target-tricore: Add instructions of RR opcode format, that have 0x4b as the first opcode

Add instructions of RR opcode format, that have 0x4b as the first opcode.
Add helper functions:
    * parity: Calculates the parity bits for every byte of a 32 int.
    * bmerge/bsplit: Merges two regs into one bitwise/Splits one reg into two bitwise.
    * unpack: unpack a IEEE 754 single precision floating point number as exponent and mantissa.
    * dvinit_b_13/131: (ISA v1.3/v1.31)Prepare operands for a divide operation,
                       where the quotient result is guaranteed to fit into 8 bit.
    * dvinit_h_13/131: (ISA v1.3/v1.31)Prepare operands for a divide operation,
                       where the quotient result is guaranteed to fit into 16 bit.
OPCM_32_RR_FLOAT -> OPCM_32_RR_DIVIDE.

Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Reviewed-by: Richard Henderson <rth@twiddle.net>

4 files changed, 390 insertions, 1 deletions

diff --git a/target-tricore/helper.h b/target-tricore/helper.h
index 3b85c1a5ec..0294512651 100644
--- a/target-tricore/helper.h
+++ b/target-tricore/helper.h
@@ -76,6 +76,17 @@ DEF_HELPER_FLAGS_2(sh, TCG_CALL_NO_RWG_SE, i32, i32, i32)
 DEF_HELPER_FLAGS_2(sh_h, TCG_CALL_NO_RWG_SE, i32, i32, i32)
 DEF_HELPER_3(sha, i32, env, i32, i32)
 DEF_HELPER_2(sha_h, i32, i32, i32)
+/* merge/split/parity */
+DEF_HELPER_FLAGS_2(bmerge, TCG_CALL_NO_RWG_SE, i32, i32, i32)
+DEF_HELPER_FLAGS_1(bsplit, TCG_CALL_NO_RWG_SE, i64, i32)
+DEF_HELPER_FLAGS_1(parity, TCG_CALL_NO_RWG_SE, i32, i32)
+/* float */
+DEF_HELPER_1(unpack, i64, i32)
+/* dvinit */
+DEF_HELPER_3(dvinit_b_13, i64, env, i32, i32)
+DEF_HELPER_3(dvinit_b_131, i64, env, i32, i32)
+DEF_HELPER_3(dvinit_h_13, i64, env, i32, i32)
+DEF_HELPER_3(dvinit_h_131, i64, env, i32, i32)
 /* CSA */
 DEF_HELPER_2(call, void, env, i32)
 DEF_HELPER_1(ret, void, env)
diff --git a/target-tricore/op_helper.c b/target-tricore/op_helper.c
index 98f28d5623..63d2d56a9d 100644
--- a/target-tricore/op_helper.c
+++ b/target-tricore/op_helper.c
@@ -1033,6 +1033,201 @@ uint32_t helper_sha_h(target_ulong r1, target_ulong r2)
     }
 }
 
+uint32_t helper_bmerge(target_ulong r1, target_ulong r2)
+{
+    uint32_t i, ret;
+
+    ret = 0;
+    for (i = 0; i < 16; i++) {
+        ret |= (r1 & 1) << (2 * i + 1);
+        ret |= (r2 & 1) << (2 * i);
+        r1 = r1 >> 1;
+        r2 = r2 >> 1;
+    }
+    return ret;
+}
+
+uint64_t helper_bsplit(uint32_t r1)
+{
+    int32_t i;
+    uint64_t ret;
+
+    ret = 0;
+    for (i = 0; i < 32; i = i + 2) {
+        /* even */
+        ret |= (r1 & 1) << (i/2);
+        r1 = r1 >> 1;
+        /* odd */
+        ret |= (uint64_t)(r1 & 1) << (i/2 + 32);
+        r1 = r1 >> 1;
+    }
+    return ret;
+}
+
+uint32_t helper_parity(target_ulong r1)
+{
+    uint32_t ret;
+    uint32_t nOnes, i;
+
+    ret = 0;
+    nOnes = 0;
+    for (i = 0; i < 8; i++) {
+        ret ^= (r1 & 1);
+        r1 = r1 >> 1;
+    }
+    /* second byte */
+    nOnes = 0;
+    for (i = 0; i < 8; i++) {
+        nOnes ^= (r1 & 1);
+        r1 = r1 >> 1;
+    }
+    ret |= nOnes << 8;
+    /* third byte */
+    nOnes = 0;
+    for (i = 0; i < 8; i++) {
+        nOnes ^= (r1 & 1);
+        r1 = r1 >> 1;
+    }
+    ret |= nOnes << 16;
+    /* fourth byte */
+    nOnes = 0;
+    for (i = 0; i < 8; i++) {
+        nOnes ^= (r1 & 1);
+        r1 = r1 >> 1;
+    }
+    ret |= nOnes << 24;
+
+    return ret;
+}
+
+uint64_t helper_unpack(target_ulong arg1)
+{
+    int32_t fp_exp  = extract32(arg1, 23, 8);
+    int32_t fp_frac = extract32(arg1, 0, 23);
+    uint64_t ret;
+    int32_t int_exp, int_mant;
+
+    if (fp_exp == 255) {
+        int_exp = 255;
+        int_mant = (fp_frac << 7);
+    } else if ((fp_exp == 0) && (fp_frac == 0)) {
+        int_exp  = -127;
+        int_mant = 0;
+    } else if ((fp_exp == 0) && (fp_frac != 0)) {
+        int_exp  = -126;
+        int_mant = (fp_frac << 7);
+    } else {
+        int_exp  = fp_exp - 127;
+        int_mant = (fp_frac << 7);
+        int_mant |= (1 << 30);
+    }
+    ret = int_exp;
+    ret = ret << 32;
+    ret |= int_mant;
+
+    return ret;
+}
+
+uint64_t helper_dvinit_b_13(CPUTriCoreState *env, uint32_t r1, uint32_t r2)
+{
+    uint64_t ret;
+    int32_t abs_sig_dividend, abs_base_dividend, abs_divisor;
+    int32_t quotient_sign;
+
+    ret = sextract32(r1, 0, 32);
+    ret = ret << 24;
+    quotient_sign = 0;
+    if (!((r1 & 0x80000000) == (r2 & 0x80000000))) {
+        ret |= 0xffffff;
+        quotient_sign = 1;
+    }
+
+    abs_sig_dividend = abs(r1) >> 7;
+    abs_base_dividend = abs(r1) & 0x7f;
+    abs_divisor = abs(r1);
+    /* calc overflow */
+    env->PSW_USB_V = 0;
+    if ((quotient_sign) && (abs_divisor)) {
+        env->PSW_USB_V = (((abs_sig_dividend == abs_divisor) &&
+                         (abs_base_dividend >= abs_divisor)) ||
+                         (abs_sig_dividend > abs_divisor));
+    } else {
+        env->PSW_USB_V = (abs_sig_dividend >= abs_divisor);
+    }
+    env->PSW_USB_V = env->PSW_USB_V << 31;
+    env->PSW_USB_SV |= env->PSW_USB_V;
+    env->PSW_USB_AV = 0;
+
+    return ret;
+}
+
+uint64_t helper_dvinit_b_131(CPUTriCoreState *env, uint32_t r1, uint32_t r2)
+{
+    uint64_t ret = sextract32(r1, 0, 32);
+
+    ret = ret << 24;
+    if (!((r1 & 0x80000000) == (r2 & 0x80000000))) {
+        ret |= 0xffffff;
+    }
+    /* calc overflow */
+    env->PSW_USB_V = ((r2 == 0) || ((r2 == 0xffffffff) && (r1 == 0xffffff80)));
+    env->PSW_USB_V = env->PSW_USB_V << 31;
+    env->PSW_USB_SV |= env->PSW_USB_V;
+    env->PSW_USB_AV = 0;
+
+    return ret;
+}
+
+uint64_t helper_dvinit_h_13(CPUTriCoreState *env, uint32_t r1, uint32_t r2)
+{
+    uint64_t ret;
+    int32_t abs_sig_dividend, abs_base_dividend, abs_divisor;
+    int32_t quotient_sign;
+
+    ret = sextract32(r1, 0, 32);
+    ret = ret << 16;
+    quotient_sign = 0;
+    if (!((r1 & 0x80000000) == (r2 & 0x80000000))) {
+        ret |= 0xffff;
+        quotient_sign = 1;
+    }
+
+    abs_sig_dividend = abs(r1) >> 7;
+    abs_base_dividend = abs(r1) & 0x7f;
+    abs_divisor = abs(r1);
+    /* calc overflow */
+    env->PSW_USB_V = 0;
+    if ((quotient_sign) && (abs_divisor)) {
+        env->PSW_USB_V = (((abs_sig_dividend == abs_divisor) &&
+                         (abs_base_dividend >= abs_divisor)) ||
+                         (abs_sig_dividend > abs_divisor));
+    } else {
+        env->PSW_USB_V = (abs_sig_dividend >= abs_divisor);
+    }
+    env->PSW_USB_V = env->PSW_USB_V << 31;
+    env->PSW_USB_SV |= env->PSW_USB_V;
+    env->PSW_USB_AV = 0;
+
+    return ret;
+}
+
+uint64_t helper_dvinit_h_131(CPUTriCoreState *env, uint32_t r1, uint32_t r2)
+{
+    uint64_t ret = sextract32(r1, 0, 32);
+
+    ret = ret << 16;
+    if (!((r1 & 0x80000000) == (r2 & 0x80000000))) {
+        ret |= 0xffff;
+    }
+    /* calc overflow */
+    env->PSW_USB_V = ((r2 == 0) || ((r2 == 0xffffffff) && (r1 == 0xffff8000)));
+    env->PSW_USB_V = env->PSW_USB_V << 31;
+    env->PSW_USB_SV |= env->PSW_USB_V;
+    env->PSW_USB_AV = 0;
+
+    return ret;
+}
+
 /* context save area (CSA) related helpers */
 
 static int cdc_increment(target_ulong *psw)
diff --git a/target-tricore/translate.c b/target-tricore/translate.c
index 92d6a19ade..1708456281 100644
--- a/target-tricore/translate.c
+++ b/target-tricore/translate.c
@@ -1356,6 +1356,56 @@ static inline void gen_insert(TCGv ret, TCGv r1, TCGv r2, TCGv width, TCGv pos)
     tcg_temp_free(temp2);
 }
 
+static inline void gen_bsplit(TCGv rl, TCGv rh, TCGv r1)
+{
+    TCGv_i64 temp = tcg_temp_new_i64();
+
+    gen_helper_bsplit(temp, r1);
+    tcg_gen_extr_i64_i32(rl, rh, temp);
+
+    tcg_temp_free_i64(temp);
+}
+
+static inline void gen_unpack(TCGv rl, TCGv rh, TCGv r1)
+{
+    TCGv_i64 temp = tcg_temp_new_i64();
+
+    gen_helper_unpack(temp, r1);
+    tcg_gen_extr_i64_i32(rl, rh, temp);
+
+    tcg_temp_free_i64(temp);
+}
+
+static inline void
+gen_dvinit_b(CPUTriCoreState *env, TCGv rl, TCGv rh, TCGv r1, TCGv r2)
+{
+    TCGv_i64 ret = tcg_temp_new_i64();
+
+    if (!tricore_feature(env, TRICORE_FEATURE_131)) {
+        gen_helper_dvinit_b_13(ret, cpu_env, r1, r2);
+    } else {
+        gen_helper_dvinit_b_131(ret, cpu_env, r1, r2);
+    }
+    tcg_gen_extr_i64_i32(rl, rh, ret);
+
+    tcg_temp_free_i64(ret);
+}
+
+static inline void
+gen_dvinit_h(CPUTriCoreState *env, TCGv rl, TCGv rh, TCGv r1, TCGv r2)
+{
+    TCGv_i64 ret = tcg_temp_new_i64();
+
+    if (!tricore_feature(env, TRICORE_FEATURE_131)) {
+        gen_helper_dvinit_h_13(ret, cpu_env, r1, r2);
+    } else {
+        gen_helper_dvinit_h_131(ret, cpu_env, r1, r2);
+    }
+    tcg_gen_extr_i64_i32(rl, rh, ret);
+
+    tcg_temp_free_i64(ret);
+}
+
 /* helpers for generating program flow micro-ops */
 
 static inline void gen_save_pc(target_ulong pc)
@@ -4362,6 +4412,136 @@ static void decode_rr_idirect(CPUTriCoreState *env, DisasContext *ctx)
     ctx->bstate = BS_BRANCH;
 }
 
+static void decode_rr_divide(CPUTriCoreState *env, DisasContext *ctx)
+{
+    uint32_t op2;
+    int r1, r2, r3;
+
+    TCGv temp, temp2;
+
+    op2 = MASK_OP_RR_OP2(ctx->opcode);
+    r3 = MASK_OP_RR_D(ctx->opcode);
+    r2 = MASK_OP_RR_S2(ctx->opcode);
+    r1 = MASK_OP_RR_S1(ctx->opcode);
+
+    switch (op2) {
+    case OPC2_32_RR_BMERGE:
+        gen_helper_bmerge(cpu_gpr_d[r3], cpu_gpr_d[r1], cpu_gpr_d[r2]);
+        break;
+    case OPC2_32_RR_BSPLIT:
+        gen_bsplit(cpu_gpr_d[r3], cpu_gpr_d[r3+1], cpu_gpr_d[r1]);
+        break;
+    case OPC2_32_RR_DVINIT_B:
+        gen_dvinit_b(env, cpu_gpr_d[r3], cpu_gpr_d[r3+1], cpu_gpr_d[r1],
+                     cpu_gpr_d[r2]);
+        break;
+    case OPC2_32_RR_DVINIT_BU:
+        temp = tcg_temp_new();
+        temp2 = tcg_temp_new();
+        /* reset av */
+        tcg_gen_movi_tl(cpu_PSW_AV, 0);
+        if (!tricore_feature(env, TRICORE_FEATURE_131)) {
+            /* overflow = (abs(D[r3+1]) >= abs(D[r2])) */
+            tcg_gen_neg_tl(temp, cpu_gpr_d[r3+1]);
+            /* use cpu_PSW_AV to compare against 0 */
+            tcg_gen_movcond_tl(TCG_COND_LT, temp, cpu_gpr_d[r3+1], cpu_PSW_AV,
+                               temp, cpu_gpr_d[r3+1]);
+            tcg_gen_neg_tl(temp2, cpu_gpr_d[r2]);
+            tcg_gen_movcond_tl(TCG_COND_LT, temp2, cpu_gpr_d[r2], cpu_PSW_AV,
+                               temp2, cpu_gpr_d[r2]);
+            tcg_gen_setcond_tl(TCG_COND_GE, cpu_PSW_V, temp, temp2);
+        } else {
+            /* overflow = (D[b] == 0) */
+            tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_PSW_V, cpu_gpr_d[r2], 0);
+        }
+        tcg_gen_shli_tl(cpu_PSW_V, cpu_PSW_V, 31);
+        /* sv */
+        tcg_gen_or_tl(cpu_PSW_SV, cpu_PSW_SV, cpu_PSW_V);
+        /* write result */
+        tcg_gen_shri_tl(temp, cpu_gpr_d[r1], 8);
+        tcg_gen_shli_tl(cpu_gpr_d[r3], cpu_gpr_d[r1], 24);
+        tcg_gen_mov_tl(cpu_gpr_d[r3+1], temp);
+
+        tcg_temp_free(temp);
+        tcg_temp_free(temp2);
+        break;
+    case OPC2_32_RR_DVINIT_H:
+        gen_dvinit_h(env, cpu_gpr_d[r3], cpu_gpr_d[r3+1], cpu_gpr_d[r1],
+                     cpu_gpr_d[r2]);
+        break;
+    case OPC2_32_RR_DVINIT_HU:
+        temp = tcg_temp_new();
+        temp2 = tcg_temp_new();
+        /* reset av */
+        tcg_gen_movi_tl(cpu_PSW_AV, 0);
+        if (!tricore_feature(env, TRICORE_FEATURE_131)) {
+            /* overflow = (abs(D[r3+1]) >= abs(D[r2])) */
+            tcg_gen_neg_tl(temp, cpu_gpr_d[r3+1]);
+            /* use cpu_PSW_AV to compare against 0 */
+            tcg_gen_movcond_tl(TCG_COND_LT, temp, cpu_gpr_d[r3+1], cpu_PSW_AV,
+                               temp, cpu_gpr_d[r3+1]);
+            tcg_gen_neg_tl(temp2, cpu_gpr_d[r2]);
+            tcg_gen_movcond_tl(TCG_COND_LT, temp2, cpu_gpr_d[r2], cpu_PSW_AV,
+                               temp2, cpu_gpr_d[r2]);
+            tcg_gen_setcond_tl(TCG_COND_GE, cpu_PSW_V, temp, temp2);
+        } else {
+            /* overflow = (D[b] == 0) */
+            tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_PSW_V, cpu_gpr_d[r2], 0);
+        }
+        tcg_gen_shli_tl(cpu_PSW_V, cpu_PSW_V, 31);
+        /* sv */
+        tcg_gen_or_tl(cpu_PSW_SV, cpu_PSW_SV, cpu_PSW_V);
+        /* write result */
+        tcg_gen_mov_tl(temp, cpu_gpr_d[r1]);
+        tcg_gen_shri_tl(cpu_gpr_d[r3+1], temp, 16);
+        tcg_gen_shli_tl(cpu_gpr_d[r3], temp, 16);
+        tcg_temp_free(temp);
+        tcg_temp_free(temp2);
+        break;
+    case OPC2_32_RR_DVINIT:
+        temp = tcg_temp_new();
+        temp2 = tcg_temp_new();
+        /* overflow = ((D[b] == 0) ||
+                      ((D[b] == 0xFFFFFFFF) && (D[a] == 0x80000000))) */
+        tcg_gen_setcondi_tl(TCG_COND_EQ, temp, cpu_gpr_d[r2], 0xffffffff);
+        tcg_gen_setcondi_tl(TCG_COND_EQ, temp2, cpu_gpr_d[r1], 0x80000000);
+        tcg_gen_and_tl(temp, temp, temp2);
+        tcg_gen_setcondi_tl(TCG_COND_EQ, temp2, cpu_gpr_d[r2], 0);
+        tcg_gen_or_tl(cpu_PSW_V, temp, temp2);
+        tcg_gen_shli_tl(cpu_PSW_V, cpu_PSW_V, 31);
+        /* sv */
+        tcg_gen_or_tl(cpu_PSW_SV, cpu_PSW_SV, cpu_PSW_V);
+        /* reset av */
+       tcg_gen_movi_tl(cpu_PSW_AV, 0);
+        /* write result */
+        tcg_gen_mov_tl(cpu_gpr_d[r3], cpu_gpr_d[r1]);
+        /* sign extend to high reg */
+        tcg_gen_sari_tl(cpu_gpr_d[r3+1], cpu_gpr_d[r1], 31);
+        tcg_temp_free(temp);
+        tcg_temp_free(temp2);
+        break;
+    case OPC2_32_RR_DVINIT_U:
+        /* overflow = (D[b] == 0) */
+        tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_PSW_V, cpu_gpr_d[r2], 0);
+        tcg_gen_shli_tl(cpu_PSW_V, cpu_PSW_V, 31);
+        /* sv */
+        tcg_gen_or_tl(cpu_PSW_SV, cpu_PSW_SV, cpu_PSW_V);
+        /* reset av */
+        tcg_gen_movi_tl(cpu_PSW_AV, 0);
+        /* write result */
+        tcg_gen_mov_tl(cpu_gpr_d[r3], cpu_gpr_d[r1]);
+        /* zero extend to high reg*/
+        tcg_gen_movi_tl(cpu_gpr_d[r3+1], 0);
+        break;
+    case OPC2_32_RR_PARITY:
+        gen_helper_parity(cpu_gpr_d[r3], cpu_gpr_d[r1]);
+        break;
+    case OPC2_32_RR_UNPACK:
+        gen_unpack(cpu_gpr_d[r3], cpu_gpr_d[r3+1], cpu_gpr_d[r1]);
+        break;
+    }
+}
+
 static void decode_32Bit_opc(CPUTriCoreState *env, DisasContext *ctx)
 {
     int op1;
@@ -4606,6 +4786,9 @@ static void decode_32Bit_opc(CPUTriCoreState *env, DisasContext *ctx)
     case OPCM_32_RR_IDIRECT:
         decode_rr_idirect(env, ctx);
         break;
+    case OPCM_32_RR_DIVIDE:
+        decode_rr_divide(env, ctx);
+        break;
     }
 }
 
diff --git a/target-tricore/tricore-opcodes.h b/target-tricore/tricore-opcodes.h
index 17d9ad950b..d66f879a39 100644
--- a/target-tricore/tricore-opcodes.h
+++ b/target-tricore/tricore-opcodes.h
@@ -496,7 +496,7 @@ enum {
     OPCM_32_RR_LOGICAL_SHIFT                         = 0x0f,
     OPCM_32_RR_ACCUMULATOR                           = 0x0b,
     OPCM_32_RR_ADRESS                                = 0x01,
-    OPCM_32_RR_FLOAT                                 = 0x4b,
+    OPCM_32_RR_DIVIDE                                = 0x4b,
     OPCM_32_RR_IDIRECT                               = 0x2d,
 /* RR1 Format */
     OPCM_32_RR1_MUL                                  = 0xb3,