Re: [PATCH v4 09/21] KVM: x86/pmu: Switch to bitmask-based KVM PMU capabilities
From: Chen, Zide
Date: Wed Nov 18 2026 - 12:45:12 EST
On 11/18/2026 10:05 AM, Yosry Ahmed wrote:
> On Thu, Nov 12, 2026 at 01:32:15PM -0700, Zide Chen wrote:
>> From: Dapeng Mi <dapeng1.mi@xxxxxxxxxxxxxxx>
>>
>> Intel platforms support non-contiguous fixed counters via CPUID.0AH:ECX
>> starting with PerfMon v5, and support non-contiguous GP counters
>> through the Architectural PerfMon Extension (CPUID leaf 23H).
>>
>> struct x86_pmu_capability now exposes {,fixed_}cntr_mask64 bitmaps,
>> which may contain sparse bits representing non-contiguous counters.
>> Switch KVM's kvm_host_pmu and kvm_pmu_cap consumers over to the new
>> bitmask fields, including iterating kvm_host_pmu bitmasks directly in
>> VMX and SVM PMU MSR intercept setup.
>>
>> CPUID.0AH:EAX[15:8] and CPUID.0AH:EDX[4:0] enumerate only contiguous
>> counters. Derive these values from kvm_pmu_cap.{,fixed_}cntr_mask64 as
>> the number of consecutive counters starting at index 0.
>>
>> Signed-off-by: Dapeng Mi <dapeng1.mi@xxxxxxxxxxxxxxx>
>> Co-developed-by: Zide Chen <zide.chen@xxxxxxxxx>
>> Signed-off-by: Zide Chen <zide.chen@xxxxxxxxx>
>> ---
>> v4:
>> - Cleanly substitute kvm_pmu_cap.num_counters_{gp,fixed} with
>> find_first_zero_bit() in intel_pmu_refresh() without throwaway
>> code churn.
>> - Use find_first_zero_bit() instead of hweight64() for AMD CPUID
>> 0x80000022 EBX[3:0] (NumCorePmc) and svm_set_cpu_caps() to enumerate
>> contiguous counters. (Yosry)
>> - Do not carry over Jim's Reviewed-by from v3 due to reworking counter
>> clamping in intel_pmu_refresh().
>>
>> v3:
>> - Iterate kvm_host_pmu bitmasks directly in VMX and SVM intercept loops.
>> - Clamp nr_{gp,fixed}_counters to the number of contiguous counters
>> supported by KVM in intel_pmu_refresh(). (Jim, Sashiko)
>> - Drop redundant !enable_mediated_pmu clamp on fixed_cntr_mask64.
>> (Sashiko)
>> ---
>> arch/x86/kvm/cpuid.c | 17 ++++++++++++++---
>> arch/x86/kvm/msrs.c | 12 ++++++------
>> arch/x86/kvm/pmu.c | 17 ++++++++---------
>> arch/x86/kvm/svm/pmu.c | 2 +-
>> arch/x86/kvm/svm/svm.c | 8 ++++----
>> arch/x86/kvm/vmx/pmu_intel.c | 18 +++++++++++++++---
>> arch/x86/kvm/vmx/vmx.c | 4 ++--
>> 7 files changed, 50 insertions(+), 28 deletions(-)
>>
>> diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c
>> index adafd6fa2c54..427871ec365f 100644
>> --- a/arch/x86/kvm/cpuid.c
>> +++ b/arch/x86/kvm/cpuid.c
>> @@ -1535,10 +1535,18 @@ static inline int __do_cpuid_func(struct kvm_cpuid_array *array, u32 function)
>> }
>>
>> eax.split.version_id = kvm_pmu_cap.version;
>> - eax.split.num_counters = kvm_pmu_cap.num_counters_gp;
>> +
>> + /* Contiguous GP counters only. */
>> + eax.split.num_counters =
>> + find_first_zero_bit(kvm_pmu_cap.cntr_mask,
>> + KVM_MAX_NR_GP_COUNTERS);
>
> Should we add helpers for the find_first_zero_bit() calls across?
> Perhaps kvm_pmu_nr_contig_gp_counters() and
> kvm_pmu_nr_contig_fixed_counters()?
Agreed, encapsulating find_first_zero_bit() makes it more readable.
static inline unsigned int kvm_pmu_nr_contig_gp_counters(void)
{
return find_first_zero_bit(kvm_pmu_cap.cntr_mask,
KVM_MAX_NR_GP_COUNTERS);
}
static inline unsigned int kvm_pmu_nr_contig_fixed_counters(void)
{
return find_first_zero_bit(kvm_pmu_cap.fixed_cntr_mask,
KVM_MAX_NR_FIXED_COUNTERS);
}