X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=libavcodec%2Fwmaprodec.c;h=5e063764c658ebefe8e9875ec75924bb5fce5511;hb=7f9f771eac0d37a632e0ed9bd89961d57fcfb7e0;hp=922b75beffd15da744b49d391aef2c634132fbe3;hpb=c5c20ae4367905f0cf3080d46aa3bed3806134b9;p=ffmpeg diff --git a/libavcodec/wmaprodec.c b/libavcodec/wmaprodec.c index 922b75beffd..5e063764c65 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: @@ -86,13 +86,19 @@ * 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 "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 @@ -100,9 +106,11 @@ #define MAX_BANDS 29 ///< max number of scale factor bands #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 +127,12 @@ 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; @@ -142,13 +150,14 @@ typedef struct { 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 @@ -162,12 +171,12 @@ 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 PutBitContext pb; ///< context for filling the frame_data buffer FFTContext mdct_ctx[WMAPRO_BLOCK_SIZES]; ///< MDCT context per block size - DECLARE_ALIGNED(16, float, tmp)[WMAPRO_BLOCK_MAX_SIZE]; ///< IMDCT output buffer + 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) */ @@ -177,7 +186,6 @@ 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 @@ -190,6 +198,7 @@ typedef struct WMAProDecodeCtx { /* packet decode state */ GetBitContext pgb; ///< 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 @@ -202,8 +211,6 @@ typedef struct WMAProDecodeCtx { uint32_t frame_num; ///< current frame number (not used for decoding) GetBitContext gb; ///< bitstream reader context int buf_bit_size; ///< buffer size in bits - 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? @@ -213,6 +220,7 @@ 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 @@ -228,10 +236,10 @@ 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); @@ -239,7 +247,7 @@ 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); } /** @@ -268,15 +276,21 @@ 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 & 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); @@ -284,39 +298,34 @@ 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"); + av_dlog(avctx, "[%x] ", avctx->extradata[i]); + av_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; @@ -325,12 +334,35 @@ 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; + } + + 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; } - s->num_channels = avctx->channels; + 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; @@ -343,14 +375,6 @@ static av_cold int decode_init(AVCodecContext *avctx) } } - if (s->num_channels < 0) { - av_log(avctx, AV_LOG_ERROR, "invalid number of channels %d\n", s->num_channels); - return AVERROR_INVALIDDATA; - } else if (s->num_channels > WMAPRO_MAX_CHANNELS) { - av_log_ask_for_sample(avctx, "unsupported number of channels\n"); - return AVERROR_PATCHWELCOME; - } - INIT_VLC_STATIC(&sf_vlc, SCALEVLCBITS, HUFF_SCALE_SIZE, scale_huffbits, 1, 1, scale_huffcodes, 2, 2, 616); @@ -414,7 +438,8 @@ 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; } } @@ -422,11 +447,11 @@ 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 win_idx = WMAPRO_BLOCK_MAX_BITS - i; ff_init_ff_sine_windows(win_idx); @@ -449,6 +474,7 @@ static av_cold int decode_init(AVCodecContext *avctx) dump_context(s); avctx->channel_layout = channel_mask; + return 0; } @@ -508,25 +534,23 @@ 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)) fixed_channel_layout = 1; @@ -535,7 +559,7 @@ 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)) @@ -552,7 +576,7 @@ 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]) { @@ -579,11 +603,11 @@ 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]" + av_dlog(s->avctx, "frame[%i] channel[%i] subframe[%i]" " len %i\n", s->frame_num, c, i, s->channel[c].subframe_len[i]); s->channel[c].subframe_offset[i] = offset; @@ -605,8 +629,8 @@ 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); @@ -659,13 +683,13 @@ 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; + avpriv_request_sample(s->avctx, + "Channel transform bit"); + return AVERROR_PATCHWELCOME; } for (s->num_chgroups = 0; remaining_channels && @@ -700,12 +724,13 @@ static int decode_channel_transform(WMAProDecodeCtx* s) 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"); + 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; @@ -726,8 +751,8 @@ static int decode_channel_transform(WMAProDecodeCtx* s) } 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], @@ -767,7 +792,7 @@ 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 int fval_tab[16] = { + static const uint32_t fval_tab[16] = { 0x00000000, 0x3f800000, 0x40000000, 0x40400000, 0x40800000, 0x40a00000, 0x40c00000, 0x40e00000, 0x41000000, 0x41100000, 0x41200000, 0x41300000, @@ -782,7 +807,7 @@ static int decode_coeffs(WMAProDecodeCtx *s, int c) const uint16_t* run; const float* level; - dprintf(s->avctx, "decode coefficients for channel %i\n", c); + av_dlog(s->avctx, "decode coefficients for channel %i\n", c); vlctable = get_bits1(&s->gb); vlc = &coef_vlc[vlctable]; @@ -797,8 +822,9 @@ 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; @@ -808,15 +834,15 @@ static int decode_coeffs(WMAProDecodeCtx *s, int c) for (i = 0; i < 4; i += 2) { idx = get_vlc2(&s->gb, vec2_vlc.table, VLCBITS, VEC2MAXDEPTH); if (idx == HUFF_VEC2_SIZE - 1) { - int v0, v1; + uint32_t v0, v1; v0 = get_vlc2(&s->gb, vec1_vlc.table, VLCBITS, VEC1MAXDEPTH); if (v0 == HUFF_VEC1_SIZE - 1) v0 += ff_wma_get_large_val(&s->gb); v1 = get_vlc2(&s->gb, vec1_vlc.table, VLCBITS, VEC1MAXDEPTH); if (v1 == HUFF_VEC1_SIZE - 1) v1 += ff_wma_get_large_val(&s->gb); - ((float*)vals)[i ] = v0; - ((float*)vals)[i+1] = v1; + vals[i ] = av_float2int(v0); + vals[i+1] = av_float2int(v1); } else { vals[i] = fval_tab[symbol_to_vec2[idx] >> 4 ]; vals[i+1] = fval_tab[symbol_to_vec2[idx] & 0xF]; @@ -832,8 +858,8 @@ static int decode_coeffs(WMAProDecodeCtx *s, int c) /** decode sign */ for (i = 0; i < 4; i++) { if (vals[i]) { - int sign = get_bits1(&s->gb) - 1; - *(uint32_t*)&ci->coeffs[cur_coeff] = vals[i] ^ sign<<31; + uint32_t sign = get_bits1(&s->gb) - 1; + AV_WN32A(&ci->coeffs[cur_coeff], vals[i] ^ sign << 31); num_zeros = 0; } else { ci->coeffs[cur_coeff] = 0; @@ -846,7 +872,7 @@ 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, @@ -994,14 +1020,14 @@ static void inverse_channel_transform(WMAProDecodeCtx *s) (*ch)[y] = sum; } } - } else if (s->num_channels == 2) { + } else if (s->avctx->channels == 2) { int len = FFMIN(sfb[1], s->subframe_len) - sfb[0]; - s->dsp.vector_fmul_scalar(ch_data[0] + sfb[0], - ch_data[0] + sfb[0], - 181.0 / 128, len); - s->dsp.vector_fmul_scalar(ch_data[1] + sfb[0], - ch_data[1] + sfb[0], - 181.0 / 128, len); + 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); } } } @@ -1026,12 +1052,12 @@ 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; } @@ -1047,7 +1073,7 @@ 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; @@ -1057,7 +1083,7 @@ static int decode_subframe(WMAProDecodeCtx *s) == 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; @@ -1066,14 +1092,14 @@ static int decode_subframe(WMAProDecodeCtx *s) } } - dprintf(s->avctx, + av_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 */ @@ -1093,7 +1119,7 @@ static int decode_subframe(WMAProDecodeCtx *s) s->parsed_all_subframes = 1; - dprintf(s->avctx, "subframe is part of %i channels\n", + av_dlog(s->avctx, "subframe is part of %i channels\n", s->channels_for_cur_subframe); /** calculate number of scale factor bands and their offsets */ @@ -1103,11 +1129,12 @@ 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; @@ -1133,8 +1160,8 @@ static int decode_subframe(WMAProDecodeCtx *s) /** 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; + avpriv_request_sample(s->avctx, "Reserved bit"); + return AVERROR_PATCHWELCOME; } @@ -1151,10 +1178,24 @@ static int decode_subframe(WMAProDecodeCtx *s) 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 = get_bits1(&s->gb))) { + 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 = get_bits(&s->gb, 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); @@ -1195,7 +1236,7 @@ static int decode_subframe(WMAProDecodeCtx *s) return AVERROR_INVALIDDATA; } - dprintf(s->avctx, "BITSTREAM: subframe header length was %i\n", + av_dlog(s->avctx, "BITSTREAM: subframe header length was %i\n", get_bits_count(&s->gb) - s->subframe_offset); /** parse coefficients */ @@ -1209,10 +1250,11 @@ static int decode_subframe(WMAProDecodeCtx *s) sizeof(*s->channel[c].coeffs) * subframe_len); } - dprintf(s->avctx, "BITSTREAM: subframe length was %i\n", + av_dlog(s->avctx, "BITSTREAM: subframe length was %i\n", get_bits_count(&s->gb) - 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++) { @@ -1232,14 +1274,13 @@ static int decode_subframe(WMAProDecodeCtx *s) s->channel[c].scale_factor_step; const float quant = pow(10.0, exp / 20.0); int start = s->cur_sfb_offsets[b]; - s->dsp.vector_fmul_scalar(s->tmp + start, - s->channel[c].coeffs + start, - quant, end - start); + 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); } } @@ -1265,27 +1306,19 @@ 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) { + AVCodecContext *avctx = s->avctx; GetBitContext* gb = &s->gb; 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 */ - av_log(s->avctx, AV_LOG_ERROR, - "not enough space for the output samples\n"); - s->packet_loss = 1; - return 0; - } + int i, ret; /** get frame length */ if (s->len_prefix) len = get_bits(gb, s->log2_frame_size); - dprintf(s->avctx, "decoding frame with length %x\n", len); + av_dlog(s->avctx, "decoding frame with length %x\n", len); /** decode tile information */ if (decode_tilehdr(s)) { @@ -1294,43 +1327,44 @@ 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 && get_bits1(gb)) { + if (get_bits1(gb)) { + for (i = 0; i < avctx->channels * avctx->channels; i++) + skip_bits(gb, 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); + av_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; + 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); + av_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); + av_dlog(s->avctx, "end skip: %i\n", skip); } } - dprintf(s->avctx, "BITSTREAM: frame header length was %i\n", + av_dlog(s->avctx, "BITSTREAM: frame header length was %i\n", get_bits_count(gb) - 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; @@ -1344,20 +1378,20 @@ 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], @@ -1366,19 +1400,29 @@ 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 != (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[%"PRIu32"] 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; + } + + /** skip the rest of the frame data */ + skip_bits_long(gb, len - (get_bits_count(gb) - s->frame_offset) - 1); + } else { + while (get_bits_count(gb) < s->num_saved_bits && get_bits1(gb) == 0) { + } + } /** decode trailer bit */ more_frames = get_bits1(gb); @@ -1403,7 +1447,7 @@ static int remaining_bits(WMAProDecodeCtx *s, GetBitContext *gb) *@param s codec context *@param gb 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, int append) @@ -1423,21 +1467,29 @@ 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), + avpriv_copy_bits(&s->pb, gb->buffer + (get_bits_count(gb) >> 3), s->num_saved_bits); } else { int align = 8 - (get_bits_count(gb) & 7); align = FFMIN(align, len); put_bits(&s->pb, align, get_bits(gb, align)); len -= align; - ff_copy_bits(&s->pb, gb->buffer + (get_bits_count(gb) >> 3), len); + avpriv_copy_bits(&s->pb, gb->buffer + (get_bits_count(gb) >> 3), len); } skip_bits_long(gb, len); @@ -1454,12 +1506,11 @@ static void save_bits(WMAProDecodeCtx *s, GetBitContext* gb, int len, *@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; @@ -1468,19 +1519,21 @@ static int decode_packet(AVCodecContext *avctx, int num_bits_prev_frame; int packet_sequence_number; - s->samples = data; - s->samples_end = (float*)((int8_t*)data + *data_size); - *data_size = 0; + *got_frame_ptr = 0; if (s->packet_done || s->packet_loss) { s->packet_done = 0; - s->buf_bit_size = buf_size << 3; /** sanity check for the buffer length */ - if (buf_size < avctx->block_align) - return 0; + 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; + } + s->next_packet_start = buf_size - avctx->block_align; buf_size = avctx->block_align; + s->buf_bit_size = buf_size << 3; /** parse packet header */ init_get_bits(gb, buf, s->buf_bit_size); @@ -1489,45 +1542,68 @@ static int decode_packet(AVCodecContext *avctx, /** 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, + av_dlog(avctx, "packet[%d]: nbpf %x\n", avctx->frame_number, num_bits_prev_frame); /** 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", + 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; if (num_bits_prev_frame > 0) { + int remaining_packet_bits = s->buf_bit_size - get_bits_count(gb); + if (num_bits_prev_frame >= remaining_packet_bits) { + num_bits_prev_frame = remaining_packet_bits; + s->packet_done = 1; + } + /** 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", + av_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); + decode_frame(s, data, got_frame_ptr); } else if (s->num_saved_bits - s->frame_offset) { - dprintf(avctx, "ignoring %x previously saved bits\n", + av_dlog(avctx, "ignoring %x previously saved bits\n", s->num_saved_bits - s->frame_offset); } - s->packet_loss = 0; + 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 << 3; + s->buf_bit_size = (avpkt->size - s->next_packet_start) << 3; init_get_bits(gb, avpkt->data, s->buf_bit_size); skip_bits(gb, s->packet_offset); - if (remaining_bits(s, gb) > s->log2_frame_size && + if (s->len_prefix && remaining_bits(s, gb) > s->log2_frame_size && (frame_size = show_bits(gb, s->log2_frame_size)) && frame_size <= remaining_bits(s, gb)) { save_bits(s, gb, frame_size, 0); - s->packet_done = !decode_frame(s); + s->packet_done = !decode_frame(s, data, got_frame_ptr); + } else if (!s->len_prefix + && s->num_saved_bits > get_bits_count(&s->gb)) { + /** 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 s->packet_done = 1; } @@ -1539,10 +1615,11 @@ static int decode_packet(AVCodecContext *avctx, save_bits(s, gb, remaining_bits(s, gb), 0); } - *data_size = (int8_t *)s->samples - (int8_t *)data; s->packet_offset = get_bits_count(gb) & 7; + if (s->packet_loss) + return AVERROR_INVALIDDATA; - return (s->packet_loss) ? AVERROR_INVALIDDATA : get_bits_count(gb) >> 3; + return get_bits_count(gb) >> 3; } /** @@ -1555,7 +1632,7 @@ 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; @@ -1565,16 +1642,17 @@ 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, - .capabilities = CODEC_CAP_SUBFRAMES, - .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 = CODEC_CAP_SUBFRAMES | CODEC_CAP_DR1, + .flush = flush, + .sample_fmts = (const enum AVSampleFormat[]) { AV_SAMPLE_FMT_FLTP, + AV_SAMPLE_FMT_NONE }, };