FFmpeg/libavcodec/pixlet.c
Vittorio Giovara b44bd7ee7f pixlet: Fix architecture-dependent code and values
The constants used in the decoder used floating point precision,
and this caused different values to be generated on different
architectures. Additionally on big endian machines, the fate test
would output bytes in native order, which is different from the one
hardcoded in the test.

So, eradicate floating point numbers and use fixed point (32.32)
arithmetics everywhere, replacing constants with precomputed integer
values, and force the pixel format output to be the same in the fate
test.

Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
2017-03-06 18:15:02 -05:00

693 lines
20 KiB
C

/*
* Apple Pixlet decoder
* Copyright (c) 2016 Paul B Mahol
*
* 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
*/
#include <stdint.h>
#include "libavutil/imgutils.h"
#include "libavutil/intmath.h"
#include "libavutil/opt.h"
#include "avcodec.h"
#include "bitstream.h"
#include "bytestream.h"
#include "internal.h"
#include "thread.h"
#include "unary.h"
#define NB_LEVELS 4
#define PIXLET_MAGIC 0xDEADBEEF
#define H 0
#define V 1
typedef struct SubBand {
size_t width, height;
size_t size;
size_t x, y;
} SubBand;
typedef struct PixletContext {
AVClass *class;
GetByteContext gb;
BitstreamContext bc;
int levels;
int depth;
size_t w, h;
int16_t *filter[2];
int16_t *prediction;
int64_t scaling[4][2][NB_LEVELS];
SubBand band[4][NB_LEVELS * 3 + 1];
} PixletContext;
static av_cold int pixlet_init(AVCodecContext *avctx)
{
avctx->pix_fmt = AV_PIX_FMT_YUV420P16;
avctx->color_range = AVCOL_RANGE_JPEG;
return 0;
}
static av_cold int pixlet_close(AVCodecContext *avctx)
{
PixletContext *ctx = avctx->priv_data;
av_freep(&ctx->filter[0]);
av_freep(&ctx->filter[1]);
av_freep(&ctx->prediction);
return 0;
}
static int init_decoder(AVCodecContext *avctx)
{
PixletContext *ctx = avctx->priv_data;
int i, plane;
ctx->filter[0] = av_malloc_array(ctx->h, sizeof(int16_t));
ctx->filter[1] = av_malloc_array(FFMAX(ctx->h, ctx->w) + 16, sizeof(int16_t));
ctx->prediction = av_malloc_array((ctx->w >> NB_LEVELS), sizeof(int16_t));
if (!ctx->filter[0] || !ctx->filter[1] || !ctx->prediction)
return AVERROR(ENOMEM);
for (plane = 0; plane < 3; plane++) {
unsigned shift = plane > 0;
size_t w = ctx->w >> shift;
size_t h = ctx->h >> shift;
ctx->band[plane][0].width = w >> NB_LEVELS;
ctx->band[plane][0].height = h >> NB_LEVELS;
ctx->band[plane][0].size = (w >> NB_LEVELS) * (h >> NB_LEVELS);
for (i = 0; i < NB_LEVELS * 3; i++) {
unsigned scale = ctx->levels - (i / 3);
ctx->band[plane][i + 1].width = w >> scale;
ctx->band[plane][i + 1].height = h >> scale;
ctx->band[plane][i + 1].size = (w >> scale) * (h >> scale);
ctx->band[plane][i + 1].x = (w >> scale) * (((i + 1) % 3) != 2);
ctx->band[plane][i + 1].y = (h >> scale) * (((i + 1) % 3) != 1);
}
}
return 0;
}
static int read_low_coeffs(AVCodecContext *avctx, int16_t *dst, size_t size,
size_t width, ptrdiff_t stride)
{
PixletContext *ctx = avctx->priv_data;
BitstreamContext *bc = &ctx->bc;
unsigned cnt1, nbits, k, j = 0, i = 0;
int64_t value, state = 3;
int rlen, escape, flag = 0;
while (i < size) {
nbits = FFMIN(ff_clz((state >> 8) + 3) ^ 0x1F, 14);
cnt1 = get_unary(bc, 0, 8);
if (cnt1 < 8) {
value = bitstream_read(bc, nbits);
if (value <= 1) {
bitstream_unget(bc, value & 1, 1);
value = 1;
}
escape = value + ((1 << nbits) - 1) * cnt1 - 1;
} else {
escape = bitstream_read(bc, 16);
}
value = -((escape + flag) & 1) | 1;
dst[j++] = value * ((escape + flag + 1) >> 1);
i++;
if (j == width) {
j = 0;
dst += stride;
}
state = 120 * (escape + flag) + state - (120 * state >> 8);
flag = 0;
if (state * 4 > 0xFF || i >= size)
continue;
nbits = ((state + 8) >> 5) + (state ? ff_clz(state) : 32) - 24;
escape = av_mod_uintp2(16383, nbits);
cnt1 = get_unary(bc, 0, 8);
if (cnt1 > 7) {
rlen = bitstream_read(bc, 16);
} else {
value = bitstream_read(bc, nbits);
if (value <= 1) {
bitstream_unget(bc, value & 1, 1);
value = 1;
}
rlen = value + escape * cnt1 - 1;
}
if (i + rlen > size)
return AVERROR_INVALIDDATA;
i += rlen;
for (k = 0; k < rlen; k++) {
dst[j++] = 0;
if (j == width) {
j = 0;
dst += stride;
}
}
state = 0;
flag = rlen < 0xFFFF ? 1 : 0;
}
bitstream_align(bc);
return bitstream_tell(bc) >> 3;
}
static int read_high_coeffs(AVCodecContext *avctx, uint8_t *src, int16_t *dst,
int size, int64_t c, int a, int64_t d,
int width, ptrdiff_t stride)
{
PixletContext *ctx = avctx->priv_data;
BitstreamContext *bc = &ctx->bc;
unsigned cnt1, shbits, rlen, nbits, length, i = 0, j = 0, k;
int ret, escape, pfx, cthulu, yflag, xflag, flag = 0;
int64_t state = 3, value, tmp;
ret = bitstream_init8(bc, src, bytestream2_get_bytes_left(&ctx->gb));
if (ret < 0)
return ret;
cthulu = (a >= 0) + (a ^ (a >> 31)) - (a >> 31);
if (cthulu != 1) {
nbits = 33 - ff_clz(cthulu - 1);
if (nbits > 16)
return AVERROR_INVALIDDATA;
} else {
nbits = 1;
}
length = 25 - nbits;
while (i < size) {
if (state >> 8 != -3)
value = ff_clz((state >> 8) + 3) ^ 0x1F;
else
value = -1;
cnt1 = get_unary(bc, 0, length);
if (cnt1 >= length) {
cnt1 = bitstream_read(bc, nbits);
} else {
pfx = 14 + (((value - 14) >> 32) & (value - 14));
cnt1 *= (1 << pfx) - 1;
shbits = bitstream_read(bc, pfx);
if (shbits <= 1) {
bitstream_unget(bc, shbits & 1, 1);
shbits = 1;
}
cnt1 += shbits - 1;
}
xflag = flag + cnt1;
yflag = xflag;
if (flag + cnt1 == 0) {
value = 0;
} else {
xflag &= 1u;
tmp = c * ((yflag + 1) >> 1) + (c >> 1);
value = xflag + (tmp ^ -xflag);
}
i++;
dst[j++] = value;
if (j == width) {
j = 0;
dst += stride;
}
state += d * yflag - (d * state >> 8);
flag = 0;
if (state * 4 > 0xFF || i >= size)
continue;
pfx = ((state + 8) >> 5) + (state ? ff_clz(state) : 32) - 24;
escape = av_mod_uintp2(16383, pfx);
cnt1 = get_unary(bc, 0, 8);
if (cnt1 < 8) {
if (pfx < 1 || pfx > 25)
return AVERROR_INVALIDDATA;
value = bitstream_read(bc, pfx);
if (value <= 1) {
bitstream_unget(bc, value & 1, 1);
value = 1;
}
rlen = value + escape * cnt1 - 1;
} else {
if (bitstream_read_bit(bc))
value = bitstream_read(bc, 16);
else
value = bitstream_read(bc, 8);
rlen = value + 8 * escape;
}
if (rlen > 0xFFFF || i + rlen > size)
return AVERROR_INVALIDDATA;
i += rlen;
for (k = 0; k < rlen; k++) {
dst[j++] = 0;
if (j == width) {
j = 0;
dst += stride;
}
}
state = 0;
flag = rlen < 0xFFFF ? 1 : 0;
}
bitstream_align(bc);
return bitstream_tell(bc) >> 3;
}
static int read_highpass(AVCodecContext *avctx, uint8_t *ptr,
int plane, AVFrame *frame)
{
PixletContext *ctx = avctx->priv_data;
ptrdiff_t stride = frame->linesize[plane] / 2;
int i, ret;
for (i = 0; i < ctx->levels * 3; i++) {
int32_t a = bytestream2_get_be32(&ctx->gb);
int32_t b = bytestream2_get_be32(&ctx->gb);
int32_t c = bytestream2_get_be32(&ctx->gb);
int32_t d = bytestream2_get_be32(&ctx->gb);
int16_t *dest = (int16_t *)frame->data[plane] +
ctx->band[plane][i + 1].x +
ctx->band[plane][i + 1].y * stride;
size_t size = ctx->band[plane][i + 1].size;
uint32_t magic = bytestream2_get_be32(&ctx->gb);
if (magic != PIXLET_MAGIC) {
av_log(avctx, AV_LOG_ERROR,
"wrong magic number: 0x%"PRIX32" for plane %d, band %d\n",
magic, plane, i);
return AVERROR_INVALIDDATA;
}
ret = read_high_coeffs(avctx, ptr + bytestream2_tell(&ctx->gb), dest,
size, c, (b >= FFABS(a)) ? b : a, d,
ctx->band[plane][i + 1].width, stride);
if (ret < 0) {
av_log(avctx, AV_LOG_ERROR,
"error in highpass coefficients for plane %d, band %d\n",
plane, i);
return ret;
}
bytestream2_skip(&ctx->gb, ret);
}
return 0;
}
static void line_add_sat_s16(int16_t *dst, const int16_t *src, size_t len)
{
int i;
for (i = 0; i < len; i++) {
int val = dst[i] + src[i];
dst[i] = av_clip_int16(val);
}
}
static void lowpass_prediction(int16_t *dst, int16_t *pred,
size_t width, size_t height, ptrdiff_t stride)
{
int i, j;
memset(pred, 0, width * sizeof(*pred));
for (i = 0; i < height; i++) {
line_add_sat_s16(pred, dst, width);
dst[0] = pred[0];
for (j = 1; j < width; j++)
dst[j] = pred[j] + dst[j - 1];
dst += stride;
}
}
static void filterfn(int16_t *dest, int16_t *tmp, size_t size, int64_t scale)
{
int16_t *low, *high, *ll, *lh, *hl, *hh;
int hsize, i, j;
int64_t value;
hsize = size >> 1;
low = tmp + 4;
high = &low[hsize + 8];
memcpy(low, dest, size);
memcpy(high, dest + hsize, size);
ll = &low[hsize];
lh = &low[hsize];
hl = &high[hsize];
hh = hl;
for (i = 4, j = 2; i; i--, j++, ll--, hh++, lh++, hl--) {
low[i - 5] = low[j - 1];
lh[0] = ll[-1];
high[i - 5] = high[j - 2];
hh[0] = hl[-2];
}
for (i = 0; i < hsize; i++) {
value = (int64_t) low [i + 1] * -INT64_C(325392907) +
(int64_t) low [i + 0] * INT64_C(3687786320) +
(int64_t) low [i - 1] * -INT64_C(325392907) +
(int64_t) high[i + 0] * INT64_C(1518500249) +
(int64_t) high[i - 1] * INT64_C(1518500249);
dest[i * 2] = av_clip_int16(((value >> 32) * scale) >> 32);
}
for (i = 0; i < hsize; i++) {
value = (int64_t) low [i + 2] * -INT64_C(65078576) +
(int64_t) low [i + 1] * INT64_C(1583578880) +
(int64_t) low [i + 0] * INT64_C(1583578880) +
(int64_t) low [i - 1] * -INT64_C(65078576) +
(int64_t) high[i + 1] * INT64_C(303700064) +
(int64_t) high[i + 0] * -INT64_C(3644400640) +
(int64_t) high[i - 1] * INT64_C(303700064);
dest[i * 2 + 1] = av_clip_int16(((value >> 32) * scale) >> 32);
}
}
static void reconstruction(AVCodecContext *avctx, int16_t *dest,
size_t width, size_t height, ptrdiff_t stride,
int64_t *scaling_h, int64_t *scaling_v)
{
PixletContext *ctx = avctx->priv_data;
unsigned scaled_width, scaled_height;
int16_t *ptr, *tmp;
int i, j, k;
scaled_width = width >> NB_LEVELS;
scaled_height = height >> NB_LEVELS;
tmp = ctx->filter[0];
for (i = 0; i < NB_LEVELS; i++) {
int64_t scale_v = scaling_v[i];
int64_t scale_h = scaling_h[i];
scaled_width <<= 1;
scaled_height <<= 1;
ptr = dest;
for (j = 0; j < scaled_height; j++) {
filterfn(ptr, ctx->filter[1], scaled_width, scale_v);
ptr += stride;
}
for (j = 0; j < scaled_width; j++) {
ptr = dest + j;
for (k = 0; k < scaled_height; k++) {
tmp[k] = *ptr;
ptr += stride;
}
filterfn(tmp, ctx->filter[1], scaled_height, scale_h);
ptr = dest + j;
for (k = 0; k < scaled_height; k++) {
*ptr = tmp[k];
ptr += stride;
}
}
}
}
static void postprocess_luma(AVFrame *frame, size_t w, size_t h, int depth)
{
uint16_t *dsty = (uint16_t *)frame->data[0];
int16_t *srcy = (int16_t *)frame->data[0];
ptrdiff_t stridey = frame->linesize[0] / 2;
int i, j;
for (j = 0; j < h; j++) {
for (i = 0; i < w; i++) {
if (srcy[i] <= 0)
dsty[i] = 0;
else if (srcy[i] > ((1 << depth) - 1))
dsty[i] = 65535;
else
dsty[i] = ((int64_t) srcy[i] * srcy[i] * 65535) /
((1 << depth) - 1) / ((1 << depth) - 1);
}
dsty += stridey;
srcy += stridey;
}
}
static void postprocess_chroma(AVFrame *frame, int w, int h, int depth)
{
uint16_t *dstu = (uint16_t *)frame->data[1];
uint16_t *dstv = (uint16_t *)frame->data[2];
int16_t *srcu = (int16_t *)frame->data[1];
int16_t *srcv = (int16_t *)frame->data[2];
ptrdiff_t strideu = frame->linesize[1] / 2;
ptrdiff_t stridev = frame->linesize[2] / 2;
const unsigned add = 1 << (depth - 1);
const unsigned shift = 16 - depth;
int i, j;
for (j = 0; j < h; j++) {
for (i = 0; i < w; i++) {
dstu[i] = av_clip_uintp2_c(add + srcu[i], depth) << shift;
dstv[i] = av_clip_uintp2_c(add + srcv[i], depth) << shift;
}
dstu += strideu;
dstv += stridev;
srcu += strideu;
srcv += stridev;
}
}
static int decode_plane(AVCodecContext *avctx, int plane,
AVPacket *avpkt, AVFrame *frame)
{
PixletContext *ctx = avctx->priv_data;
ptrdiff_t stride = frame->linesize[plane] / 2;
unsigned shift = plane > 0;
int16_t *dst;
int i, ret;
for (i = ctx->levels - 1; i >= 0; i--) {
int32_t h = sign_extend(bytestream2_get_be32(&ctx->gb), 32);
int32_t v = sign_extend(bytestream2_get_be32(&ctx->gb), 32);
if (!h || !v)
return AVERROR_INVALIDDATA;
ctx->scaling[plane][H][i] = (1000000ULL << 32) / h;
ctx->scaling[plane][V][i] = (1000000ULL << 32) / v;
}
bytestream2_skip(&ctx->gb, 4);
dst = (int16_t *)frame->data[plane];
dst[0] = sign_extend(bytestream2_get_be16(&ctx->gb), 16);
ret = bitstream_init8(&ctx->bc, avpkt->data + bytestream2_tell(&ctx->gb),
bytestream2_get_bytes_left(&ctx->gb));
if (ret < 0)
return ret;
ret = read_low_coeffs(avctx, dst + 1, ctx->band[plane][0].width - 1,
ctx->band[plane][0].width - 1, 0);
if (ret < 0) {
av_log(avctx, AV_LOG_ERROR,
"error in lowpass coefficients for plane %d, top row\n", plane);
return ret;
}
ret = read_low_coeffs(avctx, dst + stride,
ctx->band[plane][0].height - 1, 1, stride);
if (ret < 0) {
av_log(avctx, AV_LOG_ERROR,
"error in lowpass coefficients for plane %d, left column\n",
plane);
return ret;
}
ret = read_low_coeffs(avctx, dst + stride + 1,
(ctx->band[plane][0].width - 1) * (ctx->band[plane][0].height - 1),
ctx->band[plane][0].width - 1, stride);
if (ret < 0) {
av_log(avctx, AV_LOG_ERROR,
"error in lowpass coefficients for plane %d, rest\n", plane);
return ret;
}
bytestream2_skip(&ctx->gb, ret);
if (bytestream2_get_bytes_left(&ctx->gb) <= 0) {
av_log(avctx, AV_LOG_ERROR, "no bytes left\n");
return AVERROR_INVALIDDATA;
}
ret = read_highpass(avctx, avpkt->data, plane, frame);
if (ret < 0)
return ret;
lowpass_prediction(dst, ctx->prediction, ctx->band[plane][0].width,
ctx->band[plane][0].height, stride);
reconstruction(avctx, (int16_t *)frame->data[plane], ctx->w >> shift,
ctx->h >> shift, stride, ctx->scaling[plane][H],
ctx->scaling[plane][V]);
return 0;
}
static int pixlet_decode_frame(AVCodecContext *avctx, void *data,
int *got_frame, AVPacket *avpkt)
{
PixletContext *ctx = avctx->priv_data;
int i, w, h, width, height, ret, version;
AVFrame *p = data;
ThreadFrame frame = { .f = data };
uint32_t pktsize;
bytestream2_init(&ctx->gb, avpkt->data, avpkt->size);
pktsize = bytestream2_get_be32(&ctx->gb);
if (pktsize <= 44 || pktsize - 4 > bytestream2_get_bytes_left(&ctx->gb)) {
av_log(avctx, AV_LOG_ERROR, "Invalid packet size %"PRIu32".\n", pktsize);
return AVERROR_INVALIDDATA;
}
version = bytestream2_get_le32(&ctx->gb);
if (version != 1)
avpriv_request_sample(avctx, "Version %d", version);
bytestream2_skip(&ctx->gb, 4);
if (bytestream2_get_be32(&ctx->gb) != 1)
return AVERROR_INVALIDDATA;
bytestream2_skip(&ctx->gb, 4);
width = bytestream2_get_be32(&ctx->gb);
height = bytestream2_get_be32(&ctx->gb);
w = FFALIGN(width, 1 << (NB_LEVELS + 1));
h = FFALIGN(height, 1 << (NB_LEVELS + 1));
ctx->levels = bytestream2_get_be32(&ctx->gb);
if (ctx->levels != NB_LEVELS)
return AVERROR_INVALIDDATA;
ctx->depth = bytestream2_get_be32(&ctx->gb);
if (ctx->depth < 8 || ctx->depth > 15) {
avpriv_request_sample(avctx, "Depth %d", ctx->depth);
return AVERROR_INVALIDDATA;
}
ret = ff_set_dimensions(avctx, w, h);
if (ret < 0)
return ret;
avctx->width = width;
avctx->height = height;
/* reinit should dimensions change */
if (ctx->w != w || ctx->h != h) {
pixlet_close(avctx);
ctx->w = w;
ctx->h = h;
ret = init_decoder(avctx);
if (ret < 0) {
pixlet_close(avctx);
ctx->w = 0;
ctx->h = 0;
return ret;
}
}
bytestream2_skip(&ctx->gb, 8);
ret = ff_thread_get_buffer(avctx, &frame, 0);
if (ret < 0)
return ret;
for (i = 0; i < 3; i++) {
ret = decode_plane(avctx, i, avpkt, frame.f);
if (ret < 0)
return ret;
if (avctx->flags & AV_CODEC_FLAG_GRAY)
break;
}
postprocess_luma(frame.f, ctx->w, ctx->h, ctx->depth);
postprocess_chroma(frame.f, ctx->w >> 1, ctx->h >> 1, ctx->depth);
p->pict_type = AV_PICTURE_TYPE_I;
p->color_range = AVCOL_RANGE_JPEG;
p->key_frame = 1;
*got_frame = 1;
return pktsize;
}
#if HAVE_THREADS
static int pixlet_init_thread_copy(AVCodecContext *avctx)
{
PixletContext *ctx = avctx->priv_data;
ctx->filter[0] = NULL;
ctx->filter[1] = NULL;
ctx->prediction = NULL;
ctx->w = 0;
ctx->h = 0;
return 0;
}
#endif /* HAVE_THREADS */
AVCodec ff_pixlet_decoder = {
.name = "pixlet",
.long_name = NULL_IF_CONFIG_SMALL("Apple Pixlet"),
.type = AVMEDIA_TYPE_VIDEO,
.id = AV_CODEC_ID_PIXLET,
.init = pixlet_init,
.init_thread_copy = ONLY_IF_THREADS_ENABLED(pixlet_init_thread_copy),
.close = pixlet_close,
.decode = pixlet_decode_frame,
.priv_data_size = sizeof(PixletContext),
.capabilities = AV_CODEC_CAP_DR1 |
AV_CODEC_CAP_FRAME_THREADS,
.caps_internal = FF_CODEC_CAP_INIT_THREADSAFE |
FF_CODEC_CAP_INIT_CLEANUP,
};