Lines Matching refs:dst
14 void vpx_v_predictor_16x16_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_v_predictor_16x16_vsx() argument
20 for (i = 0; i < 16; i++, dst += stride) { in vpx_v_predictor_16x16_vsx()
21 vec_vsx_st(d, 0, dst); in vpx_v_predictor_16x16_vsx()
25 void vpx_v_predictor_32x32_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_v_predictor_32x32_vsx() argument
32 for (i = 0; i < 32; i++, dst += stride) { in vpx_v_predictor_32x32_vsx()
33 vec_vsx_st(d0, 0, dst); in vpx_v_predictor_32x32_vsx()
34 vec_vsx_st(d1, 16, dst); in vpx_v_predictor_32x32_vsx()
42 void vpx_h_predictor_4x4_vsx(uint8_t *dst, ptrdiff_t stride,
52 vec_vsx_st(vec_sel(v0, vec_vsx_ld(0, dst), (uint8x16_t)mask4), 0, dst);
53 dst += stride;
54 vec_vsx_st(vec_sel(v1, vec_vsx_ld(0, dst), (uint8x16_t)mask4), 0, dst);
55 dst += stride;
56 vec_vsx_st(vec_sel(v2, vec_vsx_ld(0, dst), (uint8x16_t)mask4), 0, dst);
57 dst += stride;
58 vec_vsx_st(vec_sel(v3, vec_vsx_ld(0, dst), (uint8x16_t)mask4), 0, dst);
61 void vpx_h_predictor_8x8_vsx(uint8_t *dst, ptrdiff_t stride,
76 vec_vsx_st(xxpermdi(v0, vec_vsx_ld(0, dst), 1), 0, dst);
77 dst += stride;
78 vec_vsx_st(xxpermdi(v1, vec_vsx_ld(0, dst), 1), 0, dst);
79 dst += stride;
80 vec_vsx_st(xxpermdi(v2, vec_vsx_ld(0, dst), 1), 0, dst);
81 dst += stride;
82 vec_vsx_st(xxpermdi(v3, vec_vsx_ld(0, dst), 1), 0, dst);
83 dst += stride;
84 vec_vsx_st(xxpermdi(v4, vec_vsx_ld(0, dst), 1), 0, dst);
85 dst += stride;
86 vec_vsx_st(xxpermdi(v5, vec_vsx_ld(0, dst), 1), 0, dst);
87 dst += stride;
88 vec_vsx_st(xxpermdi(v6, vec_vsx_ld(0, dst), 1), 0, dst);
89 dst += stride;
90 vec_vsx_st(xxpermdi(v7, vec_vsx_ld(0, dst), 1), 0, dst);
94 void vpx_h_predictor_16x16_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_h_predictor_16x16_vsx() argument
119 vec_vsx_st(v0, 0, dst); in vpx_h_predictor_16x16_vsx()
120 dst += stride; in vpx_h_predictor_16x16_vsx()
121 vec_vsx_st(v1, 0, dst); in vpx_h_predictor_16x16_vsx()
122 dst += stride; in vpx_h_predictor_16x16_vsx()
123 vec_vsx_st(v2, 0, dst); in vpx_h_predictor_16x16_vsx()
124 dst += stride; in vpx_h_predictor_16x16_vsx()
125 vec_vsx_st(v3, 0, dst); in vpx_h_predictor_16x16_vsx()
126 dst += stride; in vpx_h_predictor_16x16_vsx()
127 vec_vsx_st(v4, 0, dst); in vpx_h_predictor_16x16_vsx()
128 dst += stride; in vpx_h_predictor_16x16_vsx()
129 vec_vsx_st(v5, 0, dst); in vpx_h_predictor_16x16_vsx()
130 dst += stride; in vpx_h_predictor_16x16_vsx()
131 vec_vsx_st(v6, 0, dst); in vpx_h_predictor_16x16_vsx()
132 dst += stride; in vpx_h_predictor_16x16_vsx()
133 vec_vsx_st(v7, 0, dst); in vpx_h_predictor_16x16_vsx()
134 dst += stride; in vpx_h_predictor_16x16_vsx()
135 vec_vsx_st(v8, 0, dst); in vpx_h_predictor_16x16_vsx()
136 dst += stride; in vpx_h_predictor_16x16_vsx()
137 vec_vsx_st(v9, 0, dst); in vpx_h_predictor_16x16_vsx()
138 dst += stride; in vpx_h_predictor_16x16_vsx()
139 vec_vsx_st(v10, 0, dst); in vpx_h_predictor_16x16_vsx()
140 dst += stride; in vpx_h_predictor_16x16_vsx()
141 vec_vsx_st(v11, 0, dst); in vpx_h_predictor_16x16_vsx()
142 dst += stride; in vpx_h_predictor_16x16_vsx()
143 vec_vsx_st(v12, 0, dst); in vpx_h_predictor_16x16_vsx()
144 dst += stride; in vpx_h_predictor_16x16_vsx()
145 vec_vsx_st(v13, 0, dst); in vpx_h_predictor_16x16_vsx()
146 dst += stride; in vpx_h_predictor_16x16_vsx()
147 vec_vsx_st(v14, 0, dst); in vpx_h_predictor_16x16_vsx()
148 dst += stride; in vpx_h_predictor_16x16_vsx()
149 vec_vsx_st(v15, 0, dst); in vpx_h_predictor_16x16_vsx()
153 vec_vsx_st(v, 0, dst); \
154 vec_vsx_st(v, 16, dst); \
155 dst += stride
157 void vpx_h_predictor_32x32_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_h_predictor_32x32_vsx() argument
241 void vpx_tm_predictor_4x4_vsx(uint8_t *dst, ptrdiff_t stride,
249 d = vec_vsx_ld(0, dst);
252 vec_vsx_st(vec_sel(vec_packsu(val, tmp), d, (uint8x16_t)mask4), 0, dst);
253 dst += stride;
255 d = vec_vsx_ld(0, dst);
258 vec_vsx_st(vec_sel(vec_packsu(val, tmp), d, (uint8x16_t)mask4), 0, dst);
259 dst += stride;
261 d = vec_vsx_ld(0, dst);
264 vec_vsx_st(vec_sel(vec_packsu(val, tmp), d, (uint8x16_t)mask4), 0, dst);
265 dst += stride;
267 d = vec_vsx_ld(0, dst);
270 vec_vsx_st(vec_sel(vec_packsu(val, tmp), d, (uint8x16_t)mask4), 0, dst);
273 void vpx_tm_predictor_8x8_vsx(uint8_t *dst, ptrdiff_t stride,
280 tmp = unpack_to_s16_l(vec_vsx_ld(0, dst));
282 vec_vsx_st(vec_packsu(val, tmp), 0, dst);
283 dst += stride;
285 tmp = unpack_to_s16_l(vec_vsx_ld(0, dst));
287 vec_vsx_st(vec_packsu(val, tmp), 0, dst);
288 dst += stride;
290 tmp = unpack_to_s16_l(vec_vsx_ld(0, dst));
292 vec_vsx_st(vec_packsu(val, tmp), 0, dst);
293 dst += stride;
295 tmp = unpack_to_s16_l(vec_vsx_ld(0, dst));
297 vec_vsx_st(vec_packsu(val, tmp), 0, dst);
298 dst += stride;
300 tmp = unpack_to_s16_l(vec_vsx_ld(0, dst));
302 vec_vsx_st(vec_packsu(val, tmp), 0, dst);
303 dst += stride;
305 tmp = unpack_to_s16_l(vec_vsx_ld(0, dst));
307 vec_vsx_st(vec_packsu(val, tmp), 0, dst);
308 dst += stride;
310 tmp = unpack_to_s16_l(vec_vsx_ld(0, dst));
312 vec_vsx_st(vec_packsu(val, tmp), 0, dst);
313 dst += stride;
315 tmp = unpack_to_s16_l(vec_vsx_ld(0, dst));
317 vec_vsx_st(vec_packsu(val, tmp), 0, dst);
321 static void tm_predictor_16x8(uint8_t *dst, const ptrdiff_t stride, int16x8_t l, in tm_predictor_16x8() argument
328 vec_vsx_st(vec_packsu(vh, vl), 0, dst); in tm_predictor_16x8()
329 dst += stride; in tm_predictor_16x8()
334 vec_vsx_st(vec_packsu(vh, vl), 0, dst); in tm_predictor_16x8()
335 dst += stride; in tm_predictor_16x8()
340 vec_vsx_st(vec_packsu(vh, vl), 0, dst); in tm_predictor_16x8()
341 dst += stride; in tm_predictor_16x8()
346 vec_vsx_st(vec_packsu(vh, vl), 0, dst); in tm_predictor_16x8()
347 dst += stride; in tm_predictor_16x8()
352 vec_vsx_st(vec_packsu(vh, vl), 0, dst); in tm_predictor_16x8()
353 dst += stride; in tm_predictor_16x8()
358 vec_vsx_st(vec_packsu(vh, vl), 0, dst); in tm_predictor_16x8()
359 dst += stride; in tm_predictor_16x8()
364 vec_vsx_st(vec_packsu(vh, vl), 0, dst); in tm_predictor_16x8()
365 dst += stride; in tm_predictor_16x8()
370 vec_vsx_st(vec_packsu(vh, vl), 0, dst); in tm_predictor_16x8()
373 void vpx_tm_predictor_16x16_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_tm_predictor_16x16_vsx() argument
383 tm_predictor_16x8(dst, stride, lh, ah, al, tl); in vpx_tm_predictor_16x16_vsx()
385 dst += stride * 8; in vpx_tm_predictor_16x16_vsx()
387 tm_predictor_16x8(dst, stride, ll, ah, al, tl); in vpx_tm_predictor_16x16_vsx()
390 static INLINE void tm_predictor_32x1(uint8_t *dst, const int16x8_t ls, in tm_predictor_32x1() argument
398 vec_vsx_st(vec_packsu(vh, vl), 0, dst); in tm_predictor_32x1()
401 vec_vsx_st(vec_packsu(vh, vl), 16, dst); in tm_predictor_32x1()
404 static void tm_predictor_32x8(uint8_t *dst, const ptrdiff_t stride, in tm_predictor_32x8() argument
412 tm_predictor_32x1(dst, vec_splat(l, 0), a0h, a0l, a1h, a1l, tl); in tm_predictor_32x8()
413 dst += stride; in tm_predictor_32x8()
415 tm_predictor_32x1(dst, vec_splat(l, 1), a0h, a0l, a1h, a1l, tl); in tm_predictor_32x8()
416 dst += stride; in tm_predictor_32x8()
418 tm_predictor_32x1(dst, vec_splat(l, 2), a0h, a0l, a1h, a1l, tl); in tm_predictor_32x8()
419 dst += stride; in tm_predictor_32x8()
421 tm_predictor_32x1(dst, vec_splat(l, 3), a0h, a0l, a1h, a1l, tl); in tm_predictor_32x8()
422 dst += stride; in tm_predictor_32x8()
424 tm_predictor_32x1(dst, vec_splat(l, 4), a0h, a0l, a1h, a1l, tl); in tm_predictor_32x8()
425 dst += stride; in tm_predictor_32x8()
427 tm_predictor_32x1(dst, vec_splat(l, 5), a0h, a0l, a1h, a1l, tl); in tm_predictor_32x8()
428 dst += stride; in tm_predictor_32x8()
430 tm_predictor_32x1(dst, vec_splat(l, 6), a0h, a0l, a1h, a1l, tl); in tm_predictor_32x8()
431 dst += stride; in tm_predictor_32x8()
433 tm_predictor_32x1(dst, vec_splat(l, 7), a0h, a0l, a1h, a1l, tl); in tm_predictor_32x8()
436 void vpx_tm_predictor_32x32_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_tm_predictor_32x32_vsx() argument
444 tm_predictor_32x8(dst, stride, unpack_to_s16_h(l0), a0, a1, tl); in vpx_tm_predictor_32x32_vsx()
445 dst += stride * 8; in vpx_tm_predictor_32x32_vsx()
447 tm_predictor_32x8(dst, stride, unpack_to_s16_l(l0), a0, a1, tl); in vpx_tm_predictor_32x32_vsx()
448 dst += stride * 8; in vpx_tm_predictor_32x32_vsx()
450 tm_predictor_32x8(dst, stride, unpack_to_s16_h(l1), a0, a1, tl); in vpx_tm_predictor_32x32_vsx()
451 dst += stride * 8; in vpx_tm_predictor_32x32_vsx()
453 tm_predictor_32x8(dst, stride, unpack_to_s16_l(l1), a0, a1, tl); in vpx_tm_predictor_32x32_vsx()
456 static INLINE void dc_fill_predictor_8x8(uint8_t *dst, const ptrdiff_t stride, in dc_fill_predictor_8x8() argument
460 for (i = 0; i < 8; i++, dst += stride) { in dc_fill_predictor_8x8()
461 const uint8x16_t d = vec_vsx_ld(0, dst); in dc_fill_predictor_8x8()
462 vec_vsx_st(xxpermdi(val, d, 1), 0, dst); in dc_fill_predictor_8x8()
466 static INLINE void dc_fill_predictor_16x16(uint8_t *dst, const ptrdiff_t stride, in dc_fill_predictor_16x16() argument
470 for (i = 0; i < 16; i++, dst += stride) { in dc_fill_predictor_16x16()
471 vec_vsx_st(val, 0, dst); in dc_fill_predictor_16x16()
475 void vpx_dc_128_predictor_16x16_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_dc_128_predictor_16x16_vsx() argument
481 dc_fill_predictor_16x16(dst, stride, v128); in vpx_dc_128_predictor_16x16_vsx()
484 static INLINE void dc_fill_predictor_32x32(uint8_t *dst, const ptrdiff_t stride, in dc_fill_predictor_32x32() argument
488 for (i = 0; i < 32; i++, dst += stride) { in dc_fill_predictor_32x32()
489 vec_vsx_st(val, 0, dst); in dc_fill_predictor_32x32()
490 vec_vsx_st(val, 16, dst); in dc_fill_predictor_32x32()
494 void vpx_dc_128_predictor_32x32_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_dc_128_predictor_32x32_vsx() argument
500 dc_fill_predictor_32x32(dst, stride, v128); in vpx_dc_128_predictor_32x32_vsx()
513 void vpx_dc_left_predictor_16x16_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_dc_left_predictor_16x16_vsx() argument
518 dc_fill_predictor_16x16(dst, stride, avg16(left)); in vpx_dc_left_predictor_16x16_vsx()
521 void vpx_dc_top_predictor_16x16_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_dc_top_predictor_16x16_vsx() argument
525 dc_fill_predictor_16x16(dst, stride, avg16(above)); in vpx_dc_top_predictor_16x16_vsx()
541 void vpx_dc_left_predictor_32x32_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_dc_left_predictor_32x32_vsx() argument
546 dc_fill_predictor_32x32(dst, stride, avg32(left)); in vpx_dc_left_predictor_32x32_vsx()
549 void vpx_dc_top_predictor_32x32_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_dc_top_predictor_32x32_vsx() argument
553 dc_fill_predictor_32x32(dst, stride, avg32(above)); in vpx_dc_top_predictor_32x32_vsx()
587 void vpx_dc_predictor_8x8_vsx(uint8_t *dst, ptrdiff_t stride,
589 dc_fill_predictor_8x8(dst, stride, dc_avg8(above, left));
593 void vpx_dc_predictor_16x16_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_dc_predictor_16x16_vsx() argument
595 dc_fill_predictor_16x16(dst, stride, dc_avg16(above, left)); in vpx_dc_predictor_16x16_vsx()
613 void vpx_dc_predictor_32x32_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_dc_predictor_32x32_vsx() argument
615 dc_fill_predictor_32x32(dst, stride, dc_avg32(above, left)); in vpx_dc_predictor_32x32_vsx()
632 void vpx_d45_predictor_8x8_vsx(uint8_t *dst, ptrdiff_t stride,
644 const uint8x16_t d = vec_vsx_ld(0, dst);
645 vec_vsx_st(xxpermdi(row, d, 1), 0, dst);
646 dst += stride;
652 void vpx_d45_predictor_16x16_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_d45_predictor_16x16_vsx() argument
663 vec_vsx_st(row, 0, dst); in vpx_d45_predictor_16x16_vsx()
664 dst += stride; in vpx_d45_predictor_16x16_vsx()
669 void vpx_d45_predictor_32x32_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_d45_predictor_32x32_vsx() argument
684 vec_vsx_st(row0, 0, dst); in vpx_d45_predictor_32x32_vsx()
685 vec_vsx_st(row1, 16, dst); in vpx_d45_predictor_32x32_vsx()
686 dst += stride; in vpx_d45_predictor_32x32_vsx()
694 void vpx_d63_predictor_8x8_vsx(uint8_t *dst, ptrdiff_t stride,
707 const uint8x16_t d0 = vec_vsx_ld(0, dst);
708 const uint8x16_t d1 = vec_vsx_ld(0, dst + stride);
709 vec_vsx_st(xxpermdi(row0, d0, 1), 0, dst);
710 vec_vsx_st(xxpermdi(row1, d1, 1), 0, dst + stride);
711 dst += stride * 2;
718 void vpx_d63_predictor_16x16_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_d63_predictor_16x16_vsx() argument
731 vec_vsx_st(row0, 0, dst); in vpx_d63_predictor_16x16_vsx()
732 vec_vsx_st(row1, 0, dst + stride); in vpx_d63_predictor_16x16_vsx()
733 dst += stride * 2; in vpx_d63_predictor_16x16_vsx()
739 void vpx_d63_predictor_32x32_vsx(uint8_t *dst, ptrdiff_t stride, in vpx_d63_predictor_32x32_vsx() argument
757 vec_vsx_st(row0_0, 0, dst); in vpx_d63_predictor_32x32_vsx()
758 vec_vsx_st(row0_1, 16, dst); in vpx_d63_predictor_32x32_vsx()
759 vec_vsx_st(row1_0, 0, dst + stride); in vpx_d63_predictor_32x32_vsx()
760 vec_vsx_st(row1_1, 16, dst + stride); in vpx_d63_predictor_32x32_vsx()
761 dst += stride * 2; in vpx_d63_predictor_32x32_vsx()