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.

1750 lines
59 KiB

/*
* PNG image format
* Copyright (c) 2003 Fabrice Bellard
*
* 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
*/
//#define DEBUG
#include "config_components.h"
#include "libavutil/avassert.h"
#include "libavutil/bprint.h"
#include "libavutil/crc.h"
#include "libavutil/imgutils.h"
#include "libavutil/intreadwrite.h"
#include "libavutil/stereo3d.h"
#include "libavutil/mastering_display_metadata.h"
#include "avcodec.h"
#include "bytestream.h"
#include "codec_internal.h"
#include "internal.h"
#include "apng.h"
#include "png.h"
#include "pngdsp.h"
#include "thread.h"
#include "threadframe.h"
#include "zlib_wrapper.h"
#include <zlib.h>
enum PNGHeaderState {
PNG_IHDR = 1 << 0,
PNG_PLTE = 1 << 1,
};
enum PNGImageState {
PNG_IDAT = 1 << 0,
PNG_ALLIMAGE = 1 << 1,
};
typedef struct PNGDecContext {
PNGDSPContext dsp;
AVCodecContext *avctx;
GetByteContext gb;
ThreadFrame last_picture;
ThreadFrame picture;
AVDictionary *frame_metadata;
uint8_t iccp_name[82];
uint8_t *iccp_data;
size_t iccp_data_len;
int stereo_mode;
int have_chrm;
uint32_t white_point[2];
uint32_t display_primaries[3][2];
enum PNGHeaderState hdr_state;
enum PNGImageState pic_state;
int width, height;
int cur_w, cur_h;
int x_offset, y_offset;
uint8_t dispose_op, blend_op;
int bit_depth;
int color_type;
int compression_type;
int interlace_type;
int filter_type;
int channels;
int bits_per_pixel;
int bpp;
int has_trns;
uint8_t transparent_color_be[6];
uint32_t palette[256];
uint8_t *crow_buf;
uint8_t *last_row;
unsigned int last_row_size;
uint8_t *tmp_row;
unsigned int tmp_row_size;
uint8_t *buffer;
int buffer_size;
int pass;
int crow_size; /* compressed row size (include filter type) */
int row_size; /* decompressed row size */
int pass_row_size; /* decompress row size of the current pass */
int y;
FFZStream zstream;
} PNGDecContext;
/* Mask to determine which pixels are valid in a pass */
static const uint8_t png_pass_mask[NB_PASSES] = {
0x01, 0x01, 0x11, 0x11, 0x55, 0x55, 0xff,
};
/* Mask to determine which y pixels can be written in a pass */
static const uint8_t png_pass_dsp_ymask[NB_PASSES] = {
0xff, 0xff, 0x0f, 0xff, 0x33, 0xff, 0x55,
};
/* Mask to determine which pixels to overwrite while displaying */
static const uint8_t png_pass_dsp_mask[NB_PASSES] = {
0xff, 0x0f, 0xff, 0x33, 0xff, 0x55, 0xff
};
/* NOTE: we try to construct a good looking image at each pass. width
* is the original image width. We also do pixel format conversion at
* this stage */
static void png_put_interlaced_row(uint8_t *dst, int width,
int bits_per_pixel, int pass,
int color_type, const uint8_t *src)
{
int x, mask, dsp_mask, j, src_x, b, bpp;
uint8_t *d;
const uint8_t *s;
mask = png_pass_mask[pass];
dsp_mask = png_pass_dsp_mask[pass];
switch (bits_per_pixel) {
case 1:
src_x = 0;
for (x = 0; x < width; x++) {
j = (x & 7);
if ((dsp_mask << j) & 0x80) {
b = (src[src_x >> 3] >> (7 - (src_x & 7))) & 1;
dst[x >> 3] &= 0xFF7F>>j;
dst[x >> 3] |= b << (7 - j);
}
if ((mask << j) & 0x80)
src_x++;
}
break;
case 2:
src_x = 0;
for (x = 0; x < width; x++) {
int j2 = 2 * (x & 3);
j = (x & 7);
if ((dsp_mask << j) & 0x80) {
b = (src[src_x >> 2] >> (6 - 2*(src_x & 3))) & 3;
dst[x >> 2] &= 0xFF3F>>j2;
dst[x >> 2] |= b << (6 - j2);
}
if ((mask << j) & 0x80)
src_x++;
}
break;
case 4:
src_x = 0;
for (x = 0; x < width; x++) {
int j2 = 4*(x&1);
j = (x & 7);
if ((dsp_mask << j) & 0x80) {
b = (src[src_x >> 1] >> (4 - 4*(src_x & 1))) & 15;
dst[x >> 1] &= 0xFF0F>>j2;
dst[x >> 1] |= b << (4 - j2);
}
if ((mask << j) & 0x80)
src_x++;
}
break;
default:
bpp = bits_per_pixel >> 3;
d = dst;
s = src;
for (x = 0; x < width; x++) {
j = x & 7;
if ((dsp_mask << j) & 0x80) {
memcpy(d, s, bpp);
}
d += bpp;
if ((mask << j) & 0x80)
s += bpp;
}
break;
}
}
void ff_add_png_paeth_prediction(uint8_t *dst, uint8_t *src, uint8_t *top,
int w, int bpp)
{
int i;
for (i = 0; i < w; i++) {
int a, b, c, p, pa, pb, pc;
a = dst[i - bpp];
b = top[i];
c = top[i - bpp];
p = b - c;
pc = a - c;
pa = abs(p);
pb = abs(pc);
pc = abs(p + pc);
if (pa <= pb && pa <= pc)
p = a;
else if (pb <= pc)
p = b;
else
p = c;
dst[i] = p + src[i];
}
}
#define UNROLL1(bpp, op) \
{ \
r = dst[0]; \
if (bpp >= 2) \
g = dst[1]; \
if (bpp >= 3) \
b = dst[2]; \
if (bpp >= 4) \
a = dst[3]; \
for (; i <= size - bpp; i += bpp) { \
dst[i + 0] = r = op(r, src[i + 0], last[i + 0]); \
if (bpp == 1) \
continue; \
dst[i + 1] = g = op(g, src[i + 1], last[i + 1]); \
if (bpp == 2) \
continue; \
dst[i + 2] = b = op(b, src[i + 2], last[i + 2]); \
if (bpp == 3) \
continue; \
dst[i + 3] = a = op(a, src[i + 3], last[i + 3]); \
} \
}
#define UNROLL_FILTER(op) \
if (bpp == 1) { \
UNROLL1(1, op) \
} else if (bpp == 2) { \
UNROLL1(2, op) \
} else if (bpp == 3) { \
UNROLL1(3, op) \
} else if (bpp == 4) { \
UNROLL1(4, op) \
} \
for (; i < size; i++) { \
dst[i] = op(dst[i - bpp], src[i], last[i]); \
}
/* NOTE: 'dst' can be equal to 'last' */
void ff_png_filter_row(PNGDSPContext *dsp, uint8_t *dst, int filter_type,
uint8_t *src, uint8_t *last, int size, int bpp)
{
int i, p, r, g, b, a;
switch (filter_type) {
case PNG_FILTER_VALUE_NONE:
memcpy(dst, src, size);
break;
case PNG_FILTER_VALUE_SUB:
for (i = 0; i < bpp; i++)
dst[i] = src[i];
if (bpp == 4) {
p = *(int *)dst;
for (; i < size; i += bpp) {
unsigned s = *(int *)(src + i);
p = ((s & 0x7f7f7f7f) + (p & 0x7f7f7f7f)) ^ ((s ^ p) & 0x80808080);
*(int *)(dst + i) = p;
}
} else {
#define OP_SUB(x, s, l) ((x) + (s))
UNROLL_FILTER(OP_SUB);
}
break;
case PNG_FILTER_VALUE_UP:
dsp->add_bytes_l2(dst, src, last, size);
break;
case PNG_FILTER_VALUE_AVG:
for (i = 0; i < bpp; i++) {
p = (last[i] >> 1);
dst[i] = p + src[i];
}
#define OP_AVG(x, s, l) (((((x) + (l)) >> 1) + (s)) & 0xff)
UNROLL_FILTER(OP_AVG);
break;
case PNG_FILTER_VALUE_PAETH:
for (i = 0; i < bpp; i++) {
p = last[i];
dst[i] = p + src[i];
}
if (bpp > 2 && size > 4) {
/* would write off the end of the array if we let it process
* the last pixel with bpp=3 */
int w = (bpp & 3) ? size - 3 : size;
if (w > i) {
dsp->add_paeth_prediction(dst + i, src + i, last + i, size - i, bpp);
i = w;
}
}
ff_add_png_paeth_prediction(dst + i, src + i, last + i, size - i, bpp);
break;
}
}
/* This used to be called "deloco" in FFmpeg
* and is actually an inverse reversible colorspace transformation */
#define YUV2RGB(NAME, TYPE) \
static void deloco_ ## NAME(TYPE *dst, int size, int alpha) \
{ \
int i; \
for (i = 0; i < size; i += 3 + alpha) { \
int g = dst [i + 1]; \
dst[i + 0] += g; \
dst[i + 2] += g; \
} \
}
YUV2RGB(rgb8, uint8_t)
YUV2RGB(rgb16, uint16_t)
static int percent_missing(PNGDecContext *s)
{
if (s->interlace_type) {
return 100 - 100 * s->pass / (NB_PASSES - 1);
} else {
return 100 - 100 * s->y / s->cur_h;
}
}
/* process exactly one decompressed row */
static void png_handle_row(PNGDecContext *s, uint8_t *dst, ptrdiff_t dst_stride)
{
uint8_t *ptr, *last_row;
int got_line;
if (!s->interlace_type) {
ptr = dst + dst_stride * (s->y + s->y_offset) + s->x_offset * s->bpp;
if (s->y == 0)
last_row = s->last_row;
else
last_row = ptr - dst_stride;
ff_png_filter_row(&s->dsp, ptr, s->crow_buf[0], s->crow_buf + 1,
last_row, s->row_size, s->bpp);
/* loco lags by 1 row so that it doesn't interfere with top prediction */
if (s->filter_type == PNG_FILTER_TYPE_LOCO && s->y > 0) {
if (s->bit_depth == 16) {
deloco_rgb16((uint16_t *)(ptr - dst_stride), s->row_size / 2,
s->color_type == PNG_COLOR_TYPE_RGB_ALPHA);
} else {
deloco_rgb8(ptr - dst_stride, s->row_size,
s->color_type == PNG_COLOR_TYPE_RGB_ALPHA);
}
}
s->y++;
if (s->y == s->cur_h) {
s->pic_state |= PNG_ALLIMAGE;
if (s->filter_type == PNG_FILTER_TYPE_LOCO) {
if (s->bit_depth == 16) {
deloco_rgb16((uint16_t *)ptr, s->row_size / 2,
s->color_type == PNG_COLOR_TYPE_RGB_ALPHA);
} else {
deloco_rgb8(ptr, s->row_size,
s->color_type == PNG_COLOR_TYPE_RGB_ALPHA);
}
}
}
} else {
got_line = 0;
for (;;) {
ptr = dst + dst_stride * (s->y + s->y_offset) + s->x_offset * s->bpp;
if ((ff_png_pass_ymask[s->pass] << (s->y & 7)) & 0x80) {
/* if we already read one row, it is time to stop to
* wait for the next one */
if (got_line)
break;
ff_png_filter_row(&s->dsp, s->tmp_row, s->crow_buf[0], s->crow_buf + 1,
s->last_row, s->pass_row_size, s->bpp);
FFSWAP(uint8_t *, s->last_row, s->tmp_row);
FFSWAP(unsigned int, s->last_row_size, s->tmp_row_size);
got_line = 1;
}
if ((png_pass_dsp_ymask[s->pass] << (s->y & 7)) & 0x80) {
png_put_interlaced_row(ptr, s->cur_w, s->bits_per_pixel, s->pass,
s->color_type, s->last_row);
}
s->y++;
if (s->y == s->cur_h) {
memset(s->last_row, 0, s->row_size);
for (;;) {
if (s->pass == NB_PASSES - 1) {
s->pic_state |= PNG_ALLIMAGE;
goto the_end;
} else {
s->pass++;
s->y = 0;
s->pass_row_size = ff_png_pass_row_size(s->pass,
s->bits_per_pixel,
s->cur_w);
s->crow_size = s->pass_row_size + 1;
if (s->pass_row_size != 0)
break;
/* skip pass if empty row */
}
}
}
}
the_end:;
}
}
static int png_decode_idat(PNGDecContext *s, GetByteContext *gb,
uint8_t *dst, ptrdiff_t dst_stride)
{
z_stream *const zstream = &s->zstream.zstream;
int ret;
zstream->avail_in = bytestream2_get_bytes_left(gb);
zstream->next_in = gb->buffer;
/* decode one line if possible */
while (zstream->avail_in > 0) {
ret = inflate(zstream, Z_PARTIAL_FLUSH);
if (ret != Z_OK && ret != Z_STREAM_END) {
av_log(s->avctx, AV_LOG_ERROR, "inflate returned error %d\n", ret);
return AVERROR_EXTERNAL;
}
if (zstream->avail_out == 0) {
if (!(s->pic_state & PNG_ALLIMAGE)) {
png_handle_row(s, dst, dst_stride);
}
zstream->avail_out = s->crow_size;
zstream->next_out = s->crow_buf;
}
if (ret == Z_STREAM_END && zstream->avail_in > 0) {
av_log(s->avctx, AV_LOG_WARNING,
"%d undecompressed bytes left in buffer\n", zstream->avail_in);
return 0;
}
}
return 0;
}
static int decode_zbuf(AVBPrint *bp, const uint8_t *data,
const uint8_t *data_end, void *logctx)
{
FFZStream z;
z_stream *const zstream = &z.zstream;
unsigned char *buf;
unsigned buf_size;
int ret = ff_inflate_init(&z, logctx);
if (ret < 0)
return ret;
zstream->next_in = data;
zstream->avail_in = data_end - data;
av_bprint_init(bp, 0, AV_BPRINT_SIZE_UNLIMITED);
while (zstream->avail_in > 0) {
av_bprint_get_buffer(bp, 2, &buf, &buf_size);
if (buf_size < 2) {
ret = AVERROR(ENOMEM);
goto fail;
}
zstream->next_out = buf;
zstream->avail_out = buf_size - 1;
ret = inflate(zstream, Z_PARTIAL_FLUSH);
if (ret != Z_OK && ret != Z_STREAM_END) {
ret = AVERROR_EXTERNAL;
goto fail;
}
bp->len += zstream->next_out - buf;
if (ret == Z_STREAM_END)
break;
}
ff_inflate_end(&z);
bp->str[bp->len] = 0;
return 0;
fail:
ff_inflate_end(&z);
av_bprint_finalize(bp, NULL);
return ret;
}
static char *iso88591_to_utf8(const char *in, size_t size_in)
{
size_t extra = 0, i;
char *out, *q;
for (i = 0; i < size_in; i++)
extra += !!(in[i] & 0x80);
if (size_in == SIZE_MAX || extra > SIZE_MAX - size_in - 1)
return NULL;
q = out = av_malloc(size_in + extra + 1);
if (!out)
return NULL;
for (i = 0; i < size_in; i++) {
if (in[i] & 0x80) {
*(q++) = 0xC0 | (in[i] >> 6);
*(q++) = 0x80 | (in[i] & 0x3F);
} else {
*(q++) = in[i];
}
}
*(q++) = 0;
return out;
}
static int decode_text_chunk(PNGDecContext *s, GetByteContext *gb, int compressed)
{
int ret, method;
const uint8_t *data = gb->buffer;
const uint8_t *data_end = gb->buffer_end;
const char *keyword = data;
const char *keyword_end = memchr(keyword, 0, data_end - data);
char *kw_utf8 = NULL, *txt_utf8 = NULL;
const char *text;
unsigned text_len;
AVBPrint bp;
if (!keyword_end)
return AVERROR_INVALIDDATA;
data = keyword_end + 1;
if (compressed) {
if (data == data_end)
return AVERROR_INVALIDDATA;
method = *(data++);
if (method)
return AVERROR_INVALIDDATA;
if ((ret = decode_zbuf(&bp, data, data_end, s->avctx)) < 0)
return ret;
text = bp.str;
text_len = bp.len;
} else {
text = data;
text_len = data_end - data;
}
txt_utf8 = iso88591_to_utf8(text, text_len);
if (compressed)
av_bprint_finalize(&bp, NULL);
if (!txt_utf8)
return AVERROR(ENOMEM);
kw_utf8 = iso88591_to_utf8(keyword, keyword_end - keyword);
if (!kw_utf8) {
av_free(txt_utf8);
return AVERROR(ENOMEM);
}
av_dict_set(&s->frame_metadata, kw_utf8, txt_utf8,
AV_DICT_DONT_STRDUP_KEY | AV_DICT_DONT_STRDUP_VAL);
return 0;
}
static int decode_ihdr_chunk(AVCodecContext *avctx, PNGDecContext *s,
GetByteContext *gb)
{
if (bytestream2_get_bytes_left(gb) != 13)
return AVERROR_INVALIDDATA;
if (s->pic_state & PNG_IDAT) {
av_log(avctx, AV_LOG_ERROR, "IHDR after IDAT\n");
return AVERROR_INVALIDDATA;
}
if (s->hdr_state & PNG_IHDR) {
av_log(avctx, AV_LOG_ERROR, "Multiple IHDR\n");
return AVERROR_INVALIDDATA;
}
s->width = s->cur_w = bytestream2_get_be32(gb);
s->height = s->cur_h = bytestream2_get_be32(gb);
if (av_image_check_size(s->width, s->height, 0, avctx)) {
s->cur_w = s->cur_h = s->width = s->height = 0;
av_log(avctx, AV_LOG_ERROR, "Invalid image size\n");
return AVERROR_INVALIDDATA;
}
s->bit_depth = bytestream2_get_byte(gb);
if (s->bit_depth != 1 && s->bit_depth != 2 && s->bit_depth != 4 &&
s->bit_depth != 8 && s->bit_depth != 16) {
av_log(avctx, AV_LOG_ERROR, "Invalid bit depth\n");
goto error;
}
s->color_type = bytestream2_get_byte(gb);
s->compression_type = bytestream2_get_byte(gb);
if (s->compression_type) {
av_log(avctx, AV_LOG_ERROR, "Invalid compression method %d\n", s->compression_type);
goto error;
}
s->filter_type = bytestream2_get_byte(gb);
s->interlace_type = bytestream2_get_byte(gb);
s->hdr_state |= PNG_IHDR;
if (avctx->debug & FF_DEBUG_PICT_INFO)
av_log(avctx, AV_LOG_DEBUG, "width=%d height=%d depth=%d color_type=%d "
"compression_type=%d filter_type=%d interlace_type=%d\n",
s->width, s->height, s->bit_depth, s->color_type,
s->compression_type, s->filter_type, s->interlace_type);
return 0;
error:
s->cur_w = s->cur_h = s->width = s->height = 0;
s->bit_depth = 8;
return AVERROR_INVALIDDATA;
}
static int decode_phys_chunk(AVCodecContext *avctx, PNGDecContext *s,
GetByteContext *gb)
{
if (s->pic_state & PNG_IDAT) {
av_log(avctx, AV_LOG_ERROR, "pHYs after IDAT\n");
return AVERROR_INVALIDDATA;
}
avctx->sample_aspect_ratio.num = bytestream2_get_be32(gb);
avctx->sample_aspect_ratio.den = bytestream2_get_be32(gb);
if (avctx->sample_aspect_ratio.num < 0 || avctx->sample_aspect_ratio.den < 0)
avctx->sample_aspect_ratio = (AVRational){ 0, 1 };
bytestream2_skip(gb, 1); /* unit specifier */
return 0;
}
static int decode_idat_chunk(AVCodecContext *avctx, PNGDecContext *s,
GetByteContext *gb, AVFrame *p)
{
int ret;
size_t byte_depth = s->bit_depth > 8 ? 2 : 1;
if (!(s->hdr_state & PNG_IHDR)) {
av_log(avctx, AV_LOG_ERROR, "IDAT without IHDR\n");
return AVERROR_INVALIDDATA;
}
if (!(s->pic_state & PNG_IDAT)) {
/* init image info */
ret = ff_set_dimensions(avctx, s->width, s->height);
if (ret < 0)
return ret;
s->channels = ff_png_get_nb_channels(s->color_type);
s->bits_per_pixel = s->bit_depth * s->channels;
s->bpp = (s->bits_per_pixel + 7) >> 3;
s->row_size = (s->cur_w * s->bits_per_pixel + 7) >> 3;
if ((s->bit_depth == 2 || s->bit_depth == 4 || s->bit_depth == 8) &&
s->color_type == PNG_COLOR_TYPE_RGB) {
avctx->pix_fmt = AV_PIX_FMT_RGB24;
} else if ((s->bit_depth == 2 || s->bit_depth == 4 || s->bit_depth == 8) &&
s->color_type == PNG_COLOR_TYPE_RGB_ALPHA) {
avctx->pix_fmt = AV_PIX_FMT_RGBA;
} else if ((s->bit_depth == 2 || s->bit_depth == 4 || s->bit_depth == 8) &&
s->color_type == PNG_COLOR_TYPE_GRAY) {
avctx->pix_fmt = AV_PIX_FMT_GRAY8;
} else if (s->bit_depth == 16 &&
s->color_type == PNG_COLOR_TYPE_GRAY) {
avctx->pix_fmt = AV_PIX_FMT_GRAY16BE;
} else if (s->bit_depth == 16 &&
s->color_type == PNG_COLOR_TYPE_RGB) {
avctx->pix_fmt = AV_PIX_FMT_RGB48BE;
} else if (s->bit_depth == 16 &&
s->color_type == PNG_COLOR_TYPE_RGB_ALPHA) {
avctx->pix_fmt = AV_PIX_FMT_RGBA64BE;
} else if ((s->bits_per_pixel == 1 || s->bits_per_pixel == 2 || s->bits_per_pixel == 4 || s->bits_per_pixel == 8) &&
s->color_type == PNG_COLOR_TYPE_PALETTE) {
avctx->pix_fmt = avctx->codec_id == AV_CODEC_ID_APNG ? AV_PIX_FMT_RGBA : AV_PIX_FMT_PAL8;
} else if (s->bit_depth == 1 && s->bits_per_pixel == 1 && avctx->codec_id != AV_CODEC_ID_APNG) {
avctx->pix_fmt = AV_PIX_FMT_MONOBLACK;
} else if (s->bit_depth == 8 &&
s->color_type == PNG_COLOR_TYPE_GRAY_ALPHA) {
avctx->pix_fmt = AV_PIX_FMT_YA8;
} else if (s->bit_depth == 16 &&
s->color_type == PNG_COLOR_TYPE_GRAY_ALPHA) {
avctx->pix_fmt = AV_PIX_FMT_YA16BE;
} else {
avpriv_report_missing_feature(avctx,
"Bit depth %d color type %d",
s->bit_depth, s->color_type);
return AVERROR_PATCHWELCOME;
}
if (s->has_trns && s->color_type != PNG_COLOR_TYPE_PALETTE) {
switch (avctx->pix_fmt) {
case AV_PIX_FMT_RGB24:
avctx->pix_fmt = AV_PIX_FMT_RGBA;
break;
case AV_PIX_FMT_RGB48BE:
avctx->pix_fmt = AV_PIX_FMT_RGBA64BE;
break;
case AV_PIX_FMT_GRAY8:
avctx->pix_fmt = AV_PIX_FMT_YA8;
break;
case AV_PIX_FMT_GRAY16BE:
avctx->pix_fmt = AV_PIX_FMT_YA16BE;
break;
default:
avpriv_request_sample(avctx, "bit depth %d "
"and color type %d with TRNS",
s->bit_depth, s->color_type);
return AVERROR_INVALIDDATA;
}
s->bpp += byte_depth;
}
ff_thread_release_ext_buffer(avctx, &s->picture);
avcodec/pngdec: Fix APNG_DISPOSE_OP_BACKGROUND APNG works with a single reference frame and an output frame. According to the spec, decoding APNG works by decoding the current IDAT/fdAT chunks (which decodes to a rectangular subregion of the whole image region), followed by either overwriting the region of the output frame with the newly decoded data or by blending the newly decoded data with the data from the reference frame onto the current subregion of the output frame. The remainder of the output frame is just copied from the reference frame. Then the reference frame might be left untouched (APNG_DISPOSE_OP_PREVIOUS), it might be replaced by the output frame (APNG_DISPOSE_OP_NONE) or the rectangular subregion corresponding to the just decoded frame has to be reset to black (APNG_DISPOSE_OP_BACKGROUND). The latter case is not handled correctly by our decoder: It only performs resetting the rectangle in the reference frame when decoding the next frame; and since commit b593abda6c642cb0c3959752dd235c2faf66837f it does not reset the reference frame permanently, but only temporarily (i.e. it only affects decoding the frame after the frame with APNG_DISPOSE_OP_BACKGROUND). This is a problem if the frame after the APNG_DISPOSE_OP_BACKGROUND frame uses APNG_DISPOSE_OP_PREVIOUS, because then the frame after the APNG_DISPOSE_OP_PREVIOUS frame has an incorrect reference frame. (If it is not followed by an APNG_DISPOSE_OP_PREVIOUS frame, the decoder only keeps a reference to the output frame, which is ok.) This commit fixes this by being much closer to the spec than the earlier code: Resetting the background is no longer postponed until the next frame; instead it is applied to the reference frame. Fixes ticket #9602. (For multithreaded decoding it was actually already broken since commit 5663301560d77486c7f7c03c1aa5f542fab23c24.) Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2 years ago
if (s->dispose_op == APNG_DISPOSE_OP_PREVIOUS) {
/* We only need a buffer for the current picture. */
ret = ff_thread_get_buffer(avctx, p, 0);
if (ret < 0)
return ret;
} else if (s->dispose_op == APNG_DISPOSE_OP_BACKGROUND) {
/* We need a buffer for the current picture as well as
* a buffer for the reference to retain. */
ret = ff_thread_get_ext_buffer(avctx, &s->picture,
AV_GET_BUFFER_FLAG_REF);
if (ret < 0)
return ret;
ret = ff_thread_get_buffer(avctx, p, 0);
if (ret < 0)
return ret;
} else {
/* The picture output this time and the reference to retain coincide. */
if ((ret = ff_thread_get_ext_buffer(avctx, &s->picture,
AV_GET_BUFFER_FLAG_REF)) < 0)
return ret;
ret = av_frame_ref(p, s->picture.f);
if (ret < 0)
return ret;
}
p->pict_type = AV_PICTURE_TYPE_I;
p->key_frame = 1;
p->interlaced_frame = !!s->interlace_type;
ff_thread_finish_setup(avctx);
/* compute the compressed row size */
if (!s->interlace_type) {
s->crow_size = s->row_size + 1;
} else {
s->pass = 0;
s->pass_row_size = ff_png_pass_row_size(s->pass,
s->bits_per_pixel,
s->cur_w);
s->crow_size = s->pass_row_size + 1;
}
ff_dlog(avctx, "row_size=%d crow_size =%d\n",
s->row_size, s->crow_size);
/* copy the palette if needed */
if (avctx->pix_fmt == AV_PIX_FMT_PAL8)
memcpy(p->data[1], s->palette, 256 * sizeof(uint32_t));
/* empty row is used if differencing to the first row */
av_fast_padded_mallocz(&s->last_row, &s->last_row_size, s->row_size);
if (!s->last_row)
return AVERROR_INVALIDDATA;
if (s->interlace_type ||
s->color_type == PNG_COLOR_TYPE_RGB_ALPHA) {
av_fast_padded_malloc(&s->tmp_row, &s->tmp_row_size, s->row_size);
if (!s->tmp_row)
return AVERROR_INVALIDDATA;
}
/* compressed row */
av_fast_padded_malloc(&s->buffer, &s->buffer_size, s->row_size + 16);
if (!s->buffer)
return AVERROR(ENOMEM);
/* we want crow_buf+1 to be 16-byte aligned */
s->crow_buf = s->buffer + 15;
s->zstream.zstream.avail_out = s->crow_size;
s->zstream.zstream.next_out = s->crow_buf;
}
s->pic_state |= PNG_IDAT;
/* set image to non-transparent bpp while decompressing */
if (s->has_trns && s->color_type != PNG_COLOR_TYPE_PALETTE)
s->bpp -= byte_depth;
ret = png_decode_idat(s, gb, p->data[0], p->linesize[0]);
if (s->has_trns && s->color_type != PNG_COLOR_TYPE_PALETTE)
s->bpp += byte_depth;
if (ret < 0)
return ret;
return 0;
}
static int decode_plte_chunk(AVCodecContext *avctx, PNGDecContext *s,
GetByteContext *gb)
{
int length = bytestream2_get_bytes_left(gb);
int n, i, r, g, b;
if ((length % 3) != 0 || length > 256 * 3)
return AVERROR_INVALIDDATA;
/* read the palette */
n = length / 3;
for (i = 0; i < n; i++) {
r = bytestream2_get_byte(gb);
g = bytestream2_get_byte(gb);
b = bytestream2_get_byte(gb);
s->palette[i] = (0xFFU << 24) | (r << 16) | (g << 8) | b;
}
for (; i < 256; i++)
s->palette[i] = (0xFFU << 24);
s->hdr_state |= PNG_PLTE;
return 0;
}
static int decode_trns_chunk(AVCodecContext *avctx, PNGDecContext *s,
GetByteContext *gb)
{
int length = bytestream2_get_bytes_left(gb);
int v, i;
if (!(s->hdr_state & PNG_IHDR)) {
av_log(avctx, AV_LOG_ERROR, "trns before IHDR\n");
return AVERROR_INVALIDDATA;
}
if (s->pic_state & PNG_IDAT) {
av_log(avctx, AV_LOG_ERROR, "trns after IDAT\n");
return AVERROR_INVALIDDATA;
}
if (s->color_type == PNG_COLOR_TYPE_PALETTE) {
if (length > 256 || !(s->hdr_state & PNG_PLTE))
return AVERROR_INVALIDDATA;
for (i = 0; i < length; i++) {
unsigned v = bytestream2_get_byte(gb);
s->palette[i] = (s->palette[i] & 0x00ffffff) | (v << 24);
}
} else if (s->color_type == PNG_COLOR_TYPE_GRAY || s->color_type == PNG_COLOR_TYPE_RGB) {
if ((s->color_type == PNG_COLOR_TYPE_GRAY && length != 2) ||
(s->color_type == PNG_COLOR_TYPE_RGB && length != 6) ||
s->bit_depth == 1)
return AVERROR_INVALIDDATA;
for (i = 0; i < length / 2; i++) {
/* only use the least significant bits */
v = av_mod_uintp2(bytestream2_get_be16(gb), s->bit_depth);
if (s->bit_depth > 8)
AV_WB16(&s->transparent_color_be[2 * i], v);
else
s->transparent_color_be[i] = v;
}
} else {
return AVERROR_INVALIDDATA;
}
s->has_trns = 1;
return 0;
}
static int decode_iccp_chunk(PNGDecContext *s, GetByteContext *gb, AVFrame *f)
{
int ret, cnt = 0;
AVBPrint bp;
while ((s->iccp_name[cnt++] = bytestream2_get_byte(gb)) && cnt < 81);
if (cnt > 80) {
av_log(s->avctx, AV_LOG_ERROR, "iCCP with invalid name!\n");
ret = AVERROR_INVALIDDATA;
goto fail;
}
if (bytestream2_get_byte(gb) != 0) {
av_log(s->avctx, AV_LOG_ERROR, "iCCP with invalid compression!\n");
ret = AVERROR_INVALIDDATA;
goto fail;
}
if ((ret = decode_zbuf(&bp, gb->buffer, gb->buffer_end, s->avctx)) < 0)
return ret;
av_freep(&s->iccp_data);
ret = av_bprint_finalize(&bp, (char **)&s->iccp_data);
if (ret < 0)
return ret;
s->iccp_data_len = bp.len;
return 0;
fail:
s->iccp_name[0] = 0;
return ret;
}
static void handle_small_bpp(PNGDecContext *s, AVFrame *p)
{
if (s->bits_per_pixel == 1 && s->color_type == PNG_COLOR_TYPE_PALETTE) {
int i, j, k;
uint8_t *pd = p->data[0];
for (j = 0; j < s->height; j++) {
i = s->width / 8;
for (k = 7; k >= 1; k--)
if ((s->width&7) >= k)
pd[8*i + k - 1] = (pd[i]>>8-k) & 1;
for (i--; i >= 0; i--) {
pd[8*i + 7]= pd[i] & 1;
pd[8*i + 6]= (pd[i]>>1) & 1;
pd[8*i + 5]= (pd[i]>>2) & 1;
pd[8*i + 4]= (pd[i]>>3) & 1;
pd[8*i + 3]= (pd[i]>>4) & 1;
pd[8*i + 2]= (pd[i]>>5) & 1;
pd[8*i + 1]= (pd[i]>>6) & 1;
pd[8*i + 0]= pd[i]>>7;
}
pd += p->linesize[0];
}
} else if (s->bits_per_pixel == 2) {
int i, j;
uint8_t *pd = p->data[0];
for (j = 0; j < s->height; j++) {
i = s->width / 4;
if (s->color_type == PNG_COLOR_TYPE_PALETTE) {
if ((s->width&3) >= 3) pd[4*i + 2]= (pd[i] >> 2) & 3;
if ((s->width&3) >= 2) pd[4*i + 1]= (pd[i] >> 4) & 3;
if ((s->width&3) >= 1) pd[4*i + 0]= pd[i] >> 6;
for (i--; i >= 0; i--) {
pd[4*i + 3]= pd[i] & 3;
pd[4*i + 2]= (pd[i]>>2) & 3;
pd[4*i + 1]= (pd[i]>>4) & 3;
pd[4*i + 0]= pd[i]>>6;
}
} else {
if ((s->width&3) >= 3) pd[4*i + 2]= ((pd[i]>>2) & 3)*0x55;
if ((s->width&3) >= 2) pd[4*i + 1]= ((pd[i]>>4) & 3)*0x55;
if ((s->width&3) >= 1) pd[4*i + 0]= ( pd[i]>>6 )*0x55;
for (i--; i >= 0; i--) {
pd[4*i + 3]= ( pd[i] & 3)*0x55;
pd[4*i + 2]= ((pd[i]>>2) & 3)*0x55;
pd[4*i + 1]= ((pd[i]>>4) & 3)*0x55;
pd[4*i + 0]= ( pd[i]>>6 )*0x55;
}
}
pd += p->linesize[0];
}
} else if (s->bits_per_pixel == 4) {
int i, j;
uint8_t *pd = p->data[0];
for (j = 0; j < s->height; j++) {
i = s->width/2;
if (s->color_type == PNG_COLOR_TYPE_PALETTE) {
if (s->width&1) pd[2*i+0]= pd[i]>>4;
for (i--; i >= 0; i--) {
pd[2*i + 1] = pd[i] & 15;
pd[2*i + 0] = pd[i] >> 4;
}
} else {
if (s->width & 1) pd[2*i + 0]= (pd[i] >> 4) * 0x11;
for (i--; i >= 0; i--) {
pd[2*i + 1] = (pd[i] & 15) * 0x11;
pd[2*i + 0] = (pd[i] >> 4) * 0x11;
}
}
pd += p->linesize[0];
}
}
}
static int decode_fctl_chunk(AVCodecContext *avctx, PNGDecContext *s,
GetByteContext *gb)
{
uint32_t sequence_number;
int cur_w, cur_h, x_offset, y_offset, dispose_op, blend_op;
if (bytestream2_get_bytes_left(gb) != APNG_FCTL_CHUNK_SIZE)
return AVERROR_INVALIDDATA;
if (!(s->hdr_state & PNG_IHDR)) {
av_log(avctx, AV_LOG_ERROR, "fctl before IHDR\n");
return AVERROR_INVALIDDATA;
}
if (s->pic_state & PNG_IDAT) {
av_log(avctx, AV_LOG_ERROR, "fctl after IDAT\n");
return AVERROR_INVALIDDATA;
}
sequence_number = bytestream2_get_be32(gb);
cur_w = bytestream2_get_be32(gb);
cur_h = bytestream2_get_be32(gb);
x_offset = bytestream2_get_be32(gb);
y_offset = bytestream2_get_be32(gb);
bytestream2_skip(gb, 4); /* delay_num (2), delay_den (2) */
dispose_op = bytestream2_get_byte(gb);
blend_op = bytestream2_get_byte(gb);
if (sequence_number == 0 &&
(cur_w != s->width ||
cur_h != s->height ||
x_offset != 0 ||
y_offset != 0) ||
cur_w <= 0 || cur_h <= 0 ||
x_offset < 0 || y_offset < 0 ||
cur_w > s->width - x_offset|| cur_h > s->height - y_offset)
return AVERROR_INVALIDDATA;
if (blend_op != APNG_BLEND_OP_OVER && blend_op != APNG_BLEND_OP_SOURCE) {
av_log(avctx, AV_LOG_ERROR, "Invalid blend_op %d\n", blend_op);
return AVERROR_INVALIDDATA;
}
if ((sequence_number == 0 || !s->last_picture.f->data[0]) &&
dispose_op == APNG_DISPOSE_OP_PREVIOUS) {
// No previous frame to revert to for the first frame
// Spec says to just treat it as a APNG_DISPOSE_OP_BACKGROUND
dispose_op = APNG_DISPOSE_OP_BACKGROUND;
}
if (blend_op == APNG_BLEND_OP_OVER && !s->has_trns && (
avctx->pix_fmt == AV_PIX_FMT_RGB24 ||
avctx->pix_fmt == AV_PIX_FMT_RGB48BE ||
avctx->pix_fmt == AV_PIX_FMT_GRAY8 ||
avctx->pix_fmt == AV_PIX_FMT_GRAY16BE ||
avctx->pix_fmt == AV_PIX_FMT_MONOBLACK
)) {
// APNG_BLEND_OP_OVER is the same as APNG_BLEND_OP_SOURCE when there is no alpha channel
blend_op = APNG_BLEND_OP_SOURCE;
}
s->cur_w = cur_w;
s->cur_h = cur_h;
s->x_offset = x_offset;
s->y_offset = y_offset;
s->dispose_op = dispose_op;
s->blend_op = blend_op;
return 0;
}
static void handle_p_frame_png(PNGDecContext *s, AVFrame *p)
{
int i, j;
uint8_t *pd = p->data[0];
uint8_t *pd_last = s->last_picture.f->data[0];
int ls = av_image_get_linesize(p->format, s->width, 0);
ls = FFMIN(ls, s->width * s->bpp);
ff_thread_await_progress(&s->last_picture, INT_MAX, 0);
for (j = 0; j < s->height; j++) {
for (i = 0; i < ls; i++)
pd[i] += pd_last[i];
pd += p->linesize[0];
pd_last += s->last_picture.f->linesize[0];
}
}
// divide by 255 and round to nearest
// apply a fast variant: (X+127)/255 = ((X+127)*257+257)>>16 = ((X+128)*257)>>16
#define FAST_DIV255(x) ((((x) + 128) * 257) >> 16)
static int handle_p_frame_apng(AVCodecContext *avctx, PNGDecContext *s,
AVFrame *p)
{
uint8_t *dst = p->data[0];
ptrdiff_t dst_stride = p->linesize[0];
const uint8_t *src = s->last_picture.f->data[0];
ptrdiff_t src_stride = s->last_picture.f->linesize[0];
const int bpp = s->color_type == PNG_COLOR_TYPE_PALETTE ? 4 : s->bpp;
size_t x, y;
if (s->blend_op == APNG_BLEND_OP_OVER &&
avctx->pix_fmt != AV_PIX_FMT_RGBA &&
avctx->pix_fmt != AV_PIX_FMT_GRAY8A) {
avpriv_request_sample(avctx, "Blending with pixel format %s",
av_get_pix_fmt_name(avctx->pix_fmt));
return AVERROR_PATCHWELCOME;
}
ff_thread_await_progress(&s->last_picture, INT_MAX, 0);
// copy unchanged rectangles from the last frame
for (y = 0; y < s->y_offset; y++)
memcpy(dst + y * dst_stride, src + y * src_stride, p->width * bpp);
for (y = s->y_offset; y < s->y_offset + s->cur_h; y++) {
memcpy(dst + y * dst_stride, src + y * src_stride, s->x_offset * bpp);
memcpy(dst + y * dst_stride + (s->x_offset + s->cur_w) * bpp,
src + y * src_stride + (s->x_offset + s->cur_w) * bpp,
(p->width - s->cur_w - s->x_offset) * bpp);
}
for (y = s->y_offset + s->cur_h; y < p->height; y++)
memcpy(dst + y * dst_stride, src + y * src_stride, p->width * bpp);
if (s->blend_op == APNG_BLEND_OP_OVER) {
// Perform blending
for (y = s->y_offset; y < s->y_offset + s->cur_h; ++y) {
uint8_t *foreground = dst + dst_stride * y + bpp * s->x_offset;
const uint8_t *background = src + src_stride * y + bpp * s->x_offset;
for (x = s->x_offset; x < s->x_offset + s->cur_w; ++x, foreground += bpp, background += bpp) {
size_t b;
uint8_t foreground_alpha, background_alpha, output_alpha;
uint8_t output[10];
// Since we might be blending alpha onto alpha, we use the following equations:
// output_alpha = foreground_alpha + (1 - foreground_alpha) * background_alpha
// output = (foreground_alpha * foreground + (1 - foreground_alpha) * background_alpha * background) / output_alpha
switch (avctx->pix_fmt) {
case AV_PIX_FMT_RGBA:
foreground_alpha = foreground[3];
background_alpha = background[3];
break;
case AV_PIX_FMT_GRAY8A:
foreground_alpha = foreground[1];
background_alpha = background[1];
break;
}
if (foreground_alpha == 255)
continue;
if (foreground_alpha == 0) {
memcpy(foreground, background, bpp);
continue;
}
output_alpha = foreground_alpha + FAST_DIV255((255 - foreground_alpha) * background_alpha);
av_assert0(bpp <= 10);
for (b = 0; b < bpp - 1; ++b) {
if (output_alpha == 0) {
output[b] = 0;
} else if (background_alpha == 255) {
output[b] = FAST_DIV255(foreground_alpha * foreground[b] + (255 - foreground_alpha) * background[b]);
} else {
output[b] = (255 * foreground_alpha * foreground[b] + (255 - foreground_alpha) * background_alpha * background[b]) / (255 * output_alpha);
}
}
output[b] = output_alpha;
memcpy(foreground, output, bpp);
}
}
}
return 0;
}
avcodec/pngdec: Fix APNG_DISPOSE_OP_BACKGROUND APNG works with a single reference frame and an output frame. According to the spec, decoding APNG works by decoding the current IDAT/fdAT chunks (which decodes to a rectangular subregion of the whole image region), followed by either overwriting the region of the output frame with the newly decoded data or by blending the newly decoded data with the data from the reference frame onto the current subregion of the output frame. The remainder of the output frame is just copied from the reference frame. Then the reference frame might be left untouched (APNG_DISPOSE_OP_PREVIOUS), it might be replaced by the output frame (APNG_DISPOSE_OP_NONE) or the rectangular subregion corresponding to the just decoded frame has to be reset to black (APNG_DISPOSE_OP_BACKGROUND). The latter case is not handled correctly by our decoder: It only performs resetting the rectangle in the reference frame when decoding the next frame; and since commit b593abda6c642cb0c3959752dd235c2faf66837f it does not reset the reference frame permanently, but only temporarily (i.e. it only affects decoding the frame after the frame with APNG_DISPOSE_OP_BACKGROUND). This is a problem if the frame after the APNG_DISPOSE_OP_BACKGROUND frame uses APNG_DISPOSE_OP_PREVIOUS, because then the frame after the APNG_DISPOSE_OP_PREVIOUS frame has an incorrect reference frame. (If it is not followed by an APNG_DISPOSE_OP_PREVIOUS frame, the decoder only keeps a reference to the output frame, which is ok.) This commit fixes this by being much closer to the spec than the earlier code: Resetting the background is no longer postponed until the next frame; instead it is applied to the reference frame. Fixes ticket #9602. (For multithreaded decoding it was actually already broken since commit 5663301560d77486c7f7c03c1aa5f542fab23c24.) Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2 years ago
static void apng_reset_background(PNGDecContext *s, const AVFrame *p)
{
// need to reset a rectangle to black
av_unused int ret = av_frame_copy(s->picture.f, p);
const int bpp = s->color_type == PNG_COLOR_TYPE_PALETTE ? 4 : s->bpp;
const ptrdiff_t dst_stride = s->picture.f->linesize[0];
uint8_t *dst = s->picture.f->data[0] + s->y_offset * dst_stride + bpp * s->x_offset;
av_assert1(ret >= 0);
for (size_t y = 0; y < s->cur_h; y++) {
memset(dst, 0, bpp * s->cur_w);
dst += dst_stride;
}
}
static int decode_frame_common(AVCodecContext *avctx, PNGDecContext *s,
AVFrame *p, const AVPacket *avpkt)
{
const AVCRC *crc_tab = av_crc_get_table(AV_CRC_32_IEEE_LE);
uint32_t tag, length;
int decode_next_dat = 0;
int i, ret;
for (;;) {
GetByteContext gb_chunk;
length = bytestream2_get_bytes_left(&s->gb);
if (length <= 0) {
if (avctx->codec_id == AV_CODEC_ID_PNG &&
avctx->skip_frame == AVDISCARD_ALL) {
return 0;
}
if (CONFIG_APNG_DECODER && avctx->codec_id == AV_CODEC_ID_APNG && length == 0) {
if (!(s->pic_state & PNG_IDAT))
return 0;
else
goto exit_loop;
}
av_log(avctx, AV_LOG_ERROR, "%d bytes left\n", length);
if ( s->pic_state & PNG_ALLIMAGE
&& avctx->strict_std_compliance <= FF_COMPLIANCE_NORMAL)
goto exit_loop;
ret = AVERROR_INVALIDDATA;
goto fail;
}
length = bytestream2_get_be32(&s->gb);
if (length > 0x7fffffff || length + 8 > bytestream2_get_bytes_left(&s->gb)) {
av_log(avctx, AV_LOG_ERROR, "chunk too big\n");
ret = AVERROR_INVALIDDATA;
goto fail;
}
if (avctx->err_recognition & (AV_EF_CRCCHECK | AV_EF_IGNORE_ERR)) {
uint32_t crc_sig = AV_RB32(s->gb.buffer + length + 4);
uint32_t crc_cal = ~av_crc(crc_tab, UINT32_MAX, s->gb.buffer, length + 4);
if (crc_sig ^ crc_cal) {
av_log(avctx, AV_LOG_ERROR, "CRC mismatch in chunk");
if (avctx->err_recognition & AV_EF_EXPLODE) {
av_log(avctx, AV_LOG_ERROR, ", quitting\n");
ret = AVERROR_INVALIDDATA;
goto fail;
}
av_log(avctx, AV_LOG_ERROR, ", skipping\n");
bytestream2_skip(&s->gb, length + 8); /* tag */
}
}
tag = bytestream2_get_le32(&s->gb);
if (avctx->debug & FF_DEBUG_STARTCODE)
av_log(avctx, AV_LOG_DEBUG, "png: tag=%s length=%u\n",
av_fourcc2str(tag), length);
bytestream2_init(&gb_chunk, s->gb.buffer, length);
bytestream2_skip(&s->gb, length + 4);
if (avctx->codec_id == AV_CODEC_ID_PNG &&
avctx->skip_frame == AVDISCARD_ALL) {
switch(tag) {
case MKTAG('I', 'H', 'D', 'R'):
case MKTAG('p', 'H', 'Y', 's'):
case MKTAG('t', 'E', 'X', 't'):
case MKTAG('I', 'D', 'A', 'T'):
case MKTAG('t', 'R', 'N', 'S'):
break;
default:
continue;
}
}
switch (tag) {
case MKTAG('I', 'H', 'D', 'R'):
if ((ret = decode_ihdr_chunk(avctx, s, &gb_chunk)) < 0)
goto fail;
break;
case MKTAG('p', 'H', 'Y', 's'):
if ((ret = decode_phys_chunk(avctx, s, &gb_chunk)) < 0)
goto fail;
break;
case MKTAG('f', 'c', 'T', 'L'):
if (!CONFIG_APNG_DECODER || avctx->codec_id != AV_CODEC_ID_APNG)
continue;
if ((ret = decode_fctl_chunk(avctx, s, &gb_chunk)) < 0)
goto fail;
decode_next_dat = 1;
break;
case MKTAG('f', 'd', 'A', 'T'):
if (!CONFIG_APNG_DECODER || avctx->codec_id != AV_CODEC_ID_APNG)
continue;
if (!decode_next_dat || bytestream2_get_bytes_left(&gb_chunk) < 4) {
ret = AVERROR_INVALIDDATA;
goto fail;
}
bytestream2_get_be32(&gb_chunk);
/* fallthrough */
case MKTAG('I', 'D', 'A', 'T'):
if (CONFIG_APNG_DECODER && avctx->codec_id == AV_CODEC_ID_APNG && !decode_next_dat)
continue;
if ((ret = decode_idat_chunk(avctx, s, &gb_chunk, p)) < 0)
goto fail;
break;
case MKTAG('P', 'L', 'T', 'E'):
decode_plte_chunk(avctx, s, &gb_chunk);
break;
case MKTAG('t', 'R', 'N', 'S'):
decode_trns_chunk(avctx, s, &gb_chunk);
break;
case MKTAG('t', 'E', 'X', 't'):
if (decode_text_chunk(s, &gb_chunk, 0) < 0)
av_log(avctx, AV_LOG_WARNING, "Broken tEXt chunk\n");
break;
case MKTAG('z', 'T', 'X', 't'):
if (decode_text_chunk(s, &gb_chunk, 1) < 0)
av_log(avctx, AV_LOG_WARNING, "Broken zTXt chunk\n");
break;
case MKTAG('s', 'T', 'E', 'R'): {
int mode = bytestream2_get_byte(&gb_chunk);
if (mode == 0 || mode == 1) {
s->stereo_mode = mode;
} else {
av_log(avctx, AV_LOG_WARNING,
"Unknown value in sTER chunk (%d)\n", mode);
}
break;
}
case MKTAG('i', 'C', 'C', 'P'): {
if ((ret = decode_iccp_chunk(s, &gb_chunk, p)) < 0)
goto fail;
break;
}
case MKTAG('c', 'H', 'R', 'M'): {
s->have_chrm = 1;
s->white_point[0] = bytestream2_get_be32(&gb_chunk);
s->white_point[1] = bytestream2_get_be32(&gb_chunk);
/* RGB Primaries */
for (i = 0; i < 3; i++) {
s->display_primaries[i][0] = bytestream2_get_be32(&gb_chunk);
s->display_primaries[i][1] = bytestream2_get_be32(&gb_chunk);
}
break;
}
case MKTAG('g', 'A', 'M', 'A'): {
AVBPrint bp;
char *gamma_str;
int num = bytestream2_get_be32(&gb_chunk);
av_bprint_init(&bp, 0, AV_BPRINT_SIZE_UNLIMITED);
av_bprintf(&bp, "%i/%i", num, 100000);
ret = av_bprint_finalize(&bp, &gamma_str);
if (ret < 0)
return ret;
av_dict_set(&s->frame_metadata, "gamma", gamma_str, AV_DICT_DONT_STRDUP_VAL);
break;
}
case MKTAG('I', 'E', 'N', 'D'):
if (!(s->pic_state & PNG_ALLIMAGE))
av_log(avctx, AV_LOG_ERROR, "IEND without all image\n");
if (!(s->pic_state & (PNG_ALLIMAGE|PNG_IDAT))) {
ret = AVERROR_INVALIDDATA;
goto fail;
}
goto exit_loop;
}
}
exit_loop:
if (avctx->codec_id == AV_CODEC_ID_PNG &&
avctx->skip_frame == AVDISCARD_ALL) {
return 0;
}
if (percent_missing(s) > avctx->discard_damaged_percentage)
return AVERROR_INVALIDDATA;
if (s->bits_per_pixel <= 4)
handle_small_bpp(s, p);
if (s->color_type == PNG_COLOR_TYPE_PALETTE && avctx->codec_id == AV_CODEC_ID_APNG) {
for (int y = 0; y < s->height; y++) {
uint8_t *row = &p->data[0][p->linesize[0] * y];
for (int x = s->width - 1; x >= 0; x--) {
const uint8_t idx = row[x];
row[4*x+2] = s->palette[idx] & 0xFF;
row[4*x+1] = (s->palette[idx] >> 8 ) & 0xFF;
row[4*x+0] = (s->palette[idx] >> 16) & 0xFF;
row[4*x+3] = s->palette[idx] >> 24;
}
}
}
/* apply transparency if needed */
if (s->has_trns && s->color_type != PNG_COLOR_TYPE_PALETTE) {
size_t byte_depth = s->bit_depth > 8 ? 2 : 1;
size_t raw_bpp = s->bpp - byte_depth;
unsigned x, y;
av_assert0(s->bit_depth > 1);
for (y = 0; y < s->height; ++y) {
uint8_t *row = &p->data[0][p->linesize[0] * y];
if (s->bpp == 2 && byte_depth == 1) {
uint8_t *pixel = &row[2 * s->width - 1];
uint8_t *rowp = &row[1 * s->width - 1];
int tcolor = s->transparent_color_be[0];
for (x = s->width; x > 0; --x) {
*pixel-- = *rowp == tcolor ? 0 : 0xff;
*pixel-- = *rowp--;
}
} else if (s->bpp == 4 && byte_depth == 1) {
uint8_t *pixel = &row[4 * s->width - 1];
uint8_t *rowp = &row[3 * s->width - 1];
int tcolor = AV_RL24(s->transparent_color_be);
for (x = s->width; x > 0; --x) {
*pixel-- = AV_RL24(rowp-2) == tcolor ? 0 : 0xff;
*pixel-- = *rowp--;
*pixel-- = *rowp--;
*pixel-- = *rowp--;
}
} else {
/* since we're updating in-place, we have to go from right to left */
for (x = s->width; x > 0; --x) {
uint8_t *pixel = &row[s->bpp * (x - 1)];
memmove(pixel, &row[raw_bpp * (x - 1)], raw_bpp);
if (!memcmp(pixel, s->transparent_color_be, raw_bpp)) {
memset(&pixel[raw_bpp], 0, byte_depth);
} else {
memset(&pixel[raw_bpp], 0xff, byte_depth);
}
}
}
}
}
/* handle P-frames only if a predecessor frame is available */
if (s->last_picture.f->data[0]) {
if ( !(avpkt->flags & AV_PKT_FLAG_KEY) && avctx->codec_tag != AV_RL32("MPNG")
&& s->last_picture.f->width == p->width
&& s->last_picture.f->height== p->height
&& s->last_picture.f->format== p->format
) {
if (CONFIG_PNG_DECODER && avctx->codec_id != AV_CODEC_ID_APNG)
handle_p_frame_png(s, p);
else if (CONFIG_APNG_DECODER &&
avctx->codec_id == AV_CODEC_ID_APNG &&
(ret = handle_p_frame_apng(avctx, s, p)) < 0)
goto fail;
}
}
avcodec/pngdec: Fix APNG_DISPOSE_OP_BACKGROUND APNG works with a single reference frame and an output frame. According to the spec, decoding APNG works by decoding the current IDAT/fdAT chunks (which decodes to a rectangular subregion of the whole image region), followed by either overwriting the region of the output frame with the newly decoded data or by blending the newly decoded data with the data from the reference frame onto the current subregion of the output frame. The remainder of the output frame is just copied from the reference frame. Then the reference frame might be left untouched (APNG_DISPOSE_OP_PREVIOUS), it might be replaced by the output frame (APNG_DISPOSE_OP_NONE) or the rectangular subregion corresponding to the just decoded frame has to be reset to black (APNG_DISPOSE_OP_BACKGROUND). The latter case is not handled correctly by our decoder: It only performs resetting the rectangle in the reference frame when decoding the next frame; and since commit b593abda6c642cb0c3959752dd235c2faf66837f it does not reset the reference frame permanently, but only temporarily (i.e. it only affects decoding the frame after the frame with APNG_DISPOSE_OP_BACKGROUND). This is a problem if the frame after the APNG_DISPOSE_OP_BACKGROUND frame uses APNG_DISPOSE_OP_PREVIOUS, because then the frame after the APNG_DISPOSE_OP_PREVIOUS frame has an incorrect reference frame. (If it is not followed by an APNG_DISPOSE_OP_PREVIOUS frame, the decoder only keeps a reference to the output frame, which is ok.) This commit fixes this by being much closer to the spec than the earlier code: Resetting the background is no longer postponed until the next frame; instead it is applied to the reference frame. Fixes ticket #9602. (For multithreaded decoding it was actually already broken since commit 5663301560d77486c7f7c03c1aa5f542fab23c24.) Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2 years ago
if (CONFIG_APNG_DECODER && s->dispose_op == APNG_DISPOSE_OP_BACKGROUND)
apng_reset_background(s, p);
ff_thread_report_progress(&s->picture, INT_MAX, 0);
return 0;
fail:
ff_thread_report_progress(&s->picture, INT_MAX, 0);
return ret;
}
static void clear_frame_metadata(PNGDecContext *s)
{
av_freep(&s->iccp_data);
s->iccp_data_len = 0;
s->iccp_name[0] = 0;
s->stereo_mode = -1;
s->have_chrm = 0;
av_dict_free(&s->frame_metadata);
}
avcodec/pngdec: Fix APNG_DISPOSE_OP_BACKGROUND APNG works with a single reference frame and an output frame. According to the spec, decoding APNG works by decoding the current IDAT/fdAT chunks (which decodes to a rectangular subregion of the whole image region), followed by either overwriting the region of the output frame with the newly decoded data or by blending the newly decoded data with the data from the reference frame onto the current subregion of the output frame. The remainder of the output frame is just copied from the reference frame. Then the reference frame might be left untouched (APNG_DISPOSE_OP_PREVIOUS), it might be replaced by the output frame (APNG_DISPOSE_OP_NONE) or the rectangular subregion corresponding to the just decoded frame has to be reset to black (APNG_DISPOSE_OP_BACKGROUND). The latter case is not handled correctly by our decoder: It only performs resetting the rectangle in the reference frame when decoding the next frame; and since commit b593abda6c642cb0c3959752dd235c2faf66837f it does not reset the reference frame permanently, but only temporarily (i.e. it only affects decoding the frame after the frame with APNG_DISPOSE_OP_BACKGROUND). This is a problem if the frame after the APNG_DISPOSE_OP_BACKGROUND frame uses APNG_DISPOSE_OP_PREVIOUS, because then the frame after the APNG_DISPOSE_OP_PREVIOUS frame has an incorrect reference frame. (If it is not followed by an APNG_DISPOSE_OP_PREVIOUS frame, the decoder only keeps a reference to the output frame, which is ok.) This commit fixes this by being much closer to the spec than the earlier code: Resetting the background is no longer postponed until the next frame; instead it is applied to the reference frame. Fixes ticket #9602. (For multithreaded decoding it was actually already broken since commit 5663301560d77486c7f7c03c1aa5f542fab23c24.) Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2 years ago
static int output_frame(PNGDecContext *s, AVFrame *f)
{
int ret;
if (s->iccp_data) {
AVFrameSideData *sd = av_frame_new_side_data(f, AV_FRAME_DATA_ICC_PROFILE, s->iccp_data_len);
if (!sd) {
ret = AVERROR(ENOMEM);
goto fail;
}
memcpy(sd->data, s->iccp_data, s->iccp_data_len);
av_dict_set(&sd->metadata, "name", s->iccp_name, 0);
}
if (s->stereo_mode >= 0) {
AVStereo3D *stereo3d = av_stereo3d_create_side_data(f);
if (!stereo3d) {
ret = AVERROR(ENOMEM);
goto fail;
}
stereo3d->type = AV_STEREO3D_SIDEBYSIDE;
stereo3d->flags = s->stereo_mode ? 0 : AV_STEREO3D_FLAG_INVERT;
}
if (s->have_chrm) {
AVMasteringDisplayMetadata *mdm = av_mastering_display_metadata_create_side_data(f);
if (!mdm) {
ret = AVERROR(ENOMEM);
goto fail;
}
mdm->white_point[0] = av_make_q(s->white_point[0], 100000);
mdm->white_point[1] = av_make_q(s->white_point[1], 100000);
/* RGB Primaries */
for (int i = 0; i < 3; i++) {
mdm->display_primaries[i][0] = av_make_q(s->display_primaries[i][0], 100000);
mdm->display_primaries[i][1] = av_make_q(s->display_primaries[i][1], 100000);
}
mdm->has_primaries = 1;
}
FFSWAP(AVDictionary*, f->metadata, s->frame_metadata);
return 0;
fail:
av_frame_unref(f);
return ret;
}
#if CONFIG_PNG_DECODER
avcodec/pngdec: Fix APNG_DISPOSE_OP_BACKGROUND APNG works with a single reference frame and an output frame. According to the spec, decoding APNG works by decoding the current IDAT/fdAT chunks (which decodes to a rectangular subregion of the whole image region), followed by either overwriting the region of the output frame with the newly decoded data or by blending the newly decoded data with the data from the reference frame onto the current subregion of the output frame. The remainder of the output frame is just copied from the reference frame. Then the reference frame might be left untouched (APNG_DISPOSE_OP_PREVIOUS), it might be replaced by the output frame (APNG_DISPOSE_OP_NONE) or the rectangular subregion corresponding to the just decoded frame has to be reset to black (APNG_DISPOSE_OP_BACKGROUND). The latter case is not handled correctly by our decoder: It only performs resetting the rectangle in the reference frame when decoding the next frame; and since commit b593abda6c642cb0c3959752dd235c2faf66837f it does not reset the reference frame permanently, but only temporarily (i.e. it only affects decoding the frame after the frame with APNG_DISPOSE_OP_BACKGROUND). This is a problem if the frame after the APNG_DISPOSE_OP_BACKGROUND frame uses APNG_DISPOSE_OP_PREVIOUS, because then the frame after the APNG_DISPOSE_OP_PREVIOUS frame has an incorrect reference frame. (If it is not followed by an APNG_DISPOSE_OP_PREVIOUS frame, the decoder only keeps a reference to the output frame, which is ok.) This commit fixes this by being much closer to the spec than the earlier code: Resetting the background is no longer postponed until the next frame; instead it is applied to the reference frame. Fixes ticket #9602. (For multithreaded decoding it was actually already broken since commit 5663301560d77486c7f7c03c1aa5f542fab23c24.) Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2 years ago
static int decode_frame_png(AVCodecContext *avctx, AVFrame *p,
int *got_frame, AVPacket *avpkt)
{
PNGDecContext *const s = avctx->priv_data;
const uint8_t *buf = avpkt->data;
int buf_size = avpkt->size;
int64_t sig;
int ret;
clear_frame_metadata(s);
bytestream2_init(&s->gb, buf, buf_size);
/* check signature */
sig = bytestream2_get_be64(&s->gb);
if (sig != PNGSIG &&
sig != MNGSIG) {
av_log(avctx, AV_LOG_ERROR, "Invalid PNG signature 0x%08"PRIX64".\n", sig);
return AVERROR_INVALIDDATA;
}
s->y = s->has_trns = 0;
s->hdr_state = 0;
s->pic_state = 0;
/* Reset z_stream */
ret = inflateReset(&s->zstream.zstream);
if (ret != Z_OK)
return AVERROR_EXTERNAL;
if ((ret = decode_frame_common(avctx, s, p, avpkt)) < 0)
goto the_end;
Merge commit '759001c534287a96dc96d1e274665feb7059145d' * commit '759001c534287a96dc96d1e274665feb7059145d': lavc decoders: work with refcounted frames. Anton Khirnov (1): lavc decoders: work with refcounted frames. Clément Bœsch (47): lavc/ansi: reset file lavc/ansi: re-do refcounted frame changes from Anton fraps: reset file lavc/fraps: switch to refcounted frames gifdec: reset file lavc/gifdec: switch to refcounted frames dsicinav: resolve conflicts smc: resolve conflicts zmbv: resolve conflicts rpza: resolve conflicts vble: resolve conflicts xxan: resolve conflicts targa: resolve conflicts vmnc: resolve conflicts utvideodec: resolve conflicts tscc: resolve conflicts ulti: resolve conflicts ffv1dec: resolve conflicts dnxhddec: resolve conflicts v210dec: resolve conflicts vp3: resolve conflicts vcr1: resolve conflicts v210x: resolve conflicts wavpack: resolve conflicts pngdec: fix compilation roqvideodec: resolve conflicts pictordec: resolve conflicts mdec: resolve conflicts tiertexseqv: resolve conflicts smacker: resolve conflicts vb: resolve conflicts vqavideo: resolve conflicts xl: resolve conflicts tmv: resolve conflicts vmdav: resolve conflicts truemotion1: resolve conflicts truemotion2: resolve conflicts lcldec: fix compilation libcelt_dec: fix compilation qdrw: fix compilation r210dec: fix compilation rl2: fix compilation wnv1: fix compilation yop: fix compilation tiff: resolve conflicts interplayvideo: fix compilation qpeg: resolve conflicts (FIXME/TESTME). Hendrik Leppkes (33): 012v: convert to refcounted frames 8bps: fix compilation 8svx: resolve conflicts 4xm: resolve conflicts aasc: resolve conflicts bfi: fix compilation aura: fix compilation alsdec: resolve conflicts avrndec: convert to refcounted frames avuidec: convert to refcounted frames bintext: convert to refcounted frames cavsdec: resolve conflicts brender_pix: convert to refcounted frames cinepak: resolve conflicts cinepak: avoid using AVFrame struct directly in private context cljr: fix compilation cpia: convert to refcounted frames cscd: resolve conflicts iff: resolve conflicts and do proper conversion to refcounted frames 4xm: fix reference frame handling cyuv: fix compilation dxa: fix compilation eacmv: fix compilation eamad: fix compilation eatgv: fix compilation escape124: remove unused variable. escape130: convert to refcounted frames evrcdec: convert to refcounted frames exr: convert to refcounted frames mvcdec: convert to refcounted frames paf: properly free the frame data on decode close sgirle: convert to refcounted frames lavfi/moviesrc: use refcounted frames Michael Niedermayer (56): Merge commit '759001c534287a96dc96d1e274665feb7059145d' resolve conflicts in headers motion_est: resolve conflict mpeg4videodec: fix conflicts dpcm conflict fix dpx: fix conflicts indeo3: resolve confilcts kmvc: resolve conflicts kmvc: resolve conflicts h264: resolve conflicts utils: resolve conflicts rawdec: resolve conflcits mpegvideo: resolve conflicts svq1enc: resolve conflicts mpegvideo: dont clear data, fix assertion failure on fate vsynth1 with threads pthreads: resolve conflicts frame_thread_encoder: simple compilefix not yet tested snow: update to buffer refs crytsalhd: fix compile dirac: switch to new API sonic: update to new API svq1: resolve conflict, update to new API ffwavesynth: update to new buffer API g729: update to new API indeo5: fix compile j2kdec: update to new buffer API linopencore-amr: fix compile libvorbisdec: update to new API loco: fix compile paf: update to new API proresdec: update to new API vp56: update to new api / resolve conflicts xface: convert to refcounted frames xan: fix compile&fate v408: update to ref counted buffers v308: update to ref counted buffers yuv4dec: update to ref counted buffers y41p: update to ref counted frames xbm: update to refcounted frames targa_y216: update to refcounted buffers qpeg: fix fate/crash cdxl: fix fate tscc: fix reget buffer useage targa_y216dec: fix style msmpeg4: fix fate h264: ref_picture() copy fields that have been lost too update_frame_pool: use channel field h264: Put code that prevents deadlocks back mpegvideo: dont allow last == current wmalossless: fix buffer ref messup ff_alloc_picture: free tables in case of dimension mismatches h264: fix null pointer dereference and assertion failure frame_thread_encoder: update to bufrefs ec: fix used arrays snowdec: fix off by 1 error in dimensions check h264: disallow single unpaired fields as references of frames Paul B Mahol (2): lavc/vima: convert to refcounted frames sanm: convert to refcounted frames Conflicts: libavcodec/4xm.c libavcodec/8bps.c libavcodec/8svx.c libavcodec/aasc.c libavcodec/alsdec.c libavcodec/anm.c libavcodec/ansi.c libavcodec/avs.c libavcodec/bethsoftvideo.c libavcodec/bfi.c libavcodec/c93.c libavcodec/cavsdec.c libavcodec/cdgraphics.c libavcodec/cinepak.c libavcodec/cljr.c libavcodec/cscd.c libavcodec/dnxhddec.c libavcodec/dpcm.c libavcodec/dpx.c libavcodec/dsicinav.c libavcodec/dvdec.c libavcodec/dxa.c libavcodec/eacmv.c libavcodec/eamad.c libavcodec/eatgq.c libavcodec/eatgv.c libavcodec/eatqi.c libavcodec/error_resilience.c libavcodec/escape124.c libavcodec/ffv1.h libavcodec/ffv1dec.c libavcodec/flicvideo.c libavcodec/fraps.c libavcodec/frwu.c libavcodec/g723_1.c libavcodec/gifdec.c libavcodec/h264.c libavcodec/h264.h libavcodec/h264_direct.c libavcodec/h264_loopfilter.c libavcodec/h264_refs.c libavcodec/huffyuvdec.c libavcodec/idcinvideo.c libavcodec/iff.c libavcodec/indeo2.c libavcodec/indeo3.c libavcodec/internal.h libavcodec/interplayvideo.c libavcodec/ivi_common.c libavcodec/jvdec.c libavcodec/kgv1dec.c libavcodec/kmvc.c libavcodec/lagarith.c libavcodec/libopenjpegdec.c libavcodec/mdec.c libavcodec/mimic.c libavcodec/mjpegbdec.c libavcodec/mjpegdec.c libavcodec/mmvideo.c libavcodec/motion_est.c libavcodec/motionpixels.c libavcodec/mpc7.c libavcodec/mpeg12.c libavcodec/mpeg4videodec.c libavcodec/mpegvideo.c libavcodec/mpegvideo.h libavcodec/msrle.c libavcodec/msvideo1.c libavcodec/nuv.c libavcodec/options_table.h libavcodec/pcx.c libavcodec/pictordec.c libavcodec/pngdec.c libavcodec/pnmdec.c libavcodec/pthread.c libavcodec/qpeg.c libavcodec/qtrle.c libavcodec/r210dec.c libavcodec/rawdec.c libavcodec/roqvideodec.c libavcodec/rpza.c libavcodec/smacker.c libavcodec/smc.c libavcodec/svq1dec.c libavcodec/svq1enc.c libavcodec/targa.c libavcodec/tiertexseqv.c libavcodec/tiff.c libavcodec/tmv.c libavcodec/truemotion1.c libavcodec/truemotion2.c libavcodec/tscc.c libavcodec/ulti.c libavcodec/utils.c libavcodec/utvideodec.c libavcodec/v210dec.c libavcodec/v210x.c libavcodec/vb.c libavcodec/vble.c libavcodec/vcr1.c libavcodec/vmdav.c libavcodec/vmnc.c libavcodec/vp3.c libavcodec/vp56.c libavcodec/vp56.h libavcodec/vp6.c libavcodec/vqavideo.c libavcodec/wavpack.c libavcodec/xl.c libavcodec/xxan.c libavcodec/zmbv.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
12 years ago
if (avctx->skip_frame == AVDISCARD_ALL) {
*got_frame = 0;
ret = bytestream2_tell(&s->gb);
goto the_end;
}
avcodec/pngdec: Fix APNG_DISPOSE_OP_BACKGROUND APNG works with a single reference frame and an output frame. According to the spec, decoding APNG works by decoding the current IDAT/fdAT chunks (which decodes to a rectangular subregion of the whole image region), followed by either overwriting the region of the output frame with the newly decoded data or by blending the newly decoded data with the data from the reference frame onto the current subregion of the output frame. The remainder of the output frame is just copied from the reference frame. Then the reference frame might be left untouched (APNG_DISPOSE_OP_PREVIOUS), it might be replaced by the output frame (APNG_DISPOSE_OP_NONE) or the rectangular subregion corresponding to the just decoded frame has to be reset to black (APNG_DISPOSE_OP_BACKGROUND). The latter case is not handled correctly by our decoder: It only performs resetting the rectangle in the reference frame when decoding the next frame; and since commit b593abda6c642cb0c3959752dd235c2faf66837f it does not reset the reference frame permanently, but only temporarily (i.e. it only affects decoding the frame after the frame with APNG_DISPOSE_OP_BACKGROUND). This is a problem if the frame after the APNG_DISPOSE_OP_BACKGROUND frame uses APNG_DISPOSE_OP_PREVIOUS, because then the frame after the APNG_DISPOSE_OP_PREVIOUS frame has an incorrect reference frame. (If it is not followed by an APNG_DISPOSE_OP_PREVIOUS frame, the decoder only keeps a reference to the output frame, which is ok.) This commit fixes this by being much closer to the spec than the earlier code: Resetting the background is no longer postponed until the next frame; instead it is applied to the reference frame. Fixes ticket #9602. (For multithreaded decoding it was actually already broken since commit 5663301560d77486c7f7c03c1aa5f542fab23c24.) Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2 years ago
ret = output_frame(s, p);
if (ret < 0)
goto the_end;
if (!(avctx->active_thread_type & FF_THREAD_FRAME)) {
ff_thread_release_ext_buffer(avctx, &s->last_picture);
FFSWAP(ThreadFrame, s->picture, s->last_picture);
}
*got_frame = 1;
ret = bytestream2_tell(&s->gb);
the_end:
s->crow_buf = NULL;
return ret;
}
#endif
#if CONFIG_APNG_DECODER
avcodec/pngdec: Fix APNG_DISPOSE_OP_BACKGROUND APNG works with a single reference frame and an output frame. According to the spec, decoding APNG works by decoding the current IDAT/fdAT chunks (which decodes to a rectangular subregion of the whole image region), followed by either overwriting the region of the output frame with the newly decoded data or by blending the newly decoded data with the data from the reference frame onto the current subregion of the output frame. The remainder of the output frame is just copied from the reference frame. Then the reference frame might be left untouched (APNG_DISPOSE_OP_PREVIOUS), it might be replaced by the output frame (APNG_DISPOSE_OP_NONE) or the rectangular subregion corresponding to the just decoded frame has to be reset to black (APNG_DISPOSE_OP_BACKGROUND). The latter case is not handled correctly by our decoder: It only performs resetting the rectangle in the reference frame when decoding the next frame; and since commit b593abda6c642cb0c3959752dd235c2faf66837f it does not reset the reference frame permanently, but only temporarily (i.e. it only affects decoding the frame after the frame with APNG_DISPOSE_OP_BACKGROUND). This is a problem if the frame after the APNG_DISPOSE_OP_BACKGROUND frame uses APNG_DISPOSE_OP_PREVIOUS, because then the frame after the APNG_DISPOSE_OP_PREVIOUS frame has an incorrect reference frame. (If it is not followed by an APNG_DISPOSE_OP_PREVIOUS frame, the decoder only keeps a reference to the output frame, which is ok.) This commit fixes this by being much closer to the spec than the earlier code: Resetting the background is no longer postponed until the next frame; instead it is applied to the reference frame. Fixes ticket #9602. (For multithreaded decoding it was actually already broken since commit 5663301560d77486c7f7c03c1aa5f542fab23c24.) Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2 years ago
static int decode_frame_apng(AVCodecContext *avctx, AVFrame *p,
int *got_frame, AVPacket *avpkt)
{
PNGDecContext *const s = avctx->priv_data;
int ret;
clear_frame_metadata(s);
if (!(s->hdr_state & PNG_IHDR)) {
if (!avctx->extradata_size)
return AVERROR_INVALIDDATA;
if ((ret = inflateReset(&s->zstream.zstream)) != Z_OK)
return AVERROR_EXTERNAL;
bytestream2_init(&s->gb, avctx->extradata, avctx->extradata_size);
if ((ret = decode_frame_common(avctx, s, p, avpkt)) < 0)
return ret;
}
/* reset state for a new frame */
if ((ret = inflateReset(&s->zstream.zstream)) != Z_OK)
return AVERROR_EXTERNAL;
s->y = 0;
s->pic_state = 0;
bytestream2_init(&s->gb, avpkt->data, avpkt->size);
if ((ret = decode_frame_common(avctx, s, p, avpkt)) < 0)
return ret;
if (!(s->pic_state & PNG_ALLIMAGE))
av_log(avctx, AV_LOG_WARNING, "Frame did not contain a complete image\n");
if (!(s->pic_state & (PNG_ALLIMAGE|PNG_IDAT)))
return AVERROR_INVALIDDATA;
avcodec/pngdec: Fix APNG_DISPOSE_OP_BACKGROUND APNG works with a single reference frame and an output frame. According to the spec, decoding APNG works by decoding the current IDAT/fdAT chunks (which decodes to a rectangular subregion of the whole image region), followed by either overwriting the region of the output frame with the newly decoded data or by blending the newly decoded data with the data from the reference frame onto the current subregion of the output frame. The remainder of the output frame is just copied from the reference frame. Then the reference frame might be left untouched (APNG_DISPOSE_OP_PREVIOUS), it might be replaced by the output frame (APNG_DISPOSE_OP_NONE) or the rectangular subregion corresponding to the just decoded frame has to be reset to black (APNG_DISPOSE_OP_BACKGROUND). The latter case is not handled correctly by our decoder: It only performs resetting the rectangle in the reference frame when decoding the next frame; and since commit b593abda6c642cb0c3959752dd235c2faf66837f it does not reset the reference frame permanently, but only temporarily (i.e. it only affects decoding the frame after the frame with APNG_DISPOSE_OP_BACKGROUND). This is a problem if the frame after the APNG_DISPOSE_OP_BACKGROUND frame uses APNG_DISPOSE_OP_PREVIOUS, because then the frame after the APNG_DISPOSE_OP_PREVIOUS frame has an incorrect reference frame. (If it is not followed by an APNG_DISPOSE_OP_PREVIOUS frame, the decoder only keeps a reference to the output frame, which is ok.) This commit fixes this by being much closer to the spec than the earlier code: Resetting the background is no longer postponed until the next frame; instead it is applied to the reference frame. Fixes ticket #9602. (For multithreaded decoding it was actually already broken since commit 5663301560d77486c7f7c03c1aa5f542fab23c24.) Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2 years ago
ret = output_frame(s, p);
if (ret < 0)
return ret;
if (!(avctx->active_thread_type & FF_THREAD_FRAME)) {
if (s->dispose_op == APNG_DISPOSE_OP_PREVIOUS) {
ff_thread_release_ext_buffer(avctx, &s->picture);
} else {
ff_thread_release_ext_buffer(avctx, &s->last_picture);
FFSWAP(ThreadFrame, s->picture, s->last_picture);
}
}
*got_frame = 1;
return bytestream2_tell(&s->gb);
}
#endif
#if HAVE_THREADS
static int update_thread_context(AVCodecContext *dst, const AVCodecContext *src)
{
PNGDecContext *psrc = src->priv_data;
PNGDecContext *pdst = dst->priv_data;
ThreadFrame *src_frame = NULL;
int ret;
if (dst == src)
return 0;
if (CONFIG_APNG_DECODER && dst->codec_id == AV_CODEC_ID_APNG) {
pdst->width = psrc->width;
pdst->height = psrc->height;
pdst->bit_depth = psrc->bit_depth;
pdst->color_type = psrc->color_type;
pdst->compression_type = psrc->compression_type;
pdst->interlace_type = psrc->interlace_type;
pdst->filter_type = psrc->filter_type;
pdst->has_trns = psrc->has_trns;
memcpy(pdst->transparent_color_be, psrc->transparent_color_be, sizeof(pdst->transparent_color_be));
memcpy(pdst->palette, psrc->palette, sizeof(pdst->palette));
pdst->hdr_state |= psrc->hdr_state;
}
src_frame = psrc->dispose_op == APNG_DISPOSE_OP_PREVIOUS ?
&psrc->last_picture : &psrc->picture;
ff_thread_release_ext_buffer(dst, &pdst->last_picture);
if (src_frame && src_frame->f->data[0]) {
ret = ff_thread_ref_frame(&pdst->last_picture, src_frame);
if (ret < 0)
return ret;
}
return 0;
}
#endif
static av_cold int png_dec_init(AVCodecContext *avctx)
{
PNGDecContext *s = avctx->priv_data;
avctx->color_range = AVCOL_RANGE_JPEG;
s->avctx = avctx;
s->last_picture.f = av_frame_alloc();
s->picture.f = av_frame_alloc();
if (!s->last_picture.f || !s->picture.f)
return AVERROR(ENOMEM);
ff_pngdsp_init(&s->dsp);
return ff_inflate_init(&s->zstream, avctx);
}
static av_cold int png_dec_end(AVCodecContext *avctx)
{
PNGDecContext *s = avctx->priv_data;
ff_thread_release_ext_buffer(avctx, &s->last_picture);
av_frame_free(&s->last_picture.f);
ff_thread_release_ext_buffer(avctx, &s->picture);
av_frame_free(&s->picture.f);
av_freep(&s->buffer);
s->buffer_size = 0;
av_freep(&s->last_row);
s->last_row_size = 0;
av_freep(&s->tmp_row);
s->tmp_row_size = 0;
av_freep(&s->iccp_data);
av_dict_free(&s->frame_metadata);
ff_inflate_end(&s->zstream);
return 0;
}
#if CONFIG_APNG_DECODER
const FFCodec ff_apng_decoder = {
.p.name = "apng",
.p.long_name = NULL_IF_CONFIG_SMALL("APNG (Animated Portable Network Graphics) image"),
.p.type = AVMEDIA_TYPE_VIDEO,
.p.id = AV_CODEC_ID_APNG,
.priv_data_size = sizeof(PNGDecContext),
.init = png_dec_init,
.close = png_dec_end,
FF_CODEC_DECODE_CB(decode_frame_apng),
.update_thread_context = ONLY_IF_THREADS_ENABLED(update_thread_context),
.p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_FRAME_THREADS /*| AV_CODEC_CAP_DRAW_HORIZ_BAND*/,
.caps_internal = FF_CODEC_CAP_INIT_CLEANUP |
FF_CODEC_CAP_ALLOCATE_PROGRESS |
FF_CODEC_CAP_ICC_PROFILES,
};
#endif
#if CONFIG_PNG_DECODER
const FFCodec ff_png_decoder = {
.p.name = "png",
.p.long_name = NULL_IF_CONFIG_SMALL("PNG (Portable Network Graphics) image"),
.p.type = AVMEDIA_TYPE_VIDEO,
.p.id = AV_CODEC_ID_PNG,
.priv_data_size = sizeof(PNGDecContext),
.init = png_dec_init,
.close = png_dec_end,
FF_CODEC_DECODE_CB(decode_frame_png),
.update_thread_context = ONLY_IF_THREADS_ENABLED(update_thread_context),
.p.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_FRAME_THREADS /*| AV_CODEC_CAP_DRAW_HORIZ_BAND*/,
.caps_internal = FF_CODEC_CAP_SKIP_FRAME_FILL_PARAM |
FF_CODEC_CAP_ALLOCATE_PROGRESS | FF_CODEC_CAP_INIT_CLEANUP |
FF_CODEC_CAP_ICC_PROFILES,
};
#endif