| /external/libvpx/libvpx/vp8/common/arm/neon/ |
| shortidct4x4llm_neon.c | 27 int16x8_t q1s16, q2s16, q3s16, q4s16; local
38 q2s16 = vcombine_s16(d3, d5);
40 q3s16 = vqdmulhq_n_s16(q2s16, sinpi8sqrt2);
41 q4s16 = vqdmulhq_n_s16(q2s16, cospi8sqrt2minus1);
49 q3s16 = vqaddq_s16(q3s16, q2s16);
50 q4s16 = vqaddq_s16(q4s16, q2s16);
69 q2s16 = vcombine_s16(v2tmp2.val[1], v2tmp3.val[1]);
71 q3s16 = vqdmulhq_n_s16(q2s16, sinpi8sqrt2);
72 q4s16 = vqdmulhq_n_s16(q2s16, cospi8sqrt2minus1);
80 q3s16 = vqaddq_s16(q3s16, q2s16);
[all...] |
| iwalsh_neon.c | 16 int16x8_t q0s16, q1s16, q2s16, q3s16; local
31 q2s16 = vcombine_s16(d4s16, d5s16);
34 q0s16 = vaddq_s16(q2s16, q3s16);
35 q1s16 = vsubq_s16(q2s16, q3s16);
51 q2s16 = vcombine_s16(d4s16, d5s16);
56 q0s16 = vaddq_s16(q2s16, q3s16);
57 q1s16 = vsubq_s16(q2s16, q3s16);
|
| loopfiltersimplehorizontaledge_neon.c | 21 int16x8_t q2s16, q3s16, q13s16; local
52 q2s16 = vsubl_s8(vget_low_s8(vreinterpretq_s8_u8(q7u8)),
60 q2s16 = vmulq_s16(q2s16, q13s16);
66 q2s16 = vaddw_s8(q2s16, vget_low_s8(q4s8));
69 d8s8 = vqmovn_s16(q2s16);
|
| loopfiltersimpleverticaledge_neon.c | 185 int16x8_t q2s16, q13s16, q11s16; local
221 q2s16 = vsubl_s8(vget_low_s8(vreinterpretq_s8_u8(q4u8)),
229 q2s16 = vmulq_s16(q2s16, q11s16);
235 q2s16 = vaddw_s8(q2s16, vget_low_s8(q14s8));
238 d28s8 = vqmovn_s16(q2s16);
|
| mbloopfilter_neon.c | 33 int16x8_t q0s16, q2s16, q11s16, q12s16, q13s16, q14s16, q15s16; local
79 q2s16 = vsubl_s8(vget_low_s8(vreinterpretq_s8_u8(q7)),
88 q2s16 = vmulq_s16(q2s16, q11s16);
93 q2s16 = vaddw_s8(q2s16, vget_low_s8(q1s8));
99 d2 = vqmovn_s16(q2s16);
|
| vp8_loopfilter_neon.c | 32 int16x8_t q2s16, q11s16; local
74 q2s16 = vsubl_s8(vget_low_s8(vreinterpretq_s8_u8(q7)),
87 q2s16 = vmulq_s16(q2s16, vreinterpretq_s16_u16(q4u16));
94 q2s16 = vaddw_s8(q2s16, vget_low_s8(q1s8));
99 d2s8 = vqmovn_s16(q2s16);
|
| /external/libvpx/libvpx/vpx_dsp/arm/ |
| idct32x32_add_neon.c | 264 int16x8_t q2s16,
282 q4s16 = vaddq_s16(q2s16, q1s16);
285 q7s16 = vsubq_s16(q2s16, q1s16);
290 q2s16 = vaddq_s16(q10s16, q1s16);
304 q4s16 = vaddq_s16(q2s16, q1s16);
307 q7s16 = vsubq_s16(q2s16, q1s16);
312 q2s16 = vaddq_s16(q12s16, q1s16);
326 q4s16 = vaddq_s16(q2s16, q1s16);
329 q7s16 = vsubq_s16(q2s16, q1s16);
334 q2s16 = vaddq_s16(q14s16, q1s16)
466 int16x8_t q0s16, q1s16, q2s16, q3s16, q4s16, q5s16, q6s16, q7s16; local
[all...] |
| loopfilter_16_neon.c | 34 int16x8_t q2s16, q11s16; local
74 q2s16 = vsubl_s8(vget_low_s8(vreinterpretq_s8_u8(q7)),
87 q2s16 = vmulq_s16(q2s16, vreinterpretq_s16_u16(q4u16));
94 q2s16 = vaddw_s8(q2s16, vget_low_s8(q1s8));
100 d2s8 = vqmovn_s16(q2s16);
|
| idct16x16_add_neon.c | 95 int16x8_t q0s16, q1s16, q2s16, q3s16, q4s16, q5s16, q6s16, q7s16; local
238 q2s16 = vsubq_s16(q9s16, q10s16);
263 q10s16 = vaddq_s16(q2s16, q5s16);
266 q13s16 = vsubq_s16(q2s16, q5s16);
340 int16x8_t q0s16, q1s16, q2s16, q3s16, q4s16, q5s16, q6s16, q7s16; local
448 q2s16 = vcombine_s16(d4s16, d5s16);
474 q10s16 = vsubq_s16(q3s16, q2s16);
475 q11s16 = vaddq_s16(q2s16, q3s16);
529 q2s16 = vcombine_s16(d4s16, d5s16);
534 q9s16 = vaddq_s16(q1s16, q2s16);
890 int16x8_t q0s16, q1s16, q2s16, q4s16, q5s16, q6s16, q7s16; local
1034 int16x8_t q0s16, q1s16, q2s16, q3s16, q4s16, q5s16, q6s16, q7s16; local
[all...] |
| idct8x8_add_neon.c | 98 int16x8_t q0s16, q1s16, q2s16, q3s16, q4s16, q5s16, q6s16, q7s16; local
199 q2s16 = vsubq_s16(*q11s16, *q13s16);
232 *q10s16 = vaddq_s16(q2s16, q5s16);
235 *q13s16 = vsubq_s16(q2s16, q5s16);
363 int16x8_t q0s16, q1s16, q2s16, q3s16, q4s16, q5s16, q6s16, q7s16; local
413 q2s16 = vsubq_s16(q9s16, q13s16);
447 q10s16 = vaddq_s16(q2s16, q5s16);
450 q13s16 = vsubq_s16(q2s16, q5s16);
|
| /external/libvpx/libvpx/vp8/encoder/arm/neon/ |
| shortfdct_neon.c | 126 int16x8_t q0s16, q1s16, q2s16, q3s16; local
143 q2s16 = vld1q_s16(input);
148 vreinterpretq_s32_s16(q2s16));
167 q2s16 = vsubq_s16(q11s16, q12s16);
199 vreinterpretq_s32_s16(q2s16));
261 q2s16 = vcombine_s16(d4s16, d5s16);
266 vst1q_s16(output + 16, q2s16);
|
| /external/libvpx/libvpx/vp9/common/arm/neon/ |
| vp9_iht8x8_add_neon.c | 116 int16x8_t q0s16, q1s16, q2s16, q3s16, q4s16, q5s16, q6s16, q7s16; local
217 q2s16 = vsubq_s16(*q11s16, *q13s16);
250 *q10s16 = vaddq_s16(q2s16, q5s16);
253 *q13s16 = vsubq_s16(q2s16, q5s16);
272 int16x8_t q2s16, q4s16, q5s16, q6s16; local
460 q2s16 = vcombine_s16(d4s16, d5s16);
483 *q11s16 = vsubq_s16(q5s16, q2s16);
|
