Home
last modified time | relevance | path

Searched refs:sums (Results 1 – 25 of 193) sorted by relevance

12345678

/external/libvpx/libvpx/vpx_dsp/x86/
Dsad4d_avx2.c14 static INLINE void calc_final_4(const __m256i *const sums /*[4]*/, in calc_final_4() argument
16 const __m256i t0 = _mm256_hadd_epi32(sums[0], sums[1]); in calc_final_4()
17 const __m256i t1 = _mm256_hadd_epi32(sums[2], sums[3]); in calc_final_4()
29 __m256i sums[4]; in vpx_sad32x32x4d_avx2() local
35 sums[0] = _mm256_setzero_si256(); in vpx_sad32x32x4d_avx2()
36 sums[1] = _mm256_setzero_si256(); in vpx_sad32x32x4d_avx2()
37 sums[2] = _mm256_setzero_si256(); in vpx_sad32x32x4d_avx2()
38 sums[3] = _mm256_setzero_si256(); in vpx_sad32x32x4d_avx2()
57 sums[0] = _mm256_add_epi32(sums[0], r[0]); in vpx_sad32x32x4d_avx2()
58 sums[1] = _mm256_add_epi32(sums[1], r[1]); in vpx_sad32x32x4d_avx2()
[all …]
/external/fec/
Dsumsq_av.c16 vector unsigned int sums,carries,s1,s2; in sumsq_av() local
20 carries = sums = (vector unsigned int)(0); in sumsq_av()
28 sums = (vector unsigned int)vec_msum(x,x,(vector signed int)(0)); in sumsq_av()
40 carries = vec_add(carries,vec_addc(sums,s1)); in sumsq_av()
41 sums = vec_add(sums,s1); in sumsq_av()
51 carries = vec_add(carries,vec_addc(sums,s1)); in sumsq_av()
52 sums = vec_add(sums,s1); in sumsq_av()
56 s1 = vec_sro(sums,s.cv); in sumsq_av()
58 carries = vec_add(carries,vec_addc(sums,s1)); in sumsq_av()
59 sums = vec_add(sums,s1); in sumsq_av()
[all …]
/external/apache-commons-math/src/main/java/org/apache/commons/math/stat/descriptive/moment/
DVectorialCovariance.java37 private final double[] sums; field in VectorialCovariance
54 sums = new double[dimension]; in VectorialCovariance()
66 if (v.length != sums.length) { in increment()
67 throw new DimensionMismatchException(v.length, sums.length); in increment()
71 sums[i] += v[i]; in increment()
85 int dimension = sums.length; in getResult()
93 double e = c * (n * productsSums[k++] - sums[i] * sums[j]); in getResult()
117 Arrays.fill(sums, 0.0); in clear()
129 result = prime * result + Arrays.hashCode(sums); in hashCode()
147 if (!Arrays.equals(sums, other.sums)) in equals()
/external/libopus/celt/x86/
Dvq_sse2.c53 __m128 sums; in op_pvq_search_sse2() local
67 sums = _mm_setzero_ps(); in op_pvq_search_sse2()
75 sums = _mm_add_ps(sums, x4); in op_pvq_search_sse2()
82 sums = _mm_add_ps(sums, _mm_shuffle_ps(sums, sums, _MM_SHUFFLE(1, 0, 3, 2))); in op_pvq_search_sse2()
83 sums = _mm_add_ps(sums, _mm_shuffle_ps(sums, sums, _MM_SHUFFLE(2, 3, 0, 1))); in op_pvq_search_sse2()
95 opus_val32 sum = _mm_cvtss_f32(sums); in op_pvq_search_sse2()
105 sums = _mm_set_ps1(1.f); in op_pvq_search_sse2()
108 rcp4 = _mm_mul_ps(_mm_set_ps1((float)(K+.8)), _mm_rcp_ps(sums)); in op_pvq_search_sse2()
Dcelt_lpc_sse4_1.c65 opus_val32 sums[4] = {0}; in celt_fir_sse4_1() local
68 xcorr_kernel(rnum, x+i-ord, sums, ord, arch); in celt_fir_sse4_1()
70 vecSum = _mm_loadu_si128((__m128i *)sums); in celt_fir_sse4_1()
/external/llvm-project/polly/test/ScopInfo/
Dreduction_escaping_intermediate_2.ll3 ; void f(int N, int * restrict sums, int * restrict escape) {
7 ; sums[i] += 5;
8 ; escape[N-j] = escape[i] + sums[i-1];
14 ; CHECK: sums
16 ; CHECK: sums
20 ; CHECK: sums
25 define void @f(i32 %N, i32* noalias %sums, i32* noalias %escape) {
43 %arrayidx = getelementptr inbounds i32, i32* %sums, i32 %i.0
50 %arrayidx5 = getelementptr inbounds i32, i32* %sums, i32 %sub
Dreduction_escaping_intermediate.ll3 ; void f(int N, int * restrict sums, int * restrict escape) {
7 ; sums[i] += 5;
8 ; escape[N - i + j] = sums[i];
14 ; CHECK: sums
16 ; CHECK: sums
21 define void @f(i32 %N, i32* noalias %sums, i32* noalias %escape) {
39 %arrayidx = getelementptr inbounds i32, i32* %sums, i32 %i.0
Dreduction_invalid_overlapping_accesses.ll3 ; void f(int *sums) {
7 ; sums[i] += 5;
8 ; sums[i+10] *= 5;
17 define void @f(i32* %sums) {
35 %arrayidx = getelementptr inbounds i32, i32* %sums, i32 %i.0
40 %arrayidx5 = getelementptr inbounds i32, i32* %sums, i32 %add4
Dreduction_non_overlapping_chains.ll8 ; void f(int *sums) {
11 ; sums[i] += 5;
12 ; sums[i+1024] *= 5;
18 define void @f(i32* %sums) {
36 %arrayidx = getelementptr inbounds i32, i32* %sums, i32 %i.0
41 %arrayidx5 = getelementptr inbounds i32, i32* %sums, i32 %add4
/external/tensorflow/tensorflow/compiler/xla/service/
Dhlo_graph_dumper_test.cc51 std::vector<HloInstruction*> sums; in TEST_F() local
52 sums.push_back(b.AddInstruction(HloInstruction::CreateBinary( in TEST_F()
55 sums.push_back(b.AddInstruction(HloInstruction::CreateBinary( in TEST_F()
56 shape, HloOpcode::kAdd, sums[i], params[i + 2]))); in TEST_F()
65 {sums[3], sums[2], sums[1], sums[0]}, HloInstruction::FusionKind::kLoop); in TEST_F()
/external/boringssl/src/third_party/wycheproof_testvectors/
Dchacha20_poly1305_test.txt2001 # edge case intermediate sums in poly1305.
2012 # edge case intermediate sums in poly1305.
2023 # edge case intermediate sums in poly1305.
2034 # edge case intermediate sums in poly1305.
2045 # edge case intermediate sums in poly1305.
2056 # edge case intermediate sums in poly1305.
2067 # edge case intermediate sums in poly1305.
2078 # edge case intermediate sums in poly1305.
2089 # edge case intermediate sums in poly1305.
2100 # edge case intermediate sums in poly1305.
[all …]
Dxchacha20_poly1305_test.txt1864 # edge case intermediate sums in poly1305
1874 # edge case intermediate sums in poly1305
1884 # edge case intermediate sums in poly1305
1894 # edge case intermediate sums in poly1305
1904 # edge case intermediate sums in poly1305
1914 # edge case intermediate sums in poly1305
1924 # edge case intermediate sums in poly1305
1934 # edge case intermediate sums in poly1305
1944 # edge case intermediate sums in poly1305
1954 # edge case intermediate sums in poly1305
[all …]
/external/ruy/ruy/
Dpack_x86.h83 std::int32_t* sums = packed_matrix->sums;
89 std::int32_t* sums_ptr = sums ? sums + block_col : nullptr;
130 std::int32_t* sums = packed_matrix->sums;
136 std::int32_t* sums_ptr = sums ? sums + block_col : nullptr;
251 std::int32_t* sums = packed_matrix->sums;
257 std::int32_t* sums_ptr = sums ? sums + block_col : nullptr;
312 int input_xor, std::int32_t* sums);
324 std::int32_t* sums = packed_matrix->sums;
325 std::memset(sums + start_col, 0, sizeof(sums[0]) * (end_col - start_col));
338 src_matrix.layout.rows, kInputXor, sums);
[all …]
Dpack_arm.h157 std::int32_t* sums = packed_matrix->sums;
190 std::int32_t* sums_ptr = sums ? sums + block_col : nullptr;
240 std::int32_t* sums = packed_matrix->sums;
261 std::int32_t* sums_ptr = sums ? sums + block_col : nullptr;
289 std::int32_t* sums = packed_matrix->sums;
324 std::int32_t* sums_ptr = sums ? sums + block_col : nullptr;
512 std::int32_t* sums = packed_matrix->sums;
513 std::memset(sums + start_col, 0, sizeof(sums[0]) * (end_col - start_col));
564 std::int32_t* sums_ptr = sums + start_col;
589 std::int32_t* sums = packed_matrix->sums;
[all …]
Dpack.h114 SumsType* sums = packed_matrix->sums; in Run() local
127 if (sums) { in Run()
128 sums[col] = accum; in Run()
Dpack_avx2_fma.cc514 __m256i sums =
528 sums = _mm256_add_epi32(sums, sums_adjustment_v);
529 sums = _mm256_add_epi32(sums, sums_2x4_32bit_a);
530 sums = _mm256_add_epi32(sums, sums_2x4_32bit_b);
532 _mm256_storeu_si256(reinterpret_cast<__m256i*>(sums_ptr), sums);
610 int input_xor, std::int32_t* sums) {
651 _mm256_loadu_si256(reinterpret_cast<const __m256i*>(sums + col));
653 _mm256_storeu_si256(reinterpret_cast<__m256i*>(sums + col), sum);
676 if (sums) {
677 sums[col] += accum;
Dpack_avx512.cc440 __m256i sums =
449 sums = _mm256_add_epi32(sums, sums_adjustment_v);
450 sums = _mm256_add_epi32(sums, _mm512_castsi512_si256(sums_2x8_32bit));
451 sums = _mm256_add_epi32(sums, _mm512_extracti32x8_epi32(sums_2x8_32bit, 1));
453 _mm256_storeu_si256(reinterpret_cast<__m256i*>(sums_ptr), sums);
729 int input_xor, std::int32_t* sums) {
778 _mm512_loadu_si512(reinterpret_cast<const __m512i*>(sums + col));
780 _mm512_storeu_si512(reinterpret_cast<__m512i*>(sums + col), sum);
815 if (sums) {
816 sums[col] += accum;
Dpack_avx.cc656 __m256i sums =
667 sums = mm256_add_epi32(sums, sums_adjustment_v);
668 sums = mm256_add_epi32(sums, sums_2x4_32bit_a);
669 sums = mm256_add_epi32(sums, sums_2x4_32bit_b);
671 _mm256_storeu_si256(reinterpret_cast<__m256i*>(sums_ptr), sums);
752 int input_xor, std::int32_t* sums) {
793 _mm256_loadu_si256(reinterpret_cast<const __m256i*>(sums + col));
795 _mm256_storeu_si256(reinterpret_cast<__m256i*>(sums + col), sum);
818 if (sums) {
819 sums[col] += accum;
/external/libmpeg2/common/arm/
Dideint_cac_a9.s85 @ Get row sums
94 @ Both q0 and q1 have four 32 bit sums corresponding to first 4 rows
99 @ q8 now contains 8 sums
113 @ Get row sums
121 @ Both q2 and q3 have four 32 bit sums corresponding to last 4 rows
126 @ q9 now contains 8 sums
128 @ Compute absolute diff between top and bottom row sums
138 @ q10 now contains 8 absolute diff of sums above the threshold
169 @ d0 now contains 8 absolute diff of sums above the threshold
/external/llvm-project/polly/test/DependenceInfo/
Dreduction_partially_escaping_intermediate_in_other_stmt.ll6 ; void f(int N, int * restrict sums, int * restrict escape) {
9 ; sums[i] += 5;
11 ; escape[N - i + j] = sums[i];
17 define void @f(i32 %N, i32* noalias %sums, i32* noalias %escape) {
35 %arrayidx = getelementptr inbounds i32, i32* %sums, i32 %i.0
45 %arrayidx6 = getelementptr inbounds i32, i32* %sums, i32 %i.0
/external/mesa3d/src/mesa/main/
Dtexcompress_bptc_tmp.h1220 int sums[2][4]; in get_rgba_endpoints_unorm() local
1229 memset(sums, 0, sizeof sums); in get_rgba_endpoints_unorm()
1241 sums[endpoint][i] += p[i]; in get_rgba_endpoints_unorm()
1249 sums[endpoint][3] += p[3]; in get_rgba_endpoints_unorm()
1261 (sums[0][i] + sums[1][i]) / (width * height); in get_rgba_endpoints_unorm()
1264 endpoints[0][i] = sums[0][i] / rgb_left_endpoint_count; in get_rgba_endpoints_unorm()
1265 endpoints[1][i] = (sums[1][i] / in get_rgba_endpoints_unorm()
1273 (sums[0][3] + sums[1][3]) / (width * height); in get_rgba_endpoints_unorm()
1275 endpoints[0][3] = sums[0][3] / alpha_left_endpoint_count; in get_rgba_endpoints_unorm()
1276 endpoints[1][3] = (sums[1][3] / in get_rgba_endpoints_unorm()
[all …]
/external/libgav1/libgav1/src/dsp/x86/
Dconvolve_sse4.inc471 const __m128i sums = SumVerticalTaps<filter_index>(srcs, v_tap);
473 const __m128i results = Compound1DShift(sums);
478 RightShiftWithRounding_S16(sums, kFilterBits - 1);
518 const __m128i sums = SumVerticalTaps<filter_index>(srcs, v_tap);
520 const __m128i results = Compound1DShift(sums);
525 RightShiftWithRounding_S16(sums, kFilterBits - 1);
577 const __m128i sums = SumVerticalTaps<filter_index>(srcs, v_tap);
579 const __m128i results = Compound1DShift(sums);
584 RightShiftWithRounding_S16(sums, kFilterBits - 1);
648 const __m128i sums = SumVerticalTaps<filter_index>(srcs, v_tap);
[all …]
Dcdef_avx2.cc396 const __m128i sums = in Cost0Or4_Pair() local
398 cost[0] = _mm_cvtsi128_si32(sums); in Cost0Or4_Pair()
399 cost[4] = _mm_cvtsi128_si32(_mm_srli_si128(sums, 8)); in Cost0Or4_Pair()
431 const __m128i sums = in CostOdd_Pair() local
433 cost[index_a] = _mm_cvtsi128_si32(sums); in CostOdd_Pair()
434 cost[index_b] = _mm_cvtsi128_si32(_mm_srli_si128(sums, 8)); in CostOdd_Pair()
448 const __m128i sums = in Cost2And6_Pair() local
450 cost[2] = _mm_cvtsi128_si32(sums); in Cost2And6_Pair()
451 cost[6] = _mm_cvtsi128_si32(_mm_srli_si128(sums, 8)); in Cost2And6_Pair()
/external/libchrome/base/sampling_heap_profiler/
Dsampling_heap_profiler_unittest.cc106 std::map<size_t, size_t> sums; in CheckAllocationPattern() local
121 sums[it.first] += it.second; in CheckAllocationPattern()
130 for (auto sum : sums) { in CheckAllocationPattern()
/external/arm-optimized-routines/math/tools/
Dremez.jl353 sums = zeros(BigFloat, maxpow, 3)
359 sums[i,j] += x^(i-1) * y^(j-1) * weight
364 @debug("leastsquares", "sums=", repr(sums))
376 matrix[row, 1+j] = sums[1+i+j, 1]
379 matrix[row, 1+n+j] = -sums[1+i+j, 2]
381 vector[row] = sums[1+i, 2]
388 matrix[row, 1+j] = sums[1+i+j, 2]
391 matrix[row, 1+n+j] = -sums[1+i+j, 3]
393 vector[row] = sums[1+i, 3]

12345678