You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

185 lines
5.2 KiB

/*
* Copyright (c) 2018 Paul B Mahol
*
* This file is part of FFmpeg.
*
* 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.
*
* 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 FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "avcodec.h"
#include "bsf.h"
#include "get_bits.h"
#include "mlp_parse.h"
#include "mlp.h"
typedef struct AccessUnit {
uint8_t bits[4];
uint16_t offset;
uint16_t optional;
} AccessUnit;
typedef struct TrueHDCoreContext {
const AVClass *class;
MLPHeaderInfo hdr;
} TrueHDCoreContext;
static int truehd_core_filter(AVBSFContext *ctx, AVPacket *out)
{
TrueHDCoreContext *s = ctx->priv_data;
GetBitContext gbc;
AccessUnit units[MAX_SUBSTREAMS];
AVPacket *in;
truehd_core: Miscellaneous improvements 1. The loop counter of the substream_directory loop is always less than the number of substreams, yet within the loop it is checked whether it is less than FFMIN(3, s->hdr.num_substreams), although the check for < 3 would suffice. 2. In case the packet is a major sync packet, the last two bytes of the major sync structure were initialized to 0xff and then immediately overwritten afterwards without ever making use of the values just set. 3. When updating the parity_nibble during writing the new substream_directory, the parity_nibble is updated one byte at a time with bytes that might be read from the output packet's data. But one can do both bytes at the same time without resorting to the data just written by XOR'ing with the variable that contains the value that has just been written as a big endian number. This changes the intermediate value of parity_nibble, but in the end it just amounts to a reordering of the sum modulo two that will eventually be written as parity_nibble. Due to associativity and commutativity, this value is unchanged. 4. init_get_bits8 already checks that no overflow happens during the conversion of its argument from bytes to bits. ff_mlp_read_major_sync makes sure not to overread (the maximum size of a major_sync_info is 60 bytes anyway) and last_offset is < 2^13, so that no overflow in the calculation of size can happen, i.e. the check for whether size is >= 0 is unnecessary. But then size is completely unnecessary and can be removed. 5. In case the packet is just passed through, it is unnecessary to read the packet's dts. This is therefore postponed to when we know that the packet is not passed through. 6. Given that it seems overkill to use a bitreader just for one variable, the size of the input access unit is now read directly. 7. A substream's offset (of the end of the substream) is now stored as is (i.e. in units of words). These changes amount to a slight performance improvement: It improved from 5897 decicycles of ten runs with about 262144 runs each (including an insignificant amount -- about 20-25 usually of skips) to 5747 decicycles under the same conditions. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
5 years ago
int ret, i, last_offset = 0;
int in_size, out_size;
int have_header = 0;
int substream_bits = 0;
int end;
ret = ff_bsf_get_packet(ctx, &in);
if (ret < 0)
return ret;
if (in->size < 4) {
ret = AVERROR_INVALIDDATA;
goto fail;
}
truehd_core: Miscellaneous improvements 1. The loop counter of the substream_directory loop is always less than the number of substreams, yet within the loop it is checked whether it is less than FFMIN(3, s->hdr.num_substreams), although the check for < 3 would suffice. 2. In case the packet is a major sync packet, the last two bytes of the major sync structure were initialized to 0xff and then immediately overwritten afterwards without ever making use of the values just set. 3. When updating the parity_nibble during writing the new substream_directory, the parity_nibble is updated one byte at a time with bytes that might be read from the output packet's data. But one can do both bytes at the same time without resorting to the data just written by XOR'ing with the variable that contains the value that has just been written as a big endian number. This changes the intermediate value of parity_nibble, but in the end it just amounts to a reordering of the sum modulo two that will eventually be written as parity_nibble. Due to associativity and commutativity, this value is unchanged. 4. init_get_bits8 already checks that no overflow happens during the conversion of its argument from bytes to bits. ff_mlp_read_major_sync makes sure not to overread (the maximum size of a major_sync_info is 60 bytes anyway) and last_offset is < 2^13, so that no overflow in the calculation of size can happen, i.e. the check for whether size is >= 0 is unnecessary. But then size is completely unnecessary and can be removed. 5. In case the packet is just passed through, it is unnecessary to read the packet's dts. This is therefore postponed to when we know that the packet is not passed through. 6. Given that it seems overkill to use a bitreader just for one variable, the size of the input access unit is now read directly. 7. A substream's offset (of the end of the substream) is now stored as is (i.e. in units of words). These changes amount to a slight performance improvement: It improved from 5897 decicycles of ten runs with about 262144 runs each (including an insignificant amount -- about 20-25 usually of skips) to 5747 decicycles under the same conditions. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
5 years ago
in_size = (AV_RB16(in->data) & 0xFFF) * 2;
if (in_size < 4 || in_size > in->size) {
ret = AVERROR_INVALIDDATA;
goto fail;
}
ret = init_get_bits8(&gbc, in->data + 4, in->size - 4);
if (ret < 0)
goto fail;
if (show_bits_long(&gbc, 32) == 0xf8726fba) {
if ((ret = ff_mlp_read_major_sync(ctx, &s->hdr, &gbc)) < 0)
goto fail;
have_header = 1;
}
if (s->hdr.num_substreams > MAX_SUBSTREAMS) {
ret = AVERROR_INVALIDDATA;
goto fail;
}
for (i = 0; i < s->hdr.num_substreams; i++) {
for (int j = 0; j < 4; j++)
units[i].bits[j] = get_bits1(&gbc);
truehd_core: Miscellaneous improvements 1. The loop counter of the substream_directory loop is always less than the number of substreams, yet within the loop it is checked whether it is less than FFMIN(3, s->hdr.num_substreams), although the check for < 3 would suffice. 2. In case the packet is a major sync packet, the last two bytes of the major sync structure were initialized to 0xff and then immediately overwritten afterwards without ever making use of the values just set. 3. When updating the parity_nibble during writing the new substream_directory, the parity_nibble is updated one byte at a time with bytes that might be read from the output packet's data. But one can do both bytes at the same time without resorting to the data just written by XOR'ing with the variable that contains the value that has just been written as a big endian number. This changes the intermediate value of parity_nibble, but in the end it just amounts to a reordering of the sum modulo two that will eventually be written as parity_nibble. Due to associativity and commutativity, this value is unchanged. 4. init_get_bits8 already checks that no overflow happens during the conversion of its argument from bytes to bits. ff_mlp_read_major_sync makes sure not to overread (the maximum size of a major_sync_info is 60 bytes anyway) and last_offset is < 2^13, so that no overflow in the calculation of size can happen, i.e. the check for whether size is >= 0 is unnecessary. But then size is completely unnecessary and can be removed. 5. In case the packet is just passed through, it is unnecessary to read the packet's dts. This is therefore postponed to when we know that the packet is not passed through. 6. Given that it seems overkill to use a bitreader just for one variable, the size of the input access unit is now read directly. 7. A substream's offset (of the end of the substream) is now stored as is (i.e. in units of words). These changes amount to a slight performance improvement: It improved from 5897 decicycles of ten runs with about 262144 runs each (including an insignificant amount -- about 20-25 usually of skips) to 5747 decicycles under the same conditions. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
5 years ago
units[i].offset = get_bits(&gbc, 12);
if (i < 3) {
last_offset = units[i].offset * 2;
substream_bits += 16;
}
if (units[i].bits[0]) {
units[i].optional = get_bits(&gbc, 16);
truehd_core: Miscellaneous improvements 1. The loop counter of the substream_directory loop is always less than the number of substreams, yet within the loop it is checked whether it is less than FFMIN(3, s->hdr.num_substreams), although the check for < 3 would suffice. 2. In case the packet is a major sync packet, the last two bytes of the major sync structure were initialized to 0xff and then immediately overwritten afterwards without ever making use of the values just set. 3. When updating the parity_nibble during writing the new substream_directory, the parity_nibble is updated one byte at a time with bytes that might be read from the output packet's data. But one can do both bytes at the same time without resorting to the data just written by XOR'ing with the variable that contains the value that has just been written as a big endian number. This changes the intermediate value of parity_nibble, but in the end it just amounts to a reordering of the sum modulo two that will eventually be written as parity_nibble. Due to associativity and commutativity, this value is unchanged. 4. init_get_bits8 already checks that no overflow happens during the conversion of its argument from bytes to bits. ff_mlp_read_major_sync makes sure not to overread (the maximum size of a major_sync_info is 60 bytes anyway) and last_offset is < 2^13, so that no overflow in the calculation of size can happen, i.e. the check for whether size is >= 0 is unnecessary. But then size is completely unnecessary and can be removed. 5. In case the packet is just passed through, it is unnecessary to read the packet's dts. This is therefore postponed to when we know that the packet is not passed through. 6. Given that it seems overkill to use a bitreader just for one variable, the size of the input access unit is now read directly. 7. A substream's offset (of the end of the substream) is now stored as is (i.e. in units of words). These changes amount to a slight performance improvement: It improved from 5897 decicycles of ten runs with about 262144 runs each (including an insignificant amount -- about 20-25 usually of skips) to 5747 decicycles under the same conditions. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
5 years ago
if (i < 3)
substream_bits += 16;
}
}
end = get_bits_count(&gbc);
truehd_core: Miscellaneous improvements 1. The loop counter of the substream_directory loop is always less than the number of substreams, yet within the loop it is checked whether it is less than FFMIN(3, s->hdr.num_substreams), although the check for < 3 would suffice. 2. In case the packet is a major sync packet, the last two bytes of the major sync structure were initialized to 0xff and then immediately overwritten afterwards without ever making use of the values just set. 3. When updating the parity_nibble during writing the new substream_directory, the parity_nibble is updated one byte at a time with bytes that might be read from the output packet's data. But one can do both bytes at the same time without resorting to the data just written by XOR'ing with the variable that contains the value that has just been written as a big endian number. This changes the intermediate value of parity_nibble, but in the end it just amounts to a reordering of the sum modulo two that will eventually be written as parity_nibble. Due to associativity and commutativity, this value is unchanged. 4. init_get_bits8 already checks that no overflow happens during the conversion of its argument from bytes to bits. ff_mlp_read_major_sync makes sure not to overread (the maximum size of a major_sync_info is 60 bytes anyway) and last_offset is < 2^13, so that no overflow in the calculation of size can happen, i.e. the check for whether size is >= 0 is unnecessary. But then size is completely unnecessary and can be removed. 5. In case the packet is just passed through, it is unnecessary to read the packet's dts. This is therefore postponed to when we know that the packet is not passed through. 6. Given that it seems overkill to use a bitreader just for one variable, the size of the input access unit is now read directly. 7. A substream's offset (of the end of the substream) is now stored as is (i.e. in units of words). These changes amount to a slight performance improvement: It improved from 5897 decicycles of ten runs with about 262144 runs each (including an insignificant amount -- about 20-25 usually of skips) to 5747 decicycles under the same conditions. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
5 years ago
out_size = ((end + 7) >> 3) + 4 + last_offset;
if (out_size < in_size) {
int bpos = 0, reduce = (end - have_header * 28 * 8 - substream_bits) >> 4;
truehd_core: Miscellaneous improvements 1. The loop counter of the substream_directory loop is always less than the number of substreams, yet within the loop it is checked whether it is less than FFMIN(3, s->hdr.num_substreams), although the check for < 3 would suffice. 2. In case the packet is a major sync packet, the last two bytes of the major sync structure were initialized to 0xff and then immediately overwritten afterwards without ever making use of the values just set. 3. When updating the parity_nibble during writing the new substream_directory, the parity_nibble is updated one byte at a time with bytes that might be read from the output packet's data. But one can do both bytes at the same time without resorting to the data just written by XOR'ing with the variable that contains the value that has just been written as a big endian number. This changes the intermediate value of parity_nibble, but in the end it just amounts to a reordering of the sum modulo two that will eventually be written as parity_nibble. Due to associativity and commutativity, this value is unchanged. 4. init_get_bits8 already checks that no overflow happens during the conversion of its argument from bytes to bits. ff_mlp_read_major_sync makes sure not to overread (the maximum size of a major_sync_info is 60 bytes anyway) and last_offset is < 2^13, so that no overflow in the calculation of size can happen, i.e. the check for whether size is >= 0 is unnecessary. But then size is completely unnecessary and can be removed. 5. In case the packet is just passed through, it is unnecessary to read the packet's dts. This is therefore postponed to when we know that the packet is not passed through. 6. Given that it seems overkill to use a bitreader just for one variable, the size of the input access unit is now read directly. 7. A substream's offset (of the end of the substream) is now stored as is (i.e. in units of words). These changes amount to a slight performance improvement: It improved from 5897 decicycles of ten runs with about 262144 runs each (including an insignificant amount -- about 20-25 usually of skips) to 5747 decicycles under the same conditions. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
5 years ago
uint16_t parity_nibble, dts = AV_RB16(in->data + 2);
uint16_t auheader;
ret = av_new_packet(out, out_size);
if (ret < 0)
goto fail;
AV_WB16(out->data + 2, dts);
parity_nibble = dts;
out->size -= reduce * 2;
parity_nibble ^= out->size / 2;
if (have_header) {
memcpy(out->data + 4, in->data + 4, 28);
out->data[16 + 4] = (out->data[16 + 4] & 0x0c) | (FFMIN(s->hdr.num_substreams, 3) << 4);
out->data[17 + 4]&= 0x7f;
out->data[25 + 4] = out->data[25 + 4] & 0xfe;
AV_WL16(out->data + 4 + 26, ff_mlp_checksum16(out->data + 4, 26));
}
for (i = 0; i < FFMIN(s->hdr.num_substreams, 3); i++) {
uint16_t substr_hdr = 0;
substr_hdr |= (units[i].bits[0] << 15);
substr_hdr |= (units[i].bits[1] << 14);
substr_hdr |= (units[i].bits[2] << 13);
substr_hdr |= (units[i].bits[3] << 12);
truehd_core: Miscellaneous improvements 1. The loop counter of the substream_directory loop is always less than the number of substreams, yet within the loop it is checked whether it is less than FFMIN(3, s->hdr.num_substreams), although the check for < 3 would suffice. 2. In case the packet is a major sync packet, the last two bytes of the major sync structure were initialized to 0xff and then immediately overwritten afterwards without ever making use of the values just set. 3. When updating the parity_nibble during writing the new substream_directory, the parity_nibble is updated one byte at a time with bytes that might be read from the output packet's data. But one can do both bytes at the same time without resorting to the data just written by XOR'ing with the variable that contains the value that has just been written as a big endian number. This changes the intermediate value of parity_nibble, but in the end it just amounts to a reordering of the sum modulo two that will eventually be written as parity_nibble. Due to associativity and commutativity, this value is unchanged. 4. init_get_bits8 already checks that no overflow happens during the conversion of its argument from bytes to bits. ff_mlp_read_major_sync makes sure not to overread (the maximum size of a major_sync_info is 60 bytes anyway) and last_offset is < 2^13, so that no overflow in the calculation of size can happen, i.e. the check for whether size is >= 0 is unnecessary. But then size is completely unnecessary and can be removed. 5. In case the packet is just passed through, it is unnecessary to read the packet's dts. This is therefore postponed to when we know that the packet is not passed through. 6. Given that it seems overkill to use a bitreader just for one variable, the size of the input access unit is now read directly. 7. A substream's offset (of the end of the substream) is now stored as is (i.e. in units of words). These changes amount to a slight performance improvement: It improved from 5897 decicycles of ten runs with about 262144 runs each (including an insignificant amount -- about 20-25 usually of skips) to 5747 decicycles under the same conditions. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
5 years ago
substr_hdr |= units[i].offset;
AV_WB16(out->data + have_header * 28 + 4 + bpos, substr_hdr);
truehd_core: Miscellaneous improvements 1. The loop counter of the substream_directory loop is always less than the number of substreams, yet within the loop it is checked whether it is less than FFMIN(3, s->hdr.num_substreams), although the check for < 3 would suffice. 2. In case the packet is a major sync packet, the last two bytes of the major sync structure were initialized to 0xff and then immediately overwritten afterwards without ever making use of the values just set. 3. When updating the parity_nibble during writing the new substream_directory, the parity_nibble is updated one byte at a time with bytes that might be read from the output packet's data. But one can do both bytes at the same time without resorting to the data just written by XOR'ing with the variable that contains the value that has just been written as a big endian number. This changes the intermediate value of parity_nibble, but in the end it just amounts to a reordering of the sum modulo two that will eventually be written as parity_nibble. Due to associativity and commutativity, this value is unchanged. 4. init_get_bits8 already checks that no overflow happens during the conversion of its argument from bytes to bits. ff_mlp_read_major_sync makes sure not to overread (the maximum size of a major_sync_info is 60 bytes anyway) and last_offset is < 2^13, so that no overflow in the calculation of size can happen, i.e. the check for whether size is >= 0 is unnecessary. But then size is completely unnecessary and can be removed. 5. In case the packet is just passed through, it is unnecessary to read the packet's dts. This is therefore postponed to when we know that the packet is not passed through. 6. Given that it seems overkill to use a bitreader just for one variable, the size of the input access unit is now read directly. 7. A substream's offset (of the end of the substream) is now stored as is (i.e. in units of words). These changes amount to a slight performance improvement: It improved from 5897 decicycles of ten runs with about 262144 runs each (including an insignificant amount -- about 20-25 usually of skips) to 5747 decicycles under the same conditions. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
5 years ago
parity_nibble ^= substr_hdr;
bpos += 2;
if (units[i].bits[0]) {
AV_WB16(out->data + have_header * 28 + 4 + bpos, units[i].optional);
truehd_core: Miscellaneous improvements 1. The loop counter of the substream_directory loop is always less than the number of substreams, yet within the loop it is checked whether it is less than FFMIN(3, s->hdr.num_substreams), although the check for < 3 would suffice. 2. In case the packet is a major sync packet, the last two bytes of the major sync structure were initialized to 0xff and then immediately overwritten afterwards without ever making use of the values just set. 3. When updating the parity_nibble during writing the new substream_directory, the parity_nibble is updated one byte at a time with bytes that might be read from the output packet's data. But one can do both bytes at the same time without resorting to the data just written by XOR'ing with the variable that contains the value that has just been written as a big endian number. This changes the intermediate value of parity_nibble, but in the end it just amounts to a reordering of the sum modulo two that will eventually be written as parity_nibble. Due to associativity and commutativity, this value is unchanged. 4. init_get_bits8 already checks that no overflow happens during the conversion of its argument from bytes to bits. ff_mlp_read_major_sync makes sure not to overread (the maximum size of a major_sync_info is 60 bytes anyway) and last_offset is < 2^13, so that no overflow in the calculation of size can happen, i.e. the check for whether size is >= 0 is unnecessary. But then size is completely unnecessary and can be removed. 5. In case the packet is just passed through, it is unnecessary to read the packet's dts. This is therefore postponed to when we know that the packet is not passed through. 6. Given that it seems overkill to use a bitreader just for one variable, the size of the input access unit is now read directly. 7. A substream's offset (of the end of the substream) is now stored as is (i.e. in units of words). These changes amount to a slight performance improvement: It improved from 5897 decicycles of ten runs with about 262144 runs each (including an insignificant amount -- about 20-25 usually of skips) to 5747 decicycles under the same conditions. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
5 years ago
parity_nibble ^= units[i].optional;
bpos += 2;
}
}
parity_nibble ^= parity_nibble >> 8;
parity_nibble ^= parity_nibble >> 4;
parity_nibble &= 0xF;
memcpy(out->data + have_header * 28 + 4 + bpos,
in->data + 4 + (end >> 3),
out_size - (4 + (end >> 3)));
auheader = (parity_nibble ^ 0xF) << 12;
auheader |= (out->size / 2) & 0x0fff;
AV_WB16(out->data, auheader);
ret = av_packet_copy_props(out, in);
} else {
av_packet_move_ref(out, in);
}
fail:
if (ret < 0)
av_packet_unref(out);
av_packet_free(&in);
return ret;
}
static void truehd_core_flush(AVBSFContext *ctx)
{
TrueHDCoreContext *s = ctx->priv_data;
memset(&s->hdr, 0, sizeof(s->hdr));
}
static const enum AVCodecID codec_ids[] = {
AV_CODEC_ID_TRUEHD, AV_CODEC_ID_NONE,
};
const AVBitStreamFilter ff_truehd_core_bsf = {
.name = "truehd_core",
.priv_data_size = sizeof(TrueHDCoreContext),
.filter = truehd_core_filter,
.flush = truehd_core_flush,
.codec_ids = codec_ids,
};