| /external/libvpx/libvpx/vpx_dsp/x86/ |
| D | convolve_avx2.h | 61 __m256i sum1, sum2; in convolve8_16_avx2() local
66 sum1 = _mm256_add_epi16(x0, x2); in convolve8_16_avx2()
69 sum1 = _mm256_add_epi16(sum1, k_64); in convolve8_16_avx2()
70 sum1 = _mm256_adds_epi16(sum1, sum2); in convolve8_16_avx2()
72 sum1 = _mm256_srai_epi16(sum1, 7); in convolve8_16_avx2()
73 return sum1; in convolve8_16_avx2()
88 __m128i sum1, sum2; in convolve8_8_avx2() local
93 sum1 = _mm_add_epi16(x0, x2); in convolve8_8_avx2()
96 sum1 = _mm_add_epi16(sum1, k_64); in convolve8_8_avx2()
97 sum1 = _mm_adds_epi16(sum1, sum2); in convolve8_8_avx2()
[all …]
|
| D | convolve_ssse3.h | 51 __m128i sum1, sum2; in convolve8_8_ssse3() local
56 sum1 = _mm_add_epi16(x0, x2); in convolve8_8_ssse3()
59 sum1 = _mm_add_epi16(sum1, k_64); in convolve8_8_ssse3()
60 sum1 = _mm_adds_epi16(sum1, sum2); in convolve8_8_ssse3()
62 sum1 = _mm_srai_epi16(sum1, 7); in convolve8_8_ssse3()
63 return sum1; in convolve8_8_ssse3()
|
| /external/swiftshader/third_party/llvm-7.0/llvm/test/Transforms/LoopUnrollAndJam/ |
| D | unprofitable.ll | 30 %sum1 = phi i32 [ 0, %for.outer ], [ %add, %for.inner2 ]
33 %add = add i32 %0, %sum1
77 %sum1 = phi i32 [ 0, %for.outer ], [ %add, %for.inner ]
80 %add = add i32 %0, %sum1
121 %sum1 = phi i32 [ 0, %for.outer ], [ %add, %for.inner ]
124 %add = add i32 %0, %sum1
125 %add0 = add i32 %0, %sum1
126 %add1 = add i32 %0, %sum1
127 %add2 = add i32 %0, %sum1
128 %add3 = add i32 %0, %sum1
[all …]
|
| D | disable.ll | 33 %sum1.025 = phi i32 [ %0, %for.outer ], [ %add, %for.inner ]
36 %add = add i32 %1, %sum1.025
78 %sum1.029 = phi i32 [ %sum1.1, %for.inner ], [ 0, %for.outer ]
82 %sub = add i32 %sum1.029, 10
84 %sum1.1 = select i1 %tobool7, i32 %sum1.029, i32 %add
90 %sum1.1.lcssa = phi i32 [ 0, %for.outer ], [ %sum1.1, %for.inner ]
92 store i32 %sum1.1.lcssa, i32* %arrayidx11, align 4
178 %sum1.020 = phi i32 [ 0, %for.outer ], [ %add, %for.inner ]
181 %add = add i32 %0, %sum1.020
387 %sum1 = phi i32 [ 0, %for.outer ], [ %add, %for.inner ]
[all …]
|
| /external/libopus/celt/x86/ |
| D | pitch_sse4_1.c | 117 __m128i sum0, sum1, sum2, sum3, vecSum; in xcorr_kernel_sse4_1() local
123 sum1 = _mm_setzero_si128(); in xcorr_kernel_sse4_1()
136 sum1 = _mm_add_epi32(sum1, _mm_madd_epi16(vecX, vecY1)); in xcorr_kernel_sse4_1()
144 sum1 = _mm_add_epi32(sum1, _mm_unpackhi_epi64( sum1, sum1)); in xcorr_kernel_sse4_1()
145 sum1 = _mm_add_epi32(sum1, _mm_shufflelo_epi16( sum1, 0x0E)); in xcorr_kernel_sse4_1()
153 vecSum = _mm_unpacklo_epi64(_mm_unpacklo_epi32(sum0, sum1), in xcorr_kernel_sse4_1()
170 sum1 = _mm_mullo_epi32(vecX1, vecY1); in xcorr_kernel_sse4_1()
174 sum0 = _mm_add_epi32(sum0, sum1); in xcorr_kernel_sse4_1()
|
| /external/libaom/libaom/av1/common/arm/ |
| D | convolve_neon.h | 24 int32x4_t sum0, sum1; in wiener_convolve8_vert_4x8() local
43 sum1 = vmull_n_s16(vget_high_s16(ss0), filter_y[0]); in wiener_convolve8_vert_4x8()
44 sum1 = vmlal_n_s16(sum1, vget_high_s16(ss1), filter_y[1]); in wiener_convolve8_vert_4x8()
45 sum1 = vmlal_n_s16(sum1, vget_high_s16(ss2), filter_y[2]); in wiener_convolve8_vert_4x8()
46 sum1 = vmlal_n_s16(sum1, vget_high_s16(s3), filter_y[3]); in wiener_convolve8_vert_4x8()
49 sum1 = vsubq_s32(sum1, round_vec); in wiener_convolve8_vert_4x8()
53 sum1 = vrshlq_s32(sum1, round_bits); in wiener_convolve8_vert_4x8()
56 sum1 = vmaxq_s32(sum1, zero); in wiener_convolve8_vert_4x8()
60 tmp1 = vqmovn_u32(vreinterpretq_u32_s32(sum1)); in wiener_convolve8_vert_4x8()
|
| /external/flac/libFLAC/ |
| D | lpc_intrin_sse.c | 97 __m128 sum0, sum1; in FLAC__lpc_compute_autocorrelation_intrin_sse_lag_8_new() local
104 sum1 = _mm_setzero_ps(); in FLAC__lpc_compute_autocorrelation_intrin_sse_lag_8_new()
112 sum1 = _mm_add_ps(sum1, _mm_mul_ps(d1, d)); in FLAC__lpc_compute_autocorrelation_intrin_sse_lag_8_new()
127 sum1 = _mm_add_ps(sum1, _mm_mul_ps(d, d1)); in FLAC__lpc_compute_autocorrelation_intrin_sse_lag_8_new()
133 _mm_storeu_ps(autoc+4, sum1); in FLAC__lpc_compute_autocorrelation_intrin_sse_lag_8_new()
141 __m128 sum0, sum1, sum2; in FLAC__lpc_compute_autocorrelation_intrin_sse_lag_12_new() local
148 sum1 = _mm_setzero_ps(); in FLAC__lpc_compute_autocorrelation_intrin_sse_lag_12_new()
158 sum1 = _mm_add_ps(sum1, _mm_mul_ps(d1, d)); in FLAC__lpc_compute_autocorrelation_intrin_sse_lag_12_new()
178 sum1 = _mm_add_ps(sum1, _mm_mul_ps(d, d1)); in FLAC__lpc_compute_autocorrelation_intrin_sse_lag_12_new()
184 _mm_storeu_ps(autoc+4, sum1); in FLAC__lpc_compute_autocorrelation_intrin_sse_lag_12_new()
[all …]
|
| /external/python/cpython2/Modules/zlib/ |
| D | adler32.c | 148 unsigned long sum1; local
159 sum1 = adler1 & 0xffff;
160 sum2 = rem * sum1;
162 sum1 += (adler2 & 0xffff) + BASE - 1;
164 if (sum1 >= BASE) sum1 -= BASE;
165 if (sum1 >= BASE) sum1 -= BASE;
168 return sum1 | (sum2 << 16);
|
| /external/zlib/src/ |
| D | adler32.c | 148 unsigned long sum1; local
159 sum1 = adler1 & 0xffff;
160 sum2 = rem * sum1;
162 sum1 += (adler2 & 0xffff) + BASE - 1;
164 if (sum1 >= BASE) sum1 -= BASE;
165 if (sum1 >= BASE) sum1 -= BASE;
168 return sum1 | (sum2 << 16);
|
| /external/libaom/libaom/aom_dsp/x86/ |
| D | sad4d_avx2.c | 176 uint32_t sum1[4]; in aom_sad32x64x4d_avx2() local
188 aom_sad32x32x4d_avx2(src, src_stride, rf, ref_stride, sum1); in aom_sad32x64x4d_avx2()
189 res[0] = sum0[0] + sum1[0]; in aom_sad32x64x4d_avx2()
190 res[1] = sum0[1] + sum1[1]; in aom_sad32x64x4d_avx2()
191 res[2] = sum0[2] + sum1[2]; in aom_sad32x64x4d_avx2()
192 res[3] = sum0[3] + sum1[3]; in aom_sad32x64x4d_avx2()
200 uint32_t sum1[4]; in aom_sad64x32x4d_avx2() local
213 aom_sad32x32x4d_avx2(src, src_stride, rf, ref_stride, sum1); in aom_sad64x32x4d_avx2()
214 res[0] = sum0[0] + sum1[0]; in aom_sad64x32x4d_avx2()
215 res[1] = sum0[1] + sum1[1]; in aom_sad64x32x4d_avx2()
[all …]
|
| D | sad_impl_avx2.c | 98 __m128i sum0, sum1; in aom_sad64x128x4d_avx2() local
111 sad64x64x4d(src, src_stride, rf, ref_stride, &sum1); in aom_sad64x128x4d_avx2()
112 sum0 = _mm_add_epi32(sum0, sum1); in aom_sad64x128x4d_avx2()
119 __m128i sum0, sum1; in aom_sad128x64x4d_avx2() local
133 sad64x64x4d(src, src_stride, rf, ref_stride, &sum1); in aom_sad128x64x4d_avx2()
134 sum0 = _mm_add_epi32(sum0, sum1); in aom_sad128x64x4d_avx2()
143 uint32_t sum1[4]; in aom_sad128x128x4d_avx2() local
155 aom_sad128x64x4d_avx2(src, src_stride, rf, ref_stride, sum1); in aom_sad128x128x4d_avx2()
156 res[0] = sum0[0] + sum1[0]; in aom_sad128x128x4d_avx2()
157 res[1] = sum0[1] + sum1[1]; in aom_sad128x128x4d_avx2()
[all …]
|
| /external/libvpx/libvpx/vpx_dsp/arm/ |
| D | variance_neon.c | 212 int sum1, sum2; in vpx_variance32x64_neon() local
215 &sum1); in vpx_variance32x64_neon()
220 sum1 += sum2; in vpx_variance32x64_neon()
221 return *sse - (unsigned int)(((int64_t)sum1 * sum1) >> 11); in vpx_variance32x64_neon()
227 int sum1, sum2; in vpx_variance64x32_neon() local
230 &sum1); in vpx_variance64x32_neon()
235 sum1 += sum2; in vpx_variance64x32_neon()
236 return *sse - (unsigned int)(((int64_t)sum1 * sum1) >> 11); in vpx_variance64x32_neon()
242 int sum1, sum2; in vpx_variance64x64_neon() local
246 &sum1); in vpx_variance64x64_neon()
[all …]
|
| D | loopfilter_neon.c | 186 uint16x8_t *sum1) { in filter_update_16() argument
188 *sum1 = vsubw_u8(*sum1, vget_high_u8(sub0)); in filter_update_16()
190 *sum1 = vsubw_u8(*sum1, vget_high_u8(sub1)); in filter_update_16()
192 *sum1 = vaddw_u8(*sum1, vget_high_u8(add0)); in filter_update_16()
194 *sum1 = vaddw_u8(*sum1, vget_high_u8(add1)); in filter_update_16()
208 const uint8x16_t add1, uint16x8_t *sum0, uint16x8_t *sum1) { in calc_7_tap_filter_16_kernel() argument
209 filter_update_16(sub0, sub1, add0, add1, sum0, sum1); in calc_7_tap_filter_16_kernel()
210 return vcombine_u8(vrshrn_n_u16(*sum0, 3), vrshrn_n_u16(*sum1, 3)); in calc_7_tap_filter_16_kernel()
224 uint16x8_t *sum0, uint16x8_t *sum1) { in apply_15_tap_filter_16_kernel() argument
226 filter_update_16(sub0, sub1, add0, add1, sum0, sum1); in apply_15_tap_filter_16_kernel()
[all …]
|
| /external/speex/libspeexdsp/ |
| D | resample_sse.h | 104 __m128d sum1 = _mm_setzero_pd(); in interpolate_product_double() local
113 sum1 = _mm_add_pd(sum1, _mm_cvtps_pd(t)); in interpolate_product_double()
117 sum1 = _mm_add_pd(sum1, _mm_cvtps_pd(t)); in interpolate_product_double()
120 sum1 = _mm_mul_pd(f1, sum1); in interpolate_product_double()
122 sum = _mm_add_pd(sum1, sum2); in interpolate_product_double()
|
| /external/libaom/libaom/aom_dsp/arm/ |
| D | variance_neon.c | 101 int sum1, sum2; in aom_variance32x64_neon() local
103 variance_neon_w8(a, a_stride, b, b_stride, 32, 32, &sse1, &sum1); in aom_variance32x64_neon()
107 sum1 += sum2; in aom_variance32x64_neon()
108 return *sse - (unsigned int)(((int64_t)sum1 * sum1) >> 11); in aom_variance32x64_neon()
114 int sum1, sum2; in aom_variance64x32_neon() local
116 variance_neon_w8(a, a_stride, b, b_stride, 64, 16, &sse1, &sum1); in aom_variance64x32_neon()
120 sum1 += sum2; in aom_variance64x32_neon()
121 return *sse - (unsigned int)(((int64_t)sum1 * sum1) >> 11); in aom_variance64x32_neon()
127 int sum1, sum2; in aom_variance64x64_neon() local
130 variance_neon_w8(a, a_stride, b, b_stride, 64, 16, &sse1, &sum1); in aom_variance64x64_neon()
[all …]
|
| /external/webp/src/dsp/ |
| D | ssim_sse2.c | 37 const __m128i sum1 = _mm_madd_epi16(C0, C0); in SubtractAndSquare_SSE2() local
39 *sum = _mm_add_epi32(sum1, sum2); in SubtractAndSquare_SSE2()
52 __m128i sum1; in AccumulateSSE_SSE2() local
62 SubtractAndSquare_SSE2(a0, b0, &sum1); in AccumulateSSE_SSE2()
63 sum = _mm_add_epi32(sum, sum1); in AccumulateSSE_SSE2()
70 SubtractAndSquare_SSE2(a0, b0, &sum1); in AccumulateSSE_SSE2()
71 sum = _mm_add_epi32(sum, sum1); in AccumulateSSE_SSE2()
|
| /external/libavc/common/x86/ |
| D | ih264_resi_trans_quant_sse42.c | 114 __m128i sum0, sum1, sum2, cmp0, cmp1; in ih264_resi_trans_quant_4x4_sse42() local
303 sum1 = _mm_hadd_epi16(sum0, zero_8x16b); in ih264_resi_trans_quant_4x4_sse42()
304 sum2 = _mm_hadd_epi16(sum1, zero_8x16b); in ih264_resi_trans_quant_4x4_sse42()
316 sum1 = _mm_hadd_epi16(sum0, zero_8x16b); in ih264_resi_trans_quant_4x4_sse42()
317 sum2 = _mm_hadd_epi16(sum1, zero_8x16b); in ih264_resi_trans_quant_4x4_sse42()
388 __m128i cmp0, cmp1, sum0, sum1, sum2; in ih264_resi_trans_quant_chroma_4x4_sse42() local
588 sum1 = _mm_hadd_epi16(sum0, zero_8x16b); in ih264_resi_trans_quant_chroma_4x4_sse42()
589 sum2 = _mm_hadd_epi16(sum1, zero_8x16b); in ih264_resi_trans_quant_chroma_4x4_sse42()
601 sum1 = _mm_hadd_epi16(sum0, zero_8x16b); in ih264_resi_trans_quant_chroma_4x4_sse42()
602 sum2 = _mm_hadd_epi16(sum1, zero_8x16b); in ih264_resi_trans_quant_chroma_4x4_sse42()
[all …]
|
| /external/google-breakpad/src/common/mac/ |
| D | macho_id.cc | 89 #define DO1(buf,i) {sum1 += (buf)[i]; sum2 += sum1;} in UpdateCRC()
95 uint32_t sum1 = crc_ & 0xFFFF; in UpdateCRC() local
106 sum1 %= MOD_ADLER; in UpdateCRC()
118 sum1 += *bytes++; in UpdateCRC()
119 sum2 += sum1; in UpdateCRC()
121 sum1 %= MOD_ADLER; in UpdateCRC()
123 crc_ = (sum2 << 16) | sum1; in UpdateCRC()
|
| /external/swiftshader/third_party/llvm-7.0/llvm/test/CodeGen/Hexagon/ |
| D | swp-multi-loops.ll | 28 %sum1.026 = phi i32 [ %add8, %for.inc12 ], [ 0, %for.body3.lr.ph.preheader ]
49 %sum1.121 = phi i32 [ %sum1.026, %for.end ], [ %add8, %for.body6 ]
53 %add8 = add nsw i32 %1, %sum1.121
68 %sum1.0.lcssa = phi i32 [ 0, %entry ], [ %add8, %for.end14.loopexit ]
70 %add15 = add nsw i32 %sum1.0.lcssa, %sum.0.lcssa
|
| /external/llvm/test/CodeGen/Hexagon/ |
| D | swp-multi-loops.ll | 28 %sum1.026 = phi i32 [ %add8, %for.inc12 ], [ 0, %for.body3.lr.ph.preheader ]
49 %sum1.121 = phi i32 [ %sum1.026, %for.end ], [ %add8, %for.body6 ]
53 %add8 = add nsw i32 %1, %sum1.121
68 %sum1.0.lcssa = phi i32 [ 0, %entry ], [ %add8, %for.end14.loopexit ]
70 %add15 = add nsw i32 %sum1.0.lcssa, %sum.0.lcssa
|
| /external/libopus/celt/arm/ |
| D | celt_pitch_xcorr_arm_gnu.s | 361 @ r7 = opus_val32 sum1
429 @ maxcorr = max(maxcorr, sum0, sum1, sum2, sum3)
447 @ {r10, r11} = {sum0, sum1} = {0, 0}
457 SMLABT r11, r6, r8, r11 @ sum1 = MAC16_16(sum1, x_0, y_1)
461 SMLATB r11, r6, r9, r11 @ sum1 = MAC16_16(sum1, x_1, y_2)
464 SMLABT r11, r7, r9, r11 @ sum1 = MAC16_16(sum1, x_2, y_3)
467 SMLATB r11, r7, r8, r11 @ sum1 = MAC16_16(sum1, x_3, y_4)
476 SMLABT r11, r6, r8, r11 @ sum1 = MAC16_16(sum1, x_0, y_1)
480 SMLATB r11, r6, r9, r11 @ sum1 = MAC16_16(sum1, x_1, y_2)
487 SMLABT r11, r6, r8, r11 @ sum1 = MAC16_16(sum1, x_0, y_1)
[all …]
|
| /external/webrtc/webrtc/common_audio/signal_processing/ |
| D | cross_correlation_neon.c | 24 int64x2_t sum1 = vdupq_n_s64(0); in DotProductWithScaleNeon() local
40 sum1 = vpadalq_s32(sum1, tmp1); in DotProductWithScaleNeon()
53 sum0 = vaddq_s64(sum0, sum1); in DotProductWithScaleNeon()
|
| D | filter_ar_fast_q12_armv7.S | 56 mov r7, #0 @ sum1
66 smlatb r7, r10, r5, r7 @ sum1 += coefficients[j] * data_out[i - j];
76 smlabb r7, r10, r5, r7 @ sum1 += coefficients[2] * data_out[i - 2];
82 smlatb r7, r10, r5, r7 @ sum1 += coefficients[1] * data_out[i - 1];
86 sub r6, r7 @ output1 -= sum1;
97 sub r6, r8 @ output1 -= sum1;
|
| /external/swiftshader/third_party/llvm-7.0/llvm/test/CodeGen/NVPTX/ |
| D | ldg-invariant.ll | 36 %sum1 = fadd half %v1, %v2
38 %sum = fadd half %sum1, %sum2
52 %sum1 = fadd half %v1, %v2
54 %sum = fadd half %sum1, %sum2
76 %sum1 = add i32 %v1, %v2
78 %sum3 = add i32 %sum1, %sum2
|
| /external/libaom/libaom/av1/common/x86/ |
| D | selfguided_sse4.c | 147 static __m128i compute_p(__m128i sum1, __m128i sum2, int bit_depth, int n) { in compute_p() argument
155 const __m128i b = _mm_srl_epi32(_mm_add_epi32(sum1, rounding_b), shift_b); in compute_p()
161 bb = _mm_madd_epi16(sum1, sum1); in compute_p()
196 __m128i sum1 = boxsum_from_ii(Dij, buf_stride, r); in calc_ab() local
206 sum1 = _mm_and_si128(mask[idx], sum1); in calc_ab()
210 const __m128i p = compute_p(sum1, sum2, bit_depth, n); in calc_ab()
233 const __m128i b_int = _mm_mullo_epi32(a_comp_over_n, sum1); in calc_ab()
336 __m128i sum1 = boxsum_from_ii(Dij, buf_stride, r); in calc_ab_fast() local
346 sum1 = _mm_and_si128(mask[idx], sum1); in calc_ab_fast()
350 const __m128i p = compute_p(sum1, sum2, bit_depth, n); in calc_ab_fast()
[all …]
|