aboutsummaryrefslogtreecommitdiff
path: root/target-arm/neon_helper.c
diff options
context:
space:
mode:
Diffstat (limited to 'target-arm/neon_helper.c')
-rw-r--r--target-arm/neon_helper.c269
1 files changed, 124 insertions, 145 deletions
diff --git a/target-arm/neon_helper.c b/target-arm/neon_helper.c
index 002a9c11a6..f5b173aa71 100644
--- a/target-arm/neon_helper.c
+++ b/target-arm/neon_helper.c
@@ -10,39 +10,15 @@
#include <stdio.h>
#include "cpu.h"
-#include "exec-all.h"
-#include "helpers.h"
+#include "exec.h"
+#include "helper.h"
#define SIGNBIT (uint32_t)0x80000000
#define SIGNBIT64 ((uint64_t)1 << 63)
#define SET_QC() env->vfp.xregs[ARM_VFP_FPSCR] = CPSR_Q
-static float_status neon_float_status;
-#define NFS &neon_float_status
-
-/* Helper routines to perform bitwise copies between float and int. */
-static inline float32 vfp_itos(uint32_t i)
-{
- union {
- uint32_t i;
- float32 s;
- } v;
-
- v.i = i;
- return v.s;
-}
-
-static inline uint32_t vfp_stoi(float32 s)
-{
- union {
- uint32_t i;
- float32 s;
- } v;
-
- v.s = s;
- return v.i;
-}
+#define NFS (&env->vfp.standard_fp_status)
#define NEON_TYPE1(name, type) \
typedef struct \
@@ -139,10 +115,6 @@ NEON_TYPE1(u32, uint32_t)
uint32_t HELPER(glue(neon_,name))(uint32_t arg1, uint32_t arg2) \
NEON_VOP_BODY(vtype, n)
-#define NEON_VOP_ENV(name, vtype, n) \
-uint32_t HELPER(glue(neon_,name))(CPUState *env, uint32_t arg1, uint32_t arg2) \
-NEON_VOP_BODY(vtype, n)
-
/* Pairwise operations. */
/* For 32-bit elements each segment only contains a single element, so
the elementwise and pairwise operations are the same. */
@@ -191,14 +163,14 @@ uint32_t HELPER(glue(neon_,name))(uint32_t arg) \
dest = tmp; \
}} while(0)
#define NEON_FN(dest, src1, src2) NEON_USAT(dest, src1, src2, uint8_t)
-NEON_VOP_ENV(qadd_u8, neon_u8, 4)
+NEON_VOP(qadd_u8, neon_u8, 4)
#undef NEON_FN
#define NEON_FN(dest, src1, src2) NEON_USAT(dest, src1, src2, uint16_t)
-NEON_VOP_ENV(qadd_u16, neon_u16, 2)
+NEON_VOP(qadd_u16, neon_u16, 2)
#undef NEON_FN
#undef NEON_USAT
-uint32_t HELPER(neon_qadd_u32)(CPUState *env, uint32_t a, uint32_t b)
+uint32_t HELPER(neon_qadd_u32)(uint32_t a, uint32_t b)
{
uint32_t res = a + b;
if (res < a) {
@@ -208,7 +180,7 @@ uint32_t HELPER(neon_qadd_u32)(CPUState *env, uint32_t a, uint32_t b)
return res;
}
-uint64_t HELPER(neon_qadd_u64)(CPUState *env, uint64_t src1, uint64_t src2)
+uint64_t HELPER(neon_qadd_u64)(uint64_t src1, uint64_t src2)
{
uint64_t res;
@@ -233,14 +205,14 @@ uint64_t HELPER(neon_qadd_u64)(CPUState *env, uint64_t src1, uint64_t src2)
dest = tmp; \
} while(0)
#define NEON_FN(dest, src1, src2) NEON_SSAT(dest, src1, src2, int8_t)
-NEON_VOP_ENV(qadd_s8, neon_s8, 4)
+NEON_VOP(qadd_s8, neon_s8, 4)
#undef NEON_FN
#define NEON_FN(dest, src1, src2) NEON_SSAT(dest, src1, src2, int16_t)
-NEON_VOP_ENV(qadd_s16, neon_s16, 2)
+NEON_VOP(qadd_s16, neon_s16, 2)
#undef NEON_FN
#undef NEON_SSAT
-uint32_t HELPER(neon_qadd_s32)(CPUState *env, uint32_t a, uint32_t b)
+uint32_t HELPER(neon_qadd_s32)(uint32_t a, uint32_t b)
{
uint32_t res = a + b;
if (((res ^ a) & SIGNBIT) && !((a ^ b) & SIGNBIT)) {
@@ -250,7 +222,7 @@ uint32_t HELPER(neon_qadd_s32)(CPUState *env, uint32_t a, uint32_t b)
return res;
}
-uint64_t HELPER(neon_qadd_s64)(CPUState *env, uint64_t src1, uint64_t src2)
+uint64_t HELPER(neon_qadd_s64)(uint64_t src1, uint64_t src2)
{
uint64_t res;
@@ -271,14 +243,14 @@ uint64_t HELPER(neon_qadd_s64)(CPUState *env, uint64_t src1, uint64_t src2)
dest = tmp; \
}} while(0)
#define NEON_FN(dest, src1, src2) NEON_USAT(dest, src1, src2, uint8_t)
-NEON_VOP_ENV(qsub_u8, neon_u8, 4)
+NEON_VOP(qsub_u8, neon_u8, 4)
#undef NEON_FN
#define NEON_FN(dest, src1, src2) NEON_USAT(dest, src1, src2, uint16_t)
-NEON_VOP_ENV(qsub_u16, neon_u16, 2)
+NEON_VOP(qsub_u16, neon_u16, 2)
#undef NEON_FN
#undef NEON_USAT
-uint32_t HELPER(neon_qsub_u32)(CPUState *env, uint32_t a, uint32_t b)
+uint32_t HELPER(neon_qsub_u32)(uint32_t a, uint32_t b)
{
uint32_t res = a - b;
if (res > a) {
@@ -288,7 +260,7 @@ uint32_t HELPER(neon_qsub_u32)(CPUState *env, uint32_t a, uint32_t b)
return res;
}
-uint64_t HELPER(neon_qsub_u64)(CPUState *env, uint64_t src1, uint64_t src2)
+uint64_t HELPER(neon_qsub_u64)(uint64_t src1, uint64_t src2)
{
uint64_t res;
@@ -314,14 +286,14 @@ uint64_t HELPER(neon_qsub_u64)(CPUState *env, uint64_t src1, uint64_t src2)
dest = tmp; \
} while(0)
#define NEON_FN(dest, src1, src2) NEON_SSAT(dest, src1, src2, int8_t)
-NEON_VOP_ENV(qsub_s8, neon_s8, 4)
+NEON_VOP(qsub_s8, neon_s8, 4)
#undef NEON_FN
#define NEON_FN(dest, src1, src2) NEON_SSAT(dest, src1, src2, int16_t)
-NEON_VOP_ENV(qsub_s16, neon_s16, 2)
+NEON_VOP(qsub_s16, neon_s16, 2)
#undef NEON_FN
#undef NEON_SSAT
-uint32_t HELPER(neon_qsub_s32)(CPUState *env, uint32_t a, uint32_t b)
+uint32_t HELPER(neon_qsub_s32)(uint32_t a, uint32_t b)
{
uint32_t res = a - b;
if (((res ^ a) & SIGNBIT) && ((a ^ b) & SIGNBIT)) {
@@ -331,7 +303,7 @@ uint32_t HELPER(neon_qsub_s32)(CPUState *env, uint32_t a, uint32_t b)
return res;
}
-uint64_t HELPER(neon_qsub_s64)(CPUState *env, uint64_t src1, uint64_t src2)
+uint64_t HELPER(neon_qsub_s64)(uint64_t src1, uint64_t src2)
{
uint64_t res;
@@ -682,12 +654,12 @@ uint64_t HELPER(neon_rshl_u64)(uint64_t val, uint64_t shiftop)
dest = ~0; \
} \
}} while (0)
-NEON_VOP_ENV(qshl_u8, neon_u8, 4)
-NEON_VOP_ENV(qshl_u16, neon_u16, 2)
-NEON_VOP_ENV(qshl_u32, neon_u32, 1)
+NEON_VOP(qshl_u8, neon_u8, 4)
+NEON_VOP(qshl_u16, neon_u16, 2)
+NEON_VOP(qshl_u32, neon_u32, 1)
#undef NEON_FN
-uint64_t HELPER(neon_qshl_u64)(CPUState *env, uint64_t val, uint64_t shiftop)
+uint64_t HELPER(neon_qshl_u64)(uint64_t val, uint64_t shiftop)
{
int8_t shift = (int8_t)shiftop;
if (shift >= 64) {
@@ -737,12 +709,12 @@ uint64_t HELPER(neon_qshl_u64)(CPUState *env, uint64_t val, uint64_t shiftop)
} \
} \
}} while (0)
-NEON_VOP_ENV(qshl_s8, neon_s8, 4)
-NEON_VOP_ENV(qshl_s16, neon_s16, 2)
-NEON_VOP_ENV(qshl_s32, neon_s32, 1)
+NEON_VOP(qshl_s8, neon_s8, 4)
+NEON_VOP(qshl_s16, neon_s16, 2)
+NEON_VOP(qshl_s32, neon_s32, 1)
#undef NEON_FN
-uint64_t HELPER(neon_qshl_s64)(CPUState *env, uint64_t valop, uint64_t shiftop)
+uint64_t HELPER(neon_qshl_s64)(uint64_t valop, uint64_t shiftop)
{
int8_t shift = (uint8_t)shiftop;
int64_t val = valop;
@@ -792,26 +764,26 @@ uint64_t HELPER(neon_qshl_s64)(CPUState *env, uint64_t valop, uint64_t shiftop)
} \
} \
}} while (0)
-NEON_VOP_ENV(qshlu_s8, neon_u8, 4)
-NEON_VOP_ENV(qshlu_s16, neon_u16, 2)
+NEON_VOP(qshlu_s8, neon_u8, 4)
+NEON_VOP(qshlu_s16, neon_u16, 2)
#undef NEON_FN
-uint32_t HELPER(neon_qshlu_s32)(CPUState *env, uint32_t valop, uint32_t shiftop)
+uint32_t HELPER(neon_qshlu_s32)(uint32_t valop, uint32_t shiftop)
{
if ((int32_t)valop < 0) {
SET_QC();
return 0;
}
- return helper_neon_qshl_u32(env, valop, shiftop);
+ return helper_neon_qshl_u32(valop, shiftop);
}
-uint64_t HELPER(neon_qshlu_s64)(CPUState *env, uint64_t valop, uint64_t shiftop)
+uint64_t HELPER(neon_qshlu_s64)(uint64_t valop, uint64_t shiftop)
{
if ((int64_t)valop < 0) {
SET_QC();
return 0;
}
- return helper_neon_qshl_u64(env, valop, shiftop);
+ return helper_neon_qshl_u64(valop, shiftop);
}
/* FIXME: This is wrong. */
@@ -838,13 +810,13 @@ uint64_t HELPER(neon_qshlu_s64)(CPUState *env, uint64_t valop, uint64_t shiftop)
dest = ~0; \
} \
}} while (0)
-NEON_VOP_ENV(qrshl_u8, neon_u8, 4)
-NEON_VOP_ENV(qrshl_u16, neon_u16, 2)
+NEON_VOP(qrshl_u8, neon_u8, 4)
+NEON_VOP(qrshl_u16, neon_u16, 2)
#undef NEON_FN
/* The addition of the rounding constant may overflow, so we use an
* intermediate 64 bits accumulator. */
-uint32_t HELPER(neon_qrshl_u32)(CPUState *env, uint32_t val, uint32_t shiftop)
+uint32_t HELPER(neon_qrshl_u32)(uint32_t val, uint32_t shiftop)
{
uint32_t dest;
int8_t shift = (int8_t)shiftop;
@@ -874,7 +846,7 @@ uint32_t HELPER(neon_qrshl_u32)(CPUState *env, uint32_t val, uint32_t shiftop)
/* Handling addition overflow with 64 bits inputs values is more
* tricky than with 32 bits values. */
-uint64_t HELPER(neon_qrshl_u64)(CPUState *env, uint64_t val, uint64_t shiftop)
+uint64_t HELPER(neon_qrshl_u64)(uint64_t val, uint64_t shiftop)
{
int8_t shift = (int8_t)shiftop;
if (shift >= 64) {
@@ -935,13 +907,13 @@ uint64_t HELPER(neon_qrshl_u64)(CPUState *env, uint64_t val, uint64_t shiftop)
} \
} \
}} while (0)
-NEON_VOP_ENV(qrshl_s8, neon_s8, 4)
-NEON_VOP_ENV(qrshl_s16, neon_s16, 2)
+NEON_VOP(qrshl_s8, neon_s8, 4)
+NEON_VOP(qrshl_s16, neon_s16, 2)
#undef NEON_FN
/* The addition of the rounding constant may overflow, so we use an
* intermediate 64 bits accumulator. */
-uint32_t HELPER(neon_qrshl_s32)(CPUState *env, uint32_t valop, uint32_t shiftop)
+uint32_t HELPER(neon_qrshl_s32)(uint32_t valop, uint32_t shiftop)
{
int32_t dest;
int32_t val = (int32_t)valop;
@@ -970,7 +942,7 @@ uint32_t HELPER(neon_qrshl_s32)(CPUState *env, uint32_t valop, uint32_t shiftop)
/* Handling addition overflow with 64 bits inputs values is more
* tricky than with 32 bits values. */
-uint64_t HELPER(neon_qrshl_s64)(CPUState *env, uint64_t valop, uint64_t shiftop)
+uint64_t HELPER(neon_qrshl_s64)(uint64_t valop, uint64_t shiftop)
{
int8_t shift = (uint8_t)shiftop;
int64_t val = valop;
@@ -1179,10 +1151,10 @@ uint32_t HELPER(neon_cnt_u8)(uint32_t x)
dest = tmp >> 16; \
} while(0)
#define NEON_FN(dest, src1, src2) NEON_QDMULH16(dest, src1, src2, 0)
-NEON_VOP_ENV(qdmulh_s16, neon_s16, 2)
+NEON_VOP(qdmulh_s16, neon_s16, 2)
#undef NEON_FN
#define NEON_FN(dest, src1, src2) NEON_QDMULH16(dest, src1, src2, 1)
-NEON_VOP_ENV(qrdmulh_s16, neon_s16, 2)
+NEON_VOP(qrdmulh_s16, neon_s16, 2)
#undef NEON_FN
#undef NEON_QDMULH16
@@ -1205,10 +1177,10 @@ NEON_VOP_ENV(qrdmulh_s16, neon_s16, 2)
dest = tmp >> 32; \
} while(0)
#define NEON_FN(dest, src1, src2) NEON_QDMULH32(dest, src1, src2, 0)
-NEON_VOP_ENV(qdmulh_s32, neon_s32, 1)
+NEON_VOP(qdmulh_s32, neon_s32, 1)
#undef NEON_FN
#define NEON_FN(dest, src1, src2) NEON_QDMULH32(dest, src1, src2, 1)
-NEON_VOP_ENV(qrdmulh_s32, neon_s32, 1)
+NEON_VOP(qrdmulh_s32, neon_s32, 1)
#undef NEON_FN
#undef NEON_QDMULH32
@@ -1249,7 +1221,7 @@ uint32_t HELPER(neon_narrow_round_high_u16)(uint64_t x)
return ((x >> 16) & 0xffff) | ((x >> 32) & 0xffff0000);
}
-uint32_t HELPER(neon_unarrow_sat8)(CPUState *env, uint64_t x)
+uint32_t HELPER(neon_unarrow_sat8)(uint64_t x)
{
uint16_t s;
uint8_t d;
@@ -1276,7 +1248,7 @@ uint32_t HELPER(neon_unarrow_sat8)(CPUState *env, uint64_t x)
return res;
}
-uint32_t HELPER(neon_narrow_sat_u8)(CPUState *env, uint64_t x)
+uint32_t HELPER(neon_narrow_sat_u8)(uint64_t x)
{
uint16_t s;
uint8_t d;
@@ -1299,7 +1271,7 @@ uint32_t HELPER(neon_narrow_sat_u8)(CPUState *env, uint64_t x)
return res;
}
-uint32_t HELPER(neon_narrow_sat_s8)(CPUState *env, uint64_t x)
+uint32_t HELPER(neon_narrow_sat_s8)(uint64_t x)
{
int16_t s;
uint8_t d;
@@ -1322,7 +1294,7 @@ uint32_t HELPER(neon_narrow_sat_s8)(CPUState *env, uint64_t x)
return res;
}
-uint32_t HELPER(neon_unarrow_sat16)(CPUState *env, uint64_t x)
+uint32_t HELPER(neon_unarrow_sat16)(uint64_t x)
{
uint32_t high;
uint32_t low;
@@ -1345,7 +1317,7 @@ uint32_t HELPER(neon_unarrow_sat16)(CPUState *env, uint64_t x)
return low | (high << 16);
}
-uint32_t HELPER(neon_narrow_sat_u16)(CPUState *env, uint64_t x)
+uint32_t HELPER(neon_narrow_sat_u16)(uint64_t x)
{
uint32_t high;
uint32_t low;
@@ -1362,7 +1334,7 @@ uint32_t HELPER(neon_narrow_sat_u16)(CPUState *env, uint64_t x)
return low | (high << 16);
}
-uint32_t HELPER(neon_narrow_sat_s16)(CPUState *env, uint64_t x)
+uint32_t HELPER(neon_narrow_sat_s16)(uint64_t x)
{
int32_t low;
int32_t high;
@@ -1379,7 +1351,7 @@ uint32_t HELPER(neon_narrow_sat_s16)(CPUState *env, uint64_t x)
return (uint16_t)low | (high << 16);
}
-uint32_t HELPER(neon_unarrow_sat32)(CPUState *env, uint64_t x)
+uint32_t HELPER(neon_unarrow_sat32)(uint64_t x)
{
if (x & 0x8000000000000000ull) {
SET_QC();
@@ -1392,7 +1364,7 @@ uint32_t HELPER(neon_unarrow_sat32)(CPUState *env, uint64_t x)
return x;
}
-uint32_t HELPER(neon_narrow_sat_u32)(CPUState *env, uint64_t x)
+uint32_t HELPER(neon_narrow_sat_u32)(uint64_t x)
{
if (x > 0xffffffffu) {
SET_QC();
@@ -1401,7 +1373,7 @@ uint32_t HELPER(neon_narrow_sat_u32)(CPUState *env, uint64_t x)
return x;
}
-uint32_t HELPER(neon_narrow_sat_s32)(CPUState *env, uint64_t x)
+uint32_t HELPER(neon_narrow_sat_s32)(uint64_t x)
{
if ((int64_t)x != (int32_t)x) {
SET_QC();
@@ -1508,7 +1480,7 @@ uint64_t HELPER(neon_subl_u32)(uint64_t a, uint64_t b)
return (a - b) ^ mask;
}
-uint64_t HELPER(neon_addl_saturate_s32)(CPUState *env, uint64_t a, uint64_t b)
+uint64_t HELPER(neon_addl_saturate_s32)(uint64_t a, uint64_t b)
{
uint32_t x, y;
uint32_t low, high;
@@ -1530,7 +1502,7 @@ uint64_t HELPER(neon_addl_saturate_s32)(CPUState *env, uint64_t a, uint64_t b)
return low | ((uint64_t)high << 32);
}
-uint64_t HELPER(neon_addl_saturate_s64)(CPUState *env, uint64_t a, uint64_t b)
+uint64_t HELPER(neon_addl_saturate_s64)(uint64_t a, uint64_t b)
{
uint64_t result;
@@ -1542,9 +1514,13 @@ uint64_t HELPER(neon_addl_saturate_s64)(CPUState *env, uint64_t a, uint64_t b)
return result;
}
-#define DO_ABD(dest, x, y, type) do { \
- type tmp_x = x; \
- type tmp_y = y; \
+/* We have to do the arithmetic in a larger type than
+ * the input type, because for example with a signed 32 bit
+ * op the absolute difference can overflow a signed 32 bit value.
+ */
+#define DO_ABD(dest, x, y, intype, arithtype) do { \
+ arithtype tmp_x = (intype)(x); \
+ arithtype tmp_y = (intype)(y); \
dest = ((tmp_x > tmp_y) ? tmp_x - tmp_y : tmp_y - tmp_x); \
} while(0)
@@ -1552,12 +1528,12 @@ uint64_t HELPER(neon_abdl_u16)(uint32_t a, uint32_t b)
{
uint64_t tmp;
uint64_t result;
- DO_ABD(result, a, b, uint8_t);
- DO_ABD(tmp, a >> 8, b >> 8, uint8_t);
+ DO_ABD(result, a, b, uint8_t, uint32_t);
+ DO_ABD(tmp, a >> 8, b >> 8, uint8_t, uint32_t);
result |= tmp << 16;
- DO_ABD(tmp, a >> 16, b >> 16, uint8_t);
+ DO_ABD(tmp, a >> 16, b >> 16, uint8_t, uint32_t);
result |= tmp << 32;
- DO_ABD(tmp, a >> 24, b >> 24, uint8_t);
+ DO_ABD(tmp, a >> 24, b >> 24, uint8_t, uint32_t);
result |= tmp << 48;
return result;
}
@@ -1566,12 +1542,12 @@ uint64_t HELPER(neon_abdl_s16)(uint32_t a, uint32_t b)
{
uint64_t tmp;
uint64_t result;
- DO_ABD(result, a, b, int8_t);
- DO_ABD(tmp, a >> 8, b >> 8, int8_t);
+ DO_ABD(result, a, b, int8_t, int32_t);
+ DO_ABD(tmp, a >> 8, b >> 8, int8_t, int32_t);
result |= tmp << 16;
- DO_ABD(tmp, a >> 16, b >> 16, int8_t);
+ DO_ABD(tmp, a >> 16, b >> 16, int8_t, int32_t);
result |= tmp << 32;
- DO_ABD(tmp, a >> 24, b >> 24, int8_t);
+ DO_ABD(tmp, a >> 24, b >> 24, int8_t, int32_t);
result |= tmp << 48;
return result;
}
@@ -1580,8 +1556,8 @@ uint64_t HELPER(neon_abdl_u32)(uint32_t a, uint32_t b)
{
uint64_t tmp;
uint64_t result;
- DO_ABD(result, a, b, uint16_t);
- DO_ABD(tmp, a >> 16, b >> 16, uint16_t);
+ DO_ABD(result, a, b, uint16_t, uint32_t);
+ DO_ABD(tmp, a >> 16, b >> 16, uint16_t, uint32_t);
return result | (tmp << 32);
}
@@ -1589,22 +1565,22 @@ uint64_t HELPER(neon_abdl_s32)(uint32_t a, uint32_t b)
{
uint64_t tmp;
uint64_t result;
- DO_ABD(result, a, b, int16_t);
- DO_ABD(tmp, a >> 16, b >> 16, int16_t);
+ DO_ABD(result, a, b, int16_t, int32_t);
+ DO_ABD(tmp, a >> 16, b >> 16, int16_t, int32_t);
return result | (tmp << 32);
}
uint64_t HELPER(neon_abdl_u64)(uint32_t a, uint32_t b)
{
uint64_t result;
- DO_ABD(result, a, b, uint32_t);
+ DO_ABD(result, a, b, uint32_t, uint64_t);
return result;
}
uint64_t HELPER(neon_abdl_s64)(uint32_t a, uint32_t b)
{
uint64_t result;
- DO_ABD(result, a, b, int32_t);
+ DO_ABD(result, a, b, int32_t, int64_t);
return result;
}
#undef DO_ABD
@@ -1702,7 +1678,7 @@ uint64_t HELPER(neon_negl_u64)(uint64_t x)
} else if (x < 0) { \
x = -x; \
}} while (0)
-uint32_t HELPER(neon_qabs_s8)(CPUState *env, uint32_t x)
+uint32_t HELPER(neon_qabs_s8)(uint32_t x)
{
neon_s8 vec;
NEON_UNPACK(neon_s8, vec, x);
@@ -1722,7 +1698,7 @@ uint32_t HELPER(neon_qabs_s8)(CPUState *env, uint32_t x)
} else { \
x = -x; \
}} while (0)
-uint32_t HELPER(neon_qneg_s8)(CPUState *env, uint32_t x)
+uint32_t HELPER(neon_qneg_s8)(uint32_t x)
{
neon_s8 vec;
NEON_UNPACK(neon_s8, vec, x);
@@ -1742,7 +1718,7 @@ uint32_t HELPER(neon_qneg_s8)(CPUState *env, uint32_t x)
} else if (x < 0) { \
x = -x; \
}} while (0)
-uint32_t HELPER(neon_qabs_s16)(CPUState *env, uint32_t x)
+uint32_t HELPER(neon_qabs_s16)(uint32_t x)
{
neon_s16 vec;
NEON_UNPACK(neon_s16, vec, x);
@@ -1760,7 +1736,7 @@ uint32_t HELPER(neon_qabs_s16)(CPUState *env, uint32_t x)
} else { \
x = -x; \
}} while (0)
-uint32_t HELPER(neon_qneg_s16)(CPUState *env, uint32_t x)
+uint32_t HELPER(neon_qneg_s16)(uint32_t x)
{
neon_s16 vec;
NEON_UNPACK(neon_s16, vec, x);
@@ -1771,7 +1747,7 @@ uint32_t HELPER(neon_qneg_s16)(CPUState *env, uint32_t x)
}
#undef DO_QNEG16
-uint32_t HELPER(neon_qabs_s32)(CPUState *env, uint32_t x)
+uint32_t HELPER(neon_qabs_s32)(uint32_t x)
{
if (x == SIGNBIT) {
SET_QC();
@@ -1782,7 +1758,7 @@ uint32_t HELPER(neon_qabs_s32)(CPUState *env, uint32_t x)
return x;
}
-uint32_t HELPER(neon_qneg_s32)(CPUState *env, uint32_t x)
+uint32_t HELPER(neon_qneg_s32)(uint32_t x)
{
if (x == SIGNBIT) {
SET_QC();
@@ -1796,73 +1772,76 @@ uint32_t HELPER(neon_qneg_s32)(CPUState *env, uint32_t x)
/* NEON Float helpers. */
uint32_t HELPER(neon_min_f32)(uint32_t a, uint32_t b)
{
- float32 f0 = vfp_itos(a);
- float32 f1 = vfp_itos(b);
- return (float32_compare_quiet(f0, f1, NFS) == -1) ? a : b;
+ return float32_val(float32_min(make_float32(a), make_float32(b), NFS));
}
uint32_t HELPER(neon_max_f32)(uint32_t a, uint32_t b)
{
- float32 f0 = vfp_itos(a);
- float32 f1 = vfp_itos(b);
- return (float32_compare_quiet(f0, f1, NFS) == 1) ? a : b;
+ return float32_val(float32_max(make_float32(a), make_float32(b), NFS));
}
uint32_t HELPER(neon_abd_f32)(uint32_t a, uint32_t b)
{
- float32 f0 = vfp_itos(a);
- float32 f1 = vfp_itos(b);
- return vfp_stoi((float32_compare_quiet(f0, f1, NFS) == 1)
- ? float32_sub(f0, f1, NFS)
- : float32_sub(f1, f0, NFS));
+ float32 f0 = make_float32(a);
+ float32 f1 = make_float32(b);
+ return float32_val(float32_abs(float32_sub(f0, f1, NFS)));
}
uint32_t HELPER(neon_add_f32)(uint32_t a, uint32_t b)
{
- return vfp_stoi(float32_add(vfp_itos(a), vfp_itos(b), NFS));
+ return float32_val(float32_add(make_float32(a), make_float32(b), NFS));
}
uint32_t HELPER(neon_sub_f32)(uint32_t a, uint32_t b)
{
- return vfp_stoi(float32_sub(vfp_itos(a), vfp_itos(b), NFS));
+ return float32_val(float32_sub(make_float32(a), make_float32(b), NFS));
}
uint32_t HELPER(neon_mul_f32)(uint32_t a, uint32_t b)
{
- return vfp_stoi(float32_mul(vfp_itos(a), vfp_itos(b), NFS));
+ return float32_val(float32_mul(make_float32(a), make_float32(b), NFS));
}
/* Floating point comparisons produce an integer result. */
-#define NEON_VOP_FCMP(name, cmp) \
+#define NEON_VOP_FCMP(name, ok) \
uint32_t HELPER(neon_##name)(uint32_t a, uint32_t b) \
{ \
- if (float32_compare_quiet(vfp_itos(a), vfp_itos(b), NFS) cmp 0) \
- return ~0; \
- else \
- return 0; \
+ switch (float32_compare_quiet(make_float32(a), make_float32(b), NFS)) { \
+ ok return ~0; \
+ default: return 0; \
+ } \
}
-NEON_VOP_FCMP(ceq_f32, ==)
-NEON_VOP_FCMP(cge_f32, >=)
-NEON_VOP_FCMP(cgt_f32, >)
+NEON_VOP_FCMP(ceq_f32, case float_relation_equal:)
+NEON_VOP_FCMP(cge_f32, case float_relation_equal: case float_relation_greater:)
+NEON_VOP_FCMP(cgt_f32, case float_relation_greater:)
uint32_t HELPER(neon_acge_f32)(uint32_t a, uint32_t b)
{
- float32 f0 = float32_abs(vfp_itos(a));
- float32 f1 = float32_abs(vfp_itos(b));
- return (float32_compare_quiet(f0, f1,NFS) >= 0) ? ~0 : 0;
+ float32 f0 = float32_abs(make_float32(a));
+ float32 f1 = float32_abs(make_float32(b));
+ switch (float32_compare_quiet(f0, f1, NFS)) {
+ case float_relation_equal:
+ case float_relation_greater:
+ return ~0;
+ default:
+ return 0;
+ }
}
uint32_t HELPER(neon_acgt_f32)(uint32_t a, uint32_t b)
{
- float32 f0 = float32_abs(vfp_itos(a));
- float32 f1 = float32_abs(vfp_itos(b));
- return (float32_compare_quiet(f0, f1, NFS) > 0) ? ~0 : 0;
+ float32 f0 = float32_abs(make_float32(a));
+ float32 f1 = float32_abs(make_float32(b));
+ if (float32_compare_quiet(f0, f1, NFS) == float_relation_greater) {
+ return ~0;
+ }
+ return 0;
}
#define ELEM(V, N, SIZE) (((V) >> ((N) * (SIZE))) & ((1ull << (SIZE)) - 1))
-void HELPER(neon_qunzip8)(CPUState *env, uint32_t rd, uint32_t rm)
+void HELPER(neon_qunzip8)(uint32_t rd, uint32_t rm)
{
uint64_t zm0 = float64_val(env->vfp.regs[rm]);
uint64_t zm1 = float64_val(env->vfp.regs[rm + 1]);
@@ -1890,7 +1869,7 @@ void HELPER(neon_qunzip8)(CPUState *env, uint32_t rd, uint32_t rm)
env->vfp.regs[rd + 1] = make_float64(d1);
}
-void HELPER(neon_qunzip16)(CPUState *env, uint32_t rd, uint32_t rm)
+void HELPER(neon_qunzip16)(uint32_t rd, uint32_t rm)
{
uint64_t zm0 = float64_val(env->vfp.regs[rm]);
uint64_t zm1 = float64_val(env->vfp.regs[rm + 1]);
@@ -1910,7 +1889,7 @@ void HELPER(neon_qunzip16)(CPUState *env, uint32_t rd, uint32_t rm)
env->vfp.regs[rd + 1] = make_float64(d1);
}
-void HELPER(neon_qunzip32)(CPUState *env, uint32_t rd, uint32_t rm)
+void HELPER(neon_qunzip32)(uint32_t rd, uint32_t rm)
{
uint64_t zm0 = float64_val(env->vfp.regs[rm]);
uint64_t zm1 = float64_val(env->vfp.regs[rm + 1]);
@@ -1926,7 +1905,7 @@ void HELPER(neon_qunzip32)(CPUState *env, uint32_t rd, uint32_t rm)
env->vfp.regs[rd + 1] = make_float64(d1);
}
-void HELPER(neon_unzip8)(CPUState *env, uint32_t rd, uint32_t rm)
+void HELPER(neon_unzip8)(uint32_t rd, uint32_t rm)
{
uint64_t zm = float64_val(env->vfp.regs[rm]);
uint64_t zd = float64_val(env->vfp.regs[rd]);
@@ -1942,7 +1921,7 @@ void HELPER(neon_unzip8)(CPUState *env, uint32_t rd, uint32_t rm)
env->vfp.regs[rd] = make_float64(d0);
}
-void HELPER(neon_unzip16)(CPUState *env, uint32_t rd, uint32_t rm)
+void HELPER(neon_unzip16)(uint32_t rd, uint32_t rm)
{
uint64_t zm = float64_val(env->vfp.regs[rm]);
uint64_t zd = float64_val(env->vfp.regs[rd]);
@@ -1954,7 +1933,7 @@ void HELPER(neon_unzip16)(CPUState *env, uint32_t rd, uint32_t rm)
env->vfp.regs[rd] = make_float64(d0);
}
-void HELPER(neon_qzip8)(CPUState *env, uint32_t rd, uint32_t rm)
+void HELPER(neon_qzip8)(uint32_t rd, uint32_t rm)
{
uint64_t zm0 = float64_val(env->vfp.regs[rm]);
uint64_t zm1 = float64_val(env->vfp.regs[rm + 1]);
@@ -1982,7 +1961,7 @@ void HELPER(neon_qzip8)(CPUState *env, uint32_t rd, uint32_t rm)
env->vfp.regs[rd + 1] = make_float64(d1);
}
-void HELPER(neon_qzip16)(CPUState *env, uint32_t rd, uint32_t rm)
+void HELPER(neon_qzip16)(uint32_t rd, uint32_t rm)
{
uint64_t zm0 = float64_val(env->vfp.regs[rm]);
uint64_t zm1 = float64_val(env->vfp.regs[rm + 1]);
@@ -2002,7 +1981,7 @@ void HELPER(neon_qzip16)(CPUState *env, uint32_t rd, uint32_t rm)
env->vfp.regs[rd + 1] = make_float64(d1);
}
-void HELPER(neon_qzip32)(CPUState *env, uint32_t rd, uint32_t rm)
+void HELPER(neon_qzip32)(uint32_t rd, uint32_t rm)
{
uint64_t zm0 = float64_val(env->vfp.regs[rm]);
uint64_t zm1 = float64_val(env->vfp.regs[rm + 1]);
@@ -2018,7 +1997,7 @@ void HELPER(neon_qzip32)(CPUState *env, uint32_t rd, uint32_t rm)
env->vfp.regs[rd + 1] = make_float64(d1);
}
-void HELPER(neon_zip8)(CPUState *env, uint32_t rd, uint32_t rm)
+void HELPER(neon_zip8)(uint32_t rd, uint32_t rm)
{
uint64_t zm = float64_val(env->vfp.regs[rm]);
uint64_t zd = float64_val(env->vfp.regs[rd]);
@@ -2034,7 +2013,7 @@ void HELPER(neon_zip8)(CPUState *env, uint32_t rd, uint32_t rm)
env->vfp.regs[rd] = make_float64(d0);
}
-void HELPER(neon_zip16)(CPUState *env, uint32_t rd, uint32_t rm)
+void HELPER(neon_zip16)(uint32_t rd, uint32_t rm)
{
uint64_t zm = float64_val(env->vfp.regs[rm]);
uint64_t zd = float64_val(env->vfp.regs[rd]);
generated by cgit v1.2.3 (git 2.26.2) at 2024-12-19 13:29:00 +0000