KVM: PPC: Book3S: Add MMIO emulation for VMX instructions

This patch provides the MMIO load/store vector indexed
X-Form emulation.

Instructions implemented:
lvx: the quadword in storage addressed by the result of EA &
0xffff_ffff_ffff_fff0 is loaded into VRT.

stvx: the contents of VRS are stored into the quadword in storage
addressed by the result of EA & 0xffff_ffff_ffff_fff0.
Reported-by: Gopesh Kumar Chaudhary <gopchaud@in.ibm.com>
Reported-by: Balamuruhan S <bala24@linux.vnet.ibm.com>
Signed-off-by: Jose Ricardo Ziviani <joserz@linux.vnet.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>

arch/powerpc/include/asm/kvm_host.h

@@ -690,6 +690,7 @@ struct kvm_vcpu_arch {
 	u8 mmio_vsx_offset;
 	u8 mmio_vsx_copy_type;
 	u8 mmio_vsx_tx_sx_enabled;
+	u8 mmio_vmx_copy_nums;
 	u8 osi_needed;
 	u8 osi_enabled;
 	u8 papr_enabled;
@@ -804,6 +805,7 @@ struct kvm_vcpu_arch {
 #define KVM_MMIO_REG_QPR	0x0040
 #define KVM_MMIO_REG_FQPR	0x0060
 #define KVM_MMIO_REG_VSX	0x0080
+#define KVM_MMIO_REG_VMX	0x00c0
 
 #define __KVM_HAVE_ARCH_WQP
 #define __KVM_HAVE_CREATE_DEVICE

arch/powerpc/include/asm/kvm_ppc.h

@@ -81,6 +81,10 @@ extern int kvmppc_handle_loads(struct kvm_run *run, struct kvm_vcpu *vcpu,
 extern int kvmppc_handle_vsx_load(struct kvm_run *run, struct kvm_vcpu *vcpu,
 				unsigned int rt, unsigned int bytes,
 			int is_default_endian, int mmio_sign_extend);
+extern int kvmppc_handle_load128_by2x64(struct kvm_run *run,
+		struct kvm_vcpu *vcpu, unsigned int rt, int is_default_endian);
+extern int kvmppc_handle_store128_by2x64(struct kvm_run *run,
+		struct kvm_vcpu *vcpu, unsigned int rs, int is_default_endian);
 extern int kvmppc_handle_store(struct kvm_run *run, struct kvm_vcpu *vcpu,
 			       u64 val, unsigned int bytes,
 			       int is_default_endian);

arch/powerpc/include/asm/ppc-opcode.h

@@ -156,6 +156,12 @@
 #define OP_31_XOP_LFDX          599
 #define OP_31_XOP_LFDUX		631
 
+/* VMX Vector Load Instructions */
+#define OP_31_XOP_LVX           103
+
+/* VMX Vector Store Instructions */
+#define OP_31_XOP_STVX          231
+
 #define OP_LWZ  32
 #define OP_STFS 52
 #define OP_STFSU 53

arch/powerpc/kvm/emulate_loadstore.c

@@ -58,6 +58,18 @@ static bool kvmppc_check_vsx_disabled(struct kvm_vcpu *vcpu)
 }
 #endif /* CONFIG_VSX */
 
+#ifdef CONFIG_ALTIVEC
+static bool kvmppc_check_altivec_disabled(struct kvm_vcpu *vcpu)
+{
+	if (!(kvmppc_get_msr(vcpu) & MSR_VEC)) {
+		kvmppc_core_queue_vec_unavail(vcpu);
+		return true;
+	}
+
+	return false;
+}
+#endif /* CONFIG_ALTIVEC */
+
 /*
  * XXX to do:
  * lfiwax, lfiwzx
@@ -98,6 +110,7 @@ int kvmppc_emulate_loadstore(struct kvm_vcpu *vcpu)
 	vcpu->arch.mmio_vsx_copy_type = KVMPPC_VSX_COPY_NONE;
 	vcpu->arch.mmio_sp64_extend = 0;
 	vcpu->arch.mmio_sign_extend = 0;
+	vcpu->arch.mmio_vmx_copy_nums = 0;
 
 	switch (get_op(inst)) {
 	case 31:
@@ -459,6 +472,29 @@ int kvmppc_emulate_loadstore(struct kvm_vcpu *vcpu)
 							 rs, 4, 1);
 			break;
 #endif /* CONFIG_VSX */
+
+#ifdef CONFIG_ALTIVEC
+		case OP_31_XOP_LVX:
+			if (kvmppc_check_altivec_disabled(vcpu))
+				return EMULATE_DONE;
+			vcpu->arch.vaddr_accessed &= ~0xFULL;
+			vcpu->arch.paddr_accessed &= ~0xFULL;
+			vcpu->arch.mmio_vmx_copy_nums = 2;
+			emulated = kvmppc_handle_load128_by2x64(run, vcpu,
+					KVM_MMIO_REG_VMX|rt, 1);
+			break;
+
+		case OP_31_XOP_STVX:
+			if (kvmppc_check_altivec_disabled(vcpu))
+				return EMULATE_DONE;
+			vcpu->arch.vaddr_accessed &= ~0xFULL;
+			vcpu->arch.paddr_accessed &= ~0xFULL;
+			vcpu->arch.mmio_vmx_copy_nums = 2;
+			emulated = kvmppc_handle_store128_by2x64(run, vcpu,
+					rs, 1);
+			break;
+#endif /* CONFIG_ALTIVEC */
+
 		default:
 			emulated = EMULATE_FAIL;
 			break;

arch/powerpc/kvm/powerpc.c

@@ -924,6 +924,34 @@ static inline void kvmppc_set_vsr_word(struct kvm_vcpu *vcpu,
 }
 #endif /* CONFIG_VSX */
 
+#ifdef CONFIG_ALTIVEC
+static inline void kvmppc_set_vmx_dword(struct kvm_vcpu *vcpu,
+		u64 gpr)
+{
+	int index = vcpu->arch.io_gpr & KVM_MMIO_REG_MASK;
+	u32 hi, lo;
+	u32 di;
+
+#ifdef __BIG_ENDIAN
+	hi = gpr >> 32;
+	lo = gpr & 0xffffffff;
+#else
+	lo = gpr >> 32;
+	hi = gpr & 0xffffffff;
+#endif
+
+	di = 2 - vcpu->arch.mmio_vmx_copy_nums;		/* doubleword index */
+	if (di > 1)
+		return;
+
+	if (vcpu->arch.mmio_host_swabbed)
+		di = 1 - di;
+
+	VCPU_VSX_VR(vcpu, index).u[di * 2] = hi;
+	VCPU_VSX_VR(vcpu, index).u[di * 2 + 1] = lo;
+}
+#endif /* CONFIG_ALTIVEC */
+
 #ifdef CONFIG_PPC_FPU
 static inline u64 sp_to_dp(u32 fprs)
 {
@@ -1026,6 +1054,11 @@ static void kvmppc_complete_mmio_load(struct kvm_vcpu *vcpu,
 				KVMPPC_VSX_COPY_DWORD_LOAD_DUMP)
 			kvmppc_set_vsr_dword_dump(vcpu, gpr);
 		break;
+#endif
+#ifdef CONFIG_ALTIVEC
+	case KVM_MMIO_REG_VMX:
+		kvmppc_set_vmx_dword(vcpu, gpr);
+		break;
 #endif
 	default:
 		BUG();
@@ -1302,6 +1335,111 @@ static int kvmppc_emulate_mmio_vsx_loadstore(struct kvm_vcpu *vcpu,
 }
 #endif /* CONFIG_VSX */
 
+#ifdef CONFIG_ALTIVEC
+/* handle quadword load access in two halves */
+int kvmppc_handle_load128_by2x64(struct kvm_run *run, struct kvm_vcpu *vcpu,
+		unsigned int rt, int is_default_endian)
+{
+	enum emulation_result emulated;
+
+	while (vcpu->arch.mmio_vmx_copy_nums) {
+		emulated = __kvmppc_handle_load(run, vcpu, rt, 8,
+				is_default_endian, 0);
+
+		if (emulated != EMULATE_DONE)
+			break;
+
+		vcpu->arch.paddr_accessed += run->mmio.len;
+		vcpu->arch.mmio_vmx_copy_nums--;
+	}
+
+	return emulated;
+}
+
+static inline int kvmppc_get_vmx_data(struct kvm_vcpu *vcpu, int rs, u64 *val)
+{
+	vector128 vrs = VCPU_VSX_VR(vcpu, rs);
+	u32 di;
+	u64 w0, w1;
+
+	di = 2 - vcpu->arch.mmio_vmx_copy_nums;		/* doubleword index */
+	if (di > 1)
+		return -1;
+
+	if (vcpu->arch.mmio_host_swabbed)
+		di = 1 - di;
+
+	w0 = vrs.u[di * 2];
+	w1 = vrs.u[di * 2 + 1];
+
+#ifdef __BIG_ENDIAN
+	*val = (w0 << 32) | w1;
+#else
+	*val = (w1 << 32) | w0;
+#endif
+	return 0;
+}
+
+/* handle quadword store in two halves */
+int kvmppc_handle_store128_by2x64(struct kvm_run *run, struct kvm_vcpu *vcpu,
+		unsigned int rs, int is_default_endian)
+{
+	u64 val = 0;
+	enum emulation_result emulated = EMULATE_DONE;
+
+	vcpu->arch.io_gpr = rs;
+
+	while (vcpu->arch.mmio_vmx_copy_nums) {
+		if (kvmppc_get_vmx_data(vcpu, rs, &val) == -1)
+			return EMULATE_FAIL;
+
+		emulated = kvmppc_handle_store(run, vcpu, val, 8,
+				is_default_endian);
+		if (emulated != EMULATE_DONE)
+			break;
+
+		vcpu->arch.paddr_accessed += run->mmio.len;
+		vcpu->arch.mmio_vmx_copy_nums--;
+	}
+
+	return emulated;
+}
+
+static int kvmppc_emulate_mmio_vmx_loadstore(struct kvm_vcpu *vcpu,
+		struct kvm_run *run)
+{
+	enum emulation_result emulated = EMULATE_FAIL;
+	int r;
+
+	vcpu->arch.paddr_accessed += run->mmio.len;
+
+	if (!vcpu->mmio_is_write) {
+		emulated = kvmppc_handle_load128_by2x64(run, vcpu,
+				vcpu->arch.io_gpr, 1);
+	} else {
+		emulated = kvmppc_handle_store128_by2x64(run, vcpu,
+				vcpu->arch.io_gpr, 1);
+	}
+
+	switch (emulated) {
+	case EMULATE_DO_MMIO:
+		run->exit_reason = KVM_EXIT_MMIO;
+		r = RESUME_HOST;
+		break;
+	case EMULATE_FAIL:
+		pr_info("KVM: MMIO emulation failed (VMX repeat)\n");
+		run->exit_reason = KVM_EXIT_INTERNAL_ERROR;
+		run->internal.suberror = KVM_INTERNAL_ERROR_EMULATION;
+		r = RESUME_HOST;
+		break;
+	default:
+		r = RESUME_GUEST;
+		break;
+	}
+	return r;
+}
+#endif /* CONFIG_ALTIVEC */
+
 int kvm_vcpu_ioctl_get_one_reg(struct kvm_vcpu *vcpu, struct kvm_one_reg *reg)
 {
 	int r = 0;
@@ -1420,6 +1558,18 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 				return r;
 			}
 		}
+#endif
+#ifdef CONFIG_ALTIVEC
+		if (vcpu->arch.mmio_vmx_copy_nums > 0)
+			vcpu->arch.mmio_vmx_copy_nums--;
+
+		if (vcpu->arch.mmio_vmx_copy_nums > 0) {
+			r = kvmppc_emulate_mmio_vmx_loadstore(vcpu, run);
+			if (r == RESUME_HOST) {
+				vcpu->mmio_needed = 1;
+				return r;
+			}
+		}
 #endif
 	} else if (vcpu->arch.osi_needed) {
 		u64 *gprs = run->osi.gprs;