/external/libvpx/libvpx/vpx_dsp/arm/ |
D | variance_neon.c | 168 int32x4_t q8s32, q9s32, q10s32; in vpx_variance16x8_neon() local 171 q8s32 = vdupq_n_s32(0); in vpx_variance16x8_neon() 195 q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q11u16)); in vpx_variance16x8_neon() 201 q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q12u16)); in vpx_variance16x8_neon() 207 q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q13u16)); in vpx_variance16x8_neon() 213 q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q14u16)); in vpx_variance16x8_neon() 219 q0s64 = vpaddlq_s32(q8s32); in vpx_variance16x8_neon() 247 int32x4_t q8s32, q9s32, q10s32; in vpx_variance8x16_neon() local 250 q8s32 = vdupq_n_s32(0); in vpx_variance8x16_neon() 272 q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q11u16)); in vpx_variance8x16_neon() [all …]
|
D | idct8x8_add_neon.c | 99 int32x4_t q2s32, q3s32, q5s32, q6s32, q8s32, q9s32; in IDCT8x8_1D() local 182 q8s32 = vmull_s16(d20s16, d1s16); in IDCT8x8_1D() 187 q8s32 = vmlal_s16(q8s32, d28s16, d0s16); in IDCT8x8_1D() 192 d30s16 = vqrshrn_n_s32(q8s32, 14); in IDCT8x8_1D()
|
D | idct16x16_add_neon.c | 342 int32x4_t q0s32, q1s32, q2s32, q3s32, q4s32, q5s32, q6s32, q8s32, q9s32; in vpx_idct16x16_256_add_neon_pass2() local 517 q8s32 = vmull_s16(d20s16, d30s16); in vpx_idct16x16_256_add_neon_pass2() 522 q8s32 = vmlal_s16(q8s32, d26s16, d31s16); in vpx_idct16x16_256_add_neon_pass2() 527 d10s16 = vqrshrn_n_s32(q8s32, 14); in vpx_idct16x16_256_add_neon_pass2() 1036 int32x4_t q0s32, q1s32, q2s32, q3s32, q4s32, q5s32, q6s32, q8s32, q9s32; in vpx_idct16x16_10_add_neon_pass2() local 1114 q8s32 = vmull_s16(d6s16, d30s16); in vpx_idct16x16_10_add_neon_pass2() 1119 q8s32 = vmlal_s16(q8s32, d8s16, d31s16); in vpx_idct16x16_10_add_neon_pass2() 1124 d10s16 = vqrshrn_n_s32(q8s32, 14); in vpx_idct16x16_10_add_neon_pass2()
|
D | idct32x32_add_neon.c | 134 int32x4_t q8s32, q9s32, q10s32, q11s32, q12s32, q15s32; in DO_BUTTERFLY() local 145 q8s32 = vmull_s16(dCs16, d30s16); in DO_BUTTERFLY() 151 q8s32 = vsubq_s32(q8s32, q10s32); in DO_BUTTERFLY() 161 *qAs16 = vcombine_s16(vqrshrn_n_s32(q8s32, 14), in DO_BUTTERFLY()
|
/external/libvpx/libvpx/vp8/encoder/arm/neon/ |
D | vp8_shortwalsh4x4_neon.c | 30 int32x4_t qEmptys32, q0s32, q1s32, q2s32, q3s32, q8s32; in vp8_short_walsh4x4_neon() local 87 q8s32 = vaddl_s16(v2tmp1.val[0], v2tmp0.val[0]); in vp8_short_walsh4x4_neon() 92 q0s32 = vaddq_s32(q8s32, q9s32); in vp8_short_walsh4x4_neon() 95 q3s32 = vsubq_s32(q8s32, q9s32); in vp8_short_walsh4x4_neon() 102 q8s32 = vreinterpretq_s32_u32(q8u32); in vp8_short_walsh4x4_neon() 107 q0s32 = vsubq_s32(q0s32, q8s32); in vp8_short_walsh4x4_neon() 112 q8s32 = vaddq_s32(q0s32, q15s32); in vp8_short_walsh4x4_neon() 117 d0s16 = vshrn_n_s32(q8s32, 3); in vp8_short_walsh4x4_neon()
|
/external/libvpx/libvpx/vp9/common/arm/neon/ |
D | vp9_iht4x4_add_neon.c | 29 int32x4_t q8s32, q9s32; in TRANSPOSE4X4() local 36 q8s32 = vreinterpretq_s32_s16(vcombine_s16(d0x2s16.val[0], d0x2s16.val[1])); in TRANSPOSE4X4() 38 q0x2s32 = vtrnq_s32(q8s32, q9s32); in TRANSPOSE4X4() 115 int32x4_t q8s32, q9s32, q10s32, q11s32, q12s32, q13s32, q14s32, q15s32; in IADST4x4_1D() local 131 q8s32 = vmull_s16(*d4s16, d19s16); in IADST4x4_1D() 136 q10s32 = vaddq_s32(q10s32, q8s32); in IADST4x4_1D() 138 q8s32 = vdupq_n_s32(sinpi_3_9); in IADST4x4_1D() 140 q15s32 = vmulq_s32(q15s32, q8s32); in IADST4x4_1D()
|
D | vp9_iht8x8_add_neon.c | 117 int32x4_t q2s32, q3s32, q5s32, q6s32, q8s32, q9s32; in IDCT8x8_1D() local 200 q8s32 = vmull_s16(d20s16, d1s16); in IDCT8x8_1D() 205 q8s32 = vmlal_s16(q8s32, d28s16, d0s16); in IDCT8x8_1D() 210 d30s16 = vqrshrn_n_s32(q8s32, 14); in IDCT8x8_1D() 273 int32x4_t q0s32, q1s32, q2s32, q3s32, q4s32, q5s32, q6s32, q7s32, q8s32; in IADST8X8_1D() local 312 q8s32 = vmull_s16(d23s16, d31s16); in IADST8X8_1D() 317 q8s32 = vmlsl_s16(q8s32, d25s16, d30s16); in IADST8X8_1D() 329 q15s32 = vaddq_s32(q4s32, q8s32); in IADST8X8_1D() 331 q4s32 = vsubq_s32(q4s32, q8s32); in IADST8X8_1D()
|