ffmpeg/libavcodec/mv30.c

721 lines
22 KiB
C

/*
* MidiVid MV30 decoder
*
* Copyright (c) 2020 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 <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "libavutil/thread.h"
#include "avcodec.h"
#include "bytestream.h"
#include "copy_block.h"
#include "mathops.h"
#include "blockdsp.h"
#include "get_bits.h"
#include "internal.h"
#include "aandcttab.h"
typedef struct MV30Context {
GetBitContext gb;
int intra_quant;
int inter_quant;
int is_inter;
int mode_size;
int nb_mvectors;
int block[6][64];
int16_t *mvectors;
unsigned int mvectors_size;
int16_t *coeffs;
unsigned int coeffs_size;
int16_t intraq_tab[2][64];
int16_t interq_tab[2][64];
BlockDSPContext bdsp;
AVFrame *prev_frame;
} MV30Context;
static VLC cbp_tab;
static const uint8_t luma_tab[] = {
12, 12, 15, 19, 25, 34, 40, 48,
12, 12, 18, 22, 27, 44, 47, 46,
17, 18, 21, 26, 35, 46, 52, 47,
18, 20, 24, 28, 40, 61, 59, 51,
20, 24, 32, 43, 50, 72, 72, 63,
25, 31, 42, 48, 58, 72, 81, 75,
38, 46, 54, 61, 71, 84, 88, 85,
50, 61, 65, 68, 79, 78, 86, 91,
};
static const uint8_t chroma_tab[] = {
12, 16, 24, 47, 99, 99, 99, 99,
16, 21, 26, 66, 99, 99, 99, 99,
24, 26, 56, 99, 99, 99, 99, 99,
47, 66, 99, 99, 99, 99, 99, 99,
99, 99, 99, 99, 99, 99, 99, 99,
99, 99, 99, 99, 99, 99, 99, 99,
99, 99, 99, 99, 99, 99, 99, 99,
99, 99, 99, 99, 99, 99, 99, 99,
};
static const uint8_t zigzag[] = {
0, 1, 8, 9, 16, 2, 3, 10,
17, 24, 32, 25, 18, 11, 4, 5,
12, 19, 26, 33, 40, 48, 41, 34,
27, 20, 13, 6, 7, 14, 21, 28,
35, 42, 49, 56, 57, 50, 43, 36,
29, 22, 15, 23, 30, 37, 44, 51,
58, 59, 52, 45, 38, 31, 39, 46,
53, 60, 61, 54, 47, 55, 62, 63,
};
static void get_qtable(int16_t *table, int quant, const uint8_t *quant_tab)
{
int factor = quant < 50 ? 5000 / FFMAX(quant, 1) : 200 - FFMIN(quant, 100) * 2;
for (int i = 0; i < 64; i++) {
table[i] = av_clip((quant_tab[i] * factor + 0x32) / 100, 1, 0x7fff);
table[i] = ((int)ff_aanscales[i] * (int)table[i] + 0x800) >> 12;
}
}
static inline void idct_1d(unsigned *blk, int step)
{
const unsigned t0 = blk[0 * step] + blk[4 * step];
const unsigned t1 = blk[0 * step] - blk[4 * step];
const unsigned t2 = blk[2 * step] + blk[6 * step];
const unsigned t3 = ((int)((blk[2 * step] - blk[6 * step]) * 362U) >> 8) - t2;
const unsigned t4 = t0 + t2;
const unsigned t5 = t0 - t2;
const unsigned t6 = t1 + t3;
const unsigned t7 = t1 - t3;
const unsigned t8 = blk[5 * step] + blk[3 * step];
const unsigned t9 = blk[5 * step] - blk[3 * step];
const unsigned tA = blk[1 * step] + blk[7 * step];
const unsigned tB = blk[1 * step] - blk[7 * step];
const unsigned tC = t8 + tA;
const unsigned tD = (int)((tB + t9) * 473U) >> 8;
const unsigned tE = (((int)(t9 * -669U) >> 8) - tC) + tD;
const unsigned tF = ((int)((tA - t8) * 362U) >> 8) - tE;
const unsigned t10 = (((int)(tB * 277U) >> 8) - tD) + tF;
blk[0 * step] = t4 + tC;
blk[1 * step] = t6 + tE;
blk[2 * step] = t7 + tF;
blk[3 * step] = t5 - t10;
blk[4 * step] = t5 + t10;
blk[5 * step] = t7 - tF;
blk[6 * step] = t6 - tE;
blk[7 * step] = t4 - tC;
}
static void idct_put(uint8_t *dst, int stride, int *block)
{
for (int i = 0; i < 8; i++) {
if ((block[0x08 + i] |
block[0x10 + i] |
block[0x18 + i] |
block[0x20 + i] |
block[0x28 + i] |
block[0x30 + i] |
block[0x38 + i]) == 0) {
block[0x08 + i] = block[i];
block[0x10 + i] = block[i];
block[0x18 + i] = block[i];
block[0x20 + i] = block[i];
block[0x28 + i] = block[i];
block[0x30 + i] = block[i];
block[0x38 + i] = block[i];
} else {
idct_1d(block + i, 8);
}
}
for (int i = 0; i < 8; i++) {
idct_1d(block, 1);
for (int j = 0; j < 8; j++)
dst[j] = av_clip_uint8((block[j] >> 5) + 128);
block += 8;
dst += stride;
}
}
static void idct_add(uint8_t *dst, int stride,
const uint8_t *src, int in_linesize, int *block)
{
for (int i = 0; i < 8; i++) {
if ((block[0x08 + i] |
block[0x10 + i] |
block[0x18 + i] |
block[0x20 + i] |
block[0x28 + i] |
block[0x30 + i] |
block[0x38 + i]) == 0) {
block[0x08 + i] = block[i];
block[0x10 + i] = block[i];
block[0x18 + i] = block[i];
block[0x20 + i] = block[i];
block[0x28 + i] = block[i];
block[0x30 + i] = block[i];
block[0x38 + i] = block[i];
} else {
idct_1d(block + i, 8);
}
}
for (int i = 0; i < 8; i++) {
idct_1d(block, 1);
for (int j = 0; j < 8; j++)
dst[j] = av_clip_uint8((block[j] >> 5) + src[j]);
block += 8;
dst += stride;
src += in_linesize;
}
}
static inline void idct2_1d(int *blk, int step)
{
const unsigned int t0 = blk[0 * step];
const unsigned int t1 = blk[1 * step];
const unsigned int t2 = (int)(t1 * 473U) >> 8;
const unsigned int t3 = t2 - t1;
const unsigned int t4 = ((int)(t1 * 362U) >> 8) - t3;
const unsigned int t5 = (((int)(t1 * 277U) >> 8) - t2) + t4;
blk[0 * step] = t1 + t0;
blk[1 * step] = t0 + t3;
blk[2 * step] = t4 + t0;
blk[3 * step] = t0 - t5;
blk[4 * step] = t5 + t0;
blk[5 * step] = t0 - t4;
blk[6 * step] = t0 - t3;
blk[7 * step] = t0 - t1;
}
static void idct2_put(uint8_t *dst, int stride, int *block)
{
for (int i = 0; i < 2; i++) {
if ((block[0x08 + i]) == 0) {
block[0x08 + i] = block[i];
block[0x10 + i] = block[i];
block[0x18 + i] = block[i];
block[0x20 + i] = block[i];
block[0x28 + i] = block[i];
block[0x30 + i] = block[i];
block[0x38 + i] = block[i];
} else {
idct2_1d(block + i, 8);
}
}
for (int i = 0; i < 8; i++) {
if (block[1] == 0) {
for (int j = 0; j < 8; j++)
dst[j] = av_clip_uint8((block[0] >> 5) + 128);
} else {
idct2_1d(block, 1);
for (int j = 0; j < 8; j++)
dst[j] = av_clip_uint8((block[j] >> 5) + 128);
}
block += 8;
dst += stride;
}
}
static void idct2_add(uint8_t *dst, int stride,
const uint8_t *src, int in_linesize,
int *block)
{
for (int i = 0; i < 2; i++) {
if ((block[0x08 + i]) == 0) {
block[0x08 + i] = block[i];
block[0x10 + i] = block[i];
block[0x18 + i] = block[i];
block[0x20 + i] = block[i];
block[0x28 + i] = block[i];
block[0x30 + i] = block[i];
block[0x38 + i] = block[i];
} else {
idct2_1d(block + i, 8);
}
}
for (int i = 0; i < 8; i++) {
if (block[1] == 0) {
for (int j = 0; j < 8; j++)
dst[j] = av_clip_uint8((block[0] >> 5) + src[j]);
} else {
idct2_1d(block, 1);
for (int j = 0; j < 8; j++)
dst[j] = av_clip_uint8((block[j] >> 5) + src[j]);
}
block += 8;
dst += stride;
src += in_linesize;
}
}
static void update_inter_block(uint8_t *dst, int stride,
const uint8_t *src, int in_linesize,
int block)
{
for (int i = 0; i < 8; i++) {
for (int j = 0; j < 8; j++)
dst[j] = av_clip_uint8(block + src[j]);
dst += stride;
src += in_linesize;
}
}
static int decode_intra_block(AVCodecContext *avctx, int mode,
GetByteContext *gbyte, int16_t *qtab,
int *block, int *pfill,
uint8_t *dst, int linesize)
{
MV30Context *s = avctx->priv_data;
int fill;
switch (mode) {
case 0:
s->bdsp.fill_block_tab[1](dst, 128, linesize, 8);
break;
case 1:
fill = sign_extend(bytestream2_get_ne16(gbyte), 16);
pfill[0] += fill;
block[0] = ((int)((unsigned)pfill[0] * qtab[0]) >> 5) + 128;
s->bdsp.fill_block_tab[1](dst, block[0], linesize, 8);
break;
case 2:
memset(block, 0, sizeof(*block) * 64);
fill = sign_extend(bytestream2_get_ne16(gbyte), 16);
pfill[0] += fill;
block[0] = (unsigned)pfill[0] * qtab[0];
block[1] = sign_extend(bytestream2_get_ne16(gbyte), 16) * qtab[1];
block[8] = sign_extend(bytestream2_get_ne16(gbyte), 16) * qtab[8];
block[9] = sign_extend(bytestream2_get_ne16(gbyte), 16) * qtab[9];
idct2_put(dst, linesize, block);
break;
case 3:
fill = sign_extend(bytestream2_get_ne16(gbyte), 16);
pfill[0] += fill;
block[0] = (unsigned)pfill[0] * qtab[0];
for (int i = 1; i < 64; i++)
block[zigzag[i]] = sign_extend(bytestream2_get_ne16(gbyte), 16) * qtab[zigzag[i]];
idct_put(dst, linesize, block);
break;
}
return 0;
}
static int decode_inter_block(AVCodecContext *avctx, int mode,
GetByteContext *gbyte, int16_t *qtab,
int *block, int *pfill,
uint8_t *dst, int linesize,
const uint8_t *src, int in_linesize)
{
int fill;
switch (mode) {
case 0:
copy_block8(dst, src, linesize, in_linesize, 8);
break;
case 1:
fill = sign_extend(bytestream2_get_ne16(gbyte), 16);
pfill[0] += fill;
block[0] = (int)((unsigned)pfill[0] * qtab[0]) >> 5;
update_inter_block(dst, linesize, src, in_linesize, block[0]);
break;
case 2:
memset(block, 0, sizeof(*block) * 64);
fill = sign_extend(bytestream2_get_ne16(gbyte), 16);
pfill[0] += fill;
block[0] = (unsigned)pfill[0] * qtab[0];
block[1] = sign_extend(bytestream2_get_ne16(gbyte), 16) * qtab[1];
block[8] = sign_extend(bytestream2_get_ne16(gbyte), 16) * qtab[8];
block[9] = sign_extend(bytestream2_get_ne16(gbyte), 16) * qtab[9];
idct2_add(dst, linesize, src, in_linesize, block);
break;
case 3:
fill = sign_extend(bytestream2_get_ne16(gbyte), 16);
pfill[0] += fill;
block[0] = (unsigned)pfill[0] * qtab[0];
for (int i = 1; i < 64; i++)
block[zigzag[i]] = sign_extend(bytestream2_get_ne16(gbyte), 16) * qtab[zigzag[i]];
idct_add(dst, linesize, src, in_linesize, block);
break;
}
return 0;
}
static int decode_coeffs(GetBitContext *gb, int16_t *coeffs, int nb_codes)
{
memset(coeffs, 0, nb_codes * sizeof(*coeffs));
for (int i = 0; i < nb_codes;) {
int value = get_vlc2(gb, cbp_tab.table, cbp_tab.bits, 1);
if (value > 0) {
int x = get_bits(gb, value);
if (x < (1 << value) / 2) {
x = (1 << (value - 1)) + (x & ((1 << value) - 1 >> 1));
} else {
x = -(1 << (value - 1)) - (x & ((1 << value) - 1 >> 1));
}
coeffs[i++] = x;
} else {
int flag = get_bits1(gb);
i += get_bits(gb, 3 + flag * 3) + 1 + flag * 8;
}
}
return 0;
}
static int decode_intra(AVCodecContext *avctx, GetBitContext *gb, AVFrame *frame)
{
MV30Context *s = avctx->priv_data;
GetBitContext mgb;
uint8_t *dst[6];
int linesize[6];
int ret;
mgb = *gb;
if (get_bits_left(gb) < s->mode_size * 8)
return AVERROR_INVALIDDATA;
skip_bits_long(gb, s->mode_size * 8);
linesize[0] = frame->linesize[0];
linesize[1] = frame->linesize[0];
linesize[2] = frame->linesize[0];
linesize[3] = frame->linesize[0];
linesize[4] = frame->linesize[1];
linesize[5] = frame->linesize[2];
for (int y = 0; y < avctx->height; y += 16) {
GetByteContext gbyte;
int pfill[3][1] = { {0} };
int nb_codes = get_bits(gb, 16);
av_fast_padded_malloc(&s->coeffs, &s->coeffs_size, nb_codes * sizeof(*s->coeffs));
if (!s->coeffs)
return AVERROR(ENOMEM);
ret = decode_coeffs(gb, s->coeffs, nb_codes);
if (ret < 0)
return ret;
bytestream2_init(&gbyte, (uint8_t *)s->coeffs, nb_codes * sizeof(*s->coeffs));
for (int x = 0; x < avctx->width; x += 16) {
dst[0] = frame->data[0] + linesize[0] * y + x;
dst[1] = frame->data[0] + linesize[0] * y + x + 8;
dst[2] = frame->data[0] + linesize[0] * (y + 8) + x;
dst[3] = frame->data[0] + linesize[0] * (y + 8) + x + 8;
dst[4] = frame->data[1] + linesize[4] * (y >> 1) + (x >> 1);
dst[5] = frame->data[2] + linesize[5] * (y >> 1) + (x >> 1);
for (int b = 0; b < 6; b++) {
int mode = get_bits_le(&mgb, 2);
ret = decode_intra_block(avctx, mode, &gbyte, s->intraq_tab[b >= 4],
s->block[b],
pfill[(b >= 4) + (b >= 5)],
dst[b], linesize[b]);
if (ret < 0)
return ret;
}
}
}
return 0;
}
static int decode_inter(AVCodecContext *avctx, GetBitContext *gb,
AVFrame *frame, AVFrame *prev)
{
MV30Context *s = avctx->priv_data;
GetBitContext mask;
GetBitContext mgb;
GetByteContext mv;
const int mask_size = ((avctx->height >> 4) * (avctx->width >> 4) * 2 + 7) / 8;
uint8_t *dst[6], *src[6];
int in_linesize[6];
int linesize[6];
int ret, cnt = 0;
int flags = 0;
in_linesize[0] = prev->linesize[0];
in_linesize[1] = prev->linesize[0];
in_linesize[2] = prev->linesize[0];
in_linesize[3] = prev->linesize[0];
in_linesize[4] = prev->linesize[1];
in_linesize[5] = prev->linesize[2];
linesize[0] = frame->linesize[0];
linesize[1] = frame->linesize[0];
linesize[2] = frame->linesize[0];
linesize[3] = frame->linesize[0];
linesize[4] = frame->linesize[1];
linesize[5] = frame->linesize[2];
av_fast_padded_malloc(&s->mvectors, &s->mvectors_size, 2 * s->nb_mvectors * sizeof(*s->mvectors));
if (!s->mvectors) {
ret = AVERROR(ENOMEM);
goto fail;
}
mask = *gb;
skip_bits_long(gb, mask_size * 8);
mgb = *gb;
skip_bits_long(gb, s->mode_size * 8);
ret = decode_coeffs(gb, s->mvectors, 2 * s->nb_mvectors);
if (ret < 0)
goto fail;
bytestream2_init(&mv, (uint8_t *)s->mvectors, 2 * s->nb_mvectors * sizeof(*s->mvectors));
for (int y = 0; y < avctx->height; y += 16) {
GetByteContext gbyte;
int pfill[3][1] = { {0} };
int nb_codes = get_bits(gb, 16);
skip_bits(gb, 8);
if (get_bits_left(gb) < 0) {
ret = AVERROR_INVALIDDATA;
goto fail;
}
av_fast_padded_malloc(&s->coeffs, &s->coeffs_size, nb_codes * sizeof(*s->coeffs));
if (!s->coeffs) {
ret = AVERROR(ENOMEM);
goto fail;
}
ret = decode_coeffs(gb, s->coeffs, nb_codes);
if (ret < 0)
goto fail;
bytestream2_init(&gbyte, (uint8_t *)s->coeffs, nb_codes * sizeof(*s->coeffs));
for (int x = 0; x < avctx->width; x += 16) {
if (cnt >= 4)
cnt = 0;
if (cnt == 0) {
if (get_bits_left(&mask) < 8) {
ret = AVERROR_INVALIDDATA;
goto fail;
}
flags = get_bits(&mask, 8);
}
dst[0] = frame->data[0] + linesize[0] * y + x;
dst[1] = frame->data[0] + linesize[0] * y + x + 8;
dst[2] = frame->data[0] + linesize[0] * (y + 8) + x;
dst[3] = frame->data[0] + linesize[0] * (y + 8) + x + 8;
dst[4] = frame->data[1] + linesize[4] * (y >> 1) + (x >> 1);
dst[5] = frame->data[2] + linesize[5] * (y >> 1) + (x >> 1);
if ((flags >> (cnt)) & 1) {
int mv_x = sign_extend(bytestream2_get_ne16(&mv), 16);
int mv_y = sign_extend(bytestream2_get_ne16(&mv), 16);
int px = x + mv_x;
int py = y + mv_y;
if (px < 0 || px > FFALIGN(avctx->width , 16) - 16 ||
py < 0 || py > FFALIGN(avctx->height, 16) - 16)
return AVERROR_INVALIDDATA;
src[0] = prev->data[0] + in_linesize[0] * py + px;
src[1] = prev->data[0] + in_linesize[0] * py + px + 8;
src[2] = prev->data[0] + in_linesize[0] * (py + 8) + px;
src[3] = prev->data[0] + in_linesize[0] * (py + 8) + px + 8;
src[4] = prev->data[1] + in_linesize[4] * (py >> 1) + (px >> 1);
src[5] = prev->data[2] + in_linesize[5] * (py >> 1) + (px >> 1);
if ((flags >> (cnt + 4)) & 1) {
for (int b = 0; b < 6; b++)
copy_block8(dst[b], src[b], linesize[b], in_linesize[b], 8);
} else {
for (int b = 0; b < 6; b++) {
int mode = get_bits_le(&mgb, 2);
ret = decode_inter_block(avctx, mode, &gbyte, s->interq_tab[b >= 4],
s->block[b],
pfill[(b >= 4) + (b >= 5)],
dst[b], linesize[b],
src[b], in_linesize[b]);
if (ret < 0)
goto fail;
}
}
} else {
for (int b = 0; b < 6; b++) {
int mode = get_bits_le(&mgb, 2);
ret = decode_intra_block(avctx, mode, &gbyte, s->intraq_tab[b >= 4],
s->block[b],
pfill[(b >= 4) + (b >= 5)],
dst[b], linesize[b]);
if (ret < 0)
goto fail;
}
}
cnt++;
}
}
fail:
return ret;
}
static int decode_frame(AVCodecContext *avctx, void *data,
int *got_frame, AVPacket *avpkt)
{
MV30Context *s = avctx->priv_data;
GetBitContext *gb = &s->gb;
AVFrame *frame = data;
int ret;
if ((ret = init_get_bits8(gb, avpkt->data, avpkt->size)) < 0)
return ret;
if ((ret = ff_get_buffer(avctx, frame, AV_GET_BUFFER_FLAG_REF)) < 0)
return ret;
s->intra_quant = get_bits(gb, 8);
s->inter_quant = s->intra_quant + get_sbits(gb, 8);
s->is_inter = get_bits_le(gb, 16);
s->mode_size = get_bits_le(gb, 16);
if (s->is_inter)
s->nb_mvectors = get_bits_le(gb, 16);
get_qtable(s->intraq_tab[0], s->intra_quant, luma_tab);
get_qtable(s->intraq_tab[1], s->intra_quant, chroma_tab);
frame->key_frame = s->is_inter == 0;
if (frame->key_frame) {
ret = decode_intra(avctx, gb, frame);
if (ret < 0)
return ret;
} else {
get_qtable(s->interq_tab[0], s->inter_quant, luma_tab);
get_qtable(s->interq_tab[1], s->inter_quant, chroma_tab);
if (!s->prev_frame->data[0]) {
av_log(avctx, AV_LOG_ERROR, "Missing reference frame.\n");
return AVERROR_INVALIDDATA;
}
ret = decode_inter(avctx, gb, frame, s->prev_frame);
if (ret < 0)
return ret;
}
av_frame_unref(s->prev_frame);
if ((ret = av_frame_ref(s->prev_frame, frame)) < 0)
return ret;
*got_frame = 1;
return avpkt->size;
}
static const uint16_t cbp_codes[] = {
0, 1, 4, 5, 6, 0xE, 0x1E, 0x3E, 0x7E, 0xFE, 0x1FE, 0x1FF,
};
static const uint8_t cbp_bits[] = {
2, 2, 3, 3, 3, 4, 5, 6, 7, 8, 9, 9,
};
static av_cold void init_static_data(void)
{
INIT_VLC_SPARSE_STATIC(&cbp_tab, 9, FF_ARRAY_ELEMS(cbp_bits),
cbp_bits, 1, 1, cbp_codes, 2, 2, NULL, 0, 0, 512);
}
static av_cold int decode_init(AVCodecContext *avctx)
{
MV30Context *s = avctx->priv_data;
static AVOnce init_static_once = AV_ONCE_INIT;
avctx->pix_fmt = AV_PIX_FMT_YUV420P;
avctx->color_range = AVCOL_RANGE_JPEG;
ff_blockdsp_init(&s->bdsp, avctx);
s->prev_frame = av_frame_alloc();
if (!s->prev_frame)
return AVERROR(ENOMEM);
ff_thread_once(&init_static_once, init_static_data);
return 0;
}
static void decode_flush(AVCodecContext *avctx)
{
MV30Context *s = avctx->priv_data;
av_frame_unref(s->prev_frame);
}
static av_cold int decode_close(AVCodecContext *avctx)
{
MV30Context *s = avctx->priv_data;
av_frame_free(&s->prev_frame);
av_freep(&s->coeffs);
s->coeffs_size = 0;
av_freep(&s->mvectors);
s->mvectors_size = 0;
return 0;
}
AVCodec ff_mv30_decoder = {
.name = "mv30",
.long_name = NULL_IF_CONFIG_SMALL("MidiVid 3.0"),
.type = AVMEDIA_TYPE_VIDEO,
.id = AV_CODEC_ID_MV30,
.priv_data_size = sizeof(MV30Context),
.init = decode_init,
.close = decode_close,
.decode = decode_frame,
.flush = decode_flush,
.capabilities = AV_CODEC_CAP_DR1,
.caps_internal = FF_CODEC_CAP_INIT_THREADSAFE |
FF_CODEC_CAP_INIT_CLEANUP,
};