Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

Pull more KVM updates from Paolo Bonzini:
 "The guest side of the asynchronous page fault work has been delayed to
  5.9 in order to sync with Thomas's interrupt entry rework, but here's
  the rest of the KVM updates for this merge window.

  MIPS:
   - Loongson port

  PPC:
   - Fixes

  ARM:
   - Fixes

  x86:
   - KVM_SET_USER_MEMORY_REGION optimizations
   - Fixes
   - Selftest fixes"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (62 commits)
  KVM: x86: do not pass poisoned hva to __kvm_set_memory_region
  KVM: selftests: fix sync_with_host() in smm_test
  KVM: async_pf: Inject 'page ready' event only if 'page not present' was previously injected
  KVM: async_pf: Cleanup kvm_setup_async_pf()
  kvm: i8254: remove redundant assignment to pointer s
  KVM: x86: respect singlestep when emulating instruction
  KVM: selftests: Don't probe KVM_CAP_HYPERV_ENLIGHTENED_VMCS when nested VMX is unsupported
  KVM: selftests: do not substitute SVM/VMX check with KVM_CAP_NESTED_STATE check
  KVM: nVMX: Consult only the "basic" exit reason when routing nested exit
  KVM: arm64: Move hyp_symbol_addr() to kvm_asm.h
  KVM: arm64: Synchronize sysreg state on injecting an AArch32 exception
  KVM: arm64: Make vcpu_cp1x() work on Big Endian hosts
  KVM: arm64: Remove host_cpu_context member from vcpu structure
  KVM: arm64: Stop sparse from moaning at __hyp_this_cpu_ptr
  KVM: arm64: Handle PtrAuth traps early
  KVM: x86: Unexport x86_fpu_cache and make it static
  KVM: selftests: Ignore KVM 5-level paging support for VM_MODE_PXXV48_4K
  KVM: arm64: Save the host's PtrAuth keys in non-preemptible context
  KVM: arm64: Stop save/restoring ACTLR_EL1
  KVM: arm64: Add emulation for 32bit guests accessing ACTLR2
  ...

17 files changed, 1086 insertions, 167 deletions

diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index eb1e86c30d0c..26c63e8161f0 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -1403,6 +1403,7 @@ config CPU_LOONGSON64
 	select MIPS_L1_CACHE_SHIFT_6
 	select GPIOLIB
 	select SWIOTLB
+	select HAVE_KVM
 	help
 		The Loongson GSx64(GS264/GS464/GS464E/GS464V) series of processor
 		cores implements the MIPS64R2 instruction set with many extensions,
diff --git a/arch/mips/include/asm/cpu-features.h b/arch/mips/include/asm/cpu-features.h
index caecbae4b599..724dfddcab92 100644
--- a/arch/mips/include/asm/cpu-features.h
+++ b/arch/mips/include/asm/cpu-features.h
@@ -682,6 +682,9 @@
 #ifndef cpu_guest_has_htw
 #define cpu_guest_has_htw	(cpu_data[0].guest.options & MIPS_CPU_HTW)
 #endif
+#ifndef cpu_guest_has_ldpte
+#define cpu_guest_has_ldpte	(cpu_data[0].guest.options & MIPS_CPU_LDPTE)
+#endif
 #ifndef cpu_guest_has_mvh
 #define cpu_guest_has_mvh	(cpu_data[0].guest.options & MIPS_CPU_MVH)
 #endif
diff --git a/arch/mips/include/asm/kvm_host.h b/arch/mips/include/asm/kvm_host.h
index e28b5a946e26..363e7a89d173 100644
--- a/arch/mips/include/asm/kvm_host.h
+++ b/arch/mips/include/asm/kvm_host.h
@@ -23,6 +23,8 @@
 #include <asm/inst.h>
 #include <asm/mipsregs.h>
 
+#include <kvm/iodev.h>
+
 /* MIPS KVM register ids */
 #define MIPS_CP0_32(_R, _S)					\
 	(KVM_REG_MIPS_CP0 | KVM_REG_SIZE_U32 | (8 * (_R) + (_S)))
@@ -66,9 +68,11 @@
 #define KVM_REG_MIPS_CP0_CONFIG3	MIPS_CP0_32(16, 3)
 #define KVM_REG_MIPS_CP0_CONFIG4	MIPS_CP0_32(16, 4)
 #define KVM_REG_MIPS_CP0_CONFIG5	MIPS_CP0_32(16, 5)
+#define KVM_REG_MIPS_CP0_CONFIG6	MIPS_CP0_32(16, 6)
 #define KVM_REG_MIPS_CP0_CONFIG7	MIPS_CP0_32(16, 7)
 #define KVM_REG_MIPS_CP0_MAARI		MIPS_CP0_64(17, 2)
 #define KVM_REG_MIPS_CP0_XCONTEXT	MIPS_CP0_64(20, 0)
+#define KVM_REG_MIPS_CP0_DIAG		MIPS_CP0_32(22, 0)
 #define KVM_REG_MIPS_CP0_ERROREPC	MIPS_CP0_64(30, 0)
 #define KVM_REG_MIPS_CP0_KSCRATCH1	MIPS_CP0_64(31, 2)
 #define KVM_REG_MIPS_CP0_KSCRATCH2	MIPS_CP0_64(31, 3)
@@ -78,8 +82,8 @@
 #define KVM_REG_MIPS_CP0_KSCRATCH6	MIPS_CP0_64(31, 7)
 
 
-#define KVM_MAX_VCPUS		8
-#define KVM_USER_MEM_SLOTS	8
+#define KVM_MAX_VCPUS		16
+#define KVM_USER_MEM_SLOTS	16
 /* memory slots that does not exposed to userspace */
 #define KVM_PRIVATE_MEM_SLOTS	0
 
@@ -171,6 +175,9 @@ struct kvm_vcpu_stat {
 	u64 vz_ghfc_exits;
 	u64 vz_gpa_exits;
 	u64 vz_resvd_exits;
+#ifdef CONFIG_CPU_LOONGSON64
+	u64 vz_cpucfg_exits;
+#endif
 #endif
 	u64 halt_successful_poll;
 	u64 halt_attempted_poll;
@@ -183,11 +190,39 @@ struct kvm_vcpu_stat {
 struct kvm_arch_memory_slot {
 };
 
+#ifdef CONFIG_CPU_LOONGSON64
+struct ipi_state {
+	uint32_t status;
+	uint32_t en;
+	uint32_t set;
+	uint32_t clear;
+	uint64_t buf[4];
+};
+
+struct loongson_kvm_ipi;
+
+struct ipi_io_device {
+	int node_id;
+	struct loongson_kvm_ipi *ipi;
+	struct kvm_io_device device;
+};
+
+struct loongson_kvm_ipi {
+	spinlock_t lock;
+	struct kvm *kvm;
+	struct ipi_state ipistate[16];
+	struct ipi_io_device dev_ipi[4];
+};
+#endif
+
 struct kvm_arch {
 	/* Guest physical mm */
 	struct mm_struct gpa_mm;
 	/* Mask of CPUs needing GPA ASID flush */
 	cpumask_t asid_flush_mask;
+#ifdef CONFIG_CPU_LOONGSON64
+	struct loongson_kvm_ipi ipi;
+#endif
 };
 
 #define N_MIPS_COPROC_REGS	32
@@ -225,6 +260,7 @@ struct mips_coproc {
 #define MIPS_CP0_WATCH_LO	18
 #define MIPS_CP0_WATCH_HI	19
 #define MIPS_CP0_TLB_XCONTEXT	20
+#define MIPS_CP0_DIAG		22
 #define MIPS_CP0_ECC		26
 #define MIPS_CP0_CACHE_ERR	27
 #define MIPS_CP0_TAG_LO		28
@@ -276,8 +312,12 @@ enum emulation_result {
 #define MIPS3_PG_SHIFT		6
 #define MIPS3_PG_FRAME		0x3fffffc0
 
+#if defined(CONFIG_64BIT)
+#define VPN2_MASK		GENMASK(cpu_vmbits - 1, 13)
+#else
 #define VPN2_MASK		0xffffe000
-#define KVM_ENTRYHI_ASID	MIPS_ENTRYHI_ASID
+#endif
+#define KVM_ENTRYHI_ASID	cpu_asid_mask(&boot_cpu_data)
 #define TLB_IS_GLOBAL(x)	((x).tlb_lo[0] & (x).tlb_lo[1] & ENTRYLO_G)
 #define TLB_VPN2(x)		((x).tlb_hi & VPN2_MASK)
 #define TLB_ASID(x)		((x).tlb_hi & KVM_ENTRYHI_ASID)
@@ -892,6 +932,10 @@ void kvm_vz_save_guesttlb(struct kvm_mips_tlb *buf, unsigned int index,
 			  unsigned int count);
 void kvm_vz_load_guesttlb(const struct kvm_mips_tlb *buf, unsigned int index,
 			  unsigned int count);
+#ifdef CONFIG_CPU_LOONGSON64
+void kvm_loongson_clear_guest_vtlb(void);
+void kvm_loongson_clear_guest_ftlb(void);
+#endif
 #endif
 
 void kvm_mips_suspend_mm(int cpu);
@@ -1131,6 +1175,8 @@ extern int kvm_mips_trans_mtc0(union mips_instruction inst, u32 *opc,
 /* Misc */
 extern void kvm_mips_dump_stats(struct kvm_vcpu *vcpu);
 extern unsigned long kvm_mips_get_ramsize(struct kvm *kvm);
+extern int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu,
+			     struct kvm_mips_interrupt *irq);
 
 static inline void kvm_arch_hardware_unsetup(void) {}
 static inline void kvm_arch_sync_events(struct kvm *kvm) {}
diff --git a/arch/mips/include/asm/mipsregs.h b/arch/mips/include/asm/mipsregs.h
index 796dbb86575b..20d6d40c59a4 100644
--- a/arch/mips/include/asm/mipsregs.h
+++ b/arch/mips/include/asm/mipsregs.h
@@ -1038,6 +1038,8 @@
 /* Disable Branch Return Cache */
 #define R10K_DIAG_D_BRC		(_ULCAST_(1) << 22)
 
+/* Flush BTB */
+#define LOONGSON_DIAG_BTB	(_ULCAST_(1) << 1)
 /* Flush ITLB */
 #define LOONGSON_DIAG_ITLB	(_ULCAST_(1) << 2)
 /* Flush DTLB */
@@ -2874,7 +2876,9 @@ __BUILD_SET_C0(status)
 __BUILD_SET_C0(cause)
 __BUILD_SET_C0(config)
 __BUILD_SET_C0(config5)
+__BUILD_SET_C0(config6)
 __BUILD_SET_C0(config7)
+__BUILD_SET_C0(diag)
 __BUILD_SET_C0(intcontrol)
 __BUILD_SET_C0(intctl)
 __BUILD_SET_C0(srsmap)
diff --git a/arch/mips/include/uapi/asm/inst.h b/arch/mips/include/uapi/asm/inst.h
index 98f97c85e059..43d1faa02933 100644
--- a/arch/mips/include/uapi/asm/inst.h
+++ b/arch/mips/include/uapi/asm/inst.h
@@ -1012,6 +1012,16 @@ struct loongson3_lsdc2_format {	/* Loongson-3 overridden ldc2/sdc2 Load/Store fo
 	;))))))
 };
 
+struct loongson3_lscsr_format {	/* Loongson-3 CPUCFG&CSR read/write format */
+	__BITFIELD_FIELD(unsigned int opcode : 6,
+	__BITFIELD_FIELD(unsigned int rs : 5,
+	__BITFIELD_FIELD(unsigned int fr : 5,
+	__BITFIELD_FIELD(unsigned int rd : 5,
+	__BITFIELD_FIELD(unsigned int fd : 5,
+	__BITFIELD_FIELD(unsigned int func : 6,
+	;))))))
+};
+
 /*
  * MIPS16e instruction formats (16-bit length)
  */
@@ -1114,6 +1124,7 @@ union mips_instruction {
 	struct mm16_r5_format mm16_r5_format;
 	struct loongson3_lswc2_format loongson3_lswc2_format;
 	struct loongson3_lsdc2_format loongson3_lsdc2_format;
+	struct loongson3_lscsr_format loongson3_lscsr_format;
 };
 
 union mips16e_instruction {
diff --git a/arch/mips/kernel/cpu-probe.c b/arch/mips/kernel/cpu-probe.c
index 6b93162d7c5a..def1659fe262 100644
--- a/arch/mips/kernel/cpu-probe.c
+++ b/arch/mips/kernel/cpu-probe.c
@@ -2017,8 +2017,10 @@ static inline void decode_cpucfg(struct cpuinfo_mips *c)
 	if (cfg2 & LOONGSON_CFG2_LEXT2)
 		c->ases |= MIPS_ASE_LOONGSON_EXT2;
 
-	if (cfg2 & LOONGSON_CFG2_LSPW)
+	if (cfg2 & LOONGSON_CFG2_LSPW) {
 		c->options |= MIPS_CPU_LDPTE;
+		c->guest.options |= MIPS_CPU_LDPTE;
+	}
 
 	if (cfg3 & LOONGSON_CFG3_LCAMP)
 		c->ases |= MIPS_ASE_LOONGSON_CAM;
@@ -2074,6 +2076,7 @@ static inline void cpu_probe_loongson(struct cpuinfo_mips *c, unsigned int cpu)
 		c->writecombine = _CACHE_UNCACHED_ACCELERATED;
 		c->ases |= (MIPS_ASE_LOONGSON_MMI | MIPS_ASE_LOONGSON_CAM |
 			MIPS_ASE_LOONGSON_EXT | MIPS_ASE_LOONGSON_EXT2);
+		c->ases &= ~MIPS_ASE_VZ; /* VZ of Loongson-3A2000/3000 is incomplete */
 		break;
 	case PRID_IMP_LOONGSON_64G:
 		c->cputype = CPU_LOONGSON64;
diff --git a/arch/mips/kvm/Kconfig b/arch/mips/kvm/Kconfig
index b91d145aa2d5..d697752a5723 100644
--- a/arch/mips/kvm/Kconfig
+++ b/arch/mips/kvm/Kconfig
@@ -22,6 +22,7 @@ config KVM
 	select EXPORT_UASM
 	select PREEMPT_NOTIFIERS
 	select KVM_GENERIC_DIRTYLOG_READ_PROTECT
+	select HAVE_KVM_EVENTFD
 	select HAVE_KVM_VCPU_ASYNC_IOCTL
 	select KVM_MMIO
 	select MMU_NOTIFIER
diff --git a/arch/mips/kvm/Makefile b/arch/mips/kvm/Makefile
index 01affc1d21c5..506c4ac0ba1c 100644
--- a/arch/mips/kvm/Makefile
+++ b/arch/mips/kvm/Makefile
@@ -2,7 +2,7 @@
 # Makefile for KVM support for MIPS
 #
 
-common-objs-y = $(addprefix ../../../virt/kvm/, kvm_main.o coalesced_mmio.o)
+common-objs-y = $(addprefix ../../../virt/kvm/, kvm_main.o coalesced_mmio.o eventfd.o)
 
 EXTRA_CFLAGS += -Ivirt/kvm -Iarch/mips/kvm
 
@@ -13,6 +13,9 @@ kvm-objs := $(common-objs-y) mips.o emulate.o entry.o \
 	    fpu.o
 kvm-objs += hypcall.o
 kvm-objs += mmu.o
+ifdef CONFIG_CPU_LOONGSON64
+kvm-objs += loongson_ipi.o
+endif
 
 ifdef CONFIG_KVM_MIPS_VZ
 kvm-objs		+= vz.o
diff --git a/arch/mips/kvm/emulate.c b/arch/mips/kvm/emulate.c
index 7ccf9b096783..5ae82d925197 100644
--- a/arch/mips/kvm/emulate.c
+++ b/arch/mips/kvm/emulate.c
@@ -1600,9 +1600,11 @@ enum emulation_result kvm_mips_emulate_store(union mips_instruction inst,
 					     struct kvm_run *run,
 					     struct kvm_vcpu *vcpu)
 {
+	int r;
 	enum emulation_result er;
 	u32 rt;
 	void *data = run->mmio.data;
+	unsigned int imme;
 	unsigned long curr_pc;
 
 	/*
@@ -1660,15 +1662,229 @@ enum emulation_result kvm_mips_emulate_store(union mips_instruction inst,
 			  vcpu->arch.gprs[rt], *(u8 *)data);
 		break;
 
+	case swl_op:
+		run->mmio.phys_addr = kvm_mips_callbacks->gva_to_gpa(
+					vcpu->arch.host_cp0_badvaddr) & (~0x3);
+		run->mmio.len = 4;
+		imme = vcpu->arch.host_cp0_badvaddr & 0x3;
+		switch (imme) {
+		case 0:
+			*(u32 *)data = ((*(u32 *)data) & 0xffffff00) |
+					(vcpu->arch.gprs[rt] >> 24);
+			break;
+		case 1:
+			*(u32 *)data = ((*(u32 *)data) & 0xffff0000) |
+					(vcpu->arch.gprs[rt] >> 16);
+			break;
+		case 2:
+			*(u32 *)data = ((*(u32 *)data) & 0xff000000) |
+					(vcpu->arch.gprs[rt] >> 8);
+			break;
+		case 3:
+			*(u32 *)data = vcpu->arch.gprs[rt];
+			break;
+		default:
+			break;
+		}
+
+		kvm_debug("[%#lx] OP_SWL: eaddr: %#lx, gpr: %#lx, data: %#x\n",
+			  vcpu->arch.pc, vcpu->arch.host_cp0_badvaddr,
+			  vcpu->arch.gprs[rt], *(u32 *)data);
+		break;
+
+	case swr_op:
+		run->mmio.phys_addr = kvm_mips_callbacks->gva_to_gpa(
+					vcpu->arch.host_cp0_badvaddr) & (~0x3);
+		run->mmio.len = 4;
+		imme = vcpu->arch.host_cp0_badvaddr & 0x3;
+		switch (imme) {
+		case 0:
+			*(u32 *)data = vcpu->arch.gprs[rt];
+			break;
+		case 1:
+			*(u32 *)data = ((*(u32 *)data) & 0xff) |
+					(vcpu->arch.gprs[rt] << 8);
+			break;
+		case 2:
+			*(u32 *)data = ((*(u32 *)data) & 0xffff) |
+					(vcpu->arch.gprs[rt] << 16);
+			break;
+		case 3:
+			*(u32 *)data = ((*(u32 *)data) & 0xffffff) |
+					(vcpu->arch.gprs[rt] << 24);
+			break;
+		default:
+			break;
+		}
+
+		kvm_debug("[%#lx] OP_SWR: eaddr: %#lx, gpr: %#lx, data: %#x\n",
+			  vcpu->arch.pc, vcpu->arch.host_cp0_badvaddr,
+			  vcpu->arch.gprs[rt], *(u32 *)data);
+		break;
+
+	case sdl_op:
+		run->mmio.phys_addr = kvm_mips_callbacks->gva_to_gpa(
+					vcpu->arch.host_cp0_badvaddr) & (~0x7);
+
+		run->mmio.len = 8;
+		imme = vcpu->arch.host_cp0_badvaddr & 0x7;
+		switch (imme) {
+		case 0:
+			*(u64 *)data = ((*(u64 *)data) & 0xffffffffffffff00) |
+					((vcpu->arch.gprs[rt] >> 56) & 0xff);
+			break;
+		case 1:
+			*(u64 *)data = ((*(u64 *)data) & 0xffffffffffff0000) |
+					((vcpu->arch.gprs[rt] >> 48) & 0xffff);
+			break;
+		case 2:
+			*(u64 *)data = ((*(u64 *)data) & 0xffffffffff000000) |
+					((vcpu->arch.gprs[rt] >> 40) & 0xffffff);
+			break;
+		case 3:
+			*(u64 *)data = ((*(u64 *)data) & 0xffffffff00000000) |
+					((vcpu->arch.gprs[rt] >> 32) & 0xffffffff);
+			break;
+		case 4:
+			*(u64 *)data = ((*(u64 *)data) & 0xffffff0000000000) |
+					((vcpu->arch.gprs[rt] >> 24) & 0xffffffffff);
+			break;
+		case 5:
+			*(u64 *)data = ((*(u64 *)data) & 0xffff000000000000) |
+					((vcpu->arch.gprs[rt] >> 16) & 0xffffffffffff);
+			break;
+		case 6:
+			*(u64 *)data = ((*(u64 *)data) & 0xff00000000000000) |
+					((vcpu->arch.gprs[rt] >> 8) & 0xffffffffffffff);
+			break;
+		case 7:
+			*(u64 *)data = vcpu->arch.gprs[rt];
+			break;
+		default:
+			break;
+		}
+
+		kvm_debug("[%#lx] OP_SDL: eaddr: %#lx, gpr: %#lx, data: %llx\n",
+			  vcpu->arch.pc, vcpu->arch.host_cp0_badvaddr,
+			  vcpu->arch.gprs[rt], *(u64 *)data);
+		break;
+
+	case sdr_op:
+		run->mmio.phys_addr = kvm_mips_callbacks->gva_to_gpa(
+					vcpu->arch.host_cp0_badvaddr) & (~0x7);
+
+		run->mmio.len = 8;
+		imme = vcpu->arch.host_cp0_badvaddr & 0x7;
+		switch (imme) {
+		case 0:
+			*(u64 *)data = vcpu->arch.gprs[rt];
+			break;
+		case 1:
+			*(u64 *)data = ((*(u64 *)data) & 0xff) |
+					(vcpu->arch.gprs[rt] << 8);
+			break;
+		case 2:
+			*(u64 *)data = ((*(u64 *)data) & 0xffff) |
+					(vcpu->arch.gprs[rt] << 16);
+			break;
+		case 3:
+			*(u64 *)data = ((*(u64 *)data) & 0xffffff) |
+					(vcpu->arch.gprs[rt] << 24);
+			break;
+		case 4:
+			*(u64 *)data = ((*(u64 *)data) & 0xffffffff) |
+					(vcpu->arch.gprs[rt] << 32);
+			break;
+		case 5:
+			*(u64 *)data = ((*(u64 *)data) & 0xffffffffff) |
+					(vcpu->arch.gprs[rt] << 40);
+			break;
+		case 6:
+			*(u64 *)data = ((*(u64 *)data) & 0xffffffffffff) |
+					(vcpu->arch.gprs[rt] << 48);
+			break;
+		case 7:
+			*(u64 *)data = ((*(u64 *)data) & 0xffffffffffffff) |
+					(vcpu->arch.gprs[rt] << 56);
+			break;
+		default:
+			break;
+		}
+
+		kvm_debug("[%#lx] OP_SDR: eaddr: %#lx, gpr: %#lx, data: %llx\n",
+			  vcpu->arch.pc, vcpu->arch.host_cp0_badvaddr,
+			  vcpu->arch.gprs[rt], *(u64 *)data);
+		break;
+
+#ifdef CONFIG_CPU_LOONGSON64
+	case sdc2_op:
+		rt = inst.loongson3_lsdc2_format.rt;
+		switch (inst.loongson3_lsdc2_format.opcode1) {
+		/*
+		 * Loongson-3 overridden sdc2 instructions.
+		 * opcode1              instruction
+		 *   0x0          gssbx: store 1 bytes from GPR
+		 *   0x1          gsshx: store 2 bytes from GPR
+		 *   0x2          gsswx: store 4 bytes from GPR
+		 *   0x3          gssdx: store 8 bytes from GPR
+		 */
+		case 0x0:
+			run->mmio.len = 1;
+			*(u8 *)data = vcpu->arch.gprs[rt];
+
+			kvm_debug("[%#lx] OP_GSSBX: eaddr: %#lx, gpr: %#lx, data: %#x\n",
+				  vcpu->arch.pc, vcpu->arch.host_cp0_badvaddr,
+				  vcpu->arch.gprs[rt], *(u8 *)data);
+			break;
+		case 0x1:
+			run->mmio.len = 2;
+			*(u16 *)data = vcpu->arch.gprs[rt];
+
+			kvm_debug("[%#lx] OP_GSSSHX: eaddr: %#lx, gpr: %#lx, data: %#x\n",
+				  vcpu->arch.pc, vcpu->arch.host_cp0_badvaddr,
+				  vcpu->arch.gprs[rt], *(u16 *)data);
+			break;
+		case 0x2:
+			run->mmio.len = 4;
+			*(u32 *)data = vcpu->arch.gprs[rt];
+
+			kvm_debug("[%#lx] OP_GSSWX: eaddr: %#lx, gpr: %#lx, data: %#x\n",
+				  vcpu->arch.pc, vcpu->arch.host_cp0_badvaddr,
+				  vcpu->arch.gprs[rt], *(u32 *)data);
+			break;
+		case 0x3:
+			run->mmio.len = 8;
+			*(u64 *)data = vcpu->arch.gprs[rt];
+
+			kvm_debug("[%#lx] OP_GSSDX: eaddr: %#lx, gpr: %#lx, data: %#llx\n",
+				  vcpu->arch.pc, vcpu->arch.host_cp0_badvaddr,
+				  vcpu->arch.gprs[rt], *(u64 *)data);
+			break;
+		default:
+			kvm_err("Godson Exteneded GS-Store not yet supported (inst=0x%08x)\n",
+				inst.word);
+			break;
+		}
+		break;
+#endif
 	default:
 		kvm_err("Store not yet supported (inst=0x%08x)\n",
 			inst.word);
 		goto out_fail;
 	}
 
-	run->mmio.is_write = 1;
 	vcpu->mmio_needed = 1;
+	run->mmio.is_write = 1;
 	vcpu->mmio_is_write = 1;
+
+	r = kvm_io_bus_write(vcpu, KVM_MMIO_BUS,
+			run->mmio.phys_addr, run->mmio.len, data);
+
+	if (!r) {
+		vcpu->mmio_needed = 0;
+		return EMULATE_DONE;
+	}
+
 	return EMULATE_DO_MMIO;
 
 out_fail:
@@ -1681,9 +1897,11 @@ enum emulation_result kvm_mips_emulate_load(union mips_instruction inst,
 					    u32 cause, struct kvm_run *run,
 					    struct kvm_vcpu *vcpu)
 {
+	int r;
 	enum emulation_result er;
 	unsigned long curr_pc;
 	u32 op, rt;
+	unsigned int imme;
 
 	rt = inst.i_format.rt;
 	op = inst.i_format.opcode;
@@ -1736,6 +1954,162 @@ enum emulation_result kvm_mips_emulate_load(union mips_instruction inst,
 		run->mmio.len = 1;
 		break;
 
+	case lwl_op:
+		run->mmio.phys_addr = kvm_mips_callbacks->gva_to_gpa(
+					vcpu->arch.host_cp0_badvaddr) & (~0x3);
+
+		run->mmio.len = 4;
+		imme = vcpu->arch.host_cp0_badvaddr & 0x3;
+		switch (imme) {
+		case 0:
+			vcpu->mmio_needed = 3;	/* 1 byte */
+			break;
+		case 1:
+			vcpu->mmio_needed = 4;	/* 2 bytes */
+			break;
+		case 2:
+			vcpu->mmio_needed = 5;	/* 3 bytes */
+			break;
+		case 3:
+			vcpu->mmio_needed = 6;	/* 4 bytes */
+			break;
+		default:
+			break;
+		}
+		break;
+
+	case lwr_op:
+		run->mmio.phys_addr = kvm_mips_callbacks->gva_to_gpa(
+					vcpu->arch.host_cp0_badvaddr) & (~0x3);
+
+		run->mmio.len = 4;
+		imme = vcpu->arch.host_cp0_badvaddr & 0x3;
+		switch (imme) {
+		case 0:
+			vcpu->mmio_needed = 7;	/* 4 bytes */
+			break;
+		case 1:
+			vcpu->mmio_needed = 8;	/* 3 bytes */
+			break;
+		case 2:
+			vcpu->mmio_needed = 9;	/* 2 bytes */
+			break;
+		case 3:
+			vcpu->mmio_needed = 10;	/* 1 byte */
+			break;
+		default:
+			break;
+		}
+		break;
+
+	case ldl_op:
+		run->mmio.phys_addr = kvm_mips_callbacks->gva_to_gpa(
+					vcpu->arch.host_cp0_badvaddr) & (~0x7);
+
+		run->mmio.len = 8;
+		imme = vcpu->arch.host_cp0_badvaddr & 0x7;
+		switch (imme) {
+		case 0:
+			vcpu->mmio_needed = 11;	/* 1 byte */
+			break;
+		case 1:
+			vcpu->mmio_needed = 12;	/* 2 bytes */
+			break;
+		case 2:
+			vcpu->mmio_needed = 13;	/* 3 bytes */
+			break;
+		case 3:
+			vcpu->mmio_needed = 14;	/* 4 bytes */
+			break;
+		case 4:
+			vcpu->mmio_needed = 15;	/* 5 bytes */
+			break;
+		case 5:
+			vcpu->mmio_needed = 16;	/* 6 bytes */
+			break;
+		case 6:
+			vcpu->mmio_needed = 17;	/* 7 bytes */
+			break;
+		case 7:
+			vcpu->mmio_needed = 18;	/* 8 bytes */
+			break;
+		default:
+			break;
+		}
+		break;
+
+	case ldr_op:
+		run->mmio.phys_addr = kvm_mips_callbacks->gva_to_gpa(
+					vcpu->arch.host_cp0_badvaddr) & (~0x7);
+
+		run->mmio.len = 8;
+		imme = vcpu->arch.host_cp0_badvaddr & 0x7;
+		switch (imme) {
+		case 0:
+			vcpu->mmio_needed = 19;	/* 8 bytes */
+			break;
+		case 1:
+			vcpu->mmio_needed = 20;	/* 7 bytes */
+			break;
+		case 2:
+			vcpu->mmio_needed = 21;	/* 6 bytes */
+			break;
+		case 3:
+			vcpu->mmio_needed = 22;	/* 5 bytes */
+			break;
+		case 4:
+			vcpu->mmio_needed = 23;	/* 4 bytes */
+			break;
+		case 5:
+			vcpu->mmio_needed = 24;	/* 3 bytes */
+			break;
+		case 6:
+			vcpu->mmio_needed = 25;	/* 2 bytes */
+			break;
+		case 7:
+			vcpu->mmio_needed = 26;	/* 1 byte */
+			break;
+		default:
+			break;
+		}
+		break;
+
+#ifdef CONFIG_CPU_LOONGSON64
+	case ldc2_op:
+		rt = inst.loongson3_lsdc2_format.rt;
+		switch (inst.loongson3_lsdc2_format.opcode1) {
+		/*
+		 * Loongson-3 overridden ldc2 instructions.
+		 * opcode1              instruction
+		 *   0x0          gslbx: store 1 bytes from GPR
+		 *   0x1          gslhx: store 2 bytes from GPR
+		 *   0x2          gslwx: store 4 bytes from GPR
+		 *   0x3          gsldx: store 8 bytes from GPR
+		 */
+		case 0x0:
+			run->mmio.len = 1;
+			vcpu->mmio_needed = 27;	/* signed */
+			break;
+		case 0x1:
+			run->mmio.len = 2;
+			vcpu->mmio_needed = 28;	/* signed */
+			break;
+		case 0x2:
+			run->mmio.len = 4;
+			vcpu->mmio_needed = 29;	/* signed */
+			break;
+		case 0x3:
+			run->mmio.len = 8;
+			vcpu->mmio_needed = 30;	/* signed */
+			break;
+		default:
+			kvm_err("Godson Exteneded GS-Load for float not yet supported (inst=0x%08x)\n",
+				inst.word);
+			break;
+		}
+		break;
+#endif
+
 	default:
 		kvm_err("Load not yet supported (inst=0x%08x)\n",
 			inst.word);
@@ -1745,6 +2119,16 @@ enum emulation_result kvm_mips_emulate_load(union mips_instruction inst,
 
 	run->mmio.is_write = 0;
 	vcpu->mmio_is_write = 0;
+
+	r = kvm_io_bus_read(vcpu, KVM_MMIO_BUS,
+			run->mmio.phys_addr, run->mmio.len, run->mmio.data);
+
+	if (!r) {
+		kvm_mips_complete_mmio_load(vcpu, run);
+		vcpu->mmio_needed = 0;
+		return EMULATE_DONE;
+	}
+
 	return EMULATE_DO_MMIO;
 }
 
@@ -2591,28 +2975,125 @@ enum emulation_result kvm_mips_complete_mmio_load(struct kvm_vcpu *vcpu,
 
 	switch (run->mmio.len) {
 	case 8:
-		*gpr = *(s64 *)run->mmio.data;
+		switch (vcpu->mmio_needed) {
+		case 11:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffffffffffff) |
+				(((*(s64 *)run->mmio.data) & 0xff) << 56);
+			break;
+		case 12:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffffffffff) |
+				(((*(s64 *)run->mmio.data) & 0xffff) << 48);
+			break;
+		case 13:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffffffff) |
+				(((*(s64 *)run->mmio.data) & 0xffffff) << 40);
+			break;
+		case 14:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffffff) |
+				(((*(s64 *)run->mmio.data) & 0xffffffff) << 32);
+			break;
+		case 15:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffff) |
+				(((*(s64 *)run->mmio.data) & 0xffffffffff) << 24);
+			break;
+		case 16:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffff) |
+				(((*(s64 *)run->mmio.data) & 0xffffffffffff) << 16);
+			break;
+		case 17:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xff) |
+				(((*(s64 *)run->mmio.data) & 0xffffffffffffff) << 8);
+			break;
+		case 18:
+		case 19:
+			*gpr = *(s64 *)run->mmio.data;
+			break;
+		case 20:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xff00000000000000) |
+				((((*(s64 *)run->mmio.data)) >> 8) & 0xffffffffffffff);
+			break;
+		case 21:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffff000000000000) |
+				((((*(s64 *)run->mmio.data)) >> 16) & 0xffffffffffff);
+			break;
+		case 22:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffff0000000000) |
+				((((*(s64 *)run->mmio.data)) >> 24) & 0xffffffffff);
+			break;
+		case 23:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffffff00000000) |
+				((((*(s64 *)run->mmio.data)) >> 32) & 0xffffffff);
+			break;
+		case 24:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffffffff000000) |
+				((((*(s64 *)run->mmio.data)) >> 40) & 0xffffff);
+			break;
+		case 25:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffffffffff0000) |
+				((((*(s64 *)run->mmio.data)) >> 48) & 0xffff);
+			break;
+		case 26:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffffffffffff00) |
+				((((*(s64 *)run->mmio.data)) >> 56) & 0xff);
+			break;
+		default:
+			*gpr = *(s64 *)run->mmio.data;
+		}
 		break;
 
 	case 4:
-		if (vcpu->mmio_needed == 2)
-			*gpr = *(s32 *)run->mmio.data;
-		else
+		switch (vcpu->mmio_needed) {
+		case 1:
 			*gpr = *(u32 *)run->mmio.data;
+			break;
+		case 2:
+			*gpr = *(s32 *)run->mmio.data;
+			break;
+		case 3:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffff) |
+				(((*(s32 *)run->mmio.data) & 0xff) << 24);
+			break;
+		case 4:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffff) |
+				(((*(s32 *)run->mmio.data) & 0xffff) << 16);
+			break;
+		case 5:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xff) |
+				(((*(s32 *)run->mmio.data) & 0xffffff) << 8);
+			break;
+		case 6:
+		case 7:
+			*gpr = *(s32 *)run->mmio.data;
+			break;
+		case 8:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xff000000) |
+				((((*(s32 *)run->mmio.data)) >> 8) & 0xffffff);
+			break;
+		case 9:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffff0000) |
+				((((*(s32 *)run->mmio.data)) >> 16) & 0xffff);
+			break;
+		case 10:
+			*gpr = (vcpu->arch.gprs[vcpu->arch.io_gpr] & 0xffffff00) |
+				((((*(s32 *)run->mmio.data)) >> 24) & 0xff);
+			break;
+		default:
+			*gpr = *(s32 *)run->mmio.data;
+		}
 		break;
 
 	case 2:
-		if (vcpu->mmio_needed == 2)
-			*gpr = *(s16 *) run->mmio.data;
-		else
+		if (vcpu->mmio_needed == 1)
 			*gpr = *(u16 *)run->mmio.data;
+		else
+			*gpr = *(s16 *)run->mmio.data;
 
 		break;
 	case 1:
-		if (vcpu->mmio_needed == 2)
-			*gpr = *(s8 *) run->mmio.data;
+		if (vcpu->mmio_needed == 1)
+			*gpr = *(u8 *)run->mmio.data;
 		else
-			*gpr = *(u8 *) run->mmio.data;
+			*gpr = *(s8 *)run->mmio.data;
 		break;
 	}
 
diff --git a/arch/mips/kvm/entry.c b/arch/mips/kvm/entry.c
index 16e1c93b484f..fd716942e302 100644
--- a/arch/mips/kvm/entry.c
+++ b/arch/mips/kvm/entry.c
@@ -56,6 +56,7 @@
 #define C0_BADVADDR	8, 0
 #define C0_BADINSTR	8, 1
 #define C0_BADINSTRP	8, 2
+#define C0_PGD		9, 7
 #define C0_ENTRYHI	10, 0
 #define C0_GUESTCTL1	10, 4
 #define C0_STATUS	12, 0
@@ -307,7 +308,10 @@ static void *kvm_mips_build_enter_guest(void *addr)
 
 #ifdef CONFIG_KVM_MIPS_VZ
 	/* Save normal linux process pgd (VZ guarantees pgd_reg is set) */
-	UASM_i_MFC0(&p, K0, c0_kscratch(), pgd_reg);
+	if (cpu_has_ldpte)
+		UASM_i_MFC0(&p, K0, C0_PWBASE);
+	else
+		UASM_i_MFC0(&p, K0, c0_kscratch(), pgd_reg);
 	UASM_i_SW(&p, K0, offsetof(struct kvm_vcpu_arch, host_pgd), K1);
 
 	/*
@@ -469,8 +473,10 @@ void *kvm_mips_build_tlb_refill_exception(void *addr, void *handler)
 	u32 *p = addr;
 	struct uasm_label labels[2];
 	struct uasm_reloc relocs[2];
+#ifndef CONFIG_CPU_LOONGSON64
 	struct uasm_label *l = labels;
 	struct uasm_reloc *r = relocs;
+#endif
 
 	memset(labels, 0, sizeof(labels));
 	memset(relocs, 0, sizeof(relocs));
@@ -490,6 +496,16 @@ void *kvm_mips_build_tlb_refill_exception(void *addr, void *handler)
 	 */
 	preempt_disable();
 
+#ifdef CONFIG_CPU_LOONGSON64
+	UASM_i_MFC0(&p, K1, C0_PGD);
+	uasm_i_lddir(&p, K0, K1, 3);  /* global page dir */
+#ifndef __PAGETABLE_PMD_FOLDED
+	uasm_i_lddir(&p, K1, K0, 1);  /* middle page dir */
+#endif
+	uasm_i_ldpte(&p, K1, 0);      /* even */
+	uasm_i_ldpte(&p, K1, 1);      /* odd */
+	uasm_i_tlbwr(&p);
+#else
 	/*
 	 * Now for the actual refill bit. A lot of this can be common with the
 	 * Linux TLB refill handler, however we don't need to handle so many
@@ -512,6 +528,7 @@ void *kvm_mips_build_tlb_refill_exception(void *addr, void *handler)
 	build_get_ptep(&p, K0, K1);
 	build_update_entries(&p, K0, K1);
 	build_tlb_write_entry(&p, &l, &r, tlb_random);
+#endif
 
 	preempt_enable();
 
diff --git a/arch/mips/kvm/interrupt.c b/arch/mips/kvm/interrupt.c
index 7257e8b6f5a9..d28c2c9c343e 100644
--- a/arch/mips/kvm/interrupt.c
+++ b/arch/mips/kvm/interrupt.c
@@ -61,27 +61,8 @@ void kvm_mips_queue_io_int_cb(struct kvm_vcpu *vcpu,
 	 * the EXC code will be set when we are actually
 	 * delivering the interrupt:
 	 */
-	switch (intr) {
-	case 2:
-		kvm_set_c0_guest_cause(vcpu->arch.cop0, (C_IRQ0));
-		/* Queue up an INT exception for the core */
-		kvm_mips_queue_irq(vcpu, MIPS_EXC_INT_IO);
-		break;
-
-	case 3:
-		kvm_set_c0_guest_cause(vcpu->arch.cop0, (C_IRQ1));
-		kvm_mips_queue_irq(vcpu, MIPS_EXC_INT_IPI_1);
-		break;
-
-	case 4:
-		kvm_set_c0_guest_cause(vcpu->arch.cop0, (C_IRQ2));
-		kvm_mips_queue_irq(vcpu, MIPS_EXC_INT_IPI_2);
-		break;
-
-	default:
-		break;
-	}
-
+	kvm_set_c0_guest_cause(vcpu->arch.cop0, 1 << (intr + 8));
+	kvm_mips_queue_irq(vcpu, kvm_irq_to_priority(intr));
 }
 
 void kvm_mips_dequeue_io_int_cb(struct kvm_vcpu *vcpu,
@@ -89,26 +70,8 @@ void kvm_mips_dequeue_io_int_cb(struct kvm_vcpu *vcpu,
 {
 	int intr = (int)irq->irq;
 
-	switch (intr) {
-	case -2:
-		kvm_clear_c0_guest_cause(vcpu->arch.cop0, (C_IRQ0));
-		kvm_mips_dequeue_irq(vcpu, MIPS_EXC_INT_IO);
-		break;
-
-	case -3:
-		kvm_clear_c0_guest_cause(vcpu->arch.cop0, (C_IRQ1));
-		kvm_mips_dequeue_irq(vcpu, MIPS_EXC_INT_IPI_1);
-		break;
-
-	case -4:
-		kvm_clear_c0_guest_cause(vcpu->arch.cop0, (C_IRQ2));
-		kvm_mips_dequeue_irq(vcpu, MIPS_EXC_INT_IPI_2);
-		break;
-
-	default:
-		break;
-	}
-
+	kvm_clear_c0_guest_cause(vcpu->arch.cop0, 1 << (-intr + 8));
+	kvm_mips_dequeue_irq(vcpu, kvm_irq_to_priority(-intr));
 }
 
 /* Deliver the interrupt of the corresponding priority, if possible. */
@@ -116,50 +79,20 @@ int kvm_mips_irq_deliver_cb(struct kvm_vcpu *vcpu, unsigned int priority,
 			    u32 cause)
 {
 	int allowed = 0;
-	u32 exccode;
+	u32 exccode, ie;
 
 	struct kvm_vcpu_arch *arch = &vcpu->arch;
 	struct mips_coproc *cop0 = vcpu->arch.cop0;
 
-	switch (priority) {
-	case MIPS_EXC_INT_TIMER:
-		if ((kvm_read_c0_guest_status(cop0) & ST0_IE)
-		    && (!(kvm_read_c0_guest_status(cop0) & (ST0_EXL | ST0_ERL)))
-		    && (kvm_read_c0_guest_status(cop0) & IE_IRQ5)) {
-			allowed = 1;
-			exccode = EXCCODE_INT;
-		}
-		break;
-
-	case MIPS_EXC_INT_IO:
-		if ((kvm_read_c0_guest_status(cop0) & ST0_IE)
-		    && (!(kvm_read_c0_guest_status(cop0) & (ST0_EXL | ST0_ERL)))
-		    && (kvm_read_c0_guest_status(cop0) & IE_IRQ0)) {
-			allowed = 1;
-			exccode = EXCCODE_INT;
-		}
-		break;
-
-	case MIPS_EXC_INT_IPI_1:
-		if ((kvm_read_c0_guest_status(cop0) & ST0_IE)
-		    && (!(kvm_read_c0_guest_status(cop0) & (ST0_EXL | ST0_ERL)))
-		    && (kvm_read_c0_guest_status(cop0) & IE_IRQ1)) {
-			allowed = 1;
-			exccode = EXCCODE_INT;
-		}
-		break;
-
-	case MIPS_EXC_INT_IPI_2:
-		if ((kvm_read_c0_guest_status(cop0) & ST0_IE)
-		    && (!(kvm_read_c0_guest_status(cop0) & (ST0_EXL | ST0_ERL)))
-		    && (kvm_read_c0_guest_status(cop0) & IE_IRQ2)) {
-			allowed = 1;
-			exccode = EXCCODE_INT;
-		}
-		break;
+	if (priority == MIPS_EXC_MAX)
+		return 0;
 
-	default:
-		break;
+	ie = 1 << (kvm_priority_to_irq[priority] + 8);
+	if ((kvm_read_c0_guest_status(cop0) & ST0_IE)
+	    && (!(kvm_read_c0_guest_status(cop0) & (ST0_EXL | ST0_ERL)))
+	    && (kvm_read_c0_guest_status(cop0) & ie)) {
+		allowed = 1;
+		exccode = EXCCODE_INT;
 	}
 
 	/* Are we allowed to deliver the interrupt ??? */
diff --git a/arch/mips/kvm/interrupt.h b/arch/mips/kvm/interrupt.h
index 3bf0a49725e8..c3e878ca3e07 100644
--- a/arch/mips/kvm/interrupt.h
+++ b/arch/mips/kvm/interrupt.h
@@ -21,11 +21,12 @@
 #define MIPS_EXC_NMI                5
 #define MIPS_EXC_MCHK               6
 #define MIPS_EXC_INT_TIMER          7
-#define MIPS_EXC_INT_IO             8
-#define MIPS_EXC_EXECUTE            9
-#define MIPS_EXC_INT_IPI_1          10
-#define MIPS_EXC_INT_IPI_2          11
-#define MIPS_EXC_MAX                12
+#define MIPS_EXC_INT_IO_1           8
+#define MIPS_EXC_INT_IO_2           9
+#define MIPS_EXC_EXECUTE            10
+#define MIPS_EXC_INT_IPI_1          11
+#define MIPS_EXC_INT_IPI_2          12
+#define MIPS_EXC_MAX                13
 /* XXXSL More to follow */
 
 #define C_TI        (_ULCAST_(1) << 30)
@@ -38,6 +39,9 @@
 #define KVM_MIPS_IRQ_CLEAR_ALL_AT_ONCE   (0)
 #endif
 
+extern u32 *kvm_priority_to_irq;
+u32 kvm_irq_to_priority(u32 irq);
+
 void kvm_mips_queue_irq(struct kvm_vcpu *vcpu, unsigned int priority);
 void kvm_mips_dequeue_irq(struct kvm_vcpu *vcpu, unsigned int priority);
 int kvm_mips_pending_timer(struct kvm_vcpu *vcpu);
diff --git a/arch/mips/kvm/loongson_ipi.c b/arch/mips/kvm/loongson_ipi.c
new file mode 100644
index 000000000000..3681fc8fba38
--- /dev/null
+++ b/arch/mips/kvm/loongson_ipi.c
@@ -0,0 +1,214 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Loongson-3 Virtual IPI interrupt support.
+ *
+ * Copyright (C) 2019  Loongson Technologies, Inc.  All rights reserved.
+ *
+ * Authors: Chen Zhu <zhuchen@loongson.cn>
+ * Authors: Huacai Chen <chenhc@lemote.com>
+ */
+
+#include <linux/kvm_host.h>
+
+#define IPI_BASE            0x3ff01000ULL
+
+#define CORE0_STATUS_OFF       0x000
+#define CORE0_EN_OFF           0x004
+#define CORE0_SET_OFF          0x008
+#define CORE0_CLEAR_OFF        0x00c
+#define CORE0_BUF_20           0x020
+#define CORE0_BUF_28           0x028
+#define CORE0_BUF_30           0x030
+#define CORE0_BUF_38           0x038
+
+#define CORE1_STATUS_OFF       0x100
+#define CORE1_EN_OFF           0x104
+#define CORE1_SET_OFF          0x108
+#define CORE1_CLEAR_OFF        0x10c
+#define CORE1_BUF_20           0x120
+#define CORE1_BUF_28           0x128
+#define CORE1_BUF_30           0x130
+#define CORE1_BUF_38           0x138
+
+#define CORE2_STATUS_OFF       0x200
+#define CORE2_EN_OFF           0x204
+#define CORE2_SET_OFF          0x208
+#define CORE2_CLEAR_OFF        0x20c
+#define CORE2_BUF_20           0x220
+#define CORE2_BUF_28           0x228
+#define CORE2_BUF_30           0x230
+#define CORE2_BUF_38           0x238
+
+#define CORE3_STATUS_OFF       0x300
+#define CORE3_EN_OFF           0x304
+#define CORE3_SET_OFF          0x308
+#define CORE3_CLEAR_OFF        0x30c
+#define CORE3_BUF_20           0x320
+#define CORE3_BUF_28           0x328
+#define CORE3_BUF_30           0x330
+#define CORE3_BUF_38           0x338
+
+static int loongson_vipi_read(struct loongson_kvm_ipi *ipi,
+				gpa_t addr, int len, void *val)
+{
+	uint32_t core = (addr >> 8) & 3;
+	uint32_t node = (addr >> 44) & 3;
+	uint32_t id = core + node * 4;
+	uint64_t offset = addr & 0xff;
+	void *pbuf;
+	struct ipi_state *s = &(ipi->ipistate[id]);
+
+	BUG_ON(offset & (len - 1));
+
+	switch (offset) {
+	case CORE0_STATUS_OFF:
+		*(uint64_t *)val = s->status;
+		break;
+
+	case CORE0_EN_OFF:
+		*(uint64_t *)val = s->en;
+		break;
+
+	case CORE0_SET_OFF:
+		*(uint64_t *)val = 0;
+		break;
+
+	case CORE0_CLEAR_OFF:
+		*(uint64_t *)val = 0;
+		break;
+
+	case CORE0_BUF_20 ... CORE0_BUF_38:
+		pbuf = (void *)s->buf + (offset - 0x20);
+		if (len == 8)
+			*(uint64_t *)val = *(uint64_t *)pbuf;
+		else /* Assume len == 4 */
+			*(uint32_t *)val = *(uint32_t *)pbuf;
+		break;
+
+	default:
+		pr_notice("%s with unknown addr %llx\n", __func__, addr);
+		break;
+	}
+
+	return 0;
+}
+
+static int loongson_vipi_write(struct loongson_kvm_ipi *ipi,
+				gpa_t addr, int len, const void *val)
+{
+	uint32_t core = (addr >> 8) & 3;
+	uint32_t node = (addr >> 44) & 3;
+	uint32_t id = core + node * 4;
+	uint64_t data, offset = addr & 0xff;
+	void *pbuf;
+	struct kvm *kvm = ipi->kvm;
+	struct kvm_mips_interrupt irq;
+	struct ipi_state *s = &(ipi->ipistate[id]);
+
+	data = *(uint64_t *)val;
+	BUG_ON(offset & (len - 1));
+
+	switch (offset) {
+	case CORE0_STATUS_OFF:
+		break;
+
+	case CORE0_EN_OFF:
+		s->en = data;
+		break;
+
+	case CORE0_SET_OFF:
+		s->status |= data;
+		irq.cpu = id;
+		irq.irq = 6;
+		kvm_vcpu_ioctl_interrupt(kvm->vcpus[id], &irq);
+		break;
+
+	case CORE0_CLEAR_OFF:
+		s->status &= ~data;
+		if (!s->status) {
+			irq.cpu = id;
+			irq.irq = -6;
+			kvm_vcpu_ioctl_interrupt(kvm->vcpus[id], &irq);
+		}
+		break;
+
+	case CORE0_BUF_20 ... CORE0_BUF_38:
+		pbuf = (void *)s->buf + (offset - 0x20);
+		if (len == 8)
+			*(uint64_t *)pbuf = (uint64_t)data;
+		else /* Assume len == 4 */
+			*(uint32_t *)pbuf = (uint32_t)data;
+		break;
+
+	default:
+		pr_notice("%s with unknown addr %llx\n", __func__, addr);
+		break;
+	}
+
+	return 0;
+}
+
+static int kvm_ipi_read(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
+			gpa_t addr, int len, void *val)
+{
+	unsigned long flags;
+	struct loongson_kvm_ipi *ipi;
+	struct ipi_io_device *ipi_device;
+
+	ipi_device = container_of(dev, struct ipi_io_device, device);
+	ipi = ipi_device->ipi;
+
+	spin_lock_irqsave(&ipi->lock, flags);
+	loongson_vipi_read(ipi, addr, len, val);
+	spin_unlock_irqrestore(&ipi->lock, flags);
+
+	return 0;
+}
+
+static int kvm_ipi_write(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
+			gpa_t addr, int len, const void *val)
+{
+	unsigned long flags;
+	struct loongson_kvm_ipi *ipi;
+	struct ipi_io_device *ipi_device;
+
+	ipi_device = container_of(dev, struct ipi_io_device, device);
+	ipi = ipi_device->ipi;
+
+	spin_lock_irqsave(&ipi->lock, flags);
+	loongson_vipi_write(ipi, addr, len, val);
+	spin_unlock_irqrestore(&ipi->lock, flags);
+
+	return 0;
+}
+
+static const struct kvm_io_device_ops kvm_ipi_ops = {
+	.read     = kvm_ipi_read,
+	.write    = kvm_ipi_write,
+};
+
+void kvm_init_loongson_ipi(struct kvm *kvm)
+{
+	int i;
+	unsigned long addr;
+	struct loongson_kvm_ipi *s;
+	struct kvm_io_device *device;
+
+	s = &kvm->arch.ipi;
+	s->kvm = kvm;
+	spin_lock_init(&s->lock);
+
+	/*
+	 * Initialize IPI device
+	 */
+	for (i = 0; i < 4; i++) {
+		device = &s->dev_ipi[i].device;
+		kvm_iodevice_init(device, &kvm_ipi_ops);
+		addr = (((unsigned long)i) << 44) + IPI_BASE;
+		mutex_lock(&kvm->slots_lock);
+		kvm_io_bus_register_dev(kvm, KVM_MMIO_BUS, addr, 0x400, device);
+		mutex_unlock(&kvm->slots_lock);
+		s->dev_ipi[i].ipi = s;
+		s->dev_ipi[i].node_id = i;
+	}
+}
diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index 2416fa40b687..521bd5891e84 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -67,6 +67,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 	VCPU_STAT("vz_ghfc", vz_ghfc_exits),
 	VCPU_STAT("vz_gpa", vz_gpa_exits),
 	VCPU_STAT("vz_resvd", vz_resvd_exits),
+	VCPU_STAT("vz_cpucfg", vz_cpucfg_exits),
 #endif
 	VCPU_STAT("halt_successful_poll", halt_successful_poll),
 	VCPU_STAT("halt_attempted_poll", halt_attempted_poll),
@@ -129,6 +130,8 @@ int kvm_arch_check_processor_compat(void *opaque)
 	return 0;
 }
 
+extern void kvm_init_loongson_ipi(struct kvm *kvm);
+
 int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 {
 	switch (type) {
@@ -148,6 +151,10 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 	if (!kvm->arch.gpa_mm.pgd)
 		return -ENOMEM;
 
+#ifdef CONFIG_CPU_LOONGSON64
+	kvm_init_loongson_ipi(kvm);
+#endif
+
 	return 0;
 }
 
@@ -490,7 +497,10 @@ int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu,
 	int intr = (int)irq->irq;
 	struct kvm_vcpu *dvcpu = NULL;
 
-	if (intr == 3 || intr == -3 || intr == 4 || intr == -4)
+	if (intr == kvm_priority_to_irq[MIPS_EXC_INT_IPI_1] ||
+	    intr == kvm_priority_to_irq[MIPS_EXC_INT_IPI_2] ||
+	    intr == (-kvm_priority_to_irq[MIPS_EXC_INT_IPI_1]) ||
+	    intr == (-kvm_priority_to_irq[MIPS_EXC_INT_IPI_2]))
 		kvm_debug("%s: CPU: %d, INTR: %d\n", __func__, irq->cpu,
 			  (int)intr);
 
@@ -499,10 +509,10 @@ int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu,
 	else
 		dvcpu = vcpu->kvm->vcpus[irq->cpu];
 
-	if (intr == 2 || intr == 3 || intr == 4) {
+	if (intr == 2 || intr == 3 || intr == 4 || intr == 6) {
 		kvm_mips_callbacks->queue_io_int(dvcpu, irq);
 
-	} else if (intr == -2 || intr == -3 || intr == -4) {
+	} else if (intr == -2 || intr == -3 || intr == -4 || intr == -6) {
 		kvm_mips_callbacks->dequeue_io_int(dvcpu, irq);
 	} else {
 		kvm_err("%s: invalid interrupt ioctl (%d:%d)\n", __func__,
@@ -1620,6 +1630,34 @@ static struct notifier_block kvm_mips_csr_die_notifier = {
 	.notifier_call = kvm_mips_csr_die_notify,
 };
 
+static u32 kvm_default_priority_to_irq[MIPS_EXC_MAX] = {
+	[MIPS_EXC_INT_TIMER] = C_IRQ5,
+	[MIPS_EXC_INT_IO_1]  = C_IRQ0,
+	[MIPS_EXC_INT_IPI_1] = C_IRQ1,
+	[MIPS_EXC_INT_IPI_2] = C_IRQ2,
+};
+
+static u32 kvm_loongson3_priority_to_irq[MIPS_EXC_MAX] = {
+	[MIPS_EXC_INT_TIMER] = C_IRQ5,
+	[MIPS_EXC_INT_IO_1]  = C_IRQ0,
+	[MIPS_EXC_INT_IO_2]  = C_IRQ1,
+	[MIPS_EXC_INT_IPI_1] = C_IRQ4,
+};
+
+u32 *kvm_priority_to_irq = kvm_default_priority_to_irq;
+
+u32 kvm_irq_to_priority(u32 irq)
+{
+	int i;
+
+	for (i = MIPS_EXC_INT_TIMER; i < MIPS_EXC_MAX; i++) {
+		if (kvm_priority_to_irq[i] == (1 << (irq + 8)))
+			return i;
+	}
+
+	return MIPS_EXC_MAX;
+}
+
 static int __init kvm_mips_init(void)
 {
 	int ret;
@@ -1638,6 +1676,9 @@ static int __init kvm_mips_init(void)
 	if (ret)
 		return ret;
 
+	if (boot_cpu_type() == CPU_LOONGSON64)
+		kvm_priority_to_irq = kvm_loongson3_priority_to_irq;
+
 	register_die_notifier(&kvm_mips_csr_die_notifier);
 
 	return 0;
diff --git a/arch/mips/kvm/tlb.c b/arch/mips/kvm/tlb.c
index fc8aee05d10b..1c1fbce3f566 100644
--- a/arch/mips/kvm/tlb.c
+++ b/arch/mips/kvm/tlb.c
@@ -20,6 +20,7 @@
 
 #include <asm/cpu.h>
 #include <asm/bootinfo.h>
+#include <asm/mipsregs.h>
 #include <asm/mmu_context.h>
 #include <asm/cacheflush.h>
 #include <asm/tlb.h>
@@ -621,6 +622,46 @@ void kvm_vz_load_guesttlb(const struct kvm_mips_tlb *buf, unsigned int index,
 }
 EXPORT_SYMBOL_GPL(kvm_vz_load_guesttlb);
 
+#ifdef CONFIG_CPU_LOONGSON64
+void kvm_loongson_clear_guest_vtlb(void)
+{
+	int idx = read_gc0_index();
+
+	/* Set root GuestID for root probe and write of guest TLB entry */
+	set_root_gid_to_guest_gid();
+
+	write_gc0_index(0);
+	guest_tlbinvf();
+	write_gc0_index(idx);
+
+	clear_root_gid();
+	set_c0_diag(LOONGSON_DIAG_ITLB | LOONGSON_DIAG_DTLB);
+}
+EXPORT_SYMBOL_GPL(kvm_loongson_clear_guest_vtlb);
+
+void kvm_loongson_clear_guest_ftlb(void)
+{
+	int i;
+	int idx = read_gc0_index();
+
+	/* Set root GuestID for root probe and write of guest TLB entry */
+	set_root_gid_to_guest_gid();
+
+	for (i = current_cpu_data.tlbsizevtlb;
+	     i < (current_cpu_data.tlbsizevtlb +
+		     current_cpu_data.tlbsizeftlbsets);
+	     i++) {
+		write_gc0_index(i);
+		guest_tlbinvf();
+	}
+	write_gc0_index(idx);
+
+	clear_root_gid();
+	set_c0_diag(LOONGSON_DIAG_ITLB | LOONGSON_DIAG_DTLB);
+}
+EXPORT_SYMBOL_GPL(kvm_loongson_clear_guest_ftlb);
+#endif
+
 #endif
 
 /**
diff --git a/arch/mips/kvm/trap_emul.c b/arch/mips/kvm/trap_emul.c
index 60763efa723e..34ad0b46e610 100644
--- a/arch/mips/kvm/trap_emul.c
+++ b/arch/mips/kvm/trap_emul.c
@@ -529,6 +529,9 @@ static int kvm_trap_emul_check_extension(struct kvm *kvm, long ext)
 	case KVM_CAP_MIPS_TE:
 		r = 1;
 		break;
+	case KVM_CAP_IOEVENTFD:
+		r = 1;
+		break;
 	default:
 		r = 0;
 		break;
diff --git a/arch/mips/kvm/vz.c b/arch/mips/kvm/vz.c
index 51f51009a53f..d9c462c14163 100644
--- a/arch/mips/kvm/vz.c
+++ b/arch/mips/kvm/vz.c
@@ -29,6 +29,7 @@
 #include <linux/kvm_host.h>
 
 #include "interrupt.h"
+#include "loongson_regs.h"
 
 #include "trace.h"
 
@@ -126,6 +127,11 @@ static inline unsigned int kvm_vz_config5_guest_wrmask(struct kvm_vcpu *vcpu)
 	return mask;
 }
 
+static inline unsigned int kvm_vz_config6_guest_wrmask(struct kvm_vcpu *vcpu)
+{
+	return MIPS_CONF6_LOONGSON_INTIMER | MIPS_CONF6_LOONGSON_EXTIMER;
+}
+
 /*
  * VZ optionally allows these additional Config bits to be written by root:
  * Config:	M, [MT]
@@ -180,6 +186,12 @@ static inline unsigned int kvm_vz_config5_user_wrmask(struct kvm_vcpu *vcpu)
 	return kvm_vz_config5_guest_wrmask(vcpu) | MIPS_CONF5_MRP;
 }
 
+static inline unsigned int kvm_vz_config6_user_wrmask(struct kvm_vcpu *vcpu)
+{
+	return kvm_vz_config6_guest_wrmask(vcpu) |
+		MIPS_CONF6_LOONGSON_SFBEN | MIPS_CONF6_LOONGSON_FTLBDIS;
+}
+
 static gpa_t kvm_vz_gva_to_gpa_cb(gva_t gva)
 {
 	/* VZ guest has already converted gva to gpa */
@@ -225,23 +237,7 @@ static void kvm_vz_queue_io_int_cb(struct kvm_vcpu *vcpu,
 	 * interrupts are asynchronous to vcpu execution therefore defer guest
 	 * cp0 accesses
 	 */
-	switch (intr) {
-	case 2:
-		kvm_vz_queue_irq(vcpu, MIPS_EXC_INT_IO);
-		break;
-
-	case 3:
-		kvm_vz_queue_irq(vcpu, MIPS_EXC_INT_IPI_1);
-		break;
-
-	case 4:
-		kvm_vz_queue_irq(vcpu, MIPS_EXC_INT_IPI_2);
-		break;
-
-	default:
-		break;
-	}
-
+	kvm_vz_queue_irq(vcpu, kvm_irq_to_priority(intr));
 }
 
 static void kvm_vz_dequeue_io_int_cb(struct kvm_vcpu *vcpu,
@@ -253,44 +249,22 @@ static void kvm_vz_dequeue_io_int_cb(struct kvm_vcpu *vcpu,
 	 * interrupts are asynchronous to vcpu execution therefore defer guest
 	 * cp0 accesses
 	 */
-	switch (intr) {
-	case -2:
-		kvm_vz_dequeue_irq(vcpu, MIPS_EXC_INT_IO);
-		break;
-
-	case -3:
-		kvm_vz_dequeue_irq(vcpu, MIPS_EXC_INT_IPI_1);
-		break;
-
-	case -4:
-		kvm_vz_dequeue_irq(vcpu, MIPS_EXC_INT_IPI_2);
-		break;
-
-	default:
-		break;
-	}
-
+	kvm_vz_dequeue_irq(vcpu, kvm_irq_to_priority(-intr));
 }
 
-static u32 kvm_vz_priority_to_irq[MIPS_EXC_MAX] = {
-	[MIPS_EXC_INT_TIMER] = C_IRQ5,
-	[MIPS_EXC_INT_IO]    = C_IRQ0,
-	[MIPS_EXC_INT_IPI_1] = C_IRQ1,
-	[MIPS_EXC_INT_IPI_2] = C_IRQ2,
-};
-
 static int kvm_vz_irq_deliver_cb(struct kvm_vcpu *vcpu, unsigned int priority,
 				 u32 cause)
 {
 	u32 irq = (priority < MIPS_EXC_MAX) ?
-		kvm_vz_priority_to_irq[priority] : 0;
+		kvm_priority_to_irq[priority] : 0;
 
 	switch (priority) {
 	case MIPS_EXC_INT_TIMER:
 		set_gc0_cause(C_TI);
 		break;
 
-	case MIPS_EXC_INT_IO:
+	case MIPS_EXC_INT_IO_1:
+	case MIPS_EXC_INT_IO_2:
 	case MIPS_EXC_INT_IPI_1:
 	case MIPS_EXC_INT_IPI_2:
 		if (cpu_has_guestctl2)
@@ -311,7 +285,7 @@ static int kvm_vz_irq_clear_cb(struct kvm_vcpu *vcpu, unsigned int priority,
 			       u32 cause)
 {
 	u32 irq = (priority < MIPS_EXC_MAX) ?
-		kvm_vz_priority_to_irq[priority] : 0;
+		kvm_priority_to_irq[priority] : 0;
 
 	switch (priority) {
 	case MIPS_EXC_INT_TIMER:
@@ -329,7 +303,8 @@ static int kvm_vz_irq_clear_cb(struct kvm_vcpu *vcpu, unsigned int priority,
 		}
 		break;
 
-	case MIPS_EXC_INT_IO:
+	case MIPS_EXC_INT_IO_1:
+	case MIPS_EXC_INT_IO_2:
 	case MIPS_EXC_INT_IPI_1:
 	case MIPS_EXC_INT_IPI_2:
 		/* Clear GuestCtl2.VIP irq if not using Hardware Clear */
@@ -966,7 +941,8 @@ static enum emulation_result kvm_vz_gpsi_cop0(union mips_instruction inst,
 				    (sel == 2 ||	/* SRSCtl */
 				     sel == 3)) ||	/* SRSMap */
 				   (rd == MIPS_CP0_CONFIG &&
-				    (sel == 7)) ||	/* Config7 */
+				    (sel == 6 ||	/* Config6 */
+				     sel == 7)) ||	/* Config7 */
 				   (rd == MIPS_CP0_LLADDR &&
 				    (sel == 2) &&	/* MAARI */
 				    cpu_guest_has_maar &&
@@ -974,6 +950,11 @@ static enum emulation_result kvm_vz_gpsi_cop0(union mips_instruction inst,
 				   (rd == MIPS_CP0_ERRCTL &&
 				    (sel == 0))) {	/* ErrCtl */
 				val = cop0->reg[rd][sel];
+#ifdef CONFIG_CPU_LOONGSON64
+			} else if (rd == MIPS_CP0_DIAG &&
+				   (sel == 0)) {	/* Diag */
+				val = cop0->reg[rd][sel];
+#endif
 			} else {
 				val = 0;
 				er = EMULATE_FAIL;
@@ -1036,9 +1017,40 @@ static enum emulation_result kvm_vz_gpsi_cop0(union mips_instruction inst,
 				   cpu_guest_has_maar &&
 				   !cpu_guest_has_dyn_maar) {
 				kvm_write_maari(vcpu, val);
+			} else if (rd == MIPS_CP0_CONFIG &&
+				   (sel == 6)) {
+				cop0->reg[rd][sel] = (int)val;
 			} else if (rd == MIPS_CP0_ERRCTL &&
 				   (sel == 0)) {	/* ErrCtl */
 				/* ignore the written value */
+#ifdef CONFIG_CPU_LOONGSON64
+			} else if (rd == MIPS_CP0_DIAG &&
+				   (sel == 0)) {	/* Diag */
+				unsigned long flags;
+
+				local_irq_save(flags);
+				if (val & LOONGSON_DIAG_BTB) {
+					/* Flush BTB */
+					set_c0_diag(LOONGSON_DIAG_BTB);
+				}
+				if (val & LOONGSON_DIAG_ITLB) {
+					/* Flush ITLB */
+					set_c0_diag(LOONGSON_DIAG_ITLB);
+				}
+				if (val & LOONGSON_DIAG_DTLB) {
+					/* Flush DTLB */
+					set_c0_diag(LOONGSON_DIAG_DTLB);
+				}
+				if (val & LOONGSON_DIAG_VTLB) {
+					/* Flush VTLB */
+					kvm_loongson_clear_guest_vtlb();
+				}
+				if (val & LOONGSON_DIAG_FTLB) {
+					/* Flush FTLB */
+					kvm_loongson_clear_guest_ftlb();
+				}
+				local_irq_restore(flags);
+#endif
 			} else {
 				er = EMULATE_FAIL;
 			}
@@ -1129,6 +1141,77 @@ static enum emulation_result kvm_vz_gpsi_cache(union mips_instruction inst,
 	return EMULATE_FAIL;
 }
 
+#ifdef CONFIG_CPU_LOONGSON64
+static enum emulation_result kvm_vz_gpsi_lwc2(union mips_instruction inst,
+					      u32 *opc, u32 cause,
+					      struct kvm_run *run,
+					      struct kvm_vcpu *vcpu)
+{
+	unsigned int rs, rd;
+	unsigned int hostcfg;
+	unsigned long curr_pc;
+	enum emulation_result er = EMULATE_DONE;
+
+	/*
+	 * Update PC and hold onto current PC in case there is
+	 * an error and we want to rollback the PC
+	 */
+	curr_pc = vcpu->arch.pc;
+	er = update_pc(vcpu, cause);
+	if (er == EMULATE_FAIL)
+		return er;
+
+	rs = inst.loongson3_lscsr_format.rs;
+	rd = inst.loongson3_lscsr_format.rd;
+	switch (inst.loongson3_lscsr_format.fr) {
+	case 0x8:  /* Read CPUCFG */
+		++vcpu->stat.vz_cpucfg_exits;
+		hostcfg = read_cpucfg(vcpu->arch.gprs[rs]);
+
+		switch (vcpu->arch.gprs[rs]) {
+		case LOONGSON_CFG0:
+			vcpu->arch.gprs[rd] = 0x14c000;
+			break;
+		case LOONGSON_CFG1:
+			hostcfg &= (LOONGSON_CFG1_FP | LOONGSON_CFG1_MMI |
+				    LOONGSON_CFG1_MSA1 | LOONGSON_CFG1_MSA2 |
+				    LOONGSON_CFG1_SFBP);
+			vcpu->arch.gprs[rd] = hostcfg;
+			break;
+		case LOONGSON_CFG2:
+			hostcfg &= (LOONGSON_CFG2_LEXT1 | LOONGSON_CFG2_LEXT2 |
+				    LOONGSON_CFG2_LEXT3 | LOONGSON_CFG2_LSPW);
+			vcpu->arch.gprs[rd] = hostcfg;
+			break;
+		case LOONGSON_CFG3:
+			vcpu->arch.gprs[rd] = hostcfg;
+			break;
+		default:
+			/* Don't export any other advanced features to guest */
+			vcpu->arch.gprs[rd] = 0;
+			break;
+		}
+		break;
+
+	default:
+		kvm_err("lwc2 emulate not impl %d rs %lx @%lx\n",
+			inst.loongson3_lscsr_format.fr, vcpu->arch.gprs[rs], curr_pc);
+		er = EMULATE_FAIL;
+		break;
+	}
+
+	/* Rollback PC only if emulation was unsuccessful */
+	if (er == EMULATE_FAIL) {
+		kvm_err("[%#lx]%s: unsupported lwc2 instruction 0x%08x 0x%08x\n",
+			curr_pc, __func__, inst.word, inst.loongson3_lscsr_format.fr);
+
+		vcpu->arch.pc = curr_pc;
+	}
+
+	return er;
+}
+#endif
+
 static enum emulation_result kvm_trap_vz_handle_gpsi(u32 cause, u32 *opc,
 						     struct kvm_vcpu *vcpu)
 {
@@ -1158,6 +1241,11 @@ static enum emulation_result kvm_trap_vz_handle_gpsi(u32 cause, u32 *opc,
 		er = kvm_vz_gpsi_cache(inst, opc, cause, run, vcpu);
 		break;
 #endif
+#ifdef CONFIG_CPU_LOONGSON64
+	case lwc2_op:
+		er = kvm_vz_gpsi_lwc2(inst, opc, cause, run, vcpu);
+		break;
+#endif
 	case spec3_op:
 		switch (inst.spec3_format.func) {
 #ifdef CONFIG_CPU_MIPSR6
@@ -1652,6 +1740,7 @@ static u64 kvm_vz_get_one_regs[] = {
 	KVM_REG_MIPS_CP0_CONFIG3,
 	KVM_REG_MIPS_CP0_CONFIG4,
 	KVM_REG_MIPS_CP0_CONFIG5,
+	KVM_REG_MIPS_CP0_CONFIG6,
 #ifdef CONFIG_64BIT
 	KVM_REG_MIPS_CP0_XCONTEXT,
 #endif
@@ -1706,7 +1795,7 @@ static unsigned long kvm_vz_num_regs(struct kvm_vcpu *vcpu)
 		ret += ARRAY_SIZE(kvm_vz_get_one_regs_contextconfig);
 	if (cpu_guest_has_segments)
 		ret += ARRAY_SIZE(kvm_vz_get_one_regs_segments);
-	if (cpu_guest_has_htw)
+	if (cpu_guest_has_htw || cpu_guest_has_ldpte)
 		ret += ARRAY_SIZE(kvm_vz_get_one_regs_htw);
 	if (cpu_guest_has_maar && !cpu_guest_has_dyn_maar)
 		ret += 1 + ARRAY_SIZE(vcpu->arch.maar);
@@ -1755,7 +1844,7 @@ static int kvm_vz_copy_reg_indices(struct kvm_vcpu *vcpu, u64 __user *indices)
 			return -EFAULT;
 		indices += ARRAY_SIZE(kvm_vz_get_one_regs_segments);
 	}
-	if (cpu_guest_has_htw) {
+	if (cpu_guest_has_htw || cpu_guest_has_ldpte) {
 		if (copy_to_user(indices, kvm_vz_get_one_regs_htw,
 				 sizeof(kvm_vz_get_one_regs_htw)))
 			return -EFAULT;
@@ -1878,17 +1967,17 @@ static int kvm_vz_get_one_reg(struct kvm_vcpu *vcpu,
 		*v = read_gc0_segctl2();
 		break;
 	case KVM_REG_MIPS_CP0_PWBASE:
-		if (!cpu_guest_has_htw)
+		if (!cpu_guest_has_htw && !cpu_guest_has_ldpte)
 			return -EINVAL;
 		*v = read_gc0_pwbase();
 		break;
 	case KVM_REG_MIPS_CP0_PWFIELD:
-		if (!cpu_guest_has_htw)
+		if (!cpu_guest_has_htw && !cpu_guest_has_ldpte)
 			return -EINVAL;
 		*v = read_gc0_pwfield();
 		break;
 	case KVM_REG_MIPS_CP0_PWSIZE:
-		if (!cpu_guest_has_htw)
+		if (!cpu_guest_has_htw && !cpu_guest_has_ldpte)
 			return -EINVAL;
 		*v = read_gc0_pwsize();
 		break;
@@ -1896,7 +1985,7 @@ static int kvm_vz_get_one_reg(struct kvm_vcpu *vcpu,
 		*v = (long)read_gc0_wired();
 		break;
 	case KVM_REG_MIPS_CP0_PWCTL:
-		if (!cpu_guest_has_htw)
+		if (!cpu_guest_has_htw && !cpu_guest_has_ldpte)
 			return -EINVAL;
 		*v = read_gc0_pwctl();
 		break;
@@ -1979,6 +2068,9 @@ static int kvm_vz_get_one_reg(struct kvm_vcpu *vcpu,
 			return -EINVAL;
 		*v = read_gc0_config5();
 		break;
+	case KVM_REG_MIPS_CP0_CONFIG6:
+		*v = kvm_read_sw_gc0_config6(cop0);
+		break;
 	case KVM_REG_MIPS_CP0_MAAR(0) ... KVM_REG_MIPS_CP0_MAAR(0x3f):
 		if (!cpu_guest_has_maar || cpu_guest_has_dyn_maar)
 			return -EINVAL;
@@ -2101,17 +2193,17 @@ static int kvm_vz_set_one_reg(struct kvm_vcpu *vcpu,
 		write_gc0_segctl2(v);
 		break;
 	case KVM_REG_MIPS_CP0_PWBASE:
-		if (!cpu_guest_has_htw)
+		if (!cpu_guest_has_htw && !cpu_guest_has_ldpte)
 			return -EINVAL;
 		write_gc0_pwbase(v);
 		break;
 	case KVM_REG_MIPS_CP0_PWFIELD:
-		if (!cpu_guest_has_htw)
+		if (!cpu_guest_has_htw && !cpu_guest_has_ldpte)
 			return -EINVAL;
 		write_gc0_pwfield(v);
 		break;
 	case KVM_REG_MIPS_CP0_PWSIZE:
-		if (!cpu_guest_has_htw)
+		if (!cpu_guest_has_htw && !cpu_guest_has_ldpte)
 			return -EINVAL;
 		write_gc0_pwsize(v);
 		break;
@@ -2119,7 +2211,7 @@ static int kvm_vz_set_one_reg(struct kvm_vcpu *vcpu,
 		change_gc0_wired(MIPSR6_WIRED_WIRED, v);
 		break;
 	case KVM_REG_MIPS_CP0_PWCTL:
-		if (!cpu_guest_has_htw)
+		if (!cpu_guest_has_htw && !cpu_guest_has_ldpte)
 			return -EINVAL;
 		write_gc0_pwctl(v);
 		break;
@@ -2248,6 +2340,14 @@ static int kvm_vz_set_one_reg(struct kvm_vcpu *vcpu,
 			write_gc0_config5(v);
 		}
 		break;
+	case KVM_REG_MIPS_CP0_CONFIG6:
+		cur = kvm_read_sw_gc0_config6(cop0);
+		change = (cur ^ v) & kvm_vz_config6_user_wrmask(vcpu);
+		if (change) {
+			v = cur ^ change;
+			kvm_write_sw_gc0_config6(cop0, (int)v);
+		}
+		break;
 	case KVM_REG_MIPS_CP0_MAAR(0) ... KVM_REG_MIPS_CP0_MAAR(0x3f):
 		if (!cpu_guest_has_maar || cpu_guest_has_dyn_maar)
 			return -EINVAL;
@@ -2580,7 +2680,7 @@ static int kvm_vz_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 	}
 
 	/* restore HTW registers */
-	if (cpu_guest_has_htw) {
+	if (cpu_guest_has_htw || cpu_guest_has_ldpte) {
 		kvm_restore_gc0_pwbase(cop0);
 		kvm_restore_gc0_pwfield(cop0);
 		kvm_restore_gc0_pwsize(cop0);
@@ -2597,7 +2697,7 @@ static int kvm_vz_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 	 * prevents a SC on the next VCPU from succeeding by matching a LL on
 	 * the previous VCPU.
 	 */
-	if (cpu_guest_has_rw_llb)
+	if (vcpu->kvm->created_vcpus > 1)
 		write_gc0_lladdr(0);
 
 	return 0;
@@ -2685,8 +2785,8 @@ static int kvm_vz_vcpu_put(struct kvm_vcpu *vcpu, int cpu)
 	}
 
 	/* save HTW registers if enabled in guest */
-	if (cpu_guest_has_htw &&
-	    kvm_read_sw_gc0_config3(cop0) & MIPS_CONF3_PW) {
+	if (cpu_guest_has_ldpte || (cpu_guest_has_htw &&
+	    kvm_read_sw_gc0_config3(cop0) & MIPS_CONF3_PW)) {
 		kvm_save_gc0_pwbase(cop0);
 		kvm_save_gc0_pwfield(cop0);
 		kvm_save_gc0_pwsize(cop0);
@@ -2853,8 +2953,12 @@ static int kvm_vz_hardware_enable(void)
 	write_c0_guestctl0(MIPS_GCTL0_CP0 |
 			   (MIPS_GCTL0_AT_GUEST << MIPS_GCTL0_AT_SHIFT) |
 			   MIPS_GCTL0_CG | MIPS_GCTL0_CF);
-	if (cpu_has_guestctl0ext)
-		set_c0_guestctl0ext(MIPS_GCTL0EXT_CGI);
+	if (cpu_has_guestctl0ext) {
+		if (current_cpu_type() != CPU_LOONGSON64)
+			set_c0_guestctl0ext(MIPS_GCTL0EXT_CGI);
+		else
+			clear_c0_guestctl0ext(MIPS_GCTL0EXT_CGI);
+	}
 
 	if (cpu_has_guestid) {
 		write_c0_guestctl1(0);
@@ -2871,6 +2975,12 @@ static int kvm_vz_hardware_enable(void)
 	if (cpu_has_guestctl2)
 		clear_c0_guestctl2(0x3f << 10);
 
+#ifdef CONFIG_CPU_LOONGSON64
+	/* Control guest CCA attribute */
+	if (cpu_has_csr())
+		csr_writel(csr_readl(0xffffffec) | 0x1, 0xffffffec);
+#endif
+
 	return 0;
 }
 
@@ -2927,6 +3037,9 @@ static int kvm_vz_check_extension(struct kvm *kvm, long ext)
 		r = 2;
 		break;
 #endif
+	case KVM_CAP_IOEVENTFD:
+		r = 1;
+		break;
 	default:
 		r = 0;
 		break;