[Qemu-devel] [PATCH v2 4/6] target-tricore: Add instructions of RRR1 opcode format, which have 0x83 as first opcode

[Bastian Koppelmann]

Add helpers:
    * add64_ssov: adds two 64 bit values and saturates the result.
    * addr_h/_ssov: adds two halfwords with two words in q-format with rounding
                    / and saturates each result independetly.

Add microcode generator:
    * gen_add64_d: adds two 64 bit values.
    * gen_addsub64_h: adds/subtracts one halfwords with a word and adds/
                      subtracts another halftword with another word.
    * gen_madd_h/s_h: multiply four halfwords, add each result left justfied
                      to two word values / and saturate each result.
    * gen_maddm_h/s_h: multiply four halfwords, add each result left justfied
                       to two words values in q-format / and saturate each
                       result.
    * gen_maddr32/64_h/s_h: multiply four halfwords, add each result left
                            justfied to two halftwords/words values in q-format
                            / and saturate each result.

Signed-off-by: Bastian Koppelmann <address@hidden>
---
 target-tricore/helper.h    |   3 +
 target-tricore/op_helper.c | 110 ++++++++++++
 target-tricore/translate.c | 421 +++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 534 insertions(+)

diff --git a/target-tricore/helper.h b/target-tricore/helper.h
index 7405fee..a1aa12e 100644
--- a/target-tricore/helper.h
+++ b/target-tricore/helper.h
@@ -17,9 +17,11 @@
 
 /* Arithmetic */
 DEF_HELPER_3(add_ssov, i32, env, i32, i32)
+DEF_HELPER_3(add64_ssov, i64, env, i64, i64)
 DEF_HELPER_3(add_suov, i32, env, i32, i32)
 DEF_HELPER_3(add_h_ssov, i32, env, i32, i32)
 DEF_HELPER_3(add_h_suov, i32, env, i32, i32)
+DEF_HELPER_4(addr_h_ssov, i32, env, i64, i32, i32)
 DEF_HELPER_3(sub_ssov, i32, env, i32, i32)
 DEF_HELPER_3(sub_suov, i32, env, i32, i32)
 DEF_HELPER_3(sub_h_ssov, i32, env, i32, i32)
@@ -44,6 +46,7 @@ DEF_HELPER_2(abs_b, i32, env, i32)
 DEF_HELPER_2(abs_h, i32, env, i32)
 DEF_HELPER_3(absdif_b, i32, env, i32, i32)
 DEF_HELPER_3(absdif_h, i32, env, i32, i32)
+DEF_HELPER_4(addr_h, i32, env, i64, i32, i32)
 DEF_HELPER_3(add_b, i32, env, i32, i32)
 DEF_HELPER_3(add_h, i32, env, i32, i32)
 DEF_HELPER_3(sub_b, i32, env, i32, i32)
diff --git a/target-tricore/op_helper.c b/target-tricore/op_helper.c
index 08bf4ae..8cd154b 100644
--- a/target-tricore/op_helper.c
+++ b/target-tricore/op_helper.c
@@ -184,6 +184,31 @@ target_ulong helper_add_ssov(CPUTriCoreState *env, 
target_ulong r1,
     return ssov32(env, result);
 }
 
+uint64_t helper_add64_ssov(CPUTriCoreState *env, uint64_t r1, uint64_t r2)
+{
+    uint64_t result;
+    int64_t ovf;
+
+    result = r1 + r2;
+    ovf = (result ^ r1) & ~(r1 ^ r2);
+    env->PSW_USB_AV = (result ^ result * 2u) >> 32;
+    env->PSW_USB_SAV |= env->PSW_USB_AV;
+    if (ovf < 0) {
+        env->PSW_USB_V = (1 << 31);
+        env->PSW_USB_SV = (1 << 31);
+        /* ext_ret > MAX_INT */
+        if ((int64_t)r1 >= 0) {
+            result = INT64_MAX;
+        /* ext_ret < MIN_INT */
+        } else {
+            result = INT64_MIN;
+        }
+    } else {
+        env->PSW_USB_V = 0;
+    }
+    return result;
+}
+
 target_ulong helper_add_h_ssov(CPUTriCoreState *env, target_ulong r1,
                                target_ulong r2)
 {
@@ -194,6 +219,53 @@ target_ulong helper_add_h_ssov(CPUTriCoreState *env, 
target_ulong r1,
     return ssov16(env, ret_hw0, ret_hw1);
 }
 
+uint32_t helper_addr_h_ssov(CPUTriCoreState *env, uint64_t r1, uint32_t r2_l,
+                            uint32_t r2_h)
+{
+    int64_t mul_res0 = sextract64(r1, 0, 32);
+    int64_t mul_res1 = sextract64(r1, 32, 32);
+    int64_t r2_low = sextract64(r2_l, 0, 32);
+    int64_t r2_high = sextract64(r2_h, 0, 32);
+    int64_t result0, result1;
+    uint32_t ovf0, ovf1;
+    uint32_t avf0, avf1;
+
+    ovf0 = ovf1 = 0;
+
+    result0 = r2_low + mul_res0 + 0x8000;
+    result1 = r2_high + mul_res1 + 0x8000;
+
+    avf0 = result0 * 2u;
+    avf0 = result0 ^ avf0;
+    avf1 = result1 * 2u;
+    avf1 = result1 ^ avf1;
+
+    if (result0 > INT32_MAX) {
+        ovf0 = (1 << 31);
+        result0 = INT32_MAX;
+    } else if (result0 < INT32_MIN) {
+        ovf0 = (1 << 31);
+        result0 = INT32_MIN;
+    }
+
+    if (result1 > INT32_MAX) {
+        ovf1 = (1 << 31);
+        result1 = INT32_MAX;
+    } else if (result1 < INT32_MIN) {
+        ovf1 = (1 << 31);
+        result1 = INT32_MIN;
+    }
+
+    env->PSW_USB_V = ovf0 | ovf1;
+    env->PSW_USB_SV |= env->PSW_USB_V;
+
+    env->PSW_USB_AV = avf0 | avf1;
+    env->PSW_USB_SAV |= env->PSW_USB_AV;
+
+    return (result1 & 0xffff0000ULL) | ((result0 >> 16) & 0xffffULL);
+}
+
+
 target_ulong helper_add_suov(CPUTriCoreState *env, target_ulong r1,
                              target_ulong r2)
 {
@@ -619,6 +691,44 @@ uint32_t helper_absdif_h(CPUTriCoreState *env, 
target_ulong r1, target_ulong r2)
     return ret;
 }
 
+uint32_t helper_addr_h(CPUTriCoreState *env, uint64_t r1, uint32_t r2_l,
+                       uint32_t r2_h)
+{
+    int64_t mul_res0 = sextract64(r1, 0, 32);
+    int64_t mul_res1 = sextract64(r1, 32, 32);
+    int64_t r2_low = sextract64(r2_l, 0, 32);
+    int64_t r2_high = sextract64(r2_h, 0, 32);
+    int64_t result0, result1;
+    uint32_t ovf0, ovf1;
+    uint32_t avf0, avf1;
+
+    ovf0 = ovf1 = 0;
+
+    result0 = r2_low + mul_res0 + 0x8000;
+    result1 = r2_high + mul_res1 + 0x8000;
+
+    if ((result0 > INT32_MAX) || (result0 < INT32_MIN)) {
+        ovf0 = (1 << 31);
+    }
+
+    if ((result1 > INT32_MAX) || (result1 < INT32_MIN)) {
+        ovf1 = (1 << 31);
+    }
+
+    env->PSW_USB_V = ovf0 | ovf1;
+    env->PSW_USB_SV |= env->PSW_USB_V;
+
+    avf0 = result0 * 2u;
+    avf0 = result0 ^ avf0;
+    avf1 = result1 * 2u;
+    avf1 = result1 ^ avf1;
+
+    env->PSW_USB_AV = avf0 | avf1;
+    env->PSW_USB_SAV |= env->PSW_USB_AV;
+
+    return (result1 & 0xffff0000ULL) | ((result0 >> 16) & 0xffffULL);
+}
+
 uint32_t helper_add_b(CPUTriCoreState *env, target_ulong r1, target_ulong r2)
 {
     int32_t b, i;
diff --git a/target-tricore/translate.c b/target-tricore/translate.c
index cb97d5a..afa16ba 100644
--- a/target-tricore/translate.c
+++ b/target-tricore/translate.c
@@ -80,6 +80,13 @@ enum {
     BS_EXCP   = 3,
 };
 
+enum {
+    MODE_LL = 0,
+    MODE_LU = 1,
+    MODE_UL = 2,
+    MODE_UU = 3,
+};
+
 void tricore_cpu_dump_state(CPUState *cs, FILE *f,
                             fprintf_function cpu_fprintf, int flags)
 {
@@ -395,6 +402,88 @@ static inline void gen_add_d(TCGv ret, TCGv r1, TCGv r2)
     tcg_temp_free(t0);
 }
 
+static inline void
+gen_add64_d(TCGv_i64 ret, TCGv_i64 r1, TCGv_i64 r2)
+{
+    TCGv temp = tcg_temp_new();
+    TCGv_i64 t0 = tcg_temp_new_i64();
+    TCGv_i64 t1 = tcg_temp_new_i64();
+    TCGv_i64 result = tcg_temp_new_i64();
+
+    tcg_gen_add_i64(result, r1, r2);
+    /* calc v bit */
+    tcg_gen_xor_i64(t1, result, r1);
+    tcg_gen_xor_i64(t0, r1, r2);
+    tcg_gen_andc_i64(t1, t1, t0);
+    tcg_gen_trunc_shr_i64_i32(cpu_PSW_V, t1, 32);
+    /* calc SV bit */
+    tcg_gen_or_tl(cpu_PSW_SV, cpu_PSW_SV, cpu_PSW_V);
+    /* calc AV/SAV bits */
+    tcg_gen_trunc_shr_i64_i32(temp, result, 32);
+    tcg_gen_add_tl(cpu_PSW_AV, temp, temp);
+    tcg_gen_xor_tl(cpu_PSW_AV, temp, cpu_PSW_AV);
+    /* calc SAV */
+    tcg_gen_or_tl(cpu_PSW_SAV, cpu_PSW_SAV, cpu_PSW_AV);
+    /* write back result */
+    tcg_gen_mov_i64(ret, result);
+
+    tcg_temp_free(temp);
+    tcg_temp_free_i64(result);
+    tcg_temp_free_i64(t0);
+    tcg_temp_free_i64(t1);
+}
+
+static inline void
+gen_addsub64_h(TCGv ret_low, TCGv ret_high, TCGv r1_low, TCGv r1_high, TCGv r2,
+               TCGv r3, void(*op1)(TCGv, TCGv, TCGv),
+               void(*op2)(TCGv, TCGv, TCGv))
+{
+    TCGv temp = tcg_temp_new();
+    TCGv temp2 = tcg_temp_new();
+    TCGv temp3 = tcg_temp_new();
+    TCGv temp4 = tcg_temp_new();
+
+    (*op1)(temp, r1_low, r2);
+    /* calc V0 bit */
+    tcg_gen_xor_tl(temp2, temp, r1_low);
+    tcg_gen_xor_tl(temp3, r1_low, r2);
+    if (op1 == tcg_gen_add_tl) {
+        tcg_gen_andc_tl(temp2, temp2, temp3);
+    } else {
+        tcg_gen_and_tl(temp2, temp2, temp3);
+    }
+
+    (*op2)(temp3, r1_high, r3);
+    /* calc V1 bit */
+    tcg_gen_xor_tl(cpu_PSW_V, temp3, r1_high);
+    tcg_gen_xor_tl(temp4, r1_high, r3);
+    if (op2 == tcg_gen_add_tl) {
+        tcg_gen_andc_tl(cpu_PSW_V, cpu_PSW_V, temp4);
+    } else {
+        tcg_gen_andc_tl(cpu_PSW_V, cpu_PSW_V, temp4);
+    }
+    /* combine V0/V1 bits */
+    tcg_gen_or_tl(cpu_PSW_V, cpu_PSW_V, temp2);
+    /* calc sv bit */
+    tcg_gen_or_tl(cpu_PSW_SV, cpu_PSW_SV, cpu_PSW_V);
+    /* write result */
+    tcg_gen_mov_tl(ret_low, temp);
+    tcg_gen_mov_tl(ret_high, temp3);
+    /* calc AV bit */
+    tcg_gen_add_tl(temp, ret_low, ret_low);
+    tcg_gen_xor_tl(temp, temp, ret_low);
+    tcg_gen_add_tl(cpu_PSW_AV, ret_high, ret_high);
+    tcg_gen_xor_tl(cpu_PSW_AV, cpu_PSW_AV, ret_high);
+    tcg_gen_or_tl(cpu_PSW_AV, cpu_PSW_AV, temp);
+    /* calc SAV bit */
+    tcg_gen_or_tl(cpu_PSW_SAV, cpu_PSW_SAV, cpu_PSW_AV);
+
+    tcg_temp_free(temp);
+    tcg_temp_free(temp2);
+    tcg_temp_free(temp3);
+    tcg_temp_free(temp4);
+}
+
 /* ret = r2 + (r1 * r3); */
 static inline void gen_madd32_d(TCGv ret, TCGv r1, TCGv r2, TCGv r3)
 {
@@ -523,6 +612,221 @@ gen_maddui64_d(TCGv ret_low, TCGv ret_high, TCGv r1, TCGv 
r2_low, TCGv r2_high,
     tcg_temp_free(temp);
 }
 
+static inline void
+gen_madd_h(TCGv ret_low, TCGv ret_high, TCGv r1_low, TCGv r1_high, TCGv r2,
+           TCGv r3, uint32_t n, uint32_t mode)
+{
+    TCGv temp = tcg_const_i32(n);
+    TCGv temp2 = tcg_temp_new();
+    TCGv_i64 temp64 = tcg_temp_new_i64();
+    switch (mode) {
+    case MODE_LL:
+        GEN_HELPER_LL(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_LU:
+        GEN_HELPER_LU(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UL:
+        GEN_HELPER_UL(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UU:
+        GEN_HELPER_UU(mul_h, temp64, r2, r3, temp);
+        break;
+    }
+    tcg_gen_extr_i64_i32(temp, temp2, temp64);
+    gen_addsub64_h(ret_low, ret_high, r1_low, r1_high, temp, temp2,
+                   tcg_gen_add_tl, tcg_gen_add_tl);
+    tcg_temp_free(temp);
+    tcg_temp_free(temp2);
+    tcg_temp_free_i64(temp64);
+}
+
+static inline void gen_adds(TCGv ret, TCGv r1, TCGv r2);
+
+static inline void
+gen_madds_h(TCGv ret_low, TCGv ret_high, TCGv r1_low, TCGv r1_high, TCGv r2,
+           TCGv r3, uint32_t n, uint32_t mode)
+{
+    TCGv temp = tcg_const_i32(n);
+    TCGv temp2 = tcg_temp_new();
+    TCGv temp3 = tcg_temp_new();
+    TCGv_i64 temp64 = tcg_temp_new_i64();
+
+    switch (mode) {
+    case MODE_LL:
+        GEN_HELPER_LL(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_LU:
+        GEN_HELPER_LU(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UL:
+        GEN_HELPER_UL(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UU:
+        GEN_HELPER_UU(mul_h, temp64, r2, r3, temp);
+        break;
+    }
+    tcg_gen_extr_i64_i32(temp, temp2, temp64);
+    gen_adds(ret_low, r1_low, temp);
+    tcg_gen_mov_tl(temp, cpu_PSW_V);
+    tcg_gen_mov_tl(temp3, cpu_PSW_AV);
+    gen_adds(ret_high, r1_high, temp2);
+    /* combine v bits */
+    tcg_gen_or_tl(cpu_PSW_V, cpu_PSW_V, temp);
+    /* combine av bits */
+    tcg_gen_or_tl(cpu_PSW_AV, cpu_PSW_AV, temp3);
+
+    tcg_temp_free(temp);
+    tcg_temp_free(temp2);
+    tcg_temp_free(temp3);
+    tcg_temp_free_i64(temp64);
+
+}
+
+static inline void
+gen_maddm_h(TCGv ret_low, TCGv ret_high, TCGv r1_low, TCGv r1_high, TCGv r2,
+           TCGv r3, uint32_t n, uint32_t mode)
+{
+    TCGv temp = tcg_const_i32(n);
+    TCGv_i64 temp64 = tcg_temp_new_i64();
+    TCGv_i64 temp64_2 = tcg_temp_new_i64();
+    TCGv_i64 temp64_3 = tcg_temp_new_i64();
+    switch (mode) {
+    case MODE_LL:
+        GEN_HELPER_LL(mulm_h, temp64, r2, r3, temp);
+        break;
+    case MODE_LU:
+        GEN_HELPER_LU(mulm_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UL:
+        GEN_HELPER_UL(mulm_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UU:
+        GEN_HELPER_UU(mulm_h, temp64, r2, r3, temp);
+        break;
+    }
+    tcg_gen_concat_i32_i64(temp64_2, r1_low, r1_high);
+    gen_add64_d(temp64_3, temp64_2, temp64);
+    /* write back result */
+    tcg_gen_extr_i64_i32(ret_low, ret_high, temp64_3);
+
+    tcg_temp_free(temp);
+    tcg_temp_free_i64(temp64);
+    tcg_temp_free_i64(temp64_2);
+    tcg_temp_free_i64(temp64_3);
+}
+
+static inline void
+gen_maddms_h(TCGv ret_low, TCGv ret_high, TCGv r1_low, TCGv r1_high, TCGv r2,
+           TCGv r3, uint32_t n, uint32_t mode)
+{
+    TCGv temp = tcg_const_i32(n);
+    TCGv_i64 temp64 = tcg_temp_new_i64();
+    TCGv_i64 temp64_2 = tcg_temp_new_i64();
+    switch (mode) {
+    case MODE_LL:
+        GEN_HELPER_LL(mulm_h, temp64, r2, r3, temp);
+        break;
+    case MODE_LU:
+        GEN_HELPER_LU(mulm_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UL:
+        GEN_HELPER_UL(mulm_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UU:
+        GEN_HELPER_UU(mulm_h, temp64, r2, r3, temp);
+        break;
+    }
+    tcg_gen_concat_i32_i64(temp64_2, r1_low, r1_high);
+    gen_helper_add64_ssov(temp64, cpu_env, temp64_2, temp64);
+    tcg_gen_extr_i64_i32(ret_low, ret_high, temp64);
+
+    tcg_temp_free(temp);
+    tcg_temp_free_i64(temp64);
+    tcg_temp_free_i64(temp64_2);
+}
+
+static inline void
+gen_maddr64_h(TCGv ret, TCGv r1_low, TCGv r1_high, TCGv r2, TCGv r3, uint32_t 
n,
+              uint32_t mode)
+{
+    TCGv temp = tcg_const_i32(n);
+    TCGv_i64 temp64 = tcg_temp_new_i64();
+    switch (mode) {
+    case MODE_LL:
+        GEN_HELPER_LL(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_LU:
+        GEN_HELPER_LU(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UL:
+        GEN_HELPER_UL(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UU:
+        GEN_HELPER_UU(mul_h, temp64, r2, r3, temp);
+        break;
+    }
+    gen_helper_addr_h(ret, cpu_env, temp64, r1_low, r1_high);
+
+    tcg_temp_free(temp);
+    tcg_temp_free_i64(temp64);
+}
+
+static inline void
+gen_maddr32_h(TCGv ret, TCGv r1, TCGv r2, TCGv r3, uint32_t n, uint32_t mode)
+{
+    TCGv temp = tcg_temp_new();
+    TCGv temp2 = tcg_temp_new();
+
+    tcg_gen_andi_tl(temp2, r1, 0xffff0000);
+    tcg_gen_shli_tl(temp, r1, 16);
+    gen_maddr64_h(ret, temp, temp2, r2, r3, n, mode);
+
+    tcg_temp_free(temp);
+    tcg_temp_free(temp2);
+}
+
+static inline void
+gen_maddr64s_h(TCGv ret, TCGv r1_low, TCGv r1_high, TCGv r2, TCGv r3,
+               uint32_t n, uint32_t mode)
+{
+    TCGv temp = tcg_const_i32(n);
+    TCGv_i64 temp64 = tcg_temp_new_i64();
+    switch (mode) {
+    case MODE_LL:
+        GEN_HELPER_LL(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_LU:
+        GEN_HELPER_LU(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UL:
+        GEN_HELPER_UL(mul_h, temp64, r2, r3, temp);
+        break;
+    case MODE_UU:
+        GEN_HELPER_UU(mul_h, temp64, r2, r3, temp);
+        break;
+    }
+    gen_helper_addr_h_ssov(ret, cpu_env, temp64, r1_low, r1_high);
+
+    tcg_temp_free(temp);
+    tcg_temp_free_i64(temp64);
+}
+
+static inline void
+gen_maddr32s_h(TCGv ret, TCGv r1, TCGv r2, TCGv r3, uint32_t n, uint32_t mode)
+{
+    TCGv temp = tcg_temp_new();
+    TCGv temp2 = tcg_temp_new();
+
+    tcg_gen_andi_tl(temp2, r1, 0xffff0000);
+    tcg_gen_shli_tl(temp, r1, 16);
+    gen_maddr64s_h(ret, temp, temp2, r2, r3, n, mode);
+
+    tcg_temp_free(temp);
+    tcg_temp_free(temp2);
+}
+
+
 /* ret = r2 - (r1 * r3); */
 static inline void gen_msub32_d(TCGv ret, TCGv r1, TCGv r2, TCGv r3)
 {
@@ -5300,6 +5604,119 @@ static void decode_rrr2_msub(CPUTriCoreState *env, 
DisasContext *ctx)
     }
 }
 
+/* RRR1 format */
+static void decode_rrr1_madd(CPUTriCoreState *env, DisasContext *ctx)
+{
+    uint32_t op2;
+    uint32_t r1, r2, r3, r4, n;
+
+    op2 = MASK_OP_RRR1_OP2(ctx->opcode);
+    r1 = MASK_OP_RRR1_S1(ctx->opcode);
+    r2 = MASK_OP_RRR1_S2(ctx->opcode);
+    r3 = MASK_OP_RRR1_S3(ctx->opcode);
+    r4 = MASK_OP_RRR1_D(ctx->opcode);
+    n = MASK_OP_RRR1_N(ctx->opcode);
+
+    switch (op2) {
+    case OPC2_32_RRR1_MADD_H_LL:
+        gen_madd_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                   cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_LL);
+        break;
+    case OPC2_32_RRR1_MADD_H_LU:
+        gen_madd_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                   cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_LU);
+        break;
+    case OPC2_32_RRR1_MADD_H_UL:
+        gen_madd_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                   cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_UL);
+        break;
+    case OPC2_32_RRR1_MADD_H_UU:
+        gen_madd_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                   cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_UU);
+        break;
+    case OPC2_32_RRR1_MADDS_H_LL:
+        gen_madds_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                    cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_LL);
+        break;
+    case OPC2_32_RRR1_MADDS_H_LU:
+        gen_madds_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                    cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_LU);
+        break;
+    case OPC2_32_RRR1_MADDS_H_UL:
+        gen_madds_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                    cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_UL);
+        break;
+    case OPC2_32_RRR1_MADDS_H_UU:
+        gen_madds_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                    cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_UU);
+        break;
+    case OPC2_32_RRR1_MADDM_H_LL:
+        gen_maddm_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                    cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_LL);
+        break;
+    case OPC2_32_RRR1_MADDM_H_LU:
+        gen_maddm_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                    cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_LU);
+        break;
+    case OPC2_32_RRR1_MADDM_H_UL:
+        gen_maddm_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                    cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_UL);
+        break;
+    case OPC2_32_RRR1_MADDM_H_UU:
+        gen_maddm_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                    cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, MODE_UU);
+        break;
+    case OPC2_32_RRR1_MADDMS_H_LL:
+        gen_maddms_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                     cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, 
MODE_LL);
+        break;
+    case OPC2_32_RRR1_MADDMS_H_LU:
+        gen_maddms_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                     cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, 
MODE_LU);
+        break;
+    case OPC2_32_RRR1_MADDMS_H_UL:
+        gen_maddms_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                     cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, 
MODE_UL);
+        break;
+    case OPC2_32_RRR1_MADDMS_H_UU:
+        gen_maddms_h(cpu_gpr_d[r4], cpu_gpr_d[r4+1], cpu_gpr_d[r3],
+                     cpu_gpr_d[r3+1], cpu_gpr_d[r1], cpu_gpr_d[r2], n, 
MODE_UU);
+        break;
+    case OPC2_32_RRR1_MADDR_H_LL:
+        gen_maddr32_h(cpu_gpr_d[r4], cpu_gpr_d[r3], cpu_gpr_d[r1],
+                      cpu_gpr_d[r2], n, MODE_LL);
+        break;
+    case OPC2_32_RRR1_MADDR_H_LU:
+        gen_maddr32_h(cpu_gpr_d[r4], cpu_gpr_d[r3], cpu_gpr_d[r1],
+                      cpu_gpr_d[r2], n, MODE_LU);
+        break;
+    case OPC2_32_RRR1_MADDR_H_UL:
+        gen_maddr32_h(cpu_gpr_d[r4], cpu_gpr_d[r3], cpu_gpr_d[r1],
+                      cpu_gpr_d[r2], n, MODE_UL);
+        break;
+    case OPC2_32_RRR1_MADDR_H_UU:
+        gen_maddr32_h(cpu_gpr_d[r4], cpu_gpr_d[r3], cpu_gpr_d[r1],
+                      cpu_gpr_d[r2], n, MODE_UU);
+        break;
+    case OPC2_32_RRR1_MADDRS_H_LL:
+        gen_maddr32s_h(cpu_gpr_d[r4], cpu_gpr_d[r3], cpu_gpr_d[r1],
+                       cpu_gpr_d[r2], n, MODE_LL);
+        break;
+    case OPC2_32_RRR1_MADDRS_H_LU:
+        gen_maddr32s_h(cpu_gpr_d[r4], cpu_gpr_d[r3], cpu_gpr_d[r1],
+                       cpu_gpr_d[r2], n, MODE_LU);
+        break;
+    case OPC2_32_RRR1_MADDRS_H_UL:
+        gen_maddr32s_h(cpu_gpr_d[r4], cpu_gpr_d[r3], cpu_gpr_d[r1],
+                       cpu_gpr_d[r2], n, MODE_UL);
+        break;
+    case OPC2_32_RRR1_MADDRS_H_UU:
+        gen_maddr32s_h(cpu_gpr_d[r4], cpu_gpr_d[r3], cpu_gpr_d[r1],
+                       cpu_gpr_d[r2], n, MODE_UU);
+        break;
+    }
+}
+
 static void decode_32Bit_opc(CPUTriCoreState *env, DisasContext *ctx)
 {
     int op1;
@@ -5596,6 +6013,10 @@ static void decode_32Bit_opc(CPUTriCoreState *env, 
DisasContext *ctx)
     case OPCM_32_RRR2_MSUB:
         decode_rrr2_msub(env, ctx);
         break;
+/* RRR1 format */
+    case OPCM_32_RRR1_MADD:
+        decode_rrr1_madd(env, ctx);
+        break;
     }
 }
 
-- 
2.3.0