/external/libvpx/libvpx/vpx_dsp/arm/ |
D | vpx_convolve_avg_neon.c | 30 uint8x16_t q0u8, q1u8, q2u8, q3u8, q8u8, q9u8, q10u8, q11u8; in vpx_convolve_avg_neon() local 38 q1u8 = vld1q_u8(src + 16); in vpx_convolve_avg_neon() 49 q1u8 = vrhaddq_u8(q1u8, q9u8); in vpx_convolve_avg_neon() 54 vst1q_u8(dst + 16, q1u8); in vpx_convolve_avg_neon() 62 q1u8 = vld1q_u8(src + 16); in vpx_convolve_avg_neon() 75 q1u8 = vrhaddq_u8(q1u8, q9u8); in vpx_convolve_avg_neon() 80 vst1q_u8(dst + 16, q1u8); in vpx_convolve_avg_neon() 90 q1u8 = vld1q_u8(src); in vpx_convolve_avg_neon() 98 q1u8 = vrhaddq_u8(q1u8, q3u8); in vpx_convolve_avg_neon() 102 vst1q_u8(dst, q1u8); in vpx_convolve_avg_neon() [all …]
|
D | vpx_convolve_copy_neon.c | 28 uint8x16_t q0u8, q1u8, q2u8, q3u8; in vpx_convolve_copy_neon() local 35 q1u8 = vld1q_u8(src + 16); in vpx_convolve_copy_neon() 41 vst1q_u8(dst + 16, q1u8); in vpx_convolve_copy_neon() 49 q1u8 = vld1q_u8(src + 16); in vpx_convolve_copy_neon() 56 vst1q_u8(dst + 16, q1u8); in vpx_convolve_copy_neon() 66 q1u8 = vld1q_u8(src); in vpx_convolve_copy_neon() 71 vst1q_u8(dst, q1u8); in vpx_convolve_copy_neon()
|
D | intrapred_neon.c | 447 uint8x16_t q1u8 = vdupq_n_u8(0); in vpx_v_predictor_32x32_neon() local 451 q1u8 = vld1q_u8(above + 16); in vpx_v_predictor_32x32_neon() 454 vst1q_u8(dst + 16, q1u8); in vpx_v_predictor_32x32_neon() 517 uint8x16_t q1u8 = vdupq_n_u8(0); in vpx_h_predictor_16x16_neon() local 520 q1u8 = vld1q_u8(left); in vpx_h_predictor_16x16_neon() 521 d2u8 = vget_low_u8(q1u8); in vpx_h_predictor_16x16_neon() 522 for (j = 0; j < 2; j++, d2u8 = vget_high_u8(q1u8)) { in vpx_h_predictor_16x16_neon() 555 uint8x16_t q1u8 = vdupq_n_u8(0); in vpx_h_predictor_32x32_neon() local 559 q1u8 = vld1q_u8(left); in vpx_h_predictor_32x32_neon() 560 d2u8 = vget_low_u8(q1u8); in vpx_h_predictor_32x32_neon() [all …]
|
D | vpx_convolve8_avg_neon.c | 62 uint8x16_t q1u8, q3u8, q12u8, q13u8, q14u8, q15u8; in vpx_convolve8_avg_horiz_neon() local 205 q1u8 = vcombine_u8(d0x2u8.val[0], d0x2u8.val[1]); in vpx_convolve8_avg_horiz_neon() 208 q1u8 = vrhaddq_u8(q1u8, q3u8); in vpx_convolve8_avg_horiz_neon() 210 d2u32 = vreinterpret_u32_u8(vget_low_u8(q1u8)); in vpx_convolve8_avg_horiz_neon() 211 d3u32 = vreinterpret_u32_u8(vget_high_u8(q1u8)); in vpx_convolve8_avg_horiz_neon() 251 uint8x16_t q1u8, q3u8; in vpx_convolve8_avg_vert_neon() local 348 q1u8 = vcombine_u8(d2u8, d3u8); in vpx_convolve8_avg_vert_neon() 351 q1u8 = vrhaddq_u8(q1u8, q3u8); in vpx_convolve8_avg_vert_neon() 353 d2u32 = vreinterpret_u32_u8(vget_low_u8(q1u8)); in vpx_convolve8_avg_vert_neon() 354 d3u32 = vreinterpret_u32_u8(vget_high_u8(q1u8)); in vpx_convolve8_avg_vert_neon()
|
D | loopfilter_16_neon.c | 33 uint8x16_t q1u8, q2u8, q11u8, q12u8, q13u8, q14u8, q15u8; in loop_filter_neon_16() local 90 q1u8 = vandq_u8(vreinterpretq_u8_s8(q1s8), q14u8); in loop_filter_neon_16() 93 q1s8 = vreinterpretq_s8_u8(q1u8); in loop_filter_neon_16() 103 q1u8 = vandq_u8(vreinterpretq_u8_s8(q1s8), q15u8); in loop_filter_neon_16() 104 q1s8 = vreinterpretq_s8_u8(q1u8); in loop_filter_neon_16()
|
D | variance_neon.c | 166 uint8x16_t q0u8, q1u8, q2u8, q3u8; in vpx_variance16x8_neon() local 178 q1u8 = vld1q_u8(src_ptr); in vpx_variance16x8_neon() 190 q13u16 = vsubl_u8(vget_low_u8(q1u8), vget_low_u8(q3u8)); in vpx_variance16x8_neon() 191 q14u16 = vsubl_u8(vget_high_u8(q1u8), vget_high_u8(q3u8)); in vpx_variance16x8_neon() 309 uint8x16_t q0u8, q1u8, q2u8, q3u8; in vpx_mse16x16_neon() local 322 q1u8 = vld1q_u8(src_ptr); in vpx_mse16x16_neon() 331 q13u16 = vsubl_u8(vget_low_u8(q1u8), vget_low_u8(q3u8)); in vpx_mse16x16_neon() 332 q14u16 = vsubl_u8(vget_high_u8(q1u8), vget_high_u8(q3u8)); in vpx_mse16x16_neon()
|
/external/libvpx/libvpx/vp8/common/arm/neon/ |
D | bilinearpredict_neon.c | 33 uint8x16_t q1u8, q2u8; in vp8_bilinear_predict4x4_neon() local 64 q1u8 = vcombine_u8(d2u8, d3u8); in vp8_bilinear_predict4x4_neon() 70 q4u64 = vshrq_n_u64(vreinterpretq_u64_u8(q1u8), 8); in vp8_bilinear_predict4x4_neon() 74 d0u32x2 = vzip_u32(vreinterpret_u32_u8(vget_low_u8(q1u8)), in vp8_bilinear_predict4x4_neon() 75 vreinterpret_u32_u8(vget_high_u8(q1u8))); in vp8_bilinear_predict4x4_neon() 141 uint8x16_t q1u8, q2u8, q3u8, q4u8, q5u8; in vp8_bilinear_predict8x4_neon() local 152 q1u8 = vld1q_u8(src_ptr); src_ptr += src_pixels_per_line; in vp8_bilinear_predict8x4_neon() 161 q6u16 = vmull_u8(vget_low_u8(q1u8), d0u8); in vp8_bilinear_predict8x4_neon() 167 d3u8 = vext_u8(vget_low_u8(q1u8), vget_high_u8(q1u8), 1); in vp8_bilinear_predict8x4_neon() 228 uint8x16_t q1u8, q2u8, q3u8, q4u8, q5u8; in vp8_bilinear_predict8x8_neon() local [all …]
|
D | mbloopfilter_neon.c | 32 uint8x16_t q0u8, q1u8, q11u8, q12u8, q13u8, q14u8, q15u8; in vp8_mbloop_filter_neon() local 43 q1u8 = vabdq_u8(q9, q8); in vp8_mbloop_filter_neon() 48 q1u8 = vmaxq_u8(q1u8, q0u8); in vp8_mbloop_filter_neon() 56 q15u8 = vmaxq_u8(q15u8, q1u8); in vp8_mbloop_filter_neon() 60 q1u8 = vabdq_u8(q5, q8); in vp8_mbloop_filter_neon() 73 q1u8 = vshrq_n_u8(q1u8, 1); in vp8_mbloop_filter_neon() 74 q12u8 = vqaddq_u8(q12u8, q1u8); in vp8_mbloop_filter_neon() 145 q1u8 = vdupq_n_u8(0x80); in vp8_mbloop_filter_neon() 146 *q9r = veorq_u8(vreinterpretq_u8_s8(q11s8), q1u8); in vp8_mbloop_filter_neon() 147 *q8r = veorq_u8(vreinterpretq_u8_s8(q13s8), q1u8); in vp8_mbloop_filter_neon() [all …]
|
D | vp8_loopfilter_neon.c | 31 uint8x16_t q0u8, q1u8, q2u8, q11u8, q12u8, q13u8, q14u8, q15u8; in vp8_loop_filter_neon() local 90 q1u8 = vandq_u8(vreinterpretq_u8_s8(q1s8), q14u8); in vp8_loop_filter_neon() 93 q1s8 = vreinterpretq_s8_u8(q1u8); in vp8_loop_filter_neon() 102 q1u8 = vandq_u8(vreinterpretq_u8_s8(q1s8), q15u8); in vp8_loop_filter_neon() 103 q1s8 = vreinterpretq_s8_u8(q1u8); in vp8_loop_filter_neon()
|