1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-11-26 19:01:44 +02:00
FFmpeg/libavcodec/utvideo.c
Aneesh Dogra 12e984aed7 utvideo: frame multithreading.
>> time ./avconv -i file.avi -f null -
Before : real	0m7.784s
After  : real   0m3.662s

Tested on a Intel Core i3 Processor (2 cores, 4 threads).

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
2012-01-08 12:21:10 -08:00

582 lines
18 KiB
C

/*
* Ut Video decoder
* Copyright (c) 2011 Konstantin Shishkov
*
* This file is part of Libav.
*
* Libav is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* Libav is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with Libav; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* @file
* Ut Video decoder
*/
#include <stdlib.h>
#include "libavutil/intreadwrite.h"
#include "avcodec.h"
#include "bytestream.h"
#include "get_bits.h"
#include "dsputil.h"
#include "thread.h"
enum {
PRED_NONE = 0,
PRED_LEFT,
PRED_GRADIENT,
PRED_MEDIAN,
};
typedef struct UtvideoContext {
AVCodecContext *avctx;
AVFrame pic;
DSPContext dsp;
uint32_t frame_info_size, flags, frame_info;
int planes;
int slices;
int compression;
int interlaced;
int frame_pred;
uint8_t *slice_bits;
int slice_bits_size;
} UtvideoContext;
typedef struct HuffEntry {
uint8_t sym;
uint8_t len;
} HuffEntry;
static int huff_cmp(const void *a, const void *b)
{
const HuffEntry *aa = a, *bb = b;
return (aa->len - bb->len)*256 + aa->sym - bb->sym;
}
static int build_huff(const uint8_t *src, VLC *vlc, int *fsym)
{
int i;
HuffEntry he[256];
int last;
uint32_t codes[256];
uint8_t bits[256];
uint8_t syms[256];
uint32_t code;
*fsym = -1;
for (i = 0; i < 256; i++) {
he[i].sym = i;
he[i].len = *src++;
}
qsort(he, 256, sizeof(*he), huff_cmp);
if (!he[0].len) {
*fsym = he[0].sym;
return 0;
}
if (he[0].len > 32)
return -1;
last = 255;
while (he[last].len == 255 && last)
last--;
code = 1;
for (i = last; i >= 0; i--) {
codes[i] = code >> (32 - he[i].len);
bits[i] = he[i].len;
syms[i] = he[i].sym;
code += 0x80000000u >> (he[i].len - 1);
}
return init_vlc_sparse(vlc, FFMIN(he[last].len, 9), last + 1,
bits, sizeof(*bits), sizeof(*bits),
codes, sizeof(*codes), sizeof(*codes),
syms, sizeof(*syms), sizeof(*syms), 0);
}
static int decode_plane(UtvideoContext *c, int plane_no,
uint8_t *dst, int step, int stride,
int width, int height,
const uint8_t *src, int src_size, int use_pred)
{
int i, j, slice, pix;
int sstart, send;
VLC vlc;
GetBitContext gb;
int prev, fsym;
const int cmask = ~(!plane_no && c->avctx->pix_fmt == PIX_FMT_YUV420P);
if (build_huff(src, &vlc, &fsym)) {
av_log(c->avctx, AV_LOG_ERROR, "Cannot build Huffman codes\n");
return AVERROR_INVALIDDATA;
}
if (fsym >= 0) { // build_huff reported a symbol to fill slices with
send = 0;
for (slice = 0; slice < c->slices; slice++) {
uint8_t *dest;
sstart = send;
send = (height * (slice + 1) / c->slices) & cmask;
dest = dst + sstart * stride;
prev = 0x80;
for (j = sstart; j < send; j++) {
for (i = 0; i < width * step; i += step) {
pix = fsym;
if (use_pred) {
prev += pix;
pix = prev;
}
dest[i] = pix;
}
dest += stride;
}
}
return 0;
}
src += 256;
src_size -= 256;
send = 0;
for (slice = 0; slice < c->slices; slice++) {
uint8_t *dest;
int slice_data_start, slice_data_end, slice_size;
sstart = send;
send = (height * (slice + 1) / c->slices) & cmask;
dest = dst + sstart * stride;
// slice offset and size validation was done earlier
slice_data_start = slice ? AV_RL32(src + slice * 4 - 4) : 0;
slice_data_end = AV_RL32(src + slice * 4);
slice_size = slice_data_end - slice_data_start;
if (!slice_size) {
for (j = sstart; j < send; j++) {
for (i = 0; i < width * step; i += step)
dest[i] = 0x80;
dest += stride;
}
continue;
}
memcpy(c->slice_bits, src + slice_data_start + c->slices * 4, slice_size);
memset(c->slice_bits + slice_size, 0, FF_INPUT_BUFFER_PADDING_SIZE);
c->dsp.bswap_buf((uint32_t*)c->slice_bits, (uint32_t*)c->slice_bits,
(slice_data_end - slice_data_start + 3) >> 2);
init_get_bits(&gb, c->slice_bits, slice_size * 8);
prev = 0x80;
for (j = sstart; j < send; j++) {
for (i = 0; i < width * step; i += step) {
if (get_bits_left(&gb) <= 0) {
av_log(c->avctx, AV_LOG_ERROR, "Slice decoding ran out of bits\n");
goto fail;
}
pix = get_vlc2(&gb, vlc.table, vlc.bits, 4);
if (pix < 0) {
av_log(c->avctx, AV_LOG_ERROR, "Decoding error\n");
goto fail;
}
if (use_pred) {
prev += pix;
pix = prev;
}
dest[i] = pix;
}
dest += stride;
}
if (get_bits_left(&gb) > 32)
av_log(c->avctx, AV_LOG_WARNING, "%d bits left after decoding slice\n",
get_bits_left(&gb));
}
free_vlc(&vlc);
return 0;
fail:
free_vlc(&vlc);
return AVERROR_INVALIDDATA;
}
static const int rgb_order[4] = { 1, 2, 0, 3 };
static void restore_rgb_planes(uint8_t *src, int step, int stride, int width, int height)
{
int i, j;
uint8_t r, g, b;
for (j = 0; j < height; j++) {
for (i = 0; i < width * step; i += step) {
r = src[i];
g = src[i + 1];
b = src[i + 2];
src[i] = r + g - 0x80;
src[i + 2] = b + g - 0x80;
}
src += stride;
}
}
static void restore_median(uint8_t *src, int step, int stride,
int width, int height, int slices, int rmode)
{
int i, j, slice;
int A, B, C;
uint8_t *bsrc;
int slice_start, slice_height;
const int cmask = ~rmode;
for (slice = 0; slice < slices; slice++) {
slice_start = ((slice * height) / slices) & cmask;
slice_height = ((((slice + 1) * height) / slices) & cmask) - slice_start;
bsrc = src + slice_start * stride;
// first line - left neighbour prediction
bsrc[0] += 0x80;
A = bsrc[0];
for (i = step; i < width * step; i += step) {
bsrc[i] += A;
A = bsrc[i];
}
bsrc += stride;
if (slice_height == 1)
continue;
// second line - first element has top predition, the rest uses median
C = bsrc[-stride];
bsrc[0] += C;
A = bsrc[0];
for (i = step; i < width * step; i += step) {
B = bsrc[i - stride];
bsrc[i] += mid_pred(A, B, (uint8_t)(A + B - C));
C = B;
A = bsrc[i];
}
bsrc += stride;
// the rest of lines use continuous median prediction
for (j = 2; j < slice_height; j++) {
for (i = 0; i < width * step; i += step) {
B = bsrc[i - stride];
bsrc[i] += mid_pred(A, B, (uint8_t)(A + B - C));
C = B;
A = bsrc[i];
}
bsrc += stride;
}
}
}
/* UtVideo interlaced mode treats every two lines as a single one,
* so restoring function should take care of possible padding between
* two parts of the same "line".
*/
static void restore_median_il(uint8_t *src, int step, int stride,
int width, int height, int slices, int rmode)
{
int i, j, slice;
int A, B, C;
uint8_t *bsrc;
int slice_start, slice_height;
const int cmask = ~(rmode ? 3 : 1);
const int stride2 = stride << 1;
for (slice = 0; slice < slices; slice++) {
slice_start = ((slice * height) / slices) & cmask;
slice_height = ((((slice + 1) * height) / slices) & cmask) - slice_start;
slice_height >>= 1;
bsrc = src + slice_start * stride;
// first line - left neighbour prediction
bsrc[0] += 0x80;
A = bsrc[0];
for (i = step; i < width * step; i += step) {
bsrc[i] += A;
A = bsrc[i];
}
for (i = 0; i < width * step; i += step) {
bsrc[stride + i] += A;
A = bsrc[stride + i];
}
bsrc += stride2;
if (slice_height == 1)
continue;
// second line - first element has top predition, the rest uses median
C = bsrc[-stride2];
bsrc[0] += C;
A = bsrc[0];
for (i = step; i < width * step; i += step) {
B = bsrc[i - stride2];
bsrc[i] += mid_pred(A, B, (uint8_t)(A + B - C));
C = B;
A = bsrc[i];
}
for (i = 0; i < width * step; i += step) {
B = bsrc[i - stride];
bsrc[stride + i] += mid_pred(A, B, (uint8_t)(A + B - C));
C = B;
A = bsrc[stride + i];
}
bsrc += stride2;
// the rest of lines use continuous median prediction
for (j = 2; j < slice_height; j++) {
for (i = 0; i < width * step; i += step) {
B = bsrc[i - stride2];
bsrc[i] += mid_pred(A, B, (uint8_t)(A + B - C));
C = B;
A = bsrc[i];
}
for (i = 0; i < width * step; i += step) {
B = bsrc[i - stride];
bsrc[i + stride] += mid_pred(A, B, (uint8_t)(A + B - C));
C = B;
A = bsrc[i + stride];
}
bsrc += stride2;
}
}
}
static int decode_frame(AVCodecContext *avctx, void *data, int *data_size, AVPacket *avpkt)
{
const uint8_t *buf = avpkt->data;
int buf_size = avpkt->size;
const uint8_t *buf_end = buf + buf_size;
UtvideoContext *c = avctx->priv_data;
const uint8_t *ptr;
int i, j;
const uint8_t *plane_start[5];
int plane_size, max_slice_size = 0, slice_start, slice_end, slice_size;
int ret;
if (c->pic.data[0])
ff_thread_release_buffer(avctx, &c->pic);
c->pic.reference = 1;
c->pic.buffer_hints = FF_BUFFER_HINTS_VALID;
if ((ret = ff_thread_get_buffer(avctx, &c->pic)) < 0) {
av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
return ret;
}
ff_thread_finish_setup(avctx);
/* parse plane structure to retrieve frame flags and validate slice offsets */
ptr = buf;
for (i = 0; i < c->planes; i++) {
plane_start[i] = ptr;
if (buf_end - ptr < 256 + 4 * c->slices) {
av_log(avctx, AV_LOG_ERROR, "Insufficient data for a plane\n");
return AVERROR_INVALIDDATA;
}
ptr += 256;
slice_start = 0;
slice_end = 0;
for (j = 0; j < c->slices; j++) {
slice_end = bytestream_get_le32(&ptr);
slice_size = slice_end - slice_start;
if (slice_size < 0) {
av_log(avctx, AV_LOG_ERROR, "Incorrect slice size\n");
return AVERROR_INVALIDDATA;
}
slice_start = slice_end;
max_slice_size = FFMAX(max_slice_size, slice_size);
}
plane_size = slice_end;
if (buf_end - ptr < plane_size) {
av_log(avctx, AV_LOG_ERROR, "Plane size is bigger than available data\n");
return AVERROR_INVALIDDATA;
}
ptr += plane_size;
}
plane_start[c->planes] = ptr;
if (buf_end - ptr < c->frame_info_size) {
av_log(avctx, AV_LOG_ERROR, "Not enough data for frame information\n");
return AVERROR_INVALIDDATA;
}
c->frame_info = AV_RL32(ptr);
av_log(avctx, AV_LOG_DEBUG, "frame information flags %X\n", c->frame_info);
c->frame_pred = (c->frame_info >> 8) & 3;
if (c->frame_pred == PRED_GRADIENT) {
av_log_ask_for_sample(avctx, "Frame uses gradient prediction\n");
return AVERROR_PATCHWELCOME;
}
av_fast_malloc(&c->slice_bits, &c->slice_bits_size,
max_slice_size + FF_INPUT_BUFFER_PADDING_SIZE);
if (!c->slice_bits) {
av_log(avctx, AV_LOG_ERROR, "Cannot allocate temporary buffer\n");
return AVERROR(ENOMEM);
}
switch (c->avctx->pix_fmt) {
case PIX_FMT_RGB24:
case PIX_FMT_RGBA:
for (i = 0; i < c->planes; i++) {
ret = decode_plane(c, i, c->pic.data[0] + rgb_order[i], c->planes,
c->pic.linesize[0], avctx->width, avctx->height,
plane_start[i], plane_start[i + 1] - plane_start[i],
c->frame_pred == PRED_LEFT);
if (ret)
return ret;
if (c->frame_pred == PRED_MEDIAN)
restore_median(c->pic.data[0] + rgb_order[i], c->planes,
c->pic.linesize[0], avctx->width, avctx->height,
c->slices, 0);
}
restore_rgb_planes(c->pic.data[0], c->planes, c->pic.linesize[0],
avctx->width, avctx->height);
break;
case PIX_FMT_YUV420P:
for (i = 0; i < 3; i++) {
ret = decode_plane(c, i, c->pic.data[i], 1,
c->pic.linesize[i], avctx->width >> !!i, avctx->height >> !!i,
plane_start[i], plane_start[i + 1] - plane_start[i],
c->frame_pred == PRED_LEFT);
if (ret)
return ret;
if (c->frame_pred == PRED_MEDIAN) {
if (!c->interlaced) {
restore_median(c->pic.data[i], 1, c->pic.linesize[i],
avctx->width >> !!i, avctx->height >> !!i,
c->slices, !i);
} else {
restore_median_il(c->pic.data[i], 1, c->pic.linesize[i],
avctx->width >> !!i,
avctx->height >> !!i,
c->slices, !i);
}
}
}
break;
case PIX_FMT_YUV422P:
for (i = 0; i < 3; i++) {
ret = decode_plane(c, i, c->pic.data[i], 1,
c->pic.linesize[i], avctx->width >> !!i, avctx->height,
plane_start[i], plane_start[i + 1] - plane_start[i],
c->frame_pred == PRED_LEFT);
if (ret)
return ret;
if (c->frame_pred == PRED_MEDIAN) {
if (!c->interlaced) {
restore_median(c->pic.data[i], 1, c->pic.linesize[i],
avctx->width >> !!i, avctx->height,
c->slices, 0);
} else {
restore_median_il(c->pic.data[i], 1, c->pic.linesize[i],
avctx->width >> !!i, avctx->height,
c->slices, 0);
}
}
}
break;
}
*data_size = sizeof(AVFrame);
*(AVFrame*)data = c->pic;
/* always report that the buffer was completely consumed */
return buf_size;
}
static av_cold int decode_init(AVCodecContext *avctx)
{
UtvideoContext * const c = avctx->priv_data;
c->avctx = avctx;
dsputil_init(&c->dsp, avctx);
if (avctx->extradata_size < 16) {
av_log(avctx, AV_LOG_ERROR, "Insufficient extradata size %d, should be at least 16\n",
avctx->extradata_size);
return AVERROR_INVALIDDATA;
}
av_log(avctx, AV_LOG_DEBUG, "Encoder version %d.%d.%d.%d\n",
avctx->extradata[3], avctx->extradata[2],
avctx->extradata[1], avctx->extradata[0]);
av_log(avctx, AV_LOG_DEBUG, "Original format %X\n", AV_RB32(avctx->extradata + 4));
c->frame_info_size = AV_RL32(avctx->extradata + 8);
c->flags = AV_RL32(avctx->extradata + 12);
if (c->frame_info_size != 4)
av_log_ask_for_sample(avctx, "Frame info is not 4 bytes\n");
av_log(avctx, AV_LOG_DEBUG, "Encoding parameters %08X\n", c->flags);
c->slices = (c->flags >> 24) + 1;
c->compression = c->flags & 1;
c->interlaced = c->flags & 0x800;
c->slice_bits_size = 0;
switch (avctx->codec_tag) {
case MKTAG('U', 'L', 'R', 'G'):
c->planes = 3;
avctx->pix_fmt = PIX_FMT_RGB24;
break;
case MKTAG('U', 'L', 'R', 'A'):
c->planes = 4;
avctx->pix_fmt = PIX_FMT_RGBA;
break;
case MKTAG('U', 'L', 'Y', '0'):
c->planes = 3;
avctx->pix_fmt = PIX_FMT_YUV420P;
break;
case MKTAG('U', 'L', 'Y', '2'):
c->planes = 3;
avctx->pix_fmt = PIX_FMT_YUV422P;
break;
default:
av_log(avctx, AV_LOG_ERROR, "Unknown Ut Video FOURCC provided (%08X)\n",
avctx->codec_tag);
return AVERROR_INVALIDDATA;
}
return 0;
}
static av_cold int decode_end(AVCodecContext *avctx)
{
UtvideoContext * const c = avctx->priv_data;
if (c->pic.data[0])
ff_thread_release_buffer(avctx, &c->pic);
av_freep(&c->slice_bits);
return 0;
}
AVCodec ff_utvideo_decoder = {
.name = "utvideo",
.type = AVMEDIA_TYPE_VIDEO,
.id = CODEC_ID_UTVIDEO,
.priv_data_size = sizeof(UtvideoContext),
.init = decode_init,
.close = decode_end,
.decode = decode_frame,
.capabilities = CODEC_CAP_DR1 | CODEC_CAP_FRAME_THREADS,
.long_name = NULL_IF_CONFIG_SMALL("Ut Video"),
};