1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-11-26 19:01:44 +02:00
FFmpeg/libavcodec/flashsv.c
Andreas Rheinhardt dcc3f62270 avcodec/flashsv: Avoid deflating data
Currently priming the zlib decompressor involves compressing
data directly after having decompressed it and decompressing
it again in order to set the "dictionary" and to initialize
the adler32-checksum. Yet this is wasteful and can be simplified
by synthetizing the compressed data via non-compressed blocks.

This reduced the amount of allocations for the decoding part
of fate-vsynth1-flashsv2, namely from
total heap usage: 9,135 allocs, 9,135 frees, 376,503,427 bytes allocated
to
total heap usage: 2,373 allocs, 2,373 frees, 14,144,083 bytes allocated

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2022-03-15 05:45:30 +01:00

580 lines
20 KiB
C

/*
* Flash Screen Video decoder
* Copyright (C) 2004 Alex Beregszaszi
* Copyright (C) 2006 Benjamin Larsson
*
* 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
*/
/**
* @file
* Flash Screen Video decoder
* @author Alex Beregszaszi
* @author Benjamin Larsson
* @author Daniel Verkamp
* @author Konstantin Shishkov
*
* A description of the bitstream format for Flash Screen Video version 1/2
* is part of the SWF File Format Specification (version 10), which can be
* downloaded from http://www.adobe.com/devnet/swf.html.
*/
#include <stdio.h>
#include <stdlib.h>
#include <zlib.h>
#include "libavutil/intreadwrite.h"
#include "avcodec.h"
#include "bytestream.h"
#include "get_bits.h"
#include "internal.h"
typedef struct BlockInfo {
const uint8_t *pos;
int size;
} BlockInfo;
typedef struct FlashSVContext {
AVCodecContext *avctx;
AVFrame *frame;
int image_width, image_height;
int block_width, block_height;
uint8_t *tmpblock;
int block_size;
z_stream zstream;
int ver;
const uint32_t *pal;
int is_keyframe;
const uint8_t *keyframedata;
AVBufferRef *keyframedata_buf;
uint8_t *keyframe;
BlockInfo *blocks;
int color_depth;
int zlibprime_curr, zlibprime_prev;
int diff_start, diff_height;
uint8_t tmp[UINT16_MAX];
} FlashSVContext;
static int decode_hybrid(const uint8_t *sptr, const uint8_t *sptr_end, uint8_t *dptr, int dx, int dy,
int h, int w, int stride, const uint32_t *pal)
{
int x, y;
const uint8_t *orig_src = sptr;
for (y = dx + h; y > dx; y--) {
uint8_t *dst = dptr + (y * stride) + dy * 3;
for (x = 0; x < w; x++) {
if (sptr >= sptr_end)
return AVERROR_INVALIDDATA;
if (*sptr & 0x80) {
/* 15-bit color */
unsigned c = AV_RB16(sptr) & ~0x8000;
unsigned b = c & 0x1F;
unsigned g = (c >> 5) & 0x1F;
unsigned r = c >> 10;
/* 000aaabb -> aaabbaaa */
*dst++ = (b << 3) | (b >> 2);
*dst++ = (g << 3) | (g >> 2);
*dst++ = (r << 3) | (r >> 2);
sptr += 2;
} else {
/* palette index */
uint32_t c = pal[*sptr++];
bytestream_put_le24(&dst, c);
}
}
}
return sptr - orig_src;
}
static av_cold int flashsv_decode_end(AVCodecContext *avctx)
{
FlashSVContext *s = avctx->priv_data;
inflateEnd(&s->zstream);
/* release the frame if needed */
av_frame_free(&s->frame);
/* free the tmpblock */
av_freep(&s->tmpblock);
return 0;
}
static av_cold int flashsv_decode_init(AVCodecContext *avctx)
{
FlashSVContext *s = avctx->priv_data;
int zret; // Zlib return code
s->avctx = avctx;
s->zstream.zalloc = Z_NULL;
s->zstream.zfree = Z_NULL;
s->zstream.opaque = Z_NULL;
zret = inflateInit(&s->zstream);
if (zret != Z_OK) {
av_log(avctx, AV_LOG_ERROR, "Inflate init error: %d\n", zret);
return AVERROR_EXTERNAL;
}
avctx->pix_fmt = AV_PIX_FMT_BGR24;
s->frame = av_frame_alloc();
if (!s->frame) {
return AVERROR(ENOMEM);
}
return 0;
}
static int flashsv2_prime(FlashSVContext *s, const uint8_t *src, int size)
{
int zret; // Zlib return code
static const uint8_t zlib_header[] = { 0x78, 0x01 };
uint8_t *data = s->tmpblock;
unsigned remaining;
if (!src)
return AVERROR_INVALIDDATA;
s->zstream.next_in = src;
s->zstream.avail_in = size;
s->zstream.next_out = data;
s->zstream.avail_out = s->block_size * 3;
inflate(&s->zstream, Z_SYNC_FLUSH);
remaining = s->block_size * 3 - s->zstream.avail_out;
if ((zret = inflateReset(&s->zstream)) != Z_OK) {
av_log(s->avctx, AV_LOG_ERROR, "Inflate reset error: %d\n", zret);
return AVERROR_UNKNOWN;
}
/* Create input for zlib that is equivalent to encoding the output
* from above and decoding it again (the net result of this is that
* the dictionary of past decoded data is correctly primed and
* the adler32 checksum is correctly initialized).
* This is accomplished by synthetizing blocks of uncompressed data
* out of the output from above. See section 3.2.4 of RFC 1951. */
s->zstream.next_in = zlib_header;
s->zstream.avail_in = sizeof(zlib_header);
inflate(&s->zstream, Z_SYNC_FLUSH);
while (remaining > 0) {
unsigned block_size = FFMIN(UINT16_MAX, remaining);
uint8_t header[5];
/* Bit 0: Non-last-block, bits 1-2: BTYPE for uncompressed block */
header[0] = 0;
/* Block size */
AV_WL16(header + 1, block_size);
/* Block size (one's complement) */
AV_WL16(header + 3, block_size ^ 0xFFFF);
s->zstream.next_in = header;
s->zstream.avail_in = sizeof(header);
s->zstream.next_out = s->tmp;
s->zstream.avail_out = sizeof(s->tmp);
zret = inflate(&s->zstream, Z_SYNC_FLUSH);
if (zret != Z_OK)
return AVERROR_UNKNOWN;
s->zstream.next_in = data;
s->zstream.avail_in = block_size;
zret = inflate(&s->zstream, Z_SYNC_FLUSH);
if (zret != Z_OK)
return AVERROR_UNKNOWN;
data += block_size;
remaining -= block_size;
}
return 0;
}
static int flashsv_decode_block(AVCodecContext *avctx, const AVPacket *avpkt,
GetBitContext *gb, int block_size,
int width, int height, int x_pos, int y_pos,
int blk_idx)
{
struct FlashSVContext *s = avctx->priv_data;
uint8_t *line = s->tmpblock;
int k;
int ret = inflateReset(&s->zstream);
if (ret != Z_OK) {
av_log(avctx, AV_LOG_ERROR, "Inflate reset error: %d\n", ret);
return AVERROR_UNKNOWN;
}
if (s->zlibprime_curr || s->zlibprime_prev) {
ret = flashsv2_prime(s,
s->blocks[blk_idx].pos,
s->blocks[blk_idx].size);
if (ret < 0)
return ret;
}
s->zstream.next_in = avpkt->data + get_bits_count(gb) / 8;
s->zstream.avail_in = block_size;
s->zstream.next_out = s->tmpblock;
s->zstream.avail_out = s->block_size * 3;
ret = inflate(&s->zstream, Z_FINISH);
if (ret == Z_DATA_ERROR) {
av_log(avctx, AV_LOG_ERROR, "Zlib resync occurred\n");
inflateSync(&s->zstream);
ret = inflate(&s->zstream, Z_FINISH);
}
if (ret != Z_OK && ret != Z_STREAM_END) {
//return -1;
}
if (s->is_keyframe) {
s->blocks[blk_idx].pos = s->keyframedata + (get_bits_count(gb) / 8);
s->blocks[blk_idx].size = block_size;
}
y_pos += s->diff_start;
if (!s->color_depth) {
/* Flash Screen Video stores the image upside down, so copy
* lines to destination in reverse order. */
for (k = 1; k <= s->diff_height; k++) {
memcpy(s->frame->data[0] + x_pos * 3 +
(s->image_height - y_pos - k) * s->frame->linesize[0],
line, width * 3);
/* advance source pointer to next line */
line += width * 3;
}
} else {
/* hybrid 15-bit/palette mode */
ret = decode_hybrid(s->tmpblock, s->zstream.next_out,
s->frame->data[0],
s->image_height - (y_pos + 1 + s->diff_height),
x_pos, s->diff_height, width,
s->frame->linesize[0], s->pal);
if (ret < 0) {
av_log(avctx, AV_LOG_ERROR, "decode_hybrid failed\n");
return ret;
}
}
skip_bits_long(gb, 8 * block_size); /* skip the consumed bits */
return 0;
}
static int flashsv_decode_frame(AVCodecContext *avctx, void *data,
int *got_frame, AVPacket *avpkt)
{
int buf_size = avpkt->size;
FlashSVContext *s = avctx->priv_data;
int h_blocks, v_blocks, h_part, v_part, i, j, ret;
GetBitContext gb;
int last_blockwidth = s->block_width;
int last_blockheight= s->block_height;
/* no supplementary picture */
if (buf_size == 0)
return 0;
if (buf_size < 4)
return -1;
if ((ret = init_get_bits8(&gb, avpkt->data, buf_size)) < 0)
return ret;
/* start to parse the bitstream */
s->block_width = 16 * (get_bits(&gb, 4) + 1);
s->image_width = get_bits(&gb, 12);
s->block_height = 16 * (get_bits(&gb, 4) + 1);
s->image_height = get_bits(&gb, 12);
if ( last_blockwidth != s->block_width
|| last_blockheight!= s->block_height)
av_freep(&s->blocks);
if (s->ver == 2) {
skip_bits(&gb, 6);
if (get_bits1(&gb)) {
avpriv_request_sample(avctx, "iframe");
return AVERROR_PATCHWELCOME;
}
if (get_bits1(&gb)) {
avpriv_request_sample(avctx, "Custom palette");
return AVERROR_PATCHWELCOME;
}
}
/* calculate number of blocks and size of border (partial) blocks */
h_blocks = s->image_width / s->block_width;
h_part = s->image_width % s->block_width;
v_blocks = s->image_height / s->block_height;
v_part = s->image_height % s->block_height;
/* the block size could change between frames, make sure the buffer
* is large enough, if not, get a larger one */
if (s->block_size < s->block_width * s->block_height) {
int tmpblock_size = 3 * s->block_width * s->block_height, err;
if ((err = av_reallocp(&s->tmpblock, tmpblock_size)) < 0) {
s->block_size = 0;
av_log(avctx, AV_LOG_ERROR,
"Cannot allocate decompression buffer.\n");
return err;
}
}
s->block_size = s->block_width * s->block_height;
/* initialize the image size once */
if (avctx->width == 0 && avctx->height == 0) {
if ((ret = ff_set_dimensions(avctx, s->image_width, s->image_height)) < 0)
return ret;
}
/* check for changes of image width and image height */
if (avctx->width != s->image_width || avctx->height != s->image_height) {
av_log(avctx, AV_LOG_ERROR,
"Frame width or height differs from first frame!\n");
av_log(avctx, AV_LOG_ERROR, "fh = %d, fv %d vs ch = %d, cv = %d\n",
avctx->height, avctx->width, s->image_height, s->image_width);
return AVERROR_INVALIDDATA;
}
/* we care for keyframes only in Screen Video v2 */
s->is_keyframe = (avpkt->flags & AV_PKT_FLAG_KEY) && (s->ver == 2);
if (s->is_keyframe) {
int err = av_buffer_replace(&s->keyframedata_buf, avpkt->buf);
if (err < 0)
return err;
s->keyframedata = avpkt->data;
}
if(s->ver == 2 && !s->blocks)
s->blocks = av_mallocz((v_blocks + !!v_part) * (h_blocks + !!h_part) *
sizeof(s->blocks[0]));
ff_dlog(avctx, "image: %dx%d block: %dx%d num: %dx%d part: %dx%d\n",
s->image_width, s->image_height, s->block_width, s->block_height,
h_blocks, v_blocks, h_part, v_part);
if ((ret = ff_reget_buffer(avctx, s->frame, 0)) < 0)
return ret;
/* loop over all block columns */
for (j = 0; j < v_blocks + (v_part ? 1 : 0); j++) {
int y_pos = j * s->block_height; // vertical position in frame
int cur_blk_height = (j < v_blocks) ? s->block_height : v_part;
/* loop over all block rows */
for (i = 0; i < h_blocks + (h_part ? 1 : 0); i++) {
int x_pos = i * s->block_width; // horizontal position in frame
int cur_blk_width = (i < h_blocks) ? s->block_width : h_part;
int has_diff = 0;
/* get the size of the compressed zlib chunk */
int size = get_bits(&gb, 16);
s->color_depth = 0;
s->zlibprime_curr = 0;
s->zlibprime_prev = 0;
s->diff_start = 0;
s->diff_height = cur_blk_height;
if (8 * size > get_bits_left(&gb)) {
av_frame_unref(s->frame);
return AVERROR_INVALIDDATA;
}
if (s->ver == 2 && size) {
skip_bits(&gb, 3);
s->color_depth = get_bits(&gb, 2);
has_diff = get_bits1(&gb);
s->zlibprime_curr = get_bits1(&gb);
s->zlibprime_prev = get_bits1(&gb);
if (s->color_depth != 0 && s->color_depth != 2) {
av_log(avctx, AV_LOG_ERROR,
"%dx%d invalid color depth %d\n",
i, j, s->color_depth);
return AVERROR_INVALIDDATA;
}
if (has_diff) {
if (size < 3) {
av_log(avctx, AV_LOG_ERROR, "size too small for diff\n");
return AVERROR_INVALIDDATA;
}
if (!s->keyframe) {
av_log(avctx, AV_LOG_ERROR,
"Inter frame without keyframe\n");
return AVERROR_INVALIDDATA;
}
s->diff_start = get_bits(&gb, 8);
s->diff_height = get_bits(&gb, 8);
if (s->diff_start + s->diff_height > cur_blk_height) {
av_log(avctx, AV_LOG_ERROR,
"Block parameters invalid: %d + %d > %d\n",
s->diff_start, s->diff_height, cur_blk_height);
return AVERROR_INVALIDDATA;
}
av_log(avctx, AV_LOG_DEBUG,
"%dx%d diff start %d height %d\n",
i, j, s->diff_start, s->diff_height);
size -= 2;
}
if (s->zlibprime_prev)
av_log(avctx, AV_LOG_DEBUG, "%dx%d zlibprime_prev\n", i, j);
if (s->zlibprime_curr) {
int col = get_bits(&gb, 8);
int row = get_bits(&gb, 8);
av_log(avctx, AV_LOG_DEBUG, "%dx%d zlibprime_curr %dx%d\n",
i, j, col, row);
if (size < 3) {
av_log(avctx, AV_LOG_ERROR, "size too small for zlibprime_curr\n");
return AVERROR_INVALIDDATA;
}
size -= 2;
avpriv_request_sample(avctx, "zlibprime_curr");
return AVERROR_PATCHWELCOME;
}
if (!s->blocks && (s->zlibprime_curr || s->zlibprime_prev)) {
av_log(avctx, AV_LOG_ERROR,
"no data available for zlib priming\n");
return AVERROR_INVALIDDATA;
}
size--; // account for flags byte
}
if (has_diff) {
int k;
int off = (s->image_height - y_pos - 1) * s->frame->linesize[0];
for (k = 0; k < cur_blk_height; k++) {
int x = off - k * s->frame->linesize[0] + x_pos * 3;
memcpy(s->frame->data[0] + x, s->keyframe + x,
cur_blk_width * 3);
}
}
/* skip unchanged blocks, which have size 0 */
if (size) {
if (flashsv_decode_block(avctx, avpkt, &gb, size,
cur_blk_width, cur_blk_height,
x_pos, y_pos,
i + j * (h_blocks + !!h_part)))
av_log(avctx, AV_LOG_ERROR,
"error in decompression of block %dx%d\n", i, j);
}
}
}
if (s->is_keyframe && s->ver == 2) {
if (!s->keyframe) {
s->keyframe = av_malloc(s->frame->linesize[0] * avctx->height);
if (!s->keyframe) {
av_log(avctx, AV_LOG_ERROR, "Cannot allocate image data\n");
return AVERROR(ENOMEM);
}
}
memcpy(s->keyframe, s->frame->data[0],
s->frame->linesize[0] * avctx->height);
}
if ((ret = av_frame_ref(data, s->frame)) < 0)
return ret;
*got_frame = 1;
if ((get_bits_count(&gb) / 8) != buf_size)
av_log(avctx, AV_LOG_ERROR, "buffer not fully consumed (%d != %d)\n",
buf_size, (get_bits_count(&gb) / 8));
/* report that the buffer was completely consumed */
return buf_size;
}
#if CONFIG_FLASHSV_DECODER
const AVCodec ff_flashsv_decoder = {
.name = "flashsv",
.long_name = NULL_IF_CONFIG_SMALL("Flash Screen Video v1"),
.type = AVMEDIA_TYPE_VIDEO,
.id = AV_CODEC_ID_FLASHSV,
.priv_data_size = sizeof(FlashSVContext),
.init = flashsv_decode_init,
.close = flashsv_decode_end,
.decode = flashsv_decode_frame,
.capabilities = AV_CODEC_CAP_DR1,
.caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_INIT_CLEANUP,
.pix_fmts = (const enum AVPixelFormat[]) { AV_PIX_FMT_BGR24, AV_PIX_FMT_NONE },
};
#endif /* CONFIG_FLASHSV_DECODER */
#if CONFIG_FLASHSV2_DECODER
static const uint32_t ff_flashsv2_default_palette[128] = {
0x000000, 0x333333, 0x666666, 0x999999, 0xCCCCCC, 0xFFFFFF,
0x330000, 0x660000, 0x990000, 0xCC0000, 0xFF0000, 0x003300,
0x006600, 0x009900, 0x00CC00, 0x00FF00, 0x000033, 0x000066,
0x000099, 0x0000CC, 0x0000FF, 0x333300, 0x666600, 0x999900,
0xCCCC00, 0xFFFF00, 0x003333, 0x006666, 0x009999, 0x00CCCC,
0x00FFFF, 0x330033, 0x660066, 0x990099, 0xCC00CC, 0xFF00FF,
0xFFFF33, 0xFFFF66, 0xFFFF99, 0xFFFFCC, 0xFF33FF, 0xFF66FF,
0xFF99FF, 0xFFCCFF, 0x33FFFF, 0x66FFFF, 0x99FFFF, 0xCCFFFF,
0xCCCC33, 0xCCCC66, 0xCCCC99, 0xCCCCFF, 0xCC33CC, 0xCC66CC,
0xCC99CC, 0xCCFFCC, 0x33CCCC, 0x66CCCC, 0x99CCCC, 0xFFCCCC,
0x999933, 0x999966, 0x9999CC, 0x9999FF, 0x993399, 0x996699,
0x99CC99, 0x99FF99, 0x339999, 0x669999, 0xCC9999, 0xFF9999,
0x666633, 0x666699, 0x6666CC, 0x6666FF, 0x663366, 0x669966,
0x66CC66, 0x66FF66, 0x336666, 0x996666, 0xCC6666, 0xFF6666,
0x333366, 0x333399, 0x3333CC, 0x3333FF, 0x336633, 0x339933,
0x33CC33, 0x33FF33, 0x663333, 0x993333, 0xCC3333, 0xFF3333,
0x003366, 0x336600, 0x660033, 0x006633, 0x330066, 0x663300,
0x336699, 0x669933, 0x993366, 0x339966, 0x663399, 0x996633,
0x6699CC, 0x99CC66, 0xCC6699, 0x66CC99, 0x9966CC, 0xCC9966,
0x99CCFF, 0xCCFF99, 0xFF99CC, 0x99FFCC, 0xCC99FF, 0xFFCC99,
0x111111, 0x222222, 0x444444, 0x555555, 0xAAAAAA, 0xBBBBBB,
0xDDDDDD, 0xEEEEEE
};
static av_cold int flashsv2_decode_init(AVCodecContext *avctx)
{
FlashSVContext *s = avctx->priv_data;
int ret;
ret = flashsv_decode_init(avctx);
if (ret < 0)
return ret;
s->pal = ff_flashsv2_default_palette;
s->ver = 2;
return 0;
}
static av_cold int flashsv2_decode_end(AVCodecContext *avctx)
{
FlashSVContext *s = avctx->priv_data;
av_buffer_unref(&s->keyframedata_buf);
av_freep(&s->blocks);
av_freep(&s->keyframe);
flashsv_decode_end(avctx);
return 0;
}
const AVCodec ff_flashsv2_decoder = {
.name = "flashsv2",
.long_name = NULL_IF_CONFIG_SMALL("Flash Screen Video v2"),
.type = AVMEDIA_TYPE_VIDEO,
.id = AV_CODEC_ID_FLASHSV2,
.priv_data_size = sizeof(FlashSVContext),
.init = flashsv2_decode_init,
.close = flashsv2_decode_end,
.decode = flashsv_decode_frame,
.capabilities = AV_CODEC_CAP_DR1,
.caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_INIT_CLEANUP,
.pix_fmts = (const enum AVPixelFormat[]) { AV_PIX_FMT_BGR24, AV_PIX_FMT_NONE },
};
#endif /* CONFIG_FLASHSV2_DECODER */