| /external/XNNPACK/src/f32-dwconv2d-chw/gen/ |
| D | 5x5s2p2-minmax-wasmsimd-x86-loadsplat-3x4-acc2.c | 331 const v128_t vi8xGIKM = wasm_v32x4_shuffle(vi8xGHIJ, vi8xKLMN, 0, 2, 4, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_3x4_acc2() local
378 const v128_t vi8xACEG = wasm_v32x4_shuffle(vi8x8ACE, vi8xGIKM, 1, 2, 3, 4); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_3x4_acc2()
379 vi8x8ACE = vi8xGIKM; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_3x4_acc2()
|
| D | 5x5s2p2-minmax-wasmsimd-x86-loadsplat-3x4.c | 331 const v128_t vi8xGIKM = wasm_v32x4_shuffle(vi8xGHIJ, vi8xKLMN, 0, 2, 4, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_3x4() local
378 const v128_t vi8xACEG = wasm_v32x4_shuffle(vi8x8ACE, vi8xGIKM, 1, 2, 3, 4); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_3x4()
379 vi8x8ACE = vi8xGIKM; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_loadsplat_3x4()
|
| D | 5x5s2p2-minmax-wasmsimd-arm-loadsplat-3x4-acc2.c | 331 const v128_t vi8xGIKM = wasm_v32x4_shuffle(vi8xGHIJ, vi8xKLMN, 0, 2, 4, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_3x4_acc2() local
378 const v128_t vi8xACEG = wasm_v32x4_shuffle(vi8x8ACE, vi8xGIKM, 1, 2, 3, 4); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_3x4_acc2()
379 vi8x8ACE = vi8xGIKM; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_3x4_acc2()
|
| D | 5x5s2p2-minmax-wasmsimd-arm-loadsplat-3x4.c | 331 const v128_t vi8xGIKM = wasm_v32x4_shuffle(vi8xGHIJ, vi8xKLMN, 0, 2, 4, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_3x4() local
378 const v128_t vi8xACEG = wasm_v32x4_shuffle(vi8x8ACE, vi8xGIKM, 1, 2, 3, 4); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_3x4()
379 vi8x8ACE = vi8xGIKM; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_loadsplat_3x4()
|
| D | 5x5s2p2-minmax-sse-3x4.c | 339 const __m128 vi8xGIKM = _mm_shuffle_ps(vi8xGHIJ, vi8xKLMN, _MM_SHUFFLE(2, 0, 2, 0)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4() local
375 const __m128 vi8xGACE = _mm_move_ss(vi8x8ACE, vi8xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4()
376 vi8x8ACE = vi8xGIKM; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4()
|
| D | 5x5s2p2-minmax-sse-3x4-acc2.c | 339 const __m128 vi8xGIKM = _mm_shuffle_ps(vi8xGHIJ, vi8xKLMN, _MM_SHUFFLE(2, 0, 2, 0)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4_acc2() local
375 const __m128 vi8xGACE = _mm_move_ss(vi8x8ACE, vi8xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4_acc2()
376 vi8x8ACE = vi8xGIKM; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4_acc2()
|
| D | 5x5s2p2-minmax-wasmsimd-x86-splat-3x4-acc2.c | 311 const v128_t vi8xGIKM = wasm_v32x4_shuffle(vi8xGHIJ, vi8xKLMN, 0, 2, 4, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_3x4_acc2() local
358 const v128_t vi8xACEG = wasm_v32x4_shuffle(vi8x8ACE, vi8xGIKM, 1, 2, 3, 4); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_3x4_acc2()
359 vi8x8ACE = vi8xGIKM; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_3x4_acc2()
|
| D | 5x5s2p2-minmax-wasmsimd-arm-splat-3x4-acc2.c | 311 const v128_t vi8xGIKM = wasm_v32x4_shuffle(vi8xGHIJ, vi8xKLMN, 0, 2, 4, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_3x4_acc2() local
358 const v128_t vi8xACEG = wasm_v32x4_shuffle(vi8x8ACE, vi8xGIKM, 1, 2, 3, 4); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_3x4_acc2()
359 vi8x8ACE = vi8xGIKM; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_3x4_acc2()
|
| D | 5x5s2p2-minmax-wasmsimd-x86-splat-3x4.c | 311 const v128_t vi8xGIKM = wasm_v32x4_shuffle(vi8xGHIJ, vi8xKLMN, 0, 2, 4, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_3x4() local
358 const v128_t vi8xACEG = wasm_v32x4_shuffle(vi8x8ACE, vi8xGIKM, 1, 2, 3, 4); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_3x4()
359 vi8x8ACE = vi8xGIKM; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_x86_splat_3x4()
|
| D | 5x5s2p2-minmax-wasmsimd-arm-splat-3x4.c | 311 const v128_t vi8xGIKM = wasm_v32x4_shuffle(vi8xGHIJ, vi8xKLMN, 0, 2, 4, 6); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_3x4() local
358 const v128_t vi8xACEG = wasm_v32x4_shuffle(vi8x8ACE, vi8xGIKM, 1, 2, 3, 4); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_3x4()
359 vi8x8ACE = vi8xGIKM; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__wasmsimd_arm_splat_3x4()
|