!has_vhe() && (attr)->exclude_host;	\
>> +	})
>> +
>>   #if IS_ENABLED(CONFIG_HW_PERF_EVENTS) && IS_ENABLED(CONFIG_KVM)
>>   struct kvm_pmc {
>>   	u8 idx;	/* index into the pmu->pmc array */
>> @@ -68,6 +74,9 @@ int kvm_arm_pmu_v3_has_attr(struct kvm_vcpu *vcpu,
>>   int kvm_arm_pmu_v3_enable(struct kvm_vcpu *vcpu);

>>   struct kvm_pmu_events *kvm_get_pmu_events(void);
>> +void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr);
>> +void kvm_clr_pmu_events(u64 clr);
>> +bool kvm_set_pmuserenr(u64 val);
>>   void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu);
>>   void kvm_vcpu_pmu_restore_host(struct kvm_vcpu *vcpu);
>>   void kvm_vcpu_pmu_resync_el0(void);
>> @@ -161,6 +170,12 @@ static inline u64 kvm_pmu_get_pmceid(struct  
>> kvm_vcpu *vcpu, bool pmceid1)

>>   #define kvm_vcpu_has_pmu(vcpu)		({ false; })
>>   static inline void kvm_pmu_update_vcpu_events(struct kvm_vcpu *vcpu) {}
>> +static inline void kvm_set_pmu_events(u64 set, struct perf_event_attr  
>> *attr) {}
>> +static inline void kvm_clr_pmu_events(u64 clr) {}
>> +static inline bool kvm_set_pmuserenr(u64 val)
>> +{
>> +	return false;
>> +}
>>   static inline void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu) {}
>>   static inline void kvm_vcpu_pmu_restore_host(struct kvm_vcpu *vcpu) {}
>>   static inline void kvm_vcpu_reload_pmu(struct kvm_vcpu *vcpu) {}
>> diff --git a/arch/arm64/kvm/debug.c b/arch/arm64/kvm/debug.c
>> index 1a7dab333f55..a554c3e368dc 100644
>> --- a/arch/arm64/kvm/debug.c
>> +++ b/arch/arm64/kvm/debug.c
>> @@ -9,6 +9,7 @@

>>   #include <linux/kvm_host.h>
>>   #include <linux/hw_breakpoint.h>
>> +#include <linux/perf/arm_pmuv3.h>

>>   #include <asm/debug-monitors.h>
>>   #include <asm/kvm_asm.h>
>> diff --git a/arch/arm64/kvm/hyp/include/hyp/switch.h  
>> b/arch/arm64/kvm/hyp/include/hyp/switch.h
>> index 7599844908c0..825b81749972 100644
>> --- a/arch/arm64/kvm/hyp/include/hyp/switch.h
>> +++ b/arch/arm64/kvm/hyp/include/hyp/switch.h
>> @@ -14,6 +14,7 @@
>>   #include <linux/kvm_host.h>
>>   #include <linux/types.h>
>>   #include <linux/jump_label.h>
>> +#include <linux/perf/arm_pmuv3.h>
>>   #include <uapi/linux/psci.h>

>>   #include <asm/barrier.h>
>> diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c
>> index 6b48a3d16d0d..8bfc6b0a85f6 100644
>> --- a/arch/arm64/kvm/pmu.c
>> +++ b/arch/arm64/kvm/pmu.c
>> @@ -8,6 +8,8 @@
>>   #include <linux/perf/arm_pmu.h>
>>   #include <linux/perf/arm_pmuv3.h>

>> +#include <asm/kvm_pmu.h>
>> +
>>   static DEFINE_PER_CPU(struct kvm_pmu_events, kvm_pmu_events);

>>   /*
>> diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
>> index 99fdbe174202..eaff6d63ef77 100644
>> --- a/arch/arm64/kvm/sys_regs.c
>> +++ b/arch/arm64/kvm/sys_regs.c
>> @@ -18,6 +18,7 @@
>>   #include <linux/printk.h>
>>   #include <linux/uaccess.h>
>>   #include <linux/irqchip/arm-gic-v3.h>
>> +#include <linux/perf/arm_pmuv3.h>

>>   #include <asm/arm_pmuv3.h>
>>   #include <asm/cacheflush.h>
>> diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h
>> index 6dc5e0cd76ca..fb382bcd4f4b 100644
>> --- a/include/linux/perf/arm_pmu.h
>> +++ b/include/linux/perf/arm_pmu.h
>> @@ -187,6 +187,11 @@ void armpmu_free_irq(int irq, int cpu);

>>   #define ARMV8_PMU_PDEV_NAME "armv8-pmu"

>> +#else
>> +
>> +struct arm_pmu {
>> +};
>> +
>>   #endif /* CONFIG_ARM_PMU */

>>   #define ARMV8_SPE_PDEV_NAME "arm,spe-v1"
>> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
>> index e2f6344256ce..25259fcf3115 100644
>> --- a/virt/kvm/kvm_main.c
>> +++ b/virt/kvm/kvm_main.c
>> @@ -48,6 +48,7 @@
>>   #include <linux/lockdep.h>
>>   #include <linux/kthread.h>
>>   #include <linux/suspend.h>
>> +#include <linux/perf_event.h>

>>   #include <asm/processor.h>
>>   #include <asm/ioctl.h>
>> --
>> 2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 05/22] KVM: arm64: Reorganize PMU functions

[Colton Lewis]

A lot of functions in pmu-emul.c aren't specific to the emulated PMU
implementation. Move them to the more appropriate pmu.c file where
shared PMU functions should live.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/include/asm/kvm_pmu.h |   3 +
 arch/arm64/kvm/pmu-emul.c        | 672 +-----------------------------
 arch/arm64/kvm/pmu.c             | 673 +++++++++++++++++++++++++++++++
 3 files changed, 677 insertions(+), 671 deletions(-)

diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
index ad3247b46838..6c961e877804 100644
--- a/arch/arm64/include/asm/kvm_pmu.h
+++ b/arch/arm64/include/asm/kvm_pmu.h
@@ -51,13 +51,16 @@ u64 kvm_pmu_get_counter_value(struct kvm_vcpu *vcpu, u64 select_idx);
 void kvm_pmu_set_counter_value(struct kvm_vcpu *vcpu, u64 select_idx, u64 val);
 void kvm_pmu_set_counter_value_user(struct kvm_vcpu *vcpu, u64 select_idx, u64 val);
 u64 kvm_pmu_implemented_counter_mask(struct kvm_vcpu *vcpu);
+u64 kvm_pmu_hyp_counter_mask(struct kvm_vcpu *vcpu);
 u64 kvm_pmu_accessible_counter_mask(struct kvm_vcpu *vcpu);
+u32 kvm_pmu_event_mask(struct kvm *kvm);
 u64 kvm_pmu_get_pmceid(struct kvm_vcpu *vcpu, bool pmceid1);
 void kvm_pmu_vcpu_init(struct kvm_vcpu *vcpu);
 void kvm_pmu_vcpu_destroy(struct kvm_vcpu *vcpu);
 void kvm_pmu_reprogram_counter_mask(struct kvm_vcpu *vcpu, u64 val);
 void kvm_pmu_flush_hwstate(struct kvm_vcpu *vcpu);
 void kvm_pmu_sync_hwstate(struct kvm_vcpu *vcpu);
+bool kvm_pmu_overflow_status(struct kvm_vcpu *vcpu);
 bool kvm_pmu_should_notify_user(struct kvm_vcpu *vcpu);
 void kvm_pmu_update_run(struct kvm_vcpu *vcpu);
 void kvm_pmu_software_increment(struct kvm_vcpu *vcpu, u64 val);
diff --git a/arch/arm64/kvm/pmu-emul.c b/arch/arm64/kvm/pmu-emul.c
index dcdd80ffd49d..bcaa9f7a8ca2 100644
--- a/arch/arm64/kvm/pmu-emul.c
+++ b/arch/arm64/kvm/pmu-emul.c
@@ -17,19 +17,10 @@
 
 #define PERF_ATTR_CFG1_COUNTER_64BIT	BIT(0)
 
-static LIST_HEAD(arm_pmus);
-static DEFINE_MUTEX(arm_pmus_lock);
-
 static void kvm_pmu_create_perf_event(struct kvm_pmc *pmc);
 static void kvm_pmu_release_perf_event(struct kvm_pmc *pmc);
 static bool kvm_pmu_counter_is_enabled(struct kvm_pmc *pmc);
 
-bool kvm_supports_guest_pmuv3(void)
-{
-	guard(mutex)(&arm_pmus_lock);
-	return !list_empty(&arm_pmus);
-}
-
 static struct kvm_vcpu *kvm_pmc_to_vcpu(const struct kvm_pmc *pmc)
 {
 	return container_of(pmc, struct kvm_vcpu, arch.pmu.pmc[pmc->idx]);
@@ -40,46 +31,6 @@ static struct kvm_pmc *kvm_vcpu_idx_to_pmc(struct kvm_vcpu *vcpu, int cnt_idx)
 	return &vcpu->arch.pmu.pmc[cnt_idx];
 }
 
-static u32 __kvm_pmu_event_mask(unsigned int pmuver)
-{
-	switch (pmuver) {
-	case ID_AA64DFR0_EL1_PMUVer_IMP:
-		return GENMASK(9, 0);
-	case ID_AA64DFR0_EL1_PMUVer_V3P1:
-	case ID_AA64DFR0_EL1_PMUVer_V3P4:
-	case ID_AA64DFR0_EL1_PMUVer_V3P5:
-	case ID_AA64DFR0_EL1_PMUVer_V3P7:
-		return GENMASK(15, 0);
-	default:		/* Shouldn't be here, just for sanity */
-		WARN_ONCE(1, "Unknown PMU version %d\n", pmuver);
-		return 0;
-	}
-}
-
-static u32 kvm_pmu_event_mask(struct kvm *kvm)
-{
-	u64 dfr0 = kvm_read_vm_id_reg(kvm, SYS_ID_AA64DFR0_EL1);
-	u8 pmuver = SYS_FIELD_GET(ID_AA64DFR0_EL1, PMUVer, dfr0);
-
-	return __kvm_pmu_event_mask(pmuver);
-}
-
-u64 kvm_pmu_evtyper_mask(struct kvm *kvm)
-{
-	u64 mask = ARMV8_PMU_EXCLUDE_EL1 | ARMV8_PMU_EXCLUDE_EL0 |
-		   kvm_pmu_event_mask(kvm);
-
-	if (kvm_has_feat(kvm, ID_AA64PFR0_EL1, EL2, IMP))
-		mask |= ARMV8_PMU_INCLUDE_EL2;
-
-	if (kvm_has_feat(kvm, ID_AA64PFR0_EL1, EL3, IMP))
-		mask |= ARMV8_PMU_EXCLUDE_NS_EL0 |
-			ARMV8_PMU_EXCLUDE_NS_EL1 |
-			ARMV8_PMU_EXCLUDE_EL3;
-
-	return mask;
-}
-
 /**
  * kvm_pmc_is_64bit - determine if counter is 64bit
  * @pmc: counter context
@@ -272,59 +223,6 @@ void kvm_pmu_vcpu_destroy(struct kvm_vcpu *vcpu)
 	irq_work_sync(&vcpu->arch.pmu.overflow_work);
 }
 
-static u64 kvm_pmu_hyp_counter_mask(struct kvm_vcpu *vcpu)
-{
-	unsigned int hpmn, n;
-
-	if (!vcpu_has_nv(vcpu))
-		return 0;
-
-	hpmn = SYS_FIELD_GET(MDCR_EL2, HPMN, __vcpu_sys_reg(vcpu, MDCR_EL2));
-	n = vcpu->kvm->arch.nr_pmu_counters;
-
-	/*
-	 * Programming HPMN to a value greater than PMCR_EL0.N is
-	 * CONSTRAINED UNPREDICTABLE. Make the implementation choice that an
-	 * UNKNOWN number of counters (in our case, zero) are reserved for EL2.
-	 */
-	if (hpmn >= n)
-		return 0;
-
-	/*
-	 * Programming HPMN=0 is CONSTRAINED UNPREDICTABLE if FEAT_HPMN0 isn't
-	 * implemented. Since KVM's ability to emulate HPMN=0 does not directly
-	 * depend on hardware (all PMU registers are trapped), make the
-	 * implementation choice that all counters are included in the second
-	 * range reserved for EL2/EL3.
-	 */
-	return GENMASK(n - 1, hpmn);
-}
-
-bool kvm_pmu_counter_is_hyp(struct kvm_vcpu *vcpu, unsigned int idx)
-{
-	return kvm_pmu_hyp_counter_mask(vcpu) & BIT(idx);
-}
-
-u64 kvm_pmu_accessible_counter_mask(struct kvm_vcpu *vcpu)
-{
-	u64 mask = kvm_pmu_implemented_counter_mask(vcpu);
-
-	if (!vcpu_has_nv(vcpu) || vcpu_is_el2(vcpu))
-		return mask;
-
-	return mask & ~kvm_pmu_hyp_counter_mask(vcpu);
-}
-
-u64 kvm_pmu_implemented_counter_mask(struct kvm_vcpu *vcpu)
-{
-	u64 val = FIELD_GET(ARMV8_PMU_PMCR_N, kvm_vcpu_read_pmcr(vcpu));
-
-	if (val == 0)
-		return BIT(ARMV8_PMU_CYCLE_IDX);
-	else
-		return GENMASK(val - 1, 0) | BIT(ARMV8_PMU_CYCLE_IDX);
-}
-
 static void kvm_pmc_enable_perf_event(struct kvm_pmc *pmc)
 {
 	if (!pmc->perf_event) {
@@ -370,7 +268,7 @@ void kvm_pmu_reprogram_counter_mask(struct kvm_vcpu *vcpu, u64 val)
  * counter where the values of the global enable control, PMOVSSET_EL0[n], and
  * PMINTENSET_EL1[n] are all 1.
  */
-static bool kvm_pmu_overflow_status(struct kvm_vcpu *vcpu)
+bool kvm_pmu_overflow_status(struct kvm_vcpu *vcpu)
 {
 	u64 reg = __vcpu_sys_reg(vcpu, PMOVSSET_EL0);
 
@@ -393,24 +291,6 @@ static bool kvm_pmu_overflow_status(struct kvm_vcpu *vcpu)
 	return reg;
 }
 
-static void kvm_pmu_update_state(struct kvm_vcpu *vcpu)
-{
-	struct kvm_pmu *pmu = &vcpu->arch.pmu;
-	bool overflow;
-
-	overflow = kvm_pmu_overflow_status(vcpu);
-	if (pmu->irq_level == overflow)
-		return;
-
-	pmu->irq_level = overflow;
-
-	if (likely(irqchip_in_kernel(vcpu->kvm))) {
-		int ret = kvm_vgic_inject_irq(vcpu->kvm, vcpu,
-					      pmu->irq_num, overflow, pmu);
-		WARN_ON(ret);
-	}
-}
-
 bool kvm_pmu_should_notify_user(struct kvm_vcpu *vcpu)
 {
 	struct kvm_pmu *pmu = &vcpu->arch.pmu;
@@ -436,43 +316,6 @@ void kvm_pmu_update_run(struct kvm_vcpu *vcpu)
 		regs->device_irq_level |= KVM_ARM_DEV_PMU;
 }
 
-/**
- * kvm_pmu_flush_hwstate - flush pmu state to cpu
- * @vcpu: The vcpu pointer
- *
- * Check if the PMU has overflowed while we were running in the host, and inject
- * an interrupt if that was the case.
- */
-void kvm_pmu_flush_hwstate(struct kvm_vcpu *vcpu)
-{
-	kvm_pmu_update_state(vcpu);
-}
-
-/**
- * kvm_pmu_sync_hwstate - sync pmu state from cpu
- * @vcpu: The vcpu pointer
- *
- * Check if the PMU has overflowed while we were running in the guest, and
- * inject an interrupt if that was the case.
- */
-void kvm_pmu_sync_hwstate(struct kvm_vcpu *vcpu)
-{
-	kvm_pmu_update_state(vcpu);
-}
-
-/*
- * When perf interrupt is an NMI, we cannot safely notify the vcpu corresponding
- * to the event.
- * This is why we need a callback to do it once outside of the NMI context.
- */
-static void kvm_pmu_perf_overflow_notify_vcpu(struct irq_work *work)
-{
-	struct kvm_vcpu *vcpu;
-
-	vcpu = container_of(work, struct kvm_vcpu, arch.pmu.overflow_work);
-	kvm_vcpu_kick(vcpu);
-}
-
 /*
  * Perform an increment on any of the counters described in @mask,
  * generating the overflow if required, and propagate it as a chained
@@ -784,132 +627,6 @@ void kvm_pmu_set_counter_event_type(struct kvm_vcpu *vcpu, u64 data,
 	kvm_pmu_create_perf_event(pmc);
 }
 
-void kvm_host_pmu_init(struct arm_pmu *pmu)
-{
-	struct arm_pmu_entry *entry;
-
-	/*
-	 * Check the sanitised PMU version for the system, as KVM does not
-	 * support implementations where PMUv3 exists on a subset of CPUs.
-	 */
-	if (!pmuv3_implemented(kvm_arm_pmu_get_pmuver_limit()))
-		return;
-
-	guard(mutex)(&arm_pmus_lock);
-
-	entry = kmalloc(sizeof(*entry), GFP_KERNEL);
-	if (!entry)
-		return;
-
-	entry->arm_pmu = pmu;
-	list_add_tail(&entry->entry, &arm_pmus);
-}
-
-static struct arm_pmu *kvm_pmu_probe_armpmu(void)
-{
-	struct arm_pmu_entry *entry;
-	struct arm_pmu *pmu;
-	int cpu;
-
-	guard(mutex)(&arm_pmus_lock);
-
-	/*
-	 * It is safe to use a stale cpu to iterate the list of PMUs so long as
-	 * the same value is used for the entirety of the loop. Given this, and
-	 * the fact that no percpu data is used for the lookup there is no need
-	 * to disable preemption.
-	 *
-	 * It is still necessary to get a valid cpu, though, to probe for the
-	 * default PMU instance as userspace is not required to specify a PMU
-	 * type. In order to uphold the preexisting behavior KVM selects the
-	 * PMU instance for the core during vcpu init. A dependent use
-	 * case would be a user with disdain of all things big.LITTLE that
-	 * affines the VMM to a particular cluster of cores.
-	 *
-	 * In any case, userspace should just do the sane thing and use the UAPI
-	 * to select a PMU type directly. But, be wary of the baggage being
-	 * carried here.
-	 */
-	cpu = raw_smp_processor_id();
-	list_for_each_entry(entry, &arm_pmus, entry) {
-		pmu = entry->arm_pmu;
-
-		if (cpumask_test_cpu(cpu, &pmu->supported_cpus))
-			return pmu;
-	}
-
-	return NULL;
-}
-
-static u64 __compute_pmceid(struct arm_pmu *pmu, bool pmceid1)
-{
-	u32 hi[2], lo[2];
-
-	bitmap_to_arr32(lo, pmu->pmceid_bitmap, ARMV8_PMUV3_MAX_COMMON_EVENTS);
-	bitmap_to_arr32(hi, pmu->pmceid_ext_bitmap, ARMV8_PMUV3_MAX_COMMON_EVENTS);
-
-	return ((u64)hi[pmceid1] << 32) | lo[pmceid1];
-}
-
-static u64 compute_pmceid0(struct arm_pmu *pmu)
-{
-	u64 val = __compute_pmceid(pmu, 0);
-
-	/* always support SW_INCR */
-	val |= BIT(ARMV8_PMUV3_PERFCTR_SW_INCR);
-	/* always support CHAIN */
-	val |= BIT(ARMV8_PMUV3_PERFCTR_CHAIN);
-	return val;
-}
-
-static u64 compute_pmceid1(struct arm_pmu *pmu)
-{
-	u64 val = __compute_pmceid(pmu, 1);
-
-	/*
-	 * Don't advertise STALL_SLOT*, as PMMIR_EL0 is handled
-	 * as RAZ
-	 */
-	val &= ~(BIT_ULL(ARMV8_PMUV3_PERFCTR_STALL_SLOT - 32) |
-		 BIT_ULL(ARMV8_PMUV3_PERFCTR_STALL_SLOT_FRONTEND - 32) |
-		 BIT_ULL(ARMV8_PMUV3_PERFCTR_STALL_SLOT_BACKEND - 32));
-	return val;
-}
-
-u64 kvm_pmu_get_pmceid(struct kvm_vcpu *vcpu, bool pmceid1)
-{
-	struct arm_pmu *cpu_pmu = vcpu->kvm->arch.arm_pmu;
-	unsigned long *bmap = vcpu->kvm->arch.pmu_filter;
-	u64 val, mask = 0;
-	int base, i, nr_events;
-
-	if (!pmceid1) {
-		val = compute_pmceid0(cpu_pmu);
-		base = 0;
-	} else {
-		val = compute_pmceid1(cpu_pmu);
-		base = 32;
-	}
-
-	if (!bmap)
-		return val;
-
-	nr_events = kvm_pmu_event_mask(vcpu->kvm) + 1;
-
-	for (i = 0; i < 32; i += 8) {
-		u64 byte;
-
-		byte = bitmap_get_value8(bmap, base + i);
-		mask |= byte << i;
-		if (nr_events >= (0x4000 + base + 32)) {
-			byte = bitmap_get_value8(bmap, 0x4000 + base + i);
-			mask |= byte << (32 + i);
-		}
-	}
-
-	return val & mask;
-}
-
 void kvm_vcpu_reload_pmu(struct kvm_vcpu *vcpu)
 {
 	u64 mask = kvm_pmu_implemented_counter_mask(vcpu);
@@ -921,393 +638,6 @@ void kvm_vcpu_reload_pmu(struct kvm_vcpu *vcpu)
 	kvm_pmu_reprogram_counter_mask(vcpu, mask);
 }
 
-int kvm_arm_pmu_v3_enable(struct kvm_vcpu *vcpu)
-{
-	if (!vcpu->arch.pmu.created)
-		return -EINVAL;
-
-	/*
-	 * A valid interrupt configuration for the PMU is either to have a
-	 * properly configured interrupt number and using an in-kernel
-	 * irqchip, or to not have an in-kernel GIC and not set an IRQ.
-	 */
-	if (irqchip_in_kernel(vcpu->kvm)) {
-		int irq = vcpu->arch.pmu.irq_num;
-		/*
-		 * If we are using an in-kernel vgic, at this point we know
-		 * the vgic will be initialized, so we can check the PMU irq
-		 * number against the dimensions of the vgic and make sure
-		 * it's valid.
-		 */
-		if (!irq_is_ppi(irq) && !vgic_valid_spi(vcpu->kvm, irq))
-			return -EINVAL;
-	} else if (kvm_arm_pmu_irq_initialized(vcpu)) {
-		   return -EINVAL;
-	}
-
-	return 0;
-}
-
-static int kvm_arm_pmu_v3_init(struct kvm_vcpu *vcpu)
-{
-	if (irqchip_in_kernel(vcpu->kvm)) {
-		int ret;
-
-		/*
-		 * If using the PMU with an in-kernel virtual GIC
-		 * implementation, we require the GIC to be already
-		 * initialized when initializing the PMU.
-		 */
-		if (!vgic_initialized(vcpu->kvm))
-			return -ENODEV;
-
-		if (!kvm_arm_pmu_irq_initialized(vcpu))
-			return -ENXIO;
-
-		ret = kvm_vgic_set_owner(vcpu, vcpu->arch.pmu.irq_num,
-					 &vcpu->arch.pmu);
-		if (ret)
-			return ret;
-	}
-
-	init_irq_work(&vcpu->arch.pmu.overflow_work,
-		      kvm_pmu_perf_overflow_notify_vcpu);
-
-	vcpu->arch.pmu.created = true;
-	return 0;
-}
-
-/*
- * For one VM the interrupt type must be same for each vcpu.
- * As a PPI, the interrupt number is the same for all vcpus,
- * while as an SPI it must be a separate number per vcpu.
- */
-static bool pmu_irq_is_valid(struct kvm *kvm, int irq)
-{
-	unsigned long i;
-	struct kvm_vcpu *vcpu;
-
-	kvm_for_each_vcpu(i, vcpu, kvm) {
-		if (!kvm_arm_pmu_irq_initialized(vcpu))
-			continue;
-
-		if (irq_is_ppi(irq)) {
-			if (vcpu->arch.pmu.irq_num != irq)
-				return false;
-		} else {
-			if (vcpu->arch.pmu.irq_num == irq)
-				return false;
-		}
-	}
-
-	return true;
-}
-
-/**
- * kvm_arm_pmu_get_max_counters - Return the max number of PMU counters.
- * @kvm: The kvm pointer
- */
-u8 kvm_arm_pmu_get_max_counters(struct kvm *kvm)
-{
-	struct arm_pmu *arm_pmu = kvm->arch.arm_pmu;
-
-	/*
-	 * PMUv3 requires that all event counters are capable of counting any
-	 * event, though the same may not be true of non-PMUv3 hardware.
-	 */
-	if (cpus_have_final_cap(ARM64_WORKAROUND_PMUV3_IMPDEF_TRAPS))
-		return 1;
-
-	/*
-	 * The arm_pmu->cntr_mask considers the fixed counter(s) as well.
-	 * Ignore those and return only the general-purpose counters.
-	 */
-	return bitmap_weight(arm_pmu->cntr_mask, ARMV8_PMU_MAX_GENERAL_COUNTERS);
-}
-
-static void kvm_arm_set_nr_counters(struct kvm *kvm, unsigned int nr)
-{
-	kvm->arch.nr_pmu_counters = nr;
-
-	/* Reset MDCR_EL2.HPMN behind the vcpus' back... */
-	if (test_bit(KVM_ARM_VCPU_HAS_EL2, kvm->arch.vcpu_features)) {
-		struct kvm_vcpu *vcpu;
-		unsigned long i;
-
-		kvm_for_each_vcpu(i, vcpu, kvm) {
-			u64 val = __vcpu_sys_reg(vcpu, MDCR_EL2);
-			val &= ~MDCR_EL2_HPMN;
-			val |= FIELD_PREP(MDCR_EL2_HPMN, kvm->arch.nr_pmu_counters);
-			__vcpu_assign_sys_reg(vcpu, MDCR_EL2, val);
-		}
-	}
-}
-
-static void kvm_arm_set_pmu(struct kvm *kvm, struct arm_pmu *arm_pmu)
-{
-	lockdep_assert_held(&kvm->arch.config_lock);
-
-	kvm->arch.arm_pmu = arm_pmu;
-	kvm_arm_set_nr_counters(kvm, kvm_arm_pmu_get_max_counters(kvm));
-}
-
-/**
- * kvm_arm_set_default_pmu - No PMU set, get the default one.
- * @kvm: The kvm pointer
- *
- * The observant among you will notice that the supported_cpus
- * mask does not get updated for the default PMU even though it
- * is quite possible the selected instance supports only a
- * subset of cores in the system. This is intentional, and
- * upholds the preexisting behavior on heterogeneous systems
- * where vCPUs can be scheduled on any core but the guest
- * counters could stop working.
- */
-int kvm_arm_set_default_pmu(struct kvm *kvm)
-{
-	struct arm_pmu *arm_pmu = kvm_pmu_probe_armpmu();
-
-	if (!arm_pmu)
-		return -ENODEV;
-
-	kvm_arm_set_pmu(kvm, arm_pmu);
-	return 0;
-}
-
-static int kvm_arm_pmu_v3_set_pmu(struct kvm_vcpu *vcpu, int pmu_id)
-{
-	struct kvm *kvm = vcpu->kvm;
-	struct arm_pmu_entry *entry;
-	struct arm_pmu *arm_pmu;
-	int ret = -ENXIO;
-
-	lockdep_assert_held(&kvm->arch.config_lock);
-	mutex_lock(&arm_pmus_lock);
-
-	list_for_each_entry(entry, &arm_pmus, entry) {
-		arm_pmu = entry->arm_pmu;
-		if (arm_pmu->pmu.type == pmu_id) {
-			if (kvm_vm_has_ran_once(kvm) ||
-			    (kvm->arch.pmu_filter && kvm->arch.arm_pmu != arm_pmu)) {
-				ret = -EBUSY;
-				break;
-			}
-
-			kvm_arm_set_pmu(kvm, arm_pmu);
-			cpumask_copy(kvm->arch.supported_cpus, &arm_pmu->supported_cpus);
-			ret = 0;
-			break;
-		}
-	}
-
-	mutex_unlock(&arm_pmus_lock);
-	return ret;
-}
-
-static int kvm_arm_pmu_v3_set_nr_counters(struct kvm_vcpu *vcpu, unsigned int n)
-{
-	struct kvm *kvm = vcpu->kvm;
-
-	if (!kvm->arch.arm_pmu)
-		return -EINVAL;
-
-	if (n > kvm_arm_pmu_get_max_counters(kvm))
-		return -EINVAL;
-
-	kvm_arm_set_nr_counters(kvm, n);
-	return 0;
-}
-
-int kvm_arm_pmu_v3_set_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
-{
-	struct kvm *kvm = vcpu->kvm;
-
-	lockdep_assert_held(&kvm->arch.config_lock);
-
-	if (!kvm_vcpu_has_pmu(vcpu))
-		return -ENODEV;
-
-	if (vcpu->arch.pmu.created)
-		return -EBUSY;
-
-	switch (attr->attr) {
-	case KVM_ARM_VCPU_PMU_V3_IRQ: {
-		int __user *uaddr = (int __user *)(long)attr->addr;
-		int irq;
-
-		if (!irqchip_in_kernel(kvm))
-			return -EINVAL;
-
-		if (get_user(irq, uaddr))
-			return -EFAULT;
-
-		/* The PMU overflow interrupt can be a PPI or a valid SPI. */
-		if (!(irq_is_ppi(irq) || irq_is_spi(irq)))
-			return -EINVAL;
-
-		if (!pmu_irq_is_valid(kvm, irq))
-			return -EINVAL;
-
-		if (kvm_arm_pmu_irq_initialized(vcpu))
-			return -EBUSY;
-
-		kvm_debug("Set kvm ARM PMU irq: %d\n", irq);
-		vcpu->arch.pmu.irq_num = irq;
-		return 0;
-	}
-	case KVM_ARM_VCPU_PMU_V3_FILTER: {
-		u8 pmuver = kvm_arm_pmu_get_pmuver_limit();
-		struct kvm_pmu_event_filter __user *uaddr;
-		struct kvm_pmu_event_filter filter;
-		int nr_events;
-
-		/*
-		 * Allow userspace to specify an event filter for the entire
-		 * event range supported by PMUVer of the hardware, rather
-		 * than the guest's PMUVer for KVM backward compatibility.
-		 */
-		nr_events = __kvm_pmu_event_mask(pmuver) + 1;
-
-		uaddr = (struct kvm_pmu_event_filter __user *)(long)attr->addr;
-
-		if (copy_from_user(&filter, uaddr, sizeof(filter)))
-			return -EFAULT;
-
-		if (((u32)filter.base_event + filter.nevents) > nr_events ||
-		    (filter.action != KVM_PMU_EVENT_ALLOW &&
-		     filter.action != KVM_PMU_EVENT_DENY))
-			return -EINVAL;
-
-		if (kvm_vm_has_ran_once(kvm))
-			return -EBUSY;
-
-		if (!kvm->arch.pmu_filter) {
-			kvm->arch.pmu_filter = bitmap_alloc(nr_events, GFP_KERNEL_ACCOUNT);
-			if (!kvm->arch.pmu_filter)
-				return -ENOMEM;
-
-			/*
-			 * The default depends on the first applied filter.
-			 * If it allows events, the default is to deny.
-			 * Conversely, if the first filter denies a set of
-			 * events, the default is to allow.
-			 */
-			if (filter.action == KVM_PMU_EVENT_ALLOW)
-				bitmap_zero(kvm->arch.pmu_filter, nr_events);
-			else
-				bitmap_fill(kvm->arch.pmu_filter, nr_events);
-		}
-
-		if (filter.action == KVM_PMU_EVENT_ALLOW)
-			bitmap_set(kvm->arch.pmu_filter, filter.base_event, filter.nevents);
-		else
-			bitmap_clear(kvm->arch.pmu_filter, filter.base_event, filter.nevents);
-
-		return 0;
-	}
-	case KVM_ARM_VCPU_PMU_V3_SET_PMU: {
-		int __user *uaddr = (int __user *)(long)attr->addr;
-		int pmu_id;
-
-		if (get_user(pmu_id, uaddr))
-			return -EFAULT;
-
-		return kvm_arm_pmu_v3_set_pmu(vcpu, pmu_id);
-	}
-	case KVM_ARM_VCPU_PMU_V3_SET_NR_COUNTERS: {
-		unsigned int __user *uaddr = (unsigned int __user *)(long)attr->addr;
-		unsigned int n;
-
-		if (get_user(n, uaddr))
-			return -EFAULT;
-
-		return kvm_arm_pmu_v3_set_nr_counters(vcpu, n);
-	}
-	case KVM_ARM_VCPU_PMU_V3_INIT:
-		return kvm_arm_pmu_v3_init(vcpu);
-	}
-
-	return -ENXIO;
-}
-
-int kvm_arm_pmu_v3_get_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
-{
-	switch (attr->attr) {
-	case KVM_ARM_VCPU_PMU_V3_IRQ: {
-		int __user *uaddr = (int __user *)(long)attr->addr;
-		int irq;
-
-		if (!irqchip_in_kernel(vcpu->kvm))
-			return -EINVAL;
-
-		if (!kvm_vcpu_has_pmu(vcpu))
-			return -ENODEV;
-
-		if (!kvm_arm_pmu_irq_initialized(vcpu))
-			return -ENXIO;
-
-		irq = vcpu->arch.pmu.irq_num;
-		return put_user(irq, uaddr);
-	}
-	}
-
-	return -ENXIO;
-}
-
-int kvm_arm_pmu_v3_has_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
-{
-	switch (attr->attr) {
-	case KVM_ARM_VCPU_PMU_V3_IRQ:
-	case KVM_ARM_VCPU_PMU_V3_INIT:
-	case KVM_ARM_VCPU_PMU_V3_FILTER:
-	case KVM_ARM_VCPU_PMU_V3_SET_PMU:
-	case KVM_ARM_VCPU_PMU_V3_SET_NR_COUNTERS:
-		if (kvm_vcpu_has_pmu(vcpu))
-			return 0;
-	}
-
-	return -ENXIO;
-}
-
-u8 kvm_arm_pmu_get_pmuver_limit(void)
-{
-	unsigned int pmuver;
-
-	pmuver = SYS_FIELD_GET(ID_AA64DFR0_EL1, PMUVer,
-			       read_sanitised_ftr_reg(SYS_ID_AA64DFR0_EL1));
-
-	/*
-	 * Spoof a barebones PMUv3 implementation if the system supports IMPDEF
-	 * traps of the PMUv3 sysregs
-	 */
-	if (cpus_have_final_cap(ARM64_WORKAROUND_PMUV3_IMPDEF_TRAPS))
-		return ID_AA64DFR0_EL1_PMUVer_IMP;
-
-	/*
-	 * Otherwise, treat IMPLEMENTATION DEFINED functionality as
-	 * unimplemented
-	 */
-	if (pmuver == ID_AA64DFR0_EL1_PMUVer_IMP_DEF)
-		return 0;
-
-	return min(pmuver, ID_AA64DFR0_EL1_PMUVer_V3P5);
-}
-
-/**
- * kvm_vcpu_read_pmcr - Read PMCR_EL0 register for the vCPU
- * @vcpu: The vcpu pointer
- */
-u64 kvm_vcpu_read_pmcr(struct kvm_vcpu *vcpu)
-{
-	u64 pmcr = __vcpu_sys_reg(vcpu, PMCR_EL0);
-	u64 n = vcpu->kvm->arch.nr_pmu_counters;
-
-	if (vcpu_has_nv(vcpu) && !vcpu_is_el2(vcpu))
-		n = FIELD_GET(MDCR_EL2_HPMN, __vcpu_sys_reg(vcpu, MDCR_EL2));
-
-	return u64_replace_bits(pmcr, n, ARMV8_PMU_PMCR_N);
-}
-
 void kvm_pmu_nested_transition(struct kvm_vcpu *vcpu)
 {
 	bool reprogrammed = false;
diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c
index 8bfc6b0a85f6..79b7ea037153 100644
--- a/arch/arm64/kvm/pmu.c
+++ b/arch/arm64/kvm/pmu.c
@@ -8,10 +8,21 @@
 #include <linux/perf/arm_pmu.h>
 #include <linux/perf/arm_pmuv3.h>
 
+#include <asm/kvm_emulate.h>
 #include <asm/kvm_pmu.h>
 
+static LIST_HEAD(arm_pmus);
+static DEFINE_MUTEX(arm_pmus_lock);
 static DEFINE_PER_CPU(struct kvm_pmu_events, kvm_pmu_events);
 
+#define kvm_arm_pmu_irq_initialized(v)	((v)->arch.pmu.irq_num >= VGIC_NR_SGIS)
+
+bool kvm_supports_guest_pmuv3(void)
+{
+	guard(mutex)(&arm_pmus_lock);
+	return !list_empty(&arm_pmus);
+}
+
 /*
  * Given the perf event attributes and system type, determine
  * if we are going to need to switch counters at guest entry/exit.
@@ -211,3 +222,665 @@ void kvm_vcpu_pmu_resync_el0(void)
 
 	kvm_make_request(KVM_REQ_RESYNC_PMU_EL0, vcpu);
 }
+
+void kvm_host_pmu_init(struct arm_pmu *pmu)
+{
+	struct arm_pmu_entry *entry;
+
+	/*
+	 * Check the sanitised PMU version for the system, as KVM does not
+	 * support implementations where PMUv3 exists on a subset of CPUs.
+	 */
+	if (!pmuv3_implemented(kvm_arm_pmu_get_pmuver_limit()))
+		return;
+
+	guard(mutex)(&arm_pmus_lock);
+
+	entry = kmalloc(sizeof(*entry), GFP_KERNEL);
+	if (!entry)
+		return;
+
+	entry->arm_pmu = pmu;
+	list_add_tail(&entry->entry, &arm_pmus);
+}
+
+static struct arm_pmu *kvm_pmu_probe_armpmu(void)
+{
+	struct arm_pmu_entry *entry;
+	struct arm_pmu *pmu;
+	int cpu;
+
+	guard(mutex)(&arm_pmus_lock);
+
+	/*
+	 * It is safe to use a stale cpu to iterate the list of PMUs so long as
+	 * the same value is used for the entirety of the loop. Given this, and
+	 * the fact that no percpu data is used for the lookup there is no need
+	 * to disable preemption.
+	 *
+	 * It is still necessary to get a valid cpu, though, to probe for the
+	 * default PMU instance as userspace is not required to specify a PMU
+	 * type. In order to uphold the preexisting behavior KVM selects the
+	 * PMU instance for the core during vcpu init. A dependent use
+	 * case would be a user with disdain of all things big.LITTLE that
+	 * affines the VMM to a particular cluster of cores.
+	 *
+	 * In any case, userspace should just do the sane thing and use the UAPI
+	 * to select a PMU type directly. But, be wary of the baggage being
+	 * carried here.
+	 */
+	cpu = raw_smp_processor_id();
+	list_for_each_entry(entry, &arm_pmus, entry) {
+		pmu = entry->arm_pmu;
+
+		if (cpumask_test_cpu(cpu, &pmu->supported_cpus))
+			return pmu;
+	}
+
+	return NULL;
+}
+
+static u64 __compute_pmceid(struct arm_pmu *pmu, bool pmceid1)
+{
+	u32 hi[2], lo[2];
+
+	bitmap_to_arr32(lo, pmu->pmceid_bitmap, ARMV8_PMUV3_MAX_COMMON_EVENTS);
+	bitmap_to_arr32(hi, pmu->pmceid_ext_bitmap, ARMV8_PMUV3_MAX_COMMON_EVENTS);
+
+	return ((u64)hi[pmceid1] << 32) | lo[pmceid1];
+}
+
+static u64 compute_pmceid0(struct arm_pmu *pmu)
+{
+	u64 val = __compute_pmceid(pmu, 0);
+
+	/* always support SW_INCR */
+	val |= BIT(ARMV8_PMUV3_PERFCTR_SW_INCR);
+	/* always support CHAIN */
+	val |= BIT(ARMV8_PMUV3_PERFCTR_CHAIN);
+	return val;
+}
+
+static u64 compute_pmceid1(struct arm_pmu *pmu)
+{
+	u64 val = __compute_pmceid(pmu, 1);
+
+	/*
+	 * Don't advertise STALL_SLOT*, as PMMIR_EL0 is handled
+	 * as RAZ
+	 */
+	val &= ~(BIT_ULL(ARMV8_PMUV3_PERFCTR_STALL_SLOT - 32) |
+		 BIT_ULL(ARMV8_PMUV3_PERFCTR_STALL_SLOT_FRONTEND - 32) |
+		 BIT_ULL(ARMV8_PMUV3_PERFCTR_STALL_SLOT_BACKEND - 32));
+	return val;
+}
+
+u64 kvm_pmu_get_pmceid(struct kvm_vcpu *vcpu, bool pmceid1)
+{
+	struct arm_pmu *cpu_pmu = vcpu->kvm->arch.arm_pmu;
+	unsigned long *bmap = vcpu->kvm->arch.pmu_filter;
+	u64 val, mask = 0;
+	int base, i, nr_events;
+
+	if (!pmceid1) {
+		val = compute_pmceid0(cpu_pmu);
+		base = 0;
+	} else {
+		val = compute_pmceid1(cpu_pmu);
+		base = 32;
+	}
+
+	if (!bmap)
+		return val;
+
+	nr_events = kvm_pmu_event_mask(vcpu->kvm) + 1;
+
+	for (i = 0; i < 32; i += 8) {
+		u64 byte;
+
+		byte = bitmap_get_value8(bmap, base + i);
+		mask |= byte << i;
+		if (nr_events >= (0x4000 + base + 32)) {
+			byte = bitmap_get_value8(bmap, 0x4000 + base + i);
+			mask |= byte << (32 + i);
+		}
+	}
+
+	return val & mask;
+}
+
+/*
+ * When perf interrupt is an NMI, we cannot safely notify the vcpu corresponding
+ * to the event.
+ * This is why we need a callback to do it once outside of the NMI context.
+ */
+static void kvm_pmu_perf_overflow_notify_vcpu(struct irq_work *work)
+{
+	struct kvm_vcpu *vcpu;
+
+	vcpu = container_of(work, struct kvm_vcpu, arch.pmu.overflow_work);
+	kvm_vcpu_kick(vcpu);
+}
+
+static u32 __kvm_pmu_event_mask(unsigned int pmuver)
+{
+	switch (pmuver) {
+	case ID_AA64DFR0_EL1_PMUVer_IMP:
+		return GENMASK(9, 0);
+	case ID_AA64DFR0_EL1_PMUVer_V3P1:
+	case ID_AA64DFR0_EL1_PMUVer_V3P4:
+	case ID_AA64DFR0_EL1_PMUVer_V3P5:
+	case ID_AA64DFR0_EL1_PMUVer_V3P7:
+		return GENMASK(15, 0);
+	default:		/* Shouldn't be here, just for sanity */
+		WARN_ONCE(1, "Unknown PMU version %d\n", pmuver);
+		return 0;
+	}
+}
+
+u32 kvm_pmu_event_mask(struct kvm *kvm)
+{
+	u64 dfr0 = kvm_read_vm_id_reg(kvm, SYS_ID_AA64DFR0_EL1);
+	u8 pmuver = SYS_FIELD_GET(ID_AA64DFR0_EL1, PMUVer, dfr0);
+
+	return __kvm_pmu_event_mask(pmuver);
+}
+
+u64 kvm_pmu_evtyper_mask(struct kvm *kvm)
+{
+	u64 mask = ARMV8_PMU_EXCLUDE_EL1 | ARMV8_PMU_EXCLUDE_EL0 |
+		   kvm_pmu_event_mask(kvm);
+
+	if (kvm_has_feat(kvm, ID_AA64PFR0_EL1, EL2, IMP))
+		mask |= ARMV8_PMU_INCLUDE_EL2;
+
+	if (kvm_has_feat(kvm, ID_AA64PFR0_EL1, EL3, IMP))
+		mask |= ARMV8_PMU_EXCLUDE_NS_EL0 |
+			ARMV8_PMU_EXCLUDE_NS_EL1 |
+			ARMV8_PMU_EXCLUDE_EL3;
+
+	return mask;
+}
+
+static void kvm_pmu_update_state(struct kvm_vcpu *vcpu)
+{
+	struct kvm_pmu *pmu = &vcpu->arch.pmu;
+	bool overflow;
+
+	overflow = kvm_pmu_overflow_status(vcpu);
+	if (pmu->irq_level == overflow)
+		return;
+
+	pmu->irq_level = overflow;
+
+	if (likely(irqchip_in_kernel(vcpu->kvm))) {
+		int ret = kvm_vgic_inject_irq(vcpu->kvm, vcpu,
+					      pmu->irq_num, overflow, pmu);
+		WARN_ON(ret);
+	}
+}
+
+/**
+ * kvm_pmu_flush_hwstate - flush pmu state to cpu
+ * @vcpu: The vcpu pointer
+ *
+ * Check if the PMU has overflowed while we were running in the host, and inject
+ * an interrupt if that was the case.
+ */
+void kvm_pmu_flush_hwstate(struct kvm_vcpu *vcpu)
+{
+	kvm_pmu_update_state(vcpu);
+}
+
+/**
+ * kvm_pmu_sync_hwstate - sync pmu state from cpu
+ * @vcpu: The vcpu pointer
+ *
+ * Check if the PMU has overflowed while we were running in the guest, and
+ * inject an interrupt if that was the case.
+ */
+void kvm_pmu_sync_hwstate(struct kvm_vcpu *vcpu)
+{
+	kvm_pmu_update_state(vcpu);
+}
+
+int kvm_arm_pmu_v3_enable(struct kvm_vcpu *vcpu)
+{
+	if (!vcpu->arch.pmu.created)
+		return -EINVAL;
+
+	/*
+	 * A valid interrupt configuration for the PMU is either to have a
+	 * properly configured interrupt number and using an in-kernel
+	 * irqchip, or to not have an in-kernel GIC and not set an IRQ.
+	 */
+	if (irqchip_in_kernel(vcpu->kvm)) {
+		int irq = vcpu->arch.pmu.irq_num;
+		/*
+		 * If we are using an in-kernel vgic, at this point we know
+		 * the vgic will be initialized, so we can check the PMU irq
+		 * number against the dimensions of the vgic and make sure
+		 * it's valid.
+		 */
+		if (!irq_is_ppi(irq) && !vgic_valid_spi(vcpu->kvm, irq))
+			return -EINVAL;
+	} else if (kvm_arm_pmu_irq_initialized(vcpu)) {
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int kvm_arm_pmu_v3_init(struct kvm_vcpu *vcpu)
+{
+	if (irqchip_in_kernel(vcpu->kvm)) {
+		int ret;
+
+		/*
+		 * If using the PMU with an in-kernel virtual GIC
+		 * implementation, we require the GIC to be already
+		 * initialized when initializing the PMU.
+		 */
+		if (!vgic_initialized(vcpu->kvm))
+			return -ENODEV;
+
+		if (!kvm_arm_pmu_irq_initialized(vcpu))
+			return -ENXIO;
+
+		ret = kvm_vgic_set_owner(vcpu, vcpu->arch.pmu.irq_num,
+					 &vcpu->arch.pmu);
+		if (ret)
+			return ret;
+	}
+
+	init_irq_work(&vcpu->arch.pmu.overflow_work,
+		      kvm_pmu_perf_overflow_notify_vcpu);
+
+	vcpu->arch.pmu.created = true;
+	return 0;
+}
+
+/*
+ * For one VM the interrupt type must be same for each vcpu.
+ * As a PPI, the interrupt number is the same for all vcpus,
+ * while as an SPI it must be a separate number per vcpu.
+ */
+static bool pmu_irq_is_valid(struct kvm *kvm, int irq)
+{
+	unsigned long i;
+	struct kvm_vcpu *vcpu;
+
+	kvm_for_each_vcpu(i, vcpu, kvm) {
+		if (!kvm_arm_pmu_irq_initialized(vcpu))
+			continue;
+
+		if (irq_is_ppi(irq)) {
+			if (vcpu->arch.pmu.irq_num != irq)
+				return false;
+		} else {
+			if (vcpu->arch.pmu.irq_num == irq)
+				return false;
+		}
+	}
+
+	return true;
+}
+
+/**
+ * kvm_arm_pmu_get_max_counters - Return the max number of PMU counters.
+ * @kvm: The kvm pointer
+ */
+u8 kvm_arm_pmu_get_max_counters(struct kvm *kvm)
+{
+	struct arm_pmu *arm_pmu = kvm->arch.arm_pmu;
+
+	/*
+	 * PMUv3 requires that all event counters are capable of counting any
+	 * event, though the same may not be true of non-PMUv3 hardware.
+	 */
+	if (cpus_have_final_cap(ARM64_WORKAROUND_PMUV3_IMPDEF_TRAPS))
+		return 1;
+
+	/*
+	 * The arm_pmu->cntr_mask considers the fixed counter(s) as well.
+	 * Ignore those and return only the general-purpose counters.
+	 */
+	return bitmap_weight(arm_pmu->cntr_mask, ARMV8_PMU_MAX_GENERAL_COUNTERS);
+}
+
+static void kvm_arm_set_nr_counters(struct kvm *kvm, unsigned int nr)
+{
+	kvm->arch.nr_pmu_counters = nr;
+
+	/* Reset MDCR_EL2.HPMN behind the vcpus' back... */
+	if (test_bit(KVM_ARM_VCPU_HAS_EL2, kvm->arch.vcpu_features)) {
+		struct kvm_vcpu *vcpu;
+		unsigned long i;
+
+		kvm_for_each_vcpu(i, vcpu, kvm) {
+			u64 val = __vcpu_sys_reg(vcpu, MDCR_EL2);
+
+			val &= ~MDCR_EL2_HPMN;
+			val |= FIELD_PREP(MDCR_EL2_HPMN, kvm->arch.nr_pmu_counters);
+			__vcpu_assign_sys_reg(vcpu, MDCR_EL2, val);
+		}
+	}
+}
+
+static void kvm_arm_set_pmu(struct kvm *kvm, struct arm_pmu *arm_pmu)
+{
+	lockdep_assert_held(&kvm->arch.config_lock);
+
+	kvm->arch.arm_pmu = arm_pmu;
+	kvm_arm_set_nr_counters(kvm, kvm_arm_pmu_get_max_counters(kvm));
+}
+
+/**
+ * kvm_arm_set_default_pmu - No PMU set, get the default one.
+ * @kvm: The kvm pointer
+ *
+ * The observant among you will notice that the supported_cpus
+ * mask does not get updated for the default PMU even though it
+ * is quite possible the selected instance supports only a
+ * subset of cores in the system. This is intentional, and
+ * upholds the preexisting behavior on heterogeneous systems
+ * where vCPUs can be scheduled on any core but the guest
+ * counters could stop working.
+ */
+int kvm_arm_set_default_pmu(struct kvm *kvm)
+{
+	struct arm_pmu *arm_pmu = kvm_pmu_probe_armpmu();
+
+	if (!arm_pmu)
+		return -ENODEV;
+
+	kvm_arm_set_pmu(kvm, arm_pmu);
+	return 0;
+}
+
+static int kvm_arm_pmu_v3_set_pmu(struct kvm_vcpu *vcpu, int pmu_id)
+{
+	struct kvm *kvm = vcpu->kvm;
+	struct arm_pmu_entry *entry;
+	struct arm_pmu *arm_pmu;
+	int ret = -ENXIO;
+
+	lockdep_assert_held(&kvm->arch.config_lock);
+	mutex_lock(&arm_pmus_lock);
+
+	list_for_each_entry(entry, &arm_pmus, entry) {
+		arm_pmu = entry->arm_pmu;
+		if (arm_pmu->pmu.type == pmu_id) {
+			if (kvm_vm_has_ran_once(kvm) ||
+			    (kvm->arch.pmu_filter && kvm->arch.arm_pmu != arm_pmu)) {
+				ret = -EBUSY;
+				break;
+			}
+
+			kvm_arm_set_pmu(kvm, arm_pmu);
+			cpumask_copy(kvm->arch.supported_cpus, &arm_pmu->supported_cpus);
+			ret = 0;
+			break;
+		}
+	}
+
+	mutex_unlock(&arm_pmus_lock);
+	return ret;
+}
+
+static int kvm_arm_pmu_v3_set_nr_counters(struct kvm_vcpu *vcpu, unsigned int n)
+{
+	struct kvm *kvm = vcpu->kvm;
+
+	if (!kvm->arch.arm_pmu)
+		return -EINVAL;
+
+	if (n > kvm_arm_pmu_get_max_counters(kvm))
+		return -EINVAL;
+
+	kvm_arm_set_nr_counters(kvm, n);
+	return 0;
+}
+
+int kvm_arm_pmu_v3_set_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
+{
+	struct kvm *kvm = vcpu->kvm;
+
+	lockdep_assert_held(&kvm->arch.config_lock);
+
+	if (!kvm_vcpu_has_pmu(vcpu))
+		return -ENODEV;
+
+	if (vcpu->arch.pmu.created)
+		return -EBUSY;
+
+	switch (attr->attr) {
+	case KVM_ARM_VCPU_PMU_V3_IRQ: {
+		int __user *uaddr = (int __user *)(long)attr->addr;
+		int irq;
+
+		if (!irqchip_in_kernel(kvm))
+			return -EINVAL;
+
+		if (get_user(irq, uaddr))
+			return -EFAULT;
+
+		/* The PMU overflow interrupt can be a PPI or a valid SPI. */
+		if (!(irq_is_ppi(irq) || irq_is_spi(irq)))
+			return -EINVAL;
+
+		if (!pmu_irq_is_valid(kvm, irq))
+			return -EINVAL;
+
+		if (kvm_arm_pmu_irq_initialized(vcpu))
+			return -EBUSY;
+
+		kvm_debug("Set kvm ARM PMU irq: %d\n", irq);
+		vcpu->arch.pmu.irq_num = irq;
+		return 0;
+	}
+	case KVM_ARM_VCPU_PMU_V3_FILTER: {
+		u8 pmuver = kvm_arm_pmu_get_pmuver_limit();
+		struct kvm_pmu_event_filter __user *uaddr;
+		struct kvm_pmu_event_filter filter;
+		int nr_events;
+
+		/*
+		 * Allow userspace to specify an event filter for the entire
+		 * event range supported by PMUVer of the hardware, rather
+		 * than the guest's PMUVer for KVM backward compatibility.
+		 */
+		nr_events = __kvm_pmu_event_mask(pmuver) + 1;
+
+		uaddr = (struct kvm_pmu_event_filter __user *)(long)attr->addr;
+
+		if (copy_from_user(&filter, uaddr, sizeof(filter)))
+			return -EFAULT;
+
+		if (((u32)filter.base_event + filter.nevents) > nr_events ||
+		    (filter.action != KVM_PMU_EVENT_ALLOW &&
+		     filter.action != KVM_PMU_EVENT_DENY))
+			return -EINVAL;
+
+		if (kvm_vm_has_ran_once(kvm))
+			return -EBUSY;
+
+		if (!kvm->arch.pmu_filter) {
+			kvm->arch.pmu_filter = bitmap_alloc(nr_events, GFP_KERNEL_ACCOUNT);
+			if (!kvm->arch.pmu_filter)
+				return -ENOMEM;
+
+			/*
+			 * The default depends on the first applied filter.
+			 * If it allows events, the default is to deny.
+			 * Conversely, if the first filter denies a set of
+			 * events, the default is to allow.
+			 */
+			if (filter.action == KVM_PMU_EVENT_ALLOW)
+				bitmap_zero(kvm->arch.pmu_filter, nr_events);
+			else
+				bitmap_fill(kvm->arch.pmu_filter, nr_events);
+		}
+
+		if (filter.action == KVM_PMU_EVENT_ALLOW)
+			bitmap_set(kvm->arch.pmu_filter, filter.base_event, filter.nevents);
+		else
+			bitmap_clear(kvm->arch.pmu_filter, filter.base_event, filter.nevents);
+
+		return 0;
+	}
+	case KVM_ARM_VCPU_PMU_V3_SET_PMU: {
+		int __user *uaddr = (int __user *)(long)attr->addr;
+		int pmu_id;
+
+		if (get_user(pmu_id, uaddr))
+			return -EFAULT;
+
+		return kvm_arm_pmu_v3_set_pmu(vcpu, pmu_id);
+	}
+	case KVM_ARM_VCPU_PMU_V3_SET_NR_COUNTERS: {
+		unsigned int __user *uaddr = (unsigned int __user *)(long)attr->addr;
+		unsigned int n;
+
+		if (get_user(n, uaddr))
+			return -EFAULT;
+
+		return kvm_arm_pmu_v3_set_nr_counters(vcpu, n);
+	}
+	case KVM_ARM_VCPU_PMU_V3_INIT:
+		return kvm_arm_pmu_v3_init(vcpu);
+	}
+
+	return -ENXIO;
+}
+
+int kvm_arm_pmu_v3_get_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
+{
+	switch (attr->attr) {
+	case KVM_ARM_VCPU_PMU_V3_IRQ: {
+		int __user *uaddr = (int __user *)(long)attr->addr;
+		int irq;
+
+		if (!irqchip_in_kernel(vcpu->kvm))
+			return -EINVAL;
+
+		if (!kvm_vcpu_has_pmu(vcpu))
+			return -ENODEV;
+
+		if (!kvm_arm_pmu_irq_initialized(vcpu))
+			return -ENXIO;
+
+		irq = vcpu->arch.pmu.irq_num;
+		return put_user(irq, uaddr);
+	}
+	}
+
+	return -ENXIO;
+}
+
+int kvm_arm_pmu_v3_has_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
+{
+	switch (attr->attr) {
+	case KVM_ARM_VCPU_PMU_V3_IRQ:
+	case KVM_ARM_VCPU_PMU_V3_INIT:
+	case KVM_ARM_VCPU_PMU_V3_FILTER:
+	case KVM_ARM_VCPU_PMU_V3_SET_PMU:
+	case KVM_ARM_VCPU_PMU_V3_SET_NR_COUNTERS:
+		if (kvm_vcpu_has_pmu(vcpu))
+			return 0;
+	}
+
+	return -ENXIO;
+}
+
+u8 kvm_arm_pmu_get_pmuver_limit(void)
+{
+	unsigned int pmuver;
+
+	pmuver = SYS_FIELD_GET(ID_AA64DFR0_EL1, PMUVer,
+			       read_sanitised_ftr_reg(SYS_ID_AA64DFR0_EL1));
+
+	/*
+	 * Spoof a barebones PMUv3 implementation if the system supports IMPDEF
+	 * traps of the PMUv3 sysregs
+	 */
+	if (cpus_have_final_cap(ARM64_WORKAROUND_PMUV3_IMPDEF_TRAPS))
+		return ID_AA64DFR0_EL1_PMUVer_IMP;
+
+	/*
+	 * Otherwise, treat IMPLEMENTATION DEFINED functionality as
+	 * unimplemented
+	 */
+	if (pmuver == ID_AA64DFR0_EL1_PMUVer_IMP_DEF)
+		return 0;
+
+	return min(pmuver, ID_AA64DFR0_EL1_PMUVer_V3P5);
+}
+
+u64 kvm_pmu_implemented_counter_mask(struct kvm_vcpu *vcpu)
+{
+	u64 val = FIELD_GET(ARMV8_PMU_PMCR_N, kvm_vcpu_read_pmcr(vcpu));
+
+	if (val == 0)
+		return BIT(ARMV8_PMU_CYCLE_IDX);
+	else
+		return GENMASK(val - 1, 0) | BIT(ARMV8_PMU_CYCLE_IDX);
+}
+
+u64 kvm_pmu_hyp_counter_mask(struct kvm_vcpu *vcpu)
+{
+	unsigned int hpmn, n;
+
+	if (!vcpu_has_nv(vcpu))
+		return 0;
+
+	hpmn = SYS_FIELD_GET(MDCR_EL2, HPMN, __vcpu_sys_reg(vcpu, MDCR_EL2));
+	n = vcpu->kvm->arch.nr_pmu_counters;
+
+	/*
+	 * Programming HPMN to a value greater than PMCR_EL0.N is
+	 * CONSTRAINED UNPREDICTABLE. Make the implementation choice that an
+	 * UNKNOWN number of counters (in our case, zero) are reserved for EL2.
+	 */
+	if (hpmn >= n)
+		return 0;
+
+	/*
+	 * Programming HPMN=0 is CONSTRAINED UNPREDICTABLE if FEAT_HPMN0 isn't
+	 * implemented. Since KVM's ability to emulate HPMN=0 does not directly
+	 * depend on hardware (all PMU registers are trapped), make the
+	 * implementation choice that all counters are included in the second
+	 * range reserved for EL2/EL3.
+	 */
+	return GENMASK(n - 1, hpmn);
+}
+
+bool kvm_pmu_counter_is_hyp(struct kvm_vcpu *vcpu, unsigned int idx)
+{
+	return kvm_pmu_hyp_counter_mask(vcpu) & BIT(idx);
+}
+
+u64 kvm_pmu_accessible_counter_mask(struct kvm_vcpu *vcpu)
+{
+	u64 mask = kvm_pmu_implemented_counter_mask(vcpu);
+
+	if (!vcpu_has_nv(vcpu) || vcpu_is_el2(vcpu))
+		return mask;
+
+	return mask & ~kvm_pmu_hyp_counter_mask(vcpu);
+}
+
+/**
+ * kvm_vcpu_read_pmcr - Read PMCR_EL0 register for the vCPU
+ * @vcpu: The vcpu pointer
+ */
+u64 kvm_vcpu_read_pmcr(struct kvm_vcpu *vcpu)
+{
+	u64 pmcr = __vcpu_sys_reg(vcpu, PMCR_EL0);
+	u64 n = vcpu->kvm->arch.nr_pmu_counters;
+
+	if (vcpu_has_nv(vcpu) && !vcpu_is_el2(vcpu))
+		n = FIELD_GET(MDCR_EL2_HPMN, __vcpu_sys_reg(vcpu, MDCR_EL2));
+
+	return u64_replace_bits(pmcr, n, ARMV8_PMU_PMCR_N);
+}
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 06/22] perf: arm_pmuv3: Introduce method to partition the PMU

[Colton Lewis]

For PMUv3, the register field MDCR_EL2.HPMN partitiones the PMU
counters into two ranges where counters 0..HPMN-1 are accessible by
EL1 and, if allowed, EL0 while counters HPMN..N are only accessible by
EL2.

Create module parameter reserved_host_counters to reserve a number of
counters for the host. This number is set at boot because the perf
subsystem assumes the number of counters will not change after the PMU
is probed.

Introduce the function armv8pmu_partition() to modify the PMU driver's
cntr_mask of available counters to exclude the counters being reserved
for the guest and record reserved_guest_counters as the maximum
allowable value for HPMN.

Due to the difficulty this feature would create for the driver running
at EL1 on the host, partitioning is only allowed in VHE mode. Working
on nVHE mode would require a hypercall for every counter access in the
driver because the counters reserved for the host by HPMN are only
accessible to EL2.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm/include/asm/arm_pmuv3.h   | 14 ++++++
 arch/arm64/include/asm/arm_pmuv3.h |  5 ++
 arch/arm64/include/asm/kvm_pmu.h   |  6 +++
 arch/arm64/kvm/Makefile            |  2 +-
 arch/arm64/kvm/pmu-part.c          | 23 ++++++++++
 drivers/perf/arm_pmuv3.c           | 74 +++++++++++++++++++++++++++++-
 include/linux/perf/arm_pmu.h       |  1 +
 7 files changed, 122 insertions(+), 3 deletions(-)
 create mode 100644 arch/arm64/kvm/pmu-part.c

diff --git a/arch/arm/include/asm/arm_pmuv3.h b/arch/arm/include/asm/arm_pmuv3.h
index 2ec0e5e83fc9..49b1f2d7842d 100644
--- a/arch/arm/include/asm/arm_pmuv3.h
+++ b/arch/arm/include/asm/arm_pmuv3.h
@@ -221,6 +221,10 @@ static inline bool kvm_pmu_counter_deferred(struct perf_event_attr *attr)
 	return false;
 }
 
+static inline bool kvm_pmu_partition_supported(void)
+{
+	return false;
+}
 static inline bool kvm_set_pmuserenr(u64 val)
 {
 	return false;
@@ -228,6 +232,11 @@ static inline bool kvm_set_pmuserenr(u64 val)
 
 static inline void kvm_vcpu_pmu_resync_el0(void) {}
 
+static inline bool has_vhe(void)
+{
+	return false;
+}
+
 /* PMU Version in DFR Register */
 #define ARMV8_PMU_DFR_VER_NI        0
 #define ARMV8_PMU_DFR_VER_V3P1      0x4
@@ -242,6 +251,11 @@ static inline bool pmuv3_implemented(int pmuver)
 		 pmuver == ARMV8_PMU_DFR_VER_NI);
 }
 
+static inline bool is_pmuv3p1(int pmuver)
+{
+	return pmuver >= ARMV8_PMU_DFR_VER_V3P1;
+}
+
 static inline bool is_pmuv3p4(int pmuver)
 {
 	return pmuver >= ARMV8_PMU_DFR_VER_V3P4;
diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h
index 32c003a7b810..e2057365ba73 100644
--- a/arch/arm64/include/asm/arm_pmuv3.h
+++ b/arch/arm64/include/asm/arm_pmuv3.h
@@ -173,6 +173,11 @@ static inline bool pmuv3_implemented(int pmuver)
 		 pmuver == ID_AA64DFR0_EL1_PMUVer_NI);
 }
 
+static inline bool is_pmuv3p1(int pmuver)
+{
+	return pmuver >= ID_AA64DFR0_EL1_PMUVer_V3P1;
+}
+
 static inline bool is_pmuv3p4(int pmuver)
 {
 	return pmuver >= ID_AA64DFR0_EL1_PMUVer_V3P4;
diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
index 6c961e877804..8a2ed02e157d 100644
--- a/arch/arm64/include/asm/kvm_pmu.h
+++ b/arch/arm64/include/asm/kvm_pmu.h
@@ -46,6 +46,7 @@ struct arm_pmu_entry {
 };
 
 bool kvm_supports_guest_pmuv3(void);
+bool kvm_pmu_partition_supported(void);
 #define kvm_arm_pmu_irq_initialized(v)	((v)->arch.pmu.irq_num >= VGIC_NR_SGIS)
 u64 kvm_pmu_get_counter_value(struct kvm_vcpu *vcpu, u64 select_idx);
 void kvm_pmu_set_counter_value(struct kvm_vcpu *vcpu, u64 select_idx, u64 val);
@@ -115,6 +116,11 @@ static inline bool kvm_supports_guest_pmuv3(void)
 	return false;
 }
 
+static inline bool kvm_pmu_partition_supported(void)
+{
+	return false;
+}
+
 #define kvm_arm_pmu_irq_initialized(v)	(false)
 static inline u64 kvm_pmu_get_counter_value(struct kvm_vcpu *vcpu,
 					    u64 select_idx)
diff --git a/arch/arm64/kvm/Makefile b/arch/arm64/kvm/Makefile
index 86035b311269..3edbaa57bbf2 100644
--- a/arch/arm64/kvm/Makefile
+++ b/arch/arm64/kvm/Makefile
@@ -23,7 +23,7 @@ kvm-y += arm.o mmu.o mmio.o psci.o hypercalls.o pvtime.o \
 	 vgic/vgic-mmio-v3.o vgic/vgic-kvm-device.o \
 	 vgic/vgic-its.o vgic/vgic-debug.o vgic/vgic-v3-nested.o
 
-kvm-$(CONFIG_HW_PERF_EVENTS)  += pmu-emul.o pmu.o
+kvm-$(CONFIG_HW_PERF_EVENTS)  += pmu-emul.o pmu-part.o pmu.o
 kvm-$(CONFIG_ARM64_PTR_AUTH)  += pauth.o
 kvm-$(CONFIG_PTDUMP_STAGE2_DEBUGFS) += ptdump.o
 
diff --git a/arch/arm64/kvm/pmu-part.c b/arch/arm64/kvm/pmu-part.c
new file mode 100644
index 000000000000..f04c580c19c0
--- /dev/null
+++ b/arch/arm64/kvm/pmu-part.c
@@ -0,0 +1,23 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2025 Google LLC
+ * Author: Colton Lewis <coltonlewis@google.com>
+ */
+
+#include <linux/kvm_host.h>
+
+#include <asm/kvm_pmu.h>
+
+/**
+ * kvm_pmu_partition_supported() - Determine if partitioning is possible
+ *
+ * Partitioning is only supported in VHE mode (with PMUv3, assumed
+ * since we are in the PMUv3 driver)
+ *
+ * Return: True if partitioning is possible, false otherwise
+ */
+bool kvm_pmu_partition_supported(void)
+{
+	return kvm_supports_guest_pmuv3() &&
+		has_vhe();
+}
diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c
index 3db9f4ed17e8..6358de6c9fab 100644
--- a/drivers/perf/arm_pmuv3.c
+++ b/drivers/perf/arm_pmuv3.c
@@ -35,6 +35,12 @@
 #define ARMV8_THUNDER_PERFCTR_L1I_CACHE_PREF_ACCESS		0xEC
 #define ARMV8_THUNDER_PERFCTR_L1I_CACHE_PREF_MISS		0xED
 
+static int reserved_host_counters __read_mostly = -1;
+
+module_param(reserved_host_counters, int, 0);
+MODULE_PARM_DESC(reserved_host_counters,
+		 "PMU Partition: -1 = No partition; +N = Reserve N counters for the host");
+
 /*
  * ARMv8 Architectural defined events, not all of these may
  * be supported on any given implementation. Unsupported events will
@@ -500,6 +506,11 @@ static void armv8pmu_pmcr_write(u64 val)
 	write_pmcr(val);
 }
 
+static u64 armv8pmu_pmcr_n_read(void)
+{
+	return FIELD_GET(ARMV8_PMU_PMCR_N, armv8pmu_pmcr_read());
+}
+
 static int armv8pmu_has_overflowed(u64 pmovsr)
 {
 	return !!(pmovsr & ARMV8_PMU_OVERFLOWED_MASK);
@@ -1195,6 +1206,58 @@ struct armv8pmu_probe_info {
 	bool present;
 };
 
+/**
+ * armv8pmu_reservation_is_valid() - Determine if reservation is allowed
+ * @host_counters: Number of host counters to reserve
+ *
+ * Determine if the number of host counters in the argument is an
+ * allowed reservation, 0 to NR_COUNTERS inclusive.
+ *
+ * Return: True if reservation allowed, false otherwise
+ */
+static bool armv8pmu_reservation_is_valid(int host_counters)
+{
+	return host_counters >= 0 &&
+		host_counters <= armv8pmu_pmcr_n_read();
+}
+
+/**
+ * armv8pmu_partition() - Partition the PMU
+ * @pmu: Pointer to pmu being partitioned
+ * @host_counters: Number of host counters to reserve
+ *
+ * Partition the given PMU by taking a number of host counters to
+ * reserve and, if it is a valid reservation, recording the
+ * corresponding HPMN value in the hpmn_max field of the PMU and
+ * clearing the guest-reserved counters from the counter mask.
+ *
+ * Return: 0 on success, -ERROR otherwise
+ */
+static int armv8pmu_partition(struct arm_pmu *pmu, int host_counters)
+{
+	u8 nr_counters;
+	u8 hpmn;
+
+	if (!armv8pmu_reservation_is_valid(host_counters))
+		return -EINVAL;
+
+	nr_counters = armv8pmu_pmcr_n_read();
+	hpmn = nr_counters - host_counters;
+
+	pmu->hpmn_max = hpmn;
+
+	bitmap_clear(pmu->cntr_mask, 0, hpmn);
+	bitmap_set(pmu->cntr_mask, hpmn, host_counters);
+	clear_bit(ARMV8_PMU_CYCLE_IDX, pmu->cntr_mask);
+
+	if (pmuv3_has_icntr())
+		clear_bit(ARMV8_PMU_INSTR_IDX, pmu->cntr_mask);
+
+	pr_info("Partitioned PMU with HPMN %u", hpmn);
+
+	return 0;
+}
+
 static void __armv8pmu_probe_pmu(void *info)
 {
 	struct armv8pmu_probe_info *probe = info;
@@ -1209,10 +1272,10 @@ static void __armv8pmu_probe_pmu(void *info)
 
 	cpu_pmu->pmuver = pmuver;
 	probe->present = true;
+	cpu_pmu->hpmn_max = -1;
 
 	/* Read the nb of CNTx counters supported from PMNC */
-	bitmap_set(cpu_pmu->cntr_mask,
-		   0, FIELD_GET(ARMV8_PMU_PMCR_N, armv8pmu_pmcr_read()));
+	bitmap_set(cpu_pmu->cntr_mask, 0, armv8pmu_pmcr_n_read());
 
 	/* Add the CPU cycles counter */
 	set_bit(ARMV8_PMU_CYCLE_IDX, cpu_pmu->cntr_mask);
@@ -1221,6 +1284,13 @@ static void __armv8pmu_probe_pmu(void *info)
 	if (pmuv3_has_icntr())
 		set_bit(ARMV8_PMU_INSTR_IDX, cpu_pmu->cntr_mask);
 
+	if (reserved_host_counters >= 0) {
+		if (kvm_pmu_partition_supported())
+			WARN_ON(armv8pmu_partition(cpu_pmu, reserved_host_counters));
+		else
+			pr_err("PMU partition is not supported");
+	}
+
 	pmceid[0] = pmceid_raw[0] = read_pmceid0();
 	pmceid[1] = pmceid_raw[1] = read_pmceid1();
 
diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h
index fb382bcd4f4b..071cd2d1681c 100644
--- a/include/linux/perf/arm_pmu.h
+++ b/include/linux/perf/arm_pmu.h
@@ -122,6 +122,7 @@ struct arm_pmu {
 
 	/* Only to be used by ACPI probing code */
 	unsigned long acpi_cpuid;
+	int		hpmn_max; /* MDCR_EL2.HPMN: counter partition pivot */
 };
 
 #define to_arm_pmu(p) (container_of(p, struct arm_pmu, pmu))
-- 
2.50.0.727.gbf7dc18ff4-goog

Re: [PATCH v3 06/22] perf: arm_pmuv3: Introduce method to partition the PMU

[Mark Rutland]

On Thu, Jun 26, 2025 at 08:04:42PM +0000, Colton Lewis wrote:
> For PMUv3, the register field MDCR_EL2.HPMN partitiones the PMU
> counters into two ranges where counters 0..HPMN-1 are accessible by
> EL1 and, if allowed, EL0 while counters HPMN..N are only accessible by
> EL2.
> 
> Create module parameter reserved_host_counters to reserve a number of
> counters for the host. This number is set at boot because the perf
> subsystem assumes the number of counters will not change after the PMU
> is probed.
> 
> Introduce the function armv8pmu_partition() to modify the PMU driver's
> cntr_mask of available counters to exclude the counters being reserved
> for the guest and record reserved_guest_counters as the maximum
> allowable value for HPMN.
> 
> Due to the difficulty this feature would create for the driver running
> at EL1 on the host, partitioning is only allowed in VHE mode. Working
> on nVHE mode would require a hypercall for every counter access in the
> driver because the counters reserved for the host by HPMN are only
> accessible to EL2.

It would be good if we could elaborate on this last point. When exactly
do we intend to configure HPMN (e.g. is that static, dynamic at
load/put, or dynamic at finer granularity)?

I ask becuase it's not immediately clear to me how this would break nVHE
without also breaking direct userspace access on VHE, unless we flip
HPMN dynamically at load/put, and this is only broken in some transient
windows on nVHE.

> 
> Signed-off-by: Colton Lewis <coltonlewis@google.com>
> ---
>  arch/arm/include/asm/arm_pmuv3.h   | 14 ++++++
>  arch/arm64/include/asm/arm_pmuv3.h |  5 ++
>  arch/arm64/include/asm/kvm_pmu.h   |  6 +++
>  arch/arm64/kvm/Makefile            |  2 +-
>  arch/arm64/kvm/pmu-part.c          | 23 ++++++++++

Maybe I'll contradict Oliver and Marc here (and whatever they say
rules), but IMO it'd be nice to spell out "partition" rather than "part"
here for clarity.

Mark.

Re: [PATCH v3 06/22] perf: arm_pmuv3: Introduce method to partition the PMU

[Oliver Upton]

On Mon, Jul 07, 2025 at 05:57:14PM +0100, Mark Rutland wrote:
> On Thu, Jun 26, 2025 at 08:04:42PM +0000, Colton Lewis wrote:
> > For PMUv3, the register field MDCR_EL2.HPMN partitiones the PMU
> > counters into two ranges where counters 0..HPMN-1 are accessible by
> > EL1 and, if allowed, EL0 while counters HPMN..N are only accessible by
> > EL2.
> > 
> > Create module parameter reserved_host_counters to reserve a number of
> > counters for the host. This number is set at boot because the perf
> > subsystem assumes the number of counters will not change after the PMU
> > is probed.
> > 
> > Introduce the function armv8pmu_partition() to modify the PMU driver's
> > cntr_mask of available counters to exclude the counters being reserved
> > for the guest and record reserved_guest_counters as the maximum
> > allowable value for HPMN.
> > 
> > Due to the difficulty this feature would create for the driver running
> > at EL1 on the host, partitioning is only allowed in VHE mode. Working
> > on nVHE mode would require a hypercall for every counter access in the
> > driver because the counters reserved for the host by HPMN are only
> > accessible to EL2.
> 
> It would be good if we could elaborate on this last point. When exactly
> do we intend to configure HPMN (e.g. is that static, dynamic at
> load/put, or dynamic at finer granularity)?
> 
> I ask becuase it's not immediately clear to me how this would break nVHE
> without also breaking direct userspace access on VHE, unless we flip
> HPMN dynamically at load/put, and this is only broken in some transient
> windows on nVHE.

Agree that KVM's HPMN can only take effect between vcpu_load() /
vcpu_put().

The changelog isn't correct regarding the complications of nVHE, though.
In order to support a 'partitioned' PMU on nVHE we'd need to explicitly
disable guest counters on every exit and reset HPMN to place all
counters in the 'first range'. Unless someone has a use case for this
stuff on nVHE I'm not too bothered by the VHE-only limitation.

> > 
> > Signed-off-by: Colton Lewis <coltonlewis@google.com>
> > ---
> >  arch/arm/include/asm/arm_pmuv3.h   | 14 ++++++
> >  arch/arm64/include/asm/arm_pmuv3.h |  5 ++
> >  arch/arm64/include/asm/kvm_pmu.h   |  6 +++
> >  arch/arm64/kvm/Makefile            |  2 +-
> >  arch/arm64/kvm/pmu-part.c          | 23 ++++++++++
> 
> Maybe I'll contradict Oliver and Marc here (and whatever they say
> rules), but IMO it'd be nice to spell out "partition" rather than "part"
> here for clarity.

I'm not too big of a fan of the naming here either. I'd prefer something
like "pmu-direct". Partitioning is just a side effect of how we're
allocating counters currently and most of this implementation could be
reused if we pass the entire PMU to the guest in the future.

With that being said -- Colton I'd focus on getting these patches in
shape while we figure out what color we want it ;-)

Thanks,
Oliver

Re: [PATCH v3 06/22] perf: arm_pmuv3: Introduce method to partition the PMU

[Colton Lewis]

Oliver Upton <oliver.upton@linux.dev> writes:

> On Mon, Jul 07, 2025 at 05:57:14PM +0100, Mark Rutland wrote:
>> On Thu, Jun 26, 2025 at 08:04:42PM +0000, Colton Lewis wrote:
>> > For PMUv3, the register field MDCR_EL2.HPMN partitiones the PMU
>> > counters into two ranges where counters 0..HPMN-1 are accessible by
>> > EL1 and, if allowed, EL0 while counters HPMN..N are only accessible by
>> > EL2.
>> >
>> > Create module parameter reserved_host_counters to reserve a number of
>> > counters for the host. This number is set at boot because the perf
>> > subsystem assumes the number of counters will not change after the PMU
>> > is probed.
>> >
>> > Introduce the function armv8pmu_partition() to modify the PMU driver's
>> > cntr_mask of available counters to exclude the counters being reserved
>> > for the guest and record reserved_guest_counters as the maximum
>> > allowable value for HPMN.
>> >
>> > Due to the difficulty this feature would create for the driver running
>> > at EL1 on the host, partitioning is only allowed in VHE mode. Working
>> > on nVHE mode would require a hypercall for every counter access in the
>> > driver because the counters reserved for the host by HPMN are only
>> > accessible to EL2.

>> It would be good if we could elaborate on this last point. When exactly
>> do we intend to configure HPMN (e.g. is that static, dynamic at
>> load/put, or dynamic at finer granularity)?

>> I ask becuase it's not immediately clear to me how this would break nVHE
>> without also breaking direct userspace access on VHE, unless we flip
>> HPMN dynamically at load/put, and this is only broken in some transient
>> windows on nVHE.

> Agree that KVM's HPMN can only take effect between vcpu_load() /
> vcpu_put().

> The changelog isn't correct regarding the complications of nVHE, though.
> In order to support a 'partitioned' PMU on nVHE we'd need to explicitly
> disable guest counters on every exit and reset HPMN to place all
> counters in the 'first range'. Unless someone has a use case for this
> stuff on nVHE I'm not too bothered by the VHE-only limitation.

I'll fix this.

>> >
>> > Signed-off-by: Colton Lewis <coltonlewis@google.com>
>> > ---
>> >  arch/arm/include/asm/arm_pmuv3.h   | 14 ++++++
>> >  arch/arm64/include/asm/arm_pmuv3.h |  5 ++
>> >  arch/arm64/include/asm/kvm_pmu.h   |  6 +++
>> >  arch/arm64/kvm/Makefile            |  2 +-
>> >  arch/arm64/kvm/pmu-part.c          | 23 ++++++++++

>> Maybe I'll contradict Oliver and Marc here (and whatever they say
>> rules), but IMO it'd be nice to spell out "partition" rather than "part"
>> here for clarity.

> I'm not too big of a fan of the naming here either. I'd prefer something
> like "pmu-direct". Partitioning is just a side effect of how we're
> allocating counters currently and most of this implementation could be
> reused if we pass the entire PMU to the guest in the future.

Sure.

> With that being said -- Colton I'd focus on getting these patches in
> shape while we figure out what color we want it ;-)

> Thanks,
> Oliver

Trust me I'm working on it.

Re: [PATCH v3 06/22] perf: arm_pmuv3: Introduce method to partition the PMU

[Oliver Upton]

On Tue, Jul 08, 2025 at 10:38:35PM +0000, Colton Lewis wrote:
> Oliver Upton <oliver.upton@linux.dev> writes:
> 
> > On Mon, Jul 07, 2025 at 05:57:14PM +0100, Mark Rutland wrote:
> > > On Thu, Jun 26, 2025 at 08:04:42PM +0000, Colton Lewis wrote:
> > > > For PMUv3, the register field MDCR_EL2.HPMN partitiones the PMU
> > > > counters into two ranges where counters 0..HPMN-1 are accessible by
> > > > EL1 and, if allowed, EL0 while counters HPMN..N are only accessible by
> > > > EL2.
> > > >
> > > > Create module parameter reserved_host_counters to reserve a number of
> > > > counters for the host. This number is set at boot because the perf
> > > > subsystem assumes the number of counters will not change after the PMU
> > > > is probed.
> > > >
> > > > Introduce the function armv8pmu_partition() to modify the PMU driver's
> > > > cntr_mask of available counters to exclude the counters being reserved
> > > > for the guest and record reserved_guest_counters as the maximum
> > > > allowable value for HPMN.
> > > >
> > > > Due to the difficulty this feature would create for the driver running
> > > > at EL1 on the host, partitioning is only allowed in VHE mode. Working
> > > > on nVHE mode would require a hypercall for every counter access in the
> > > > driver because the counters reserved for the host by HPMN are only
> > > > accessible to EL2.
> 
> > > It would be good if we could elaborate on this last point. When exactly
> > > do we intend to configure HPMN (e.g. is that static, dynamic at
> > > load/put, or dynamic at finer granularity)?
> 
> > > I ask becuase it's not immediately clear to me how this would break nVHE
> > > without also breaking direct userspace access on VHE, unless we flip
> > > HPMN dynamically at load/put, and this is only broken in some transient
> > > windows on nVHE.
> 
> > Agree that KVM's HPMN can only take effect between vcpu_load() /
> > vcpu_put().
> 
> > The changelog isn't correct regarding the complications of nVHE, though.
> > In order to support a 'partitioned' PMU on nVHE we'd need to explicitly
> > disable guest counters on every exit and reset HPMN to place all
> > counters in the 'first range'. Unless someone has a use case for this
> > stuff on nVHE I'm not too bothered by the VHE-only limitation.
> 
> I'll fix this.
> 
> > > >
> > > > Signed-off-by: Colton Lewis <coltonlewis@google.com>
> > > > ---
> > > >  arch/arm/include/asm/arm_pmuv3.h   | 14 ++++++
> > > >  arch/arm64/include/asm/arm_pmuv3.h |  5 ++
> > > >  arch/arm64/include/asm/kvm_pmu.h   |  6 +++
> > > >  arch/arm64/kvm/Makefile            |  2 +-
> > > >  arch/arm64/kvm/pmu-part.c          | 23 ++++++++++
> 
> > > Maybe I'll contradict Oliver and Marc here (and whatever they say
> > > rules), but IMO it'd be nice to spell out "partition" rather than "part"
> > > here for clarity.
> 
> > I'm not too big of a fan of the naming here either. I'd prefer something
> > like "pmu-direct". Partitioning is just a side effect of how we're
> > allocating counters currently and most of this implementation could be
> > reused if we pass the entire PMU to the guest in the future.
> 
> Sure.
> 
> > With that being said -- Colton I'd focus on getting these patches in
> > shape while we figure out what color we want it ;-)
> 
> > Thanks,
> > Oliver
> 
> Trust me I'm working on it.

I know you are -- this comment wasn't meant to add pressure to you. Just
noting that we might bikeshed a bit further on the naming here but I
wouldn't worry much about it for the time being.

sed isn't _that_ difficult :)

Thanks,
Oliver

[PATCH v3 07/22] perf: arm_pmuv3: Generalize counter bitmasks

[Colton Lewis]

The OVSR bitmasks are valid for enable and interrupt registers as well as
overflow registers. Generalize the names.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 drivers/perf/arm_pmuv3.c       |  4 ++--
 include/linux/perf/arm_pmuv3.h | 14 +++++++-------
 2 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c
index 6358de6c9fab..3bc016afea34 100644
--- a/drivers/perf/arm_pmuv3.c
+++ b/drivers/perf/arm_pmuv3.c
@@ -513,7 +513,7 @@ static u64 armv8pmu_pmcr_n_read(void)
 
 static int armv8pmu_has_overflowed(u64 pmovsr)
 {
-	return !!(pmovsr & ARMV8_PMU_OVERFLOWED_MASK);
+	return !!(pmovsr & ARMV8_PMU_CNT_MASK_ALL);
 }
 
 static int armv8pmu_counter_has_overflowed(u64 pmnc, int idx)
@@ -749,7 +749,7 @@ static u64 armv8pmu_getreset_flags(void)
 	value = read_pmovsclr();
 
 	/* Write to clear flags */
-	value &= ARMV8_PMU_OVERFLOWED_MASK;
+	value &= ARMV8_PMU_CNT_MASK_ALL;
 	write_pmovsclr(value);
 
 	return value;
diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h
index d698efba28a2..fd2a34b4a64d 100644
--- a/include/linux/perf/arm_pmuv3.h
+++ b/include/linux/perf/arm_pmuv3.h
@@ -224,14 +224,14 @@
 				 ARMV8_PMU_PMCR_LC | ARMV8_PMU_PMCR_LP)
 
 /*
- * PMOVSR: counters overflow flag status reg
+ * Counter bitmask layouts for overflow, enable, and interrupts
  */
-#define ARMV8_PMU_OVSR_P		GENMASK(30, 0)
-#define ARMV8_PMU_OVSR_C		BIT(31)
-#define ARMV8_PMU_OVSR_F		BIT_ULL(32) /* arm64 only */
-/* Mask for writable bits is both P and C fields */
-#define ARMV8_PMU_OVERFLOWED_MASK	(ARMV8_PMU_OVSR_P | ARMV8_PMU_OVSR_C | \
-					ARMV8_PMU_OVSR_F)
+#define ARMV8_PMU_CNT_MASK_P		GENMASK(30, 0)
+#define ARMV8_PMU_CNT_MASK_C		BIT(31)
+#define ARMV8_PMU_CNT_MASK_F		BIT_ULL(32) /* arm64 only */
+#define ARMV8_PMU_CNT_MASK_ALL		(ARMV8_PMU_CNT_MASK_P | \
+					 ARMV8_PMU_CNT_MASK_C | \
+					 ARMV8_PMU_CNT_MASK_F)
 
 /*
  * PMXEVTYPER: Event selection reg
-- 
2.50.0.727.gbf7dc18ff4-goog

Re: [PATCH v3 07/22] perf: arm_pmuv3: Generalize counter bitmasks

[Mark Rutland]

On Thu, Jun 26, 2025 at 08:04:43PM +0000, Colton Lewis wrote:
> The OVSR bitmasks are valid for enable and interrupt registers as well as
> overflow registers. Generalize the names.
> 
> Signed-off-by: Colton Lewis <coltonlewis@google.com>

FWIW, this looks fine to me, so:

Acked-by: Mark Rutland <mark.rutland@arm.com>

Mark.

> ---
>  drivers/perf/arm_pmuv3.c       |  4 ++--
>  include/linux/perf/arm_pmuv3.h | 14 +++++++-------
>  2 files changed, 9 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c
> index 6358de6c9fab..3bc016afea34 100644
> --- a/drivers/perf/arm_pmuv3.c
> +++ b/drivers/perf/arm_pmuv3.c
> @@ -513,7 +513,7 @@ static u64 armv8pmu_pmcr_n_read(void)
>  
>  static int armv8pmu_has_overflowed(u64 pmovsr)
>  {
> -	return !!(pmovsr & ARMV8_PMU_OVERFLOWED_MASK);
> +	return !!(pmovsr & ARMV8_PMU_CNT_MASK_ALL);
>  }
>  
>  static int armv8pmu_counter_has_overflowed(u64 pmnc, int idx)
> @@ -749,7 +749,7 @@ static u64 armv8pmu_getreset_flags(void)
>  	value = read_pmovsclr();
>  
>  	/* Write to clear flags */
> -	value &= ARMV8_PMU_OVERFLOWED_MASK;
> +	value &= ARMV8_PMU_CNT_MASK_ALL;
>  	write_pmovsclr(value);
>  
>  	return value;
> diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h
> index d698efba28a2..fd2a34b4a64d 100644
> --- a/include/linux/perf/arm_pmuv3.h
> +++ b/include/linux/perf/arm_pmuv3.h
> @@ -224,14 +224,14 @@
>  				 ARMV8_PMU_PMCR_LC | ARMV8_PMU_PMCR_LP)
>  
>  /*
> - * PMOVSR: counters overflow flag status reg
> + * Counter bitmask layouts for overflow, enable, and interrupts
>   */
> -#define ARMV8_PMU_OVSR_P		GENMASK(30, 0)
> -#define ARMV8_PMU_OVSR_C		BIT(31)
> -#define ARMV8_PMU_OVSR_F		BIT_ULL(32) /* arm64 only */
> -/* Mask for writable bits is both P and C fields */
> -#define ARMV8_PMU_OVERFLOWED_MASK	(ARMV8_PMU_OVSR_P | ARMV8_PMU_OVSR_C | \
> -					ARMV8_PMU_OVSR_F)
> +#define ARMV8_PMU_CNT_MASK_P		GENMASK(30, 0)
> +#define ARMV8_PMU_CNT_MASK_C		BIT(31)
> +#define ARMV8_PMU_CNT_MASK_F		BIT_ULL(32) /* arm64 only */
> +#define ARMV8_PMU_CNT_MASK_ALL		(ARMV8_PMU_CNT_MASK_P | \
> +					 ARMV8_PMU_CNT_MASK_C | \
> +					 ARMV8_PMU_CNT_MASK_F)
>  
>  /*
>   * PMXEVTYPER: Event selection reg
> -- 
> 2.50.0.727.gbf7dc18ff4-goog
> 

Re: [PATCH v3 07/22] perf: arm_pmuv3: Generalize counter bitmasks

[Colton Lewis]

Mark Rutland <mark.rutland@arm.com> writes:

> On Thu, Jun 26, 2025 at 08:04:43PM +0000, Colton Lewis wrote:
>> The OVSR bitmasks are valid for enable and interrupt registers as well as
>> overflow registers. Generalize the names.

>> Signed-off-by: Colton Lewis <coltonlewis@google.com>

> FWIW, this looks fine to me, so:

> Acked-by: Mark Rutland <mark.rutland@arm.com>

> Mark.

Thanks.
>> ---
>>   drivers/perf/arm_pmuv3.c       |  4 ++--
>>   include/linux/perf/arm_pmuv3.h | 14 +++++++-------
>>   2 files changed, 9 insertions(+), 9 deletions(-)

>> diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c
>> index 6358de6c9fab..3bc016afea34 100644
>> --- a/drivers/perf/arm_pmuv3.c
>> +++ b/drivers/perf/arm_pmuv3.c
>> @@ -513,7 +513,7 @@ static u64 armv8pmu_pmcr_n_read(void)

>>   static int armv8pmu_has_overflowed(u64 pmovsr)
>>   {
>> -	return !!(pmovsr & ARMV8_PMU_OVERFLOWED_MASK);
>> +	return !!(pmovsr & ARMV8_PMU_CNT_MASK_ALL);
>>   }

>>   static int armv8pmu_counter_has_overflowed(u64 pmnc, int idx)
>> @@ -749,7 +749,7 @@ static u64 armv8pmu_getreset_flags(void)
>>   	value = read_pmovsclr();

>>   	/* Write to clear flags */
>> -	value &= ARMV8_PMU_OVERFLOWED_MASK;
>> +	value &= ARMV8_PMU_CNT_MASK_ALL;
>>   	write_pmovsclr(value);

>>   	return value;
>> diff --git a/include/linux/perf/arm_pmuv3.h  
>> b/include/linux/perf/arm_pmuv3.h
>> index d698efba28a2..fd2a34b4a64d 100644
>> --- a/include/linux/perf/arm_pmuv3.h
>> +++ b/include/linux/perf/arm_pmuv3.h
>> @@ -224,14 +224,14 @@
>>   				 ARMV8_PMU_PMCR_LC | ARMV8_PMU_PMCR_LP)

>>   /*
>> - * PMOVSR: counters overflow flag status reg
>> + * Counter bitmask layouts for overflow, enable, and interrupts
>>    */
>> -#define ARMV8_PMU_OVSR_P		GENMASK(30, 0)
>> -#define ARMV8_PMU_OVSR_C		BIT(31)
>> -#define ARMV8_PMU_OVSR_F		BIT_ULL(32) /* arm64 only */
>> -/* Mask for writable bits is both P and C fields */
>> -#define ARMV8_PMU_OVERFLOWED_MASK	(ARMV8_PMU_OVSR_P | ARMV8_PMU_OVSR_C  
>> | \
>> -					ARMV8_PMU_OVSR_F)
>> +#define ARMV8_PMU_CNT_MASK_P		GENMASK(30, 0)
>> +#define ARMV8_PMU_CNT_MASK_C		BIT(31)
>> +#define ARMV8_PMU_CNT_MASK_F		BIT_ULL(32) /* arm64 only */
>> +#define ARMV8_PMU_CNT_MASK_ALL		(ARMV8_PMU_CNT_MASK_P | \
>> +					 ARMV8_PMU_CNT_MASK_C | \
>> +					 ARMV8_PMU_CNT_MASK_F)

>>   /*
>>    * PMXEVTYPER: Event selection reg
>> --
>> 2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 08/22] perf: arm_pmuv3: Keep out of guest counter partition

[Colton Lewis]

If the PMU is partitioned, keep the driver out of the guest counter
partition and only use the host counter partition. Partitioning is
defined by the MDCR_EL2.HPMN register field and the maximum value KVM
can use is saved in cpu_pmu->hpmn_max. The range 0..HPMN-1 is
accessible by EL1 and EL0 while HPMN..PMCR.N is reserved for EL2.

Define some functions that take HPMN as an argument and construct
mutually exclusive bitmaps for testing which partition a particular
counter is in. Note that despite their different position in the
bitmap, the cycle and instruction counters are always in the guest
partition.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm/include/asm/arm_pmuv3.h | 18 +++++++
 arch/arm64/include/asm/kvm_pmu.h | 24 +++++++++
 arch/arm64/kvm/pmu-part.c        | 84 ++++++++++++++++++++++++++++++++
 drivers/perf/arm_pmuv3.c         | 36 ++++++++++++--
 4 files changed, 158 insertions(+), 4 deletions(-)

diff --git a/arch/arm/include/asm/arm_pmuv3.h b/arch/arm/include/asm/arm_pmuv3.h
index 49b1f2d7842d..5f6269039f44 100644
--- a/arch/arm/include/asm/arm_pmuv3.h
+++ b/arch/arm/include/asm/arm_pmuv3.h
@@ -231,6 +231,24 @@ static inline bool kvm_set_pmuserenr(u64 val)
 }
 
 static inline void kvm_vcpu_pmu_resync_el0(void) {}
+static inline void kvm_pmu_host_counters_enable(void) {}
+static inline void kvm_pmu_host_counters_disable(void) {}
+
+static inline bool kvm_pmu_is_partitioned(struct arm_pmu *pmu)
+{
+	return false;
+}
+
+static inline u64 kvm_pmu_host_counter_mask(struct arm_pmu *pmu)
+{
+	return ~0;
+}
+
+static inline u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu)
+{
+	return ~0;
+}
+
 
 static inline bool has_vhe(void)
 {
diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
index 8a2ed02e157d..6328e90952ba 100644
--- a/arch/arm64/include/asm/kvm_pmu.h
+++ b/arch/arm64/include/asm/kvm_pmu.h
@@ -88,6 +88,12 @@ void kvm_vcpu_pmu_resync_el0(void);
 #define kvm_vcpu_has_pmu(vcpu)					\
 	(vcpu_has_feature(vcpu, KVM_ARM_VCPU_PMU_V3))
 
+bool kvm_pmu_is_partitioned(struct arm_pmu *pmu);
+u64 kvm_pmu_host_counter_mask(struct arm_pmu *pmu);
+u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu);
+void kvm_pmu_host_counters_enable(void);
+void kvm_pmu_host_counters_disable(void);
+
 /*
  * Updates the vcpu's view of the pmu events for this cpu.
  * Must be called before every vcpu run after disabling interrupts, to ensure
@@ -220,6 +226,24 @@ static inline bool kvm_pmu_counter_is_hyp(struct kvm_vcpu *vcpu, unsigned int id
 
 static inline void kvm_pmu_nested_transition(struct kvm_vcpu *vcpu) {}
 
+static inline bool kvm_pmu_is_partitioned(struct arm_pmu *pmu)
+{
+	return false;
+}
+
+static inline u64 kvm_pmu_host_counter_mask(struct arm_pmu *pmu)
+{
+	return ~0;
+}
+
+static inline u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu)
+{
+	return ~0;
+}
+
+static inline void kvm_pmu_host_counters_enable(void) {}
+static inline void kvm_pmu_host_counters_disable(void) {}
+
 #endif
 
 #endif
diff --git a/arch/arm64/kvm/pmu-part.c b/arch/arm64/kvm/pmu-part.c
index f04c580c19c0..4f06a48175e2 100644
--- a/arch/arm64/kvm/pmu-part.c
+++ b/arch/arm64/kvm/pmu-part.c
@@ -5,7 +5,10 @@
  */
 
 #include <linux/kvm_host.h>
+#include <linux/perf/arm_pmu.h>
+#include <linux/perf/arm_pmuv3.h>
 
+#include <asm/arm_pmuv3.h>
 #include <asm/kvm_pmu.h>
 
 /**
@@ -21,3 +24,84 @@ bool kvm_pmu_partition_supported(void)
 	return kvm_supports_guest_pmuv3() &&
 		has_vhe();
 }
+
+/**
+ * kvm_pmu_is_partitioned() - Determine if given PMU is partitioned
+ * @pmu: Pointer to arm_pmu struct
+ *
+ * Determine if given PMU is partitioned by looking at hpmn field. The
+ * PMU is partitioned if this field is less than the number of
+ * counters in the system.
+ *
+ * Return: True if the PMU is partitioned, false otherwise
+ */
+bool kvm_pmu_is_partitioned(struct arm_pmu *pmu)
+{
+	return pmu->hpmn_max >= 0 &&
+		pmu->hpmn_max <= *host_data_ptr(nr_event_counters);
+}
+
+/**
+ * kvm_pmu_host_counter_mask() - Compute bitmask of host-reserved counters
+ * @pmu: Pointer to arm_pmu struct
+ *
+ * Compute the bitmask that selects the host-reserved counters in the
+ * {PMCNTEN,PMINTEN,PMOVS}{SET,CLR} registers. These are the counters
+ * in HPMN..N
+ *
+ * Assumes pmu is partitioned and hpmn_max is a valid value.
+ *
+ * Return: Bitmask
+ */
+u64 kvm_pmu_host_counter_mask(struct arm_pmu *pmu)
+{
+	u8 nr_counters = *host_data_ptr(nr_event_counters);
+
+	return GENMASK(nr_counters - 1, pmu->hpmn_max);
+}
+
+/**
+ * kvm_pmu_guest_counter_mask() - Compute bitmask of guest-reserved counters
+ *
+ * Compute the bitmask that selects the guest-reserved counters in the
+ * {PMCNTEN,PMINTEN,PMOVS}{SET,CLR} registers. These are the counters
+ * in 0..HPMN and the cycle and instruction counters.
+ *
+ * Assumes pmu is partitioned and hpmn_max is a valid value.
+ *
+ * Return: Bitmask
+ */
+u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu)
+{
+	return ARMV8_PMU_CNT_MASK_ALL & ~kvm_pmu_host_counter_mask(pmu);
+}
+
+/**
+ * kvm_pmu_host_counters_enable() - Enable host-reserved counters
+ *
+ * When partitioned the enable bit for host-reserved counters is
+ * MDCR_EL2.HPME instead of the typical PMCR_EL0.E, which now
+ * exclusively controls the guest-reserved counters. Enable that bit.
+ */
+void kvm_pmu_host_counters_enable(void)
+{
+	u64 mdcr = read_sysreg(mdcr_el2);
+
+	mdcr |= MDCR_EL2_HPME;
+	write_sysreg(mdcr, mdcr_el2);
+}
+
+/**
+ * kvm_pmu_host_counters_disable() - Disable host-reserved counters
+ *
+ * When partitioned the disable bit for host-reserved counters is
+ * MDCR_EL2.HPME instead of the typical PMCR_EL0.E, which now
+ * exclusively controls the guest-reserved counters. Disable that bit.
+ */
+void kvm_pmu_host_counters_disable(void)
+{
+	u64 mdcr = read_sysreg(mdcr_el2);
+
+	mdcr &= ~MDCR_EL2_HPME;
+	write_sysreg(mdcr, mdcr_el2);
+}
diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c
index 3bc016afea34..5836672a3dd9 100644
--- a/drivers/perf/arm_pmuv3.c
+++ b/drivers/perf/arm_pmuv3.c
@@ -834,12 +834,18 @@ static void armv8pmu_start(struct arm_pmu *cpu_pmu)
 	kvm_vcpu_pmu_resync_el0();
 
 	/* Enable all counters */
+	if (kvm_pmu_is_partitioned(cpu_pmu))
+		kvm_pmu_host_counters_enable();
+
 	armv8pmu_pmcr_write(armv8pmu_pmcr_read() | ARMV8_PMU_PMCR_E);
 }
 
 static void armv8pmu_stop(struct arm_pmu *cpu_pmu)
 {
 	/* Disable all counters */
+	if (kvm_pmu_is_partitioned(cpu_pmu))
+		kvm_pmu_host_counters_disable();
+
 	armv8pmu_pmcr_write(armv8pmu_pmcr_read() & ~ARMV8_PMU_PMCR_E);
 }
 
@@ -949,6 +955,7 @@ static int armv8pmu_get_event_idx(struct pmu_hw_events *cpuc,
 
 	/* Always prefer to place a cycle counter into the cycle counter. */
 	if ((evtype == ARMV8_PMUV3_PERFCTR_CPU_CYCLES) &&
+	    !kvm_pmu_is_partitioned(cpu_pmu) &&
 	    !armv8pmu_event_get_threshold(&event->attr)) {
 		if (!test_and_set_bit(ARMV8_PMU_CYCLE_IDX, cpuc->used_mask))
 			return ARMV8_PMU_CYCLE_IDX;
@@ -964,6 +971,7 @@ static int armv8pmu_get_event_idx(struct pmu_hw_events *cpuc,
 	 * may not know how to handle it.
 	 */
 	if ((evtype == ARMV8_PMUV3_PERFCTR_INST_RETIRED) &&
+	    !kvm_pmu_is_partitioned(cpu_pmu) &&
 	    !armv8pmu_event_get_threshold(&event->attr) &&
 	    test_bit(ARMV8_PMU_INSTR_IDX, cpu_pmu->cntr_mask) &&
 	    !armv8pmu_event_want_user_access(event)) {
@@ -975,7 +983,7 @@ static int armv8pmu_get_event_idx(struct pmu_hw_events *cpuc,
 	 * Otherwise use events counters
 	 */
 	if (armv8pmu_event_is_chained(event))
-		return	armv8pmu_get_chain_idx(cpuc, cpu_pmu);
+		return armv8pmu_get_chain_idx(cpuc, cpu_pmu);
 	else
 		return armv8pmu_get_single_idx(cpuc, cpu_pmu);
 }
@@ -1067,6 +1075,14 @@ static int armv8pmu_set_event_filter(struct hw_perf_event *event,
 	return 0;
 }
 
+static void armv8pmu_reset_host_counters(struct arm_pmu *cpu_pmu)
+{
+	int idx;
+
+	for_each_set_bit(idx, cpu_pmu->cntr_mask, ARMV8_PMU_MAX_GENERAL_COUNTERS)
+		armv8pmu_write_evcntr(idx, 0);
+}
+
 static void armv8pmu_reset(void *info)
 {
 	struct arm_pmu *cpu_pmu = (struct arm_pmu *)info;
@@ -1074,6 +1090,9 @@ static void armv8pmu_reset(void *info)
 
 	bitmap_to_arr64(&mask, cpu_pmu->cntr_mask, ARMPMU_MAX_HWEVENTS);
 
+	if (kvm_pmu_is_partitioned(cpu_pmu))
+		mask &= kvm_pmu_host_counter_mask(cpu_pmu);
+
 	/* The counter and interrupt enable registers are unknown at reset. */
 	armv8pmu_disable_counter(mask);
 	armv8pmu_disable_intens(mask);
@@ -1081,11 +1100,20 @@ static void armv8pmu_reset(void *info)
 	/* Clear the counters we flip at guest entry/exit */
 	kvm_clr_pmu_events(mask);
 
+
+	pmcr = ARMV8_PMU_PMCR_LC;
+
 	/*
-	 * Initialize & Reset PMNC. Request overflow interrupt for
-	 * 64 bit cycle counter but cheat in armv8pmu_write_counter().
+	 * Initialize & Reset PMNC. Request overflow interrupt for 64
+	 * bit cycle counter but cheat in armv8pmu_write_counter().
+	 *
+	 * When partitioned, there is no single bit to reset only the
+	 * host counters. so reset them individually.
 	 */
-	pmcr = ARMV8_PMU_PMCR_P | ARMV8_PMU_PMCR_C | ARMV8_PMU_PMCR_LC;
+	if (kvm_pmu_is_partitioned(cpu_pmu))
+		armv8pmu_reset_host_counters(cpu_pmu);
+	else
+		pmcr = ARMV8_PMU_PMCR_P | ARMV8_PMU_PMCR_C;
 
 	/* Enable long event counter support where available */
 	if (armv8pmu_has_long_event(cpu_pmu))
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 09/22] KVM: arm64: Correct kvm_arm_pmu_get_max_counters()

[Colton Lewis]

Since cntr_mask is modified when the PMU is partitioned to remove some
bits, make sure the missing counters are added back to get the right
total.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/kvm/pmu.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c
index 79b7ea037153..67216451b8ce 100644
--- a/arch/arm64/kvm/pmu.c
+++ b/arch/arm64/kvm/pmu.c
@@ -533,6 +533,8 @@ static bool pmu_irq_is_valid(struct kvm *kvm, int irq)
 u8 kvm_arm_pmu_get_max_counters(struct kvm *kvm)
 {
 	struct arm_pmu *arm_pmu = kvm->arch.arm_pmu;
+	u8 counters;
+
 
 	/*
 	 * PMUv3 requires that all event counters are capable of counting any
@@ -545,7 +547,12 @@ u8 kvm_arm_pmu_get_max_counters(struct kvm *kvm)
 	 * The arm_pmu->cntr_mask considers the fixed counter(s) as well.
 	 * Ignore those and return only the general-purpose counters.
 	 */
-	return bitmap_weight(arm_pmu->cntr_mask, ARMV8_PMU_MAX_GENERAL_COUNTERS);
+	counters = bitmap_weight(arm_pmu->cntr_mask, ARMV8_PMU_MAX_GENERAL_COUNTERS);
+
+	if (kvm_pmu_is_partitioned(arm_pmu))
+		counters += arm_pmu->hpmn_max;
+
+	return counters;
 }
 
 static void kvm_arm_set_nr_counters(struct kvm *kvm, unsigned int nr)
-- 
2.50.0.727.gbf7dc18ff4-goog

Re: [PATCH v3 09/22] KVM: arm64: Correct kvm_arm_pmu_get_max_counters()

[Marc Zyngier]

On Thu, 26 Jun 2025 21:04:45 +0100,
Colton Lewis <coltonlewis@google.com> wrote:
> 
> Since cntr_mask is modified when the PMU is partitioned to remove some
> bits, make sure the missing counters are added back to get the right
> total.

Please fix the subject of the patch to be more descriptive. It is
worded like a bug fix, while it really is only a step in the patch
series.

Something like "Take partitioning into account for max number of
counters" would go a long way.

> 
> Signed-off-by: Colton Lewis <coltonlewis@google.com>
> ---
>  arch/arm64/kvm/pmu.c | 9 ++++++++-
>  1 file changed, 8 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c
> index 79b7ea037153..67216451b8ce 100644
> --- a/arch/arm64/kvm/pmu.c
> +++ b/arch/arm64/kvm/pmu.c
> @@ -533,6 +533,8 @@ static bool pmu_irq_is_valid(struct kvm *kvm, int irq)
>  u8 kvm_arm_pmu_get_max_counters(struct kvm *kvm)
>  {
>  	struct arm_pmu *arm_pmu = kvm->arch.arm_pmu;
> +	u8 counters;
> +

nit: superfluous blank line.
>  
>  	/*
>  	 * PMUv3 requires that all event counters are capable of counting any
> @@ -545,7 +547,12 @@ u8 kvm_arm_pmu_get_max_counters(struct kvm *kvm)
>  	 * The arm_pmu->cntr_mask considers the fixed counter(s) as well.
>  	 * Ignore those and return only the general-purpose counters.
>  	 */
> -	return bitmap_weight(arm_pmu->cntr_mask, ARMV8_PMU_MAX_GENERAL_COUNTERS);
> +	counters = bitmap_weight(arm_pmu->cntr_mask, ARMV8_PMU_MAX_GENERAL_COUNTERS);
> +
> +	if (kvm_pmu_is_partitioned(arm_pmu))
> +		counters += arm_pmu->hpmn_max;

Why the check? Why can't we rely on hpmn_max to always give us the
correct value?

Thanks,

	M.

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

Re: [PATCH v3 09/22] KVM: arm64: Correct kvm_arm_pmu_get_max_counters()

[Colton Lewis]

Marc Zyngier <maz@kernel.org> writes:

> On Thu, 26 Jun 2025 21:04:45 +0100,
> Colton Lewis <coltonlewis@google.com> wrote:

>> Since cntr_mask is modified when the PMU is partitioned to remove some
>> bits, make sure the missing counters are added back to get the right
>> total.

> Please fix the subject of the patch to be more descriptive. It is
> worded like a bug fix, while it really is only a step in the patch
> series.

> Something like "Take partitioning into account for max number of
> counters" would go a long way.

Done.


>> Signed-off-by: Colton Lewis <coltonlewis@google.com>
>> ---
>>   arch/arm64/kvm/pmu.c | 9 ++++++++-
>>   1 file changed, 8 insertions(+), 1 deletion(-)

>> diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c
>> index 79b7ea037153..67216451b8ce 100644
>> --- a/arch/arm64/kvm/pmu.c
>> +++ b/arch/arm64/kvm/pmu.c
>> @@ -533,6 +533,8 @@ static bool pmu_irq_is_valid(struct kvm *kvm, int  
>> irq)
>>   u8 kvm_arm_pmu_get_max_counters(struct kvm *kvm)
>>   {
>>   	struct arm_pmu *arm_pmu = kvm->arch.arm_pmu;
>> +	u8 counters;
>> +

> nit: superfluous blank line.

I'll remove.


>>   	/*
>>   	 * PMUv3 requires that all event counters are capable of counting any
>> @@ -545,7 +547,12 @@ u8 kvm_arm_pmu_get_max_counters(struct kvm *kvm)
>>   	 * The arm_pmu->cntr_mask considers the fixed counter(s) as well.
>>   	 * Ignore those and return only the general-purpose counters.
>>   	 */
>> -	return bitmap_weight(arm_pmu->cntr_mask,  
>> ARMV8_PMU_MAX_GENERAL_COUNTERS);
>> +	counters = bitmap_weight(arm_pmu->cntr_mask,  
>> ARMV8_PMU_MAX_GENERAL_COUNTERS);
>> +
>> +	if (kvm_pmu_is_partitioned(arm_pmu))
>> +		counters += arm_pmu->hpmn_max;

> Why the check? Why can't we rely on hpmn_max to always give us the
> correct value?

Imagine no partitioning. Then the value of hpmn_max that would be give a
correct result here is 0 but 0 is also a valid value (anything from 0 to
the number of host counters inclusive) indicating we are partitioned and
want to give only the cycle counter to the guest.

My predicate to determine if the PMU is partitoned relies on hpmn_max
being a valid value only when the PMU is partitioned.

Making hpmn_max 0 here would require introducing another flag to check
partitioning.

> Thanks,

> 	M.

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

[PATCH v3 10/22] KVM: arm64: Set up FGT for Partitioned PMU

[Colton Lewis]

In order to gain the best performance benefit from partitioning the
PMU, utilize fine grain traps (FEAT_FGT and FEAT_FGT2) to avoid
trapping common PMU register accesses by the guest to remove that
overhead.

There should be no information leaks between guests as all these
registers are context swapped by a later patch in this series.

Untrapped:
* PMCR_EL0
* PMUSERENR_EL0
* PMSELR_EL0
* PMCCNTR_EL0
* PMINTEN_EL0
* PMEVCNTRn_EL0

Trapped:
* PMOVS_EL0
* PMEVTYPERn_EL0
* PMCCFILTR_EL0
* PMICNTR_EL0
* PMICFILTR_EL0

PMOVS remains trapped so KVM can track overflow IRQs that will need to
be injected into the guest.

PMICNTR remains trapped because KVM is not handling that yet.

PMEVTYPERn remains trapped so KVM can limit which events guests can
count, such as disallowing counting at EL2. PMCCFILTR and PMCIFILTR
are the same.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/include/asm/kvm_pmu.h        | 23 ++++++++++
 arch/arm64/kvm/hyp/include/hyp/switch.h | 58 +++++++++++++++++++++++++
 arch/arm64/kvm/pmu-part.c               | 32 ++++++++++++++
 3 files changed, 113 insertions(+)

diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
index 6328e90952ba..73b7161e3f4e 100644
--- a/arch/arm64/include/asm/kvm_pmu.h
+++ b/arch/arm64/include/asm/kvm_pmu.h
@@ -94,6 +94,21 @@ u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu);
 void kvm_pmu_host_counters_enable(void);
 void kvm_pmu_host_counters_disable(void);
 
+#if !defined(__KVM_NVHE_HYPERVISOR__)
+bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu);
+bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu);
+#else
+static inline bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu)
+{
+	return false;
+}
+
+static inline bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu)
+{
+	return false;
+}
+#endif
+
 /*
  * Updates the vcpu's view of the pmu events for this cpu.
  * Must be called before every vcpu run after disabling interrupts, to ensure
@@ -133,6 +148,14 @@ static inline u64 kvm_pmu_get_counter_value(struct kvm_vcpu *vcpu,
 {
 	return 0;
 }
+static inline bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu)
+{
+	return false;
+}
+static inline bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu)
+{
+	return false;
+}
 static inline void kvm_pmu_set_counter_value(struct kvm_vcpu *vcpu,
 					     u64 select_idx, u64 val) {}
 static inline void kvm_pmu_set_counter_value_user(struct kvm_vcpu *vcpu,
diff --git a/arch/arm64/kvm/hyp/include/hyp/switch.h b/arch/arm64/kvm/hyp/include/hyp/switch.h
index 825b81749972..47d2db8446df 100644
--- a/arch/arm64/kvm/hyp/include/hyp/switch.h
+++ b/arch/arm64/kvm/hyp/include/hyp/switch.h
@@ -191,6 +191,61 @@ static inline bool cpu_has_amu(void)
                ID_AA64PFR0_EL1_AMU_SHIFT);
 }
 
+/**
+ * __activate_pmu_fgt() - Activate fine grain traps for partitioned PMU
+ * @vcpu: Pointer to struct kvm_vcpu
+ *
+ * Clear the most commonly accessed registers for a partitioned
+ * PMU. Trap the rest.
+ */
+static inline void __activate_pmu_fgt(struct kvm_vcpu *vcpu)
+{
+	struct kvm_cpu_context *hctxt = host_data_ptr(host_ctxt);
+	struct kvm *kvm = kern_hyp_va(vcpu->kvm);
+	u64 set;
+	u64 clr;
+
+	set = HDFGRTR_EL2_PMOVS
+		| HDFGRTR_EL2_PMCCFILTR_EL0
+		| HDFGRTR_EL2_PMEVTYPERn_EL0;
+	clr = HDFGRTR_EL2_PMUSERENR_EL0
+		| HDFGRTR_EL2_PMSELR_EL0
+		| HDFGRTR_EL2_PMINTEN
+		| HDFGRTR_EL2_PMCNTEN
+		| HDFGRTR_EL2_PMCCNTR_EL0
+		| HDFGRTR_EL2_PMEVCNTRn_EL0;
+
+	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGRTR_EL2, clr, set);
+
+	set = HDFGWTR_EL2_PMOVS
+		| HDFGWTR_EL2_PMCCFILTR_EL0
+		| HDFGWTR_EL2_PMEVTYPERn_EL0;
+	clr = HDFGWTR_EL2_PMUSERENR_EL0
+		| HDFGWTR_EL2_PMCR_EL0
+		| HDFGWTR_EL2_PMSELR_EL0
+		| HDFGWTR_EL2_PMINTEN
+		| HDFGWTR_EL2_PMCNTEN
+		| HDFGWTR_EL2_PMCCNTR_EL0
+		| HDFGWTR_EL2_PMEVCNTRn_EL0;
+
+	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGWTR_EL2, clr, set);
+
+	if (!cpus_have_final_cap(ARM64_HAS_FGT2))
+		return;
+
+	set = HDFGRTR2_EL2_nPMICFILTR_EL0
+		| HDFGRTR2_EL2_nPMICNTR_EL0;
+	clr = 0;
+
+	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGRTR2_EL2, clr, set);
+
+	set = HDFGWTR2_EL2_nPMICFILTR_EL0
+		| HDFGWTR2_EL2_nPMICNTR_EL0;
+	clr = 0;
+
+	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGWTR2_EL2, clr, set);
+}
+
 static inline void __activate_traps_hfgxtr(struct kvm_vcpu *vcpu)
 {
 	struct kvm_cpu_context *hctxt = host_data_ptr(host_ctxt);
@@ -210,6 +265,9 @@ static inline void __activate_traps_hfgxtr(struct kvm_vcpu *vcpu)
 	if (cpu_has_amu())
 		update_fgt_traps(hctxt, vcpu, kvm, HAFGRTR_EL2);
 
+	if (kvm_vcpu_pmu_use_fgt(vcpu))
+		__activate_pmu_fgt(vcpu);
+
 	if (!cpus_have_final_cap(ARM64_HAS_FGT2))
 	    return;
 
diff --git a/arch/arm64/kvm/pmu-part.c b/arch/arm64/kvm/pmu-part.c
index 4f06a48175e2..92775e19cbf6 100644
--- a/arch/arm64/kvm/pmu-part.c
+++ b/arch/arm64/kvm/pmu-part.c
@@ -41,6 +41,38 @@ bool kvm_pmu_is_partitioned(struct arm_pmu *pmu)
 		pmu->hpmn_max <= *host_data_ptr(nr_event_counters);
 }
 
+/**
+ * kvm_vcpu_pmu_is_partitioned() - Determine if given VCPU has a partitioned PMU
+ * @vcpu: Pointer to kvm_vcpu struct
+ *
+ * Determine if given VCPU has a partitioned PMU by extracting that
+ * field and passing it to :c:func:`kvm_pmu_is_partitioned`
+ *
+ * Return: True if the VCPU PMU is partitioned, false otherwise
+ */
+bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu)
+{
+	return kvm_pmu_is_partitioned(vcpu->kvm->arch.arm_pmu);
+}
+
+/**
+ * kvm_vcpu_pmu_use_fgt() - Determine if we can use FGT
+ * @vcpu: Pointer to struct kvm_vcpu
+ *
+ * Determine if we can use FGT for direct access to registers. We can
+ * if capabilities permit the number of guest counters requested.
+ *
+ * Return: True if we can use FGT, false otherwise
+ */
+bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu)
+{
+	u8 hpmn = vcpu->kvm->arch.nr_pmu_counters;
+
+	return kvm_vcpu_pmu_is_partitioned(vcpu) &&
+		cpus_have_final_cap(ARM64_HAS_FGT) &&
+		(hpmn != 0 || cpus_have_final_cap(ARM64_HAS_HPMN0));
+}
+
 /**
  * kvm_pmu_host_counter_mask() - Compute bitmask of host-reserved counters
  * @pmu: Pointer to arm_pmu struct
-- 
2.50.0.727.gbf7dc18ff4-goog

Re: [PATCH v3 10/22] KVM: arm64: Set up FGT for Partitioned PMU

[Marc Zyngier]

On Thu, 26 Jun 2025 21:04:46 +0100,
Colton Lewis <coltonlewis@google.com> wrote:
> 
> In order to gain the best performance benefit from partitioning the
> PMU, utilize fine grain traps (FEAT_FGT and FEAT_FGT2) to avoid
> trapping common PMU register accesses by the guest to remove that
> overhead.
> 
> There should be no information leaks between guests as all these
> registers are context swapped by a later patch in this series.
> 
> Untrapped:
> * PMCR_EL0
> * PMUSERENR_EL0
> * PMSELR_EL0
> * PMCCNTR_EL0
> * PMINTEN_EL0
> * PMEVCNTRn_EL0
> 
> Trapped:
> * PMOVS_EL0
> * PMEVTYPERn_EL0
> * PMCCFILTR_EL0
> * PMICNTR_EL0
> * PMICFILTR_EL0
> 
> PMOVS remains trapped so KVM can track overflow IRQs that will need to
> be injected into the guest.
> 
> PMICNTR remains trapped because KVM is not handling that yet.
> 
> PMEVTYPERn remains trapped so KVM can limit which events guests can
> count, such as disallowing counting at EL2. PMCCFILTR and PMCIFILTR
> are the same.

I'd rather you explain why it is safe not to trap the rest.

> 
> Signed-off-by: Colton Lewis <coltonlewis@google.com>
> ---
>  arch/arm64/include/asm/kvm_pmu.h        | 23 ++++++++++
>  arch/arm64/kvm/hyp/include/hyp/switch.h | 58 +++++++++++++++++++++++++
>  arch/arm64/kvm/pmu-part.c               | 32 ++++++++++++++
>  3 files changed, 113 insertions(+)
> 
> diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
> index 6328e90952ba..73b7161e3f4e 100644
> --- a/arch/arm64/include/asm/kvm_pmu.h
> +++ b/arch/arm64/include/asm/kvm_pmu.h
> @@ -94,6 +94,21 @@ u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu);
>  void kvm_pmu_host_counters_enable(void);
>  void kvm_pmu_host_counters_disable(void);
>  
> +#if !defined(__KVM_NVHE_HYPERVISOR__)
> +bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu);
> +bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu);
> +#else
> +static inline bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu)
> +{
> +	return false;
> +}
> +
> +static inline bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu)
> +{
> +	return false;
> +}
> +#endif
> +
>  /*
>   * Updates the vcpu's view of the pmu events for this cpu.
>   * Must be called before every vcpu run after disabling interrupts, to ensure
> @@ -133,6 +148,14 @@ static inline u64 kvm_pmu_get_counter_value(struct kvm_vcpu *vcpu,
>  {
>  	return 0;
>  }
> +static inline bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu)
> +{
> +	return false;
> +}
> +static inline bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu)
> +{
> +	return false;
> +}
>  static inline void kvm_pmu_set_counter_value(struct kvm_vcpu *vcpu,
>  					     u64 select_idx, u64 val) {}
>  static inline void kvm_pmu_set_counter_value_user(struct kvm_vcpu *vcpu,
> diff --git a/arch/arm64/kvm/hyp/include/hyp/switch.h b/arch/arm64/kvm/hyp/include/hyp/switch.h
> index 825b81749972..47d2db8446df 100644
> --- a/arch/arm64/kvm/hyp/include/hyp/switch.h
> +++ b/arch/arm64/kvm/hyp/include/hyp/switch.h
> @@ -191,6 +191,61 @@ static inline bool cpu_has_amu(void)
>                 ID_AA64PFR0_EL1_AMU_SHIFT);
>  }
>  
> +/**
> + * __activate_pmu_fgt() - Activate fine grain traps for partitioned PMU
> + * @vcpu: Pointer to struct kvm_vcpu
> + *
> + * Clear the most commonly accessed registers for a partitioned
> + * PMU. Trap the rest.
> + */
> +static inline void __activate_pmu_fgt(struct kvm_vcpu *vcpu)
> +{
> +	struct kvm_cpu_context *hctxt = host_data_ptr(host_ctxt);
> +	struct kvm *kvm = kern_hyp_va(vcpu->kvm);
> +	u64 set;
> +	u64 clr;
> +
> +	set = HDFGRTR_EL2_PMOVS
> +		| HDFGRTR_EL2_PMCCFILTR_EL0
> +		| HDFGRTR_EL2_PMEVTYPERn_EL0;
> +	clr = HDFGRTR_EL2_PMUSERENR_EL0
> +		| HDFGRTR_EL2_PMSELR_EL0
> +		| HDFGRTR_EL2_PMINTEN
> +		| HDFGRTR_EL2_PMCNTEN
> +		| HDFGRTR_EL2_PMCCNTR_EL0
> +		| HDFGRTR_EL2_PMEVCNTRn_EL0;
> +
> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGRTR_EL2, clr, set);
> +
> +	set = HDFGWTR_EL2_PMOVS
> +		| HDFGWTR_EL2_PMCCFILTR_EL0
> +		| HDFGWTR_EL2_PMEVTYPERn_EL0;
> +	clr = HDFGWTR_EL2_PMUSERENR_EL0
> +		| HDFGWTR_EL2_PMCR_EL0
> +		| HDFGWTR_EL2_PMSELR_EL0
> +		| HDFGWTR_EL2_PMINTEN
> +		| HDFGWTR_EL2_PMCNTEN
> +		| HDFGWTR_EL2_PMCCNTR_EL0
> +		| HDFGWTR_EL2_PMEVCNTRn_EL0;
> +
> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGWTR_EL2, clr, set);
> +
> +	if (!cpus_have_final_cap(ARM64_HAS_FGT2))
> +		return;
> +
> +	set = HDFGRTR2_EL2_nPMICFILTR_EL0
> +		| HDFGRTR2_EL2_nPMICNTR_EL0;
> +	clr = 0;
> +
> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGRTR2_EL2, clr, set);
> +
> +	set = HDFGWTR2_EL2_nPMICFILTR_EL0
> +		| HDFGWTR2_EL2_nPMICNTR_EL0;
> +	clr = 0;
> +
> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGWTR2_EL2, clr, set);

This feels wrong. There should be one place to populate the FGTs that
apply to a guest as set from the host, not two or more.

There is such a construct in the SME series, and maybe you could have
a look at it, specially if the trap configuration is this static.

	M.

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

Re: [PATCH v3 10/22] KVM: arm64: Set up FGT for Partitioned PMU

[Colton Lewis]

Marc Zyngier <maz@kernel.org> writes:

> On Thu, 26 Jun 2025 21:04:46 +0100,
> Colton Lewis <coltonlewis@google.com> wrote:

>> In order to gain the best performance benefit from partitioning the
>> PMU, utilize fine grain traps (FEAT_FGT and FEAT_FGT2) to avoid
>> trapping common PMU register accesses by the guest to remove that
>> overhead.

>> There should be no information leaks between guests as all these
>> registers are context swapped by a later patch in this series.

>> Untrapped:
>> * PMCR_EL0
>> * PMUSERENR_EL0
>> * PMSELR_EL0
>> * PMCCNTR_EL0
>> * PMINTEN_EL0
>> * PMEVCNTRn_EL0

>> Trapped:
>> * PMOVS_EL0
>> * PMEVTYPERn_EL0
>> * PMCCFILTR_EL0
>> * PMICNTR_EL0
>> * PMICFILTR_EL0

>> PMOVS remains trapped so KVM can track overflow IRQs that will need to
>> be injected into the guest.

>> PMICNTR remains trapped because KVM is not handling that yet.

>> PMEVTYPERn remains trapped so KVM can limit which events guests can
>> count, such as disallowing counting at EL2. PMCCFILTR and PMCIFILTR
>> are the same.

> I'd rather you explain why it is safe not to trap the rest.

Okay, I will reverse my explanation.


>> Signed-off-by: Colton Lewis <coltonlewis@google.com>
>> ---
>>   arch/arm64/include/asm/kvm_pmu.h        | 23 ++++++++++
>>   arch/arm64/kvm/hyp/include/hyp/switch.h | 58 +++++++++++++++++++++++++
>>   arch/arm64/kvm/pmu-part.c               | 32 ++++++++++++++
>>   3 files changed, 113 insertions(+)

>> diff --git a/arch/arm64/include/asm/kvm_pmu.h  
>> b/arch/arm64/include/asm/kvm_pmu.h
>> index 6328e90952ba..73b7161e3f4e 100644
>> --- a/arch/arm64/include/asm/kvm_pmu.h
>> +++ b/arch/arm64/include/asm/kvm_pmu.h
>> @@ -94,6 +94,21 @@ u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu);
>>   void kvm_pmu_host_counters_enable(void);
>>   void kvm_pmu_host_counters_disable(void);

>> +#if !defined(__KVM_NVHE_HYPERVISOR__)
>> +bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu);
>> +bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu);
>> +#else
>> +static inline bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu)
>> +{
>> +	return false;
>> +}
>> +
>> +static inline bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu)
>> +{
>> +	return false;
>> +}
>> +#endif
>> +
>>   /*
>>    * Updates the vcpu's view of the pmu events for this cpu.
>>    * Must be called before every vcpu run after disabling interrupts, to  
>> ensure
>> @@ -133,6 +148,14 @@ static inline u64 kvm_pmu_get_counter_value(struct  
>> kvm_vcpu *vcpu,
>>   {
>>   	return 0;
>>   }
>> +static inline bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu)
>> +{
>> +	return false;
>> +}
>> +static inline bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu)
>> +{
>> +	return false;
>> +}
>>   static inline void kvm_pmu_set_counter_value(struct kvm_vcpu *vcpu,
>>   					     u64 select_idx, u64 val) {}
>>   static inline void kvm_pmu_set_counter_value_user(struct kvm_vcpu *vcpu,
>> diff --git a/arch/arm64/kvm/hyp/include/hyp/switch.h  
>> b/arch/arm64/kvm/hyp/include/hyp/switch.h
>> index 825b81749972..47d2db8446df 100644
>> --- a/arch/arm64/kvm/hyp/include/hyp/switch.h
>> +++ b/arch/arm64/kvm/hyp/include/hyp/switch.h
>> @@ -191,6 +191,61 @@ static inline bool cpu_has_amu(void)
>>                  ID_AA64PFR0_EL1_AMU_SHIFT);
>>   }

>> +/**
>> + * __activate_pmu_fgt() - Activate fine grain traps for partitioned PMU
>> + * @vcpu: Pointer to struct kvm_vcpu
>> + *
>> + * Clear the most commonly accessed registers for a partitioned
>> + * PMU. Trap the rest.
>> + */
>> +static inline void __activate_pmu_fgt(struct kvm_vcpu *vcpu)
>> +{
>> +	struct kvm_cpu_context *hctxt = host_data_ptr(host_ctxt);
>> +	struct kvm *kvm = kern_hyp_va(vcpu->kvm);
>> +	u64 set;
>> +	u64 clr;
>> +
>> +	set = HDFGRTR_EL2_PMOVS
>> +		| HDFGRTR_EL2_PMCCFILTR_EL0
>> +		| HDFGRTR_EL2_PMEVTYPERn_EL0;
>> +	clr = HDFGRTR_EL2_PMUSERENR_EL0
>> +		| HDFGRTR_EL2_PMSELR_EL0
>> +		| HDFGRTR_EL2_PMINTEN
>> +		| HDFGRTR_EL2_PMCNTEN
>> +		| HDFGRTR_EL2_PMCCNTR_EL0
>> +		| HDFGRTR_EL2_PMEVCNTRn_EL0;
>> +
>> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGRTR_EL2, clr, set);
>> +
>> +	set = HDFGWTR_EL2_PMOVS
>> +		| HDFGWTR_EL2_PMCCFILTR_EL0
>> +		| HDFGWTR_EL2_PMEVTYPERn_EL0;
>> +	clr = HDFGWTR_EL2_PMUSERENR_EL0
>> +		| HDFGWTR_EL2_PMCR_EL0
>> +		| HDFGWTR_EL2_PMSELR_EL0
>> +		| HDFGWTR_EL2_PMINTEN
>> +		| HDFGWTR_EL2_PMCNTEN
>> +		| HDFGWTR_EL2_PMCCNTR_EL0
>> +		| HDFGWTR_EL2_PMEVCNTRn_EL0;
>> +
>> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGWTR_EL2, clr, set);
>> +
>> +	if (!cpus_have_final_cap(ARM64_HAS_FGT2))
>> +		return;
>> +
>> +	set = HDFGRTR2_EL2_nPMICFILTR_EL0
>> +		| HDFGRTR2_EL2_nPMICNTR_EL0;
>> +	clr = 0;
>> +
>> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGRTR2_EL2, clr, set);
>> +
>> +	set = HDFGWTR2_EL2_nPMICFILTR_EL0
>> +		| HDFGWTR2_EL2_nPMICNTR_EL0;
>> +	clr = 0;
>> +
>> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGWTR2_EL2, clr, set);

> This feels wrong. There should be one place to populate the FGTs that
> apply to a guest as set from the host, not two or more.

> There is such a construct in the SME series, and maybe you could have
> a look at it, specially if the trap configuration is this static.

> 	M.

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

I'm assuming you are referring to Mark Brown's series [1], specifically
patches 5 and 18 and I see what you mean.

You are probably thinking configuration should happen from
sys_regs.c:kvm_calculate_traps or thereabout and should be setting bits
in the existing kvm->arch.fgt array.

Correct me if I'm mistaken.

[1]  
https://lore.kernel.org/kvm/20250625-kvm-arm64-sme-v6-0-114cff4ffe04@kernel.org/

Re: [PATCH v3 10/22] KVM: arm64: Set up FGT for Partitioned PMU

[Marc Zyngier]

On Fri, 27 Jun 2025 21:45:57 +0100,
Colton Lewis <coltonlewis@google.com> wrote:
> 
> Marc Zyngier <maz@kernel.org> writes:
> 
> > On Thu, 26 Jun 2025 21:04:46 +0100,
> > Colton Lewis <coltonlewis@google.com> wrote:
> 
> >> +static inline void __activate_pmu_fgt(struct kvm_vcpu *vcpu)
> >> +{
> >> +	struct kvm_cpu_context *hctxt = host_data_ptr(host_ctxt);
> >> +	struct kvm *kvm = kern_hyp_va(vcpu->kvm);
> >> +	u64 set;
> >> +	u64 clr;
> >> +
> >> +	set = HDFGRTR_EL2_PMOVS
> >> +		| HDFGRTR_EL2_PMCCFILTR_EL0
> >> +		| HDFGRTR_EL2_PMEVTYPERn_EL0;
> >> +	clr = HDFGRTR_EL2_PMUSERENR_EL0
> >> +		| HDFGRTR_EL2_PMSELR_EL0
> >> +		| HDFGRTR_EL2_PMINTEN
> >> +		| HDFGRTR_EL2_PMCNTEN
> >> +		| HDFGRTR_EL2_PMCCNTR_EL0
> >> +		| HDFGRTR_EL2_PMEVCNTRn_EL0;
> >> +
> >> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGRTR_EL2, clr, set);
> >> +
> >> +	set = HDFGWTR_EL2_PMOVS
> >> +		| HDFGWTR_EL2_PMCCFILTR_EL0
> >> +		| HDFGWTR_EL2_PMEVTYPERn_EL0;
> >> +	clr = HDFGWTR_EL2_PMUSERENR_EL0
> >> +		| HDFGWTR_EL2_PMCR_EL0
> >> +		| HDFGWTR_EL2_PMSELR_EL0
> >> +		| HDFGWTR_EL2_PMINTEN
> >> +		| HDFGWTR_EL2_PMCNTEN
> >> +		| HDFGWTR_EL2_PMCCNTR_EL0
> >> +		| HDFGWTR_EL2_PMEVCNTRn_EL0;
> >> +
> >> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGWTR_EL2, clr, set);
> >> +
> >> +	if (!cpus_have_final_cap(ARM64_HAS_FGT2))
> >> +		return;
> >> +
> >> +	set = HDFGRTR2_EL2_nPMICFILTR_EL0
> >> +		| HDFGRTR2_EL2_nPMICNTR_EL0;
> >> +	clr = 0;
> >> +
> >> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGRTR2_EL2, clr, set);
> >> +
> >> +	set = HDFGWTR2_EL2_nPMICFILTR_EL0
> >> +		| HDFGWTR2_EL2_nPMICNTR_EL0;
> >> +	clr = 0;
> >> +
> >> +	update_fgt_traps_cs(hctxt, vcpu, kvm, HDFGWTR2_EL2, clr, set);
> 
> > This feels wrong. There should be one place to populate the FGTs that
> > apply to a guest as set from the host, not two or more.
> 
> > There is such a construct in the SME series, and maybe you could have
> > a look at it, specially if the trap configuration is this static.
> 
> > 	M.
> 
> > --
> > Without deviation from the norm, progress is not possible.
> 
> I'm assuming you are referring to Mark Brown's series [1], specifically
> patches 5 and 18 and I see what you mean.
> 
> You are probably thinking configuration should happen from
> sys_regs.c:kvm_calculate_traps or thereabout and should be setting bits
> in the existing kvm->arch.fgt array.
> 
> Correct me if I'm mistaken.

I'm saying there should be exactly one place where we write to the
individual trap registers, and that the source of these settings
should be equally unique when they are immutable in the lifetime of
the guest.

That's the existing pattern for most trap configuration, including
HCR_EL2, ICH_HCR_EL2, HCRX_EL2, and the FGU configuration that
trickles into the actual trap registers, and I want to stick with it
if at all possible.

The way it is done in the SME series may be reasonable, but I haven't
reviewed this series at all. I'm merely pointing out that similar
constructs exist for other features.

	M.

-- 
Jazz isn't dead. It just smells funny.

[PATCH v3 11/22] KVM: arm64: Writethrough trapped PMEVTYPER register

[Colton Lewis]

With FGT in place, the remaining trapped registers need to be written
through to the underlying physical registers as well as the virtual
ones. Failing to do this means delaying when guest writes take effect.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/kvm/sys_regs.c | 29 ++++++++++++++++++++++++++++-
 1 file changed, 28 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
index eaff6d63ef77..49e8e3dcd306 100644
--- a/arch/arm64/kvm/sys_regs.c
+++ b/arch/arm64/kvm/sys_regs.c
@@ -18,6 +18,7 @@
 #include <linux/printk.h>
 #include <linux/uaccess.h>
 #include <linux/irqchip/arm-gic-v3.h>
+#include <linux/perf/arm_pmu.h>
 #include <linux/perf/arm_pmuv3.h>
 
 #include <asm/arm_pmuv3.h>
@@ -1037,6 +1038,30 @@ static bool access_pmu_evcntr(struct kvm_vcpu *vcpu,
 	return true;
 }
 
+static bool writethrough_pmevtyper(struct kvm_vcpu *vcpu, struct sys_reg_params *p,
+				   u64 reg, u64 idx)
+{
+	u64 eventsel;
+
+	if (idx == ARMV8_PMU_CYCLE_IDX)
+		eventsel = ARMV8_PMUV3_PERFCTR_CPU_CYCLES;
+	else
+		eventsel = p->regval & kvm_pmu_evtyper_mask(vcpu->kvm);
+
+	if (vcpu->kvm->arch.pmu_filter &&
+	    !test_bit(eventsel, vcpu->kvm->arch.pmu_filter))
+		return false;
+
+	__vcpu_assign_sys_reg(vcpu, reg, eventsel);
+
+	if (idx == ARMV8_PMU_CYCLE_IDX)
+		write_pmccfiltr(eventsel);
+	else
+		write_pmevtypern(idx, eventsel);
+
+	return true;
+}
+
 static bool access_pmu_evtyper(struct kvm_vcpu *vcpu, struct sys_reg_params *p,
 			       const struct sys_reg_desc *r)
 {
@@ -1063,7 +1088,9 @@ static bool access_pmu_evtyper(struct kvm_vcpu *vcpu, struct sys_reg_params *p,
 	if (!pmu_counter_idx_valid(vcpu, idx))
 		return false;
 
-	if (p->is_write) {
+	if (kvm_vcpu_pmu_is_partitioned(vcpu) && p->is_write) {
+		writethrough_pmevtyper(vcpu, p, reg, idx);
+	} else if (p->is_write) {
 		kvm_pmu_set_counter_event_type(vcpu, p->regval, idx);
 		kvm_vcpu_pmu_restore_guest(vcpu);
 	} else {
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 12/22] KVM: arm64: Use physical PMSELR for PMXEVTYPER if partitioned

[Colton Lewis]

Because PMXEVTYPER is trapped and PMSELR is not, it is not appropriate
to use the virtual PMSELR register when it could be outdated and lead
to an invalid write. Use the physical register.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/include/asm/arm_pmuv3.h | 7 ++++++-
 arch/arm64/kvm/sys_regs.c          | 9 +++++++--
 2 files changed, 13 insertions(+), 3 deletions(-)

diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h
index e2057365ba73..1880e426a559 100644
--- a/arch/arm64/include/asm/arm_pmuv3.h
+++ b/arch/arm64/include/asm/arm_pmuv3.h
@@ -72,11 +72,16 @@ static inline u64 read_pmcr(void)
 	return read_sysreg(pmcr_el0);
 }
 
-static inline void write_pmselr(u32 val)
+static inline void write_pmselr(u64 val)
 {
 	write_sysreg(val, pmselr_el0);
 }
 
+static inline u64 read_pmselr(void)
+{
+	return read_sysreg(pmselr_el0);
+}
+
 static inline void write_pmccntr(u64 val)
 {
 	write_sysreg(val, pmccntr_el0);
diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
index 49e8e3dcd306..771d73451b9a 100644
--- a/arch/arm64/kvm/sys_regs.c
+++ b/arch/arm64/kvm/sys_regs.c
@@ -1065,14 +1065,19 @@ static bool writethrough_pmevtyper(struct kvm_vcpu *vcpu, struct sys_reg_params
 static bool access_pmu_evtyper(struct kvm_vcpu *vcpu, struct sys_reg_params *p,
 			       const struct sys_reg_desc *r)
 {
-	u64 idx, reg;
+	u64 idx, reg, pmselr;
 
 	if (pmu_access_el0_disabled(vcpu))
 		return false;
 
 	if (r->CRn == 9 && r->CRm == 13 && r->Op2 == 1) {
 		/* PMXEVTYPER_EL0 */
-		idx = SYS_FIELD_GET(PMSELR_EL0, SEL, __vcpu_sys_reg(vcpu, PMSELR_EL0));
+		if (kvm_vcpu_pmu_is_partitioned(vcpu))
+			pmselr = read_pmselr();
+		else
+			pmselr = __vcpu_sys_reg(vcpu, PMSELR_EL0);
+
+		idx = SYS_FIELD_GET(PMSELR_EL0, SEL, pmselr);
 		reg = PMEVTYPER0_EL0 + idx;
 	} else if (r->CRn == 14 && (r->CRm & 12) == 12) {
 		idx = ((r->CRm & 3) << 3) | (r->Op2 & 7);
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 13/22] KVM: arm64: Writethrough trapped PMOVS register

[Colton Lewis]

With FGT in place, the remaining trapped registers need to be written
through to the underlying physical registers as well as the virtual
ones. Failing to do this means delaying when guest writes take effect.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/include/asm/arm_pmuv3.h | 10 ++++++++++
 arch/arm64/kvm/sys_regs.c          | 17 ++++++++++++++++-
 2 files changed, 26 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h
index 1880e426a559..3bddde5f4ebb 100644
--- a/arch/arm64/include/asm/arm_pmuv3.h
+++ b/arch/arm64/include/asm/arm_pmuv3.h
@@ -142,6 +142,16 @@ static inline u64 read_pmicfiltr(void)
 	return read_sysreg_s(SYS_PMICFILTR_EL0);
 }
 
+static inline void write_pmovsset(u64 val)
+{
+	write_sysreg(val, pmovsset_el0);
+}
+
+static inline u64 read_pmovsset(void)
+{
+	return read_sysreg(pmovsset_el0);
+}
+
 static inline void write_pmovsclr(u64 val)
 {
 	write_sysreg(val, pmovsclr_el0);
diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
index 771d73451b9a..cfbce4537b4c 100644
--- a/arch/arm64/kvm/sys_regs.c
+++ b/arch/arm64/kvm/sys_regs.c
@@ -1173,6 +1173,19 @@ static bool access_pminten(struct kvm_vcpu *vcpu, struct sys_reg_params *p,
 	return true;
 }
 
+static void writethrough_pmovs(struct kvm_vcpu *vcpu, struct sys_reg_params *p, bool set)
+{
+	u64 mask = kvm_pmu_accessible_counter_mask(vcpu);
+
+	if (set) {
+		__vcpu_rmw_sys_reg(vcpu, PMOVSSET_EL0, |=, (p->regval & mask));
+		write_pmovsset(p->regval & mask);
+	} else {
+		__vcpu_rmw_sys_reg(vcpu, PMOVSSET_EL0, &=, ~(p->regval & mask));
+		write_pmovsclr(p->regval & mask);
+	}
+}
+
 static bool access_pmovs(struct kvm_vcpu *vcpu, struct sys_reg_params *p,
 			 const struct sys_reg_desc *r)
 {
@@ -1181,7 +1194,9 @@ static bool access_pmovs(struct kvm_vcpu *vcpu, struct sys_reg_params *p,
 	if (pmu_access_el0_disabled(vcpu))
 		return false;
 
-	if (p->is_write) {
+	if (kvm_vcpu_pmu_is_partitioned(vcpu) && p->is_write) {
+		writethrough_pmovs(vcpu, p, r->CRm & 0x2);
+	} else if (p->is_write) {
 		if (r->CRm & 0x2)
 			/* accessing PMOVSSET_EL0 */
 			__vcpu_rmw_sys_reg(vcpu, PMOVSSET_EL0, |=, (p->regval & mask));
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 14/22] KVM: arm64: Write fast path PMU register handlers

[Colton Lewis]

We may want a partitioned PMU but not have FEAT_FGT to untrap the
specific registers that would normally be. Add a handler for those
registers in the fast path so we can still get a performance boost
from partitioning.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/include/asm/arm_pmuv3.h      |  37 ++++-
 arch/arm64/include/asm/kvm_pmu.h        |  10 ++
 arch/arm64/kvm/hyp/include/hyp/switch.h | 174 ++++++++++++++++++++++++
 arch/arm64/kvm/pmu.c                    |  16 +++
 arch/arm64/kvm/sys_regs.c               |  16 ---
 5 files changed, 236 insertions(+), 17 deletions(-)

diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h
index 3bddde5f4ebb..12004fd04018 100644
--- a/arch/arm64/include/asm/arm_pmuv3.h
+++ b/arch/arm64/include/asm/arm_pmuv3.h
@@ -41,6 +41,16 @@ static inline unsigned long read_pmevtypern(int n)
 	return 0;
 }
 
+static inline void write_pmxevcntr(u64 val)
+{
+	write_sysreg(val, pmxevcntr_el0);
+}
+
+static inline u64 read_pmxevcntr(void)
+{
+	return read_sysreg(pmxevcntr_el0);
+}
+
 static inline unsigned long read_pmmir(void)
 {
 	return read_cpuid(PMMIR_EL1);
@@ -107,21 +117,41 @@ static inline void write_pmcntenset(u64 val)
 	write_sysreg(val, pmcntenset_el0);
 }
 
+static inline u64 read_pmcntenset(void)
+{
+	return read_sysreg(pmcntenset_el0);
+}
+
 static inline void write_pmcntenclr(u64 val)
 {
 	write_sysreg(val, pmcntenclr_el0);
 }
 
+static inline u64 read_pmcntenclr(void)
+{
+	return read_sysreg(pmcntenclr_el0);
+}
+
 static inline void write_pmintenset(u64 val)
 {
 	write_sysreg(val, pmintenset_el1);
 }
 
+static inline u64 read_pmintenset(void)
+{
+	return read_sysreg(pmintenset_el1);
+}
+
 static inline void write_pmintenclr(u64 val)
 {
 	write_sysreg(val, pmintenclr_el1);
 }
 
+static inline u64 read_pmintenclr(void)
+{
+	return read_sysreg(pmintenclr_el1);
+}
+
 static inline void write_pmccfiltr(u64 val)
 {
 	write_sysreg(val, pmccfiltr_el0);
@@ -162,11 +192,16 @@ static inline u64 read_pmovsclr(void)
 	return read_sysreg(pmovsclr_el0);
 }
 
-static inline void write_pmuserenr(u32 val)
+static inline void write_pmuserenr(u64 val)
 {
 	write_sysreg(val, pmuserenr_el0);
 }
 
+static inline u64 read_pmuserenr(void)
+{
+	return read_sysreg(pmuserenr_el0);
+}
+
 static inline void write_pmuacr(u64 val)
 {
 	write_sysreg_s(val, SYS_PMUACR_EL1);
diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
index 73b7161e3f4e..62c8032a548f 100644
--- a/arch/arm64/include/asm/kvm_pmu.h
+++ b/arch/arm64/include/asm/kvm_pmu.h
@@ -81,6 +81,8 @@ struct kvm_pmu_events *kvm_get_pmu_events(void);
 void kvm_set_pmu_events(u64 set, struct perf_event_attr *attr);
 void kvm_clr_pmu_events(u64 clr);
 bool kvm_set_pmuserenr(u64 val);
+bool check_pmu_access_disabled(struct kvm_vcpu *vcpu, u64 flags);
+bool pmu_access_el0_disabled(struct kvm_vcpu *vcpu);
 void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu);
 void kvm_vcpu_pmu_restore_host(struct kvm_vcpu *vcpu);
 void kvm_vcpu_pmu_resync_el0(void);
@@ -214,6 +216,14 @@ static inline bool kvm_set_pmuserenr(u64 val)
 {
 	return false;
 }
+static inline bool check_pmu_access_disabled(struct kvm_vcpu *vcpu, u64 flags)
+{
+	return false;
+}
+static inline bool pmu_access_el0_disabled(struct kvm_vcpu *vcpu)
+{
+	return false;
+}
 static inline void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu) {}
 static inline void kvm_vcpu_pmu_restore_host(struct kvm_vcpu *vcpu) {}
 static inline void kvm_vcpu_reload_pmu(struct kvm_vcpu *vcpu) {}
diff --git a/arch/arm64/kvm/hyp/include/hyp/switch.h b/arch/arm64/kvm/hyp/include/hyp/switch.h
index 47d2db8446df..4920b7da9ce8 100644
--- a/arch/arm64/kvm/hyp/include/hyp/switch.h
+++ b/arch/arm64/kvm/hyp/include/hyp/switch.h
@@ -25,12 +25,14 @@
 #include <asm/kvm_emulate.h>
 #include <asm/kvm_hyp.h>
 #include <asm/kvm_mmu.h>
+#include <asm/kvm_pmu.h>
 #include <asm/kvm_nested.h>
 #include <asm/fpsimd.h>
 #include <asm/debug-monitors.h>
 #include <asm/processor.h>
 #include <asm/traps.h>
 
+#include <../../sys_regs.h>
 #include "arm_psci.h"
 
 struct kvm_exception_table_entry {
@@ -782,6 +784,175 @@ static bool handle_ampere1_tcr(struct kvm_vcpu *vcpu)
 	return true;
 }
 
+/**
+ * handle_pmu_reg() - Handle fast access to most PMU regs
+ * @vcpu: Ponter to kvm_vcpu struct
+ * @p: System register parameters (read/write, Op0, Op1, CRm, CRn, Op2)
+ * @reg: VCPU register identifier
+ * @rt: Target general register
+ * @val: Value to write
+ * @readfn: Sysreg read function
+ * @writefn: Sysreg write function
+ *
+ * Handle fast access to most PMU regs. Writethrough to the physical
+ * register. This function is a wrapper for the simplest case, but
+ * sadly there aren't many of those.
+ *
+ * Always return true. The boolean makes usage more consistent with
+ * similar functions.
+ *
+ * Return: True
+ */
+static bool handle_pmu_reg(struct kvm_vcpu *vcpu, struct sys_reg_params *p,
+			   enum vcpu_sysreg reg, u8 rt, u64 val,
+			   u64 (*readfn)(void), void (*writefn)(u64))
+{
+	if (p->is_write) {
+		__vcpu_assign_sys_reg(vcpu, reg, val);
+		writefn(val);
+	} else {
+		vcpu_set_reg(vcpu, rt, readfn());
+	}
+
+	return true;
+}
+
+/**
+ * kvm_hyp_handle_pmu_regs() - Fast handler for PMU registers
+ * @vcpu: Pointer to vcpu struct
+ *
+ * This handler immediately writes through certain PMU registers when
+ * we have a partitioned PMU (that is, MDCR_EL2.HPMN is set to reserve
+ * a range of counters for the guest) but the machine does not have
+ * FEAT_FGT to selectively untrap the registers we want.
+ *
+ * Return: True if the exception was successfully handled, false otherwise
+ */
+static bool kvm_hyp_handle_pmu_regs(struct kvm_vcpu *vcpu)
+{
+	struct sys_reg_params p;
+	u64 esr;
+	u32 sysreg;
+	u8 rt;
+	u64 val;
+	u8 idx;
+	bool ret;
+
+	if (!kvm_vcpu_pmu_is_partitioned(vcpu)
+	    || pmu_access_el0_disabled(vcpu))
+		return false;
+
+	esr = kvm_vcpu_get_esr(vcpu);
+	p = esr_sys64_to_params(esr);
+	sysreg = esr_sys64_to_sysreg(esr);
+	rt = kvm_vcpu_sys_get_rt(vcpu);
+	val = vcpu_get_reg(vcpu, rt);
+
+	switch (sysreg) {
+	case SYS_PMCR_EL0:
+		val &= ARMV8_PMU_PMCR_MASK;
+
+		if (p.is_write) {
+			write_pmcr(val);
+			__vcpu_assign_sys_reg(vcpu, PMCR_EL0, read_pmcr());
+		} else {
+			val = u64_replace_bits(
+				read_pmcr(),
+				vcpu->kvm->arch.nr_pmu_counters,
+				ARMV8_PMU_PMCR_N);
+			vcpu_set_reg(vcpu, rt, val);
+		}
+
+		ret = true;
+		break;
+	case SYS_PMUSERENR_EL0:
+		val &= ARMV8_PMU_USERENR_MASK;
+		ret = handle_pmu_reg(vcpu, &p, PMUSERENR_EL0, rt, val,
+				     &read_pmuserenr, &write_pmuserenr);
+		break;
+	case SYS_PMSELR_EL0:
+		val &= PMSELR_EL0_SEL_MASK;
+		ret = handle_pmu_reg(vcpu, &p, PMSELR_EL0, rt, val,
+				     &read_pmselr, &write_pmselr);
+		break;
+	case SYS_PMINTENCLR_EL1:
+		val &= kvm_pmu_accessible_counter_mask(vcpu);
+		if (p.is_write) {
+			__vcpu_rmw_sys_reg(vcpu, PMINTENSET_EL1, &=, ~val);
+			write_pmintenclr(val);
+		} else {
+			vcpu_set_reg(vcpu, rt, read_pmintenclr());
+		}
+		ret = true;
+		break;
+	case SYS_PMINTENSET_EL1:
+		val &= kvm_pmu_accessible_counter_mask(vcpu);
+		if (p.is_write) {
+			__vcpu_rmw_sys_reg(vcpu, PMINTENSET_EL1, |=, val);
+			write_pmintenset(val);
+		} else {
+			vcpu_set_reg(vcpu, rt, read_pmintenset());
+		}
+		ret = true;
+		break;
+	case SYS_PMCNTENCLR_EL0:
+		val &= kvm_pmu_accessible_counter_mask(vcpu);
+		if (p.is_write) {
+			__vcpu_rmw_sys_reg(vcpu, PMCNTENSET_EL0, &=, ~val);
+			write_pmcntenclr(val);
+		} else {
+			vcpu_set_reg(vcpu, rt, read_pmcntenclr());
+		}
+		ret = true;
+		break;
+	case SYS_PMCNTENSET_EL0:
+		val &= kvm_pmu_accessible_counter_mask(vcpu);
+		if (p.is_write) {
+			__vcpu_rmw_sys_reg(vcpu, PMCNTENSET_EL0, |=, val);
+			write_pmcntenset(val);
+		} else {
+			vcpu_set_reg(vcpu, rt, read_pmcntenset());
+		}
+		ret = true;
+		break;
+	case SYS_PMCCNTR_EL0:
+		ret = handle_pmu_reg(vcpu, &p, PMCCNTR_EL0, rt, val,
+				     &read_pmccntr, &write_pmccntr);
+		break;
+	case SYS_PMXEVCNTR_EL0:
+		idx = FIELD_GET(PMSELR_EL0_SEL, read_pmselr());
+
+		if (idx >= vcpu->kvm->arch.nr_pmu_counters)
+			return false;
+
+		ret = handle_pmu_reg(vcpu, &p, PMEVCNTR0_EL0 + idx, rt, val,
+				     &read_pmxevcntr, &write_pmxevcntr);
+		break;
+	case SYS_PMEVCNTRn_EL0(0) ... SYS_PMEVCNTRn_EL0(30):
+		idx = ((p.CRm & 3) << 3) | (p.Op2 & 7);
+
+		if (idx >= vcpu->kvm->arch.nr_pmu_counters)
+			return false;
+
+		if (p.is_write) {
+			write_pmevcntrn(idx, val);
+			__vcpu_assign_sys_reg(vcpu, PMEVCNTR0_EL0 + idx, val);
+		} else {
+			vcpu_set_reg(vcpu, rt, read_pmevcntrn(idx));
+		}
+
+		ret = true;
+		break;
+	default:
+		ret = false;
+	}
+
+	if (ret)
+		__kvm_skip_instr(vcpu);
+
+	return ret;
+}
+
 static inline bool kvm_hyp_handle_sysreg(struct kvm_vcpu *vcpu, u64 *exit_code)
 {
 	if (cpus_have_final_cap(ARM64_WORKAROUND_CAVIUM_TX2_219_TVM) &&
@@ -799,6 +970,9 @@ static inline bool kvm_hyp_handle_sysreg(struct kvm_vcpu *vcpu, u64 *exit_code)
 	if (kvm_handle_cntxct(vcpu))
 		return true;
 
+	if (kvm_hyp_handle_pmu_regs(vcpu))
+		return true;
+
 	return false;
 }
 
diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c
index 67216451b8ce..338e7eebf0d1 100644
--- a/arch/arm64/kvm/pmu.c
+++ b/arch/arm64/kvm/pmu.c
@@ -891,3 +891,19 @@ u64 kvm_vcpu_read_pmcr(struct kvm_vcpu *vcpu)
 
 	return u64_replace_bits(pmcr, n, ARMV8_PMU_PMCR_N);
 }
+
+bool check_pmu_access_disabled(struct kvm_vcpu *vcpu, u64 flags)
+{
+	u64 reg = __vcpu_sys_reg(vcpu, PMUSERENR_EL0);
+	bool enabled = (reg & flags) || vcpu_mode_priv(vcpu);
+
+	if (!enabled)
+		kvm_inject_undefined(vcpu);
+
+	return !enabled;
+}
+
+bool pmu_access_el0_disabled(struct kvm_vcpu *vcpu)
+{
+	return check_pmu_access_disabled(vcpu, ARMV8_PMU_USERENR_EN);
+}
diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
index cfbce4537b4c..b80cf6194fa3 100644
--- a/arch/arm64/kvm/sys_regs.c
+++ b/arch/arm64/kvm/sys_regs.c
@@ -842,22 +842,6 @@ static u64 reset_pmcr(struct kvm_vcpu *vcpu, const struct sys_reg_desc *r)
 	return __vcpu_sys_reg(vcpu, r->reg);
 }
 
-static bool check_pmu_access_disabled(struct kvm_vcpu *vcpu, u64 flags)
-{
-	u64 reg = __vcpu_sys_reg(vcpu, PMUSERENR_EL0);
-	bool enabled = (reg & flags) || vcpu_mode_priv(vcpu);
-
-	if (!enabled)
-		kvm_inject_undefined(vcpu);
-
-	return !enabled;
-}
-
-static bool pmu_access_el0_disabled(struct kvm_vcpu *vcpu)
-{
-	return check_pmu_access_disabled(vcpu, ARMV8_PMU_USERENR_EN);
-}
-
 static bool pmu_write_swinc_el0_disabled(struct kvm_vcpu *vcpu)
 {
 	return check_pmu_access_disabled(vcpu, ARMV8_PMU_USERENR_SW | ARMV8_PMU_USERENR_EN);
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 15/22] KVM: arm64: Setup MDCR_EL2 to handle a partitioned PMU

[Colton Lewis]

Setup MDCR_EL2 to handle a partitioned PMU. That means calculate an
appropriate value for HPMN instead of the maximum setting the host
allows (which implies no partition) so hardware enforces that a guest
will only see the counters in the guest partition.

With HPMN set, we can now leave the TPM and TPMCR bits unset unless
FGT is not available, in which case we need to fall back to that.

Also, if available, set the filtering bits HPMD and HCCD to be extra
sure nothing counts at EL2.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/include/asm/kvm_pmu.h | 11 ++++++
 arch/arm64/kvm/debug.c           | 23 ++++++++++---
 arch/arm64/kvm/pmu-part.c        | 57 ++++++++++++++++++++++++++++++++
 3 files changed, 86 insertions(+), 5 deletions(-)

diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
index 62c8032a548f..35674879aae0 100644
--- a/arch/arm64/include/asm/kvm_pmu.h
+++ b/arch/arm64/include/asm/kvm_pmu.h
@@ -96,6 +96,9 @@ u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu);
 void kvm_pmu_host_counters_enable(void);
 void kvm_pmu_host_counters_disable(void);
 
+u8 kvm_pmu_guest_num_counters(struct kvm_vcpu *vcpu);
+u8 kvm_pmu_hpmn(struct kvm_vcpu *vcpu);
+
 #if !defined(__KVM_NVHE_HYPERVISOR__)
 bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu);
 bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu);
@@ -158,6 +161,14 @@ static inline bool kvm_vcpu_pmu_use_fgt(struct kvm_vcpu *vcpu)
 {
 	return false;
 }
+static inline u8 kvm_pmu_guest_num_counters(struct kvm_vcpu *vcpu)
+{
+	return 0;
+}
+static inline u8 kvm_pmu_hpmn(struct kvm_vcpu *vcpu)
+{
+	return 0;
+}
 static inline void kvm_pmu_set_counter_value(struct kvm_vcpu *vcpu,
 					     u64 select_idx, u64 val) {}
 static inline void kvm_pmu_set_counter_value_user(struct kvm_vcpu *vcpu,
diff --git a/arch/arm64/kvm/debug.c b/arch/arm64/kvm/debug.c
index a554c3e368dc..b420fec3c754 100644
--- a/arch/arm64/kvm/debug.c
+++ b/arch/arm64/kvm/debug.c
@@ -37,15 +37,28 @@ static void kvm_arm_setup_mdcr_el2(struct kvm_vcpu *vcpu)
 	 * This also clears MDCR_EL2_E2PB_MASK and MDCR_EL2_E2TB_MASK
 	 * to disable guest access to the profiling and trace buffers
 	 */
-	vcpu->arch.mdcr_el2 = FIELD_PREP(MDCR_EL2_HPMN,
-					 *host_data_ptr(nr_event_counters));
-	vcpu->arch.mdcr_el2 |= (MDCR_EL2_TPM |
-				MDCR_EL2_TPMS |
-				MDCR_EL2_TTRF |
+	vcpu->arch.mdcr_el2 = FIELD_PREP(MDCR_EL2_HPMN, kvm_pmu_hpmn(vcpu));
+	vcpu->arch.mdcr_el2 |= (MDCR_EL2_TTRF |
 				MDCR_EL2_TPMCR |
 				MDCR_EL2_TDRA |
 				MDCR_EL2_TDOSA);
 
+	if (kvm_vcpu_pmu_is_partitioned(vcpu)
+	    && is_pmuv3p1(read_pmuver())) {
+		/*
+		 * Filtering these should be redundant because we trap
+		 * all the TYPER and FILTR registers anyway and ensure
+		 * they filter EL2, but set the bits if they are here.
+		 */
+		vcpu->arch.mdcr_el2 |= MDCR_EL2_HPMD;
+
+		if (is_pmuv3p5(read_pmuver()))
+			vcpu->arch.mdcr_el2 |= MDCR_EL2_HCCD;
+	}
+
+	if (!kvm_vcpu_pmu_use_fgt(vcpu))
+		vcpu->arch.mdcr_el2 |= MDCR_EL2_TPM | MDCR_EL2_TPMCR;
+
 	/* Is the VM being debugged by userspace? */
 	if (vcpu->guest_debug)
 		/* Route all software debug exceptions to EL2 */
diff --git a/arch/arm64/kvm/pmu-part.c b/arch/arm64/kvm/pmu-part.c
index 92775e19cbf6..f954d2d29314 100644
--- a/arch/arm64/kvm/pmu-part.c
+++ b/arch/arm64/kvm/pmu-part.c
@@ -137,3 +137,60 @@ void kvm_pmu_host_counters_disable(void)
 	mdcr &= ~MDCR_EL2_HPME;
 	write_sysreg(mdcr, mdcr_el2);
 }
+
+/**
+ * kvm_pmu_guest_num_counters() - Number of counters to show to guest
+ * @vcpu: Pointer to struct kvm_vcpu
+ *
+ * Calculate the number of counters to show to the guest via
+ * PMCR_EL0.N, making sure to respect the maximum the host allows,
+ * which is hpmn_max if partitioned and host_max otherwise.
+ *
+ * Return: Valid value for PMCR_EL0.N
+ */
+u8 kvm_pmu_guest_num_counters(struct kvm_vcpu *vcpu)
+{
+	u8 nr_cnt = vcpu->kvm->arch.nr_pmu_counters;
+	int hpmn_max = vcpu->kvm->arch.arm_pmu->hpmn_max;
+	u8 host_max = *host_data_ptr(nr_event_counters);
+
+	if (kvm_vcpu_pmu_is_partitioned(vcpu)) {
+		if (nr_cnt <= hpmn_max && nr_cnt <= host_max)
+			return nr_cnt;
+		if (hpmn_max <= host_max)
+			return hpmn_max;
+	}
+
+	if (nr_cnt <= host_max)
+		return nr_cnt;
+
+	return host_max;
+}
+
+/**
+ * kvm_pmu_hpmn() - Calculate HPMN field value
+ * @vcpu: Pointer to struct kvm_vcpu
+ *
+ * Calculate the appropriate value to set for MDCR_EL2.HPMN, ensuring
+ * it always stays below the number of counters on the current CPU and
+ * above 0 unless the CPU has FEAT_HPMN0.
+ *
+ * This function works whether or not the PMU is partitioned.
+ *
+ * Return: A valid HPMN value
+ */
+u8 kvm_pmu_hpmn(struct kvm_vcpu *vcpu)
+{
+	u8 hpmn = kvm_pmu_guest_num_counters(vcpu);
+	int hpmn_max = vcpu->kvm->arch.arm_pmu->hpmn_max;
+	u8 host_max = *host_data_ptr(nr_event_counters);
+
+	if (hpmn == 0 && !cpus_have_final_cap(ARM64_HAS_HPMN0)) {
+		if (kvm_vcpu_pmu_is_partitioned(vcpu))
+			return hpmn_max;
+		else
+			return host_max;
+	}
+
+	return hpmn;
+}
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 16/22] KVM: arm64: Account for partitioning in PMCR_EL0 access

[Colton Lewis]

For some reason unknown to me, KVM allows writes to PMCR_EL0.N even
though the architecture specifies that field as RO. Make sure these
accesses conform to additional constraints imposed when the PMU is
partitioned.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/kvm/pmu.c      | 2 +-
 arch/arm64/kvm/sys_regs.c | 4 +++-
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c
index 338e7eebf0d1..9469f1e0a0b6 100644
--- a/arch/arm64/kvm/pmu.c
+++ b/arch/arm64/kvm/pmu.c
@@ -884,7 +884,7 @@ u64 kvm_pmu_accessible_counter_mask(struct kvm_vcpu *vcpu)
 u64 kvm_vcpu_read_pmcr(struct kvm_vcpu *vcpu)
 {
 	u64 pmcr = __vcpu_sys_reg(vcpu, PMCR_EL0);
-	u64 n = vcpu->kvm->arch.nr_pmu_counters;
+	u64 n = kvm_pmu_guest_num_counters(vcpu);
 
 	if (vcpu_has_nv(vcpu) && !vcpu_is_el2(vcpu))
 		n = FIELD_GET(MDCR_EL2_HPMN, __vcpu_sys_reg(vcpu, MDCR_EL2));
diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c
index b80cf6194fa3..e3d53f2da60b 100644
--- a/arch/arm64/kvm/sys_regs.c
+++ b/arch/arm64/kvm/sys_regs.c
@@ -1249,7 +1249,9 @@ static int set_pmcr(struct kvm_vcpu *vcpu, const struct sys_reg_desc *r,
 	 */
 	if (!kvm_vm_has_ran_once(kvm) &&
 	    !vcpu_has_nv(vcpu)	      &&
-	    new_n <= kvm_arm_pmu_get_max_counters(kvm))
+	    new_n <= kvm_arm_pmu_get_max_counters(kvm) &&
+	    (!kvm_vcpu_pmu_is_partitioned(vcpu) ||
+	     new_n <= kvm_pmu_hpmn(vcpu)))
 		kvm->arch.nr_pmu_counters = new_n;
 
 	mutex_unlock(&kvm->arch.config_lock);
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 17/22] KVM: arm64: Context swap Partitioned PMU guest registers

[Colton Lewis]

Save and restore newly untrapped registers that can be directly
accessed by the guest when the PMU is partitioned.

* PMEVCNTRn_EL0
* PMCCNTR_EL0
* PMICNTR_EL0
* PMUSERENR_EL0
* PMSELR_EL0
* PMCR_EL0
* PMCNTEN_EL0
* PMINTEN_EL1

If the PMU is not partitioned or MDCR_EL2.TPM is set, all PMU
registers are trapped so return immediately.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/include/asm/kvm_pmu.h |   4 ++
 arch/arm64/kvm/arm.c             |   2 +
 arch/arm64/kvm/pmu-part.c        | 101 +++++++++++++++++++++++++++++++
 3 files changed, 107 insertions(+)

diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
index 35674879aae0..4f0741bf6779 100644
--- a/arch/arm64/include/asm/kvm_pmu.h
+++ b/arch/arm64/include/asm/kvm_pmu.h
@@ -98,6 +98,8 @@ void kvm_pmu_host_counters_disable(void);
 
 u8 kvm_pmu_guest_num_counters(struct kvm_vcpu *vcpu);
 u8 kvm_pmu_hpmn(struct kvm_vcpu *vcpu);
+void kvm_pmu_load(struct kvm_vcpu *vcpu);
+void kvm_pmu_put(struct kvm_vcpu *vcpu);
 
 #if !defined(__KVM_NVHE_HYPERVISOR__)
 bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu);
@@ -169,6 +171,8 @@ static inline u8 kvm_pmu_hpmn(struct kvm_vcpu *vcpu)
 {
 	return 0;
 }
+static inline void kvm_pmu_load(struct kvm_vcpu *vcpu) {}
+static inline void kvm_pmu_put(struct kvm_vcpu *vcpu) {}
 static inline void kvm_pmu_set_counter_value(struct kvm_vcpu *vcpu,
 					     u64 select_idx, u64 val) {}
 static inline void kvm_pmu_set_counter_value_user(struct kvm_vcpu *vcpu,
diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c
index e452aba1a3b2..7c007ee44ecb 100644
--- a/arch/arm64/kvm/arm.c
+++ b/arch/arm64/kvm/arm.c
@@ -616,6 +616,7 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 		kvm_vcpu_load_vhe(vcpu);
 	kvm_arch_vcpu_load_fp(vcpu);
 	kvm_vcpu_pmu_restore_guest(vcpu);
+	kvm_pmu_load(vcpu);
 	if (kvm_arm_is_pvtime_enabled(&vcpu->arch))
 		kvm_make_request(KVM_REQ_RECORD_STEAL, vcpu);
 
@@ -658,6 +659,7 @@ void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
 	kvm_timer_vcpu_put(vcpu);
 	kvm_vgic_put(vcpu);
 	kvm_vcpu_pmu_restore_host(vcpu);
+	kvm_pmu_put(vcpu);
 	if (vcpu_has_nv(vcpu))
 		kvm_vcpu_put_hw_mmu(vcpu);
 	kvm_arm_vmid_clear_active();
diff --git a/arch/arm64/kvm/pmu-part.c b/arch/arm64/kvm/pmu-part.c
index f954d2d29314..5eb53c6409e7 100644
--- a/arch/arm64/kvm/pmu-part.c
+++ b/arch/arm64/kvm/pmu-part.c
@@ -9,6 +9,7 @@
 #include <linux/perf/arm_pmuv3.h>
 
 #include <asm/arm_pmuv3.h>
+#include <asm/kvm_emulate.h>
 #include <asm/kvm_pmu.h>
 
 /**
@@ -194,3 +195,103 @@ u8 kvm_pmu_hpmn(struct kvm_vcpu *vcpu)
 
 	return hpmn;
 }
+
+/**
+ * kvm_pmu_load() - Load untrapped PMU registers
+ * @vcpu: Pointer to struct kvm_vcpu
+ *
+ * Load all untrapped PMU registers from the VCPU into the PCPU. Mask
+ * to only bits belonging to guest-reserved counters and leave
+ * host-reserved counters alone in bitmask registers.
+ */
+void kvm_pmu_load(struct kvm_vcpu *vcpu)
+{
+	struct arm_pmu *pmu = vcpu->kvm->arch.arm_pmu;
+	u64 mask = kvm_pmu_guest_counter_mask(pmu);
+	u8 i;
+	u64 val;
+
+	/*
+	 * If the PMU is not partitioned or we have MDCR_EL2_TPM,
+	 * every PMU access is trapped so don't bother with the swap.
+	 */
+	if (!kvm_pmu_is_partitioned(pmu) || (vcpu->arch.mdcr_el2 & MDCR_EL2_TPM))
+		return;
+
+	for (i = 0; i < pmu->hpmn_max; i++) {
+		val = __vcpu_sys_reg(vcpu, PMEVCNTR0_EL0 + i);
+		write_pmevcntrn(i, val);
+	}
+
+	val = __vcpu_sys_reg(vcpu, PMCCNTR_EL0);
+	write_pmccntr(val);
+
+	val = __vcpu_sys_reg(vcpu, PMUSERENR_EL0);
+	write_pmuserenr(val);
+
+	val = __vcpu_sys_reg(vcpu, PMSELR_EL0);
+	write_pmselr(val);
+
+	val = __vcpu_sys_reg(vcpu, PMCR_EL0);
+	write_pmcr(val);
+
+	/*
+	 * Loading these registers is tricky because of
+	 * 1. Applying only the bits for guest counters (indicated by mask)
+	 * 2. Setting and clearing are different registers
+	 */
+	val = __vcpu_sys_reg(vcpu, PMCNTENSET_EL0);
+	write_pmcntenset(val & mask);
+	write_pmcntenclr(~val & mask);
+
+	val = __vcpu_sys_reg(vcpu, PMINTENSET_EL1);
+	write_pmintenset(val & mask);
+	write_pmintenclr(~val & mask);
+}
+
+/**
+ * kvm_pmu_put() - Put untrapped PMU registers
+ * @vcpu: Pointer to struct kvm_vcpu
+ *
+ * Put all untrapped PMU registers from the VCPU into the PCPU. Mask
+ * to only bits belonging to guest-reserved counters and leave
+ * host-reserved counters alone in bitmask registers.
+ */
+void kvm_pmu_put(struct kvm_vcpu *vcpu)
+{
+	struct arm_pmu *pmu = vcpu->kvm->arch.arm_pmu;
+	u64 mask = kvm_pmu_guest_counter_mask(pmu);
+	u8 i;
+	u64 val;
+
+	/*
+	 * If the PMU is not partitioned or we have MDCR_EL2_TPM,
+	 * every PMU access is trapped so don't bother with the swap.
+	 */
+	if (!kvm_pmu_is_partitioned(pmu) || (vcpu->arch.mdcr_el2 & MDCR_EL2_TPM))
+		return;
+
+	for (i = 0; i < pmu->hpmn_max; i++) {
+		val = read_pmevcntrn(i);
+		__vcpu_assign_sys_reg(vcpu, PMEVCNTR0_EL0 + i, val);
+	}
+
+	val = read_pmccntr();
+	__vcpu_assign_sys_reg(vcpu, PMCCNTR_EL0, val);
+
+	val = read_pmuserenr();
+	__vcpu_assign_sys_reg(vcpu, PMUSERENR_EL0, val);
+
+	val = read_pmselr();
+	__vcpu_assign_sys_reg(vcpu, PMSELR_EL0, val);
+
+	val = read_pmcr();
+	__vcpu_assign_sys_reg(vcpu, PMCR_EL0, val);
+
+	/* Mask these to only save the guest relevant bits. */
+	val = read_pmcntenset();
+	__vcpu_assign_sys_reg(vcpu, PMCNTENSET_EL0, val & mask);
+
+	val = read_pmintenset();
+	__vcpu_assign_sys_reg(vcpu, PMINTENSET_EL1, val & mask);
+}
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 18/22] KVM: arm64: Enforce PMU event filter at vcpu_load()

[Colton Lewis]

The KVM API for event filtering says that counters do not count when
blocked by the event filter. To enforce that, the event filter must be
rechecked on every load. If the event is filtered, exclude counting at
all exception levels before writing the hardware.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/kvm/pmu-part.c | 43 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 43 insertions(+)

diff --git a/arch/arm64/kvm/pmu-part.c b/arch/arm64/kvm/pmu-part.c
index 5eb53c6409e7..1451870757e1 100644
--- a/arch/arm64/kvm/pmu-part.c
+++ b/arch/arm64/kvm/pmu-part.c
@@ -196,6 +196,47 @@ u8 kvm_pmu_hpmn(struct kvm_vcpu *vcpu)
 	return hpmn;
 }
 
+/**
+ * kvm_pmu_apply_event_filter()
+ * @vcpu: Pointer to vcpu struct
+ *
+ * To uphold the guarantee of the KVM PMU event filter, we must ensure
+ * no counter counts if the event is filtered. Accomplish this by
+ * filtering all exception levels if the event is filtered.
+ */
+static void kvm_pmu_apply_event_filter(struct kvm_vcpu *vcpu)
+{
+	struct arm_pmu *pmu = vcpu->kvm->arch.arm_pmu;
+	u64 evtyper_set = kvm_pmu_evtyper_mask(vcpu->kvm)
+		& ~kvm_pmu_event_mask(vcpu->kvm)
+		& ~ARMV8_PMU_INCLUDE_EL2;
+	u64 evtyper_clr = ARMV8_PMU_INCLUDE_EL2;
+	u8 i;
+	u64 val;
+
+	for (i = 0; i < pmu->hpmn_max; i++) {
+		val = __vcpu_sys_reg(vcpu, PMEVTYPER0_EL0 + i);
+
+		if (vcpu->kvm->arch.pmu_filter &&
+		    !test_bit(val, vcpu->kvm->arch.pmu_filter)) {
+			val |= evtyper_set;
+			val &= ~evtyper_clr;
+		}
+
+		write_pmevtypern(i, val);
+	}
+
+	val = __vcpu_sys_reg(vcpu, PMCCFILTR_EL0);
+
+	if (vcpu->kvm->arch.pmu_filter &&
+	    !test_bit(ARMV8_PMUV3_PERFCTR_CPU_CYCLES, vcpu->kvm->arch.pmu_filter)) {
+		val |= evtyper_set;
+		val &= ~evtyper_clr;
+	}
+
+	write_pmccfiltr(val);
+}
+
 /**
  * kvm_pmu_load() - Load untrapped PMU registers
  * @vcpu: Pointer to struct kvm_vcpu
@@ -218,6 +259,8 @@ void kvm_pmu_load(struct kvm_vcpu *vcpu)
 	if (!kvm_pmu_is_partitioned(pmu) || (vcpu->arch.mdcr_el2 & MDCR_EL2_TPM))
 		return;
 
+	kvm_pmu_apply_event_filter(vcpu);
+
 	for (i = 0; i < pmu->hpmn_max; i++) {
 		val = __vcpu_sys_reg(vcpu, PMEVCNTR0_EL0 + i);
 		write_pmevcntrn(i, val);
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 19/22] perf: arm_pmuv3: Handle IRQs for Partitioned PMU guest counters

[Colton Lewis]

Guest counters will still trigger interrupts that need to be handled
by the host PMU interrupt handler. Clear the overflow flags in
hardware to handle the interrupt as normal, but record which guest
overflow flags were set in the virtual overflow register for later
injecting the interrupt into the guest.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm/include/asm/arm_pmuv3.h |  6 ++++++
 arch/arm64/include/asm/kvm_pmu.h |  2 ++
 arch/arm64/kvm/pmu-part.c        | 17 +++++++++++++++++
 drivers/perf/arm_pmuv3.c         |  9 +++++++++
 4 files changed, 34 insertions(+)

diff --git a/arch/arm/include/asm/arm_pmuv3.h b/arch/arm/include/asm/arm_pmuv3.h
index 5f6269039f44..36638efe4258 100644
--- a/arch/arm/include/asm/arm_pmuv3.h
+++ b/arch/arm/include/asm/arm_pmuv3.h
@@ -180,6 +180,11 @@ static inline void write_pmintenset(u32 val)
 	write_sysreg(val, PMINTENSET);
 }
 
+static inline u32 read_pmintenset(void)
+{
+	return read_sysreg(PMINTENSET);
+}
+
 static inline void write_pmintenclr(u32 val)
 {
 	write_sysreg(val, PMINTENCLR);
@@ -249,6 +254,7 @@ static inline u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu)
 	return ~0;
 }
 
+static inline void kvm_pmu_handle_guest_irq(u64 govf) {}
 
 static inline bool has_vhe(void)
 {
diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
index 4f0741bf6779..03e3bd318e4b 100644
--- a/arch/arm64/include/asm/kvm_pmu.h
+++ b/arch/arm64/include/asm/kvm_pmu.h
@@ -95,6 +95,7 @@ u64 kvm_pmu_host_counter_mask(struct arm_pmu *pmu);
 u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu);
 void kvm_pmu_host_counters_enable(void);
 void kvm_pmu_host_counters_disable(void);
+void kvm_pmu_handle_guest_irq(u64 govf);
 
 u8 kvm_pmu_guest_num_counters(struct kvm_vcpu *vcpu);
 u8 kvm_pmu_hpmn(struct kvm_vcpu *vcpu);
@@ -291,6 +292,7 @@ static inline u64 kvm_pmu_guest_counter_mask(struct arm_pmu *pmu)
 
 static inline void kvm_pmu_host_counters_enable(void) {}
 static inline void kvm_pmu_host_counters_disable(void) {}
+static inline void kvm_pmu_handle_guest_irq(u64 govf) {}
 
 #endif
 
diff --git a/arch/arm64/kvm/pmu-part.c b/arch/arm64/kvm/pmu-part.c
index 1451870757e1..bd04c37b19b9 100644
--- a/arch/arm64/kvm/pmu-part.c
+++ b/arch/arm64/kvm/pmu-part.c
@@ -338,3 +338,20 @@ void kvm_pmu_put(struct kvm_vcpu *vcpu)
 	val = read_pmintenset();
 	__vcpu_assign_sys_reg(vcpu, PMINTENSET_EL1, val & mask);
 }
+
+/**
+ * kvm_pmu_handle_guest_irq() - Record IRQs in guest counters
+ * @govf: Bitmask of guest overflowed counters
+ *
+ * Record IRQs from overflows in guest-reserved counters in the VCPU
+ * register for the guest to clear later.
+ */
+void kvm_pmu_handle_guest_irq(u64 govf)
+{
+	struct kvm_vcpu *vcpu = kvm_get_running_vcpu();
+
+	if (!vcpu)
+		return;
+
+	__vcpu_rmw_sys_reg(vcpu, PMOVSSET_EL0, |=, govf);
+}
diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c
index 5836672a3dd9..dba7dcee8258 100644
--- a/drivers/perf/arm_pmuv3.c
+++ b/drivers/perf/arm_pmuv3.c
@@ -750,6 +750,8 @@ static u64 armv8pmu_getreset_flags(void)
 
 	/* Write to clear flags */
 	value &= ARMV8_PMU_CNT_MASK_ALL;
+	/* Only reset interrupt enabled counters. */
+	value &= read_pmintenset();
 	write_pmovsclr(value);
 
 	return value;
@@ -852,6 +854,7 @@ static void armv8pmu_stop(struct arm_pmu *cpu_pmu)
 static irqreturn_t armv8pmu_handle_irq(struct arm_pmu *cpu_pmu)
 {
 	u64 pmovsr;
+	u64 govf;
 	struct perf_sample_data data;
 	struct pmu_hw_events *cpuc = this_cpu_ptr(cpu_pmu->hw_events);
 	struct pt_regs *regs;
@@ -906,6 +909,12 @@ static irqreturn_t armv8pmu_handle_irq(struct arm_pmu *cpu_pmu)
 		 */
 		perf_event_overflow(event, &data, regs);
 	}
+
+	govf = pmovsr & kvm_pmu_guest_counter_mask(cpu_pmu);
+
+	if (kvm_pmu_is_partitioned(cpu_pmu) && govf)
+		kvm_pmu_handle_guest_irq(govf);
+
 	armv8pmu_start(cpu_pmu);
 
 	return IRQ_HANDLED;
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 20/22] KVM: arm64: Inject recorded guest interrupts

[Colton Lewis]

When we re-enter the VM after handling a PMU interrupt, calculate
whether it was any of the guest counters that overflowed and inject an
interrupt into the guest if so.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 arch/arm64/include/asm/kvm_pmu.h |  2 ++
 arch/arm64/kvm/pmu-emul.c        |  4 ++--
 arch/arm64/kvm/pmu-part.c        | 22 +++++++++++++++++++++-
 arch/arm64/kvm/pmu.c             |  6 +++++-
 4 files changed, 30 insertions(+), 4 deletions(-)

diff --git a/arch/arm64/include/asm/kvm_pmu.h b/arch/arm64/include/asm/kvm_pmu.h
index 03e3bd318e4b..d047def897bc 100644
--- a/arch/arm64/include/asm/kvm_pmu.h
+++ b/arch/arm64/include/asm/kvm_pmu.h
@@ -86,6 +86,8 @@ bool pmu_access_el0_disabled(struct kvm_vcpu *vcpu);
 void kvm_vcpu_pmu_restore_guest(struct kvm_vcpu *vcpu);
 void kvm_vcpu_pmu_restore_host(struct kvm_vcpu *vcpu);
 void kvm_vcpu_pmu_resync_el0(void);
+bool kvm_pmu_emul_overflow_status(struct kvm_vcpu *vcpu);
+bool kvm_pmu_part_overflow_status(struct kvm_vcpu *vcpu);
 
 #define kvm_vcpu_has_pmu(vcpu)					\
 	(vcpu_has_feature(vcpu, KVM_ARM_VCPU_PMU_V3))
diff --git a/arch/arm64/kvm/pmu-emul.c b/arch/arm64/kvm/pmu-emul.c
index bcaa9f7a8ca2..6f41fc3e3f74 100644
--- a/arch/arm64/kvm/pmu-emul.c
+++ b/arch/arm64/kvm/pmu-emul.c
@@ -268,7 +268,7 @@ void kvm_pmu_reprogram_counter_mask(struct kvm_vcpu *vcpu, u64 val)
  * counter where the values of the global enable control, PMOVSSET_EL0[n], and
  * PMINTENSET_EL1[n] are all 1.
  */
-bool kvm_pmu_overflow_status(struct kvm_vcpu *vcpu)
+bool kvm_pmu_emul_overflow_status(struct kvm_vcpu *vcpu)
 {
 	u64 reg = __vcpu_sys_reg(vcpu, PMOVSSET_EL0);
 
@@ -405,7 +405,7 @@ static void kvm_pmu_perf_overflow(struct perf_event *perf_event,
 		kvm_pmu_counter_increment(vcpu, BIT(idx + 1),
 					  ARMV8_PMUV3_PERFCTR_CHAIN);
 
-	if (kvm_pmu_overflow_status(vcpu)) {
+	if (kvm_pmu_emul_overflow_status(vcpu)) {
 		kvm_make_request(KVM_REQ_IRQ_PENDING, vcpu);
 
 		if (!in_nmi())
diff --git a/arch/arm64/kvm/pmu-part.c b/arch/arm64/kvm/pmu-part.c
index bd04c37b19b9..165d1eae2634 100644
--- a/arch/arm64/kvm/pmu-part.c
+++ b/arch/arm64/kvm/pmu-part.c
@@ -279,7 +279,7 @@ void kvm_pmu_load(struct kvm_vcpu *vcpu)
 	write_pmcr(val);
 
 	/*
-	 * Loading these registers is tricky because of
+	 * Loading these registers is more intricate because of
 	 * 1. Applying only the bits for guest counters (indicated by mask)
 	 * 2. Setting and clearing are different registers
 	 */
@@ -355,3 +355,23 @@ void kvm_pmu_handle_guest_irq(u64 govf)
 
 	__vcpu_rmw_sys_reg(vcpu, PMOVSSET_EL0, |=, govf);
 }
+
+/**
+ * kvm_pmu_part_overflow_status() - Determine if any guest counters have overflowed
+ * @vcpu: Ponter to struct kvm_vcpu
+ *
+ * Determine if any guest counters have overflowed and therefore an
+ * IRQ needs to be injected into the guest.
+ *
+ * Return: True if there was an overflow, false otherwise
+ */
+bool kvm_pmu_part_overflow_status(struct kvm_vcpu *vcpu)
+{
+	struct arm_pmu *pmu = vcpu->kvm->arch.arm_pmu;
+	u64 mask = kvm_pmu_guest_counter_mask(pmu);
+	u64 pmovs = __vcpu_sys_reg(vcpu, PMOVSSET_EL0);
+	u64 pmint = read_pmintenset();
+	u64 pmcr = read_pmcr();
+
+	return (pmcr & ARMV8_PMU_PMCR_E) && (mask & pmovs & pmint);
+}
diff --git a/arch/arm64/kvm/pmu.c b/arch/arm64/kvm/pmu.c
index 9469f1e0a0b6..6ab0d23f9251 100644
--- a/arch/arm64/kvm/pmu.c
+++ b/arch/arm64/kvm/pmu.c
@@ -407,7 +407,11 @@ static void kvm_pmu_update_state(struct kvm_vcpu *vcpu)
 	struct kvm_pmu *pmu = &vcpu->arch.pmu;
 	bool overflow;
 
-	overflow = kvm_pmu_overflow_status(vcpu);
+	if (kvm_vcpu_pmu_is_partitioned(vcpu))
+		overflow = kvm_pmu_part_overflow_status(vcpu);
+	else
+		overflow = kvm_pmu_emul_overflow_status(vcpu);
+
 	if (pmu->irq_level == overflow)
 		return;
 
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 21/22] KVM: arm64: Add ioctl to partition the PMU when supported

[Colton Lewis]

Add KVM_ARM_PMU_PARTITION to enable the partitioned PMU for a given
vCPU. Add a corresponding KVM_CAP_ARM_PMU_PARTITION to check for this
ability. This capability is allowed on an initialized vCPU where PMUv3
and VHE are supported.

However, because the underlying ability relies on the driver being
passed some command line arguments to configure the hardware partition
at boot, enabling the partitioned PMU will not be allowed without the
underlying driver configuration even though the capability exists.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 Documentation/virt/kvm/api.rst    | 21 +++++++++++++++++++++
 arch/arm64/include/asm/kvm_host.h |  3 +++
 arch/arm64/kvm/arm.c              | 20 ++++++++++++++++++++
 arch/arm64/kvm/pmu-part.c         |  3 ++-
 include/uapi/linux/kvm.h          |  4 ++++
 5 files changed, 50 insertions(+), 1 deletion(-)

diff --git a/Documentation/virt/kvm/api.rst b/Documentation/virt/kvm/api.rst
index 4ef3d8482000..7e76f7c87598 100644
--- a/Documentation/virt/kvm/api.rst
+++ b/Documentation/virt/kvm/api.rst
@@ -6478,6 +6478,27 @@ the capability to be present.
 
 `flags` must currently be zero.
 
+4.144 KVM_ARM_PARTITION_PMU
+---------------------------
+
+:Capability: KVM_CAP_ARM_PARTITION_PMU
+:Architectures: arm64
+:Type: vcpu ioctl
+:Parameters: arg[0] is a boolean to enable the partitioned PMU
+
+This API controls the PMU implementation used for VMs. The capability
+is only available if the host PMUv3 driver was configured for
+partitioning via the module parameters `arm-pmuv3.partition_pmu=y` and
+`arm-pmuv3.reserved_guest_counters=[0-$NR_COUNTERS]`. When enabled,
+VMs are configured to have direct hardware access to the most
+frequently used registers for the counters configured by the
+aforementioned module parameters, bypassing the KVM traps in the
+standard emulated PMU implementation and reducing overhead of any
+guest software that uses PMU capabilities such as `perf`.
+
+If the host driver was configured for partitioning but the partitioned
+PMU is disabled through this interface, the VM will use the legacy PMU
+that shares the host partition.
 
 .. _kvm_run:
 
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 2df76689381a..3f835b28c855 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -369,6 +369,9 @@ struct kvm_arch {
 	/* Maximum number of counters for the guest */
 	u8 nr_pmu_counters;
 
+	/* Whether this guest uses the partitioned PMU */
+	bool partitioned_pmu_enable;
+
 	/* Iterator for idreg debugfs */
 	u8	idreg_debugfs_iter;
 
diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c
index 7c007ee44ecb..cb4380f39d98 100644
--- a/arch/arm64/kvm/arm.c
+++ b/arch/arm64/kvm/arm.c
@@ -21,6 +21,7 @@
 #include <linux/irqbypass.h>
 #include <linux/sched/stat.h>
 #include <linux/psci.h>
+#include <linux/perf/arm_pmu.h>
 #include <trace/events/kvm.h>
 
 #define CREATE_TRACE_POINTS
@@ -38,6 +39,7 @@
 #include <asm/kvm_emulate.h>
 #include <asm/kvm_mmu.h>
 #include <asm/kvm_nested.h>
+#include <asm/kvm_pmu.h>
 #include <asm/kvm_pkvm.h>
 #include <asm/kvm_pmu.h>
 #include <asm/kvm_ptrauth.h>
@@ -383,6 +385,9 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
 	case KVM_CAP_ARM_PMU_V3:
 		r = kvm_supports_guest_pmuv3();
 		break;
+	case KVM_CAP_ARM_PARTITION_PMU:
+		r = kvm_pmu_partition_supported();
+		break;
 	case KVM_CAP_ARM_INJECT_SERROR_ESR:
 		r = cpus_have_final_cap(ARM64_HAS_RAS_EXTN);
 		break;
@@ -1810,6 +1815,21 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 
 		return kvm_arm_vcpu_finalize(vcpu, what);
 	}
+	case KVM_ARM_PARTITION_PMU: {
+		bool enable;
+
+		if (unlikely(!kvm_vcpu_initialized(vcpu)))
+			return -ENOEXEC;
+
+		if (!kvm_pmu_is_partitioned(vcpu->kvm->arch.arm_pmu))
+			return -EPERM;
+
+		if (copy_from_user(&enable, argp, sizeof(enable)))
+			return -EFAULT;
+
+		vcpu->kvm->arch.partitioned_pmu_enable = enable;
+		return 0;
+	}
 	default:
 		r = -EINVAL;
 	}
diff --git a/arch/arm64/kvm/pmu-part.c b/arch/arm64/kvm/pmu-part.c
index 165d1eae2634..7b5c5fcc6623 100644
--- a/arch/arm64/kvm/pmu-part.c
+++ b/arch/arm64/kvm/pmu-part.c
@@ -53,7 +53,8 @@ bool kvm_pmu_is_partitioned(struct arm_pmu *pmu)
  */
 bool kvm_vcpu_pmu_is_partitioned(struct kvm_vcpu *vcpu)
 {
-	return kvm_pmu_is_partitioned(vcpu->kvm->arch.arm_pmu);
+	return kvm_pmu_is_partitioned(vcpu->kvm->arch.arm_pmu)
+		&& vcpu->kvm->arch.partitioned_pmu_enable;
 }
 
 /**
diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
index c74cf8f73337..2f8a8d4cfe3c 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -935,6 +935,7 @@ struct kvm_enable_cap {
 #define KVM_CAP_ARM_EL2_E2H0 241
 #define KVM_CAP_RISCV_MP_STATE_RESET 242
 #define KVM_CAP_GMEM_SHARED_MEM 243
+#define KVM_CAP_ARM_PARTITION_PMU 244
 
 struct kvm_irq_routing_irqchip {
 	__u32 irqchip;
@@ -1413,6 +1414,9 @@ struct kvm_enc_region {
 #define KVM_GET_SREGS2             _IOR(KVMIO,  0xcc, struct kvm_sregs2)
 #define KVM_SET_SREGS2             _IOW(KVMIO,  0xcd, struct kvm_sregs2)
 
+/* Available with KVM_CAP_ARM_PARTITION_PMU */
+#define KVM_ARM_PARTITION_PMU	_IOWR(KVMIO, 0xce, bool)
+
 #define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE    (1 << 0)
 #define KVM_DIRTY_LOG_INITIALLY_SET            (1 << 1)
 
-- 
2.50.0.727.gbf7dc18ff4-goog

[PATCH v3 22/22] KVM: arm64: selftests: Add test case for partitioned PMU

[Colton Lewis]

Run separate a test case for a partitioned PMU in vpmu_counter_access.

An enum is created specifying whether we are testing the emulated or
partitioned PMU and all the test functions are modified to take the
implementation as an argument and make the difference in setup
appropriately.

Because the test should still succeed even if we are on a machine
where we have the capability but the ioctl fails because the driver
was never configured properly, use __vcpu_ioctl to avoid checking the
return code.

Signed-off-by: Colton Lewis <coltonlewis@google.com>
---
 tools/include/uapi/linux/kvm.h                |  2 +
 .../selftests/kvm/arm64/vpmu_counter_access.c | 62 +++++++++++++------
 2 files changed, 46 insertions(+), 18 deletions(-)

diff --git a/tools/include/uapi/linux/kvm.h b/tools/include/uapi/linux/kvm.h
index b6ae8ad8934b..21a2c37528c8 100644
--- a/tools/include/uapi/linux/kvm.h
+++ b/tools/include/uapi/linux/kvm.h
@@ -930,6 +930,7 @@ struct kvm_enable_cap {
 #define KVM_CAP_X86_APIC_BUS_CYCLES_NS 237
 #define KVM_CAP_X86_GUEST_MODE 238
 #define KVM_CAP_ARM_WRITABLE_IMP_ID_REGS 239
+#define KVM_CAP_ARM_PARTITION_PMU 244
 
 struct kvm_irq_routing_irqchip {
 	__u32 irqchip;
@@ -1356,6 +1357,7 @@ struct kvm_vfio_spapr_tce {
 #define KVM_S390_SET_CMMA_BITS      _IOW(KVMIO, 0xb9, struct kvm_s390_cmma_log)
 /* Memory Encryption Commands */
 #define KVM_MEMORY_ENCRYPT_OP      _IOWR(KVMIO, 0xba, unsigned long)
+#define KVM_ARM_PARTITION_PMU	_IOWR(KVMIO, 0xce, bool)
 
 struct kvm_enc_region {
 	__u64 addr;
diff --git a/tools/testing/selftests/kvm/arm64/vpmu_counter_access.c b/tools/testing/selftests/kvm/arm64/vpmu_counter_access.c
index f16b3b27e32e..92e665516bc8 100644
--- a/tools/testing/selftests/kvm/arm64/vpmu_counter_access.c
+++ b/tools/testing/selftests/kvm/arm64/vpmu_counter_access.c
@@ -25,6 +25,16 @@
 /* The cycle counter bit position that's common among the PMU registers */
 #define ARMV8_PMU_CYCLE_IDX		31
 
+enum pmu_impl {
+	EMULATED,
+	PARTITIONED
+};
+
+const char *pmu_impl_str[] = {
+	"Emulated",
+	"Partitioned"
+};
+
 struct vpmu_vm {
 	struct kvm_vm *vm;
 	struct kvm_vcpu *vcpu;
@@ -405,7 +415,7 @@ static void guest_code(uint64_t expected_pmcr_n)
 }
 
 /* Create a VM that has one vCPU with PMUv3 configured. */
-static void create_vpmu_vm(void *guest_code)
+static void create_vpmu_vm(void *guest_code, enum pmu_impl impl)
 {
 	struct kvm_vcpu_init init;
 	uint8_t pmuver, ec;
@@ -419,6 +429,7 @@ static void create_vpmu_vm(void *guest_code)
 		.group = KVM_ARM_VCPU_PMU_V3_CTRL,
 		.attr = KVM_ARM_VCPU_PMU_V3_INIT,
 	};
+	bool partition = (impl == PARTITIONED);
 
 	/* The test creates the vpmu_vm multiple times. Ensure a clean state */
 	memset(&vpmu_vm, 0, sizeof(vpmu_vm));
@@ -449,6 +460,9 @@ static void create_vpmu_vm(void *guest_code)
 	/* Initialize vPMU */
 	vcpu_ioctl(vpmu_vm.vcpu, KVM_SET_DEVICE_ATTR, &irq_attr);
 	vcpu_ioctl(vpmu_vm.vcpu, KVM_SET_DEVICE_ATTR, &init_attr);
+
+	if (kvm_has_cap(KVM_CAP_ARM_PARTITION_PMU))
+		__vcpu_ioctl(vpmu_vm.vcpu, KVM_ARM_PARTITION_PMU, &partition);
 }
 
 static void destroy_vpmu_vm(void)
@@ -475,12 +489,12 @@ static void run_vcpu(struct kvm_vcpu *vcpu, uint64_t pmcr_n)
 	}
 }
 
-static void test_create_vpmu_vm_with_pmcr_n(uint64_t pmcr_n, bool expect_fail)
+static void test_create_vpmu_vm_with_pmcr_n(uint64_t pmcr_n, enum pmu_impl impl, bool expect_fail)
 {
 	struct kvm_vcpu *vcpu;
 	uint64_t pmcr, pmcr_orig;
 
-	create_vpmu_vm(guest_code);
+	create_vpmu_vm(guest_code, impl);
 	vcpu = vpmu_vm.vcpu;
 
 	pmcr_orig = vcpu_get_reg(vcpu, KVM_ARM64_SYS_REG(SYS_PMCR_EL0));
@@ -508,7 +522,7 @@ static void test_create_vpmu_vm_with_pmcr_n(uint64_t pmcr_n, bool expect_fail)
  * Create a guest with one vCPU, set the PMCR_EL0.N for the vCPU to @pmcr_n,
  * and run the test.
  */
-static void run_access_test(uint64_t pmcr_n)
+static void run_access_test(uint64_t pmcr_n, enum pmu_impl impl)
 {
 	uint64_t sp;
 	struct kvm_vcpu *vcpu;
@@ -516,7 +530,7 @@ static void run_access_test(uint64_t pmcr_n)
 
 	pr_debug("Test with pmcr_n %lu\n", pmcr_n);
 
-	test_create_vpmu_vm_with_pmcr_n(pmcr_n, false);
+	test_create_vpmu_vm_with_pmcr_n(pmcr_n, impl, false);
 	vcpu = vpmu_vm.vcpu;
 
 	/* Save the initial sp to restore them later to run the guest again */
@@ -550,14 +564,14 @@ static struct pmreg_sets validity_check_reg_sets[] = {
  * Create a VM, and check if KVM handles the userspace accesses of
  * the PMU register sets in @validity_check_reg_sets[] correctly.
  */
-static void run_pmregs_validity_test(uint64_t pmcr_n)
+static void run_pmregs_validity_test(uint64_t pmcr_n, enum pmu_impl impl)
 {
 	int i;
 	struct kvm_vcpu *vcpu;
 	uint64_t set_reg_id, clr_reg_id, reg_val;
 	uint64_t valid_counters_mask, max_counters_mask;
 
-	test_create_vpmu_vm_with_pmcr_n(pmcr_n, false);
+	test_create_vpmu_vm_with_pmcr_n(pmcr_n, impl, false);
 	vcpu = vpmu_vm.vcpu;
 
 	valid_counters_mask = get_counters_mask(pmcr_n);
@@ -607,11 +621,11 @@ static void run_pmregs_validity_test(uint64_t pmcr_n)
  * the vCPU to @pmcr_n, which is larger than the host value.
  * The attempt should fail as @pmcr_n is too big to set for the vCPU.
  */
-static void run_error_test(uint64_t pmcr_n)
+static void run_error_test(uint64_t pmcr_n, enum pmu_impl impl)
 {
-	pr_debug("Error test with pmcr_n %lu (larger than the host)\n", pmcr_n);
+	pr_debug("Error test with pmcr_n %lu (larger than the host allows)\n", pmcr_n);
 
-	test_create_vpmu_vm_with_pmcr_n(pmcr_n, true);
+	test_create_vpmu_vm_with_pmcr_n(pmcr_n, impl, true);
 	destroy_vpmu_vm();
 }
 
@@ -619,30 +633,42 @@ static void run_error_test(uint64_t pmcr_n)
  * Return the default number of implemented PMU event counters excluding
  * the cycle counter (i.e. PMCR_EL0.N value) for the guest.
  */
-static uint64_t get_pmcr_n_limit(void)
+static uint64_t get_pmcr_n_limit(enum pmu_impl impl)
 {
 	uint64_t pmcr;
 
-	create_vpmu_vm(guest_code);
+	create_vpmu_vm(guest_code, impl);
 	pmcr = vcpu_get_reg(vpmu_vm.vcpu, KVM_ARM64_SYS_REG(SYS_PMCR_EL0));
 	destroy_vpmu_vm();
 	return get_pmcr_n(pmcr);
 }
 
-int main(void)
+void test_pmu(enum pmu_impl impl)
 {
 	uint64_t i, pmcr_n;
 
-	TEST_REQUIRE(kvm_has_cap(KVM_CAP_ARM_PMU_V3));
+	pr_info("Testing PMU: Implementation = %s\n", pmu_impl_str[impl]);
+
+	pmcr_n = get_pmcr_n_limit(impl);
+	pr_debug("PMCR_EL0.N: Limit = %lu\n", pmcr_n);
 
-	pmcr_n = get_pmcr_n_limit();
 	for (i = 0; i <= pmcr_n; i++) {
-		run_access_test(i);
-		run_pmregs_validity_test(i);
+		run_access_test(i, impl);
+		run_pmregs_validity_test(i, impl);
 	}
 
 	for (i = pmcr_n + 1; i < ARMV8_PMU_MAX_COUNTERS; i++)
-		run_error_test(i);
+		run_error_test(i, impl);
+}
+
+int main(void)
+{
+	TEST_REQUIRE(kvm_has_cap(KVM_CAP_ARM_PMU_V3));
+
+	test_pmu(EMULATED);
+
+	if (kvm_has_cap(KVM_CAP_ARM_PARTITION_PMU))
+		test_pmu(PARTITIONED);
 
 	return 0;
 }
-- 
2.50.0.727.gbf7dc18ff4-goog