Lines Matching refs:v16i8
33 src0 = (v16u8)__msa_ld_b((v16i8*)src_argb, 0); in ScaleARGBRowDown2_MSA()
34 src1 = (v16u8)__msa_ld_b((v16i8*)src_argb, 16); in ScaleARGBRowDown2_MSA()
51 src0 = (v16u8)__msa_ld_b((v16i8*)src_argb, 0); in ScaleARGBRowDown2Linear_MSA()
52 src1 = (v16u8)__msa_ld_b((v16i8*)src_argb, 16); in ScaleARGBRowDown2Linear_MSA()
71 v16i8 shuffler = {0, 4, 1, 5, 2, 6, 3, 7, 8, 12, 9, 13, 10, 14, 11, 15}; in ScaleARGBRowDown2Box_MSA()
74 src0 = (v16u8)__msa_ld_b((v16i8*)s, 0); in ScaleARGBRowDown2Box_MSA()
75 src1 = (v16u8)__msa_ld_b((v16i8*)s, 16); in ScaleARGBRowDown2Box_MSA()
76 src2 = (v16u8)__msa_ld_b((v16i8*)t, 0); in ScaleARGBRowDown2Box_MSA()
77 src3 = (v16u8)__msa_ld_b((v16i8*)t, 16); in ScaleARGBRowDown2Box_MSA()
78 vec0 = (v16u8)__msa_vshf_b(shuffler, (v16i8)src0, (v16i8)src0); in ScaleARGBRowDown2Box_MSA()
79 vec1 = (v16u8)__msa_vshf_b(shuffler, (v16i8)src1, (v16i8)src1); in ScaleARGBRowDown2Box_MSA()
80 vec2 = (v16u8)__msa_vshf_b(shuffler, (v16i8)src2, (v16i8)src2); in ScaleARGBRowDown2Box_MSA()
81 vec3 = (v16u8)__msa_vshf_b(shuffler, (v16i8)src3, (v16i8)src3); in ScaleARGBRowDown2Box_MSA()
90 dst0 = (v16u8)__msa_pckev_b((v16i8)reg1, (v16i8)reg0); in ScaleARGBRowDown2Box_MSA()
153 vec0 = (v16u8)__msa_ilvr_b((v16i8)src2, (v16i8)src0); in ScaleARGBRowDownEvenBox_MSA()
154 vec1 = (v16u8)__msa_ilvr_b((v16i8)src3, (v16i8)src1); in ScaleARGBRowDownEvenBox_MSA()
155 vec2 = (v16u8)__msa_ilvl_b((v16i8)src2, (v16i8)src0); in ScaleARGBRowDownEvenBox_MSA()
156 vec3 = (v16u8)__msa_ilvl_b((v16i8)src3, (v16i8)src1); in ScaleARGBRowDownEvenBox_MSA()
169 dst0 = (v16u8)__msa_pckev_b((v16i8)reg5, (v16i8)reg4); in ScaleARGBRowDownEvenBox_MSA()
186 src0 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 0); in ScaleRowDown2_MSA()
187 src1 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 16); in ScaleRowDown2_MSA()
188 src2 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 32); in ScaleRowDown2_MSA()
189 src3 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 48); in ScaleRowDown2_MSA()
190 dst0 = (v16u8)__msa_pckod_b((v16i8)src1, (v16i8)src0); in ScaleRowDown2_MSA()
191 dst1 = (v16u8)__msa_pckod_b((v16i8)src3, (v16i8)src2); in ScaleRowDown2_MSA()
207 src0 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 0); in ScaleRowDown2Linear_MSA()
208 src1 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 16); in ScaleRowDown2Linear_MSA()
209 src2 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 32); in ScaleRowDown2Linear_MSA()
210 src3 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 48); in ScaleRowDown2Linear_MSA()
211 vec0 = (v16u8)__msa_pckev_b((v16i8)src1, (v16i8)src0); in ScaleRowDown2Linear_MSA()
212 vec2 = (v16u8)__msa_pckev_b((v16i8)src3, (v16i8)src2); in ScaleRowDown2Linear_MSA()
213 vec1 = (v16u8)__msa_pckod_b((v16i8)src1, (v16i8)src0); in ScaleRowDown2Linear_MSA()
214 vec3 = (v16u8)__msa_pckod_b((v16i8)src3, (v16i8)src2); in ScaleRowDown2Linear_MSA()
234 src0 = (v16u8)__msa_ld_b((v16i8*)s, 0); in ScaleRowDown2Box_MSA()
235 src1 = (v16u8)__msa_ld_b((v16i8*)s, 16); in ScaleRowDown2Box_MSA()
236 src2 = (v16u8)__msa_ld_b((v16i8*)s, 32); in ScaleRowDown2Box_MSA()
237 src3 = (v16u8)__msa_ld_b((v16i8*)s, 48); in ScaleRowDown2Box_MSA()
238 src4 = (v16u8)__msa_ld_b((v16i8*)t, 0); in ScaleRowDown2Box_MSA()
239 src5 = (v16u8)__msa_ld_b((v16i8*)t, 16); in ScaleRowDown2Box_MSA()
240 src6 = (v16u8)__msa_ld_b((v16i8*)t, 32); in ScaleRowDown2Box_MSA()
241 src7 = (v16u8)__msa_ld_b((v16i8*)t, 48); in ScaleRowDown2Box_MSA()
254 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ScaleRowDown2Box_MSA()
255 dst1 = (v16u8)__msa_pckev_b((v16i8)vec3, (v16i8)vec2); in ScaleRowDown2Box_MSA()
272 src0 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 0); in ScaleRowDown4_MSA()
273 src1 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 16); in ScaleRowDown4_MSA()
274 src2 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 32); in ScaleRowDown4_MSA()
275 src3 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 48); in ScaleRowDown4_MSA()
276 vec0 = (v16u8)__msa_pckev_b((v16i8)src1, (v16i8)src0); in ScaleRowDown4_MSA()
277 vec1 = (v16u8)__msa_pckev_b((v16i8)src3, (v16i8)src2); in ScaleRowDown4_MSA()
278 dst0 = (v16u8)__msa_pckod_b((v16i8)vec1, (v16i8)vec0); in ScaleRowDown4_MSA()
299 src0 = (v16u8)__msa_ld_b((v16i8*)s, 0); in ScaleRowDown4Box_MSA()
300 src1 = (v16u8)__msa_ld_b((v16i8*)s, 16); in ScaleRowDown4Box_MSA()
301 src2 = (v16u8)__msa_ld_b((v16i8*)s, 32); in ScaleRowDown4Box_MSA()
302 src3 = (v16u8)__msa_ld_b((v16i8*)s, 48); in ScaleRowDown4Box_MSA()
303 src4 = (v16u8)__msa_ld_b((v16i8*)t0, 0); in ScaleRowDown4Box_MSA()
304 src5 = (v16u8)__msa_ld_b((v16i8*)t0, 16); in ScaleRowDown4Box_MSA()
305 src6 = (v16u8)__msa_ld_b((v16i8*)t0, 32); in ScaleRowDown4Box_MSA()
306 src7 = (v16u8)__msa_ld_b((v16i8*)t0, 48); in ScaleRowDown4Box_MSA()
315 src0 = (v16u8)__msa_ld_b((v16i8*)t1, 0); in ScaleRowDown4Box_MSA()
316 src1 = (v16u8)__msa_ld_b((v16i8*)t1, 16); in ScaleRowDown4Box_MSA()
317 src2 = (v16u8)__msa_ld_b((v16i8*)t1, 32); in ScaleRowDown4Box_MSA()
318 src3 = (v16u8)__msa_ld_b((v16i8*)t1, 48); in ScaleRowDown4Box_MSA()
319 src4 = (v16u8)__msa_ld_b((v16i8*)t2, 0); in ScaleRowDown4Box_MSA()
320 src5 = (v16u8)__msa_ld_b((v16i8*)t2, 16); in ScaleRowDown4Box_MSA()
321 src6 = (v16u8)__msa_ld_b((v16i8*)t2, 32); in ScaleRowDown4Box_MSA()
322 src7 = (v16u8)__msa_ld_b((v16i8*)t2, 48); in ScaleRowDown4Box_MSA()
341 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ScaleRowDown4Box_MSA()
359 v16i8 mask = {0, 3, 6, 8, 11, 14, 16, 19, 22, 24, 27, 30, 0, 0, 0, 0}; in ScaleRowDown38_MSA()
366 src0 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 0); in ScaleRowDown38_MSA()
367 src1 = (v16u8)__msa_ld_b((v16i8*)src_ptr, 16); in ScaleRowDown38_MSA()
368 vec0 = (v16u8)__msa_vshf_b(mask, (v16i8)src1, (v16i8)src0); in ScaleRowDown38_MSA()
392 v16i8 dst_mask = {0, 2, 16, 4, 6, 18, 8, 10, 20, 12, 14, 22, 0, 0, 0, 0}; in ScaleRowDown38_2_Box_MSA()
400 src0 = (v16u8)__msa_ld_b((v16i8*)s, 0); in ScaleRowDown38_2_Box_MSA()
401 src1 = (v16u8)__msa_ld_b((v16i8*)s, 16); in ScaleRowDown38_2_Box_MSA()
402 src2 = (v16u8)__msa_ld_b((v16i8*)t, 0); in ScaleRowDown38_2_Box_MSA()
403 src3 = (v16u8)__msa_ld_b((v16i8*)t, 16); in ScaleRowDown38_2_Box_MSA()
404 vec0 = (v8u16)__msa_ilvr_b((v16i8)src2, (v16i8)src0); in ScaleRowDown38_2_Box_MSA()
405 vec1 = (v8u16)__msa_ilvl_b((v16i8)src2, (v16i8)src0); in ScaleRowDown38_2_Box_MSA()
406 vec2 = (v8u16)__msa_ilvr_b((v16i8)src3, (v16i8)src1); in ScaleRowDown38_2_Box_MSA()
407 vec3 = (v8u16)__msa_ilvl_b((v16i8)src3, (v16i8)src1); in ScaleRowDown38_2_Box_MSA()
436 out = (v16u8)__msa_vshf_b(dst_mask, (v16i8)vec1, (v16i8)vec0); in ScaleRowDown38_2_Box_MSA()
462 v16i8 dst_mask = {0, 2, 16, 4, 6, 18, 8, 10, 20, 12, 14, 22, 0, 0, 0, 0}; in ScaleRowDown38_3_Box_MSA()
470 src0 = (v16u8)__msa_ld_b((v16i8*)s, 0); in ScaleRowDown38_3_Box_MSA()
471 src1 = (v16u8)__msa_ld_b((v16i8*)s, 16); in ScaleRowDown38_3_Box_MSA()
472 src2 = (v16u8)__msa_ld_b((v16i8*)t0, 0); in ScaleRowDown38_3_Box_MSA()
473 src3 = (v16u8)__msa_ld_b((v16i8*)t0, 16); in ScaleRowDown38_3_Box_MSA()
474 src4 = (v16u8)__msa_ld_b((v16i8*)t1, 0); in ScaleRowDown38_3_Box_MSA()
475 src5 = (v16u8)__msa_ld_b((v16i8*)t1, 16); in ScaleRowDown38_3_Box_MSA()
476 vec0 = (v8u16)__msa_ilvr_b((v16i8)src2, (v16i8)src0); in ScaleRowDown38_3_Box_MSA()
477 vec1 = (v8u16)__msa_ilvl_b((v16i8)src2, (v16i8)src0); in ScaleRowDown38_3_Box_MSA()
478 vec2 = (v8u16)__msa_ilvr_b((v16i8)src3, (v16i8)src1); in ScaleRowDown38_3_Box_MSA()
479 vec3 = (v8u16)__msa_ilvl_b((v16i8)src3, (v16i8)src1); in ScaleRowDown38_3_Box_MSA()
480 vec4 = (v8u16)__msa_ilvr_b((v16i8)zero, (v16i8)src4); in ScaleRowDown38_3_Box_MSA()
481 vec5 = (v8u16)__msa_ilvl_b((v16i8)zero, (v16i8)src4); in ScaleRowDown38_3_Box_MSA()
482 vec6 = (v8u16)__msa_ilvr_b((v16i8)zero, (v16i8)src5); in ScaleRowDown38_3_Box_MSA()
483 vec7 = (v8u16)__msa_ilvl_b((v16i8)zero, (v16i8)src5); in ScaleRowDown38_3_Box_MSA()
516 out = (v16u8)__msa_vshf_b(dst_mask, (v16i8)vec1, (v16i8)vec0); in ScaleRowDown38_3_Box_MSA()
532 v16i8 zero = {0}; in ScaleAddRow_MSA()
540 dst0 += (v8u16)__msa_ilvr_b(zero, (v16i8)src0); in ScaleAddRow_MSA()
541 dst1 += (v8u16)__msa_ilvl_b(zero, (v16i8)src0); in ScaleAddRow_MSA()