@@ -794,6 +794,8 @@ struct kvm_x86_ops {
int (*check_nested_events)(struct kvm_vcpu *vcpu, bool external_intr);
void (*sched_in)(struct kvm_vcpu *kvm, int cpu);
+
+ struct kvm_pmu_ops *(*get_pmu_ops)(void);
};
struct kvm_arch_async_pf {
@@ -804,6 +806,7 @@ struct kvm_arch_async_pf {
};
extern struct kvm_x86_ops *kvm_x86_ops;
+extern struct kvm_pmu_ops *kvm_pmu_ops;
static inline void adjust_tsc_offset_guest(struct kvm_vcpu *vcpu,
s64 adjustment)
@@ -1104,16 +1107,6 @@ void kvm_complete_insn_gp(struct kvm_vcpu *vcpu, int err);
int kvm_is_in_guest(void);
-void kvm_pmu_init(struct kvm_vcpu *vcpu);
-void kvm_pmu_destroy(struct kvm_vcpu *vcpu);
-void kvm_pmu_reset(struct kvm_vcpu *vcpu);
-void kvm_pmu_cpuid_update(struct kvm_vcpu *vcpu);
-bool kvm_pmu_msr(struct kvm_vcpu *vcpu, u32 msr);
-int kvm_pmu_get_msr(struct kvm_vcpu *vcpu, u32 msr, u64 *data);
-int kvm_pmu_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info);
-int kvm_pmu_check_pmc(struct kvm_vcpu *vcpu, unsigned pmc);
-int kvm_pmu_read_pmc(struct kvm_vcpu *vcpu, unsigned pmc, u64 *data);
-void kvm_handle_pmu_event(struct kvm_vcpu *vcpu);
-void kvm_deliver_pmi(struct kvm_vcpu *vcpu);
-
+extern struct kvm_pmu_ops intel_pmu_ops;
+extern struct kvm_pmu_ops amd_pmu_ops;
#endif /* _ASM_X86_KVM_HOST_H */
@@ -14,9 +14,9 @@ kvm-$(CONFIG_KVM_DEVICE_ASSIGNMENT) += $(KVM)/assigned-dev.o $(KVM)/iommu.o
kvm-$(CONFIG_KVM_ASYNC_PF) += $(KVM)/async_pf.o
kvm-y += x86.o mmu.o emulate.o i8259.o irq.o lapic.o \
- i8254.o cpuid.o pmu.o
-kvm-intel-y += vmx.o
-kvm-amd-y += svm.o
+ i8254.o cpuid.o
+kvm-intel-y += vmx.o pmu_intel.o
+kvm-amd-y += svm.o pmu_amd.o
obj-$(CONFIG_KVM) += kvm.o
obj-$(CONFIG_KVM_INTEL) += kvm-intel.o
@@ -97,7 +97,7 @@ int kvm_update_cpuid(struct kvm_vcpu *vcpu)
((best->eax & 0xff00) >> 8) != 0)
return -EINVAL;
- kvm_pmu_cpuid_update(vcpu);
+ kvm_pmu_ops->pmu_cpuid_update(vcpu);
return 0;
}
@@ -1520,6 +1520,7 @@ int kvm_apic_local_deliver(struct kvm_lapic *apic, int lvt_type)
}
return 0;
}
+EXPORT_SYMBOL_GPL(kvm_apic_local_deliver);
void kvm_apic_nmi_wd_deliver(struct kvm_vcpu *vcpu)
{
deleted file mode 100644
@@ -1,571 +0,0 @@
-/*
- * Kernel-based Virtual Machine -- Performance Monitoring Unit support
- *
- * Copyright 2011 Red Hat, Inc. and/or its affiliates.
- *
- * Authors:
- * Avi Kivity <avi@redhat.com>
- * Gleb Natapov <gleb@redhat.com>
- *
- * This work is licensed under the terms of the GNU GPL, version 2. See
- * the COPYING file in the top-level directory.
- *
- */
-
-#include <linux/types.h>
-#include <linux/kvm_host.h>
-#include <linux/perf_event.h>
-#include <asm/perf_event.h>
-#include "x86.h"
-#include "cpuid.h"
-#include "lapic.h"
-
-struct kvm_event_hw_type_mapping intel_arch_events[] = {
- /* Index must match CPUID 0x0A.EBX bit vector */
- [0] = { 0x3c, 0x00, PERF_COUNT_HW_CPU_CYCLES },
- [1] = { 0xc0, 0x00, PERF_COUNT_HW_INSTRUCTIONS },
- [2] = { 0x3c, 0x01, PERF_COUNT_HW_BUS_CYCLES },
- [3] = { 0x2e, 0x4f, PERF_COUNT_HW_CACHE_REFERENCES },
- [4] = { 0x2e, 0x41, PERF_COUNT_HW_CACHE_MISSES },
- [5] = { 0xc4, 0x00, PERF_COUNT_HW_BRANCH_INSTRUCTIONS },
- [6] = { 0xc5, 0x00, PERF_COUNT_HW_BRANCH_MISSES },
- [7] = { 0x00, 0x30, PERF_COUNT_HW_REF_CPU_CYCLES },
-};
-
-/* mapping between fixed pmc index and intel_arch_events array */
-int fixed_pmc_events[] = {1, 0, 7};
-
-static bool pmc_is_gp(struct kvm_pmc *pmc)
-{
- return pmc->type == KVM_PMC_GP;
-}
-
-static inline u64 pmc_bitmask(struct kvm_pmc *pmc)
-{
- struct kvm_pmu *pmu = &pmc->vcpu->arch.pmu;
-
- return pmu->counter_bitmask[pmc->type];
-}
-
-static inline bool pmc_enabled(struct kvm_pmc *pmc)
-{
- struct kvm_pmu *pmu = &pmc->vcpu->arch.pmu;
- return test_bit(pmc->idx, (unsigned long *)&pmu->global_ctrl);
-}
-
-static inline struct kvm_pmc *get_gp_pmc(struct kvm_pmu *pmu, u32 msr,
- u32 base)
-{
- if (msr >= base && msr < base + pmu->nr_arch_gp_counters)
- return &pmu->gp_counters[msr - base];
- return NULL;
-}
-
-static inline struct kvm_pmc *get_fixed_pmc(struct kvm_pmu *pmu, u32 msr)
-{
- int base = MSR_CORE_PERF_FIXED_CTR0;
- if (msr >= base && msr < base + pmu->nr_arch_fixed_counters)
- return &pmu->fixed_counters[msr - base];
- return NULL;
-}
-
-static inline struct kvm_pmc *get_fixed_pmc_idx(struct kvm_pmu *pmu, int idx)
-{
- return get_fixed_pmc(pmu, MSR_CORE_PERF_FIXED_CTR0 + idx);
-}
-
-static struct kvm_pmc *global_idx_to_pmc(struct kvm_pmu *pmu, int idx)
-{
- if (idx < INTEL_PMC_IDX_FIXED)
- return get_gp_pmc(pmu, MSR_P6_EVNTSEL0 + idx, MSR_P6_EVNTSEL0);
- else
- return get_fixed_pmc_idx(pmu, idx - INTEL_PMC_IDX_FIXED);
-}
-
-void kvm_deliver_pmi(struct kvm_vcpu *vcpu)
-{
- if (vcpu->arch.apic)
- kvm_apic_local_deliver(vcpu->arch.apic, APIC_LVTPC);
-}
-
-static void trigger_pmi(struct irq_work *irq_work)
-{
- struct kvm_pmu *pmu = container_of(irq_work, struct kvm_pmu,
- irq_work);
- struct kvm_vcpu *vcpu = container_of(pmu, struct kvm_vcpu,
- arch.pmu);
-
- kvm_deliver_pmi(vcpu);
-}
-
-static void kvm_perf_overflow(struct perf_event *perf_event,
- struct perf_sample_data *data,
- struct pt_regs *regs)
-{
- struct kvm_pmc *pmc = perf_event->overflow_handler_context;
- struct kvm_pmu *pmu = &pmc->vcpu->arch.pmu;
- if (!test_and_set_bit(pmc->idx, (unsigned long *)&pmu->reprogram_pmi)) {
- __set_bit(pmc->idx, (unsigned long *)&pmu->global_status);
- kvm_make_request(KVM_REQ_PMU, pmc->vcpu);
- }
-}
-
-static void kvm_perf_overflow_intr(struct perf_event *perf_event,
- struct perf_sample_data *data, struct pt_regs *regs)
-{
- struct kvm_pmc *pmc = perf_event->overflow_handler_context;
- struct kvm_pmu *pmu = &pmc->vcpu->arch.pmu;
- if (!test_and_set_bit(pmc->idx, (unsigned long *)&pmu->reprogram_pmi)) {
- __set_bit(pmc->idx, (unsigned long *)&pmu->global_status);
- kvm_make_request(KVM_REQ_PMU, pmc->vcpu);
- /*
- * Inject PMI. If vcpu was in a guest mode during NMI PMI
- * can be ejected on a guest mode re-entry. Otherwise we can't
- * be sure that vcpu wasn't executing hlt instruction at the
- * time of vmexit and is not going to re-enter guest mode until,
- * woken up. So we should wake it, but this is impossible from
- * NMI context. Do it from irq work instead.
- */
- if (!kvm_is_in_guest())
- irq_work_queue(&pmc->vcpu->arch.pmu.irq_work);
- else
- kvm_make_request(KVM_REQ_PMI, pmc->vcpu);
- }
-}
-
-static u64 read_pmc(struct kvm_pmc *pmc)
-{
- u64 counter, enabled, running;
-
- counter = pmc->counter;
-
- if (pmc->perf_event)
- counter += perf_event_read_value(pmc->perf_event,
- &enabled, &running);
-
- /* FIXME: Scaling needed? */
-
- return counter & pmc_bitmask(pmc);
-}
-
-static void stop_counter(struct kvm_pmc *pmc)
-{
- if (pmc->perf_event) {
- pmc->counter = read_pmc(pmc);
- perf_event_release_kernel(pmc->perf_event);
- pmc->perf_event = NULL;
- }
-}
-
-static void reprogram_counter(struct kvm_pmc *pmc, u32 type,
- unsigned config, bool exclude_user, bool exclude_kernel,
- bool intr, bool in_tx, bool in_tx_cp)
-{
- struct perf_event *event;
- struct perf_event_attr attr = {
- .type = type,
- .size = sizeof(attr),
- .pinned = true,
- .exclude_idle = true,
- .exclude_host = 1,
- .exclude_user = exclude_user,
- .exclude_kernel = exclude_kernel,
- .config = config,
- };
- if (in_tx)
- attr.config |= HSW_IN_TX;
- if (in_tx_cp)
- attr.config |= HSW_IN_TX_CHECKPOINTED;
-
- attr.sample_period = (-pmc->counter) & pmc_bitmask(pmc);
-
- event = perf_event_create_kernel_counter(&attr, -1, current,
- intr ? kvm_perf_overflow_intr :
- kvm_perf_overflow, pmc);
- if (IS_ERR(event)) {
- printk_once("kvm: pmu event creation failed %ld\n",
- PTR_ERR(event));
- return;
- }
-
- pmc->perf_event = event;
- clear_bit(pmc->idx, (unsigned long*)&pmc->vcpu->arch.pmu.reprogram_pmi);
-}
-
-static unsigned find_arch_event(struct kvm_pmu *pmu, u8 event_select,
- u8 unit_mask)
-{
- int i;
-
- for (i = 0; i < ARRAY_SIZE(intel_arch_events); i++)
- if (intel_arch_events[i].eventsel == event_select
- && intel_arch_events[i].unit_mask == unit_mask
- && (pmu->available_event_types & (1 << i)))
- break;
-
- if (i == ARRAY_SIZE(intel_arch_events))
- return PERF_COUNT_HW_MAX;
-
- return intel_arch_events[i].event_type;
-}
-
-static void reprogram_gp_counter(struct kvm_pmc *pmc, u64 eventsel)
-{
- unsigned config, type = PERF_TYPE_RAW;
- u8 event_select, unit_mask;
-
- if (eventsel & ARCH_PERFMON_EVENTSEL_PIN_CONTROL)
- printk_once("kvm pmu: pin control bit is ignored\n");
-
- pmc->eventsel = eventsel;
-
- stop_counter(pmc);
-
- if (!(eventsel & ARCH_PERFMON_EVENTSEL_ENABLE) || !pmc_enabled(pmc))
- return;
-
- event_select = eventsel & ARCH_PERFMON_EVENTSEL_EVENT;
- unit_mask = (eventsel & ARCH_PERFMON_EVENTSEL_UMASK) >> 8;
-
- if (!(eventsel & (ARCH_PERFMON_EVENTSEL_EDGE |
- ARCH_PERFMON_EVENTSEL_INV |
- ARCH_PERFMON_EVENTSEL_CMASK |
- HSW_IN_TX |
- HSW_IN_TX_CHECKPOINTED))) {
- config = find_arch_event(&pmc->vcpu->arch.pmu, event_select,
- unit_mask);
- if (config != PERF_COUNT_HW_MAX)
- type = PERF_TYPE_HARDWARE;
- }
-
- if (type == PERF_TYPE_RAW)
- config = eventsel & X86_RAW_EVENT_MASK;
-
- reprogram_counter(pmc, type, config,
- !(eventsel & ARCH_PERFMON_EVENTSEL_USR),
- !(eventsel & ARCH_PERFMON_EVENTSEL_OS),
- eventsel & ARCH_PERFMON_EVENTSEL_INT,
- (eventsel & HSW_IN_TX),
- (eventsel & HSW_IN_TX_CHECKPOINTED));
-}
-
-static void reprogram_fixed_counter(struct kvm_pmc *pmc, u8 en_pmi, int idx)
-{
- unsigned en = en_pmi & 0x3;
- bool pmi = en_pmi & 0x8;
-
- stop_counter(pmc);
-
- if (!en || !pmc_enabled(pmc))
- return;
-
- reprogram_counter(pmc, PERF_TYPE_HARDWARE,
- intel_arch_events[fixed_pmc_events[idx]].event_type,
- !(en & 0x2), /* exclude user */
- !(en & 0x1), /* exclude kernel */
- pmi, false, false);
-}
-
-static inline u8 fixed_en_pmi(u64 ctrl, int idx)
-{
- return (ctrl >> (idx * 4)) & 0xf;
-}
-
-static void reprogram_fixed_counters(struct kvm_pmu *pmu, u64 data)
-{
- int i;
-
- for (i = 0; i < pmu->nr_arch_fixed_counters; i++) {
- u8 en_pmi = fixed_en_pmi(data, i);
- struct kvm_pmc *pmc = get_fixed_pmc_idx(pmu, i);
-
- if (fixed_en_pmi(pmu->fixed_ctr_ctrl, i) == en_pmi)
- continue;
-
- reprogram_fixed_counter(pmc, en_pmi, i);
- }
-
- pmu->fixed_ctr_ctrl = data;
-}
-
-static void reprogram_idx(struct kvm_pmu *pmu, int idx)
-{
- struct kvm_pmc *pmc = global_idx_to_pmc(pmu, idx);
-
- if (!pmc)
- return;
-
- if (pmc_is_gp(pmc))
- reprogram_gp_counter(pmc, pmc->eventsel);
- else {
- int fidx = idx - INTEL_PMC_IDX_FIXED;
- reprogram_fixed_counter(pmc,
- fixed_en_pmi(pmu->fixed_ctr_ctrl, fidx), fidx);
- }
-}
-
-static void global_ctrl_changed(struct kvm_pmu *pmu, u64 data)
-{
- int bit;
- u64 diff = pmu->global_ctrl ^ data;
-
- pmu->global_ctrl = data;
-
- for_each_set_bit(bit, (unsigned long *)&diff, X86_PMC_IDX_MAX)
- reprogram_idx(pmu, bit);
-}
-
-bool kvm_pmu_msr(struct kvm_vcpu *vcpu, u32 msr)
-{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- int ret;
-
- switch (msr) {
- case MSR_CORE_PERF_FIXED_CTR_CTRL:
- case MSR_CORE_PERF_GLOBAL_STATUS:
- case MSR_CORE_PERF_GLOBAL_CTRL:
- case MSR_CORE_PERF_GLOBAL_OVF_CTRL:
- ret = pmu->version > 1;
- break;
- default:
- ret = get_gp_pmc(pmu, msr, MSR_IA32_PERFCTR0)
- || get_gp_pmc(pmu, msr, MSR_P6_EVNTSEL0)
- || get_fixed_pmc(pmu, msr);
- break;
- }
- return ret;
-}
-
-int kvm_pmu_get_msr(struct kvm_vcpu *vcpu, u32 index, u64 *data)
-{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- struct kvm_pmc *pmc;
-
- switch (index) {
- case MSR_CORE_PERF_FIXED_CTR_CTRL:
- *data = pmu->fixed_ctr_ctrl;
- return 0;
- case MSR_CORE_PERF_GLOBAL_STATUS:
- *data = pmu->global_status;
- return 0;
- case MSR_CORE_PERF_GLOBAL_CTRL:
- *data = pmu->global_ctrl;
- return 0;
- case MSR_CORE_PERF_GLOBAL_OVF_CTRL:
- *data = pmu->global_ovf_ctrl;
- return 0;
- default:
- if ((pmc = get_gp_pmc(pmu, index, MSR_IA32_PERFCTR0)) ||
- (pmc = get_fixed_pmc(pmu, index))) {
- *data = read_pmc(pmc);
- return 0;
- } else if ((pmc = get_gp_pmc(pmu, index, MSR_P6_EVNTSEL0))) {
- *data = pmc->eventsel;
- return 0;
- }
- }
- return 1;
-}
-
-int kvm_pmu_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
-{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- struct kvm_pmc *pmc;
- u32 index = msr_info->index;
- u64 data = msr_info->data;
-
- switch (index) {
- case MSR_CORE_PERF_FIXED_CTR_CTRL:
- if (pmu->fixed_ctr_ctrl == data)
- return 0;
- if (!(data & 0xfffffffffffff444ull)) {
- reprogram_fixed_counters(pmu, data);
- return 0;
- }
- break;
- case MSR_CORE_PERF_GLOBAL_STATUS:
- if (msr_info->host_initiated) {
- pmu->global_status = data;
- return 0;
- }
- break; /* RO MSR */
- case MSR_CORE_PERF_GLOBAL_CTRL:
- if (pmu->global_ctrl == data)
- return 0;
- if (!(data & pmu->global_ctrl_mask)) {
- global_ctrl_changed(pmu, data);
- return 0;
- }
- break;
- case MSR_CORE_PERF_GLOBAL_OVF_CTRL:
- if (!(data & (pmu->global_ctrl_mask & ~(3ull<<62)))) {
- if (!msr_info->host_initiated)
- pmu->global_status &= ~data;
- pmu->global_ovf_ctrl = data;
- return 0;
- }
- break;
- default:
- if ((pmc = get_gp_pmc(pmu, index, MSR_IA32_PERFCTR0)) ||
- (pmc = get_fixed_pmc(pmu, index))) {
- if (!msr_info->host_initiated)
- data = (s64)(s32)data;
- pmc->counter += data - read_pmc(pmc);
- return 0;
- } else if ((pmc = get_gp_pmc(pmu, index, MSR_P6_EVNTSEL0))) {
- if (data == pmc->eventsel)
- return 0;
- if (!(data & pmu->reserved_bits)) {
- reprogram_gp_counter(pmc, data);
- return 0;
- }
- }
- }
- return 1;
-}
-
-int kvm_pmu_check_pmc(struct kvm_vcpu *vcpu, unsigned pmc)
-{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- bool fixed = pmc & (1u << 30);
- pmc &= ~(3u << 30);
- return (!fixed && pmc >= pmu->nr_arch_gp_counters) ||
- (fixed && pmc >= pmu->nr_arch_fixed_counters);
-}
-
-int kvm_pmu_read_pmc(struct kvm_vcpu *vcpu, unsigned pmc, u64 *data)
-{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- bool fast_mode = pmc & (1u << 31);
- bool fixed = pmc & (1u << 30);
- struct kvm_pmc *counters;
- u64 ctr;
-
- pmc &= ~(3u << 30);
- if (!fixed && pmc >= pmu->nr_arch_gp_counters)
- return 1;
- if (fixed && pmc >= pmu->nr_arch_fixed_counters)
- return 1;
- counters = fixed ? pmu->fixed_counters : pmu->gp_counters;
- ctr = read_pmc(&counters[pmc]);
- if (fast_mode)
- ctr = (u32)ctr;
- *data = ctr;
-
- return 0;
-}
-
-void kvm_pmu_cpuid_update(struct kvm_vcpu *vcpu)
-{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- struct kvm_cpuid_entry2 *entry;
- union cpuid10_eax eax;
- union cpuid10_edx edx;
-
- pmu->nr_arch_gp_counters = 0;
- pmu->nr_arch_fixed_counters = 0;
- pmu->counter_bitmask[KVM_PMC_GP] = 0;
- pmu->counter_bitmask[KVM_PMC_FIXED] = 0;
- pmu->version = 0;
- pmu->reserved_bits = 0xffffffff00200000ull;
-
- entry = kvm_find_cpuid_entry(vcpu, 0xa, 0);
- if (!entry)
- return;
- eax.full = entry->eax;
- edx.full = entry->edx;
-
- pmu->version = eax.split.version_id;
- if (!pmu->version)
- return;
-
- pmu->nr_arch_gp_counters = min_t(int, eax.split.num_counters,
- INTEL_PMC_MAX_GENERIC);
- pmu->counter_bitmask[KVM_PMC_GP] = ((u64)1 << eax.split.bit_width) - 1;
- pmu->available_event_types = ~entry->ebx &
- ((1ull << eax.split.mask_length) - 1);
-
- if (pmu->version == 1) {
- pmu->nr_arch_fixed_counters = 0;
- } else {
- pmu->nr_arch_fixed_counters =
- min_t(int, edx.split.num_counters_fixed,
- INTEL_PMC_MAX_FIXED);
- pmu->counter_bitmask[KVM_PMC_FIXED] =
- ((u64)1 << edx.split.bit_width_fixed) - 1;
- }
-
- pmu->global_ctrl = ((1 << pmu->nr_arch_gp_counters) - 1) |
- (((1ull << pmu->nr_arch_fixed_counters) - 1) << INTEL_PMC_IDX_FIXED);
- pmu->global_ctrl_mask = ~pmu->global_ctrl;
-
- entry = kvm_find_cpuid_entry(vcpu, 7, 0);
- if (entry &&
- (boot_cpu_has(X86_FEATURE_HLE) || boot_cpu_has(X86_FEATURE_RTM)) &&
- (entry->ebx & (X86_FEATURE_HLE|X86_FEATURE_RTM)))
- pmu->reserved_bits ^= HSW_IN_TX|HSW_IN_TX_CHECKPOINTED;
-}
-
-void kvm_pmu_init(struct kvm_vcpu *vcpu)
-{
- int i;
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
-
- memset(pmu, 0, sizeof(*pmu));
- for (i = 0; i < INTEL_PMC_MAX_GENERIC; i++) {
- pmu->gp_counters[i].type = KVM_PMC_GP;
- pmu->gp_counters[i].vcpu = vcpu;
- pmu->gp_counters[i].idx = i;
- }
- for (i = 0; i < INTEL_PMC_MAX_FIXED; i++) {
- pmu->fixed_counters[i].type = KVM_PMC_FIXED;
- pmu->fixed_counters[i].vcpu = vcpu;
- pmu->fixed_counters[i].idx = i + INTEL_PMC_IDX_FIXED;
- }
- init_irq_work(&pmu->irq_work, trigger_pmi);
- kvm_pmu_cpuid_update(vcpu);
-}
-
-void kvm_pmu_reset(struct kvm_vcpu *vcpu)
-{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- int i;
-
- irq_work_sync(&pmu->irq_work);
- for (i = 0; i < INTEL_PMC_MAX_GENERIC; i++) {
- struct kvm_pmc *pmc = &pmu->gp_counters[i];
- stop_counter(pmc);
- pmc->counter = pmc->eventsel = 0;
- }
-
- for (i = 0; i < INTEL_PMC_MAX_FIXED; i++)
- stop_counter(&pmu->fixed_counters[i]);
-
- pmu->fixed_ctr_ctrl = pmu->global_ctrl = pmu->global_status =
- pmu->global_ovf_ctrl = 0;
-}
-
-void kvm_pmu_destroy(struct kvm_vcpu *vcpu)
-{
- kvm_pmu_reset(vcpu);
-}
-
-void kvm_handle_pmu_event(struct kvm_vcpu *vcpu)
-{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- u64 bitmask;
- int bit;
-
- bitmask = pmu->reprogram_pmi;
-
- for_each_set_bit(bit, (unsigned long *)&bitmask, X86_PMC_IDX_MAX) {
- struct kvm_pmc *pmc = global_idx_to_pmc(pmu, bit);
-
- if (unlikely(!pmc || !pmc->perf_event)) {
- clear_bit(bit, (unsigned long *)&pmu->reprogram_pmi);
- continue;
- }
-
- reprogram_idx(pmu, bit);
- }
-}
new file mode 100644
@@ -0,0 +1,86 @@
+/*
+ * KVM -- Performance Monitoring Unit support for AMD
+ *
+ * Copyright 2014, Red Hat, Inc. and/or its affiliates.
+ *
+ * Author:
+ * Wei Huang <wei@redhat.com>
+ *
+ * Implementation is based on pmu_intel.c file
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2. See
+ * the COPYING file in the top-level directory.
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/kvm_host.h>
+#include <linux/perf_event.h>
+#include <asm/perf_event.h>
+#include "x86.h"
+#include "cpuid.h"
+#include "lapic.h"
+
+void amd_pmu_cpuid_update(struct kvm_vcpu *vcpu)
+{
+}
+
+int amd_pmu_check_pmc(struct kvm_vcpu *vcpu, unsigned pmc)
+{
+ return 1;
+}
+
+int amd_pmu_read_pmc(struct kvm_vcpu *vcpu, unsigned pmc, u64 *data)
+{
+ return 1;
+}
+
+int amd_pmu_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
+{
+ return 1;
+}
+
+int amd_pmu_get_msr(struct kvm_vcpu *vcpu, u32 index, u64 *data)
+{
+ return 1;
+}
+
+bool amd_is_pmu_msr(struct kvm_vcpu *vcpu, u32 msr)
+{
+ return 0;
+}
+
+void amd_deliver_pmi(struct kvm_vcpu *vcpu)
+{
+}
+
+void amd_handle_pmu_event(struct kvm_vcpu *vcpu)
+{
+}
+
+void amd_pmu_reset(struct kvm_vcpu *vcpu)
+{
+}
+
+void amd_pmu_init(struct kvm_vcpu *vcpu)
+{
+}
+
+void amd_pmu_destroy(struct kvm_vcpu *vcpu)
+{
+}
+
+struct kvm_pmu_ops amd_pmu_ops = {
+ .pmu_cpuid_update = amd_pmu_cpuid_update,
+ .pmu_check_pmc = amd_pmu_check_pmc,
+ .pmu_read_pmc = amd_pmu_read_pmc,
+ .pmu_set_msr = amd_pmu_set_msr,
+ .pmu_get_msr = amd_pmu_get_msr,
+ .is_pmu_msr = amd_is_pmu_msr,
+ .pmu_deliver_pmi = amd_deliver_pmi,
+ .pmu_handle_event = amd_handle_pmu_event,
+ .pmu_reset = amd_pmu_reset,
+ .pmu_init = amd_pmu_init,
+ .pmu_destroy = amd_pmu_destroy,
+};
+EXPORT_SYMBOL_GPL(amd_pmu_ops);
new file mode 100644
@@ -0,0 +1,586 @@
+/*
+ * KVM -- Performance Monitoring Unit support for Intel
+ *
+ * Copyright 2011 - 2014, Red Hat, Inc. and/or its affiliates.
+ *
+ * Authors:
+ * Avi Kivity <avi@redhat.com>
+ * Gleb Natapov <gleb@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2. See
+ * the COPYING file in the top-level directory.
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/kvm_host.h>
+#include <linux/perf_event.h>
+#include <asm/perf_event.h>
+#include "x86.h"
+#include "cpuid.h"
+#include "lapic.h"
+
+struct kvm_event_hw_type_mapping intel_arch_events[] = {
+ /* Index must match CPUID 0x0A.EBX bit vector */
+ [0] = { 0x3c, 0x00, PERF_COUNT_HW_CPU_CYCLES },
+ [1] = { 0xc0, 0x00, PERF_COUNT_HW_INSTRUCTIONS },
+ [2] = { 0x3c, 0x01, PERF_COUNT_HW_BUS_CYCLES },
+ [3] = { 0x2e, 0x4f, PERF_COUNT_HW_CACHE_REFERENCES },
+ [4] = { 0x2e, 0x41, PERF_COUNT_HW_CACHE_MISSES },
+ [5] = { 0xc4, 0x00, PERF_COUNT_HW_BRANCH_INSTRUCTIONS },
+ [6] = { 0xc5, 0x00, PERF_COUNT_HW_BRANCH_MISSES },
+ [7] = { 0x00, 0x30, PERF_COUNT_HW_REF_CPU_CYCLES },
+};
+
+/* mapping between fixed pmc index and intel_arch_events array */
+int fixed_pmc_events[] = {1, 0, 7};
+
+static bool pmc_is_gp(struct kvm_pmc *pmc)
+{
+ return pmc->type == KVM_PMC_GP;
+}
+
+static inline u64 pmc_bitmask(struct kvm_pmc *pmc)
+{
+ struct kvm_pmu *pmu = &pmc->vcpu->arch.pmu;
+
+ return pmu->counter_bitmask[pmc->type];
+}
+
+static inline bool pmc_enabled(struct kvm_pmc *pmc)
+{
+ struct kvm_pmu *pmu = &pmc->vcpu->arch.pmu;
+ return test_bit(pmc->idx, (unsigned long *)&pmu->global_ctrl);
+}
+
+static inline struct kvm_pmc *get_gp_pmc(struct kvm_pmu *pmu, u32 msr,
+ u32 base)
+{
+ if (msr >= base && msr < base + pmu->nr_arch_gp_counters)
+ return &pmu->gp_counters[msr - base];
+ return NULL;
+}
+
+static inline struct kvm_pmc *get_fixed_pmc(struct kvm_pmu *pmu, u32 msr)
+{
+ int base = MSR_CORE_PERF_FIXED_CTR0;
+ if (msr >= base && msr < base + pmu->nr_arch_fixed_counters)
+ return &pmu->fixed_counters[msr - base];
+ return NULL;
+}
+
+static inline struct kvm_pmc *get_fixed_pmc_idx(struct kvm_pmu *pmu, int idx)
+{
+ return get_fixed_pmc(pmu, MSR_CORE_PERF_FIXED_CTR0 + idx);
+}
+
+static struct kvm_pmc *global_idx_to_pmc(struct kvm_pmu *pmu, int idx)
+{
+ if (idx < INTEL_PMC_IDX_FIXED)
+ return get_gp_pmc(pmu, MSR_P6_EVNTSEL0 + idx, MSR_P6_EVNTSEL0);
+ else
+ return get_fixed_pmc_idx(pmu, idx - INTEL_PMC_IDX_FIXED);
+}
+
+void intel_deliver_pmi(struct kvm_vcpu *vcpu)
+{
+ if (vcpu->arch.apic)
+ kvm_apic_local_deliver(vcpu->arch.apic, APIC_LVTPC);
+}
+
+static void trigger_pmi(struct irq_work *irq_work)
+{
+ struct kvm_pmu *pmu = container_of(irq_work, struct kvm_pmu,
+ irq_work);
+ struct kvm_vcpu *vcpu = container_of(pmu, struct kvm_vcpu,
+ arch.pmu);
+
+ intel_deliver_pmi(vcpu);
+}
+
+static void kvm_perf_overflow(struct perf_event *perf_event,
+ struct perf_sample_data *data,
+ struct pt_regs *regs)
+{
+ struct kvm_pmc *pmc = perf_event->overflow_handler_context;
+ struct kvm_pmu *pmu = &pmc->vcpu->arch.pmu;
+ if (!test_and_set_bit(pmc->idx, (unsigned long *)&pmu->reprogram_pmi)) {
+ __set_bit(pmc->idx, (unsigned long *)&pmu->global_status);
+ kvm_make_request(KVM_REQ_PMU, pmc->vcpu);
+ }
+}
+
+static void kvm_perf_overflow_intr(struct perf_event *perf_event,
+ struct perf_sample_data *data, struct pt_regs *regs)
+{
+ struct kvm_pmc *pmc = perf_event->overflow_handler_context;
+ struct kvm_pmu *pmu = &pmc->vcpu->arch.pmu;
+ if (!test_and_set_bit(pmc->idx, (unsigned long *)&pmu->reprogram_pmi)) {
+ __set_bit(pmc->idx, (unsigned long *)&pmu->global_status);
+ kvm_make_request(KVM_REQ_PMU, pmc->vcpu);
+ /*
+ * Inject PMI. If vcpu was in a guest mode during NMI PMI
+ * can be ejected on a guest mode re-entry. Otherwise we can't
+ * be sure that vcpu wasn't executing hlt instruction at the
+ * time of vmexit and is not going to re-enter guest mode until,
+ * woken up. So we should wake it, but this is impossible from
+ * NMI context. Do it from irq work instead.
+ */
+ if (!kvm_is_in_guest())
+ irq_work_queue(&pmc->vcpu->arch.pmu.irq_work);
+ else
+ kvm_make_request(KVM_REQ_PMI, pmc->vcpu);
+ }
+}
+
+static u64 read_pmc(struct kvm_pmc *pmc)
+{
+ u64 counter, enabled, running;
+
+ counter = pmc->counter;
+
+ if (pmc->perf_event)
+ counter += perf_event_read_value(pmc->perf_event,
+ &enabled, &running);
+
+ /* FIXME: Scaling needed? */
+
+ return counter & pmc_bitmask(pmc);
+}
+
+static void stop_counter(struct kvm_pmc *pmc)
+{
+ if (pmc->perf_event) {
+ pmc->counter = read_pmc(pmc);
+ perf_event_release_kernel(pmc->perf_event);
+ pmc->perf_event = NULL;
+ }
+}
+
+static void reprogram_counter(struct kvm_pmc *pmc, u32 type,
+ unsigned config, bool exclude_user, bool exclude_kernel,
+ bool intr, bool in_tx, bool in_tx_cp)
+{
+ struct perf_event *event;
+ struct perf_event_attr attr = {
+ .type = type,
+ .size = sizeof(attr),
+ .pinned = true,
+ .exclude_idle = true,
+ .exclude_host = 1,
+ .exclude_user = exclude_user,
+ .exclude_kernel = exclude_kernel,
+ .config = config,
+ };
+ if (in_tx)
+ attr.config |= HSW_IN_TX;
+ if (in_tx_cp)
+ attr.config |= HSW_IN_TX_CHECKPOINTED;
+
+ attr.sample_period = (-pmc->counter) & pmc_bitmask(pmc);
+
+ event = perf_event_create_kernel_counter(&attr, -1, current,
+ intr ? kvm_perf_overflow_intr :
+ kvm_perf_overflow, pmc);
+ if (IS_ERR(event)) {
+ printk_once("kvm: pmu event creation failed %ld\n",
+ PTR_ERR(event));
+ return;
+ }
+
+ pmc->perf_event = event;
+ clear_bit(pmc->idx, (unsigned long*)&pmc->vcpu->arch.pmu.reprogram_pmi);
+}
+
+static unsigned find_arch_event(struct kvm_pmu *pmu, u8 event_select,
+ u8 unit_mask)
+{
+ int i;
+
+ for (i = 0; i < ARRAY_SIZE(intel_arch_events); i++)
+ if (intel_arch_events[i].eventsel == event_select
+ && intel_arch_events[i].unit_mask == unit_mask
+ && (pmu->available_event_types & (1 << i)))
+ break;
+
+ if (i == ARRAY_SIZE(intel_arch_events))
+ return PERF_COUNT_HW_MAX;
+
+ return intel_arch_events[i].event_type;
+}
+
+static void reprogram_gp_counter(struct kvm_pmc *pmc, u64 eventsel)
+{
+ unsigned config, type = PERF_TYPE_RAW;
+ u8 event_select, unit_mask;
+
+ if (eventsel & ARCH_PERFMON_EVENTSEL_PIN_CONTROL)
+ printk_once("kvm pmu: pin control bit is ignored\n");
+
+ pmc->eventsel = eventsel;
+
+ stop_counter(pmc);
+
+ if (!(eventsel & ARCH_PERFMON_EVENTSEL_ENABLE) || !pmc_enabled(pmc))
+ return;
+
+ event_select = eventsel & ARCH_PERFMON_EVENTSEL_EVENT;
+ unit_mask = (eventsel & ARCH_PERFMON_EVENTSEL_UMASK) >> 8;
+
+ if (!(eventsel & (ARCH_PERFMON_EVENTSEL_EDGE |
+ ARCH_PERFMON_EVENTSEL_INV |
+ ARCH_PERFMON_EVENTSEL_CMASK |
+ HSW_IN_TX |
+ HSW_IN_TX_CHECKPOINTED))) {
+ config = find_arch_event(&pmc->vcpu->arch.pmu, event_select,
+ unit_mask);
+ if (config != PERF_COUNT_HW_MAX)
+ type = PERF_TYPE_HARDWARE;
+ }
+
+ if (type == PERF_TYPE_RAW)
+ config = eventsel & X86_RAW_EVENT_MASK;
+
+ reprogram_counter(pmc, type, config,
+ !(eventsel & ARCH_PERFMON_EVENTSEL_USR),
+ !(eventsel & ARCH_PERFMON_EVENTSEL_OS),
+ eventsel & ARCH_PERFMON_EVENTSEL_INT,
+ (eventsel & HSW_IN_TX),
+ (eventsel & HSW_IN_TX_CHECKPOINTED));
+}
+
+static void reprogram_fixed_counter(struct kvm_pmc *pmc, u8 en_pmi, int idx)
+{
+ unsigned en = en_pmi & 0x3;
+ bool pmi = en_pmi & 0x8;
+
+ stop_counter(pmc);
+
+ if (!en || !pmc_enabled(pmc))
+ return;
+
+ reprogram_counter(pmc, PERF_TYPE_HARDWARE,
+ intel_arch_events[fixed_pmc_events[idx]].event_type,
+ !(en & 0x2), /* exclude user */
+ !(en & 0x1), /* exclude kernel */
+ pmi, false, false);
+}
+
+static inline u8 fixed_en_pmi(u64 ctrl, int idx)
+{
+ return (ctrl >> (idx * 4)) & 0xf;
+}
+
+static void reprogram_fixed_counters(struct kvm_pmu *pmu, u64 data)
+{
+ int i;
+
+ for (i = 0; i < pmu->nr_arch_fixed_counters; i++) {
+ u8 en_pmi = fixed_en_pmi(data, i);
+ struct kvm_pmc *pmc = get_fixed_pmc_idx(pmu, i);
+
+ if (fixed_en_pmi(pmu->fixed_ctr_ctrl, i) == en_pmi)
+ continue;
+
+ reprogram_fixed_counter(pmc, en_pmi, i);
+ }
+
+ pmu->fixed_ctr_ctrl = data;
+}
+
+static void reprogram_idx(struct kvm_pmu *pmu, int idx)
+{
+ struct kvm_pmc *pmc = global_idx_to_pmc(pmu, idx);
+
+ if (!pmc)
+ return;
+
+ if (pmc_is_gp(pmc))
+ reprogram_gp_counter(pmc, pmc->eventsel);
+ else {
+ int fidx = idx - INTEL_PMC_IDX_FIXED;
+ reprogram_fixed_counter(pmc,
+ fixed_en_pmi(pmu->fixed_ctr_ctrl, fidx), fidx);
+ }
+}
+
+static void global_ctrl_changed(struct kvm_pmu *pmu, u64 data)
+{
+ int bit;
+ u64 diff = pmu->global_ctrl ^ data;
+
+ pmu->global_ctrl = data;
+
+ for_each_set_bit(bit, (unsigned long *)&diff, X86_PMC_IDX_MAX)
+ reprogram_idx(pmu, bit);
+}
+
+bool intel_is_pmu_msr(struct kvm_vcpu *vcpu, u32 msr)
+{
+ struct kvm_pmu *pmu = &vcpu->arch.pmu;
+ int ret;
+
+ switch (msr) {
+ case MSR_CORE_PERF_FIXED_CTR_CTRL:
+ case MSR_CORE_PERF_GLOBAL_STATUS:
+ case MSR_CORE_PERF_GLOBAL_CTRL:
+ case MSR_CORE_PERF_GLOBAL_OVF_CTRL:
+ ret = pmu->version > 1;
+ break;
+ default:
+ ret = get_gp_pmc(pmu, msr, MSR_IA32_PERFCTR0)
+ || get_gp_pmc(pmu, msr, MSR_P6_EVNTSEL0)
+ || get_fixed_pmc(pmu, msr);
+ break;
+ }
+ return ret;
+}
+
+int intel_pmu_get_msr(struct kvm_vcpu *vcpu, u32 index, u64 *data)
+{
+ struct kvm_pmu *pmu = &vcpu->arch.pmu;
+ struct kvm_pmc *pmc;
+
+ switch (index) {
+ case MSR_CORE_PERF_FIXED_CTR_CTRL:
+ *data = pmu->fixed_ctr_ctrl;
+ return 0;
+ case MSR_CORE_PERF_GLOBAL_STATUS:
+ *data = pmu->global_status;
+ return 0;
+ case MSR_CORE_PERF_GLOBAL_CTRL:
+ *data = pmu->global_ctrl;
+ return 0;
+ case MSR_CORE_PERF_GLOBAL_OVF_CTRL:
+ *data = pmu->global_ovf_ctrl;
+ return 0;
+ default:
+ if ((pmc = get_gp_pmc(pmu, index, MSR_IA32_PERFCTR0)) ||
+ (pmc = get_fixed_pmc(pmu, index))) {
+ *data = read_pmc(pmc);
+ return 0;
+ } else if ((pmc = get_gp_pmc(pmu, index, MSR_P6_EVNTSEL0))) {
+ *data = pmc->eventsel;
+ return 0;
+ }
+ }
+ return 1;
+}
+
+int intel_pmu_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
+{
+ struct kvm_pmu *pmu = &vcpu->arch.pmu;
+ struct kvm_pmc *pmc;
+ u32 index = msr_info->index;
+ u64 data = msr_info->data;
+
+ switch (index) {
+ case MSR_CORE_PERF_FIXED_CTR_CTRL:
+ if (pmu->fixed_ctr_ctrl == data)
+ return 0;
+ if (!(data & 0xfffffffffffff444ull)) {
+ reprogram_fixed_counters(pmu, data);
+ return 0;
+ }
+ break;
+ case MSR_CORE_PERF_GLOBAL_STATUS:
+ if (msr_info->host_initiated) {
+ pmu->global_status = data;
+ return 0;
+ }
+ break; /* RO MSR */
+ case MSR_CORE_PERF_GLOBAL_CTRL:
+ if (pmu->global_ctrl == data)
+ return 0;
+ if (!(data & pmu->global_ctrl_mask)) {
+ global_ctrl_changed(pmu, data);
+ return 0;
+ }
+ break;
+ case MSR_CORE_PERF_GLOBAL_OVF_CTRL:
+ if (!(data & (pmu->global_ctrl_mask & ~(3ull<<62)))) {
+ if (!msr_info->host_initiated)
+ pmu->global_status &= ~data;
+ pmu->global_ovf_ctrl = data;
+ return 0;
+ }
+ break;
+ default:
+ if ((pmc = get_gp_pmc(pmu, index, MSR_IA32_PERFCTR0)) ||
+ (pmc = get_fixed_pmc(pmu, index))) {
+ if (!msr_info->host_initiated)
+ data = (s64)(s32)data;
+ pmc->counter += data - read_pmc(pmc);
+ return 0;
+ } else if ((pmc = get_gp_pmc(pmu, index, MSR_P6_EVNTSEL0))) {
+ if (data == pmc->eventsel)
+ return 0;
+ if (!(data & pmu->reserved_bits)) {
+ reprogram_gp_counter(pmc, data);
+ return 0;
+ }
+ }
+ }
+ return 1;
+}
+
+int intel_pmu_check_pmc(struct kvm_vcpu *vcpu, unsigned pmc)
+{
+ struct kvm_pmu *pmu = &vcpu->arch.pmu;
+ bool fixed = pmc & (1u << 30);
+ pmc &= ~(3u << 30);
+ return (!fixed && pmc >= pmu->nr_arch_gp_counters) ||
+ (fixed && pmc >= pmu->nr_arch_fixed_counters);
+}
+
+int intel_pmu_read_pmc(struct kvm_vcpu *vcpu, unsigned pmc, u64 *data)
+{
+ struct kvm_pmu *pmu = &vcpu->arch.pmu;
+ bool fast_mode = pmc & (1u << 31);
+ bool fixed = pmc & (1u << 30);
+ struct kvm_pmc *counters;
+ u64 ctr;
+
+ pmc &= ~(3u << 30);
+ if (!fixed && pmc >= pmu->nr_arch_gp_counters)
+ return 1;
+ if (fixed && pmc >= pmu->nr_arch_fixed_counters)
+ return 1;
+ counters = fixed ? pmu->fixed_counters : pmu->gp_counters;
+ ctr = read_pmc(&counters[pmc]);
+ if (fast_mode)
+ ctr = (u32)ctr;
+ *data = ctr;
+
+ return 0;
+}
+
+void intel_pmu_cpuid_update(struct kvm_vcpu *vcpu)
+{
+ struct kvm_pmu *pmu = &vcpu->arch.pmu;
+ struct kvm_cpuid_entry2 *entry;
+ union cpuid10_eax eax;
+ union cpuid10_edx edx;
+
+ pmu->nr_arch_gp_counters = 0;
+ pmu->nr_arch_fixed_counters = 0;
+ pmu->counter_bitmask[KVM_PMC_GP] = 0;
+ pmu->counter_bitmask[KVM_PMC_FIXED] = 0;
+ pmu->version = 0;
+ pmu->reserved_bits = 0xffffffff00200000ull;
+
+ entry = kvm_find_cpuid_entry(vcpu, 0xa, 0);
+ if (!entry)
+ return;
+ eax.full = entry->eax;
+ edx.full = entry->edx;
+
+ pmu->version = eax.split.version_id;
+ if (!pmu->version)
+ return;
+
+ pmu->nr_arch_gp_counters = min_t(int, eax.split.num_counters,
+ INTEL_PMC_MAX_GENERIC);
+ pmu->counter_bitmask[KVM_PMC_GP] = ((u64)1 << eax.split.bit_width) - 1;
+ pmu->available_event_types = ~entry->ebx &
+ ((1ull << eax.split.mask_length) - 1);
+
+ if (pmu->version == 1) {
+ pmu->nr_arch_fixed_counters = 0;
+ } else {
+ pmu->nr_arch_fixed_counters =
+ min_t(int, edx.split.num_counters_fixed,
+ INTEL_PMC_MAX_FIXED);
+ pmu->counter_bitmask[KVM_PMC_FIXED] =
+ ((u64)1 << edx.split.bit_width_fixed) - 1;
+ }
+
+ pmu->global_ctrl = ((1 << pmu->nr_arch_gp_counters) - 1) |
+ (((1ull << pmu->nr_arch_fixed_counters) - 1) << INTEL_PMC_IDX_FIXED);
+ pmu->global_ctrl_mask = ~pmu->global_ctrl;
+
+ entry = kvm_find_cpuid_entry(vcpu, 7, 0);
+ if (entry &&
+ (boot_cpu_has(X86_FEATURE_HLE) || boot_cpu_has(X86_FEATURE_RTM)) &&
+ (entry->ebx & (X86_FEATURE_HLE|X86_FEATURE_RTM)))
+ pmu->reserved_bits ^= HSW_IN_TX|HSW_IN_TX_CHECKPOINTED;
+}
+
+void intel_pmu_init(struct kvm_vcpu *vcpu)
+{
+ int i;
+ struct kvm_pmu *pmu = &vcpu->arch.pmu;
+
+ memset(pmu, 0, sizeof(*pmu));
+ for (i = 0; i < INTEL_PMC_MAX_GENERIC; i++) {
+ pmu->gp_counters[i].type = KVM_PMC_GP;
+ pmu->gp_counters[i].vcpu = vcpu;
+ pmu->gp_counters[i].idx = i;
+ }
+ for (i = 0; i < INTEL_PMC_MAX_FIXED; i++) {
+ pmu->fixed_counters[i].type = KVM_PMC_FIXED;
+ pmu->fixed_counters[i].vcpu = vcpu;
+ pmu->fixed_counters[i].idx = i + INTEL_PMC_IDX_FIXED;
+ }
+ init_irq_work(&pmu->irq_work, trigger_pmi);
+ intel_pmu_cpuid_update(vcpu);
+}
+
+void intel_pmu_reset(struct kvm_vcpu *vcpu)
+{
+ struct kvm_pmu *pmu = &vcpu->arch.pmu;
+ int i;
+
+ irq_work_sync(&pmu->irq_work);
+ for (i = 0; i < INTEL_PMC_MAX_GENERIC; i++) {
+ struct kvm_pmc *pmc = &pmu->gp_counters[i];
+ stop_counter(pmc);
+ pmc->counter = pmc->eventsel = 0;
+ }
+
+ for (i = 0; i < INTEL_PMC_MAX_FIXED; i++)
+ stop_counter(&pmu->fixed_counters[i]);
+
+ pmu->fixed_ctr_ctrl = pmu->global_ctrl = pmu->global_status =
+ pmu->global_ovf_ctrl = 0;
+}
+
+void intel_pmu_destroy(struct kvm_vcpu *vcpu)
+{
+ intel_pmu_reset(vcpu);
+}
+
+void intel_handle_pmu_event(struct kvm_vcpu *vcpu)
+{
+ struct kvm_pmu *pmu = &vcpu->arch.pmu;
+ u64 bitmask;
+ int bit;
+
+ bitmask = pmu->reprogram_pmi;
+
+ for_each_set_bit(bit, (unsigned long *)&bitmask, X86_PMC_IDX_MAX) {
+ struct kvm_pmc *pmc = global_idx_to_pmc(pmu, bit);
+
+ if (unlikely(!pmc || !pmc->perf_event)) {
+ clear_bit(bit, (unsigned long *)&pmu->reprogram_pmi);
+ continue;
+ }
+
+ reprogram_idx(pmu, bit);
+ }
+}
+
+struct kvm_pmu_ops intel_pmu_ops = {
+ .pmu_cpuid_update = intel_pmu_cpuid_update,
+ .pmu_check_pmc = intel_pmu_check_pmc,
+ .pmu_read_pmc = intel_pmu_read_pmc,
+ .pmu_set_msr = intel_pmu_set_msr,
+ .pmu_get_msr = intel_pmu_get_msr,
+ .is_pmu_msr = intel_is_pmu_msr,
+ .pmu_deliver_pmi = intel_deliver_pmi,
+ .pmu_handle_event = intel_handle_pmu_event,
+ .pmu_reset = intel_pmu_reset,
+ .pmu_init = intel_pmu_init,
+ .pmu_destroy = intel_pmu_destroy,
+};
+EXPORT_SYMBOL_GPL(intel_pmu_ops);
@@ -4326,6 +4326,11 @@ static void svm_sched_in(struct kvm_vcpu *vcpu, int cpu)
{
}
+struct kvm_pmu_ops *svm_get_pmu_ops(void)
+{
+ return &amd_pmu_ops;
+}
+
static struct kvm_x86_ops svm_x86_ops = {
.cpu_has_kvm_support = has_svm,
.disabled_by_bios = is_disabled,
@@ -4428,6 +4433,8 @@ static struct kvm_x86_ops svm_x86_ops = {
.handle_external_intr = svm_handle_external_intr,
.sched_in = svm_sched_in,
+
+ .get_pmu_ops = svm_get_pmu_ops,
};
static int __init svm_init(void)
@@ -9073,6 +9073,11 @@ static void vmx_sched_in(struct kvm_vcpu *vcpu, int cpu)
shrink_ple_window(vcpu);
}
+static struct kvm_pmu_ops *vmx_get_pmu_ops(void)
+{
+ return &intel_pmu_ops;
+}
+
static struct kvm_x86_ops vmx_x86_ops = {
.cpu_has_kvm_support = cpu_has_kvm_support,
.disabled_by_bios = vmx_disabled_by_bios,
@@ -9180,6 +9185,8 @@ static struct kvm_x86_ops vmx_x86_ops = {
.check_nested_events = vmx_check_nested_events,
.sched_in = vmx_sched_in,
+
+ .get_pmu_ops = vmx_get_pmu_ops,
};
static int __init vmx_init(void)
@@ -91,6 +91,7 @@ static void __kvm_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags);
struct kvm_x86_ops *kvm_x86_ops;
EXPORT_SYMBOL_GPL(kvm_x86_ops);
+struct kvm_pmu_ops *kvm_pmu_ops;
static bool ignore_msrs = 0;
module_param(ignore_msrs, bool, S_IRUGO | S_IWUSR);
@@ -893,7 +894,7 @@ bool kvm_rdpmc(struct kvm_vcpu *vcpu)
u64 data;
int err;
- err = kvm_pmu_read_pmc(vcpu, ecx, &data);
+ err = kvm_pmu_ops->pmu_read_pmc(vcpu, ecx, &data);
if (err)
return err;
kvm_register_write(vcpu, VCPU_REGS_RAX, (u32)data);
@@ -2248,8 +2249,8 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
pr = true;
case MSR_P6_EVNTSEL0:
case MSR_P6_EVNTSEL1:
- if (kvm_pmu_msr(vcpu, msr))
- return kvm_pmu_set_msr(vcpu, msr_info);
+ if (kvm_pmu_ops->is_pmu_msr(vcpu, msr))
+ return kvm_pmu_ops->pmu_set_msr(vcpu, msr_info);
if (pr || data != 0)
vcpu_unimpl(vcpu, "disabled perfctr wrmsr: "
@@ -2294,8 +2295,8 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
default:
if (msr && (msr == vcpu->kvm->arch.xen_hvm_config.msr))
return xen_hvm_config(vcpu, data);
- if (kvm_pmu_msr(vcpu, msr))
- return kvm_pmu_set_msr(vcpu, msr_info);
+ if (kvm_pmu_ops->is_pmu_msr(vcpu, msr))
+ return kvm_pmu_ops->pmu_set_msr(vcpu, msr_info);
if (!ignore_msrs) {
vcpu_unimpl(vcpu, "unhandled wrmsr: 0x%x data %llx\n",
msr, data);
@@ -2487,8 +2488,8 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata)
case MSR_P6_PERFCTR1:
case MSR_P6_EVNTSEL0:
case MSR_P6_EVNTSEL1:
- if (kvm_pmu_msr(vcpu, msr))
- return kvm_pmu_get_msr(vcpu, msr, pdata);
+ if (kvm_pmu_ops->is_pmu_msr(vcpu, msr))
+ return kvm_pmu_ops->pmu_get_msr(vcpu, msr, pdata);
data = 0;
break;
case MSR_IA32_UCODE_REV:
@@ -2610,8 +2611,8 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata)
data = vcpu->arch.osvw.status;
break;
default:
- if (kvm_pmu_msr(vcpu, msr))
- return kvm_pmu_get_msr(vcpu, msr, pdata);
+ if (kvm_pmu_ops->is_pmu_msr(vcpu, msr))
+ return kvm_pmu_ops->pmu_get_msr(vcpu, msr, pdata);
if (!ignore_msrs) {
vcpu_unimpl(vcpu, "unhandled rdmsr: 0x%x\n", msr);
return 1;
@@ -4848,19 +4849,19 @@ static int emulator_set_msr(struct x86_emulate_ctxt *ctxt,
msr.data = data;
msr.index = msr_index;
msr.host_initiated = false;
- return kvm_set_msr(emul_to_vcpu(ctxt), &msr);
+ return kvm_pmu_ops->pmu_set_msr(emul_to_vcpu(ctxt), &msr);
}
static int emulator_check_pmc(struct x86_emulate_ctxt *ctxt,
u32 pmc)
{
- return kvm_pmu_check_pmc(emul_to_vcpu(ctxt), pmc);
+ return kvm_pmu_ops->pmu_check_pmc(emul_to_vcpu(ctxt), pmc);
}
static int emulator_read_pmc(struct x86_emulate_ctxt *ctxt,
u32 pmc, u64 *pdata)
{
- return kvm_pmu_read_pmc(emul_to_vcpu(ctxt), pmc, pdata);
+ return kvm_pmu_ops->pmu_read_pmc(emul_to_vcpu(ctxt), pmc, pdata);
}
static void emulator_halt(struct x86_emulate_ctxt *ctxt)
@@ -5557,6 +5558,7 @@ int kvm_is_in_guest(void)
{
return __this_cpu_read(current_vcpu) != NULL;
}
+EXPORT_SYMBOL_GPL(kvm_is_in_guest);
static int kvm_is_user_mode(void)
{
@@ -5713,6 +5715,9 @@ int kvm_arch_init(void *opaque)
kvm_timer_init();
+ /* init pointers of PMU operations */
+ kvm_pmu_ops = ops->get_pmu_ops();
+
perf_register_guest_info_callbacks(&kvm_guest_cbs);
if (cpu_has_xsave)
@@ -5743,6 +5748,7 @@ void kvm_arch_exit(void)
pvclock_gtod_unregister_notifier(&pvclock_gtod_notifier);
#endif
kvm_x86_ops = NULL;
+ kvm_pmu_ops = NULL;
kvm_mmu_module_exit();
free_percpu(shared_msrs);
}
@@ -6143,9 +6149,9 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
if (kvm_check_request(KVM_REQ_NMI, vcpu))
process_nmi(vcpu);
if (kvm_check_request(KVM_REQ_PMU, vcpu))
- kvm_handle_pmu_event(vcpu);
+ kvm_pmu_ops->pmu_handle_event(vcpu);
if (kvm_check_request(KVM_REQ_PMI, vcpu))
- kvm_deliver_pmi(vcpu);
+ kvm_pmu_ops->pmu_deliver_pmi(vcpu);
if (kvm_check_request(KVM_REQ_SCAN_IOAPIC, vcpu))
vcpu_scan_ioapic(vcpu);
if (kvm_check_request(KVM_REQ_APIC_PAGE_RELOAD, vcpu))
@@ -7015,7 +7021,7 @@ void kvm_vcpu_reset(struct kvm_vcpu *vcpu)
kvm_async_pf_hash_reset(vcpu);
vcpu->arch.apf.halted = false;
- kvm_pmu_reset(vcpu);
+ kvm_pmu_ops->pmu_reset(vcpu);
memset(vcpu->arch.regs, 0, sizeof(vcpu->arch.regs));
vcpu->arch.regs_avail = ~0;
@@ -7214,7 +7220,7 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
vcpu->arch.guest_xstate_size = XSAVE_HDR_SIZE + XSAVE_HDR_OFFSET;
kvm_async_pf_hash_reset(vcpu);
- kvm_pmu_init(vcpu);
+ kvm_pmu_ops->pmu_init(vcpu);
return 0;
fail_free_wbinvd_dirty_mask:
@@ -7235,7 +7241,7 @@ void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
{
int idx;
- kvm_pmu_destroy(vcpu);
+ kvm_pmu_ops->pmu_destroy(vcpu);
kfree(vcpu->arch.mce_banks);
kvm_free_lapic(vcpu);
idx = srcu_read_lock(&vcpu->kvm->srcu);
This patch converts existing pmu.c into Intel specific code and hooks up with the PMU interface using the following steps: - Convert pmu.c to pmu_intel.c; All public PMU functions are renamed and hooked up with the newly defined intel_pmu_ops. - Create a corresponding pmu_amd.c file with empty functions for AMD arch. - The PMU function pointer, kvm_pmu_ops, is initialized by calling kvm_x86_ops->get_pmu_ops(). - To reduce the code size, Intel and AMD modules are now generated from their corrponding arch and PMU files; In the meanwhile, due to this arrangement several, functions are exposed as public ones to allow calling from PMU code. Signed-off-by: Wei Huang <wei@redhat.com> --- arch/x86/include/asm/kvm_host.h | 17 +- arch/x86/kvm/Makefile | 6 +- arch/x86/kvm/cpuid.c | 2 +- arch/x86/kvm/lapic.c | 1 + arch/x86/kvm/pmu.c | 571 --------------------------------------- arch/x86/kvm/pmu_amd.c | 86 ++++++ arch/x86/kvm/pmu_intel.c | 586 ++++++++++++++++++++++++++++++++++++++++ arch/x86/kvm/svm.c | 7 + arch/x86/kvm/vmx.c | 7 + arch/x86/kvm/x86.c | 40 +-- 10 files changed, 719 insertions(+), 604 deletions(-) delete mode 100644 arch/x86/kvm/pmu.c create mode 100644 arch/x86/kvm/pmu_amd.c create mode 100644 arch/x86/kvm/pmu_intel.c