/external/XNNPACK/src/f32-dwconv2d-chw/gen/ |
D | 5x5s2p2-minmax-sse-1x4-acc4.c | 224 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vi1xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc4() local 234 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc4() 321 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vzero); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc4() local 327 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc4()
|
D | 5x5s2p2-minmax-sse-1x4-acc2.c | 224 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vi1xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc2() local 234 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc2() 319 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vzero); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc2() local 325 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc2()
|
D | 5x5s2p2-minmax-sse-1x4.c | 224 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vi1xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4() local 234 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4() 318 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vzero); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4() local 324 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4()
|
D | 5x5s2p2-minmax-sse-1x4-acc5.c | 224 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vi1xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc5() local 234 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc5() 322 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vzero); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc5() local 328 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc5()
|
D | 5x5s2p2-minmax-sse-1x4-acc3.c | 224 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vi1xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc3() local 234 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc3() 320 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vzero); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc3() local 326 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_1x4_acc3()
|
D | 5x5s2p2-minmax-sse-2x4-acc2.c | 293 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vi1xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4_acc2() local 307 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4_acc2() 436 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vzero); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4_acc2() local 444 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4_acc2()
|
D | 5x5s2p2-minmax-sse-2x4.c | 293 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vi1xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4() local 307 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4() 434 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vzero); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4() local 442 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4()
|
D | 5x5s2p2-minmax-sse-2x4-acc3.c | 293 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vi1xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4_acc3() local 307 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4_acc3() 438 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vzero); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4_acc3() local 446 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_2x4_acc3()
|
D | 5x5s2p2-minmax-sse-3x4.c | 361 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vi1xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4() local 379 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4() 549 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vzero); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4() local 559 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4()
|
D | 5x5s2p2-minmax-sse-3x4-acc2.c | 361 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vi1xGIKM); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4_acc2() local 379 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4_acc2() 552 const __m128 vi1xGACE = _mm_move_ss(vi1x8ACE, vzero); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4_acc2() local 562 const __m128 vi1xACEG = _mm_shuffle_ps(vi1xGACE, vi1xGACE, _MM_SHUFFLE(0, 3, 2, 1)); in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__sse_3x4_acc2()
|