diff mbox series

[RFC,v3,49/58] KVM: x86/pmu/svm: Set passthrough capability for vcpus

Message ID 20240801045907.4010984-50-mizhang@google.com (mailing list archive)
State New, archived
Headers show
Series Mediated Passthrough vPMU 3.0 for x86 | expand

Commit Message

Mingwei Zhang Aug. 1, 2024, 4:58 a.m. UTC 
From: Sandipan Das <sandipan.das@amd.com>

Pass on the passthrough PMU setting from kvm->arch into kvm_pmu for each
vcpu. As long as the host supports PerfMonV2, the guest PMU version does
not matter.

Note that guest vCPUs without a local APIC do not allocate an instance of
struct kvm_lapic because of which reading the guest LVTPC before switching
over to the PMI vector results in a NULL pointer dereference. Such vCPUs
also cannot receive PMIs. Hence, disable passthrough mode in such cases.

Signed-off-by: Sandipan Das <sandipan.das@amd.com>
Signed-off-by: Mingwei Zhang <mizhang@google.com>
---
 arch/x86/kvm/svm/pmu.c | 2 ++
 1 file changed, 2 insertions(+)
diff mbox series

Patch

diff --git a/arch/x86/kvm/svm/pmu.c b/arch/x86/kvm/svm/pmu.c
index 64060cbd8210..0a16f0eb2511 100644
--- a/arch/x86/kvm/svm/pmu.c
+++ b/arch/x86/kvm/svm/pmu.c
@@ -211,6 +211,8 @@  static void amd_pmu_refresh(struct kvm_vcpu *vcpu)
 	pmu->counter_bitmask[KVM_PMC_FIXED] = 0;
 	pmu->nr_arch_fixed_counters = 0;
 	bitmap_set(pmu->all_valid_pmc_idx, 0, pmu->nr_arch_gp_counters);
+	pmu->passthrough = vcpu->kvm->arch.enable_passthrough_pmu &&
+			   lapic_in_kernel(vcpu);
 
 	if (pmu->version > 1 || guest_cpuid_has(vcpu, X86_FEATURE_PERFCTR_CORE)) {
 		for (i = 0; i < pmu->nr_arch_gp_counters; i++) {