KVM: nVMX: Add preemption timer support

Provided the host has this feature, it's straightforward to offer it to
the guest as well. We just need to load to timer value on L2 entry if
the feature was enabled by L1 and watch out for the corresponding exit
reason.
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Gleb Natapov <gleb@redhat.com>

arch/x86/include/asm/vmx.h

@@ -70,6 +70,7 @@
 #define PIN_BASED_EXT_INTR_MASK                 0x00000001
 #define PIN_BASED_NMI_EXITING                   0x00000008
 #define PIN_BASED_VIRTUAL_NMIS                  0x00000020
+#define PIN_BASED_VMX_PREEMPTION_TIMER          0x00000040
 
 #define PIN_BASED_ALWAYSON_WITHOUT_TRUE_MSR	0x00000016
 
@@ -95,6 +96,7 @@
 
 #define VM_ENTRY_ALWAYSON_WITHOUT_TRUE_MSR	0x000011ff
 
+#define VMX_MISC_PREEMPTION_TIMER_RATE_MASK	0x0000001f
 #define VMX_MISC_SAVE_EFER_LMA			0x00000020
 
 /* VMCS Encodings */
@@ -217,6 +219,7 @@ enum vmcs_field {
 	GUEST_INTERRUPTIBILITY_INFO     = 0x00004824,
 	GUEST_ACTIVITY_STATE            = 0X00004826,
 	GUEST_SYSENTER_CS               = 0x0000482A,
+	VMX_PREEMPTION_TIMER_VALUE      = 0x0000482E,
 	HOST_IA32_SYSENTER_CS           = 0x00004c00,
 	CR0_GUEST_HOST_MASK             = 0x00006000,
 	CR4_GUEST_HOST_MASK             = 0x00006002,

arch/x86/include/uapi/asm/vmx.h

@@ -65,6 +65,7 @@
 #define EXIT_REASON_EOI_INDUCED         45
 #define EXIT_REASON_EPT_VIOLATION       48
 #define EXIT_REASON_EPT_MISCONFIG       49
+#define EXIT_REASON_PREEMPTION_TIMER    52
 #define EXIT_REASON_WBINVD              54
 #define EXIT_REASON_XSETBV              55
 #define EXIT_REASON_APIC_WRITE          56
@@ -110,7 +111,7 @@
 	{ EXIT_REASON_EOI_INDUCED,           "EOI_INDUCED" }, \
 	{ EXIT_REASON_INVALID_STATE,         "INVALID_STATE" }, \
 	{ EXIT_REASON_INVD,                  "INVD" }, \
-	{ EXIT_REASON_INVPCID,               "INVPCID" }
-
+	{ EXIT_REASON_INVPCID,               "INVPCID" }, \
+	{ EXIT_REASON_PREEMPTION_TIMER,      "PREEMPTION_TIMER" }
 
 #endif /* _UAPIVMX_H */

arch/x86/kvm/vmx.c

@@ -298,7 +298,8 @@ struct __packed vmcs12 {
 	u32 guest_activity_state;
 	u32 guest_sysenter_cs;
 	u32 host_ia32_sysenter_cs;
-	u32 padding32[8]; /* room for future expansion */
+	u32 vmx_preemption_timer_value;
+	u32 padding32[7]; /* room for future expansion */
 	u16 virtual_processor_id;
 	u16 guest_es_selector;
 	u16 guest_cs_selector;
@@ -537,6 +538,7 @@ static const unsigned short vmcs_field_to_offset_table[] = {
 	FIELD(GUEST_ACTIVITY_STATE, guest_activity_state),
 	FIELD(GUEST_SYSENTER_CS, guest_sysenter_cs),
 	FIELD(HOST_IA32_SYSENTER_CS, host_ia32_sysenter_cs),
+	FIELD(VMX_PREEMPTION_TIMER_VALUE, vmx_preemption_timer_value),
 	FIELD(CR0_GUEST_HOST_MASK, cr0_guest_host_mask),
 	FIELD(CR4_GUEST_HOST_MASK, cr4_guest_host_mask),
 	FIELD(CR0_READ_SHADOW, cr0_read_shadow),
@@ -2049,7 +2051,8 @@ static __init void nested_vmx_setup_ctls_msrs(void)
 	 */
 	nested_vmx_pinbased_ctls_low |= PIN_BASED_ALWAYSON_WITHOUT_TRUE_MSR;
 	nested_vmx_pinbased_ctls_high &= PIN_BASED_EXT_INTR_MASK |
-		PIN_BASED_NMI_EXITING | PIN_BASED_VIRTUAL_NMIS;
+		PIN_BASED_NMI_EXITING | PIN_BASED_VIRTUAL_NMIS |
+		PIN_BASED_VMX_PREEMPTION_TIMER;
 	nested_vmx_pinbased_ctls_high |= PIN_BASED_ALWAYSON_WITHOUT_TRUE_MSR;
 
 	/*
@@ -2110,7 +2113,8 @@ static __init void nested_vmx_setup_ctls_msrs(void)
 
 	/* miscellaneous data */
 	rdmsr(MSR_IA32_VMX_MISC, nested_vmx_misc_low, nested_vmx_misc_high);
-	nested_vmx_misc_low &= VMX_MISC_SAVE_EFER_LMA;
+	nested_vmx_misc_low &= VMX_MISC_PREEMPTION_TIMER_RATE_MASK |
+		VMX_MISC_SAVE_EFER_LMA;
 	nested_vmx_misc_high = 0;
 }
 
@@ -6190,6 +6194,9 @@ static bool nested_vmx_exit_handled(struct kvm_vcpu *vcpu)
 	case EXIT_REASON_EPT_VIOLATION:
 	case EXIT_REASON_EPT_MISCONFIG:
 		return 0;
+	case EXIT_REASON_PREEMPTION_TIMER:
+		return vmcs12->pin_based_vm_exec_control &
+			PIN_BASED_VMX_PREEMPTION_TIMER;
 	case EXIT_REASON_WBINVD:
 		return nested_cpu_has2(vmcs12, SECONDARY_EXEC_WBINVD_EXITING);
 	case EXIT_REASON_XSETBV:
@@ -7011,6 +7018,10 @@ static void prepare_vmcs02(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12)
 		(vmcs_config.pin_based_exec_ctrl |
 		 vmcs12->pin_based_vm_exec_control));
 
+	if (vmcs12->pin_based_vm_exec_control & PIN_BASED_VMX_PREEMPTION_TIMER)
+		vmcs_write32(VMX_PREEMPTION_TIMER_VALUE,
+			     vmcs12->vmx_preemption_timer_value);
+
 	/*
 	 * Whether page-faults are trapped is determined by a combination of
 	 * 3 settings: PFEC_MASK, PFEC_MATCH and EXCEPTION_BITMAP.PF.