OLD | NEW |
1 /* | 1 /* |
2 * Copyright (c) 2014 The WebM project authors. All Rights Reserved. | 2 * Copyright (c) 2014 The WebM project authors. All Rights Reserved. |
3 * | 3 * |
4 * Use of this source code is governed by a BSD-style license | 4 * Use of this source code is governed by a BSD-style license |
5 * that can be found in the LICENSE file in the root of the source | 5 * that can be found in the LICENSE file in the root of the source |
6 * tree. An additional intellectual property rights grant can be found | 6 * tree. An additional intellectual property rights grant can be found |
7 * in the file PATENTS. All contributing project authors may | 7 * in the file PATENTS. All contributing project authors may |
8 * be found in the AUTHORS file in the root of the source tree. | 8 * be found in the AUTHORS file in the root of the source tree. |
9 */ | 9 */ |
10 | 10 |
(...skipping 14 matching lines...) Expand all Loading... |
25 { | 25 { |
26 const int32x4_t a = vpaddlq_s16(sum); | 26 const int32x4_t a = vpaddlq_s16(sum); |
27 const int64x2_t b = vpaddlq_s32(a); | 27 const int64x2_t b = vpaddlq_s32(a); |
28 const int32x2_t c = vadd_s32(vreinterpret_s32_s64(vget_low_s64(b)), | 28 const int32x2_t c = vadd_s32(vreinterpret_s32_s64(vget_low_s64(b)), |
29 vreinterpret_s32_s64(vget_high_s64(b))); | 29 vreinterpret_s32_s64(vget_high_s64(b))); |
30 output[0] = vget_lane_s16(vreinterpret_s16_s32(c), 0); | 30 output[0] = vget_lane_s16(vreinterpret_s16_s32(c), 0); |
31 output[1] = 0; | 31 output[1] = 0; |
32 } | 32 } |
33 } | 33 } |
34 | 34 |
35 void vp9_fdct8x8_quant_neon(const int16_t *input, int stride, | |
36 int16_t* coeff_ptr, intptr_t n_coeffs, | |
37 int skip_block, const int16_t* zbin_ptr, | |
38 const int16_t* round_ptr, const int16_t* quant_ptr, | |
39 const int16_t* quant_shift_ptr, | |
40 int16_t* qcoeff_ptr, int16_t* dqcoeff_ptr, | |
41 const int16_t* dequant_ptr, uint16_t* eob_ptr, | |
42 const int16_t* scan_ptr, | |
43 const int16_t* iscan_ptr) { | |
44 int16_t temp_buffer[64]; | |
45 (void)coeff_ptr; | |
46 | |
47 vp9_fdct8x8_neon(input, temp_buffer, stride); | |
48 vp9_quantize_fp_neon(temp_buffer, n_coeffs, skip_block, zbin_ptr, round_ptr, | |
49 quant_ptr, quant_shift_ptr, qcoeff_ptr, dqcoeff_ptr, | |
50 dequant_ptr, eob_ptr, scan_ptr, iscan_ptr); | |
51 } | |
52 | |
53 void vp9_fdct8x8_neon(const int16_t *input, int16_t *final_output, int stride) { | 35 void vp9_fdct8x8_neon(const int16_t *input, int16_t *final_output, int stride) { |
54 int i; | 36 int i; |
55 // stage 1 | 37 // stage 1 |
56 int16x8_t input_0 = vshlq_n_s16(vld1q_s16(&input[0 * stride]), 2); | 38 int16x8_t input_0 = vshlq_n_s16(vld1q_s16(&input[0 * stride]), 2); |
57 int16x8_t input_1 = vshlq_n_s16(vld1q_s16(&input[1 * stride]), 2); | 39 int16x8_t input_1 = vshlq_n_s16(vld1q_s16(&input[1 * stride]), 2); |
58 int16x8_t input_2 = vshlq_n_s16(vld1q_s16(&input[2 * stride]), 2); | 40 int16x8_t input_2 = vshlq_n_s16(vld1q_s16(&input[2 * stride]), 2); |
59 int16x8_t input_3 = vshlq_n_s16(vld1q_s16(&input[3 * stride]), 2); | 41 int16x8_t input_3 = vshlq_n_s16(vld1q_s16(&input[3 * stride]), 2); |
60 int16x8_t input_4 = vshlq_n_s16(vld1q_s16(&input[4 * stride]), 2); | 42 int16x8_t input_4 = vshlq_n_s16(vld1q_s16(&input[4 * stride]), 2); |
61 int16x8_t input_5 = vshlq_n_s16(vld1q_s16(&input[5 * stride]), 2); | 43 int16x8_t input_5 = vshlq_n_s16(vld1q_s16(&input[5 * stride]), 2); |
62 int16x8_t input_6 = vshlq_n_s16(vld1q_s16(&input[6 * stride]), 2); | 44 int16x8_t input_6 = vshlq_n_s16(vld1q_s16(&input[6 * stride]), 2); |
(...skipping 169 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... |
232 vst1q_s16(&final_output[1 * 8], input_1); | 214 vst1q_s16(&final_output[1 * 8], input_1); |
233 vst1q_s16(&final_output[2 * 8], input_2); | 215 vst1q_s16(&final_output[2 * 8], input_2); |
234 vst1q_s16(&final_output[3 * 8], input_3); | 216 vst1q_s16(&final_output[3 * 8], input_3); |
235 vst1q_s16(&final_output[4 * 8], input_4); | 217 vst1q_s16(&final_output[4 * 8], input_4); |
236 vst1q_s16(&final_output[5 * 8], input_5); | 218 vst1q_s16(&final_output[5 * 8], input_5); |
237 vst1q_s16(&final_output[6 * 8], input_6); | 219 vst1q_s16(&final_output[6 * 8], input_6); |
238 vst1q_s16(&final_output[7 * 8], input_7); | 220 vst1q_s16(&final_output[7 * 8], input_7); |
239 } | 221 } |
240 } | 222 } |
241 | 223 |
OLD | NEW |