target/riscv: vector single-width saturating add and subtract

Signed-off-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Message-Id: <20200701152549.1218-25-zhiwei_liu@c-sky.com>
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>

1 files changed, 385 insertions, 0 deletions

diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 92d90d07d6..1277aa1c10 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -2112,3 +2112,388 @@ GEN_VEXT_VMERGE_VX(vmerge_vxm_b, int8_t,  H1, clearb)
 GEN_VEXT_VMERGE_VX(vmerge_vxm_h, int16_t, H2, clearh)
 GEN_VEXT_VMERGE_VX(vmerge_vxm_w, int32_t, H4, clearl)
 GEN_VEXT_VMERGE_VX(vmerge_vxm_d, int64_t, H8, clearq)
+
+/*
+ *** Vector Fixed-Point Arithmetic Instructions
+ */
+
+/* Vector Single-Width Saturating Add and Subtract */
+
+/*
+ * As fixed point instructions probably have round mode and saturation,
+ * define common macros for fixed point here.
+ */
+typedef void opivv2_rm_fn(void *vd, void *vs1, void *vs2, int i,
+                          CPURISCVState *env, int vxrm);
+
+#define OPIVV2_RM(NAME, TD, T1, T2, TX1, TX2, HD, HS1, HS2, OP)     \
+static inline void                                                  \
+do_##NAME(void *vd, void *vs1, void *vs2, int i,                    \
+          CPURISCVState *env, int vxrm)                             \
+{                                                                   \
+    TX1 s1 = *((T1 *)vs1 + HS1(i));                                 \
+    TX2 s2 = *((T2 *)vs2 + HS2(i));                                 \
+    *((TD *)vd + HD(i)) = OP(env, vxrm, s2, s1);                    \
+}
+
+static inline void
+vext_vv_rm_1(void *vd, void *v0, void *vs1, void *vs2,
+             CPURISCVState *env,
+             uint32_t vl, uint32_t vm, uint32_t mlen, int vxrm,
+             opivv2_rm_fn *fn)
+{
+    for (uint32_t i = 0; i < vl; i++) {
+        if (!vm && !vext_elem_mask(v0, mlen, i)) {
+            continue;
+        }
+        fn(vd, vs1, vs2, i, env, vxrm);
+    }
+}
+
+static inline void
+vext_vv_rm_2(void *vd, void *v0, void *vs1, void *vs2,
+             CPURISCVState *env,
+             uint32_t desc, uint32_t esz, uint32_t dsz,
+             opivv2_rm_fn *fn, clear_fn *clearfn)
+{
+    uint32_t vlmax = vext_maxsz(desc) / esz;
+    uint32_t mlen = vext_mlen(desc);
+    uint32_t vm = vext_vm(desc);
+    uint32_t vl = env->vl;
+
+    switch (env->vxrm) {
+    case 0: /* rnu */
+        vext_vv_rm_1(vd, v0, vs1, vs2,
+                     env, vl, vm, mlen, 0, fn);
+        break;
+    case 1: /* rne */
+        vext_vv_rm_1(vd, v0, vs1, vs2,
+                     env, vl, vm, mlen, 1, fn);
+        break;
+    case 2: /* rdn */
+        vext_vv_rm_1(vd, v0, vs1, vs2,
+                     env, vl, vm, mlen, 2, fn);
+        break;
+    default: /* rod */
+        vext_vv_rm_1(vd, v0, vs1, vs2,
+                     env, vl, vm, mlen, 3, fn);
+        break;
+    }
+
+    clearfn(vd, vl, vl * dsz,  vlmax * dsz);
+}
+
+/* generate helpers for fixed point instructions with OPIVV format */
+#define GEN_VEXT_VV_RM(NAME, ESZ, DSZ, CLEAR_FN)                \
+void HELPER(NAME)(void *vd, void *v0, void *vs1, void *vs2,     \
+                  CPURISCVState *env, uint32_t desc)            \
+{                                                               \
+    vext_vv_rm_2(vd, v0, vs1, vs2, env, desc, ESZ, DSZ,         \
+                 do_##NAME, CLEAR_FN);                          \
+}
+
+static inline uint8_t saddu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
+{
+    uint8_t res = a + b;
+    if (res < a) {
+        res = UINT8_MAX;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline uint16_t saddu16(CPURISCVState *env, int vxrm, uint16_t a,
+                               uint16_t b)
+{
+    uint16_t res = a + b;
+    if (res < a) {
+        res = UINT16_MAX;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline uint32_t saddu32(CPURISCVState *env, int vxrm, uint32_t a,
+                               uint32_t b)
+{
+    uint32_t res = a + b;
+    if (res < a) {
+        res = UINT32_MAX;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline uint64_t saddu64(CPURISCVState *env, int vxrm, uint64_t a,
+                               uint64_t b)
+{
+    uint64_t res = a + b;
+    if (res < a) {
+        res = UINT64_MAX;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+RVVCALL(OPIVV2_RM, vsaddu_vv_b, OP_UUU_B, H1, H1, H1, saddu8)
+RVVCALL(OPIVV2_RM, vsaddu_vv_h, OP_UUU_H, H2, H2, H2, saddu16)
+RVVCALL(OPIVV2_RM, vsaddu_vv_w, OP_UUU_W, H4, H4, H4, saddu32)
+RVVCALL(OPIVV2_RM, vsaddu_vv_d, OP_UUU_D, H8, H8, H8, saddu64)
+GEN_VEXT_VV_RM(vsaddu_vv_b, 1, 1, clearb)
+GEN_VEXT_VV_RM(vsaddu_vv_h, 2, 2, clearh)
+GEN_VEXT_VV_RM(vsaddu_vv_w, 4, 4, clearl)
+GEN_VEXT_VV_RM(vsaddu_vv_d, 8, 8, clearq)
+
+typedef void opivx2_rm_fn(void *vd, target_long s1, void *vs2, int i,
+                          CPURISCVState *env, int vxrm);
+
+#define OPIVX2_RM(NAME, TD, T1, T2, TX1, TX2, HD, HS2, OP)          \
+static inline void                                                  \
+do_##NAME(void *vd, target_long s1, void *vs2, int i,               \
+          CPURISCVState *env, int vxrm)                             \
+{                                                                   \
+    TX2 s2 = *((T2 *)vs2 + HS2(i));                                 \
+    *((TD *)vd + HD(i)) = OP(env, vxrm, s2, (TX1)(T1)s1);           \
+}
+
+static inline void
+vext_vx_rm_1(void *vd, void *v0, target_long s1, void *vs2,
+             CPURISCVState *env,
+             uint32_t vl, uint32_t vm, uint32_t mlen, int vxrm,
+             opivx2_rm_fn *fn)
+{
+    for (uint32_t i = 0; i < vl; i++) {
+        if (!vm && !vext_elem_mask(v0, mlen, i)) {
+            continue;
+        }
+        fn(vd, s1, vs2, i, env, vxrm);
+    }
+}
+
+static inline void
+vext_vx_rm_2(void *vd, void *v0, target_long s1, void *vs2,
+             CPURISCVState *env,
+             uint32_t desc, uint32_t esz, uint32_t dsz,
+             opivx2_rm_fn *fn, clear_fn *clearfn)
+{
+    uint32_t vlmax = vext_maxsz(desc) / esz;
+    uint32_t mlen = vext_mlen(desc);
+    uint32_t vm = vext_vm(desc);
+    uint32_t vl = env->vl;
+
+    switch (env->vxrm) {
+    case 0: /* rnu */
+        vext_vx_rm_1(vd, v0, s1, vs2,
+                     env, vl, vm, mlen, 0, fn);
+        break;
+    case 1: /* rne */
+        vext_vx_rm_1(vd, v0, s1, vs2,
+                     env, vl, vm, mlen, 1, fn);
+        break;
+    case 2: /* rdn */
+        vext_vx_rm_1(vd, v0, s1, vs2,
+                     env, vl, vm, mlen, 2, fn);
+        break;
+    default: /* rod */
+        vext_vx_rm_1(vd, v0, s1, vs2,
+                     env, vl, vm, mlen, 3, fn);
+        break;
+    }
+
+    clearfn(vd, vl, vl * dsz,  vlmax * dsz);
+}
+
+/* generate helpers for fixed point instructions with OPIVX format */
+#define GEN_VEXT_VX_RM(NAME, ESZ, DSZ, CLEAR_FN)          \
+void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
+        void *vs2, CPURISCVState *env, uint32_t desc)     \
+{                                                         \
+    vext_vx_rm_2(vd, v0, s1, vs2, env, desc, ESZ, DSZ,    \
+                 do_##NAME, CLEAR_FN);                    \
+}
+
+RVVCALL(OPIVX2_RM, vsaddu_vx_b, OP_UUU_B, H1, H1, saddu8)
+RVVCALL(OPIVX2_RM, vsaddu_vx_h, OP_UUU_H, H2, H2, saddu16)
+RVVCALL(OPIVX2_RM, vsaddu_vx_w, OP_UUU_W, H4, H4, saddu32)
+RVVCALL(OPIVX2_RM, vsaddu_vx_d, OP_UUU_D, H8, H8, saddu64)
+GEN_VEXT_VX_RM(vsaddu_vx_b, 1, 1, clearb)
+GEN_VEXT_VX_RM(vsaddu_vx_h, 2, 2, clearh)
+GEN_VEXT_VX_RM(vsaddu_vx_w, 4, 4, clearl)
+GEN_VEXT_VX_RM(vsaddu_vx_d, 8, 8, clearq)
+
+static inline int8_t sadd8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
+{
+    int8_t res = a + b;
+    if ((res ^ a) & (res ^ b) & INT8_MIN) {
+        res = a > 0 ? INT8_MAX : INT8_MIN;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline int16_t sadd16(CPURISCVState *env, int vxrm, int16_t a, int16_t b)
+{
+    int16_t res = a + b;
+    if ((res ^ a) & (res ^ b) & INT16_MIN) {
+        res = a > 0 ? INT16_MAX : INT16_MIN;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline int32_t sadd32(CPURISCVState *env, int vxrm, int32_t a, int32_t b)
+{
+    int32_t res = a + b;
+    if ((res ^ a) & (res ^ b) & INT32_MIN) {
+        res = a > 0 ? INT32_MAX : INT32_MIN;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline int64_t sadd64(CPURISCVState *env, int vxrm, int64_t a, int64_t b)
+{
+    int64_t res = a + b;
+    if ((res ^ a) & (res ^ b) & INT64_MIN) {
+        res = a > 0 ? INT64_MAX : INT64_MIN;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+RVVCALL(OPIVV2_RM, vsadd_vv_b, OP_SSS_B, H1, H1, H1, sadd8)
+RVVCALL(OPIVV2_RM, vsadd_vv_h, OP_SSS_H, H2, H2, H2, sadd16)
+RVVCALL(OPIVV2_RM, vsadd_vv_w, OP_SSS_W, H4, H4, H4, sadd32)
+RVVCALL(OPIVV2_RM, vsadd_vv_d, OP_SSS_D, H8, H8, H8, sadd64)
+GEN_VEXT_VV_RM(vsadd_vv_b, 1, 1, clearb)
+GEN_VEXT_VV_RM(vsadd_vv_h, 2, 2, clearh)
+GEN_VEXT_VV_RM(vsadd_vv_w, 4, 4, clearl)
+GEN_VEXT_VV_RM(vsadd_vv_d, 8, 8, clearq)
+
+RVVCALL(OPIVX2_RM, vsadd_vx_b, OP_SSS_B, H1, H1, sadd8)
+RVVCALL(OPIVX2_RM, vsadd_vx_h, OP_SSS_H, H2, H2, sadd16)
+RVVCALL(OPIVX2_RM, vsadd_vx_w, OP_SSS_W, H4, H4, sadd32)
+RVVCALL(OPIVX2_RM, vsadd_vx_d, OP_SSS_D, H8, H8, sadd64)
+GEN_VEXT_VX_RM(vsadd_vx_b, 1, 1, clearb)
+GEN_VEXT_VX_RM(vsadd_vx_h, 2, 2, clearh)
+GEN_VEXT_VX_RM(vsadd_vx_w, 4, 4, clearl)
+GEN_VEXT_VX_RM(vsadd_vx_d, 8, 8, clearq)
+
+static inline uint8_t ssubu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
+{
+    uint8_t res = a - b;
+    if (res > a) {
+        res = 0;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline uint16_t ssubu16(CPURISCVState *env, int vxrm, uint16_t a,
+                               uint16_t b)
+{
+    uint16_t res = a - b;
+    if (res > a) {
+        res = 0;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline uint32_t ssubu32(CPURISCVState *env, int vxrm, uint32_t a,
+                               uint32_t b)
+{
+    uint32_t res = a - b;
+    if (res > a) {
+        res = 0;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline uint64_t ssubu64(CPURISCVState *env, int vxrm, uint64_t a,
+                               uint64_t b)
+{
+    uint64_t res = a - b;
+    if (res > a) {
+        res = 0;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+RVVCALL(OPIVV2_RM, vssubu_vv_b, OP_UUU_B, H1, H1, H1, ssubu8)
+RVVCALL(OPIVV2_RM, vssubu_vv_h, OP_UUU_H, H2, H2, H2, ssubu16)
+RVVCALL(OPIVV2_RM, vssubu_vv_w, OP_UUU_W, H4, H4, H4, ssubu32)
+RVVCALL(OPIVV2_RM, vssubu_vv_d, OP_UUU_D, H8, H8, H8, ssubu64)
+GEN_VEXT_VV_RM(vssubu_vv_b, 1, 1, clearb)
+GEN_VEXT_VV_RM(vssubu_vv_h, 2, 2, clearh)
+GEN_VEXT_VV_RM(vssubu_vv_w, 4, 4, clearl)
+GEN_VEXT_VV_RM(vssubu_vv_d, 8, 8, clearq)
+
+RVVCALL(OPIVX2_RM, vssubu_vx_b, OP_UUU_B, H1, H1, ssubu8)
+RVVCALL(OPIVX2_RM, vssubu_vx_h, OP_UUU_H, H2, H2, ssubu16)
+RVVCALL(OPIVX2_RM, vssubu_vx_w, OP_UUU_W, H4, H4, ssubu32)
+RVVCALL(OPIVX2_RM, vssubu_vx_d, OP_UUU_D, H8, H8, ssubu64)
+GEN_VEXT_VX_RM(vssubu_vx_b, 1, 1, clearb)
+GEN_VEXT_VX_RM(vssubu_vx_h, 2, 2, clearh)
+GEN_VEXT_VX_RM(vssubu_vx_w, 4, 4, clearl)
+GEN_VEXT_VX_RM(vssubu_vx_d, 8, 8, clearq)
+
+static inline int8_t ssub8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
+{
+    int8_t res = a - b;
+    if ((res ^ a) & (a ^ b) & INT8_MIN) {
+        res = a > 0 ? INT8_MAX : INT8_MIN;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline int16_t ssub16(CPURISCVState *env, int vxrm, int16_t a, int16_t b)
+{
+    int16_t res = a - b;
+    if ((res ^ a) & (a ^ b) & INT16_MIN) {
+        res = a > 0 ? INT16_MAX : INT16_MIN;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline int32_t ssub32(CPURISCVState *env, int vxrm, int32_t a, int32_t b)
+{
+    int32_t res = a - b;
+    if ((res ^ a) & (a ^ b) & INT32_MIN) {
+        res = a > 0 ? INT32_MAX : INT32_MIN;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+static inline int64_t ssub64(CPURISCVState *env, int vxrm, int64_t a, int64_t b)
+{
+    int64_t res = a - b;
+    if ((res ^ a) & (a ^ b) & INT64_MIN) {
+        res = a > 0 ? INT64_MAX : INT64_MIN;
+        env->vxsat = 0x1;
+    }
+    return res;
+}
+
+RVVCALL(OPIVV2_RM, vssub_vv_b, OP_SSS_B, H1, H1, H1, ssub8)
+RVVCALL(OPIVV2_RM, vssub_vv_h, OP_SSS_H, H2, H2, H2, ssub16)
+RVVCALL(OPIVV2_RM, vssub_vv_w, OP_SSS_W, H4, H4, H4, ssub32)
+RVVCALL(OPIVV2_RM, vssub_vv_d, OP_SSS_D, H8, H8, H8, ssub64)
+GEN_VEXT_VV_RM(vssub_vv_b, 1, 1, clearb)
+GEN_VEXT_VV_RM(vssub_vv_h, 2, 2, clearh)
+GEN_VEXT_VV_RM(vssub_vv_w, 4, 4, clearl)
+GEN_VEXT_VV_RM(vssub_vv_d, 8, 8, clearq)
+
+RVVCALL(OPIVX2_RM, vssub_vx_b, OP_SSS_B, H1, H1, ssub8)
+RVVCALL(OPIVX2_RM, vssub_vx_h, OP_SSS_H, H2, H2, ssub16)
+RVVCALL(OPIVX2_RM, vssub_vx_w, OP_SSS_W, H4, H4, ssub32)
+RVVCALL(OPIVX2_RM, vssub_vx_d, OP_SSS_D, H8, H8, ssub64)
+GEN_VEXT_VX_RM(vssub_vx_b, 1, 1, clearb)
+GEN_VEXT_VX_RM(vssub_vx_h, 2, 2, clearh)
+GEN_VEXT_VX_RM(vssub_vx_w, 4, 4, clearl)
+GEN_VEXT_VX_RM(vssub_vx_d, 8, 8, clearq)