yading@10: /*
yading@10:  * Copyright (c) 2011 Mans Rullgard <mans@mansr.com>
yading@10:  *
yading@10:  * This file is part of FFmpeg.
yading@10:  *
yading@10:  * FFmpeg is free software; you can redistribute it and/or
yading@10:  * modify it under the terms of the GNU Lesser General Public
yading@10:  * License as published by the Free Software Foundation; either
yading@10:  * version 2.1 of the License, or (at your option) any later version.
yading@10:  *
yading@10:  * FFmpeg is distributed in the hope that it will be useful,
yading@10:  * but WITHOUT ANY WARRANTY; without even the implied warranty of
yading@10:  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
yading@10:  * Lesser General Public License for more details.
yading@10:  *
yading@10:  * You should have received a copy of the GNU Lesser General Public
yading@10:  * License along with FFmpeg; if not, write to the Free Software
yading@10:  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
yading@10:  */
yading@10: 
yading@10: #include <stdint.h>
yading@10: 
yading@10: #include "libavutil/arm/cpu.h"
yading@10: #include "libavutil/attributes.h"
yading@10: #include "libavcodec/ac3dsp.h"
yading@10: #include "config.h"
yading@10: 
yading@10: void ff_ac3_exponent_min_neon(uint8_t *exp, int num_reuse_blocks, int nb_coefs);
yading@10: int ff_ac3_max_msb_abs_int16_neon(const int16_t *src, int len);
yading@10: void ff_ac3_lshift_int16_neon(int16_t *src, unsigned len, unsigned shift);
yading@10: void ff_ac3_rshift_int32_neon(int32_t *src, unsigned len, unsigned shift);
yading@10: void ff_float_to_fixed24_neon(int32_t *dst, const float *src, unsigned int len);
yading@10: void ff_ac3_extract_exponents_neon(uint8_t *exp, int32_t *coef, int nb_coefs);
yading@10: void ff_ac3_sum_square_butterfly_int32_neon(int64_t sum[4],
yading@10:                                             const int32_t *coef0,
yading@10:                                             const int32_t *coef1,
yading@10:                                             int len);
yading@10: void ff_ac3_sum_square_butterfly_float_neon(float sum[4],
yading@10:                                             const float *coef0,
yading@10:                                             const float *coef1,
yading@10:                                             int len);
yading@10: 
yading@10: void ff_ac3_bit_alloc_calc_bap_armv6(int16_t *mask, int16_t *psd,
yading@10:                                      int start, int end,
yading@10:                                      int snr_offset, int floor,
yading@10:                                      const uint8_t *bap_tab, uint8_t *bap);
yading@10: 
yading@10: void ff_ac3_update_bap_counts_arm(uint16_t mant_cnt[16], uint8_t *bap, int len);
yading@10: 
yading@10: av_cold void ff_ac3dsp_init_arm(AC3DSPContext *c, int bit_exact)
yading@10: {
yading@10:     int cpu_flags = av_get_cpu_flags();
yading@10: 
yading@10:     c->update_bap_counts         = ff_ac3_update_bap_counts_arm;
yading@10: 
yading@10:     if (have_armv6(cpu_flags)) {
yading@10:         c->bit_alloc_calc_bap    = ff_ac3_bit_alloc_calc_bap_armv6;
yading@10:     }
yading@10: 
yading@10:     if (have_neon(cpu_flags)) {
yading@10:         c->ac3_exponent_min      = ff_ac3_exponent_min_neon;
yading@10:         c->ac3_max_msb_abs_int16 = ff_ac3_max_msb_abs_int16_neon;
yading@10:         c->ac3_lshift_int16      = ff_ac3_lshift_int16_neon;
yading@10:         c->ac3_rshift_int32      = ff_ac3_rshift_int32_neon;
yading@10:         c->float_to_fixed24      = ff_float_to_fixed24_neon;
yading@10:         c->extract_exponents     = ff_ac3_extract_exponents_neon;
yading@10:         c->sum_square_butterfly_int32 = ff_ac3_sum_square_butterfly_int32_neon;
yading@10:         c->sum_square_butterfly_float = ff_ac3_sum_square_butterfly_float_neon;
yading@10:     }
yading@10: }