The saturate functions about add,subtract and shift functions can
be used in packed extension.Therefore hoist them up.
The endianess process macro is also be hoisted.
Signed-off-by: LIU Zhiwei <zhiwei_...@c-sky.com>
Reviewed-by: Alistair Francis <alistair.fran...@wdc.com>
---
target/riscv/internals.h | 50 ++++++++++++++++++++++
target/riscv/vector_helper.c | 82 +++++++++++-------------------------
2 files changed, 74 insertions(+), 58 deletions(-)
diff --git a/target/riscv/internals.h b/target/riscv/internals.h
index b15ad394bb..698158e116 100644
--- a/target/riscv/internals.h
+++ b/target/riscv/internals.h
@@ -58,4 +58,54 @@ static inline float32 check_nanbox_s(uint64_t f)
}
}
+/*
+ * Note that vector data is stored in host-endian 64-bit chunks,
+ * so addressing units smaller than that needs a host-endian fixup.
+ */
+#ifdef HOST_WORDS_BIGENDIAN
+#define H1(x) ((x) ^ 7)
+#define H1_2(x) ((x) ^ 6)
+#define H1_4(x) ((x) ^ 4)
+#define H2(x) ((x) ^ 3)
+#define H4(x) ((x) ^ 1)
+#define H8(x) ((x))
+#else
+#define H1(x) (x)
+#define H1_2(x) (x)
+#define H1_4(x) (x)
+#define H2(x) (x)
+#define H4(x) (x)
+#define H8(x) (x)
+#endif
+
+/* share functions about saturation */
+int8_t sadd8(CPURISCVState *, int vxrm, int8_t, int8_t);
+int16_t sadd16(CPURISCVState *, int vxrm, int16_t, int16_t);
+int32_t sadd32(CPURISCVState *, int vxrm, int32_t, int32_t);
+int64_t sadd64(CPURISCVState *, int vxrm, int64_t, int64_t);
+
+uint8_t saddu8(CPURISCVState *, int vxrm, uint8_t, uint8_t);
+uint16_t saddu16(CPURISCVState *, int vxrm, uint16_t, uint16_t);
+uint32_t saddu32(CPURISCVState *, int vxrm, uint32_t, uint32_t);
+uint64_t saddu64(CPURISCVState *, int vxrm, uint64_t, uint64_t);
+
+int8_t ssub8(CPURISCVState *, int vxrm, int8_t, int8_t);
+int16_t ssub16(CPURISCVState *, int vxrm, int16_t, int16_t);
+int32_t ssub32(CPURISCVState *, int vxrm, int32_t, int32_t);
+int64_t ssub64(CPURISCVState *, int vxrm, int64_t, int64_t);
+
+uint8_t ssubu8(CPURISCVState *, int vxrm, uint8_t, uint8_t);
+uint16_t ssubu16(CPURISCVState *, int vxrm, uint16_t, uint16_t);
+uint32_t ssubu32(CPURISCVState *, int vxrm, uint32_t, uint32_t);
+uint64_t ssubu64(CPURISCVState *, int vxrm, uint64_t, uint64_t);
+
+/* share shift functions */
+int8_t vssra8(CPURISCVState *env, int vxrm, int8_t a, int8_t b);
+int16_t vssra16(CPURISCVState *env, int vxrm, int16_t a, int16_t b);
+int32_t vssra32(CPURISCVState *env, int vxrm, int32_t a, int32_t b);
+int64_t vssra64(CPURISCVState *env, int vxrm, int64_t a, int64_t b);
+uint8_t vssrl8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b);
+uint16_t vssrl16(CPURISCVState *env, int vxrm, uint16_t a, uint16_t b);
+uint32_t vssrl32(CPURISCVState *env, int vxrm, uint32_t a, uint32_t b);
+uint64_t vssrl64(CPURISCVState *env, int vxrm, uint64_t a, uint64_t b);
#endif
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 12c31aa4b4..c720e7b1fc 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -56,26 +56,6 @@ target_ulong HELPER(vsetvl)(CPURISCVState *env, target_ulong
s1,
return vl;
}
-/*
- * Note that vector data is stored in host-endian 64-bit chunks,
- * so addressing units smaller than that needs a host-endian fixup.
- */
-#ifdef HOST_WORDS_BIGENDIAN
-#define H1(x) ((x) ^ 7)
-#define H1_2(x) ((x) ^ 6)
-#define H1_4(x) ((x) ^ 4)
-#define H2(x) ((x) ^ 3)
-#define H4(x) ((x) ^ 1)
-#define H8(x) ((x))
-#else
-#define H1(x) (x)
-#define H1_2(x) (x)
-#define H1_4(x) (x)
-#define H2(x) (x)
-#define H4(x) (x)
-#define H8(x) (x)
-#endif
-
static inline uint32_t vext_nf(uint32_t desc)
{
return FIELD_EX32(simd_data(desc), VDATA, NF);
@@ -2195,7 +2175,7 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1, void
*vs2, \
do_##NAME, CLEAR_FN); \
}
-static inline uint8_t saddu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t
b)
+uint8_t saddu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
{
uint8_t res = a + b;
if (res < a) {
@@ -2205,8 +2185,7 @@ static inline uint8_t saddu8(CPURISCVState *env, int
vxrm, uint8_t a, uint8_t b)
return res;
}
-static inline uint16_t saddu16(CPURISCVState *env, int vxrm, uint16_t a,
- uint16_t b)
+uint16_t saddu16(CPURISCVState *env, int vxrm, uint16_t a, uint16_t b)
{
uint16_t res = a + b;
if (res < a) {
@@ -2216,8 +2195,7 @@ static inline uint16_t saddu16(CPURISCVState *env, int
vxrm, uint16_t a,
return res;
}
-static inline uint32_t saddu32(CPURISCVState *env, int vxrm, uint32_t a,
- uint32_t b)
+uint32_t saddu32(CPURISCVState *env, int vxrm, uint32_t a, uint32_t b)
{
uint32_t res = a + b;
if (res < a) {
@@ -2227,8 +2205,7 @@ static inline uint32_t saddu32(CPURISCVState *env, int
vxrm, uint32_t a,
return res;
}
-static inline uint64_t saddu64(CPURISCVState *env, int vxrm, uint64_t a,
- uint64_t b)
+uint64_t saddu64(CPURISCVState *env, int vxrm, uint64_t a, uint64_t b)
{
uint64_t res = a + b;
if (res < a) {
@@ -2324,7 +2301,7 @@ 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 sadd8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
{
int8_t res = a + b;
if ((res ^ a) & (res ^ b) & INT8_MIN) {
@@ -2334,7 +2311,7 @@ static inline int8_t sadd8(CPURISCVState *env, int vxrm,
int8_t a, int8_t b)
return res;
}
-static inline int16_t sadd16(CPURISCVState *env, int vxrm, int16_t a, int16_t
b)
+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) {
@@ -2344,7 +2321,7 @@ static inline int16_t sadd16(CPURISCVState *env, int
vxrm, int16_t a, int16_t b)
return res;
}
-static inline int32_t sadd32(CPURISCVState *env, int vxrm, int32_t a, int32_t
b)
+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) {
@@ -2354,7 +2331,7 @@ static inline int32_t sadd32(CPURISCVState *env, int
vxrm, int32_t a, int32_t b)
return res;
}
-static inline int64_t sadd64(CPURISCVState *env, int vxrm, int64_t a, int64_t
b)
+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) {
@@ -2382,7 +2359,7 @@ 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 ssubu8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
{
uint8_t res = a - b;
if (res > a) {
@@ -2392,8 +2369,7 @@ static inline uint8_t ssubu8(CPURISCVState *env, int
vxrm, uint8_t a, uint8_t b)
return res;
}
-static inline uint16_t ssubu16(CPURISCVState *env, int vxrm, uint16_t a,
- uint16_t b)
+uint16_t ssubu16(CPURISCVState *env, int vxrm, uint16_t a, uint16_t b)
{
uint16_t res = a - b;
if (res > a) {
@@ -2403,8 +2379,7 @@ static inline uint16_t ssubu16(CPURISCVState *env, int
vxrm, uint16_t a,
return res;
}
-static inline uint32_t ssubu32(CPURISCVState *env, int vxrm, uint32_t a,
- uint32_t b)
+uint32_t ssubu32(CPURISCVState *env, int vxrm, uint32_t a, uint32_t b)
{
uint32_t res = a - b;
if (res > a) {
@@ -2414,8 +2389,7 @@ static inline uint32_t ssubu32(CPURISCVState *env, int
vxrm, uint32_t a,
return res;
}
-static inline uint64_t ssubu64(CPURISCVState *env, int vxrm, uint64_t a,
- uint64_t b)
+uint64_t ssubu64(CPURISCVState *env, int vxrm, uint64_t a, uint64_t b)
{
uint64_t res = a - b;
if (res > a) {
@@ -2443,7 +2417,7 @@ 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 ssub8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
{
int8_t res = a - b;
if ((res ^ a) & (a ^ b) & INT8_MIN) {
@@ -2453,7 +2427,7 @@ static inline int8_t ssub8(CPURISCVState *env, int vxrm,
int8_t a, int8_t b)
return res;
}
-static inline int16_t ssub16(CPURISCVState *env, int vxrm, int16_t a, int16_t
b)
+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) {
@@ -2463,7 +2437,7 @@ static inline int16_t ssub16(CPURISCVState *env, int
vxrm, int16_t a, int16_t b)
return res;
}
-static inline int32_t ssub32(CPURISCVState *env, int vxrm, int32_t a, int32_t
b)
+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) {
@@ -2473,7 +2447,7 @@ static inline int32_t ssub32(CPURISCVState *env, int
vxrm, int32_t a, int32_t b)
return res;
}
-static inline int64_t ssub64(CPURISCVState *env, int vxrm, int64_t a, int64_t
b)
+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) {
@@ -2914,8 +2888,7 @@ GEN_VEXT_VX_RM(vwsmaccus_vx_h, 2, 4, clearl)
GEN_VEXT_VX_RM(vwsmaccus_vx_w, 4, 8, clearq)
/* Vector Single-Width Scaling Shift Instructions */
-static inline uint8_t
-vssrl8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
+uint8_t vssrl8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
{
uint8_t round, shift = b & 0x7;
uint8_t res;
@@ -2924,8 +2897,7 @@ vssrl8(CPURISCVState *env, int vxrm, uint8_t a, uint8_t b)
res = (a >> shift) + round;
return res;
}
-static inline uint16_t
-vssrl16(CPURISCVState *env, int vxrm, uint16_t a, uint16_t b)
+uint16_t vssrl16(CPURISCVState *env, int vxrm, uint16_t a, uint16_t b)
{
uint8_t round, shift = b & 0xf;
uint16_t res;
@@ -2934,8 +2906,7 @@ vssrl16(CPURISCVState *env, int vxrm, uint16_t a,
uint16_t b)
res = (a >> shift) + round;
return res;
}
-static inline uint32_t
-vssrl32(CPURISCVState *env, int vxrm, uint32_t a, uint32_t b)
+uint32_t vssrl32(CPURISCVState *env, int vxrm, uint32_t a, uint32_t b)
{
uint8_t round, shift = b & 0x1f;
uint32_t res;
@@ -2944,8 +2915,7 @@ vssrl32(CPURISCVState *env, int vxrm, uint32_t a,
uint32_t b)
res = (a >> shift) + round;
return res;
}
-static inline uint64_t
-vssrl64(CPURISCVState *env, int vxrm, uint64_t a, uint64_t b)
+uint64_t vssrl64(CPURISCVState *env, int vxrm, uint64_t a, uint64_t b)
{
uint8_t round, shift = b & 0x3f;
uint64_t res;
@@ -2972,8 +2942,7 @@ GEN_VEXT_VX_RM(vssrl_vx_h, 2, 2, clearh)
GEN_VEXT_VX_RM(vssrl_vx_w, 4, 4, clearl)
GEN_VEXT_VX_RM(vssrl_vx_d, 8, 8, clearq)
-static inline int8_t
-vssra8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
+int8_t vssra8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
{
uint8_t round, shift = b & 0x7;
int8_t res;
@@ -2982,8 +2951,7 @@ vssra8(CPURISCVState *env, int vxrm, int8_t a, int8_t b)
res = (a >> shift) + round;
return res;
}
-static inline int16_t
-vssra16(CPURISCVState *env, int vxrm, int16_t a, int16_t b)
+int16_t vssra16(CPURISCVState *env, int vxrm, int16_t a, int16_t b)
{
uint8_t round, shift = b & 0xf;
int16_t res;
@@ -2992,8 +2960,7 @@ vssra16(CPURISCVState *env, int vxrm, int16_t a, int16_t
b)
res = (a >> shift) + round;
return res;
}
-static inline int32_t
-vssra32(CPURISCVState *env, int vxrm, int32_t a, int32_t b)
+int32_t vssra32(CPURISCVState *env, int vxrm, int32_t a, int32_t b)
{
uint8_t round, shift = b & 0x1f;
int32_t res;
@@ -3002,8 +2969,7 @@ vssra32(CPURISCVState *env, int vxrm, int32_t a, int32_t
b)
res = (a >> shift) + round;
return res;
}
-static inline int64_t
-vssra64(CPURISCVState *env, int vxrm, int64_t a, int64_t b)
+int64_t vssra64(CPURISCVState *env, int vxrm, int64_t a, int64_t b)
{
uint8_t round, shift = b & 0x3f;
int64_t res;
--
2.17.1
