[MIPS] Import MIPS target support

From v0.12.5 tag at git://git.sv.gnu.org/qemu.git
CommitID: 174f225e9d62e8f3002e274e4f718bd2a967fbf4

Change-Id: I35b49a4319cee4b69cf9da4e5af1f43327e21056
Signed-off-by: Chris Dearman <chris@mips.com>

10 files changed, 14250 insertions, 0 deletions

diff --git a/target-mips/TODO b/target-mips/TODO
new file mode 100644
index 0000000..4769e2a
--- /dev/null
+++ b/target-mips/TODO
@@ -0,0 +1,53 @@
+Unsolved issues/bugs in the mips/mipsel backend
+-----------------------------------------------
+
+General
+-------
+- Unimplemented ASEs:
+  - MIPS16
+  - MDMX
+  - SmartMIPS
+  - DSP r1
+  - DSP r2
+- MT ASE only partially implemented and not functional
+- Shadow register support only partially implemented,
+  lacks set switching on interrupt/exception.
+- 34K ITC not implemented.
+- A general lack of documentation, especially for technical internals.
+  Existing documentation is x86-centric.
+- Reverse endianness bit not implemented
+- The TLB emulation is very inefficient:
+  Qemu's softmmu implements a x86-style MMU, with separate entries
+  for read/write/execute, a TLB index which is just a modulo of the
+  virtual address, and a set of TLBs for each user/kernel/supervisor
+  MMU mode.
+  MIPS has a single entry for read/write/execute and only one MMU mode.
+  But it is fully associative with randomized entry indices, and uses
+  up to 256 ASID tags as additional matching criterion (which roughly
+  equates to 256 MMU modes). It also has a global flag which causes
+  entries to match regardless of ASID.
+  To cope with these differences, Qemu currently flushes the TLB at
+  each ASID change. Using the MMU modes to implement ASIDs hinges on
+  implementing the global bit efficiently.
+- save/restore of the CPU state is not implemented (see machine.c).
+
+MIPS64
+------
+- Userland emulation (both n32 and n64) not functional.
+
+"Generic" 4Kc system emulation
+------------------------------
+- Doesn't correspond to any real hardware. Should be removed some day,
+  U-Boot is the last remaining user.
+
+PICA 61 system emulation
+------------------------
+- No framebuffer support yet.
+
+MALTA system emulation
+----------------------
+- We fake firmware support instead of doing the real thing
+- Real firmware (YAMON) falls over when trying to init RAM, presumably
+  due to lacking system controller emulation.
+- Bonito system controller not implemented
+- MSC1 system controller not implemented
diff --git a/target-mips/cpu.h b/target-mips/cpu.h
new file mode 100644
index 0000000..89c28f1
--- /dev/null
+++ b/target-mips/cpu.h
@@ -0,0 +1,616 @@
+#if !defined (__MIPS_CPU_H__)
+#define __MIPS_CPU_H__
+
+#define TARGET_HAS_ICE 1
+
+#define ELF_MACHINE	EM_MIPS
+
+#define CPUState struct CPUMIPSState
+
+#include "config.h"
+#include "mips-defs.h"
+#include "cpu-defs.h"
+#include "softfloat.h"
+
+// uint_fast8_t and uint_fast16_t not in <sys/int_types.h>
+// XXX: move that elsewhere
+#if defined(CONFIG_SOLARIS) && CONFIG_SOLARIS_VERSION < 10
+typedef unsigned char           uint_fast8_t;
+typedef unsigned int            uint_fast16_t;
+#endif
+
+struct CPUMIPSState;
+
+typedef struct r4k_tlb_t r4k_tlb_t;
+struct r4k_tlb_t {
+    target_ulong VPN;
+    uint32_t PageMask;
+    uint_fast8_t ASID;
+    uint_fast16_t G:1;
+    uint_fast16_t C0:3;
+    uint_fast16_t C1:3;
+    uint_fast16_t V0:1;
+    uint_fast16_t V1:1;
+    uint_fast16_t D0:1;
+    uint_fast16_t D1:1;
+    target_ulong PFN[2];
+};
+
+typedef struct CPUMIPSTLBContext CPUMIPSTLBContext;
+struct CPUMIPSTLBContext {
+    uint32_t nb_tlb;
+    uint32_t tlb_in_use;
+    int (*map_address) (struct CPUMIPSState *env, target_phys_addr_t *physical, int *prot, target_ulong address, int rw, int access_type);
+    void (*helper_tlbwi) (void);
+    void (*helper_tlbwr) (void);
+    void (*helper_tlbp) (void);
+    void (*helper_tlbr) (void);
+    union {
+        struct {
+            r4k_tlb_t tlb[MIPS_TLB_MAX];
+        } r4k;
+    } mmu;
+};
+
+typedef union fpr_t fpr_t;
+union fpr_t {
+    float64  fd;   /* ieee double precision */
+    float32  fs[2];/* ieee single precision */
+    uint64_t d;    /* binary double fixed-point */
+    uint32_t w[2]; /* binary single fixed-point */
+};
+/* define FP_ENDIAN_IDX to access the same location
+ * in the fpr_t union regardless of the host endianess
+ */
+#if defined(HOST_WORDS_BIGENDIAN)
+#  define FP_ENDIAN_IDX 1
+#else
+#  define FP_ENDIAN_IDX 0
+#endif
+
+typedef struct CPUMIPSFPUContext CPUMIPSFPUContext;
+struct CPUMIPSFPUContext {
+    /* Floating point registers */
+    fpr_t fpr[32];
+    float_status fp_status;
+    /* fpu implementation/revision register (fir) */
+    uint32_t fcr0;
+#define FCR0_F64 22
+#define FCR0_L 21
+#define FCR0_W 20
+#define FCR0_3D 19
+#define FCR0_PS 18
+#define FCR0_D 17
+#define FCR0_S 16
+#define FCR0_PRID 8
+#define FCR0_REV 0
+    /* fcsr */
+    uint32_t fcr31;
+#define SET_FP_COND(num,env)     do { ((env).fcr31) |= ((num) ? (1 << ((num) + 24)) : (1 << 23)); } while(0)
+#define CLEAR_FP_COND(num,env)   do { ((env).fcr31) &= ~((num) ? (1 << ((num) + 24)) : (1 << 23)); } while(0)
+#define GET_FP_COND(env)         ((((env).fcr31 >> 24) & 0xfe) | (((env).fcr31 >> 23) & 0x1))
+#define GET_FP_CAUSE(reg)        (((reg) >> 12) & 0x3f)
+#define GET_FP_ENABLE(reg)       (((reg) >>  7) & 0x1f)
+#define GET_FP_FLAGS(reg)        (((reg) >>  2) & 0x1f)
+#define SET_FP_CAUSE(reg,v)      do { (reg) = ((reg) & ~(0x3f << 12)) | ((v & 0x3f) << 12); } while(0)
+#define SET_FP_ENABLE(reg,v)     do { (reg) = ((reg) & ~(0x1f <<  7)) | ((v & 0x1f) << 7); } while(0)
+#define SET_FP_FLAGS(reg,v)      do { (reg) = ((reg) & ~(0x1f <<  2)) | ((v & 0x1f) << 2); } while(0)
+#define UPDATE_FP_FLAGS(reg,v)   do { (reg) |= ((v & 0x1f) << 2); } while(0)
+#define FP_INEXACT        1
+#define FP_UNDERFLOW      2
+#define FP_OVERFLOW       4
+#define FP_DIV0           8
+#define FP_INVALID        16
+#define FP_UNIMPLEMENTED  32
+};
+
+#define NB_MMU_MODES 3
+
+typedef struct CPUMIPSMVPContext CPUMIPSMVPContext;
+struct CPUMIPSMVPContext {
+    int32_t CP0_MVPControl;
+#define CP0MVPCo_CPA	3
+#define CP0MVPCo_STLB	2
+#define CP0MVPCo_VPC	1
+#define CP0MVPCo_EVP	0
+    int32_t CP0_MVPConf0;
+#define CP0MVPC0_M	31
+#define CP0MVPC0_TLBS	29
+#define CP0MVPC0_GS	28
+#define CP0MVPC0_PCP	27
+#define CP0MVPC0_PTLBE	16
+#define CP0MVPC0_TCA	15
+#define CP0MVPC0_PVPE	10
+#define CP0MVPC0_PTC	0
+    int32_t CP0_MVPConf1;
+#define CP0MVPC1_CIM	31
+#define CP0MVPC1_CIF	30
+#define CP0MVPC1_PCX	20
+#define CP0MVPC1_PCP2	10
+#define CP0MVPC1_PCP1	0
+};
+
+typedef struct mips_def_t mips_def_t;
+
+#define MIPS_SHADOW_SET_MAX 16
+#define MIPS_TC_MAX 5
+#define MIPS_FPU_MAX 1
+#define MIPS_DSP_ACC 4
+
+typedef struct TCState TCState;
+struct TCState {
+    target_ulong gpr[32];
+    target_ulong PC;
+    target_ulong HI[MIPS_DSP_ACC];
+    target_ulong LO[MIPS_DSP_ACC];
+    target_ulong ACX[MIPS_DSP_ACC];
+    target_ulong DSPControl;
+    int32_t CP0_TCStatus;
+#define CP0TCSt_TCU3	31
+#define CP0TCSt_TCU2	30
+#define CP0TCSt_TCU1	29
+#define CP0TCSt_TCU0	28
+#define CP0TCSt_TMX	27
+#define CP0TCSt_RNST	23
+#define CP0TCSt_TDS	21
+#define CP0TCSt_DT	20
+#define CP0TCSt_DA	15
+#define CP0TCSt_A	13
+#define CP0TCSt_TKSU	11
+#define CP0TCSt_IXMT	10
+#define CP0TCSt_TASID	0
+    int32_t CP0_TCBind;
+#define CP0TCBd_CurTC	21
+#define CP0TCBd_TBE	17
+#define CP0TCBd_CurVPE	0
+    target_ulong CP0_TCHalt;
+    target_ulong CP0_TCContext;
+    target_ulong CP0_TCSchedule;
+    target_ulong CP0_TCScheFBack;
+    int32_t CP0_Debug_tcstatus;
+};
+
+typedef struct CPUMIPSState CPUMIPSState;
+struct CPUMIPSState {
+    TCState active_tc;
+    CPUMIPSFPUContext active_fpu;
+
+    uint32_t current_tc;
+    uint32_t current_fpu;
+
+    uint32_t SEGBITS;
+    uint32_t PABITS;
+    target_ulong SEGMask;
+    target_ulong PAMask;
+
+    int32_t CP0_Index;
+    /* CP0_MVP* are per MVP registers. */
+    int32_t CP0_Random;
+    int32_t CP0_VPEControl;
+#define CP0VPECo_YSI	21
+#define CP0VPECo_GSI	20
+#define CP0VPECo_EXCPT	16
+#define CP0VPECo_TE	15
+#define CP0VPECo_TargTC	0
+    int32_t CP0_VPEConf0;
+#define CP0VPEC0_M	31
+#define CP0VPEC0_XTC	21
+#define CP0VPEC0_TCS	19
+#define CP0VPEC0_SCS	18
+#define CP0VPEC0_DSC	17
+#define CP0VPEC0_ICS	16
+#define CP0VPEC0_MVP	1
+#define CP0VPEC0_VPA	0
+    int32_t CP0_VPEConf1;
+#define CP0VPEC1_NCX	20
+#define CP0VPEC1_NCP2	10
+#define CP0VPEC1_NCP1	0
+    target_ulong CP0_YQMask;
+    target_ulong CP0_VPESchedule;
+    target_ulong CP0_VPEScheFBack;
+    int32_t CP0_VPEOpt;
+#define CP0VPEOpt_IWX7	15
+#define CP0VPEOpt_IWX6	14
+#define CP0VPEOpt_IWX5	13
+#define CP0VPEOpt_IWX4	12
+#define CP0VPEOpt_IWX3	11
+#define CP0VPEOpt_IWX2	10
+#define CP0VPEOpt_IWX1	9
+#define CP0VPEOpt_IWX0	8
+#define CP0VPEOpt_DWX7	7
+#define CP0VPEOpt_DWX6	6
+#define CP0VPEOpt_DWX5	5
+#define CP0VPEOpt_DWX4	4
+#define CP0VPEOpt_DWX3	3
+#define CP0VPEOpt_DWX2	2
+#define CP0VPEOpt_DWX1	1
+#define CP0VPEOpt_DWX0	0
+    target_ulong CP0_EntryLo0;
+    target_ulong CP0_EntryLo1;
+    target_ulong CP0_Context;
+    int32_t CP0_PageMask;
+    int32_t CP0_PageGrain;
+    int32_t CP0_Wired;
+    int32_t CP0_SRSConf0_rw_bitmask;
+    int32_t CP0_SRSConf0;
+#define CP0SRSC0_M	31
+#define CP0SRSC0_SRS3	20
+#define CP0SRSC0_SRS2	10
+#define CP0SRSC0_SRS1	0
+    int32_t CP0_SRSConf1_rw_bitmask;
+    int32_t CP0_SRSConf1;
+#define CP0SRSC1_M	31
+#define CP0SRSC1_SRS6	20
+#define CP0SRSC1_SRS5	10
+#define CP0SRSC1_SRS4	0
+    int32_t CP0_SRSConf2_rw_bitmask;
+    int32_t CP0_SRSConf2;
+#define CP0SRSC2_M	31
+#define CP0SRSC2_SRS9	20
+#define CP0SRSC2_SRS8	10
+#define CP0SRSC2_SRS7	0
+    int32_t CP0_SRSConf3_rw_bitmask;
+    int32_t CP0_SRSConf3;
+#define CP0SRSC3_M	31
+#define CP0SRSC3_SRS12	20
+#define CP0SRSC3_SRS11	10
+#define CP0SRSC3_SRS10	0
+    int32_t CP0_SRSConf4_rw_bitmask;
+    int32_t CP0_SRSConf4;
+#define CP0SRSC4_SRS15	20
+#define CP0SRSC4_SRS14	10
+#define CP0SRSC4_SRS13	0
+    int32_t CP0_HWREna;
+    target_ulong CP0_BadVAddr;
+    int32_t CP0_Count;
+    target_ulong CP0_EntryHi;
+    int32_t CP0_Compare;
+    int32_t CP0_Status;
+#define CP0St_CU3   31
+#define CP0St_CU2   30
+#define CP0St_CU1   29
+#define CP0St_CU0   28
+#define CP0St_RP    27
+#define CP0St_FR    26
+#define CP0St_RE    25
+#define CP0St_MX    24
+#define CP0St_PX    23
+#define CP0St_BEV   22
+#define CP0St_TS    21
+#define CP0St_SR    20
+#define CP0St_NMI   19
+#define CP0St_IM    8
+#define CP0St_KX    7
+#define CP0St_SX    6
+#define CP0St_UX    5
+#define CP0St_KSU   3
+#define CP0St_ERL   2
+#define CP0St_EXL   1
+#define CP0St_IE    0
+    int32_t CP0_IntCtl;
+#define CP0IntCtl_IPTI 29
+#define CP0IntCtl_IPPC1 26
+#define CP0IntCtl_VS 5
+    int32_t CP0_SRSCtl;
+#define CP0SRSCtl_HSS 26
+#define CP0SRSCtl_EICSS 18
+#define CP0SRSCtl_ESS 12
+#define CP0SRSCtl_PSS 6
+#define CP0SRSCtl_CSS 0
+    int32_t CP0_SRSMap;
+#define CP0SRSMap_SSV7 28
+#define CP0SRSMap_SSV6 24
+#define CP0SRSMap_SSV5 20
+#define CP0SRSMap_SSV4 16
+#define CP0SRSMap_SSV3 12
+#define CP0SRSMap_SSV2 8
+#define CP0SRSMap_SSV1 4
+#define CP0SRSMap_SSV0 0
+    int32_t CP0_Cause;
+#define CP0Ca_BD   31
+#define CP0Ca_TI   30
+#define CP0Ca_CE   28
+#define CP0Ca_DC   27
+#define CP0Ca_PCI  26
+#define CP0Ca_IV   23
+#define CP0Ca_WP   22
+#define CP0Ca_IP    8
+#define CP0Ca_IP_mask 0x0000FF00
+#define CP0Ca_EC    2
+    target_ulong CP0_EPC;
+    int32_t CP0_PRid;
+    int32_t CP0_EBase;
+    int32_t CP0_Config0;
+#define CP0C0_M    31
+#define CP0C0_K23  28
+#define CP0C0_KU   25
+#define CP0C0_MDU  20
+#define CP0C0_MM   17
+#define CP0C0_BM   16
+#define CP0C0_BE   15
+#define CP0C0_AT   13
+#define CP0C0_AR   10
+#define CP0C0_MT   7
+#define CP0C0_VI   3
+#define CP0C0_K0   0
+    int32_t CP0_Config1;
+#define CP0C1_M    31
+#define CP0C1_MMU  25
+#define CP0C1_IS   22
+#define CP0C1_IL   19
+#define CP0C1_IA   16
+#define CP0C1_DS   13
+#define CP0C1_DL   10
+#define CP0C1_DA   7
+#define CP0C1_C2   6
+#define CP0C1_MD   5
+#define CP0C1_PC   4
+#define CP0C1_WR   3
+#define CP0C1_CA   2
+#define CP0C1_EP   1
+#define CP0C1_FP   0
+    int32_t CP0_Config2;
+#define CP0C2_M    31
+#define CP0C2_TU   28
+#define CP0C2_TS   24
+#define CP0C2_TL   20
+#define CP0C2_TA   16
+#define CP0C2_SU   12
+#define CP0C2_SS   8
+#define CP0C2_SL   4
+#define CP0C2_SA   0
+    int32_t CP0_Config3;
+#define CP0C3_M    31
+#define CP0C3_DSPP 10
+#define CP0C3_LPA  7
+#define CP0C3_VEIC 6
+#define CP0C3_VInt 5
+#define CP0C3_SP   4
+#define CP0C3_MT   2
+#define CP0C3_SM   1
+#define CP0C3_TL   0
+    int32_t CP0_Config6;
+    int32_t CP0_Config7;
+    /* XXX: Maybe make LLAddr per-TC? */
+    target_ulong lladdr;
+    target_ulong llval;
+    target_ulong llnewval;
+    target_ulong llreg;
+    target_ulong CP0_LLAddr_rw_bitmask;
+    int CP0_LLAddr_shift;
+    target_ulong CP0_WatchLo[8];
+    int32_t CP0_WatchHi[8];
+    target_ulong CP0_XContext;
+    int32_t CP0_Framemask;
+    int32_t CP0_Debug;
+#define CP0DB_DBD  31
+#define CP0DB_DM   30
+#define CP0DB_LSNM 28
+#define CP0DB_Doze 27
+#define CP0DB_Halt 26
+#define CP0DB_CNT  25
+#define CP0DB_IBEP 24
+#define CP0DB_DBEP 21
+#define CP0DB_IEXI 20
+#define CP0DB_VER  15
+#define CP0DB_DEC  10
+#define CP0DB_SSt  8
+#define CP0DB_DINT 5
+#define CP0DB_DIB  4
+#define CP0DB_DDBS 3
+#define CP0DB_DDBL 2
+#define CP0DB_DBp  1
+#define CP0DB_DSS  0
+    target_ulong CP0_DEPC;
+    int32_t CP0_Performance0;
+    int32_t CP0_TagLo;
+    int32_t CP0_DataLo;
+    int32_t CP0_TagHi;
+    int32_t CP0_DataHi;
+    target_ulong CP0_ErrorEPC;
+    int32_t CP0_DESAVE;
+    /* We waste some space so we can handle shadow registers like TCs. */
+    TCState tcs[MIPS_SHADOW_SET_MAX];
+    CPUMIPSFPUContext fpus[MIPS_FPU_MAX];
+    /* Qemu */
+    int error_code;
+    uint32_t hflags;    /* CPU State */
+    /* TMASK defines different execution modes */
+#define MIPS_HFLAG_TMASK  0x03FF
+#define MIPS_HFLAG_MODE   0x0007 /* execution modes                    */
+    /* The KSU flags must be the lowest bits in hflags. The flag order
+       must be the same as defined for CP0 Status. This allows to use
+       the bits as the value of mmu_idx. */
+#define MIPS_HFLAG_KSU    0x0003 /* kernel/supervisor/user mode mask   */
+#define MIPS_HFLAG_UM       0x0002 /* user mode flag */
+#define MIPS_HFLAG_SM       0x0001 /* supervisor mode flag */
+#define MIPS_HFLAG_KM       0x0000 /* kernel mode flag */
+#define MIPS_HFLAG_DM     0x0004 /* Debug mode                         */
+#define MIPS_HFLAG_64     0x0008 /* 64-bit instructions enabled        */
+#define MIPS_HFLAG_CP0    0x0010 /* CP0 enabled                        */
+#define MIPS_HFLAG_FPU    0x0020 /* FPU enabled                        */
+#define MIPS_HFLAG_F64    0x0040 /* 64-bit FPU enabled                 */
+    /* True if the MIPS IV COP1X instructions can be used.  This also
+       controls the non-COP1X instructions RECIP.S, RECIP.D, RSQRT.S
+       and RSQRT.D.  */
+#define MIPS_HFLAG_COP1X  0x0080 /* COP1X instructions enabled         */
+#define MIPS_HFLAG_RE     0x0100 /* Reversed endianness                */
+#define MIPS_HFLAG_UX     0x0200 /* 64-bit user mode                   */
+    /* If translation is interrupted between the branch instruction and
+     * the delay slot, record what type of branch it is so that we can
+     * resume translation properly.  It might be possible to reduce
+     * this from three bits to two.  */
+#define MIPS_HFLAG_BMASK  0x1C00
+#define MIPS_HFLAG_B      0x0400 /* Unconditional branch               */
+#define MIPS_HFLAG_BC     0x0800 /* Conditional branch                 */
+#define MIPS_HFLAG_BL     0x0C00 /* Likely branch                      */
+#define MIPS_HFLAG_BR     0x1000 /* branch to register (can't link TB) */
+    target_ulong btarget;        /* Jump / branch target               */
+    target_ulong bcond;          /* Branch condition (if needed)       */
+
+    int SYNCI_Step; /* Address step size for SYNCI */
+    int CCRes; /* Cycle count resolution/divisor */
+    uint32_t CP0_Status_rw_bitmask; /* Read/write bits in CP0_Status */
+    uint32_t CP0_TCStatus_rw_bitmask; /* Read/write bits in CP0_TCStatus */
+    int insn_flags; /* Supported instruction set */
+
+    target_ulong tls_value; /* For usermode emulation */
+
+    CPU_COMMON
+
+    CPUMIPSMVPContext *mvp;
+    CPUMIPSTLBContext *tlb;
+
+    const mips_def_t *cpu_model;
+    void *irq[8];
+    struct QEMUTimer *timer; /* Internal timer */
+};
+
+int no_mmu_map_address (CPUMIPSState *env, target_phys_addr_t *physical, int *prot,
+                        target_ulong address, int rw, int access_type);
+int fixed_mmu_map_address (CPUMIPSState *env, target_phys_addr_t *physical, int *prot,
+                           target_ulong address, int rw, int access_type);
+int r4k_map_address (CPUMIPSState *env, target_phys_addr_t *physical, int *prot,
+                     target_ulong address, int rw, int access_type);
+void r4k_helper_tlbwi (void);
+void r4k_helper_tlbwr (void);
+void r4k_helper_tlbp (void);
+void r4k_helper_tlbr (void);
+void mips_cpu_list (FILE *f, int (*cpu_fprintf)(FILE *f, const char *fmt, ...));
+
+void do_unassigned_access(target_phys_addr_t addr, int is_write, int is_exec,
+                          int unused, int size);
+
+#define cpu_init cpu_mips_init
+#define cpu_exec cpu_mips_exec
+#define cpu_gen_code cpu_mips_gen_code
+#define cpu_signal_handler cpu_mips_signal_handler
+#define cpu_list mips_cpu_list
+
+#define CPU_SAVE_VERSION 3
+
+/* MMU modes definitions. We carefully match the indices with our
+   hflags layout. */
+#define MMU_MODE0_SUFFIX _kernel
+#define MMU_MODE1_SUFFIX _super
+#define MMU_MODE2_SUFFIX _user
+#define MMU_USER_IDX 2
+static inline int cpu_mmu_index (CPUState *env)
+{
+    return env->hflags & MIPS_HFLAG_KSU;
+}
+
+static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
+{
+    if (newsp)
+        env->active_tc.gpr[29] = newsp;
+    env->active_tc.gpr[7] = 0;
+    env->active_tc.gpr[2] = 0;
+}
+
+#include "cpu-all.h"
+#include "exec-all.h"
+
+/* Memory access type :
+ * may be needed for precise access rights control and precise exceptions.
+ */
+enum {
+    /* 1 bit to define user level / supervisor access */
+    ACCESS_USER  = 0x00,
+    ACCESS_SUPER = 0x01,
+    /* 1 bit to indicate direction */
+    ACCESS_STORE = 0x02,
+    /* Type of instruction that generated the access */
+    ACCESS_CODE  = 0x10, /* Code fetch access                */
+    ACCESS_INT   = 0x20, /* Integer load/store access        */
+    ACCESS_FLOAT = 0x30, /* floating point load/store access */
+};
+
+/* Exceptions */
+enum {
+    EXCP_NONE          = -1,
+    EXCP_RESET         = 0,
+    EXCP_SRESET,
+    EXCP_DSS,
+    EXCP_DINT,
+    EXCP_DDBL,
+    EXCP_DDBS,
+    EXCP_NMI,
+    EXCP_MCHECK,
+    EXCP_EXT_INTERRUPT, /* 8 */
+    EXCP_DFWATCH,
+    EXCP_DIB,
+    EXCP_IWATCH,
+    EXCP_AdEL,
+    EXCP_AdES,
+    EXCP_TLBF,
+    EXCP_IBE,
+    EXCP_DBp, /* 16 */
+    EXCP_SYSCALL,
+    EXCP_BREAK,
+    EXCP_CpU,
+    EXCP_RI,
+    EXCP_OVERFLOW,
+    EXCP_TRAP,
+    EXCP_FPE,
+    EXCP_DWATCH, /* 24 */
+    EXCP_LTLBL,
+    EXCP_TLBL,
+    EXCP_TLBS,
+    EXCP_DBE,
+    EXCP_THREAD,
+    EXCP_MDMX,
+    EXCP_C2E,
+    EXCP_CACHE, /* 32 */
+
+    EXCP_LAST = EXCP_CACHE,
+};
+/* Dummy exception for conditional stores.  */
+#define EXCP_SC 0x100
+
+int cpu_mips_exec(CPUMIPSState *s);
+CPUMIPSState *cpu_mips_init(const char *cpu_model);
+//~ uint32_t cpu_mips_get_clock (void);
+int cpu_mips_signal_handler(int host_signum, void *pinfo, void *puc);
+
+/* mips_timer.c */
+uint32_t cpu_mips_get_random (CPUState *env);
+uint32_t cpu_mips_get_count (CPUState *env);
+void cpu_mips_store_count (CPUState *env, uint32_t value);
+void cpu_mips_store_compare (CPUState *env, uint32_t value);
+void cpu_mips_start_count(CPUState *env);
+void cpu_mips_stop_count(CPUState *env);
+
+/* mips_int.c */
+void cpu_mips_update_irq (CPUState *env);
+
+/* helper.c */
+int cpu_mips_handle_mmu_fault (CPUState *env, target_ulong address, int rw,
+                               int mmu_idx, int is_softmmu);
+#define cpu_handle_mmu_fault cpu_mips_handle_mmu_fault
+void do_interrupt (CPUState *env);
+void r4k_invalidate_tlb (CPUState *env, int idx, int use_extra);
+target_phys_addr_t cpu_mips_translate_address (CPUState *env, target_ulong address,
+		                               int rw);
+
+static inline void cpu_pc_from_tb(CPUState *env, TranslationBlock *tb)
+{
+    env->active_tc.PC = tb->pc;
+    env->hflags &= ~MIPS_HFLAG_BMASK;
+    env->hflags |= tb->flags & MIPS_HFLAG_BMASK;
+}
+
+static inline void cpu_get_tb_cpu_state(CPUState *env, target_ulong *pc,
+                                        target_ulong *cs_base, int *flags)
+{
+    *pc = env->active_tc.PC;
+    *cs_base = 0;
+    *flags = env->hflags & (MIPS_HFLAG_TMASK | MIPS_HFLAG_BMASK);
+}
+
+static inline void cpu_set_tls(CPUState *env, target_ulong newtls)
+{
+    env->tls_value = newtls;
+}
+
+#endif /* !defined (__MIPS_CPU_H__) */
diff --git a/target-mips/exec.h b/target-mips/exec.h
new file mode 100644
index 0000000..8a118bb
--- /dev/null
+++ b/target-mips/exec.h
@@ -0,0 +1,95 @@
+#if !defined(__QEMU_MIPS_EXEC_H__)
+#define __QEMU_MIPS_EXEC_H__
+
+//#define DEBUG_OP
+
+#include "config.h"
+#include "mips-defs.h"
+#include "dyngen-exec.h"
+#include "cpu-defs.h"
+
+register struct CPUMIPSState *env asm(AREG0);
+
+#include "cpu.h"
+#include "exec-all.h"
+
+#if !defined(CONFIG_USER_ONLY)
+#include "softmmu_exec.h"
+#endif /* !defined(CONFIG_USER_ONLY) */
+
+void dump_fpu(CPUState *env);
+void fpu_dump_state(CPUState *env, FILE *f,
+                    int (*fpu_fprintf)(FILE *f, const char *fmt, ...),
+                    int flags);
+
+void cpu_mips_clock_init (CPUState *env);
+void cpu_mips_tlb_flush (CPUState *env, int flush_global);
+
+static inline void env_to_regs(void)
+{
+}
+
+static inline void regs_to_env(void)
+{
+}
+
+static inline int cpu_has_work(CPUState *env)
+{
+    return (env->interrupt_request &
+            (CPU_INTERRUPT_HARD | CPU_INTERRUPT_TIMER));
+}
+
+
+static inline int cpu_halted(CPUState *env)
+{
+    if (!env->halted)
+        return 0;
+    if (cpu_has_work(env)) {
+        env->halted = 0;
+        return 0;
+    }
+    return EXCP_HALTED;
+}
+
+static inline void compute_hflags(CPUState *env)
+{
+    env->hflags &= ~(MIPS_HFLAG_COP1X | MIPS_HFLAG_64 | MIPS_HFLAG_CP0 |
+                     MIPS_HFLAG_F64 | MIPS_HFLAG_FPU | MIPS_HFLAG_KSU |
+                     MIPS_HFLAG_UX);
+    if (!(env->CP0_Status & (1 << CP0St_EXL)) &&
+        !(env->CP0_Status & (1 << CP0St_ERL)) &&
+        !(env->hflags & MIPS_HFLAG_DM)) {
+        env->hflags |= (env->CP0_Status >> CP0St_KSU) & MIPS_HFLAG_KSU;
+    }
+#if defined(TARGET_MIPS64)
+    if (((env->hflags & MIPS_HFLAG_KSU) != MIPS_HFLAG_UM) ||
+        (env->CP0_Status & (1 << CP0St_PX)) ||
+        (env->CP0_Status & (1 << CP0St_UX)))
+        env->hflags |= MIPS_HFLAG_64;
+    if (env->CP0_Status & (1 << CP0St_UX))
+        env->hflags |= MIPS_HFLAG_UX;
+#endif
+    if ((env->CP0_Status & (1 << CP0St_CU0)) ||
+        !(env->hflags & MIPS_HFLAG_KSU))
+        env->hflags |= MIPS_HFLAG_CP0;
+    if (env->CP0_Status & (1 << CP0St_CU1))
+        env->hflags |= MIPS_HFLAG_FPU;
+    if (env->CP0_Status & (1 << CP0St_FR))
+        env->hflags |= MIPS_HFLAG_F64;
+    if (env->insn_flags & ISA_MIPS32R2) {
+        if (env->active_fpu.fcr0 & (1 << FCR0_F64))
+            env->hflags |= MIPS_HFLAG_COP1X;
+    } else if (env->insn_flags & ISA_MIPS32) {
+        if (env->hflags & MIPS_HFLAG_64)
+            env->hflags |= MIPS_HFLAG_COP1X;
+    } else if (env->insn_flags & ISA_MIPS4) {
+        /* All supported MIPS IV CPUs use the XX (CU3) to enable
+           and disable the MIPS IV extensions to the MIPS III ISA.
+           Some other MIPS IV CPUs ignore the bit, so the check here
+           would be too restrictive for them.  */
+        if (env->CP0_Status & (1 << CP0St_CU3))
+            env->hflags |= MIPS_HFLAG_COP1X;
+    }
+}
+
+#endif /* !defined(__QEMU_MIPS_EXEC_H__) */
diff --git a/target-mips/helper.c b/target-mips/helper.c
new file mode 100644
index 0000000..ec87114
--- /dev/null
+++ b/target-mips/helper.c
@@ -0,0 +1,647 @@
+/*
+ *  MIPS emulation helpers for qemu.
+ *
+ *  Copyright (c) 2004-2005 Jocelyn Mayer
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+#include <stdarg.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <inttypes.h>
+#include <signal.h>
+
+#include "cpu.h"
+#include "exec-all.h"
+
+enum {
+    TLBRET_DIRTY = -4,
+    TLBRET_INVALID = -3,
+    TLBRET_NOMATCH = -2,
+    TLBRET_BADADDR = -1,
+    TLBRET_MATCH = 0
+};
+
+/* no MMU emulation */
+int no_mmu_map_address (CPUState *env, target_phys_addr_t *physical, int *prot,
+                        target_ulong address, int rw, int access_type)
+{
+    *physical = address;
+    *prot = PAGE_READ | PAGE_WRITE;
+    return TLBRET_MATCH;
+}
+
+/* fixed mapping MMU emulation */
+int fixed_mmu_map_address (CPUState *env, target_phys_addr_t *physical, int *prot,
+                           target_ulong address, int rw, int access_type)
+{
+    if (address <= (int32_t)0x7FFFFFFFUL) {
+        if (!(env->CP0_Status & (1 << CP0St_ERL)))
+            *physical = address + 0x40000000UL;
+        else
+            *physical = address;
+    } else if (address <= (int32_t)0xBFFFFFFFUL)
+        *physical = address & 0x1FFFFFFF;
+    else
+        *physical = address;
+
+    *prot = PAGE_READ | PAGE_WRITE;
+    return TLBRET_MATCH;
+}
+
+/* MIPS32/MIPS64 R4000-style MMU emulation */
+int r4k_map_address (CPUState *env, target_phys_addr_t *physical, int *prot,
+                     target_ulong address, int rw, int access_type)
+{
+    uint8_t ASID = env->CP0_EntryHi & 0xFF;
+    int i;
+
+    for (i = 0; i < env->tlb->tlb_in_use; i++) {
+        r4k_tlb_t *tlb = &env->tlb->mmu.r4k.tlb[i];
+        /* 1k pages are not supported. */
+        target_ulong mask = tlb->PageMask | ~(TARGET_PAGE_MASK << 1);
+        target_ulong tag = address & ~mask;
+        target_ulong VPN = tlb->VPN & ~mask;
+#if defined(TARGET_MIPS64)
+        tag &= env->SEGMask;
+#endif
+
+        /* Check ASID, virtual page number & size */
+        if ((tlb->G == 1 || tlb->ASID == ASID) && VPN == tag) {
+            /* TLB match */
+            int n = !!(address & mask & ~(mask >> 1));
+            /* Check access rights */
+            if (!(n ? tlb->V1 : tlb->V0))
+                return TLBRET_INVALID;
+            if (rw == 0 || (n ? tlb->D1 : tlb->D0)) {
+                *physical = tlb->PFN[n] | (address & (mask >> 1));
+                *prot = PAGE_READ;
+                if (n ? tlb->D1 : tlb->D0)
+                    *prot |= PAGE_WRITE;
+                return TLBRET_MATCH;
+            }
+            return TLBRET_DIRTY;
+        }
+    }
+    return TLBRET_NOMATCH;
+}
+
+#if !defined(CONFIG_USER_ONLY)
+static int get_physical_address (CPUState *env, target_phys_addr_t *physical,
+                                int *prot, target_ulong address,
+                                int rw, int access_type)
+{
+    /* User mode can only access useg/xuseg */
+    int user_mode = (env->hflags & MIPS_HFLAG_MODE) == MIPS_HFLAG_UM;
+    int supervisor_mode = (env->hflags & MIPS_HFLAG_MODE) == MIPS_HFLAG_SM;
+    int kernel_mode = !user_mode && !supervisor_mode;
+#if defined(TARGET_MIPS64)
+    int UX = (env->CP0_Status & (1 << CP0St_UX)) != 0;
+    int SX = (env->CP0_Status & (1 << CP0St_SX)) != 0;
+    int KX = (env->CP0_Status & (1 << CP0St_KX)) != 0;
+#endif
+    int ret = TLBRET_MATCH;
+
+#if 0
+    qemu_log("user mode %d h %08x\n", user_mode, env->hflags);
+#endif
+
+    if (address <= (int32_t)0x7FFFFFFFUL) {
+        /* useg */
+        if (env->CP0_Status & (1 << CP0St_ERL)) {
+            *physical = address & 0xFFFFFFFF;
+            *prot = PAGE_READ | PAGE_WRITE;
+        } else {
+            ret = env->tlb->map_address(env, physical, prot, address, rw, access_type);
+        }
+#if defined(TARGET_MIPS64)
+    } else if (address < 0x4000000000000000ULL) {
+        /* xuseg */
+        if (UX && address <= (0x3FFFFFFFFFFFFFFFULL & env->SEGMask)) {
+            ret = env->tlb->map_address(env, physical, prot, address, rw, access_type);
+        } else {
+            ret = TLBRET_BADADDR;
+        }
+    } else if (address < 0x8000000000000000ULL) {
+        /* xsseg */
+        if ((supervisor_mode || kernel_mode) &&
+            SX && address <= (0x7FFFFFFFFFFFFFFFULL & env->SEGMask)) {
+            ret = env->tlb->map_address(env, physical, prot, address, rw, access_type);
+        } else {
+            ret = TLBRET_BADADDR;
+        }
+    } else if (address < 0xC000000000000000ULL) {
+        /* xkphys */
+        if (kernel_mode && KX &&
+            (address & 0x07FFFFFFFFFFFFFFULL) <= env->PAMask) {
+            *physical = address & env->PAMask;
+            *prot = PAGE_READ | PAGE_WRITE;
+        } else {
+            ret = TLBRET_BADADDR;
+        }
+    } else if (address < 0xFFFFFFFF80000000ULL) {
+        /* xkseg */
+        if (kernel_mode && KX &&
+            address <= (0xFFFFFFFF7FFFFFFFULL & env->SEGMask)) {
+            ret = env->tlb->map_address(env, physical, prot, address, rw, access_type);
+        } else {
+            ret = TLBRET_BADADDR;
+        }
+#endif
+    } else if (address < (int32_t)0xA0000000UL) {
+        /* kseg0 */
+        if (kernel_mode) {
+            *physical = address - (int32_t)0x80000000UL;
+            *prot = PAGE_READ | PAGE_WRITE;
+        } else {
+            ret = TLBRET_BADADDR;
+        }
+    } else if (address < (int32_t)0xC0000000UL) {
+        /* kseg1 */
+        if (kernel_mode) {
+            *physical = address - (int32_t)0xA0000000UL;
+            *prot = PAGE_READ | PAGE_WRITE;
+        } else {
+            ret = TLBRET_BADADDR;
+        }
+    } else if (address < (int32_t)0xE0000000UL) {
+        /* sseg (kseg2) */
+        if (supervisor_mode || kernel_mode) {
+            ret = env->tlb->map_address(env, physical, prot, address, rw, access_type);
+        } else {
+            ret = TLBRET_BADADDR;
+        }
+    } else {
+        /* kseg3 */
+        /* XXX: debug segment is not emulated */
+        if (kernel_mode) {
+            ret = env->tlb->map_address(env, physical, prot, address, rw, access_type);
+        } else {
+            ret = TLBRET_BADADDR;
+        }
+    }
+#if 0
+    qemu_log(TARGET_FMT_lx " %d %d => " TARGET_FMT_lx " %d (%d)\n",
+            address, rw, access_type, *physical, *prot, ret);
+#endif
+
+    return ret;
+}
+#endif
+
+static void raise_mmu_exception(CPUState *env, target_ulong address,
+                                int rw, int tlb_error)
+{
+    int exception = 0, error_code = 0;
+
+    switch (tlb_error) {
+    default:
+    case TLBRET_BADADDR:
+        /* Reference to kernel address from user mode or supervisor mode */
+        /* Reference to supervisor address from user mode */
+        if (rw)
+            exception = EXCP_AdES;
+        else
+            exception = EXCP_AdEL;
+        break;
+    case TLBRET_NOMATCH:
+        /* No TLB match for a mapped address */
+        if (rw)
+            exception = EXCP_TLBS;
+        else
+            exception = EXCP_TLBL;
+        error_code = 1;
+        break;
+    case TLBRET_INVALID:
+        /* TLB match with no valid bit */
+        if (rw)
+            exception = EXCP_TLBS;
+        else
+            exception = EXCP_TLBL;
+        break;
+    case TLBRET_DIRTY:
+        /* TLB match but 'D' bit is cleared */
+        exception = EXCP_LTLBL;
+        break;
+
+    }
+    /* Raise exception */
+    env->CP0_BadVAddr = address;
+    env->CP0_Context = (env->CP0_Context & ~0x007fffff) |
+                       ((address >> 9) & 0x007ffff0);
+    env->CP0_EntryHi =
+        (env->CP0_EntryHi & 0xFF) | (address & (TARGET_PAGE_MASK << 1));
+#if defined(TARGET_MIPS64)
+    env->CP0_EntryHi &= env->SEGMask;
+    env->CP0_XContext = (env->CP0_XContext & ((~0ULL) << (env->SEGBITS - 7))) |
+                        ((address & 0xC00000000000ULL) >> (55 - env->SEGBITS)) |
+                        ((address & ((1ULL << env->SEGBITS) - 1) & 0xFFFFFFFFFFFFE000ULL) >> 9);
+#endif
+    env->exception_index = exception;
+    env->error_code = error_code;
+}
+
+target_phys_addr_t cpu_get_phys_page_debug(CPUState *env, target_ulong addr)
+{
+#if defined(CONFIG_USER_ONLY)
+    return addr;
+#else
+    target_phys_addr_t phys_addr;
+    int prot;
+
+    if (get_physical_address(env, &phys_addr, &prot, addr, 0, ACCESS_INT) != 0)
+        return -1;
+    return phys_addr;
+#endif
+}
+
+int cpu_mips_handle_mmu_fault (CPUState *env, target_ulong address, int rw,
+                               int mmu_idx, int is_softmmu)
+{
+#if !defined(CONFIG_USER_ONLY)
+    target_phys_addr_t physical;
+    int prot;
+#endif
+    int access_type;
+    int ret = 0;
+
+#if 0
+    log_cpu_state(env, 0);
+#endif
+    qemu_log("%s pc " TARGET_FMT_lx " ad " TARGET_FMT_lx " rw %d mmu_idx %d smmu %d\n",
+              __func__, env->active_tc.PC, address, rw, mmu_idx, is_softmmu);
+
+    rw &= 1;
+
+    /* data access */
+    /* XXX: put correct access by using cpu_restore_state()
+       correctly */
+    access_type = ACCESS_INT;
+#if defined(CONFIG_USER_ONLY)
+    ret = TLBRET_NOMATCH;
+#else
+    ret = get_physical_address(env, &physical, &prot,
+                               address, rw, access_type);
+    qemu_log("%s address=" TARGET_FMT_lx " ret %d physical " TARGET_FMT_plx " prot %d\n",
+              __func__, address, ret, physical, prot);
+    if (ret == TLBRET_MATCH) {
+       ret = tlb_set_page(env, address & TARGET_PAGE_MASK,
+                          physical & TARGET_PAGE_MASK, prot,
+                          mmu_idx, is_softmmu);
+    } else if (ret < 0)
+#endif
+    {
+        raise_mmu_exception(env, address, rw, ret);
+        ret = 1;
+    }
+
+    return ret;
+}
+
+#if !defined(CONFIG_USER_ONLY)
+target_phys_addr_t cpu_mips_translate_address(CPUState *env, target_ulong address, int rw)
+{
+    target_phys_addr_t physical;
+    int prot;
+    int access_type;
+    int ret = 0;
+
+    rw &= 1;
+
+    /* data access */
+    access_type = ACCESS_INT;
+    ret = get_physical_address(env, &physical, &prot,
+                               address, rw, access_type);
+    if (ret != TLBRET_MATCH) {
+        raise_mmu_exception(env, address, rw, ret);
+        return -1LL;
+    } else {
+        return physical;
+    }
+}
+#endif
+
+static const char * const excp_names[EXCP_LAST + 1] = {
+    [EXCP_RESET] = "reset",
+    [EXCP_SRESET] = "soft reset",
+    [EXCP_DSS] = "debug single step",
+    [EXCP_DINT] = "debug interrupt",
+    [EXCP_NMI] = "non-maskable interrupt",
+    [EXCP_MCHECK] = "machine check",
+    [EXCP_EXT_INTERRUPT] = "interrupt",
+    [EXCP_DFWATCH] = "deferred watchpoint",
+    [EXCP_DIB] = "debug instruction breakpoint",
+    [EXCP_IWATCH] = "instruction fetch watchpoint",
+    [EXCP_AdEL] = "address error load",
+    [EXCP_AdES] = "address error store",
+    [EXCP_TLBF] = "TLB refill",
+    [EXCP_IBE] = "instruction bus error",
+    [EXCP_DBp] = "debug breakpoint",
+    [EXCP_SYSCALL] = "syscall",
+    [EXCP_BREAK] = "break",
+    [EXCP_CpU] = "coprocessor unusable",
+    [EXCP_RI] = "reserved instruction",
+    [EXCP_OVERFLOW] = "arithmetic overflow",
+    [EXCP_TRAP] = "trap",
+    [EXCP_FPE] = "floating point",
+    [EXCP_DDBS] = "debug data break store",
+    [EXCP_DWATCH] = "data watchpoint",
+    [EXCP_LTLBL] = "TLB modify",
+    [EXCP_TLBL] = "TLB load",
+    [EXCP_TLBS] = "TLB store",
+    [EXCP_DBE] = "data bus error",
+    [EXCP_DDBL] = "debug data break load",
+    [EXCP_THREAD] = "thread",
+    [EXCP_MDMX] = "MDMX",
+    [EXCP_C2E] = "precise coprocessor 2",
+    [EXCP_CACHE] = "cache error",
+};
+
+void do_interrupt (CPUState *env)
+{
+#if !defined(CONFIG_USER_ONLY)
+    target_ulong offset;
+    int cause = -1;
+    const char *name;
+
+    if (qemu_log_enabled() && env->exception_index != EXCP_EXT_INTERRUPT) {
+        if (env->exception_index < 0 || env->exception_index > EXCP_LAST)
+            name = "unknown";
+        else
+            name = excp_names[env->exception_index];
+
+        qemu_log("%s enter: PC " TARGET_FMT_lx " EPC " TARGET_FMT_lx " %s exception\n",
+                 __func__, env->active_tc.PC, env->CP0_EPC, name);
+    }
+    if (env->exception_index == EXCP_EXT_INTERRUPT &&
+        (env->hflags & MIPS_HFLAG_DM))
+        env->exception_index = EXCP_DINT;
+    offset = 0x180;
+    switch (env->exception_index) {
+    case EXCP_DSS:
+        env->CP0_Debug |= 1 << CP0DB_DSS;
+        /* Debug single step cannot be raised inside a delay slot and
+           resume will always occur on the next instruction
+           (but we assume the pc has always been updated during
+           code translation). */
+        env->CP0_DEPC = env->active_tc.PC;
+        goto enter_debug_mode;
+    case EXCP_DINT:
+        env->CP0_Debug |= 1 << CP0DB_DINT;
+        goto set_DEPC;
+    case EXCP_DIB:
+        env->CP0_Debug |= 1 << CP0DB_DIB;
+        goto set_DEPC;
+    case EXCP_DBp:
+        env->CP0_Debug |= 1 << CP0DB_DBp;
+        goto set_DEPC;
+    case EXCP_DDBS:
+        env->CP0_Debug |= 1 << CP0DB_DDBS;
+        goto set_DEPC;
+    case EXCP_DDBL:
+        env->CP0_Debug |= 1 << CP0DB_DDBL;
+    set_DEPC:
+        if (env->hflags & MIPS_HFLAG_BMASK) {
+            /* If the exception was raised from a delay slot,
+               come back to the jump.  */
+            env->CP0_DEPC = env->active_tc.PC - 4;
+            env->hflags &= ~MIPS_HFLAG_BMASK;
+        } else {
+            env->CP0_DEPC = env->active_tc.PC;
+        }
+ enter_debug_mode:
+        env->hflags |= MIPS_HFLAG_DM | MIPS_HFLAG_64 | MIPS_HFLAG_CP0;
+        env->hflags &= ~(MIPS_HFLAG_KSU);
+        /* EJTAG probe trap enable is not implemented... */
+        if (!(env->CP0_Status & (1 << CP0St_EXL)))
+            env->CP0_Cause &= ~(1 << CP0Ca_BD);
+        env->active_tc.PC = (int32_t)0xBFC00480;
+        break;
+    case EXCP_RESET:
+        cpu_reset(env);
+        break;
+    case EXCP_SRESET:
+        env->CP0_Status |= (1 << CP0St_SR);
+        memset(env->CP0_WatchLo, 0, sizeof(*env->CP0_WatchLo));
+        goto set_error_EPC;
+    case EXCP_NMI:
+        env->CP0_Status |= (1 << CP0St_NMI);
+ set_error_EPC:
+        if (env->hflags & MIPS_HFLAG_BMASK) {
+            /* If the exception was raised from a delay slot,
+               come back to the jump.  */
+            env->CP0_ErrorEPC = env->active_tc.PC - 4;
+            env->hflags &= ~MIPS_HFLAG_BMASK;
+        } else {
+            env->CP0_ErrorEPC = env->active_tc.PC;
+        }
+        env->CP0_Status |= (1 << CP0St_ERL) | (1 << CP0St_BEV);
+        env->hflags |= MIPS_HFLAG_64 | MIPS_HFLAG_CP0;
+        env->hflags &= ~(MIPS_HFLAG_KSU);
+        if (!(env->CP0_Status & (1 << CP0St_EXL)))
+            env->CP0_Cause &= ~(1 << CP0Ca_BD);
+        env->active_tc.PC = (int32_t)0xBFC00000;
+        break;
+    case EXCP_EXT_INTERRUPT:
+        cause = 0;
+        if (env->CP0_Cause & (1 << CP0Ca_IV))
+            offset = 0x200;
+        goto set_EPC;
+    case EXCP_LTLBL:
+        cause = 1;
+        goto set_EPC;
+    case EXCP_TLBL:
+        cause = 2;
+        if (env->error_code == 1 && !(env->CP0_Status & (1 << CP0St_EXL))) {
+#if defined(TARGET_MIPS64)
+            int R = env->CP0_BadVAddr >> 62;
+            int UX = (env->CP0_Status & (1 << CP0St_UX)) != 0;
+            int SX = (env->CP0_Status & (1 << CP0St_SX)) != 0;
+            int KX = (env->CP0_Status & (1 << CP0St_KX)) != 0;
+
+            if ((R == 0 && UX) || (R == 1 && SX) || (R == 3 && KX))
+                offset = 0x080;
+            else
+#endif
+                offset = 0x000;
+        }
+        goto set_EPC;
+    case EXCP_TLBS:
+        cause = 3;
+        if (env->error_code == 1 && !(env->CP0_Status & (1 << CP0St_EXL))) {
+#if defined(TARGET_MIPS64)
+            int R = env->CP0_BadVAddr >> 62;
+            int UX = (env->CP0_Status & (1 << CP0St_UX)) != 0;
+            int SX = (env->CP0_Status & (1 << CP0St_SX)) != 0;
+            int KX = (env->CP0_Status & (1 << CP0St_KX)) != 0;
+
+            if ((R == 0 && UX) || (R == 1 && SX) || (R == 3 && KX))
+                offset = 0x080;
+            else
+#endif
+                offset = 0x000;
+        }
+        goto set_EPC;
+    case EXCP_AdEL:
+        cause = 4;
+        goto set_EPC;
+    case EXCP_AdES:
+        cause = 5;
+        goto set_EPC;
+    case EXCP_IBE:
+        cause = 6;
+        goto set_EPC;
+    case EXCP_DBE:
+        cause = 7;
+        goto set_EPC;
+    case EXCP_SYSCALL:
+        cause = 8;
+        goto set_EPC;
+    case EXCP_BREAK:
+        cause = 9;
+        goto set_EPC;
+    case EXCP_RI:
+        cause = 10;
+        goto set_EPC;
+    case EXCP_CpU:
+        cause = 11;
+        env->CP0_Cause = (env->CP0_Cause & ~(0x3 << CP0Ca_CE)) |
+                         (env->error_code << CP0Ca_CE);
+        goto set_EPC;
+    case EXCP_OVERFLOW:
+        cause = 12;
+        goto set_EPC;
+    case EXCP_TRAP:
+        cause = 13;
+        goto set_EPC;
+    case EXCP_FPE:
+        cause = 15;
+        goto set_EPC;
+    case EXCP_C2E:
+        cause = 18;
+        goto set_EPC;
+    case EXCP_MDMX:
+        cause = 22;
+        goto set_EPC;
+    case EXCP_DWATCH:
+        cause = 23;
+        /* XXX: TODO: manage defered watch exceptions */
+        goto set_EPC;
+    case EXCP_MCHECK:
+        cause = 24;
+        goto set_EPC;
+    case EXCP_THREAD:
+        cause = 25;
+        goto set_EPC;
+    case EXCP_CACHE:
+        cause = 30;
+        if (env->CP0_Status & (1 << CP0St_BEV)) {
+            offset = 0x100;
+        } else {
+            offset = 0x20000100;
+        }
+ set_EPC:
+        if (!(env->CP0_Status & (1 << CP0St_EXL))) {
+            if (env->hflags & MIPS_HFLAG_BMASK) {
+                /* If the exception was raised from a delay slot,
+                   come back to the jump.  */
+                env->CP0_EPC = env->active_tc.PC - 4;
+                env->CP0_Cause |= (1 << CP0Ca_BD);
+            } else {
+                env->CP0_EPC = env->active_tc.PC;
+                env->CP0_Cause &= ~(1 << CP0Ca_BD);
+            }
+            env->CP0_Status |= (1 << CP0St_EXL);
+            env->hflags |= MIPS_HFLAG_64 | MIPS_HFLAG_CP0;
+            env->hflags &= ~(MIPS_HFLAG_KSU);
+        }
+        env->hflags &= ~MIPS_HFLAG_BMASK;
+        if (env->CP0_Status & (1 << CP0St_BEV)) {
+            env->active_tc.PC = (int32_t)0xBFC00200;
+        } else {
+            env->active_tc.PC = (int32_t)(env->CP0_EBase & ~0x3ff);
+        }
+        env->active_tc.PC += offset;
+        env->CP0_Cause = (env->CP0_Cause & ~(0x1f << CP0Ca_EC)) | (cause << CP0Ca_EC);
+        break;
+    default:
+        qemu_log("Invalid MIPS exception %d. Exiting\n", env->exception_index);
+        printf("Invalid MIPS exception %d. Exiting\n", env->exception_index);
+        exit(1);
+    }
+    if (qemu_log_enabled() && env->exception_index != EXCP_EXT_INTERRUPT) {
+        qemu_log("%s: PC " TARGET_FMT_lx " EPC " TARGET_FMT_lx " cause %d\n"
+                "    S %08x C %08x A " TARGET_FMT_lx " D " TARGET_FMT_lx "\n",
+                __func__, env->active_tc.PC, env->CP0_EPC, cause,
+                env->CP0_Status, env->CP0_Cause, env->CP0_BadVAddr,
+                env->CP0_DEPC);
+    }
+#endif
+    env->exception_index = EXCP_NONE;
+}
+
+void r4k_invalidate_tlb (CPUState *env, int idx, int use_extra)
+{
+    r4k_tlb_t *tlb;
+    target_ulong addr;
+    target_ulong end;
+    uint8_t ASID = env->CP0_EntryHi & 0xFF;
+    target_ulong mask;
+
+    tlb = &env->tlb->mmu.r4k.tlb[idx];
+    /* The qemu TLB is flushed when the ASID changes, so no need to
+       flush these entries again.  */
+    if (tlb->G == 0 && tlb->ASID != ASID) {
+        return;
+    }
+
+    if (use_extra && env->tlb->tlb_in_use < MIPS_TLB_MAX) {
+        /* For tlbwr, we can shadow the discarded entry into
+           a new (fake) TLB entry, as long as the guest can not
+           tell that it's there.  */
+        env->tlb->mmu.r4k.tlb[env->tlb->tlb_in_use] = *tlb;
+        env->tlb->tlb_in_use++;
+        return;
+    }
+
+    /* 1k pages are not supported. */
+    mask = tlb->PageMask | ~(TARGET_PAGE_MASK << 1);
+    if (tlb->V0) {
+        addr = tlb->VPN & ~mask;
+#if defined(TARGET_MIPS64)
+        if (addr >= (0xFFFFFFFF80000000ULL & env->SEGMask)) {
+            addr |= 0x3FFFFF0000000000ULL;
+        }
+#endif
+        end = addr | (mask >> 1);
+        while (addr < end) {
+            tlb_flush_page (env, addr);
+            addr += TARGET_PAGE_SIZE;
+        }
+    }
+    if (tlb->V1) {
+        addr = (tlb->VPN & ~mask) | ((mask >> 1) + 1);
+#if defined(TARGET_MIPS64)
+        if (addr >= (0xFFFFFFFF80000000ULL & env->SEGMask)) {
+            addr |= 0x3FFFFF0000000000ULL;
+        }
+#endif
+        end = addr | mask;
+        while (addr - 1 < end) {
+            tlb_flush_page (env, addr);
+            addr += TARGET_PAGE_SIZE;
+        }
+    }
+}
diff --git a/target-mips/helper.h b/target-mips/helper.h
new file mode 100644
index 0000000..ab47b1a
--- /dev/null
+++ b/target-mips/helper.h
@@ -0,0 +1,282 @@
+#include "def-helper.h"
+
+DEF_HELPER_2(raise_exception_err, void, i32, int)
+DEF_HELPER_1(raise_exception, void, i32)
+DEF_HELPER_0(interrupt_restart, void)
+
+#ifdef TARGET_MIPS64
+DEF_HELPER_3(ldl, tl, tl, tl, int)
+DEF_HELPER_3(ldr, tl, tl, tl, int)
+DEF_HELPER_3(sdl, void, tl, tl, int)
+DEF_HELPER_3(sdr, void, tl, tl, int)
+#endif
+DEF_HELPER_3(lwl, tl, tl, tl, int)
+DEF_HELPER_3(lwr, tl, tl, tl, int)
+DEF_HELPER_3(swl, void, tl, tl, int)
+DEF_HELPER_3(swr, void, tl, tl, int)
+
+#ifndef CONFIG_USER_ONLY
+DEF_HELPER_2(ll, tl, tl, int)
+DEF_HELPER_3(sc, tl, tl, tl, int)
+#ifdef TARGET_MIPS64
+DEF_HELPER_2(lld, tl, tl, int)
+DEF_HELPER_3(scd, tl, tl, tl, int)
+#endif
+#endif
+
+DEF_HELPER_FLAGS_1(clo, TCG_CALL_CONST | TCG_CALL_PURE, tl, tl)
+DEF_HELPER_FLAGS_1(clz, TCG_CALL_CONST | TCG_CALL_PURE, tl, tl)
+#ifdef TARGET_MIPS64
+DEF_HELPER_FLAGS_1(dclo, TCG_CALL_CONST | TCG_CALL_PURE, tl, tl)
+DEF_HELPER_FLAGS_1(dclz, TCG_CALL_CONST | TCG_CALL_PURE, tl, tl)
+DEF_HELPER_2(dmult, void, tl, tl)
+DEF_HELPER_2(dmultu, void, tl, tl)
+#endif
+
+DEF_HELPER_2(muls, tl, tl, tl)
+DEF_HELPER_2(mulsu, tl, tl, tl)
+DEF_HELPER_2(macc, tl, tl, tl)
+DEF_HELPER_2(maccu, tl, tl, tl)
+DEF_HELPER_2(msac, tl, tl, tl)
+DEF_HELPER_2(msacu, tl, tl, tl)
+DEF_HELPER_2(mulhi, tl, tl, tl)
+DEF_HELPER_2(mulhiu, tl, tl, tl)
+DEF_HELPER_2(mulshi, tl, tl, tl)
+DEF_HELPER_2(mulshiu, tl, tl, tl)
+DEF_HELPER_2(macchi, tl, tl, tl)
+DEF_HELPER_2(macchiu, tl, tl, tl)
+DEF_HELPER_2(msachi, tl, tl, tl)
+DEF_HELPER_2(msachiu, tl, tl, tl)
+
+#ifndef CONFIG_USER_ONLY
+/* CP0 helpers */
+DEF_HELPER_0(mfc0_mvpcontrol, tl)
+DEF_HELPER_0(mfc0_mvpconf0, tl)
+DEF_HELPER_0(mfc0_mvpconf1, tl)
+DEF_HELPER_0(mfc0_random, tl)
+DEF_HELPER_0(mfc0_tcstatus, tl)
+DEF_HELPER_0(mftc0_tcstatus, tl)
+DEF_HELPER_0(mfc0_tcbind, tl)
+DEF_HELPER_0(mftc0_tcbind, tl)
+DEF_HELPER_0(mfc0_tcrestart, tl)
+DEF_HELPER_0(mftc0_tcrestart, tl)
+DEF_HELPER_0(mfc0_tchalt, tl)
+DEF_HELPER_0(mftc0_tchalt, tl)
+DEF_HELPER_0(mfc0_tccontext, tl)
+DEF_HELPER_0(mftc0_tccontext, tl)
+DEF_HELPER_0(mfc0_tcschedule, tl)
+DEF_HELPER_0(mftc0_tcschedule, tl)
+DEF_HELPER_0(mfc0_tcschefback, tl)
+DEF_HELPER_0(mftc0_tcschefback, tl)
+DEF_HELPER_0(mfc0_count, tl)
+DEF_HELPER_0(mftc0_entryhi, tl)
+DEF_HELPER_0(mftc0_status, tl)
+DEF_HELPER_0(mfc0_lladdr, tl)
+DEF_HELPER_1(mfc0_watchlo, tl, i32)
+DEF_HELPER_1(mfc0_watchhi, tl, i32)
+DEF_HELPER_0(mfc0_debug, tl)
+DEF_HELPER_0(mftc0_debug, tl)
+#ifdef TARGET_MIPS64
+DEF_HELPER_0(dmfc0_tcrestart, tl)
+DEF_HELPER_0(dmfc0_tchalt, tl)
+DEF_HELPER_0(dmfc0_tccontext, tl)
+DEF_HELPER_0(dmfc0_tcschedule, tl)
+DEF_HELPER_0(dmfc0_tcschefback, tl)
+DEF_HELPER_0(dmfc0_lladdr, tl)
+DEF_HELPER_1(dmfc0_watchlo, tl, i32)
+#endif /* TARGET_MIPS64 */
+
+DEF_HELPER_1(mtc0_index, void, tl)
+DEF_HELPER_1(mtc0_mvpcontrol, void, tl)
+DEF_HELPER_1(mtc0_vpecontrol, void, tl)
+DEF_HELPER_1(mtc0_vpeconf0, void, tl)
+DEF_HELPER_1(mtc0_vpeconf1, void, tl)
+DEF_HELPER_1(mtc0_yqmask, void, tl)
+DEF_HELPER_1(mtc0_vpeopt, void, tl)
+DEF_HELPER_1(mtc0_entrylo0, void, tl)
+DEF_HELPER_1(mtc0_tcstatus, void, tl)
+DEF_HELPER_1(mttc0_tcstatus, void, tl)
+DEF_HELPER_1(mtc0_tcbind, void, tl)
+DEF_HELPER_1(mttc0_tcbind, void, tl)
+DEF_HELPER_1(mtc0_tcrestart, void, tl)
+DEF_HELPER_1(mttc0_tcrestart, void, tl)
+DEF_HELPER_1(mtc0_tchalt, void, tl)
+DEF_HELPER_1(mttc0_tchalt, void, tl)
+DEF_HELPER_1(mtc0_tccontext, void, tl)
+DEF_HELPER_1(mttc0_tccontext, void, tl)
+DEF_HELPER_1(mtc0_tcschedule, void, tl)
+DEF_HELPER_1(mttc0_tcschedule, void, tl)
+DEF_HELPER_1(mtc0_tcschefback, void, tl)
+DEF_HELPER_1(mttc0_tcschefback, void, tl)
+DEF_HELPER_1(mtc0_entrylo1, void, tl)
+DEF_HELPER_1(mtc0_context, void, tl)
+DEF_HELPER_1(mtc0_pagemask, void, tl)
+DEF_HELPER_1(mtc0_pagegrain, void, tl)
+DEF_HELPER_1(mtc0_wired, void, tl)
+DEF_HELPER_1(mtc0_srsconf0, void, tl)
+DEF_HELPER_1(mtc0_srsconf1, void, tl)
+DEF_HELPER_1(mtc0_srsconf2, void, tl)
+DEF_HELPER_1(mtc0_srsconf3, void, tl)
+DEF_HELPER_1(mtc0_srsconf4, void, tl)
+DEF_HELPER_1(mtc0_hwrena, void, tl)
+DEF_HELPER_1(mtc0_count, void, tl)
+DEF_HELPER_1(mtc0_entryhi, void, tl)
+DEF_HELPER_1(mttc0_entryhi, void, tl)
+DEF_HELPER_1(mtc0_compare, void, tl)
+DEF_HELPER_1(mtc0_status, void, tl)
+DEF_HELPER_1(mttc0_status, void, tl)
+DEF_HELPER_1(mtc0_intctl, void, tl)
+DEF_HELPER_1(mtc0_srsctl, void, tl)
+DEF_HELPER_1(mtc0_cause, void, tl)
+DEF_HELPER_1(mtc0_ebase, void, tl)
+DEF_HELPER_1(mtc0_config0, void, tl)
+DEF_HELPER_1(mtc0_config2, void, tl)
+DEF_HELPER_1(mtc0_lladdr, void, tl)
+DEF_HELPER_2(mtc0_watchlo, void, tl, i32)
+DEF_HELPER_2(mtc0_watchhi, void, tl, i32)
+DEF_HELPER_1(mtc0_xcontext, void, tl)
+DEF_HELPER_1(mtc0_framemask, void, tl)
+DEF_HELPER_1(mtc0_debug, void, tl)
+DEF_HELPER_1(mttc0_debug, void, tl)
+DEF_HELPER_1(mtc0_performance0, void, tl)
+DEF_HELPER_1(mtc0_taglo, void, tl)
+DEF_HELPER_1(mtc0_datalo, void, tl)
+DEF_HELPER_1(mtc0_taghi, void, tl)
+DEF_HELPER_1(mtc0_datahi, void, tl)
+
+/* MIPS MT functions */
+DEF_HELPER_1(mftgpr, tl, i32);
+DEF_HELPER_1(mftlo, tl, i32)
+DEF_HELPER_1(mfthi, tl, i32)
+DEF_HELPER_1(mftacx, tl, i32)
+DEF_HELPER_0(mftdsp, tl)
+DEF_HELPER_2(mttgpr, void, tl, i32)
+DEF_HELPER_2(mttlo, void, tl, i32)
+DEF_HELPER_2(mtthi, void, tl, i32)
+DEF_HELPER_2(mttacx, void, tl, i32)
+DEF_HELPER_1(mttdsp, void, tl)
+DEF_HELPER_1(dmt, tl, tl)
+DEF_HELPER_1(emt, tl, tl)
+DEF_HELPER_1(dvpe, tl, tl)
+DEF_HELPER_1(evpe, tl, tl)
+#endif /* !CONFIG_USER_ONLY */
+DEF_HELPER_2(fork, void, tl, tl)
+DEF_HELPER_1(yield, tl, tl)
+
+/* CP1 functions */
+DEF_HELPER_1(cfc1, tl, i32)
+DEF_HELPER_2(ctc1, void, tl, i32)
+
+DEF_HELPER_1(float_cvtd_s, i64, i32)
+DEF_HELPER_1(float_cvtd_w, i64, i32)
+DEF_HELPER_1(float_cvtd_l, i64, i64)
+DEF_HELPER_1(float_cvtl_d, i64, i64)
+DEF_HELPER_1(float_cvtl_s, i64, i32)
+DEF_HELPER_1(float_cvtps_pw, i64, i64)
+DEF_HELPER_1(float_cvtpw_ps, i64, i64)
+DEF_HELPER_1(float_cvts_d, i32, i64)
+DEF_HELPER_1(float_cvts_w, i32, i32)
+DEF_HELPER_1(float_cvts_l, i32, i64)
+DEF_HELPER_1(float_cvts_pl, i32, i32)
+DEF_HELPER_1(float_cvts_pu, i32, i32)
+DEF_HELPER_1(float_cvtw_s, i32, i32)
+DEF_HELPER_1(float_cvtw_d, i32, i64)
+
+DEF_HELPER_2(float_addr_ps, i64, i64, i64)
+DEF_HELPER_2(float_mulr_ps, i64, i64, i64)
+
+#define FOP_PROTO(op)                       \
+DEF_HELPER_1(float_ ## op ## l_s, i64, i32) \
+DEF_HELPER_1(float_ ## op ## l_d, i64, i64) \
+DEF_HELPER_1(float_ ## op ## w_s, i32, i32) \
+DEF_HELPER_1(float_ ## op ## w_d, i32, i64)
+FOP_PROTO(round)
+FOP_PROTO(trunc)
+FOP_PROTO(ceil)
+FOP_PROTO(floor)
+#undef FOP_PROTO
+
+#define FOP_PROTO(op)                       \
+DEF_HELPER_1(float_ ## op ## _s, i32, i32)  \
+DEF_HELPER_1(float_ ## op ## _d, i64, i64)
+FOP_PROTO(sqrt)
+FOP_PROTO(rsqrt)
+FOP_PROTO(recip)
+#undef FOP_PROTO
+
+#define FOP_PROTO(op)                       \
+DEF_HELPER_1(float_ ## op ## _s, i32, i32)  \
+DEF_HELPER_1(float_ ## op ## _d, i64, i64)  \
+DEF_HELPER_1(float_ ## op ## _ps, i64, i64)
+FOP_PROTO(abs)
+FOP_PROTO(chs)
+FOP_PROTO(recip1)
+FOP_PROTO(rsqrt1)
+#undef FOP_PROTO
+
+#define FOP_PROTO(op)                             \
+DEF_HELPER_2(float_ ## op ## _s, i32, i32, i32)   \
+DEF_HELPER_2(float_ ## op ## _d, i64, i64, i64)   \
+DEF_HELPER_2(float_ ## op ## _ps, i64, i64, i64)
+FOP_PROTO(add)
+FOP_PROTO(sub)
+FOP_PROTO(mul)
+FOP_PROTO(div)
+FOP_PROTO(recip2)
+FOP_PROTO(rsqrt2)
+#undef FOP_PROTO
+
+#define FOP_PROTO(op)                                 \
+DEF_HELPER_3(float_ ## op ## _s, i32, i32, i32, i32)  \
+DEF_HELPER_3(float_ ## op ## _d, i64, i64, i64, i64)  \
+DEF_HELPER_3(float_ ## op ## _ps, i64, i64, i64, i64)
+FOP_PROTO(muladd)
+FOP_PROTO(mulsub)
+FOP_PROTO(nmuladd)
+FOP_PROTO(nmulsub)
+#undef FOP_PROTO
+
+#define FOP_PROTO(op)                               \
+DEF_HELPER_3(cmp_d_ ## op, void, i64, i64, int)     \
+DEF_HELPER_3(cmpabs_d_ ## op, void, i64, i64, int)  \
+DEF_HELPER_3(cmp_s_ ## op, void, i32, i32, int)     \
+DEF_HELPER_3(cmpabs_s_ ## op, void, i32, i32, int)  \
+DEF_HELPER_3(cmp_ps_ ## op, void, i64, i64, int)    \
+DEF_HELPER_3(cmpabs_ps_ ## op, void, i64, i64, int)
+FOP_PROTO(f)
+FOP_PROTO(un)
+FOP_PROTO(eq)
+FOP_PROTO(ueq)
+FOP_PROTO(olt)
+FOP_PROTO(ult)
+FOP_PROTO(ole)
+FOP_PROTO(ule)
+FOP_PROTO(sf)
+FOP_PROTO(ngle)
+FOP_PROTO(seq)
+FOP_PROTO(ngl)
+FOP_PROTO(lt)
+FOP_PROTO(nge)
+FOP_PROTO(le)
+FOP_PROTO(ngt)
+#undef FOP_PROTO
+
+/* Special functions */
+#ifndef CONFIG_USER_ONLY
+DEF_HELPER_0(tlbwi, void)
+DEF_HELPER_0(tlbwr, void)
+DEF_HELPER_0(tlbp, void)
+DEF_HELPER_0(tlbr, void)
+DEF_HELPER_0(di, tl)
+DEF_HELPER_0(ei, tl)
+DEF_HELPER_0(eret, void)
+DEF_HELPER_0(deret, void)
+#endif /* !CONFIG_USER_ONLY */
+DEF_HELPER_0(rdhwr_cpunum, tl)
+DEF_HELPER_0(rdhwr_synci_step, tl)
+DEF_HELPER_0(rdhwr_cc, tl)
+DEF_HELPER_0(rdhwr_ccres, tl)
+DEF_HELPER_1(pmon, void, int)
+DEF_HELPER_0(wait, void)
+
+#include "def-helper.h"
diff --git a/target-mips/machine.c b/target-mips/machine.c
new file mode 100644
index 0000000..9ffac71
--- /dev/null
+++ b/target-mips/machine.c
@@ -0,0 +1,308 @@
+#include "hw/hw.h"
+#include "hw/boards.h"
+
+#include "exec-all.h"
+
+static void save_tc(QEMUFile *f, TCState *tc)
+{
+    int i;
+
+    /* Save active TC */
+    for(i = 0; i < 32; i++)
+        qemu_put_betls(f, &tc->gpr[i]);
+    qemu_put_betls(f, &tc->PC);
+    for(i = 0; i < MIPS_DSP_ACC; i++)
+        qemu_put_betls(f, &tc->HI[i]);
+    for(i = 0; i < MIPS_DSP_ACC; i++)
+        qemu_put_betls(f, &tc->LO[i]);
+    for(i = 0; i < MIPS_DSP_ACC; i++)
+        qemu_put_betls(f, &tc->ACX[i]);
+    qemu_put_betls(f, &tc->DSPControl);
+    qemu_put_sbe32s(f, &tc->CP0_TCStatus);
+    qemu_put_sbe32s(f, &tc->CP0_TCBind);
+    qemu_put_betls(f, &tc->CP0_TCHalt);
+    qemu_put_betls(f, &tc->CP0_TCContext);
+    qemu_put_betls(f, &tc->CP0_TCSchedule);
+    qemu_put_betls(f, &tc->CP0_TCScheFBack);
+    qemu_put_sbe32s(f, &tc->CP0_Debug_tcstatus);
+}
+
+static void save_fpu(QEMUFile *f, CPUMIPSFPUContext *fpu)
+{
+    int i;
+
+    for(i = 0; i < 32; i++)
+        qemu_put_be64s(f, &fpu->fpr[i].d);
+    qemu_put_s8s(f, &fpu->fp_status.float_detect_tininess);
+    qemu_put_s8s(f, &fpu->fp_status.float_rounding_mode);
+    qemu_put_s8s(f, &fpu->fp_status.float_exception_flags);
+    qemu_put_be32s(f, &fpu->fcr0);
+    qemu_put_be32s(f, &fpu->fcr31);
+}
+
+void cpu_save(QEMUFile *f, void *opaque)
+{
+    CPUState *env = opaque;
+    int i;
+
+    /* Save active TC */
+    save_tc(f, &env->active_tc);
+
+    /* Save active FPU */
+    save_fpu(f, &env->active_fpu);
+
+    /* Save MVP */
+    qemu_put_sbe32s(f, &env->mvp->CP0_MVPControl);
+    qemu_put_sbe32s(f, &env->mvp->CP0_MVPConf0);
+    qemu_put_sbe32s(f, &env->mvp->CP0_MVPConf1);
+
+    /* Save TLB */
+    qemu_put_be32s(f, &env->tlb->nb_tlb);
+    qemu_put_be32s(f, &env->tlb->tlb_in_use);
+    for(i = 0; i < MIPS_TLB_MAX; i++) {
+        uint16_t flags = ((env->tlb->mmu.r4k.tlb[i].G << 10) |
+                          (env->tlb->mmu.r4k.tlb[i].C0 << 7) |
+                          (env->tlb->mmu.r4k.tlb[i].C1 << 4) |
+                          (env->tlb->mmu.r4k.tlb[i].V0 << 3) |
+                          (env->tlb->mmu.r4k.tlb[i].V1 << 2) |
+                          (env->tlb->mmu.r4k.tlb[i].D0 << 1) |
+                          (env->tlb->mmu.r4k.tlb[i].D1 << 0));
+        uint8_t asid;
+
+        qemu_put_betls(f, &env->tlb->mmu.r4k.tlb[i].VPN);
+        qemu_put_be32s(f, &env->tlb->mmu.r4k.tlb[i].PageMask);
+        asid = env->tlb->mmu.r4k.tlb[i].ASID;
+        qemu_put_8s(f, &asid);
+        qemu_put_be16s(f, &flags);
+        qemu_put_betls(f, &env->tlb->mmu.r4k.tlb[i].PFN[0]);
+        qemu_put_betls(f, &env->tlb->mmu.r4k.tlb[i].PFN[1]);
+    }
+
+    /* Save CPU metastate */
+    qemu_put_be32s(f, &env->current_tc);
+    qemu_put_be32s(f, &env->current_fpu);
+    qemu_put_sbe32s(f, &env->error_code);
+    qemu_put_be32s(f, &env->hflags);
+    qemu_put_betls(f, &env->btarget);
+    i = env->bcond;
+    qemu_put_sbe32s(f, &i);
+
+    /* Save remaining CP1 registers */
+    qemu_put_sbe32s(f, &env->CP0_Index);
+    qemu_put_sbe32s(f, &env->CP0_Random);
+    qemu_put_sbe32s(f, &env->CP0_VPEControl);
+    qemu_put_sbe32s(f, &env->CP0_VPEConf0);
+    qemu_put_sbe32s(f, &env->CP0_VPEConf1);
+    qemu_put_betls(f, &env->CP0_YQMask);
+    qemu_put_betls(f, &env->CP0_VPESchedule);
+    qemu_put_betls(f, &env->CP0_VPEScheFBack);
+    qemu_put_sbe32s(f, &env->CP0_VPEOpt);
+    qemu_put_betls(f, &env->CP0_EntryLo0);
+    qemu_put_betls(f, &env->CP0_EntryLo1);
+    qemu_put_betls(f, &env->CP0_Context);
+    qemu_put_sbe32s(f, &env->CP0_PageMask);
+    qemu_put_sbe32s(f, &env->CP0_PageGrain);
+    qemu_put_sbe32s(f, &env->CP0_Wired);
+    qemu_put_sbe32s(f, &env->CP0_SRSConf0);
+    qemu_put_sbe32s(f, &env->CP0_SRSConf1);
+    qemu_put_sbe32s(f, &env->CP0_SRSConf2);
+    qemu_put_sbe32s(f, &env->CP0_SRSConf3);
+    qemu_put_sbe32s(f, &env->CP0_SRSConf4);
+    qemu_put_sbe32s(f, &env->CP0_HWREna);
+    qemu_put_betls(f, &env->CP0_BadVAddr);
+    qemu_put_sbe32s(f, &env->CP0_Count);
+    qemu_put_betls(f, &env->CP0_EntryHi);
+    qemu_put_sbe32s(f, &env->CP0_Compare);
+    qemu_put_sbe32s(f, &env->CP0_Status);
+    qemu_put_sbe32s(f, &env->CP0_IntCtl);
+    qemu_put_sbe32s(f, &env->CP0_SRSCtl);
+    qemu_put_sbe32s(f, &env->CP0_SRSMap);
+    qemu_put_sbe32s(f, &env->CP0_Cause);
+    qemu_put_betls(f, &env->CP0_EPC);
+    qemu_put_sbe32s(f, &env->CP0_PRid);
+    qemu_put_sbe32s(f, &env->CP0_EBase);
+    qemu_put_sbe32s(f, &env->CP0_Config0);
+    qemu_put_sbe32s(f, &env->CP0_Config1);
+    qemu_put_sbe32s(f, &env->CP0_Config2);
+    qemu_put_sbe32s(f, &env->CP0_Config3);
+    qemu_put_sbe32s(f, &env->CP0_Config6);
+    qemu_put_sbe32s(f, &env->CP0_Config7);
+    qemu_put_betls(f, &env->lladdr);
+    for(i = 0; i < 8; i++)
+        qemu_put_betls(f, &env->CP0_WatchLo[i]);
+    for(i = 0; i < 8; i++)
+        qemu_put_sbe32s(f, &env->CP0_WatchHi[i]);
+    qemu_put_betls(f, &env->CP0_XContext);
+    qemu_put_sbe32s(f, &env->CP0_Framemask);
+    qemu_put_sbe32s(f, &env->CP0_Debug);
+    qemu_put_betls(f, &env->CP0_DEPC);
+    qemu_put_sbe32s(f, &env->CP0_Performance0);
+    qemu_put_sbe32s(f, &env->CP0_TagLo);
+    qemu_put_sbe32s(f, &env->CP0_DataLo);
+    qemu_put_sbe32s(f, &env->CP0_TagHi);
+    qemu_put_sbe32s(f, &env->CP0_DataHi);
+    qemu_put_betls(f, &env->CP0_ErrorEPC);
+    qemu_put_sbe32s(f, &env->CP0_DESAVE);
+
+    /* Save inactive TC state */
+    for (i = 0; i < MIPS_SHADOW_SET_MAX; i++)
+        save_tc(f, &env->tcs[i]);
+    for (i = 0; i < MIPS_FPU_MAX; i++)
+        save_fpu(f, &env->fpus[i]);
+}
+
+static void load_tc(QEMUFile *f, TCState *tc)
+{
+    int i;
+
+    /* Save active TC */
+    for(i = 0; i < 32; i++)
+        qemu_get_betls(f, &tc->gpr[i]);
+    qemu_get_betls(f, &tc->PC);
+    for(i = 0; i < MIPS_DSP_ACC; i++)
+        qemu_get_betls(f, &tc->HI[i]);
+    for(i = 0; i < MIPS_DSP_ACC; i++)
+        qemu_get_betls(f, &tc->LO[i]);
+    for(i = 0; i < MIPS_DSP_ACC; i++)
+        qemu_get_betls(f, &tc->ACX[i]);
+    qemu_get_betls(f, &tc->DSPControl);
+    qemu_get_sbe32s(f, &tc->CP0_TCStatus);
+    qemu_get_sbe32s(f, &tc->CP0_TCBind);
+    qemu_get_betls(f, &tc->CP0_TCHalt);
+    qemu_get_betls(f, &tc->CP0_TCContext);
+    qemu_get_betls(f, &tc->CP0_TCSchedule);
+    qemu_get_betls(f, &tc->CP0_TCScheFBack);
+    qemu_get_sbe32s(f, &tc->CP0_Debug_tcstatus);
+}
+
+static void load_fpu(QEMUFile *f, CPUMIPSFPUContext *fpu)
+{
+    int i;
+
+    for(i = 0; i < 32; i++)
+        qemu_get_be64s(f, &fpu->fpr[i].d);
+    qemu_get_s8s(f, &fpu->fp_status.float_detect_tininess);
+    qemu_get_s8s(f, &fpu->fp_status.float_rounding_mode);
+    qemu_get_s8s(f, &fpu->fp_status.float_exception_flags);
+    qemu_get_be32s(f, &fpu->fcr0);
+    qemu_get_be32s(f, &fpu->fcr31);
+}
+
+int cpu_load(QEMUFile *f, void *opaque, int version_id)
+{
+    CPUState *env = opaque;
+    int i;
+
+    if (version_id != 3)
+        return -EINVAL;
+
+    /* Load active TC */
+    load_tc(f, &env->active_tc);
+
+    /* Load active FPU */
+    load_fpu(f, &env->active_fpu);
+
+    /* Load MVP */
+    qemu_get_sbe32s(f, &env->mvp->CP0_MVPControl);
+    qemu_get_sbe32s(f, &env->mvp->CP0_MVPConf0);
+    qemu_get_sbe32s(f, &env->mvp->CP0_MVPConf1);
+
+    /* Load TLB */
+    qemu_get_be32s(f, &env->tlb->nb_tlb);
+    qemu_get_be32s(f, &env->tlb->tlb_in_use);
+    for(i = 0; i < MIPS_TLB_MAX; i++) {
+        uint16_t flags;
+        uint8_t asid;
+
+        qemu_get_betls(f, &env->tlb->mmu.r4k.tlb[i].VPN);
+        qemu_get_be32s(f, &env->tlb->mmu.r4k.tlb[i].PageMask);
+        qemu_get_8s(f, &asid);
+        env->tlb->mmu.r4k.tlb[i].ASID = asid;
+        qemu_get_be16s(f, &flags);
+        env->tlb->mmu.r4k.tlb[i].G = (flags >> 10) & 1;
+        env->tlb->mmu.r4k.tlb[i].C0 = (flags >> 7) & 3;
+        env->tlb->mmu.r4k.tlb[i].C1 = (flags >> 4) & 3;
+        env->tlb->mmu.r4k.tlb[i].V0 = (flags >> 3) & 1;
+        env->tlb->mmu.r4k.tlb[i].V1 = (flags >> 2) & 1;
+        env->tlb->mmu.r4k.tlb[i].D0 = (flags >> 1) & 1;
+        env->tlb->mmu.r4k.tlb[i].D1 = (flags >> 0) & 1;
+        qemu_get_betls(f, &env->tlb->mmu.r4k.tlb[i].PFN[0]);
+        qemu_get_betls(f, &env->tlb->mmu.r4k.tlb[i].PFN[1]);
+    }
+
+    /* Load CPU metastate */
+    qemu_get_be32s(f, &env->current_tc);
+    qemu_get_be32s(f, &env->current_fpu);
+    qemu_get_sbe32s(f, &env->error_code);
+    qemu_get_be32s(f, &env->hflags);
+    qemu_get_betls(f, &env->btarget);
+    qemu_get_sbe32s(f, &i);
+    env->bcond = i;
+
+    /* Load remaining CP1 registers */
+    qemu_get_sbe32s(f, &env->CP0_Index);
+    qemu_get_sbe32s(f, &env->CP0_Random);
+    qemu_get_sbe32s(f, &env->CP0_VPEControl);
+    qemu_get_sbe32s(f, &env->CP0_VPEConf0);
+    qemu_get_sbe32s(f, &env->CP0_VPEConf1);
+    qemu_get_betls(f, &env->CP0_YQMask);
+    qemu_get_betls(f, &env->CP0_VPESchedule);
+    qemu_get_betls(f, &env->CP0_VPEScheFBack);
+    qemu_get_sbe32s(f, &env->CP0_VPEOpt);
+    qemu_get_betls(f, &env->CP0_EntryLo0);
+    qemu_get_betls(f, &env->CP0_EntryLo1);
+    qemu_get_betls(f, &env->CP0_Context);
+    qemu_get_sbe32s(f, &env->CP0_PageMask);
+    qemu_get_sbe32s(f, &env->CP0_PageGrain);
+    qemu_get_sbe32s(f, &env->CP0_Wired);
+    qemu_get_sbe32s(f, &env->CP0_SRSConf0);
+    qemu_get_sbe32s(f, &env->CP0_SRSConf1);
+    qemu_get_sbe32s(f, &env->CP0_SRSConf2);
+    qemu_get_sbe32s(f, &env->CP0_SRSConf3);
+    qemu_get_sbe32s(f, &env->CP0_SRSConf4);
+    qemu_get_sbe32s(f, &env->CP0_HWREna);
+    qemu_get_betls(f, &env->CP0_BadVAddr);
+    qemu_get_sbe32s(f, &env->CP0_Count);
+    qemu_get_betls(f, &env->CP0_EntryHi);
+    qemu_get_sbe32s(f, &env->CP0_Compare);
+    qemu_get_sbe32s(f, &env->CP0_Status);
+    qemu_get_sbe32s(f, &env->CP0_IntCtl);
+    qemu_get_sbe32s(f, &env->CP0_SRSCtl);
+    qemu_get_sbe32s(f, &env->CP0_SRSMap);
+    qemu_get_sbe32s(f, &env->CP0_Cause);
+    qemu_get_betls(f, &env->CP0_EPC);
+    qemu_get_sbe32s(f, &env->CP0_PRid);
+    qemu_get_sbe32s(f, &env->CP0_EBase);
+    qemu_get_sbe32s(f, &env->CP0_Config0);
+    qemu_get_sbe32s(f, &env->CP0_Config1);
+    qemu_get_sbe32s(f, &env->CP0_Config2);
+    qemu_get_sbe32s(f, &env->CP0_Config3);
+    qemu_get_sbe32s(f, &env->CP0_Config6);
+    qemu_get_sbe32s(f, &env->CP0_Config7);
+    qemu_get_betls(f, &env->lladdr);
+    for(i = 0; i < 8; i++)
+        qemu_get_betls(f, &env->CP0_WatchLo[i]);
+    for(i = 0; i < 8; i++)
+        qemu_get_sbe32s(f, &env->CP0_WatchHi[i]);
+    qemu_get_betls(f, &env->CP0_XContext);
+    qemu_get_sbe32s(f, &env->CP0_Framemask);
+    qemu_get_sbe32s(f, &env->CP0_Debug);
+    qemu_get_betls(f, &env->CP0_DEPC);
+    qemu_get_sbe32s(f, &env->CP0_Performance0);
+    qemu_get_sbe32s(f, &env->CP0_TagLo);
+    qemu_get_sbe32s(f, &env->CP0_DataLo);
+    qemu_get_sbe32s(f, &env->CP0_TagHi);
+    qemu_get_sbe32s(f, &env->CP0_DataHi);
+    qemu_get_betls(f, &env->CP0_ErrorEPC);
+    qemu_get_sbe32s(f, &env->CP0_DESAVE);
+
+    /* Load inactive TC state */
+    for (i = 0; i < MIPS_SHADOW_SET_MAX; i++)
+        load_tc(f, &env->tcs[i]);
+    for (i = 0; i < MIPS_FPU_MAX; i++)
+        load_fpu(f, &env->fpus[i]);
+
+    /* XXX: ensure compatiblity for halted bit ? */
+    tlb_flush(env, 1);
+    return 0;
+}
diff --git a/target-mips/mips-defs.h b/target-mips/mips-defs.h
new file mode 100644
index 0000000..54e80f1
--- /dev/null
+++ b/target-mips/mips-defs.h
@@ -0,0 +1,63 @@
+#if !defined (__QEMU_MIPS_DEFS_H__)
+#define __QEMU_MIPS_DEFS_H__
+
+/* If we want to use host float regs... */
+//#define USE_HOST_FLOAT_REGS
+
+/* Real pages are variable size... */
+#define TARGET_PAGE_BITS 12
+#define MIPS_TLB_MAX 128
+
+#if defined(TARGET_MIPS64)
+#define TARGET_LONG_BITS 64
+#else
+#define TARGET_LONG_BITS 32
+#endif
+
+/* Masks used to mark instructions to indicate which ISA level they
+   were introduced in. */
+#define		ISA_MIPS1	0x00000001
+#define		ISA_MIPS2	0x00000002
+#define		ISA_MIPS3	0x00000004
+#define		ISA_MIPS4	0x00000008
+#define		ISA_MIPS5	0x00000010
+#define		ISA_MIPS32	0x00000020
+#define		ISA_MIPS32R2	0x00000040
+#define		ISA_MIPS64	0x00000080
+#define		ISA_MIPS64R2	0x00000100
+
+/* MIPS ASEs. */
+#define		ASE_MIPS16	0x00001000
+#define		ASE_MIPS3D	0x00002000
+#define		ASE_MDMX	0x00004000
+#define		ASE_DSP		0x00008000
+#define		ASE_DSPR2	0x00010000
+#define		ASE_MT		0x00020000
+#define		ASE_SMARTMIPS	0x00040000
+
+/* Chip specific instructions. */
+#define		INSN_VR54XX	0x80000000
+
+/* MIPS CPU defines. */
+#define		CPU_MIPS1	(ISA_MIPS1)
+#define		CPU_MIPS2	(CPU_MIPS1 | ISA_MIPS2)
+#define		CPU_MIPS3	(CPU_MIPS2 | ISA_MIPS3)
+#define		CPU_MIPS4	(CPU_MIPS3 | ISA_MIPS4)
+#define		CPU_VR54XX	(CPU_MIPS4 | INSN_VR54XX)
+
+#define		CPU_MIPS5	(CPU_MIPS4 | ISA_MIPS5)
+
+/* MIPS Technologies "Release 1" */
+#define		CPU_MIPS32	(CPU_MIPS2 | ISA_MIPS32)
+#define		CPU_MIPS64	(CPU_MIPS5 | CPU_MIPS32 | ISA_MIPS64)
+
+/* MIPS Technologies "Release 2" */
+#define		CPU_MIPS32R2	(CPU_MIPS32 | ISA_MIPS32R2)
+#define		CPU_MIPS64R2	(CPU_MIPS64 | CPU_MIPS32R2 | ISA_MIPS64R2)
+
+/* Strictly follow the architecture standard:
+   - Disallow "special" instruction handling for PMON/SPIM.
+   Note that we still maintain Count/Compare to match the host clock. */
+//#define MIPS_STRICT_STANDARD 1
+
+#endif /* !defined (__QEMU_MIPS_DEFS_H__) */
diff --git a/target-mips/op_helper.c b/target-mips/op_helper.c
new file mode 100644
index 0000000..dda9e82
--- /dev/null
+++ b/target-mips/op_helper.c
@@ -0,0 +1,2918 @@
+/*
+ *  MIPS emulation helpers for qemu.
+ *
+ *  Copyright (c) 2004-2005 Jocelyn Mayer
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+#include <stdlib.h>
+#include "exec.h"
+
+#include "host-utils.h"
+
+#include "helper.h"
+/*****************************************************************************/
+/* Exceptions processing helpers */
+
+void helper_raise_exception_err (uint32_t exception, int error_code)
+{
+#if 1
+    if (exception < 0x100)
+        qemu_log("%s: %d %d\n", __func__, exception, error_code);
+#endif
+    env->exception_index = exception;
+    env->error_code = error_code;
+    cpu_loop_exit();
+}
+
+void helper_raise_exception (uint32_t exception)
+{
+    helper_raise_exception_err(exception, 0);
+}
+
+void helper_interrupt_restart (void)
+{
+    if (!(env->CP0_Status & (1 << CP0St_EXL)) &&
+        !(env->CP0_Status & (1 << CP0St_ERL)) &&
+        !(env->hflags & MIPS_HFLAG_DM) &&
+        (env->CP0_Status & (1 << CP0St_IE)) &&
+        (env->CP0_Status & env->CP0_Cause & CP0Ca_IP_mask)) {
+        env->CP0_Cause &= ~(0x1f << CP0Ca_EC);
+        helper_raise_exception(EXCP_EXT_INTERRUPT);
+    }
+}
+
+#if !defined(CONFIG_USER_ONLY)
+static void do_restore_state (void *pc_ptr)
+{
+    TranslationBlock *tb;
+    unsigned long pc = (unsigned long) pc_ptr;
+    
+    tb = tb_find_pc (pc);
+    if (tb) {
+        cpu_restore_state (tb, env, pc, NULL);
+    }
+}
+#endif
+
+#if defined(CONFIG_USER_ONLY)
+#define HELPER_LD(name, insn, type)                                     \
+static inline type do_##name(target_ulong addr, int mem_idx)            \
+{                                                                       \
+    return (type) insn##_raw(addr);                                     \
+}
+#else
+#define HELPER_LD(name, insn, type)                                     \
+static inline type do_##name(target_ulong addr, int mem_idx)            \
+{                                                                       \
+    switch (mem_idx)                                                    \
+    {                                                                   \
+    case 0: return (type) insn##_kernel(addr); break;                   \
+    case 1: return (type) insn##_super(addr); break;                    \
+    default:                                                            \
+    case 2: return (type) insn##_user(addr); break;                     \
+    }                                                                   \
+}
+#endif
+HELPER_LD(lbu, ldub, uint8_t)
+HELPER_LD(lw, ldl, int32_t)
+#ifdef TARGET_MIPS64
+HELPER_LD(ld, ldq, int64_t)
+#endif
+#undef HELPER_LD
+
+#if defined(CONFIG_USER_ONLY)
+#define HELPER_ST(name, insn, type)                                     \
+static inline void do_##name(target_ulong addr, type val, int mem_idx)  \
+{                                                                       \
+    insn##_raw(addr, val);                                              \
+}
+#else
+#define HELPER_ST(name, insn, type)                                     \
+static inline void do_##name(target_ulong addr, type val, int mem_idx)  \
+{                                                                       \
+    switch (mem_idx)                                                    \
+    {                                                                   \
+    case 0: insn##_kernel(addr, val); break;                            \
+    case 1: insn##_super(addr, val); break;                             \
+    default:                                                            \
+    case 2: insn##_user(addr, val); break;                              \
+    }                                                                   \
+}
+#endif
+HELPER_ST(sb, stb, uint8_t)
+HELPER_ST(sw, stl, uint32_t)
+#ifdef TARGET_MIPS64
+HELPER_ST(sd, stq, uint64_t)
+#endif
+#undef HELPER_ST
+
+target_ulong helper_clo (target_ulong arg1)
+{
+    return clo32(arg1);
+}
+
+target_ulong helper_clz (target_ulong arg1)
+{
+    return clz32(arg1);
+}
+
+#if defined(TARGET_MIPS64)
+target_ulong helper_dclo (target_ulong arg1)
+{
+    return clo64(arg1);
+}
+
+target_ulong helper_dclz (target_ulong arg1)
+{
+    return clz64(arg1);
+}
+#endif /* TARGET_MIPS64 */
+
+/* 64 bits arithmetic for 32 bits hosts */
+static inline uint64_t get_HILO (void)
+{
+    return ((uint64_t)(env->active_tc.HI[0]) << 32) | (uint32_t)env->active_tc.LO[0];
+}
+
+static inline void set_HILO (uint64_t HILO)
+{
+    env->active_tc.LO[0] = (int32_t)HILO;
+    env->active_tc.HI[0] = (int32_t)(HILO >> 32);
+}
+
+static inline void set_HIT0_LO (target_ulong arg1, uint64_t HILO)
+{
+    env->active_tc.LO[0] = (int32_t)(HILO & 0xFFFFFFFF);
+    arg1 = env->active_tc.HI[0] = (int32_t)(HILO >> 32);
+}
+
+static inline void set_HI_LOT0 (target_ulong arg1, uint64_t HILO)
+{
+    arg1 = env->active_tc.LO[0] = (int32_t)(HILO & 0xFFFFFFFF);
+    env->active_tc.HI[0] = (int32_t)(HILO >> 32);
+}
+
+/* Multiplication variants of the vr54xx. */
+target_ulong helper_muls (target_ulong arg1, target_ulong arg2)
+{
+    set_HI_LOT0(arg1, 0 - ((int64_t)(int32_t)arg1 * (int64_t)(int32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_mulsu (target_ulong arg1, target_ulong arg2)
+{
+    set_HI_LOT0(arg1, 0 - ((uint64_t)(uint32_t)arg1 * (uint64_t)(uint32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_macc (target_ulong arg1, target_ulong arg2)
+{
+    set_HI_LOT0(arg1, ((int64_t)get_HILO()) + ((int64_t)(int32_t)arg1 * (int64_t)(int32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_macchi (target_ulong arg1, target_ulong arg2)
+{
+    set_HIT0_LO(arg1, ((int64_t)get_HILO()) + ((int64_t)(int32_t)arg1 * (int64_t)(int32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_maccu (target_ulong arg1, target_ulong arg2)
+{
+    set_HI_LOT0(arg1, ((uint64_t)get_HILO()) + ((uint64_t)(uint32_t)arg1 * (uint64_t)(uint32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_macchiu (target_ulong arg1, target_ulong arg2)
+{
+    set_HIT0_LO(arg1, ((uint64_t)get_HILO()) + ((uint64_t)(uint32_t)arg1 * (uint64_t)(uint32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_msac (target_ulong arg1, target_ulong arg2)
+{
+    set_HI_LOT0(arg1, ((int64_t)get_HILO()) - ((int64_t)(int32_t)arg1 * (int64_t)(int32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_msachi (target_ulong arg1, target_ulong arg2)
+{
+    set_HIT0_LO(arg1, ((int64_t)get_HILO()) - ((int64_t)(int32_t)arg1 * (int64_t)(int32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_msacu (target_ulong arg1, target_ulong arg2)
+{
+    set_HI_LOT0(arg1, ((uint64_t)get_HILO()) - ((uint64_t)(uint32_t)arg1 * (uint64_t)(uint32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_msachiu (target_ulong arg1, target_ulong arg2)
+{
+    set_HIT0_LO(arg1, ((uint64_t)get_HILO()) - ((uint64_t)(uint32_t)arg1 * (uint64_t)(uint32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_mulhi (target_ulong arg1, target_ulong arg2)
+{
+    set_HIT0_LO(arg1, (int64_t)(int32_t)arg1 * (int64_t)(int32_t)arg2);
+
+    return arg1;
+}
+
+target_ulong helper_mulhiu (target_ulong arg1, target_ulong arg2)
+{
+    set_HIT0_LO(arg1, (uint64_t)(uint32_t)arg1 * (uint64_t)(uint32_t)arg2);
+
+    return arg1;
+}
+
+target_ulong helper_mulshi (target_ulong arg1, target_ulong arg2)
+{
+    set_HIT0_LO(arg1, 0 - ((int64_t)(int32_t)arg1 * (int64_t)(int32_t)arg2));
+
+    return arg1;
+}
+
+target_ulong helper_mulshiu (target_ulong arg1, target_ulong arg2)
+{
+    set_HIT0_LO(arg1, 0 - ((uint64_t)(uint32_t)arg1 * (uint64_t)(uint32_t)arg2));
+
+    return arg1;
+}
+
+#ifdef TARGET_MIPS64
+void helper_dmult (target_ulong arg1, target_ulong arg2)
+{
+    muls64(&(env->active_tc.LO[0]), &(env->active_tc.HI[0]), arg1, arg2);
+}
+
+void helper_dmultu (target_ulong arg1, target_ulong arg2)
+{
+    mulu64(&(env->active_tc.LO[0]), &(env->active_tc.HI[0]), arg1, arg2);
+}
+#endif
+
+#ifndef CONFIG_USER_ONLY
+
+static inline target_phys_addr_t do_translate_address(target_ulong address, int rw)
+{
+    target_phys_addr_t lladdr;
+
+    lladdr = cpu_mips_translate_address(env, address, rw);
+
+    if (lladdr == -1LL) {
+        cpu_loop_exit();
+    } else {
+        return lladdr;
+    }
+}
+
+#define HELPER_LD_ATOMIC(name, insn)                                          \
+target_ulong helper_##name(target_ulong arg, int mem_idx)                     \
+{                                                                             \
+    env->lladdr = do_translate_address(arg, 0);                               \
+    env->llval = do_##insn(arg, mem_idx);                                     \
+    return env->llval;                                                        \
+}
+HELPER_LD_ATOMIC(ll, lw)
+#ifdef TARGET_MIPS64
+HELPER_LD_ATOMIC(lld, ld)
+#endif
+#undef HELPER_LD_ATOMIC
+
+#define HELPER_ST_ATOMIC(name, ld_insn, st_insn, almask)                      \
+target_ulong helper_##name(target_ulong arg1, target_ulong arg2, int mem_idx) \
+{                                                                             \
+    target_long tmp;                                                          \
+                                                                              \
+    if (arg2 & almask) {                                                      \
+        env->CP0_BadVAddr = arg2;                                             \
+        helper_raise_exception(EXCP_AdES);                                    \
+    }                                                                         \
+    if (do_translate_address(arg2, 1) == env->lladdr) {                       \
+        tmp = do_##ld_insn(arg2, mem_idx);                                    \
+        if (tmp == env->llval) {                                              \
+            do_##st_insn(arg2, arg1, mem_idx);                                \
+            return 1;                                                         \
+        }                                                                     \
+    }                                                                         \
+    return 0;                                                                 \
+}
+HELPER_ST_ATOMIC(sc, lw, sw, 0x3)
+#ifdef TARGET_MIPS64
+HELPER_ST_ATOMIC(scd, ld, sd, 0x7)
+#endif
+#undef HELPER_ST_ATOMIC
+#endif
+
+#ifdef TARGET_WORDS_BIGENDIAN
+#define GET_LMASK(v) ((v) & 3)
+#define GET_OFFSET(addr, offset) (addr + (offset))
+#else
+#define GET_LMASK(v) (((v) & 3) ^ 3)
+#define GET_OFFSET(addr, offset) (addr - (offset))
+#endif
+
+target_ulong helper_lwl(target_ulong arg1, target_ulong arg2, int mem_idx)
+{
+    target_ulong tmp;
+
+    tmp = do_lbu(arg2, mem_idx);
+    arg1 = (arg1 & 0x00FFFFFF) | (tmp << 24);
+
+    if (GET_LMASK(arg2) <= 2) {
+        tmp = do_lbu(GET_OFFSET(arg2, 1), mem_idx);
+        arg1 = (arg1 & 0xFF00FFFF) | (tmp << 16);
+    }
+
+    if (GET_LMASK(arg2) <= 1) {
+        tmp = do_lbu(GET_OFFSET(arg2, 2), mem_idx);
+        arg1 = (arg1 & 0xFFFF00FF) | (tmp << 8);
+    }
+
+    if (GET_LMASK(arg2) == 0) {
+        tmp = do_lbu(GET_OFFSET(arg2, 3), mem_idx);
+        arg1 = (arg1 & 0xFFFFFF00) | tmp;
+    }
+    return (int32_t)arg1;
+}
+
+target_ulong helper_lwr(target_ulong arg1, target_ulong arg2, int mem_idx)
+{
+    target_ulong tmp;
+
+    tmp = do_lbu(arg2, mem_idx);
+    arg1 = (arg1 & 0xFFFFFF00) | tmp;
+
+    if (GET_LMASK(arg2) >= 1) {
+        tmp = do_lbu(GET_OFFSET(arg2, -1), mem_idx);
+        arg1 = (arg1 & 0xFFFF00FF) | (tmp << 8);
+    }
+
+    if (GET_LMASK(arg2) >= 2) {
+        tmp = do_lbu(GET_OFFSET(arg2, -2), mem_idx);
+        arg1 = (arg1 & 0xFF00FFFF) | (tmp << 16);
+    }
+
+    if (GET_LMASK(arg2) == 3) {
+        tmp = do_lbu(GET_OFFSET(arg2, -3), mem_idx);
+        arg1 = (arg1 & 0x00FFFFFF) | (tmp << 24);
+    }
+    return (int32_t)arg1;
+}
+
+void helper_swl(target_ulong arg1, target_ulong arg2, int mem_idx)
+{
+    do_sb(arg2, (uint8_t)(arg1 >> 24), mem_idx);
+
+    if (GET_LMASK(arg2) <= 2)
+        do_sb(GET_OFFSET(arg2, 1), (uint8_t)(arg1 >> 16), mem_idx);
+
+    if (GET_LMASK(arg2) <= 1)
+        do_sb(GET_OFFSET(arg2, 2), (uint8_t)(arg1 >> 8), mem_idx);
+
+    if (GET_LMASK(arg2) == 0)
+        do_sb(GET_OFFSET(arg2, 3), (uint8_t)arg1, mem_idx);
+}
+
+void helper_swr(target_ulong arg1, target_ulong arg2, int mem_idx)
+{
+    do_sb(arg2, (uint8_t)arg1, mem_idx);
+
+    if (GET_LMASK(arg2) >= 1)
+        do_sb(GET_OFFSET(arg2, -1), (uint8_t)(arg1 >> 8), mem_idx);
+
+    if (GET_LMASK(arg2) >= 2)
+        do_sb(GET_OFFSET(arg2, -2), (uint8_t)(arg1 >> 16), mem_idx);
+
+    if (GET_LMASK(arg2) == 3)
+        do_sb(GET_OFFSET(arg2, -3), (uint8_t)(arg1 >> 24), mem_idx);
+}
+
+#if defined(TARGET_MIPS64)
+/* "half" load and stores.  We must do the memory access inline,
+   or fault handling won't work.  */
+
+#ifdef TARGET_WORDS_BIGENDIAN
+#define GET_LMASK64(v) ((v) & 7)
+#else
+#define GET_LMASK64(v) (((v) & 7) ^ 7)
+#endif
+
+target_ulong helper_ldl(target_ulong arg1, target_ulong arg2, int mem_idx)
+{
+    uint64_t tmp;
+
+    tmp = do_lbu(arg2, mem_idx);
+    arg1 = (arg1 & 0x00FFFFFFFFFFFFFFULL) | (tmp << 56);
+
+    if (GET_LMASK64(arg2) <= 6) {
+        tmp = do_lbu(GET_OFFSET(arg2, 1), mem_idx);
+        arg1 = (arg1 & 0xFF00FFFFFFFFFFFFULL) | (tmp << 48);
+    }
+
+    if (GET_LMASK64(arg2) <= 5) {
+        tmp = do_lbu(GET_OFFSET(arg2, 2), mem_idx);
+        arg1 = (arg1 & 0xFFFF00FFFFFFFFFFULL) | (tmp << 40);
+    }
+
+    if (GET_LMASK64(arg2) <= 4) {
+        tmp = do_lbu(GET_OFFSET(arg2, 3), mem_idx);
+        arg1 = (arg1 & 0xFFFFFF00FFFFFFFFULL) | (tmp << 32);
+    }
+
+    if (GET_LMASK64(arg2) <= 3) {
+        tmp = do_lbu(GET_OFFSET(arg2, 4), mem_idx);
+        arg1 = (arg1 & 0xFFFFFFFF00FFFFFFULL) | (tmp << 24);
+    }
+
+    if (GET_LMASK64(arg2) <= 2) {
+        tmp = do_lbu(GET_OFFSET(arg2, 5), mem_idx);
+        arg1 = (arg1 & 0xFFFFFFFFFF00FFFFULL) | (tmp << 16);
+    }
+
+    if (GET_LMASK64(arg2) <= 1) {
+        tmp = do_lbu(GET_OFFSET(arg2, 6), mem_idx);
+        arg1 = (arg1 & 0xFFFFFFFFFFFF00FFULL) | (tmp << 8);
+    }
+
+    if (GET_LMASK64(arg2) == 0) {
+        tmp = do_lbu(GET_OFFSET(arg2, 7), mem_idx);
+        arg1 = (arg1 & 0xFFFFFFFFFFFFFF00ULL) | tmp;
+    }
+
+    return arg1;
+}
+
+target_ulong helper_ldr(target_ulong arg1, target_ulong arg2, int mem_idx)
+{
+    uint64_t tmp;
+
+    tmp = do_lbu(arg2, mem_idx);
+    arg1 = (arg1 & 0xFFFFFFFFFFFFFF00ULL) | tmp;
+
+    if (GET_LMASK64(arg2) >= 1) {
+        tmp = do_lbu(GET_OFFSET(arg2, -1), mem_idx);
+        arg1 = (arg1 & 0xFFFFFFFFFFFF00FFULL) | (tmp  << 8);
+    }
+
+    if (GET_LMASK64(arg2) >= 2) {
+        tmp = do_lbu(GET_OFFSET(arg2, -2), mem_idx);
+        arg1 = (arg1 & 0xFFFFFFFFFF00FFFFULL) | (tmp << 16);
+    }
+
+    if (GET_LMASK64(arg2) >= 3) {
+        tmp = do_lbu(GET_OFFSET(arg2, -3), mem_idx);
+        arg1 = (arg1 & 0xFFFFFFFF00FFFFFFULL) | (tmp << 24);
+    }
+
+    if (GET_LMASK64(arg2) >= 4) {
+        tmp = do_lbu(GET_OFFSET(arg2, -4), mem_idx);
+        arg1 = (arg1 & 0xFFFFFF00FFFFFFFFULL) | (tmp << 32);
+    }
+
+    if (GET_LMASK64(arg2) >= 5) {
+        tmp = do_lbu(GET_OFFSET(arg2, -5), mem_idx);
+        arg1 = (arg1 & 0xFFFF00FFFFFFFFFFULL) | (tmp << 40);
+    }
+
+    if (GET_LMASK64(arg2) >= 6) {
+        tmp = do_lbu(GET_OFFSET(arg2, -6), mem_idx);
+        arg1 = (arg1 & 0xFF00FFFFFFFFFFFFULL) | (tmp << 48);
+    }
+
+    if (GET_LMASK64(arg2) == 7) {
+        tmp = do_lbu(GET_OFFSET(arg2, -7), mem_idx);
+        arg1 = (arg1 & 0x00FFFFFFFFFFFFFFULL) | (tmp << 56);
+    }
+
+    return arg1;
+}
+
+void helper_sdl(target_ulong arg1, target_ulong arg2, int mem_idx)
+{
+    do_sb(arg2, (uint8_t)(arg1 >> 56), mem_idx);
+
+    if (GET_LMASK64(arg2) <= 6)
+        do_sb(GET_OFFSET(arg2, 1), (uint8_t)(arg1 >> 48), mem_idx);
+
+    if (GET_LMASK64(arg2) <= 5)
+        do_sb(GET_OFFSET(arg2, 2), (uint8_t)(arg1 >> 40), mem_idx);
+
+    if (GET_LMASK64(arg2) <= 4)
+        do_sb(GET_OFFSET(arg2, 3), (uint8_t)(arg1 >> 32), mem_idx);
+
+    if (GET_LMASK64(arg2) <= 3)
+        do_sb(GET_OFFSET(arg2, 4), (uint8_t)(arg1 >> 24), mem_idx);
+
+    if (GET_LMASK64(arg2) <= 2)
+        do_sb(GET_OFFSET(arg2, 5), (uint8_t)(arg1 >> 16), mem_idx);
+
+    if (GET_LMASK64(arg2) <= 1)
+        do_sb(GET_OFFSET(arg2, 6), (uint8_t)(arg1 >> 8), mem_idx);
+
+    if (GET_LMASK64(arg2) <= 0)
+        do_sb(GET_OFFSET(arg2, 7), (uint8_t)arg1, mem_idx);
+}
+
+void helper_sdr(target_ulong arg1, target_ulong arg2, int mem_idx)
+{
+    do_sb(arg2, (uint8_t)arg1, mem_idx);
+
+    if (GET_LMASK64(arg2) >= 1)
+        do_sb(GET_OFFSET(arg2, -1), (uint8_t)(arg1 >> 8), mem_idx);
+
+    if (GET_LMASK64(arg2) >= 2)
+        do_sb(GET_OFFSET(arg2, -2), (uint8_t)(arg1 >> 16), mem_idx);
+
+    if (GET_LMASK64(arg2) >= 3)
+        do_sb(GET_OFFSET(arg2, -3), (uint8_t)(arg1 >> 24), mem_idx);
+
+    if (GET_LMASK64(arg2) >= 4)
+        do_sb(GET_OFFSET(arg2, -4), (uint8_t)(arg1 >> 32), mem_idx);
+
+    if (GET_LMASK64(arg2) >= 5)
+        do_sb(GET_OFFSET(arg2, -5), (uint8_t)(arg1 >> 40), mem_idx);
+
+    if (GET_LMASK64(arg2) >= 6)
+        do_sb(GET_OFFSET(arg2, -6), (uint8_t)(arg1 >> 48), mem_idx);
+
+    if (GET_LMASK64(arg2) == 7)
+        do_sb(GET_OFFSET(arg2, -7), (uint8_t)(arg1 >> 56), mem_idx);
+}
+#endif /* TARGET_MIPS64 */
+
+#ifndef CONFIG_USER_ONLY
+/* CP0 helpers */
+target_ulong helper_mfc0_mvpcontrol (void)
+{
+    return env->mvp->CP0_MVPControl;
+}
+
+target_ulong helper_mfc0_mvpconf0 (void)
+{
+    return env->mvp->CP0_MVPConf0;
+}
+
+target_ulong helper_mfc0_mvpconf1 (void)
+{
+    return env->mvp->CP0_MVPConf1;
+}
+
+target_ulong helper_mfc0_random (void)
+{
+    return (int32_t)cpu_mips_get_random(env);
+}
+
+target_ulong helper_mfc0_tcstatus (void)
+{
+    return env->active_tc.CP0_TCStatus;
+}
+
+target_ulong helper_mftc0_tcstatus(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.CP0_TCStatus;
+    else
+        return env->tcs[other_tc].CP0_TCStatus;
+}
+
+target_ulong helper_mfc0_tcbind (void)
+{
+    return env->active_tc.CP0_TCBind;
+}
+
+target_ulong helper_mftc0_tcbind(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.CP0_TCBind;
+    else
+        return env->tcs[other_tc].CP0_TCBind;
+}
+
+target_ulong helper_mfc0_tcrestart (void)
+{
+    return env->active_tc.PC;
+}
+
+target_ulong helper_mftc0_tcrestart(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.PC;
+    else
+        return env->tcs[other_tc].PC;
+}
+
+target_ulong helper_mfc0_tchalt (void)
+{
+    return env->active_tc.CP0_TCHalt;
+}
+
+target_ulong helper_mftc0_tchalt(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.CP0_TCHalt;
+    else
+        return env->tcs[other_tc].CP0_TCHalt;
+}
+
+target_ulong helper_mfc0_tccontext (void)
+{
+    return env->active_tc.CP0_TCContext;
+}
+
+target_ulong helper_mftc0_tccontext(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.CP0_TCContext;
+    else
+        return env->tcs[other_tc].CP0_TCContext;
+}
+
+target_ulong helper_mfc0_tcschedule (void)
+{
+    return env->active_tc.CP0_TCSchedule;
+}
+
+target_ulong helper_mftc0_tcschedule(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.CP0_TCSchedule;
+    else
+        return env->tcs[other_tc].CP0_TCSchedule;
+}
+
+target_ulong helper_mfc0_tcschefback (void)
+{
+    return env->active_tc.CP0_TCScheFBack;
+}
+
+target_ulong helper_mftc0_tcschefback(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.CP0_TCScheFBack;
+    else
+        return env->tcs[other_tc].CP0_TCScheFBack;
+}
+
+target_ulong helper_mfc0_count (void)
+{
+    return (int32_t)cpu_mips_get_count(env);
+}
+
+target_ulong helper_mftc0_entryhi(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+    int32_t tcstatus;
+
+    if (other_tc == env->current_tc)
+        tcstatus = env->active_tc.CP0_TCStatus;
+    else
+        tcstatus = env->tcs[other_tc].CP0_TCStatus;
+
+    return (env->CP0_EntryHi & ~0xff) | (tcstatus & 0xff);
+}
+
+target_ulong helper_mftc0_status(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+    target_ulong t0;
+    int32_t tcstatus;
+
+    if (other_tc == env->current_tc)
+        tcstatus = env->active_tc.CP0_TCStatus;
+    else
+        tcstatus = env->tcs[other_tc].CP0_TCStatus;
+
+    t0 = env->CP0_Status & ~0xf1000018;
+    t0 |= tcstatus & (0xf << CP0TCSt_TCU0);
+    t0 |= (tcstatus & (1 << CP0TCSt_TMX)) >> (CP0TCSt_TMX - CP0St_MX);
+    t0 |= (tcstatus & (0x3 << CP0TCSt_TKSU)) >> (CP0TCSt_TKSU - CP0St_KSU);
+
+    return t0;
+}
+
+target_ulong helper_mfc0_lladdr (void)
+{
+    return (int32_t)(env->lladdr >> env->CP0_LLAddr_shift);
+}
+
+target_ulong helper_mfc0_watchlo (uint32_t sel)
+{
+    return (int32_t)env->CP0_WatchLo[sel];
+}
+
+target_ulong helper_mfc0_watchhi (uint32_t sel)
+{
+    return env->CP0_WatchHi[sel];
+}
+
+target_ulong helper_mfc0_debug (void)
+{
+    target_ulong t0 = env->CP0_Debug;
+    if (env->hflags & MIPS_HFLAG_DM)
+        t0 |= 1 << CP0DB_DM;
+
+    return t0;
+}
+
+target_ulong helper_mftc0_debug(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+    int32_t tcstatus;
+
+    if (other_tc == env->current_tc)
+        tcstatus = env->active_tc.CP0_Debug_tcstatus;
+    else
+        tcstatus = env->tcs[other_tc].CP0_Debug_tcstatus;
+
+    /* XXX: Might be wrong, check with EJTAG spec. */
+    return (env->CP0_Debug & ~((1 << CP0DB_SSt) | (1 << CP0DB_Halt))) |
+            (tcstatus & ((1 << CP0DB_SSt) | (1 << CP0DB_Halt)));
+}
+
+#if defined(TARGET_MIPS64)
+target_ulong helper_dmfc0_tcrestart (void)
+{
+    return env->active_tc.PC;
+}
+
+target_ulong helper_dmfc0_tchalt (void)
+{
+    return env->active_tc.CP0_TCHalt;
+}
+
+target_ulong helper_dmfc0_tccontext (void)
+{
+    return env->active_tc.CP0_TCContext;
+}
+
+target_ulong helper_dmfc0_tcschedule (void)
+{
+    return env->active_tc.CP0_TCSchedule;
+}
+
+target_ulong helper_dmfc0_tcschefback (void)
+{
+    return env->active_tc.CP0_TCScheFBack;
+}
+
+target_ulong helper_dmfc0_lladdr (void)
+{
+    return env->lladdr >> env->CP0_LLAddr_shift;
+}
+
+target_ulong helper_dmfc0_watchlo (uint32_t sel)
+{
+    return env->CP0_WatchLo[sel];
+}
+#endif /* TARGET_MIPS64 */
+
+void helper_mtc0_index (target_ulong arg1)
+{
+    int num = 1;
+    unsigned int tmp = env->tlb->nb_tlb;
+
+    do {
+        tmp >>= 1;
+        num <<= 1;
+    } while (tmp);
+    env->CP0_Index = (env->CP0_Index & 0x80000000) | (arg1 & (num - 1));
+}
+
+void helper_mtc0_mvpcontrol (target_ulong arg1)
+{
+    uint32_t mask = 0;
+    uint32_t newval;
+
+    if (env->CP0_VPEConf0 & (1 << CP0VPEC0_MVP))
+        mask |= (1 << CP0MVPCo_CPA) | (1 << CP0MVPCo_VPC) |
+                (1 << CP0MVPCo_EVP);
+    if (env->mvp->CP0_MVPControl & (1 << CP0MVPCo_VPC))
+        mask |= (1 << CP0MVPCo_STLB);
+    newval = (env->mvp->CP0_MVPControl & ~mask) | (arg1 & mask);
+
+    // TODO: Enable/disable shared TLB, enable/disable VPEs.
+
+    env->mvp->CP0_MVPControl = newval;
+}
+
+void helper_mtc0_vpecontrol (target_ulong arg1)
+{
+    uint32_t mask;
+    uint32_t newval;
+
+    mask = (1 << CP0VPECo_YSI) | (1 << CP0VPECo_GSI) |
+           (1 << CP0VPECo_TE) | (0xff << CP0VPECo_TargTC);
+    newval = (env->CP0_VPEControl & ~mask) | (arg1 & mask);
+
+    /* Yield scheduler intercept not implemented. */
+    /* Gating storage scheduler intercept not implemented. */
+
+    // TODO: Enable/disable TCs.
+
+    env->CP0_VPEControl = newval;
+}
+
+void helper_mtc0_vpeconf0 (target_ulong arg1)
+{
+    uint32_t mask = 0;
+    uint32_t newval;
+
+    if (env->CP0_VPEConf0 & (1 << CP0VPEC0_MVP)) {
+        if (env->CP0_VPEConf0 & (1 << CP0VPEC0_VPA))
+            mask |= (0xff << CP0VPEC0_XTC);
+        mask |= (1 << CP0VPEC0_MVP) | (1 << CP0VPEC0_VPA);
+    }
+    newval = (env->CP0_VPEConf0 & ~mask) | (arg1 & mask);
+
+    // TODO: TC exclusive handling due to ERL/EXL.
+
+    env->CP0_VPEConf0 = newval;
+}
+
+void helper_mtc0_vpeconf1 (target_ulong arg1)
+{
+    uint32_t mask = 0;
+    uint32_t newval;
+
+    if (env->mvp->CP0_MVPControl & (1 << CP0MVPCo_VPC))
+        mask |= (0xff << CP0VPEC1_NCX) | (0xff << CP0VPEC1_NCP2) |
+                (0xff << CP0VPEC1_NCP1);
+    newval = (env->CP0_VPEConf1 & ~mask) | (arg1 & mask);
+
+    /* UDI not implemented. */
+    /* CP2 not implemented. */
+
+    // TODO: Handle FPU (CP1) binding.
+
+    env->CP0_VPEConf1 = newval;
+}
+
+void helper_mtc0_yqmask (target_ulong arg1)
+{
+    /* Yield qualifier inputs not implemented. */
+    env->CP0_YQMask = 0x00000000;
+}
+
+void helper_mtc0_vpeopt (target_ulong arg1)
+{
+    env->CP0_VPEOpt = arg1 & 0x0000ffff;
+}
+
+void helper_mtc0_entrylo0 (target_ulong arg1)
+{
+    /* Large physaddr (PABITS) not implemented */
+    /* 1k pages not implemented */
+    env->CP0_EntryLo0 = arg1 & 0x3FFFFFFF;
+}
+
+void helper_mtc0_tcstatus (target_ulong arg1)
+{
+    uint32_t mask = env->CP0_TCStatus_rw_bitmask;
+    uint32_t newval;
+
+    newval = (env->active_tc.CP0_TCStatus & ~mask) | (arg1 & mask);
+
+    // TODO: Sync with CP0_Status.
+
+    env->active_tc.CP0_TCStatus = newval;
+}
+
+void helper_mttc0_tcstatus (target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    // TODO: Sync with CP0_Status.
+
+    if (other_tc == env->current_tc)
+        env->active_tc.CP0_TCStatus = arg1;
+    else
+        env->tcs[other_tc].CP0_TCStatus = arg1;
+}
+
+void helper_mtc0_tcbind (target_ulong arg1)
+{
+    uint32_t mask = (1 << CP0TCBd_TBE);
+    uint32_t newval;
+
+    if (env->mvp->CP0_MVPControl & (1 << CP0MVPCo_VPC))
+        mask |= (1 << CP0TCBd_CurVPE);
+    newval = (env->active_tc.CP0_TCBind & ~mask) | (arg1 & mask);
+    env->active_tc.CP0_TCBind = newval;
+}
+
+void helper_mttc0_tcbind (target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+    uint32_t mask = (1 << CP0TCBd_TBE);
+    uint32_t newval;
+
+    if (env->mvp->CP0_MVPControl & (1 << CP0MVPCo_VPC))
+        mask |= (1 << CP0TCBd_CurVPE);
+    if (other_tc == env->current_tc) {
+        newval = (env->active_tc.CP0_TCBind & ~mask) | (arg1 & mask);
+        env->active_tc.CP0_TCBind = newval;
+    } else {
+        newval = (env->tcs[other_tc].CP0_TCBind & ~mask) | (arg1 & mask);
+        env->tcs[other_tc].CP0_TCBind = newval;
+    }
+}
+
+void helper_mtc0_tcrestart (target_ulong arg1)
+{
+    env->active_tc.PC = arg1;
+    env->active_tc.CP0_TCStatus &= ~(1 << CP0TCSt_TDS);
+    env->lladdr = 0ULL;
+    /* MIPS16 not implemented. */
+}
+
+void helper_mttc0_tcrestart (target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc) {
+        env->active_tc.PC = arg1;
+        env->active_tc.CP0_TCStatus &= ~(1 << CP0TCSt_TDS);
+        env->lladdr = 0ULL;
+        /* MIPS16 not implemented. */
+    } else {
+        env->tcs[other_tc].PC = arg1;
+        env->tcs[other_tc].CP0_TCStatus &= ~(1 << CP0TCSt_TDS);
+        env->lladdr = 0ULL;
+        /* MIPS16 not implemented. */
+    }
+}
+
+void helper_mtc0_tchalt (target_ulong arg1)
+{
+    env->active_tc.CP0_TCHalt = arg1 & 0x1;
+
+    // TODO: Halt TC / Restart (if allocated+active) TC.
+}
+
+void helper_mttc0_tchalt (target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    // TODO: Halt TC / Restart (if allocated+active) TC.
+
+    if (other_tc == env->current_tc)
+        env->active_tc.CP0_TCHalt = arg1;
+    else
+        env->tcs[other_tc].CP0_TCHalt = arg1;
+}
+
+void helper_mtc0_tccontext (target_ulong arg1)
+{
+    env->active_tc.CP0_TCContext = arg1;
+}
+
+void helper_mttc0_tccontext (target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        env->active_tc.CP0_TCContext = arg1;
+    else
+        env->tcs[other_tc].CP0_TCContext = arg1;
+}
+
+void helper_mtc0_tcschedule (target_ulong arg1)
+{
+    env->active_tc.CP0_TCSchedule = arg1;
+}
+
+void helper_mttc0_tcschedule (target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        env->active_tc.CP0_TCSchedule = arg1;
+    else
+        env->tcs[other_tc].CP0_TCSchedule = arg1;
+}
+
+void helper_mtc0_tcschefback (target_ulong arg1)
+{
+    env->active_tc.CP0_TCScheFBack = arg1;
+}
+
+void helper_mttc0_tcschefback (target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        env->active_tc.CP0_TCScheFBack = arg1;
+    else
+        env->tcs[other_tc].CP0_TCScheFBack = arg1;
+}
+
+void helper_mtc0_entrylo1 (target_ulong arg1)
+{
+    /* Large physaddr (PABITS) not implemented */
+    /* 1k pages not implemented */
+    env->CP0_EntryLo1 = arg1 & 0x3FFFFFFF;
+}
+
+void helper_mtc0_context (target_ulong arg1)
+{
+    env->CP0_Context = (env->CP0_Context & 0x007FFFFF) | (arg1 & ~0x007FFFFF);
+}
+
+void helper_mtc0_pagemask (target_ulong arg1)
+{
+    /* 1k pages not implemented */
+    env->CP0_PageMask = arg1 & (0x1FFFFFFF & (TARGET_PAGE_MASK << 1));
+}
+
+void helper_mtc0_pagegrain (target_ulong arg1)
+{
+    /* SmartMIPS not implemented */
+    /* Large physaddr (PABITS) not implemented */
+    /* 1k pages not implemented */
+    env->CP0_PageGrain = 0;
+}
+
+void helper_mtc0_wired (target_ulong arg1)
+{
+    env->CP0_Wired = arg1 % env->tlb->nb_tlb;
+}
+
+void helper_mtc0_srsconf0 (target_ulong arg1)
+{
+    env->CP0_SRSConf0 |= arg1 & env->CP0_SRSConf0_rw_bitmask;
+}
+
+void helper_mtc0_srsconf1 (target_ulong arg1)
+{
+    env->CP0_SRSConf1 |= arg1 & env->CP0_SRSConf1_rw_bitmask;
+}
+
+void helper_mtc0_srsconf2 (target_ulong arg1)
+{
+    env->CP0_SRSConf2 |= arg1 & env->CP0_SRSConf2_rw_bitmask;
+}
+
+void helper_mtc0_srsconf3 (target_ulong arg1)
+{
+    env->CP0_SRSConf3 |= arg1 & env->CP0_SRSConf3_rw_bitmask;
+}
+
+void helper_mtc0_srsconf4 (target_ulong arg1)
+{
+    env->CP0_SRSConf4 |= arg1 & env->CP0_SRSConf4_rw_bitmask;
+}
+
+void helper_mtc0_hwrena (target_ulong arg1)
+{
+    env->CP0_HWREna = arg1 & 0x0000000F;
+}
+
+void helper_mtc0_count (target_ulong arg1)
+{
+    cpu_mips_store_count(env, arg1);
+}
+
+void helper_mtc0_entryhi (target_ulong arg1)
+{
+    target_ulong old, val;
+
+    /* 1k pages not implemented */
+    val = arg1 & ((TARGET_PAGE_MASK << 1) | 0xFF);
+#if defined(TARGET_MIPS64)
+    val &= env->SEGMask;
+#endif
+    old = env->CP0_EntryHi;
+    env->CP0_EntryHi = val;
+    if (env->CP0_Config3 & (1 << CP0C3_MT)) {
+        uint32_t tcst = env->active_tc.CP0_TCStatus & ~0xff;
+        env->active_tc.CP0_TCStatus = tcst | (val & 0xff);
+    }
+    /* If the ASID changes, flush qemu's TLB.  */
+    if ((old & 0xFF) != (val & 0xFF))
+        cpu_mips_tlb_flush(env, 1);
+}
+
+void helper_mttc0_entryhi(target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+    int32_t tcstatus;
+
+    env->CP0_EntryHi = (env->CP0_EntryHi & 0xff) | (arg1 & ~0xff);
+    if (other_tc == env->current_tc) {
+        tcstatus = (env->active_tc.CP0_TCStatus & ~0xff) | (arg1 & 0xff);
+        env->active_tc.CP0_TCStatus = tcstatus;
+    } else {
+        tcstatus = (env->tcs[other_tc].CP0_TCStatus & ~0xff) | (arg1 & 0xff);
+        env->tcs[other_tc].CP0_TCStatus = tcstatus;
+    }
+}
+
+void helper_mtc0_compare (target_ulong arg1)
+{
+    cpu_mips_store_compare(env, arg1);
+}
+
+void helper_mtc0_status (target_ulong arg1)
+{
+    uint32_t val, old;
+    uint32_t mask = env->CP0_Status_rw_bitmask;
+
+    val = arg1 & mask;
+    old = env->CP0_Status;
+    env->CP0_Status = (env->CP0_Status & ~mask) | val;
+    compute_hflags(env);
+    if (qemu_loglevel_mask(CPU_LOG_EXEC)) {
+        qemu_log("Status %08x (%08x) => %08x (%08x) Cause %08x",
+                old, old & env->CP0_Cause & CP0Ca_IP_mask,
+                val, val & env->CP0_Cause & CP0Ca_IP_mask,
+                env->CP0_Cause);
+        switch (env->hflags & MIPS_HFLAG_KSU) {
+        case MIPS_HFLAG_UM: qemu_log(", UM\n"); break;
+        case MIPS_HFLAG_SM: qemu_log(", SM\n"); break;
+        case MIPS_HFLAG_KM: qemu_log("\n"); break;
+        default: cpu_abort(env, "Invalid MMU mode!\n"); break;
+        }
+    }
+    cpu_mips_update_irq(env);
+}
+
+void helper_mttc0_status(target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+    int32_t tcstatus = env->tcs[other_tc].CP0_TCStatus;
+
+    env->CP0_Status = arg1 & ~0xf1000018;
+    tcstatus = (tcstatus & ~(0xf << CP0TCSt_TCU0)) | (arg1 & (0xf << CP0St_CU0));
+    tcstatus = (tcstatus & ~(1 << CP0TCSt_TMX)) | ((arg1 & (1 << CP0St_MX)) << (CP0TCSt_TMX - CP0St_MX));
+    tcstatus = (tcstatus & ~(0x3 << CP0TCSt_TKSU)) | ((arg1 & (0x3 << CP0St_KSU)) << (CP0TCSt_TKSU - CP0St_KSU));
+    if (other_tc == env->current_tc)
+        env->active_tc.CP0_TCStatus = tcstatus;
+    else
+        env->tcs[other_tc].CP0_TCStatus = tcstatus;
+}
+
+void helper_mtc0_intctl (target_ulong arg1)
+{
+    /* vectored interrupts not implemented, no performance counters. */
+    env->CP0_IntCtl = (env->CP0_IntCtl & ~0x000002e0) | (arg1 & 0x000002e0);
+}
+
+void helper_mtc0_srsctl (target_ulong arg1)
+{
+    uint32_t mask = (0xf << CP0SRSCtl_ESS) | (0xf << CP0SRSCtl_PSS);
+    env->CP0_SRSCtl = (env->CP0_SRSCtl & ~mask) | (arg1 & mask);
+}
+
+void helper_mtc0_cause (target_ulong arg1)
+{
+    uint32_t mask = 0x00C00300;
+    uint32_t old = env->CP0_Cause;
+
+    if (env->insn_flags & ISA_MIPS32R2)
+        mask |= 1 << CP0Ca_DC;
+
+    env->CP0_Cause = (env->CP0_Cause & ~mask) | (arg1 & mask);
+
+    if ((old ^ env->CP0_Cause) & (1 << CP0Ca_DC)) {
+        if (env->CP0_Cause & (1 << CP0Ca_DC))
+            cpu_mips_stop_count(env);
+        else
+            cpu_mips_start_count(env);
+    }
+
+    /* Handle the software interrupt as an hardware one, as they
+       are very similar */
+    if (arg1 & CP0Ca_IP_mask) {
+        cpu_mips_update_irq(env);
+    }
+}
+
+void helper_mtc0_ebase (target_ulong arg1)
+{
+    /* vectored interrupts not implemented */
+    /* Multi-CPU not implemented */
+    env->CP0_EBase = 0x80000000 | (arg1 & 0x3FFFF000);
+}
+
+void helper_mtc0_config0 (target_ulong arg1)
+{
+    env->CP0_Config0 = (env->CP0_Config0 & 0x81FFFFF8) | (arg1 & 0x00000007);
+}
+
+void helper_mtc0_config2 (target_ulong arg1)
+{
+    /* tertiary/secondary caches not implemented */
+    env->CP0_Config2 = (env->CP0_Config2 & 0x8FFF0FFF);
+}
+
+void helper_mtc0_lladdr (target_ulong arg1)
+{
+    target_long mask = env->CP0_LLAddr_rw_bitmask;
+    arg1 = arg1 << env->CP0_LLAddr_shift;
+    env->lladdr = (env->lladdr & ~mask) | (arg1 & mask);
+}
+
+void helper_mtc0_watchlo (target_ulong arg1, uint32_t sel)
+{
+    /* Watch exceptions for instructions, data loads, data stores
+       not implemented. */
+    env->CP0_WatchLo[sel] = (arg1 & ~0x7);
+}
+
+void helper_mtc0_watchhi (target_ulong arg1, uint32_t sel)
+{
+    env->CP0_WatchHi[sel] = (arg1 & 0x40FF0FF8);
+    env->CP0_WatchHi[sel] &= ~(env->CP0_WatchHi[sel] & arg1 & 0x7);
+}
+
+void helper_mtc0_xcontext (target_ulong arg1)
+{
+    target_ulong mask = (1ULL << (env->SEGBITS - 7)) - 1;
+    env->CP0_XContext = (env->CP0_XContext & mask) | (arg1 & ~mask);
+}
+
+void helper_mtc0_framemask (target_ulong arg1)
+{
+    env->CP0_Framemask = arg1; /* XXX */
+}
+
+void helper_mtc0_debug (target_ulong arg1)
+{
+    env->CP0_Debug = (env->CP0_Debug & 0x8C03FC1F) | (arg1 & 0x13300120);
+    if (arg1 & (1 << CP0DB_DM))
+        env->hflags |= MIPS_HFLAG_DM;
+    else
+        env->hflags &= ~MIPS_HFLAG_DM;
+}
+
+void helper_mttc0_debug(target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+    uint32_t val = arg1 & ((1 << CP0DB_SSt) | (1 << CP0DB_Halt));
+
+    /* XXX: Might be wrong, check with EJTAG spec. */
+    if (other_tc == env->current_tc)
+        env->active_tc.CP0_Debug_tcstatus = val;
+    else
+        env->tcs[other_tc].CP0_Debug_tcstatus = val;
+    env->CP0_Debug = (env->CP0_Debug & ((1 << CP0DB_SSt) | (1 << CP0DB_Halt))) |
+                     (arg1 & ~((1 << CP0DB_SSt) | (1 << CP0DB_Halt)));
+}
+
+void helper_mtc0_performance0 (target_ulong arg1)
+{
+    env->CP0_Performance0 = arg1 & 0x000007ff;
+}
+
+void helper_mtc0_taglo (target_ulong arg1)
+{
+    env->CP0_TagLo = arg1 & 0xFFFFFCF6;
+}
+
+void helper_mtc0_datalo (target_ulong arg1)
+{
+    env->CP0_DataLo = arg1; /* XXX */
+}
+
+void helper_mtc0_taghi (target_ulong arg1)
+{
+    env->CP0_TagHi = arg1; /* XXX */
+}
+
+void helper_mtc0_datahi (target_ulong arg1)
+{
+    env->CP0_DataHi = arg1; /* XXX */
+}
+
+/* MIPS MT functions */
+target_ulong helper_mftgpr(uint32_t sel)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.gpr[sel];
+    else
+        return env->tcs[other_tc].gpr[sel];
+}
+
+target_ulong helper_mftlo(uint32_t sel)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.LO[sel];
+    else
+        return env->tcs[other_tc].LO[sel];
+}
+
+target_ulong helper_mfthi(uint32_t sel)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.HI[sel];
+    else
+        return env->tcs[other_tc].HI[sel];
+}
+
+target_ulong helper_mftacx(uint32_t sel)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.ACX[sel];
+    else
+        return env->tcs[other_tc].ACX[sel];
+}
+
+target_ulong helper_mftdsp(void)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        return env->active_tc.DSPControl;
+    else
+        return env->tcs[other_tc].DSPControl;
+}
+
+void helper_mttgpr(target_ulong arg1, uint32_t sel)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        env->active_tc.gpr[sel] = arg1;
+    else
+        env->tcs[other_tc].gpr[sel] = arg1;
+}
+
+void helper_mttlo(target_ulong arg1, uint32_t sel)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        env->active_tc.LO[sel] = arg1;
+    else
+        env->tcs[other_tc].LO[sel] = arg1;
+}
+
+void helper_mtthi(target_ulong arg1, uint32_t sel)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        env->active_tc.HI[sel] = arg1;
+    else
+        env->tcs[other_tc].HI[sel] = arg1;
+}
+
+void helper_mttacx(target_ulong arg1, uint32_t sel)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        env->active_tc.ACX[sel] = arg1;
+    else
+        env->tcs[other_tc].ACX[sel] = arg1;
+}
+
+void helper_mttdsp(target_ulong arg1)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+
+    if (other_tc == env->current_tc)
+        env->active_tc.DSPControl = arg1;
+    else
+        env->tcs[other_tc].DSPControl = arg1;
+}
+
+/* MIPS MT functions */
+target_ulong helper_dmt(target_ulong arg1)
+{
+    // TODO
+    arg1 = 0;
+    // rt = arg1
+
+    return arg1;
+}
+
+target_ulong helper_emt(target_ulong arg1)
+{
+    // TODO
+    arg1 = 0;
+    // rt = arg1
+
+    return arg1;
+}
+
+target_ulong helper_dvpe(target_ulong arg1)
+{
+    // TODO
+    arg1 = 0;
+    // rt = arg1
+
+    return arg1;
+}
+
+target_ulong helper_evpe(target_ulong arg1)
+{
+    // TODO
+    arg1 = 0;
+    // rt = arg1
+
+    return arg1;
+}
+#endif /* !CONFIG_USER_ONLY */
+
+void helper_fork(target_ulong arg1, target_ulong arg2)
+{
+    // arg1 = rt, arg2 = rs
+    arg1 = 0;
+    // TODO: store to TC register
+}
+
+target_ulong helper_yield(target_ulong arg1)
+{
+    if (arg1 < 0) {
+        /* No scheduling policy implemented. */
+        if (arg1 != -2) {
+            if (env->CP0_VPEControl & (1 << CP0VPECo_YSI) &&
+                env->active_tc.CP0_TCStatus & (1 << CP0TCSt_DT)) {
+                env->CP0_VPEControl &= ~(0x7 << CP0VPECo_EXCPT);
+                env->CP0_VPEControl |= 4 << CP0VPECo_EXCPT;
+                helper_raise_exception(EXCP_THREAD);
+            }
+        }
+    } else if (arg1 == 0) {
+        if (0 /* TODO: TC underflow */) {
+            env->CP0_VPEControl &= ~(0x7 << CP0VPECo_EXCPT);
+            helper_raise_exception(EXCP_THREAD);
+        } else {
+            // TODO: Deallocate TC
+        }
+    } else if (arg1 > 0) {
+        /* Yield qualifier inputs not implemented. */
+        env->CP0_VPEControl &= ~(0x7 << CP0VPECo_EXCPT);
+        env->CP0_VPEControl |= 2 << CP0VPECo_EXCPT;
+        helper_raise_exception(EXCP_THREAD);
+    }
+    return env->CP0_YQMask;
+}
+
+#ifndef CONFIG_USER_ONLY
+/* TLB management */
+void cpu_mips_tlb_flush (CPUState *env, int flush_global)
+{
+    /* Flush qemu's TLB and discard all shadowed entries.  */
+    tlb_flush (env, flush_global);
+    env->tlb->tlb_in_use = env->tlb->nb_tlb;
+}
+
+static void r4k_mips_tlb_flush_extra (CPUState *env, int first)
+{
+    /* Discard entries from env->tlb[first] onwards.  */
+    while (env->tlb->tlb_in_use > first) {
+        r4k_invalidate_tlb(env, --env->tlb->tlb_in_use, 0);
+    }
+}
+
+static void r4k_fill_tlb (int idx)
+{
+    r4k_tlb_t *tlb;
+
+    /* XXX: detect conflicting TLBs and raise a MCHECK exception when needed */
+    tlb = &env->tlb->mmu.r4k.tlb[idx];
+    tlb->VPN = env->CP0_EntryHi & (TARGET_PAGE_MASK << 1);
+#if defined(TARGET_MIPS64)
+    tlb->VPN &= env->SEGMask;
+#endif
+    tlb->ASID = env->CP0_EntryHi & 0xFF;
+    tlb->PageMask = env->CP0_PageMask;
+    tlb->G = env->CP0_EntryLo0 & env->CP0_EntryLo1 & 1;
+    tlb->V0 = (env->CP0_EntryLo0 & 2) != 0;
+    tlb->D0 = (env->CP0_EntryLo0 & 4) != 0;
+    tlb->C0 = (env->CP0_EntryLo0 >> 3) & 0x7;
+    tlb->PFN[0] = (env->CP0_EntryLo0 >> 6) << 12;
+    tlb->V1 = (env->CP0_EntryLo1 & 2) != 0;
+    tlb->D1 = (env->CP0_EntryLo1 & 4) != 0;
+    tlb->C1 = (env->CP0_EntryLo1 >> 3) & 0x7;
+    tlb->PFN[1] = (env->CP0_EntryLo1 >> 6) << 12;
+}
+
+void r4k_helper_tlbwi (void)
+{
+    int idx;
+
+    idx = (env->CP0_Index & ~0x80000000) % env->tlb->nb_tlb;
+
+    /* Discard cached TLB entries.  We could avoid doing this if the
+       tlbwi is just upgrading access permissions on the current entry;
+       that might be a further win.  */
+    r4k_mips_tlb_flush_extra (env, env->tlb->nb_tlb);
+
+    r4k_invalidate_tlb(env, idx, 0);
+    r4k_fill_tlb(idx);
+}
+
+void r4k_helper_tlbwr (void)
+{
+    int r = cpu_mips_get_random(env);
+
+    r4k_invalidate_tlb(env, r, 1);
+    r4k_fill_tlb(r);
+}
+
+void r4k_helper_tlbp (void)
+{
+    r4k_tlb_t *tlb;
+    target_ulong mask;
+    target_ulong tag;
+    target_ulong VPN;
+    uint8_t ASID;
+    int i;
+
+    ASID = env->CP0_EntryHi & 0xFF;
+    for (i = 0; i < env->tlb->nb_tlb; i++) {
+        tlb = &env->tlb->mmu.r4k.tlb[i];
+        /* 1k pages are not supported. */
+        mask = tlb->PageMask | ~(TARGET_PAGE_MASK << 1);
+        tag = env->CP0_EntryHi & ~mask;
+        VPN = tlb->VPN & ~mask;
+        /* Check ASID, virtual page number & size */
+        if ((tlb->G == 1 || tlb->ASID == ASID) && VPN == tag) {
+            /* TLB match */
+            env->CP0_Index = i;
+            break;
+        }
+    }
+    if (i == env->tlb->nb_tlb) {
+        /* No match.  Discard any shadow entries, if any of them match.  */
+        for (i = env->tlb->nb_tlb; i < env->tlb->tlb_in_use; i++) {
+            tlb = &env->tlb->mmu.r4k.tlb[i];
+            /* 1k pages are not supported. */
+            mask = tlb->PageMask | ~(TARGET_PAGE_MASK << 1);
+            tag = env->CP0_EntryHi & ~mask;
+            VPN = tlb->VPN & ~mask;
+            /* Check ASID, virtual page number & size */
+            if ((tlb->G == 1 || tlb->ASID == ASID) && VPN == tag) {
+                r4k_mips_tlb_flush_extra (env, i);
+                break;
+            }
+        }
+
+        env->CP0_Index |= 0x80000000;
+    }
+}
+
+void r4k_helper_tlbr (void)
+{
+    r4k_tlb_t *tlb;
+    uint8_t ASID;
+    int idx;
+
+    ASID = env->CP0_EntryHi & 0xFF;
+    idx = (env->CP0_Index & ~0x80000000) % env->tlb->nb_tlb;
+    tlb = &env->tlb->mmu.r4k.tlb[idx];
+
+    /* If this will change the current ASID, flush qemu's TLB.  */
+    if (ASID != tlb->ASID)
+        cpu_mips_tlb_flush (env, 1);
+
+    r4k_mips_tlb_flush_extra(env, env->tlb->nb_tlb);
+
+    env->CP0_EntryHi = tlb->VPN | tlb->ASID;
+    env->CP0_PageMask = tlb->PageMask;
+    env->CP0_EntryLo0 = tlb->G | (tlb->V0 << 1) | (tlb->D0 << 2) |
+                        (tlb->C0 << 3) | (tlb->PFN[0] >> 6);
+    env->CP0_EntryLo1 = tlb->G | (tlb->V1 << 1) | (tlb->D1 << 2) |
+                        (tlb->C1 << 3) | (tlb->PFN[1] >> 6);
+}
+
+void helper_tlbwi(void)
+{
+    env->tlb->helper_tlbwi();
+}
+
+void helper_tlbwr(void)
+{
+    env->tlb->helper_tlbwr();
+}
+
+void helper_tlbp(void)
+{
+    env->tlb->helper_tlbp();
+}
+
+void helper_tlbr(void)
+{
+    env->tlb->helper_tlbr();
+}
+
+/* Specials */
+target_ulong helper_di (void)
+{
+    target_ulong t0 = env->CP0_Status;
+
+    env->CP0_Status = t0 & ~(1 << CP0St_IE);
+    cpu_mips_update_irq(env);
+
+    return t0;
+}
+
+target_ulong helper_ei (void)
+{
+    target_ulong t0 = env->CP0_Status;
+
+    env->CP0_Status = t0 | (1 << CP0St_IE);
+    cpu_mips_update_irq(env);
+
+    return t0;
+}
+
+static void debug_pre_eret (void)
+{
+    if (qemu_loglevel_mask(CPU_LOG_EXEC)) {
+        qemu_log("ERET: PC " TARGET_FMT_lx " EPC " TARGET_FMT_lx,
+                env->active_tc.PC, env->CP0_EPC);
+        if (env->CP0_Status & (1 << CP0St_ERL))
+            qemu_log(" ErrorEPC " TARGET_FMT_lx, env->CP0_ErrorEPC);
+        if (env->hflags & MIPS_HFLAG_DM)
+            qemu_log(" DEPC " TARGET_FMT_lx, env->CP0_DEPC);
+        qemu_log("\n");
+    }
+}
+
+static void debug_post_eret (void)
+{
+    if (qemu_loglevel_mask(CPU_LOG_EXEC)) {
+        qemu_log("  =>  PC " TARGET_FMT_lx " EPC " TARGET_FMT_lx,
+                env->active_tc.PC, env->CP0_EPC);
+        if (env->CP0_Status & (1 << CP0St_ERL))
+            qemu_log(" ErrorEPC " TARGET_FMT_lx, env->CP0_ErrorEPC);
+        if (env->hflags & MIPS_HFLAG_DM)
+            qemu_log(" DEPC " TARGET_FMT_lx, env->CP0_DEPC);
+        switch (env->hflags & MIPS_HFLAG_KSU) {
+        case MIPS_HFLAG_UM: qemu_log(", UM\n"); break;
+        case MIPS_HFLAG_SM: qemu_log(", SM\n"); break;
+        case MIPS_HFLAG_KM: qemu_log("\n"); break;
+        default: cpu_abort(env, "Invalid MMU mode!\n"); break;
+        }
+    }
+}
+
+void helper_eret (void)
+{
+    debug_pre_eret();
+    if (env->CP0_Status & (1 << CP0St_ERL)) {
+        env->active_tc.PC = env->CP0_ErrorEPC;
+        env->CP0_Status &= ~(1 << CP0St_ERL);
+    } else {
+        env->active_tc.PC = env->CP0_EPC;
+        env->CP0_Status &= ~(1 << CP0St_EXL);
+    }
+    compute_hflags(env);
+    debug_post_eret();
+    env->lladdr = 1;
+}
+
+void helper_deret (void)
+{
+    debug_pre_eret();
+    env->active_tc.PC = env->CP0_DEPC;
+    env->hflags &= MIPS_HFLAG_DM;
+    compute_hflags(env);
+    debug_post_eret();
+    env->lladdr = 1;
+}
+#endif /* !CONFIG_USER_ONLY */
+
+target_ulong helper_rdhwr_cpunum(void)
+{
+    if ((env->hflags & MIPS_HFLAG_CP0) ||
+        (env->CP0_HWREna & (1 << 0)))
+        return env->CP0_EBase & 0x3ff;
+    else
+        helper_raise_exception(EXCP_RI);
+
+    return 0;
+}
+
+target_ulong helper_rdhwr_synci_step(void)
+{
+    if ((env->hflags & MIPS_HFLAG_CP0) ||
+        (env->CP0_HWREna & (1 << 1)))
+        return env->SYNCI_Step;
+    else
+        helper_raise_exception(EXCP_RI);
+
+    return 0;
+}
+
+target_ulong helper_rdhwr_cc(void)
+{
+    if ((env->hflags & MIPS_HFLAG_CP0) ||
+        (env->CP0_HWREna & (1 << 2)))
+        return env->CP0_Count;
+    else
+        helper_raise_exception(EXCP_RI);
+
+    return 0;
+}
+
+target_ulong helper_rdhwr_ccres(void)
+{
+    if ((env->hflags & MIPS_HFLAG_CP0) ||
+        (env->CP0_HWREna & (1 << 3)))
+        return env->CCRes;
+    else
+        helper_raise_exception(EXCP_RI);
+
+    return 0;
+}
+
+void helper_pmon (int function)
+{
+    function /= 2;
+    switch (function) {
+    case 2: /* TODO: char inbyte(int waitflag); */
+        if (env->active_tc.gpr[4] == 0)
+            env->active_tc.gpr[2] = -1;
+        /* Fall through */
+    case 11: /* TODO: char inbyte (void); */
+        env->active_tc.gpr[2] = -1;
+        break;
+    case 3:
+    case 12:
+        printf("%c", (char)(env->active_tc.gpr[4] & 0xFF));
+        break;
+    case 17:
+        break;
+    case 158:
+        {
+            unsigned char *fmt = (void *)(unsigned long)env->active_tc.gpr[4];
+            printf("%s", fmt);
+        }
+        break;
+    }
+}
+
+void helper_wait (void)
+{
+    env->halted = 1;
+    helper_raise_exception(EXCP_HLT);
+}
+
+#if !defined(CONFIG_USER_ONLY)
+
+static void do_unaligned_access (target_ulong addr, int is_write, int is_user, void *retaddr);
+
+#define MMUSUFFIX _mmu
+#define ALIGNED_ONLY
+
+#define SHIFT 0
+#include "softmmu_template.h"
+
+#define SHIFT 1
+#include "softmmu_template.h"
+
+#define SHIFT 2
+#include "softmmu_template.h"
+
+#define SHIFT 3
+#include "softmmu_template.h"
+
+static void do_unaligned_access (target_ulong addr, int is_write, int is_user, void *retaddr)
+{
+    env->CP0_BadVAddr = addr;
+    do_restore_state (retaddr);
+    helper_raise_exception ((is_write == 1) ? EXCP_AdES : EXCP_AdEL);
+}
+
+void tlb_fill (target_ulong addr, int is_write, int mmu_idx, void *retaddr)
+{
+    TranslationBlock *tb;
+    CPUState *saved_env;
+    unsigned long pc;
+    int ret;
+
+    /* XXX: hack to restore env in all cases, even if not called from
+       generated code */
+    saved_env = env;
+    env = cpu_single_env;
+    ret = cpu_mips_handle_mmu_fault(env, addr, is_write, mmu_idx, 1);
+    if (ret) {
+        if (retaddr) {
+            /* now we have a real cpu fault */
+            pc = (unsigned long)retaddr;
+            tb = tb_find_pc(pc);
+            if (tb) {
+                /* the PC is inside the translated code. It means that we have
+                   a virtual CPU fault */
+                cpu_restore_state(tb, env, pc, NULL);
+            }
+        }
+        helper_raise_exception_err(env->exception_index, env->error_code);
+    }
+    env = saved_env;
+}
+
+void do_unassigned_access(target_phys_addr_t addr, int is_write, int is_exec,
+                          int unused, int size)
+{
+    if (is_exec)
+        helper_raise_exception(EXCP_IBE);
+    else
+        helper_raise_exception(EXCP_DBE);
+}
+#endif /* !CONFIG_USER_ONLY */
+
+/* Complex FPU operations which may need stack space. */
+
+#define FLOAT_ONE32 make_float32(0x3f8 << 20)
+#define FLOAT_ONE64 make_float64(0x3ffULL << 52)
+#define FLOAT_TWO32 make_float32(1 << 30)
+#define FLOAT_TWO64 make_float64(1ULL << 62)
+#define FLOAT_QNAN32 0x7fbfffff
+#define FLOAT_QNAN64 0x7ff7ffffffffffffULL
+#define FLOAT_SNAN32 0x7fffffff
+#define FLOAT_SNAN64 0x7fffffffffffffffULL
+
+/* convert MIPS rounding mode in FCR31 to IEEE library */
+static unsigned int ieee_rm[] = {
+    float_round_nearest_even,
+    float_round_to_zero,
+    float_round_up,
+    float_round_down
+};
+
+#define RESTORE_ROUNDING_MODE \
+    set_float_rounding_mode(ieee_rm[env->active_fpu.fcr31 & 3], &env->active_fpu.fp_status)
+
+#define RESTORE_FLUSH_MODE \
+    set_flush_to_zero((env->active_fpu.fcr31 & (1 << 24)) != 0, &env->active_fpu.fp_status);
+
+target_ulong helper_cfc1 (uint32_t reg)
+{
+    target_ulong arg1;
+
+    switch (reg) {
+    case 0:
+        arg1 = (int32_t)env->active_fpu.fcr0;
+        break;
+    case 25:
+        arg1 = ((env->active_fpu.fcr31 >> 24) & 0xfe) | ((env->active_fpu.fcr31 >> 23) & 0x1);
+        break;
+    case 26:
+        arg1 = env->active_fpu.fcr31 & 0x0003f07c;
+        break;
+    case 28:
+        arg1 = (env->active_fpu.fcr31 & 0x00000f83) | ((env->active_fpu.fcr31 >> 22) & 0x4);
+        break;
+    default:
+        arg1 = (int32_t)env->active_fpu.fcr31;
+        break;
+    }
+
+    return arg1;
+}
+
+void helper_ctc1 (target_ulong arg1, uint32_t reg)
+{
+    switch(reg) {
+    case 25:
+        if (arg1 & 0xffffff00)
+            return;
+        env->active_fpu.fcr31 = (env->active_fpu.fcr31 & 0x017fffff) | ((arg1 & 0xfe) << 24) |
+                     ((arg1 & 0x1) << 23);
+        break;
+    case 26:
+        if (arg1 & 0x007c0000)
+            return;
+        env->active_fpu.fcr31 = (env->active_fpu.fcr31 & 0xfffc0f83) | (arg1 & 0x0003f07c);
+        break;
+    case 28:
+        if (arg1 & 0x007c0000)
+            return;
+        env->active_fpu.fcr31 = (env->active_fpu.fcr31 & 0xfefff07c) | (arg1 & 0x00000f83) |
+                     ((arg1 & 0x4) << 22);
+        break;
+    case 31:
+        if (arg1 & 0x007c0000)
+            return;
+        env->active_fpu.fcr31 = arg1;
+        break;
+    default:
+        return;
+    }
+    /* set rounding mode */
+    RESTORE_ROUNDING_MODE;
+    /* set flush-to-zero mode */
+    RESTORE_FLUSH_MODE;
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    if ((GET_FP_ENABLE(env->active_fpu.fcr31) | 0x20) & GET_FP_CAUSE(env->active_fpu.fcr31))
+        helper_raise_exception(EXCP_FPE);
+}
+
+static inline char ieee_ex_to_mips(char xcpt)
+{
+    return (xcpt & float_flag_inexact) >> 5 |
+           (xcpt & float_flag_underflow) >> 3 |
+           (xcpt & float_flag_overflow) >> 1 |
+           (xcpt & float_flag_divbyzero) << 1 |
+           (xcpt & float_flag_invalid) << 4;
+}
+
+static inline char mips_ex_to_ieee(char xcpt)
+{
+    return (xcpt & FP_INEXACT) << 5 |
+           (xcpt & FP_UNDERFLOW) << 3 |
+           (xcpt & FP_OVERFLOW) << 1 |
+           (xcpt & FP_DIV0) >> 1 |
+           (xcpt & FP_INVALID) >> 4;
+}
+
+static inline void update_fcr31(void)
+{
+    int tmp = ieee_ex_to_mips(get_float_exception_flags(&env->active_fpu.fp_status));
+
+    SET_FP_CAUSE(env->active_fpu.fcr31, tmp);
+    if (GET_FP_ENABLE(env->active_fpu.fcr31) & tmp)
+        helper_raise_exception(EXCP_FPE);
+    else
+        UPDATE_FP_FLAGS(env->active_fpu.fcr31, tmp);
+}
+
+/* Float support.
+   Single precition routines have a "s" suffix, double precision a
+   "d" suffix, 32bit integer "w", 64bit integer "l", paired single "ps",
+   paired single lower "pl", paired single upper "pu".  */
+
+/* unary operations, modifying fp status  */
+uint64_t helper_float_sqrt_d(uint64_t fdt0)
+{
+    return float64_sqrt(fdt0, &env->active_fpu.fp_status);
+}
+
+uint32_t helper_float_sqrt_s(uint32_t fst0)
+{
+    return float32_sqrt(fst0, &env->active_fpu.fp_status);
+}
+
+uint64_t helper_float_cvtd_s(uint32_t fst0)
+{
+    uint64_t fdt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fdt2 = float32_to_float64(fst0, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fdt2;
+}
+
+uint64_t helper_float_cvtd_w(uint32_t wt0)
+{
+    uint64_t fdt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fdt2 = int32_to_float64(wt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fdt2;
+}
+
+uint64_t helper_float_cvtd_l(uint64_t dt0)
+{
+    uint64_t fdt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fdt2 = int64_to_float64(dt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fdt2;
+}
+
+uint64_t helper_float_cvtl_d(uint64_t fdt0)
+{
+    uint64_t dt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    dt2 = float64_to_int64(fdt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        dt2 = FLOAT_SNAN64;
+    return dt2;
+}
+
+uint64_t helper_float_cvtl_s(uint32_t fst0)
+{
+    uint64_t dt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    dt2 = float32_to_int64(fst0, &env->active_fpu.fp_status);
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        dt2 = FLOAT_SNAN64;
+    return dt2;
+}
+
+uint64_t helper_float_cvtps_pw(uint64_t dt0)
+{
+    uint32_t fst2;
+    uint32_t fsth2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = int32_to_float32(dt0 & 0XFFFFFFFF, &env->active_fpu.fp_status);
+    fsth2 = int32_to_float32(dt0 >> 32, &env->active_fpu.fp_status);
+    update_fcr31();
+    return ((uint64_t)fsth2 << 32) | fst2;
+}
+
+uint64_t helper_float_cvtpw_ps(uint64_t fdt0)
+{
+    uint32_t wt2;
+    uint32_t wth2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    wt2 = float32_to_int32(fdt0 & 0XFFFFFFFF, &env->active_fpu.fp_status);
+    wth2 = float32_to_int32(fdt0 >> 32, &env->active_fpu.fp_status);
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID)) {
+        wt2 = FLOAT_SNAN32;
+        wth2 = FLOAT_SNAN32;
+    }
+    return ((uint64_t)wth2 << 32) | wt2;
+}
+
+uint32_t helper_float_cvts_d(uint64_t fdt0)
+{
+    uint32_t fst2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float64_to_float32(fdt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fst2;
+}
+
+uint32_t helper_float_cvts_w(uint32_t wt0)
+{
+    uint32_t fst2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = int32_to_float32(wt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fst2;
+}
+
+uint32_t helper_float_cvts_l(uint64_t dt0)
+{
+    uint32_t fst2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = int64_to_float32(dt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fst2;
+}
+
+uint32_t helper_float_cvts_pl(uint32_t wt0)
+{
+    uint32_t wt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    wt2 = wt0;
+    update_fcr31();
+    return wt2;
+}
+
+uint32_t helper_float_cvts_pu(uint32_t wth0)
+{
+    uint32_t wt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    wt2 = wth0;
+    update_fcr31();
+    return wt2;
+}
+
+uint32_t helper_float_cvtw_s(uint32_t fst0)
+{
+    uint32_t wt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    wt2 = float32_to_int32(fst0, &env->active_fpu.fp_status);
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        wt2 = FLOAT_SNAN32;
+    return wt2;
+}
+
+uint32_t helper_float_cvtw_d(uint64_t fdt0)
+{
+    uint32_t wt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    wt2 = float64_to_int32(fdt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        wt2 = FLOAT_SNAN32;
+    return wt2;
+}
+
+uint64_t helper_float_roundl_d(uint64_t fdt0)
+{
+    uint64_t dt2;
+
+    set_float_rounding_mode(float_round_nearest_even, &env->active_fpu.fp_status);
+    dt2 = float64_to_int64(fdt0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        dt2 = FLOAT_SNAN64;
+    return dt2;
+}
+
+uint64_t helper_float_roundl_s(uint32_t fst0)
+{
+    uint64_t dt2;
+
+    set_float_rounding_mode(float_round_nearest_even, &env->active_fpu.fp_status);
+    dt2 = float32_to_int64(fst0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        dt2 = FLOAT_SNAN64;
+    return dt2;
+}
+
+uint32_t helper_float_roundw_d(uint64_t fdt0)
+{
+    uint32_t wt2;
+
+    set_float_rounding_mode(float_round_nearest_even, &env->active_fpu.fp_status);
+    wt2 = float64_to_int32(fdt0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        wt2 = FLOAT_SNAN32;
+    return wt2;
+}
+
+uint32_t helper_float_roundw_s(uint32_t fst0)
+{
+    uint32_t wt2;
+
+    set_float_rounding_mode(float_round_nearest_even, &env->active_fpu.fp_status);
+    wt2 = float32_to_int32(fst0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        wt2 = FLOAT_SNAN32;
+    return wt2;
+}
+
+uint64_t helper_float_truncl_d(uint64_t fdt0)
+{
+    uint64_t dt2;
+
+    dt2 = float64_to_int64_round_to_zero(fdt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        dt2 = FLOAT_SNAN64;
+    return dt2;
+}
+
+uint64_t helper_float_truncl_s(uint32_t fst0)
+{
+    uint64_t dt2;
+
+    dt2 = float32_to_int64_round_to_zero(fst0, &env->active_fpu.fp_status);
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        dt2 = FLOAT_SNAN64;
+    return dt2;
+}
+
+uint32_t helper_float_truncw_d(uint64_t fdt0)
+{
+    uint32_t wt2;
+
+    wt2 = float64_to_int32_round_to_zero(fdt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        wt2 = FLOAT_SNAN32;
+    return wt2;
+}
+
+uint32_t helper_float_truncw_s(uint32_t fst0)
+{
+    uint32_t wt2;
+
+    wt2 = float32_to_int32_round_to_zero(fst0, &env->active_fpu.fp_status);
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        wt2 = FLOAT_SNAN32;
+    return wt2;
+}
+
+uint64_t helper_float_ceill_d(uint64_t fdt0)
+{
+    uint64_t dt2;
+
+    set_float_rounding_mode(float_round_up, &env->active_fpu.fp_status);
+    dt2 = float64_to_int64(fdt0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        dt2 = FLOAT_SNAN64;
+    return dt2;
+}
+
+uint64_t helper_float_ceill_s(uint32_t fst0)
+{
+    uint64_t dt2;
+
+    set_float_rounding_mode(float_round_up, &env->active_fpu.fp_status);
+    dt2 = float32_to_int64(fst0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        dt2 = FLOAT_SNAN64;
+    return dt2;
+}
+
+uint32_t helper_float_ceilw_d(uint64_t fdt0)
+{
+    uint32_t wt2;
+
+    set_float_rounding_mode(float_round_up, &env->active_fpu.fp_status);
+    wt2 = float64_to_int32(fdt0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        wt2 = FLOAT_SNAN32;
+    return wt2;
+}
+
+uint32_t helper_float_ceilw_s(uint32_t fst0)
+{
+    uint32_t wt2;
+
+    set_float_rounding_mode(float_round_up, &env->active_fpu.fp_status);
+    wt2 = float32_to_int32(fst0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        wt2 = FLOAT_SNAN32;
+    return wt2;
+}
+
+uint64_t helper_float_floorl_d(uint64_t fdt0)
+{
+    uint64_t dt2;
+
+    set_float_rounding_mode(float_round_down, &env->active_fpu.fp_status);
+    dt2 = float64_to_int64(fdt0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        dt2 = FLOAT_SNAN64;
+    return dt2;
+}
+
+uint64_t helper_float_floorl_s(uint32_t fst0)
+{
+    uint64_t dt2;
+
+    set_float_rounding_mode(float_round_down, &env->active_fpu.fp_status);
+    dt2 = float32_to_int64(fst0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        dt2 = FLOAT_SNAN64;
+    return dt2;
+}
+
+uint32_t helper_float_floorw_d(uint64_t fdt0)
+{
+    uint32_t wt2;
+
+    set_float_rounding_mode(float_round_down, &env->active_fpu.fp_status);
+    wt2 = float64_to_int32(fdt0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        wt2 = FLOAT_SNAN32;
+    return wt2;
+}
+
+uint32_t helper_float_floorw_s(uint32_t fst0)
+{
+    uint32_t wt2;
+
+    set_float_rounding_mode(float_round_down, &env->active_fpu.fp_status);
+    wt2 = float32_to_int32(fst0, &env->active_fpu.fp_status);
+    RESTORE_ROUNDING_MODE;
+    update_fcr31();
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & (FP_OVERFLOW | FP_INVALID))
+        wt2 = FLOAT_SNAN32;
+    return wt2;
+}
+
+/* unary operations, not modifying fp status  */
+#define FLOAT_UNOP(name)                                       \
+uint64_t helper_float_ ## name ## _d(uint64_t fdt0)                \
+{                                                              \
+    return float64_ ## name(fdt0);                             \
+}                                                              \
+uint32_t helper_float_ ## name ## _s(uint32_t fst0)                \
+{                                                              \
+    return float32_ ## name(fst0);                             \
+}                                                              \
+uint64_t helper_float_ ## name ## _ps(uint64_t fdt0)               \
+{                                                              \
+    uint32_t wt0;                                              \
+    uint32_t wth0;                                             \
+                                                               \
+    wt0 = float32_ ## name(fdt0 & 0XFFFFFFFF);                 \
+    wth0 = float32_ ## name(fdt0 >> 32);                       \
+    return ((uint64_t)wth0 << 32) | wt0;                       \
+}
+FLOAT_UNOP(abs)
+FLOAT_UNOP(chs)
+#undef FLOAT_UNOP
+
+/* MIPS specific unary operations */
+uint64_t helper_float_recip_d(uint64_t fdt0)
+{
+    uint64_t fdt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fdt2 = float64_div(FLOAT_ONE64, fdt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fdt2;
+}
+
+uint32_t helper_float_recip_s(uint32_t fst0)
+{
+    uint32_t fst2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_div(FLOAT_ONE32, fst0, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fst2;
+}
+
+uint64_t helper_float_rsqrt_d(uint64_t fdt0)
+{
+    uint64_t fdt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fdt2 = float64_sqrt(fdt0, &env->active_fpu.fp_status);
+    fdt2 = float64_div(FLOAT_ONE64, fdt2, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fdt2;
+}
+
+uint32_t helper_float_rsqrt_s(uint32_t fst0)
+{
+    uint32_t fst2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_sqrt(fst0, &env->active_fpu.fp_status);
+    fst2 = float32_div(FLOAT_ONE32, fst2, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fst2;
+}
+
+uint64_t helper_float_recip1_d(uint64_t fdt0)
+{
+    uint64_t fdt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fdt2 = float64_div(FLOAT_ONE64, fdt0, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fdt2;
+}
+
+uint32_t helper_float_recip1_s(uint32_t fst0)
+{
+    uint32_t fst2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_div(FLOAT_ONE32, fst0, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fst2;
+}
+
+uint64_t helper_float_recip1_ps(uint64_t fdt0)
+{
+    uint32_t fst2;
+    uint32_t fsth2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_div(FLOAT_ONE32, fdt0 & 0XFFFFFFFF, &env->active_fpu.fp_status);
+    fsth2 = float32_div(FLOAT_ONE32, fdt0 >> 32, &env->active_fpu.fp_status);
+    update_fcr31();
+    return ((uint64_t)fsth2 << 32) | fst2;
+}
+
+uint64_t helper_float_rsqrt1_d(uint64_t fdt0)
+{
+    uint64_t fdt2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fdt2 = float64_sqrt(fdt0, &env->active_fpu.fp_status);
+    fdt2 = float64_div(FLOAT_ONE64, fdt2, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fdt2;
+}
+
+uint32_t helper_float_rsqrt1_s(uint32_t fst0)
+{
+    uint32_t fst2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_sqrt(fst0, &env->active_fpu.fp_status);
+    fst2 = float32_div(FLOAT_ONE32, fst2, &env->active_fpu.fp_status);
+    update_fcr31();
+    return fst2;
+}
+
+uint64_t helper_float_rsqrt1_ps(uint64_t fdt0)
+{
+    uint32_t fst2;
+    uint32_t fsth2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_sqrt(fdt0 & 0XFFFFFFFF, &env->active_fpu.fp_status);
+    fsth2 = float32_sqrt(fdt0 >> 32, &env->active_fpu.fp_status);
+    fst2 = float32_div(FLOAT_ONE32, fst2, &env->active_fpu.fp_status);
+    fsth2 = float32_div(FLOAT_ONE32, fsth2, &env->active_fpu.fp_status);
+    update_fcr31();
+    return ((uint64_t)fsth2 << 32) | fst2;
+}
+
+#define FLOAT_OP(name, p) void helper_float_##name##_##p(void)
+
+/* binary operations */
+#define FLOAT_BINOP(name)                                          \
+uint64_t helper_float_ ## name ## _d(uint64_t fdt0, uint64_t fdt1)     \
+{                                                                  \
+    uint64_t dt2;                                                  \
+                                                                   \
+    set_float_exception_flags(0, &env->active_fpu.fp_status);            \
+    dt2 = float64_ ## name (fdt0, fdt1, &env->active_fpu.fp_status);     \
+    update_fcr31();                                                \
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & FP_INVALID)                \
+        dt2 = FLOAT_QNAN64;                                        \
+    return dt2;                                                    \
+}                                                                  \
+                                                                   \
+uint32_t helper_float_ ## name ## _s(uint32_t fst0, uint32_t fst1)     \
+{                                                                  \
+    uint32_t wt2;                                                  \
+                                                                   \
+    set_float_exception_flags(0, &env->active_fpu.fp_status);            \
+    wt2 = float32_ ## name (fst0, fst1, &env->active_fpu.fp_status);     \
+    update_fcr31();                                                \
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & FP_INVALID)                \
+        wt2 = FLOAT_QNAN32;                                        \
+    return wt2;                                                    \
+}                                                                  \
+                                                                   \
+uint64_t helper_float_ ## name ## _ps(uint64_t fdt0, uint64_t fdt1)    \
+{                                                                  \
+    uint32_t fst0 = fdt0 & 0XFFFFFFFF;                             \
+    uint32_t fsth0 = fdt0 >> 32;                                   \
+    uint32_t fst1 = fdt1 & 0XFFFFFFFF;                             \
+    uint32_t fsth1 = fdt1 >> 32;                                   \
+    uint32_t wt2;                                                  \
+    uint32_t wth2;                                                 \
+                                                                   \
+    set_float_exception_flags(0, &env->active_fpu.fp_status);            \
+    wt2 = float32_ ## name (fst0, fst1, &env->active_fpu.fp_status);     \
+    wth2 = float32_ ## name (fsth0, fsth1, &env->active_fpu.fp_status);  \
+    update_fcr31();                                                \
+    if (GET_FP_CAUSE(env->active_fpu.fcr31) & FP_INVALID) {              \
+        wt2 = FLOAT_QNAN32;                                        \
+        wth2 = FLOAT_QNAN32;                                       \
+    }                                                              \
+    return ((uint64_t)wth2 << 32) | wt2;                           \
+}
+
+FLOAT_BINOP(add)
+FLOAT_BINOP(sub)
+FLOAT_BINOP(mul)
+FLOAT_BINOP(div)
+#undef FLOAT_BINOP
+
+/* ternary operations */
+#define FLOAT_TERNOP(name1, name2)                                        \
+uint64_t helper_float_ ## name1 ## name2 ## _d(uint64_t fdt0, uint64_t fdt1,  \
+                                           uint64_t fdt2)                 \
+{                                                                         \
+    fdt0 = float64_ ## name1 (fdt0, fdt1, &env->active_fpu.fp_status);          \
+    return float64_ ## name2 (fdt0, fdt2, &env->active_fpu.fp_status);          \
+}                                                                         \
+                                                                          \
+uint32_t helper_float_ ## name1 ## name2 ## _s(uint32_t fst0, uint32_t fst1,  \
+                                           uint32_t fst2)                 \
+{                                                                         \
+    fst0 = float32_ ## name1 (fst0, fst1, &env->active_fpu.fp_status);          \
+    return float32_ ## name2 (fst0, fst2, &env->active_fpu.fp_status);          \
+}                                                                         \
+                                                                          \
+uint64_t helper_float_ ## name1 ## name2 ## _ps(uint64_t fdt0, uint64_t fdt1, \
+                                            uint64_t fdt2)                \
+{                                                                         \
+    uint32_t fst0 = fdt0 & 0XFFFFFFFF;                                    \
+    uint32_t fsth0 = fdt0 >> 32;                                          \
+    uint32_t fst1 = fdt1 & 0XFFFFFFFF;                                    \
+    uint32_t fsth1 = fdt1 >> 32;                                          \
+    uint32_t fst2 = fdt2 & 0XFFFFFFFF;                                    \
+    uint32_t fsth2 = fdt2 >> 32;                                          \
+                                                                          \
+    fst0 = float32_ ## name1 (fst0, fst1, &env->active_fpu.fp_status);          \
+    fsth0 = float32_ ## name1 (fsth0, fsth1, &env->active_fpu.fp_status);       \
+    fst2 = float32_ ## name2 (fst0, fst2, &env->active_fpu.fp_status);          \
+    fsth2 = float32_ ## name2 (fsth0, fsth2, &env->active_fpu.fp_status);       \
+    return ((uint64_t)fsth2 << 32) | fst2;                                \
+}
+
+FLOAT_TERNOP(mul, add)
+FLOAT_TERNOP(mul, sub)
+#undef FLOAT_TERNOP
+
+/* negated ternary operations */
+#define FLOAT_NTERNOP(name1, name2)                                       \
+uint64_t helper_float_n ## name1 ## name2 ## _d(uint64_t fdt0, uint64_t fdt1, \
+                                           uint64_t fdt2)                 \
+{                                                                         \
+    fdt0 = float64_ ## name1 (fdt0, fdt1, &env->active_fpu.fp_status);          \
+    fdt2 = float64_ ## name2 (fdt0, fdt2, &env->active_fpu.fp_status);          \
+    return float64_chs(fdt2);                                             \
+}                                                                         \
+                                                                          \
+uint32_t helper_float_n ## name1 ## name2 ## _s(uint32_t fst0, uint32_t fst1, \
+                                           uint32_t fst2)                 \
+{                                                                         \
+    fst0 = float32_ ## name1 (fst0, fst1, &env->active_fpu.fp_status);          \
+    fst2 = float32_ ## name2 (fst0, fst2, &env->active_fpu.fp_status);          \
+    return float32_chs(fst2);                                             \
+}                                                                         \
+                                                                          \
+uint64_t helper_float_n ## name1 ## name2 ## _ps(uint64_t fdt0, uint64_t fdt1,\
+                                           uint64_t fdt2)                 \
+{                                                                         \
+    uint32_t fst0 = fdt0 & 0XFFFFFFFF;                                    \
+    uint32_t fsth0 = fdt0 >> 32;                                          \
+    uint32_t fst1 = fdt1 & 0XFFFFFFFF;                                    \
+    uint32_t fsth1 = fdt1 >> 32;                                          \
+    uint32_t fst2 = fdt2 & 0XFFFFFFFF;                                    \
+    uint32_t fsth2 = fdt2 >> 32;                                          \
+                                                                          \
+    fst0 = float32_ ## name1 (fst0, fst1, &env->active_fpu.fp_status);          \
+    fsth0 = float32_ ## name1 (fsth0, fsth1, &env->active_fpu.fp_status);       \
+    fst2 = float32_ ## name2 (fst0, fst2, &env->active_fpu.fp_status);          \
+    fsth2 = float32_ ## name2 (fsth0, fsth2, &env->active_fpu.fp_status);       \
+    fst2 = float32_chs(fst2);                                             \
+    fsth2 = float32_chs(fsth2);                                           \
+    return ((uint64_t)fsth2 << 32) | fst2;                                \
+}
+
+FLOAT_NTERNOP(mul, add)
+FLOAT_NTERNOP(mul, sub)
+#undef FLOAT_NTERNOP
+
+/* MIPS specific binary operations */
+uint64_t helper_float_recip2_d(uint64_t fdt0, uint64_t fdt2)
+{
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fdt2 = float64_mul(fdt0, fdt2, &env->active_fpu.fp_status);
+    fdt2 = float64_chs(float64_sub(fdt2, FLOAT_ONE64, &env->active_fpu.fp_status));
+    update_fcr31();
+    return fdt2;
+}
+
+uint32_t helper_float_recip2_s(uint32_t fst0, uint32_t fst2)
+{
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_mul(fst0, fst2, &env->active_fpu.fp_status);
+    fst2 = float32_chs(float32_sub(fst2, FLOAT_ONE32, &env->active_fpu.fp_status));
+    update_fcr31();
+    return fst2;
+}
+
+uint64_t helper_float_recip2_ps(uint64_t fdt0, uint64_t fdt2)
+{
+    uint32_t fst0 = fdt0 & 0XFFFFFFFF;
+    uint32_t fsth0 = fdt0 >> 32;
+    uint32_t fst2 = fdt2 & 0XFFFFFFFF;
+    uint32_t fsth2 = fdt2 >> 32;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_mul(fst0, fst2, &env->active_fpu.fp_status);
+    fsth2 = float32_mul(fsth0, fsth2, &env->active_fpu.fp_status);
+    fst2 = float32_chs(float32_sub(fst2, FLOAT_ONE32, &env->active_fpu.fp_status));
+    fsth2 = float32_chs(float32_sub(fsth2, FLOAT_ONE32, &env->active_fpu.fp_status));
+    update_fcr31();
+    return ((uint64_t)fsth2 << 32) | fst2;
+}
+
+uint64_t helper_float_rsqrt2_d(uint64_t fdt0, uint64_t fdt2)
+{
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fdt2 = float64_mul(fdt0, fdt2, &env->active_fpu.fp_status);
+    fdt2 = float64_sub(fdt2, FLOAT_ONE64, &env->active_fpu.fp_status);
+    fdt2 = float64_chs(float64_div(fdt2, FLOAT_TWO64, &env->active_fpu.fp_status));
+    update_fcr31();
+    return fdt2;
+}
+
+uint32_t helper_float_rsqrt2_s(uint32_t fst0, uint32_t fst2)
+{
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_mul(fst0, fst2, &env->active_fpu.fp_status);
+    fst2 = float32_sub(fst2, FLOAT_ONE32, &env->active_fpu.fp_status);
+    fst2 = float32_chs(float32_div(fst2, FLOAT_TWO32, &env->active_fpu.fp_status));
+    update_fcr31();
+    return fst2;
+}
+
+uint64_t helper_float_rsqrt2_ps(uint64_t fdt0, uint64_t fdt2)
+{
+    uint32_t fst0 = fdt0 & 0XFFFFFFFF;
+    uint32_t fsth0 = fdt0 >> 32;
+    uint32_t fst2 = fdt2 & 0XFFFFFFFF;
+    uint32_t fsth2 = fdt2 >> 32;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_mul(fst0, fst2, &env->active_fpu.fp_status);
+    fsth2 = float32_mul(fsth0, fsth2, &env->active_fpu.fp_status);
+    fst2 = float32_sub(fst2, FLOAT_ONE32, &env->active_fpu.fp_status);
+    fsth2 = float32_sub(fsth2, FLOAT_ONE32, &env->active_fpu.fp_status);
+    fst2 = float32_chs(float32_div(fst2, FLOAT_TWO32, &env->active_fpu.fp_status));
+    fsth2 = float32_chs(float32_div(fsth2, FLOAT_TWO32, &env->active_fpu.fp_status));
+    update_fcr31();
+    return ((uint64_t)fsth2 << 32) | fst2;
+}
+
+uint64_t helper_float_addr_ps(uint64_t fdt0, uint64_t fdt1)
+{
+    uint32_t fst0 = fdt0 & 0XFFFFFFFF;
+    uint32_t fsth0 = fdt0 >> 32;
+    uint32_t fst1 = fdt1 & 0XFFFFFFFF;
+    uint32_t fsth1 = fdt1 >> 32;
+    uint32_t fst2;
+    uint32_t fsth2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_add (fst0, fsth0, &env->active_fpu.fp_status);
+    fsth2 = float32_add (fst1, fsth1, &env->active_fpu.fp_status);
+    update_fcr31();
+    return ((uint64_t)fsth2 << 32) | fst2;
+}
+
+uint64_t helper_float_mulr_ps(uint64_t fdt0, uint64_t fdt1)
+{
+    uint32_t fst0 = fdt0 & 0XFFFFFFFF;
+    uint32_t fsth0 = fdt0 >> 32;
+    uint32_t fst1 = fdt1 & 0XFFFFFFFF;
+    uint32_t fsth1 = fdt1 >> 32;
+    uint32_t fst2;
+    uint32_t fsth2;
+
+    set_float_exception_flags(0, &env->active_fpu.fp_status);
+    fst2 = float32_mul (fst0, fsth0, &env->active_fpu.fp_status);
+    fsth2 = float32_mul (fst1, fsth1, &env->active_fpu.fp_status);
+    update_fcr31();
+    return ((uint64_t)fsth2 << 32) | fst2;
+}
+
+/* compare operations */
+#define FOP_COND_D(op, cond)                                   \
+void helper_cmp_d_ ## op (uint64_t fdt0, uint64_t fdt1, int cc)    \
+{                                                              \
+    int c = cond;                                              \
+    update_fcr31();                                            \
+    if (c)                                                     \
+        SET_FP_COND(cc, env->active_fpu);                      \
+    else                                                       \
+        CLEAR_FP_COND(cc, env->active_fpu);                    \
+}                                                              \
+void helper_cmpabs_d_ ## op (uint64_t fdt0, uint64_t fdt1, int cc) \
+{                                                              \
+    int c;                                                     \
+    fdt0 = float64_abs(fdt0);                                  \
+    fdt1 = float64_abs(fdt1);                                  \
+    c = cond;                                                  \
+    update_fcr31();                                            \
+    if (c)                                                     \
+        SET_FP_COND(cc, env->active_fpu);                      \
+    else                                                       \
+        CLEAR_FP_COND(cc, env->active_fpu);                    \
+}
+
+static int float64_is_unordered(int sig, float64 a, float64 b STATUS_PARAM)
+{
+    if (float64_is_signaling_nan(a) ||
+        float64_is_signaling_nan(b) ||
+        (sig && (float64_is_nan(a) || float64_is_nan(b)))) {
+        float_raise(float_flag_invalid, status);
+        return 1;
+    } else if (float64_is_nan(a) || float64_is_nan(b)) {
+        return 1;
+    } else {
+        return 0;
+    }
+}
+
+/* NOTE: the comma operator will make "cond" to eval to false,
+ * but float*_is_unordered() is still called. */
+FOP_COND_D(f,   (float64_is_unordered(0, fdt1, fdt0, &env->active_fpu.fp_status), 0))
+FOP_COND_D(un,  float64_is_unordered(0, fdt1, fdt0, &env->active_fpu.fp_status))
+FOP_COND_D(eq,  !float64_is_unordered(0, fdt1, fdt0, &env->active_fpu.fp_status) && float64_eq(fdt0, fdt1, &env->active_fpu.fp_status))
+FOP_COND_D(ueq, float64_is_unordered(0, fdt1, fdt0, &env->active_fpu.fp_status)  || float64_eq(fdt0, fdt1, &env->active_fpu.fp_status))
+FOP_COND_D(olt, !float64_is_unordered(0, fdt1, fdt0, &env->active_fpu.fp_status) && float64_lt(fdt0, fdt1, &env->active_fpu.fp_status))
+FOP_COND_D(ult, float64_is_unordered(0, fdt1, fdt0, &env->active_fpu.fp_status)  || float64_lt(fdt0, fdt1, &env->active_fpu.fp_status))
+FOP_COND_D(ole, !float64_is_unordered(0, fdt1, fdt0, &env->active_fpu.fp_status) && float64_le(fdt0, fdt1, &env->active_fpu.fp_status))
+FOP_COND_D(ule, float64_is_unordered(0, fdt1, fdt0, &env->active_fpu.fp_status)  || float64_le(fdt0, fdt1, &env->active_fpu.fp_status))
+/* NOTE: the comma operator will make "cond" to eval to false,
+ * but float*_is_unordered() is still called. */
+FOP_COND_D(sf,  (float64_is_unordered(1, fdt1, fdt0, &env->active_fpu.fp_status), 0))
+FOP_COND_D(ngle,float64_is_unordered(1, fdt1, fdt0, &env->active_fpu.fp_status))
+FOP_COND_D(seq, !float64_is_unordered(1, fdt1, fdt0, &env->active_fpu.fp_status) && float64_eq(fdt0, fdt1, &env->active_fpu.fp_status))
+FOP_COND_D(ngl, float64_is_unordered(1, fdt1, fdt0, &env->active_fpu.fp_status)  || float64_eq(fdt0, fdt1, &env->active_fpu.fp_status))
+FOP_COND_D(lt,  !float64_is_unordered(1, fdt1, fdt0, &env->active_fpu.fp_status) && float64_lt(fdt0, fdt1, &env->active_fpu.fp_status))
+FOP_COND_D(nge, float64_is_unordered(1, fdt1, fdt0, &env->active_fpu.fp_status)  || float64_lt(fdt0, fdt1, &env->active_fpu.fp_status))
+FOP_COND_D(le,  !float64_is_unordered(1, fdt1, fdt0, &env->active_fpu.fp_status) && float64_le(fdt0, fdt1, &env->active_fpu.fp_status))
+FOP_COND_D(ngt, float64_is_unordered(1, fdt1, fdt0, &env->active_fpu.fp_status)  || float64_le(fdt0, fdt1, &env->active_fpu.fp_status))
+
+#define FOP_COND_S(op, cond)                                   \
+void helper_cmp_s_ ## op (uint32_t fst0, uint32_t fst1, int cc)    \
+{                                                              \
+    int c = cond;                                              \
+    update_fcr31();                                            \
+    if (c)                                                     \
+        SET_FP_COND(cc, env->active_fpu);                      \
+    else                                                       \
+        CLEAR_FP_COND(cc, env->active_fpu);                    \
+}                                                              \
+void helper_cmpabs_s_ ## op (uint32_t fst0, uint32_t fst1, int cc) \
+{                                                              \
+    int c;                                                     \
+    fst0 = float32_abs(fst0);                                  \
+    fst1 = float32_abs(fst1);                                  \
+    c = cond;                                                  \
+    update_fcr31();                                            \
+    if (c)                                                     \
+        SET_FP_COND(cc, env->active_fpu);                      \
+    else                                                       \
+        CLEAR_FP_COND(cc, env->active_fpu);                    \
+}
+
+static flag float32_is_unordered(int sig, float32 a, float32 b STATUS_PARAM)
+{
+    if (float32_is_signaling_nan(a) ||
+        float32_is_signaling_nan(b) ||
+        (sig && (float32_is_nan(a) || float32_is_nan(b)))) {
+        float_raise(float_flag_invalid, status);
+        return 1;
+    } else if (float32_is_nan(a) || float32_is_nan(b)) {
+        return 1;
+    } else {
+        return 0;
+    }
+}
+
+/* NOTE: the comma operator will make "cond" to eval to false,
+ * but float*_is_unordered() is still called. */
+FOP_COND_S(f,   (float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status), 0))
+FOP_COND_S(un,  float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status))
+FOP_COND_S(eq,  !float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status) && float32_eq(fst0, fst1, &env->active_fpu.fp_status))
+FOP_COND_S(ueq, float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status)  || float32_eq(fst0, fst1, &env->active_fpu.fp_status))
+FOP_COND_S(olt, !float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status) && float32_lt(fst0, fst1, &env->active_fpu.fp_status))
+FOP_COND_S(ult, float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status)  || float32_lt(fst0, fst1, &env->active_fpu.fp_status))
+FOP_COND_S(ole, !float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status) && float32_le(fst0, fst1, &env->active_fpu.fp_status))
+FOP_COND_S(ule, float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status)  || float32_le(fst0, fst1, &env->active_fpu.fp_status))
+/* NOTE: the comma operator will make "cond" to eval to false,
+ * but float*_is_unordered() is still called. */
+FOP_COND_S(sf,  (float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status), 0))
+FOP_COND_S(ngle,float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status))
+FOP_COND_S(seq, !float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status) && float32_eq(fst0, fst1, &env->active_fpu.fp_status))
+FOP_COND_S(ngl, float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status)  || float32_eq(fst0, fst1, &env->active_fpu.fp_status))
+FOP_COND_S(lt,  !float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status) && float32_lt(fst0, fst1, &env->active_fpu.fp_status))
+FOP_COND_S(nge, float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status)  || float32_lt(fst0, fst1, &env->active_fpu.fp_status))
+FOP_COND_S(le,  !float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status) && float32_le(fst0, fst1, &env->active_fpu.fp_status))
+FOP_COND_S(ngt, float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status)  || float32_le(fst0, fst1, &env->active_fpu.fp_status))
+
+#define FOP_COND_PS(op, condl, condh)                           \
+void helper_cmp_ps_ ## op (uint64_t fdt0, uint64_t fdt1, int cc)    \
+{                                                               \
+    uint32_t fst0 = float32_abs(fdt0 & 0XFFFFFFFF);             \
+    uint32_t fsth0 = float32_abs(fdt0 >> 32);                   \
+    uint32_t fst1 = float32_abs(fdt1 & 0XFFFFFFFF);             \
+    uint32_t fsth1 = float32_abs(fdt1 >> 32);                   \
+    int cl = condl;                                             \
+    int ch = condh;                                             \
+                                                                \
+    update_fcr31();                                             \
+    if (cl)                                                     \
+        SET_FP_COND(cc, env->active_fpu);                       \
+    else                                                        \
+        CLEAR_FP_COND(cc, env->active_fpu);                     \
+    if (ch)                                                     \
+        SET_FP_COND(cc + 1, env->active_fpu);                   \
+    else                                                        \
+        CLEAR_FP_COND(cc + 1, env->active_fpu);                 \
+}                                                               \
+void helper_cmpabs_ps_ ## op (uint64_t fdt0, uint64_t fdt1, int cc) \
+{                                                               \
+    uint32_t fst0 = float32_abs(fdt0 & 0XFFFFFFFF);             \
+    uint32_t fsth0 = float32_abs(fdt0 >> 32);                   \
+    uint32_t fst1 = float32_abs(fdt1 & 0XFFFFFFFF);             \
+    uint32_t fsth1 = float32_abs(fdt1 >> 32);                   \
+    int cl = condl;                                             \
+    int ch = condh;                                             \
+                                                                \
+    update_fcr31();                                             \
+    if (cl)                                                     \
+        SET_FP_COND(cc, env->active_fpu);                       \
+    else                                                        \
+        CLEAR_FP_COND(cc, env->active_fpu);                     \
+    if (ch)                                                     \
+        SET_FP_COND(cc + 1, env->active_fpu);                   \
+    else                                                        \
+        CLEAR_FP_COND(cc + 1, env->active_fpu);                 \
+}
+
+/* NOTE: the comma operator will make "cond" to eval to false,
+ * but float*_is_unordered() is still called. */
+FOP_COND_PS(f,   (float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status), 0),
+                 (float32_is_unordered(0, fsth1, fsth0, &env->active_fpu.fp_status), 0))
+FOP_COND_PS(un,  float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status),
+                 float32_is_unordered(0, fsth1, fsth0, &env->active_fpu.fp_status))
+FOP_COND_PS(eq,  !float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status)   && float32_eq(fst0, fst1, &env->active_fpu.fp_status),
+                 !float32_is_unordered(0, fsth1, fsth0, &env->active_fpu.fp_status) && float32_eq(fsth0, fsth1, &env->active_fpu.fp_status))
+FOP_COND_PS(ueq, float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status)    || float32_eq(fst0, fst1, &env->active_fpu.fp_status),
+                 float32_is_unordered(0, fsth1, fsth0, &env->active_fpu.fp_status)  || float32_eq(fsth0, fsth1, &env->active_fpu.fp_status))
+FOP_COND_PS(olt, !float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status)   && float32_lt(fst0, fst1, &env->active_fpu.fp_status),
+                 !float32_is_unordered(0, fsth1, fsth0, &env->active_fpu.fp_status) && float32_lt(fsth0, fsth1, &env->active_fpu.fp_status))
+FOP_COND_PS(ult, float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status)    || float32_lt(fst0, fst1, &env->active_fpu.fp_status),
+                 float32_is_unordered(0, fsth1, fsth0, &env->active_fpu.fp_status)  || float32_lt(fsth0, fsth1, &env->active_fpu.fp_status))
+FOP_COND_PS(ole, !float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status)   && float32_le(fst0, fst1, &env->active_fpu.fp_status),
+                 !float32_is_unordered(0, fsth1, fsth0, &env->active_fpu.fp_status) && float32_le(fsth0, fsth1, &env->active_fpu.fp_status))
+FOP_COND_PS(ule, float32_is_unordered(0, fst1, fst0, &env->active_fpu.fp_status)    || float32_le(fst0, fst1, &env->active_fpu.fp_status),
+                 float32_is_unordered(0, fsth1, fsth0, &env->active_fpu.fp_status)  || float32_le(fsth0, fsth1, &env->active_fpu.fp_status))
+/* NOTE: the comma operator will make "cond" to eval to false,
+ * but float*_is_unordered() is still called. */
+FOP_COND_PS(sf,  (float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status), 0),
+                 (float32_is_unordered(1, fsth1, fsth0, &env->active_fpu.fp_status), 0))
+FOP_COND_PS(ngle,float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status),
+                 float32_is_unordered(1, fsth1, fsth0, &env->active_fpu.fp_status))
+FOP_COND_PS(seq, !float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status)   && float32_eq(fst0, fst1, &env->active_fpu.fp_status),
+                 !float32_is_unordered(1, fsth1, fsth0, &env->active_fpu.fp_status) && float32_eq(fsth0, fsth1, &env->active_fpu.fp_status))
+FOP_COND_PS(ngl, float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status)    || float32_eq(fst0, fst1, &env->active_fpu.fp_status),
+                 float32_is_unordered(1, fsth1, fsth0, &env->active_fpu.fp_status)  || float32_eq(fsth0, fsth1, &env->active_fpu.fp_status))
+FOP_COND_PS(lt,  !float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status)   && float32_lt(fst0, fst1, &env->active_fpu.fp_status),
+                 !float32_is_unordered(1, fsth1, fsth0, &env->active_fpu.fp_status) && float32_lt(fsth0, fsth1, &env->active_fpu.fp_status))
+FOP_COND_PS(nge, float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status)    || float32_lt(fst0, fst1, &env->active_fpu.fp_status),
+                 float32_is_unordered(1, fsth1, fsth0, &env->active_fpu.fp_status)  || float32_lt(fsth0, fsth1, &env->active_fpu.fp_status))
+FOP_COND_PS(le,  !float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status)   && float32_le(fst0, fst1, &env->active_fpu.fp_status),
+                 !float32_is_unordered(1, fsth1, fsth0, &env->active_fpu.fp_status) && float32_le(fsth0, fsth1, &env->active_fpu.fp_status))
+FOP_COND_PS(ngt, float32_is_unordered(1, fst1, fst0, &env->active_fpu.fp_status)    || float32_le(fst0, fst1, &env->active_fpu.fp_status),
+                 float32_is_unordered(1, fsth1, fsth0, &env->active_fpu.fp_status)  || float32_le(fsth0, fsth1, &env->active_fpu.fp_status))
diff --git a/target-mips/translate.c b/target-mips/translate.c
new file mode 100644
index 0000000..d1d6f23
--- /dev/null
+++ b/target-mips/translate.c
@@ -0,0 +1,8714 @@
+/*
+ *  MIPS32 emulation for qemu: main translation routines.
+ *
+ *  Copyright (c) 2004-2005 Jocelyn Mayer
+ *  Copyright (c) 2006 Marius Groeger (FPU operations)
+ *  Copyright (c) 2006 Thiemo Seufer (MIPS32R2 support)
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <stdarg.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <inttypes.h>
+
+#include "cpu.h"
+#include "exec-all.h"
+#include "disas.h"
+#include "tcg-op.h"
+#include "qemu-common.h"
+
+#include "helper.h"
+#define GEN_HELPER 1
+#include "helper.h"
+
+//#define MIPS_DEBUG_DISAS
+//#define MIPS_DEBUG_SIGN_EXTENSIONS
+
+/* MIPS major opcodes */
+#define MASK_OP_MAJOR(op)  (op & (0x3F << 26))
+
+enum {
+    /* indirect opcode tables */
+    OPC_SPECIAL  = (0x00 << 26),
+    OPC_REGIMM   = (0x01 << 26),
+    OPC_CP0      = (0x10 << 26),
+    OPC_CP1      = (0x11 << 26),
+    OPC_CP2      = (0x12 << 26),
+    OPC_CP3      = (0x13 << 26),
+    OPC_SPECIAL2 = (0x1C << 26),
+    OPC_SPECIAL3 = (0x1F << 26),
+    /* arithmetic with immediate */
+    OPC_ADDI     = (0x08 << 26),
+    OPC_ADDIU    = (0x09 << 26),
+    OPC_SLTI     = (0x0A << 26),
+    OPC_SLTIU    = (0x0B << 26),
+    /* logic with immediate */
+    OPC_ANDI     = (0x0C << 26),
+    OPC_ORI      = (0x0D << 26),
+    OPC_XORI     = (0x0E << 26),
+    OPC_LUI      = (0x0F << 26),
+    /* arithmetic with immediate */
+    OPC_DADDI    = (0x18 << 26),
+    OPC_DADDIU   = (0x19 << 26),
+    /* Jump and branches */
+    OPC_J        = (0x02 << 26),
+    OPC_JAL      = (0x03 << 26),
+    OPC_BEQ      = (0x04 << 26),  /* Unconditional if rs = rt = 0 (B) */
+    OPC_BEQL     = (0x14 << 26),
+    OPC_BNE      = (0x05 << 26),
+    OPC_BNEL     = (0x15 << 26),
+    OPC_BLEZ     = (0x06 << 26),
+    OPC_BLEZL    = (0x16 << 26),
+    OPC_BGTZ     = (0x07 << 26),
+    OPC_BGTZL    = (0x17 << 26),
+    OPC_JALX     = (0x1D << 26),  /* MIPS 16 only */
+    /* Load and stores */
+    OPC_LDL      = (0x1A << 26),
+    OPC_LDR      = (0x1B << 26),
+    OPC_LB       = (0x20 << 26),
+    OPC_LH       = (0x21 << 26),
+    OPC_LWL      = (0x22 << 26),
+    OPC_LW       = (0x23 << 26),
+    OPC_LBU      = (0x24 << 26),
+    OPC_LHU      = (0x25 << 26),
+    OPC_LWR      = (0x26 << 26),
+    OPC_LWU      = (0x27 << 26),
+    OPC_SB       = (0x28 << 26),
+    OPC_SH       = (0x29 << 26),
+    OPC_SWL      = (0x2A << 26),
+    OPC_SW       = (0x2B << 26),
+    OPC_SDL      = (0x2C << 26),
+    OPC_SDR      = (0x2D << 26),
+    OPC_SWR      = (0x2E << 26),
+    OPC_LL       = (0x30 << 26),
+    OPC_LLD      = (0x34 << 26),
+    OPC_LD       = (0x37 << 26),
+    OPC_SC       = (0x38 << 26),
+    OPC_SCD      = (0x3C << 26),
+    OPC_SD       = (0x3F << 26),
+    /* Floating point load/store */
+    OPC_LWC1     = (0x31 << 26),
+    OPC_LWC2     = (0x32 << 26),
+    OPC_LDC1     = (0x35 << 26),
+    OPC_LDC2     = (0x36 << 26),
+    OPC_SWC1     = (0x39 << 26),
+    OPC_SWC2     = (0x3A << 26),
+    OPC_SDC1     = (0x3D << 26),
+    OPC_SDC2     = (0x3E << 26),
+    /* MDMX ASE specific */
+    OPC_MDMX     = (0x1E << 26),
+    /* Cache and prefetch */
+    OPC_CACHE    = (0x2F << 26),
+    OPC_PREF     = (0x33 << 26),
+    /* Reserved major opcode */
+    OPC_MAJOR3B_RESERVED = (0x3B << 26),
+};
+
+/* MIPS special opcodes */
+#define MASK_SPECIAL(op)   MASK_OP_MAJOR(op) | (op & 0x3F)
+
+enum {
+    /* Shifts */
+    OPC_SLL      = 0x00 | OPC_SPECIAL,
+    /* NOP is SLL r0, r0, 0   */
+    /* SSNOP is SLL r0, r0, 1 */
+    /* EHB is SLL r0, r0, 3 */
+    OPC_SRL      = 0x02 | OPC_SPECIAL, /* also ROTR */
+    OPC_SRA      = 0x03 | OPC_SPECIAL,
+    OPC_SLLV     = 0x04 | OPC_SPECIAL,
+    OPC_SRLV     = 0x06 | OPC_SPECIAL, /* also ROTRV */
+    OPC_SRAV     = 0x07 | OPC_SPECIAL,
+    OPC_DSLLV    = 0x14 | OPC_SPECIAL,
+    OPC_DSRLV    = 0x16 | OPC_SPECIAL, /* also DROTRV */
+    OPC_DSRAV    = 0x17 | OPC_SPECIAL,
+    OPC_DSLL     = 0x38 | OPC_SPECIAL,
+    OPC_DSRL     = 0x3A | OPC_SPECIAL, /* also DROTR */
+    OPC_DSRA     = 0x3B | OPC_SPECIAL,
+    OPC_DSLL32   = 0x3C | OPC_SPECIAL,
+    OPC_DSRL32   = 0x3E | OPC_SPECIAL, /* also DROTR32 */
+    OPC_DSRA32   = 0x3F | OPC_SPECIAL,
+    /* Multiplication / division */
+    OPC_MULT     = 0x18 | OPC_SPECIAL,
+    OPC_MULTU    = 0x19 | OPC_SPECIAL,
+    OPC_DIV      = 0x1A | OPC_SPECIAL,
+    OPC_DIVU     = 0x1B | OPC_SPECIAL,
+    OPC_DMULT    = 0x1C | OPC_SPECIAL,
+    OPC_DMULTU   = 0x1D | OPC_SPECIAL,
+    OPC_DDIV     = 0x1E | OPC_SPECIAL,
+    OPC_DDIVU    = 0x1F | OPC_SPECIAL,
+    /* 2 registers arithmetic / logic */
+    OPC_ADD      = 0x20 | OPC_SPECIAL,
+    OPC_ADDU     = 0x21 | OPC_SPECIAL,
+    OPC_SUB      = 0x22 | OPC_SPECIAL,
+    OPC_SUBU     = 0x23 | OPC_SPECIAL,
+    OPC_AND      = 0x24 | OPC_SPECIAL,
+    OPC_OR       = 0x25 | OPC_SPECIAL,
+    OPC_XOR      = 0x26 | OPC_SPECIAL,
+    OPC_NOR      = 0x27 | OPC_SPECIAL,
+    OPC_SLT      = 0x2A | OPC_SPECIAL,
+    OPC_SLTU     = 0x2B | OPC_SPECIAL,
+    OPC_DADD     = 0x2C | OPC_SPECIAL,
+    OPC_DADDU    = 0x2D | OPC_SPECIAL,
+    OPC_DSUB     = 0x2E | OPC_SPECIAL,
+    OPC_DSUBU    = 0x2F | OPC_SPECIAL,
+    /* Jumps */
+    OPC_JR       = 0x08 | OPC_SPECIAL, /* Also JR.HB */
+    OPC_JALR     = 0x09 | OPC_SPECIAL, /* Also JALR.HB */
+    /* Traps */
+    OPC_TGE      = 0x30 | OPC_SPECIAL,
+    OPC_TGEU     = 0x31 | OPC_SPECIAL,
+    OPC_TLT      = 0x32 | OPC_SPECIAL,
+    OPC_TLTU     = 0x33 | OPC_SPECIAL,
+    OPC_TEQ      = 0x34 | OPC_SPECIAL,
+    OPC_TNE      = 0x36 | OPC_SPECIAL,
+    /* HI / LO registers load & stores */
+    OPC_MFHI     = 0x10 | OPC_SPECIAL,
+    OPC_MTHI     = 0x11 | OPC_SPECIAL,
+    OPC_MFLO     = 0x12 | OPC_SPECIAL,
+    OPC_MTLO     = 0x13 | OPC_SPECIAL,
+    /* Conditional moves */
+    OPC_MOVZ     = 0x0A | OPC_SPECIAL,
+    OPC_MOVN     = 0x0B | OPC_SPECIAL,
+
+    OPC_MOVCI    = 0x01 | OPC_SPECIAL,
+
+    /* Special */
+    OPC_PMON     = 0x05 | OPC_SPECIAL, /* unofficial */
+    OPC_SYSCALL  = 0x0C | OPC_SPECIAL,
+    OPC_BREAK    = 0x0D | OPC_SPECIAL,
+    OPC_SPIM     = 0x0E | OPC_SPECIAL, /* unofficial */
+    OPC_SYNC     = 0x0F | OPC_SPECIAL,
+
+    OPC_SPECIAL15_RESERVED = 0x15 | OPC_SPECIAL,
+    OPC_SPECIAL28_RESERVED = 0x28 | OPC_SPECIAL,
+    OPC_SPECIAL29_RESERVED = 0x29 | OPC_SPECIAL,
+    OPC_SPECIAL35_RESERVED = 0x35 | OPC_SPECIAL,
+    OPC_SPECIAL37_RESERVED = 0x37 | OPC_SPECIAL,
+    OPC_SPECIAL39_RESERVED = 0x39 | OPC_SPECIAL,
+    OPC_SPECIAL3D_RESERVED = 0x3D | OPC_SPECIAL,
+};
+
+/* Multiplication variants of the vr54xx. */
+#define MASK_MUL_VR54XX(op)   MASK_SPECIAL(op) | (op & (0x1F << 6))
+
+enum {
+    OPC_VR54XX_MULS    = (0x03 << 6) | OPC_MULT,
+    OPC_VR54XX_MULSU   = (0x03 << 6) | OPC_MULTU,
+    OPC_VR54XX_MACC    = (0x05 << 6) | OPC_MULT,
+    OPC_VR54XX_MACCU   = (0x05 << 6) | OPC_MULTU,
+    OPC_VR54XX_MSAC    = (0x07 << 6) | OPC_MULT,
+    OPC_VR54XX_MSACU   = (0x07 << 6) | OPC_MULTU,
+    OPC_VR54XX_MULHI   = (0x09 << 6) | OPC_MULT,
+    OPC_VR54XX_MULHIU  = (0x09 << 6) | OPC_MULTU,
+    OPC_VR54XX_MULSHI  = (0x0B << 6) | OPC_MULT,
+    OPC_VR54XX_MULSHIU = (0x0B << 6) | OPC_MULTU,
+    OPC_VR54XX_MACCHI  = (0x0D << 6) | OPC_MULT,
+    OPC_VR54XX_MACCHIU = (0x0D << 6) | OPC_MULTU,
+    OPC_VR54XX_MSACHI  = (0x0F << 6) | OPC_MULT,
+    OPC_VR54XX_MSACHIU = (0x0F << 6) | OPC_MULTU,
+};
+
+/* REGIMM (rt field) opcodes */
+#define MASK_REGIMM(op)    MASK_OP_MAJOR(op) | (op & (0x1F << 16))
+
+enum {
+    OPC_BLTZ     = (0x00 << 16) | OPC_REGIMM,
+    OPC_BLTZL    = (0x02 << 16) | OPC_REGIMM,
+    OPC_BGEZ     = (0x01 << 16) | OPC_REGIMM,
+    OPC_BGEZL    = (0x03 << 16) | OPC_REGIMM,
+    OPC_BLTZAL   = (0x10 << 16) | OPC_REGIMM,
+    OPC_BLTZALL  = (0x12 << 16) | OPC_REGIMM,
+    OPC_BGEZAL   = (0x11 << 16) | OPC_REGIMM,
+    OPC_BGEZALL  = (0x13 << 16) | OPC_REGIMM,
+    OPC_TGEI     = (0x08 << 16) | OPC_REGIMM,
+    OPC_TGEIU    = (0x09 << 16) | OPC_REGIMM,
+    OPC_TLTI     = (0x0A << 16) | OPC_REGIMM,
+    OPC_TLTIU    = (0x0B << 16) | OPC_REGIMM,
+    OPC_TEQI     = (0x0C << 16) | OPC_REGIMM,
+    OPC_TNEI     = (0x0E << 16) | OPC_REGIMM,
+    OPC_SYNCI    = (0x1F << 16) | OPC_REGIMM,
+};
+
+/* Special2 opcodes */
+#define MASK_SPECIAL2(op)  MASK_OP_MAJOR(op) | (op & 0x3F)
+
+enum {
+    /* Multiply & xxx operations */
+    OPC_MADD     = 0x00 | OPC_SPECIAL2,
+    OPC_MADDU    = 0x01 | OPC_SPECIAL2,
+    OPC_MUL      = 0x02 | OPC_SPECIAL2,
+    OPC_MSUB     = 0x04 | OPC_SPECIAL2,
+    OPC_MSUBU    = 0x05 | OPC_SPECIAL2,
+    /* Misc */
+    OPC_CLZ      = 0x20 | OPC_SPECIAL2,
+    OPC_CLO      = 0x21 | OPC_SPECIAL2,
+    OPC_DCLZ     = 0x24 | OPC_SPECIAL2,
+    OPC_DCLO     = 0x25 | OPC_SPECIAL2,
+    /* Special */
+    OPC_SDBBP    = 0x3F | OPC_SPECIAL2,
+};
+
+/* Special3 opcodes */
+#define MASK_SPECIAL3(op)  MASK_OP_MAJOR(op) | (op & 0x3F)
+
+enum {
+    OPC_EXT      = 0x00 | OPC_SPECIAL3,
+    OPC_DEXTM    = 0x01 | OPC_SPECIAL3,
+    OPC_DEXTU    = 0x02 | OPC_SPECIAL3,
+    OPC_DEXT     = 0x03 | OPC_SPECIAL3,
+    OPC_INS      = 0x04 | OPC_SPECIAL3,
+    OPC_DINSM    = 0x05 | OPC_SPECIAL3,
+    OPC_DINSU    = 0x06 | OPC_SPECIAL3,
+    OPC_DINS     = 0x07 | OPC_SPECIAL3,
+    OPC_FORK     = 0x08 | OPC_SPECIAL3,
+    OPC_YIELD    = 0x09 | OPC_SPECIAL3,
+    OPC_BSHFL    = 0x20 | OPC_SPECIAL3,
+    OPC_DBSHFL   = 0x24 | OPC_SPECIAL3,
+    OPC_RDHWR    = 0x3B | OPC_SPECIAL3,
+};
+
+/* BSHFL opcodes */
+#define MASK_BSHFL(op)     MASK_SPECIAL3(op) | (op & (0x1F << 6))
+
+enum {
+    OPC_WSBH     = (0x02 << 6) | OPC_BSHFL,
+    OPC_SEB      = (0x10 << 6) | OPC_BSHFL,
+    OPC_SEH      = (0x18 << 6) | OPC_BSHFL,
+};
+
+/* DBSHFL opcodes */
+#define MASK_DBSHFL(op)    MASK_SPECIAL3(op) | (op & (0x1F << 6))
+
+enum {
+    OPC_DSBH     = (0x02 << 6) | OPC_DBSHFL,
+    OPC_DSHD     = (0x05 << 6) | OPC_DBSHFL,
+};
+
+/* Coprocessor 0 (rs field) */
+#define MASK_CP0(op)       MASK_OP_MAJOR(op) | (op & (0x1F << 21))
+
+enum {
+    OPC_MFC0     = (0x00 << 21) | OPC_CP0,
+    OPC_DMFC0    = (0x01 << 21) | OPC_CP0,
+    OPC_MTC0     = (0x04 << 21) | OPC_CP0,
+    OPC_DMTC0    = (0x05 << 21) | OPC_CP0,
+    OPC_MFTR     = (0x08 << 21) | OPC_CP0,
+    OPC_RDPGPR   = (0x0A << 21) | OPC_CP0,
+    OPC_MFMC0    = (0x0B << 21) | OPC_CP0,
+    OPC_MTTR     = (0x0C << 21) | OPC_CP0,
+    OPC_WRPGPR   = (0x0E << 21) | OPC_CP0,
+    OPC_C0       = (0x10 << 21) | OPC_CP0,
+    OPC_C0_FIRST = (0x10 << 21) | OPC_CP0,
+    OPC_C0_LAST  = (0x1F << 21) | OPC_CP0,
+};
+
+/* MFMC0 opcodes */
+#define MASK_MFMC0(op)     MASK_CP0(op) | (op & 0xFFFF)
+
+enum {
+    OPC_DMT      = 0x01 | (0 << 5) | (0x0F << 6) | (0x01 << 11) | OPC_MFMC0,
+    OPC_EMT      = 0x01 | (1 << 5) | (0x0F << 6) | (0x01 << 11) | OPC_MFMC0,
+    OPC_DVPE     = 0x01 | (0 << 5) | OPC_MFMC0,
+    OPC_EVPE     = 0x01 | (1 << 5) | OPC_MFMC0,
+    OPC_DI       = (0 << 5) | (0x0C << 11) | OPC_MFMC0,
+    OPC_EI       = (1 << 5) | (0x0C << 11) | OPC_MFMC0,
+};
+
+/* Coprocessor 0 (with rs == C0) */
+#define MASK_C0(op)        MASK_CP0(op) | (op & 0x3F)
+
+enum {
+    OPC_TLBR     = 0x01 | OPC_C0,
+    OPC_TLBWI    = 0x02 | OPC_C0,
+    OPC_TLBWR    = 0x06 | OPC_C0,
+    OPC_TLBP     = 0x08 | OPC_C0,
+    OPC_RFE      = 0x10 | OPC_C0,
+    OPC_ERET     = 0x18 | OPC_C0,
+    OPC_DERET    = 0x1F | OPC_C0,
+    OPC_WAIT     = 0x20 | OPC_C0,
+};
+
+/* Coprocessor 1 (rs field) */
+#define MASK_CP1(op)       MASK_OP_MAJOR(op) | (op & (0x1F << 21))
+
+enum {
+    OPC_MFC1     = (0x00 << 21) | OPC_CP1,
+    OPC_DMFC1    = (0x01 << 21) | OPC_CP1,
+    OPC_CFC1     = (0x02 << 21) | OPC_CP1,
+    OPC_MFHC1    = (0x03 << 21) | OPC_CP1,
+    OPC_MTC1     = (0x04 << 21) | OPC_CP1,
+    OPC_DMTC1    = (0x05 << 21) | OPC_CP1,
+    OPC_CTC1     = (0x06 << 21) | OPC_CP1,
+    OPC_MTHC1    = (0x07 << 21) | OPC_CP1,
+    OPC_BC1      = (0x08 << 21) | OPC_CP1, /* bc */
+    OPC_BC1ANY2  = (0x09 << 21) | OPC_CP1,
+    OPC_BC1ANY4  = (0x0A << 21) | OPC_CP1,
+    OPC_S_FMT    = (0x10 << 21) | OPC_CP1, /* 16: fmt=single fp */
+    OPC_D_FMT    = (0x11 << 21) | OPC_CP1, /* 17: fmt=double fp */
+    OPC_E_FMT    = (0x12 << 21) | OPC_CP1, /* 18: fmt=extended fp */
+    OPC_Q_FMT    = (0x13 << 21) | OPC_CP1, /* 19: fmt=quad fp */
+    OPC_W_FMT    = (0x14 << 21) | OPC_CP1, /* 20: fmt=32bit fixed */
+    OPC_L_FMT    = (0x15 << 21) | OPC_CP1, /* 21: fmt=64bit fixed */
+    OPC_PS_FMT   = (0x16 << 21) | OPC_CP1, /* 22: fmt=paired single fp */
+};
+
+#define MASK_CP1_FUNC(op)       MASK_CP1(op) | (op & 0x3F)
+#define MASK_BC1(op)            MASK_CP1(op) | (op & (0x3 << 16))
+
+enum {
+    OPC_BC1F     = (0x00 << 16) | OPC_BC1,
+    OPC_BC1T     = (0x01 << 16) | OPC_BC1,
+    OPC_BC1FL    = (0x02 << 16) | OPC_BC1,
+    OPC_BC1TL    = (0x03 << 16) | OPC_BC1,
+};
+
+enum {
+    OPC_BC1FANY2     = (0x00 << 16) | OPC_BC1ANY2,
+    OPC_BC1TANY2     = (0x01 << 16) | OPC_BC1ANY2,
+};
+
+enum {
+    OPC_BC1FANY4     = (0x00 << 16) | OPC_BC1ANY4,
+    OPC_BC1TANY4     = (0x01 << 16) | OPC_BC1ANY4,
+};
+
+#define MASK_CP2(op)       MASK_OP_MAJOR(op) | (op & (0x1F << 21))
+
+enum {
+    OPC_MFC2    = (0x00 << 21) | OPC_CP2,
+    OPC_DMFC2   = (0x01 << 21) | OPC_CP2,
+    OPC_CFC2    = (0x02 << 21) | OPC_CP2,
+    OPC_MFHC2   = (0x03 << 21) | OPC_CP2,
+    OPC_MTC2    = (0x04 << 21) | OPC_CP2,
+    OPC_DMTC2   = (0x05 << 21) | OPC_CP2,
+    OPC_CTC2    = (0x06 << 21) | OPC_CP2,
+    OPC_MTHC2   = (0x07 << 21) | OPC_CP2,
+    OPC_BC2     = (0x08 << 21) | OPC_CP2,
+};
+
+#define MASK_CP3(op)       MASK_OP_MAJOR(op) | (op & 0x3F)
+
+enum {
+    OPC_LWXC1   = 0x00 | OPC_CP3,
+    OPC_LDXC1   = 0x01 | OPC_CP3,
+    OPC_LUXC1   = 0x05 | OPC_CP3,
+    OPC_SWXC1   = 0x08 | OPC_CP3,
+    OPC_SDXC1   = 0x09 | OPC_CP3,
+    OPC_SUXC1   = 0x0D | OPC_CP3,
+    OPC_PREFX   = 0x0F | OPC_CP3,
+    OPC_ALNV_PS = 0x1E | OPC_CP3,
+    OPC_MADD_S  = 0x20 | OPC_CP3,
+    OPC_MADD_D  = 0x21 | OPC_CP3,
+    OPC_MADD_PS = 0x26 | OPC_CP3,
+    OPC_MSUB_S  = 0x28 | OPC_CP3,
+    OPC_MSUB_D  = 0x29 | OPC_CP3,
+    OPC_MSUB_PS = 0x2E | OPC_CP3,
+    OPC_NMADD_S = 0x30 | OPC_CP3,
+    OPC_NMADD_D = 0x31 | OPC_CP3,
+    OPC_NMADD_PS= 0x36 | OPC_CP3,
+    OPC_NMSUB_S = 0x38 | OPC_CP3,
+    OPC_NMSUB_D = 0x39 | OPC_CP3,
+    OPC_NMSUB_PS= 0x3E | OPC_CP3,
+};
+
+/* global register indices */
+static TCGv_ptr cpu_env;
+static TCGv cpu_gpr[32], cpu_PC;
+static TCGv cpu_HI[MIPS_DSP_ACC], cpu_LO[MIPS_DSP_ACC], cpu_ACX[MIPS_DSP_ACC];
+static TCGv cpu_dspctrl, btarget, bcond;
+static TCGv_i32 hflags;
+static TCGv_i32 fpu_fcr0, fpu_fcr31;
+
+#include "gen-icount.h"
+
+#define gen_helper_0i(name, arg) do {                             \
+    TCGv_i32 helper_tmp = tcg_const_i32(arg);                     \
+    gen_helper_##name(helper_tmp);                                \
+    tcg_temp_free_i32(helper_tmp);                                \
+    } while(0)
+
+#define gen_helper_1i(name, arg1, arg2) do {                      \
+    TCGv_i32 helper_tmp = tcg_const_i32(arg2);                    \
+    gen_helper_##name(arg1, helper_tmp);                          \
+    tcg_temp_free_i32(helper_tmp);                                \
+    } while(0)
+
+#define gen_helper_2i(name, arg1, arg2, arg3) do {                \
+    TCGv_i32 helper_tmp = tcg_const_i32(arg3);                    \
+    gen_helper_##name(arg1, arg2, helper_tmp);                    \
+    tcg_temp_free_i32(helper_tmp);                                \
+    } while(0)
+
+#define gen_helper_3i(name, arg1, arg2, arg3, arg4) do {          \
+    TCGv_i32 helper_tmp = tcg_const_i32(arg4);                    \
+    gen_helper_##name(arg1, arg2, arg3, helper_tmp);              \
+    tcg_temp_free_i32(helper_tmp);                                \
+    } while(0)
+
+typedef struct DisasContext {
+    struct TranslationBlock *tb;
+    target_ulong pc, saved_pc;
+    uint32_t opcode;
+    int singlestep_enabled;
+    /* Routine used to access memory */
+    int mem_idx;
+    uint32_t hflags, saved_hflags;
+    int bstate;
+    target_ulong btarget;
+} DisasContext;
+
+enum {
+    BS_NONE     = 0, /* We go out of the TB without reaching a branch or an
+                      * exception condition */
+    BS_STOP     = 1, /* We want to stop translation for any reason */
+    BS_BRANCH   = 2, /* We reached a branch condition     */
+    BS_EXCP     = 3, /* We reached an exception condition */
+};
+
+static const char *regnames[] =
+    { "r0", "at", "v0", "v1", "a0", "a1", "a2", "a3",
+      "t0", "t1", "t2", "t3", "t4", "t5", "t6", "t7",
+      "s0", "s1", "s2", "s3", "s4", "s5", "s6", "s7",
+      "t8", "t9", "k0", "k1", "gp", "sp", "s8", "ra", };
+
+static const char *regnames_HI[] =
+    { "HI0", "HI1", "HI2", "HI3", };
+
+static const char *regnames_LO[] =
+    { "LO0", "LO1", "LO2", "LO3", };
+
+static const char *regnames_ACX[] =
+    { "ACX0", "ACX1", "ACX2", "ACX3", };
+
+static const char *fregnames[] =
+    { "f0",  "f1",  "f2",  "f3",  "f4",  "f5",  "f6",  "f7",
+      "f8",  "f9",  "f10", "f11", "f12", "f13", "f14", "f15",
+      "f16", "f17", "f18", "f19", "f20", "f21", "f22", "f23",
+      "f24", "f25", "f26", "f27", "f28", "f29", "f30", "f31", };
+
+#ifdef MIPS_DEBUG_DISAS
+#define MIPS_DEBUG(fmt, ...)                         \
+        qemu_log_mask(CPU_LOG_TB_IN_ASM,                \
+                       TARGET_FMT_lx ": %08x " fmt "\n", \
+                       ctx->pc, ctx->opcode , ## __VA_ARGS__)
+#define LOG_DISAS(...) qemu_log_mask(CPU_LOG_TB_IN_ASM, ## __VA_ARGS__)
+#else
+#define MIPS_DEBUG(fmt, ...) do { } while(0)
+#define LOG_DISAS(...) do { } while (0)
+#endif
+
+#define MIPS_INVAL(op)                                                        \
+do {                                                                          \
+    MIPS_DEBUG("Invalid %s %03x %03x %03x", op, ctx->opcode >> 26,            \
+               ctx->opcode & 0x3F, ((ctx->opcode >> 16) & 0x1F));             \
+} while (0)
+
+/* General purpose registers moves. */
+static inline void gen_load_gpr (TCGv t, int reg)
+{
+    if (reg == 0)
+        tcg_gen_movi_tl(t, 0);
+    else
+        tcg_gen_mov_tl(t, cpu_gpr[reg]);
+}
+
+static inline void gen_store_gpr (TCGv t, int reg)
+{
+    if (reg != 0)
+        tcg_gen_mov_tl(cpu_gpr[reg], t);
+}
+
+/* Moves to/from ACX register.  */
+static inline void gen_load_ACX (TCGv t, int reg)
+{
+    tcg_gen_mov_tl(t, cpu_ACX[reg]);
+}
+
+static inline void gen_store_ACX (TCGv t, int reg)
+{
+    tcg_gen_mov_tl(cpu_ACX[reg], t);
+}
+
+/* Moves to/from shadow registers. */
+static inline void gen_load_srsgpr (int from, int to)
+{
+    TCGv t0 = tcg_temp_new();
+
+    if (from == 0)
+        tcg_gen_movi_tl(t0, 0);
+    else {
+        TCGv_i32 t2 = tcg_temp_new_i32();
+        TCGv_ptr addr = tcg_temp_new_ptr();
+
+        tcg_gen_ld_i32(t2, cpu_env, offsetof(CPUState, CP0_SRSCtl));
+        tcg_gen_shri_i32(t2, t2, CP0SRSCtl_PSS);
+        tcg_gen_andi_i32(t2, t2, 0xf);
+        tcg_gen_muli_i32(t2, t2, sizeof(target_ulong) * 32);
+        tcg_gen_ext_i32_ptr(addr, t2);
+        tcg_gen_add_ptr(addr, cpu_env, addr);
+
+        tcg_gen_ld_tl(t0, addr, sizeof(target_ulong) * from);
+        tcg_temp_free_ptr(addr);
+        tcg_temp_free_i32(t2);
+    }
+    gen_store_gpr(t0, to);
+    tcg_temp_free(t0);
+}
+
+static inline void gen_store_srsgpr (int from, int to)
+{
+    if (to != 0) {
+        TCGv t0 = tcg_temp_new();
+        TCGv_i32 t2 = tcg_temp_new_i32();
+        TCGv_ptr addr = tcg_temp_new_ptr();
+
+        gen_load_gpr(t0, from);
+        tcg_gen_ld_i32(t2, cpu_env, offsetof(CPUState, CP0_SRSCtl));
+        tcg_gen_shri_i32(t2, t2, CP0SRSCtl_PSS);
+        tcg_gen_andi_i32(t2, t2, 0xf);
+        tcg_gen_muli_i32(t2, t2, sizeof(target_ulong) * 32);
+        tcg_gen_ext_i32_ptr(addr, t2);
+        tcg_gen_add_ptr(addr, cpu_env, addr);
+
+        tcg_gen_st_tl(t0, addr, sizeof(target_ulong) * to);
+        tcg_temp_free_ptr(addr);
+        tcg_temp_free_i32(t2);
+        tcg_temp_free(t0);
+    }
+}
+
+/* Floating point register moves. */
+static inline void gen_load_fpr32 (TCGv_i32 t, int reg)
+{
+    tcg_gen_ld_i32(t, cpu_env, offsetof(CPUState, active_fpu.fpr[reg].w[FP_ENDIAN_IDX]));
+}
+
+static inline void gen_store_fpr32 (TCGv_i32 t, int reg)
+{
+    tcg_gen_st_i32(t, cpu_env, offsetof(CPUState, active_fpu.fpr[reg].w[FP_ENDIAN_IDX]));
+}
+
+static inline void gen_load_fpr32h (TCGv_i32 t, int reg)
+{
+    tcg_gen_ld_i32(t, cpu_env, offsetof(CPUState, active_fpu.fpr[reg].w[!FP_ENDIAN_IDX]));
+}
+
+static inline void gen_store_fpr32h (TCGv_i32 t, int reg)
+{
+    tcg_gen_st_i32(t, cpu_env, offsetof(CPUState, active_fpu.fpr[reg].w[!FP_ENDIAN_IDX]));
+}
+
+static inline void gen_load_fpr64 (DisasContext *ctx, TCGv_i64 t, int reg)
+{
+    if (ctx->hflags & MIPS_HFLAG_F64) {
+        tcg_gen_ld_i64(t, cpu_env, offsetof(CPUState, active_fpu.fpr[reg].d));
+    } else {
+        TCGv_i32 t0 = tcg_temp_new_i32();
+        TCGv_i32 t1 = tcg_temp_new_i32();
+        gen_load_fpr32(t0, reg & ~1);
+        gen_load_fpr32(t1, reg | 1);
+        tcg_gen_concat_i32_i64(t, t0, t1);
+        tcg_temp_free_i32(t0);
+        tcg_temp_free_i32(t1);
+    }
+}
+
+static inline void gen_store_fpr64 (DisasContext *ctx, TCGv_i64 t, int reg)
+{
+    if (ctx->hflags & MIPS_HFLAG_F64) {
+        tcg_gen_st_i64(t, cpu_env, offsetof(CPUState, active_fpu.fpr[reg].d));
+    } else {
+        TCGv_i64 t0 = tcg_temp_new_i64();
+        TCGv_i32 t1 = tcg_temp_new_i32();
+        tcg_gen_trunc_i64_i32(t1, t);
+        gen_store_fpr32(t1, reg & ~1);
+        tcg_gen_shri_i64(t0, t, 32);
+        tcg_gen_trunc_i64_i32(t1, t0);
+        gen_store_fpr32(t1, reg | 1);
+        tcg_temp_free_i32(t1);
+        tcg_temp_free_i64(t0);
+    }
+}
+
+static inline int get_fp_bit (int cc)
+{
+    if (cc)
+        return 24 + cc;
+    else
+        return 23;
+}
+
+#define FOP_CONDS(type, fmt, bits)                                            \
+static inline void gen_cmp ## type ## _ ## fmt(int n, TCGv_i##bits a,         \
+                                               TCGv_i##bits b, int cc)        \
+{                                                                             \
+    switch (n) {                                                              \
+    case  0: gen_helper_2i(cmp ## type ## _ ## fmt ## _f, a, b, cc);    break;\
+    case  1: gen_helper_2i(cmp ## type ## _ ## fmt ## _un, a, b, cc);   break;\
+    case  2: gen_helper_2i(cmp ## type ## _ ## fmt ## _eq, a, b, cc);   break;\
+    case  3: gen_helper_2i(cmp ## type ## _ ## fmt ## _ueq, a, b, cc);  break;\
+    case  4: gen_helper_2i(cmp ## type ## _ ## fmt ## _olt, a, b, cc);  break;\
+    case  5: gen_helper_2i(cmp ## type ## _ ## fmt ## _ult, a, b, cc);  break;\
+    case  6: gen_helper_2i(cmp ## type ## _ ## fmt ## _ole, a, b, cc);  break;\
+    case  7: gen_helper_2i(cmp ## type ## _ ## fmt ## _ule, a, b, cc);  break;\
+    case  8: gen_helper_2i(cmp ## type ## _ ## fmt ## _sf, a, b, cc);   break;\
+    case  9: gen_helper_2i(cmp ## type ## _ ## fmt ## _ngle, a, b, cc); break;\
+    case 10: gen_helper_2i(cmp ## type ## _ ## fmt ## _seq, a, b, cc);  break;\
+    case 11: gen_helper_2i(cmp ## type ## _ ## fmt ## _ngl, a, b, cc);  break;\
+    case 12: gen_helper_2i(cmp ## type ## _ ## fmt ## _lt, a, b, cc);   break;\
+    case 13: gen_helper_2i(cmp ## type ## _ ## fmt ## _nge, a, b, cc);  break;\
+    case 14: gen_helper_2i(cmp ## type ## _ ## fmt ## _le, a, b, cc);   break;\
+    case 15: gen_helper_2i(cmp ## type ## _ ## fmt ## _ngt, a, b, cc);  break;\
+    default: abort();                                                         \
+    }                                                                         \
+}
+
+FOP_CONDS(, d, 64)
+FOP_CONDS(abs, d, 64)
+FOP_CONDS(, s, 32)
+FOP_CONDS(abs, s, 32)
+FOP_CONDS(, ps, 64)
+FOP_CONDS(abs, ps, 64)
+#undef FOP_CONDS
+
+/* Tests */
+#define OP_COND(name, cond)                                         \
+static inline void glue(gen_op_, name) (TCGv ret, TCGv t0, TCGv t1) \
+{                                                                   \
+    int l1 = gen_new_label();                                       \
+    int l2 = gen_new_label();                                       \
+                                                                    \
+    tcg_gen_brcond_tl(cond, t0, t1, l1);                            \
+    tcg_gen_movi_tl(ret, 0);                                        \
+    tcg_gen_br(l2);                                                 \
+    gen_set_label(l1);                                              \
+    tcg_gen_movi_tl(ret, 1);                                        \
+    gen_set_label(l2);                                              \
+}
+OP_COND(eq, TCG_COND_EQ);
+OP_COND(ne, TCG_COND_NE);
+OP_COND(ge, TCG_COND_GE);
+OP_COND(geu, TCG_COND_GEU);
+OP_COND(lt, TCG_COND_LT);
+OP_COND(ltu, TCG_COND_LTU);
+#undef OP_COND
+
+#define OP_CONDI(name, cond)                                                 \
+static inline void glue(gen_op_, name) (TCGv ret, TCGv t0, target_ulong val) \
+{                                                                            \
+    int l1 = gen_new_label();                                                \
+    int l2 = gen_new_label();                                                \
+                                                                             \
+    tcg_gen_brcondi_tl(cond, t0, val, l1);                                   \
+    tcg_gen_movi_tl(ret, 0);                                                 \
+    tcg_gen_br(l2);                                                          \
+    gen_set_label(l1);                                                       \
+    tcg_gen_movi_tl(ret, 1);                                                 \
+    gen_set_label(l2);                                                       \
+}
+OP_CONDI(lti, TCG_COND_LT);
+OP_CONDI(ltiu, TCG_COND_LTU);
+#undef OP_CONDI
+
+#define OP_CONDZ(name, cond)                                  \
+static inline void glue(gen_op_, name) (TCGv ret, TCGv t0)    \
+{                                                             \
+    int l1 = gen_new_label();                                 \
+    int l2 = gen_new_label();                                 \
+                                                              \
+    tcg_gen_brcondi_tl(cond, t0, 0, l1);                      \
+    tcg_gen_movi_tl(ret, 0);                                  \
+    tcg_gen_br(l2);                                           \
+    gen_set_label(l1);                                        \
+    tcg_gen_movi_tl(ret, 1);                                  \
+    gen_set_label(l2);                                        \
+}
+OP_CONDZ(gez, TCG_COND_GE);
+OP_CONDZ(gtz, TCG_COND_GT);
+OP_CONDZ(lez, TCG_COND_LE);
+OP_CONDZ(ltz, TCG_COND_LT);
+#undef OP_CONDZ
+
+static inline void gen_save_pc(target_ulong pc)
+{
+    tcg_gen_movi_tl(cpu_PC, pc);
+}
+
+static inline void save_cpu_state (DisasContext *ctx, int do_save_pc)
+{
+    LOG_DISAS("hflags %08x saved %08x\n", ctx->hflags, ctx->saved_hflags);
+    if (do_save_pc && ctx->pc != ctx->saved_pc) {
+        gen_save_pc(ctx->pc);
+        ctx->saved_pc = ctx->pc;
+    }
+    if (ctx->hflags != ctx->saved_hflags) {
+        tcg_gen_movi_i32(hflags, ctx->hflags);
+        ctx->saved_hflags = ctx->hflags;
+        switch (ctx->hflags & MIPS_HFLAG_BMASK) {
+        case MIPS_HFLAG_BR:
+            break;
+        case MIPS_HFLAG_BC:
+        case MIPS_HFLAG_BL:
+        case MIPS_HFLAG_B:
+            tcg_gen_movi_tl(btarget, ctx->btarget);
+            break;
+        }
+    }
+}
+
+static inline void restore_cpu_state (CPUState *env, DisasContext *ctx)
+{
+    ctx->saved_hflags = ctx->hflags;
+    switch (ctx->hflags & MIPS_HFLAG_BMASK) {
+    case MIPS_HFLAG_BR:
+        break;
+    case MIPS_HFLAG_BC:
+    case MIPS_HFLAG_BL:
+    case MIPS_HFLAG_B:
+        ctx->btarget = env->btarget;
+        break;
+    }
+}
+
+static inline void
+generate_exception_err (DisasContext *ctx, int excp, int err)
+{
+    TCGv_i32 texcp = tcg_const_i32(excp);
+    TCGv_i32 terr = tcg_const_i32(err);
+    save_cpu_state(ctx, 1);
+    gen_helper_raise_exception_err(texcp, terr);
+    tcg_temp_free_i32(terr);
+    tcg_temp_free_i32(texcp);
+}
+
+static inline void
+generate_exception (DisasContext *ctx, int excp)
+{
+    save_cpu_state(ctx, 1);
+    gen_helper_0i(raise_exception, excp);
+}
+
+/* Addresses computation */
+static inline void gen_op_addr_add (DisasContext *ctx, TCGv ret, TCGv arg0, TCGv arg1)
+{
+    tcg_gen_add_tl(ret, arg0, arg1);
+
+#if defined(TARGET_MIPS64)
+    /* For compatibility with 32-bit code, data reference in user mode
+       with Status_UX = 0 should be casted to 32-bit and sign extended.
+       See the MIPS64 PRA manual, section 4.10. */
+    if (((ctx->hflags & MIPS_HFLAG_KSU) == MIPS_HFLAG_UM) &&
+        !(ctx->hflags & MIPS_HFLAG_UX)) {
+        tcg_gen_ext32s_i64(ret, ret);
+    }
+#endif
+}
+
+static inline void check_cp0_enabled(DisasContext *ctx)
+{
+    if (unlikely(!(ctx->hflags & MIPS_HFLAG_CP0)))
+        generate_exception_err(ctx, EXCP_CpU, 0);
+}
+
+static inline void check_cp1_enabled(DisasContext *ctx)
+{
+    if (unlikely(!(ctx->hflags & MIPS_HFLAG_FPU)))
+        generate_exception_err(ctx, EXCP_CpU, 1);
+}
+
+/* Verify that the processor is running with COP1X instructions enabled.
+   This is associated with the nabla symbol in the MIPS32 and MIPS64
+   opcode tables.  */
+
+static inline void check_cop1x(DisasContext *ctx)
+{
+    if (unlikely(!(ctx->hflags & MIPS_HFLAG_COP1X)))
+        generate_exception(ctx, EXCP_RI);
+}
+
+/* Verify that the processor is running with 64-bit floating-point
+   operations enabled.  */
+
+static inline void check_cp1_64bitmode(DisasContext *ctx)
+{
+    if (unlikely(~ctx->hflags & (MIPS_HFLAG_F64 | MIPS_HFLAG_COP1X)))
+        generate_exception(ctx, EXCP_RI);
+}
+
+/*
+ * Verify if floating point register is valid; an operation is not defined
+ * if bit 0 of any register specification is set and the FR bit in the
+ * Status register equals zero, since the register numbers specify an
+ * even-odd pair of adjacent coprocessor general registers. When the FR bit
+ * in the Status register equals one, both even and odd register numbers
+ * are valid. This limitation exists only for 64 bit wide (d,l,ps) registers.
+ *
+ * Multiple 64 bit wide registers can be checked by calling
+ * gen_op_cp1_registers(freg1 | freg2 | ... | fregN);
+ */
+static inline void check_cp1_registers(DisasContext *ctx, int regs)
+{
+    if (unlikely(!(ctx->hflags & MIPS_HFLAG_F64) && (regs & 1)))
+        generate_exception(ctx, EXCP_RI);
+}
+
+/* This code generates a "reserved instruction" exception if the
+   CPU does not support the instruction set corresponding to flags. */
+static inline void check_insn(CPUState *env, DisasContext *ctx, int flags)
+{
+    if (unlikely(!(env->insn_flags & flags)))
+        generate_exception(ctx, EXCP_RI);
+}
+
+/* This code generates a "reserved instruction" exception if 64-bit
+   instructions are not enabled. */
+static inline void check_mips_64(DisasContext *ctx)
+{
+    if (unlikely(!(ctx->hflags & MIPS_HFLAG_64)))
+        generate_exception(ctx, EXCP_RI);
+}
+
+/* load/store instructions. */
+#define OP_LD(insn,fname)                                                 \
+static inline void op_ldst_##insn(TCGv ret, TCGv arg1, DisasContext *ctx) \
+{                                                                         \
+    tcg_gen_qemu_##fname(ret, arg1, ctx->mem_idx);                        \
+}
+OP_LD(lb,ld8s);
+OP_LD(lbu,ld8u);
+OP_LD(lh,ld16s);
+OP_LD(lhu,ld16u);
+OP_LD(lw,ld32s);
+#if defined(TARGET_MIPS64)
+OP_LD(lwu,ld32u);
+OP_LD(ld,ld64);
+#endif
+#undef OP_LD
+
+#define OP_ST(insn,fname)                                                  \
+static inline void op_ldst_##insn(TCGv arg1, TCGv arg2, DisasContext *ctx) \
+{                                                                          \
+    tcg_gen_qemu_##fname(arg1, arg2, ctx->mem_idx);                        \
+}
+OP_ST(sb,st8);
+OP_ST(sh,st16);
+OP_ST(sw,st32);
+#if defined(TARGET_MIPS64)
+OP_ST(sd,st64);
+#endif
+#undef OP_ST
+
+#ifdef CONFIG_USER_ONLY
+#define OP_LD_ATOMIC(insn,fname)                                           \
+static inline void op_ldst_##insn(TCGv ret, TCGv arg1, DisasContext *ctx)  \
+{                                                                          \
+    TCGv t0 = tcg_temp_new();                                              \
+    tcg_gen_mov_tl(t0, arg1);                                              \
+    tcg_gen_qemu_##fname(ret, arg1, ctx->mem_idx);                         \
+    tcg_gen_st_tl(t0, cpu_env, offsetof(CPUState, lladdr));                \
+    tcg_gen_st_tl(ret, cpu_env, offsetof(CPUState, llval));                \
+    tcg_temp_free(t0);                                                     \
+}
+#else
+#define OP_LD_ATOMIC(insn,fname)                                           \
+static inline void op_ldst_##insn(TCGv ret, TCGv arg1, DisasContext *ctx)  \
+{                                                                          \
+    gen_helper_2i(insn, ret, arg1, ctx->mem_idx);                          \
+}
+#endif
+OP_LD_ATOMIC(ll,ld32s);
+#if defined(TARGET_MIPS64)
+OP_LD_ATOMIC(lld,ld64);
+#endif
+#undef OP_LD_ATOMIC
+
+#ifdef CONFIG_USER_ONLY
+#define OP_ST_ATOMIC(insn,fname,ldname,almask)                               \
+static inline void op_ldst_##insn(TCGv arg1, TCGv arg2, int rt, DisasContext *ctx) \
+{                                                                            \
+    TCGv t0 = tcg_temp_new();                                                \
+    int l1 = gen_new_label();                                                \
+    int l2 = gen_new_label();                                                \
+                                                                             \
+    tcg_gen_andi_tl(t0, arg2, almask);                                       \
+    tcg_gen_brcondi_tl(TCG_COND_EQ, t0, 0, l1);                              \
+    tcg_gen_st_tl(arg2, cpu_env, offsetof(CPUState, CP0_BadVAddr));          \
+    generate_exception(ctx, EXCP_AdES);                                      \
+    gen_set_label(l1);                                                       \
+    tcg_gen_ld_tl(t0, cpu_env, offsetof(CPUState, lladdr));                  \
+    tcg_gen_brcond_tl(TCG_COND_NE, arg2, t0, l2);                            \
+    tcg_gen_movi_tl(t0, rt | ((almask << 3) & 0x20));                        \
+    tcg_gen_st_tl(t0, cpu_env, offsetof(CPUState, llreg));                   \
+    tcg_gen_st_tl(arg1, cpu_env, offsetof(CPUState, llnewval));              \
+    gen_helper_0i(raise_exception, EXCP_SC);                                 \
+    gen_set_label(l2);                                                       \
+    tcg_gen_movi_tl(t0, 0);                                                  \
+    gen_store_gpr(t0, rt);                                                   \
+    tcg_temp_free(t0);                                                       \
+}
+#else
+#define OP_ST_ATOMIC(insn,fname,ldname,almask)                               \
+static inline void op_ldst_##insn(TCGv arg1, TCGv arg2, int rt, DisasContext *ctx) \
+{                                                                            \
+    TCGv t0 = tcg_temp_new();                                                \
+    gen_helper_3i(insn, t0, arg1, arg2, ctx->mem_idx);                       \
+    gen_store_gpr(t0, rt);                                                   \
+    tcg_temp_free(t0);                                                       \
+}
+#endif
+OP_ST_ATOMIC(sc,st32,ld32s,0x3);
+#if defined(TARGET_MIPS64)
+OP_ST_ATOMIC(scd,st64,ld64,0x7);
+#endif
+#undef OP_ST_ATOMIC
+
+/* Load and store */
+static void gen_ldst (DisasContext *ctx, uint32_t opc, int rt,
+                      int base, int16_t offset)
+{
+    const char *opn = "ldst";
+    TCGv t0 = tcg_temp_new();
+    TCGv t1 = tcg_temp_new();
+
+    if (base == 0) {
+        tcg_gen_movi_tl(t0, offset);
+    } else if (offset == 0) {
+        gen_load_gpr(t0, base);
+    } else {
+        tcg_gen_movi_tl(t0, offset);
+        gen_op_addr_add(ctx, t0, cpu_gpr[base], t0);
+    }
+    /* Don't do NOP if destination is zero: we must perform the actual
+       memory access. */
+    switch (opc) {
+#if defined(TARGET_MIPS64)
+    case OPC_LWU:
+        save_cpu_state(ctx, 0);
+        op_ldst_lwu(t0, t0, ctx);
+        gen_store_gpr(t0, rt);
+        opn = "lwu";
+        break;
+    case OPC_LD:
+        save_cpu_state(ctx, 0);
+        op_ldst_ld(t0, t0, ctx);
+        gen_store_gpr(t0, rt);
+        opn = "ld";
+        break;
+    case OPC_LLD:
+        save_cpu_state(ctx, 0);
+        op_ldst_lld(t0, t0, ctx);
+        gen_store_gpr(t0, rt);
+        opn = "lld";
+        break;
+    case OPC_SD:
+        save_cpu_state(ctx, 0);
+        gen_load_gpr(t1, rt);
+        op_ldst_sd(t1, t0, ctx);
+        opn = "sd";
+        break;
+    case OPC_LDL:
+        save_cpu_state(ctx, 1);
+        gen_load_gpr(t1, rt);
+        gen_helper_3i(ldl, t1, t1, t0, ctx->mem_idx);
+        gen_store_gpr(t1, rt);
+        opn = "ldl";
+        break;
+    case OPC_SDL:
+        save_cpu_state(ctx, 1);
+        gen_load_gpr(t1, rt);
+        gen_helper_2i(sdl, t1, t0, ctx->mem_idx);
+        opn = "sdl";
+        break;
+    case OPC_LDR:
+        save_cpu_state(ctx, 1);
+        gen_load_gpr(t1, rt);
+        gen_helper_3i(ldr, t1, t1, t0, ctx->mem_idx);
+        gen_store_gpr(t1, rt);
+        opn = "ldr";
+        break;
+    case OPC_SDR:
+        save_cpu_state(ctx, 1);
+        gen_load_gpr(t1, rt);
+        gen_helper_2i(sdr, t1, t0, ctx->mem_idx);
+        opn = "sdr";
+        break;
+#endif
+    case OPC_LW:
+        save_cpu_state(ctx, 0);
+        op_ldst_lw(t0, t0, ctx);
+        gen_store_gpr(t0, rt);
+        opn = "lw";
+        break;
+    case OPC_SW:
+        save_cpu_state(ctx, 0);
+        gen_load_gpr(t1, rt);
+        op_ldst_sw(t1, t0, ctx);
+        opn = "sw";
+        break;
+    case OPC_LH:
+        save_cpu_state(ctx, 0);
+        op_ldst_lh(t0, t0, ctx);
+        gen_store_gpr(t0, rt);
+        opn = "lh";
+        break;
+    case OPC_SH:
+        save_cpu_state(ctx, 0);
+        gen_load_gpr(t1, rt);
+        op_ldst_sh(t1, t0, ctx);
+        opn = "sh";
+        break;
+    case OPC_LHU:
+        save_cpu_state(ctx, 0);
+        op_ldst_lhu(t0, t0, ctx);
+        gen_store_gpr(t0, rt);
+        opn = "lhu";
+        break;
+    case OPC_LB:
+        save_cpu_state(ctx, 0);
+        op_ldst_lb(t0, t0, ctx);
+        gen_store_gpr(t0, rt);
+        opn = "lb";
+        break;
+    case OPC_SB:
+        save_cpu_state(ctx, 0);
+        gen_load_gpr(t1, rt);
+        op_ldst_sb(t1, t0, ctx);
+        opn = "sb";
+        break;
+    case OPC_LBU:
+        save_cpu_state(ctx, 0);
+        op_ldst_lbu(t0, t0, ctx);
+        gen_store_gpr(t0, rt);
+        opn = "lbu";
+        break;
+    case OPC_LWL:
+        save_cpu_state(ctx, 1);
+        gen_load_gpr(t1, rt);
+        gen_helper_3i(lwl, t1, t1, t0, ctx->mem_idx);
+        gen_store_gpr(t1, rt);
+        opn = "lwl";
+        break;
+    case OPC_SWL:
+        save_cpu_state(ctx, 1);
+        gen_load_gpr(t1, rt);
+        gen_helper_2i(swl, t1, t0, ctx->mem_idx);
+        opn = "swr";
+        break;
+    case OPC_LWR:
+        save_cpu_state(ctx, 1);
+        gen_load_gpr(t1, rt);
+        gen_helper_3i(lwr, t1, t1, t0, ctx->mem_idx);
+        gen_store_gpr(t1, rt);
+        opn = "lwr";
+        break;
+    case OPC_SWR:
+        save_cpu_state(ctx, 1);
+        gen_load_gpr(t1, rt);
+        gen_helper_2i(swr, t1, t0, ctx->mem_idx);
+        opn = "swr";
+        break;
+    case OPC_LL:
+        save_cpu_state(ctx, 1);
+        op_ldst_ll(t0, t0, ctx);
+        gen_store_gpr(t0, rt);
+        opn = "ll";
+        break;
+    }
+    MIPS_DEBUG("%s %s, %d(%s)", opn, regnames[rt], offset, regnames[base]);
+    tcg_temp_free(t0);
+    tcg_temp_free(t1);
+}
+
+/* Store conditional */
+static void gen_st_cond (DisasContext *ctx, uint32_t opc, int rt,
+                         int base, int16_t offset)
+{
+    const char *opn = "st_cond";
+    TCGv t0, t1;
+
+    t0 = tcg_temp_local_new();
+
+    if (base == 0) {
+        tcg_gen_movi_tl(t0, offset);
+    } else if (offset == 0) {
+        gen_load_gpr(t0, base);
+    } else {
+        tcg_gen_movi_tl(t0, offset);
+        gen_op_addr_add(ctx, t0, cpu_gpr[base], t0);
+    }
+    /* Don't do NOP if destination is zero: we must perform the actual
+       memory access. */
+
+    t1 = tcg_temp_local_new();
+    gen_load_gpr(t1, rt);
+    switch (opc) {
+#if defined(TARGET_MIPS64)
+    case OPC_SCD:
+        save_cpu_state(ctx, 0);
+        op_ldst_scd(t1, t0, rt, ctx);
+        opn = "scd";
+        break;
+#endif
+    case OPC_SC:
+        save_cpu_state(ctx, 1);
+        op_ldst_sc(t1, t0, rt, ctx);
+        opn = "sc";
+        break;
+    }
+    MIPS_DEBUG("%s %s, %d(%s)", opn, regnames[rt], offset, regnames[base]);
+    tcg_temp_free(t1);
+    tcg_temp_free(t0);
+}
+
+/* Load and store */
+static void gen_flt_ldst (DisasContext *ctx, uint32_t opc, int ft,
+                          int base, int16_t offset)
+{
+    const char *opn = "flt_ldst";
+    TCGv t0 = tcg_temp_new();
+
+    if (base == 0) {
+        tcg_gen_movi_tl(t0, offset);
+    } else if (offset == 0) {
+        gen_load_gpr(t0, base);
+    } else {
+        tcg_gen_movi_tl(t0, offset);
+        gen_op_addr_add(ctx, t0, cpu_gpr[base], t0);
+    }
+    /* Don't do NOP if destination is zero: we must perform the actual
+       memory access. */
+    switch (opc) {
+    case OPC_LWC1:
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            tcg_gen_qemu_ld32s(t0, t0, ctx->mem_idx);
+            tcg_gen_trunc_tl_i32(fp0, t0);
+            gen_store_fpr32(fp0, ft);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "lwc1";
+        break;
+    case OPC_SWC1:
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv t1 = tcg_temp_new();
+
+            gen_load_fpr32(fp0, ft);
+            tcg_gen_extu_i32_tl(t1, fp0);
+            tcg_gen_qemu_st32(t1, t0, ctx->mem_idx);
+            tcg_temp_free(t1);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "swc1";
+        break;
+    case OPC_LDC1:
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            tcg_gen_qemu_ld64(fp0, t0, ctx->mem_idx);
+            gen_store_fpr64(ctx, fp0, ft);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "ldc1";
+        break;
+    case OPC_SDC1:
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, ft);
+            tcg_gen_qemu_st64(fp0, t0, ctx->mem_idx);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "sdc1";
+        break;
+    default:
+        MIPS_INVAL(opn);
+        generate_exception(ctx, EXCP_RI);
+        goto out;
+    }
+    MIPS_DEBUG("%s %s, %d(%s)", opn, fregnames[ft], offset, regnames[base]);
+ out:
+    tcg_temp_free(t0);
+}
+
+/* Arithmetic with immediate operand */
+static void gen_arith_imm (CPUState *env, DisasContext *ctx, uint32_t opc,
+                           int rt, int rs, int16_t imm)
+{
+    target_ulong uimm = (target_long)imm; /* Sign extend to 32/64 bits */
+    const char *opn = "imm arith";
+
+    if (rt == 0 && opc != OPC_ADDI && opc != OPC_DADDI) {
+        /* If no destination, treat it as a NOP.
+           For addi, we must generate the overflow exception when needed. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+    switch (opc) {
+    case OPC_ADDI:
+        {
+            TCGv t0 = tcg_temp_local_new();
+            TCGv t1 = tcg_temp_new();
+            TCGv t2 = tcg_temp_new();
+            int l1 = gen_new_label();
+
+            gen_load_gpr(t1, rs);
+            tcg_gen_addi_tl(t0, t1, uimm);
+            tcg_gen_ext32s_tl(t0, t0);
+
+            tcg_gen_xori_tl(t1, t1, ~uimm);
+            tcg_gen_xori_tl(t2, t0, uimm);
+            tcg_gen_and_tl(t1, t1, t2);
+            tcg_temp_free(t2);
+            tcg_gen_brcondi_tl(TCG_COND_GE, t1, 0, l1);
+            tcg_temp_free(t1);
+            /* operands of same sign, result different sign */
+            generate_exception(ctx, EXCP_OVERFLOW);
+            gen_set_label(l1);
+            tcg_gen_ext32s_tl(t0, t0);
+            gen_store_gpr(t0, rt);
+            tcg_temp_free(t0);
+        }
+        opn = "addi";
+        break;
+    case OPC_ADDIU:
+        if (rs != 0) {
+            tcg_gen_addi_tl(cpu_gpr[rt], cpu_gpr[rs], uimm);
+            tcg_gen_ext32s_tl(cpu_gpr[rt], cpu_gpr[rt]);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rt], uimm);
+        }
+        opn = "addiu";
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DADDI:
+        {
+            TCGv t0 = tcg_temp_local_new();
+            TCGv t1 = tcg_temp_new();
+            TCGv t2 = tcg_temp_new();
+            int l1 = gen_new_label();
+
+            gen_load_gpr(t1, rs);
+            tcg_gen_addi_tl(t0, t1, uimm);
+
+            tcg_gen_xori_tl(t1, t1, ~uimm);
+            tcg_gen_xori_tl(t2, t0, uimm);
+            tcg_gen_and_tl(t1, t1, t2);
+            tcg_temp_free(t2);
+            tcg_gen_brcondi_tl(TCG_COND_GE, t1, 0, l1);
+            tcg_temp_free(t1);
+            /* operands of same sign, result different sign */
+            generate_exception(ctx, EXCP_OVERFLOW);
+            gen_set_label(l1);
+            gen_store_gpr(t0, rt);
+            tcg_temp_free(t0);
+        }
+        opn = "daddi";
+        break;
+    case OPC_DADDIU:
+        if (rs != 0) {
+            tcg_gen_addi_tl(cpu_gpr[rt], cpu_gpr[rs], uimm);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rt], uimm);
+        }
+        opn = "daddiu";
+        break;
+#endif
+    }
+    MIPS_DEBUG("%s %s, %s, " TARGET_FMT_lx, opn, regnames[rt], regnames[rs], uimm);
+}
+
+/* Logic with immediate operand */
+static void gen_logic_imm (CPUState *env, uint32_t opc, int rt, int rs, int16_t imm)
+{
+    target_ulong uimm;
+    const char *opn = "imm logic";
+
+    if (rt == 0) {
+        /* If no destination, treat it as a NOP. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+    uimm = (uint16_t)imm;
+    switch (opc) {
+    case OPC_ANDI:
+        if (likely(rs != 0))
+            tcg_gen_andi_tl(cpu_gpr[rt], cpu_gpr[rs], uimm);
+        else
+            tcg_gen_movi_tl(cpu_gpr[rt], 0);
+        opn = "andi";
+        break;
+    case OPC_ORI:
+        if (rs != 0)
+            tcg_gen_ori_tl(cpu_gpr[rt], cpu_gpr[rs], uimm);
+        else
+            tcg_gen_movi_tl(cpu_gpr[rt], uimm);
+        opn = "ori";
+        break;
+    case OPC_XORI:
+        if (likely(rs != 0))
+            tcg_gen_xori_tl(cpu_gpr[rt], cpu_gpr[rs], uimm);
+        else
+            tcg_gen_movi_tl(cpu_gpr[rt], uimm);
+        opn = "xori";
+        break;
+    case OPC_LUI:
+        tcg_gen_movi_tl(cpu_gpr[rt], imm << 16);
+        opn = "lui";
+        break;
+    }
+    MIPS_DEBUG("%s %s, %s, " TARGET_FMT_lx, opn, regnames[rt], regnames[rs], uimm);
+}
+
+/* Set on less than with immediate operand */
+static void gen_slt_imm (CPUState *env, uint32_t opc, int rt, int rs, int16_t imm)
+{
+    target_ulong uimm = (target_long)imm; /* Sign extend to 32/64 bits */
+    const char *opn = "imm arith";
+    TCGv t0;
+
+    if (rt == 0) {
+        /* If no destination, treat it as a NOP. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+    t0 = tcg_temp_new();
+    gen_load_gpr(t0, rs);
+    switch (opc) {
+    case OPC_SLTI:
+        gen_op_lti(cpu_gpr[rt], t0, uimm);
+        opn = "slti";
+        break;
+    case OPC_SLTIU:
+        gen_op_ltiu(cpu_gpr[rt], t0, uimm);
+        opn = "sltiu";
+        break;
+    }
+    MIPS_DEBUG("%s %s, %s, " TARGET_FMT_lx, opn, regnames[rt], regnames[rs], uimm);
+    tcg_temp_free(t0);
+}
+
+/* Shifts with immediate operand */
+static void gen_shift_imm(CPUState *env, DisasContext *ctx, uint32_t opc,
+                          int rt, int rs, int16_t imm)
+{
+    target_ulong uimm = ((uint16_t)imm) & 0x1f;
+    const char *opn = "imm shift";
+    TCGv t0;
+
+    if (rt == 0) {
+        /* If no destination, treat it as a NOP. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+
+    t0 = tcg_temp_new();
+    gen_load_gpr(t0, rs);
+    switch (opc) {
+    case OPC_SLL:
+        tcg_gen_shli_tl(t0, t0, uimm);
+        tcg_gen_ext32s_tl(cpu_gpr[rt], t0);
+        opn = "sll";
+        break;
+    case OPC_SRA:
+        tcg_gen_ext32s_tl(t0, t0);
+        tcg_gen_sari_tl(cpu_gpr[rt], t0, uimm);
+        opn = "sra";
+        break;
+    case OPC_SRL:
+        switch ((ctx->opcode >> 21) & 0x1f) {
+        case 0:
+            if (uimm != 0) {
+                tcg_gen_ext32u_tl(t0, t0);
+                tcg_gen_shri_tl(cpu_gpr[rt], t0, uimm);
+            } else {
+                tcg_gen_ext32s_tl(cpu_gpr[rt], t0);
+            }
+            opn = "srl";
+            break;
+        case 1:
+            /* rotr is decoded as srl on non-R2 CPUs */
+            if (env->insn_flags & ISA_MIPS32R2) {
+                if (uimm != 0) {
+                    TCGv_i32 t1 = tcg_temp_new_i32();
+
+                    tcg_gen_trunc_tl_i32(t1, t0);
+                    tcg_gen_rotri_i32(t1, t1, uimm);
+                    tcg_gen_ext_i32_tl(cpu_gpr[rt], t1);
+                    tcg_temp_free_i32(t1);
+                } else {
+                    tcg_gen_ext32s_tl(cpu_gpr[rt], t0);
+                }
+                opn = "rotr";
+            } else {
+                if (uimm != 0) {
+                    tcg_gen_ext32u_tl(t0, t0);
+                    tcg_gen_shri_tl(cpu_gpr[rt], t0, uimm);
+                } else {
+                    tcg_gen_ext32s_tl(cpu_gpr[rt], t0);
+                }
+                opn = "srl";
+            }
+            break;
+        default:
+            MIPS_INVAL("invalid srl flag");
+            generate_exception(ctx, EXCP_RI);
+            break;
+        }
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DSLL:
+        tcg_gen_shli_tl(cpu_gpr[rt], t0, uimm);
+        opn = "dsll";
+        break;
+    case OPC_DSRA:
+        tcg_gen_sari_tl(cpu_gpr[rt], t0, uimm);
+        opn = "dsra";
+        break;
+    case OPC_DSRL:
+        switch ((ctx->opcode >> 21) & 0x1f) {
+        case 0:
+            tcg_gen_shri_tl(cpu_gpr[rt], t0, uimm);
+            opn = "dsrl";
+            break;
+        case 1:
+            /* drotr is decoded as dsrl on non-R2 CPUs */
+            if (env->insn_flags & ISA_MIPS32R2) {
+                if (uimm != 0) {
+                    tcg_gen_rotri_tl(cpu_gpr[rt], t0, uimm);
+                } else {
+                    tcg_gen_mov_tl(cpu_gpr[rt], t0);
+                }
+                opn = "drotr";
+            } else {
+                tcg_gen_shri_tl(cpu_gpr[rt], t0, uimm);
+                opn = "dsrl";
+            }
+            break;
+        default:
+            MIPS_INVAL("invalid dsrl flag");
+            generate_exception(ctx, EXCP_RI);
+            break;
+        }
+        break;
+    case OPC_DSLL32:
+        tcg_gen_shli_tl(cpu_gpr[rt], t0, uimm + 32);
+        opn = "dsll32";
+        break;
+    case OPC_DSRA32:
+        tcg_gen_sari_tl(cpu_gpr[rt], t0, uimm + 32);
+        opn = "dsra32";
+        break;
+    case OPC_DSRL32:
+        switch ((ctx->opcode >> 21) & 0x1f) {
+        case 0:
+            tcg_gen_shri_tl(cpu_gpr[rt], t0, uimm + 32);
+            opn = "dsrl32";
+            break;
+        case 1:
+            /* drotr32 is decoded as dsrl32 on non-R2 CPUs */
+            if (env->insn_flags & ISA_MIPS32R2) {
+                tcg_gen_rotri_tl(cpu_gpr[rt], t0, uimm + 32);
+                opn = "drotr32";
+            } else {
+                tcg_gen_shri_tl(cpu_gpr[rt], t0, uimm + 32);
+                opn = "dsrl32";
+            }
+            break;
+        default:
+            MIPS_INVAL("invalid dsrl32 flag");
+            generate_exception(ctx, EXCP_RI);
+            break;
+        }
+        break;
+#endif
+    }
+    MIPS_DEBUG("%s %s, %s, " TARGET_FMT_lx, opn, regnames[rt], regnames[rs], uimm);
+    tcg_temp_free(t0);
+}
+
+/* Arithmetic */
+static void gen_arith (CPUState *env, DisasContext *ctx, uint32_t opc,
+                       int rd, int rs, int rt)
+{
+    const char *opn = "arith";
+
+    if (rd == 0 && opc != OPC_ADD && opc != OPC_SUB
+       && opc != OPC_DADD && opc != OPC_DSUB) {
+        /* If no destination, treat it as a NOP.
+           For add & sub, we must generate the overflow exception when needed. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+
+    switch (opc) {
+    case OPC_ADD:
+        {
+            TCGv t0 = tcg_temp_local_new();
+            TCGv t1 = tcg_temp_new();
+            TCGv t2 = tcg_temp_new();
+            int l1 = gen_new_label();
+
+            gen_load_gpr(t1, rs);
+            gen_load_gpr(t2, rt);
+            tcg_gen_add_tl(t0, t1, t2);
+            tcg_gen_ext32s_tl(t0, t0);
+            tcg_gen_xor_tl(t1, t1, t2);
+            tcg_gen_not_tl(t1, t1);
+            tcg_gen_xor_tl(t2, t0, t2);
+            tcg_gen_and_tl(t1, t1, t2);
+            tcg_temp_free(t2);
+            tcg_gen_brcondi_tl(TCG_COND_GE, t1, 0, l1);
+            tcg_temp_free(t1);
+            /* operands of same sign, result different sign */
+            generate_exception(ctx, EXCP_OVERFLOW);
+            gen_set_label(l1);
+            gen_store_gpr(t0, rd);
+            tcg_temp_free(t0);
+        }
+        opn = "add";
+        break;
+    case OPC_ADDU:
+        if (rs != 0 && rt != 0) {
+            tcg_gen_add_tl(cpu_gpr[rd], cpu_gpr[rs], cpu_gpr[rt]);
+            tcg_gen_ext32s_tl(cpu_gpr[rd], cpu_gpr[rd]);
+        } else if (rs == 0 && rt != 0) {
+            tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rt]);
+        } else if (rs != 0 && rt == 0) {
+            tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rs]);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rd], 0);
+        }
+        opn = "addu";
+        break;
+    case OPC_SUB:
+        {
+            TCGv t0 = tcg_temp_local_new();
+            TCGv t1 = tcg_temp_new();
+            TCGv t2 = tcg_temp_new();
+            int l1 = gen_new_label();
+
+            gen_load_gpr(t1, rs);
+            gen_load_gpr(t2, rt);
+            tcg_gen_sub_tl(t0, t1, t2);
+            tcg_gen_ext32s_tl(t0, t0);
+            tcg_gen_xor_tl(t2, t1, t2);
+            tcg_gen_xor_tl(t1, t0, t1);
+            tcg_gen_and_tl(t1, t1, t2);
+            tcg_temp_free(t2);
+            tcg_gen_brcondi_tl(TCG_COND_GE, t1, 0, l1);
+            tcg_temp_free(t1);
+            /* operands of different sign, first operand and result different sign */
+            generate_exception(ctx, EXCP_OVERFLOW);
+            gen_set_label(l1);
+            gen_store_gpr(t0, rd);
+            tcg_temp_free(t0);
+        }
+        opn = "sub";
+        break;
+    case OPC_SUBU:
+        if (rs != 0 && rt != 0) {
+            tcg_gen_sub_tl(cpu_gpr[rd], cpu_gpr[rs], cpu_gpr[rt]);
+            tcg_gen_ext32s_tl(cpu_gpr[rd], cpu_gpr[rd]);
+        } else if (rs == 0 && rt != 0) {
+            tcg_gen_neg_tl(cpu_gpr[rd], cpu_gpr[rt]);
+            tcg_gen_ext32s_tl(cpu_gpr[rd], cpu_gpr[rd]);
+        } else if (rs != 0 && rt == 0) {
+            tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rs]);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rd], 0);
+        }
+        opn = "subu";
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DADD:
+        {
+            TCGv t0 = tcg_temp_local_new();
+            TCGv t1 = tcg_temp_new();
+            TCGv t2 = tcg_temp_new();
+            int l1 = gen_new_label();
+
+            gen_load_gpr(t1, rs);
+            gen_load_gpr(t2, rt);
+            tcg_gen_add_tl(t0, t1, t2);
+            tcg_gen_xor_tl(t1, t1, t2);
+            tcg_gen_not_tl(t1, t1);
+            tcg_gen_xor_tl(t2, t0, t2);
+            tcg_gen_and_tl(t1, t1, t2);
+            tcg_temp_free(t2);
+            tcg_gen_brcondi_tl(TCG_COND_GE, t1, 0, l1);
+            tcg_temp_free(t1);
+            /* operands of same sign, result different sign */
+            generate_exception(ctx, EXCP_OVERFLOW);
+            gen_set_label(l1);
+            gen_store_gpr(t0, rd);
+            tcg_temp_free(t0);
+        }
+        opn = "dadd";
+        break;
+    case OPC_DADDU:
+        if (rs != 0 && rt != 0) {
+            tcg_gen_add_tl(cpu_gpr[rd], cpu_gpr[rs], cpu_gpr[rt]);
+        } else if (rs == 0 && rt != 0) {
+            tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rt]);
+        } else if (rs != 0 && rt == 0) {
+            tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rs]);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rd], 0);
+        }
+        opn = "daddu";
+        break;
+    case OPC_DSUB:
+        {
+            TCGv t0 = tcg_temp_local_new();
+            TCGv t1 = tcg_temp_new();
+            TCGv t2 = tcg_temp_new();
+            int l1 = gen_new_label();
+
+            gen_load_gpr(t1, rs);
+            gen_load_gpr(t2, rt);
+            tcg_gen_sub_tl(t0, t1, t2);
+            tcg_gen_xor_tl(t2, t1, t2);
+            tcg_gen_xor_tl(t1, t0, t1);
+            tcg_gen_and_tl(t1, t1, t2);
+            tcg_temp_free(t2);
+            tcg_gen_brcondi_tl(TCG_COND_GE, t1, 0, l1);
+            tcg_temp_free(t1);
+            /* operands of different sign, first operand and result different sign */
+            generate_exception(ctx, EXCP_OVERFLOW);
+            gen_set_label(l1);
+            gen_store_gpr(t0, rd);
+            tcg_temp_free(t0);
+        }
+        opn = "dsub";
+        break;
+    case OPC_DSUBU:
+        if (rs != 0 && rt != 0) {
+            tcg_gen_sub_tl(cpu_gpr[rd], cpu_gpr[rs], cpu_gpr[rt]);
+        } else if (rs == 0 && rt != 0) {
+            tcg_gen_neg_tl(cpu_gpr[rd], cpu_gpr[rt]);
+        } else if (rs != 0 && rt == 0) {
+            tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rs]);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rd], 0);
+        }
+        opn = "dsubu";
+        break;
+#endif
+    case OPC_MUL:
+        if (likely(rs != 0 && rt != 0)) {
+            tcg_gen_mul_tl(cpu_gpr[rd], cpu_gpr[rs], cpu_gpr[rt]);
+            tcg_gen_ext32s_tl(cpu_gpr[rd], cpu_gpr[rd]);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rd], 0);
+        }
+        opn = "mul";
+        break;
+    }
+    MIPS_DEBUG("%s %s, %s, %s", opn, regnames[rd], regnames[rs], regnames[rt]);
+}
+
+/* Conditional move */
+static void gen_cond_move (CPUState *env, uint32_t opc, int rd, int rs, int rt)
+{
+    const char *opn = "cond move";
+    int l1;
+
+    if (rd == 0) {
+        /* If no destination, treat it as a NOP.
+           For add & sub, we must generate the overflow exception when needed. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+
+    l1 = gen_new_label();
+    switch (opc) {
+    case OPC_MOVN:
+        if (likely(rt != 0))
+            tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_gpr[rt], 0, l1);
+        else
+            tcg_gen_br(l1);
+        opn = "movn";
+        break;
+    case OPC_MOVZ:
+        if (likely(rt != 0))
+            tcg_gen_brcondi_tl(TCG_COND_NE, cpu_gpr[rt], 0, l1);
+        opn = "movz";
+        break;
+    }
+    if (rs != 0)
+        tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rs]);
+    else
+        tcg_gen_movi_tl(cpu_gpr[rd], 0);
+    gen_set_label(l1);
+
+    MIPS_DEBUG("%s %s, %s, %s", opn, regnames[rd], regnames[rs], regnames[rt]);
+}
+
+/* Logic */
+static void gen_logic (CPUState *env, uint32_t opc, int rd, int rs, int rt)
+{
+    const char *opn = "logic";
+
+    if (rd == 0) {
+        /* If no destination, treat it as a NOP. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+
+    switch (opc) {
+    case OPC_AND:
+        if (likely(rs != 0 && rt != 0)) {
+            tcg_gen_and_tl(cpu_gpr[rd], cpu_gpr[rs], cpu_gpr[rt]);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rd], 0);
+        }
+        opn = "and";
+        break;
+    case OPC_NOR:
+        if (rs != 0 && rt != 0) {
+            tcg_gen_nor_tl(cpu_gpr[rd], cpu_gpr[rs], cpu_gpr[rt]);
+        } else if (rs == 0 && rt != 0) {
+            tcg_gen_not_tl(cpu_gpr[rd], cpu_gpr[rt]);
+        } else if (rs != 0 && rt == 0) {
+            tcg_gen_not_tl(cpu_gpr[rd], cpu_gpr[rs]);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rd], ~((target_ulong)0));
+        }
+        opn = "nor";
+        break;
+    case OPC_OR:
+        if (likely(rs != 0 && rt != 0)) {
+            tcg_gen_or_tl(cpu_gpr[rd], cpu_gpr[rs], cpu_gpr[rt]);
+        } else if (rs == 0 && rt != 0) {
+            tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rt]);
+        } else if (rs != 0 && rt == 0) {
+            tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rs]);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rd], 0);
+        }
+        opn = "or";
+        break;
+    case OPC_XOR:
+        if (likely(rs != 0 && rt != 0)) {
+            tcg_gen_xor_tl(cpu_gpr[rd], cpu_gpr[rs], cpu_gpr[rt]);
+        } else if (rs == 0 && rt != 0) {
+            tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rt]);
+        } else if (rs != 0 && rt == 0) {
+            tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rs]);
+        } else {
+            tcg_gen_movi_tl(cpu_gpr[rd], 0);
+        }
+        opn = "xor";
+        break;
+    }
+    MIPS_DEBUG("%s %s, %s, %s", opn, regnames[rd], regnames[rs], regnames[rt]);
+}
+
+/* Set on lower than */
+static void gen_slt (CPUState *env, uint32_t opc, int rd, int rs, int rt)
+{
+    const char *opn = "slt";
+    TCGv t0, t1;
+
+    if (rd == 0) {
+        /* If no destination, treat it as a NOP. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+
+    t0 = tcg_temp_new();
+    t1 = tcg_temp_new();
+    gen_load_gpr(t0, rs);
+    gen_load_gpr(t1, rt);
+    switch (opc) {
+    case OPC_SLT:
+        gen_op_lt(cpu_gpr[rd], t0, t1);
+        opn = "slt";
+        break;
+    case OPC_SLTU:
+        gen_op_ltu(cpu_gpr[rd], t0, t1);
+        opn = "sltu";
+        break;
+    }
+    MIPS_DEBUG("%s %s, %s, %s", opn, regnames[rd], regnames[rs], regnames[rt]);
+    tcg_temp_free(t0);
+    tcg_temp_free(t1);
+}
+
+/* Shifts */
+static void gen_shift (CPUState *env, DisasContext *ctx, uint32_t opc,
+                       int rd, int rs, int rt)
+{
+    const char *opn = "shifts";
+    TCGv t0, t1;
+
+    if (rd == 0) {
+        /* If no destination, treat it as a NOP.
+           For add & sub, we must generate the overflow exception when needed. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+
+    t0 = tcg_temp_new();
+    t1 = tcg_temp_new();
+    gen_load_gpr(t0, rs);
+    gen_load_gpr(t1, rt);
+    switch (opc) {
+    case OPC_SLLV:
+        tcg_gen_andi_tl(t0, t0, 0x1f);
+        tcg_gen_shl_tl(t0, t1, t0);
+        tcg_gen_ext32s_tl(cpu_gpr[rd], t0);
+        opn = "sllv";
+        break;
+    case OPC_SRAV:
+        tcg_gen_ext32s_tl(t1, t1);
+        tcg_gen_andi_tl(t0, t0, 0x1f);
+        tcg_gen_sar_tl(cpu_gpr[rd], t1, t0);
+        opn = "srav";
+        break;
+    case OPC_SRLV:
+        switch ((ctx->opcode >> 6) & 0x1f) {
+        case 0:
+            tcg_gen_ext32u_tl(t1, t1);
+            tcg_gen_andi_tl(t0, t0, 0x1f);
+            tcg_gen_shr_tl(t0, t1, t0);
+            tcg_gen_ext32s_tl(cpu_gpr[rd], t0);
+            opn = "srlv";
+            break;
+        case 1:
+            /* rotrv is decoded as srlv on non-R2 CPUs */
+            if (env->insn_flags & ISA_MIPS32R2) {
+                TCGv_i32 t2 = tcg_temp_new_i32();
+                TCGv_i32 t3 = tcg_temp_new_i32();
+
+                tcg_gen_trunc_tl_i32(t2, t0);
+                tcg_gen_trunc_tl_i32(t3, t1);
+                tcg_gen_andi_i32(t2, t2, 0x1f);
+                tcg_gen_rotr_i32(t2, t3, t2);
+                tcg_gen_ext_i32_tl(cpu_gpr[rd], t2);
+                tcg_temp_free_i32(t2);
+                tcg_temp_free_i32(t3);
+                opn = "rotrv";
+            } else {
+                tcg_gen_ext32u_tl(t1, t1);
+                tcg_gen_andi_tl(t0, t0, 0x1f);
+                tcg_gen_shr_tl(t0, t1, t0);
+                tcg_gen_ext32s_tl(cpu_gpr[rd], t0);
+                opn = "srlv";
+            }
+            break;
+        default:
+            MIPS_INVAL("invalid srlv flag");
+            generate_exception(ctx, EXCP_RI);
+            break;
+        }
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DSLLV:
+        tcg_gen_andi_tl(t0, t0, 0x3f);
+        tcg_gen_shl_tl(cpu_gpr[rd], t1, t0);
+        opn = "dsllv";
+        break;
+    case OPC_DSRAV:
+        tcg_gen_andi_tl(t0, t0, 0x3f);
+        tcg_gen_sar_tl(cpu_gpr[rd], t1, t0);
+        opn = "dsrav";
+        break;
+    case OPC_DSRLV:
+        switch ((ctx->opcode >> 6) & 0x1f) {
+        case 0:
+            tcg_gen_andi_tl(t0, t0, 0x3f);
+            tcg_gen_shr_tl(cpu_gpr[rd], t1, t0);
+            opn = "dsrlv";
+            break;
+        case 1:
+            /* drotrv is decoded as dsrlv on non-R2 CPUs */
+            if (env->insn_flags & ISA_MIPS32R2) {
+                tcg_gen_andi_tl(t0, t0, 0x3f);
+                tcg_gen_rotr_tl(cpu_gpr[rd], t1, t0);
+                opn = "drotrv";
+            } else {
+                tcg_gen_andi_tl(t0, t0, 0x3f);
+                tcg_gen_shr_tl(t0, t1, t0);
+                opn = "dsrlv";
+            }
+            break;
+        default:
+            MIPS_INVAL("invalid dsrlv flag");
+            generate_exception(ctx, EXCP_RI);
+            break;
+        }
+        break;
+#endif
+    }
+    MIPS_DEBUG("%s %s, %s, %s", opn, regnames[rd], regnames[rs], regnames[rt]);
+    tcg_temp_free(t0);
+    tcg_temp_free(t1);
+}
+
+/* Arithmetic on HI/LO registers */
+static void gen_HILO (DisasContext *ctx, uint32_t opc, int reg)
+{
+    const char *opn = "hilo";
+
+    if (reg == 0 && (opc == OPC_MFHI || opc == OPC_MFLO)) {
+        /* Treat as NOP. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+    switch (opc) {
+    case OPC_MFHI:
+        tcg_gen_mov_tl(cpu_gpr[reg], cpu_HI[0]);
+        opn = "mfhi";
+        break;
+    case OPC_MFLO:
+        tcg_gen_mov_tl(cpu_gpr[reg], cpu_LO[0]);
+        opn = "mflo";
+        break;
+    case OPC_MTHI:
+        if (reg != 0)
+            tcg_gen_mov_tl(cpu_HI[0], cpu_gpr[reg]);
+        else
+            tcg_gen_movi_tl(cpu_HI[0], 0);
+        opn = "mthi";
+        break;
+    case OPC_MTLO:
+        if (reg != 0)
+            tcg_gen_mov_tl(cpu_LO[0], cpu_gpr[reg]);
+        else
+            tcg_gen_movi_tl(cpu_LO[0], 0);
+        opn = "mtlo";
+        break;
+    }
+    MIPS_DEBUG("%s %s", opn, regnames[reg]);
+}
+
+static void gen_muldiv (DisasContext *ctx, uint32_t opc,
+                        int rs, int rt)
+{
+    const char *opn = "mul/div";
+    TCGv t0, t1;
+
+    switch (opc) {
+    case OPC_DIV:
+    case OPC_DIVU:
+#if defined(TARGET_MIPS64)
+    case OPC_DDIV:
+    case OPC_DDIVU:
+#endif
+        t0 = tcg_temp_local_new();
+        t1 = tcg_temp_local_new();
+        break;
+    default:
+        t0 = tcg_temp_new();
+        t1 = tcg_temp_new();
+        break;
+    }
+
+    gen_load_gpr(t0, rs);
+    gen_load_gpr(t1, rt);
+    switch (opc) {
+    case OPC_DIV:
+        {
+            int l1 = gen_new_label();
+            int l2 = gen_new_label();
+
+            tcg_gen_ext32s_tl(t0, t0);
+            tcg_gen_ext32s_tl(t1, t1);
+            tcg_gen_brcondi_tl(TCG_COND_EQ, t1, 0, l1);
+            tcg_gen_brcondi_tl(TCG_COND_NE, t0, INT_MIN, l2);
+            tcg_gen_brcondi_tl(TCG_COND_NE, t1, -1, l2);
+
+            tcg_gen_mov_tl(cpu_LO[0], t0);
+            tcg_gen_movi_tl(cpu_HI[0], 0);
+            tcg_gen_br(l1);
+            gen_set_label(l2);
+            tcg_gen_div_tl(cpu_LO[0], t0, t1);
+            tcg_gen_rem_tl(cpu_HI[0], t0, t1);
+            tcg_gen_ext32s_tl(cpu_LO[0], cpu_LO[0]);
+            tcg_gen_ext32s_tl(cpu_HI[0], cpu_HI[0]);
+            gen_set_label(l1);
+        }
+        opn = "div";
+        break;
+    case OPC_DIVU:
+        {
+            int l1 = gen_new_label();
+
+            tcg_gen_ext32u_tl(t0, t0);
+            tcg_gen_ext32u_tl(t1, t1);
+            tcg_gen_brcondi_tl(TCG_COND_EQ, t1, 0, l1);
+            tcg_gen_divu_tl(cpu_LO[0], t0, t1);
+            tcg_gen_remu_tl(cpu_HI[0], t0, t1);
+            tcg_gen_ext32s_tl(cpu_LO[0], cpu_LO[0]);
+            tcg_gen_ext32s_tl(cpu_HI[0], cpu_HI[0]);
+            gen_set_label(l1);
+        }
+        opn = "divu";
+        break;
+    case OPC_MULT:
+        {
+            TCGv_i64 t2 = tcg_temp_new_i64();
+            TCGv_i64 t3 = tcg_temp_new_i64();
+
+            tcg_gen_ext_tl_i64(t2, t0);
+            tcg_gen_ext_tl_i64(t3, t1);
+            tcg_gen_mul_i64(t2, t2, t3);
+            tcg_temp_free_i64(t3);
+            tcg_gen_trunc_i64_tl(t0, t2);
+            tcg_gen_shri_i64(t2, t2, 32);
+            tcg_gen_trunc_i64_tl(t1, t2);
+            tcg_temp_free_i64(t2);
+            tcg_gen_ext32s_tl(cpu_LO[0], t0);
+            tcg_gen_ext32s_tl(cpu_HI[0], t1);
+        }
+        opn = "mult";
+        break;
+    case OPC_MULTU:
+        {
+            TCGv_i64 t2 = tcg_temp_new_i64();
+            TCGv_i64 t3 = tcg_temp_new_i64();
+
+            tcg_gen_ext32u_tl(t0, t0);
+            tcg_gen_ext32u_tl(t1, t1);
+            tcg_gen_extu_tl_i64(t2, t0);
+            tcg_gen_extu_tl_i64(t3, t1);
+            tcg_gen_mul_i64(t2, t2, t3);
+            tcg_temp_free_i64(t3);
+            tcg_gen_trunc_i64_tl(t0, t2);
+            tcg_gen_shri_i64(t2, t2, 32);
+            tcg_gen_trunc_i64_tl(t1, t2);
+            tcg_temp_free_i64(t2);
+            tcg_gen_ext32s_tl(cpu_LO[0], t0);
+            tcg_gen_ext32s_tl(cpu_HI[0], t1);
+        }
+        opn = "multu";
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DDIV:
+        {
+            int l1 = gen_new_label();
+            int l2 = gen_new_label();
+
+            tcg_gen_brcondi_tl(TCG_COND_EQ, t1, 0, l1);
+            tcg_gen_brcondi_tl(TCG_COND_NE, t0, -1LL << 63, l2);
+            tcg_gen_brcondi_tl(TCG_COND_NE, t1, -1LL, l2);
+            tcg_gen_mov_tl(cpu_LO[0], t0);
+            tcg_gen_movi_tl(cpu_HI[0], 0);
+            tcg_gen_br(l1);
+            gen_set_label(l2);
+            tcg_gen_div_i64(cpu_LO[0], t0, t1);
+            tcg_gen_rem_i64(cpu_HI[0], t0, t1);
+            gen_set_label(l1);
+        }
+        opn = "ddiv";
+        break;
+    case OPC_DDIVU:
+        {
+            int l1 = gen_new_label();
+
+            tcg_gen_brcondi_tl(TCG_COND_EQ, t1, 0, l1);
+            tcg_gen_divu_i64(cpu_LO[0], t0, t1);
+            tcg_gen_remu_i64(cpu_HI[0], t0, t1);
+            gen_set_label(l1);
+        }
+        opn = "ddivu";
+        break;
+    case OPC_DMULT:
+        gen_helper_dmult(t0, t1);
+        opn = "dmult";
+        break;
+    case OPC_DMULTU:
+        gen_helper_dmultu(t0, t1);
+        opn = "dmultu";
+        break;
+#endif
+    case OPC_MADD:
+        {
+            TCGv_i64 t2 = tcg_temp_new_i64();
+            TCGv_i64 t3 = tcg_temp_new_i64();
+
+            tcg_gen_ext_tl_i64(t2, t0);
+            tcg_gen_ext_tl_i64(t3, t1);
+            tcg_gen_mul_i64(t2, t2, t3);
+            tcg_gen_concat_tl_i64(t3, cpu_LO[0], cpu_HI[0]);
+            tcg_gen_add_i64(t2, t2, t3);
+            tcg_temp_free_i64(t3);
+            tcg_gen_trunc_i64_tl(t0, t2);
+            tcg_gen_shri_i64(t2, t2, 32);
+            tcg_gen_trunc_i64_tl(t1, t2);
+            tcg_temp_free_i64(t2);
+            tcg_gen_ext32s_tl(cpu_LO[0], t0);
+            tcg_gen_ext32s_tl(cpu_HI[0], t1);
+        }
+        opn = "madd";
+        break;
+    case OPC_MADDU:
+       {
+            TCGv_i64 t2 = tcg_temp_new_i64();
+            TCGv_i64 t3 = tcg_temp_new_i64();
+
+            tcg_gen_ext32u_tl(t0, t0);
+            tcg_gen_ext32u_tl(t1, t1);
+            tcg_gen_extu_tl_i64(t2, t0);
+            tcg_gen_extu_tl_i64(t3, t1);
+            tcg_gen_mul_i64(t2, t2, t3);
+            tcg_gen_concat_tl_i64(t3, cpu_LO[0], cpu_HI[0]);
+            tcg_gen_add_i64(t2, t2, t3);
+            tcg_temp_free_i64(t3);
+            tcg_gen_trunc_i64_tl(t0, t2);
+            tcg_gen_shri_i64(t2, t2, 32);
+            tcg_gen_trunc_i64_tl(t1, t2);
+            tcg_temp_free_i64(t2);
+            tcg_gen_ext32s_tl(cpu_LO[0], t0);
+            tcg_gen_ext32s_tl(cpu_HI[0], t1);
+        }
+        opn = "maddu";
+        break;
+    case OPC_MSUB:
+        {
+            TCGv_i64 t2 = tcg_temp_new_i64();
+            TCGv_i64 t3 = tcg_temp_new_i64();
+
+            tcg_gen_ext_tl_i64(t2, t0);
+            tcg_gen_ext_tl_i64(t3, t1);
+            tcg_gen_mul_i64(t2, t2, t3);
+            tcg_gen_concat_tl_i64(t3, cpu_LO[0], cpu_HI[0]);
+            tcg_gen_sub_i64(t2, t3, t2);
+            tcg_temp_free_i64(t3);
+            tcg_gen_trunc_i64_tl(t0, t2);
+            tcg_gen_shri_i64(t2, t2, 32);
+            tcg_gen_trunc_i64_tl(t1, t2);
+            tcg_temp_free_i64(t2);
+            tcg_gen_ext32s_tl(cpu_LO[0], t0);
+            tcg_gen_ext32s_tl(cpu_HI[0], t1);
+        }
+        opn = "msub";
+        break;
+    case OPC_MSUBU:
+        {
+            TCGv_i64 t2 = tcg_temp_new_i64();
+            TCGv_i64 t3 = tcg_temp_new_i64();
+
+            tcg_gen_ext32u_tl(t0, t0);
+            tcg_gen_ext32u_tl(t1, t1);
+            tcg_gen_extu_tl_i64(t2, t0);
+            tcg_gen_extu_tl_i64(t3, t1);
+            tcg_gen_mul_i64(t2, t2, t3);
+            tcg_gen_concat_tl_i64(t3, cpu_LO[0], cpu_HI[0]);
+            tcg_gen_sub_i64(t2, t3, t2);
+            tcg_temp_free_i64(t3);
+            tcg_gen_trunc_i64_tl(t0, t2);
+            tcg_gen_shri_i64(t2, t2, 32);
+            tcg_gen_trunc_i64_tl(t1, t2);
+            tcg_temp_free_i64(t2);
+            tcg_gen_ext32s_tl(cpu_LO[0], t0);
+            tcg_gen_ext32s_tl(cpu_HI[0], t1);
+        }
+        opn = "msubu";
+        break;
+    default:
+        MIPS_INVAL(opn);
+        generate_exception(ctx, EXCP_RI);
+        goto out;
+    }
+    MIPS_DEBUG("%s %s %s", opn, regnames[rs], regnames[rt]);
+ out:
+    tcg_temp_free(t0);
+    tcg_temp_free(t1);
+}
+
+static void gen_mul_vr54xx (DisasContext *ctx, uint32_t opc,
+                            int rd, int rs, int rt)
+{
+    const char *opn = "mul vr54xx";
+    TCGv t0 = tcg_temp_new();
+    TCGv t1 = tcg_temp_new();
+
+    gen_load_gpr(t0, rs);
+    gen_load_gpr(t1, rt);
+
+    switch (opc) {
+    case OPC_VR54XX_MULS:
+        gen_helper_muls(t0, t0, t1);
+        opn = "muls";
+        break;
+    case OPC_VR54XX_MULSU:
+        gen_helper_mulsu(t0, t0, t1);
+        opn = "mulsu";
+        break;
+    case OPC_VR54XX_MACC:
+        gen_helper_macc(t0, t0, t1);
+        opn = "macc";
+        break;
+    case OPC_VR54XX_MACCU:
+        gen_helper_maccu(t0, t0, t1);
+        opn = "maccu";
+        break;
+    case OPC_VR54XX_MSAC:
+        gen_helper_msac(t0, t0, t1);
+        opn = "msac";
+        break;
+    case OPC_VR54XX_MSACU:
+        gen_helper_msacu(t0, t0, t1);
+        opn = "msacu";
+        break;
+    case OPC_VR54XX_MULHI:
+        gen_helper_mulhi(t0, t0, t1);
+        opn = "mulhi";
+        break;
+    case OPC_VR54XX_MULHIU:
+        gen_helper_mulhiu(t0, t0, t1);
+        opn = "mulhiu";
+        break;
+    case OPC_VR54XX_MULSHI:
+        gen_helper_mulshi(t0, t0, t1);
+        opn = "mulshi";
+        break;
+    case OPC_VR54XX_MULSHIU:
+        gen_helper_mulshiu(t0, t0, t1);
+        opn = "mulshiu";
+        break;
+    case OPC_VR54XX_MACCHI:
+        gen_helper_macchi(t0, t0, t1);
+        opn = "macchi";
+        break;
+    case OPC_VR54XX_MACCHIU:
+        gen_helper_macchiu(t0, t0, t1);
+        opn = "macchiu";
+        break;
+    case OPC_VR54XX_MSACHI:
+        gen_helper_msachi(t0, t0, t1);
+        opn = "msachi";
+        break;
+    case OPC_VR54XX_MSACHIU:
+        gen_helper_msachiu(t0, t0, t1);
+        opn = "msachiu";
+        break;
+    default:
+        MIPS_INVAL("mul vr54xx");
+        generate_exception(ctx, EXCP_RI);
+        goto out;
+    }
+    gen_store_gpr(t0, rd);
+    MIPS_DEBUG("%s %s, %s, %s", opn, regnames[rd], regnames[rs], regnames[rt]);
+
+ out:
+    tcg_temp_free(t0);
+    tcg_temp_free(t1);
+}
+
+static void gen_cl (DisasContext *ctx, uint32_t opc,
+                    int rd, int rs)
+{
+    const char *opn = "CLx";
+    TCGv t0;
+
+    if (rd == 0) {
+        /* Treat as NOP. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+    t0 = tcg_temp_new();
+    gen_load_gpr(t0, rs);
+    switch (opc) {
+    case OPC_CLO:
+        gen_helper_clo(cpu_gpr[rd], t0);
+        opn = "clo";
+        break;
+    case OPC_CLZ:
+        gen_helper_clz(cpu_gpr[rd], t0);
+        opn = "clz";
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DCLO:
+        gen_helper_dclo(cpu_gpr[rd], t0);
+        opn = "dclo";
+        break;
+    case OPC_DCLZ:
+        gen_helper_dclz(cpu_gpr[rd], t0);
+        opn = "dclz";
+        break;
+#endif
+    }
+    MIPS_DEBUG("%s %s, %s", opn, regnames[rd], regnames[rs]);
+    tcg_temp_free(t0);
+}
+
+/* Traps */
+static void gen_trap (DisasContext *ctx, uint32_t opc,
+                      int rs, int rt, int16_t imm)
+{
+    int cond;
+    TCGv t0 = tcg_temp_new();
+    TCGv t1 = tcg_temp_new();
+
+    cond = 0;
+    /* Load needed operands */
+    switch (opc) {
+    case OPC_TEQ:
+    case OPC_TGE:
+    case OPC_TGEU:
+    case OPC_TLT:
+    case OPC_TLTU:
+    case OPC_TNE:
+        /* Compare two registers */
+        if (rs != rt) {
+            gen_load_gpr(t0, rs);
+            gen_load_gpr(t1, rt);
+            cond = 1;
+        }
+        break;
+    case OPC_TEQI:
+    case OPC_TGEI:
+    case OPC_TGEIU:
+    case OPC_TLTI:
+    case OPC_TLTIU:
+    case OPC_TNEI:
+        /* Compare register to immediate */
+        if (rs != 0 || imm != 0) {
+            gen_load_gpr(t0, rs);
+            tcg_gen_movi_tl(t1, (int32_t)imm);
+            cond = 1;
+        }
+        break;
+    }
+    if (cond == 0) {
+        switch (opc) {
+        case OPC_TEQ:   /* rs == rs */
+        case OPC_TEQI:  /* r0 == 0  */
+        case OPC_TGE:   /* rs >= rs */
+        case OPC_TGEI:  /* r0 >= 0  */
+        case OPC_TGEU:  /* rs >= rs unsigned */
+        case OPC_TGEIU: /* r0 >= 0  unsigned */
+            /* Always trap */
+            generate_exception(ctx, EXCP_TRAP);
+            break;
+        case OPC_TLT:   /* rs < rs           */
+        case OPC_TLTI:  /* r0 < 0            */
+        case OPC_TLTU:  /* rs < rs unsigned  */
+        case OPC_TLTIU: /* r0 < 0  unsigned  */
+        case OPC_TNE:   /* rs != rs          */
+        case OPC_TNEI:  /* r0 != 0           */
+            /* Never trap: treat as NOP. */
+            break;
+        }
+    } else {
+        int l1 = gen_new_label();
+
+        switch (opc) {
+        case OPC_TEQ:
+        case OPC_TEQI:
+            tcg_gen_brcond_tl(TCG_COND_NE, t0, t1, l1);
+            break;
+        case OPC_TGE:
+        case OPC_TGEI:
+            tcg_gen_brcond_tl(TCG_COND_LT, t0, t1, l1);
+            break;
+        case OPC_TGEU:
+        case OPC_TGEIU:
+            tcg_gen_brcond_tl(TCG_COND_LTU, t0, t1, l1);
+            break;
+        case OPC_TLT:
+        case OPC_TLTI:
+            tcg_gen_brcond_tl(TCG_COND_GE, t0, t1, l1);
+            break;
+        case OPC_TLTU:
+        case OPC_TLTIU:
+            tcg_gen_brcond_tl(TCG_COND_GEU, t0, t1, l1);
+            break;
+        case OPC_TNE:
+        case OPC_TNEI:
+            tcg_gen_brcond_tl(TCG_COND_EQ, t0, t1, l1);
+            break;
+        }
+        generate_exception(ctx, EXCP_TRAP);
+        gen_set_label(l1);
+    }
+    tcg_temp_free(t0);
+    tcg_temp_free(t1);
+}
+
+static inline void gen_goto_tb(DisasContext *ctx, int n, target_ulong dest)
+{
+    TranslationBlock *tb;
+    tb = ctx->tb;
+    if ((tb->pc & TARGET_PAGE_MASK) == (dest & TARGET_PAGE_MASK) &&
+        likely(!ctx->singlestep_enabled)) {
+        tcg_gen_goto_tb(n);
+        gen_save_pc(dest);
+        tcg_gen_exit_tb((long)tb + n);
+    } else {
+        gen_save_pc(dest);
+        if (ctx->singlestep_enabled) {
+            save_cpu_state(ctx, 0);
+            gen_helper_0i(raise_exception, EXCP_DEBUG);
+        }
+        tcg_gen_exit_tb(0);
+    }
+}
+
+/* Branches (before delay slot) */
+static void gen_compute_branch (DisasContext *ctx, uint32_t opc,
+                                int rs, int rt, int32_t offset)
+{
+    target_ulong btgt = -1;
+    int blink = 0;
+    int bcond_compute = 0;
+    TCGv t0 = tcg_temp_new();
+    TCGv t1 = tcg_temp_new();
+
+    if (ctx->hflags & MIPS_HFLAG_BMASK) {
+#ifdef MIPS_DEBUG_DISAS
+        LOG_DISAS("Branch in delay slot at PC 0x" TARGET_FMT_lx "\n", ctx->pc);
+#endif
+        generate_exception(ctx, EXCP_RI);
+        goto out;
+    }
+
+    /* Load needed operands */
+    switch (opc) {
+    case OPC_BEQ:
+    case OPC_BEQL:
+    case OPC_BNE:
+    case OPC_BNEL:
+        /* Compare two registers */
+        if (rs != rt) {
+            gen_load_gpr(t0, rs);
+            gen_load_gpr(t1, rt);
+            bcond_compute = 1;
+        }
+        btgt = ctx->pc + 4 + offset;
+        break;
+    case OPC_BGEZ:
+    case OPC_BGEZAL:
+    case OPC_BGEZALL:
+    case OPC_BGEZL:
+    case OPC_BGTZ:
+    case OPC_BGTZL:
+    case OPC_BLEZ:
+    case OPC_BLEZL:
+    case OPC_BLTZ:
+    case OPC_BLTZAL:
+    case OPC_BLTZALL:
+    case OPC_BLTZL:
+        /* Compare to zero */
+        if (rs != 0) {
+            gen_load_gpr(t0, rs);
+            bcond_compute = 1;
+        }
+        btgt = ctx->pc + 4 + offset;
+        break;
+    case OPC_J:
+    case OPC_JAL:
+        /* Jump to immediate */
+        btgt = ((ctx->pc + 4) & (int32_t)0xF0000000) | (uint32_t)offset;
+        break;
+    case OPC_JR:
+    case OPC_JALR:
+        /* Jump to register */
+        if (offset != 0 && offset != 16) {
+            /* Hint = 0 is JR/JALR, hint 16 is JR.HB/JALR.HB, the
+               others are reserved. */
+            MIPS_INVAL("jump hint");
+            generate_exception(ctx, EXCP_RI);
+            goto out;
+        }
+        gen_load_gpr(btarget, rs);
+        break;
+    default:
+        MIPS_INVAL("branch/jump");
+        generate_exception(ctx, EXCP_RI);
+        goto out;
+    }
+    if (bcond_compute == 0) {
+        /* No condition to be computed */
+        switch (opc) {
+        case OPC_BEQ:     /* rx == rx        */
+        case OPC_BEQL:    /* rx == rx likely */
+        case OPC_BGEZ:    /* 0 >= 0          */
+        case OPC_BGEZL:   /* 0 >= 0 likely   */
+        case OPC_BLEZ:    /* 0 <= 0          */
+        case OPC_BLEZL:   /* 0 <= 0 likely   */
+            /* Always take */
+            ctx->hflags |= MIPS_HFLAG_B;
+            MIPS_DEBUG("balways");
+            break;
+        case OPC_BGEZAL:  /* 0 >= 0          */
+        case OPC_BGEZALL: /* 0 >= 0 likely   */
+            /* Always take and link */
+            blink = 31;
+            ctx->hflags |= MIPS_HFLAG_B;
+            MIPS_DEBUG("balways and link");
+            break;
+        case OPC_BNE:     /* rx != rx        */
+        case OPC_BGTZ:    /* 0 > 0           */
+        case OPC_BLTZ:    /* 0 < 0           */
+            /* Treat as NOP. */
+            MIPS_DEBUG("bnever (NOP)");
+            goto out;
+        case OPC_BLTZAL:  /* 0 < 0           */
+            tcg_gen_movi_tl(cpu_gpr[31], ctx->pc + 8);
+            MIPS_DEBUG("bnever and link");
+            goto out;
+        case OPC_BLTZALL: /* 0 < 0 likely */
+            tcg_gen_movi_tl(cpu_gpr[31], ctx->pc + 8);
+            /* Skip the instruction in the delay slot */
+            MIPS_DEBUG("bnever, link and skip");
+            ctx->pc += 4;
+            goto out;
+        case OPC_BNEL:    /* rx != rx likely */
+        case OPC_BGTZL:   /* 0 > 0 likely */
+        case OPC_BLTZL:   /* 0 < 0 likely */
+            /* Skip the instruction in the delay slot */
+            MIPS_DEBUG("bnever and skip");
+            ctx->pc += 4;
+            goto out;
+        case OPC_J:
+            ctx->hflags |= MIPS_HFLAG_B;
+            MIPS_DEBUG("j " TARGET_FMT_lx, btgt);
+            break;
+        case OPC_JAL:
+            blink = 31;
+            ctx->hflags |= MIPS_HFLAG_B;
+            MIPS_DEBUG("jal " TARGET_FMT_lx, btgt);
+            break;
+        case OPC_JR:
+            ctx->hflags |= MIPS_HFLAG_BR;
+            MIPS_DEBUG("jr %s", regnames[rs]);
+            break;
+        case OPC_JALR:
+            blink = rt;
+            ctx->hflags |= MIPS_HFLAG_BR;
+            MIPS_DEBUG("jalr %s, %s", regnames[rt], regnames[rs]);
+            break;
+        default:
+            MIPS_INVAL("branch/jump");
+            generate_exception(ctx, EXCP_RI);
+            goto out;
+        }
+    } else {
+        switch (opc) {
+        case OPC_BEQ:
+            gen_op_eq(bcond, t0, t1);
+            MIPS_DEBUG("beq %s, %s, " TARGET_FMT_lx,
+                       regnames[rs], regnames[rt], btgt);
+            goto not_likely;
+        case OPC_BEQL:
+            gen_op_eq(bcond, t0, t1);
+            MIPS_DEBUG("beql %s, %s, " TARGET_FMT_lx,
+                       regnames[rs], regnames[rt], btgt);
+            goto likely;
+        case OPC_BNE:
+            gen_op_ne(bcond, t0, t1);
+            MIPS_DEBUG("bne %s, %s, " TARGET_FMT_lx,
+                       regnames[rs], regnames[rt], btgt);
+            goto not_likely;
+        case OPC_BNEL:
+            gen_op_ne(bcond, t0, t1);
+            MIPS_DEBUG("bnel %s, %s, " TARGET_FMT_lx,
+                       regnames[rs], regnames[rt], btgt);
+            goto likely;
+        case OPC_BGEZ:
+            gen_op_gez(bcond, t0);
+            MIPS_DEBUG("bgez %s, " TARGET_FMT_lx, regnames[rs], btgt);
+            goto not_likely;
+        case OPC_BGEZL:
+            gen_op_gez(bcond, t0);
+            MIPS_DEBUG("bgezl %s, " TARGET_FMT_lx, regnames[rs], btgt);
+            goto likely;
+        case OPC_BGEZAL:
+            gen_op_gez(bcond, t0);
+            MIPS_DEBUG("bgezal %s, " TARGET_FMT_lx, regnames[rs], btgt);
+            blink = 31;
+            goto not_likely;
+        case OPC_BGEZALL:
+            gen_op_gez(bcond, t0);
+            blink = 31;
+            MIPS_DEBUG("bgezall %s, " TARGET_FMT_lx, regnames[rs], btgt);
+            goto likely;
+        case OPC_BGTZ:
+            gen_op_gtz(bcond, t0);
+            MIPS_DEBUG("bgtz %s, " TARGET_FMT_lx, regnames[rs], btgt);
+            goto not_likely;
+        case OPC_BGTZL:
+            gen_op_gtz(bcond, t0);
+            MIPS_DEBUG("bgtzl %s, " TARGET_FMT_lx, regnames[rs], btgt);
+            goto likely;
+        case OPC_BLEZ:
+            gen_op_lez(bcond, t0);
+            MIPS_DEBUG("blez %s, " TARGET_FMT_lx, regnames[rs], btgt);
+            goto not_likely;
+        case OPC_BLEZL:
+            gen_op_lez(bcond, t0);
+            MIPS_DEBUG("blezl %s, " TARGET_FMT_lx, regnames[rs], btgt);
+            goto likely;
+        case OPC_BLTZ:
+            gen_op_ltz(bcond, t0);
+            MIPS_DEBUG("bltz %s, " TARGET_FMT_lx, regnames[rs], btgt);
+            goto not_likely;
+        case OPC_BLTZL:
+            gen_op_ltz(bcond, t0);
+            MIPS_DEBUG("bltzl %s, " TARGET_FMT_lx, regnames[rs], btgt);
+            goto likely;
+        case OPC_BLTZAL:
+            gen_op_ltz(bcond, t0);
+            blink = 31;
+            MIPS_DEBUG("bltzal %s, " TARGET_FMT_lx, regnames[rs], btgt);
+        not_likely:
+            ctx->hflags |= MIPS_HFLAG_BC;
+            break;
+        case OPC_BLTZALL:
+            gen_op_ltz(bcond, t0);
+            blink = 31;
+            MIPS_DEBUG("bltzall %s, " TARGET_FMT_lx, regnames[rs], btgt);
+        likely:
+            ctx->hflags |= MIPS_HFLAG_BL;
+            break;
+        default:
+            MIPS_INVAL("conditional branch/jump");
+            generate_exception(ctx, EXCP_RI);
+            goto out;
+        }
+    }
+    MIPS_DEBUG("enter ds: link %d cond %02x target " TARGET_FMT_lx,
+               blink, ctx->hflags, btgt);
+
+    ctx->btarget = btgt;
+    if (blink > 0) {
+        tcg_gen_movi_tl(cpu_gpr[blink], ctx->pc + 8);
+    }
+
+ out:
+    tcg_temp_free(t0);
+    tcg_temp_free(t1);
+}
+
+/* special3 bitfield operations */
+static void gen_bitops (DisasContext *ctx, uint32_t opc, int rt,
+                        int rs, int lsb, int msb)
+{
+    TCGv t0 = tcg_temp_new();
+    TCGv t1 = tcg_temp_new();
+    target_ulong mask;
+
+    gen_load_gpr(t1, rs);
+    switch (opc) {
+    case OPC_EXT:
+        if (lsb + msb > 31)
+            goto fail;
+        tcg_gen_shri_tl(t0, t1, lsb);
+        if (msb != 31) {
+            tcg_gen_andi_tl(t0, t0, (1 << (msb + 1)) - 1);
+        } else {
+            tcg_gen_ext32s_tl(t0, t0);
+        }
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DEXTM:
+        tcg_gen_shri_tl(t0, t1, lsb);
+        if (msb != 31) {
+            tcg_gen_andi_tl(t0, t0, (1ULL << (msb + 1 + 32)) - 1);
+        }
+        break;
+    case OPC_DEXTU:
+        tcg_gen_shri_tl(t0, t1, lsb + 32);
+        tcg_gen_andi_tl(t0, t0, (1ULL << (msb + 1)) - 1);
+        break;
+    case OPC_DEXT:
+        tcg_gen_shri_tl(t0, t1, lsb);
+        tcg_gen_andi_tl(t0, t0, (1ULL << (msb + 1)) - 1);
+        break;
+#endif
+    case OPC_INS:
+        if (lsb > msb)
+            goto fail;
+        mask = ((msb - lsb + 1 < 32) ? ((1 << (msb - lsb + 1)) - 1) : ~0) << lsb;
+        gen_load_gpr(t0, rt);
+        tcg_gen_andi_tl(t0, t0, ~mask);
+        tcg_gen_shli_tl(t1, t1, lsb);
+        tcg_gen_andi_tl(t1, t1, mask);
+        tcg_gen_or_tl(t0, t0, t1);
+        tcg_gen_ext32s_tl(t0, t0);
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DINSM:
+        if (lsb > msb)
+            goto fail;
+        mask = ((msb - lsb + 1 + 32 < 64) ? ((1ULL << (msb - lsb + 1 + 32)) - 1) : ~0ULL) << lsb;
+        gen_load_gpr(t0, rt);
+        tcg_gen_andi_tl(t0, t0, ~mask);
+        tcg_gen_shli_tl(t1, t1, lsb);
+        tcg_gen_andi_tl(t1, t1, mask);
+        tcg_gen_or_tl(t0, t0, t1);
+        break;
+    case OPC_DINSU:
+        if (lsb > msb)
+            goto fail;
+        mask = ((1ULL << (msb - lsb + 1)) - 1) << (lsb + 32);
+        gen_load_gpr(t0, rt);
+        tcg_gen_andi_tl(t0, t0, ~mask);
+        tcg_gen_shli_tl(t1, t1, lsb + 32);
+        tcg_gen_andi_tl(t1, t1, mask);
+        tcg_gen_or_tl(t0, t0, t1);
+        break;
+    case OPC_DINS:
+        if (lsb > msb)
+            goto fail;
+        gen_load_gpr(t0, rt);
+        mask = ((1ULL << (msb - lsb + 1)) - 1) << lsb;
+        gen_load_gpr(t0, rt);
+        tcg_gen_andi_tl(t0, t0, ~mask);
+        tcg_gen_shli_tl(t1, t1, lsb);
+        tcg_gen_andi_tl(t1, t1, mask);
+        tcg_gen_or_tl(t0, t0, t1);
+        break;
+#endif
+    default:
+fail:
+        MIPS_INVAL("bitops");
+        generate_exception(ctx, EXCP_RI);
+        tcg_temp_free(t0);
+        tcg_temp_free(t1);
+        return;
+    }
+    gen_store_gpr(t0, rt);
+    tcg_temp_free(t0);
+    tcg_temp_free(t1);
+}
+
+static void gen_bshfl (DisasContext *ctx, uint32_t op2, int rt, int rd)
+{
+    TCGv t0;
+
+    if (rd == 0) {
+        /* If no destination, treat it as a NOP. */
+        MIPS_DEBUG("NOP");
+        return;
+    }
+
+    t0 = tcg_temp_new();
+    gen_load_gpr(t0, rt);
+    switch (op2) {
+    case OPC_WSBH:
+        {
+            TCGv t1 = tcg_temp_new();
+
+            tcg_gen_shri_tl(t1, t0, 8);
+            tcg_gen_andi_tl(t1, t1, 0x00FF00FF);
+            tcg_gen_shli_tl(t0, t0, 8);
+            tcg_gen_andi_tl(t0, t0, ~0x00FF00FF);
+            tcg_gen_or_tl(t0, t0, t1);
+            tcg_temp_free(t1);
+            tcg_gen_ext32s_tl(cpu_gpr[rd], t0);
+        }
+        break;
+    case OPC_SEB:
+        tcg_gen_ext8s_tl(cpu_gpr[rd], t0);
+        break;
+    case OPC_SEH:
+        tcg_gen_ext16s_tl(cpu_gpr[rd], t0);
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DSBH:
+        {
+            TCGv t1 = tcg_temp_new();
+
+            tcg_gen_shri_tl(t1, t0, 8);
+            tcg_gen_andi_tl(t1, t1, 0x00FF00FF00FF00FFULL);
+            tcg_gen_shli_tl(t0, t0, 8);
+            tcg_gen_andi_tl(t0, t0, ~0x00FF00FF00FF00FFULL);
+            tcg_gen_or_tl(cpu_gpr[rd], t0, t1);
+            tcg_temp_free(t1);
+        }
+        break;
+    case OPC_DSHD:
+        {
+            TCGv t1 = tcg_temp_new();
+
+            tcg_gen_shri_tl(t1, t0, 16);
+            tcg_gen_andi_tl(t1, t1, 0x0000FFFF0000FFFFULL);
+            tcg_gen_shli_tl(t0, t0, 16);
+            tcg_gen_andi_tl(t0, t0, ~0x0000FFFF0000FFFFULL);
+            tcg_gen_or_tl(t0, t0, t1);
+            tcg_gen_shri_tl(t1, t0, 32);
+            tcg_gen_shli_tl(t0, t0, 32);
+            tcg_gen_or_tl(cpu_gpr[rd], t0, t1);
+            tcg_temp_free(t1);
+        }
+        break;
+#endif
+    default:
+        MIPS_INVAL("bsfhl");
+        generate_exception(ctx, EXCP_RI);
+        tcg_temp_free(t0);
+        return;
+    }
+    tcg_temp_free(t0);
+}
+
+#ifndef CONFIG_USER_ONLY
+/* CP0 (MMU and control) */
+static inline void gen_mfc0_load32 (TCGv arg, target_ulong off)
+{
+    TCGv_i32 t0 = tcg_temp_new_i32();
+
+    tcg_gen_ld_i32(t0, cpu_env, off);
+    tcg_gen_ext_i32_tl(arg, t0);
+    tcg_temp_free_i32(t0);
+}
+
+static inline void gen_mfc0_load64 (TCGv arg, target_ulong off)
+{
+    tcg_gen_ld_tl(arg, cpu_env, off);
+    tcg_gen_ext32s_tl(arg, arg);
+}
+
+static inline void gen_mtc0_store32 (TCGv arg, target_ulong off)
+{
+    TCGv_i32 t0 = tcg_temp_new_i32();
+
+    tcg_gen_trunc_tl_i32(t0, arg);
+    tcg_gen_st_i32(t0, cpu_env, off);
+    tcg_temp_free_i32(t0);
+}
+
+static inline void gen_mtc0_store64 (TCGv arg, target_ulong off)
+{
+    tcg_gen_ext32s_tl(arg, arg);
+    tcg_gen_st_tl(arg, cpu_env, off);
+}
+
+static void gen_mfc0 (CPUState *env, DisasContext *ctx, TCGv arg, int reg, int sel)
+{
+    const char *rn = "invalid";
+
+    if (sel != 0)
+        check_insn(env, ctx, ISA_MIPS32);
+
+    switch (reg) {
+    case 0:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Index));
+            rn = "Index";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_mvpcontrol(arg);
+            rn = "MVPControl";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_mvpconf0(arg);
+            rn = "MVPConf0";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_mvpconf1(arg);
+            rn = "MVPConf1";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 1:
+        switch (sel) {
+        case 0:
+            gen_helper_mfc0_random(arg);
+            rn = "Random";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_VPEControl));
+            rn = "VPEControl";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_VPEConf0));
+            rn = "VPEConf0";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_VPEConf1));
+            rn = "VPEConf1";
+            break;
+        case 4:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load64(arg, offsetof(CPUState, CP0_YQMask));
+            rn = "YQMask";
+            break;
+        case 5:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load64(arg, offsetof(CPUState, CP0_VPESchedule));
+            rn = "VPESchedule";
+            break;
+        case 6:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load64(arg, offsetof(CPUState, CP0_VPEScheFBack));
+            rn = "VPEScheFBack";
+            break;
+        case 7:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_VPEOpt));
+            rn = "VPEOpt";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 2:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_EntryLo0));
+            tcg_gen_ext32s_tl(arg, arg);
+            rn = "EntryLo0";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_tcstatus(arg);
+            rn = "TCStatus";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_tcbind(arg);
+            rn = "TCBind";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_tcrestart(arg);
+            rn = "TCRestart";
+            break;
+        case 4:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_tchalt(arg);
+            rn = "TCHalt";
+            break;
+        case 5:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_tccontext(arg);
+            rn = "TCContext";
+            break;
+        case 6:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_tcschedule(arg);
+            rn = "TCSchedule";
+            break;
+        case 7:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_tcschefback(arg);
+            rn = "TCScheFBack";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 3:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_EntryLo1));
+            tcg_gen_ext32s_tl(arg, arg);
+            rn = "EntryLo1";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 4:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_Context));
+            tcg_gen_ext32s_tl(arg, arg);
+            rn = "Context";
+            break;
+        case 1:
+//            gen_helper_mfc0_contextconfig(arg); /* SmartMIPS ASE */
+            rn = "ContextConfig";
+//            break;
+        default:
+            goto die;
+        }
+        break;
+    case 5:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_PageMask));
+            rn = "PageMask";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_PageGrain));
+            rn = "PageGrain";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 6:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Wired));
+            rn = "Wired";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSConf0));
+            rn = "SRSConf0";
+            break;
+        case 2:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSConf1));
+            rn = "SRSConf1";
+            break;
+        case 3:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSConf2));
+            rn = "SRSConf2";
+            break;
+        case 4:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSConf3));
+            rn = "SRSConf3";
+            break;
+        case 5:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSConf4));
+            rn = "SRSConf4";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 7:
+        switch (sel) {
+        case 0:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_HWREna));
+            rn = "HWREna";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 8:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_BadVAddr));
+            tcg_gen_ext32s_tl(arg, arg);
+            rn = "BadVAddr";
+            break;
+        default:
+            goto die;
+       }
+        break;
+    case 9:
+        switch (sel) {
+        case 0:
+            /* Mark as an IO operation because we read the time.  */
+            if (use_icount)
+                gen_io_start();
+            gen_helper_mfc0_count(arg);
+            if (use_icount) {
+                gen_io_end();
+                ctx->bstate = BS_STOP;
+            }
+            rn = "Count";
+            break;
+        /* 6,7 are implementation dependent */
+        default:
+            goto die;
+        }
+        break;
+    case 10:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_EntryHi));
+            tcg_gen_ext32s_tl(arg, arg);
+            rn = "EntryHi";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 11:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Compare));
+            rn = "Compare";
+            break;
+        /* 6,7 are implementation dependent */
+        default:
+            goto die;
+        }
+        break;
+    case 12:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Status));
+            rn = "Status";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_IntCtl));
+            rn = "IntCtl";
+            break;
+        case 2:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSCtl));
+            rn = "SRSCtl";
+            break;
+        case 3:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSMap));
+            rn = "SRSMap";
+            break;
+        default:
+            goto die;
+       }
+        break;
+    case 13:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Cause));
+            rn = "Cause";
+            break;
+        default:
+            goto die;
+       }
+        break;
+    case 14:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_EPC));
+            tcg_gen_ext32s_tl(arg, arg);
+            rn = "EPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 15:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_PRid));
+            rn = "PRid";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_EBase));
+            rn = "EBase";
+            break;
+        default:
+            goto die;
+       }
+        break;
+    case 16:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config0));
+            rn = "Config";
+            break;
+        case 1:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config1));
+            rn = "Config1";
+            break;
+        case 2:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config2));
+            rn = "Config2";
+            break;
+        case 3:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config3));
+            rn = "Config3";
+            break;
+        /* 4,5 are reserved */
+        /* 6,7 are implementation dependent */
+        case 6:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config6));
+            rn = "Config6";
+            break;
+        case 7:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config7));
+            rn = "Config7";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 17:
+        switch (sel) {
+        case 0:
+            gen_helper_mfc0_lladdr(arg);
+            rn = "LLAddr";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 18:
+        switch (sel) {
+        case 0 ... 7:
+            gen_helper_1i(mfc0_watchlo, arg, sel);
+            rn = "WatchLo";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 19:
+        switch (sel) {
+        case 0 ...7:
+            gen_helper_1i(mfc0_watchhi, arg, sel);
+            rn = "WatchHi";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 20:
+        switch (sel) {
+        case 0:
+#if defined(TARGET_MIPS64)
+            check_insn(env, ctx, ISA_MIPS3);
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_XContext));
+            tcg_gen_ext32s_tl(arg, arg);
+            rn = "XContext";
+            break;
+#endif
+        default:
+            goto die;
+        }
+        break;
+    case 21:
+       /* Officially reserved, but sel 0 is used for R1x000 framemask */
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Framemask));
+            rn = "Framemask";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 22:
+        tcg_gen_movi_tl(arg, 0); /* unimplemented */
+        rn = "'Diagnostic"; /* implementation dependent */
+        break;
+    case 23:
+        switch (sel) {
+        case 0:
+            gen_helper_mfc0_debug(arg); /* EJTAG support */
+            rn = "Debug";
+            break;
+        case 1:
+//            gen_helper_mfc0_tracecontrol(arg); /* PDtrace support */
+            rn = "TraceControl";
+//            break;
+        case 2:
+//            gen_helper_mfc0_tracecontrol2(arg); /* PDtrace support */
+            rn = "TraceControl2";
+//            break;
+        case 3:
+//            gen_helper_mfc0_usertracedata(arg); /* PDtrace support */
+            rn = "UserTraceData";
+//            break;
+        case 4:
+//            gen_helper_mfc0_tracebpc(arg); /* PDtrace support */
+            rn = "TraceBPC";
+//            break;
+        default:
+            goto die;
+        }
+        break;
+    case 24:
+        switch (sel) {
+        case 0:
+            /* EJTAG support */
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_DEPC));
+            tcg_gen_ext32s_tl(arg, arg);
+            rn = "DEPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 25:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Performance0));
+            rn = "Performance0";
+            break;
+        case 1:
+//            gen_helper_mfc0_performance1(arg);
+            rn = "Performance1";
+//            break;
+        case 2:
+//            gen_helper_mfc0_performance2(arg);
+            rn = "Performance2";
+//            break;
+        case 3:
+//            gen_helper_mfc0_performance3(arg);
+            rn = "Performance3";
+//            break;
+        case 4:
+//            gen_helper_mfc0_performance4(arg);
+            rn = "Performance4";
+//            break;
+        case 5:
+//            gen_helper_mfc0_performance5(arg);
+            rn = "Performance5";
+//            break;
+        case 6:
+//            gen_helper_mfc0_performance6(arg);
+            rn = "Performance6";
+//            break;
+        case 7:
+//            gen_helper_mfc0_performance7(arg);
+            rn = "Performance7";
+//            break;
+        default:
+            goto die;
+        }
+        break;
+    case 26:
+        tcg_gen_movi_tl(arg, 0); /* unimplemented */
+        rn = "ECC";
+        break;
+    case 27:
+        switch (sel) {
+        case 0 ... 3:
+            tcg_gen_movi_tl(arg, 0); /* unimplemented */
+            rn = "CacheErr";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 28:
+        switch (sel) {
+        case 0:
+        case 2:
+        case 4:
+        case 6:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_TagLo));
+            rn = "TagLo";
+            break;
+        case 1:
+        case 3:
+        case 5:
+        case 7:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_DataLo));
+            rn = "DataLo";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 29:
+        switch (sel) {
+        case 0:
+        case 2:
+        case 4:
+        case 6:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_TagHi));
+            rn = "TagHi";
+            break;
+        case 1:
+        case 3:
+        case 5:
+        case 7:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_DataHi));
+            rn = "DataHi";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 30:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_ErrorEPC));
+            tcg_gen_ext32s_tl(arg, arg);
+            rn = "ErrorEPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 31:
+        switch (sel) {
+        case 0:
+            /* EJTAG support */
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_DESAVE));
+            rn = "DESAVE";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    default:
+       goto die;
+    }
+    LOG_DISAS("mfc0 %s (reg %d sel %d)\n", rn, reg, sel);
+    return;
+
+die:
+    LOG_DISAS("mfc0 %s (reg %d sel %d)\n", rn, reg, sel);
+    generate_exception(ctx, EXCP_RI);
+}
+
+static void gen_mtc0 (CPUState *env, DisasContext *ctx, TCGv arg, int reg, int sel)
+{
+    const char *rn = "invalid";
+
+    if (sel != 0)
+        check_insn(env, ctx, ISA_MIPS32);
+
+    if (use_icount)
+        gen_io_start();
+
+    switch (reg) {
+    case 0:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_index(arg);
+            rn = "Index";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_mvpcontrol(arg);
+            rn = "MVPControl";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            /* ignored */
+            rn = "MVPConf0";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            /* ignored */
+            rn = "MVPConf1";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 1:
+        switch (sel) {
+        case 0:
+            /* ignored */
+            rn = "Random";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_vpecontrol(arg);
+            rn = "VPEControl";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_vpeconf0(arg);
+            rn = "VPEConf0";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_vpeconf1(arg);
+            rn = "VPEConf1";
+            break;
+        case 4:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_yqmask(arg);
+            rn = "YQMask";
+            break;
+        case 5:
+            check_insn(env, ctx, ASE_MT);
+            gen_mtc0_store64(arg, offsetof(CPUState, CP0_VPESchedule));
+            rn = "VPESchedule";
+            break;
+        case 6:
+            check_insn(env, ctx, ASE_MT);
+            gen_mtc0_store64(arg, offsetof(CPUState, CP0_VPEScheFBack));
+            rn = "VPEScheFBack";
+            break;
+        case 7:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_vpeopt(arg);
+            rn = "VPEOpt";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 2:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_entrylo0(arg);
+            rn = "EntryLo0";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tcstatus(arg);
+            rn = "TCStatus";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tcbind(arg);
+            rn = "TCBind";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tcrestart(arg);
+            rn = "TCRestart";
+            break;
+        case 4:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tchalt(arg);
+            rn = "TCHalt";
+            break;
+        case 5:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tccontext(arg);
+            rn = "TCContext";
+            break;
+        case 6:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tcschedule(arg);
+            rn = "TCSchedule";
+            break;
+        case 7:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tcschefback(arg);
+            rn = "TCScheFBack";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 3:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_entrylo1(arg);
+            rn = "EntryLo1";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 4:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_context(arg);
+            rn = "Context";
+            break;
+        case 1:
+//            gen_helper_mtc0_contextconfig(arg); /* SmartMIPS ASE */
+            rn = "ContextConfig";
+//            break;
+        default:
+            goto die;
+        }
+        break;
+    case 5:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_pagemask(arg);
+            rn = "PageMask";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_pagegrain(arg);
+            rn = "PageGrain";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 6:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_wired(arg);
+            rn = "Wired";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsconf0(arg);
+            rn = "SRSConf0";
+            break;
+        case 2:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsconf1(arg);
+            rn = "SRSConf1";
+            break;
+        case 3:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsconf2(arg);
+            rn = "SRSConf2";
+            break;
+        case 4:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsconf3(arg);
+            rn = "SRSConf3";
+            break;
+        case 5:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsconf4(arg);
+            rn = "SRSConf4";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 7:
+        switch (sel) {
+        case 0:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_hwrena(arg);
+            rn = "HWREna";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 8:
+        /* ignored */
+        rn = "BadVAddr";
+        break;
+    case 9:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_count(arg);
+            rn = "Count";
+            break;
+        /* 6,7 are implementation dependent */
+        default:
+            goto die;
+        }
+        break;
+    case 10:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_entryhi(arg);
+            rn = "EntryHi";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 11:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_compare(arg);
+            rn = "Compare";
+            break;
+        /* 6,7 are implementation dependent */
+        default:
+            goto die;
+        }
+        break;
+    case 12:
+        switch (sel) {
+        case 0:
+            save_cpu_state(ctx, 1);
+            gen_helper_mtc0_status(arg);
+            /* BS_STOP isn't good enough here, hflags may have changed. */
+            gen_save_pc(ctx->pc + 4);
+            ctx->bstate = BS_EXCP;
+            rn = "Status";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_intctl(arg);
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "IntCtl";
+            break;
+        case 2:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsctl(arg);
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "SRSCtl";
+            break;
+        case 3:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mtc0_store32(arg, offsetof(CPUState, CP0_SRSMap));
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "SRSMap";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 13:
+        switch (sel) {
+        case 0:
+            save_cpu_state(ctx, 1);
+            gen_helper_mtc0_cause(arg);
+            rn = "Cause";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 14:
+        switch (sel) {
+        case 0:
+            gen_mtc0_store64(arg, offsetof(CPUState, CP0_EPC));
+            rn = "EPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 15:
+        switch (sel) {
+        case 0:
+            /* ignored */
+            rn = "PRid";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_ebase(arg);
+            rn = "EBase";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 16:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_config0(arg);
+            rn = "Config";
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            break;
+        case 1:
+            /* ignored, read only */
+            rn = "Config1";
+            break;
+        case 2:
+            gen_helper_mtc0_config2(arg);
+            rn = "Config2";
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            break;
+        case 3:
+            /* ignored, read only */
+            rn = "Config3";
+            break;
+        /* 4,5 are reserved */
+        /* 6,7 are implementation dependent */
+        case 6:
+            /* ignored */
+            rn = "Config6";
+            break;
+        case 7:
+            /* ignored */
+            rn = "Config7";
+            break;
+        default:
+            rn = "Invalid config selector";
+            goto die;
+        }
+        break;
+    case 17:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_lladdr(arg);
+            rn = "LLAddr";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 18:
+        switch (sel) {
+        case 0 ... 7:
+            gen_helper_1i(mtc0_watchlo, arg, sel);
+            rn = "WatchLo";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 19:
+        switch (sel) {
+        case 0 ... 7:
+            gen_helper_1i(mtc0_watchhi, arg, sel);
+            rn = "WatchHi";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 20:
+        switch (sel) {
+        case 0:
+#if defined(TARGET_MIPS64)
+            check_insn(env, ctx, ISA_MIPS3);
+            gen_helper_mtc0_xcontext(arg);
+            rn = "XContext";
+            break;
+#endif
+        default:
+            goto die;
+        }
+        break;
+    case 21:
+       /* Officially reserved, but sel 0 is used for R1x000 framemask */
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_framemask(arg);
+            rn = "Framemask";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 22:
+        /* ignored */
+        rn = "Diagnostic"; /* implementation dependent */
+        break;
+    case 23:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_debug(arg); /* EJTAG support */
+            /* BS_STOP isn't good enough here, hflags may have changed. */
+            gen_save_pc(ctx->pc + 4);
+            ctx->bstate = BS_EXCP;
+            rn = "Debug";
+            break;
+        case 1:
+//            gen_helper_mtc0_tracecontrol(arg); /* PDtrace support */
+            rn = "TraceControl";
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+//            break;
+        case 2:
+//            gen_helper_mtc0_tracecontrol2(arg); /* PDtrace support */
+            rn = "TraceControl2";
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+//            break;
+        case 3:
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+//            gen_helper_mtc0_usertracedata(arg); /* PDtrace support */
+            rn = "UserTraceData";
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+//            break;
+        case 4:
+//            gen_helper_mtc0_tracebpc(arg); /* PDtrace support */
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "TraceBPC";
+//            break;
+        default:
+            goto die;
+        }
+        break;
+    case 24:
+        switch (sel) {
+        case 0:
+            /* EJTAG support */
+            gen_mtc0_store64(arg, offsetof(CPUState, CP0_DEPC));
+            rn = "DEPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 25:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_performance0(arg);
+            rn = "Performance0";
+            break;
+        case 1:
+//            gen_helper_mtc0_performance1(arg);
+            rn = "Performance1";
+//            break;
+        case 2:
+//            gen_helper_mtc0_performance2(arg);
+            rn = "Performance2";
+//            break;
+        case 3:
+//            gen_helper_mtc0_performance3(arg);
+            rn = "Performance3";
+//            break;
+        case 4:
+//            gen_helper_mtc0_performance4(arg);
+            rn = "Performance4";
+//            break;
+        case 5:
+//            gen_helper_mtc0_performance5(arg);
+            rn = "Performance5";
+//            break;
+        case 6:
+//            gen_helper_mtc0_performance6(arg);
+            rn = "Performance6";
+//            break;
+        case 7:
+//            gen_helper_mtc0_performance7(arg);
+            rn = "Performance7";
+//            break;
+        default:
+            goto die;
+        }
+       break;
+    case 26:
+        /* ignored */
+        rn = "ECC";
+        break;
+    case 27:
+        switch (sel) {
+        case 0 ... 3:
+            /* ignored */
+            rn = "CacheErr";
+            break;
+        default:
+            goto die;
+        }
+       break;
+    case 28:
+        switch (sel) {
+        case 0:
+        case 2:
+        case 4:
+        case 6:
+            gen_helper_mtc0_taglo(arg);
+            rn = "TagLo";
+            break;
+        case 1:
+        case 3:
+        case 5:
+        case 7:
+            gen_helper_mtc0_datalo(arg);
+            rn = "DataLo";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 29:
+        switch (sel) {
+        case 0:
+        case 2:
+        case 4:
+        case 6:
+            gen_helper_mtc0_taghi(arg);
+            rn = "TagHi";
+            break;
+        case 1:
+        case 3:
+        case 5:
+        case 7:
+            gen_helper_mtc0_datahi(arg);
+            rn = "DataHi";
+            break;
+        default:
+            rn = "invalid sel";
+            goto die;
+        }
+       break;
+    case 30:
+        switch (sel) {
+        case 0:
+            gen_mtc0_store64(arg, offsetof(CPUState, CP0_ErrorEPC));
+            rn = "ErrorEPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 31:
+        switch (sel) {
+        case 0:
+            /* EJTAG support */
+            gen_mtc0_store32(arg, offsetof(CPUState, CP0_DESAVE));
+            rn = "DESAVE";
+            break;
+        default:
+            goto die;
+        }
+        /* Stop translation as we may have switched the execution mode */
+        ctx->bstate = BS_STOP;
+        break;
+    default:
+       goto die;
+    }
+    LOG_DISAS("mtc0 %s (reg %d sel %d)\n", rn, reg, sel);
+    /* For simplicity assume that all writes can cause interrupts.  */
+    if (use_icount) {
+        gen_io_end();
+        ctx->bstate = BS_STOP;
+    }
+    return;
+
+die:
+    LOG_DISAS("mtc0 %s (reg %d sel %d)\n", rn, reg, sel);
+    generate_exception(ctx, EXCP_RI);
+}
+
+#if defined(TARGET_MIPS64)
+static void gen_dmfc0 (CPUState *env, DisasContext *ctx, TCGv arg, int reg, int sel)
+{
+    const char *rn = "invalid";
+
+    if (sel != 0)
+        check_insn(env, ctx, ISA_MIPS64);
+
+    switch (reg) {
+    case 0:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Index));
+            rn = "Index";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_mvpcontrol(arg);
+            rn = "MVPControl";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_mvpconf0(arg);
+            rn = "MVPConf0";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_mvpconf1(arg);
+            rn = "MVPConf1";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 1:
+        switch (sel) {
+        case 0:
+            gen_helper_mfc0_random(arg);
+            rn = "Random";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_VPEControl));
+            rn = "VPEControl";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_VPEConf0));
+            rn = "VPEConf0";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_VPEConf1));
+            rn = "VPEConf1";
+            break;
+        case 4:
+            check_insn(env, ctx, ASE_MT);
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_YQMask));
+            rn = "YQMask";
+            break;
+        case 5:
+            check_insn(env, ctx, ASE_MT);
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_VPESchedule));
+            rn = "VPESchedule";
+            break;
+        case 6:
+            check_insn(env, ctx, ASE_MT);
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_VPEScheFBack));
+            rn = "VPEScheFBack";
+            break;
+        case 7:
+            check_insn(env, ctx, ASE_MT);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_VPEOpt));
+            rn = "VPEOpt";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 2:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_EntryLo0));
+            rn = "EntryLo0";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_tcstatus(arg);
+            rn = "TCStatus";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mfc0_tcbind(arg);
+            rn = "TCBind";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_dmfc0_tcrestart(arg);
+            rn = "TCRestart";
+            break;
+        case 4:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_dmfc0_tchalt(arg);
+            rn = "TCHalt";
+            break;
+        case 5:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_dmfc0_tccontext(arg);
+            rn = "TCContext";
+            break;
+        case 6:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_dmfc0_tcschedule(arg);
+            rn = "TCSchedule";
+            break;
+        case 7:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_dmfc0_tcschefback(arg);
+            rn = "TCScheFBack";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 3:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_EntryLo1));
+            rn = "EntryLo1";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 4:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_Context));
+            rn = "Context";
+            break;
+        case 1:
+//            gen_helper_dmfc0_contextconfig(arg); /* SmartMIPS ASE */
+            rn = "ContextConfig";
+//            break;
+        default:
+            goto die;
+        }
+        break;
+    case 5:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_PageMask));
+            rn = "PageMask";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_PageGrain));
+            rn = "PageGrain";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 6:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Wired));
+            rn = "Wired";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSConf0));
+            rn = "SRSConf0";
+            break;
+        case 2:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSConf1));
+            rn = "SRSConf1";
+            break;
+        case 3:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSConf2));
+            rn = "SRSConf2";
+            break;
+        case 4:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSConf3));
+            rn = "SRSConf3";
+            break;
+        case 5:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSConf4));
+            rn = "SRSConf4";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 7:
+        switch (sel) {
+        case 0:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_HWREna));
+            rn = "HWREna";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 8:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_BadVAddr));
+            rn = "BadVAddr";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 9:
+        switch (sel) {
+        case 0:
+            /* Mark as an IO operation because we read the time.  */
+            if (use_icount)
+                gen_io_start();
+            gen_helper_mfc0_count(arg);
+            if (use_icount) {
+                gen_io_end();
+                ctx->bstate = BS_STOP;
+            }
+            rn = "Count";
+            break;
+        /* 6,7 are implementation dependent */
+        default:
+            goto die;
+        }
+        break;
+    case 10:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_EntryHi));
+            rn = "EntryHi";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 11:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Compare));
+            rn = "Compare";
+            break;
+        /* 6,7 are implementation dependent */
+        default:
+            goto die;
+        }
+        break;
+    case 12:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Status));
+            rn = "Status";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_IntCtl));
+            rn = "IntCtl";
+            break;
+        case 2:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSCtl));
+            rn = "SRSCtl";
+            break;
+        case 3:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_SRSMap));
+            rn = "SRSMap";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 13:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Cause));
+            rn = "Cause";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 14:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_EPC));
+            rn = "EPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 15:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_PRid));
+            rn = "PRid";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_EBase));
+            rn = "EBase";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 16:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config0));
+            rn = "Config";
+            break;
+        case 1:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config1));
+            rn = "Config1";
+            break;
+        case 2:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config2));
+            rn = "Config2";
+            break;
+        case 3:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config3));
+            rn = "Config3";
+            break;
+       /* 6,7 are implementation dependent */
+        case 6:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config6));
+            rn = "Config6";
+            break;
+        case 7:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Config7));
+            rn = "Config7";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 17:
+        switch (sel) {
+        case 0:
+            gen_helper_dmfc0_lladdr(arg);
+            rn = "LLAddr";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 18:
+        switch (sel) {
+        case 0 ... 7:
+            gen_helper_1i(dmfc0_watchlo, arg, sel);
+            rn = "WatchLo";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 19:
+        switch (sel) {
+        case 0 ... 7:
+            gen_helper_1i(mfc0_watchhi, arg, sel);
+            rn = "WatchHi";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 20:
+        switch (sel) {
+        case 0:
+            check_insn(env, ctx, ISA_MIPS3);
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_XContext));
+            rn = "XContext";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 21:
+       /* Officially reserved, but sel 0 is used for R1x000 framemask */
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Framemask));
+            rn = "Framemask";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 22:
+        tcg_gen_movi_tl(arg, 0); /* unimplemented */
+        rn = "'Diagnostic"; /* implementation dependent */
+        break;
+    case 23:
+        switch (sel) {
+        case 0:
+            gen_helper_mfc0_debug(arg); /* EJTAG support */
+            rn = "Debug";
+            break;
+        case 1:
+//            gen_helper_dmfc0_tracecontrol(arg); /* PDtrace support */
+            rn = "TraceControl";
+//            break;
+        case 2:
+//            gen_helper_dmfc0_tracecontrol2(arg); /* PDtrace support */
+            rn = "TraceControl2";
+//            break;
+        case 3:
+//            gen_helper_dmfc0_usertracedata(arg); /* PDtrace support */
+            rn = "UserTraceData";
+//            break;
+        case 4:
+//            gen_helper_dmfc0_tracebpc(arg); /* PDtrace support */
+            rn = "TraceBPC";
+//            break;
+        default:
+            goto die;
+        }
+        break;
+    case 24:
+        switch (sel) {
+        case 0:
+            /* EJTAG support */
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_DEPC));
+            rn = "DEPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 25:
+        switch (sel) {
+        case 0:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_Performance0));
+            rn = "Performance0";
+            break;
+        case 1:
+//            gen_helper_dmfc0_performance1(arg);
+            rn = "Performance1";
+//            break;
+        case 2:
+//            gen_helper_dmfc0_performance2(arg);
+            rn = "Performance2";
+//            break;
+        case 3:
+//            gen_helper_dmfc0_performance3(arg);
+            rn = "Performance3";
+//            break;
+        case 4:
+//            gen_helper_dmfc0_performance4(arg);
+            rn = "Performance4";
+//            break;
+        case 5:
+//            gen_helper_dmfc0_performance5(arg);
+            rn = "Performance5";
+//            break;
+        case 6:
+//            gen_helper_dmfc0_performance6(arg);
+            rn = "Performance6";
+//            break;
+        case 7:
+//            gen_helper_dmfc0_performance7(arg);
+            rn = "Performance7";
+//            break;
+        default:
+            goto die;
+        }
+        break;
+    case 26:
+        tcg_gen_movi_tl(arg, 0); /* unimplemented */
+        rn = "ECC";
+        break;
+    case 27:
+        switch (sel) {
+        /* ignored */
+        case 0 ... 3:
+            tcg_gen_movi_tl(arg, 0); /* unimplemented */
+            rn = "CacheErr";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 28:
+        switch (sel) {
+        case 0:
+        case 2:
+        case 4:
+        case 6:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_TagLo));
+            rn = "TagLo";
+            break;
+        case 1:
+        case 3:
+        case 5:
+        case 7:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_DataLo));
+            rn = "DataLo";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 29:
+        switch (sel) {
+        case 0:
+        case 2:
+        case 4:
+        case 6:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_TagHi));
+            rn = "TagHi";
+            break;
+        case 1:
+        case 3:
+        case 5:
+        case 7:
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_DataHi));
+            rn = "DataHi";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 30:
+        switch (sel) {
+        case 0:
+            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUState, CP0_ErrorEPC));
+            rn = "ErrorEPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 31:
+        switch (sel) {
+        case 0:
+            /* EJTAG support */
+            gen_mfc0_load32(arg, offsetof(CPUState, CP0_DESAVE));
+            rn = "DESAVE";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    default:
+        goto die;
+    }
+    LOG_DISAS("dmfc0 %s (reg %d sel %d)\n", rn, reg, sel);
+    return;
+
+die:
+    LOG_DISAS("dmfc0 %s (reg %d sel %d)\n", rn, reg, sel);
+    generate_exception(ctx, EXCP_RI);
+}
+
+static void gen_dmtc0 (CPUState *env, DisasContext *ctx, TCGv arg, int reg, int sel)
+{
+    const char *rn = "invalid";
+
+    if (sel != 0)
+        check_insn(env, ctx, ISA_MIPS64);
+
+    if (use_icount)
+        gen_io_start();
+
+    switch (reg) {
+    case 0:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_index(arg);
+            rn = "Index";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_mvpcontrol(arg);
+            rn = "MVPControl";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            /* ignored */
+            rn = "MVPConf0";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            /* ignored */
+            rn = "MVPConf1";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 1:
+        switch (sel) {
+        case 0:
+            /* ignored */
+            rn = "Random";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_vpecontrol(arg);
+            rn = "VPEControl";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_vpeconf0(arg);
+            rn = "VPEConf0";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_vpeconf1(arg);
+            rn = "VPEConf1";
+            break;
+        case 4:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_yqmask(arg);
+            rn = "YQMask";
+            break;
+        case 5:
+            check_insn(env, ctx, ASE_MT);
+            tcg_gen_st_tl(arg, cpu_env, offsetof(CPUState, CP0_VPESchedule));
+            rn = "VPESchedule";
+            break;
+        case 6:
+            check_insn(env, ctx, ASE_MT);
+            tcg_gen_st_tl(arg, cpu_env, offsetof(CPUState, CP0_VPEScheFBack));
+            rn = "VPEScheFBack";
+            break;
+        case 7:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_vpeopt(arg);
+            rn = "VPEOpt";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 2:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_entrylo0(arg);
+            rn = "EntryLo0";
+            break;
+        case 1:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tcstatus(arg);
+            rn = "TCStatus";
+            break;
+        case 2:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tcbind(arg);
+            rn = "TCBind";
+            break;
+        case 3:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tcrestart(arg);
+            rn = "TCRestart";
+            break;
+        case 4:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tchalt(arg);
+            rn = "TCHalt";
+            break;
+        case 5:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tccontext(arg);
+            rn = "TCContext";
+            break;
+        case 6:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tcschedule(arg);
+            rn = "TCSchedule";
+            break;
+        case 7:
+            check_insn(env, ctx, ASE_MT);
+            gen_helper_mtc0_tcschefback(arg);
+            rn = "TCScheFBack";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 3:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_entrylo1(arg);
+            rn = "EntryLo1";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 4:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_context(arg);
+            rn = "Context";
+            break;
+        case 1:
+//           gen_helper_mtc0_contextconfig(arg); /* SmartMIPS ASE */
+            rn = "ContextConfig";
+//           break;
+        default:
+            goto die;
+        }
+        break;
+    case 5:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_pagemask(arg);
+            rn = "PageMask";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_pagegrain(arg);
+            rn = "PageGrain";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 6:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_wired(arg);
+            rn = "Wired";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsconf0(arg);
+            rn = "SRSConf0";
+            break;
+        case 2:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsconf1(arg);
+            rn = "SRSConf1";
+            break;
+        case 3:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsconf2(arg);
+            rn = "SRSConf2";
+            break;
+        case 4:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsconf3(arg);
+            rn = "SRSConf3";
+            break;
+        case 5:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsconf4(arg);
+            rn = "SRSConf4";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 7:
+        switch (sel) {
+        case 0:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_hwrena(arg);
+            rn = "HWREna";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 8:
+        /* ignored */
+        rn = "BadVAddr";
+        break;
+    case 9:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_count(arg);
+            rn = "Count";
+            break;
+        /* 6,7 are implementation dependent */
+        default:
+            goto die;
+        }
+        /* Stop translation as we may have switched the execution mode */
+        ctx->bstate = BS_STOP;
+        break;
+    case 10:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_entryhi(arg);
+            rn = "EntryHi";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 11:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_compare(arg);
+            rn = "Compare";
+            break;
+        /* 6,7 are implementation dependent */
+        default:
+            goto die;
+        }
+        /* Stop translation as we may have switched the execution mode */
+        ctx->bstate = BS_STOP;
+        break;
+    case 12:
+        switch (sel) {
+        case 0:
+            save_cpu_state(ctx, 1);
+            gen_helper_mtc0_status(arg);
+            /* BS_STOP isn't good enough here, hflags may have changed. */
+            gen_save_pc(ctx->pc + 4);
+            ctx->bstate = BS_EXCP;
+            rn = "Status";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_intctl(arg);
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "IntCtl";
+            break;
+        case 2:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_srsctl(arg);
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "SRSCtl";
+            break;
+        case 3:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_mtc0_store32(arg, offsetof(CPUState, CP0_SRSMap));
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "SRSMap";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 13:
+        switch (sel) {
+        case 0:
+            save_cpu_state(ctx, 1);
+            gen_helper_mtc0_cause(arg);
+            rn = "Cause";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 14:
+        switch (sel) {
+        case 0:
+            tcg_gen_st_tl(arg, cpu_env, offsetof(CPUState, CP0_EPC));
+            rn = "EPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 15:
+        switch (sel) {
+        case 0:
+            /* ignored */
+            rn = "PRid";
+            break;
+        case 1:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_helper_mtc0_ebase(arg);
+            rn = "EBase";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 16:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_config0(arg);
+            rn = "Config";
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            break;
+        case 1:
+            /* ignored, read only */
+            rn = "Config1";
+            break;
+        case 2:
+            gen_helper_mtc0_config2(arg);
+            rn = "Config2";
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            break;
+        case 3:
+            /* ignored */
+            rn = "Config3";
+            break;
+        /* 6,7 are implementation dependent */
+        default:
+            rn = "Invalid config selector";
+            goto die;
+        }
+        break;
+    case 17:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_lladdr(arg);
+            rn = "LLAddr";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 18:
+        switch (sel) {
+        case 0 ... 7:
+            gen_helper_1i(mtc0_watchlo, arg, sel);
+            rn = "WatchLo";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 19:
+        switch (sel) {
+        case 0 ... 7:
+            gen_helper_1i(mtc0_watchhi, arg, sel);
+            rn = "WatchHi";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 20:
+        switch (sel) {
+        case 0:
+            check_insn(env, ctx, ISA_MIPS3);
+            gen_helper_mtc0_xcontext(arg);
+            rn = "XContext";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 21:
+       /* Officially reserved, but sel 0 is used for R1x000 framemask */
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_framemask(arg);
+            rn = "Framemask";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 22:
+        /* ignored */
+        rn = "Diagnostic"; /* implementation dependent */
+        break;
+    case 23:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_debug(arg); /* EJTAG support */
+            /* BS_STOP isn't good enough here, hflags may have changed. */
+            gen_save_pc(ctx->pc + 4);
+            ctx->bstate = BS_EXCP;
+            rn = "Debug";
+            break;
+        case 1:
+//            gen_helper_mtc0_tracecontrol(arg); /* PDtrace support */
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "TraceControl";
+//            break;
+        case 2:
+//            gen_helper_mtc0_tracecontrol2(arg); /* PDtrace support */
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "TraceControl2";
+//            break;
+        case 3:
+//            gen_helper_mtc0_usertracedata(arg); /* PDtrace support */
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "UserTraceData";
+//            break;
+        case 4:
+//            gen_helper_mtc0_tracebpc(arg); /* PDtrace support */
+            /* Stop translation as we may have switched the execution mode */
+            ctx->bstate = BS_STOP;
+            rn = "TraceBPC";
+//            break;
+        default:
+            goto die;
+        }
+        break;
+    case 24:
+        switch (sel) {
+        case 0:
+            /* EJTAG support */
+            tcg_gen_st_tl(arg, cpu_env, offsetof(CPUState, CP0_DEPC));
+            rn = "DEPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 25:
+        switch (sel) {
+        case 0:
+            gen_helper_mtc0_performance0(arg);
+            rn = "Performance0";
+            break;
+        case 1:
+//            gen_helper_mtc0_performance1(arg);
+            rn = "Performance1";
+//            break;
+        case 2:
+//            gen_helper_mtc0_performance2(arg);
+            rn = "Performance2";
+//            break;
+        case 3:
+//            gen_helper_mtc0_performance3(arg);
+            rn = "Performance3";
+//            break;
+        case 4:
+//            gen_helper_mtc0_performance4(arg);
+            rn = "Performance4";
+//            break;
+        case 5:
+//            gen_helper_mtc0_performance5(arg);
+            rn = "Performance5";
+//            break;
+        case 6:
+//            gen_helper_mtc0_performance6(arg);
+            rn = "Performance6";
+//            break;
+        case 7:
+//            gen_helper_mtc0_performance7(arg);
+            rn = "Performance7";
+//            break;
+        default:
+            goto die;
+        }
+        break;
+    case 26:
+        /* ignored */
+        rn = "ECC";
+        break;
+    case 27:
+        switch (sel) {
+        case 0 ... 3:
+            /* ignored */
+            rn = "CacheErr";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 28:
+        switch (sel) {
+        case 0:
+        case 2:
+        case 4:
+        case 6:
+            gen_helper_mtc0_taglo(arg);
+            rn = "TagLo";
+            break;
+        case 1:
+        case 3:
+        case 5:
+        case 7:
+            gen_helper_mtc0_datalo(arg);
+            rn = "DataLo";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 29:
+        switch (sel) {
+        case 0:
+        case 2:
+        case 4:
+        case 6:
+            gen_helper_mtc0_taghi(arg);
+            rn = "TagHi";
+            break;
+        case 1:
+        case 3:
+        case 5:
+        case 7:
+            gen_helper_mtc0_datahi(arg);
+            rn = "DataHi";
+            break;
+        default:
+            rn = "invalid sel";
+            goto die;
+        }
+        break;
+    case 30:
+        switch (sel) {
+        case 0:
+            tcg_gen_st_tl(arg, cpu_env, offsetof(CPUState, CP0_ErrorEPC));
+            rn = "ErrorEPC";
+            break;
+        default:
+            goto die;
+        }
+        break;
+    case 31:
+        switch (sel) {
+        case 0:
+            /* EJTAG support */
+            gen_mtc0_store32(arg, offsetof(CPUState, CP0_DESAVE));
+            rn = "DESAVE";
+            break;
+        default:
+            goto die;
+        }
+        /* Stop translation as we may have switched the execution mode */
+        ctx->bstate = BS_STOP;
+        break;
+    default:
+        goto die;
+    }
+    LOG_DISAS("dmtc0 %s (reg %d sel %d)\n", rn, reg, sel);
+    /* For simplicity assume that all writes can cause interrupts.  */
+    if (use_icount) {
+        gen_io_end();
+        ctx->bstate = BS_STOP;
+    }
+    return;
+
+die:
+    LOG_DISAS("dmtc0 %s (reg %d sel %d)\n", rn, reg, sel);
+    generate_exception(ctx, EXCP_RI);
+}
+#endif /* TARGET_MIPS64 */
+
+static void gen_mftr(CPUState *env, DisasContext *ctx, int rt, int rd,
+                     int u, int sel, int h)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+    TCGv t0 = tcg_temp_local_new();
+
+    if ((env->CP0_VPEConf0 & (1 << CP0VPEC0_MVP)) == 0 &&
+        ((env->tcs[other_tc].CP0_TCBind & (0xf << CP0TCBd_CurVPE)) !=
+         (env->active_tc.CP0_TCBind & (0xf << CP0TCBd_CurVPE))))
+        tcg_gen_movi_tl(t0, -1);
+    else if ((env->CP0_VPEControl & (0xff << CP0VPECo_TargTC)) >
+             (env->mvp->CP0_MVPConf0 & (0xff << CP0MVPC0_PTC)))
+        tcg_gen_movi_tl(t0, -1);
+    else if (u == 0) {
+        switch (rt) {
+        case 2:
+            switch (sel) {
+            case 1:
+                gen_helper_mftc0_tcstatus(t0);
+                break;
+            case 2:
+                gen_helper_mftc0_tcbind(t0);
+                break;
+            case 3:
+                gen_helper_mftc0_tcrestart(t0);
+                break;
+            case 4:
+                gen_helper_mftc0_tchalt(t0);
+                break;
+            case 5:
+                gen_helper_mftc0_tccontext(t0);
+                break;
+            case 6:
+                gen_helper_mftc0_tcschedule(t0);
+                break;
+            case 7:
+                gen_helper_mftc0_tcschefback(t0);
+                break;
+            default:
+                gen_mfc0(env, ctx, t0, rt, sel);
+                break;
+            }
+            break;
+        case 10:
+            switch (sel) {
+            case 0:
+                gen_helper_mftc0_entryhi(t0);
+                break;
+            default:
+                gen_mfc0(env, ctx, t0, rt, sel);
+                break;
+            }
+        case 12:
+            switch (sel) {
+            case 0:
+                gen_helper_mftc0_status(t0);
+                break;
+            default:
+                gen_mfc0(env, ctx, t0, rt, sel);
+                break;
+            }
+        case 23:
+            switch (sel) {
+            case 0:
+                gen_helper_mftc0_debug(t0);
+                break;
+            default:
+                gen_mfc0(env, ctx, t0, rt, sel);
+                break;
+            }
+            break;
+        default:
+            gen_mfc0(env, ctx, t0, rt, sel);
+        }
+    } else switch (sel) {
+    /* GPR registers. */
+    case 0:
+        gen_helper_1i(mftgpr, t0, rt);
+        break;
+    /* Auxiliary CPU registers */
+    case 1:
+        switch (rt) {
+        case 0:
+            gen_helper_1i(mftlo, t0, 0);
+            break;
+        case 1:
+            gen_helper_1i(mfthi, t0, 0);
+            break;
+        case 2:
+            gen_helper_1i(mftacx, t0, 0);
+            break;
+        case 4:
+            gen_helper_1i(mftlo, t0, 1);
+            break;
+        case 5:
+            gen_helper_1i(mfthi, t0, 1);
+            break;
+        case 6:
+            gen_helper_1i(mftacx, t0, 1);
+            break;
+        case 8:
+            gen_helper_1i(mftlo, t0, 2);
+            break;
+        case 9:
+            gen_helper_1i(mfthi, t0, 2);
+            break;
+        case 10:
+            gen_helper_1i(mftacx, t0, 2);
+            break;
+        case 12:
+            gen_helper_1i(mftlo, t0, 3);
+            break;
+        case 13:
+            gen_helper_1i(mfthi, t0, 3);
+            break;
+        case 14:
+            gen_helper_1i(mftacx, t0, 3);
+            break;
+        case 16:
+            gen_helper_mftdsp(t0);
+            break;
+        default:
+            goto die;
+        }
+        break;
+    /* Floating point (COP1). */
+    case 2:
+        /* XXX: For now we support only a single FPU context. */
+        if (h == 0) {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, rt);
+            tcg_gen_ext_i32_tl(t0, fp0);
+            tcg_temp_free_i32(fp0);
+        } else {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32h(fp0, rt);
+            tcg_gen_ext_i32_tl(t0, fp0);
+            tcg_temp_free_i32(fp0);
+        }
+        break;
+    case 3:
+        /* XXX: For now we support only a single FPU context. */
+        gen_helper_1i(cfc1, t0, rt);
+        break;
+    /* COP2: Not implemented. */
+    case 4:
+    case 5:
+        /* fall through */
+    default:
+        goto die;
+    }
+    LOG_DISAS("mftr (reg %d u %d sel %d h %d)\n", rt, u, sel, h);
+    gen_store_gpr(t0, rd);
+    tcg_temp_free(t0);
+    return;
+
+die:
+    tcg_temp_free(t0);
+    LOG_DISAS("mftr (reg %d u %d sel %d h %d)\n", rt, u, sel, h);
+    generate_exception(ctx, EXCP_RI);
+}
+
+static void gen_mttr(CPUState *env, DisasContext *ctx, int rd, int rt,
+                     int u, int sel, int h)
+{
+    int other_tc = env->CP0_VPEControl & (0xff << CP0VPECo_TargTC);
+    TCGv t0 = tcg_temp_local_new();
+
+    gen_load_gpr(t0, rt);
+    if ((env->CP0_VPEConf0 & (1 << CP0VPEC0_MVP)) == 0 &&
+        ((env->tcs[other_tc].CP0_TCBind & (0xf << CP0TCBd_CurVPE)) !=
+         (env->active_tc.CP0_TCBind & (0xf << CP0TCBd_CurVPE))))
+        /* NOP */ ;
+    else if ((env->CP0_VPEControl & (0xff << CP0VPECo_TargTC)) >
+             (env->mvp->CP0_MVPConf0 & (0xff << CP0MVPC0_PTC)))
+        /* NOP */ ;
+    else if (u == 0) {
+        switch (rd) {
+        case 2:
+            switch (sel) {
+            case 1:
+                gen_helper_mttc0_tcstatus(t0);
+                break;
+            case 2:
+                gen_helper_mttc0_tcbind(t0);
+                break;
+            case 3:
+                gen_helper_mttc0_tcrestart(t0);
+                break;
+            case 4:
+                gen_helper_mttc0_tchalt(t0);
+                break;
+            case 5:
+                gen_helper_mttc0_tccontext(t0);
+                break;
+            case 6:
+                gen_helper_mttc0_tcschedule(t0);
+                break;
+            case 7:
+                gen_helper_mttc0_tcschefback(t0);
+                break;
+            default:
+                gen_mtc0(env, ctx, t0, rd, sel);
+                break;
+            }
+            break;
+        case 10:
+            switch (sel) {
+            case 0:
+                gen_helper_mttc0_entryhi(t0);
+                break;
+            default:
+                gen_mtc0(env, ctx, t0, rd, sel);
+                break;
+            }
+        case 12:
+            switch (sel) {
+            case 0:
+                gen_helper_mttc0_status(t0);
+                break;
+            default:
+                gen_mtc0(env, ctx, t0, rd, sel);
+                break;
+            }
+        case 23:
+            switch (sel) {
+            case 0:
+                gen_helper_mttc0_debug(t0);
+                break;
+            default:
+                gen_mtc0(env, ctx, t0, rd, sel);
+                break;
+            }
+            break;
+        default:
+            gen_mtc0(env, ctx, t0, rd, sel);
+        }
+    } else switch (sel) {
+    /* GPR registers. */
+    case 0:
+        gen_helper_1i(mttgpr, t0, rd);
+        break;
+    /* Auxiliary CPU registers */
+    case 1:
+        switch (rd) {
+        case 0:
+            gen_helper_1i(mttlo, t0, 0);
+            break;
+        case 1:
+            gen_helper_1i(mtthi, t0, 0);
+            break;
+        case 2:
+            gen_helper_1i(mttacx, t0, 0);
+            break;
+        case 4:
+            gen_helper_1i(mttlo, t0, 1);
+            break;
+        case 5:
+            gen_helper_1i(mtthi, t0, 1);
+            break;
+        case 6:
+            gen_helper_1i(mttacx, t0, 1);
+            break;
+        case 8:
+            gen_helper_1i(mttlo, t0, 2);
+            break;
+        case 9:
+            gen_helper_1i(mtthi, t0, 2);
+            break;
+        case 10:
+            gen_helper_1i(mttacx, t0, 2);
+            break;
+        case 12:
+            gen_helper_1i(mttlo, t0, 3);
+            break;
+        case 13:
+            gen_helper_1i(mtthi, t0, 3);
+            break;
+        case 14:
+            gen_helper_1i(mttacx, t0, 3);
+            break;
+        case 16:
+            gen_helper_mttdsp(t0);
+            break;
+        default:
+            goto die;
+        }
+        break;
+    /* Floating point (COP1). */
+    case 2:
+        /* XXX: For now we support only a single FPU context. */
+        if (h == 0) {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            tcg_gen_trunc_tl_i32(fp0, t0);
+            gen_store_fpr32(fp0, rd);
+            tcg_temp_free_i32(fp0);
+        } else {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            tcg_gen_trunc_tl_i32(fp0, t0);
+            gen_store_fpr32h(fp0, rd);
+            tcg_temp_free_i32(fp0);
+        }
+        break;
+    case 3:
+        /* XXX: For now we support only a single FPU context. */
+        gen_helper_1i(ctc1, t0, rd);
+        break;
+    /* COP2: Not implemented. */
+    case 4:
+    case 5:
+        /* fall through */
+    default:
+        goto die;
+    }
+    LOG_DISAS("mttr (reg %d u %d sel %d h %d)\n", rd, u, sel, h);
+    tcg_temp_free(t0);
+    return;
+
+die:
+    tcg_temp_free(t0);
+    LOG_DISAS("mttr (reg %d u %d sel %d h %d)\n", rd, u, sel, h);
+    generate_exception(ctx, EXCP_RI);
+}
+
+static void gen_cp0 (CPUState *env, DisasContext *ctx, uint32_t opc, int rt, int rd)
+{
+    const char *opn = "ldst";
+
+    switch (opc) {
+    case OPC_MFC0:
+        if (rt == 0) {
+            /* Treat as NOP. */
+            return;
+        }
+        gen_mfc0(env, ctx, cpu_gpr[rt], rd, ctx->opcode & 0x7);
+        opn = "mfc0";
+        break;
+    case OPC_MTC0:
+        {
+            TCGv t0 = tcg_temp_new();
+
+            gen_load_gpr(t0, rt);
+            gen_mtc0(env, ctx, t0, rd, ctx->opcode & 0x7);
+            tcg_temp_free(t0);
+        }
+        opn = "mtc0";
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DMFC0:
+        check_insn(env, ctx, ISA_MIPS3);
+        if (rt == 0) {
+            /* Treat as NOP. */
+            return;
+        }
+        gen_dmfc0(env, ctx, cpu_gpr[rt], rd, ctx->opcode & 0x7);
+        opn = "dmfc0";
+        break;
+    case OPC_DMTC0:
+        check_insn(env, ctx, ISA_MIPS3);
+        {
+            TCGv t0 = tcg_temp_new();
+
+            gen_load_gpr(t0, rt);
+            gen_dmtc0(env, ctx, t0, rd, ctx->opcode & 0x7);
+            tcg_temp_free(t0);
+        }
+        opn = "dmtc0";
+        break;
+#endif
+    case OPC_MFTR:
+        check_insn(env, ctx, ASE_MT);
+        if (rd == 0) {
+            /* Treat as NOP. */
+            return;
+        }
+        gen_mftr(env, ctx, rt, rd, (ctx->opcode >> 5) & 1,
+                 ctx->opcode & 0x7, (ctx->opcode >> 4) & 1);
+        opn = "mftr";
+        break;
+    case OPC_MTTR:
+        check_insn(env, ctx, ASE_MT);
+        gen_mttr(env, ctx, rd, rt, (ctx->opcode >> 5) & 1,
+                 ctx->opcode & 0x7, (ctx->opcode >> 4) & 1);
+        opn = "mttr";
+        break;
+    case OPC_TLBWI:
+        opn = "tlbwi";
+        if (!env->tlb->helper_tlbwi)
+            goto die;
+        gen_helper_tlbwi();
+        break;
+    case OPC_TLBWR:
+        opn = "tlbwr";
+        if (!env->tlb->helper_tlbwr)
+            goto die;
+        gen_helper_tlbwr();
+        break;
+    case OPC_TLBP:
+        opn = "tlbp";
+        if (!env->tlb->helper_tlbp)
+            goto die;
+        gen_helper_tlbp();
+        break;
+    case OPC_TLBR:
+        opn = "tlbr";
+        if (!env->tlb->helper_tlbr)
+            goto die;
+        gen_helper_tlbr();
+        break;
+    case OPC_ERET:
+        opn = "eret";
+        check_insn(env, ctx, ISA_MIPS2);
+        gen_helper_eret();
+        ctx->bstate = BS_EXCP;
+        break;
+    case OPC_DERET:
+        opn = "deret";
+        check_insn(env, ctx, ISA_MIPS32);
+        if (!(ctx->hflags & MIPS_HFLAG_DM)) {
+            MIPS_INVAL(opn);
+            generate_exception(ctx, EXCP_RI);
+        } else {
+            gen_helper_deret();
+            ctx->bstate = BS_EXCP;
+        }
+        break;
+    case OPC_WAIT:
+        opn = "wait";
+        check_insn(env, ctx, ISA_MIPS3 | ISA_MIPS32);
+        /* If we get an exception, we want to restart at next instruction */
+        ctx->pc += 4;
+        save_cpu_state(ctx, 1);
+        ctx->pc -= 4;
+        gen_helper_wait();
+        ctx->bstate = BS_EXCP;
+        break;
+    default:
+ die:
+        MIPS_INVAL(opn);
+        generate_exception(ctx, EXCP_RI);
+        return;
+    }
+    MIPS_DEBUG("%s %s %d", opn, regnames[rt], rd);
+}
+#endif /* !CONFIG_USER_ONLY */
+
+/* CP1 Branches (before delay slot) */
+static void gen_compute_branch1 (CPUState *env, DisasContext *ctx, uint32_t op,
+                                 int32_t cc, int32_t offset)
+{
+    target_ulong btarget;
+    const char *opn = "cp1 cond branch";
+    TCGv_i32 t0 = tcg_temp_new_i32();
+
+    if (cc != 0)
+        check_insn(env, ctx, ISA_MIPS4 | ISA_MIPS32);
+
+    btarget = ctx->pc + 4 + offset;
+
+    switch (op) {
+    case OPC_BC1F:
+        tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc));
+        tcg_gen_not_i32(t0, t0);
+        tcg_gen_andi_i32(t0, t0, 1);
+        tcg_gen_extu_i32_tl(bcond, t0);
+        opn = "bc1f";
+        goto not_likely;
+    case OPC_BC1FL:
+        tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc));
+        tcg_gen_not_i32(t0, t0);
+        tcg_gen_andi_i32(t0, t0, 1);
+        tcg_gen_extu_i32_tl(bcond, t0);
+        opn = "bc1fl";
+        goto likely;
+    case OPC_BC1T:
+        tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc));
+        tcg_gen_andi_i32(t0, t0, 1);
+        tcg_gen_extu_i32_tl(bcond, t0);
+        opn = "bc1t";
+        goto not_likely;
+    case OPC_BC1TL:
+        tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc));
+        tcg_gen_andi_i32(t0, t0, 1);
+        tcg_gen_extu_i32_tl(bcond, t0);
+        opn = "bc1tl";
+    likely:
+        ctx->hflags |= MIPS_HFLAG_BL;
+        break;
+    case OPC_BC1FANY2:
+        {
+            TCGv_i32 t1 = tcg_temp_new_i32();
+            tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc));
+            tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1));
+            tcg_gen_or_i32(t0, t0, t1);
+            tcg_temp_free_i32(t1);
+            tcg_gen_not_i32(t0, t0);
+            tcg_gen_andi_i32(t0, t0, 1);
+            tcg_gen_extu_i32_tl(bcond, t0);
+        }
+        opn = "bc1any2f";
+        goto not_likely;
+    case OPC_BC1TANY2:
+        {
+            TCGv_i32 t1 = tcg_temp_new_i32();
+            tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc));
+            tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1));
+            tcg_gen_or_i32(t0, t0, t1);
+            tcg_temp_free_i32(t1);
+            tcg_gen_andi_i32(t0, t0, 1);
+            tcg_gen_extu_i32_tl(bcond, t0);
+        }
+        opn = "bc1any2t";
+        goto not_likely;
+    case OPC_BC1FANY4:
+        {
+            TCGv_i32 t1 = tcg_temp_new_i32();
+            tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc));
+            tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1));
+            tcg_gen_or_i32(t0, t0, t1);
+            tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+2));
+            tcg_gen_or_i32(t0, t0, t1);
+            tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+3));
+            tcg_gen_or_i32(t0, t0, t1);
+            tcg_temp_free_i32(t1);
+            tcg_gen_not_i32(t0, t0);
+            tcg_gen_andi_i32(t0, t0, 1);
+            tcg_gen_extu_i32_tl(bcond, t0);
+        }
+        opn = "bc1any4f";
+        goto not_likely;
+    case OPC_BC1TANY4:
+        {
+            TCGv_i32 t1 = tcg_temp_new_i32();
+            tcg_gen_shri_i32(t0, fpu_fcr31, get_fp_bit(cc));
+            tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+1));
+            tcg_gen_or_i32(t0, t0, t1);
+            tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+2));
+            tcg_gen_or_i32(t0, t0, t1);
+            tcg_gen_shri_i32(t1, fpu_fcr31, get_fp_bit(cc+3));
+            tcg_gen_or_i32(t0, t0, t1);
+            tcg_temp_free_i32(t1);
+            tcg_gen_andi_i32(t0, t0, 1);
+            tcg_gen_extu_i32_tl(bcond, t0);
+        }
+        opn = "bc1any4t";
+    not_likely:
+        ctx->hflags |= MIPS_HFLAG_BC;
+        break;
+    default:
+        MIPS_INVAL(opn);
+        generate_exception (ctx, EXCP_RI);
+        goto out;
+    }
+    MIPS_DEBUG("%s: cond %02x target " TARGET_FMT_lx, opn,
+               ctx->hflags, btarget);
+    ctx->btarget = btarget;
+
+ out:
+    tcg_temp_free_i32(t0);
+}
+
+/* Coprocessor 1 (FPU) */
+
+#define FOP(func, fmt) (((fmt) << 21) | (func))
+
+static void gen_cp1 (DisasContext *ctx, uint32_t opc, int rt, int fs)
+{
+    const char *opn = "cp1 move";
+    TCGv t0 = tcg_temp_new();
+
+    switch (opc) {
+    case OPC_MFC1:
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            tcg_gen_ext_i32_tl(t0, fp0);
+            tcg_temp_free_i32(fp0);
+        }
+        gen_store_gpr(t0, rt);
+        opn = "mfc1";
+        break;
+    case OPC_MTC1:
+        gen_load_gpr(t0, rt);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            tcg_gen_trunc_tl_i32(fp0, t0);
+            gen_store_fpr32(fp0, fs);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "mtc1";
+        break;
+    case OPC_CFC1:
+        gen_helper_1i(cfc1, t0, fs);
+        gen_store_gpr(t0, rt);
+        opn = "cfc1";
+        break;
+    case OPC_CTC1:
+        gen_load_gpr(t0, rt);
+        gen_helper_1i(ctc1, t0, fs);
+        opn = "ctc1";
+        break;
+#if defined(TARGET_MIPS64)
+    case OPC_DMFC1:
+        gen_load_fpr64(ctx, t0, fs);
+        gen_store_gpr(t0, rt);
+        opn = "dmfc1";
+        break;
+    case OPC_DMTC1:
+        gen_load_gpr(t0, rt);
+        gen_store_fpr64(ctx, t0, fs);
+        opn = "dmtc1";
+        break;
+#endif
+    case OPC_MFHC1:
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32h(fp0, fs);
+            tcg_gen_ext_i32_tl(t0, fp0);
+            tcg_temp_free_i32(fp0);
+        }
+        gen_store_gpr(t0, rt);
+        opn = "mfhc1";
+        break;
+    case OPC_MTHC1:
+        gen_load_gpr(t0, rt);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            tcg_gen_trunc_tl_i32(fp0, t0);
+            gen_store_fpr32h(fp0, fs);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "mthc1";
+        break;
+    default:
+        MIPS_INVAL(opn);
+        generate_exception (ctx, EXCP_RI);
+        goto out;
+    }
+    MIPS_DEBUG("%s %s %s", opn, regnames[rt], fregnames[fs]);
+
+ out:
+    tcg_temp_free(t0);
+}
+
+static void gen_movci (DisasContext *ctx, int rd, int rs, int cc, int tf)
+{
+    int l1;
+    TCGCond cond;
+    TCGv_i32 t0;
+
+    if (rd == 0) {
+        /* Treat as NOP. */
+        return;
+    }
+
+    if (tf)
+        cond = TCG_COND_EQ;
+    else
+        cond = TCG_COND_NE;
+
+    l1 = gen_new_label();
+    t0 = tcg_temp_new_i32();
+    tcg_gen_andi_i32(t0, fpu_fcr31, 1 << get_fp_bit(cc));
+    tcg_gen_brcondi_i32(cond, t0, 0, l1);
+    tcg_temp_free_i32(t0);
+    if (rs == 0) {
+        tcg_gen_movi_tl(cpu_gpr[rd], 0);
+    } else {
+        tcg_gen_mov_tl(cpu_gpr[rd], cpu_gpr[rs]);
+    }
+    gen_set_label(l1);
+}
+
+static inline void gen_movcf_s (int fs, int fd, int cc, int tf)
+{
+    int cond;
+    TCGv_i32 t0 = tcg_temp_new_i32();
+    int l1 = gen_new_label();
+
+    if (tf)
+        cond = TCG_COND_EQ;
+    else
+        cond = TCG_COND_NE;
+
+    tcg_gen_andi_i32(t0, fpu_fcr31, 1 << get_fp_bit(cc));
+    tcg_gen_brcondi_i32(cond, t0, 0, l1);
+    gen_load_fpr32(t0, fs);
+    gen_store_fpr32(t0, fd);
+    gen_set_label(l1);
+    tcg_temp_free_i32(t0);
+}
+
+static inline void gen_movcf_d (DisasContext *ctx, int fs, int fd, int cc, int tf)
+{
+    int cond;
+    TCGv_i32 t0 = tcg_temp_new_i32();
+    TCGv_i64 fp0;
+    int l1 = gen_new_label();
+
+    if (tf)
+        cond = TCG_COND_EQ;
+    else
+        cond = TCG_COND_NE;
+
+    tcg_gen_andi_i32(t0, fpu_fcr31, 1 << get_fp_bit(cc));
+    tcg_gen_brcondi_i32(cond, t0, 0, l1);
+    tcg_temp_free_i32(t0);
+    fp0 = tcg_temp_new_i64();
+    gen_load_fpr64(ctx, fp0, fs);
+    gen_store_fpr64(ctx, fp0, fd);
+    tcg_temp_free_i64(fp0);
+    gen_set_label(l1);
+}
+
+static inline void gen_movcf_ps (int fs, int fd, int cc, int tf)
+{
+    int cond;
+    TCGv_i32 t0 = tcg_temp_new_i32();
+    int l1 = gen_new_label();
+    int l2 = gen_new_label();
+
+    if (tf)
+        cond = TCG_COND_EQ;
+    else
+        cond = TCG_COND_NE;
+
+    tcg_gen_andi_i32(t0, fpu_fcr31, 1 << get_fp_bit(cc));
+    tcg_gen_brcondi_i32(cond, t0, 0, l1);
+    gen_load_fpr32(t0, fs);
+    gen_store_fpr32(t0, fd);
+    gen_set_label(l1);
+
+    tcg_gen_andi_i32(t0, fpu_fcr31, 1 << get_fp_bit(cc+1));
+    tcg_gen_brcondi_i32(cond, t0, 0, l2);
+    gen_load_fpr32h(t0, fs);
+    gen_store_fpr32h(t0, fd);
+    tcg_temp_free_i32(t0);
+    gen_set_label(l2);
+}
+
+
+static void gen_farith (DisasContext *ctx, uint32_t op1,
+                        int ft, int fs, int fd, int cc)
+{
+    const char *opn = "farith";
+    const char *condnames[] = {
+            "c.f",
+            "c.un",
+            "c.eq",
+            "c.ueq",
+            "c.olt",
+            "c.ult",
+            "c.ole",
+            "c.ule",
+            "c.sf",
+            "c.ngle",
+            "c.seq",
+            "c.ngl",
+            "c.lt",
+            "c.nge",
+            "c.le",
+            "c.ngt",
+    };
+    const char *condnames_abs[] = {
+            "cabs.f",
+            "cabs.un",
+            "cabs.eq",
+            "cabs.ueq",
+            "cabs.olt",
+            "cabs.ult",
+            "cabs.ole",
+            "cabs.ule",
+            "cabs.sf",
+            "cabs.ngle",
+            "cabs.seq",
+            "cabs.ngl",
+            "cabs.lt",
+            "cabs.nge",
+            "cabs.le",
+            "cabs.ngt",
+    };
+    enum { BINOP, CMPOP, OTHEROP } optype = OTHEROP;
+    uint32_t func = ctx->opcode & 0x3f;
+
+    switch (ctx->opcode & FOP(0x3f, 0x1f)) {
+    case FOP(0, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_helper_float_add_s(fp0, fp0, fp1);
+            tcg_temp_free_i32(fp1);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "add.s";
+        optype = BINOP;
+        break;
+    case FOP(1, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_helper_float_sub_s(fp0, fp0, fp1);
+            tcg_temp_free_i32(fp1);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "sub.s";
+        optype = BINOP;
+        break;
+    case FOP(2, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_helper_float_mul_s(fp0, fp0, fp1);
+            tcg_temp_free_i32(fp1);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "mul.s";
+        optype = BINOP;
+        break;
+    case FOP(3, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_helper_float_div_s(fp0, fp0, fp1);
+            tcg_temp_free_i32(fp1);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "div.s";
+        optype = BINOP;
+        break;
+    case FOP(4, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_sqrt_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "sqrt.s";
+        break;
+    case FOP(5, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_abs_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "abs.s";
+        break;
+    case FOP(6, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "mov.s";
+        break;
+    case FOP(7, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_chs_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "neg.s";
+        break;
+    case FOP(8, 16):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr32(fp32, fs);
+            gen_helper_float_roundl_s(fp64, fp32);
+            tcg_temp_free_i32(fp32);
+            gen_store_fpr64(ctx, fp64, fd);
+            tcg_temp_free_i64(fp64);
+        }
+        opn = "round.l.s";
+        break;
+    case FOP(9, 16):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr32(fp32, fs);
+            gen_helper_float_truncl_s(fp64, fp32);
+            tcg_temp_free_i32(fp32);
+            gen_store_fpr64(ctx, fp64, fd);
+            tcg_temp_free_i64(fp64);
+        }
+        opn = "trunc.l.s";
+        break;
+    case FOP(10, 16):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr32(fp32, fs);
+            gen_helper_float_ceill_s(fp64, fp32);
+            tcg_temp_free_i32(fp32);
+            gen_store_fpr64(ctx, fp64, fd);
+            tcg_temp_free_i64(fp64);
+        }
+        opn = "ceil.l.s";
+        break;
+    case FOP(11, 16):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr32(fp32, fs);
+            gen_helper_float_floorl_s(fp64, fp32);
+            tcg_temp_free_i32(fp32);
+            gen_store_fpr64(ctx, fp64, fd);
+            tcg_temp_free_i64(fp64);
+        }
+        opn = "floor.l.s";
+        break;
+    case FOP(12, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_roundw_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "round.w.s";
+        break;
+    case FOP(13, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_truncw_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "trunc.w.s";
+        break;
+    case FOP(14, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_ceilw_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "ceil.w.s";
+        break;
+    case FOP(15, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_floorw_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "floor.w.s";
+        break;
+    case FOP(17, 16):
+        gen_movcf_s(fs, fd, (ft >> 2) & 0x7, ft & 0x1);
+        opn = "movcf.s";
+        break;
+    case FOP(18, 16):
+        {
+            int l1 = gen_new_label();
+            TCGv_i32 fp0;
+
+            if (ft != 0) {
+                tcg_gen_brcondi_tl(TCG_COND_NE, cpu_gpr[ft], 0, l1);
+            }
+            fp0 = tcg_temp_new_i32();
+            gen_load_fpr32(fp0, fs);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+            gen_set_label(l1);
+        }
+        opn = "movz.s";
+        break;
+    case FOP(19, 16):
+        {
+            int l1 = gen_new_label();
+            TCGv_i32 fp0;
+
+            if (ft != 0) {
+                tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_gpr[ft], 0, l1);
+                fp0 = tcg_temp_new_i32();
+                gen_load_fpr32(fp0, fs);
+                gen_store_fpr32(fp0, fd);
+                tcg_temp_free_i32(fp0);
+                gen_set_label(l1);
+            }
+        }
+        opn = "movn.s";
+        break;
+    case FOP(21, 16):
+        check_cop1x(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_recip_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "recip.s";
+        break;
+    case FOP(22, 16):
+        check_cop1x(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_rsqrt_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "rsqrt.s";
+        break;
+    case FOP(28, 16):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, fd);
+            gen_helper_float_recip2_s(fp0, fp0, fp1);
+            tcg_temp_free_i32(fp1);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "recip2.s";
+        break;
+    case FOP(29, 16):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_recip1_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "recip1.s";
+        break;
+    case FOP(30, 16):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_rsqrt1_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "rsqrt1.s";
+        break;
+    case FOP(31, 16):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_helper_float_rsqrt2_s(fp0, fp0, fp1);
+            tcg_temp_free_i32(fp1);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "rsqrt2.s";
+        break;
+    case FOP(33, 16):
+        check_cp1_registers(ctx, fd);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr32(fp32, fs);
+            gen_helper_float_cvtd_s(fp64, fp32);
+            tcg_temp_free_i32(fp32);
+            gen_store_fpr64(ctx, fp64, fd);
+            tcg_temp_free_i64(fp64);
+        }
+        opn = "cvt.d.s";
+        break;
+    case FOP(36, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_cvtw_s(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "cvt.w.s";
+        break;
+    case FOP(37, 16):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr32(fp32, fs);
+            gen_helper_float_cvtl_s(fp64, fp32);
+            tcg_temp_free_i32(fp32);
+            gen_store_fpr64(ctx, fp64, fd);
+            tcg_temp_free_i64(fp64);
+        }
+        opn = "cvt.l.s";
+        break;
+    case FOP(38, 16):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+            TCGv_i32 fp32_0 = tcg_temp_new_i32();
+            TCGv_i32 fp32_1 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp32_0, fs);
+            gen_load_fpr32(fp32_1, ft);
+            tcg_gen_concat_i32_i64(fp64, fp32_0, fp32_1);
+            tcg_temp_free_i32(fp32_1);
+            tcg_temp_free_i32(fp32_0);
+            gen_store_fpr64(ctx, fp64, fd);
+            tcg_temp_free_i64(fp64);
+        }
+        opn = "cvt.ps.s";
+        break;
+    case FOP(48, 16):
+    case FOP(49, 16):
+    case FOP(50, 16):
+    case FOP(51, 16):
+    case FOP(52, 16):
+    case FOP(53, 16):
+    case FOP(54, 16):
+    case FOP(55, 16):
+    case FOP(56, 16):
+    case FOP(57, 16):
+    case FOP(58, 16):
+    case FOP(59, 16):
+    case FOP(60, 16):
+    case FOP(61, 16):
+    case FOP(62, 16):
+    case FOP(63, 16):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            if (ctx->opcode & (1 << 6)) {
+                check_cop1x(ctx);
+                gen_cmpabs_s(func-48, fp0, fp1, cc);
+                opn = condnames_abs[func-48];
+            } else {
+                gen_cmp_s(func-48, fp0, fp1, cc);
+                opn = condnames[func-48];
+            }
+            tcg_temp_free_i32(fp0);
+            tcg_temp_free_i32(fp1);
+        }
+        break;
+    case FOP(0, 17):
+        check_cp1_registers(ctx, fs | ft | fd);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_helper_float_add_d(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "add.d";
+        optype = BINOP;
+        break;
+    case FOP(1, 17):
+        check_cp1_registers(ctx, fs | ft | fd);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_helper_float_sub_d(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "sub.d";
+        optype = BINOP;
+        break;
+    case FOP(2, 17):
+        check_cp1_registers(ctx, fs | ft | fd);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_helper_float_mul_d(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "mul.d";
+        optype = BINOP;
+        break;
+    case FOP(3, 17):
+        check_cp1_registers(ctx, fs | ft | fd);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_helper_float_div_d(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "div.d";
+        optype = BINOP;
+        break;
+    case FOP(4, 17):
+        check_cp1_registers(ctx, fs | fd);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_sqrt_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "sqrt.d";
+        break;
+    case FOP(5, 17):
+        check_cp1_registers(ctx, fs | fd);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_abs_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "abs.d";
+        break;
+    case FOP(6, 17):
+        check_cp1_registers(ctx, fs | fd);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "mov.d";
+        break;
+    case FOP(7, 17):
+        check_cp1_registers(ctx, fs | fd);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_chs_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "neg.d";
+        break;
+    case FOP(8, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_roundl_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "round.l.d";
+        break;
+    case FOP(9, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_truncl_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "trunc.l.d";
+        break;
+    case FOP(10, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_ceill_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "ceil.l.d";
+        break;
+    case FOP(11, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_floorl_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "floor.l.d";
+        break;
+    case FOP(12, 17):
+        check_cp1_registers(ctx, fs);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp64, fs);
+            gen_helper_float_roundw_d(fp32, fp64);
+            tcg_temp_free_i64(fp64);
+            gen_store_fpr32(fp32, fd);
+            tcg_temp_free_i32(fp32);
+        }
+        opn = "round.w.d";
+        break;
+    case FOP(13, 17):
+        check_cp1_registers(ctx, fs);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp64, fs);
+            gen_helper_float_truncw_d(fp32, fp64);
+            tcg_temp_free_i64(fp64);
+            gen_store_fpr32(fp32, fd);
+            tcg_temp_free_i32(fp32);
+        }
+        opn = "trunc.w.d";
+        break;
+    case FOP(14, 17):
+        check_cp1_registers(ctx, fs);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp64, fs);
+            gen_helper_float_ceilw_d(fp32, fp64);
+            tcg_temp_free_i64(fp64);
+            gen_store_fpr32(fp32, fd);
+            tcg_temp_free_i32(fp32);
+        }
+        opn = "ceil.w.d";
+        break;
+    case FOP(15, 17):
+        check_cp1_registers(ctx, fs);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp64, fs);
+            gen_helper_float_floorw_d(fp32, fp64);
+            tcg_temp_free_i64(fp64);
+            gen_store_fpr32(fp32, fd);
+            tcg_temp_free_i32(fp32);
+        }
+        opn = "floor.w.d";
+        break;
+    case FOP(17, 17):
+        gen_movcf_d(ctx, fs, fd, (ft >> 2) & 0x7, ft & 0x1);
+        opn = "movcf.d";
+        break;
+    case FOP(18, 17):
+        {
+            int l1 = gen_new_label();
+            TCGv_i64 fp0;
+
+            if (ft != 0) {
+                tcg_gen_brcondi_tl(TCG_COND_NE, cpu_gpr[ft], 0, l1);
+            }
+            fp0 = tcg_temp_new_i64();
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+            gen_set_label(l1);
+        }
+        opn = "movz.d";
+        break;
+    case FOP(19, 17):
+        {
+            int l1 = gen_new_label();
+            TCGv_i64 fp0;
+
+            if (ft != 0) {
+                tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_gpr[ft], 0, l1);
+                fp0 = tcg_temp_new_i64();
+                gen_load_fpr64(ctx, fp0, fs);
+                gen_store_fpr64(ctx, fp0, fd);
+                tcg_temp_free_i64(fp0);
+                gen_set_label(l1);
+            }
+        }
+        opn = "movn.d";
+        break;
+    case FOP(21, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_recip_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "recip.d";
+        break;
+    case FOP(22, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_rsqrt_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "rsqrt.d";
+        break;
+    case FOP(28, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_helper_float_recip2_d(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "recip2.d";
+        break;
+    case FOP(29, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_recip1_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "recip1.d";
+        break;
+    case FOP(30, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_rsqrt1_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "rsqrt1.d";
+        break;
+    case FOP(31, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_helper_float_rsqrt2_d(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "rsqrt2.d";
+        break;
+    case FOP(48, 17):
+    case FOP(49, 17):
+    case FOP(50, 17):
+    case FOP(51, 17):
+    case FOP(52, 17):
+    case FOP(53, 17):
+    case FOP(54, 17):
+    case FOP(55, 17):
+    case FOP(56, 17):
+    case FOP(57, 17):
+    case FOP(58, 17):
+    case FOP(59, 17):
+    case FOP(60, 17):
+    case FOP(61, 17):
+    case FOP(62, 17):
+    case FOP(63, 17):
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            if (ctx->opcode & (1 << 6)) {
+                check_cop1x(ctx);
+                check_cp1_registers(ctx, fs | ft);
+                gen_cmpabs_d(func-48, fp0, fp1, cc);
+                opn = condnames_abs[func-48];
+            } else {
+                check_cp1_registers(ctx, fs | ft);
+                gen_cmp_d(func-48, fp0, fp1, cc);
+                opn = condnames[func-48];
+            }
+            tcg_temp_free_i64(fp0);
+            tcg_temp_free_i64(fp1);
+        }
+        break;
+    case FOP(32, 17):
+        check_cp1_registers(ctx, fs);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp64, fs);
+            gen_helper_float_cvts_d(fp32, fp64);
+            tcg_temp_free_i64(fp64);
+            gen_store_fpr32(fp32, fd);
+            tcg_temp_free_i32(fp32);
+        }
+        opn = "cvt.s.d";
+        break;
+    case FOP(36, 17):
+        check_cp1_registers(ctx, fs);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp64, fs);
+            gen_helper_float_cvtw_d(fp32, fp64);
+            tcg_temp_free_i64(fp64);
+            gen_store_fpr32(fp32, fd);
+            tcg_temp_free_i32(fp32);
+        }
+        opn = "cvt.w.d";
+        break;
+    case FOP(37, 17):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_cvtl_d(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "cvt.l.d";
+        break;
+    case FOP(32, 20):
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_cvts_w(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "cvt.s.w";
+        break;
+    case FOP(33, 20):
+        check_cp1_registers(ctx, fd);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr32(fp32, fs);
+            gen_helper_float_cvtd_w(fp64, fp32);
+            tcg_temp_free_i32(fp32);
+            gen_store_fpr64(ctx, fp64, fd);
+            tcg_temp_free_i64(fp64);
+        }
+        opn = "cvt.d.w";
+        break;
+    case FOP(32, 21):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp32 = tcg_temp_new_i32();
+            TCGv_i64 fp64 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp64, fs);
+            gen_helper_float_cvts_l(fp32, fp64);
+            tcg_temp_free_i64(fp64);
+            gen_store_fpr32(fp32, fd);
+            tcg_temp_free_i32(fp32);
+        }
+        opn = "cvt.s.l";
+        break;
+    case FOP(33, 21):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_cvtd_l(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "cvt.d.l";
+        break;
+    case FOP(38, 20):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_cvtps_pw(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "cvt.ps.pw";
+        break;
+    case FOP(0, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_helper_float_add_ps(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "add.ps";
+        break;
+    case FOP(1, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_helper_float_sub_ps(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "sub.ps";
+        break;
+    case FOP(2, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_helper_float_mul_ps(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "mul.ps";
+        break;
+    case FOP(5, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_abs_ps(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "abs.ps";
+        break;
+    case FOP(6, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "mov.ps";
+        break;
+    case FOP(7, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_chs_ps(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "neg.ps";
+        break;
+    case FOP(17, 22):
+        check_cp1_64bitmode(ctx);
+        gen_movcf_ps(fs, fd, (ft >> 2) & 0x7, ft & 0x1);
+        opn = "movcf.ps";
+        break;
+    case FOP(18, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            int l1 = gen_new_label();
+            TCGv_i64 fp0;
+
+            if (ft != 0)
+                tcg_gen_brcondi_tl(TCG_COND_NE, cpu_gpr[ft], 0, l1);
+            fp0 = tcg_temp_new_i64();
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+            gen_set_label(l1);
+        }
+        opn = "movz.ps";
+        break;
+    case FOP(19, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            int l1 = gen_new_label();
+            TCGv_i64 fp0;
+
+            if (ft != 0) {
+                tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_gpr[ft], 0, l1);
+                fp0 = tcg_temp_new_i64();
+                gen_load_fpr64(ctx, fp0, fs);
+                gen_store_fpr64(ctx, fp0, fd);
+                tcg_temp_free_i64(fp0);
+                gen_set_label(l1);
+            }
+        }
+        opn = "movn.ps";
+        break;
+    case FOP(24, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, ft);
+            gen_load_fpr64(ctx, fp1, fs);
+            gen_helper_float_addr_ps(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "addr.ps";
+        break;
+    case FOP(26, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, ft);
+            gen_load_fpr64(ctx, fp1, fs);
+            gen_helper_float_mulr_ps(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "mulr.ps";
+        break;
+    case FOP(28, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, fd);
+            gen_helper_float_recip2_ps(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "recip2.ps";
+        break;
+    case FOP(29, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_recip1_ps(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "recip1.ps";
+        break;
+    case FOP(30, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_rsqrt1_ps(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "rsqrt1.ps";
+        break;
+    case FOP(31, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_helper_float_rsqrt2_ps(fp0, fp0, fp1);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "rsqrt2.ps";
+        break;
+    case FOP(32, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32h(fp0, fs);
+            gen_helper_float_cvts_pu(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "cvt.s.pu";
+        break;
+    case FOP(36, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_helper_float_cvtpw_ps(fp0, fp0);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "cvt.pw.ps";
+        break;
+    case FOP(40, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_helper_float_cvts_pl(fp0, fp0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "cvt.s.pl";
+        break;
+    case FOP(44, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_store_fpr32h(fp0, fd);
+            gen_store_fpr32(fp1, fd);
+            tcg_temp_free_i32(fp0);
+            tcg_temp_free_i32(fp1);
+        }
+        opn = "pll.ps";
+        break;
+    case FOP(45, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32h(fp1, ft);
+            gen_store_fpr32(fp1, fd);
+            gen_store_fpr32h(fp0, fd);
+            tcg_temp_free_i32(fp0);
+            tcg_temp_free_i32(fp1);
+        }
+        opn = "plu.ps";
+        break;
+    case FOP(46, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32h(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_store_fpr32(fp1, fd);
+            gen_store_fpr32h(fp0, fd);
+            tcg_temp_free_i32(fp0);
+            tcg_temp_free_i32(fp1);
+        }
+        opn = "pul.ps";
+        break;
+    case FOP(47, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+
+            gen_load_fpr32h(fp0, fs);
+            gen_load_fpr32h(fp1, ft);
+            gen_store_fpr32(fp1, fd);
+            gen_store_fpr32h(fp0, fd);
+            tcg_temp_free_i32(fp0);
+            tcg_temp_free_i32(fp1);
+        }
+        opn = "puu.ps";
+        break;
+    case FOP(48, 22):
+    case FOP(49, 22):
+    case FOP(50, 22):
+    case FOP(51, 22):
+    case FOP(52, 22):
+    case FOP(53, 22):
+    case FOP(54, 22):
+    case FOP(55, 22):
+    case FOP(56, 22):
+    case FOP(57, 22):
+    case FOP(58, 22):
+    case FOP(59, 22):
+    case FOP(60, 22):
+    case FOP(61, 22):
+    case FOP(62, 22):
+    case FOP(63, 22):
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            if (ctx->opcode & (1 << 6)) {
+                gen_cmpabs_ps(func-48, fp0, fp1, cc);
+                opn = condnames_abs[func-48];
+            } else {
+                gen_cmp_ps(func-48, fp0, fp1, cc);
+                opn = condnames[func-48];
+            }
+            tcg_temp_free_i64(fp0);
+            tcg_temp_free_i64(fp1);
+        }
+        break;
+    default:
+        MIPS_INVAL(opn);
+        generate_exception (ctx, EXCP_RI);
+        return;
+    }
+    switch (optype) {
+    case BINOP:
+        MIPS_DEBUG("%s %s, %s, %s", opn, fregnames[fd], fregnames[fs], fregnames[ft]);
+        break;
+    case CMPOP:
+        MIPS_DEBUG("%s %s,%s", opn, fregnames[fs], fregnames[ft]);
+        break;
+    default:
+        MIPS_DEBUG("%s %s,%s", opn, fregnames[fd], fregnames[fs]);
+        break;
+    }
+}
+
+/* Coprocessor 3 (FPU) */
+static void gen_flt3_ldst (DisasContext *ctx, uint32_t opc,
+                           int fd, int fs, int base, int index)
+{
+    const char *opn = "extended float load/store";
+    int store = 0;
+    TCGv t0 = tcg_temp_new();
+
+    if (base == 0) {
+        gen_load_gpr(t0, index);
+    } else if (index == 0) {
+        gen_load_gpr(t0, base);
+    } else {
+        gen_load_gpr(t0, index);
+        gen_op_addr_add(ctx, t0, cpu_gpr[base], t0);
+    }
+    /* Don't do NOP if destination is zero: we must perform the actual
+       memory access. */
+    save_cpu_state(ctx, 0);
+    switch (opc) {
+    case OPC_LWXC1:
+        check_cop1x(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+
+            tcg_gen_qemu_ld32s(t0, t0, ctx->mem_idx);
+            tcg_gen_trunc_tl_i32(fp0, t0);
+            gen_store_fpr32(fp0, fd);
+            tcg_temp_free_i32(fp0);
+        }
+        opn = "lwxc1";
+        break;
+    case OPC_LDXC1:
+        check_cop1x(ctx);
+        check_cp1_registers(ctx, fd);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            tcg_gen_qemu_ld64(fp0, t0, ctx->mem_idx);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "ldxc1";
+        break;
+    case OPC_LUXC1:
+        check_cp1_64bitmode(ctx);
+        tcg_gen_andi_tl(t0, t0, ~0x7);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            tcg_gen_qemu_ld64(fp0, t0, ctx->mem_idx);
+            gen_store_fpr64(ctx, fp0, fd);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "luxc1";
+        break;
+    case OPC_SWXC1:
+        check_cop1x(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv t1 = tcg_temp_new();
+
+            gen_load_fpr32(fp0, fs);
+            tcg_gen_extu_i32_tl(t1, fp0);
+            tcg_gen_qemu_st32(t1, t0, ctx->mem_idx);
+            tcg_temp_free_i32(fp0);
+            tcg_temp_free(t1);
+        }
+        opn = "swxc1";
+        store = 1;
+        break;
+    case OPC_SDXC1:
+        check_cop1x(ctx);
+        check_cp1_registers(ctx, fs);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            tcg_gen_qemu_st64(fp0, t0, ctx->mem_idx);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "sdxc1";
+        store = 1;
+        break;
+    case OPC_SUXC1:
+        check_cp1_64bitmode(ctx);
+        tcg_gen_andi_tl(t0, t0, ~0x7);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            tcg_gen_qemu_st64(fp0, t0, ctx->mem_idx);
+            tcg_temp_free_i64(fp0);
+        }
+        opn = "suxc1";
+        store = 1;
+        break;
+    }
+    tcg_temp_free(t0);
+    MIPS_DEBUG("%s %s, %s(%s)", opn, fregnames[store ? fs : fd],
+               regnames[index], regnames[base]);
+}
+
+static void gen_flt3_arith (DisasContext *ctx, uint32_t opc,
+                            int fd, int fr, int fs, int ft)
+{
+    const char *opn = "flt3_arith";
+
+    switch (opc) {
+    case OPC_ALNV_PS:
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv t0 = tcg_temp_local_new();
+            TCGv_i32 fp = tcg_temp_new_i32();
+            TCGv_i32 fph = tcg_temp_new_i32();
+            int l1 = gen_new_label();
+            int l2 = gen_new_label();
+
+            gen_load_gpr(t0, fr);
+            tcg_gen_andi_tl(t0, t0, 0x7);
+
+            tcg_gen_brcondi_tl(TCG_COND_NE, t0, 0, l1);
+            gen_load_fpr32(fp, fs);
+            gen_load_fpr32h(fph, fs);
+            gen_store_fpr32(fp, fd);
+            gen_store_fpr32h(fph, fd);
+            tcg_gen_br(l2);
+            gen_set_label(l1);
+            tcg_gen_brcondi_tl(TCG_COND_NE, t0, 4, l2);
+            tcg_temp_free(t0);
+#ifdef TARGET_WORDS_BIGENDIAN
+            gen_load_fpr32(fp, fs);
+            gen_load_fpr32h(fph, ft);
+            gen_store_fpr32h(fp, fd);
+            gen_store_fpr32(fph, fd);
+#else
+            gen_load_fpr32h(fph, fs);
+            gen_load_fpr32(fp, ft);
+            gen_store_fpr32(fph, fd);
+            gen_store_fpr32h(fp, fd);
+#endif
+            gen_set_label(l2);
+            tcg_temp_free_i32(fp);
+            tcg_temp_free_i32(fph);
+        }
+        opn = "alnv.ps";
+        break;
+    case OPC_MADD_S:
+        check_cop1x(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+            TCGv_i32 fp2 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(fp2, fr);
+            gen_helper_float_muladd_s(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i32(fp0);
+            tcg_temp_free_i32(fp1);
+            gen_store_fpr32(fp2, fd);
+            tcg_temp_free_i32(fp2);
+        }
+        opn = "madd.s";
+        break;
+    case OPC_MADD_D:
+        check_cop1x(ctx);
+        check_cp1_registers(ctx, fd | fs | ft | fr);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+            TCGv_i64 fp2 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_load_fpr64(ctx, fp2, fr);
+            gen_helper_float_muladd_d(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i64(fp0);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp2, fd);
+            tcg_temp_free_i64(fp2);
+        }
+        opn = "madd.d";
+        break;
+    case OPC_MADD_PS:
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+            TCGv_i64 fp2 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_load_fpr64(ctx, fp2, fr);
+            gen_helper_float_muladd_ps(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i64(fp0);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp2, fd);
+            tcg_temp_free_i64(fp2);
+        }
+        opn = "madd.ps";
+        break;
+    case OPC_MSUB_S:
+        check_cop1x(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+            TCGv_i32 fp2 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(fp2, fr);
+            gen_helper_float_mulsub_s(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i32(fp0);
+            tcg_temp_free_i32(fp1);
+            gen_store_fpr32(fp2, fd);
+            tcg_temp_free_i32(fp2);
+        }
+        opn = "msub.s";
+        break;
+    case OPC_MSUB_D:
+        check_cop1x(ctx);
+        check_cp1_registers(ctx, fd | fs | ft | fr);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+            TCGv_i64 fp2 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_load_fpr64(ctx, fp2, fr);
+            gen_helper_float_mulsub_d(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i64(fp0);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp2, fd);
+            tcg_temp_free_i64(fp2);
+        }
+        opn = "msub.d";
+        break;
+    case OPC_MSUB_PS:
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+            TCGv_i64 fp2 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_load_fpr64(ctx, fp2, fr);
+            gen_helper_float_mulsub_ps(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i64(fp0);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp2, fd);
+            tcg_temp_free_i64(fp2);
+        }
+        opn = "msub.ps";
+        break;
+    case OPC_NMADD_S:
+        check_cop1x(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+            TCGv_i32 fp2 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(fp2, fr);
+            gen_helper_float_nmuladd_s(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i32(fp0);
+            tcg_temp_free_i32(fp1);
+            gen_store_fpr32(fp2, fd);
+            tcg_temp_free_i32(fp2);
+        }
+        opn = "nmadd.s";
+        break;
+    case OPC_NMADD_D:
+        check_cop1x(ctx);
+        check_cp1_registers(ctx, fd | fs | ft | fr);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+            TCGv_i64 fp2 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_load_fpr64(ctx, fp2, fr);
+            gen_helper_float_nmuladd_d(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i64(fp0);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp2, fd);
+            tcg_temp_free_i64(fp2);
+        }
+        opn = "nmadd.d";
+        break;
+    case OPC_NMADD_PS:
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+            TCGv_i64 fp2 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_load_fpr64(ctx, fp2, fr);
+            gen_helper_float_nmuladd_ps(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i64(fp0);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp2, fd);
+            tcg_temp_free_i64(fp2);
+        }
+        opn = "nmadd.ps";
+        break;
+    case OPC_NMSUB_S:
+        check_cop1x(ctx);
+        {
+            TCGv_i32 fp0 = tcg_temp_new_i32();
+            TCGv_i32 fp1 = tcg_temp_new_i32();
+            TCGv_i32 fp2 = tcg_temp_new_i32();
+
+            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(fp2, fr);
+            gen_helper_float_nmulsub_s(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i32(fp0);
+            tcg_temp_free_i32(fp1);
+            gen_store_fpr32(fp2, fd);
+            tcg_temp_free_i32(fp2);
+        }
+        opn = "nmsub.s";
+        break;
+    case OPC_NMSUB_D:
+        check_cop1x(ctx);
+        check_cp1_registers(ctx, fd | fs | ft | fr);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+            TCGv_i64 fp2 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_load_fpr64(ctx, fp2, fr);
+            gen_helper_float_nmulsub_d(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i64(fp0);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp2, fd);
+            tcg_temp_free_i64(fp2);
+        }
+        opn = "nmsub.d";
+        break;
+    case OPC_NMSUB_PS:
+        check_cp1_64bitmode(ctx);
+        {
+            TCGv_i64 fp0 = tcg_temp_new_i64();
+            TCGv_i64 fp1 = tcg_temp_new_i64();
+            TCGv_i64 fp2 = tcg_temp_new_i64();
+
+            gen_load_fpr64(ctx, fp0, fs);
+            gen_load_fpr64(ctx, fp1, ft);
+            gen_load_fpr64(ctx, fp2, fr);
+            gen_helper_float_nmulsub_ps(fp2, fp0, fp1, fp2);
+            tcg_temp_free_i64(fp0);
+            tcg_temp_free_i64(fp1);
+            gen_store_fpr64(ctx, fp2, fd);
+            tcg_temp_free_i64(fp2);
+        }
+        opn = "nmsub.ps";
+        break;
+    default:
+        MIPS_INVAL(opn);
+        generate_exception (ctx, EXCP_RI);
+        return;
+    }
+    MIPS_DEBUG("%s %s, %s, %s, %s", opn, fregnames[fd], fregnames[fr],
+               fregnames[fs], fregnames[ft]);
+}
+
+/* ISA extensions (ASEs) */
+/* MIPS16 extension to MIPS32 */
+/* SmartMIPS extension to MIPS32 */
+
+#if defined(TARGET_MIPS64)
+
+/* MDMX extension to MIPS64 */
+
+#endif
+
+static void decode_opc (CPUState *env, DisasContext *ctx)
+{
+    int32_t offset;
+    int rs, rt, rd, sa;
+    uint32_t op, op1, op2;
+    int16_t imm;
+
+    /* make sure instructions are on a word boundary */
+    if (ctx->pc & 0x3) {
+        env->CP0_BadVAddr = ctx->pc;
+        generate_exception(ctx, EXCP_AdEL);
+        return;
+    }
+
+    /* Handle blikely not taken case */
+    if ((ctx->hflags & MIPS_HFLAG_BMASK) == MIPS_HFLAG_BL) {
+        int l1 = gen_new_label();
+
+        MIPS_DEBUG("blikely condition (" TARGET_FMT_lx ")", ctx->pc + 4);
+        tcg_gen_brcondi_tl(TCG_COND_NE, bcond, 0, l1);
+        tcg_gen_movi_i32(hflags, ctx->hflags & ~MIPS_HFLAG_BMASK);
+        gen_goto_tb(ctx, 1, ctx->pc + 4);
+        gen_set_label(l1);
+    }
+
+    if (unlikely(qemu_loglevel_mask(CPU_LOG_TB_OP)))
+        tcg_gen_debug_insn_start(ctx->pc);
+
+    op = MASK_OP_MAJOR(ctx->opcode);
+    rs = (ctx->opcode >> 21) & 0x1f;
+    rt = (ctx->opcode >> 16) & 0x1f;
+    rd = (ctx->opcode >> 11) & 0x1f;
+    sa = (ctx->opcode >> 6) & 0x1f;
+    imm = (int16_t)ctx->opcode;
+    switch (op) {
+    case OPC_SPECIAL:
+        op1 = MASK_SPECIAL(ctx->opcode);
+        switch (op1) {
+        case OPC_SLL:          /* Shift with immediate */
+        case OPC_SRA:
+        case OPC_SRL:
+            gen_shift_imm(env, ctx, op1, rd, rt, sa);
+            break;
+        case OPC_MOVN:         /* Conditional move */
+        case OPC_MOVZ:
+            check_insn(env, ctx, ISA_MIPS4 | ISA_MIPS32);
+            gen_cond_move(env, op1, rd, rs, rt);
+            break;
+        case OPC_ADD ... OPC_SUBU:
+            gen_arith(env, ctx, op1, rd, rs, rt);
+            break;
+        case OPC_SLLV:         /* Shifts */
+        case OPC_SRLV:
+        case OPC_SRAV:
+            gen_shift(env, ctx, op1, rd, rs, rt);
+            break;
+        case OPC_SLT:          /* Set on less than */
+        case OPC_SLTU:
+            gen_slt(env, op1, rd, rs, rt);
+            break;
+        case OPC_AND:          /* Logic*/
+        case OPC_OR:
+        case OPC_NOR:
+        case OPC_XOR:
+            gen_logic(env, op1, rd, rs, rt);
+            break;
+        case OPC_MULT ... OPC_DIVU:
+            if (sa) {
+                check_insn(env, ctx, INSN_VR54XX);
+                op1 = MASK_MUL_VR54XX(ctx->opcode);
+                gen_mul_vr54xx(ctx, op1, rd, rs, rt);
+            } else
+                gen_muldiv(ctx, op1, rs, rt);
+            break;
+        case OPC_JR ... OPC_JALR:
+            gen_compute_branch(ctx, op1, rs, rd, sa);
+            return;
+        case OPC_TGE ... OPC_TEQ: /* Traps */
+        case OPC_TNE:
+            gen_trap(ctx, op1, rs, rt, -1);
+            break;
+        case OPC_MFHI:          /* Move from HI/LO */
+        case OPC_MFLO:
+            gen_HILO(ctx, op1, rd);
+            break;
+        case OPC_MTHI:
+        case OPC_MTLO:          /* Move to HI/LO */
+            gen_HILO(ctx, op1, rs);
+            break;
+        case OPC_PMON:          /* Pmon entry point, also R4010 selsl */
+#ifdef MIPS_STRICT_STANDARD
+            MIPS_INVAL("PMON / selsl");
+            generate_exception(ctx, EXCP_RI);
+#else
+            gen_helper_0i(pmon, sa);
+#endif
+            break;
+        case OPC_SYSCALL:
+            generate_exception(ctx, EXCP_SYSCALL);
+            ctx->bstate = BS_STOP;
+            break;
+        case OPC_BREAK:
+            generate_exception(ctx, EXCP_BREAK);
+            break;
+        case OPC_SPIM:
+#ifdef MIPS_STRICT_STANDARD
+            MIPS_INVAL("SPIM");
+            generate_exception(ctx, EXCP_RI);
+#else
+           /* Implemented as RI exception for now. */
+            MIPS_INVAL("spim (unofficial)");
+            generate_exception(ctx, EXCP_RI);
+#endif
+            break;
+        case OPC_SYNC:
+            /* Treat as NOP. */
+            break;
+
+        case OPC_MOVCI:
+            check_insn(env, ctx, ISA_MIPS4 | ISA_MIPS32);
+            if (env->CP0_Config1 & (1 << CP0C1_FP)) {
+                check_cp1_enabled(ctx);
+                gen_movci(ctx, rd, rs, (ctx->opcode >> 18) & 0x7,
+                          (ctx->opcode >> 16) & 1);
+            } else {
+                generate_exception_err(ctx, EXCP_CpU, 1);
+            }
+            break;
+
+#if defined(TARGET_MIPS64)
+       /* MIPS64 specific opcodes */
+        case OPC_DSLL:
+        case OPC_DSRA:
+        case OPC_DSRL:
+        case OPC_DSLL32:
+        case OPC_DSRA32:
+        case OPC_DSRL32:
+            check_insn(env, ctx, ISA_MIPS3);
+            check_mips_64(ctx);
+            gen_shift_imm(env, ctx, op1, rd, rt, sa);
+            break;
+        case OPC_DADD ... OPC_DSUBU:
+            check_insn(env, ctx, ISA_MIPS3);
+            check_mips_64(ctx);
+            gen_arith(env, ctx, op1, rd, rs, rt);
+            break;
+        case OPC_DSLLV:
+        case OPC_DSRAV:
+        case OPC_DSRLV:
+            check_insn(env, ctx, ISA_MIPS3);
+            check_mips_64(ctx);
+            gen_shift(env, ctx, op1, rd, rs, rt);
+            break;
+        case OPC_DMULT ... OPC_DDIVU:
+            check_insn(env, ctx, ISA_MIPS3);
+            check_mips_64(ctx);
+            gen_muldiv(ctx, op1, rs, rt);
+            break;
+#endif
+        default:            /* Invalid */
+            MIPS_INVAL("special");
+            generate_exception(ctx, EXCP_RI);
+            break;
+        }
+        break;
+    case OPC_SPECIAL2:
+        op1 = MASK_SPECIAL2(ctx->opcode);
+        switch (op1) {
+        case OPC_MADD ... OPC_MADDU: /* Multiply and add/sub */
+        case OPC_MSUB ... OPC_MSUBU:
+            check_insn(env, ctx, ISA_MIPS32);
+            gen_muldiv(ctx, op1, rs, rt);
+            break;
+        case OPC_MUL:
+            gen_arith(env, ctx, op1, rd, rs, rt);
+            break;
+        case OPC_CLO:
+        case OPC_CLZ:
+            check_insn(env, ctx, ISA_MIPS32);
+            gen_cl(ctx, op1, rd, rs);
+            break;
+        case OPC_SDBBP:
+            /* XXX: not clear which exception should be raised
+             *      when in debug mode...
+             */
+            check_insn(env, ctx, ISA_MIPS32);
+            if (!(ctx->hflags & MIPS_HFLAG_DM)) {
+                generate_exception(ctx, EXCP_DBp);
+            } else {
+                generate_exception(ctx, EXCP_DBp);
+            }
+            /* Treat as NOP. */
+            break;
+#if defined(TARGET_MIPS64)
+        case OPC_DCLO:
+        case OPC_DCLZ:
+            check_insn(env, ctx, ISA_MIPS64);
+            check_mips_64(ctx);
+            gen_cl(ctx, op1, rd, rs);
+            break;
+#endif
+        default:            /* Invalid */
+            MIPS_INVAL("special2");
+            generate_exception(ctx, EXCP_RI);
+            break;
+        }
+        break;
+    case OPC_SPECIAL3:
+        op1 = MASK_SPECIAL3(ctx->opcode);
+        switch (op1) {
+        case OPC_EXT:
+        case OPC_INS:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_bitops(ctx, op1, rt, rs, sa, rd);
+            break;
+        case OPC_BSHFL:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            op2 = MASK_BSHFL(ctx->opcode);
+            gen_bshfl(ctx, op2, rt, rd);
+            break;
+        case OPC_RDHWR:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            {
+                TCGv t0 = tcg_temp_new();
+
+                switch (rd) {
+                case 0:
+                    save_cpu_state(ctx, 1);
+                    gen_helper_rdhwr_cpunum(t0);
+                    gen_store_gpr(t0, rt);
+                    break;
+                case 1:
+                    save_cpu_state(ctx, 1);
+                    gen_helper_rdhwr_synci_step(t0);
+                    gen_store_gpr(t0, rt);
+                    break;
+                case 2:
+                    save_cpu_state(ctx, 1);
+                    gen_helper_rdhwr_cc(t0);
+                    gen_store_gpr(t0, rt);
+                    break;
+                case 3:
+                    save_cpu_state(ctx, 1);
+                    gen_helper_rdhwr_ccres(t0);
+                    gen_store_gpr(t0, rt);
+                    break;
+                case 29:
+#if defined(CONFIG_USER_ONLY)
+                    tcg_gen_ld_tl(t0, cpu_env, offsetof(CPUState, tls_value));
+                    gen_store_gpr(t0, rt);
+                    break;
+#else
+                    /* XXX: Some CPUs implement this in hardware.
+                       Not supported yet. */
+#endif
+                default:            /* Invalid */
+                    MIPS_INVAL("rdhwr");
+                    generate_exception(ctx, EXCP_RI);
+                    break;
+                }
+                tcg_temp_free(t0);
+            }
+            break;
+        case OPC_FORK:
+            check_insn(env, ctx, ASE_MT);
+            {
+                TCGv t0 = tcg_temp_new();
+                TCGv t1 = tcg_temp_new();
+
+                gen_load_gpr(t0, rt);
+                gen_load_gpr(t1, rs);
+                gen_helper_fork(t0, t1);
+                tcg_temp_free(t0);
+                tcg_temp_free(t1);
+            }
+            break;
+        case OPC_YIELD:
+            check_insn(env, ctx, ASE_MT);
+            {
+                TCGv t0 = tcg_temp_new();
+
+                save_cpu_state(ctx, 1);
+                gen_load_gpr(t0, rs);
+                gen_helper_yield(t0, t0);
+                gen_store_gpr(t0, rd);
+                tcg_temp_free(t0);
+            }
+            break;
+#if defined(TARGET_MIPS64)
+        case OPC_DEXTM ... OPC_DEXT:
+        case OPC_DINSM ... OPC_DINS:
+            check_insn(env, ctx, ISA_MIPS64R2);
+            check_mips_64(ctx);
+            gen_bitops(ctx, op1, rt, rs, sa, rd);
+            break;
+        case OPC_DBSHFL:
+            check_insn(env, ctx, ISA_MIPS64R2);
+            check_mips_64(ctx);
+            op2 = MASK_DBSHFL(ctx->opcode);
+            gen_bshfl(ctx, op2, rt, rd);
+            break;
+#endif
+        default:            /* Invalid */
+            MIPS_INVAL("special3");
+            generate_exception(ctx, EXCP_RI);
+            break;
+        }
+        break;
+    case OPC_REGIMM:
+        op1 = MASK_REGIMM(ctx->opcode);
+        switch (op1) {
+        case OPC_BLTZ ... OPC_BGEZL: /* REGIMM branches */
+        case OPC_BLTZAL ... OPC_BGEZALL:
+            gen_compute_branch(ctx, op1, rs, -1, imm << 2);
+            return;
+        case OPC_TGEI ... OPC_TEQI: /* REGIMM traps */
+        case OPC_TNEI:
+            gen_trap(ctx, op1, rs, -1, imm);
+            break;
+        case OPC_SYNCI:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            /* Treat as NOP. */
+            break;
+        default:            /* Invalid */
+            MIPS_INVAL("regimm");
+            generate_exception(ctx, EXCP_RI);
+            break;
+        }
+        break;
+    case OPC_CP0:
+        check_cp0_enabled(ctx);
+        op1 = MASK_CP0(ctx->opcode);
+        switch (op1) {
+        case OPC_MFC0:
+        case OPC_MTC0:
+        case OPC_MFTR:
+        case OPC_MTTR:
+#if defined(TARGET_MIPS64)
+        case OPC_DMFC0:
+        case OPC_DMTC0:
+#endif
+#ifndef CONFIG_USER_ONLY
+            gen_cp0(env, ctx, op1, rt, rd);
+#endif /* !CONFIG_USER_ONLY */
+            break;
+        case OPC_C0_FIRST ... OPC_C0_LAST:
+#ifndef CONFIG_USER_ONLY
+            gen_cp0(env, ctx, MASK_C0(ctx->opcode), rt, rd);
+#endif /* !CONFIG_USER_ONLY */
+            break;
+        case OPC_MFMC0:
+#ifndef CONFIG_USER_ONLY
+            {
+                TCGv t0 = tcg_temp_new();
+
+                op2 = MASK_MFMC0(ctx->opcode);
+                switch (op2) {
+                case OPC_DMT:
+                    check_insn(env, ctx, ASE_MT);
+                    gen_helper_dmt(t0, t0);
+                    gen_store_gpr(t0, rt);
+                    break;
+                case OPC_EMT:
+                    check_insn(env, ctx, ASE_MT);
+                    gen_helper_emt(t0, t0);
+                    gen_store_gpr(t0, rt);
+                    break;
+                case OPC_DVPE:
+                    check_insn(env, ctx, ASE_MT);
+                    gen_helper_dvpe(t0, t0);
+                    gen_store_gpr(t0, rt);
+                    break;
+                case OPC_EVPE:
+                    check_insn(env, ctx, ASE_MT);
+                    gen_helper_evpe(t0, t0);
+                    gen_store_gpr(t0, rt);
+                    break;
+                case OPC_DI:
+                    check_insn(env, ctx, ISA_MIPS32R2);
+                    save_cpu_state(ctx, 1);
+                    gen_helper_di(t0);
+                    gen_store_gpr(t0, rt);
+                    /* Stop translation as we may have switched the execution mode */
+                    ctx->bstate = BS_STOP;
+                    break;
+                case OPC_EI:
+                    check_insn(env, ctx, ISA_MIPS32R2);
+                    save_cpu_state(ctx, 1);
+                    gen_helper_ei(t0);
+                    gen_store_gpr(t0, rt);
+                    /* Stop translation as we may have switched the execution mode */
+                    ctx->bstate = BS_STOP;
+                    break;
+                default:            /* Invalid */
+                    MIPS_INVAL("mfmc0");
+                    generate_exception(ctx, EXCP_RI);
+                    break;
+                }
+                tcg_temp_free(t0);
+            }
+#endif /* !CONFIG_USER_ONLY */
+            break;
+        case OPC_RDPGPR:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_load_srsgpr(rt, rd);
+            break;
+        case OPC_WRPGPR:
+            check_insn(env, ctx, ISA_MIPS32R2);
+            gen_store_srsgpr(rt, rd);
+            break;
+        default:
+            MIPS_INVAL("cp0");
+            generate_exception(ctx, EXCP_RI);
+            break;
+        }
+        break;
+    case OPC_ADDI: /* Arithmetic with immediate opcode */
+    case OPC_ADDIU:
+         gen_arith_imm(env, ctx, op, rt, rs, imm);
+         break;
+    case OPC_SLTI: /* Set on less than with immediate opcode */
+    case OPC_SLTIU:
+         gen_slt_imm(env, op, rt, rs, imm);
+         break;
+    case OPC_ANDI: /* Arithmetic with immediate opcode */
+    case OPC_LUI:
+    case OPC_ORI:
+    case OPC_XORI:
+         gen_logic_imm(env, op, rt, rs, imm);
+         break;
+    case OPC_J ... OPC_JAL: /* Jump */
+         offset = (int32_t)(ctx->opcode & 0x3FFFFFF) << 2;
+         gen_compute_branch(ctx, op, rs, rt, offset);
+         return;
+    case OPC_BEQ ... OPC_BGTZ: /* Branch */
+    case OPC_BEQL ... OPC_BGTZL:
+         gen_compute_branch(ctx, op, rs, rt, imm << 2);
+         return;
+    case OPC_LB ... OPC_LWR: /* Load and stores */
+    case OPC_SB ... OPC_SW:
+    case OPC_SWR:
+    case OPC_LL:
+         gen_ldst(ctx, op, rt, rs, imm);
+         break;
+    case OPC_SC:
+         gen_st_cond(ctx, op, rt, rs, imm);
+         break;
+    case OPC_CACHE:
+        check_insn(env, ctx, ISA_MIPS3 | ISA_MIPS32);
+        /* Treat as NOP. */
+        break;
+    case OPC_PREF:
+        check_insn(env, ctx, ISA_MIPS4 | ISA_MIPS32);
+        /* Treat as NOP. */
+        break;
+
+    /* Floating point (COP1). */
+    case OPC_LWC1:
+    case OPC_LDC1:
+    case OPC_SWC1:
+    case OPC_SDC1:
+        if (env->CP0_Config1 & (1 << CP0C1_FP)) {
+            check_cp1_enabled(ctx);
+            gen_flt_ldst(ctx, op, rt, rs, imm);
+        } else {
+            generate_exception_err(ctx, EXCP_CpU, 1);
+        }
+        break;
+
+    case OPC_CP1:
+        if (env->CP0_Config1 & (1 << CP0C1_FP)) {
+            check_cp1_enabled(ctx);
+            op1 = MASK_CP1(ctx->opcode);
+            switch (op1) {
+            case OPC_MFHC1:
+            case OPC_MTHC1:
+                check_insn(env, ctx, ISA_MIPS32R2);
+            case OPC_MFC1:
+            case OPC_CFC1:
+            case OPC_MTC1:
+            case OPC_CTC1:
+                gen_cp1(ctx, op1, rt, rd);
+                break;
+#if defined(TARGET_MIPS64)
+            case OPC_DMFC1:
+            case OPC_DMTC1:
+                check_insn(env, ctx, ISA_MIPS3);
+                gen_cp1(ctx, op1, rt, rd);
+                break;
+#endif
+            case OPC_BC1ANY2:
+            case OPC_BC1ANY4:
+                check_cop1x(ctx);
+                check_insn(env, ctx, ASE_MIPS3D);
+                /* fall through */
+            case OPC_BC1:
+                gen_compute_branch1(env, ctx, MASK_BC1(ctx->opcode),
+                                    (rt >> 2) & 0x7, imm << 2);
+                return;
+            case OPC_S_FMT:
+            case OPC_D_FMT:
+            case OPC_W_FMT:
+            case OPC_L_FMT:
+            case OPC_PS_FMT:
+                gen_farith(ctx, MASK_CP1_FUNC(ctx->opcode), rt, rd, sa,
+                           (imm >> 8) & 0x7);
+                break;
+            default:
+                MIPS_INVAL("cp1");
+                generate_exception (ctx, EXCP_RI);
+                break;
+            }
+        } else {
+            generate_exception_err(ctx, EXCP_CpU, 1);
+        }
+        break;
+
+    /* COP2.  */
+    case OPC_LWC2:
+    case OPC_LDC2:
+    case OPC_SWC2:
+    case OPC_SDC2:
+    case OPC_CP2:
+        /* COP2: Not implemented. */
+        generate_exception_err(ctx, EXCP_CpU, 2);
+        break;
+
+    case OPC_CP3:
+        if (env->CP0_Config1 & (1 << CP0C1_FP)) {
+            check_cp1_enabled(ctx);
+            op1 = MASK_CP3(ctx->opcode);
+            switch (op1) {
+            case OPC_LWXC1:
+            case OPC_LDXC1:
+            case OPC_LUXC1:
+            case OPC_SWXC1:
+            case OPC_SDXC1:
+            case OPC_SUXC1:
+                gen_flt3_ldst(ctx, op1, sa, rd, rs, rt);
+                break;
+            case OPC_PREFX:
+                /* Treat as NOP. */
+                break;
+            case OPC_ALNV_PS:
+            case OPC_MADD_S:
+            case OPC_MADD_D:
+            case OPC_MADD_PS:
+            case OPC_MSUB_S:
+            case OPC_MSUB_D:
+            case OPC_MSUB_PS:
+            case OPC_NMADD_S:
+            case OPC_NMADD_D:
+            case OPC_NMADD_PS:
+            case OPC_NMSUB_S:
+            case OPC_NMSUB_D:
+            case OPC_NMSUB_PS:
+                gen_flt3_arith(ctx, op1, sa, rs, rd, rt);
+                break;
+            default:
+                MIPS_INVAL("cp3");
+                generate_exception (ctx, EXCP_RI);
+                break;
+            }
+        } else {
+            generate_exception_err(ctx, EXCP_CpU, 1);
+        }
+        break;
+
+#if defined(TARGET_MIPS64)
+    /* MIPS64 opcodes */
+    case OPC_LWU:
+    case OPC_LDL ... OPC_LDR:
+    case OPC_SDL ... OPC_SDR:
+    case OPC_LLD:
+    case OPC_LD:
+    case OPC_SD:
+        check_insn(env, ctx, ISA_MIPS3);
+        check_mips_64(ctx);
+        gen_ldst(ctx, op, rt, rs, imm);
+        break;
+    case OPC_SCD:
+        check_insn(env, ctx, ISA_MIPS3);
+        check_mips_64(ctx);
+        gen_st_cond(ctx, op, rt, rs, imm);
+        break;
+    case OPC_DADDI:
+    case OPC_DADDIU:
+        check_insn(env, ctx, ISA_MIPS3);
+        check_mips_64(ctx);
+        gen_arith_imm(env, ctx, op, rt, rs, imm);
+        break;
+#endif
+    case OPC_JALX:
+        check_insn(env, ctx, ASE_MIPS16);
+        /* MIPS16: Not implemented. */
+    case OPC_MDMX:
+        check_insn(env, ctx, ASE_MDMX);
+        /* MDMX: Not implemented. */
+    default:            /* Invalid */
+        MIPS_INVAL("major opcode");
+        generate_exception(ctx, EXCP_RI);
+        break;
+    }
+    if (ctx->hflags & MIPS_HFLAG_BMASK) {
+        int hflags = ctx->hflags & MIPS_HFLAG_BMASK;
+        /* Branches completion */
+        ctx->hflags &= ~MIPS_HFLAG_BMASK;
+        ctx->bstate = BS_BRANCH;
+        save_cpu_state(ctx, 0);
+        /* FIXME: Need to clear can_do_io.  */
+        switch (hflags) {
+        case MIPS_HFLAG_B:
+            /* unconditional branch */
+            MIPS_DEBUG("unconditional branch");
+            gen_goto_tb(ctx, 0, ctx->btarget);
+            break;
+        case MIPS_HFLAG_BL:
+            /* blikely taken case */
+            MIPS_DEBUG("blikely branch taken");
+            gen_goto_tb(ctx, 0, ctx->btarget);
+            break;
+        case MIPS_HFLAG_BC:
+            /* Conditional branch */
+            MIPS_DEBUG("conditional branch");
+            {
+                int l1 = gen_new_label();
+
+                tcg_gen_brcondi_tl(TCG_COND_NE, bcond, 0, l1);
+                gen_goto_tb(ctx, 1, ctx->pc + 4);
+                gen_set_label(l1);
+                gen_goto_tb(ctx, 0, ctx->btarget);
+            }
+            break;
+        case MIPS_HFLAG_BR:
+            /* unconditional branch to register */
+            MIPS_DEBUG("branch to register");
+            tcg_gen_mov_tl(cpu_PC, btarget);
+            if (ctx->singlestep_enabled) {
+                save_cpu_state(ctx, 0);
+                gen_helper_0i(raise_exception, EXCP_DEBUG);
+            }
+            tcg_gen_exit_tb(0);
+            break;
+        default:
+            MIPS_DEBUG("unknown branch");
+            break;
+        }
+    }
+}
+
+static inline void
+gen_intermediate_code_internal (CPUState *env, TranslationBlock *tb,
+                                int search_pc)
+{
+    DisasContext ctx;
+    target_ulong pc_start;
+    uint16_t *gen_opc_end;
+    CPUBreakpoint *bp;
+    int j, lj = -1;
+    int num_insns;
+    int max_insns;
+
+    if (search_pc)
+        qemu_log("search pc %d\n", search_pc);
+
+    pc_start = tb->pc;
+    gen_opc_end = gen_opc_buf + OPC_MAX_SIZE;
+    ctx.pc = pc_start;
+    ctx.saved_pc = -1;
+    ctx.singlestep_enabled = env->singlestep_enabled;
+    ctx.tb = tb;
+    ctx.bstate = BS_NONE;
+    /* Restore delay slot state from the tb context.  */
+    ctx.hflags = (uint32_t)tb->flags; /* FIXME: maybe use 64 bits here? */
+    restore_cpu_state(env, &ctx);
+#ifdef CONFIG_USER_ONLY
+        ctx.mem_idx = MIPS_HFLAG_UM;
+#else
+        ctx.mem_idx = ctx.hflags & MIPS_HFLAG_KSU;
+#endif
+    num_insns = 0;
+    max_insns = tb->cflags & CF_COUNT_MASK;
+    if (max_insns == 0)
+        max_insns = CF_COUNT_MASK;
+#ifdef DEBUG_DISAS
+    qemu_log_mask(CPU_LOG_TB_CPU, "------------------------------------------------\n");
+    /* FIXME: This may print out stale hflags from env... */
+    log_cpu_state_mask(CPU_LOG_TB_CPU, env, 0);
+#endif
+    LOG_DISAS("\ntb %p idx %d hflags %04x\n", tb, ctx.mem_idx, ctx.hflags);
+    gen_icount_start();
+    while (ctx.bstate == BS_NONE) {
+        if (unlikely(!QTAILQ_EMPTY(&env->breakpoints))) {
+            QTAILQ_FOREACH(bp, &env->breakpoints, entry) {
+                if (bp->pc == ctx.pc) {
+                    save_cpu_state(&ctx, 1);
+                    ctx.bstate = BS_BRANCH;
+                    gen_helper_0i(raise_exception, EXCP_DEBUG);
+                    /* Include the breakpoint location or the tb won't
+                     * be flushed when it must be.  */
+                    ctx.pc += 4;
+                    goto done_generating;
+                }
+            }
+        }
+
+        if (search_pc) {
+            j = gen_opc_ptr - gen_opc_buf;
+            if (lj < j) {
+                lj++;
+                while (lj < j)
+                    gen_opc_instr_start[lj++] = 0;
+            }
+            gen_opc_pc[lj] = ctx.pc;
+            gen_opc_hflags[lj] = ctx.hflags & MIPS_HFLAG_BMASK;
+            gen_opc_instr_start[lj] = 1;
+            gen_opc_icount[lj] = num_insns;
+        }
+        if (num_insns + 1 == max_insns && (tb->cflags & CF_LAST_IO))
+            gen_io_start();
+        ctx.opcode = ldl_code(ctx.pc);
+        decode_opc(env, &ctx);
+        ctx.pc += 4;
+        num_insns++;
+
+        /* Execute a branch and its delay slot as a single instruction.
+           This is what GDB expects and is consistent with what the
+           hardware does (e.g. if a delay slot instruction faults, the
+           reported PC is the PC of the branch).  */
+        if (env->singlestep_enabled && (ctx.hflags & MIPS_HFLAG_BMASK) == 0)
+            break;
+
+        if ((ctx.pc & (TARGET_PAGE_SIZE - 1)) == 0)
+            break;
+
+        if (gen_opc_ptr >= gen_opc_end)
+            break;
+
+        if (num_insns >= max_insns)
+            break;
+
+        if (singlestep)
+            break;
+    }
+    if (tb->cflags & CF_LAST_IO)
+        gen_io_end();
+    if (env->singlestep_enabled && ctx.bstate != BS_BRANCH) {
+        save_cpu_state(&ctx, ctx.bstate == BS_NONE);
+        gen_helper_0i(raise_exception, EXCP_DEBUG);
+    } else {
+        switch (ctx.bstate) {
+        case BS_STOP:
+            gen_helper_interrupt_restart();
+            gen_goto_tb(&ctx, 0, ctx.pc);
+            break;
+        case BS_NONE:
+            save_cpu_state(&ctx, 0);
+            gen_goto_tb(&ctx, 0, ctx.pc);
+            break;
+        case BS_EXCP:
+            gen_helper_interrupt_restart();
+            tcg_gen_exit_tb(0);
+            break;
+        case BS_BRANCH:
+        default:
+            break;
+        }
+    }
+done_generating:
+    gen_icount_end(tb, num_insns);
+    *gen_opc_ptr = INDEX_op_end;
+    if (search_pc) {
+        j = gen_opc_ptr - gen_opc_buf;
+        lj++;
+        while (lj <= j)
+            gen_opc_instr_start[lj++] = 0;
+    } else {
+        tb->size = ctx.pc - pc_start;
+        tb->icount = num_insns;
+    }
+#ifdef DEBUG_DISAS
+    LOG_DISAS("\n");
+    if (qemu_loglevel_mask(CPU_LOG_TB_IN_ASM)) {
+        qemu_log("IN: %s\n", lookup_symbol(pc_start));
+        log_target_disas(pc_start, ctx.pc - pc_start, 0);
+        qemu_log("\n");
+    }
+    qemu_log_mask(CPU_LOG_TB_CPU, "---------------- %d %08x\n", ctx.bstate, ctx.hflags);
+#endif
+}
+
+void gen_intermediate_code (CPUState *env, struct TranslationBlock *tb)
+{
+    gen_intermediate_code_internal(env, tb, 0);
+}
+
+void gen_intermediate_code_pc (CPUState *env, struct TranslationBlock *tb)
+{
+    gen_intermediate_code_internal(env, tb, 1);
+}
+
+static void fpu_dump_state(CPUState *env, FILE *f,
+                           int (*fpu_fprintf)(FILE *f, const char *fmt, ...),
+                           int flags)
+{
+    int i;
+    int is_fpu64 = !!(env->hflags & MIPS_HFLAG_F64);
+
+#define printfpr(fp)                                                        \
+    do {                                                                    \
+        if (is_fpu64)                                                       \
+            fpu_fprintf(f, "w:%08x d:%016lx fd:%13g fs:%13g psu: %13g\n",   \
+                        (fp)->w[FP_ENDIAN_IDX], (fp)->d, (fp)->fd,          \
+                        (fp)->fs[FP_ENDIAN_IDX], (fp)->fs[!FP_ENDIAN_IDX]); \
+        else {                                                              \
+            fpr_t tmp;                                                      \
+            tmp.w[FP_ENDIAN_IDX] = (fp)->w[FP_ENDIAN_IDX];                  \
+            tmp.w[!FP_ENDIAN_IDX] = ((fp) + 1)->w[FP_ENDIAN_IDX];           \
+            fpu_fprintf(f, "w:%08x d:%016lx fd:%13g fs:%13g psu:%13g\n",    \
+                        tmp.w[FP_ENDIAN_IDX], tmp.d, tmp.fd,                \
+                        tmp.fs[FP_ENDIAN_IDX], tmp.fs[!FP_ENDIAN_IDX]);     \
+        }                                                                   \
+    } while(0)
+
+
+    fpu_fprintf(f, "CP1 FCR0 0x%08x  FCR31 0x%08x  SR.FR %d  fp_status 0x%08x(0x%02x)\n",
+                env->active_fpu.fcr0, env->active_fpu.fcr31, is_fpu64, env->active_fpu.fp_status,
+                get_float_exception_flags(&env->active_fpu.fp_status));
+    for (i = 0; i < 32; (is_fpu64) ? i++ : (i += 2)) {
+        fpu_fprintf(f, "%3s: ", fregnames[i]);
+        printfpr(&env->active_fpu.fpr[i]);
+    }
+
+#undef printfpr
+}
+
+#if defined(TARGET_MIPS64) && defined(MIPS_DEBUG_SIGN_EXTENSIONS)
+/* Debug help: The architecture requires 32bit code to maintain proper
+   sign-extended values on 64bit machines.  */
+
+#define SIGN_EXT_P(val) ((((val) & ~0x7fffffff) == 0) || (((val) & ~0x7fffffff) == ~0x7fffffff))
+
+static void
+cpu_mips_check_sign_extensions (CPUState *env, FILE *f,
+                                int (*cpu_fprintf)(FILE *f, const char *fmt, ...),
+                                int flags)
+{
+    int i;
+
+    if (!SIGN_EXT_P(env->active_tc.PC))
+        cpu_fprintf(f, "BROKEN: pc=0x" TARGET_FMT_lx "\n", env->active_tc.PC);
+    if (!SIGN_EXT_P(env->active_tc.HI[0]))
+        cpu_fprintf(f, "BROKEN: HI=0x" TARGET_FMT_lx "\n", env->active_tc.HI[0]);
+    if (!SIGN_EXT_P(env->active_tc.LO[0]))
+        cpu_fprintf(f, "BROKEN: LO=0x" TARGET_FMT_lx "\n", env->active_tc.LO[0]);
+    if (!SIGN_EXT_P(env->btarget))
+        cpu_fprintf(f, "BROKEN: btarget=0x" TARGET_FMT_lx "\n", env->btarget);
+
+    for (i = 0; i < 32; i++) {
+        if (!SIGN_EXT_P(env->active_tc.gpr[i]))
+            cpu_fprintf(f, "BROKEN: %s=0x" TARGET_FMT_lx "\n", regnames[i], env->active_tc.gpr[i]);
+    }
+
+    if (!SIGN_EXT_P(env->CP0_EPC))
+        cpu_fprintf(f, "BROKEN: EPC=0x" TARGET_FMT_lx "\n", env->CP0_EPC);
+    if (!SIGN_EXT_P(env->lladdr))
+        cpu_fprintf(f, "BROKEN: LLAddr=0x" TARGET_FMT_lx "\n", env->lladdr);
+}
+#endif
+
+void cpu_dump_state (CPUState *env, FILE *f,
+                     int (*cpu_fprintf)(FILE *f, const char *fmt, ...),
+                     int flags)
+{
+    int i;
+
+    cpu_fprintf(f, "pc=0x" TARGET_FMT_lx " HI=0x" TARGET_FMT_lx " LO=0x" TARGET_FMT_lx " ds %04x " TARGET_FMT_lx " %d\n",
+                env->active_tc.PC, env->active_tc.HI[0], env->active_tc.LO[0],
+                env->hflags, env->btarget, env->bcond);
+    for (i = 0; i < 32; i++) {
+        if ((i & 3) == 0)
+            cpu_fprintf(f, "GPR%02d:", i);
+        cpu_fprintf(f, " %s " TARGET_FMT_lx, regnames[i], env->active_tc.gpr[i]);
+        if ((i & 3) == 3)
+            cpu_fprintf(f, "\n");
+    }
+
+    cpu_fprintf(f, "CP0 Status  0x%08x Cause   0x%08x EPC    0x" TARGET_FMT_lx "\n",
+                env->CP0_Status, env->CP0_Cause, env->CP0_EPC);
+    cpu_fprintf(f, "    Config0 0x%08x Config1 0x%08x LLAddr 0x" TARGET_FMT_lx "\n",
+                env->CP0_Config0, env->CP0_Config1, env->lladdr);
+    if (env->hflags & MIPS_HFLAG_FPU)
+        fpu_dump_state(env, f, cpu_fprintf, flags);
+#if defined(TARGET_MIPS64) && defined(MIPS_DEBUG_SIGN_EXTENSIONS)
+    cpu_mips_check_sign_extensions(env, f, cpu_fprintf, flags);
+#endif
+}
+
+static void mips_tcg_init(void)
+{
+    int i;
+    static int inited;
+
+    /* Initialize various static tables. */
+    if (inited)
+        return;
+
+    cpu_env = tcg_global_reg_new_ptr(TCG_AREG0, "env");
+    TCGV_UNUSED(cpu_gpr[0]);
+    for (i = 1; i < 32; i++)
+        cpu_gpr[i] = tcg_global_mem_new(TCG_AREG0,
+                                        offsetof(CPUState, active_tc.gpr[i]),
+                                        regnames[i]);
+    cpu_PC = tcg_global_mem_new(TCG_AREG0,
+                                offsetof(CPUState, active_tc.PC), "PC");
+    for (i = 0; i < MIPS_DSP_ACC; i++) {
+        cpu_HI[i] = tcg_global_mem_new(TCG_AREG0,
+                                       offsetof(CPUState, active_tc.HI[i]),
+                                       regnames_HI[i]);
+        cpu_LO[i] = tcg_global_mem_new(TCG_AREG0,
+                                       offsetof(CPUState, active_tc.LO[i]),
+                                       regnames_LO[i]);
+        cpu_ACX[i] = tcg_global_mem_new(TCG_AREG0,
+                                        offsetof(CPUState, active_tc.ACX[i]),
+                                        regnames_ACX[i]);
+    }
+    cpu_dspctrl = tcg_global_mem_new(TCG_AREG0,
+                                     offsetof(CPUState, active_tc.DSPControl),
+                                     "DSPControl");
+    bcond = tcg_global_mem_new(TCG_AREG0,
+                               offsetof(CPUState, bcond), "bcond");
+    btarget = tcg_global_mem_new(TCG_AREG0,
+                                 offsetof(CPUState, btarget), "btarget");
+    hflags = tcg_global_mem_new_i32(TCG_AREG0,
+                                    offsetof(CPUState, hflags), "hflags");
+
+    fpu_fcr0 = tcg_global_mem_new_i32(TCG_AREG0,
+                                      offsetof(CPUState, active_fpu.fcr0),
+                                      "fcr0");
+    fpu_fcr31 = tcg_global_mem_new_i32(TCG_AREG0,
+                                       offsetof(CPUState, active_fpu.fcr31),
+                                       "fcr31");
+
+    /* register helpers */
+#define GEN_HELPER 2
+#include "helper.h"
+
+    inited = 1;
+}
+
+#include "translate_init.c"
+
+CPUMIPSState *cpu_mips_init (const char *cpu_model)
+{
+    CPUMIPSState *env;
+    const mips_def_t *def;
+
+    def = cpu_mips_find_by_name(cpu_model);
+    if (!def)
+        return NULL;
+    env = qemu_mallocz(sizeof(CPUMIPSState));
+    env->cpu_model = def;
+    env->cpu_model_str = cpu_model;
+
+    cpu_exec_init(env);
+#ifndef CONFIG_USER_ONLY
+    mmu_init(env, def);
+#endif
+    mvp_init(env, def);
+    mips_tcg_init();
+    cpu_reset(env);
+    qemu_init_vcpu(env);
+    return env;
+}
+
+void cpu_reset (CPUMIPSState *env)
+{
+    if (qemu_loglevel_mask(CPU_LOG_RESET)) {
+        qemu_log("CPU Reset (CPU %d)\n", env->cpu_index);
+        log_cpu_state(env, 0);
+    }
+
+    memset(env, 0, offsetof(CPUMIPSState, breakpoints));
+    tlb_flush(env, 1);
+
+    /* Reset registers to their default values */
+    env->CP0_PRid = env->cpu_model->CP0_PRid;
+    env->CP0_Config0 = env->cpu_model->CP0_Config0;
+#ifdef TARGET_WORDS_BIGENDIAN
+    env->CP0_Config0 |= (1 << CP0C0_BE);
+#endif
+    env->CP0_Config1 = env->cpu_model->CP0_Config1;
+    env->CP0_Config2 = env->cpu_model->CP0_Config2;
+    env->CP0_Config3 = env->cpu_model->CP0_Config3;
+    env->CP0_Config6 = env->cpu_model->CP0_Config6;
+    env->CP0_Config7 = env->cpu_model->CP0_Config7;
+    env->CP0_LLAddr_rw_bitmask = env->cpu_model->CP0_LLAddr_rw_bitmask
+                                 << env->cpu_model->CP0_LLAddr_shift;
+    env->CP0_LLAddr_shift = env->cpu_model->CP0_LLAddr_shift;
+    env->SYNCI_Step = env->cpu_model->SYNCI_Step;
+    env->CCRes = env->cpu_model->CCRes;
+    env->CP0_Status_rw_bitmask = env->cpu_model->CP0_Status_rw_bitmask;
+    env->CP0_TCStatus_rw_bitmask = env->cpu_model->CP0_TCStatus_rw_bitmask;
+    env->CP0_SRSCtl = env->cpu_model->CP0_SRSCtl;
+    env->current_tc = 0;
+    env->SEGBITS = env->cpu_model->SEGBITS;
+    env->SEGMask = (target_ulong)((1ULL << env->cpu_model->SEGBITS) - 1);
+#if defined(TARGET_MIPS64)
+    if (env->cpu_model->insn_flags & ISA_MIPS3) {
+        env->SEGMask |= 3ULL << 62;
+    }
+#endif
+    env->PABITS = env->cpu_model->PABITS;
+    env->PAMask = (target_ulong)((1ULL << env->cpu_model->PABITS) - 1);
+    env->CP0_SRSConf0_rw_bitmask = env->cpu_model->CP0_SRSConf0_rw_bitmask;
+    env->CP0_SRSConf0 = env->cpu_model->CP0_SRSConf0;
+    env->CP0_SRSConf1_rw_bitmask = env->cpu_model->CP0_SRSConf1_rw_bitmask;
+    env->CP0_SRSConf1 = env->cpu_model->CP0_SRSConf1;
+    env->CP0_SRSConf2_rw_bitmask = env->cpu_model->CP0_SRSConf2_rw_bitmask;
+    env->CP0_SRSConf2 = env->cpu_model->CP0_SRSConf2;
+    env->CP0_SRSConf3_rw_bitmask = env->cpu_model->CP0_SRSConf3_rw_bitmask;
+    env->CP0_SRSConf3 = env->cpu_model->CP0_SRSConf3;
+    env->CP0_SRSConf4_rw_bitmask = env->cpu_model->CP0_SRSConf4_rw_bitmask;
+    env->CP0_SRSConf4 = env->cpu_model->CP0_SRSConf4;
+    env->insn_flags = env->cpu_model->insn_flags;
+
+    fpu_init(env, env->cpu_model);
+
+#if defined(CONFIG_USER_ONLY)
+    env->hflags = MIPS_HFLAG_UM;
+    /* Enable access to the SYNCI_Step register.  */
+    env->CP0_HWREna |= (1 << 1);
+    if (env->CP0_Config1 & (1 << CP0C1_FP)) {
+        env->hflags |= MIPS_HFLAG_FPU;
+    }
+#ifdef TARGET_MIPS64
+    if (env->active_fpu.fcr0 & (1 << FCR0_F64)) {
+        env->hflags |= MIPS_HFLAG_F64;
+    }
+#endif
+#else
+    if (env->hflags & MIPS_HFLAG_BMASK) {
+        /* If the exception was raised from a delay slot,
+           come back to the jump.  */
+        env->CP0_ErrorEPC = env->active_tc.PC - 4;
+    } else {
+        env->CP0_ErrorEPC = env->active_tc.PC;
+    }
+    env->active_tc.PC = (int32_t)0xBFC00000;
+    env->CP0_Random = env->tlb->nb_tlb - 1;
+    env->tlb->tlb_in_use = env->tlb->nb_tlb;
+    env->CP0_Wired = 0;
+    /* SMP not implemented */
+    env->CP0_EBase = 0x80000000;
+    env->CP0_Status = (1 << CP0St_BEV) | (1 << CP0St_ERL);
+    /* vectored interrupts not implemented, timer on int 7,
+       no performance counters. */
+    env->CP0_IntCtl = 0xe0000000;
+    {
+        int i;
+
+        for (i = 0; i < 7; i++) {
+            env->CP0_WatchLo[i] = 0;
+            env->CP0_WatchHi[i] = 0x80000000;
+        }
+        env->CP0_WatchLo[7] = 0;
+        env->CP0_WatchHi[7] = 0;
+    }
+    /* Count register increments in debug mode, EJTAG version 1 */
+    env->CP0_Debug = (1 << CP0DB_CNT) | (0x1 << CP0DB_VER);
+    env->hflags = MIPS_HFLAG_CP0;
+#endif
+#if defined(TARGET_MIPS64)
+    if (env->cpu_model->insn_flags & ISA_MIPS3) {
+        env->hflags |= MIPS_HFLAG_64;
+    }
+#endif
+    env->exception_index = EXCP_NONE;
+}
+
+void gen_pc_load(CPUState *env, TranslationBlock *tb,
+                unsigned long searched_pc, int pc_pos, void *puc)
+{
+    env->active_tc.PC = gen_opc_pc[pc_pos];
+    env->hflags &= ~MIPS_HFLAG_BMASK;
+    env->hflags |= gen_opc_hflags[pc_pos];
+}
diff --git a/target-mips/translate_init.c b/target-mips/translate_init.c
new file mode 100644
index 0000000..3978908
--- /dev/null
+++ b/target-mips/translate_init.c
@@ -0,0 +1,554 @@
+/*
+ *  MIPS emulation for qemu: CPU initialisation routines.
+ *
+ *  Copyright (c) 2004-2005 Jocelyn Mayer
+ *  Copyright (c) 2007 Herve Poussineau
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+/* CPU / CPU family specific config register values. */
+
+/* Have config1, uncached coherency */
+#define MIPS_CONFIG0                                              \
+  ((1 << CP0C0_M) | (0x2 << CP0C0_K0))
+
+/* Have config2, no coprocessor2 attached, no MDMX support attached,
+   no performance counters, watch registers present,
+   no code compression, EJTAG present, no FPU */
+#define MIPS_CONFIG1                                              \
+((1 << CP0C1_M) |                                                 \
+ (0 << CP0C1_C2) | (0 << CP0C1_MD) | (0 << CP0C1_PC) |            \
+ (1 << CP0C1_WR) | (0 << CP0C1_CA) | (1 << CP0C1_EP) |            \
+ (0 << CP0C1_FP))
+
+/* Have config3, no tertiary/secondary caches implemented */
+#define MIPS_CONFIG2                                              \
+((1 << CP0C2_M))
+
+/* No config4, no DSP ASE, no large physaddr (PABITS),
+   no external interrupt controller, no vectored interupts,
+   no 1kb pages, no SmartMIPS ASE, no trace logic */
+#define MIPS_CONFIG3                                              \
+((0 << CP0C3_M) | (0 << CP0C3_DSPP) | (0 << CP0C3_LPA) |          \
+ (0 << CP0C3_VEIC) | (0 << CP0C3_VInt) | (0 << CP0C3_SP) |        \
+ (0 << CP0C3_SM) | (0 << CP0C3_TL))
+
+/* Define a implementation number of 1.
+   Define a major version 1, minor version 0. */
+#define MIPS_FCR0 ((0 << FCR0_S) | (0x1 << FCR0_PRID) | (0x10 << FCR0_REV))
+
+/* MMU types, the first four entries have the same layout as the
+   CP0C0_MT field.  */
+enum mips_mmu_types {
+    MMU_TYPE_NONE,
+    MMU_TYPE_R4000,
+    MMU_TYPE_RESERVED,
+    MMU_TYPE_FMT,
+    MMU_TYPE_R3000,
+    MMU_TYPE_R6000,
+    MMU_TYPE_R8000
+};
+
+struct mips_def_t {
+    const char *name;
+    int32_t CP0_PRid;
+    int32_t CP0_Config0;
+    int32_t CP0_Config1;
+    int32_t CP0_Config2;
+    int32_t CP0_Config3;
+    int32_t CP0_Config6;
+    int32_t CP0_Config7;
+    target_ulong CP0_LLAddr_rw_bitmask;
+    int CP0_LLAddr_shift;
+    int32_t SYNCI_Step;
+    int32_t CCRes;
+    int32_t CP0_Status_rw_bitmask;
+    int32_t CP0_TCStatus_rw_bitmask;
+    int32_t CP0_SRSCtl;
+    int32_t CP1_fcr0;
+    int32_t SEGBITS;
+    int32_t PABITS;
+    int32_t CP0_SRSConf0_rw_bitmask;
+    int32_t CP0_SRSConf0;
+    int32_t CP0_SRSConf1_rw_bitmask;
+    int32_t CP0_SRSConf1;
+    int32_t CP0_SRSConf2_rw_bitmask;
+    int32_t CP0_SRSConf2;
+    int32_t CP0_SRSConf3_rw_bitmask;
+    int32_t CP0_SRSConf3;
+    int32_t CP0_SRSConf4_rw_bitmask;
+    int32_t CP0_SRSConf4;
+    int insn_flags;
+    enum mips_mmu_types mmu_type;
+};
+
+/*****************************************************************************/
+/* MIPS CPU definitions */
+static const mips_def_t mips_defs[] =
+{
+    {
+        .name = "4Kc",
+        .CP0_PRid = 0x00018000,
+        .CP0_Config0 = MIPS_CONFIG0 | (MMU_TYPE_R4000 << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 | (15 << CP0C1_MMU) |
+                       (0 << CP0C1_IS) | (3 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (0 << CP0C1_DS) | (3 << CP0C1_DL) | (1 << CP0C1_DA),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3,
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x1278FF17,
+        .SEGBITS = 32,
+        .PABITS = 32,
+        .insn_flags = CPU_MIPS32 | ASE_MIPS16,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+    {
+        .name = "4Km",
+        .CP0_PRid = 0x00018300,
+        /* Config1 implemented, fixed mapping MMU,
+           no virtual icache, uncached coherency. */
+        .CP0_Config0 = MIPS_CONFIG0 | (MMU_TYPE_FMT << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 |
+                       (0 << CP0C1_IS) | (3 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (0 << CP0C1_DS) | (3 << CP0C1_DL) | (1 << CP0C1_DA),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3,
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x1258FF17,
+        .SEGBITS = 32,
+        .PABITS = 32,
+        .insn_flags = CPU_MIPS32 | ASE_MIPS16,
+        .mmu_type = MMU_TYPE_FMT,
+    },
+    {
+        .name = "4KEcR1",
+        .CP0_PRid = 0x00018400,
+        .CP0_Config0 = MIPS_CONFIG0 | (MMU_TYPE_R4000 << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 | (15 << CP0C1_MMU) |
+                       (0 << CP0C1_IS) | (3 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (0 << CP0C1_DS) | (3 << CP0C1_DL) | (1 << CP0C1_DA),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3,
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x1278FF17,
+        .SEGBITS = 32,
+        .PABITS = 32,
+        .insn_flags = CPU_MIPS32 | ASE_MIPS16,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+    {
+        .name = "4KEmR1",
+        .CP0_PRid = 0x00018500,
+        .CP0_Config0 = MIPS_CONFIG0 | (MMU_TYPE_FMT << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 |
+                       (0 << CP0C1_IS) | (3 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (0 << CP0C1_DS) | (3 << CP0C1_DL) | (1 << CP0C1_DA),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3,
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x1258FF17,
+        .SEGBITS = 32,
+        .PABITS = 32,
+        .insn_flags = CPU_MIPS32 | ASE_MIPS16,
+        .mmu_type = MMU_TYPE_FMT,
+    },
+    {
+        .name = "4KEc",
+        .CP0_PRid = 0x00019000,
+        .CP0_Config0 = MIPS_CONFIG0 | (0x1 << CP0C0_AR) |
+                    (MMU_TYPE_R4000 << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 | (15 << CP0C1_MMU) |
+                       (0 << CP0C1_IS) | (3 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (0 << CP0C1_DS) | (3 << CP0C1_DL) | (1 << CP0C1_DA),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3 | (0 << CP0C3_VInt),
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x1278FF17,
+        .SEGBITS = 32,
+        .PABITS = 32,
+        .insn_flags = CPU_MIPS32R2 | ASE_MIPS16,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+    {
+        .name = "4KEm",
+        .CP0_PRid = 0x00019100,
+        .CP0_Config0 = MIPS_CONFIG0 | (0x1 << CP0C0_AR) |
+                       (MMU_TYPE_FMT << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 |
+                       (0 << CP0C1_IS) | (3 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (0 << CP0C1_DS) | (3 << CP0C1_DL) | (1 << CP0C1_DA),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3,
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x1258FF17,
+        .SEGBITS = 32,
+        .PABITS = 32,
+        .insn_flags = CPU_MIPS32R2 | ASE_MIPS16,
+        .mmu_type = MMU_TYPE_FMT,
+    },
+    {
+        .name = "24Kc",
+        .CP0_PRid = 0x00019300,
+        .CP0_Config0 = MIPS_CONFIG0 | (0x1 << CP0C0_AR) |
+                       (MMU_TYPE_R4000 << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 | (15 << CP0C1_MMU) |
+                       (0 << CP0C1_IS) | (3 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (0 << CP0C1_DS) | (3 << CP0C1_DL) | (1 << CP0C1_DA),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3 | (0 << CP0C3_VInt),
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        /* No DSP implemented. */
+        .CP0_Status_rw_bitmask = 0x1278FF1F,
+        .SEGBITS = 32,
+        .PABITS = 32,
+        .insn_flags = CPU_MIPS32R2 | ASE_MIPS16,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+    {
+        .name = "24Kf",
+        .CP0_PRid = 0x00019300,
+        .CP0_Config0 = MIPS_CONFIG0 | (0x1 << CP0C0_AR) |
+                    (MMU_TYPE_R4000 << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 | (1 << CP0C1_FP) | (15 << CP0C1_MMU) |
+                       (0 << CP0C1_IS) | (3 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (0 << CP0C1_DS) | (3 << CP0C1_DL) | (1 << CP0C1_DA),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3 | (0 << CP0C3_VInt),
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        /* No DSP implemented. */
+        .CP0_Status_rw_bitmask = 0x3678FF1F,
+        .CP1_fcr0 = (1 << FCR0_F64) | (1 << FCR0_L) | (1 << FCR0_W) |
+                    (1 << FCR0_D) | (1 << FCR0_S) | (0x93 << FCR0_PRID),
+        .SEGBITS = 32,
+        .PABITS = 32,
+        .insn_flags = CPU_MIPS32R2 | ASE_MIPS16,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+    {
+        .name = "34Kf",
+        .CP0_PRid = 0x00019500,
+        .CP0_Config0 = MIPS_CONFIG0 | (0x1 << CP0C0_AR) |
+                       (MMU_TYPE_R4000 << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 | (1 << CP0C1_FP) | (15 << CP0C1_MMU) |
+                       (0 << CP0C1_IS) | (3 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (0 << CP0C1_DS) | (3 << CP0C1_DL) | (1 << CP0C1_DA),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3 | (0 << CP0C3_VInt) | (1 << CP0C3_MT),
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 0,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        /* No DSP implemented. */
+        .CP0_Status_rw_bitmask = 0x3678FF1F,
+        /* No DSP implemented. */
+        .CP0_TCStatus_rw_bitmask = (0 << CP0TCSt_TCU3) | (0 << CP0TCSt_TCU2) |
+                    (1 << CP0TCSt_TCU1) | (1 << CP0TCSt_TCU0) |
+                    (0 << CP0TCSt_TMX) | (1 << CP0TCSt_DT) |
+                    (1 << CP0TCSt_DA) | (1 << CP0TCSt_A) |
+                    (0x3 << CP0TCSt_TKSU) | (1 << CP0TCSt_IXMT) |
+                    (0xff << CP0TCSt_TASID),
+        .CP1_fcr0 = (1 << FCR0_F64) | (1 << FCR0_L) | (1 << FCR0_W) |
+                    (1 << FCR0_D) | (1 << FCR0_S) | (0x95 << FCR0_PRID),
+        .CP0_SRSCtl = (0xf << CP0SRSCtl_HSS),
+        .CP0_SRSConf0_rw_bitmask = 0x3fffffff,
+        .CP0_SRSConf0 = (1 << CP0SRSC0_M) | (0x3fe << CP0SRSC0_SRS3) |
+                    (0x3fe << CP0SRSC0_SRS2) | (0x3fe << CP0SRSC0_SRS1),
+        .CP0_SRSConf1_rw_bitmask = 0x3fffffff,
+        .CP0_SRSConf1 = (1 << CP0SRSC1_M) | (0x3fe << CP0SRSC1_SRS6) |
+                    (0x3fe << CP0SRSC1_SRS5) | (0x3fe << CP0SRSC1_SRS4),
+        .CP0_SRSConf2_rw_bitmask = 0x3fffffff,
+        .CP0_SRSConf2 = (1 << CP0SRSC2_M) | (0x3fe << CP0SRSC2_SRS9) |
+                    (0x3fe << CP0SRSC2_SRS8) | (0x3fe << CP0SRSC2_SRS7),
+        .CP0_SRSConf3_rw_bitmask = 0x3fffffff,
+        .CP0_SRSConf3 = (1 << CP0SRSC3_M) | (0x3fe << CP0SRSC3_SRS12) |
+                    (0x3fe << CP0SRSC3_SRS11) | (0x3fe << CP0SRSC3_SRS10),
+        .CP0_SRSConf4_rw_bitmask = 0x3fffffff,
+        .CP0_SRSConf4 = (0x3fe << CP0SRSC4_SRS15) |
+                    (0x3fe << CP0SRSC4_SRS14) | (0x3fe << CP0SRSC4_SRS13),
+        .SEGBITS = 32,
+        .PABITS = 32,
+        .insn_flags = CPU_MIPS32R2 | ASE_MIPS16 | ASE_DSP | ASE_MT,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+#if defined(TARGET_MIPS64)
+    {
+        .name = "R4000",
+        .CP0_PRid = 0x00000400,
+        /* No L2 cache, icache size 8k, dcache size 8k, uncached coherency. */
+        .CP0_Config0 = (1 << 17) | (0x1 << 9) | (0x1 << 6) | (0x2 << CP0C0_K0),
+        /* Note: Config1 is only used internally, the R4000 has only Config0. */
+        .CP0_Config1 = (1 << CP0C1_FP) | (47 << CP0C1_MMU),
+        .CP0_LLAddr_rw_bitmask = 0xFFFFFFFF,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 16,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x3678FFFF,
+        /* The R4000 has a full 64bit FPU but doesn't use the fcr0 bits. */
+        .CP1_fcr0 = (0x5 << FCR0_PRID) | (0x0 << FCR0_REV),
+        .SEGBITS = 40,
+        .PABITS = 36,
+        .insn_flags = CPU_MIPS3,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+    {
+        .name = "VR5432",
+        .CP0_PRid = 0x00005400,
+        /* No L2 cache, icache size 8k, dcache size 8k, uncached coherency. */
+        .CP0_Config0 = (1 << 17) | (0x1 << 9) | (0x1 << 6) | (0x2 << CP0C0_K0),
+        .CP0_Config1 = (1 << CP0C1_FP) | (47 << CP0C1_MMU),
+        .CP0_LLAddr_rw_bitmask = 0xFFFFFFFFL,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 16,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x3678FFFF,
+        /* The VR5432 has a full 64bit FPU but doesn't use the fcr0 bits. */
+        .CP1_fcr0 = (0x54 << FCR0_PRID) | (0x0 << FCR0_REV),
+        .SEGBITS = 40,
+        .PABITS = 32,
+        .insn_flags = CPU_VR54XX,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+    {
+        .name = "5Kc",
+        .CP0_PRid = 0x00018100,
+        .CP0_Config0 = MIPS_CONFIG0 | (0x2 << CP0C0_AT) |
+                       (MMU_TYPE_R4000 << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 | (31 << CP0C1_MMU) |
+                       (1 << CP0C1_IS) | (4 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (1 << CP0C1_DS) | (4 << CP0C1_DL) | (1 << CP0C1_DA) |
+                       (1 << CP0C1_PC) | (1 << CP0C1_WR) | (1 << CP0C1_EP),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3,
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x32F8FFFF,
+        .SEGBITS = 42,
+        .PABITS = 36,
+        .insn_flags = CPU_MIPS64,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+    {
+        .name = "5Kf",
+        .CP0_PRid = 0x00018100,
+        .CP0_Config0 = MIPS_CONFIG0 | (0x2 << CP0C0_AT) |
+                       (MMU_TYPE_R4000 << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 | (1 << CP0C1_FP) | (31 << CP0C1_MMU) |
+                       (1 << CP0C1_IS) | (4 << CP0C1_IL) | (1 << CP0C1_IA) |
+                       (1 << CP0C1_DS) | (4 << CP0C1_DL) | (1 << CP0C1_DA) |
+                       (1 << CP0C1_PC) | (1 << CP0C1_WR) | (1 << CP0C1_EP),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3,
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 4,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x36F8FFFF,
+        /* The 5Kf has F64 / L / W but doesn't use the fcr0 bits. */
+        .CP1_fcr0 = (1 << FCR0_D) | (1 << FCR0_S) |
+                    (0x81 << FCR0_PRID) | (0x0 << FCR0_REV),
+        .SEGBITS = 42,
+        .PABITS = 36,
+        .insn_flags = CPU_MIPS64,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+    {
+        .name = "20Kc",
+        /* We emulate a later version of the 20Kc, earlier ones had a broken
+           WAIT instruction. */
+        .CP0_PRid = 0x000182a0,
+        .CP0_Config0 = MIPS_CONFIG0 | (0x2 << CP0C0_AT) |
+                    (MMU_TYPE_R4000 << CP0C0_MT) | (1 << CP0C0_VI),
+        .CP0_Config1 = MIPS_CONFIG1 | (1 << CP0C1_FP) | (47 << CP0C1_MMU) |
+                       (2 << CP0C1_IS) | (4 << CP0C1_IL) | (3 << CP0C1_IA) |
+                       (2 << CP0C1_DS) | (4 << CP0C1_DL) | (3 << CP0C1_DA) |
+                       (1 << CP0C1_PC) | (1 << CP0C1_WR) | (1 << CP0C1_EP),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3,
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 0,
+        .SYNCI_Step = 32,
+        .CCRes = 1,
+        .CP0_Status_rw_bitmask = 0x36FBFFFF,
+        /* The 20Kc has F64 / L / W but doesn't use the fcr0 bits. */
+        .CP1_fcr0 = (1 << FCR0_3D) | (1 << FCR0_PS) |
+                    (1 << FCR0_D) | (1 << FCR0_S) |
+                    (0x82 << FCR0_PRID) | (0x0 << FCR0_REV),
+        .SEGBITS = 40,
+        .PABITS = 36,
+        .insn_flags = CPU_MIPS64 | ASE_MIPS3D,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+    {
+        /* A generic CPU providing MIPS64 Release 2 features.
+           FIXME: Eventually this should be replaced by a real CPU model. */
+        .name = "MIPS64R2-generic",
+        .CP0_PRid = 0x00010000,
+        .CP0_Config0 = MIPS_CONFIG0 | (0x1 << CP0C0_AR) | (0x2 << CP0C0_AT) |
+                       (MMU_TYPE_R4000 << CP0C0_MT),
+        .CP0_Config1 = MIPS_CONFIG1 | (1 << CP0C1_FP) | (63 << CP0C1_MMU) |
+                       (2 << CP0C1_IS) | (4 << CP0C1_IL) | (3 << CP0C1_IA) |
+                       (2 << CP0C1_DS) | (4 << CP0C1_DL) | (3 << CP0C1_DA) |
+                       (1 << CP0C1_PC) | (1 << CP0C1_WR) | (1 << CP0C1_EP),
+        .CP0_Config2 = MIPS_CONFIG2,
+        .CP0_Config3 = MIPS_CONFIG3 | (1 << CP0C3_LPA),
+        .CP0_LLAddr_rw_bitmask = 0,
+        .CP0_LLAddr_shift = 0,
+        .SYNCI_Step = 32,
+        .CCRes = 2,
+        .CP0_Status_rw_bitmask = 0x36FBFFFF,
+        .CP1_fcr0 = (1 << FCR0_F64) | (1 << FCR0_3D) | (1 << FCR0_PS) |
+                    (1 << FCR0_L) | (1 << FCR0_W) | (1 << FCR0_D) |
+                    (1 << FCR0_S) | (0x00 << FCR0_PRID) | (0x0 << FCR0_REV),
+        .SEGBITS = 42,
+        /* The architectural limit is 59, but we have hardcoded 36 bit
+           in some places...
+        .PABITS = 59, */ /* the architectural limit */
+        .PABITS = 36,
+        .insn_flags = CPU_MIPS64R2 | ASE_MIPS3D,
+        .mmu_type = MMU_TYPE_R4000,
+    },
+#endif
+};
+
+static const mips_def_t *cpu_mips_find_by_name (const char *name)
+{
+    int i;
+
+    for (i = 0; i < ARRAY_SIZE(mips_defs); i++) {
+        if (strcasecmp(name, mips_defs[i].name) == 0) {
+            return &mips_defs[i];
+        }
+    }
+    return NULL;
+}
+
+void mips_cpu_list (FILE *f, int (*cpu_fprintf)(FILE *f, const char *fmt, ...))
+{
+    int i;
+
+    for (i = 0; i < ARRAY_SIZE(mips_defs); i++) {
+        (*cpu_fprintf)(f, "MIPS '%s'\n",
+                       mips_defs[i].name);
+    }
+}
+
+#ifndef CONFIG_USER_ONLY
+static void no_mmu_init (CPUMIPSState *env, const mips_def_t *def)
+{
+    env->tlb->nb_tlb = 1;
+    env->tlb->map_address = &no_mmu_map_address;
+}
+
+static void fixed_mmu_init (CPUMIPSState *env, const mips_def_t *def)
+{
+    env->tlb->nb_tlb = 1;
+    env->tlb->map_address = &fixed_mmu_map_address;
+}
+
+static void r4k_mmu_init (CPUMIPSState *env, const mips_def_t *def)
+{
+    env->tlb->nb_tlb = 1 + ((def->CP0_Config1 >> CP0C1_MMU) & 63);
+    env->tlb->map_address = &r4k_map_address;
+    env->tlb->helper_tlbwi = r4k_helper_tlbwi;
+    env->tlb->helper_tlbwr = r4k_helper_tlbwr;
+    env->tlb->helper_tlbp = r4k_helper_tlbp;
+    env->tlb->helper_tlbr = r4k_helper_tlbr;
+}
+
+static void mmu_init (CPUMIPSState *env, const mips_def_t *def)
+{
+    env->tlb = qemu_mallocz(sizeof(CPUMIPSTLBContext));
+
+    switch (def->mmu_type) {
+        case MMU_TYPE_NONE:
+            no_mmu_init(env, def);
+            break;
+        case MMU_TYPE_R4000:
+            r4k_mmu_init(env, def);
+            break;
+        case MMU_TYPE_FMT:
+            fixed_mmu_init(env, def);
+            break;
+        case MMU_TYPE_R3000:
+        case MMU_TYPE_R6000:
+        case MMU_TYPE_R8000:
+        default:
+            cpu_abort(env, "MMU type not supported\n");
+    }
+}
+#endif /* CONFIG_USER_ONLY */
+
+static void fpu_init (CPUMIPSState *env, const mips_def_t *def)
+{
+    int i;
+
+    for (i = 0; i < MIPS_FPU_MAX; i++)
+        env->fpus[i].fcr0 = def->CP1_fcr0;
+
+    memcpy(&env->active_fpu, &env->fpus[0], sizeof(env->active_fpu));
+}
+
+static void mvp_init (CPUMIPSState *env, const mips_def_t *def)
+{
+    env->mvp = qemu_mallocz(sizeof(CPUMIPSMVPContext));
+
+    /* MVPConf1 implemented, TLB sharable, no gating storage support,
+       programmable cache partitioning implemented, number of allocatable
+       and sharable TLB entries, MVP has allocatable TCs, 2 VPEs
+       implemented, 5 TCs implemented. */
+    env->mvp->CP0_MVPConf0 = (1 << CP0MVPC0_M) | (1 << CP0MVPC0_TLBS) |
+                             (0 << CP0MVPC0_GS) | (1 << CP0MVPC0_PCP) |
+// TODO: actually do 2 VPEs.
+//                             (1 << CP0MVPC0_TCA) | (0x1 << CP0MVPC0_PVPE) |
+//                             (0x04 << CP0MVPC0_PTC);
+                             (1 << CP0MVPC0_TCA) | (0x0 << CP0MVPC0_PVPE) |
+                             (0x04 << CP0MVPC0_PTC);
+#if !defined(CONFIG_USER_ONLY)
+    /* Usermode has no TLB support */
+    env->mvp->CP0_MVPConf0 |= (env->tlb->nb_tlb << CP0MVPC0_PTLBE);
+#endif
+
+    /* Allocatable CP1 have media extensions, allocatable CP1 have FP support,
+       no UDI implemented, no CP2 implemented, 1 CP1 implemented. */
+    env->mvp->CP0_MVPConf1 = (1 << CP0MVPC1_CIM) | (1 << CP0MVPC1_CIF) |
+                             (0x0 << CP0MVPC1_PCX) | (0x0 << CP0MVPC1_PCP2) |
+                             (0x1 << CP0MVPC1_PCP1);
+}