KVM: x86: Save&restore interrupt shadow mask

The interrupt shadow created by STI or MOV-SS-like operations is part of
the VCPU state and must be preserved across migration. Transfer it in
the spare padding field of kvm_vcpu_events.interrupt.

As a side effect we now have to make vmx_set_interrupt_shadow robust
against both shadow types being set. Give MOV SS a higher priority and
skip STI in that case to avoid that VMX throws a fault on next entry.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>

Documentation/kvm/api.txt

@@ -656,6 +656,7 @@ struct kvm_clock_data {
 4.29 KVM_GET_VCPU_EVENTS
 
 Capability: KVM_CAP_VCPU_EVENTS
+Extended by: KVM_CAP_INTR_SHADOW
 Architectures: x86
 Type: vm ioctl
 Parameters: struct kvm_vcpu_event (out)
@@ -676,7 +677,7 @@ struct kvm_vcpu_events {
 		__u8 injected;
 		__u8 nr;
 		__u8 soft;
-		__u8 pad;
+		__u8 shadow;
 	} interrupt;
 	struct {
 		__u8 injected;
@@ -688,9 +689,13 @@ struct kvm_vcpu_events {
 	__u32 flags;
 };
 
+KVM_VCPUEVENT_VALID_SHADOW may be set in the flags field to signal that
+interrupt.shadow contains a valid state. Otherwise, this field is undefined.
+
 4.30 KVM_SET_VCPU_EVENTS
 
 Capability: KVM_CAP_VCPU_EVENTS
+Extended by: KVM_CAP_INTR_SHADOW
 Architectures: x86
 Type: vm ioctl
 Parameters: struct kvm_vcpu_event (in)
@@ -709,6 +714,10 @@ current in-kernel state. The bits are:
 KVM_VCPUEVENT_VALID_NMI_PENDING - transfer nmi.pending to the kernel
 KVM_VCPUEVENT_VALID_SIPI_VECTOR - transfer sipi_vector
 
+If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in
+the flags field to signal that interrupt.shadow contains a valid state and
+shall be written into the VCPU.
+
 
 5. The kvm_run structure
 

arch/x86/include/asm/kvm.h

@@ -257,6 +257,11 @@ struct kvm_reinject_control {
 /* When set in flags, include corresponding fields on KVM_SET_VCPU_EVENTS */
 #define KVM_VCPUEVENT_VALID_NMI_PENDING	0x00000001
 #define KVM_VCPUEVENT_VALID_SIPI_VECTOR	0x00000002
+#define KVM_VCPUEVENT_VALID_SHADOW	0x00000004
+
+/* Interrupt shadow states */
+#define KVM_X86_SHADOW_INT_MOV_SS	0x01
+#define KVM_X86_SHADOW_INT_STI		0x02
 
 /* for KVM_GET/SET_VCPU_EVENTS */
 struct kvm_vcpu_events {
@@ -271,7 +276,7 @@ struct kvm_vcpu_events {
 		__u8 injected;
 		__u8 nr;
 		__u8 soft;
-		__u8 pad;
+		__u8 shadow;
 	} interrupt;
 	struct {
 		__u8 injected;

arch/x86/include/asm/kvm_emulate.h

@@ -153,9 +153,6 @@ struct decode_cache {
 	struct fetch_cache fetch;
 };
 
-#define X86_SHADOW_INT_MOV_SS  1
-#define X86_SHADOW_INT_STI     2
-
 struct x86_emulate_ctxt {
 	/* Register state before/after emulation. */
 	struct kvm_vcpu *vcpu;

arch/x86/kvm/emulate.c

@@ -2128,7 +2128,7 @@ x86_emulate_insn(struct x86_emulate_ctxt *ctxt, struct x86_emulate_ops *ops)
 		}
 
 		if (c->modrm_reg == VCPU_SREG_SS)
-			toggle_interruptibility(ctxt, X86_SHADOW_INT_MOV_SS);
+			toggle_interruptibility(ctxt, KVM_X86_SHADOW_INT_MOV_SS);
 
 		rc = kvm_load_segment_descriptor(ctxt->vcpu, sel, c->modrm_reg);
 
@@ -2366,7 +2366,7 @@ x86_emulate_insn(struct x86_emulate_ctxt *ctxt, struct x86_emulate_ops *ops)
 		if (emulator_bad_iopl(ctxt))
 			kvm_inject_gp(ctxt->vcpu, 0);
 		else {
-			toggle_interruptibility(ctxt, X86_SHADOW_INT_STI);
+			toggle_interruptibility(ctxt, KVM_X86_SHADOW_INT_STI);
 			ctxt->eflags |= X86_EFLAGS_IF;
 			c->dst.type = OP_NONE;	/* Disable writeback. */
 		}

arch/x86/kvm/svm.c

@@ -265,7 +265,7 @@ static u32 svm_get_interrupt_shadow(struct kvm_vcpu *vcpu, int mask)
 	u32 ret = 0;
 
 	if (svm->vmcb->control.int_state & SVM_INTERRUPT_SHADOW_MASK)
-		ret |= X86_SHADOW_INT_STI | X86_SHADOW_INT_MOV_SS;
+		ret |= KVM_X86_SHADOW_INT_STI | KVM_X86_SHADOW_INT_MOV_SS;
 	return ret & mask;
 }
 

arch/x86/kvm/vmx.c

@@ -846,9 +846,9 @@ static u32 vmx_get_interrupt_shadow(struct kvm_vcpu *vcpu, int mask)
 	int ret = 0;
 
 	if (interruptibility & GUEST_INTR_STATE_STI)
-		ret |= X86_SHADOW_INT_STI;
+		ret |= KVM_X86_SHADOW_INT_STI;
 	if (interruptibility & GUEST_INTR_STATE_MOV_SS)
-		ret |= X86_SHADOW_INT_MOV_SS;
+		ret |= KVM_X86_SHADOW_INT_MOV_SS;
 
 	return ret & mask;
 }
@@ -860,9 +860,9 @@ static void vmx_set_interrupt_shadow(struct kvm_vcpu *vcpu, int mask)
 
 	interruptibility &= ~(GUEST_INTR_STATE_STI | GUEST_INTR_STATE_MOV_SS);
 
-	if (mask & X86_SHADOW_INT_MOV_SS)
+	if (mask & KVM_X86_SHADOW_INT_MOV_SS)
 		interruptibility |= GUEST_INTR_STATE_MOV_SS;
-	if (mask & X86_SHADOW_INT_STI)
+	else if (mask & KVM_X86_SHADOW_INT_STI)
 		interruptibility |= GUEST_INTR_STATE_STI;
 
 	if ((interruptibility != interruptibility_old))

arch/x86/kvm/x86.c

@@ -2111,6 +2111,9 @@ static void kvm_vcpu_ioctl_x86_get_vcpu_events(struct kvm_vcpu *vcpu,
 		vcpu->arch.interrupt.pending && !vcpu->arch.interrupt.soft;
 	events->interrupt.nr = vcpu->arch.interrupt.nr;
 	events->interrupt.soft = 0;
+	events->interrupt.shadow =
+		kvm_x86_ops->get_interrupt_shadow(vcpu,
+			KVM_X86_SHADOW_INT_MOV_SS | KVM_X86_SHADOW_INT_STI);
 
 	events->nmi.injected = vcpu->arch.nmi_injected;
 	events->nmi.pending = vcpu->arch.nmi_pending;
@@ -2119,7 +2122,8 @@ static void kvm_vcpu_ioctl_x86_get_vcpu_events(struct kvm_vcpu *vcpu,
 	events->sipi_vector = vcpu->arch.sipi_vector;
 
 	events->flags = (KVM_VCPUEVENT_VALID_NMI_PENDING
-			 | KVM_VCPUEVENT_VALID_SIPI_VECTOR);
+			 | KVM_VCPUEVENT_VALID_SIPI_VECTOR
+			 | KVM_VCPUEVENT_VALID_SHADOW);
 
 	vcpu_put(vcpu);
 }
@@ -2128,7 +2132,8 @@ static int kvm_vcpu_ioctl_x86_set_vcpu_events(struct kvm_vcpu *vcpu,
 					      struct kvm_vcpu_events *events)
 {
 	if (events->flags & ~(KVM_VCPUEVENT_VALID_NMI_PENDING
-			      | KVM_VCPUEVENT_VALID_SIPI_VECTOR))
+			      | KVM_VCPUEVENT_VALID_SIPI_VECTOR
+			      | KVM_VCPUEVENT_VALID_SHADOW))
 		return -EINVAL;
 
 	vcpu_load(vcpu);
@@ -2143,6 +2148,9 @@ static int kvm_vcpu_ioctl_x86_set_vcpu_events(struct kvm_vcpu *vcpu,
 	vcpu->arch.interrupt.soft = events->interrupt.soft;
 	if (vcpu->arch.interrupt.pending && irqchip_in_kernel(vcpu->kvm))
 		kvm_pic_clear_isr_ack(vcpu->kvm);
+	if (events->flags & KVM_VCPUEVENT_VALID_SHADOW)
+		kvm_x86_ops->set_interrupt_shadow(vcpu,
+						  events->interrupt.shadow);
 
 	vcpu->arch.nmi_injected = events->nmi.injected;
 	if (events->flags & KVM_VCPUEVENT_VALID_NMI_PENDING)

include/linux/kvm.h

@@ -502,6 +502,7 @@ struct kvm_ioeventfd {
 #define KVM_CAP_HYPERV_SPIN 46
 #define KVM_CAP_PCI_SEGMENT 47
 #define KVM_CAP_PPC_PAIRED_SINGLES 48
+#define KVM_CAP_INTR_SHADOW 49
 #define KVM_CAP_X86_ROBUST_SINGLESTEP 51
 
 #ifdef KVM_CAP_IRQ_ROUTING