summaryrefslogtreecommitdiffstats
path: root/libavcodec/wmalosslessdec.c
diff options
context:
space:
mode:
Diffstat (limited to 'libavcodec/wmalosslessdec.c')
-rw-r--r--libavcodec/wmalosslessdec.c136
1 files changed, 74 insertions, 62 deletions
diff --git a/libavcodec/wmalosslessdec.c b/libavcodec/wmalosslessdec.c
index 6ee27d3..8e2ac5f 100644
--- a/libavcodec/wmalosslessdec.c
+++ b/libavcodec/wmalosslessdec.c
@@ -5,20 +5,20 @@
* Copyright (c) 2011 Andreas Öman
* Copyright (c) 2011 - 2012 Mashiat Sarker Shakkhar
*
- * This file is part of Libav.
+ * This file is part of FFmpeg.
*
- * Libav is free software; you can redistribute it and/or
+ * FFmpeg 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.
*
- * Libav is distributed in the hope that it will be useful,
+ * FFmpeg 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 Libav; if not, write to the Free Software
+ * License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
@@ -31,6 +31,7 @@
#include "internal.h"
#include "get_bits.h"
#include "put_bits.h"
+#include "lossless_audiodsp.h"
#include "wma.h"
#include "wma_common.h"
@@ -46,6 +47,7 @@
#define WMALL_BLOCK_MAX_SIZE (1 << WMALL_BLOCK_MAX_BITS) ///< maximum block size
#define WMALL_BLOCK_SIZES (WMALL_BLOCK_MAX_BITS - WMALL_BLOCK_MIN_BITS + 1) ///< possible block sizes
+#define WMALL_COEFF_PAD_SIZE 16 ///< pad coef buffers with 0 for use with SIMD
/**
* @brief frame-specific decoder context for a single channel
@@ -69,6 +71,7 @@ typedef struct WmallDecodeCtx {
/* generic decoder variables */
AVCodecContext *avctx;
AVFrame *frame;
+ LLAudDSPContext dsp; ///< accelerated DSP functions
uint8_t frame_data[MAX_FRAMESIZE + FF_INPUT_BUFFER_PADDING_SIZE]; ///< compressed frame data
PutBitContext pb; ///< context for filling the frame_data buffer
@@ -125,12 +128,12 @@ typedef struct WmallDecodeCtx {
int8_t acfilter_order;
int8_t acfilter_scaling;
int64_t acfilter_coeffs[16];
- int acfilter_prevvalues[2][16];
+ int acfilter_prevvalues[WMALL_MAX_CHANNELS][16];
int8_t mclms_order;
int8_t mclms_scaling;
- int16_t mclms_coeffs[128];
- int16_t mclms_coeffs_cur[4];
+ int16_t mclms_coeffs[WMALL_MAX_CHANNELS * WMALL_MAX_CHANNELS * 32];
+ int16_t mclms_coeffs_cur[WMALL_MAX_CHANNELS * WMALL_MAX_CHANNELS];
int16_t mclms_prevvalues[WMALL_MAX_CHANNELS * 2 * 32];
int16_t mclms_updates[WMALL_MAX_CHANNELS * 2 * 32];
int mclms_recent;
@@ -143,33 +146,33 @@ typedef struct WmallDecodeCtx {
int scaling;
int coefsend;
int bitsend;
- int16_t coefs[MAX_ORDER];
- int16_t lms_prevvalues[MAX_ORDER * 2];
- int16_t lms_updates[MAX_ORDER * 2];
+ DECLARE_ALIGNED(16, int16_t, coefs)[MAX_ORDER + WMALL_COEFF_PAD_SIZE/sizeof(int16_t)];
+ DECLARE_ALIGNED(16, int16_t, lms_prevvalues)[MAX_ORDER * 2];
+ DECLARE_ALIGNED(16, int16_t, lms_updates)[MAX_ORDER * 2];
int recent;
- } cdlms[2][9];
+ } cdlms[WMALL_MAX_CHANNELS][9];
- int cdlms_ttl[2];
+ int cdlms_ttl[WMALL_MAX_CHANNELS];
int bV3RTM;
- int is_channel_coded[2];
- int update_speed[2];
+ int is_channel_coded[WMALL_MAX_CHANNELS];
+ int update_speed[WMALL_MAX_CHANNELS];
- int transient[2];
- int transient_pos[2];
+ int transient[WMALL_MAX_CHANNELS];
+ int transient_pos[WMALL_MAX_CHANNELS];
int seekable_tile;
- int ave_sum[2];
+ int ave_sum[WMALL_MAX_CHANNELS];
- int channel_residues[2][WMALL_BLOCK_MAX_SIZE];
+ int channel_residues[WMALL_MAX_CHANNELS][WMALL_BLOCK_MAX_SIZE];
- int lpc_coefs[2][40];
+ int lpc_coefs[WMALL_MAX_CHANNELS][40];
int lpc_order;
int lpc_scaling;
int lpc_intbits;
- int channel_coeffs[2][WMALL_BLOCK_MAX_SIZE];
+ int channel_coeffs[WMALL_MAX_CHANNELS][WMALL_BLOCK_MAX_SIZE];
} WmallDecodeCtx;
@@ -180,7 +183,13 @@ static av_cold int decode_init(AVCodecContext *avctx)
unsigned int channel_mask;
int i, log2_max_num_subframes;
+ if (!avctx->block_align) {
+ av_log(avctx, AV_LOG_ERROR, "block_align is not set\n");
+ return AVERROR(EINVAL);
+ }
+
s->avctx = avctx;
+ ff_llauddsp_init(&s->dsp);
init_put_bits(&s->pb, s->frame_data, MAX_FRAMESIZE);
if (avctx->extradata_size >= 18) {
@@ -345,11 +354,11 @@ static int decode_tilehdr(WmallDecodeCtx *s)
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] = in_use = 1;
+ contains_subframe[c] = 1;
} else {
- if (get_bits1(&s->gb))
- contains_subframe[c] = in_use = 1;
+ contains_subframe[c] = get_bits1(&s->gb);
}
+ in_use |= contains_subframe[c];
} else
contains_subframe[c] = 0;
}
@@ -454,6 +463,13 @@ static int decode_cdlms(WmallDecodeCtx *s)
s->cdlms[0][0].order = 0;
return AVERROR_INVALIDDATA;
}
+ if(s->cdlms[c][i].order & 8) {
+ static int warned;
+ if(!warned)
+ avpriv_request_sample(s->avctx, "CDLMS of order %d",
+ s->cdlms[c][i].order);
+ warned = 1;
+ }
}
for (i = 0; i < s->cdlms_ttl[c]; i++)
@@ -479,6 +495,10 @@ static int decode_cdlms(WmallDecodeCtx *s)
(get_bits(&s->gb, s->cdlms[c][i].bitsend) << shift_l) >> shift_r;
}
}
+
+ for (i = 0; i < s->cdlms_ttl[c]; i++)
+ memset(s->cdlms[c][i].coefs + s->cdlms[c][i].order,
+ 0, WMALL_COEFF_PAD_SIZE);
}
return 0;
@@ -505,9 +525,9 @@ static int decode_channel_residues(WmallDecodeCtx *s, int ch, int tile_size)
if (s->seekable_tile) {
if (s->do_inter_ch_decorr)
- s->channel_residues[ch][0] = get_sbits(&s->gb, s->bits_per_sample + 1);
+ s->channel_residues[ch][0] = get_sbits_long(&s->gb, s->bits_per_sample + 1);
else
- s->channel_residues[ch][0] = get_sbits(&s->gb, s->bits_per_sample);
+ s->channel_residues[ch][0] = get_sbits_long(&s->gb, s->bits_per_sample);
i++;
}
for (; i < tile_size; i++) {
@@ -525,7 +545,7 @@ static int decode_channel_residues(WmallDecodeCtx *s, int ch, int tile_size)
residue = quo;
else {
rem_bits = av_ceil_log2(ave_mean);
- rem = rem_bits ? get_bits_long(&s->gb, rem_bits) : 0;
+ rem = get_bits_long(&s->gb, rem_bits);
residue = (quo << rem_bits) + rem;
}
@@ -649,10 +669,10 @@ static void mclms_update(WmallDecodeCtx *s, int icoef, int *pred)
if (s->mclms_recent == 0) {
memcpy(&s->mclms_prevvalues[order * num_channels],
s->mclms_prevvalues,
- 2 * order * num_channels);
+ sizeof(int16_t) * order * num_channels);
memcpy(&s->mclms_updates[order * num_channels],
s->mclms_updates,
- 2 * order * num_channels);
+ sizeof(int16_t) * order * num_channels);
s->mclms_recent = num_channels * order;
}
}
@@ -688,35 +708,11 @@ static void revert_mclms(WmallDecodeCtx *s, int tile_size)
}
}
-static int lms_predict(WmallDecodeCtx *s, int ich, int ilms)
-{
- int pred = 0, icoef;
- int recent = s->cdlms[ich][ilms].recent;
-
- for (icoef = 0; icoef < s->cdlms[ich][ilms].order; icoef++)
- pred += s->cdlms[ich][ilms].coefs[icoef] *
- s->cdlms[ich][ilms].lms_prevvalues[icoef + recent];
-
- return pred;
-}
-
-static void lms_update(WmallDecodeCtx *s, int ich, int ilms,
- int input, int residue)
+static void lms_update(WmallDecodeCtx *s, int ich, int ilms, int input)
{
- int icoef;
int recent = s->cdlms[ich][ilms].recent;
int range = 1 << s->bits_per_sample - 1;
- if (residue < 0) {
- for (icoef = 0; icoef < s->cdlms[ich][ilms].order; icoef++)
- s->cdlms[ich][ilms].coefs[icoef] -=
- s->cdlms[ich][ilms].lms_updates[icoef + recent];
- } else if (residue > 0) {
- for (icoef = 0; icoef < s->cdlms[ich][ilms].order; icoef++)
- s->cdlms[ich][ilms].coefs[icoef] +=
- s->cdlms[ich][ilms].lms_updates[icoef + recent];
- }
-
if (recent)
recent--;
else {
@@ -777,6 +773,9 @@ static void use_normal_update_speed(WmallDecodeCtx *s, int ich)
s->update_speed[ich] = 8;
}
+/** Get sign of integer (1 for positive, -1 for negative and 0 for zero) */
+#define WMASIGN(x) ((x > 0) - (x < 0))
+
static void revert_cdlms(WmallDecodeCtx *s, int ch,
int coef_begin, int coef_end)
{
@@ -787,12 +786,19 @@ static void revert_cdlms(WmallDecodeCtx *s, int ch,
for (icoef = coef_begin; icoef < coef_end; icoef++) {
pred = 1 << (s->cdlms[ch][ilms].scaling - 1);
residue = s->channel_residues[ch][icoef];
- pred += lms_predict(s, ch, ilms);
+ pred += s->dsp.scalarproduct_and_madd_int16(s->cdlms[ch][ilms].coefs,
+ s->cdlms[ch][ilms].lms_prevvalues
+ + s->cdlms[ch][ilms].recent,
+ s->cdlms[ch][ilms].lms_updates
+ + s->cdlms[ch][ilms].recent,
+ s->cdlms[ch][ilms].order,
+ WMASIGN(residue));
input = residue + (pred >> s->cdlms[ch][ilms].scaling);
- lms_update(s, ch, ilms, input, residue);
+ lms_update(s, ch, ilms, input);
s->channel_residues[ch][icoef] = input;
}
}
+ emms_c();
}
static void revert_inter_ch_decorr(WmallDecodeCtx *s, int tile_size)
@@ -955,7 +961,7 @@ static int decode_subframe(WmallDecodeCtx *s)
bits * s->num_channels * subframe_len, get_bits_count(&s->gb));
for (i = 0; i < s->num_channels; i++)
for (j = 0; j < subframe_len; j++)
- s->channel_coeffs[i][j] = get_sbits(&s->gb, bits);
+ s->channel_coeffs[i][j] = get_sbits_long(&s->gb, bits);
} else {
for (i = 0; i < s->num_channels; i++)
if (s->is_channel_coded[i]) {
@@ -1022,8 +1028,6 @@ static int decode_frame(WmallDecodeCtx *s)
s->frame->nb_samples = s->samples_per_frame;
if ((ret = ff_get_buffer(s->avctx, s->frame, 0)) < 0) {
/* 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 ret;
}
@@ -1037,9 +1041,10 @@ static int decode_frame(WmallDecodeCtx *s)
len = get_bits(gb, s->log2_frame_size);
/* decode tile information */
- if (decode_tilehdr(s)) {
+ if ((ret = decode_tilehdr(s))) {
s->packet_loss = 1;
- return 0;
+ av_frame_unref(s->frame);
+ return ret;
}
/* read drc info */
@@ -1074,8 +1079,11 @@ static int decode_frame(WmallDecodeCtx *s)
/* decode all subframes */
while (!s->parsed_all_subframes) {
+ int decoded_samples = s->channel[0].decoded_samples;
if (decode_subframe(s) < 0) {
s->packet_loss = 1;
+ if (s->frame->nb_samples)
+ s->frame->nb_samples = decoded_samples;
return 0;
}
}
@@ -1182,9 +1190,13 @@ static int decode_packet(AVCodecContext *avctx, void *data, int *got_frame_ptr,
if (s->packet_done || s->packet_loss) {
s->packet_done = 0;
- /* sanity check for the buffer length */
- if (buf_size < avctx->block_align)
+ if (!buf_size)
return 0;
+ /* sanity check for the buffer length */
+ if (buf_size < avctx->block_align) {
+ av_log(avctx, AV_LOG_ERROR, "buf size %d invalid\n", buf_size);
+ return AVERROR_INVALIDDATA;
+ }
s->next_packet_start = buf_size - avctx->block_align;
buf_size = avctx->block_align;
OpenPOWER on IntegriCloud