@@ -855,6 +855,8 @@ struct kvm_x86_ops {
void (*enable_log_dirty_pt_masked)(struct kvm *kvm,
struct kvm_memory_slot *slot,
gfn_t offset, unsigned long mask);
+
+ struct kvm_pmu_ops *(*get_pmu_ops)(void);
};
struct kvm_arch_async_pf {
@@ -865,6 +867,7 @@ 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)
@@ -14,8 +14,8 @@ kvm-$(CONFIG_KVM_ASYNC_PF) += $(KVM)/async_pf.o
kvm-y += x86.o mmu.o emulate.o i8259.o irq.o lapic.o \
i8254.o ioapic.o irq_comm.o cpuid.o pmu.o
kvm-$(CONFIG_KVM_DEVICE_ASSIGNMENT) += assigned-dev.o iommu.o
-kvm-intel-y += vmx.o
-kvm-amd-y += svm.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
@@ -22,6 +22,7 @@
#include "lapic.h"
#include "mmu.h"
#include "trace.h"
+#include "pmu.h"
static u32 xstate_required_size(u64 xstate_bv, bool compacted)
{
@@ -107,7 +108,7 @@ int kvm_update_cpuid(struct kvm_vcpu *vcpu)
/* Update physical-address width */
vcpu->arch.maxphyaddr = cpuid_query_maxphyaddr(vcpu);
- kvm_pmu_cpuid_update(vcpu);
+ kvm_pmu_refresh(vcpu);
return 0;
}
@@ -1,17 +1,16 @@
/*
* Kernel-based Virtual Machine -- Performance Monitoring Unit support
*
- * Copyright 2011 Red Hat, Inc. and/or its affiliates.
+ * Copyright 2015 Red Hat, Inc. and/or its affiliates.
*
* Authors:
* Avi Kivity <avi@redhat.com>
* Gleb Natapov <gleb@redhat.com>
+ * Wei Huang <wei@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>
@@ -19,83 +18,41 @@
#include "x86.h"
#include "cpuid.h"
#include "lapic.h"
+#include "pmu.h"
+
+/* NOTE:
+ * - Each perf counter is defined as "struct kvm_pmc";
+ * - There are two types of perf counters: general purpose (gp) and fixed.
+ * gp counters are stored in gp_counters[] and fixed counters are stored
+ * in fixed_counters[] respectively. Both of them are part of "struct
+ * kvm_pmu";
+ * - pmu.c understands the difference between gp counters and fixed counters.
+ * However AMD doesn't support fixed-counters;
+ * - There are three types of index to access perf counters (PMC):
+ * 1. MSR (named msr): For example Intel has MSR_IA32_PERFCTRn and AMD
+ * has MSR_K7_PERFCTRn.
+ * 2. MSR Index (named idx): This normally is used by RDPMC instruction.
+ * For instance AMD RDPMC instruction uses 0000_0003h in ECX to access
+ * C001_0007h (MSR_K7_PERCTR3). Intel has a similar mechanism, except
+ * that it also supports fixed counters. idx can be used to as index to
+ * gp and fixed counters.
+ * 3. Global PMC Index (named pmc_idx): pmc_idx is an index specific to PMU
+ * code. Each pmc_idx, stored in kvm_pmc.idx field, is unique across
+ * all perf counters (both gp and fixed). The mapping relationship
+ * between pmc_idx and perf counters is as the following:
+ * * Intel: [0 .. INTEL_PMC_MAX_GENERIC-1] <=> gp counters
+ * [INTEL_PMC_IDX_FIXED .. INTEL_PMC_IDX_FIXED + 2] <=> fixed
+ * * AMD: [0 .. AMD64_NUM_COUNTERS-1] <=> gp counters
+ */
-static 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 */
-static 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);
-}
+struct kvm_pmu_ops *kvm_pmu_ops;
-static void trigger_pmi(struct irq_work *irq_work)
+static void kvm_pmi_trigger_fn(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);
+ struct kvm_pmu *pmu = container_of(irq_work, struct kvm_pmu, irq_work);
+ struct kvm_vcpu *vcpu = pmu_to_vcpu(pmu);
- kvm_deliver_pmi(vcpu);
+ kvm_pmu_deliver_pmi(vcpu);
}
static void kvm_perf_overflow(struct perf_event *perf_event,
@@ -103,26 +60,32 @@ static void kvm_perf_overflow(struct perf_event *perf_event,
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)) {
+ struct kvm_pmu *pmu = pmc_to_pmu(pmc);
+
+ 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 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)) {
+ struct kvm_pmu *pmu = pmc_to_pmu(pmc);
+
+ 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,
+ * 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.
*/
@@ -133,33 +96,10 @@ static void kvm_perf_overflow_intr(struct perf_event *perf_event,
}
}
-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)
+static void pmc_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 = {
@@ -172,6 +112,7 @@ static void reprogram_counter(struct kvm_pmc *pmc, u32 type,
.exclude_kernel = exclude_kernel,
.config = config,
};
+
if (in_tx)
attr.config |= HSW_IN_TX;
if (in_tx_cp)
@@ -183,8 +124,8 @@ static void reprogram_counter(struct kvm_pmc *pmc, u32 type,
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));
+ printk_once("kvm_pmu: event creation failed %ld\n",
+ PTR_ERR(event));
return;
}
@@ -192,24 +133,7 @@ static void reprogram_counter(struct kvm_pmc *pmc, u32 type,
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)
+void reprogram_gp_counter(struct kvm_pmc *pmc, u64 eventsel)
{
unsigned config, type = PERF_TYPE_RAW;
u8 event_select, unit_mask;
@@ -219,21 +143,22 @@ static void reprogram_gp_counter(struct kvm_pmc *pmc, u64 eventsel)
pmc->eventsel = eventsel;
- stop_counter(pmc);
+ pmc_stop_counter(pmc);
- if (!(eventsel & ARCH_PERFMON_EVENTSEL_ENABLE) || !pmc_enabled(pmc))
+ if (!(eventsel & ARCH_PERFMON_EVENTSEL_ENABLE) || !pmc_is_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);
+ ARCH_PERFMON_EVENTSEL_INV |
+ ARCH_PERFMON_EVENTSEL_CMASK |
+ HSW_IN_TX |
+ HSW_IN_TX_CHECKPOINTED))) {
+ config = kvm_pmu_ops->find_arch_event(&pmc->vcpu->arch.pmu,
+ event_select,
+ unit_mask);
if (config != PERF_COUNT_HW_MAX)
type = PERF_TYPE_HARDWARE;
}
@@ -241,56 +166,36 @@ static void reprogram_gp_counter(struct kvm_pmc *pmc, u64 eventsel)
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));
+ pmc_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));
}
+EXPORT_SYMBOL_GPL(reprogram_gp_counter);
-static void reprogram_fixed_counter(struct kvm_pmc *pmc, u8 en_pmi, int idx)
+void reprogram_fixed_counter(struct kvm_pmc *pmc, u8 ctrl, int idx)
{
- unsigned en = en_pmi & 0x3;
- bool pmi = en_pmi & 0x8;
+ unsigned en_field = ctrl & 0x3;
+ bool pmi = ctrl & 0x8;
- stop_counter(pmc);
+ pmc_stop_counter(pmc);
- if (!en || !pmc_enabled(pmc))
+ if (!en_field || !pmc_is_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);
+ pmc_reprogram_counter(pmc, PERF_TYPE_HARDWARE,
+ kvm_pmu_ops->find_fixed_event(idx),
+ !(en_field & 0x2), /* exclude user */
+ !(en_field & 0x1), /* exclude kernel */
+ pmi, false, false);
}
+EXPORT_SYMBOL_GPL(reprogram_fixed_counter);
-static inline u8 fixed_en_pmi(u64 ctrl, int idx)
+void reprogram_counter(struct kvm_pmu *pmu, int pmc_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);
+ struct kvm_pmc *pmc = kvm_pmu_ops->pmc_idx_to_pmc(pmu, pmc_idx);
if (!pmc)
return;
@@ -298,251 +203,104 @@ static void reprogram_idx(struct kvm_pmu *pmu, int idx)
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);
+ int idx = pmc_idx - INTEL_PMC_IDX_FIXED;
+ u8 ctrl = fixed_ctrl_field(pmu->fixed_ctr_ctrl, idx);
+
+ reprogram_fixed_counter(pmc, ctrl, idx);
}
}
+EXPORT_SYMBOL_GPL(reprogram_counter);
-static void global_ctrl_changed(struct kvm_pmu *pmu, u64 data)
+void kvm_pmu_handle_event(struct kvm_vcpu *vcpu)
{
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
+ u64 bitmask;
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;
-}
+ bitmask = pmu->reprogram_pmi;
-int kvm_pmu_get_msr(struct kvm_vcpu *vcpu, u32 index, u64 *data)
-{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- struct kvm_pmc *pmc;
+ for_each_set_bit(bit, (unsigned long *)&bitmask, X86_PMC_IDX_MAX) {
+ struct kvm_pmc *pmc = kvm_pmu_ops->pmc_idx_to_pmc(pmu, bit);
- 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;
+ if (unlikely(!pmc || !pmc->perf_event)) {
+ clear_bit(bit, (unsigned long *)&pmu->reprogram_pmi);
+ continue;
}
- }
- 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;
- }
- }
+ reprogram_counter(pmu, bit);
}
- return 1;
}
-int kvm_pmu_check_pmc(struct kvm_vcpu *vcpu, unsigned pmc)
+/* check if idx is a valid index to access PMU */
+int kvm_pmu_check_msr(struct kvm_vcpu *vcpu, unsigned idx)
{
- 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);
+ return kvm_pmu_ops->check_msr(vcpu, idx);
}
-int kvm_pmu_read_pmc(struct kvm_vcpu *vcpu, unsigned pmc, u64 *data)
+int kvm_pmu_rdpmc(struct kvm_vcpu *vcpu, unsigned idx, 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)
+ bool fast_mode = idx & (1u << 31);
+ struct kvm_pmc *pmc;
+ u64 ctr_val;
+
+ pmc = kvm_pmu_ops->msr_to_pmc(vcpu, idx);
+ if (!pmc)
return 1;
- counters = fixed ? pmu->fixed_counters : pmu->gp_counters;
- ctr = read_pmc(&counters[pmc]);
+
+ ctr_val = pmc_read_counter(pmc);
if (fast_mode)
- ctr = (u32)ctr;
- *data = ctr;
+ ctr_val = (u32)ctr_val;
+ *data = ctr_val;
return 0;
}
-void kvm_pmu_cpuid_update(struct kvm_vcpu *vcpu)
+void kvm_pmu_deliver_pmi(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;
- }
+ if (vcpu->arch.apic)
+ kvm_apic_local_deliver(vcpu->arch.apic, APIC_LVTPC);
+}
- 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;
+bool kvm_pmu_msr(struct kvm_vcpu *vcpu, u32 msr)
+{
+ return kvm_pmu_ops->is_pmu_msr(vcpu, msr);
+}
- 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;
+int kvm_pmu_get_msr(struct kvm_vcpu *vcpu, u32 msr, u64 *data)
+{
+ return kvm_pmu_ops->get_msr(vcpu, msr, data);
}
-void kvm_pmu_init(struct kvm_vcpu *vcpu)
+int kvm_pmu_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
{
- int i;
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
+ return kvm_pmu_ops->set_msr(vcpu, msr_info);
+}
- 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);
+/* refresh PMU settings. This function generally is called when underlying
+ * settings are changed (such as changes of PMU CPUID by guest VMs), which
+ * should rarely happen.
+ */
+void kvm_pmu_refresh(struct kvm_vcpu *vcpu)
+{
+ kvm_pmu_ops->refresh(vcpu);
}
void kvm_pmu_reset(struct kvm_vcpu *vcpu)
{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- int i;
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
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;
- }
+ kvm_pmu_ops->reset(vcpu);
+}
- for (i = 0; i < INTEL_PMC_MAX_FIXED; i++)
- stop_counter(&pmu->fixed_counters[i]);
+void kvm_pmu_init(struct kvm_vcpu *vcpu)
+{
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
- pmu->fixed_ctr_ctrl = pmu->global_ctrl = pmu->global_status =
- pmu->global_ovf_ctrl = 0;
+ memset(pmu, 0, sizeof(*pmu));
+ kvm_pmu_ops->init(vcpu);
+ init_irq_work(&pmu->irq_work, kvm_pmi_trigger_fn);
+ kvm_pmu_refresh(vcpu);
}
void kvm_pmu_destroy(struct kvm_vcpu *vcpu)
@@ -550,22 +308,20 @@ void kvm_pmu_destroy(struct kvm_vcpu *vcpu)
kvm_pmu_reset(vcpu);
}
-void kvm_handle_pmu_event(struct kvm_vcpu *vcpu)
+/* called before using any PMU functions above */
+int kvm_pmu_arch_init(struct kvm_x86_ops *x86_ops)
{
- struct kvm_pmu *pmu = &vcpu->arch.pmu;
- u64 bitmask;
- int bit;
-
- bitmask = pmu->reprogram_pmi;
+ if (x86_ops == NULL)
+ return -EINVAL;
- for_each_set_bit(bit, (unsigned long *)&bitmask, X86_PMC_IDX_MAX) {
- struct kvm_pmc *pmc = global_idx_to_pmc(pmu, bit);
+ kvm_pmu_ops = x86_ops->get_pmu_ops();
+ if (kvm_pmu_ops == NULL)
+ return -EINVAL;
- if (unlikely(!pmc || !pmc->perf_event)) {
- clear_bit(bit, (unsigned long *)&pmu->reprogram_pmi);
- continue;
- }
+ return 0;
+}
- reprogram_idx(pmu, bit);
- }
+void kvm_pmu_arch_exit(void)
+{
+ kvm_pmu_ops = NULL;
}
new file mode 100644
@@ -0,0 +1,98 @@
+#ifndef __KVM_X86_PMU_H
+#define __KVM_X86_PMU_H
+
+#define vcpu_to_pmu(vcpu) (&(vcpu)->arch.pmu)
+#define pmu_to_vcpu(pmu) (container_of((pmu), struct kvm_vcpu, arch.pmu))
+#define pmc_to_pmu(pmc) (&(pmc)->vcpu->arch.pmu)
+
+/* retrieve the control fields of IA32_FIXED_CTR_CTRL */
+#define fixed_ctrl_field(ctrl_reg, idx) (((ctrl_reg) >> ((idx)*4)) & 0xf)
+
+static inline u64 pmc_bitmask(struct kvm_pmc *pmc)
+{
+ struct kvm_pmu *pmu = pmc_to_pmu(pmc);
+
+ return pmu->counter_bitmask[pmc->type];
+}
+
+static inline u64 pmc_read_counter(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 inline void pmc_stop_counter(struct kvm_pmc *pmc)
+{
+ if (pmc->perf_event) {
+ pmc->counter = pmc_read_counter(pmc);
+ perf_event_release_kernel(pmc->perf_event);
+ pmc->perf_event = NULL;
+ }
+}
+
+static inline bool pmc_is_gp(struct kvm_pmc *pmc)
+{
+ return pmc->type == KVM_PMC_GP;
+}
+
+static inline bool pmc_is_fixed(struct kvm_pmc *pmc)
+{
+ return pmc->type == KVM_PMC_FIXED;
+}
+
+static inline bool pmc_is_enabled(struct kvm_pmc *pmc)
+{
+ return kvm_pmu_ops->pmc_enabled(pmc);
+}
+
+/* returns general purpose PMC with the specified MSR. Note that it can be
+ * used for both PERFCTRn and EVNTSELn; that is why it accepts base as a
+ * paramenter to tell them apart.
+ */
+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;
+}
+
+/* returns fixed PMC with the specified MSR */
+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;
+}
+
+void reprogram_gp_counter(struct kvm_pmc *pmc, u64 eventsel);
+void reprogram_fixed_counter(struct kvm_pmc *pmc, u8 ctrl, int fixed_idx);
+void reprogram_counter(struct kvm_pmu *pmu, int pmc_idx);
+
+void kvm_pmu_handle_event(struct kvm_vcpu *vcpu);
+int kvm_pmu_check_msr(struct kvm_vcpu *vcpu, unsigned idx);
+int kvm_pmu_rdpmc(struct kvm_vcpu *vcpu, unsigned pmc, u64 *data);
+void kvm_pmu_deliver_pmi(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);
+void kvm_pmu_refresh(struct kvm_vcpu *vcpu);
+void kvm_pmu_reset(struct kvm_vcpu *vcpu);
+void kvm_pmu_init(struct kvm_vcpu *vcpu);
+void kvm_pmu_destroy(struct kvm_vcpu *vcpu);
+int kvm_pmu_arch_init(struct kvm_x86_ops *x86_ops);
+void kvm_pmu_arch_exit(void);
+
+extern struct kvm_pmu_ops intel_pmu_ops;
+extern struct kvm_pmu_ops amd_pmu_ops;
+#endif /* __KVM_X86_PMU_H */
new file mode 100644
@@ -0,0 +1,97 @@
+/*
+ * KVM PMU support for AMD
+ *
+ * Copyright 2015, Red Hat, Inc. and/or its affiliates.
+ *
+ * Author:
+ * Wei Huang <wei@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2. See
+ * the COPYING file in the top-level directory.
+ *
+ * Implementation is based on pmu_intel.c file
+ */
+#include <linux/types.h>
+#include <linux/kvm_host.h>
+#include <linux/perf_event.h>
+#include "x86.h"
+#include "cpuid.h"
+#include "lapic.h"
+#include "pmu.h"
+
+static unsigned amd_find_arch_event(struct kvm_pmu *pmu,
+ u8 event_select,
+ u8 unit_mask)
+{
+ return PERF_COUNT_HW_MAX;
+}
+
+/* return PERF_COUNT_HW_MAX as AMD doesn't have fixed events */
+static unsigned amd_find_fixed_event(int idx)
+{
+ return PERF_COUNT_HW_MAX;
+}
+
+static bool amd_pmc_enabled(struct kvm_pmc *pmc)
+{
+ return false;
+}
+
+static struct kvm_pmc *amd_pmc_idx_to_pmc(struct kvm_pmu *pmu, int pmc_idx)
+{
+ return NULL;
+}
+
+/* returns 0 if idx's corresponding MSR exists; otherwise returns 1. */
+static int amd_check_msr(struct kvm_vcpu *vcpu, unsigned idx)
+{
+ return 1;
+}
+
+/* idx is the ECX register of RDPMC instruction */
+static struct kvm_pmc *amd_msr_to_pmc(struct kvm_vcpu *vcpu, unsigned idx)
+{
+ return NULL;
+}
+
+static bool amd_is_pmu_msr(struct kvm_vcpu *vcpu, u32 msr)
+{
+ return false;
+}
+
+static int amd_pmu_get_msr(struct kvm_vcpu *vcpu, u32 msr, u64 *data)
+{
+ return 1;
+}
+
+static int amd_pmu_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
+{
+ return 1;
+}
+
+static void amd_pmu_refresh(struct kvm_vcpu *vcpu)
+{
+}
+
+static void amd_pmu_init(struct kvm_vcpu *vcpu)
+{
+}
+
+static void amd_pmu_reset(struct kvm_vcpu *vcpu)
+{
+}
+
+struct kvm_pmu_ops amd_pmu_ops = {
+ .find_arch_event = amd_find_arch_event,
+ .find_fixed_event = amd_find_fixed_event,
+ .pmc_enabled = amd_pmc_enabled,
+ .pmc_idx_to_pmc = amd_pmc_idx_to_pmc,
+ .check_msr = amd_check_msr,
+ .msr_to_pmc = amd_msr_to_pmc,
+ .is_pmu_msr = amd_is_pmu_msr,
+ .get_msr = amd_pmu_get_msr,
+ .set_msr = amd_pmu_set_msr,
+ .refresh = amd_pmu_refresh,
+ .init = amd_pmu_init,
+ .reset = amd_pmu_reset,
+};
new file mode 100644
@@ -0,0 +1,359 @@
+/*
+ * KVM PMU support for Intel CPUs
+ *
+ * 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"
+#include "pmu.h"
+
+static 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 */
+static int fixed_pmc_events[] = {1, 0, 7};
+
+static void reprogram_fixed_counters(struct kvm_pmu *pmu, u64 data)
+{
+ int i;
+
+ for (i = 0; i < pmu->nr_arch_fixed_counters; i++) {
+ u8 new_ctrl = fixed_ctrl_field(data, i);
+ u8 old_ctrl = fixed_ctrl_field(pmu->fixed_ctr_ctrl, i);
+ struct kvm_pmc *pmc;
+
+ pmc = get_fixed_pmc(pmu, MSR_CORE_PERF_FIXED_CTR0 + i);
+
+ if (old_ctrl == new_ctrl)
+ continue;
+
+ reprogram_fixed_counter(pmc, new_ctrl, i);
+ }
+
+ pmu->fixed_ctr_ctrl = data;
+}
+
+/* function is called when global control register has been updated. */
+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_counter(pmu, bit);
+ }
+}
+
+static unsigned intel_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 unsigned intel_find_fixed_event(int idx)
+{
+
+ if (idx >= ARRAY_SIZE(fixed_pmc_events))
+ return PERF_COUNT_HW_MAX;
+
+ return intel_arch_events[fixed_pmc_events[idx]].event_type;
+}
+
+/* check if a PMC is enabled by comparising it with globl_ctrl bits. */
+static bool intel_pmc_enabled(struct kvm_pmc *pmc)
+{
+ struct kvm_pmu *pmu = pmc_to_pmu(pmc);
+
+ return test_bit(pmc->idx, (unsigned long *)&pmu->global_ctrl);
+}
+
+static struct kvm_pmc *intel_pmc_idx_to_pmc(struct kvm_pmu *pmu, int pmc_idx)
+{
+ if (pmc_idx < INTEL_PMC_IDX_FIXED)
+ return get_gp_pmc(pmu, MSR_P6_EVNTSEL0 + pmc_idx,
+ MSR_P6_EVNTSEL0);
+ else {
+ u32 idx = pmc_idx - INTEL_PMC_IDX_FIXED;
+
+ return get_fixed_pmc(pmu, idx + MSR_CORE_PERF_FIXED_CTR0);
+ }
+}
+
+/* returns 0 if idx's corresponding MSR exists; otherwise returns 1. */
+static int intel_check_msr(struct kvm_vcpu *vcpu, unsigned idx)
+{
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
+ bool fixed = idx & (1u << 30);
+
+ idx &= ~(3u << 30);
+
+ return (!fixed && idx >= pmu->nr_arch_gp_counters) ||
+ (fixed && idx >= pmu->nr_arch_fixed_counters);
+}
+
+static struct kvm_pmc *intel_msr_to_pmc(struct kvm_vcpu *vcpu, unsigned idx)
+{
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
+ bool fixed = idx & (1u << 30);
+ struct kvm_pmc *counters;
+
+ idx &= ~(3u << 30);
+ if (!fixed && idx >= pmu->nr_arch_gp_counters)
+ return NULL;
+ if (fixed && idx >= pmu->nr_arch_fixed_counters)
+ return NULL;
+ counters = fixed ? pmu->fixed_counters : pmu->gp_counters;
+
+ return &counters[idx];
+}
+
+static bool intel_is_pmu_msr(struct kvm_vcpu *vcpu, u32 msr)
+{
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
+ 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;
+}
+
+static int intel_pmu_get_msr(struct kvm_vcpu *vcpu, u32 msr, u64 *data)
+{
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
+ struct kvm_pmc *pmc;
+
+ switch (msr) {
+ 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, msr, MSR_IA32_PERFCTR0)) ||
+ (pmc = get_fixed_pmc(pmu, msr))) {
+ *data = pmc_read_counter(pmc);
+ return 0;
+ } else if ((pmc = get_gp_pmc(pmu, msr, MSR_P6_EVNTSEL0))) {
+ *data = pmc->eventsel;
+ return 0;
+ }
+ }
+
+ return 1;
+}
+
+static int intel_pmu_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
+{
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
+ struct kvm_pmc *pmc;
+ u32 msr = msr_info->index;
+ u64 data = msr_info->data;
+
+ switch (msr) {
+ 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, msr, MSR_IA32_PERFCTR0)) ||
+ (pmc = get_fixed_pmc(pmu, msr))) {
+ if (!msr_info->host_initiated)
+ data = (s64)(s32)data;
+ pmc->counter += data - pmc_read_counter(pmc);
+ return 0;
+ } else if ((pmc = get_gp_pmc(pmu, msr, MSR_P6_EVNTSEL0))) {
+ if (data == pmc->eventsel)
+ return 0;
+ if (!(data & pmu->reserved_bits)) {
+ reprogram_gp_counter(pmc, data);
+ return 0;
+ }
+ }
+ }
+
+ return 1;
+}
+
+static void intel_pmu_refresh(struct kvm_vcpu *vcpu)
+{
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
+ 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;
+}
+
+static void intel_pmu_init(struct kvm_vcpu *vcpu)
+{
+ int i;
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
+
+ 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;
+ }
+}
+
+static void intel_pmu_reset(struct kvm_vcpu *vcpu)
+{
+ struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
+ int i;
+
+ for (i = 0; i < INTEL_PMC_MAX_GENERIC; i++) {
+ struct kvm_pmc *pmc = &pmu->gp_counters[i];
+
+ pmc_stop_counter(pmc);
+ pmc->counter = pmc->eventsel = 0;
+ }
+
+ for (i = 0; i < INTEL_PMC_MAX_FIXED; i++)
+ pmc_stop_counter(&pmu->fixed_counters[i]);
+
+ pmu->fixed_ctr_ctrl = pmu->global_ctrl = pmu->global_status =
+ pmu->global_ovf_ctrl = 0;
+}
+
+struct kvm_pmu_ops intel_pmu_ops = {
+ .find_arch_event = intel_find_arch_event,
+ .find_fixed_event = intel_find_fixed_event,
+ .pmc_enabled = intel_pmc_enabled,
+ .pmc_idx_to_pmc = intel_pmc_idx_to_pmc,
+ .check_msr = intel_check_msr,
+ .msr_to_pmc = intel_msr_to_pmc,
+ .is_pmu_msr = intel_is_pmu_msr,
+ .get_msr = intel_pmu_get_msr,
+ .set_msr = intel_pmu_set_msr,
+ .refresh = intel_pmu_refresh,
+ .init = intel_pmu_init,
+ .reset = intel_pmu_reset,
+};
@@ -21,6 +21,7 @@
#include "kvm_cache_regs.h"
#include "x86.h"
#include "cpuid.h"
+#include "pmu.h"
#include <linux/module.h>
#include <linux/mod_devicetable.h>
@@ -4337,6 +4338,11 @@ static void svm_sched_in(struct kvm_vcpu *vcpu, int cpu)
{
}
+static 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,
@@ -4439,6 +4445,8 @@ 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)
@@ -48,6 +48,7 @@
#include <asm/apic.h>
#include "trace.h"
+#include "pmu.h"
#define __ex(x) __kvm_handle_fault_on_reboot(x)
#define __ex_clear(x, reg) \
@@ -10133,6 +10134,11 @@ static void vmx_enable_log_dirty_pt_masked(struct kvm *kvm,
kvm_mmu_clear_dirty_pt_masked(kvm, memslot, offset, mask);
}
+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,
@@ -10246,6 +10252,8 @@ static struct kvm_x86_ops vmx_x86_ops = {
.slot_disable_log_dirty = vmx_slot_disable_log_dirty,
.flush_log_dirty = vmx_flush_log_dirty,
.enable_log_dirty_pt_masked = vmx_enable_log_dirty_pt_masked,
+
+ .get_pmu_ops = vmx_get_pmu_ops,
};
static int __init vmx_init(void)
@@ -28,6 +28,7 @@
#include "x86.h"
#include "cpuid.h"
#include "assigned-dev.h"
+#include "pmu.h"
#include <linux/clocksource.h>
#include <linux/interrupt.h>
@@ -910,7 +911,7 @@ bool kvm_rdpmc(struct kvm_vcpu *vcpu)
u64 data;
int err;
- err = kvm_pmu_read_pmc(vcpu, ecx, &data);
+ err = kvm_pmu_rdpmc(vcpu, ecx, &data);
if (err)
return err;
kvm_register_write(vcpu, VCPU_REGS_RAX, (u32)data);
@@ -4942,13 +4943,13 @@ static int emulator_set_msr(struct x86_emulate_ctxt *ctxt,
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_check_msr(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_rdpmc(emul_to_vcpu(ctxt), pmc, pdata);
}
static void emulator_halt(struct x86_emulate_ctxt *ctxt)
@@ -5805,6 +5806,11 @@ int kvm_arch_init(void *opaque)
kvm_timer_init();
+ /* init pointers of PMU operations */
+ r = kvm_pmu_arch_init(ops);
+ if (r)
+ goto out_free_percpu;
+
perf_register_guest_info_callbacks(&kvm_guest_cbs);
if (cpu_has_xsave)
@@ -5834,6 +5840,7 @@ void kvm_arch_exit(void)
#ifdef CONFIG_X86_64
pvclock_gtod_unregister_notifier(&pvclock_gtod_notifier);
#endif
+ kvm_pmu_arch_exit();
kvm_x86_ops = NULL;
kvm_mmu_module_exit();
free_percpu(shared_msrs);
@@ -6250,9 +6257,9 @@ 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_handle_event(vcpu);
if (kvm_check_request(KVM_REQ_PMI, vcpu))
- kvm_deliver_pmi(vcpu);
+ kvm_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))
This patch splits existing vPMU code into a common vPMU interface (pmc.c) and Intel specific vPMU code (pmu_intel.c) using the following steps: - Part of arechitectural vPMU code is extracted and moved to pmu_intel.c file. They are hooked up with the newly-defined intel_pmu_ops, which will be called from pmu interface; - Create a dummy pmu_amd.c file for AMD SVM with empty functions; All architectural vPMU functions are now called via PMU function dispatcher (kvm_pmu_ops). This function dispatcher is initialized by calling kvm_x86_ops->get_pmu_ops() at the beginning. Also note that Intel and AMD modules are now generated by combinig their corresponding arch files (vmx.c/svm.c) and pmu files (pmu_intel.c/pmu_amd.c). Signed-off-by: Wei Huang <wei@redhat.com> --- arch/x86/include/asm/kvm_host.h | 3 + arch/x86/kvm/Makefile | 4 +- arch/x86/kvm/cpuid.c | 3 +- arch/x86/kvm/pmu.c | 552 +++++++++++----------------------------- arch/x86/kvm/pmu.h | 98 +++++++ arch/x86/kvm/pmu_amd.c | 97 +++++++ arch/x86/kvm/pmu_intel.c | 359 ++++++++++++++++++++++++++ arch/x86/kvm/svm.c | 8 + arch/x86/kvm/vmx.c | 8 + arch/x86/kvm/x86.c | 17 +- 10 files changed, 743 insertions(+), 406 deletions(-) create mode 100644 arch/x86/kvm/pmu.h create mode 100644 arch/x86/kvm/pmu_amd.c create mode 100644 arch/x86/kvm/pmu_intel.c