/external/libvpx/libvpx/vp8/encoder/mips/msa/ |
denoising_msa.c | 31 v16u8 src0, src1, src2, src3, src4, src5, src6, src7; 32 v16u8 src8, src9, src10, src11, src12, src13, src14, src15; 33 v16u8 mc_running_avg_y0, running_avg_y, sig0; 34 v16u8 mc_running_avg_y1, running_avg_y1, sig1; 35 v16u8 coeff0, coeff1; 114 adjust0 = (v8i16)__msa_bmnz_v((v16u8)adjust0, (v16u8)diff0, 115 (v16u8)temp2_h); 116 adjust1 = (v8i16)__msa_bmnz_v((v16u8)adjust1, (v16u8)diff1 [all...] |
encodeopt_msa.c | 54 v16u8 zero = { 0 }; 55 v16u8 mask0 = (v16u8)__msa_ldi_b(255); 59 mask0 = (v16u8)__msa_insve_w((v4i32)mask0, 0, (v4i32)zero); 86 diff0 = (v4i32)__msa_bmnz_v(zero, (v16u8)diff0, mask0); 98 diff0 = (v4i32)__msa_bmnz_v(zero, (v16u8)diff0, mask0);
|
/external/libvpx/libvpx/vpx_dsp/mips/ |
loopfilter_4_msa.c | 19 v16u8 mask, hev, flat, thresh, b_limit, limit; 20 v16u8 p3, p2, p1, p0, q3, q2, q1, q0, p1_out, p0_out, q0_out, q1_out; 27 thresh = (v16u8)__msa_fill_b(*thresh_ptr); 28 b_limit = (v16u8)__msa_fill_b(*b_limit_ptr); 29 limit = (v16u8)__msa_fill_b(*limit_ptr); 49 v16u8 mask, hev, flat, thresh0, b_limit0, limit0, thresh1, b_limit1, limit1; 50 v16u8 p3, p2, p1, p0, q3, q2, q1, q0; 55 thresh0 = (v16u8)__msa_fill_b(*thresh0_ptr); 56 thresh1 = (v16u8)__msa_fill_b(*thresh1_ptr); 57 thresh0 = (v16u8)__msa_ilvr_d((v2i64)thresh1, (v2i64)thresh0) [all...] |
loopfilter_8_msa.c | 19 v16u8 mask, hev, flat, thresh, b_limit, limit; 20 v16u8 p3, p2, p1, p0, q3, q2, q1, q0; 21 v16u8 p2_out, p1_out, p0_out, q0_out, q1_out, q2_out; 31 thresh = (v16u8)__msa_fill_b(*thresh_ptr); 32 b_limit = (v16u8)__msa_fill_b(*b_limit_ptr); 33 limit = (v16u8)__msa_fill_b(*limit_ptr); 40 flat = (v16u8)__msa_ilvr_d((v2i64)zero, (v2i64)flat); 62 p2_out = __msa_bmnz_v(p2, (v16u8)p2_filter8, flat); 63 p1_out = __msa_bmnz_v(p1_out, (v16u8)p1_filter8, flat); 64 p0_out = __msa_bmnz_v(p0_out, (v16u8)p0_filter8, flat) [all...] |
loopfilter_16_msa.c | 19 v16u8 p3, p2, p1, p0, q3, q2, q1, q0; 20 v16u8 p2_out, p1_out, p0_out, q0_out, q1_out, q2_out; 21 v16u8 flat, mask, hev, thresh, b_limit, limit; 26 v16u8 zero = { 0 }; 31 thresh = (v16u8)__msa_fill_b(*thresh_ptr); 32 b_limit = (v16u8)__msa_fill_b(*b_limit_ptr); 33 limit = (v16u8)__msa_fill_b(*limit_ptr); 65 p2_out = __msa_bmnz_v(p2, (v16u8)p2_filt8_r, flat); 66 p1_out = __msa_bmnz_v(p1_out, (v16u8)p1_filt8_r, flat); 67 p0_out = __msa_bmnz_v(p0_out, (v16u8)p0_filt8_r, flat) [all...] |
loopfilter_msa.h | 52 q0_out = __msa_xori_b((v16u8)q0_m, 0x80); \ 54 p0_out = __msa_xori_b((v16u8)p0_m, 0x80); \ 57 hev_in = __msa_xori_b((v16u8)hev_in, 0xff); \ 61 q1_out = __msa_xori_b((v16u8)q1_m, 0x80); \ 63 p1_out = __msa_xori_b((v16u8)p1_m, 0x80); \ 109 q0_out = __msa_xori_b((v16u8)q0_m, 0x80); \ 111 p0_out = __msa_xori_b((v16u8)p0_m, 0x80); \ 114 hev_in = __msa_xori_b((v16u8)hev_in, 0xff); \ 118 q1_out = __msa_xori_b((v16u8)q1_m, 0x80); \ 120 p1_out = __msa_xori_b((v16u8)p1_m, 0x80); [all...] |
macros_msa.h | 20 #define LD_UB(...) LD_B(v16u8, __VA_ARGS__) 31 #define ST_UB(...) ST_B(v16u8, __VA_ARGS__) 300 #define LD_UB2(...) LD_B2(v16u8, __VA_ARGS__) 307 #define LD_UB3(...) LD_B3(v16u8, __VA_ARGS__) 313 #define LD_UB4(...) LD_B4(v16u8, __VA_ARGS__) 320 #define LD_UB5(...) LD_B5(v16u8, __VA_ARGS__) 335 #define LD_UB8(...) LD_B8(v16u8, __VA_ARGS__) 404 #define ST_UB2(...) ST_B2(v16u8, __VA_ARGS__) 410 #define ST_UB4(...) ST_B4(v16u8, __VA_ARGS__) 417 #define ST_UB8(...) ST_B8(v16u8, __VA_ARGS__ [all...] |
sub_pixel_variance_msa.c | 28 v16u8 src_l0_m, src_l1_m; \ 54 v16u8 pred, src = { 0 }; 55 v16u8 ref = { 0 }; 88 v16u8 src0, src1, src2, src3; 89 v16u8 ref0, ref1, ref2, ref3; 90 v16u8 pred0, pred1; 123 v16u8 src, ref, pred; 179 v16u8 src0, src1, ref0, ref1, pred0, pred1; 238 v16u8 src0, src1, ref0, ref1, pred0, pred1; 299 v16u8 src0, src1, src2, src3 [all...] |
intrapred_msa.c | 46 v16u8 src0; 59 v16u8 src1, src2; 104 v16u8 src0, src1, src2, src3; 113 src0 = (v16u8)__msa_fill_b(inp0); 114 src1 = (v16u8)__msa_fill_b(inp1); 115 src2 = (v16u8)__msa_fill_b(inp2); 116 src3 = (v16u8)__msa_fill_b(inp3); 127 v16u8 src0, src1, src2, src3; 136 src0 = (v16u8)__msa_fill_b(inp0); 137 src1 = (v16u8)__msa_fill_b(inp1) [all...] |
vpx_convolve_avg_msa.c | 17 v16u8 src0, src1, src2, src3; 18 v16u8 dst0, dst1, dst2, dst3; 60 v16u8 src0, src1, src2, src3; 61 v16u8 dst0, dst1, dst2, dst3; 83 v16u8 src0, src1, src2, src3, src4, src5, src6, src7; 84 v16u8 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; 104 v16u8 src0, src1, src2, src3, src4, src5, src6, src7; 105 v16u8 src8, src9, src10, src11, src12, src13, src14, src15; 106 v16u8 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; 107 v16u8 dst8, dst9, dst10, dst11, dst12, dst13, dst14, dst15 [all...] |
sad_msa.c | 20 #define SAD_INSVE_W4_UB(...) SAD_INSVE_W4(v16u8, __VA_ARGS__) 27 v16u8 src = { 0 }; 28 v16u8 ref = { 0 }; 29 v16u8 diff; 52 v16u8 src0, src1, src2, src3, ref0, ref1, ref2, ref3; 73 v16u8 src0, src1, ref0, ref1; 97 v16u8 src0, src1, ref0, ref1; 134 v16u8 src0, src1, src2, src3; 135 v16u8 ref0, ref1, ref2, ref3; 166 v16u8 src = { 0 } [all...] |
vpx_convolve_msa.h | 94 v16u8 tmp_m; \ 97 tmp_m = __msa_aver_u_b(tmp_m, (v16u8)dst); \ 102 v16u8 tmp_m; \ 104 tmp_m = (v16u8)__msa_pckev_b((v16i8)in0, (v16i8)in1); \ 105 tmp_m = __msa_aver_u_b(tmp_m, (v16u8)dst); \ 111 v16u8 tmp0_m, tmp1_m, tmp2_m, tmp3_m; \
|
vpx_convolve8_avg_vert_msa.c | 23 v16u8 dst0, dst1, dst2, dst3, out; 62 dst0 = (v16u8)__msa_ilvr_d((v2i64)dst2, (v2i64)dst0); 83 v16u8 dst0, dst1, dst2, dst3; 148 v16u8 dst0, dst1, dst2, dst3, tmp0, tmp1, tmp2, tmp3; 265 v16u8 dst0, dst1, dst2, dst3, out, filt0, src2110, src4332; 271 filt0 = (v16u8)__msa_splati_h(filt, 0); 281 dst0 = (v16u8)__msa_ilvr_d((v2i64)dst1, (v2i64)dst0); 288 out = (v16u8)__msa_pckev_b((v16i8)tmp1, (v16i8)tmp0); 299 v16u8 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; 302 v16u8 src2110, src4332, src6554, src8776, filt0 [all...] |
vpx_convolve8_msa.c | 31 v16u8 mask0, mask1, mask2, mask3, out; 105 v16u8 mask0, mask1, mask2, mask3, vec0, vec1; 238 v16u8 filt_vt, filt_hz, vec0, vec1, res0, res1; 245 filt_hz = (v16u8)__msa_splati_h((v8i16)filt, 0); 248 filt_vt = (v16u8)__msa_splati_h((v8i16)filt, 0); 270 v16u8 filt_hz, filt_vt, vec0, vec1, vec2, vec3; 278 filt_hz = (v16u8)__msa_splati_h((v8i16)filt, 0); 281 filt_vt = (v16u8)__msa_splati_h((v8i16)filt, 0); 326 v16u8 filt_hz, filt_vt, vec0, vec1, vec2, vec3; 334 filt_hz = (v16u8)__msa_splati_h(filt, 0) [all...] |
variance_msa.c | 15 v16u8 src_l0_m, src_l1_m; \ 24 v16u8 src_l0_m, src_l1_m; \ 46 v16u8 src = { 0 }; 47 v16u8 ref = { 0 }; 72 v16u8 src0, src1, src2, src3; 73 v16u8 ref0, ref1, ref2, ref3; 99 v16u8 src, ref; 139 v16u8 src0, src1, ref0, ref1; 183 v16u8 src0, src1, ref0, ref1; 229 v16u8 src0, src1, src2, src3 [all...] |
vpx_convolve8_avg_msa.c | 24 v16u8 dst0, dst1, dst2, dst3, mask0, mask1, mask2, mask3, tmp0, tmp1; 108 v16u8 dst0, dst1, dst2, dst3, mask0, mask1, mask2, mask3; 252 v16u8 filt_hz, filt_vt, vec0, vec1; 253 v16u8 dst0, dst1, dst2, dst3, res0, res1; 260 filt_hz = (v16u8)__msa_splati_h((v8i16)filt, 0); 263 filt_vt = (v16u8)__msa_splati_h((v8i16)filt, 0); 290 v16u8 filt_hz, filt_vt, vec0, vec1, vec2, vec3, res0, res1, res2, res3; 291 v16u8 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; 300 filt_hz = (v16u8)__msa_splati_h(filt, 0); 303 filt_vt = (v16u8)__msa_splati_h(filt, 0) [all...] |
vpx_convolve8_vert_msa.c | 23 v16u8 out; 73 v16u8 tmp0, tmp1; 129 v16u8 tmp0, tmp1, tmp2, tmp3; 210 v16u8 tmp0, tmp1, tmp2, tmp3; 305 v16u8 filt0; 310 filt0 = (v16u8)__msa_splati_h(filt, 0); 331 v16u8 filt0; 335 filt0 = (v16u8)__msa_splati_h(filt, 0); 370 v16u8 src0, src1, src2, src3, src4, vec0, vec1, vec2, vec3, filt0; 377 filt0 = (v16u8)__msa_splati_h(filt, 0) [all...] |
vpx_convolve8_horiz_msa.c | 18 v16u8 mask0, mask1, mask2, mask3, out; 48 v16u8 mask0, mask1, mask2, mask3, out; 94 v16u8 mask0, mask1, mask2, mask3, tmp0, tmp1; 125 v16u8 mask0, mask1, mask2, mask3, tmp0, tmp1; 170 v16u8 mask0, mask1, mask2, mask3, out; 208 v16u8 mask0, mask1, mask2, mask3, out; 266 v16u8 mask0, mask1, mask2, mask3, out; 321 v16u8 filt0, vec0, vec1, res0, res1; 328 filt0 = (v16u8) __msa_splati_h((v8i16) filt, 0); 341 v16u8 vec0, vec1, vec2, vec3, filt0 [all...] |
vpx_convolve8_avg_horiz_msa.c | 21 v16u8 dst0, dst1, dst2, dst3, res2, res3; 22 v16u8 mask0, mask1, mask2, mask3; 56 v16u8 mask0, mask1, mask2, mask3, res0, res1, res2, res3; 57 v16u8 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; 115 v16u8 mask0, mask1, mask2, mask3, dst0, dst1, dst2, dst3; 153 v16u8 mask0, mask1, mask2, mask3, dst0, dst1; 209 v16u8 dst1, dst2, mask0, mask1, mask2, mask3; 266 v16u8 dst1, dst2, mask0, mask1, mask2, mask3; 326 v16u8 filt0, dst0, dst1, dst2, dst3, vec0, vec1, res0, res1; 333 filt0 = (v16u8)__msa_splati_h((v8i16)filt, 0) [all...] |
/external/libvpx/libvpx/vp8/common/mips/msa/ |
loopfilter_filters_msa.c | 17 v16u8 p1_a_sub_q1, p0_a_sub_q0; \ 21 p1_a_sub_q1 = (v16u8)__msa_srli_b((v16i8)p1_a_sub_q1, 1); \ 24 mask = ((v16u8)mask <= b_limit); \ 71 q0_in_out = __msa_xori_b((v16u8)q0_m, 0x80); \ 73 p0_in_out = __msa_xori_b((v16u8)p0_m, 0x80); \ 76 hev_in = __msa_xori_b((v16u8)hev_in, 0xff); \ 80 q1_in_out = __msa_xori_b((v16u8)q1_m, 0x80); \ 82 p1_in_out = __msa_xori_b((v16u8)p1_m, 0x80); \ 128 q0_in = __msa_xori_b((v16u8)q0_m, 0x80); \ 129 p0_in = __msa_xori_b((v16u8)p0_m, 0x80); [all...] |
postproc_msa.c | 84 out0 = (v16u8)temp6; \ 85 out2 = (v16u8)temp7; \ 86 out4 = (v16u8)temp8; \ 87 out6 = (v16u8)temp9; \ 88 out9 = (v16u8)__msa_ilvl_d((v2i64)out8, (v2i64)out8); \ 89 out11 = (v16u8)__msa_ilvl_d((v2i64)out10, (v2i64)out10); \ 90 out13 = (v16u8)__msa_ilvl_d((v2i64)out12, (v2i64)out12); \ 91 out15 = (v16u8)__msa_ilvl_d((v2i64)out14, (v2i64)out14); \ 92 out1 = (v16u8)__msa_ilvl_d((v2i64)out0, (v2i64)out0); \ 93 out3 = (v16u8)__msa_ilvl_d((v2i64)out2, (v2i64)out2); [all...] |
vp8_macros_msa.h | 20 #define LD_UB(...) LD_B(v16u8, __VA_ARGS__) 32 #define ST_UB(...) ST_B(v16u8, __VA_ARGS__) 291 #define LD_UB2(...) LD_B2(v16u8, __VA_ARGS__) 299 #define LD_UB3(...) LD_B3(v16u8, __VA_ARGS__) 307 #define LD_UB4(...) LD_B4(v16u8, __VA_ARGS__) 315 #define LD_UB5(...) LD_B5(v16u8, __VA_ARGS__) 324 #define LD_UB8(...) LD_B8(v16u8, __VA_ARGS__) 368 #define ST_UB2(...) ST_B2(v16u8, __VA_ARGS__) 375 #define ST_UB4(...) ST_B4(v16u8, __VA_ARGS__) 384 #define ST_UB8(...) ST_B8(v16u8, __VA_ARGS__ [all...] |
bilinear_filter_msa.c | 42 v16u8 filt0, vec0, vec1, res0, res1; 48 filt0 = (v16u8)__msa_splati_h((v8i16)filt, 0); 62 v16u8 vec0, vec1, vec2, vec3, filt0; 70 filt0 = (v16u8)__msa_splati_h((v8i16)filt, 0); 103 v16u8 filt0; 110 filt0 = (v16u8)__msa_splati_h((v8i16)filt, 0); 126 v16u8 filt0; 133 filt0 = (v16u8)__msa_splati_h((v8i16)filt, 0); 206 v16u8 filt0, vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7; 214 filt0 = (v16u8)__msa_splati_h((v8i16)filt, 0) [all...] |
copymem_msa.c | 40 v16u8 src0, src1, src2, src3, src4, src5, src6, src7; 41 v16u8 src8, src9, src10, src11, src12, src13, src14, src15;
|
/external/libvpx/libvpx/vp9/encoder/mips/msa/ |
vp9_avg_msa.c | 16 v16u8 src0, src1, src2, src3, src4, src5, src6, src7; 39 v16u8 vec = { 0 };
|