80e9e63c94
* 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>
533 lines
16 KiB
C
533 lines
16 KiB
C
/*
|
|
* VMware Screen Codec (VMnc) decoder
|
|
* Copyright (c) 2006 Konstantin Shishkov
|
|
*
|
|
* 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
|
|
* VMware Screen Codec (VMnc) decoder
|
|
* As Alex Beregszaszi discovered, this is effectively RFB data dump
|
|
*/
|
|
|
|
#include <stdio.h>
|
|
#include <stdlib.h>
|
|
|
|
#include "libavutil/common.h"
|
|
#include "libavutil/intreadwrite.h"
|
|
#include "avcodec.h"
|
|
#include "internal.h"
|
|
|
|
enum EncTypes {
|
|
MAGIC_WMVd = 0x574D5664,
|
|
MAGIC_WMVe,
|
|
MAGIC_WMVf,
|
|
MAGIC_WMVg,
|
|
MAGIC_WMVh,
|
|
MAGIC_WMVi,
|
|
MAGIC_WMVj
|
|
};
|
|
|
|
enum HexTile_Flags {
|
|
HT_RAW = 1, // tile is raw
|
|
HT_BKG = 2, // background color is present
|
|
HT_FG = 4, // foreground color is present
|
|
HT_SUB = 8, // subrects are present
|
|
HT_CLR = 16 // each subrect has own color
|
|
};
|
|
|
|
/*
|
|
* Decoder context
|
|
*/
|
|
typedef struct VmncContext {
|
|
AVCodecContext *avctx;
|
|
AVFrame pic;
|
|
|
|
int bpp;
|
|
int bpp2;
|
|
int bigendian;
|
|
uint8_t pal[768];
|
|
int width, height;
|
|
|
|
/* cursor data */
|
|
int cur_w, cur_h;
|
|
int cur_x, cur_y;
|
|
int cur_hx, cur_hy;
|
|
uint8_t* curbits, *curmask;
|
|
uint8_t* screendta;
|
|
} VmncContext;
|
|
|
|
/* read pixel value from stream */
|
|
static av_always_inline int vmnc_get_pixel(const uint8_t* buf, int bpp, int be) {
|
|
switch(bpp * 2 + be) {
|
|
case 2:
|
|
case 3: return *buf;
|
|
case 4: return AV_RL16(buf);
|
|
case 5: return AV_RB16(buf);
|
|
case 8: return AV_RL32(buf);
|
|
case 9: return AV_RB32(buf);
|
|
default: return 0;
|
|
}
|
|
}
|
|
|
|
static void load_cursor(VmncContext *c, const uint8_t *src)
|
|
{
|
|
int i, j, p;
|
|
const int bpp = c->bpp2;
|
|
uint8_t *dst8 = c->curbits;
|
|
uint16_t *dst16 = (uint16_t*)c->curbits;
|
|
uint32_t *dst32 = (uint32_t*)c->curbits;
|
|
|
|
for(j = 0; j < c->cur_h; j++) {
|
|
for(i = 0; i < c->cur_w; i++) {
|
|
p = vmnc_get_pixel(src, bpp, c->bigendian);
|
|
src += bpp;
|
|
if(bpp == 1) *dst8++ = p;
|
|
if(bpp == 2) *dst16++ = p;
|
|
if(bpp == 4) *dst32++ = p;
|
|
}
|
|
}
|
|
dst8 = c->curmask;
|
|
dst16 = (uint16_t*)c->curmask;
|
|
dst32 = (uint32_t*)c->curmask;
|
|
for(j = 0; j < c->cur_h; j++) {
|
|
for(i = 0; i < c->cur_w; i++) {
|
|
p = vmnc_get_pixel(src, bpp, c->bigendian);
|
|
src += bpp;
|
|
if(bpp == 1) *dst8++ = p;
|
|
if(bpp == 2) *dst16++ = p;
|
|
if(bpp == 4) *dst32++ = p;
|
|
}
|
|
}
|
|
}
|
|
|
|
static void put_cursor(uint8_t *dst, int stride, VmncContext *c, int dx, int dy)
|
|
{
|
|
int i, j;
|
|
int w, h, x, y;
|
|
w = c->cur_w;
|
|
if(c->width < c->cur_x + c->cur_w) w = c->width - c->cur_x;
|
|
h = c->cur_h;
|
|
if(c->height < c->cur_y + c->cur_h) h = c->height - c->cur_y;
|
|
x = c->cur_x;
|
|
y = c->cur_y;
|
|
if(x < 0) {
|
|
w += x;
|
|
x = 0;
|
|
}
|
|
if(y < 0) {
|
|
h += y;
|
|
y = 0;
|
|
}
|
|
|
|
if((w < 1) || (h < 1)) return;
|
|
dst += x * c->bpp2 + y * stride;
|
|
|
|
if(c->bpp2 == 1) {
|
|
uint8_t* cd = c->curbits, *msk = c->curmask;
|
|
for(j = 0; j < h; j++) {
|
|
for(i = 0; i < w; i++)
|
|
dst[i] = (dst[i] & cd[i]) ^ msk[i];
|
|
msk += c->cur_w;
|
|
cd += c->cur_w;
|
|
dst += stride;
|
|
}
|
|
} else if(c->bpp2 == 2) {
|
|
uint16_t* cd = (uint16_t*)c->curbits, *msk = (uint16_t*)c->curmask;
|
|
uint16_t* dst2;
|
|
for(j = 0; j < h; j++) {
|
|
dst2 = (uint16_t*)dst;
|
|
for(i = 0; i < w; i++)
|
|
dst2[i] = (dst2[i] & cd[i]) ^ msk[i];
|
|
msk += c->cur_w;
|
|
cd += c->cur_w;
|
|
dst += stride;
|
|
}
|
|
} else if(c->bpp2 == 4) {
|
|
uint32_t* cd = (uint32_t*)c->curbits, *msk = (uint32_t*)c->curmask;
|
|
uint32_t* dst2;
|
|
for(j = 0; j < h; j++) {
|
|
dst2 = (uint32_t*)dst;
|
|
for(i = 0; i < w; i++)
|
|
dst2[i] = (dst2[i] & cd[i]) ^ msk[i];
|
|
msk += c->cur_w;
|
|
cd += c->cur_w;
|
|
dst += stride;
|
|
}
|
|
}
|
|
}
|
|
|
|
/* fill rectangle with given color */
|
|
static av_always_inline void paint_rect(uint8_t *dst, int dx, int dy, int w, int h, int color, int bpp, int stride)
|
|
{
|
|
int i, j;
|
|
dst += dx * bpp + dy * stride;
|
|
if(bpp == 1){
|
|
for(j = 0; j < h; j++) {
|
|
memset(dst, color, w);
|
|
dst += stride;
|
|
}
|
|
}else if(bpp == 2){
|
|
uint16_t* dst2;
|
|
for(j = 0; j < h; j++) {
|
|
dst2 = (uint16_t*)dst;
|
|
for(i = 0; i < w; i++) {
|
|
*dst2++ = color;
|
|
}
|
|
dst += stride;
|
|
}
|
|
}else if(bpp == 4){
|
|
uint32_t* dst2;
|
|
for(j = 0; j < h; j++) {
|
|
dst2 = (uint32_t*)dst;
|
|
for(i = 0; i < w; i++) {
|
|
dst2[i] = color;
|
|
}
|
|
dst += stride;
|
|
}
|
|
}
|
|
}
|
|
|
|
static av_always_inline void paint_raw(uint8_t *dst, int w, int h, const uint8_t* src, int bpp, int be, int stride)
|
|
{
|
|
int i, j, p;
|
|
for(j = 0; j < h; j++) {
|
|
for(i = 0; i < w; i++) {
|
|
p = vmnc_get_pixel(src, bpp, be);
|
|
src += bpp;
|
|
switch(bpp){
|
|
case 1:
|
|
dst[i] = p;
|
|
break;
|
|
case 2:
|
|
((uint16_t*)dst)[i] = p;
|
|
break;
|
|
case 4:
|
|
((uint32_t*)dst)[i] = p;
|
|
break;
|
|
}
|
|
}
|
|
dst += stride;
|
|
}
|
|
}
|
|
|
|
static int decode_hextile(VmncContext *c, uint8_t* dst, const uint8_t* src, int ssize, int w, int h, int stride)
|
|
{
|
|
int i, j, k;
|
|
int bg = 0, fg = 0, rects, color, flags, xy, wh;
|
|
const int bpp = c->bpp2;
|
|
uint8_t *dst2;
|
|
int bw = 16, bh = 16;
|
|
const uint8_t *ssrc=src;
|
|
|
|
for(j = 0; j < h; j += 16) {
|
|
dst2 = dst;
|
|
bw = 16;
|
|
if(j + 16 > h) bh = h - j;
|
|
for(i = 0; i < w; i += 16, dst2 += 16 * bpp) {
|
|
if(src - ssrc >= ssize) {
|
|
av_log(c->avctx, AV_LOG_ERROR, "Premature end of data!\n");
|
|
return -1;
|
|
}
|
|
if(i + 16 > w) bw = w - i;
|
|
flags = *src++;
|
|
if(flags & HT_RAW) {
|
|
if(src - ssrc > ssize - bw * bh * bpp) {
|
|
av_log(c->avctx, AV_LOG_ERROR, "Premature end of data!\n");
|
|
return -1;
|
|
}
|
|
paint_raw(dst2, bw, bh, src, bpp, c->bigendian, stride);
|
|
src += bw * bh * bpp;
|
|
} else {
|
|
if(flags & HT_BKG) {
|
|
bg = vmnc_get_pixel(src, bpp, c->bigendian); src += bpp;
|
|
}
|
|
if(flags & HT_FG) {
|
|
fg = vmnc_get_pixel(src, bpp, c->bigendian); src += bpp;
|
|
}
|
|
rects = 0;
|
|
if(flags & HT_SUB)
|
|
rects = *src++;
|
|
color = !!(flags & HT_CLR);
|
|
|
|
paint_rect(dst2, 0, 0, bw, bh, bg, bpp, stride);
|
|
|
|
if(src - ssrc > ssize - rects * (color * bpp + 2)) {
|
|
av_log(c->avctx, AV_LOG_ERROR, "Premature end of data!\n");
|
|
return -1;
|
|
}
|
|
for(k = 0; k < rects; k++) {
|
|
if(color) {
|
|
fg = vmnc_get_pixel(src, bpp, c->bigendian); src += bpp;
|
|
}
|
|
xy = *src++;
|
|
wh = *src++;
|
|
paint_rect(dst2, xy >> 4, xy & 0xF, (wh>>4)+1, (wh & 0xF)+1, fg, bpp, stride);
|
|
}
|
|
}
|
|
}
|
|
dst += stride * 16;
|
|
}
|
|
return src - ssrc;
|
|
}
|
|
|
|
static int decode_frame(AVCodecContext *avctx, void *data, int *got_frame,
|
|
AVPacket *avpkt)
|
|
{
|
|
const uint8_t *buf = avpkt->data;
|
|
int buf_size = avpkt->size;
|
|
VmncContext * const c = avctx->priv_data;
|
|
uint8_t *outptr;
|
|
const uint8_t *src = buf;
|
|
int dx, dy, w, h, depth, enc, chunks, res, size_left, ret;
|
|
|
|
if ((ret = ff_reget_buffer(avctx, &c->pic)) < 0) {
|
|
av_log(avctx, AV_LOG_ERROR, "reget_buffer() failed\n");
|
|
return ret;
|
|
}
|
|
|
|
c->pic.key_frame = 0;
|
|
c->pic.pict_type = AV_PICTURE_TYPE_P;
|
|
|
|
//restore screen after cursor
|
|
if(c->screendta) {
|
|
int i;
|
|
w = c->cur_w;
|
|
if(c->width < c->cur_x + w) w = c->width - c->cur_x;
|
|
h = c->cur_h;
|
|
if(c->height < c->cur_y + h) h = c->height - c->cur_y;
|
|
dx = c->cur_x;
|
|
if(dx < 0) {
|
|
w += dx;
|
|
dx = 0;
|
|
}
|
|
dy = c->cur_y;
|
|
if(dy < 0) {
|
|
h += dy;
|
|
dy = 0;
|
|
}
|
|
if((w > 0) && (h > 0)) {
|
|
outptr = c->pic.data[0] + dx * c->bpp2 + dy * c->pic.linesize[0];
|
|
for(i = 0; i < h; i++) {
|
|
memcpy(outptr, c->screendta + i * c->cur_w * c->bpp2, w * c->bpp2);
|
|
outptr += c->pic.linesize[0];
|
|
}
|
|
}
|
|
}
|
|
src += 2;
|
|
chunks = AV_RB16(src); src += 2;
|
|
while(chunks--) {
|
|
if(buf_size - (src - buf) < 12) {
|
|
av_log(avctx, AV_LOG_ERROR, "Premature end of data!\n");
|
|
return -1;
|
|
}
|
|
dx = AV_RB16(src); src += 2;
|
|
dy = AV_RB16(src); src += 2;
|
|
w = AV_RB16(src); src += 2;
|
|
h = AV_RB16(src); src += 2;
|
|
enc = AV_RB32(src); src += 4;
|
|
outptr = c->pic.data[0] + dx * c->bpp2 + dy * c->pic.linesize[0];
|
|
size_left = buf_size - (src - buf);
|
|
switch(enc) {
|
|
case MAGIC_WMVd: // cursor
|
|
if (w*(int64_t)h*c->bpp2 > INT_MAX/2 - 2) {
|
|
av_log(avctx, AV_LOG_ERROR, "dimensions too large\n");
|
|
return AVERROR_INVALIDDATA;
|
|
}
|
|
if(size_left < 2 + w * h * c->bpp2 * 2) {
|
|
av_log(avctx, AV_LOG_ERROR, "Premature end of data! (need %i got %i)\n", 2 + w * h * c->bpp2 * 2, size_left);
|
|
return -1;
|
|
}
|
|
src += 2;
|
|
c->cur_w = w;
|
|
c->cur_h = h;
|
|
c->cur_hx = dx;
|
|
c->cur_hy = dy;
|
|
if((c->cur_hx > c->cur_w) || (c->cur_hy > c->cur_h)) {
|
|
av_log(avctx, AV_LOG_ERROR, "Cursor hot spot is not in image: %ix%i of %ix%i cursor size\n", c->cur_hx, c->cur_hy, c->cur_w, c->cur_h);
|
|
c->cur_hx = c->cur_hy = 0;
|
|
}
|
|
c->curbits = av_realloc(c->curbits, c->cur_w * c->cur_h * c->bpp2);
|
|
c->curmask = av_realloc(c->curmask, c->cur_w * c->cur_h * c->bpp2);
|
|
c->screendta = av_realloc(c->screendta, c->cur_w * c->cur_h * c->bpp2);
|
|
load_cursor(c, src);
|
|
src += w * h * c->bpp2 * 2;
|
|
break;
|
|
case MAGIC_WMVe: // unknown
|
|
src += 2;
|
|
break;
|
|
case MAGIC_WMVf: // update cursor position
|
|
c->cur_x = dx - c->cur_hx;
|
|
c->cur_y = dy - c->cur_hy;
|
|
break;
|
|
case MAGIC_WMVg: // unknown
|
|
src += 10;
|
|
break;
|
|
case MAGIC_WMVh: // unknown
|
|
src += 4;
|
|
break;
|
|
case MAGIC_WMVi: // ServerInitialization struct
|
|
c->pic.key_frame = 1;
|
|
c->pic.pict_type = AV_PICTURE_TYPE_I;
|
|
depth = *src++;
|
|
if(depth != c->bpp) {
|
|
av_log(avctx, AV_LOG_INFO, "Depth mismatch. Container %i bpp, Frame data: %i bpp\n", c->bpp, depth);
|
|
}
|
|
src++;
|
|
c->bigendian = *src++;
|
|
if(c->bigendian & (~1)) {
|
|
av_log(avctx, AV_LOG_INFO, "Invalid header: bigendian flag = %i\n", c->bigendian);
|
|
return -1;
|
|
}
|
|
//skip the rest of pixel format data
|
|
src += 13;
|
|
break;
|
|
case MAGIC_WMVj: // unknown
|
|
src += 2;
|
|
break;
|
|
case 0x00000000: // raw rectangle data
|
|
if((dx + w > c->width) || (dy + h > c->height)) {
|
|
av_log(avctx, AV_LOG_ERROR, "Incorrect frame size: %ix%i+%ix%i of %ix%i\n", w, h, dx, dy, c->width, c->height);
|
|
return -1;
|
|
}
|
|
if(size_left < w * h * c->bpp2) {
|
|
av_log(avctx, AV_LOG_ERROR, "Premature end of data! (need %i got %i)\n", w * h * c->bpp2, size_left);
|
|
return -1;
|
|
}
|
|
paint_raw(outptr, w, h, src, c->bpp2, c->bigendian, c->pic.linesize[0]);
|
|
src += w * h * c->bpp2;
|
|
break;
|
|
case 0x00000005: // HexTile encoded rectangle
|
|
if((dx + w > c->width) || (dy + h > c->height)) {
|
|
av_log(avctx, AV_LOG_ERROR, "Incorrect frame size: %ix%i+%ix%i of %ix%i\n", w, h, dx, dy, c->width, c->height);
|
|
return -1;
|
|
}
|
|
res = decode_hextile(c, outptr, src, size_left, w, h, c->pic.linesize[0]);
|
|
if(res < 0)
|
|
return -1;
|
|
src += res;
|
|
break;
|
|
default:
|
|
av_log(avctx, AV_LOG_ERROR, "Unsupported block type 0x%08X\n", enc);
|
|
chunks = 0; // leave chunks decoding loop
|
|
}
|
|
}
|
|
if(c->screendta){
|
|
int i;
|
|
//save screen data before painting cursor
|
|
w = c->cur_w;
|
|
if(c->width < c->cur_x + w) w = c->width - c->cur_x;
|
|
h = c->cur_h;
|
|
if(c->height < c->cur_y + h) h = c->height - c->cur_y;
|
|
dx = c->cur_x;
|
|
if(dx < 0) {
|
|
w += dx;
|
|
dx = 0;
|
|
}
|
|
dy = c->cur_y;
|
|
if(dy < 0) {
|
|
h += dy;
|
|
dy = 0;
|
|
}
|
|
if((w > 0) && (h > 0)) {
|
|
outptr = c->pic.data[0] + dx * c->bpp2 + dy * c->pic.linesize[0];
|
|
for(i = 0; i < h; i++) {
|
|
memcpy(c->screendta + i * c->cur_w * c->bpp2, outptr, w * c->bpp2);
|
|
outptr += c->pic.linesize[0];
|
|
}
|
|
outptr = c->pic.data[0];
|
|
put_cursor(outptr, c->pic.linesize[0], c, c->cur_x, c->cur_y);
|
|
}
|
|
}
|
|
*got_frame = 1;
|
|
if ((ret = av_frame_ref(data, &c->pic)) < 0)
|
|
return ret;
|
|
|
|
/* always report that the buffer was completely consumed */
|
|
return buf_size;
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
*
|
|
* Init VMnc decoder
|
|
*
|
|
*/
|
|
static av_cold int decode_init(AVCodecContext *avctx)
|
|
{
|
|
VmncContext * const c = avctx->priv_data;
|
|
|
|
c->avctx = avctx;
|
|
|
|
c->width = avctx->width;
|
|
c->height = avctx->height;
|
|
|
|
c->bpp = avctx->bits_per_coded_sample;
|
|
c->bpp2 = c->bpp/8;
|
|
avcodec_get_frame_defaults(&c->pic);
|
|
|
|
switch(c->bpp){
|
|
case 8:
|
|
avctx->pix_fmt = AV_PIX_FMT_PAL8;
|
|
break;
|
|
case 16:
|
|
avctx->pix_fmt = AV_PIX_FMT_RGB555;
|
|
break;
|
|
case 32:
|
|
avctx->pix_fmt = AV_PIX_FMT_RGB32;
|
|
break;
|
|
default:
|
|
av_log(avctx, AV_LOG_ERROR, "Unsupported bitdepth %i\n", c->bpp);
|
|
return AVERROR_INVALIDDATA;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
*
|
|
* Uninit VMnc decoder
|
|
*
|
|
*/
|
|
static av_cold int decode_end(AVCodecContext *avctx)
|
|
{
|
|
VmncContext * const c = avctx->priv_data;
|
|
|
|
av_frame_unref(&c->pic);
|
|
|
|
av_free(c->curbits);
|
|
av_free(c->curmask);
|
|
av_free(c->screendta);
|
|
return 0;
|
|
}
|
|
|
|
AVCodec ff_vmnc_decoder = {
|
|
.name = "vmnc",
|
|
.type = AVMEDIA_TYPE_VIDEO,
|
|
.id = AV_CODEC_ID_VMNC,
|
|
.priv_data_size = sizeof(VmncContext),
|
|
.init = decode_init,
|
|
.close = decode_end,
|
|
.decode = decode_frame,
|
|
.capabilities = CODEC_CAP_DR1,
|
|
.long_name = NULL_IF_CONFIG_SMALL("VMware Screen Codec / VMware Video"),
|
|
};
|