diff options
Diffstat (limited to 'tests32/test17.c')
| -rw-r--r-- | tests32/test17.c | 611 |
1 files changed, 611 insertions, 0 deletions
diff --git a/tests32/test17.c b/tests32/test17.c new file mode 100644 index 00000000..70902fb6 --- /dev/null +++ b/tests32/test17.c @@ -0,0 +1,611 @@ +// build with gcc -O0 -g -msse -msse2 -mssse3 -msse4.1 test17.c -o test17 +// and -m32 for 32bits version +#include <inttypes.h> +#include <string.h> +#include <stdio.h> +#include <stddef.h> +#include <stdlib.h> +#include <stdint.h> +#include <math.h> +#include <pmmintrin.h> +#include <immintrin.h> + +typedef unsigned char u8x16 __attribute__ ((vector_size (16))); +typedef unsigned short u16x8 __attribute__ ((vector_size (16))); +typedef unsigned int u32x4 __attribute__ ((vector_size (16))); +typedef unsigned long long int u64x2 __attribute__ ((vector_size (16))); +typedef float f32x4 __attribute__ ((vector_size (16))); +typedef double d64x2 __attribute__ ((vector_size (16))); + +typedef union { + __m128i mm; + __m128 mf; + __m128d md; + u8x16 u8; + u16x8 u16; + u32x4 u32; + u64x2 u64; + f32x4 f32; + d64x2 d64; +} v128; + +uint64_t _ucomiss_(float a, float b) +{ + uint64_t ret = 0x202; + v128 va, vb; + va.f32[0] = a; + vb.f32[0] = b; + if(_mm_ucomigt_ss(va.mf, vb.mf)) + ret |= 0x000; + else if(_mm_ucomilt_ss(va.mf, vb.mf)) + ret |= 0x001; + else if(_mm_ucomieq_ss(va.mf, vb.mf)) + ret |= 0x040; + else + ret |= 0x045; + return ret; +} + +uint64_t _minss_(float a, float b) +{ + v128 va, vb, ret; + va.f32[0] = a; + vb.f32[0] = b; + ret.mf = _mm_min_ss(va.mf, vb.mf); + return ret.u64[0]; +} +uint64_t _maxss_(float a, float b) +{ + v128 va, vb, ret; + va.f32[0] = a; + vb.f32[0] = b; + ret.mf = _mm_max_ss(va.mf, vb.mf); + return ret.u64[0]; +} + +#define CMPSS(A, B) \ +uint64_t _cmpss_##A(float a, float b) \ +{ \ + v128 va, vb, ret; \ + va.f32[0] = a; \ + vb.f32[0] = b; \ + ret.mf = _mm_cmp##B##_ss(va.mf, vb.mf); \ + return ret.u64[0]; \ +} +CMPSS(0, eq) +CMPSS(1, lt) +CMPSS(2, le) +CMPSS(3, unord) +CMPSS(4, neq) +CMPSS(5, nlt) +CMPSS(6, nle) +CMPSS(7, ord) +#undef CMPSS + +const v128 a128_8 = {.u8 = { + 0xff, 0x80, 0x7f, 0x00, 0x01, 0x02, 0x03, 0x81, + 0xfe, 0x84, 0x72, 0x52, 0xa5, 0x00, 0xc0, 0x32 +}}; +const v128 a128_16 = {.u16 = { + 0xffff, 0x8000, 0x7fff, 0x0000, 0x0001, 0x0002, 0x0003, 0x8001 +}}; +const v128 a128_32 = {.u32 = { + 0xffffffff, 0x80000000, 0x7fffffff, 0x00000000 +}}; +const v128 a128_64 = {.u64 = { + 0xffffffffffffffffLL, 0x8000000000000000LL +}}; + +const v128 b128_8 = {.u8 = { + 0x00, 0x01, 0x05, 0x15, 0x20, 0x80, 0xff, 0x00, + 0x08, 0x07, 0x81, 0x06, 0x0a, 0x0f, 0x10, 0x01 +}}; +const v128 b128_16 = {.u16 = { + 0x8000, 0x7fff, 0xffff, 0xffff, 0x0050, 0x9000, 0xfffe, 0x8001 +}}; +const v128 b128_32 = {.u32 = { + 0x00000001, 0x80000000, 0x00000005, 0xfffffffe +}}; +const v128 b128_64 = {.u64 = { + 0x0000000000000001LL, 0x8000000000000000LL +}}; +const v128 c128_32 = {.u32 = { + 0x00000001, 0x80000000, 0x80000005, 0x0000fffe +}}; + +const v128 a128_pd = {.d64 = { 1.0, 2.0}}; +const v128 b128_pd = {.d64 = { 0.0, -2.0}}; +const v128 c128_pd = {.d64 = { INFINITY, -INFINITY}}; +const v128 d128_pd = {.d64 = { NAN, -0.0}}; +const v128 a128_ps = {.f32 = { 1.0, 2.0, 3.0, -4.0}}; +const v128 b128_ps = {.f32 = { 0.0, -2.0, -10.0, 0.5}}; +const v128 c128_ps = {.f32 = { INFINITY, -INFINITY, -INFINITY, 1.0}}; +const v128 d128_ps = {.f32 = { NAN, -0.0, -NAN, INFINITY}}; + +v128 reverse_pd(v128 a) { + v128 ret; + ret.md = _mm_shuffle_pd(a.md, a.md, 1); + return ret; +} + +void print_8(v128 v) { + for(int i=0; i<16; ++i) + printf("0x%x ", v.u8[i]); +} +void print_16(v128 v) { + for(int i=0; i<8; ++i) + printf("0x%x ", v.u16[i]); +} +void print_32(v128 v) { + for(int i=0; i<4; ++i) + printf("0x%x ", v.u32[i]); +} +void print_64(v128 v) { + for(int i=0; i<2; ++i) + printf("0x%"PRIx64" ", v.u64[i]); +} +#define print_128 print_64 +void print_ps(v128 v) { + for(int i=0; i<4; ++i) + if(isnanf(v.f32[i])) + printf("nan "); + else + printf("%g ", v.f32[i]); +} +void print_pd(v128 v) { + for(int i=0; i<2; ++i) + if(isnan(v.d64[i])) + printf("0x%"PRIx64" ", v.u64[i]); + else + printf("%g ", v.d64[i]); +} +#define print_sd print_pd + +int main(int argc, const char** argv) +{ + float a, b; + uint32_t flags; + uint32_t maxf = 0x7f7fffff; + uint32_t minf = 0xff7fffff; + uint32_t r; + +#define GO1(A, N) \ +a = 1.0f; b = 2.0f; \ +flags = A(a, b); \ +printf(N " %f, %f => 0x%"PRIx32"\n", a, b, flags); \ +flags = A(b, a); \ +printf(N " %f, %f => 0x%"PRIx32"\n", b, a, flags); \ +b = INFINITY; \ +flags = A(a, b); \ +printf(N " %f, %f => 0x%"PRIx32"\n", a, b, flags); \ +flags = A(b, a); \ +printf(N " %f, %f => 0x%"PRIx32"\n", b, a, flags); \ +b = -INFINITY; \ +flags = A(a, b); \ +printf(N " %f, %f => 0x%"PRIx32"\n", a, b, flags); \ +flags = A(b, a); \ +printf(N " %f, %f => 0x%"PRIx32"\n", b, a, flags); \ +b = NAN; \ +flags = A(a, b); \ +printf(N " %f, %f => 0x%"PRIx32"\n", a, b, flags); \ +flags = A(b, a); \ +printf(N " %f, %f => 0x%"PRIx32"\n", b, a, flags); \ +b = a; \ +flags = A(a, b); \ +printf(N " %f, %f => 0x%"PRIx32"\n", a, b, flags); \ +flags = A(b, a); \ +printf(N " %f, %f => 0x%"PRIx32"\n", b, a, flags); \ +a = b = INFINITY; \ +flags = A(a, b); \ +printf(N " %f, %f => 0x%"PRIx32"\n", a, b, flags); \ +a = -INFINITY; \ +flags = A(a, b); \ +printf(N " %f, %f => 0x%"PRIx32"\n", a, b, flags); \ +flags = A(b, a); \ +printf(N " %f, %f => 0x%"PRIx32"\n", b, a, flags); \ +a = b = NAN; \ +flags = A(a, b); \ +printf(N " %f, %f => 0x%"PRIx32"\n", a, b, flags); + +#define GO2(A, N) \ +a = 1.0f; b = 2.0f; \ +r = A(a, b); \ +printf(N " %g, %g => %g\n", a, b, *(float*)&r); \ +r = A(b, a); \ +printf(N " %g, %g => %g\n", b, a, *(float*)&r); \ +a = -INFINITY; \ +r = A(a, b); \ +printf(N " %g, %g => %g\n", a, b, *(float*)&r); \ +r = A(b, a); \ +printf(N " %g, %g => %g\n", b, a, *(float*)&r); \ +a = +INFINITY; \ +r = A(a, b); \ +printf(N " %g, %g => %g\n", a, b, *(float*)&r); \ +r = A(b, a); \ +printf(N " %g, %g => %g\n", b, a, *(float*)&r); \ +a = NAN; \ +r = A(a, b); \ +printf(N " %g, %g => %g\n", a, b, *(float*)&r); \ +r = A(b, a); \ +printf(N " %g, %g => %g\n", b, a, *(float*)&r); \ +b = *(float*)&maxf; \ +r = A(a, b); \ +printf(N " %g, %g => %g\n", a, b, *(float*)&r); \ +r = A(b, a); \ +printf(N " %g, %g => %g\n", b, a, *(float*)&r); \ +a = -INFINITY; \ +r = A(a, b); \ +printf(N " %g, %g => %g\n", a, b, *(float*)&r); \ +r = A(b, a); \ +printf(N " %g, %g => %g\n", b, a, *(float*)&r); \ +a = +INFINITY; \ +r = A(a, b); \ +printf(N " %g, %g => %g\n", a, b, *(float*)&r); \ +r = A(b, a); \ +printf(N " %g, %g => %g\n", b, a, *(float*)&r); + + GO1(_ucomiss_, "ucomiss") + GO2(_minss_, "minss") + GO2(_maxss_, "maxss") + GO1(_cmpss_0, "cmpss 0") + GO1(_cmpss_1, "cmpss 1") + GO1(_cmpss_2, "cmpss 2") + GO1(_cmpss_3, "cmpss 3") + GO1(_cmpss_4, "cmpss 4") + GO1(_cmpss_5, "cmpss 5") + GO1(_cmpss_6, "cmpss 6") + GO1(_cmpss_7, "cmpss 7") + + #undef GO1 + #undef GO2 + v128 a128; + int i; + + #define GO1(A, N, C) \ + a128.mm = _mm_##A##_epi##N(a128_##N.mm); \ + printf("%s(", #C); print_##N(a128_##N); \ + printf(") = "); print_##N(a128); printf("\n"); + #define GO1C(A, N, C, A1, I) \ + a128.mm = _mm_##A##_epi##N(A1.mm, I); \ + printf("%s(", #C); print_##N(A1); \ + printf("%d) = ", I); print_##N(a128); printf("\n"); + #define GO2(A, N, C, A1, A2) \ + a128.mm = _mm_##A##_epi##N(A1.mm, A2.mm); \ + printf("%s(", #C); print_##N(A1); \ + printf(", "); print_##N(A2); \ + printf(") = "); print_##N(a128); printf("\n"); + #define GO2u(A, N, C, A1, A2) \ + a128.mm = _mm_##A##_epu##N(A1.mm, A2.mm); \ + printf("%s(", #C); print_##N(A1); \ + printf(", "); print_##N(A2); \ + printf(") = "); print_##N(a128); printf("\n"); + #define GO2f(A, C, A1, A2) \ + a128.mm = _mm_##A##_si128(A1.mm, A2.mm); \ + printf("%s(", #C); print_128(A1); \ + printf(", "); print_128(A2); \ + printf(") = "); print_128(a128); printf("\n"); + #define GO2C(A, N, C, A1, A2, I) \ + a128.mm = _mm_##A##_epi##N(A1.mm, A2.mm, I); \ + printf("%s(", #C); print_##N(A1); \ + printf(", "); print_##N(A2); \ + printf("%d) = ", I); print_##N(a128); printf("\n"); + #define GO2i(A, A1, A2) \ + i = _mm_##A##_si128(A1.mm, A2.mm); \ + printf("p%s(", #A); print_64(A1); \ + printf(", "); print_64(A2); \ + printf(") = %d\n", i); + #define GO3PS(A, N, A1, A2, A3) \ + a128.mf = _mm_##A##_ps(A1.mf, A2.mf, A3.mf); \ + printf("p%s%s(", #A, "ps"); print_##N(A1); \ + printf(", "); print_##N(A2); \ + printf(", "); print_##N(A3); \ + printf(") = "); print_##N(a128); printf("\n"); + #define GO1ipd(A, C, A1) \ + i = _mm_##A##_pd(A1.md); \ + printf("%s(", #C); print_64(A1); \ + printf(") = 0x%x\n", i); + #define GO1pd(A, C, A1) \ + a128.md = _mm_##A##_pd(A1.md); \ + printf("%s(", #C); print_pd(A1); \ + printf(") = "); print_pd(a128); printf("\n"); + #define GO2pd(A, C, A1, A2) \ + a128.md = _mm_##A##_pd(A1.md, A2.md); \ + printf("%s(", #C); print_pd(A1); \ + printf(", "); print_pd(A2); \ + printf(") = "); print_pd(a128); printf("\n"); + #define GO2Cpd(A, C, A1, A2, I) \ + a128.md = _mm_##A##_pd(A1.md, A2.md, I); \ + printf("%s(", #C); print_pd(A1); \ + printf(", "); print_pd(A2); \ + printf(", %d) = ", I); print_pd(a128); printf("\n"); + #define GO1isd(A, C, A1) \ + i = _mm_##A##_sd(A1.md); \ + printf("%s(", #C); print_64(A1); \ + printf(") = 0x%x\n", i); + #define GO1sd(A, C, A1) \ + a128.md = _mm_##A##_sd(A1.md); \ + printf("%s(", #C); print_sd(A1); \ + printf(") = "); print_sd(a128); printf("\n"); + #define GO2sd(A, C, A1, A2) \ + a128.md = _mm_##A##_sd(A1.md, A2.md); \ + printf("%s(", #C); print_sd(A1); \ + printf(", "); print_sd(A2); \ + printf(") = "); print_sd(a128); printf("\n"); + #define GO2Csd(A, C, A1, A2, I) \ + a128.md = _mm_##A##_sd(A1.md, A2.md, I); \ + printf("%s(", #C); print_sd(A1); \ + printf(", "); print_sd(A2); \ + printf(", %d) = ", I); print_sd(a128); printf("\n"); + #define GO1ips(A, C, A1) \ + i = _mm_##A##_ps(A1.mf); \ + printf("%s(", #C); print_32(A1); \ + printf(") = 0x%x\n", i); + #define GO1ps(A, C, A1) \ + a128.mf = _mm_##A##_ps(A1.mf); \ + printf("%s(", #C); print_ps(A1); \ + printf(") = "); print_ps(a128); printf("\n"); + #define GO2ps(A, C, A1, A2) \ + a128.mf = _mm_##A##_ps(A1.mf, A2.mf); \ + printf("%s(", #C); print_ps(A1); \ + printf(", "); print_ps(A2); \ + printf(") = "); print_ps(a128); printf("\n"); + #define GO2Cps(A, C, A1, A2, I) \ + a128.mf = _mm_##A##_ps(A1.mf, A2.mf, I); \ + printf("%s(", #C); print_ps(A1); \ + printf(", "); print_ps(A2); \ + printf(", %d) = ", I); print_ps(a128); printf("\n"); + #define GO1ps2dq(A, C, A1) \ + a128.mm = _mm_##A##_epi32(A1.mf); \ + printf("%s(", #C); print_ps(A1); \ + printf(") = "); print_32(a128); printf("\n"); + + #define MULITGO2pd(A, B) \ + GO2pd(A, B, a128_pd, b128_pd) \ + GO2pd(A, B, b128_pd, c128_pd) \ + GO2pd(A, B, a128_pd, d128_pd) \ + GO2pd(A, B, b128_pd, d128_pd) \ + GO2pd(A, B, c128_pd, d128_pd) \ + GO2pd(A, B, d128_pd, d128_pd) + + #define MULITGO2Cpd(A, B, I) \ + GO2Cpd(A, B, a128_pd, b128_pd, I) \ + GO2Cpd(A, B, b128_pd, c128_pd, I) \ + GO2Cpd(A, B, a128_pd, d128_pd, I) \ + GO2Cpd(A, B, b128_pd, d128_pd, I) \ + GO2Cpd(A, B, c128_pd, d128_pd, I) \ + GO2Cpd(A, B, d128_pd, d128_pd, I) + + #define MULITGO2ps(A, B) \ + GO2ps(A, B, a128_ps, b128_ps) \ + GO2ps(A, B, b128_ps, c128_ps) \ + GO2ps(A, B, a128_ps, d128_ps) \ + GO2ps(A, B, b128_ps, d128_ps) \ + GO2ps(A, B, c128_ps, d128_ps) \ + GO2ps(A, B, d128_ps, d128_ps) + + #define MULTIGO1ps2dq(A, B) \ + GO1ps2dq(A, B, a128_ps) \ + GO1ps2dq(A, B, b128_ps) \ + GO1ps2dq(A, B, c128_ps) \ + GO1ps2dq(A, B, d128_ps) + + #define MULITGO2Cps(A, B, I) \ + GO2Cps(A, B, a128_ps, b128_ps, I) \ + GO2Cps(A, B, b128_ps, c128_ps, I) \ + GO2Cps(A, B, a128_ps, d128_ps, I) \ + GO2Cps(A, B, b128_ps, d128_ps, I) \ + GO2Cps(A, B, c128_ps, d128_ps, I) \ + GO2Cps(A, B, d128_ps, d128_ps, I) + + #define MULTIGO2sd(A, B) \ + GO2sd(A, B, a128_pd, a128_pd) \ + GO2sd(A, B, a128_pd, b128_pd) \ + GO2sd(A, B, a128_pd, c128_pd) \ + GO2sd(A, B, a128_pd, d128_pd) \ + GO2sd(A, B, b128_pd, d128_pd) \ + GO2sd(A, B, c128_pd, d128_pd) \ + GO2sd(A, B, a128_pd, reverse_pd(a128_pd)) \ + GO2sd(A, B, a128_pd, reverse_pd(b128_pd)) \ + GO2sd(A, B, a128_pd, reverse_pd(c128_pd)) \ + GO2sd(A, B, a128_pd, reverse_pd(d128_pd)) \ + GO2sd(A, B, b128_pd, reverse_pd(d128_pd)) \ + GO2sd(A, B, b128_pd, reverse_pd(d128_pd)) + + + GO2(shuffle, 8, pshufb, a128_8, b128_8) + GO2(hadd, 16, phaddw, a128_16, b128_16) + GO2(hadd, 32, phaddd, a128_32, b128_32) + GO2(hadds, 16, phaddsw, a128_16, b128_16) + GO2(maddubs, 16, pmaddubsw, a128_8, b128_8) + GO2(hsub, 16, phsubw, a128_16, b128_16) + GO2(sign, 8, psignb, a128_8, b128_8) + GO2(sign, 16, psignw, a128_16, b128_16) + GO2(sign, 32, psignd, a128_32, b128_32) + GO2(mulhrs, 16, pmulhrsw, a128_16, b128_16) + GO3PS(blendv, 32, a128_32, b128_32, c128_32) + GO2i(testz, a128_32, b128_32) + GO2i(testc, a128_32, b128_32) + GO2i(testnzc, a128_32, b128_32) + GO1(abs, 8, pabsb) + GO1(abs, 16, pabsw) + GO1(abs, 32, pabsd) + GO1(cvtepi8, 16, pmovsxbw); + GO1(cvtepi8, 32, pmovsxbd); + GO1(cvtepi8, 64, pmovsxbq); + GO1(cvtepi16, 32, pmovsxwd); + GO1(cvtepi16, 64, pmovsxwq); + GO1(cvtepi32, 64, pmovsxdq); + GO1(cvtepu8, 16, pmovzxbw); + GO1(cvtepu8, 32, pmovzxbd); + GO1(cvtepu8, 64, pmovzxbq); + GO1(cvtepu16, 32, pmovzxwd); + GO1(cvtepu16, 64, pmovzxwq); + GO1(cvtepu32, 64, pmovzxdq); + GO2(min, 32, pminsd, a128_32, b128_32) + GO2(max, 32, pmaxsd, a128_32, b128_32) + GO2C(blend, 16, pblendw, a128_16, b128_16, 0) + GO2C(blend, 16, pblendw, a128_16, b128_16, 0xff) + GO2C(blend, 16, pblendw, a128_16, b128_16, 0xaa) + GO2C(blend, 16, pblendw, a128_16, b128_16, 2) + GO2C(alignr, 8, palignr, a128_8, b128_8, 0) + GO2C(alignr, 8, palignr, a128_8, b128_8, 2) + GO2C(alignr, 8, palignr, a128_8, b128_8, 7) + GO2C(alignr, 8, palignr, a128_8, b128_8, 15) + GO2C(alignr, 8, palignr, a128_8, b128_8, 16) + GO2C(alignr, 8, palignr, a128_8, b128_8, 0xff) + GO1ipd(movemask, movmskpd, a128_64) + GO1pd(sqrt, psqrtpd, a128_pd) + GO1pd(sqrt, psqrtpd, b128_pd) + GO1pd(sqrt, psqrtpd, c128_pd) + GO1pd(sqrt, psqrtpd, d128_pd) + MULITGO2pd(and, andpd) + MULITGO2pd(andnot, andnpd) + MULITGO2pd(or, orpd) + MULITGO2pd(xor, xorpd) + MULITGO2pd(add, addpd) + MULITGO2pd(mul, mulpd) + MULITGO2pd(sub, subpd) + MULITGO2pd(min, minpd) + MULITGO2pd(div, divpd) + MULITGO2pd(max, maxpd) + GO2(unpacklo, 8, punpcklbw, a128_8, b128_8) + GO2(unpacklo, 16, punpcklwd, a128_16, b128_16) + GO2(unpacklo, 32, punpckldq, a128_32, b128_32) + GO2(packs, 16, ppacksswb, a128_16, b128_16) + GO2(cmpgt, 8, pcmpgtb, a128_8, b128_8) + GO2(cmpgt, 16, pcmpgtw, a128_16, b128_16) + GO2(cmpgt, 32, pcmpgtd, a128_32, b128_32) + GO2(packus, 16, packuswb, a128_16, b128_16) + GO2(unpackhi, 8, punpckhbw, a128_8, b128_8) + GO2(unpackhi, 16, punpckhwd, a128_16, b128_16) + GO2(unpackhi, 32, punpckhdq, a128_32, b128_32) + GO2(packs, 32, ppackssdw, a128_32, b128_32) + GO2(unpacklo, 64, punpcklqdq, a128_64, b128_64) + GO2(unpackhi, 64, punpckhqdq, a128_64, b128_64) + GO1C(shuffle, 32, pshufd, a128_32, 0) + GO1C(shuffle, 32, pshufd, a128_32, 0xff) + GO1C(shuffle, 32, pshufd, a128_32, 0xaa) + GO1C(shuffle, 32, pshufd, a128_32, 2) + GO1C(srli, 16, psrlw, a128_16, 0) + GO1C(srli, 16, psrlw, a128_16, 0xff) + GO1C(srli, 16, psrlw, a128_16, 0xaa) + GO1C(srli, 16, psrlw, a128_16, 2) + GO1C(srli, 32, psrld, a128_32, 0) + GO1C(srli, 32, psrld, a128_32, 0xff) + GO1C(srli, 32, psrld, a128_32, 0xaa) + GO1C(srli, 32, psrld, a128_32, 2) + GO1C(srli, 64, psrlq, a128_64, 0) + GO1C(srli, 64, psrlq, a128_64, 0xff) + GO1C(srli, 64, psrlq, a128_64, 0xaa) + GO1C(srli, 64, psrlq, a128_64, 2) + GO1C(srai, 16, psraw, a128_16, 0) + GO1C(srai, 16, psraw, a128_16, 0xff) + GO1C(srai, 16, psraw, a128_16, 0xaa) + GO1C(srai, 16, psraw, a128_16, 2) + GO1C(srai, 32, psrad, a128_32, 0) + GO1C(srai, 32, psrad, a128_32, 0xff) + GO1C(srai, 32, psrad, a128_32, 0xaa) + GO1C(srai, 32, psrad, a128_32, 2) + GO1C(slli, 16, psllw, a128_16, 0) + GO1C(slli, 16, psllw, a128_16, 0xff) + GO1C(slli, 16, psllw, a128_16, 0xaa) + GO1C(slli, 16, psllw, a128_16, 2) + GO1C(slli, 32, pslld, a128_32, 0) + GO1C(slli, 32, pslld, a128_32, 0xff) + GO1C(slli, 32, pslld, a128_32, 0xaa) + GO1C(slli, 32, pslld, a128_32, 2) + GO1C(slli, 64, psllq, a128_64, 0) + GO1C(slli, 64, psllq, a128_64, 0xff) + GO1C(slli, 64, psllq, a128_64, 0xaa) + GO1C(slli, 64, psllq, a128_64, 2) + GO2(cmpeq, 8, pcmpeqb, a128_8, b128_8) + GO2(cmpeq, 16, pcmpeqw, a128_16, b128_16) + GO2(cmpeq, 32, pcmpeqd, a128_32, b128_32) + MULITGO2pd(hadd, haddpd) + GO2(srl, 16, psrlw, a128_16, b128_16) + GO2(srl, 32, psrld, a128_32, b128_32) + GO2(srl, 64, psrlq, a128_64, b128_64) + GO2(add, 64, paddq, a128_64, b128_64) + GO2(mullo, 16, pmullw, a128_16, b128_16) + GO2u(subs, 8, psubusb, a128_8, b128_8) + GO2u(subs, 16, psubusw, a128_16, b128_16) + GO2u(min, 8, pminub, a128_8, b128_8) + GO2f(and, pand, a128_8, b128_8) + GO2u(adds, 8, paddusb, a128_8, b128_8) + GO2u(adds, 16, paddusw, a128_16, b128_16) + GO2u(max, 8, pmaxub, a128_8, b128_8) + GO2f(andnot, pandn, a128_8, b128_8) + GO2u(avg, 8, pavgb, a128_8, b128_8) + GO2(sra, 16, psraw, a128_16, b128_16) + GO2(sra, 32, psrad, a128_32, b128_32) + GO2u(avg, 16, pavgb, a128_16, b128_16) + GO2u(mulhi, 16, pmulhuw, a128_16, b128_16) + GO2(mulhi, 16, pmulhw, a128_16, b128_16) + GO2(subs, 8, psubsb, a128_8, b128_8) + GO2(subs, 16, psubsw, a128_16, b128_16) + GO2(min, 16, pminsw, a128_16, b128_16) + GO2f(or, por, a128_8, b128_8) + GO2(adds, 8, paddusb, a128_8, b128_8) + GO2(adds, 16, paddusw, a128_16, b128_16) + GO2(max, 16, pmaxsw, a128_16, b128_16) + GO2f(xor, pxor, a128_8, b128_8) + GO2(sll, 16, psllw, a128_16, b128_16) + GO2(sll, 32, pslld, a128_32, b128_32) + GO2(sll, 64, psllq, a128_64, b128_64) + GO2u(mul, 32, pmuludq, a128_32, b128_32) + GO2(madd, 16, pmaddwd, a128_16, b128_16) + GO2u(sad, 8, psadbw, a128_8, b128_8) + GO2(sub, 8, psubb, a128_8, b128_8) + GO2(sub, 16, psubw, a128_16, b128_16) + GO2(sub, 32, psubd, a128_32, b128_32) + GO2(sub, 64, psubq, a128_64, b128_64) + GO2(add, 8, paddb, a128_8, b128_8) + GO2(add, 16, paddw, a128_16, b128_16) + GO2(add, 32, paddd, a128_32, b128_32) + GO2ps(movehl, pmovhlps, a128_ps, b128_ps) + GO2ps(unpacklo, unpcklps, a128_ps, b128_ps) + GO2ps(unpackhi, unpckhps, a128_ps, b128_ps) + GO2ps(movelh, pmovhps, a128_ps, b128_ps) + GO1ps(sqrt, psqrtps, a128_ps) + GO1ps(sqrt, psqrtps, b128_ps) + GO1ps(sqrt, psqrtps, c128_ps) + GO1ps(sqrt, psqrtps, d128_ps) + //GO1ps(rsqrt, prsqrtps, a128_ps) // difference in precision + //GO1ps(rsqrt, prsqrtps, b128_ps) // same + //GO1ps(rsqrt, prsqrtps, c128_ps) // same + //GO1ps(rsqrt, prsqrtps, d128_ps) // difference in the handling of NAN, (-)0, and INF in Dynarec + //GO1ps(rcp, prcpps, a128_ps) // deference in precision + //GO1ps(rcp, prcpps, b128_ps) // deference in precision + //GO1ps(rcp, prcpps, c128_ps) // deference in precision + GO1ps(rcp, prcpps, d128_ps) + MULITGO2ps(and, andps) + MULITGO2ps(andnot, andnps) + MULITGO2ps(or, orps) + MULITGO2ps(xor, xorps) + MULITGO2ps(add, addps) + MULITGO2ps(mul, mulps) + MULITGO2ps(sub, subps) + MULITGO2ps(min, minps) + MULITGO2ps(div, divps) + MULITGO2ps(max, maxps) + //MULITGO2Cps(cmp, cmpps, 0) // use avx for some reason + MULITGO2Cps(shuffle, shufps, 0) + MULITGO2Cps(shuffle, shufps, 0x15) + MULITGO2Cps(shuffle, shufps, 0xff) + MULITGO2Cps(shuffle, shufps, 0x02) + MULTIGO2sd(sqrt, sqrtsd) + MULTIGO2sd(add, addsd) + MULTIGO2sd(mul, mulsd) + MULTIGO2sd(sub, subsd) + MULTIGO2sd(min, minsd) + MULTIGO2sd(div, divsd) + MULTIGO2sd(max, maxsd) + MULTIGO1ps2dq(cvtps, cvtps2dq) + MULITGO2Cps(dp, dpps, 0xff) + MULITGO2Cps(dp, dpps, 0x3f) + MULITGO2Cps(dp, dpps, 0xf3) + MULITGO2Cps(dp, dpps, 0x53) + + return 0; +} + |