/external/XNNPACK/src/f32-dwconv2d-chw/gen/ |
D | 5x5p2-minmax-neonfma-1x4-acc2.c | 100 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc2() local 208 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc2() local 313 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc2() local
|
D | 5x5p2-minmax-neon-1x4.c | 100 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4() local 207 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4() local 311 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4() local
|
D | 5x5p2-minmax-wasmsimd-x86-splat-1x4.c | 103 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4() local 209 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4() local 312 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4() local
|
D | 5x5p2-minmax-wasmsimd-arm-splat-1x4-acc2.c | 103 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc2() local 210 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc2() local 314 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc2() local
|
D | 5x5p2-minmax-neonfma-1x4.c | 100 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4() local 207 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4() local 311 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4() local
|
D | 5x5p2-minmax-neon-1x4-acc2.c | 100 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc2() local 208 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc2() local 313 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc2() local
|
D | 5x5p2-minmax-wasmsimd-arm-splat-1x4.c | 103 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4() local 209 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4() local 312 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4() local
|
D | 5x5p2-minmax-wasmsimd-x86-splat-1x4-acc2.c | 103 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc2() local 210 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc2() local 314 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc2() local
|
D | 5x5p2-minmax-neonfma-1x4-acc3.c | 100 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc3() local 209 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc3() local 315 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc3() local
|
D | 5x5p2-minmax-wasmsimd-x86-splat-1x4-acc4.c | 103 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc4() local 212 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc4() local 318 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc4() local
|
D | 5x5p2-minmax-neonfma-1x4-acc4.c | 100 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc4() local 210 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc4() local 317 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc4() local
|
D | 5x5p2-minmax-wasmsimd-arm-splat-1x4-acc4.c | 103 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc4() local 212 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc4() local 318 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc4() local
|
D | 5x5p2-minmax-wasmsimd-arm-splat-1x4-acc5.c | 103 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc5() local 213 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc5() local 320 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc5() local
|
D | 5x5p2-minmax-neon-1x4-acc3.c | 100 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc3() local 209 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc3() local 315 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc3() local
|
D | 5x5p2-minmax-wasmsimd-x86-splat-1x4-acc3.c | 103 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc3() local 211 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc3() local 316 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc3() local
|
D | 5x5p2-minmax-neon-1x4-acc4.c | 100 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc4() local 210 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc4() local 317 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc4() local
|
D | 5x5p2-minmax-wasmsimd-arm-splat-1x4-acc3.c | 103 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc3() local 211 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc3() local 316 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_arm_splat_1x4_acc3() local
|
D | 5x5p2-minmax-wasmsimd-x86-splat-1x4-acc5.c | 103 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc5() local 213 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc5() local 320 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__wasmsimd_x86_splat_1x4_acc5() local
|
D | 3x3p1-minmax-wasmsimd-x86-splat-3x4.c | 110 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__wasmsimd_x86_splat_3x4() local 196 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__wasmsimd_x86_splat_3x4() local
|
D | 3x3p1-minmax-wasmsimd-arm-splat-3x4.c | 110 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__wasmsimd_arm_splat_3x4() local 196 const v128_t vi4x3456 = wasm_v32x4_shuffle(vi4x0123, vi4x4567, 3, 4, 5, 6); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__wasmsimd_arm_splat_3x4() local
|
D | 3x3p1-minmax-neonfma-3x4.c | 107 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__neonfma_3x4() local 194 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__neonfma_3x4() local
|
D | 3x3p1-minmax-neon-3x4.c | 107 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__neon_3x4() local 194 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__neon_3x4() local
|
D | 5x5p2-minmax-neonfma-1x4-acc5.c | 100 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc5() local 211 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc5() local 319 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neonfma_1x4_acc5() local
|
D | 5x5p2-minmax-neon-1x4-acc5.c | 100 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc5() local 211 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc5() local 319 const float32x4_t vi4x3456 = vextq_f32(vi4x0123, vi4x4567, 3); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__neon_1x4_acc5() local
|
D | 3x3p1-minmax-ssse3-3x4.c | 118 …const __m128 vi4x3456 = _mm_castsi128_ps(_mm_alignr_epi8(_mm_castps_si128(vi4x4567), _mm_castps_si… in xnn_f32_dwconv2d_chw_ukernel_3x3p1__ssse3_3x4() local 198 …const __m128 vi4x3456 = _mm_castsi128_ps(_mm_alignr_epi8(_mm_castps_si128(vi4x4567), _mm_castps_si… in xnn_f32_dwconv2d_chw_ukernel_3x3p1__ssse3_3x4() local
|