Lines Matching refs:val
26 b0.val[0] = vcombine_s16(vreinterpret_s16_s32(vget_low_s32(a0)), in vpx_vtrnq_s64_to_s16()
28 b0.val[1] = vcombine_s16(vreinterpret_s16_s32(vget_high_s32(a0)), in vpx_vtrnq_s64_to_s16()
35 b0.val[0] = vcombine_s32(vget_low_s32(a0), vget_low_s32(a1)); in vpx_vtrnq_s64_to_s32()
36 b0.val[1] = vcombine_s32(vget_high_s32(a0), vget_high_s32(a1)); in vpx_vtrnq_s64_to_s32()
42 b0.val[0] = vcombine_s64(vreinterpret_s64_s32(vget_low_s32(a0)), in vpx_vtrnq_s64()
44 b0.val[1] = vcombine_s64(vreinterpret_s64_s32(vget_high_s32(a0)), in vpx_vtrnq_s64()
51 b0.val[0] = vcombine_u8(vreinterpret_u8_u32(vget_low_u32(a0)), in vpx_vtrnq_u64_to_u8()
53 b0.val[1] = vcombine_u8(vreinterpret_u8_u32(vget_high_u32(a0)), in vpx_vtrnq_u64_to_u8()
60 b0.val[0] = vcombine_u16(vreinterpret_u16_u32(vget_low_u32(a0)), in vpx_vtrnq_u64_to_u16()
62 b0.val[1] = vcombine_u16(vreinterpret_u16_u32(vget_high_u32(a0)), in vpx_vtrnq_u64_to_u16()
82 const uint32x2x2_t c0 = vtrn_u32(vreinterpret_u32_u16(b0.val[0]), in transpose_u8_4x4()
83 vreinterpret_u32_u16(b0.val[1])); in transpose_u8_4x4()
90 vtrn_u8(vreinterpret_u8_u32(c0.val[0]), vreinterpret_u8_u32(c0.val[1])); in transpose_u8_4x4()
92 *a0 = d0.val[0]; in transpose_u8_4x4()
93 *a1 = d0.val[1]; in transpose_u8_4x4()
118 const int32x2x2_t c0 = vtrn_s32(vreinterpret_s32_s16(b0.val[0]), in transpose_s16_4x4d()
119 vreinterpret_s32_s16(b1.val[0])); in transpose_s16_4x4d()
120 const int32x2x2_t c1 = vtrn_s32(vreinterpret_s32_s16(b0.val[1]), in transpose_s16_4x4d()
121 vreinterpret_s32_s16(b1.val[1])); in transpose_s16_4x4d()
123 *a0 = vreinterpret_s16_s32(c0.val[0]); in transpose_s16_4x4d()
124 *a1 = vreinterpret_s16_s32(c1.val[0]); in transpose_s16_4x4d()
125 *a2 = vreinterpret_s16_s32(c0.val[1]); in transpose_s16_4x4d()
126 *a3 = vreinterpret_s16_s32(c1.val[1]); in transpose_s16_4x4d()
145 vcombine_s32(vget_low_s32(b0.val[0]), vget_low_s32(b0.val[1])); in transpose_s16_4x4q()
147 vcombine_s32(vget_high_s32(b0.val[0]), vget_high_s32(b0.val[1])); in transpose_s16_4x4q()
156 *a0 = d0.val[0]; in transpose_s16_4x4q()
157 *a1 = d0.val[1]; in transpose_s16_4x4q()
176 vcombine_u32(vget_low_u32(b0.val[0]), vget_low_u32(b0.val[1])); in transpose_u16_4x4q()
178 vcombine_u32(vget_high_u32(b0.val[0]), vget_high_u32(b0.val[1])); in transpose_u16_4x4q()
187 *a0 = d0.val[0]; in transpose_u16_4x4q()
188 *a1 = d0.val[1]; in transpose_u16_4x4q()
225 const uint16x4x2_t c0 = vtrn_u16(vreinterpret_u16_u32(b0.val[0]), in transpose_u8_4x8()
226 vreinterpret_u16_u32(b2.val[0])); in transpose_u8_4x8()
227 const uint16x4x2_t c1 = vtrn_u16(vreinterpret_u16_u32(b1.val[0]), in transpose_u8_4x8()
228 vreinterpret_u16_u32(b3.val[0])); in transpose_u8_4x8()
237 vtrn_u8(vreinterpret_u8_u16(c0.val[0]), vreinterpret_u8_u16(c1.val[0])); in transpose_u8_4x8()
239 vtrn_u8(vreinterpret_u8_u16(c0.val[1]), vreinterpret_u8_u16(c1.val[1])); in transpose_u8_4x8()
241 *a0 = d0.val[0]; in transpose_u8_4x8()
242 *a1 = d0.val[1]; in transpose_u8_4x8()
243 *a2 = d1.val[0]; in transpose_u8_4x8()
244 *a3 = d1.val[1]; in transpose_u8_4x8()
269 const int32x4x2_t c0 = vpx_vtrnq_s64_to_s32(b0.val[0], b1.val[0]); in transpose_s32_4x4()
270 const int32x4x2_t c1 = vpx_vtrnq_s64_to_s32(b0.val[1], b1.val[1]); in transpose_s32_4x4()
272 *a0 = c0.val[0]; in transpose_s32_4x4()
273 *a1 = c1.val[0]; in transpose_s32_4x4()
274 *a2 = c0.val[1]; in transpose_s32_4x4()
275 *a3 = c1.val[1]; in transpose_s32_4x4()
318 const int32x2x2_t c0 = vtrn_s32(vreinterpret_s32_s16(b0.val[0]), in transpose_s16_4x8()
319 vreinterpret_s32_s16(b1.val[0])); in transpose_s16_4x8()
320 const int32x2x2_t c1 = vtrn_s32(vreinterpret_s32_s16(b0.val[1]), in transpose_s16_4x8()
321 vreinterpret_s32_s16(b1.val[1])); in transpose_s16_4x8()
322 const int32x2x2_t c2 = vtrn_s32(vreinterpret_s32_s16(b2.val[0]), in transpose_s16_4x8()
323 vreinterpret_s32_s16(b3.val[0])); in transpose_s16_4x8()
324 const int32x2x2_t c3 = vtrn_s32(vreinterpret_s32_s16(b2.val[1]), in transpose_s16_4x8()
325 vreinterpret_s32_s16(b3.val[1])); in transpose_s16_4x8()
333 *o0 = vcombine_s16(vreinterpret_s16_s32(c0.val[0]), in transpose_s16_4x8()
334 vreinterpret_s16_s32(c2.val[0])); in transpose_s16_4x8()
335 *o1 = vcombine_s16(vreinterpret_s16_s32(c1.val[0]), in transpose_s16_4x8()
336 vreinterpret_s16_s32(c3.val[0])); in transpose_s16_4x8()
337 *o2 = vcombine_s16(vreinterpret_s16_s32(c0.val[1]), in transpose_s16_4x8()
338 vreinterpret_s16_s32(c2.val[1])); in transpose_s16_4x8()
339 *o3 = vcombine_s16(vreinterpret_s16_s32(c1.val[1]), in transpose_s16_4x8()
340 vreinterpret_s16_s32(c3.val[1])); in transpose_s16_4x8()
381 const int64x2x2_t c0 = vpx_vtrnq_s64(b0.val[0], b1.val[0]); in transpose_s32_4x8()
382 const int64x2x2_t c1 = vpx_vtrnq_s64(b0.val[1], b1.val[1]); in transpose_s32_4x8()
383 const int64x2x2_t c2 = vpx_vtrnq_s64(b2.val[0], b3.val[0]); in transpose_s32_4x8()
384 const int64x2x2_t c3 = vpx_vtrnq_s64(b2.val[1], b3.val[1]); in transpose_s32_4x8()
386 *a0 = vreinterpretq_s32_s64(c0.val[0]); in transpose_s32_4x8()
387 *a1 = vreinterpretq_s32_s64(c2.val[0]); in transpose_s32_4x8()
388 *a2 = vreinterpretq_s32_s64(c1.val[0]); in transpose_s32_4x8()
389 *a3 = vreinterpretq_s32_s64(c3.val[0]); in transpose_s32_4x8()
390 *a4 = vreinterpretq_s32_s64(c0.val[1]); in transpose_s32_4x8()
391 *a5 = vreinterpretq_s32_s64(c2.val[1]); in transpose_s32_4x8()
392 *a6 = vreinterpretq_s32_s64(c1.val[1]); in transpose_s32_4x8()
393 *a7 = vreinterpretq_s32_s64(c3.val[1]); in transpose_s32_4x8()
419 vtrn_u16(vreinterpret_u16_u8(b0.val[0]), vreinterpret_u16_u8(b1.val[0])); in transpose_u8_8x4()
421 vtrn_u16(vreinterpret_u16_u8(b0.val[1]), vreinterpret_u16_u8(b1.val[1])); in transpose_u8_8x4()
423 *a0 = vreinterpret_u8_u16(c0.val[0]); in transpose_u8_8x4()
424 *a1 = vreinterpret_u8_u16(c1.val[0]); in transpose_u8_8x4()
425 *a2 = vreinterpret_u8_u16(c0.val[1]); in transpose_u8_8x4()
426 *a3 = vreinterpret_u8_u16(c1.val[1]); in transpose_u8_8x4()
451 const uint32x4x2_t c0 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[0]), in transpose_u16_8x4()
452 vreinterpretq_u32_u16(b1.val[0])); in transpose_u16_8x4()
453 const uint32x4x2_t c1 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[1]), in transpose_u16_8x4()
454 vreinterpretq_u32_u16(b1.val[1])); in transpose_u16_8x4()
456 *a0 = vreinterpretq_u16_u32(c0.val[0]); in transpose_u16_8x4()
457 *a1 = vreinterpretq_u16_u32(c1.val[0]); in transpose_u16_8x4()
458 *a2 = vreinterpretq_u16_u32(c0.val[1]); in transpose_u16_8x4()
459 *a3 = vreinterpretq_u16_u32(c1.val[1]); in transpose_u16_8x4()
500 const int64x2x2_t c0 = vpx_vtrnq_s64(b0.val[0], b2.val[0]); in transpose_s32_8x4()
501 const int64x2x2_t c1 = vpx_vtrnq_s64(b0.val[1], b2.val[1]); in transpose_s32_8x4()
502 const int64x2x2_t c2 = vpx_vtrnq_s64(b1.val[0], b3.val[0]); in transpose_s32_8x4()
503 const int64x2x2_t c3 = vpx_vtrnq_s64(b1.val[1], b3.val[1]); in transpose_s32_8x4()
505 *a0 = vreinterpretq_s32_s64(c0.val[0]); in transpose_s32_8x4()
506 *a1 = vreinterpretq_s32_s64(c1.val[0]); in transpose_s32_8x4()
507 *a2 = vreinterpretq_s32_s64(c0.val[1]); in transpose_s32_8x4()
508 *a3 = vreinterpretq_s32_s64(c1.val[1]); in transpose_s32_8x4()
509 *a4 = vreinterpretq_s32_s64(c2.val[0]); in transpose_s32_8x4()
510 *a5 = vreinterpretq_s32_s64(c3.val[0]); in transpose_s32_8x4()
511 *a6 = vreinterpretq_s32_s64(c2.val[1]); in transpose_s32_8x4()
512 *a7 = vreinterpretq_s32_s64(c3.val[1]); in transpose_s32_8x4()
546 const uint16x8x2_t c0 = vtrnq_u16(vreinterpretq_u16_u8(b0.val[0]), in transpose_u8_8x8()
547 vreinterpretq_u16_u8(b1.val[0])); in transpose_u8_8x8()
548 const uint16x8x2_t c1 = vtrnq_u16(vreinterpretq_u16_u8(b0.val[1]), in transpose_u8_8x8()
549 vreinterpretq_u16_u8(b1.val[1])); in transpose_u8_8x8()
556 const uint32x4x2_t d0 = vuzpq_u32(vreinterpretq_u32_u16(c0.val[0]), in transpose_u8_8x8()
557 vreinterpretq_u32_u16(c1.val[0])); in transpose_u8_8x8()
558 const uint32x4x2_t d1 = vuzpq_u32(vreinterpretq_u32_u16(c0.val[1]), in transpose_u8_8x8()
559 vreinterpretq_u32_u16(c1.val[1])); in transpose_u8_8x8()
561 *a0 = vreinterpret_u8_u32(vget_low_u32(d0.val[0])); in transpose_u8_8x8()
562 *a1 = vreinterpret_u8_u32(vget_high_u32(d0.val[0])); in transpose_u8_8x8()
563 *a2 = vreinterpret_u8_u32(vget_low_u32(d1.val[0])); in transpose_u8_8x8()
564 *a3 = vreinterpret_u8_u32(vget_high_u32(d1.val[0])); in transpose_u8_8x8()
565 *a4 = vreinterpret_u8_u32(vget_low_u32(d0.val[1])); in transpose_u8_8x8()
566 *a5 = vreinterpret_u8_u32(vget_high_u32(d0.val[1])); in transpose_u8_8x8()
567 *a6 = vreinterpret_u8_u32(vget_low_u32(d1.val[1])); in transpose_u8_8x8()
568 *a7 = vreinterpret_u8_u32(vget_high_u32(d1.val[1])); in transpose_u8_8x8()
580 const int32x4x2_t d0 = vtrnq_s32(vreinterpretq_s32_s16(c0.val[0]), in transpose_s16_8x8_new()
581 vreinterpretq_s32_s16(c1.val[0])); in transpose_s16_8x8_new()
582 const int32x4x2_t d1 = vtrnq_s32(vreinterpretq_s32_s16(c0.val[1]), in transpose_s16_8x8_new()
583 vreinterpretq_s32_s16(c1.val[1])); in transpose_s16_8x8_new()
584 const int32x4x2_t d2 = vtrnq_s32(vreinterpretq_s32_s16(c2.val[0]), in transpose_s16_8x8_new()
585 vreinterpretq_s32_s16(c3.val[0])); in transpose_s16_8x8_new()
586 const int32x4x2_t d3 = vtrnq_s32(vreinterpretq_s32_s16(c2.val[1]), in transpose_s16_8x8_new()
587 vreinterpretq_s32_s16(c3.val[1])); in transpose_s16_8x8_new()
590 const int16x8x2_t e0 = vpx_vtrnq_s64_to_s16(d0.val[0], d2.val[0]); in transpose_s16_8x8_new()
591 const int16x8x2_t e1 = vpx_vtrnq_s64_to_s16(d1.val[0], d3.val[0]); in transpose_s16_8x8_new()
592 const int16x8x2_t e2 = vpx_vtrnq_s64_to_s16(d0.val[1], d2.val[1]); in transpose_s16_8x8_new()
593 const int16x8x2_t e3 = vpx_vtrnq_s64_to_s16(d1.val[1], d3.val[1]); in transpose_s16_8x8_new()
595 b[0] = e0.val[0]; in transpose_s16_8x8_new()
596 b[1] = e1.val[0]; in transpose_s16_8x8_new()
597 b[2] = e2.val[0]; in transpose_s16_8x8_new()
598 b[3] = e3.val[0]; in transpose_s16_8x8_new()
599 b[4] = e0.val[1]; in transpose_s16_8x8_new()
600 b[5] = e1.val[1]; in transpose_s16_8x8_new()
601 b[6] = e2.val[1]; in transpose_s16_8x8_new()
602 b[7] = e3.val[1]; in transpose_s16_8x8_new()
643 const int32x4x2_t c0 = vtrnq_s32(vreinterpretq_s32_s16(b0.val[0]), in transpose_s16_8x8()
644 vreinterpretq_s32_s16(b1.val[0])); in transpose_s16_8x8()
645 const int32x4x2_t c1 = vtrnq_s32(vreinterpretq_s32_s16(b0.val[1]), in transpose_s16_8x8()
646 vreinterpretq_s32_s16(b1.val[1])); in transpose_s16_8x8()
647 const int32x4x2_t c2 = vtrnq_s32(vreinterpretq_s32_s16(b2.val[0]), in transpose_s16_8x8()
648 vreinterpretq_s32_s16(b3.val[0])); in transpose_s16_8x8()
649 const int32x4x2_t c3 = vtrnq_s32(vreinterpretq_s32_s16(b2.val[1]), in transpose_s16_8x8()
650 vreinterpretq_s32_s16(b3.val[1])); in transpose_s16_8x8()
661 const int16x8x2_t d0 = vpx_vtrnq_s64_to_s16(c0.val[0], c2.val[0]); in transpose_s16_8x8()
662 const int16x8x2_t d1 = vpx_vtrnq_s64_to_s16(c1.val[0], c3.val[0]); in transpose_s16_8x8()
663 const int16x8x2_t d2 = vpx_vtrnq_s64_to_s16(c0.val[1], c2.val[1]); in transpose_s16_8x8()
664 const int16x8x2_t d3 = vpx_vtrnq_s64_to_s16(c1.val[1], c3.val[1]); in transpose_s16_8x8()
666 *a0 = d0.val[0]; in transpose_s16_8x8()
667 *a1 = d1.val[0]; in transpose_s16_8x8()
668 *a2 = d2.val[0]; in transpose_s16_8x8()
669 *a3 = d3.val[0]; in transpose_s16_8x8()
670 *a4 = d0.val[1]; in transpose_s16_8x8()
671 *a5 = d1.val[1]; in transpose_s16_8x8()
672 *a6 = d2.val[1]; in transpose_s16_8x8()
673 *a7 = d3.val[1]; in transpose_s16_8x8()
714 const uint32x4x2_t c0 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[0]), in transpose_u16_8x8()
715 vreinterpretq_u32_u16(b1.val[0])); in transpose_u16_8x8()
716 const uint32x4x2_t c1 = vtrnq_u32(vreinterpretq_u32_u16(b0.val[1]), in transpose_u16_8x8()
717 vreinterpretq_u32_u16(b1.val[1])); in transpose_u16_8x8()
718 const uint32x4x2_t c2 = vtrnq_u32(vreinterpretq_u32_u16(b2.val[0]), in transpose_u16_8x8()
719 vreinterpretq_u32_u16(b3.val[0])); in transpose_u16_8x8()
720 const uint32x4x2_t c3 = vtrnq_u32(vreinterpretq_u32_u16(b2.val[1]), in transpose_u16_8x8()
721 vreinterpretq_u32_u16(b3.val[1])); in transpose_u16_8x8()
732 const uint16x8x2_t d0 = vpx_vtrnq_u64_to_u16(c0.val[0], c2.val[0]); in transpose_u16_8x8()
733 const uint16x8x2_t d1 = vpx_vtrnq_u64_to_u16(c1.val[0], c3.val[0]); in transpose_u16_8x8()
734 const uint16x8x2_t d2 = vpx_vtrnq_u64_to_u16(c0.val[1], c2.val[1]); in transpose_u16_8x8()
735 const uint16x8x2_t d3 = vpx_vtrnq_u64_to_u16(c1.val[1], c3.val[1]); in transpose_u16_8x8()
737 *a0 = d0.val[0]; in transpose_u16_8x8()
738 *a1 = d1.val[0]; in transpose_u16_8x8()
739 *a2 = d2.val[0]; in transpose_u16_8x8()
740 *a3 = d3.val[0]; in transpose_u16_8x8()
741 *a4 = d0.val[1]; in transpose_u16_8x8()
742 *a5 = d1.val[1]; in transpose_u16_8x8()
743 *a6 = d2.val[1]; in transpose_u16_8x8()
744 *a7 = d3.val[1]; in transpose_u16_8x8()
770 const int32x4x2_t b0 = vtrnq_s32(a0->val[0], a1->val[0]); in transpose_s32_8x8()
771 const int32x4x2_t b1 = vtrnq_s32(a2->val[0], a3->val[0]); in transpose_s32_8x8()
772 const int32x4x2_t b2 = vtrnq_s32(a4->val[0], a5->val[0]); in transpose_s32_8x8()
773 const int32x4x2_t b3 = vtrnq_s32(a6->val[0], a7->val[0]); in transpose_s32_8x8()
774 const int32x4x2_t b4 = vtrnq_s32(a0->val[1], a1->val[1]); in transpose_s32_8x8()
775 const int32x4x2_t b5 = vtrnq_s32(a2->val[1], a3->val[1]); in transpose_s32_8x8()
776 const int32x4x2_t b6 = vtrnq_s32(a4->val[1], a5->val[1]); in transpose_s32_8x8()
777 const int32x4x2_t b7 = vtrnq_s32(a6->val[1], a7->val[1]); in transpose_s32_8x8()
788 const int32x4x2_t c0 = vpx_vtrnq_s64_to_s32(b0.val[0], b1.val[0]); in transpose_s32_8x8()
789 const int32x4x2_t c1 = vpx_vtrnq_s64_to_s32(b0.val[1], b1.val[1]); in transpose_s32_8x8()
790 const int32x4x2_t c2 = vpx_vtrnq_s64_to_s32(b2.val[0], b3.val[0]); in transpose_s32_8x8()
791 const int32x4x2_t c3 = vpx_vtrnq_s64_to_s32(b2.val[1], b3.val[1]); in transpose_s32_8x8()
792 const int32x4x2_t c4 = vpx_vtrnq_s64_to_s32(b4.val[0], b5.val[0]); in transpose_s32_8x8()
793 const int32x4x2_t c5 = vpx_vtrnq_s64_to_s32(b4.val[1], b5.val[1]); in transpose_s32_8x8()
794 const int32x4x2_t c6 = vpx_vtrnq_s64_to_s32(b6.val[0], b7.val[0]); in transpose_s32_8x8()
795 const int32x4x2_t c7 = vpx_vtrnq_s64_to_s32(b6.val[1], b7.val[1]); in transpose_s32_8x8()
806 a0->val[0] = c0.val[0]; in transpose_s32_8x8()
807 a0->val[1] = c2.val[0]; in transpose_s32_8x8()
808 a1->val[0] = c1.val[0]; in transpose_s32_8x8()
809 a1->val[1] = c3.val[0]; in transpose_s32_8x8()
810 a2->val[0] = c0.val[1]; in transpose_s32_8x8()
811 a2->val[1] = c2.val[1]; in transpose_s32_8x8()
812 a3->val[0] = c1.val[1]; in transpose_s32_8x8()
813 a3->val[1] = c3.val[1]; in transpose_s32_8x8()
814 a4->val[0] = c4.val[0]; in transpose_s32_8x8()
815 a4->val[1] = c6.val[0]; in transpose_s32_8x8()
816 a5->val[0] = c5.val[0]; in transpose_s32_8x8()
817 a5->val[1] = c7.val[0]; in transpose_s32_8x8()
818 a6->val[0] = c4.val[1]; in transpose_s32_8x8()
819 a6->val[1] = c6.val[1]; in transpose_s32_8x8()
820 a7->val[0] = c5.val[1]; in transpose_s32_8x8()
821 a7->val[1] = c7.val[1]; in transpose_s32_8x8()
831 out[0].val[0] = left[0]; in transpose_s32_8x8_2()
832 out[0].val[1] = right[0]; in transpose_s32_8x8_2()
833 out[1].val[0] = left[1]; in transpose_s32_8x8_2()
834 out[1].val[1] = right[1]; in transpose_s32_8x8_2()
835 out[2].val[0] = left[2]; in transpose_s32_8x8_2()
836 out[2].val[1] = right[2]; in transpose_s32_8x8_2()
837 out[3].val[0] = left[3]; in transpose_s32_8x8_2()
838 out[3].val[1] = right[3]; in transpose_s32_8x8_2()
839 out[4].val[0] = left[4]; in transpose_s32_8x8_2()
840 out[4].val[1] = right[4]; in transpose_s32_8x8_2()
841 out[5].val[0] = left[5]; in transpose_s32_8x8_2()
842 out[5].val[1] = right[5]; in transpose_s32_8x8_2()
843 out[6].val[0] = left[6]; in transpose_s32_8x8_2()
844 out[6].val[1] = right[6]; in transpose_s32_8x8_2()
845 out[7].val[0] = left[7]; in transpose_s32_8x8_2()
846 out[7].val[1] = right[7]; in transpose_s32_8x8_2()
851 out_left[0] = out[0].val[0]; in transpose_s32_8x8_2()
852 out_left[1] = out[1].val[0]; in transpose_s32_8x8_2()
853 out_left[2] = out[2].val[0]; in transpose_s32_8x8_2()
854 out_left[3] = out[3].val[0]; in transpose_s32_8x8_2()
855 out_left[4] = out[4].val[0]; in transpose_s32_8x8_2()
856 out_left[5] = out[5].val[0]; in transpose_s32_8x8_2()
857 out_left[6] = out[6].val[0]; in transpose_s32_8x8_2()
858 out_left[7] = out[7].val[0]; in transpose_s32_8x8_2()
859 out_right[0] = out[0].val[1]; in transpose_s32_8x8_2()
860 out_right[1] = out[1].val[1]; in transpose_s32_8x8_2()
861 out_right[2] = out[2].val[1]; in transpose_s32_8x8_2()
862 out_right[3] = out[3].val[1]; in transpose_s32_8x8_2()
863 out_right[4] = out[4].val[1]; in transpose_s32_8x8_2()
864 out_right[5] = out[5].val[1]; in transpose_s32_8x8_2()
865 out_right[6] = out[6].val[1]; in transpose_s32_8x8_2()
866 out_right[7] = out[7].val[1]; in transpose_s32_8x8_2()
908 const uint16x8x2_t c0 = vtrnq_u16(vreinterpretq_u16_u8(b0.val[0]), in transpose_u8_16x8()
909 vreinterpretq_u16_u8(b1.val[0])); in transpose_u8_16x8()
910 const uint16x8x2_t c1 = vtrnq_u16(vreinterpretq_u16_u8(b0.val[1]), in transpose_u8_16x8()
911 vreinterpretq_u16_u8(b1.val[1])); in transpose_u8_16x8()
912 const uint16x8x2_t c2 = vtrnq_u16(vreinterpretq_u16_u8(b2.val[0]), in transpose_u8_16x8()
913 vreinterpretq_u16_u8(b3.val[0])); in transpose_u8_16x8()
914 const uint16x8x2_t c3 = vtrnq_u16(vreinterpretq_u16_u8(b2.val[1]), in transpose_u8_16x8()
915 vreinterpretq_u16_u8(b3.val[1])); in transpose_u8_16x8()
926 const uint32x4x2_t d0 = vtrnq_u32(vreinterpretq_u32_u16(c0.val[0]), in transpose_u8_16x8()
927 vreinterpretq_u32_u16(c2.val[0])); in transpose_u8_16x8()
928 const uint32x4x2_t d1 = vtrnq_u32(vreinterpretq_u32_u16(c0.val[1]), in transpose_u8_16x8()
929 vreinterpretq_u32_u16(c2.val[1])); in transpose_u8_16x8()
930 const uint32x4x2_t d2 = vtrnq_u32(vreinterpretq_u32_u16(c1.val[0]), in transpose_u8_16x8()
931 vreinterpretq_u32_u16(c3.val[0])); in transpose_u8_16x8()
932 const uint32x4x2_t d3 = vtrnq_u32(vreinterpretq_u32_u16(c1.val[1]), in transpose_u8_16x8()
933 vreinterpretq_u32_u16(c3.val[1])); in transpose_u8_16x8()
952 *o0 = vget_low_u8(vreinterpretq_u8_u32(d0.val[0])); in transpose_u8_16x8()
953 *o1 = vget_low_u8(vreinterpretq_u8_u32(d2.val[0])); in transpose_u8_16x8()
954 *o2 = vget_low_u8(vreinterpretq_u8_u32(d1.val[0])); in transpose_u8_16x8()
955 *o3 = vget_low_u8(vreinterpretq_u8_u32(d3.val[0])); in transpose_u8_16x8()
956 *o4 = vget_low_u8(vreinterpretq_u8_u32(d0.val[1])); in transpose_u8_16x8()
957 *o5 = vget_low_u8(vreinterpretq_u8_u32(d2.val[1])); in transpose_u8_16x8()
958 *o6 = vget_low_u8(vreinterpretq_u8_u32(d1.val[1])); in transpose_u8_16x8()
959 *o7 = vget_low_u8(vreinterpretq_u8_u32(d3.val[1])); in transpose_u8_16x8()
960 *o8 = vget_high_u8(vreinterpretq_u8_u32(d0.val[0])); in transpose_u8_16x8()
961 *o9 = vget_high_u8(vreinterpretq_u8_u32(d2.val[0])); in transpose_u8_16x8()
962 *o10 = vget_high_u8(vreinterpretq_u8_u32(d1.val[0])); in transpose_u8_16x8()
963 *o11 = vget_high_u8(vreinterpretq_u8_u32(d3.val[0])); in transpose_u8_16x8()
964 *o12 = vget_high_u8(vreinterpretq_u8_u32(d0.val[1])); in transpose_u8_16x8()
965 *o13 = vget_high_u8(vreinterpretq_u8_u32(d2.val[1])); in transpose_u8_16x8()
966 *o14 = vget_high_u8(vreinterpretq_u8_u32(d1.val[1])); in transpose_u8_16x8()
967 *o15 = vget_high_u8(vreinterpretq_u8_u32(d3.val[1])); in transpose_u8_16x8()
1037 const uint16x8x2_t c0 = vtrnq_u16(vreinterpretq_u16_u8(b0.val[0]), in transpose_u8_8x16()
1038 vreinterpretq_u16_u8(b1.val[0])); in transpose_u8_8x16()
1039 const uint16x8x2_t c1 = vtrnq_u16(vreinterpretq_u16_u8(b0.val[1]), in transpose_u8_8x16()
1040 vreinterpretq_u16_u8(b1.val[1])); in transpose_u8_8x16()
1041 const uint16x8x2_t c2 = vtrnq_u16(vreinterpretq_u16_u8(b2.val[0]), in transpose_u8_8x16()
1042 vreinterpretq_u16_u8(b3.val[0])); in transpose_u8_8x16()
1043 const uint16x8x2_t c3 = vtrnq_u16(vreinterpretq_u16_u8(b2.val[1]), in transpose_u8_8x16()
1044 vreinterpretq_u16_u8(b3.val[1])); in transpose_u8_8x16()
1055 const uint32x4x2_t d0 = vtrnq_u32(vreinterpretq_u32_u16(c0.val[0]), in transpose_u8_8x16()
1056 vreinterpretq_u32_u16(c2.val[0])); in transpose_u8_8x16()
1057 const uint32x4x2_t d1 = vtrnq_u32(vreinterpretq_u32_u16(c0.val[1]), in transpose_u8_8x16()
1058 vreinterpretq_u32_u16(c2.val[1])); in transpose_u8_8x16()
1059 const uint32x4x2_t d2 = vtrnq_u32(vreinterpretq_u32_u16(c1.val[0]), in transpose_u8_8x16()
1060 vreinterpretq_u32_u16(c3.val[0])); in transpose_u8_8x16()
1061 const uint32x4x2_t d3 = vtrnq_u32(vreinterpretq_u32_u16(c1.val[1]), in transpose_u8_8x16()
1062 vreinterpretq_u32_u16(c3.val[1])); in transpose_u8_8x16()
1073 *o0 = vreinterpretq_u8_u32(d0.val[0]); in transpose_u8_8x16()
1074 *o1 = vreinterpretq_u8_u32(d2.val[0]); in transpose_u8_8x16()
1075 *o2 = vreinterpretq_u8_u32(d1.val[0]); in transpose_u8_8x16()
1076 *o3 = vreinterpretq_u8_u32(d3.val[0]); in transpose_u8_8x16()
1077 *o4 = vreinterpretq_u8_u32(d0.val[1]); in transpose_u8_8x16()
1078 *o5 = vreinterpretq_u8_u32(d2.val[1]); in transpose_u8_8x16()
1079 *o6 = vreinterpretq_u8_u32(d1.val[1]); in transpose_u8_8x16()
1080 *o7 = vreinterpretq_u8_u32(d3.val[1]); in transpose_u8_8x16()
1154 const uint16x8x2_t c0 = vtrnq_u16(vreinterpretq_u16_u8(b0.val[0]), in transpose_u8_16x16()
1155 vreinterpretq_u16_u8(b1.val[0])); in transpose_u8_16x16()
1156 const uint16x8x2_t c1 = vtrnq_u16(vreinterpretq_u16_u8(b0.val[1]), in transpose_u8_16x16()
1157 vreinterpretq_u16_u8(b1.val[1])); in transpose_u8_16x16()
1158 const uint16x8x2_t c2 = vtrnq_u16(vreinterpretq_u16_u8(b2.val[0]), in transpose_u8_16x16()
1159 vreinterpretq_u16_u8(b3.val[0])); in transpose_u8_16x16()
1160 const uint16x8x2_t c3 = vtrnq_u16(vreinterpretq_u16_u8(b2.val[1]), in transpose_u8_16x16()
1161 vreinterpretq_u16_u8(b3.val[1])); in transpose_u8_16x16()
1162 const uint16x8x2_t c4 = vtrnq_u16(vreinterpretq_u16_u8(b4.val[0]), in transpose_u8_16x16()
1163 vreinterpretq_u16_u8(b5.val[0])); in transpose_u8_16x16()
1164 const uint16x8x2_t c5 = vtrnq_u16(vreinterpretq_u16_u8(b4.val[1]), in transpose_u8_16x16()
1165 vreinterpretq_u16_u8(b5.val[1])); in transpose_u8_16x16()
1166 const uint16x8x2_t c6 = vtrnq_u16(vreinterpretq_u16_u8(b6.val[0]), in transpose_u8_16x16()
1167 vreinterpretq_u16_u8(b7.val[0])); in transpose_u8_16x16()
1168 const uint16x8x2_t c7 = vtrnq_u16(vreinterpretq_u16_u8(b6.val[1]), in transpose_u8_16x16()
1169 vreinterpretq_u16_u8(b7.val[1])); in transpose_u8_16x16()
1188 const uint32x4x2_t d0 = vtrnq_u32(vreinterpretq_u32_u16(c0.val[0]), in transpose_u8_16x16()
1189 vreinterpretq_u32_u16(c2.val[0])); in transpose_u8_16x16()
1190 const uint32x4x2_t d1 = vtrnq_u32(vreinterpretq_u32_u16(c0.val[1]), in transpose_u8_16x16()
1191 vreinterpretq_u32_u16(c2.val[1])); in transpose_u8_16x16()
1192 const uint32x4x2_t d2 = vtrnq_u32(vreinterpretq_u32_u16(c1.val[0]), in transpose_u8_16x16()
1193 vreinterpretq_u32_u16(c3.val[0])); in transpose_u8_16x16()
1194 const uint32x4x2_t d3 = vtrnq_u32(vreinterpretq_u32_u16(c1.val[1]), in transpose_u8_16x16()
1195 vreinterpretq_u32_u16(c3.val[1])); in transpose_u8_16x16()
1196 const uint32x4x2_t d4 = vtrnq_u32(vreinterpretq_u32_u16(c4.val[0]), in transpose_u8_16x16()
1197 vreinterpretq_u32_u16(c6.val[0])); in transpose_u8_16x16()
1198 const uint32x4x2_t d5 = vtrnq_u32(vreinterpretq_u32_u16(c4.val[1]), in transpose_u8_16x16()
1199 vreinterpretq_u32_u16(c6.val[1])); in transpose_u8_16x16()
1200 const uint32x4x2_t d6 = vtrnq_u32(vreinterpretq_u32_u16(c5.val[0]), in transpose_u8_16x16()
1201 vreinterpretq_u32_u16(c7.val[0])); in transpose_u8_16x16()
1202 const uint32x4x2_t d7 = vtrnq_u32(vreinterpretq_u32_u16(c5.val[1]), in transpose_u8_16x16()
1203 vreinterpretq_u32_u16(c7.val[1])); in transpose_u8_16x16()
1222 const uint8x16x2_t e0 = vpx_vtrnq_u64_to_u8(d0.val[0], d4.val[0]); in transpose_u8_16x16()
1223 const uint8x16x2_t e1 = vpx_vtrnq_u64_to_u8(d2.val[0], d6.val[0]); in transpose_u8_16x16()
1224 const uint8x16x2_t e2 = vpx_vtrnq_u64_to_u8(d1.val[0], d5.val[0]); in transpose_u8_16x16()
1225 const uint8x16x2_t e3 = vpx_vtrnq_u64_to_u8(d3.val[0], d7.val[0]); in transpose_u8_16x16()
1226 const uint8x16x2_t e4 = vpx_vtrnq_u64_to_u8(d0.val[1], d4.val[1]); in transpose_u8_16x16()
1227 const uint8x16x2_t e5 = vpx_vtrnq_u64_to_u8(d2.val[1], d6.val[1]); in transpose_u8_16x16()
1228 const uint8x16x2_t e6 = vpx_vtrnq_u64_to_u8(d1.val[1], d5.val[1]); in transpose_u8_16x16()
1229 const uint8x16x2_t e7 = vpx_vtrnq_u64_to_u8(d3.val[1], d7.val[1]); in transpose_u8_16x16()
1248 *o0 = e0.val[0]; in transpose_u8_16x16()
1249 *o1 = e1.val[0]; in transpose_u8_16x16()
1250 *o2 = e2.val[0]; in transpose_u8_16x16()
1251 *o3 = e3.val[0]; in transpose_u8_16x16()
1252 *o4 = e4.val[0]; in transpose_u8_16x16()
1253 *o5 = e5.val[0]; in transpose_u8_16x16()
1254 *o6 = e6.val[0]; in transpose_u8_16x16()
1255 *o7 = e7.val[0]; in transpose_u8_16x16()
1256 *o8 = e0.val[1]; in transpose_u8_16x16()
1257 *o9 = e1.val[1]; in transpose_u8_16x16()
1258 *o10 = e2.val[1]; in transpose_u8_16x16()
1259 *o11 = e3.val[1]; in transpose_u8_16x16()
1260 *o12 = e4.val[1]; in transpose_u8_16x16()
1261 *o13 = e5.val[1]; in transpose_u8_16x16()
1262 *o14 = e6.val[1]; in transpose_u8_16x16()
1263 *o15 = e7.val[1]; in transpose_u8_16x16()
1408 a0->val[0] = vld1q_s32(a); in load_and_transpose_s32_8x8()
1409 a0->val[1] = vld1q_s32(a + 4); in load_and_transpose_s32_8x8()
1411 a1->val[0] = vld1q_s32(a); in load_and_transpose_s32_8x8()
1412 a1->val[1] = vld1q_s32(a + 4); in load_and_transpose_s32_8x8()
1414 a2->val[0] = vld1q_s32(a); in load_and_transpose_s32_8x8()
1415 a2->val[1] = vld1q_s32(a + 4); in load_and_transpose_s32_8x8()
1417 a3->val[0] = vld1q_s32(a); in load_and_transpose_s32_8x8()
1418 a3->val[1] = vld1q_s32(a + 4); in load_and_transpose_s32_8x8()
1420 a4->val[0] = vld1q_s32(a); in load_and_transpose_s32_8x8()
1421 a4->val[1] = vld1q_s32(a + 4); in load_and_transpose_s32_8x8()
1423 a5->val[0] = vld1q_s32(a); in load_and_transpose_s32_8x8()
1424 a5->val[1] = vld1q_s32(a + 4); in load_and_transpose_s32_8x8()
1426 a6->val[0] = vld1q_s32(a); in load_and_transpose_s32_8x8()
1427 a6->val[1] = vld1q_s32(a + 4); in load_and_transpose_s32_8x8()
1429 a7->val[0] = vld1q_s32(a); in load_and_transpose_s32_8x8()
1430 a7->val[1] = vld1q_s32(a + 4); in load_and_transpose_s32_8x8()