| /external/XNNPACK/src/f32-gavgpool-cw/ |
| D | neon-x4.c | 38 float32x4_t vsum2 = vmovq_n_f32(0.0f); in xnn_f32_gavgpool_cw_ukernel__neon_x4() local
49 vsum2 = vaddq_f32(vsum2, vi2); in xnn_f32_gavgpool_cw_ukernel__neon_x4()
67 vsum2 = vaddq_f32(vsum2, vi2); in xnn_f32_gavgpool_cw_ukernel__neon_x4()
75 const float32x4_t vsum23 = vpaddq_f32(vsum2, vsum3); in xnn_f32_gavgpool_cw_ukernel__neon_x4()
80 const float32x4_t vsum23 = vcombine_f32(vadd_f32(vget_low_f32(vsum2), vget_high_f32(vsum2)), in xnn_f32_gavgpool_cw_ukernel__neon_x4()
|
| D | sse-x4.c | 38 __m128 vsum2 = _mm_setzero_ps(); in xnn_f32_gavgpool_cw_ukernel__sse_x4() local
53 vsum2 = _mm_add_ps(vsum2, vi2); in xnn_f32_gavgpool_cw_ukernel__sse_x4()
70 vsum2 = _mm_add_ps(vsum2, vi2); in xnn_f32_gavgpool_cw_ukernel__sse_x4()
77 const __m128 vsum23 = _mm_add_ps(_mm_unpacklo_ps(vsum2, vsum3), _mm_unpackhi_ps(vsum2, vsum3)); in xnn_f32_gavgpool_cw_ukernel__sse_x4()
|
| D | wasmsimd-x86-x4.c | 38 v128_t vsum2 = vsum0; in xnn_f32_gavgpool_cw_ukernel__wasmsimd_x86_x4() local
53 vsum2 = wasm_f32x4_add(vsum2, vi2); in xnn_f32_gavgpool_cw_ukernel__wasmsimd_x86_x4()
70 vsum2 = wasm_f32x4_add(vsum2, vi2); in xnn_f32_gavgpool_cw_ukernel__wasmsimd_x86_x4()
77 … v128_t vsum23 = wasm_f32x4_add(wasm_v32x4_shuffle(vsum2, vsum3, 0, 2, 4, 6), wasm_v32x4_shuffle(v… in xnn_f32_gavgpool_cw_ukernel__wasmsimd_x86_x4()
|
| D | wasmsimd-arm-x4.c | 38 v128_t vsum2 = vsum0; in xnn_f32_gavgpool_cw_ukernel__wasmsimd_arm_x4() local
53 vsum2 = wasm_f32x4_add(vsum2, vi2); in xnn_f32_gavgpool_cw_ukernel__wasmsimd_arm_x4()
70 vsum2 = wasm_f32x4_add(vsum2, vi2); in xnn_f32_gavgpool_cw_ukernel__wasmsimd_arm_x4()
77 … v128_t vsum23 = wasm_f32x4_add(wasm_v32x4_shuffle(vsum2, vsum3, 0, 2, 4, 6), wasm_v32x4_shuffle(v… in xnn_f32_gavgpool_cw_ukernel__wasmsimd_arm_x4()
|
| D | scalar-x1.c | 32 float vsum2 = 0.f; in xnn_f32_gavgpool_cw_ukernel__scalar_x1() local
38 vsum2 += i0[2]; in xnn_f32_gavgpool_cw_ukernel__scalar_x1()
50 float vout = ( (vsum0 + vsum1) + (vsum2 + vsum3) ) * vmultiplier; in xnn_f32_gavgpool_cw_ukernel__scalar_x1()
|
| /external/XNNPACK/src/f16-gavgpool-cw/ |
| D | neonfp16arith-x8.c | 40 float16x8_t vsum2 = vmovq_n_f16(0); in xnn_f16_gavgpool_cw_ukernel__neonfp16arith_x8() local
51 vsum2 = vaddq_f16(vsum2, vi2); in xnn_f16_gavgpool_cw_ukernel__neonfp16arith_x8()
69 vsum2 = vaddq_f16(vsum2, vi2); in xnn_f16_gavgpool_cw_ukernel__neonfp16arith_x8()
77 const float16x8_t vsum23 = vpaddq_f16(vsum2, vsum3); in xnn_f16_gavgpool_cw_ukernel__neonfp16arith_x8()
83 const float16x4_t vsum2_lo = vadd_f16(vget_low_f16(vsum2), vget_high_f16(vsum2)); in xnn_f16_gavgpool_cw_ukernel__neonfp16arith_x8()
|
| D | neonfp16arith-x4.c | 39 float16x4_t vsum2 = vmov_n_f16(0); in xnn_f16_gavgpool_cw_ukernel__neonfp16arith_x4() local
50 vsum2 = vadd_f16(vsum2, vi2); in xnn_f16_gavgpool_cw_ukernel__neonfp16arith_x4()
68 vsum2 = vadd_f16(vsum2, vi2); in xnn_f16_gavgpool_cw_ukernel__neonfp16arith_x4()
75 const float16x4_t vsum23 = vpadd_f16(vsum2, vsum3); in xnn_f16_gavgpool_cw_ukernel__neonfp16arith_x4()
|
| /external/llvm/test/CodeGen/X86/ |
| D | peephole-multiple-folds.ll | 15 %vsum2 = phi <8 x float> [ %vsum2.next, %loopbody ], [ zeroinitializer, %entry ]
19 …%vsum2.next = tail call <8 x float> @llvm.x86.fma.vfmadd.ps.256(<8 x float> %m2, <8 x float> zeroi…
25 %r = fadd <8 x float> %vsum1.next, %vsum2.next
|
| /external/libaom/av1/encoder/x86/ |
| D | highbd_temporal_filter_sse2.c | 75 static int32_t xx_mask_and_hadd(__m128i vsum1, __m128i vsum2, int i) { in xx_mask_and_hadd() argument
79 vecb = _mm_and_si128(vsum2, *(__m128i *)sse_bytemask_2x4[i][1]); in xx_mask_and_hadd()
137 __m128i vsum2 = _mm_add_epi32(vsum23, vsrc[4][1]); in highbd_apply_temporal_filter() local
152 acc_5x5_sse[row][col] = xx_mask_and_hadd(vsum1, vsum2, 0); in highbd_apply_temporal_filter()
153 acc_5x5_sse[row][col + 1] = xx_mask_and_hadd(vsum1, vsum2, 1); in highbd_apply_temporal_filter()
154 acc_5x5_sse[row][col + 2] = xx_mask_and_hadd(vsum1, vsum2, 2); in highbd_apply_temporal_filter()
155 acc_5x5_sse[row][col + 3] = xx_mask_and_hadd(vsum1, vsum2, 3); in highbd_apply_temporal_filter()
166 __m128i vsum2 = _mm_add_epi32(vsum23, vsrc[4][1]); in highbd_apply_temporal_filter() local
177 acc_5x5_sse[row][col] = xx_mask_and_hadd(vsum1, vsum2, 0); in highbd_apply_temporal_filter()
178 acc_5x5_sse[row][col + 1] = xx_mask_and_hadd(vsum1, vsum2, 1); in highbd_apply_temporal_filter()
[all …]
|
| D | highbd_temporal_filter_avx2.c | 183 __m256i vsum2 = _mm256_add_epi32(vsrc[2], vsrc[3]); in highbd_apply_temporal_filter() local
184 __m256i vsum3 = _mm256_add_epi32(vsum1, vsum2); in highbd_apply_temporal_filter()
201 __m256i vsum2 = _mm256_add_epi32(vsrc[2], vsrc[3]); in highbd_apply_temporal_filter() local
202 __m256i vsum3 = _mm256_add_epi32(vsum1, vsum2); in highbd_apply_temporal_filter()
229 __m256i vsum2 = _mm256_add_epi32(vsrc[2], vsrc[3]); in highbd_apply_temporal_filter() local
230 __m256i vsum3 = _mm256_add_epi32(vsum1, vsum2); in highbd_apply_temporal_filter()
248 __m256i vsum2 = _mm256_add_epi32(vsrc[2], vsrc[3]); in highbd_apply_temporal_filter() local
249 __m256i vsum3 = _mm256_add_epi32(vsum1, vsum2); in highbd_apply_temporal_filter()
277 __m256i vsum2 = _mm256_add_epi32(vsrc[2], vsrc[3]); in highbd_apply_temporal_filter() local
278 __m256i vsum3 = _mm256_add_epi32(vsum1, vsum2); in highbd_apply_temporal_filter()
[all …]
|
| D | temporal_filter_sse2.c | 87 static int32_t xx_mask_and_hadd(__m128i vsum1, __m128i vsum2, int i) { in xx_mask_and_hadd() argument
91 vecb = _mm_and_si128(vsum2, *(__m128i *)sse_bytemask_2x4[i][1]); in xx_mask_and_hadd()
142 __m128i vsum2 = _mm_setzero_si128(); in apply_temporal_filter() local
147 vsum2 = _mm_add_epi32(vsrc[i][1], vsum2); in apply_temporal_filter()
168 acc_5x5_sse[row][col + i] = xx_mask_and_hadd(vsum1, vsum2, i); in apply_temporal_filter()
|
| /external/cronet/tot/third_party/llvm-libc/src/AOR_v20.02/networking/aarch64/ |
| D | chksum_simd.c | 81 uint64x2_t vsum2 = { 0, 0 }; in __chksum_aarch64_simd() local
93 vsum2 = vpadalq_u32(vsum2, vtmp2); in __chksum_aarch64_simd()
100 vsum0 = vpadalq_u32(vsum0, vreinterpretq_u32_u64(vsum2)); in __chksum_aarch64_simd()
|
| /external/cronet/stable/third_party/llvm-libc/src/AOR_v20.02/networking/aarch64/ |
| D | chksum_simd.c | 81 uint64x2_t vsum2 = { 0, 0 }; in __chksum_aarch64_simd() local
93 vsum2 = vpadalq_u32(vsum2, vtmp2); in __chksum_aarch64_simd()
100 vsum0 = vpadalq_u32(vsum0, vreinterpretq_u32_u64(vsum2)); in __chksum_aarch64_simd()
|
| /external/arm-optimized-routines/networking/aarch64/ |
| D | chksum_simd.c | 80 uint64x2_t vsum2 = { 0, 0 }; in __chksum_aarch64_simd() local
92 vsum2 = vpadalq_u32(vsum2, vtmp2); in __chksum_aarch64_simd()
99 vsum0 = vpadalq_u32(vsum0, vreinterpretq_u32_u64(vsum2)); in __chksum_aarch64_simd()
|
| /external/cronet/tot/third_party/llvm-libc/src/AOR_v20.02/networking/arm/ |
| D | chksum_simd.c | 55 uint64x2_t vsum2 = { 0, 0 }; in __chksum_arm_simd() local
66 vsum2 = vpadalq_u32(vsum2, vtmp2); in __chksum_arm_simd()
73 vsum0 = vpadalq_u32(vsum0, vreinterpretq_u32_u64(vsum2)); in __chksum_arm_simd()
|
| /external/cronet/stable/third_party/llvm-libc/src/AOR_v20.02/networking/arm/ |
| D | chksum_simd.c | 55 uint64x2_t vsum2 = { 0, 0 }; in __chksum_arm_simd() local
66 vsum2 = vpadalq_u32(vsum2, vtmp2); in __chksum_arm_simd()
73 vsum0 = vpadalq_u32(vsum0, vreinterpretq_u32_u64(vsum2)); in __chksum_arm_simd()
|
| /external/arm-optimized-routines/networking/arm/ |
| D | chksum_simd.c | 54 uint64x2_t vsum2 = { 0, 0 }; in __chksum_arm_simd() local
65 vsum2 = vpadalq_u32(vsum2, vtmp2); in __chksum_arm_simd()
72 vsum0 = vpadalq_u32(vsum0, vreinterpretq_u32_u64(vsum2)); in __chksum_arm_simd()
|
| /external/XNNPACK/src/amalgam/ |
| D | sse.c | 4191 __m128 vsum2 = _mm_setzero_ps(); in xnn_f32_gavgpool_cw_ukernel__sse_x4() local
4206 vsum2 = _mm_add_ps(vsum2, vi2); in xnn_f32_gavgpool_cw_ukernel__sse_x4()
4223 vsum2 = _mm_add_ps(vsum2, vi2); in xnn_f32_gavgpool_cw_ukernel__sse_x4()
4230 const __m128 vsum23 = _mm_add_ps(_mm_unpacklo_ps(vsum2, vsum3), _mm_unpackhi_ps(vsum2, vsum3)); in xnn_f32_gavgpool_cw_ukernel__sse_x4()
|
| /external/swiftshader/third_party/llvm-16.0/configs/common/lib/Target/PowerPC/ |
| D | PPCGenAsmMatcher.inc | 5199 "vsubudm\007vsubuhm\007vsubuhs\007vsubuqm\007vsubuwm\007vsubuws\010vsum2"
|