X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=libavcodec%2Fwmaprodec.c;h=4eaeed610ee241c634f3cf6dcab9daa020fd598d;hb=5085f25ace1e74846a0de3369bedd0e22d1a1bdc;hp=6ed55fd24ca99fae794248859068768878d55339;hpb=42ea968188db0cfc141779215b463bc9351be371;p=ffmpeg diff --git a/libavcodec/wmaprodec.c b/libavcodec/wmaprodec.c index 6ed55fd24ca..4eaeed610ee 100644 --- a/libavcodec/wmaprodec.c +++ b/libavcodec/wmaprodec.c @@ -1,27 +1,27 @@ /* * Wmapro compatible decoder * Copyright (c) 2007 Baptiste Coudurier, Benjamin Larsson, Ulion - * Copyright (c) 2008 - 2009 Sascha Sommer, Benjamin Larsson + * Copyright (c) 2008 - 2011 Sascha Sommer, Benjamin Larsson * - * This file is part of FFmpeg. + * This file is part of Libav. * - * FFmpeg is free software; you can redistribute it and/or + * Libav is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public * License as published by the Free Software Foundation; either * version 2.1 of the License, or (at your option) any later version. * - * FFmpeg is distributed in the hope that it will be useful, + * Libav is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public - * License along with FFmpeg; if not, write to the Free Software + * License along with Libav; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA */ /** - * @file libavcodec/wmaprodec.c + * @file * @brief wmapro decoder implementation * Wmapro is an MDCT based codec comparable to wma standard or AAC. * The decoding therefore consists of the following steps: @@ -86,23 +86,32 @@ * subframe in order to reconstruct the output samples. */ +#include + +#include "libavutil/float_dsp.h" +#include "libavutil/intfloat.h" +#include "libavutil/intreadwrite.h" + #include "avcodec.h" +#include "bitstream.h" #include "internal.h" -#include "get_bits.h" #include "put_bits.h" #include "wmaprodata.h" -#include "dsputil.h" +#include "sinewin.h" #include "wma.h" +#include "wma_common.h" /** current decoder limitations */ #define WMAPRO_MAX_CHANNELS 8 ///< max number of handled channels #define MAX_SUBFRAMES 32 ///< max number of subframes per channel #define MAX_BANDS 29 ///< max number of scale factor bands -#define MAX_FRAMESIZE 16384 ///< maximum compressed frame size +#define MAX_FRAMESIZE 32768 ///< maximum compressed frame size -#define WMAPRO_BLOCK_MAX_BITS 12 ///< log2 of max block size +#define WMAPRO_BLOCK_MIN_BITS 6 ///< log2 of min block size +#define WMAPRO_BLOCK_MAX_BITS 13 ///< log2 of max block size +#define WMAPRO_BLOCK_MIN_SIZE (1 << WMAPRO_BLOCK_MIN_BITS) ///< minimum block size #define WMAPRO_BLOCK_MAX_SIZE (1 << WMAPRO_BLOCK_MAX_BITS) ///< maximum block size -#define WMAPRO_BLOCK_SIZES (WMAPRO_BLOCK_MAX_BITS - BLOCK_MIN_BITS + 1) ///< possible block sizes +#define WMAPRO_BLOCK_SIZES (WMAPRO_BLOCK_MAX_BITS - WMAPRO_BLOCK_MIN_BITS + 1) ///< possible block sizes #define VLCBITS 9 @@ -119,12 +128,12 @@ static VLC vec4_vlc; ///< 4 coefficients per symbol static VLC vec2_vlc; ///< 2 coefficients per symbol static VLC vec1_vlc; ///< 1 coefficient per symbol static VLC coef_vlc[2]; ///< coefficient run length vlc codes -static float sin64[33]; ///< sinus table for decorrelation +static float sin64[33]; ///< sine table for decorrelation /** * @brief frame specific decoder context for a single channel */ -typedef struct { +typedef struct WMAProChannelCtx { int16_t prev_block_len; ///< length of the previous block uint8_t transmit_coefs; uint8_t num_subframes; @@ -137,17 +146,19 @@ typedef struct { int8_t reuse_sf; ///< share scale factors between subframes int8_t scale_factor_step; ///< scaling step for the current subframe int max_scale_factor; ///< maximum scale factor for the current subframe - int scale_factors[MAX_BANDS]; ///< scale factor values for the current subframe - int saved_scale_factors[MAX_BANDS]; ///< scale factors from a previous subframe + int saved_scale_factors[2][MAX_BANDS]; ///< resampled and (previously) transmitted scale factor values + int8_t scale_factor_idx; ///< index for the transmitted scale factor values (used for resampling) + int* scale_factors; ///< pointer to the scale factor values used for decoding uint8_t table_idx; ///< index in sf_offsets for the scale factor reference block float* coeffs; ///< pointer to the subframe decode buffer - DECLARE_ALIGNED_16(float, out[WMAPRO_BLOCK_MAX_SIZE + WMAPRO_BLOCK_MAX_SIZE / 2]); ///< output buffer + uint16_t num_vec_coeffs; ///< number of vector coded coefficients + DECLARE_ALIGNED(32, float, out)[WMAPRO_BLOCK_MAX_SIZE + WMAPRO_BLOCK_MAX_SIZE / 2]; ///< output buffer } WMAProChannelCtx; /** * @brief channel group for channel transformations */ -typedef struct { +typedef struct WMAProChannelGrp { uint8_t num_channels; ///< number of channels in the group int8_t transform; ///< transform on / off int8_t transform_band[MAX_BANDS]; ///< controls if the transform is enabled for a certain band @@ -161,12 +172,12 @@ typedef struct { typedef struct WMAProDecodeCtx { /* generic decoder variables */ AVCodecContext* avctx; ///< codec context for av_log - DSPContext dsp; ///< accelerated DSP functions + AVFloatDSPContext fdsp; uint8_t frame_data[MAX_FRAMESIZE + - FF_INPUT_BUFFER_PADDING_SIZE];///< compressed frame data + AV_INPUT_BUFFER_PADDING_SIZE];///< compressed frame data PutBitContext pb; ///< context for filling the frame_data buffer - MDCTContext mdct_ctx[WMAPRO_BLOCK_SIZES]; ///< MDCT context per block size - DECLARE_ALIGNED_16(float, tmp[WMAPRO_BLOCK_MAX_SIZE]); ///< IMDCT output buffer + FFTContext mdct_ctx[WMAPRO_BLOCK_SIZES]; ///< MDCT context per block size + DECLARE_ALIGNED(32, float, tmp)[WMAPRO_BLOCK_MAX_SIZE]; ///< IMDCT output buffer float* windows[WMAPRO_BLOCK_SIZES]; ///< windows for the different block sizes /* frame size dependent frame information (set during initialization) */ @@ -176,7 +187,6 @@ typedef struct WMAProDecodeCtx { uint8_t bits_per_sample; ///< integer audio sample size for the unscaled IMDCT output (used to scale to [-1.0, 1.0]) uint16_t samples_per_frame; ///< number of samples to output uint16_t log2_frame_size; - int8_t num_channels; ///< number of channels in the stream (same as AVCodecContext.num_channels) int8_t lfe_channel; ///< lfe channel index uint8_t max_num_subframes; uint8_t subframe_len_bits; ///< number of bits used for the subframe length @@ -188,21 +198,20 @@ typedef struct WMAProDecodeCtx { int16_t subwoofer_cutoffs[WMAPRO_BLOCK_SIZES]; ///< subwoofer cutoff values /* packet decode state */ - GetBitContext pgb; ///< bitstream reader context for the packet + BitstreamContext pbc; ///< bitstream reader context for the packet + int next_packet_start; ///< start offset of the next wma packet in the demuxer packet + uint8_t packet_offset; ///< frame offset in the packet uint8_t packet_sequence_number; ///< current packet number int num_saved_bits; ///< saved number of bits int frame_offset; ///< frame offset in the bit reservoir int subframe_offset; ///< subframe offset in the bit reservoir uint8_t packet_loss; ///< set in case of bitstream error - uint8_t output_buffer_full; ///< flag indicating that the output buffer is full + uint8_t packet_done; ///< set when a packet is fully decoded /* frame decode state */ uint32_t frame_num; ///< current frame number (not used for decoding) - GetBitContext gb; ///< bitstream reader context + BitstreamContext bc; ///< bitstream reader context int buf_bit_size; ///< buffer size in bits - float* samples_start; ///< start samplebuffer pointer - float* samples; ///< current samplebuffer pointer - float* samples_end; ///< maximum samplebuffer pointer uint8_t drc_gain; ///< gain for the DRC tool int8_t skip_frame; ///< skip output step int8_t parsed_all_subframes; ///< all subframes decoded? @@ -212,6 +221,7 @@ typedef struct WMAProDecodeCtx { int8_t channels_for_cur_subframe; ///< number of channels that contain the subframe int8_t channel_indexes_for_cur_subframe[WMAPRO_MAX_CHANNELS]; int8_t num_bands; ///< number of scale factor bands + int8_t transmit_num_vec_coeffs; ///< number of vector coded coefficients is part of the bitstream int16_t* cur_sfb_offsets; ///< sfb offsets for the current block uint8_t table_idx; ///< index for the num_sfb, sfb_offsets, sf_offsets and subwoofer_cutoffs tables int8_t esc_len; ///< length of escaped coefficients @@ -227,10 +237,10 @@ typedef struct WMAProDecodeCtx { *@brief helper function to print the most important members of the context *@param s context */ -static void av_cold dump_context(WMAProDecodeCtx *s) +static av_cold void dump_context(WMAProDecodeCtx *s) { #define PRINT(a, b) av_log(s->avctx, AV_LOG_DEBUG, " %s = %d\n", a, b); -#define PRINT_HEX(a, b) av_log(s->avctx, AV_LOG_DEBUG, " %s = %x\n", a, b); +#define PRINT_HEX(a, b) av_log(s->avctx, AV_LOG_DEBUG, " %s = %"PRIx32"\n", a, b); PRINT("ed sample bit depth", s->bits_per_sample); PRINT_HEX("ed decode flags", s->decode_flags); @@ -238,7 +248,7 @@ static void av_cold dump_context(WMAProDecodeCtx *s) PRINT("log2 frame size", s->log2_frame_size); PRINT("max num subframes", s->max_num_subframes); PRINT("len prefix", s->len_prefix); - PRINT("num channels", s->num_channels); + PRINT("num channels", s->avctx->channels); } /** @@ -267,15 +277,21 @@ static av_cold int decode_init(AVCodecContext *avctx) WMAProDecodeCtx *s = avctx->priv_data; uint8_t *edata_ptr = avctx->extradata; unsigned int channel_mask; - int i; + int i, bits; int log2_max_num_subframes; int num_possible_block_sizes; + if (!avctx->block_align) { + av_log(avctx, AV_LOG_ERROR, "block_align is not set\n"); + return AVERROR(EINVAL); + } + s->avctx = avctx; - dsputil_init(&s->dsp, avctx); + avpriv_float_dsp_init(&s->fdsp, avctx->flags & AV_CODEC_FLAG_BITEXACT); + init_put_bits(&s->pb, s->frame_data, MAX_FRAMESIZE); - avctx->sample_fmt = SAMPLE_FMT_FLT; + avctx->sample_fmt = AV_SAMPLE_FMT_FLTP; if (avctx->extradata_size >= 18) { s->decode_flags = AV_RL16(edata_ptr+14); @@ -283,39 +299,34 @@ static av_cold int decode_init(AVCodecContext *avctx) s->bits_per_sample = AV_RL16(edata_ptr); /** dump the extradata */ for (i = 0; i < avctx->extradata_size; i++) - dprintf(avctx, "[%x] ", avctx->extradata[i]); - dprintf(avctx, "\n"); + ff_dlog(avctx, "[%x] ", avctx->extradata[i]); + ff_dlog(avctx, "\n"); } else { - av_log_ask_for_sample(avctx, "Unknown extradata size\n"); - return AVERROR_INVALIDDATA; + avpriv_request_sample(avctx, "Unknown extradata size"); + return AVERROR_PATCHWELCOME; } /** generic init */ s->log2_frame_size = av_log2(avctx->block_align) + 4; /** frame info */ - s->skip_frame = 1; /** skip first frame */ + s->skip_frame = 1; /* skip first frame */ s->packet_loss = 1; s->len_prefix = (s->decode_flags & 0x40); - if (!s->len_prefix) { - av_log_ask_for_sample(avctx, "no length prefix\n"); - return AVERROR_INVALIDDATA; - } - /** get frame len */ - s->samples_per_frame = 1 << ff_wma_get_frame_len_bits(avctx->sample_rate, - 3, s->decode_flags); - - /** init previous block len */ - for (i = 0; i < avctx->channels; i++) - s->channel[i].prev_block_len = s->samples_per_frame; + bits = ff_wma_get_frame_len_bits(avctx->sample_rate, 3, s->decode_flags); + if (bits > WMAPRO_BLOCK_MAX_BITS) { + avpriv_request_sample(avctx, "14-bit block sizes"); + return AVERROR_PATCHWELCOME; + } + s->samples_per_frame = 1 << bits; /** subframe info */ log2_max_num_subframes = ((s->decode_flags & 0x38) >> 3); s->max_num_subframes = 1 << log2_max_num_subframes; - if (s->max_num_subframes == 16) + if (s->max_num_subframes == 16 || s->max_num_subframes == 4) s->max_subframe_len_bit = 1; s->subframe_len_bits = av_log2(log2_max_num_subframes) + 1; @@ -324,12 +335,35 @@ static av_cold int decode_init(AVCodecContext *avctx) s->dynamic_range_compression = (s->decode_flags & 0x80); if (s->max_num_subframes > MAX_SUBFRAMES) { - av_log(avctx, AV_LOG_ERROR, "invalid number of subframes %i\n", + av_log(avctx, AV_LOG_ERROR, "invalid number of subframes %"PRId8"\n", s->max_num_subframes); return AVERROR_INVALIDDATA; } - s->num_channels = avctx->channels; + if (s->min_samples_per_subframe < WMAPRO_BLOCK_MIN_SIZE) { + av_log(avctx, AV_LOG_ERROR, "Invalid minimum block size %"PRId8"\n", + s->max_num_subframes); + return AVERROR_INVALIDDATA; + } + + if (s->avctx->sample_rate <= 0) { + av_log(avctx, AV_LOG_ERROR, "invalid sample rate\n"); + return AVERROR_INVALIDDATA; + } + + if (avctx->channels < 0) { + av_log(avctx, AV_LOG_ERROR, "invalid number of channels %d\n", + avctx->channels); + return AVERROR_INVALIDDATA; + } else if (avctx->channels > WMAPRO_MAX_CHANNELS) { + avpriv_request_sample(avctx, + "More than %d channels", WMAPRO_MAX_CHANNELS); + return AVERROR_PATCHWELCOME; + } + + /** init previous block len */ + for (i = 0; i < avctx->channels; i++) + s->channel[i].prev_block_len = s->samples_per_frame; /** extract lfe channel position */ s->lfe_channel = -1; @@ -342,11 +376,6 @@ static av_cold int decode_init(AVCodecContext *avctx) } } - if (s->num_channels < 0 || s->num_channels > WMAPRO_MAX_CHANNELS) { - av_log_ask_for_sample(avctx, "invalid number of channels\n"); - return AVERROR_NOTSUPP; - } - INIT_VLC_STATIC(&sf_vlc, SCALEVLCBITS, HUFF_SCALE_SIZE, scale_huffbits, 1, 1, scale_huffcodes, 2, 2, 616); @@ -410,7 +439,8 @@ static av_cold int decode_init(AVCodecContext *avctx) for (x = 0; x < num_possible_block_sizes; x++) { int v = 0; while (s->sfb_offsets[x][v + 1] << x < offset) - ++v; + if (++v >= MAX_BANDS) + return AVERROR_INVALIDDATA; s->sf_offsets[i][x][b] = v; } } @@ -418,15 +448,14 @@ static av_cold int decode_init(AVCodecContext *avctx) /** init MDCT, FIXME: only init needed sizes */ for (i = 0; i < WMAPRO_BLOCK_SIZES; i++) - ff_mdct_init(&s->mdct_ctx[i], BLOCK_MIN_BITS+1+i, 1, - 1.0 / (1 << (BLOCK_MIN_BITS + i - 1)) + ff_mdct_init(&s->mdct_ctx[i], WMAPRO_BLOCK_MIN_BITS+1+i, 1, + 1.0 / (1 << (WMAPRO_BLOCK_MIN_BITS + i - 1)) / (1 << (s->bits_per_sample - 1))); - /** init MDCT windows: simple sinus window */ + /** init MDCT windows: simple sine window */ for (i = 0; i < WMAPRO_BLOCK_SIZES; i++) { - const int n = 1 << (WMAPRO_BLOCK_MAX_BITS - i); - const int win_idx = WMAPRO_BLOCK_MAX_BITS - i - 7; - ff_sine_window_init(ff_sine_windows[win_idx], n); + const int win_idx = WMAPRO_BLOCK_MAX_BITS - i; + ff_init_ff_sine_windows(win_idx); s->windows[WMAPRO_BLOCK_SIZES - i - 1] = ff_sine_windows[win_idx]; } @@ -446,6 +475,7 @@ static av_cold int decode_init(AVCodecContext *avctx) dump_context(s); avctx->channel_layout = channel_mask; + return 0; } @@ -466,10 +496,11 @@ static int decode_subframe_length(WMAProDecodeCtx *s, int offset) /** 1 bit indicates if the subframe is of maximum length */ if (s->max_subframe_len_bit) { - if (get_bits1(&s->gb)) - frame_len_shift = 1 + get_bits(&s->gb, s->subframe_len_bits-1); + if (bitstream_read_bit(&s->bc)) + frame_len_shift = 1 + bitstream_read(&s->bc, + s->subframe_len_bits - 1); } else - frame_len_shift = get_bits(&s->gb, s->subframe_len_bits); + frame_len_shift = bitstream_read(&s->bc, s->subframe_len_bits); subframe_len = s->samples_per_frame >> frame_len_shift; @@ -505,26 +536,24 @@ static int decode_subframe_length(WMAProDecodeCtx *s, int offset) */ static int decode_tilehdr(WMAProDecodeCtx *s) { - uint16_t num_samples[WMAPRO_MAX_CHANNELS]; /** sum of samples for all currently known subframes of a channel */ - uint8_t contains_subframe[WMAPRO_MAX_CHANNELS]; /** flag indicating if a channel contains the current subframe */ - int channels_for_cur_subframe = s->num_channels; /** number of channels that contain the current subframe */ - int fixed_channel_layout = 0; /** flag indicating that all channels use the same subframe offsets and sizes */ - int min_channel_len = 0; /** smallest sum of samples (channels with this length will be processed first) */ + uint16_t num_samples[WMAPRO_MAX_CHANNELS] = { 0 };/**< sum of samples for all currently known subframes of a channel */ + uint8_t contains_subframe[WMAPRO_MAX_CHANNELS]; /**< flag indicating if a channel contains the current subframe */ + int channels_for_cur_subframe = s->avctx->channels; /**< number of channels that contain the current subframe */ + int fixed_channel_layout = 0; /**< flag indicating that all channels use the same subframe offsets and sizes */ + int min_channel_len = 0; /**< smallest sum of samples (channels with this length will be processed first) */ int c; /* Should never consume more than 3073 bits (256 iterations for the - * while loop when always the minimum amount of 128 samples is substracted + * while loop when always the minimum amount of 128 samples is subtracted * from missing samples in the 8 channel case). * 1 + BLOCK_MAX_SIZE * MAX_CHANNELS / BLOCK_MIN_SIZE * (MAX_CHANNELS + 4) */ /** reset tiling information */ - for (c = 0; c < s->num_channels; c++) + for (c = 0; c < s->avctx->channels; c++) s->channel[c].num_subframes = 0; - memset(num_samples, 0, sizeof(num_samples)); - - if (s->max_num_subframes == 1 || get_bits1(&s->gb)) + if (s->max_num_subframes == 1 || bitstream_read_bit(&s->bc)) fixed_channel_layout = 1; /** loop until the frame data is split between the subframes */ @@ -532,13 +561,13 @@ static int decode_tilehdr(WMAProDecodeCtx *s) int subframe_len; /** check which channels contain the subframe */ - for (c = 0; c < s->num_channels; c++) { + for (c = 0; c < s->avctx->channels; c++) { if (num_samples[c] == min_channel_len) { if (fixed_channel_layout || channels_for_cur_subframe == 1 || (min_channel_len == s->samples_per_frame - s->min_samples_per_subframe)) contains_subframe[c] = 1; else - contains_subframe[c] = get_bits1(&s->gb); + contains_subframe[c] = bitstream_read_bit(&s->bc); } else contains_subframe[c] = 0; } @@ -549,7 +578,7 @@ static int decode_tilehdr(WMAProDecodeCtx *s) /** add subframes to the individual channels and find new min_channel_len */ min_channel_len += subframe_len; - for (c = 0; c < s->num_channels; c++) { + for (c = 0; c < s->avctx->channels; c++) { WMAProChannelCtx* chan = &s->channel[c]; if (contains_subframe[c]) { @@ -576,12 +605,12 @@ static int decode_tilehdr(WMAProDecodeCtx *s) } } while (min_channel_len < s->samples_per_frame); - for (c = 0; c < s->num_channels; c++) { + for (c = 0; c < s->avctx->channels; c++) { int i; int offset = 0; for (i = 0; i < s->channel[c].num_subframes; i++) { - dprintf(s->avctx, "frame[%i] channel[%i] subframe[%i]" - " len %i\n", s->frame_num, c, i, + ff_dlog(s->avctx, "frame[%"PRIi32"] channel[%i] subframe[%i]" + " len %"PRIu16"\n", s->frame_num, c, i, s->channel[c].subframe_len[i]); s->channel[c].subframe_offset[i] = offset; offset += s->channel[c].subframe_len[i]; @@ -602,15 +631,15 @@ static void decode_decorrelation_matrix(WMAProDecodeCtx *s, int i; int offset = 0; int8_t rotation_offset[WMAPRO_MAX_CHANNELS * WMAPRO_MAX_CHANNELS]; - memset(chgroup->decorrelation_matrix, 0, s->num_channels * - s->num_channels * sizeof(*chgroup->decorrelation_matrix)); + memset(chgroup->decorrelation_matrix, 0, s->avctx->channels * + s->avctx->channels * sizeof(*chgroup->decorrelation_matrix)); for (i = 0; i < chgroup->num_channels * (chgroup->num_channels - 1) >> 1; i++) - rotation_offset[i] = get_bits(&s->gb, 6); + rotation_offset[i] = bitstream_read(&s->bc, 6); for (i = 0; i < chgroup->num_channels; i++) chgroup->decorrelation_matrix[chgroup->num_channels * i + i] = - get_bits1(&s->gb) ? 1.0 : -1.0; + bitstream_read_bit(&s->bc) ? 1.0 : -1.0; for (i = 1; i < chgroup->num_channels; i++) { int x; @@ -656,13 +685,13 @@ static int decode_channel_transform(WMAProDecodeCtx* s) /** in the one channel case channel transforms are pointless */ s->num_chgroups = 0; - if (s->num_channels > 1) { + if (s->avctx->channels > 1) { int remaining_channels = s->channels_for_cur_subframe; - if (get_bits1(&s->gb)) { - av_log_ask_for_sample(s->avctx, - "unsupported channel transform bit\n"); - return AVERROR_INVALIDDATA; + if (bitstream_read_bit(&s->bc)) { + avpriv_request_sample(s->avctx, + "Channel transform bit"); + return AVERROR_PATCHWELCOME; } for (s->num_chgroups = 0; remaining_channels && @@ -677,7 +706,7 @@ static int decode_channel_transform(WMAProDecodeCtx* s) for (i = 0; i < s->channels_for_cur_subframe; i++) { int channel_idx = s->channel_indexes_for_cur_subframe[i]; if (!s->channel[channel_idx].grouped - && get_bits1(&s->gb)) { + && bitstream_read_bit(&s->bc)) { ++chgroup->num_channels; s->channel[channel_idx].grouped = 1; *channel_data++ = s->channel[channel_idx].coeffs; @@ -695,14 +724,15 @@ static int decode_channel_transform(WMAProDecodeCtx* s) /** decode transform type */ if (chgroup->num_channels == 2) { - if (get_bits1(&s->gb)) { - if (get_bits1(&s->gb)) { - av_log_ask_for_sample(s->avctx, - "unsupported channel transform type\n"); + if (bitstream_read_bit(&s->bc)) { + if (bitstream_read_bit(&s->bc)) { + avpriv_request_sample(s->avctx, + "Unknown channel transform type"); + return AVERROR_PATCHWELCOME; } } else { chgroup->transform = 1; - if (s->num_channels == 2) { + if (s->avctx->channels == 2) { chgroup->decorrelation_matrix[0] = 1.0; chgroup->decorrelation_matrix[1] = -1.0; chgroup->decorrelation_matrix[2] = 1.0; @@ -716,15 +746,15 @@ static int decode_channel_transform(WMAProDecodeCtx* s) } } } else if (chgroup->num_channels > 2) { - if (get_bits1(&s->gb)) { + if (bitstream_read_bit(&s->bc)) { chgroup->transform = 1; - if (get_bits1(&s->gb)) { + if (bitstream_read_bit(&s->bc)) { decode_decorrelation_matrix(s, chgroup); } else { /** FIXME: more than 6 coupled channels not supported */ if (chgroup->num_channels > 6) { - av_log_ask_for_sample(s->avctx, - "coupled channels > 6\n"); + avpriv_request_sample(s->avctx, + "Coupled channels > 6"); } else { memcpy(chgroup->decorrelation_matrix, default_decorrelation[chgroup->num_channels], @@ -737,11 +767,11 @@ static int decode_channel_transform(WMAProDecodeCtx* s) /** decode transform on / off */ if (chgroup->transform) { - if (!get_bits1(&s->gb)) { + if (!bitstream_read_bit(&s->bc)) { int i; /** transform can be enabled for individual bands */ for (i = 0; i < s->num_bands; i++) { - chgroup->transform_band[i] = get_bits1(&s->gb); + chgroup->transform_band[i] = bitstream_read_bit(&s->bc); } } else { memset(chgroup->transform_band, 1, s->num_bands); @@ -761,6 +791,15 @@ static int decode_channel_transform(WMAProDecodeCtx* s) */ static int decode_coeffs(WMAProDecodeCtx *s, int c) { + /* Integers 0..15 as single-precision floats. The table saves a + costly int to float conversion, and storing the values as + integers allows fast sign-flipping. */ + static const uint32_t fval_tab[16] = { + 0x00000000, 0x3f800000, 0x40000000, 0x40400000, + 0x40800000, 0x40a00000, 0x40c00000, 0x40e00000, + 0x41000000, 0x41100000, 0x41200000, 0x41300000, + 0x41400000, 0x41500000, 0x41600000, 0x41700000, + }; int vlctable; VLC* vlc; WMAProChannelCtx* ci = &s->channel[c]; @@ -768,11 +807,11 @@ static int decode_coeffs(WMAProDecodeCtx *s, int c) int cur_coeff = 0; int num_zeros = 0; const uint16_t* run; - const uint16_t* level; + const float* level; - dprintf(s->avctx, "decode coefficients for channel %i\n", c); + ff_dlog(s->avctx, "decode coefficients for channel %i\n", c); - vlctable = get_bits1(&s->gb); + vlctable = bitstream_read_bit(&s->bc); vlc = &coef_vlc[vlctable]; if (vlctable) { @@ -785,40 +824,44 @@ static int decode_coeffs(WMAProDecodeCtx *s, int c) /** decode vector coefficients (consumes up to 167 bits per iteration for 4 vector coded large values) */ - while (!rl_mode && cur_coeff + 3 < s->subframe_len) { - int vals[4]; + while ((s->transmit_num_vec_coeffs || !rl_mode) && + (cur_coeff + 3 < ci->num_vec_coeffs)) { + uint32_t vals[4]; int i; unsigned int idx; - idx = get_vlc2(&s->gb, vec4_vlc.table, VLCBITS, VEC4MAXDEPTH); + idx = bitstream_read_vlc(&s->bc, vec4_vlc.table, VLCBITS, VEC4MAXDEPTH); if (idx == HUFF_VEC4_SIZE - 1) { for (i = 0; i < 4; i += 2) { - idx = get_vlc2(&s->gb, vec2_vlc.table, VLCBITS, VEC2MAXDEPTH); + idx = bitstream_read_vlc(&s->bc, vec2_vlc.table, VLCBITS, VEC2MAXDEPTH); if (idx == HUFF_VEC2_SIZE - 1) { - vals[i] = get_vlc2(&s->gb, vec1_vlc.table, VLCBITS, VEC1MAXDEPTH); - if (vals[i] == HUFF_VEC1_SIZE - 1) - vals[i] += ff_wma_get_large_val(&s->gb); - vals[i+1] = get_vlc2(&s->gb, vec1_vlc.table, VLCBITS, VEC1MAXDEPTH); - if (vals[i+1] == HUFF_VEC1_SIZE - 1) - vals[i+1] += ff_wma_get_large_val(&s->gb); + uint32_t v0, v1; + v0 = bitstream_read_vlc(&s->bc, vec1_vlc.table, VLCBITS, VEC1MAXDEPTH); + if (v0 == HUFF_VEC1_SIZE - 1) + v0 += ff_wma_get_large_val(&s->bc); + v1 = bitstream_read_vlc(&s->bc, vec1_vlc.table, VLCBITS, VEC1MAXDEPTH); + if (v1 == HUFF_VEC1_SIZE - 1) + v1 += ff_wma_get_large_val(&s->bc); + vals[i ] = av_float2int(v0); + vals[i+1] = av_float2int(v1); } else { - vals[i] = symbol_to_vec2[idx] >> 4; - vals[i+1] = symbol_to_vec2[idx] & 0xF; + vals[i] = fval_tab[symbol_to_vec2[idx] >> 4 ]; + vals[i+1] = fval_tab[symbol_to_vec2[idx] & 0xF]; } } } else { - vals[0] = symbol_to_vec4[idx] >> 12; - vals[1] = (symbol_to_vec4[idx] >> 8) & 0xF; - vals[2] = (symbol_to_vec4[idx] >> 4) & 0xF; - vals[3] = symbol_to_vec4[idx] & 0xF; + vals[0] = fval_tab[ symbol_to_vec4[idx] >> 12 ]; + vals[1] = fval_tab[(symbol_to_vec4[idx] >> 8) & 0xF]; + vals[2] = fval_tab[(symbol_to_vec4[idx] >> 4) & 0xF]; + vals[3] = fval_tab[ symbol_to_vec4[idx] & 0xF]; } /** decode sign */ for (i = 0; i < 4; i++) { if (vals[i]) { - int sign = get_bits1(&s->gb) - 1; - ci->coeffs[cur_coeff] = (vals[i] ^ sign) - sign; + uint32_t sign = bitstream_read_bit(&s->bc) - 1; + AV_WN32A(&ci->coeffs[cur_coeff], vals[i] ^ sign << 31); num_zeros = 0; } else { ci->coeffs[cur_coeff] = 0; @@ -831,10 +874,10 @@ static int decode_coeffs(WMAProDecodeCtx *s, int c) } /** decode run level coded coefficients */ - if (rl_mode) { + if (cur_coeff < s->subframe_len) { memset(&ci->coeffs[cur_coeff], 0, sizeof(*ci->coeffs) * (s->subframe_len - cur_coeff)); - if (ff_wma_run_level_decode(s->avctx, &s->gb, vlc, + if (ff_wma_run_level_decode(s->avctx, &s->bc, vlc, level, run, 1, ci->coeffs, cur_coeff, s->subframe_len, s->subframe_len, s->esc_len, 0)) @@ -860,7 +903,9 @@ static int decode_scale_factors(WMAProDecodeCtx* s) for (i = 0; i < s->channels_for_cur_subframe; i++) { int c = s->channel_indexes_for_cur_subframe[i]; int* sf; - int* sf_end = s->channel[c].scale_factors + s->num_bands; + int* sf_end; + s->channel[c].scale_factors = s->channel[c].saved_scale_factors[!s->channel[c].scale_factor_idx]; + sf_end = s->channel[c].scale_factors + s->num_bands; /** resample scale factors for the new block size * as the scale factors might need to be resampled several times @@ -872,18 +917,17 @@ static int decode_scale_factors(WMAProDecodeCtx* s) int b; for (b = 0; b < s->num_bands; b++) s->channel[c].scale_factors[b] = - s->channel[c].saved_scale_factors[*sf_offsets++]; + s->channel[c].saved_scale_factors[s->channel[c].scale_factor_idx][*sf_offsets++]; } - if (!s->channel[c].cur_subframe || get_bits1(&s->gb)) { - + if (!s->channel[c].cur_subframe || bitstream_read_bit(&s->bc)) { if (!s->channel[c].reuse_sf) { int val; /** decode DPCM coded scale factors */ - s->channel[c].scale_factor_step = get_bits(&s->gb, 2) + 1; + s->channel[c].scale_factor_step = bitstream_read(&s->bc, 2) + 1; val = 45 / s->channel[c].scale_factor_step; for (sf = s->channel[c].scale_factors; sf < sf_end; sf++) { - val += get_vlc2(&s->gb, sf_vlc.table, SCALEVLCBITS, SCALEMAXDEPTH) - 60; + val += bitstream_read_vlc(&s->bc, sf_vlc.table, SCALEVLCBITS, SCALEMAXDEPTH) - 60; *sf = val; } } else { @@ -895,10 +939,10 @@ static int decode_scale_factors(WMAProDecodeCtx* s) int val; int sign; - idx = get_vlc2(&s->gb, sf_rl_vlc.table, VLCBITS, SCALERLMAXDEPTH); + idx = bitstream_read_vlc(&s->bc, sf_rl_vlc.table, VLCBITS, SCALERLMAXDEPTH); if (!idx) { - uint32_t code = get_bits(&s->gb, 14); + uint32_t code = bitstream_read(&s->bc, 14); val = code >> 6; sign = (code & 1) - 1; skip = (code & 0x3f) >> 1; @@ -907,7 +951,7 @@ static int decode_scale_factors(WMAProDecodeCtx* s) } else { skip = scale_rl_run[idx]; val = scale_rl_level[idx]; - sign = get_bits1(&s->gb)-1; + sign = bitstream_read_bit(&s->bc)-1; } i += skip; @@ -919,12 +963,8 @@ static int decode_scale_factors(WMAProDecodeCtx* s) s->channel[c].scale_factors[i] += (val ^ sign) - sign; } } - - /** save transmitted scale factors so that they can be reused for - the next subframe */ - memcpy(s->channel[c].saved_scale_factors, - s->channel[c].scale_factors, s->num_bands * - sizeof(*s->channel[c].saved_scale_factors)); + /** swap buffers */ + s->channel[c].scale_factor_idx = !s->channel[c].scale_factor_idx; s->channel[c].table_idx = s->table_idx; s->channel[c].reuse_sf = 1; } @@ -981,11 +1021,14 @@ static void inverse_channel_transform(WMAProDecodeCtx *s) (*ch)[y] = sum; } } - } else if (s->num_channels == 2) { - for (y = sfb[0]; y < FFMIN(sfb[1], s->subframe_len); y++) { - ch_data[0][y] *= 181.0 / 128; - ch_data[1][y] *= 181.0 / 128; - } + } else if (s->avctx->channels == 2) { + int len = FFMIN(sfb[1], s->subframe_len) - sfb[0]; + s->fdsp.vector_fmul_scalar(ch_data[0] + sfb[0], + ch_data[0] + sfb[0], + 181.0 / 128, len); + s->fdsp.vector_fmul_scalar(ch_data[1] + sfb[0], + ch_data[1] + sfb[0], + 181.0 / 128, len); } } } @@ -1010,12 +1053,12 @@ static void wmapro_window(WMAProDecodeCtx *s) winlen = s->subframe_len; } - window = s->windows[av_log2(winlen) - BLOCK_MIN_BITS]; + window = s->windows[av_log2(winlen) - WMAPRO_BLOCK_MIN_BITS]; winlen >>= 1; - s->dsp.vector_fmul_window(start, start, start + winlen, - window, 0, winlen); + s->fdsp.vector_fmul_window(start, start, start + winlen, + window, winlen); s->channel[c].prev_block_len = s->subframe_len; } @@ -1031,17 +1074,17 @@ static int decode_subframe(WMAProDecodeCtx *s) int offset = s->samples_per_frame; int subframe_len = s->samples_per_frame; int i; - int total_samples = s->samples_per_frame * s->num_channels; + int total_samples = s->samples_per_frame * s->avctx->channels; int transmit_coeffs = 0; int cur_subwoofer_cutoff; - s->subframe_offset = get_bits_count(&s->gb); + s->subframe_offset = bitstream_tell(&s->bc); /** reset channel context and find the next block offset and size == the next block of the channel with the smallest number of decoded samples */ - for (i = 0; i < s->num_channels; i++) { + for (i = 0; i < s->avctx->channels; i++) { s->channel[i].grouped = 0; if (offset > s->channel[i].decoded_samples) { offset = s->channel[i].decoded_samples; @@ -1050,14 +1093,14 @@ static int decode_subframe(WMAProDecodeCtx *s) } } - dprintf(s->avctx, + ff_dlog(s->avctx, "processing subframe with offset %i len %i\n", offset, subframe_len); /** get a list of all channels that contain the estimated block */ s->channels_for_cur_subframe = 0; - for (i = 0; i < s->num_channels; i++) { + for (i = 0; i < s->avctx->channels; i++) { const int cur_subframe = s->channel[i].cur_subframe; - /** substract already processed samples */ + /** subtract already processed samples */ total_samples -= s->channel[i].decoded_samples; /** and count if there are multiple subframes that match our profile */ @@ -1077,7 +1120,7 @@ static int decode_subframe(WMAProDecodeCtx *s) s->parsed_all_subframes = 1; - dprintf(s->avctx, "subframe is part of %i channels\n", + ff_dlog(s->avctx, "subframe is part of %i channels\n", s->channels_for_cur_subframe); /** calculate number of scale factor bands and their offsets */ @@ -1087,38 +1130,39 @@ static int decode_subframe(WMAProDecodeCtx *s) cur_subwoofer_cutoff = s->subwoofer_cutoffs[s->table_idx]; /** configure the decoder for the current subframe */ + offset += s->samples_per_frame >> 1; + for (i = 0; i < s->channels_for_cur_subframe; i++) { int c = s->channel_indexes_for_cur_subframe[i]; - s->channel[c].coeffs = &s->channel[c].out[(s->samples_per_frame >> 1) - + offset]; + s->channel[c].coeffs = &s->channel[c].out[offset]; } s->subframe_len = subframe_len; s->esc_len = av_log2(s->subframe_len - 1) + 1; /** skip extended header if any */ - if (get_bits1(&s->gb)) { + if (bitstream_read_bit(&s->bc)) { int num_fill_bits; - if (!(num_fill_bits = get_bits(&s->gb, 2))) { - int len = get_bits(&s->gb, 4); - num_fill_bits = get_bits(&s->gb, len) + 1; + if (!(num_fill_bits = bitstream_read(&s->bc, 2))) { + int len = bitstream_read(&s->bc, 4); + num_fill_bits = bitstream_read(&s->bc, len) + 1; } if (num_fill_bits >= 0) { - if (get_bits_count(&s->gb) + num_fill_bits > s->num_saved_bits) { + if (bitstream_tell(&s->bc) + num_fill_bits > s->num_saved_bits) { av_log(s->avctx, AV_LOG_ERROR, "invalid number of fill bits\n"); return AVERROR_INVALIDDATA; } - skip_bits_long(&s->gb, num_fill_bits); + bitstream_skip(&s->bc, num_fill_bits); } } /** no idea for what the following bit is used */ - if (get_bits1(&s->gb)) { - av_log_ask_for_sample(s->avctx, "reserved bit set\n"); - return AVERROR_INVALIDDATA; + if (bitstream_read_bit(&s->bc)) { + avpriv_request_sample(s->avctx, "Reserved bit"); + return AVERROR_PATCHWELCOME; } @@ -1128,26 +1172,40 @@ static int decode_subframe(WMAProDecodeCtx *s) for (i = 0; i < s->channels_for_cur_subframe; i++) { int c = s->channel_indexes_for_cur_subframe[i]; - if ((s->channel[c].transmit_coefs = get_bits1(&s->gb))) + if ((s->channel[c].transmit_coefs = bitstream_read_bit(&s->bc))) transmit_coeffs = 1; } if (transmit_coeffs) { int step; int quant_step = 90 * s->bits_per_sample >> 4; - if ((get_bits1(&s->gb))) { - /** FIXME: might change run level mode decision */ - av_log_ask_for_sample(s->avctx, "unsupported quant step coding\n"); - return AVERROR_INVALIDDATA; + + /** decode number of vector coded coefficients */ + if ((s->transmit_num_vec_coeffs = bitstream_read_bit(&s->bc))) { + int num_bits = av_log2((s->subframe_len + 3)/4) + 1; + for (i = 0; i < s->channels_for_cur_subframe; i++) { + int c = s->channel_indexes_for_cur_subframe[i]; + int num_vec_coeffs = bitstream_read(&s->bc, num_bits) << 2; + if (num_vec_coeffs + offset > FF_ARRAY_ELEMS(s->channel[c].out)) { + av_log(s->avctx, AV_LOG_ERROR, "num_vec_coeffs %d is too large\n", num_vec_coeffs); + return AVERROR_INVALIDDATA; + } + s->channel[c].num_vec_coeffs = num_vec_coeffs; + } + } else { + for (i = 0; i < s->channels_for_cur_subframe; i++) { + int c = s->channel_indexes_for_cur_subframe[i]; + s->channel[c].num_vec_coeffs = s->subframe_len; + } } /** decode quantization step */ - step = get_sbits(&s->gb, 6); + step = bitstream_read_signed(&s->bc, 6); quant_step += step; if (step == -32 || step == 31) { const int sign = (step == 31) - 1; int quant = 0; - while (get_bits_count(&s->gb) + 5 < s->num_saved_bits && - (step = get_bits(&s->gb, 5)) == 31) { + while (bitstream_tell(&s->bc) + 5 < s->num_saved_bits && + (step = bitstream_read(&s->bc, 5)) == 31) { quant += 31; } quant_step += ((quant + step) ^ sign) - sign; @@ -1161,13 +1219,13 @@ static int decode_subframe(WMAProDecodeCtx *s) if (s->channels_for_cur_subframe == 1) { s->channel[s->channel_indexes_for_cur_subframe[0]].quant_step = quant_step; } else { - int modifier_len = get_bits(&s->gb, 3); + int modifier_len = bitstream_read(&s->bc, 3); for (i = 0; i < s->channels_for_cur_subframe; i++) { int c = s->channel_indexes_for_cur_subframe[i]; s->channel[c].quant_step = quant_step; - if (get_bits1(&s->gb)) { + if (bitstream_read_bit(&s->bc)) { if (modifier_len) { - s->channel[c].quant_step += get_bits(&s->gb, modifier_len) + 1; + s->channel[c].quant_step += bitstream_read(&s->bc, modifier_len) + 1; } else ++s->channel[c].quant_step; } @@ -1179,24 +1237,25 @@ static int decode_subframe(WMAProDecodeCtx *s) return AVERROR_INVALIDDATA; } - dprintf(s->avctx, "BITSTREAM: subframe header length was %i\n", - get_bits_count(&s->gb) - s->subframe_offset); + ff_dlog(s->avctx, "BITSTREAM: subframe header length was %i\n", + bitstream_tell(&s->bc) - s->subframe_offset); /** parse coefficients */ for (i = 0; i < s->channels_for_cur_subframe; i++) { int c = s->channel_indexes_for_cur_subframe[i]; if (s->channel[c].transmit_coefs && - get_bits_count(&s->gb) < s->num_saved_bits) { + bitstream_tell(&s->bc) < s->num_saved_bits) { decode_coeffs(s, c); } else memset(s->channel[c].coeffs, 0, sizeof(*s->channel[c].coeffs) * subframe_len); } - dprintf(s->avctx, "BITSTREAM: subframe length was %i\n", - get_bits_count(&s->gb) - s->subframe_offset); + ff_dlog(s->avctx, "BITSTREAM: subframe length was %i\n", + bitstream_tell(&s->bc) - s->subframe_offset); if (transmit_coeffs) { + FFTContext *mdct = &s->mdct_ctx[av_log2(subframe_len) - WMAPRO_BLOCK_MIN_BITS]; /** reconstruct the per channel data */ inverse_channel_transform(s); for (i = 0; i < s->channels_for_cur_subframe; i++) { @@ -1215,15 +1274,14 @@ static int decode_subframe(WMAProDecodeCtx *s) (s->channel[c].max_scale_factor - *sf++) * s->channel[c].scale_factor_step; const float quant = pow(10.0, exp / 20.0); - int start; - - for (start = s->cur_sfb_offsets[b]; start < end; start++) - s->tmp[start] = s->channel[c].coeffs[start] * quant; + int start = s->cur_sfb_offsets[b]; + s->fdsp.vector_fmul_scalar(s->tmp + start, + s->channel[c].coeffs + start, + quant, end - start); } - /** apply imdct (ff_imdct_half == DCTIV with reverse) */ - ff_imdct_half(&s->mdct_ctx[av_log2(subframe_len) - BLOCK_MIN_BITS], - s->channel[c].coeffs, s->tmp); + /** apply imdct (imdct_half == DCTIV with reverse) */ + mdct->imdct_half(mdct, s->channel[c].coeffs, s->tmp); } } @@ -1249,30 +1307,19 @@ static int decode_subframe(WMAProDecodeCtx *s) *@return 0 if the trailer bit indicates that this is the last frame, * 1 if there are additional frames */ -static int decode_frame(WMAProDecodeCtx *s) +static int decode_frame(WMAProDecodeCtx *s, AVFrame *frame, int *got_frame_ptr) { - GetBitContext* gb = &s->gb; + AVCodecContext *avctx = s->avctx; + BitstreamContext *bc = &s->bc; int more_frames = 0; int len = 0; - int i; - - /** check for potential output buffer overflow */ - if (s->num_channels * s->samples_per_frame > s->samples_end - s->samples) { - /** return an error if no frame could be decoded at all */ - if (s->samples_start == s->samples) { - av_log(s->avctx, AV_LOG_ERROR, - "not enough space for the output samples\n"); - s->packet_loss = 1; - } else - s->output_buffer_full = 1; - return 0; - } + int i, ret; /** get frame length */ if (s->len_prefix) - len = get_bits(gb, s->log2_frame_size); + len = bitstream_read(bc, s->log2_frame_size); - dprintf(s->avctx, "decoding frame with length %x\n", len); + ff_dlog(s->avctx, "decoding frame with length %x\n", len); /** decode tile information */ if (decode_tilehdr(s)) { @@ -1281,43 +1328,44 @@ static int decode_frame(WMAProDecodeCtx *s) } /** read postproc transform */ - if (s->num_channels > 1 && get_bits1(gb)) { - av_log_ask_for_sample(s->avctx, "Unsupported postproc transform found\n"); - s->packet_loss = 1; - return 0; + if (s->avctx->channels > 1 && bitstream_read_bit(bc)) { + if (bitstream_read_bit(bc)) { + for (i = 0; i < avctx->channels * avctx->channels; i++) + bitstream_skip(bc, 4); + } } /** read drc info */ if (s->dynamic_range_compression) { - s->drc_gain = get_bits(gb, 8); - dprintf(s->avctx, "drc_gain %i\n", s->drc_gain); + s->drc_gain = bitstream_read(bc, 8); + ff_dlog(s->avctx, "drc_gain %i\n", s->drc_gain); } /** no idea what these are for, might be the number of samples that need to be skipped at the beginning or end of a stream */ - if (get_bits1(gb)) { - int skip; + if (bitstream_read_bit(bc)) { + int av_unused skip; /** usually true for the first frame */ - if (get_bits1(gb)) { - skip = get_bits(gb, av_log2(s->samples_per_frame * 2)); - dprintf(s->avctx, "start skip: %i\n", skip); + if (bitstream_read_bit(bc)) { + skip = bitstream_read(bc, av_log2(s->samples_per_frame * 2)); + ff_dlog(s->avctx, "start skip: %i\n", skip); } /** sometimes true for the last frame */ - if (get_bits1(gb)) { - skip = get_bits(gb, av_log2(s->samples_per_frame * 2)); - dprintf(s->avctx, "end skip: %i\n", skip); + if (bitstream_read_bit(bc)) { + skip = bitstream_read(bc, av_log2(s->samples_per_frame * 2)); + ff_dlog(s->avctx, "end skip: %i\n", skip); } } - dprintf(s->avctx, "BITSTREAM: frame header length was %i\n", - get_bits_count(gb) - s->frame_offset); + ff_dlog(s->avctx, "BITSTREAM: frame header length was %i\n", + bitstream_tell(bc) - s->frame_offset); /** reset subframe states */ s->parsed_all_subframes = 0; - for (i = 0; i < s->num_channels; i++) { + for (i = 0; i < avctx->channels; i++) { s->channel[i].decoded_samples = 0; s->channel[i].cur_subframe = 0; s->channel[i].reuse_sf = 0; @@ -1331,20 +1379,20 @@ static int decode_frame(WMAProDecodeCtx *s) } } - /** interleave samples and write them to the output buffer */ - for (i = 0; i < s->num_channels; i++) { - float* ptr; - int incr = s->num_channels; - float* iptr = s->channel[i].out; - int x; - - ptr = s->samples + i; + /* get output buffer */ + frame->nb_samples = s->samples_per_frame; + if ((ret = ff_get_buffer(avctx, frame, 0)) < 0) { + av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n"); + s->packet_loss = 1; + return 0; + } - for (x = 0; x < s->samples_per_frame; x++) { - *ptr = av_clipf(*iptr++, -1.0, 32767.0 / 32768.0); - ptr += incr; - } + /** copy samples to the output buffer */ + for (i = 0; i < avctx->channels; i++) + memcpy(frame->extended_data[i], s->channel[i].out, + s->samples_per_frame * sizeof(*s->channel[i].out)); + for (i = 0; i < avctx->channels; i++) { /** reuse second half of the IMDCT output for the next frame */ memcpy(&s->channel[i].out[0], &s->channel[i].out[s->samples_per_frame], @@ -1353,22 +1401,32 @@ static int decode_frame(WMAProDecodeCtx *s) if (s->skip_frame) { s->skip_frame = 0; - } else - s->samples += s->num_channels * s->samples_per_frame; - - if (len != (get_bits_count(gb) - s->frame_offset) + 2) { - /** FIXME: not sure if this is always an error */ - av_log(s->avctx, AV_LOG_ERROR, "frame[%i] would have to skip %i bits\n", - s->frame_num, len - (get_bits_count(gb) - s->frame_offset) - 1); - s->packet_loss = 1; - return 0; + *got_frame_ptr = 0; + av_frame_unref(frame); + } else { + *got_frame_ptr = 1; } - /** skip the rest of the frame data */ - skip_bits_long(gb, len - (get_bits_count(gb) - s->frame_offset) - 1); + if (s->len_prefix) { + if (len != (bitstream_tell(bc) - s->frame_offset) + 2) { + /** FIXME: not sure if this is always an error */ + av_log(s->avctx, AV_LOG_ERROR, + "frame[%"PRIu32"] would have to skip %i bits\n", + s->frame_num, + len - (bitstream_tell(bc) - s->frame_offset) - 1); + s->packet_loss = 1; + return 0; + } + + /** skip the rest of the frame data */ + bitstream_skip(bc, len - (bitstream_tell(bc) - s->frame_offset) - 1); + } else { + while (bitstream_tell(bc) < s->num_saved_bits && bitstream_read_bit(bc) == 0) { + } + } /** decode trailer bit */ - more_frames = get_bits1(gb); + more_frames = bitstream_read_bit(bc); ++s->frame_num; return more_frames; @@ -1377,32 +1435,32 @@ static int decode_frame(WMAProDecodeCtx *s) /** *@brief Calculate remaining input buffer length. *@param s codec context - *@param gb bitstream reader context + *@param bc bitstream reader context *@return remaining size in bits */ -static int remaining_bits(WMAProDecodeCtx *s, GetBitContext *gb) +static int remaining_bits(WMAProDecodeCtx *s, BitstreamContext *bc) { - return s->buf_bit_size - get_bits_count(gb); + return s->buf_bit_size - bitstream_tell(bc); } /** *@brief Fill the bit reservoir with a (partial) frame. *@param s codec context - *@param gb bitstream reader context + *@param bc bitstream reader context *@param len length of the partial frame - *@param append decides wether to reset the buffer or not + *@param append decides whether to reset the buffer or not */ -static void save_bits(WMAProDecodeCtx *s, GetBitContext* gb, int len, +static void save_bits(WMAProDecodeCtx *s, BitstreamContext *bc, int len, int append) { int buflen; /** when the frame data does not need to be concatenated, the input buffer - is resetted and additional bits from the previous frame are copyed + is reset and additional bits from the previous frame are copied and skipped later so that a fast byte copy is possible */ if (!append) { - s->frame_offset = get_bits_count(gb) & 7; + s->frame_offset = bitstream_tell(bc) & 7; s->num_saved_bits = s->frame_offset; init_put_bits(&s->pb, s->frame_data, MAX_FRAMESIZE); } @@ -1410,134 +1468,159 @@ static void save_bits(WMAProDecodeCtx *s, GetBitContext* gb, int len, buflen = (s->num_saved_bits + len + 8) >> 3; if (len <= 0 || buflen > MAX_FRAMESIZE) { - av_log_ask_for_sample(s->avctx, "input buffer too small\n"); + avpriv_request_sample(s->avctx, "Too small input buffer"); + s->packet_loss = 1; + return; + } + + if (len > put_bits_left(&s->pb)) { + av_log(s->avctx, AV_LOG_ERROR, + "Cannot append %d bits, only %d bits available.\n", + len, put_bits_left(&s->pb)); s->packet_loss = 1; return; } s->num_saved_bits += len; if (!append) { - ff_copy_bits(&s->pb, gb->buffer + (get_bits_count(gb) >> 3), - s->num_saved_bits); + avpriv_copy_bits(&s->pb, bc->buffer + (bitstream_tell(bc) >> 3), + s->num_saved_bits); } else { - int align = 8 - (get_bits_count(gb) & 7); + int align = 8 - (bitstream_tell(bc) & 7); align = FFMIN(align, len); - put_bits(&s->pb, align, get_bits(gb, align)); + put_bits(&s->pb, align, bitstream_read(bc, align)); len -= align; - ff_copy_bits(&s->pb, gb->buffer + (get_bits_count(gb) >> 3), len); + avpriv_copy_bits(&s->pb, bc->buffer + (bitstream_tell(bc) >> 3), len); } - skip_bits_long(gb, len); + bitstream_skip(bc, len); { PutBitContext tmp = s->pb; flush_put_bits(&tmp); } - init_get_bits(&s->gb, s->frame_data, s->num_saved_bits); - skip_bits(&s->gb, s->frame_offset); + bitstream_init(&s->bc, s->frame_data, s->num_saved_bits); + bitstream_skip(&s->bc, s->frame_offset); } /** *@brief Decode a single WMA packet. *@param avctx codec context *@param data the output buffer - *@param data_size number of bytes that were written to the output buffer *@param avpkt input packet *@return number of bytes that were read from the input buffer */ -static int decode_packet(AVCodecContext *avctx, - void *data, int *data_size, AVPacket* avpkt) +static int decode_packet(AVCodecContext *avctx, void *data, + int *got_frame_ptr, AVPacket* avpkt) { WMAProDecodeCtx *s = avctx->priv_data; - GetBitContext* gb = &s->pgb; - const uint8_t* buf = avpkt->data; - int buf_size = avpkt->size; - int more_frames = 1; + BitstreamContext *bc = &s->pbc; + const uint8_t* buf = avpkt->data; + int buf_size = avpkt->size; int num_bits_prev_frame; int packet_sequence_number; - s->samples = data; - s->samples_start = data; - s->samples_end = (float*)((int8_t*)data + *data_size); - *data_size = 0; - - if (!s->output_buffer_full) { - s->buf_bit_size = buf_size << 3; - - /** sanity check for the buffer length */ - if (buf_size < avctx->block_align) - return 0; + *got_frame_ptr = 0; - buf_size = avctx->block_align; + if (s->packet_done || s->packet_loss) { + s->packet_done = 0; - /** parse packet header */ - init_get_bits(gb, buf, s->buf_bit_size); - packet_sequence_number = get_bits(gb, 4); - skip_bits(gb, 2); + /** sanity check for the buffer length */ + if (buf_size < avctx->block_align) { + av_log(avctx, AV_LOG_ERROR, "Input packet too small (%d < %d)\n", + buf_size, avctx->block_align); + return AVERROR_INVALIDDATA; + } - /** get number of bits that need to be added to the previous frame */ - num_bits_prev_frame = get_bits(gb, s->log2_frame_size); - dprintf(avctx, "packet[%d]: nbpf %x\n", avctx->frame_number, - num_bits_prev_frame); + s->next_packet_start = buf_size - avctx->block_align; + buf_size = avctx->block_align; + s->buf_bit_size = buf_size << 3; - /** check for packet loss */ - if (!s->packet_loss && - ((s->packet_sequence_number + 1) & 0xF) != packet_sequence_number) { - s->packet_loss = 1; - av_log(avctx, AV_LOG_ERROR, "Packet loss detected! seq %x vs %x\n", - s->packet_sequence_number, packet_sequence_number); - } - s->packet_sequence_number = packet_sequence_number; - - if (num_bits_prev_frame > 0) { - /** append the previous frame data to the remaining data from the - previous packet to create a full frame */ - save_bits(s, gb, num_bits_prev_frame, 1); - dprintf(avctx, "accumulated %x bits of frame data\n", - s->num_saved_bits - s->frame_offset); - - /** decode the cross packet frame if it is valid */ - if (!s->packet_loss) - decode_frame(s); - } else if (s->num_saved_bits - s->frame_offset) { - dprintf(avctx, "ignoring %x previously saved bits\n", - s->num_saved_bits - s->frame_offset); - } + /** parse packet header */ + bitstream_init(bc, buf, s->buf_bit_size); + packet_sequence_number = bitstream_read(bc, 4); + bitstream_skip(bc, 2); - s->packet_loss = 0; + /** get number of bits that need to be added to the previous frame */ + num_bits_prev_frame = bitstream_read(bc, s->log2_frame_size); + ff_dlog(avctx, "packet[%d]: nbpf %x\n", avctx->frame_number, + num_bits_prev_frame); - } else { - /** continue decoding */ - s->output_buffer_full = 0; - more_frames = decode_frame(s); - } + /** check for packet loss */ + if (!s->packet_loss && + ((s->packet_sequence_number + 1) & 0xF) != packet_sequence_number) { + s->packet_loss = 1; + av_log(avctx, AV_LOG_ERROR, + "Packet loss detected! seq %"PRIx8" vs %x\n", + s->packet_sequence_number, packet_sequence_number); + } + s->packet_sequence_number = packet_sequence_number; - /** decode the rest of the packet */ - while (!s->packet_loss && !s->output_buffer_full && more_frames && - remaining_bits(s, gb) > s->log2_frame_size) { - int frame_size = show_bits(gb, s->log2_frame_size); + if (num_bits_prev_frame > 0) { + int remaining_packet_bits = s->buf_bit_size - bitstream_tell(bc); + if (num_bits_prev_frame >= remaining_packet_bits) { + num_bits_prev_frame = remaining_packet_bits; + s->packet_done = 1; + } - /** there is enough data for a full frame */ - if (remaining_bits(s, gb) >= frame_size && frame_size > 0) { - save_bits(s, gb, frame_size, 0); + /** append the previous frame data to the remaining data from the + previous packet to create a full frame */ + save_bits(s, bc, num_bits_prev_frame, 1); + ff_dlog(avctx, "accumulated %x bits of frame data\n", + s->num_saved_bits - s->frame_offset); + + /** decode the cross packet frame if it is valid */ + if (!s->packet_loss) + decode_frame(s, data, got_frame_ptr); + } else if (s->num_saved_bits - s->frame_offset) { + ff_dlog(avctx, "ignoring %x previously saved bits\n", + s->num_saved_bits - s->frame_offset); + } - /** decode the frame */ - more_frames = decode_frame(s); + if (s->packet_loss) { + /** reset number of saved bits so that the decoder + does not start to decode incomplete frames in the + s->len_prefix == 0 case */ + s->num_saved_bits = 0; + s->packet_loss = 0; + } + } else { + int frame_size; + s->buf_bit_size = (avpkt->size - s->next_packet_start) << 3; + bitstream_init(bc, avpkt->data, s->buf_bit_size); + bitstream_skip(bc, s->packet_offset); + if (s->len_prefix && remaining_bits(s, bc) > s->log2_frame_size && + (frame_size = bitstream_peek(bc, s->log2_frame_size)) && + frame_size <= remaining_bits(s, bc)) { + save_bits(s, bc, frame_size, 0); + s->packet_done = !decode_frame(s, data, got_frame_ptr); + } else if (!s->len_prefix + && s->num_saved_bits > bitstream_tell(&s->bc)) { + /** when the frames do not have a length prefix, we don't know + the compressed length of the individual frames + however, we know what part of a new packet belongs to the + previous frame + therefore we save the incoming packet first, then we append + the "previous frame" data from the next packet so that + we get a buffer that only contains full frames */ + s->packet_done = !decode_frame(s, data, got_frame_ptr); } else - more_frames = 0; + s->packet_done = 1; } - if (!s->output_buffer_full && !s->packet_loss && - remaining_bits(s, gb) > 0) { + if (s->packet_done && !s->packet_loss && + remaining_bits(s, bc) > 0) { /** save the rest of the data so that it can be decoded with the next packet */ - save_bits(s, gb, remaining_bits(s, gb), 0); + save_bits(s, bc, remaining_bits(s, bc), 0); } - *data_size = (int8_t *)s->samples - (int8_t *)data; + s->packet_offset = bitstream_tell(bc) & 7; + if (s->packet_loss) + return AVERROR_INVALIDDATA; - return (s->output_buffer_full)?0: avctx->block_align; + return bitstream_tell(bc) >> 3; } /** @@ -1550,7 +1633,7 @@ static void flush(AVCodecContext *avctx) int i; /** reset output buffer as a part of it is used during the windowing of a new frame */ - for (i = 0; i < s->num_channels; i++) + for (i = 0; i < avctx->channels; i++) memset(s->channel[i].out, 0, s->samples_per_frame * sizeof(*s->channel[i].out)); s->packet_loss = 1; @@ -1560,15 +1643,17 @@ static void flush(AVCodecContext *avctx) /** *@brief wmapro decoder */ -AVCodec wmapro_decoder = { - "wmapro", - CODEC_TYPE_AUDIO, - CODEC_ID_WMAPRO, - sizeof(WMAProDecodeCtx), - decode_init, - NULL, - decode_end, - decode_packet, - .flush= flush, - .long_name = NULL_IF_CONFIG_SMALL("Windows Media Audio 9 Professional"), +AVCodec ff_wmapro_decoder = { + .name = "wmapro", + .long_name = NULL_IF_CONFIG_SMALL("Windows Media Audio 9 Professional"), + .type = AVMEDIA_TYPE_AUDIO, + .id = AV_CODEC_ID_WMAPRO, + .priv_data_size = sizeof(WMAProDecodeCtx), + .init = decode_init, + .close = decode_end, + .decode = decode_packet, + .capabilities = AV_CODEC_CAP_SUBFRAMES | AV_CODEC_CAP_DR1, + .flush = flush, + .sample_fmts = (const enum AVSampleFormat[]) { AV_SAMPLE_FMT_FLTP, + AV_SAMPLE_FMT_NONE }, };