Home
last modified time | relevance | path

Searched refs:vi5x3012 (Results 1 – 11 of 11) sorted by relevance

/external/XNNPACK/src/f32-dwconv2d-chw/gen/
D5x5p2-minmax-sse-2x4.c94 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4() local
147 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4()
170 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4()
171 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4()
283 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4()
306 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4()
307 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4()
416 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4()
434 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4()
D5x5p2-minmax-sse-2x4-acc2.c94 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc2() local
147 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc2()
170 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc2()
171 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc2()
285 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc2()
308 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc2()
309 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc2()
420 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc2()
438 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc2()
D5x5p2-minmax-sse-2x4-acc3.c94 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc3() local
147 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc3()
170 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc3()
171 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc3()
287 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc3()
310 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc3()
311 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc3()
424 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc3()
442 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_2x4_acc3()
D5x5p2-minmax-sse-3x4-acc2.c100 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4_acc2() local
164 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4_acc2()
193 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4_acc2()
194 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4_acc2()
342 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4_acc2()
371 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4_acc2()
372 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4_acc2()
516 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4_acc2()
540 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4_acc2()
D5x5p2-minmax-sse-3x4.c100 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4() local
164 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4()
193 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4()
194 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4()
339 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4()
368 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4()
369 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4()
510 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4()
534 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_3x4()
D3x3p1-minmax-sse-4x4.c91 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_4x4() local
164 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_4x4()
184 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_4x4()
300 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_4x4()
D3x3p1-minmax-sse-5x4.c97 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_5x4() local
182 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_5x4()
207 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_5x4()
342 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_5x4()
D5x5p2-minmax-sse-4x4.c106 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4() local
181 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4()
216 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4()
217 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4()
395 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4()
430 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4()
431 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4()
604 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4()
634 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4()
D5x5p2-minmax-sse-4x4-acc2.c106 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4_acc2() local
181 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4_acc2()
216 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4_acc2()
217 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4_acc2()
399 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4_acc2()
434 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4_acc2()
435 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4_acc2()
612 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4_acc2()
642 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_4x4_acc2()
D3x3p1-minmax-sse-6x4.c103 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_6x4() local
200 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_6x4()
230 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_6x4()
384 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_3x3p1__sse_6x4()
D5x5p2-minmax-sse-5x4.c112 __m128 vi5x3012 = _mm_setzero_ps(); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_5x4() local
198 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_5x4()
239 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_5x4()
240 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_5x4()
451 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_5x4()
492 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_5x4()
493 vi5x3012 = vi5x7456; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_5x4()
698 const __m128 vi5x3456 = _mm_move_ss(vi5x7456, vi5x3012); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_5x4()
734 const __m128 vi5x2345 = _mm_shuffle_ps(vi5x3012, vi5x7456, _MM_SHUFFLE(2, 1, 0, 3)); in xnn_f32_dwconv2d_chw_ukernel_5x5p2__sse_5x4()