/external/libvpx/libvpx/vp8/encoder/arm/neon/ |
D | shortfdct_neon.c | 17 int16x4_t d0s16, d1s16, d2s16, d3s16, d4s16, d5s16, d6s16, d7s16; in vp8_short_fdct4x4_neon() local 36 d1s16 = vld1_s16(input); in vp8_short_fdct4x4_neon() 44 v2tmp3 = vtrn_s32(vreinterpret_s32_s16(d1s16), in vp8_short_fdct4x4_neon() 69 d1s16 = vshrn_n_s32(q9s32, 12); in vp8_short_fdct4x4_neon() 75 v2tmp3 = vtrn_s32(vreinterpret_s32_s16(d1s16), in vp8_short_fdct4x4_neon() 106 d1s16 = vshrn_n_s32(q11s32, 16); in vp8_short_fdct4x4_neon() 107 d1s16 = vsub_s16(d1s16, vreinterpret_s16_u16(d4u16)); in vp8_short_fdct4x4_neon() 110 q0s16 = vcombine_s16(d0s16, d1s16); in vp8_short_fdct4x4_neon() 122 int16x4_t d0s16, d1s16, d2s16, d3s16, d4s16, d5s16, d6s16, d7s16; in vp8_short_fdct8x4_neon() local 221 d1s16 = vget_high_s16(q0s16); in vp8_short_fdct8x4_neon() [all …]
|
D | vp8_shortwalsh4x4_neon.c | 29 int16x4_t dEmptys16, d0s16, d1s16, d2s16, d3s16, d4s16, d5s16, d6s16, d7s16; in vp8_short_walsh4x4_neon() local 42 d1s16 = vld1_s16(input); in vp8_short_walsh4x4_neon() 50 v2tmp3 = vtrn_s32(vreinterpret_s32_s16(d1s16), in vp8_short_walsh4x4_neon() 72 d1s16 = vadd_s16(d7s16, d6s16); in vp8_short_walsh4x4_neon() 78 v2tmp2 = vtrn_s32(vreinterpret_s32_s16(d1s16), in vp8_short_walsh4x4_neon() 118 d1s16 = vshrn_n_s32(q9s32, 3); in vp8_short_walsh4x4_neon() 122 q0s16 = vcombine_s16(d0s16, d1s16); in vp8_short_walsh4x4_neon()
|
/external/libvpx/libvpx/vp9/common/arm/neon/ |
D | vp9_iht4x4_add_neon.c | 47 int16x4_t *d1s16, in GENERATE_COSINE_CONSTANTS() argument 50 *d1s16 = vdup_n_s16(cospi_16_64); in GENERATE_COSINE_CONSTANTS() 69 int16x4_t *d1s16, in IDCT4x4_1D() argument 88 q13s32 = vmull_s16(d23s16, *d1s16); in IDCT4x4_1D() 89 q14s32 = vmull_s16(d24s16, *d1s16); in IDCT4x4_1D() 160 int16x4_t d0s16, d1s16, d2s16, d3s16, d4s16, d5s16; in vp9_iht4x4_16_add_neon() local 179 GENERATE_COSINE_CONSTANTS(&d0s16, &d1s16, &d2s16); in vp9_iht4x4_16_add_neon() 183 IDCT4x4_1D(&d0s16, &d1s16, &d2s16, &q8s16, &q9s16); in vp9_iht4x4_16_add_neon() 193 GENERATE_COSINE_CONSTANTS(&d0s16, &d1s16, &d2s16); in vp9_iht4x4_16_add_neon() 203 IDCT4x4_1D(&d0s16, &d1s16, &d2s16, &q8s16, &q9s16); in vp9_iht4x4_16_add_neon()
|
D | vp9_iht8x8_add_neon.c | 112 int16x4_t d0s16, d1s16, d2s16, d3s16; in IDCT8x8_1D() local 121 d1s16 = vdup_n_s16(cospi_4_64); in IDCT8x8_1D() 147 q2s32 = vmlsl_s16(q2s32, d30s16, d1s16); in IDCT8x8_1D() 148 q3s32 = vmlsl_s16(q3s32, d31s16, d1s16); in IDCT8x8_1D() 159 q2s32 = vmull_s16(d18s16, d1s16); in IDCT8x8_1D() 160 q3s32 = vmull_s16(d19s16, d1s16); in IDCT8x8_1D() 189 d1s16 = vdup_n_s16(cospi_8_64); in IDCT8x8_1D() 200 q8s32 = vmull_s16(d20s16, d1s16); in IDCT8x8_1D() 201 q12s32 = vmull_s16(d21s16, d1s16); in IDCT8x8_1D() 203 q2s32 = vmlsl_s16(q2s32, d28s16, d1s16); in IDCT8x8_1D() [all …]
|
/external/libvpx/libvpx/vpx_dsp/arm/ |
D | idct8x8_add_neon.c | 94 int16x4_t d0s16, d1s16, d2s16, d3s16; in IDCT8x8_1D() local 103 d1s16 = vdup_n_s16(cospi_4_64); in IDCT8x8_1D() 129 q2s32 = vmlsl_s16(q2s32, d30s16, d1s16); in IDCT8x8_1D() 130 q3s32 = vmlsl_s16(q3s32, d31s16, d1s16); in IDCT8x8_1D() 141 q2s32 = vmull_s16(d18s16, d1s16); in IDCT8x8_1D() 142 q3s32 = vmull_s16(d19s16, d1s16); in IDCT8x8_1D() 171 d1s16 = vdup_n_s16(cospi_8_64); in IDCT8x8_1D() 182 q8s32 = vmull_s16(d20s16, d1s16); in IDCT8x8_1D() 183 q12s32 = vmull_s16(d21s16, d1s16); in IDCT8x8_1D() 185 q2s32 = vmlsl_s16(q2s32, d28s16, d1s16); in IDCT8x8_1D() [all …]
|
D | vpx_convolve8_neon.c | 30 int16x4_t d0s16, d1s16; in MULTIPLY_BY_Q0() local 33 d1s16 = vget_high_s16(q0s16); in MULTIPLY_BY_Q0() 39 qdst = vmlal_lane_s16(qdst, dsrc4, d1s16, 0); in MULTIPLY_BY_Q0() 40 qdst = vmlal_lane_s16(qdst, dsrc5, d1s16, 1); in MULTIPLY_BY_Q0() 41 qdst = vmlal_lane_s16(qdst, dsrc6, d1s16, 2); in MULTIPLY_BY_Q0() 42 qdst = vmlal_lane_s16(qdst, dsrc7, d1s16, 3); in MULTIPLY_BY_Q0()
|
D | vpx_convolve8_avg_neon.c | 30 int16x4_t d0s16, d1s16; in MULTIPLY_BY_Q0() local 33 d1s16 = vget_high_s16(q0s16); in MULTIPLY_BY_Q0() 39 qdst = vmlal_lane_s16(qdst, dsrc4, d1s16, 0); in MULTIPLY_BY_Q0() 40 qdst = vmlal_lane_s16(qdst, dsrc5, d1s16, 1); in MULTIPLY_BY_Q0() 41 qdst = vmlal_lane_s16(qdst, dsrc6, d1s16, 2); in MULTIPLY_BY_Q0() 42 qdst = vmlal_lane_s16(qdst, dsrc7, d1s16, 3); in MULTIPLY_BY_Q0()
|
D | idct16x16_add_neon.c | 89 int16x4_t d0s16, d1s16, d2s16, d3s16; in vpx_idct16x16_256_add_neon_pass1() local 147 d1s16 = vdup_n_s16(cospi_4_64); in vpx_idct16x16_256_add_neon_pass1() 151 q5s32 = vmull_s16(d18s16, d1s16); in vpx_idct16x16_256_add_neon_pass1() 152 q6s32 = vmull_s16(d19s16, d1s16); in vpx_idct16x16_256_add_neon_pass1() 154 q2s32 = vmlsl_s16(q2s32, d30s16, d1s16); in vpx_idct16x16_256_add_neon_pass1() 155 q3s32 = vmlsl_s16(q3s32, d31s16, d1s16); in vpx_idct16x16_256_add_neon_pass1() 332 int16x4_t d0s16, d1s16, d2s16, d3s16, d4s16, d5s16, d6s16, d7s16; in vpx_idct16x16_256_add_neon_pass2() local 405 d1s16 = vqrshrn_n_s32(q3s32, 14); in vpx_idct16x16_256_add_neon_pass2() 408 q0s16 = vcombine_s16(d0s16, d1s16); in vpx_idct16x16_256_add_neon_pass2() 1027 int16x4_t d0s16, d1s16, d2s16, d3s16, d4s16, d5s16, d6s16, d7s16; in vpx_idct16x16_10_add_neon_pass2() local [all …]
|