KVM: stats: Support binary stats retrieval for a VCPU

Add a VCPU ioctl to get a statistics file descriptor by which a read
functionality is provided for userspace to read out VCPU stats header,
descriptors and data.
Define VCPU statistics descriptors and header for all architectures.
Reviewed-by: David Matlack <dmatlack@google.com>
Reviewed-by: Ricardo Koller <ricarkol@google.com>
Reviewed-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
Reviewed-by: Fuad Tabba <tabba@google.com>
Tested-by: Fuad Tabba <tabba@google.com> #arm64
Signed-off-by: Jing Zhang <jingzhangos@google.com>
Message-Id: <20210618222709.1858088-5-jingzhangos@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

arch/arm64/kvm/guest.c

@@ -43,6 +43,27 @@ const struct kvm_stats_header kvm_vm_stats_header = {
 		       sizeof(kvm_vm_stats_desc),
 };
 
+const struct _kvm_stats_desc kvm_vcpu_stats_desc[] = {
+	KVM_GENERIC_VCPU_STATS(),
+	STATS_DESC_COUNTER(VCPU, hvc_exit_stat),
+	STATS_DESC_COUNTER(VCPU, wfe_exit_stat),
+	STATS_DESC_COUNTER(VCPU, wfi_exit_stat),
+	STATS_DESC_COUNTER(VCPU, mmio_exit_user),
+	STATS_DESC_COUNTER(VCPU, mmio_exit_kernel),
+	STATS_DESC_COUNTER(VCPU, exits)
+};
+static_assert(ARRAY_SIZE(kvm_vcpu_stats_desc) ==
+		sizeof(struct kvm_vcpu_stat) / sizeof(u64));
+
+const struct kvm_stats_header kvm_vcpu_stats_header = {
+	.name_size = KVM_STATS_NAME_SIZE,
+	.num_desc = ARRAY_SIZE(kvm_vcpu_stats_desc),
+	.id_offset = sizeof(struct kvm_stats_header),
+	.desc_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE,
+	.data_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE +
+		       sizeof(kvm_vcpu_stats_desc),
+};
+
 struct kvm_stats_debugfs_item debugfs_entries[] = {
 	VCPU_STAT("halt_successful_poll", generic.halt_successful_poll),
 	VCPU_STAT("halt_attempted_poll", generic.halt_attempted_poll),

arch/mips/kvm/mips.c

@@ -53,6 +53,50 @@ const struct kvm_stats_header kvm_vm_stats_header = {
 		       sizeof(kvm_vm_stats_desc),
 };
 
+const struct _kvm_stats_desc kvm_vcpu_stats_desc[] = {
+	KVM_GENERIC_VCPU_STATS(),
+	STATS_DESC_COUNTER(VCPU, wait_exits),
+	STATS_DESC_COUNTER(VCPU, cache_exits),
+	STATS_DESC_COUNTER(VCPU, signal_exits),
+	STATS_DESC_COUNTER(VCPU, int_exits),
+	STATS_DESC_COUNTER(VCPU, cop_unusable_exits),
+	STATS_DESC_COUNTER(VCPU, tlbmod_exits),
+	STATS_DESC_COUNTER(VCPU, tlbmiss_ld_exits),
+	STATS_DESC_COUNTER(VCPU, tlbmiss_st_exits),
+	STATS_DESC_COUNTER(VCPU, addrerr_st_exits),
+	STATS_DESC_COUNTER(VCPU, addrerr_ld_exits),
+	STATS_DESC_COUNTER(VCPU, syscall_exits),
+	STATS_DESC_COUNTER(VCPU, resvd_inst_exits),
+	STATS_DESC_COUNTER(VCPU, break_inst_exits),
+	STATS_DESC_COUNTER(VCPU, trap_inst_exits),
+	STATS_DESC_COUNTER(VCPU, msa_fpe_exits),
+	STATS_DESC_COUNTER(VCPU, fpe_exits),
+	STATS_DESC_COUNTER(VCPU, msa_disabled_exits),
+	STATS_DESC_COUNTER(VCPU, flush_dcache_exits),
+	STATS_DESC_COUNTER(VCPU, vz_gpsi_exits),
+	STATS_DESC_COUNTER(VCPU, vz_gsfc_exits),
+	STATS_DESC_COUNTER(VCPU, vz_hc_exits),
+	STATS_DESC_COUNTER(VCPU, vz_grr_exits),
+	STATS_DESC_COUNTER(VCPU, vz_gva_exits),
+	STATS_DESC_COUNTER(VCPU, vz_ghfc_exits),
+	STATS_DESC_COUNTER(VCPU, vz_gpa_exits),
+	STATS_DESC_COUNTER(VCPU, vz_resvd_exits),
+#ifdef CONFIG_CPU_LOONGSON64
+	STATS_DESC_COUNTER(VCPU, vz_cpucfg_exits),
+#endif
+};
+static_assert(ARRAY_SIZE(kvm_vcpu_stats_desc) ==
+		sizeof(struct kvm_vcpu_stat) / sizeof(u64));
+
+const struct kvm_stats_header kvm_vcpu_stats_header = {
+	.name_size = KVM_STATS_NAME_SIZE,
+	.num_desc = ARRAY_SIZE(kvm_vcpu_stats_desc),
+	.id_offset = sizeof(struct kvm_stats_header),
+	.desc_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE,
+	.data_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE +
+		       sizeof(kvm_vcpu_stats_desc),
+};
+
 struct kvm_stats_debugfs_item debugfs_entries[] = {
 	VCPU_STAT("wait", wait_exits),
 	VCPU_STAT("cache", cache_exits),

arch/powerpc/kvm/book3s.c

@@ -55,6 +55,51 @@ const struct kvm_stats_header kvm_vm_stats_header = {
 		       sizeof(kvm_vm_stats_desc),
 };
 
+const struct _kvm_stats_desc kvm_vcpu_stats_desc[] = {
+	KVM_GENERIC_VCPU_STATS(),
+	STATS_DESC_COUNTER(VCPU, sum_exits),
+	STATS_DESC_COUNTER(VCPU, mmio_exits),
+	STATS_DESC_COUNTER(VCPU, signal_exits),
+	STATS_DESC_COUNTER(VCPU, light_exits),
+	STATS_DESC_COUNTER(VCPU, itlb_real_miss_exits),
+	STATS_DESC_COUNTER(VCPU, itlb_virt_miss_exits),
+	STATS_DESC_COUNTER(VCPU, dtlb_real_miss_exits),
+	STATS_DESC_COUNTER(VCPU, dtlb_virt_miss_exits),
+	STATS_DESC_COUNTER(VCPU, syscall_exits),
+	STATS_DESC_COUNTER(VCPU, isi_exits),
+	STATS_DESC_COUNTER(VCPU, dsi_exits),
+	STATS_DESC_COUNTER(VCPU, emulated_inst_exits),
+	STATS_DESC_COUNTER(VCPU, dec_exits),
+	STATS_DESC_COUNTER(VCPU, ext_intr_exits),
+	STATS_DESC_TIME_NSEC(VCPU, halt_wait_ns),
+	STATS_DESC_COUNTER(VCPU, halt_successful_wait),
+	STATS_DESC_COUNTER(VCPU, dbell_exits),
+	STATS_DESC_COUNTER(VCPU, gdbell_exits),
+	STATS_DESC_COUNTER(VCPU, ld),
+	STATS_DESC_COUNTER(VCPU, st),
+	STATS_DESC_COUNTER(VCPU, pf_storage),
+	STATS_DESC_COUNTER(VCPU, pf_instruc),
+	STATS_DESC_COUNTER(VCPU, sp_storage),
+	STATS_DESC_COUNTER(VCPU, sp_instruc),
+	STATS_DESC_COUNTER(VCPU, queue_intr),
+	STATS_DESC_COUNTER(VCPU, ld_slow),
+	STATS_DESC_COUNTER(VCPU, st_slow),
+	STATS_DESC_COUNTER(VCPU, pthru_all),
+	STATS_DESC_COUNTER(VCPU, pthru_host),
+	STATS_DESC_COUNTER(VCPU, pthru_bad_aff)
+};
+static_assert(ARRAY_SIZE(kvm_vcpu_stats_desc) ==
+		sizeof(struct kvm_vcpu_stat) / sizeof(u64));
+
+const struct kvm_stats_header kvm_vcpu_stats_header = {
+	.name_size = KVM_STATS_NAME_SIZE,
+	.num_desc = ARRAY_SIZE(kvm_vcpu_stats_desc),
+	.id_offset = sizeof(struct kvm_stats_header),
+	.desc_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE,
+	.data_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE +
+		       sizeof(kvm_vcpu_stats_desc),
+};
+
 struct kvm_stats_debugfs_item debugfs_entries[] = {
 	VCPU_STAT("exits", sum_exits),
 	VCPU_STAT("mmio", mmio_exits),

arch/powerpc/kvm/booke.c

@@ -53,6 +53,44 @@ const struct kvm_stats_header kvm_vm_stats_header = {
 		       sizeof(kvm_vm_stats_desc),
 };
 
+const struct _kvm_stats_desc kvm_vcpu_stats_desc[] = {
+	KVM_GENERIC_VCPU_STATS(),
+	STATS_DESC_COUNTER(VCPU, sum_exits),
+	STATS_DESC_COUNTER(VCPU, mmio_exits),
+	STATS_DESC_COUNTER(VCPU, signal_exits),
+	STATS_DESC_COUNTER(VCPU, light_exits),
+	STATS_DESC_COUNTER(VCPU, itlb_real_miss_exits),
+	STATS_DESC_COUNTER(VCPU, itlb_virt_miss_exits),
+	STATS_DESC_COUNTER(VCPU, dtlb_real_miss_exits),
+	STATS_DESC_COUNTER(VCPU, dtlb_virt_miss_exits),
+	STATS_DESC_COUNTER(VCPU, syscall_exits),
+	STATS_DESC_COUNTER(VCPU, isi_exits),
+	STATS_DESC_COUNTER(VCPU, dsi_exits),
+	STATS_DESC_COUNTER(VCPU, emulated_inst_exits),
+	STATS_DESC_COUNTER(VCPU, dec_exits),
+	STATS_DESC_COUNTER(VCPU, ext_intr_exits),
+	STATS_DESC_TIME_NSEC(VCPU, halt_wait_ns),
+	STATS_DESC_COUNTER(VCPU, halt_successful_wait),
+	STATS_DESC_COUNTER(VCPU, dbell_exits),
+	STATS_DESC_COUNTER(VCPU, gdbell_exits),
+	STATS_DESC_COUNTER(VCPU, ld),
+	STATS_DESC_COUNTER(VCPU, st),
+	STATS_DESC_COUNTER(VCPU, pthru_all),
+	STATS_DESC_COUNTER(VCPU, pthru_host),
+	STATS_DESC_COUNTER(VCPU, pthru_bad_aff)
+};
+static_assert(ARRAY_SIZE(kvm_vcpu_stats_desc) ==
+		sizeof(struct kvm_vcpu_stat) / sizeof(u64));
+
+const struct kvm_stats_header kvm_vcpu_stats_header = {
+	.name_size = KVM_STATS_NAME_SIZE,
+	.num_desc = ARRAY_SIZE(kvm_vcpu_stats_desc),
+	.id_offset = sizeof(struct kvm_stats_header),
+	.desc_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE,
+	.data_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE +
+		       sizeof(kvm_vcpu_stats_desc),
+};
+
 struct kvm_stats_debugfs_item debugfs_entries[] = {
 	VCPU_STAT("mmio", mmio_exits),
 	VCPU_STAT("sig", signal_exits),

arch/s390/kvm/kvm-s390.c

@@ -78,6 +78,114 @@ const struct kvm_stats_header kvm_vm_stats_header = {
 		       sizeof(kvm_vm_stats_desc),
 };
 
+const struct _kvm_stats_desc kvm_vcpu_stats_desc[] = {
+	KVM_GENERIC_VCPU_STATS(),
+	STATS_DESC_COUNTER(VCPU, exit_userspace),
+	STATS_DESC_COUNTER(VCPU, exit_null),
+	STATS_DESC_COUNTER(VCPU, exit_external_request),
+	STATS_DESC_COUNTER(VCPU, exit_io_request),
+	STATS_DESC_COUNTER(VCPU, exit_external_interrupt),
+	STATS_DESC_COUNTER(VCPU, exit_stop_request),
+	STATS_DESC_COUNTER(VCPU, exit_validity),
+	STATS_DESC_COUNTER(VCPU, exit_instruction),
+	STATS_DESC_COUNTER(VCPU, exit_pei),
+	STATS_DESC_COUNTER(VCPU, halt_no_poll_steal),
+	STATS_DESC_COUNTER(VCPU, instruction_lctl),
+	STATS_DESC_COUNTER(VCPU, instruction_lctlg),
+	STATS_DESC_COUNTER(VCPU, instruction_stctl),
+	STATS_DESC_COUNTER(VCPU, instruction_stctg),
+	STATS_DESC_COUNTER(VCPU, exit_program_interruption),
+	STATS_DESC_COUNTER(VCPU, exit_instr_and_program),
+	STATS_DESC_COUNTER(VCPU, exit_operation_exception),
+	STATS_DESC_COUNTER(VCPU, deliver_ckc),
+	STATS_DESC_COUNTER(VCPU, deliver_cputm),
+	STATS_DESC_COUNTER(VCPU, deliver_external_call),
+	STATS_DESC_COUNTER(VCPU, deliver_emergency_signal),
+	STATS_DESC_COUNTER(VCPU, deliver_service_signal),
+	STATS_DESC_COUNTER(VCPU, deliver_virtio),
+	STATS_DESC_COUNTER(VCPU, deliver_stop_signal),
+	STATS_DESC_COUNTER(VCPU, deliver_prefix_signal),
+	STATS_DESC_COUNTER(VCPU, deliver_restart_signal),
+	STATS_DESC_COUNTER(VCPU, deliver_program),
+	STATS_DESC_COUNTER(VCPU, deliver_io),
+	STATS_DESC_COUNTER(VCPU, deliver_machine_check),
+	STATS_DESC_COUNTER(VCPU, exit_wait_state),
+	STATS_DESC_COUNTER(VCPU, inject_ckc),
+	STATS_DESC_COUNTER(VCPU, inject_cputm),
+	STATS_DESC_COUNTER(VCPU, inject_external_call),
+	STATS_DESC_COUNTER(VCPU, inject_emergency_signal),
+	STATS_DESC_COUNTER(VCPU, inject_mchk),
+	STATS_DESC_COUNTER(VCPU, inject_pfault_init),
+	STATS_DESC_COUNTER(VCPU, inject_program),
+	STATS_DESC_COUNTER(VCPU, inject_restart),
+	STATS_DESC_COUNTER(VCPU, inject_set_prefix),
+	STATS_DESC_COUNTER(VCPU, inject_stop_signal),
+	STATS_DESC_COUNTER(VCPU, instruction_epsw),
+	STATS_DESC_COUNTER(VCPU, instruction_gs),
+	STATS_DESC_COUNTER(VCPU, instruction_io_other),
+	STATS_DESC_COUNTER(VCPU, instruction_lpsw),
+	STATS_DESC_COUNTER(VCPU, instruction_lpswe),
+	STATS_DESC_COUNTER(VCPU, instruction_pfmf),
+	STATS_DESC_COUNTER(VCPU, instruction_ptff),
+	STATS_DESC_COUNTER(VCPU, instruction_sck),
+	STATS_DESC_COUNTER(VCPU, instruction_sckpf),
+	STATS_DESC_COUNTER(VCPU, instruction_stidp),
+	STATS_DESC_COUNTER(VCPU, instruction_spx),
+	STATS_DESC_COUNTER(VCPU, instruction_stpx),
+	STATS_DESC_COUNTER(VCPU, instruction_stap),
+	STATS_DESC_COUNTER(VCPU, instruction_iske),
+	STATS_DESC_COUNTER(VCPU, instruction_ri),
+	STATS_DESC_COUNTER(VCPU, instruction_rrbe),
+	STATS_DESC_COUNTER(VCPU, instruction_sske),
+	STATS_DESC_COUNTER(VCPU, instruction_ipte_interlock),
+	STATS_DESC_COUNTER(VCPU, instruction_stsi),
+	STATS_DESC_COUNTER(VCPU, instruction_stfl),
+	STATS_DESC_COUNTER(VCPU, instruction_tb),
+	STATS_DESC_COUNTER(VCPU, instruction_tpi),
+	STATS_DESC_COUNTER(VCPU, instruction_tprot),
+	STATS_DESC_COUNTER(VCPU, instruction_tsch),
+	STATS_DESC_COUNTER(VCPU, instruction_sie),
+	STATS_DESC_COUNTER(VCPU, instruction_essa),
+	STATS_DESC_COUNTER(VCPU, instruction_sthyi),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_sense),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_sense_running),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_external_call),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_emergency),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_cond_emergency),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_start),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_stop),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_stop_store_status),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_store_status),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_store_adtl_status),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_arch),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_prefix),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_restart),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_init_cpu_reset),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_cpu_reset),
+	STATS_DESC_COUNTER(VCPU, instruction_sigp_unknown),
+	STATS_DESC_COUNTER(VCPU, diagnose_10),
+	STATS_DESC_COUNTER(VCPU, diagnose_44),
+	STATS_DESC_COUNTER(VCPU, diagnose_9c),
+	STATS_DESC_COUNTER(VCPU, diagnose_9c_ignored),
+	STATS_DESC_COUNTER(VCPU, diagnose_9c_forward),
+	STATS_DESC_COUNTER(VCPU, diagnose_258),
+	STATS_DESC_COUNTER(VCPU, diagnose_308),
+	STATS_DESC_COUNTER(VCPU, diagnose_500),
+	STATS_DESC_COUNTER(VCPU, diagnose_other),
+	STATS_DESC_COUNTER(VCPU, pfault_sync)
+};
+static_assert(ARRAY_SIZE(kvm_vcpu_stats_desc) ==
+		sizeof(struct kvm_vcpu_stat) / sizeof(u64));
+
+const struct kvm_stats_header kvm_vcpu_stats_header = {
+	.name_size = KVM_STATS_NAME_SIZE,
+	.num_desc = ARRAY_SIZE(kvm_vcpu_stats_desc),
+	.id_offset = sizeof(struct kvm_stats_header),
+	.desc_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE,
+	.data_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE +
+		       sizeof(kvm_vcpu_stats_desc),
+};
+
 struct kvm_stats_debugfs_item debugfs_entries[] = {
 	VCPU_STAT("userspace_handled", exit_userspace),
 	VCPU_STAT("exit_null", exit_null),

arch/x86/kvm/x86.c

@@ -248,6 +248,47 @@ const struct kvm_stats_header kvm_vm_stats_header = {
 		       sizeof(kvm_vm_stats_desc),
 };
 
+const struct _kvm_stats_desc kvm_vcpu_stats_desc[] = {
+	KVM_GENERIC_VCPU_STATS(),
+	STATS_DESC_COUNTER(VCPU, pf_fixed),
+	STATS_DESC_COUNTER(VCPU, pf_guest),
+	STATS_DESC_COUNTER(VCPU, tlb_flush),
+	STATS_DESC_COUNTER(VCPU, invlpg),
+	STATS_DESC_COUNTER(VCPU, exits),
+	STATS_DESC_COUNTER(VCPU, io_exits),
+	STATS_DESC_COUNTER(VCPU, mmio_exits),
+	STATS_DESC_COUNTER(VCPU, signal_exits),
+	STATS_DESC_COUNTER(VCPU, irq_window_exits),
+	STATS_DESC_COUNTER(VCPU, nmi_window_exits),
+	STATS_DESC_COUNTER(VCPU, l1d_flush),
+	STATS_DESC_COUNTER(VCPU, halt_exits),
+	STATS_DESC_COUNTER(VCPU, request_irq_exits),
+	STATS_DESC_COUNTER(VCPU, irq_exits),
+	STATS_DESC_COUNTER(VCPU, host_state_reload),
+	STATS_DESC_COUNTER(VCPU, fpu_reload),
+	STATS_DESC_COUNTER(VCPU, insn_emulation),
+	STATS_DESC_COUNTER(VCPU, insn_emulation_fail),
+	STATS_DESC_COUNTER(VCPU, hypercalls),
+	STATS_DESC_COUNTER(VCPU, irq_injections),
+	STATS_DESC_COUNTER(VCPU, nmi_injections),
+	STATS_DESC_COUNTER(VCPU, req_event),
+	STATS_DESC_COUNTER(VCPU, nested_run),
+	STATS_DESC_COUNTER(VCPU, directed_yield_attempted),
+	STATS_DESC_COUNTER(VCPU, directed_yield_successful),
+	STATS_DESC_ICOUNTER(VCPU, guest_mode)
+};
+static_assert(ARRAY_SIZE(kvm_vcpu_stats_desc) ==
+		sizeof(struct kvm_vcpu_stat) / sizeof(u64));
+
+const struct kvm_stats_header kvm_vcpu_stats_header = {
+	.name_size = KVM_STATS_NAME_SIZE,
+	.num_desc = ARRAY_SIZE(kvm_vcpu_stats_desc),
+	.id_offset = sizeof(struct kvm_stats_header),
+	.desc_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE,
+	.data_offset = sizeof(struct kvm_stats_header) + KVM_STATS_NAME_SIZE +
+		       sizeof(kvm_vcpu_stats_desc),
+};
+
 struct kvm_stats_debugfs_item debugfs_entries[] = {
 	VCPU_STAT("pf_fixed", pf_fixed),
 	VCPU_STAT("pf_guest", pf_guest),

include/linux/kvm_host.h

@@ -305,7 +305,6 @@ struct kvm_vcpu {
 	struct pid __rcu *pid;
 	int sigset_active;
 	sigset_t sigset;
-	struct kvm_vcpu_stat stat;
 	unsigned int halt_poll_ns;
 	bool valid_wakeup;
 
@@ -342,6 +341,8 @@ struct kvm_vcpu {
 	bool preempted;
 	bool ready;
 	struct kvm_vcpu_arch arch;
+	struct kvm_vcpu_stat stat;
+	char stats_id[KVM_STATS_NAME_SIZE];
 	struct kvm_dirty_ring dirty_ring;
 };
 
@@ -1358,6 +1359,14 @@ struct _kvm_stats_desc {
 #define KVM_GENERIC_VM_STATS()						       \
 	STATS_DESC_COUNTER(VM_GENERIC, remote_tlb_flush)
 
+#define KVM_GENERIC_VCPU_STATS()					       \
+	STATS_DESC_COUNTER(VCPU_GENERIC, halt_successful_poll),		       \
+	STATS_DESC_COUNTER(VCPU_GENERIC, halt_attempted_poll),		       \
+	STATS_DESC_COUNTER(VCPU_GENERIC, halt_poll_invalid),		       \
+	STATS_DESC_COUNTER(VCPU_GENERIC, halt_wakeup),			       \
+	STATS_DESC_TIME_NSEC(VCPU_GENERIC, halt_poll_success_ns),	       \
+	STATS_DESC_TIME_NSEC(VCPU_GENERIC, halt_poll_fail_ns)
+
 extern struct kvm_stats_debugfs_item debugfs_entries[];
 extern struct dentry *kvm_debugfs_dir;
 ssize_t kvm_stats_read(char *id, const struct kvm_stats_header *header,
@@ -1366,6 +1375,8 @@ ssize_t kvm_stats_read(char *id, const struct kvm_stats_header *header,
 		       char __user *user_buffer, size_t size, loff_t *offset);
 extern const struct kvm_stats_header kvm_vm_stats_header;
 extern const struct _kvm_stats_desc kvm_vm_stats_desc[];
+extern const struct kvm_stats_header kvm_vcpu_stats_header;
+extern const struct _kvm_stats_desc kvm_vcpu_stats_desc[];
 
 #if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
 static inline int mmu_notifier_retry(struct kvm *kvm, unsigned long mmu_seq)

virt/kvm/kvm_main.c

@@ -3448,6 +3448,10 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, u32 id)
 	vcpu->vcpu_idx = atomic_read(&kvm->online_vcpus);
 	BUG_ON(kvm->vcpus[vcpu->vcpu_idx]);
 
+	/* Fill the stats id string for the vcpu */
+	snprintf(vcpu->stats_id, sizeof(vcpu->stats_id), "kvm-%d/vcpu-%d",
+		 task_pid_nr(current), id);
+
 	/* Now it's all set up, let userspace reach it */
 	kvm_get_kvm(kvm);
 	r = create_vcpu_fd(vcpu);
@@ -3497,6 +3501,44 @@ static int kvm_vcpu_ioctl_set_sigmask(struct kvm_vcpu *vcpu, sigset_t *sigset)
 	return 0;
 }
 
+static ssize_t kvm_vcpu_stats_read(struct file *file, char __user *user_buffer,
+			      size_t size, loff_t *offset)
+{
+	struct kvm_vcpu *vcpu = file->private_data;
+
+	return kvm_stats_read(vcpu->stats_id, &kvm_vcpu_stats_header,
+			&kvm_vcpu_stats_desc[0], &vcpu->stat,
+			sizeof(vcpu->stat), user_buffer, size, offset);
+}
+
+static const struct file_operations kvm_vcpu_stats_fops = {
+	.read = kvm_vcpu_stats_read,
+	.llseek = noop_llseek,
+};
+
+static int kvm_vcpu_ioctl_get_stats_fd(struct kvm_vcpu *vcpu)
+{
+	int fd;
+	struct file *file;
+	char name[15 + ITOA_MAX_LEN + 1];
+
+	snprintf(name, sizeof(name), "kvm-vcpu-stats:%d", vcpu->vcpu_id);
+
+	fd = get_unused_fd_flags(O_CLOEXEC);
+	if (fd < 0)
+		return fd;
+
+	file = anon_inode_getfile(name, &kvm_vcpu_stats_fops, vcpu, O_RDONLY);
+	if (IS_ERR(file)) {
+		put_unused_fd(fd);
+		return PTR_ERR(file);
+	}
+	file->f_mode |= FMODE_PREAD;
+	fd_install(fd, file);
+
+	return fd;
+}
+
 static long kvm_vcpu_ioctl(struct file *filp,
 			   unsigned int ioctl, unsigned long arg)
 {
@@ -3694,6 +3736,10 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = kvm_arch_vcpu_ioctl_set_fpu(vcpu, fpu);
 		break;
 	}
+	case KVM_GET_STATS_FD: {
+		r = kvm_vcpu_ioctl_get_stats_fd(vcpu);
+		break;
+	}
 	default:
 		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
 	}
@@ -3952,6 +3998,8 @@ static long kvm_vm_ioctl_check_extension_generic(struct kvm *kvm, long arg)
 #else
 		return 0;
 #endif
+	case KVM_CAP_BINARY_STATS_FD:
+		return 1;
 	default:
 		break;
 	}
@@ -5254,7 +5302,8 @@ int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align,
 		kmem_cache_create_usercopy("kvm_vcpu", vcpu_size, vcpu_align,
 					   SLAB_ACCOUNT,
 					   offsetof(struct kvm_vcpu, arch),
-					   sizeof_field(struct kvm_vcpu, arch),
+					   offsetofend(struct kvm_vcpu, stats_id)
+					   - offsetof(struct kvm_vcpu, arch),
 					   NULL);
 	if (!kvm_vcpu_cache) {
 		r = -ENOMEM;