/*
 *  S/390 translation
 *
 *  Copyright (c) 2009 Ulrich Hecht
 *  Copyright (c) 2010 Alexander Graf
 *
 * 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 2.1 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/>.
 */

/* #define DEBUG_INLINE_BRANCHES */
#define S390X_DEBUG_DISAS
/* #define S390X_DEBUG_DISAS_VERBOSE */

#ifdef S390X_DEBUG_DISAS_VERBOSE
#  define LOG_DISAS(...) qemu_log(__VA_ARGS__)
#else
#  define LOG_DISAS(...) do { } while (0)
#endif

#include "qemu/osdep.h"
#include "cpu.h"
#include "s390x-internal.h"
#include "disas/disas.h"
#include "exec/exec-all.h"
#include "tcg/tcg-op.h"
#include "tcg/tcg-op-gvec.h"
#include "qemu/log.h"
#include "qemu/host-utils.h"
#include "exec/cpu_ldst.h"
#include "exec/gen-icount.h"
#include "exec/helper-proto.h"
#include "exec/helper-gen.h"

#include "exec/translator.h"
#include "exec/log.h"
#include "qemu/atomic128.h"


/* Information that (most) every instruction needs to manipulate.  */
typedef struct DisasContext DisasContext;
typedef struct DisasInsn DisasInsn;
typedef struct DisasFields DisasFields;

/*
 * Define a structure to hold the decoded fields.  We'll store each inside
 * an array indexed by an enum.  In order to conserve memory, we'll arrange
 * for fields that do not exist at the same time to overlap, thus the "C"
 * for compact.  For checking purposes there is an "O" for original index
 * as well that will be applied to availability bitmaps.
 */

enum DisasFieldIndexO {
    FLD_O_r1,
    FLD_O_r2,
    FLD_O_r3,
    FLD_O_m1,
    FLD_O_m3,
    FLD_O_m4,
    FLD_O_m5,
    FLD_O_m6,
    FLD_O_b1,
    FLD_O_b2,
    FLD_O_b4,
    FLD_O_d1,
    FLD_O_d2,
    FLD_O_d4,
    FLD_O_x2,
    FLD_O_l1,
    FLD_O_l2,
    FLD_O_i1,
    FLD_O_i2,
    FLD_O_i3,
    FLD_O_i4,
    FLD_O_i5,
    FLD_O_v1,
    FLD_O_v2,
    FLD_O_v3,
    FLD_O_v4,
};

enum DisasFieldIndexC {
    FLD_C_r1 = 0,
    FLD_C_m1 = 0,
    FLD_C_b1 = 0,
    FLD_C_i1 = 0,
    FLD_C_v1 = 0,

    FLD_C_r2 = 1,
    FLD_C_b2 = 1,
    FLD_C_i2 = 1,

    FLD_C_r3 = 2,
    FLD_C_m3 = 2,
    FLD_C_i3 = 2,
    FLD_C_v3 = 2,

    FLD_C_m4 = 3,
    FLD_C_b4 = 3,
    FLD_C_i4 = 3,
    FLD_C_l1 = 3,
    FLD_C_v4 = 3,

    FLD_C_i5 = 4,
    FLD_C_d1 = 4,
    FLD_C_m5 = 4,

    FLD_C_d2 = 5,
    FLD_C_m6 = 5,

    FLD_C_d4 = 6,
    FLD_C_x2 = 6,
    FLD_C_l2 = 6,
    FLD_C_v2 = 6,

    NUM_C_FIELD = 7
};

struct DisasFields {
    uint64_t raw_insn;
    unsigned op:8;
    unsigned op2:8;
    unsigned presentC:16;
    unsigned int presentO;
    int c[NUM_C_FIELD];
};

struct DisasContext {
    DisasContextBase base;
    const DisasInsn *insn;
    TCGOp *insn_start;
    DisasFields fields;
    uint64_t ex_value;
    /*
     * During translate_one(), pc_tmp is used to determine the instruction
     * to be executed after base.pc_next - e.g. next sequential instruction
     * or a branch target.
     */
    uint64_t pc_tmp;
    uint32_t ilen;
    enum cc_op cc_op;
};

/* Information carried about a condition to be evaluated.  */
typedef struct {
    TCGCond cond:8;
    bool is_64;
    bool g1;
    bool g2;
    union {
        struct { TCGv_i64 a, b; } s64;
        struct { TCGv_i32 a, b; } s32;
    } u;
} DisasCompare;

#ifdef DEBUG_INLINE_BRANCHES
static uint64_t inline_branch_hit[CC_OP_MAX];
static uint64_t inline_branch_miss[CC_OP_MAX];
#endif

static void pc_to_link_info(TCGv_i64 out, DisasContext *s, uint64_t pc)
{
    TCGv_i64 tmp;

    if (s->base.tb->flags & FLAG_MASK_32) {
        if (s->base.tb->flags & FLAG_MASK_64) {
            tcg_gen_movi_i64(out, pc);
            return;
        }
        pc |= 0x80000000;
    }
    assert(!(s->base.tb->flags & FLAG_MASK_64));
    tmp = tcg_const_i64(pc);
    tcg_gen_deposit_i64(out, out, tmp, 0, 32);
    tcg_temp_free_i64(tmp);
}

static TCGv_i64 psw_addr;
static TCGv_i64 psw_mask;
static TCGv_i64 gbea;

static TCGv_i32 cc_op;
static TCGv_i64 cc_src;
static TCGv_i64 cc_dst;
static TCGv_i64 cc_vr;

static char cpu_reg_names[16][4];
static TCGv_i64 regs[16];

void s390x_translate_init(void)
{
    int i;

    psw_addr = tcg_global_mem_new_i64(cpu_env,
                                      offsetof(CPUS390XState, psw.addr),
                                      "psw_addr");
    psw_mask = tcg_global_mem_new_i64(cpu_env,
                                      offsetof(CPUS390XState, psw.mask),
                                      "psw_mask");
    gbea = tcg_global_mem_new_i64(cpu_env,
                                  offsetof(CPUS390XState, gbea),
                                  "gbea");

    cc_op = tcg_global_mem_new_i32(cpu_env, offsetof(CPUS390XState, cc_op),
                                   "cc_op");
    cc_src = tcg_global_mem_new_i64(cpu_env, offsetof(CPUS390XState, cc_src),
                                    "cc_src");
    cc_dst = tcg_global_mem_new_i64(cpu_env, offsetof(CPUS390XState, cc_dst),
                                    "cc_dst");
    cc_vr = tcg_global_mem_new_i64(cpu_env, offsetof(CPUS390XState, cc_vr),
                                   "cc_vr");

    for (i = 0; i < 16; i++) {
        snprintf(cpu_reg_names[i], sizeof(cpu_reg_names[0]), "r%d", i);
        regs[i] = tcg_global_mem_new(cpu_env,
                                     offsetof(CPUS390XState, regs[i]),
                                     cpu_reg_names[i]);
    }
}

static inline int vec_full_reg_offset(uint8_t reg)
{
    g_assert(reg < 32);
    return offsetof(CPUS390XState, vregs[reg][0]);
}

static inline int vec_reg_offset(uint8_t reg, uint8_t enr, MemOp es)
{
    /* Convert element size (es) - e.g. MO_8 - to bytes */
    const uint8_t bytes = 1 << es;
    int offs = enr * bytes;

    /*
     * vregs[n][0] is the lowest 8 byte and vregs[n][1] the highest 8 byte
     * of the 16 byte vector, on both, little and big endian systems.
     *
     * Big Endian (target/possible host)
     * B:  [ 0][ 1][ 2][ 3][ 4][ 5][ 6][ 7] - [ 8][ 9][10][11][12][13][14][15]
     * HW: [     0][     1][     2][     3] - [     4][     5][     6][     7]
     * W:  [             0][             1] - [             2][             3]
     * DW: [                             0] - [                             1]
     *
     * Little Endian (possible host)
     * B:  [ 7][ 6][ 5][ 4][ 3][ 2][ 1][ 0] - [15][14][13][12][11][10][ 9][ 8]
     * HW: [     3][     2][     1][     0] - [     7][     6][     5][     4]
     * W:  [             1][             0] - [             3][             2]
     * DW: [                             0] - [                             1]
     *
     * For 16 byte elements, the two 8 byte halves will not form a host
     * int128 if the host is little endian, since they're in the wrong order.
     * Some operations (e.g. xor) do not care. For operations like addition,
     * the two 8 byte elements have to be loaded separately. Let's force all
     * 16 byte operations to handle it in a special way.
     */
    g_assert(es <= MO_64);
#ifndef HOST_WORDS_BIGENDIAN
    offs ^= (8 - bytes);
#endif
    return offs + vec_full_reg_offset(reg);
}

static inline int freg64_offset(uint8_t reg)
{
    g_assert(reg < 16);
    return vec_reg_offset(reg, 0, MO_64);
}

static inline int freg32_offset(uint8_t reg)
{
    g_assert(reg < 16);
    return vec_reg_offset(reg, 0, MO_32);
}

static TCGv_i64 load_reg(int reg)
{
    TCGv_i64 r = tcg_temp_new_i64();
    tcg_gen_mov_i64(r, regs[reg]);
    return r;
}

static TCGv_i64 load_freg(int reg)
{
    TCGv_i64 r = tcg_temp_new_i64();

    tcg_gen_ld_i64(r, cpu_env, freg64_offset(reg));
    return r;
}

static TCGv_i64 load_freg32_i64(int reg)
{
    TCGv_i64 r = tcg_temp_new_i64();

    tcg_gen_ld32u_i64(r, cpu_env, freg32_offset(reg));
    return r;
}

static void store_reg(int reg, TCGv_i64 v)
{
    tcg_gen_mov_i64(regs[reg], v);
}

static void store_freg(int reg, TCGv_i64 v)
{
    tcg_gen_st_i64(v, cpu_env, freg64_offset(reg));
}

static void store_reg32_i64(int reg, TCGv_i64 v)
{
    /* 32 bit register writes keep the upper half */
    tcg_gen_deposit_i64(regs[reg], regs[reg], v, 0, 32);
}

static void store_reg32h_i64(int reg, TCGv_i64 v)
{
    tcg_gen_deposit_i64(regs[reg], regs[reg], v, 32, 32);
}

static void store_freg32_i64(int reg, TCGv_i64 v)
{
    tcg_gen_st32_i64(v, cpu_env, freg32_offset(reg));
}

static void return_low128(TCGv_i64 dest)
{
    tcg_gen_ld_i64(dest, cpu_env, offsetof(CPUS390XState, retxl));
}

static void update_psw_addr(DisasContext *s)
{
    /* psw.addr */
    tcg_gen_movi_i64(psw_addr, s->base.pc_next);
}

static void per_branch(DisasContext *s, bool to_next)
{
#ifndef CONFIG_USER_ONLY
    tcg_gen_movi_i64(gbea, s->base.pc_next);

    if (s->base.tb->flags & FLAG_MASK_PER) {
        TCGv_i64 next_pc = to_next ? tcg_const_i64(s->pc_tmp) : psw_addr;
        gen_helper_per_branch(cpu_env, gbea, next_pc);
        if (to_next) {
            tcg_temp_free_i64(next_pc);
        }
    }
#endif
}

static void per_branch_cond(DisasContext *s, TCGCond cond,
                            TCGv_i64 arg1, TCGv_i64 arg2)
{
#ifndef CONFIG_USER_ONLY
    if (s->base.tb->flags & FLAG_MASK_PER) {
        TCGLabel *lab = gen_new_label();
        tcg_gen_brcond_i64(tcg_invert_cond(cond), arg1, arg2, lab);

        tcg_gen_movi_i64(gbea, s->base.pc_next);
        gen_helper_per_branch(cpu_env, gbea, psw_addr);

        gen_set_label(lab);
    } else {
        TCGv_i64 pc = tcg_const_i64(s->base.pc_next);
        tcg_gen_movcond_i64(cond, gbea, arg1, arg2, gbea, pc);
        tcg_temp_free_i64(pc);
    }
#endif
}

static void per_breaking_event(DisasContext *s)
{
    tcg_gen_movi_i64(gbea, s->base.pc_next);
}

static void update_cc_op(DisasContext *s)
{
    if (s->cc_op != CC_OP_DYNAMIC && s->cc_op != CC_OP_STATIC) {
        tcg_gen_movi_i32(cc_op, s->cc_op);
    }
}

static inline uint64_t ld_code2(CPUS390XState *env, DisasContext *s,
                                uint64_t pc)
{
    return (uint64_t)translator_lduw(env, &s->base, pc);
}

static inline uint64_t ld_code4(CPUS390XState *env, DisasContext *s,
                                uint64_t pc)
{
    return (uint64_t)(uint32_t)translator_ldl(env, &s->base, pc);
}

static int get_mem_index(DisasContext *s)
{
#ifdef CONFIG_USER_ONLY
    return MMU_USER_IDX;
#else
    if (!(s->base.tb->flags & FLAG_MASK_DAT)) {
        return MMU_REAL_IDX;
    }

    switch (s->base.tb->flags & FLAG_MASK_ASC) {
    case PSW_ASC_PRIMARY >> FLAG_MASK_PSW_SHIFT:
        return MMU_PRIMARY_IDX;
    case PSW_ASC_SECONDARY >> FLAG_MASK_PSW_SHIFT:
        return MMU_SECONDARY_IDX;
    case PSW_ASC_HOME >> FLAG_MASK_PSW_SHIFT:
        return MMU_HOME_IDX;
    default:
        tcg_abort();
        break;
    }
#endif
}

static void gen_exception(int excp)
{
    TCGv_i32 tmp = tcg_const_i32(excp);
    gen_helper_exception(cpu_env, tmp);
    tcg_temp_free_i32(tmp);
}

static void gen_program_exception(DisasContext *s, int code)
{
    TCGv_i32 tmp;

    /* Remember what pgm exeption this was.  */
    tmp = tcg_const_i32(code);
    tcg_gen_st_i32(tmp, cpu_env, offsetof(CPUS390XState, int_pgm_code));
    tcg_temp_free_i32(tmp);

    tmp = tcg_const_i32(s->ilen);
    tcg_gen_st_i32(tmp, cpu_env, offsetof(CPUS390XState, int_pgm_ilen));
    tcg_temp_free_i32(tmp);

    /* update the psw */
    update_psw_addr(s);

    /* Save off cc.  */
    update_cc_op(s);

    /* Trigger exception.  */
    gen_exception(EXCP_PGM);
}

static inline void gen_illegal_opcode(DisasContext *s)
{
    gen_program_exception(s, PGM_OPERATION);
}

static inline void gen_data_exception(uint8_t dxc)
{
    TCGv_i32 tmp = tcg_const_i32(dxc);
    gen_helper_data_exception(cpu_env, tmp);
    tcg_temp_free_i32(tmp);
}

static inline void gen_trap(DisasContext *s)
{
    /* Set DXC to 0xff */
    gen_data_exception(0xff);
}

static void gen_addi_and_wrap_i64(DisasContext *s, TCGv_i64 dst, TCGv_i64 src,
                                  int64_t imm)
{
    tcg_gen_addi_i64(dst, src, imm);
    if (!(s->base.tb->flags & FLAG_MASK_64)) {
        if (s->base.tb->flags & FLAG_MASK_32) {
            tcg_gen_andi_i64(dst, dst, 0x7fffffff);
        } else {
            tcg_gen_andi_i64(dst, dst, 0x00ffffff);
        }
    }
}

static TCGv_i64 get_address(DisasContext *s, int x2, int b2, int d2)
{
    TCGv_i64 tmp = tcg_temp_new_i64();

    /*
     * Note that d2 is limited to 20 bits, signed.  If we crop negative
     * displacements early we create larger immedate addends.
     */
    if (b2 && x2) {
        tcg_gen_add_i64(tmp, regs[b2], regs[x2]);
        gen_addi_and_wrap_i64(s, tmp, tmp, d2);
    } else if (b2) {
        gen_addi_and_wrap_i64(s, tmp, regs[b2], d2);
    } else if (x2) {
        gen_addi_and_wrap_i64(s, tmp, regs[x2], d2);
    } else if (!(s->base.tb->flags & FLAG_MASK_64)) {
        if (s->base.tb->flags & FLAG_MASK_32) {
            tcg_gen_movi_i64(tmp, d2 & 0x7fffffff);
        } else {
            tcg_gen_movi_i64(tmp, d2 & 0x00ffffff);
        }
    } else {
        tcg_gen_movi_i64(tmp, d2);
    }

    return tmp;
}

static inline bool live_cc_data(DisasContext *s)
{
    return (s->cc_op != CC_OP_DYNAMIC
            && s->cc_op != CC_OP_STATIC
            && s->cc_op > 3);
}

static inline void gen_op_movi_cc(DisasContext *s, uint32_t val)
{
    if (live_cc_data(s)) {
        tcg_gen_discard_i64(cc_src);
        tcg_gen_discard_i64(cc_dst);
        tcg_gen_discard_i64(cc_vr);
    }
    s->cc_op = CC_OP_CONST0 + val;
}

static void gen_op_update1_cc_i64(DisasContext *s, enum cc_op op, TCGv_i64 dst)
{
    if (live_cc_data(s)) {
        tcg_gen_discard_i64(cc_src);
        tcg_gen_discard_i64(cc_vr);
    }
    tcg_gen_mov_i64(cc_dst, dst);
    s->cc_op = op;
}

static void gen_op_update2_cc_i64(DisasContext *s, enum cc_op op, TCGv_i64 src,
                                  TCGv_i64 dst)
{
    if (live_cc_data(s)) {
        tcg_gen_discard_i64(cc_vr);
    }
    tcg_gen_mov_i64(cc_src, src);
    tcg_gen_mov_i64(cc_dst, dst);
    s->cc_op = op;
}

static void gen_op_update3_cc_i64(DisasContext *s, enum cc_op op, TCGv_i64 src,
                                  TCGv_i64 dst, TCGv_i64 vr)
{
    tcg_gen_mov_i64(cc_src, src);
    tcg_gen_mov_i64(cc_dst, dst);
    tcg_gen_mov_i64(cc_vr, vr);
    s->cc_op = op;
}

static void set_cc_nz_u64(DisasContext *s, TCGv_i64 val)
{
    gen_op_update1_cc_i64(s, CC_OP_NZ, val);
}

/* CC value is in env->cc_op */
static void set_cc_static(DisasContext *s)
{
    if (live_cc_data(s)) {
        tcg_gen_discard_i64(cc_src);
        tcg_gen_discard_i64(cc_dst);
        tcg_gen_discard_i64(cc_vr);
    }
    s->cc_op = CC_OP_STATIC;
}

/* calculates cc into cc_op */
static void gen_op_calc_cc(DisasContext *s)
{
    TCGv_i32 local_cc_op = NULL;
    TCGv_i64 dummy = NULL;

    switch (s->cc_op) {
    default:
        dummy = tcg_const_i64(0);
        /* FALLTHRU */
    case CC_OP_ADD_64:
    case CC_OP_SUB_64:
    case CC_OP_ADD_32:
    case CC_OP_SUB_32:
        local_cc_op = tcg_const_i32(s->cc_op);
        break;
    case CC_OP_CONST0:
    case CC_OP_CONST1:
    case CC_OP_CONST2:
    case CC_OP_CONST3:
    case CC_OP_STATIC:
    case CC_OP_DYNAMIC:
        break;
    }

    switch (s->cc_op) {
    case CC_OP_CONST0:
    case CC_OP_CONST1:
    case CC_OP_CONST2:
    case CC_OP_CONST3:
        /* s->cc_op is the cc value */
        tcg_gen_movi_i32(cc_op, s->cc_op - CC_OP_CONST0);
        break;
    case CC_OP_STATIC:
        /* env->cc_op already is the cc value */
        break;
    case CC_OP_NZ:
    case CC_OP_ABS_64:
    case CC_OP_NABS_64:
    case CC_OP_ABS_32:
    case CC_OP_NABS_32:
    case CC_OP_LTGT0_32:
    case CC_OP_LTGT0_64:
    case CC_OP_COMP_32:
    case CC_OP_COMP_64:
    case CC_OP_NZ_F32:
    case CC_OP_NZ_F64:
    case CC_OP_FLOGR:
    case CC_OP_LCBB:
    case CC_OP_MULS_32:
        /* 1 argument */
        gen_helper_calc_cc(cc_op, cpu_env, local_cc_op, dummy, cc_dst, dummy);
        break;
    case CC_OP_ADDU:
    case CC_OP_ICM:
    case CC_OP_LTGT_32:
    case CC_OP_LTGT_64:
    case CC_OP_LTUGTU_32:
    case CC_OP_LTUGTU_64:
    case CC_OP_TM_32:
    case CC_OP_TM_64:
    case CC_OP_SLA:
    case CC_OP_SUBU:
    case CC_OP_NZ_F128:
    case CC_OP_VC:
    case CC_OP_MULS_64:
        /* 2 arguments */
        gen_helper_calc_cc(cc_op, cpu_env, local_cc_op, cc_src, cc_dst, dummy);
        break;
    case CC_OP_ADD_64:
    case CC_OP_SUB_64:
    case CC_OP_ADD_32:
    case CC_OP_SUB_32:
        /* 3 arguments */
        gen_helper_calc_cc(cc_op, cpu_env, local_cc_op, cc_src, cc_dst, cc_vr);
        break;
    case CC_OP_DYNAMIC:
        /* unknown operation - assume 3 arguments and cc_op in env */
        gen_helper_calc_cc(cc_op, cpu_env, cc_op, cc_src, cc_dst, cc_vr);
        break;
    default:
        tcg_abort();
    }

    if (local_cc_op) {
        tcg_temp_free_i32(local_cc_op);
    }
    if (dummy) {
        tcg_temp_free_i64(dummy);
    }

    /* We now have cc in cc_op as constant */
    set_cc_static(s);
}

static bool use_goto_tb(DisasContext *s, uint64_t dest)
{
    if (unlikely(s->base.tb->flags & FLAG_MASK_PER)) {
        return false;
    }
    return translator_use_goto_tb(&s->base, dest);
}

static void account_noninline_branch(DisasContext *s, int cc_op)
{
#ifdef DEBUG_INLINE_BRANCHES
    inline_branch_miss[cc_op]++;
#endif
}

static void account_inline_branch(DisasContext *s, int cc_op)
{
#ifdef DEBUG_INLINE_BRANCHES
    inline_branch_hit[cc_op]++;
#endif
}

/* Table of mask values to comparison codes, given a comparison as input.
   For such, CC=3 should not be possible.  */
static const TCGCond ltgt_cond[16] = {
    TCG_COND_NEVER,  TCG_COND_NEVER,     /*    |    |    | x */
    TCG_COND_GT,     TCG_COND_GT,        /*    |    | GT | x */
    TCG_COND_LT,     TCG_COND_LT,        /*    | LT |    | x */
    TCG_COND_NE,     TCG_COND_NE,        /*    | LT | GT | x */
    TCG_COND_EQ,     TCG_COND_EQ,        /* EQ |    |    | x */
    TCG_COND_GE,     TCG_COND_GE,        /* EQ |    | GT | x */
    TCG_COND_LE,     TCG_COND_LE,        /* EQ | LT |    | x */
    TCG_COND_ALWAYS, TCG_COND_ALWAYS,    /* EQ | LT | GT | x */
};

/* Table of mask values to comparison codes, given a logic op as input.
   For such, only CC=0 and CC=1 should be possible.  */
static const TCGCond nz_cond[16] = {
    TCG_COND_NEVER, TCG_COND_NEVER,      /*    |    | x | x */
    TCG_COND_NEVER, TCG_COND_NEVER,
    TCG_COND_NE, TCG_COND_NE,            /*    | NE | x | x */
    TCG_COND_NE, TCG_COND_NE,
    TCG_COND_EQ, TCG_COND_EQ,            /* EQ |    | x | x */
    TCG_COND_EQ, TCG_COND_EQ,
    TCG_COND_ALWAYS, TCG_COND_ALWAYS,    /* EQ | NE | x | x */
    TCG_COND_ALWAYS, TCG_COND_ALWAYS,
};

/* Interpret MASK in terms of S->CC_OP, and fill in C with all the
   details required to generate a TCG comparison.  */
static void disas_jcc(DisasContext *s, DisasCompare *c, uint32_t mask)
{
    TCGCond cond;
    enum cc_op old_cc_op = s->cc_op;

    if (mask == 15 || mask == 0) {
        c->cond = (mask ? TCG_COND_ALWAYS : TCG_COND_NEVER);
        c->u.s32.a = cc_op;
        c->u.s32.b = cc_op;
        c->g1 = c->g2 = true;
        c->is_64 = false;
        return;
    }

    /* Find the TCG condition for the mask + cc op.  */
    switch (old_cc_op) {
    case CC_OP_LTGT0_32:
    case CC_OP_LTGT0_64:
    case CC_OP_LTGT_32:
    case CC_OP_LTGT_64:
        cond = ltgt_cond[mask];
        if (cond == TCG_COND_NEVER) {
            goto do_dynamic;
        }
        account_inline_branch(s, old_cc_op);
        break;

    case CC_OP_LTUGTU_32:
    case CC_OP_LTUGTU_64:
        cond = tcg_unsigned_cond(ltgt_cond[mask]);
        if (cond == TCG_COND_NEVER) {
            goto do_dynamic;
        }
        account_inline_branch(s, old_cc_op);
        break;

    case CC_OP_NZ:
        cond = nz_cond[mask];
        if (cond == TCG_COND_NEVER) {
            goto do_dynamic;
        }
        account_inline_branch(s, old_cc_op);
        break;

    case CC_OP_TM_32:
    case CC_OP_TM_64:
        switch (mask) {
        case 8:
            cond = TCG_COND_EQ;
            break;
        case 4 | 2 | 1:
            cond = TCG_COND_NE;
            break;
        default:
            goto do_dynamic;
        }
        account_inline_branch(s, old_cc_op);
        break;

    case CC_OP_ICM:
        switch (mask) {
        case 8:
            cond = TCG_COND_EQ;
            break;
        case 4 | 2 | 1:
        case 4 | 2:
            cond = TCG_COND_NE;
            break;
        default:
            goto do_dynamic;
        }
        account_inline_branch(s, old_cc_op);
        break;

    case CC_OP_FLOGR:
        switch (mask & 0xa) {
        case 8: /* src == 0 -> no one bit found */
            cond = TCG_COND_EQ;
            break;
        case 2: /* src != 0 -> one bit found */
            cond = TCG_COND_NE;
            break;
        default:
            goto do_dynamic;
        }
        account_inline_branch(s, old_cc_op);
        break;

    case CC_OP_ADDU:
    case CC_OP_SUBU:
        switch (mask) {
        case 8 | 2: /* result == 0 */
            cond = TCG_COND_EQ;
            break;
        case 4 | 1: /* result != 0 */
            cond = TCG_COND_NE;
            break;
        case 8 | 4: /* !carry (borrow) */
            cond = old_cc_op == CC_OP_ADDU ? TCG_COND_EQ : TCG_COND_NE;
            break;
        case 2 | 1: /* carry (!borrow) */
            cond = old_cc_op == CC_OP_ADDU ? TCG_COND_NE : TCG_COND_EQ;
            break;
        default:
            goto do_dynamic;
        }
        account_inline_branch(s, old_cc_op);
        break;

    default:
    do_dynamic:
        /* Calculate cc value.  */
        gen_op_calc_cc(s);
        /* FALLTHRU */

    case CC_OP_STATIC:
        /* Jump based on CC.  We'll load up the real cond below;
           the assignment here merely avoids a compiler warning.  */
        account_noninline_branch(s, old_cc_op);
        old_cc_op = CC_OP_STATIC;
        cond = TCG_COND_NEVER;
        break;
    }

    /* Load up the arguments of the comparison.  */
    c->is_64 = true;
    c->g1 = c->g2 = false;
    switch (old_cc_op) {
    case CC_OP_LTGT0_32:
        c->is_64 = false;
        c->u.s32.a = tcg_temp_new_i32();
        tcg_gen_extrl_i64_i32(c->u.s32.a, cc_dst);
        c->u.s32.b = tcg_const_i32(0);
        break;
    case CC_OP_LTGT_32:
    case CC_OP_LTUGTU_32:
        c->is_64 = false;
        c->u.s32.a = tcg_temp_new_i32();
        tcg_gen_extrl_i64_i32(c->u.s32.a, cc_src);
        c->u.s32.b = tcg_temp_new_i32();
        tcg_gen_extrl_i64_i32(c->u.s32.b, cc_dst);
        break;

    case CC_OP_LTGT0_64:
    case CC_OP_NZ:
    case CC_OP_FLOGR:
        c->u.s64.a = cc_dst;
        c->u.s64.b = tcg_const_i64(0);
        c->g1 = true;
        break;
    case CC_OP_LTGT_64:
    case CC_OP_LTUGTU_64:
        c->u.s64.a = cc_src;
        c->u.s64.b = cc_dst;
        c->g1 = c->g2 = true;
        break;

    case CC_OP_TM_32:
    case CC_OP_TM_64:
    case CC_OP_ICM:
        c->u.s64.a = tcg_temp_new_i64();
        c->u.s64.b = tcg_const_i64(0);
        tcg_gen_and_i64(c->u.s64.a, cc_src, cc_dst);
        break;

    case CC_OP_ADDU:
    case CC_OP_SUBU:
        c->is_64 = true;
        c->u.s64.b = tcg_const_i64(0);
        c->g1 = true;
        switch (mask) {
        case 8 | 2:
        case 4 | 1: /* result */
            c->u.s64.a = cc_dst;
            break;
        case 8 | 4:
        case 2 | 1: /* carry */
            c->u.s64.a = cc_src;
            break;
        default:
            g_assert_not_reached();
        }
        break;

    case CC_OP_STATIC:
        c->is_64 = false;
        c->u.s32.a = cc_op;
        c->g1 = true;
        switch (mask) {
        case 0x8 | 0x4 | 0x2: /* cc != 3 */
            cond = TCG_COND_NE;
            c->u.s32.b = tcg_const_i32(3);
            break;
        case 0x8 | 0x4 | 0x1: /* cc != 2 */
            cond = TCG_COND_NE;
            c->u.s32.b = tcg_const_i32(2);
            break;
        case 0x8 | 0x2 | 0x1: /* cc != 1 */
            cond = TCG_COND_NE;
            c->u.s32.b = tcg_const_i32(1);
            break;
        case 0x8 | 0x2: /* cc == 0 || cc == 2 => (cc & 1) == 0 */
            cond = TCG_COND_EQ;
            c->g1 = false;
            c->u.s32.a = tcg_temp_new_i32();
            c->u.s32.b = tcg_const_i32(0);
            tcg_gen_andi_i32(c->u.s32.a, cc_op, 1);
            break;
        case 0x8 | 0x4: /* cc < 2 */
            cond = TCG_COND_LTU;
            c->u.s32.b = tcg_const_i32(2);
            break;
        case 0x8: /* cc == 0 */
            cond = TCG_COND_EQ;
            c->u.s32.b = tcg_const_i32(0);
            break;
        case 0x4 | 0x2 | 0x1: /* cc != 0 */
            cond = TCG_COND_NE;
            c->u.s32.b = tcg_const_i32(0);
            break;
        case 0x4 | 0x1: /* cc == 1 || cc == 3 => (cc & 1) != 0 */
            cond = TCG_COND_NE;
            c->g1 = false;
            c->u.s32.a = tcg_temp_new_i32();
            c->u.s32.b = tcg_const_i32(0);
            tcg_gen_andi_i32(c->u.s32.a, cc_op, 1);
            break;
        case 0x4: /* cc == 1 */
            cond = TCG_COND_EQ;
            c->u.s32.b = tcg_const_i32(1);
            break;
        case 0x2 | 0x1: /* cc > 1 */
            cond = TCG_COND_GTU;
            c->u.s32.b = tcg_const_i32(1);
            break;
        case 0x2: /* cc == 2 */
            cond = TCG_COND_EQ;
            c->u.s32.b = tcg_const_i32(2);
            break;
        case 0x1: /* cc == 3 */
            cond = TCG_COND_EQ;
            c->u.s32.b = tcg_const_i32(3);
            break;
        default:
            /* CC is masked by something else: (8 >> cc) & mask.  */
            cond = TCG_COND_NE;
            c->g1 = false;
            c->u.s32.a = tcg_const_i32(8);
            c->u.s32.b = tcg_const_i32(0);
            tcg_gen_shr_i32(c->u.s32.a, c->u.s32.a, cc_op);
            tcg_gen_andi_i32(c->u.s32.a, c->u.s32.a, mask);
            break;
        }
        break;

    default:
        abort();
    }
    c->cond = cond;
}

static void free_compare(DisasCompare *c)
{
    if (!c->g1) {
        if (c->is_64) {
            tcg_temp_free_i64(c->u.s64.a);
        } else {
            tcg_temp_free_i32(c->u.s32.a);
        }
    }
    if (!c->g2) {
        if (c->is_64) {
            tcg_temp_free_i64(c->u.s64.b);
        } else {
            tcg_temp_free_i32(c->u.s32.b);
        }
    }
}

/* ====================================================================== */
/* Define the insn format enumeration.  */
#define F0(N)                         FMT_##N,
#define F1(N, X1)                     F0(N)
#define F2(N, X1, X2)                 F0(N)
#define F3(N, X1, X2, X3)             F0(N)
#define F4(N, X1, X2, X3, X4)         F0(N)
#define F5(N, X1, X2, X3, X4, X5)     F0(N)
#define F6(N, X1, X2, X3, X4, X5, X6) F0(N)

typedef enum {
#include "insn-format.def"
} DisasFormat;

#undef F0
#undef F1
#undef F2
#undef F3
#undef F4
#undef F5
#undef F6

/* This is the way fields are to be accessed out of DisasFields.  */
#define have_field(S, F)  have_field1((S), FLD_O_##F)
#define get_field(S, F)   get_field1((S), FLD_O_##F, FLD_C_##F)

static bool have_field1(const DisasContext *s, enum DisasFieldIndexO c)
{
    return (s->fields.presentO >> c) & 1;
}

static int get_field1(const DisasContext *s, enum DisasFieldIndexO o,
                      enum DisasFieldIndexC c)
{
    assert(have_field1(s, o));
    return s->fields.c[c];
}

/* Describe the layout of each field in each format.  */
typedef struct DisasField {
    unsigned int beg:8;
    unsigned int size:8;
    unsigned int type:2;
    unsigned int indexC:6;
    enum DisasFieldIndexO indexO:8;
} DisasField;

typedef struct DisasFormatInfo {
    DisasField op[NUM_C_FIELD];
} DisasFormatInfo;

#define R(N, B)       {  B,  4, 0, FLD_C_r##N, FLD_O_r##N }
#define M(N, B)       {  B,  4, 0, FLD_C_m##N, FLD_O_m##N }
#define V(N, B)       {  B,  4, 3, FLD_C_v##N, FLD_O_v##N }
#define BD(N, BB, BD) { BB,  4, 0, FLD_C_b##N, FLD_O_b##N }, \
                      { BD, 12, 0, FLD_C_d##N, FLD_O_d##N }
#define BXD(N)        { 16,  4, 0, FLD_C_b##N, FLD_O_b##N }, \
                      { 12,  4, 0, FLD_C_x##N, FLD_O_x##N }, \
                      { 20, 12, 0, FLD_C_d##N, FLD_O_d##N }
#define BDL(N)        { 16,  4, 0, FLD_C_b##N, FLD_O_b##N }, \
                      { 20, 20, 2, FLD_C_d##N, FLD_O_d##N }
#define BXDL(N)       { 16,  4, 0, FLD_C_b##N, FLD_O_b##N }, \
                      { 12,  4, 0, FLD_C_x##N, FLD_O_x##N }, \
                      { 20, 20, 2, FLD_C_d##N, FLD_O_d##N }
#define I(N, B, S)    {  B,  S, 1, FLD_C_i##N, FLD_O_i##N }
#define L(N, B, S)    {  B,  S, 0, FLD_C_l##N, FLD_O_l##N }

#define F0(N)                     { { } },
#define F1(N, X1)                 { { X1 } },
#define F2(N, X1, X2)             { { X1, X2 } },
#define F3(N, X1, X2, X3)         { { X1, X2, X3 } },
#define F4(N, X1, X2, X3, X4)     { { X1, X2, X3, X4 } },
#define F5(N, X1, X2, X3, X4, X5) { { X1, X2, X3, X4, X5 } },
#define F6(N, X1, X2, X3, X4, X5, X6)       { { X1, X2, X3, X4, X5, X6 } },

static const DisasFormatInfo format_info[] = {
#include "insn-format.def"
};

#undef F0
#undef F1
#undef F2
#undef F3
#undef F4
#undef F5
#undef F6
#undef R
#undef M
#undef V
#undef BD
#undef BXD
#undef BDL
#undef BXDL
#undef I
#undef L

/* Generally, we'll extract operands into this structures, operate upon
   them, and store them back.  See the "in1", "in2", "prep", "wout" sets
   of routines below for more details.  */
typedef struct {
    bool g_out, g_out2, g_in1, g_in2;
    TCGv_i64 out, out2, in1, in2;
    TCGv_i64 addr1;
} DisasOps;

/* Instructions can place constraints on their operands, raising specification
   exceptions if they are violated.  To make this easy to automate, each "in1",
   "in2", "prep", "wout" helper will have a SPEC_<name> define that equals one
   of the following, or 0.  To make this easy to document, we'll put the
   SPEC_<name> defines next to <name>.  */

#define SPEC_r1_even    1
#define SPEC_r2_even    2
#define SPEC_r3_even    4
#define SPEC_r1_f128    8
#define SPEC_r2_f128    16

/* Return values from translate_one, indicating the state of the TB.  */

/* We are not using a goto_tb (for whatever reason), but have updated
   the PC (for whatever reason), so there's no need to do it again on
   exiting the TB.  */
#define DISAS_PC_UPDATED        DISAS_TARGET_0

/* We have emitted one or more goto_tb.  No fixup required.  */
#define DISAS_GOTO_TB           DISAS_TARGET_1

/* We have updated the PC and CC values.  */
#define DISAS_PC_CC_UPDATED     DISAS_TARGET_2

/* We are exiting the TB, but have neither emitted a goto_tb, nor
   updated the PC for the next instruction to be executed.  */
#define DISAS_PC_STALE          DISAS_TARGET_3

/* We are exiting the TB to the main loop.  */
#define DISAS_PC_STALE_NOCHAIN  DISAS_TARGET_4


/* Instruction flags */
#define IF_AFP1     0x0001      /* r1 is a fp reg for HFP/FPS instructions */
#define IF_AFP2     0x0002      /* r2 is a fp reg for HFP/FPS instructions */
#define IF_AFP3     0x0004      /* r3 is a fp reg for HFP/FPS instructions */
#define IF_BFP      0x0008      /* binary floating point instruction */
#define IF_DFP      0x0010      /* decimal floating point instruction */
#define IF_PRIV     0x0020      /* privileged instruction */
#define IF_VEC      0x0040      /* vector instruction */
#define IF_IO       0x0080      /* input/output instruction */

struct DisasInsn {
    unsigned opc:16;
    unsigned flags:16;
    DisasFormat fmt:8;
    unsigned fac:8;
    unsigned spec:8;

    const char *name;

    /* Pre-process arguments before HELP_OP.  */
    void (*help_in1)(DisasContext *, DisasOps *);
    void (*help_in2)(DisasContext *, DisasOps *);
    void (*help_prep)(DisasContext *, DisasOps *);

    /*
     * Post-process output after HELP_OP.
     * Note that these are not called if HELP_OP returns DISAS_NORETURN.
     */
    void (*help_wout)(DisasContext *, DisasOps *);
    void (*help_cout)(DisasContext *, DisasOps *);

    /* Implement the operation itself.  */
    DisasJumpType (*help_op)(DisasContext *, DisasOps *);

    uint64_t data;
};

/* ====================================================================== */
/* Miscellaneous helpers, used by several operations.  */

static DisasJumpType help_goto_direct(DisasContext *s, uint64_t dest)
{
    if (dest == s->pc_tmp) {
        per_branch(s, true);
        return DISAS_NEXT;
    }
    if (use_goto_tb(s, dest)) {
        update_cc_op(s);
        per_breaking_event(s);
        tcg_gen_goto_tb(0);
        tcg_gen_movi_i64(psw_addr, dest);
        tcg_gen_exit_tb(s->base.tb, 0);
        return DISAS_GOTO_TB;
    } else {
        tcg_gen_movi_i64(psw_addr, dest);
        per_branch(s, false);
        return DISAS_PC_UPDATED;
    }
}

static DisasJumpType help_branch(DisasContext *s, DisasCompare *c,
                                 bool is_imm, int imm, TCGv_i64 cdest)
{
    DisasJumpType ret;
    uint64_t dest = s->base.pc_next + (int64_t)imm * 2;
    TCGLabel *lab;

    /* Take care of the special cases first.  */
    if (c->cond == TCG_COND_NEVER) {
        ret = DISAS_NEXT;
        goto egress;
    }
    if (is_imm) {
        if (dest == s->pc_tmp) {
            /* Branch to next.  */
            per_branch(s, true);
            ret = DISAS_NEXT;
            goto egress;
        }
        if (c->cond == TCG_COND_ALWAYS) {
            ret = help_goto_direct(s, dest);
            goto egress;
        }
    } else {
        if (!cdest) {
            /* E.g. bcr %r0 -> no branch.  */
            ret = DISAS_NEXT;
            goto egress;
        }
        if (c->cond == TCG_COND_ALWAYS) {
            tcg_gen_mov_i64(psw_addr, cdest);
            per_branch(s, false);
            ret = DISAS_PC_UPDATED;
            goto egress;
        }
    }

    if (use_goto_tb(s, s->pc_tmp)) {
        if (is_imm && use_goto_tb(s, dest)) {
            /* Both exits can use goto_tb.  */
            update_cc_op(s);

            lab = gen_new_label();
            if (c->is_64) {
                tcg_gen_brcond_i64(c->cond, c->u.s64.a, c->u.s64.b, lab);
            } else {
                tcg_gen_brcond_i32(c->cond, c->u.s32.a, c->u.s32.b, lab);
            }

            /* Branch not taken.  */
            tcg_gen_goto_tb(0);
            tcg_gen_movi_i64(psw_addr, s->pc_tmp);
            tcg_gen_exit_tb(s->base.tb, 0);

            /* Branch taken.  */
            gen_set_label(lab);
            per_breaking_event(s);
            tcg_gen_goto_tb(1);
            tcg_gen_movi_i64(psw_addr, dest);
            tcg_gen_exit_tb(s->base.tb, 1);

            ret = DISAS_GOTO_TB;
        } else {
            /* Fallthru can use goto_tb, but taken branch cannot.  */
            /* Store taken branch destination before the brcond.  This
               avoids having to allocate a new local temp to hold it.
               We'll overwrite this in the not taken case anyway.  */
            if (!is_imm) {
                tcg_gen_mov_i64(psw_addr, cdest);
            }

            lab = gen_new_label();
            if (c->is_64) {
                tcg_gen_brcond_i64(c->cond, c->u.s64.a, c->u.s64.b, lab);
            } else {
                tcg_gen_brcond_i32(c->cond, c->u.s32.a, c->u.s32.b, lab);
            }

            /* Branch not taken.  */
            update_cc_op(s);
            tcg_gen_goto_tb(0);
            tcg_gen_movi_i64(psw_addr, s->pc_tmp);
            tcg_gen_exit_tb(s->base.tb, 0);

            gen_set_label(lab);
            if (is_imm) {
                tcg_gen_movi_i64(psw_addr, dest);
            }
            per_breaking_event(s);
            ret = DISAS_PC_UPDATED;
        }
    } else {
        /* Fallthru cannot use goto_tb.  This by itself is vanishingly rare.
           Most commonly we're single-stepping or some other condition that
           disables all use of goto_tb.  Just update the PC and exit.  */

        TCGv_i64 next = tcg_const_i64(s->pc_tmp);
        if (is_imm) {
            cdest = tcg_const_i64(dest);
        }

        if (c->is_64) {
            tcg_gen_movcond_i64(c->cond, psw_addr, c->u.s64.a, c->u.s64.b,
                                cdest, next);
            per_branch_cond(s, c->cond, c->u.s64.a, c->u.s64.b);
        } else {
            TCGv_i32 t0 = tcg_temp_new_i32();
            TCGv_i64 t1 = tcg_temp_new_i64();
            TCGv_i64 z = tcg_const_i64(0);
            tcg_gen_setcond_i32(c->cond, t0, c->u.s32.a, c->u.s32.b);
            tcg_gen_extu_i32_i64(t1, t0);
            tcg_temp_free_i32(t0);
            tcg_gen_movcond_i64(TCG_COND_NE, psw_addr, t1, z, cdest, next);
            per_branch_cond(s, TCG_COND_NE, t1, z);
            tcg_temp_free_i64(t1);
            tcg_temp_free_i64(z);
        }

        if (is_imm) {
            tcg_temp_free_i64(cdest);
        }
        tcg_temp_free_i64(next);

        ret = DISAS_PC_UPDATED;
    }

 egress:
    free_compare(c);
    return ret;
}

/* ====================================================================== */
/* The operations.  These perform the bulk of the work for any insn,
   usually after the operands have been loaded and output initialized.  */

static DisasJumpType op_abs(DisasContext *s, DisasOps *o)
{
    tcg_gen_abs_i64(o->out, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_absf32(DisasContext *s, DisasOps *o)
{
    tcg_gen_andi_i64(o->out, o->in2, 0x7fffffffull);
    return DISAS_NEXT;
}

static DisasJumpType op_absf64(DisasContext *s, DisasOps *o)
{
    tcg_gen_andi_i64(o->out, o->in2, 0x7fffffffffffffffull);
    return DISAS_NEXT;
}

static DisasJumpType op_absf128(DisasContext *s, DisasOps *o)
{
    tcg_gen_andi_i64(o->out, o->in1, 0x7fffffffffffffffull);
    tcg_gen_mov_i64(o->out2, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_add(DisasContext *s, DisasOps *o)
{
    tcg_gen_add_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_addu64(DisasContext *s, DisasOps *o)
{
    tcg_gen_movi_i64(cc_src, 0);
    tcg_gen_add2_i64(o->out, cc_src, o->in1, cc_src, o->in2, cc_src);
    return DISAS_NEXT;
}

/* Compute carry into cc_src. */
static void compute_carry(DisasContext *s)
{
    switch (s->cc_op) {
    case CC_OP_ADDU:
        /* The carry value is already in cc_src (1,0). */
        break;
    case CC_OP_SUBU:
        tcg_gen_addi_i64(cc_src, cc_src, 1);
        break;
    default:
        gen_op_calc_cc(s);
        /* fall through */
    case CC_OP_STATIC:
        /* The carry flag is the msb of CC; compute into cc_src. */
        tcg_gen_extu_i32_i64(cc_src, cc_op);
        tcg_gen_shri_i64(cc_src, cc_src, 1);
        break;
    }
}

static DisasJumpType op_addc32(DisasContext *s, DisasOps *o)
{
    compute_carry(s);
    tcg_gen_add_i64(o->out, o->in1, o->in2);
    tcg_gen_add_i64(o->out, o->out, cc_src);
    return DISAS_NEXT;
}

static DisasJumpType op_addc64(DisasContext *s, DisasOps *o)
{
    compute_carry(s);

    TCGv_i64 zero = tcg_const_i64(0);
    tcg_gen_add2_i64(o->out, cc_src, o->in1, zero, cc_src, zero);
    tcg_gen_add2_i64(o->out, cc_src, o->out, cc_src, o->in2, zero);
    tcg_temp_free_i64(zero);

    return DISAS_NEXT;
}

static DisasJumpType op_asi(DisasContext *s, DisasOps *o)
{
    bool non_atomic = !s390_has_feat(S390_FEAT_STFLE_45);

    o->in1 = tcg_temp_new_i64();
    if (non_atomic) {
        tcg_gen_qemu_ld_tl(o->in1, o->addr1, get_mem_index(s), s->insn->data);
    } else {
        /* Perform the atomic addition in memory. */
        tcg_gen_atomic_fetch_add_i64(o->in1, o->addr1, o->in2, get_mem_index(s),
                                     s->insn->data);
    }

    /* Recompute also for atomic case: needed for setting CC. */
    tcg_gen_add_i64(o->out, o->in1, o->in2);

    if (non_atomic) {
        tcg_gen_qemu_st_tl(o->out, o->addr1, get_mem_index(s), s->insn->data);
    }
    return DISAS_NEXT;
}

static DisasJumpType op_asiu64(DisasContext *s, DisasOps *o)
{
    bool non_atomic = !s390_has_feat(S390_FEAT_STFLE_45);

    o->in1 = tcg_temp_new_i64();
    if (non_atomic) {
        tcg_gen_qemu_ld_tl(o->in1, o->addr1, get_mem_index(s), s->insn->data);
    } else {
        /* Perform the atomic addition in memory. */
        tcg_gen_atomic_fetch_add_i64(o->in1, o->addr1, o->in2, get_mem_index(s),
                                     s->insn->data);
    }

    /* Recompute also for atomic case: needed for setting CC. */
    tcg_gen_movi_i64(cc_src, 0);
    tcg_gen_add2_i64(o->out, cc_src, o->in1, cc_src, o->in2, cc_src);

    if (non_atomic) {
        tcg_gen_qemu_st_tl(o->out, o->addr1, get_mem_index(s), s->insn->data);
    }
    return DISAS_NEXT;
}

static DisasJumpType op_aeb(DisasContext *s, DisasOps *o)
{
    gen_helper_aeb(o->out, cpu_env, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_adb(DisasContext *s, DisasOps *o)
{
    gen_helper_adb(o->out, cpu_env, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_axb(DisasContext *s, DisasOps *o)
{
    gen_helper_axb(o->out, cpu_env, o->out, o->out2, o->in1, o->in2);
    return_low128(o->out2);
    return DISAS_NEXT;
}

static DisasJumpType op_and(DisasContext *s, DisasOps *o)
{
    tcg_gen_and_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_andi(DisasContext *s, DisasOps *o)
{
    int shift = s->insn->data & 0xff;
    int size = s->insn->data >> 8;
    uint64_t mask = ((1ull << size) - 1) << shift;

    assert(!o->g_in2);
    tcg_gen_shli_i64(o->in2, o->in2, shift);
    tcg_gen_ori_i64(o->in2, o->in2, ~mask);
    tcg_gen_and_i64(o->out, o->in1, o->in2);

    /* Produce the CC from only the bits manipulated.  */
    tcg_gen_andi_i64(cc_dst, o->out, mask);
    set_cc_nz_u64(s, cc_dst);
    return DISAS_NEXT;
}

static DisasJumpType op_andc(DisasContext *s, DisasOps *o)
{
    tcg_gen_andc_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_orc(DisasContext *s, DisasOps *o)
{
    tcg_gen_orc_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_nand(DisasContext *s, DisasOps *o)
{
    tcg_gen_nand_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_nor(DisasContext *s, DisasOps *o)
{
    tcg_gen_nor_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_nxor(DisasContext *s, DisasOps *o)
{
    tcg_gen_eqv_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_ni(DisasContext *s, DisasOps *o)
{
    o->in1 = tcg_temp_new_i64();

    if (!s390_has_feat(S390_FEAT_INTERLOCKED_ACCESS_2)) {
        tcg_gen_qemu_ld_tl(o->in1, o->addr1, get_mem_index(s), s->insn->data);
    } else {
        /* Perform the atomic operation in memory. */
        tcg_gen_atomic_fetch_and_i64(o->in1, o->addr1, o->in2, get_mem_index(s),
                                     s->insn->data);
    }

    /* Recompute also for atomic case: needed for setting CC. */
    tcg_gen_and_i64(o->out, o->in1, o->in2);

    if (!s390_has_feat(S390_FEAT_INTERLOCKED_ACCESS_2)) {
        tcg_gen_qemu_st_tl(o->out, o->addr1, get_mem_index(s), s->insn->data);
    }
    return DISAS_NEXT;
}

static DisasJumpType op_bas(DisasContext *s, DisasOps *o)
{
    pc_to_link_info(o->out, s, s->pc_tmp);
    if (o->in2) {
        tcg_gen_mov_i64(psw_addr, o->in2);
        per_branch(s, false);
        return DISAS_PC_UPDATED;
    } else {
        return DISAS_NEXT;
    }
}

static void save_link_info(DisasContext *s, DisasOps *o)
{
    TCGv_i64 t;

    if (s->base.tb->flags & (FLAG_MASK_32 | FLAG_MASK_64)) {
        pc_to_link_info(o->out, s, s->pc_tmp);
        return;
    }
    gen_op_calc_cc(s);
    tcg_gen_andi_i64(o->out, o->out, 0xffffffff00000000ull);
    tcg_gen_ori_i64(o->out, o->out, ((s->ilen / 2) << 30) | s->pc_tmp);
    t = tcg_temp_new_i64();
    tcg_gen_shri_i64(t, psw_mask, 16);
    tcg_gen_andi_i64(t, t, 0x0f000000);
    tcg_gen_or_i64(o->out, o->out, t);
    tcg_gen_extu_i32_i64(t, cc_op);
    tcg_gen_shli_i64(t, t, 28);
    tcg_gen_or_i64(o->out, o->out, t);
    tcg_temp_free_i64(t);
}

static DisasJumpType op_bal(DisasContext *s, DisasOps *o)
{
    save_link_info(s, o);
    if (o->in2) {
        tcg_gen_mov_i64(psw_addr, o->in2);
        per_branch(s, false);
        return DISAS_PC_UPDATED;
    } else {
        return DISAS_NEXT;
    }
}

static DisasJumpType op_basi(DisasContext *s, DisasOps *o)
{
    pc_to_link_info(o->out, s, s->pc_tmp);
    return help_goto_direct(s, s->base.pc_next + (int64_t)get_field(s, i2) * 2);
}

static DisasJumpType op_bc(DisasContext *s, DisasOps *o)
{
    int m1 = get_field(s, m1);
    bool is_imm = have_field(s, i2);
    int imm = is_imm ? get_field(s, i2) : 0;
    DisasCompare c;

    /* BCR with R2 = 0 causes no branching */
    if (have_field(s, r2) && get_field(s, r2) == 0) {
        if (m1 == 14) {
            /* Perform serialization */
            /* FIXME: check for fast-BCR-serialization facility */
            tcg_gen_mb(TCG_MO_ALL | TCG_BAR_SC);
        }
        if (m1 == 15) {
            /* Perform serialization */
            /* FIXME: perform checkpoint-synchronisation */
            tcg_gen_mb(TCG_MO_ALL | TCG_BAR_SC);
        }
        return DISAS_NEXT;
    }

    disas_jcc(s, &c, m1);
    return help_branch(s, &c, is_imm, imm, o->in2);
}

static DisasJumpType op_bct32(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    bool is_imm = have_field(s, i2);
    int imm = is_imm ? get_field(s, i2) : 0;
    DisasCompare c;
    TCGv_i64 t;

    c.cond = TCG_COND_NE;
    c.is_64 = false;
    c.g1 = false;
    c.g2 = false;

    t = tcg_temp_new_i64();
    tcg_gen_subi_i64(t, regs[r1], 1);
    store_reg32_i64(r1, t);
    c.u.s32.a = tcg_temp_new_i32();
    c.u.s32.b = tcg_const_i32(0);
    tcg_gen_extrl_i64_i32(c.u.s32.a, t);
    tcg_temp_free_i64(t);

    return help_branch(s, &c, is_imm, imm, o->in2);
}

static DisasJumpType op_bcth(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    int imm = get_field(s, i2);
    DisasCompare c;
    TCGv_i64 t;

    c.cond = TCG_COND_NE;
    c.is_64 = false;
    c.g1 = false;
    c.g2 = false;

    t = tcg_temp_new_i64();
    tcg_gen_shri_i64(t, regs[r1], 32);
    tcg_gen_subi_i64(t, t, 1);
    store_reg32h_i64(r1, t);
    c.u.s32.a = tcg_temp_new_i32();
    c.u.s32.b = tcg_const_i32(0);
    tcg_gen_extrl_i64_i32(c.u.s32.a, t);
    tcg_temp_free_i64(t);

    return help_branch(s, &c, 1, imm, o->in2);
}

static DisasJumpType op_bct64(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    bool is_imm = have_field(s, i2);
    int imm = is_imm ? get_field(s, i2) : 0;
    DisasCompare c;

    c.cond = TCG_COND_NE;
    c.is_64 = true;
    c.g1 = true;
    c.g2 = false;

    tcg_gen_subi_i64(regs[r1], regs[r1], 1);
    c.u.s64.a = regs[r1];
    c.u.s64.b = tcg_const_i64(0);

    return help_branch(s, &c, is_imm, imm, o->in2);
}

static DisasJumpType op_bx32(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    int r3 = get_field(s, r3);
    bool is_imm = have_field(s, i2);
    int imm = is_imm ? get_field(s, i2) : 0;
    DisasCompare c;
    TCGv_i64 t;

    c.cond = (s->insn->data ? TCG_COND_LE : TCG_COND_GT);
    c.is_64 = false;
    c.g1 = false;
    c.g2 = false;

    t = tcg_temp_new_i64();
    tcg_gen_add_i64(t, regs[r1], regs[r3]);
    c.u.s32.a = tcg_temp_new_i32();
    c.u.s32.b = tcg_temp_new_i32();
    tcg_gen_extrl_i64_i32(c.u.s32.a, t);
    tcg_gen_extrl_i64_i32(c.u.s32.b, regs[r3 | 1]);
    store_reg32_i64(r1, t);
    tcg_temp_free_i64(t);

    return help_branch(s, &c, is_imm, imm, o->in2);
}

static DisasJumpType op_bx64(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    int r3 = get_field(s, r3);
    bool is_imm = have_field(s, i2);
    int imm = is_imm ? get_field(s, i2) : 0;
    DisasCompare c;

    c.cond = (s->insn->data ? TCG_COND_LE : TCG_COND_GT);
    c.is_64 = true;

    if (r1 == (r3 | 1)) {
        c.u.s64.b = load_reg(r3 | 1);
        c.g2 = false;
    } else {
        c.u.s64.b = regs[r3 | 1];
        c.g2 = true;
    }

    tcg_gen_add_i64(regs[r1], regs[r1], regs[r3]);
    c.u.s64.a = regs[r1];
    c.g1 = true;

    return help_branch(s, &c, is_imm, imm, o->in2);
}

static DisasJumpType op_cj(DisasContext *s, DisasOps *o)
{
    int imm, m3 = get_field(s, m3);
    bool is_imm;
    DisasCompare c;

    c.cond = ltgt_cond[m3];
    if (s->insn->data) {
        c.cond = tcg_unsigned_cond(c.cond);
    }
    c.is_64 = c.g1 = c.g2 = true;
    c.u.s64.a = o->in1;
    c.u.s64.b = o->in2;

    is_imm = have_field(s, i4);
    if (is_imm) {
        imm = get_field(s, i4);
    } else {
        imm = 0;
        o->out = get_address(s, 0, get_field(s, b4),
                             get_field(s, d4));
    }

    return help_branch(s, &c, is_imm, imm, o->out);
}

static DisasJumpType op_ceb(DisasContext *s, DisasOps *o)
{
    gen_helper_ceb(cc_op, cpu_env, o->in1, o->in2);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_cdb(DisasContext *s, DisasOps *o)
{
    gen_helper_cdb(cc_op, cpu_env, o->in1, o->in2);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_cxb(DisasContext *s, DisasOps *o)
{
    gen_helper_cxb(cc_op, cpu_env, o->out, o->out2, o->in1, o->in2);
    set_cc_static(s);
    return DISAS_NEXT;
}

static TCGv_i32 fpinst_extract_m34(DisasContext *s, bool m3_with_fpe,
                                   bool m4_with_fpe)
{
    const bool fpe = s390_has_feat(S390_FEAT_FLOATING_POINT_EXT);
    uint8_t m3 = get_field(s, m3);
    uint8_t m4 = get_field(s, m4);

    /* m3 field was introduced with FPE */
    if (!fpe && m3_with_fpe) {
        m3 = 0;
    }
    /* m4 field was introduced with FPE */
    if (!fpe && m4_with_fpe) {
        m4 = 0;
    }

    /* Check for valid rounding modes. Mode 3 was introduced later. */
    if (m3 == 2 || m3 > 7 || (!fpe && m3 == 3)) {
        gen_program_exception(s, PGM_SPECIFICATION);
        return NULL;
    }

    return tcg_const_i32(deposit32(m3, 4, 4, m4));
}

static DisasJumpType op_cfeb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cfeb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_cfdb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cfdb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_cfxb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cfxb(o->out, cpu_env, o->in1, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_cgeb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cgeb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_cgdb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cgdb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_cgxb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cgxb(o->out, cpu_env, o->in1, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_clfeb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, false);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_clfeb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_clfdb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, false);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_clfdb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_clfxb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, false);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_clfxb(o->out, cpu_env, o->in1, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_clgeb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, false);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_clgeb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_clgdb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, false);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_clgdb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_clgxb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, false);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_clgxb(o->out, cpu_env, o->in1, o->in2, m34);
    tcg_temp_free_i32(m34);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_cegb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, true, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cegb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    return DISAS_NEXT;
}

static DisasJumpType op_cdgb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, true, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cdgb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    return DISAS_NEXT;
}

static DisasJumpType op_cxgb(DisasContext *s, DisasOps *o)

{
    TCGv_i32 m34 = fpinst_extract_m34(s, true, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cxgb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    return_low128(o->out2);
    return DISAS_NEXT;
}

static DisasJumpType op_celgb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, false);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_celgb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    return DISAS_NEXT;
}

static DisasJumpType op_cdlgb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, false);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cdlgb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    return DISAS_NEXT;
}

static DisasJumpType op_cxlgb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, false);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_cxlgb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    return_low128(o->out2);
    return DISAS_NEXT;
}

static DisasJumpType op_cksm(DisasContext *s, DisasOps *o)
{
    int r2 = get_field(s, r2);
    TCGv_i64 len = tcg_temp_new_i64();

    gen_helper_cksm(len, cpu_env, o->in1, o->in2, regs[r2 + 1]);
    set_cc_static(s);
    return_low128(o->out);

    tcg_gen_add_i64(regs[r2], regs[r2], len);
    tcg_gen_sub_i64(regs[r2 + 1], regs[r2 + 1], len);
    tcg_temp_free_i64(len);

    return DISAS_NEXT;
}

static DisasJumpType op_clc(DisasContext *s, DisasOps *o)
{
    int l = get_field(s, l1);
    TCGv_i32 vl;

    switch (l + 1) {
    case 1:
        tcg_gen_qemu_ld8u(cc_src, o->addr1, get_mem_index(s));
        tcg_gen_qemu_ld8u(cc_dst, o->in2, get_mem_index(s));
        break;
    case 2:
        tcg_gen_qemu_ld16u(cc_src, o->addr1, get_mem_index(s));
        tcg_gen_qemu_ld16u(cc_dst, o->in2, get_mem_index(s));
        break;
    case 4:
        tcg_gen_qemu_ld32u(cc_src, o->addr1, get_mem_index(s));
        tcg_gen_qemu_ld32u(cc_dst, o->in2, get_mem_index(s));
        break;
    case 8:
        tcg_gen_qemu_ld64(cc_src, o->addr1, get_mem_index(s));
        tcg_gen_qemu_ld64(cc_dst, o->in2, get_mem_index(s));
        break;
    default:
        vl = tcg_const_i32(l);
        gen_helper_clc(cc_op, cpu_env, vl, o->addr1, o->in2);
        tcg_temp_free_i32(vl);
        set_cc_static(s);
        return DISAS_NEXT;
    }
    gen_op_update2_cc_i64(s, CC_OP_LTUGTU_64, cc_src, cc_dst);
    return DISAS_NEXT;
}

static DisasJumpType op_clcl(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    int r2 = get_field(s, r2);
    TCGv_i32 t1, t2;

    /* r1 and r2 must be even.  */
    if (r1 & 1 || r2 & 1) {
        gen_program_exception(s, PGM_SPECIFICATION);
        return DISAS_NORETURN;
    }

    t1 = tcg_const_i32(r1);
    t2 = tcg_const_i32(r2);
    gen_helper_clcl(cc_op, cpu_env, t1, t2);
    tcg_temp_free_i32(t1);
    tcg_temp_free_i32(t2);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_clcle(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    int r3 = get_field(s, r3);
    TCGv_i32 t1, t3;

    /* r1 and r3 must be even.  */
    if (r1 & 1 || r3 & 1) {
        gen_program_exception(s, PGM_SPECIFICATION);
        return DISAS_NORETURN;
    }

    t1 = tcg_const_i32(r1);
    t3 = tcg_const_i32(r3);
    gen_helper_clcle(cc_op, cpu_env, t1, o->in2, t3);
    tcg_temp_free_i32(t1);
    tcg_temp_free_i32(t3);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_clclu(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    int r3 = get_field(s, r3);
    TCGv_i32 t1, t3;

    /* r1 and r3 must be even.  */
    if (r1 & 1 || r3 & 1) {
        gen_program_exception(s, PGM_SPECIFICATION);
        return DISAS_NORETURN;
    }

    t1 = tcg_const_i32(r1);
    t3 = tcg_const_i32(r3);
    gen_helper_clclu(cc_op, cpu_env, t1, o->in2, t3);
    tcg_temp_free_i32(t1);
    tcg_temp_free_i32(t3);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_clm(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m3 = tcg_const_i32(get_field(s, m3));
    TCGv_i32 t1 = tcg_temp_new_i32();
    tcg_gen_extrl_i64_i32(t1, o->in1);
    gen_helper_clm(cc_op, cpu_env, t1, m3, o->in2);
    set_cc_static(s);
    tcg_temp_free_i32(t1);
    tcg_temp_free_i32(m3);
    return DISAS_NEXT;
}

static DisasJumpType op_clst(DisasContext *s, DisasOps *o)
{
    gen_helper_clst(o->in1, cpu_env, regs[0], o->in1, o->in2);
    set_cc_static(s);
    return_low128(o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_cps(DisasContext *s, DisasOps *o)
{
    TCGv_i64 t = tcg_temp_new_i64();
    tcg_gen_andi_i64(t, o->in1, 0x8000000000000000ull);
    tcg_gen_andi_i64(o->out, o->in2, 0x7fffffffffffffffull);
    tcg_gen_or_i64(o->out, o->out, t);
    tcg_temp_free_i64(t);
    return DISAS_NEXT;
}

static DisasJumpType op_cs(DisasContext *s, DisasOps *o)
{
    int d2 = get_field(s, d2);
    int b2 = get_field(s, b2);
    TCGv_i64 addr, cc;

    /* Note that in1 = R3 (new value) and
       in2 = (zero-extended) R1 (expected value).  */

    addr = get_address(s, 0, b2, d2);
    tcg_gen_atomic_cmpxchg_i64(o->out, addr, o->in2, o->in1,
                               get_mem_index(s), s->insn->data | MO_ALIGN);
    tcg_temp_free_i64(addr);

    /* Are the memory and expected values (un)equal?  Note that this setcond
       produces the output CC value, thus the NE sense of the test.  */
    cc = tcg_temp_new_i64();
    tcg_gen_setcond_i64(TCG_COND_NE, cc, o->in2, o->out);
    tcg_gen_extrl_i64_i32(cc_op, cc);
    tcg_temp_free_i64(cc);
    set_cc_static(s);

    return DISAS_NEXT;
}

static DisasJumpType op_cdsg(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    int r3 = get_field(s, r3);
    int d2 = get_field(s, d2);
    int b2 = get_field(s, b2);
    DisasJumpType ret = DISAS_NEXT;
    TCGv_i64 addr;
    TCGv_i32 t_r1, t_r3;

    /* Note that R1:R1+1 = expected value and R3:R3+1 = new value.  */
    addr = get_address(s, 0, b2, d2);
    t_r1 = tcg_const_i32(r1);
    t_r3 = tcg_const_i32(r3);
    if (!(tb_cflags(s->base.tb) & CF_PARALLEL)) {
        gen_helper_cdsg(cpu_env, addr, t_r1, t_r3);
    } else if (HAVE_CMPXCHG128) {
        gen_helper_cdsg_parallel(cpu_env, addr, t_r1, t_r3);
    } else {
        gen_helper_exit_atomic(cpu_env);
        ret = DISAS_NORETURN;
    }
    tcg_temp_free_i64(addr);
    tcg_temp_free_i32(t_r1);
    tcg_temp_free_i32(t_r3);

    set_cc_static(s);
    return ret;
}

static DisasJumpType op_csst(DisasContext *s, DisasOps *o)
{
    int r3 = get_field(s, r3);
    TCGv_i32 t_r3 = tcg_const_i32(r3);

    if (tb_cflags(s->base.tb) & CF_PARALLEL) {
        gen_helper_csst_parallel(cc_op, cpu_env, t_r3, o->addr1, o->in2);
    } else {
        gen_helper_csst(cc_op, cpu_env, t_r3, o->addr1, o->in2);
    }
    tcg_temp_free_i32(t_r3);

    set_cc_static(s);
    return DISAS_NEXT;
}

#ifndef CONFIG_USER_ONLY
static DisasJumpType op_csp(DisasContext *s, DisasOps *o)
{
    MemOp mop = s->insn->data;
    TCGv_i64 addr, old, cc;
    TCGLabel *lab = gen_new_label();

    /* Note that in1 = R1 (zero-extended expected value),
       out = R1 (original reg), out2 = R1+1 (new value).  */

    addr = tcg_temp_new_i64();
    old = tcg_temp_new_i64();
    tcg_gen_andi_i64(addr, o->in2, -1ULL << (mop & MO_SIZE));
    tcg_gen_atomic_cmpxchg_i64(old, addr, o->in1, o->out2,
                               get_mem_index(s), mop | MO_ALIGN);
    tcg_temp_free_i64(addr);

    /* Are the memory and expected values (un)equal?  */
    cc = tcg_temp_new_i64();
    tcg_gen_setcond_i64(TCG_COND_NE, cc, o->in1, old);
    tcg_gen_extrl_i64_i32(cc_op, cc);

    /* Write back the output now, so that it happens before the
       following branch, so that we don't need local temps.  */
    if ((mop & MO_SIZE) == MO_32) {
        tcg_gen_deposit_i64(o->out, o->out, old, 0, 32);
    } else {
        tcg_gen_mov_i64(o->out, old);
    }
    tcg_temp_free_i64(old);

    /* If the comparison was equal, and the LSB of R2 was set,
       then we need to flush the TLB (for all cpus).  */
    tcg_gen_xori_i64(cc, cc, 1);
    tcg_gen_and_i64(cc, cc, o->in2);
    tcg_gen_brcondi_i64(TCG_COND_EQ, cc, 0, lab);
    tcg_temp_free_i64(cc);

    gen_helper_purge(cpu_env);
    gen_set_label(lab);

    return DISAS_NEXT;
}
#endif

static DisasJumpType op_cvd(DisasContext *s, DisasOps *o)
{
    TCGv_i64 t1 = tcg_temp_new_i64();
    TCGv_i32 t2 = tcg_temp_new_i32();
    tcg_gen_extrl_i64_i32(t2, o->in1);
    gen_helper_cvd(t1, t2);
    tcg_temp_free_i32(t2);
    tcg_gen_qemu_st64(t1, o->in2, get_mem_index(s));
    tcg_temp_free_i64(t1);
    return DISAS_NEXT;
}

static DisasJumpType op_ct(DisasContext *s, DisasOps *o)
{
    int m3 = get_field(s, m3);
    TCGLabel *lab = gen_new_label();
    TCGCond c;

    c = tcg_invert_cond(ltgt_cond[m3]);
    if (s->insn->data) {
        c = tcg_unsigned_cond(c);
    }
    tcg_gen_brcond_i64(c, o->in1, o->in2, lab);

    /* Trap.  */
    gen_trap(s);

    gen_set_label(lab);
    return DISAS_NEXT;
}

static DisasJumpType op_cuXX(DisasContext *s, DisasOps *o)
{
    int m3 = get_field(s, m3);
    int r1 = get_field(s, r1);
    int r2 = get_field(s, r2);
    TCGv_i32 tr1, tr2, chk;

    /* R1 and R2 must both be even.  */
    if ((r1 | r2) & 1) {
        gen_program_exception(s, PGM_SPECIFICATION);
        return DISAS_NORETURN;
    }
    if (!s390_has_feat(S390_FEAT_ETF3_ENH)) {
        m3 = 0;
    }

    tr1 = tcg_const_i32(r1);
    tr2 = tcg_const_i32(r2);
    chk = tcg_const_i32(m3);

    switch (s->insn->data) {
    case 12:
        gen_helper_cu12(cc_op, cpu_env, tr1, tr2, chk);
        break;
    case 14:
        gen_helper_cu14(cc_op, cpu_env, tr1, tr2, chk);
        break;
    case 21:
        gen_helper_cu21(cc_op, cpu_env, tr1, tr2, chk);
        break;
    case 24:
        gen_helper_cu24(cc_op, cpu_env, tr1, tr2, chk);
        break;
    case 41:
        gen_helper_cu41(cc_op, cpu_env, tr1, tr2, chk);
        break;
    case 42:
        gen_helper_cu42(cc_op, cpu_env, tr1, tr2, chk);
        break;
    default:
        g_assert_not_reached();
    }

    tcg_temp_free_i32(tr1);
    tcg_temp_free_i32(tr2);
    tcg_temp_free_i32(chk);
    set_cc_static(s);
    return DISAS_NEXT;
}

#ifndef CONFIG_USER_ONLY
static DisasJumpType op_diag(DisasContext *s, DisasOps *o)
{
    TCGv_i32 r1 = tcg_const_i32(get_field(s, r1));
    TCGv_i32 r3 = tcg_const_i32(get_field(s, r3));
    TCGv_i32 func_code = tcg_const_i32(get_field(s, i2));

    gen_helper_diag(cpu_env, r1, r3, func_code);

    tcg_temp_free_i32(func_code);
    tcg_temp_free_i32(r3);
    tcg_temp_free_i32(r1);
    return DISAS_NEXT;
}
#endif

static DisasJumpType op_divs32(DisasContext *s, DisasOps *o)
{
    gen_helper_divs32(o->out2, cpu_env, o->in1, o->in2);
    return_low128(o->out);
    return DISAS_NEXT;
}

static DisasJumpType op_divu32(DisasContext *s, DisasOps *o)
{
    gen_helper_divu32(o->out2, cpu_env, o->in1, o->in2);
    return_low128(o->out);
    return DISAS_NEXT;
}

static DisasJumpType op_divs64(DisasContext *s, DisasOps *o)
{
    gen_helper_divs64(o->out2, cpu_env, o->in1, o->in2);
    return_low128(o->out);
    return DISAS_NEXT;
}

static DisasJumpType op_divu64(DisasContext *s, DisasOps *o)
{
    gen_helper_divu64(o->out2, cpu_env, o->out, o->out2, o->in2);
    return_low128(o->out);
    return DISAS_NEXT;
}

static DisasJumpType op_deb(DisasContext *s, DisasOps *o)
{
    gen_helper_deb(o->out, cpu_env, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_ddb(DisasContext *s, DisasOps *o)
{
    gen_helper_ddb(o->out, cpu_env, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_dxb(DisasContext *s, DisasOps *o)
{
    gen_helper_dxb(o->out, cpu_env, o->out, o->out2, o->in1, o->in2);
    return_low128(o->out2);
    return DISAS_NEXT;
}

static DisasJumpType op_ear(DisasContext *s, DisasOps *o)
{
    int r2 = get_field(s, r2);
    tcg_gen_ld32u_i64(o->out, cpu_env, offsetof(CPUS390XState, aregs[r2]));
    return DISAS_NEXT;
}

static DisasJumpType op_ecag(DisasContext *s, DisasOps *o)
{
    /* No cache information provided.  */
    tcg_gen_movi_i64(o->out, -1);
    return DISAS_NEXT;
}

static DisasJumpType op_efpc(DisasContext *s, DisasOps *o)
{
    tcg_gen_ld32u_i64(o->out, cpu_env, offsetof(CPUS390XState, fpc));
    return DISAS_NEXT;
}

static DisasJumpType op_epsw(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    int r2 = get_field(s, r2);
    TCGv_i64 t = tcg_temp_new_i64();

    /* Note the "subsequently" in the PoO, which implies a defined result
       if r1 == r2.  Thus we cannot defer these writes to an output hook.  */
    tcg_gen_shri_i64(t, psw_mask, 32);
    store_reg32_i64(r1, t);
    if (r2 != 0) {
        store_reg32_i64(r2, psw_mask);
    }

    tcg_temp_free_i64(t);
    return DISAS_NEXT;
}

static DisasJumpType op_ex(DisasContext *s, DisasOps *o)
{
    int r1 = get_field(s, r1);
    TCGv_i32 ilen;
    TCGv_i64 v1;

    /* Nested EXECUTE is not allowed.  */
    if (unlikely(s->ex_value)) {
        gen_program_exception(s, PGM_EXECUTE);
        return DISAS_NORETURN;
    }

    update_psw_addr(s);
    update_cc_op(s);

    if (r1 == 0) {
        v1 = tcg_const_i64(0);
    } else {
        v1 = regs[r1];
    }

    ilen = tcg_const_i32(s->ilen);
    gen_helper_ex(cpu_env, ilen, v1, o->in2);
    tcg_temp_free_i32(ilen);

    if (r1 == 0) {
        tcg_temp_free_i64(v1);
    }

    return DISAS_PC_CC_UPDATED;
}

static DisasJumpType op_fieb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_fieb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    return DISAS_NEXT;
}

static DisasJumpType op_fidb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_fidb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    return DISAS_NEXT;
}

static DisasJumpType op_fixb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, false, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_fixb(o->out, cpu_env, o->in1, o->in2, m34);
    return_low128(o->out2);
    tcg_temp_free_i32(m34);
    return DISAS_NEXT;
}

static DisasJumpType op_flogr(DisasContext *s, DisasOps *o)
{
    /* We'll use the original input for cc computation, since we get to
       compare that against 0, which ought to be better than comparing
       the real output against 64.  It also lets cc_dst be a convenient
       temporary during our computation.  */
    gen_op_update1_cc_i64(s, CC_OP_FLOGR, o->in2);

    /* R1 = IN ? CLZ(IN) : 64.  */
    tcg_gen_clzi_i64(o->out, o->in2, 64);

    /* R1+1 = IN & ~(found bit).  Note that we may attempt to shift this
       value by 64, which is undefined.  But since the shift is 64 iff the
       input is zero, we still get the correct result after and'ing.  */
    tcg_gen_movi_i64(o->out2, 0x8000000000000000ull);
    tcg_gen_shr_i64(o->out2, o->out2, o->out);
    tcg_gen_andc_i64(o->out2, cc_dst, o->out2);
    return DISAS_NEXT;
}

static DisasJumpType op_icm(DisasContext *s, DisasOps *o)
{
    int m3 = get_field(s, m3);
    int pos, len, base = s->insn->data;
    TCGv_i64 tmp = tcg_temp_new_i64();
    uint64_t ccm;

    switch (m3) {
    case 0xf:
        /* Effectively a 32-bit load.  */
        tcg_gen_qemu_ld32u(tmp, o->in2, get_mem_index(s));
        len = 32;
        goto one_insert;

    case 0xc:
    case 0x6:
    case 0x3:
        /* Effectively a 16-bit load.  */
        tcg_gen_qemu_ld16u(tmp, o->in2, get_mem_index(s));
        len = 16;
        goto one_insert;

    case 0x8:
    case 0x4:
    case 0x2:
    case 0x1:
        /* Effectively an 8-bit load.  */
        tcg_gen_qemu_ld8u(tmp, o->in2, get_mem_index(s));
        len = 8;
        goto one_insert;

    one_insert:
        pos = base + ctz32(m3) * 8;
        tcg_gen_deposit_i64(o->out, o->out, tmp, pos, len);
        ccm = ((1ull << len) - 1) << pos;
        break;

    default:
        /* This is going to be a sequence of loads and inserts.  */
        pos = base + 32 - 8;
        ccm = 0;
        while (m3) {
            if (m3 & 0x8) {
                tcg_gen_qemu_ld8u(tmp, o->in2, get_mem_index(s));
                tcg_gen_addi_i64(o->in2, o->in2, 1);
                tcg_gen_deposit_i64(o->out, o->out, tmp, pos, 8);
                ccm |= 0xff << pos;
            }
            m3 = (m3 << 1) & 0xf;
            pos -= 8;
        }
        break;
    }

    tcg_gen_movi_i64(tmp, ccm);
    gen_op_update2_cc_i64(s, CC_OP_ICM, tmp, o->out);
    tcg_temp_free_i64(tmp);
    return DISAS_NEXT;
}

static DisasJumpType op_insi(DisasContext *s, DisasOps *o)
{
    int shift = s->insn->data & 0xff;
    int size = s->insn->data >> 8;
    tcg_gen_deposit_i64(o->out, o->in1, o->in2, shift, size);
    return DISAS_NEXT;
}

static DisasJumpType op_ipm(DisasContext *s, DisasOps *o)
{
    TCGv_i64 t1, t2;

    gen_op_calc_cc(s);
    t1 = tcg_temp_new_i64();
    tcg_gen_extract_i64(t1, psw_mask, 40, 4);
    t2 = tcg_temp_new_i64();
    tcg_gen_extu_i32_i64(t2, cc_op);
    tcg_gen_deposit_i64(t1, t1, t2, 4, 60);
    tcg_gen_deposit_i64(o->out, o->out, t1, 24, 8);
    tcg_temp_free_i64(t1);
    tcg_temp_free_i64(t2);
    return DISAS_NEXT;
}

#ifndef CONFIG_USER_ONLY
static DisasJumpType op_idte(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m4;

    if (s390_has_feat(S390_FEAT_LOCAL_TLB_CLEARING)) {
        m4 = tcg_const_i32(get_field(s, m4));
    } else {
        m4 = tcg_const_i32(0);
    }
    gen_helper_idte(cpu_env, o->in1, o->in2, m4);
    tcg_temp_free_i32(m4);
    return DISAS_NEXT;
}

static DisasJumpType op_ipte(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m4;

    if (s390_has_feat(S390_FEAT_LOCAL_TLB_CLEARING)) {
        m4 = tcg_const_i32(get_field(s, m4));
    } else {
        m4 = tcg_const_i32(0);
    }
    gen_helper_ipte(cpu_env, o->in1, o->in2, m4);
    tcg_temp_free_i32(m4);
    return DISAS_NEXT;
}

static DisasJumpType op_iske(DisasContext *s, DisasOps *o)
{
    gen_helper_iske(o->out, cpu_env, o->in2);
    return DISAS_NEXT;
}
#endif

static DisasJumpType op_msa(DisasContext *s, DisasOps *o)
{
    int r1 = have_field(s, r1) ? get_field(s, r1) : 0;
    int r2 = have_field(s, r2) ? get_field(s, r2) : 0;
    int r3 = have_field(s, r3) ? get_field(s, r3) : 0;
    TCGv_i32 t_r1, t_r2, t_r3, type;

    switch (s->insn->data) {
    case S390_FEAT_TYPE_KMA:
        if (r3 == r1 || r3 == r2) {
            gen_program_exception(s, PGM_SPECIFICATION);
            return DISAS_NORETURN;
        }
        /* FALL THROUGH */
    case S390_FEAT_TYPE_KMCTR:
        if (r3 & 1 || !r3) {
            gen_program_exception(s, PGM_SPECIFICATION);
            return DISAS_NORETURN;
        }
        /* FALL THROUGH */
    case S390_FEAT_TYPE_PPNO:
    case S390_FEAT_TYPE_KMF:
    case S390_FEAT_TYPE_KMC:
    case S390_FEAT_TYPE_KMO:
    case S390_FEAT_TYPE_KM:
        if (r1 & 1 || !r1) {
            gen_program_exception(s, PGM_SPECIFICATION);
            return DISAS_NORETURN;
        }
        /* FALL THROUGH */
    case S390_FEAT_TYPE_KMAC:
    case S390_FEAT_TYPE_KIMD:
    case S390_FEAT_TYPE_KLMD:
        if (r2 & 1 || !r2) {
            gen_program_exception(s, PGM_SPECIFICATION);
            return DISAS_NORETURN;
        }
        /* FALL THROUGH */
    case S390_FEAT_TYPE_PCKMO:
    case S390_FEAT_TYPE_PCC:
        break;
    default:
        g_assert_not_reached();
    };

    t_r1 = tcg_const_i32(r1);
    t_r2 = tcg_const_i32(r2);
    t_r3 = tcg_const_i32(r3);
    type = tcg_const_i32(s->insn->data);
    gen_helper_msa(cc_op, cpu_env, t_r1, t_r2, t_r3, type);
    set_cc_static(s);
    tcg_temp_free_i32(t_r1);
    tcg_temp_free_i32(t_r2);
    tcg_temp_free_i32(t_r3);
    tcg_temp_free_i32(type);
    return DISAS_NEXT;
}

static DisasJumpType op_keb(DisasContext *s, DisasOps *o)
{
    gen_helper_keb(cc_op, cpu_env, o->in1, o->in2);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_kdb(DisasContext *s, DisasOps *o)
{
    gen_helper_kdb(cc_op, cpu_env, o->in1, o->in2);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_kxb(DisasContext *s, DisasOps *o)
{
    gen_helper_kxb(cc_op, cpu_env, o->out, o->out2, o->in1, o->in2);
    set_cc_static(s);
    return DISAS_NEXT;
}

static DisasJumpType op_laa(DisasContext *s, DisasOps *o)
{
    /* The real output is indeed the original value in memory;
       recompute the addition for the computation of CC.  */
    tcg_gen_atomic_fetch_add_i64(o->in2, o->in2, o->in1, get_mem_index(s),
                                 s->insn->data | MO_ALIGN);
    /* However, we need to recompute the addition for setting CC.  */
    tcg_gen_add_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_lan(DisasContext *s, DisasOps *o)
{
    /* The real output is indeed the original value in memory;
       recompute the addition for the computation of CC.  */
    tcg_gen_atomic_fetch_and_i64(o->in2, o->in2, o->in1, get_mem_index(s),
                                 s->insn->data | MO_ALIGN);
    /* However, we need to recompute the operation for setting CC.  */
    tcg_gen_and_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_lao(DisasContext *s, DisasOps *o)
{
    /* The real output is indeed the original value in memory;
       recompute the addition for the computation of CC.  */
    tcg_gen_atomic_fetch_or_i64(o->in2, o->in2, o->in1, get_mem_index(s),
                                s->insn->data | MO_ALIGN);
    /* However, we need to recompute the operation for setting CC.  */
    tcg_gen_or_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_lax(DisasContext *s, DisasOps *o)
{
    /* The real output is indeed the original value in memory;
       recompute the addition for the computation of CC.  */
    tcg_gen_atomic_fetch_xor_i64(o->in2, o->in2, o->in1, get_mem_index(s),
                                 s->insn->data | MO_ALIGN);
    /* However, we need to recompute the operation for setting CC.  */
    tcg_gen_xor_i64(o->out, o->in1, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_ldeb(DisasContext *s, DisasOps *o)
{
    gen_helper_ldeb(o->out, cpu_env, o->in2);
    return DISAS_NEXT;
}

static DisasJumpType op_ledb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, true, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_ledb(o->out, cpu_env, o->in2, m34);
    tcg_temp_free_i32(m34);
    return DISAS_NEXT;
}

static DisasJumpType op_ldxb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, true, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_ldxb(o->out, cpu_env, o->in1, o->in2, m34);
    tcg_temp_free_i32(m34);
    return DISAS_NEXT;
}

static DisasJumpType op_lexb(DisasContext *s, DisasOps *o)
{
    TCGv_i32 m34 = fpinst_extract_m34(s, true, true);

    if (!m34) {
        return DISAS_NORETURN;
    }
    gen_helper_lexb(o->out, cpu_env, o->in1, o->in2, m34);
    tcg_temp_free_i32(m34);
    return DISAS_NEXT;
}

static DisasJumpType op_lxdb(DisasContext *s, DisasOps *o)
{
    gen_helper_lxdb(o->out, cpu_env, o->in2);
    return_low128(o->out2);
    return DISAS_NEXT;
}

static DisasJumpType op_lxeb(DisasContext *s, DisasOps *o)
{
    gen_helper_lxeb(o->out, cpu_env, o->in2);
    return_low128(o->out2);
    return DISAS_NEXT;
}

static DisasJumpType op_lde(DisasContext *s, DisasOps *o)
{
    tcg_gen_shli_i64(o->out, o->in2, 32);
    return DISAS_NEXT;
}

static DisasJumpType op_llgt(DisasContext *s, DisasOps *o)
{
    tcg_gen_andi_i64(o->out, o->in2, 0x7fffffff);
    return DISAS_NEXT;
}

static DisasJumpType op_ld8s(DisasContext *s, DisasOps *o)
{
    tcg_gen_qemu_ld8s(o->out, o->in2, get_mem_index(s));
    return DISAS_NEXT;
}

static DisasJumpType op_ld8u(DisasContext *s, DisasOps *o)
{
    tcg_gen_qemu_ld8u(o->out, o->in2, get_mem_index(s));
    return DISAS_NEXT;
}

static DisasJumpType op_ld16s(DisasContext *s, DisasOps *o)
{
    tcg_gen_qemu_ld16s(o->out, o->in2, get_mem_index(s));
    return DISAS_NEXT;
}

static DisasJumpType op_ld16u(DisasContext *s, DisasOps *o)
{
    tcg_gen_qemu_ld16u(o->out, o->in2, get_mem_index(s));
    return DISAS_NEXT;
}

static DisasJumpType op_ld32s(DisasContext *s, DisasOps *o)
{
    tcg_gen_qemu_ld32s(o->out, o->in2, get_mem_index(s));
    return DISAS_NEXT;
}

static DisasJumpType op_ld32u(DisasContext *s, DisasOps *o)
{
    tcg_gen_qemu_ld32u(o->out, o->in2, get_mem_index(s));
    return DISAS_NEXT;
}

static DisasJumpType op_ld64(DisasContext *s, DisasOps *o)
{
    tcg_gen_qemu_ld64(o->out, o->in2, get_mem_index(s));
    return DISAS_NEXT;
}

static DisasJumpType op_lat(DisasContext *s, DisasOps *o)
{
    TCGLabel *lab = gen_new_label();
    store_reg32_i64(get_field(s, r1), o->in2);
    /* The value is stored even in case of trap. */
    tcg_gen_brcondi_i64(TCG_COND_NE, o->in2, 0, lab);
    gen_trap(s);
    gen_set_label(lab);
    return DISAS_NEXT;
}

static DisasJumpType op_lgat(DisasContext *s, DisasOps *o)
{
    TCGLabel *lab = gen_new_label();
    tcg_gen_qemu_ld64(o->out, o->in2, get_mem_index(s));
    /* The value is stored even in case of trap. */
    tcg_gen_brcondi_i64(TCG_COND_NE, o->out, 0, lab);
    gen_trap(s);
    gen_set_label(lab);
    return DISAS_NEXT;
}

static DisasJumpType op_lfhat(DisasContext *s, DisasOps *o)
{
    TCGLabel *lab = gen_new_label();
    store_reg32h_i64(get_field(s, r1), o->in2);
    /* The value is stored even in case of trap. */
    tcg_gen_brcondi_i64(TCG_COND_NE, o->in2, 0, lab);
    gen_trap(s);
    gen_set_label(lab);
    return DISAS_NEXT;
}

static DisasJumpType op_llgfat(DisasContext *s, DisasOps *o)
{
    TCGLabel *lab = gen_new_label();
    tcg_gen_qemu_ld32u(o->out, o->in2, get_mem_index(s));
    /* The value is stored even in case of trap. */
    tcg_gen_brcondi_i64(TCG_COND_NE, o->out, 0, lab);
    gen_trap(s);
    gen_set_label(lab);
    return DISAS_NEXT;
}

static DisasJumpType op_llgtat(DisasContext *s, DisasOps *o)
{
    TCGLabel *lab = gen_new_label();
    tcg_gen_andi_i64(o->out, o->in2, 0x7fffffff);
    /* The value is stored even in case of trap. */
    tcg_gen_brcondi_i64(TCG_COND_NE, o->out, 0, lab);
    gen_trap(s);
    gen_set_label(lab);
    return DISAS_NEXT;
}

static DisasJumpType op_loc(DisasContext *s, DisasOps *o)
{
    DisasCompare c;

    if (have_field(s, m3)) {
        /* LOAD * ON CONDITION */
        disas_jcc(s, &c, get_field(s, m3));
    } else {
        /* SELECT */
        disas_jcc(s, &c, get_field(s, m4));
    }

    if (c.is_64) {
        tcg_gen_movcond_i64(c.cond, o->out, c.u.s64.a, c.u.s64.b,