/external/libvpx/libvpx/vp8/common/arm/neon/ |
D | sixtappredict_neon.c | 37 uint16x8_t q8u16, q9u16, q10u16, q11u16, q12u16; in vp8_sixtap_predict4x4_neon() local 178 q8u16 = vmull_u8(vreinterpret_u8_u32(d1u32x2.val[0]), d5u8); in vp8_sixtap_predict4x4_neon() 191 q8u16 = vmlal_u8(q8u16, vreinterpret_u8_u32(d1u32x2.val[0]), d0u8); in vp8_sixtap_predict4x4_neon() 200 q8u16 = vmlsl_u8(q8u16, vreinterpret_u8_u32(d1u32x2.val[0]), d1u8); in vp8_sixtap_predict4x4_neon() 209 q8u16 = vmlsl_u8(q8u16, vreinterpret_u8_u32(d1u32x2.val[0]), d4u8); in vp8_sixtap_predict4x4_neon() 218 q8u16 = vmlal_u8(q8u16, vreinterpret_u8_u32(d1u32x2.val[0]), d2u8); in vp8_sixtap_predict4x4_neon() 228 q8s16 = vreinterpretq_s16_u16(q8u16); in vp8_sixtap_predict4x4_neon() 274 q8u16 = vmull_u8(vreinterpret_u8_u32(d1u32x2.val[0]), d5u8); in vp8_sixtap_predict4x4_neon() 287 q8u16 = vmlal_u8(q8u16, vreinterpret_u8_u32(d1u32x2.val[0]), d0u8); in vp8_sixtap_predict4x4_neon() 298 q8u16 = vmlsl_u8(q8u16, vreinterpret_u8_u32(d1u32x2.val[0]), d1u8); in vp8_sixtap_predict4x4_neon() [all …]
|
D | bilinearpredict_neon.c | 35 uint16x8_t q7u16, q8u16, q9u16; in vp8_bilinear_predict4x4_neon() local 84 q8u16 = vmull_u8(vreinterpret_u8_u32(d1u32x2.val[0]), d0u8); in vp8_bilinear_predict4x4_neon() 88 q8u16 = vmlal_u8(q8u16, vreinterpret_u8_u32(d3u32x2.val[0]), d1u8); in vp8_bilinear_predict4x4_neon() 92 d29u8 = vqrshrn_n_u16(q8u16, 7); in vp8_bilinear_predict4x4_neon() 143 uint16x8_t q6u16, q7u16, q8u16, q9u16, q10u16; in vp8_bilinear_predict8x4_neon() local 163 q8u16 = vmull_u8(vget_low_u8(q3u8), d0u8); in vp8_bilinear_predict8x4_neon() 175 q8u16 = vmlal_u8(q8u16, d7u8, d1u8); in vp8_bilinear_predict8x4_neon() 181 d24u8 = vqrshrn_n_u16(q8u16, 7); in vp8_bilinear_predict8x4_neon() 230 uint16x8_t q6u16, q7u16, q8u16, q9u16, q10u16; in vp8_bilinear_predict8x8_neon() local 253 q8u16 = vmull_u8(vget_low_u8(q3u8), d0u8); in vp8_bilinear_predict8x8_neon() [all …]
|
/external/libvpx/libvpx/vpx_dsp/arm/ |
D | vpx_convolve8_neon.c | 67 uint16x8_t q1u16, q2u16, q8u16, q9u16, q10u16, q11u16, q12u16, q13u16; in vpx_convolve8_horiz_neon() local 108 q8u16 = vmovl_u8(d0x2u8.val[0]); in vpx_convolve8_horiz_neon() 113 d16u16 = vget_low_u16(q8u16); in vpx_convolve8_horiz_neon() 114 d17u16 = vget_high_u16(q8u16); in vpx_convolve8_horiz_neon() 117 q8u16 = vcombine_u16(d16u16, d18u16); // vswp 17 18 in vpx_convolve8_horiz_neon() 159 d16s16 = vreinterpret_s16_u16(vget_low_u16(q8u16)); in vpx_convolve8_horiz_neon() 160 d17s16 = vreinterpret_s16_u16(vget_high_u16(q8u16)); in vpx_convolve8_horiz_neon() 210 q8u16 = q9u16; in vpx_convolve8_horiz_neon() 240 uint16x8_t q1u16, q2u16, q8u16, q9u16, q10u16, q11u16, q12u16, q13u16; in vpx_convolve8_vert_neon() local 264 q8u16 = vmovl_u8(vreinterpret_u8_u32(d16u32)); in vpx_convolve8_vert_neon() [all …]
|
D | vpx_convolve8_avg_neon.c | 67 uint16x8_t q1u16, q2u16, q8u16, q9u16, q10u16, q11u16, q12u16, q13u16; in vpx_convolve8_avg_horiz_neon() local 105 q8u16 = vmovl_u8(d0x2u8.val[0]); in vpx_convolve8_avg_horiz_neon() 111 d16u16 = vget_low_u16(q8u16); in vpx_convolve8_avg_horiz_neon() 112 d17u16 = vget_high_u16(q8u16); in vpx_convolve8_avg_horiz_neon() 115 q8u16 = vcombine_u16(d16u16, d18u16); // vswp 17 18 in vpx_convolve8_avg_horiz_neon() 166 d16s16 = vreinterpret_s16_u16(vget_low_u16(q8u16)); in vpx_convolve8_avg_horiz_neon() 167 d17s16 = vreinterpret_s16_u16(vget_high_u16(q8u16)); in vpx_convolve8_avg_horiz_neon() 222 q8u16 = q9u16; in vpx_convolve8_avg_horiz_neon() 256 uint16x8_t q1u16, q2u16, q8u16, q9u16, q10u16, q11u16, q12u16, q13u16; in vpx_convolve8_avg_vert_neon() local 280 q8u16 = vmovl_u8(vreinterpret_u8_u32(d16u32)); in vpx_convolve8_avg_vert_neon() [all …]
|
D | idct4x4_1_add_neon.c | 22 uint16x8_t q8u16; in vpx_idct4x4_1_add_neon() local 40 q8u16 = vaddw_u8(vreinterpretq_u16_s16(q0s16), in vpx_idct4x4_1_add_neon() 42 d6u8 = vqmovun_s16(vreinterpretq_s16_u16(q8u16)); in vpx_idct4x4_1_add_neon()
|
D | idct8x8_add_neon.c | 249 uint16x8_t q8u16, q9u16, q10u16, q11u16; in vpx_idct8x8_64_add_neon() local 292 q8u16 = vaddw_u8(vreinterpretq_u16_s16(q8s16), in vpx_idct8x8_64_add_neon() 301 d0u8 = vqmovun_s16(vreinterpretq_s16_u16(q8u16)); in vpx_idct8x8_64_add_neon() 329 q8u16 = vaddw_u8(vreinterpretq_u16_s16(q8s16), in vpx_idct8x8_64_add_neon() 338 d0u8 = vqmovun_s16(vreinterpretq_s16_u16(q8u16)); in vpx_idct8x8_64_add_neon() 365 uint16x8_t q8u16, q9u16, q10u16, q11u16; in vpx_idct8x8_12_add_neon() local 480 q8u16 = vaddw_u8(vreinterpretq_u16_s16(q8s16), in vpx_idct8x8_12_add_neon() 489 d0u8 = vqmovun_s16(vreinterpretq_s16_u16(q8u16)); in vpx_idct8x8_12_add_neon() 517 q8u16 = vaddw_u8(vreinterpretq_u16_s16(q8s16), in vpx_idct8x8_12_add_neon() 526 d0u8 = vqmovun_s16(vreinterpretq_s16_u16(q8u16)); in vpx_idct8x8_12_add_neon()
|
D | intrapred_neon.c | 663 uint16x8_t q0u16, q2u16, q3u16, q8u16, q10u16; in vpx_tm_predictor_16x16_neon() local 679 q8u16 = vdupq_lane_u16(d20u16, 1); in vpx_tm_predictor_16x16_neon() 684 q11s16 = vaddq_s16(vreinterpretq_s16_u16(q8u16), in vpx_tm_predictor_16x16_neon() 686 q8s16 = vaddq_s16(vreinterpretq_s16_u16(q8u16), in vpx_tm_predictor_16x16_neon() 700 q8u16 = vdupq_lane_u16(d20u16, 3); in vpx_tm_predictor_16x16_neon() 705 q11s16 = vaddq_s16(vreinterpretq_s16_u16(q8u16), in vpx_tm_predictor_16x16_neon() 707 q8s16 = vaddq_s16(vreinterpretq_s16_u16(q8u16), in vpx_tm_predictor_16x16_neon() 726 uint16x8_t q0u16, q3u16, q8u16, q9u16, q10u16, q11u16; in vpx_tm_predictor_32x32_neon() local 735 q8u16 = vsubl_u8(vget_low_u8(q1u8), vget_low_u8(q0u8)); in vpx_tm_predictor_32x32_neon() 746 vreinterpretq_s16_u16(q8u16)); in vpx_tm_predictor_32x32_neon() [all …]
|
D | idct4x4_add_neon.c | 19 uint16x8_t q8u16, q9u16; in vpx_idct4x4_16_add_neon() local 134 q8u16 = vaddw_u8(vreinterpretq_u16_s16(q8s16), in vpx_idct4x4_16_add_neon() 139 d26u8 = vqmovun_s16(vreinterpretq_s16_u16(q8u16)); in vpx_idct4x4_16_add_neon()
|
D | idct16x16_add_neon.c | 338 uint16x8_t q2u16, q3u16, q4u16, q5u16, q8u16; in vpx_idct16x16_256_add_neon_pass2() local 696 q8u16 = vaddw_u8(vreinterpretq_u16_s16(q8s16), in vpx_idct16x16_256_add_neon_pass2() 698 d12u8 = vqmovun_s16(vreinterpretq_s16_u16(q8u16)); in vpx_idct16x16_256_add_neon_pass2()
|
/external/libvpx/libvpx/vp9/common/arm/neon/ |
D | vp9_iht4x4_add_neon.c | 163 uint16x8_t q8u16, q9u16; in vp9_iht4x4_16_add_neon() local 234 q8u16 = vaddw_u8(vreinterpretq_u16_s16(q8s16), vreinterpret_u8_u32(d26u32)); in vp9_iht4x4_16_add_neon() 237 d26u8 = vqmovun_s16(vreinterpretq_s16_u16(q8u16)); in vp9_iht4x4_16_add_neon()
|
D | vp9_iht8x8_add_neon.c | 496 uint16x8_t q8u16, q9u16, q10u16, q11u16; in vp9_iht8x8_64_add_neon() local 600 q8u16 = vaddw_u8(vreinterpretq_u16_s16(q8s16), in vp9_iht8x8_64_add_neon() 609 d0u8 = vqmovun_s16(vreinterpretq_s16_u16(q8u16)); in vp9_iht8x8_64_add_neon()
|