|
|
|
@ -18,21 +18,22 @@ |
|
|
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
|
|
|
|
*/ |
|
|
|
|
|
|
|
|
|
#include <assert.h> |
|
|
|
|
#include <inttypes.h> |
|
|
|
|
#include <string.h> |
|
|
|
|
#include <math.h> |
|
|
|
|
#include <stdio.h> |
|
|
|
|
#include "config.h" |
|
|
|
|
#include <assert.h> |
|
|
|
|
#include "swscale.h" |
|
|
|
|
#include "swscale_internal.h" |
|
|
|
|
#include "rgb2rgb.h" |
|
|
|
|
#include "libavutil/intreadwrite.h" |
|
|
|
|
#include "libavutil/cpu.h" |
|
|
|
|
#include <string.h> |
|
|
|
|
|
|
|
|
|
#include "libavutil/avutil.h" |
|
|
|
|
#include "libavutil/mathematics.h" |
|
|
|
|
#include "libavutil/bswap.h" |
|
|
|
|
#include "libavutil/cpu.h" |
|
|
|
|
#include "libavutil/intreadwrite.h" |
|
|
|
|
#include "libavutil/mathematics.h" |
|
|
|
|
#include "libavutil/pixdesc.h" |
|
|
|
|
#include "config.h" |
|
|
|
|
#include "rgb2rgb.h" |
|
|
|
|
#include "swscale_internal.h" |
|
|
|
|
#include "swscale.h" |
|
|
|
|
|
|
|
|
|
DECLARE_ALIGNED(8, const uint8_t, dither_8x8_128)[8][8] = { |
|
|
|
|
{ 36, 68, 60, 92, 34, 66, 58, 90, }, |
|
|
|
@ -44,12 +45,13 @@ DECLARE_ALIGNED(8, const uint8_t, dither_8x8_128)[8][8] = { |
|
|
|
|
{ 48, 80, 40, 72, 54, 86, 46, 78, }, |
|
|
|
|
{ 112, 16, 104, 8, 118, 22, 110, 14, }, |
|
|
|
|
}; |
|
|
|
|
DECLARE_ALIGNED(8, const uint8_t, ff_sws_pb_64)[8] = |
|
|
|
|
{ 64, 64, 64, 64, 64, 64, 64, 64 }; |
|
|
|
|
|
|
|
|
|
static av_always_inline void fillPlane(uint8_t* plane, int stride, |
|
|
|
|
int width, int height, |
|
|
|
|
int y, uint8_t val) |
|
|
|
|
DECLARE_ALIGNED(8, const uint8_t, ff_sws_pb_64)[8] = { |
|
|
|
|
64, 64, 64, 64, 64, 64, 64, 64 |
|
|
|
|
}; |
|
|
|
|
|
|
|
|
|
static av_always_inline void fillPlane(uint8_t *plane, int stride, int width, |
|
|
|
|
int height, int y, uint8_t val) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
|
uint8_t *ptr = plane + stride * y; |
|
|
|
@ -59,8 +61,8 @@ static av_always_inline void fillPlane(uint8_t* plane, int stride, |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void hScale16To19_c(SwsContext *c, int16_t *_dst, int dstW, const uint8_t *_src, |
|
|
|
|
const int16_t *filter, |
|
|
|
|
static void hScale16To19_c(SwsContext *c, int16_t *_dst, int dstW, |
|
|
|
|
const uint8_t *_src, const int16_t *filter, |
|
|
|
|
const int32_t *filterPos, int filterSize) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
@ -82,8 +84,8 @@ static void hScale16To19_c(SwsContext *c, int16_t *_dst, int dstW, const uint8_t |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void hScale16To15_c(SwsContext *c, int16_t *dst, int dstW, const uint8_t *_src, |
|
|
|
|
const int16_t *filter, |
|
|
|
|
static void hScale16To15_c(SwsContext *c, int16_t *dst, int dstW, |
|
|
|
|
const uint8_t *_src, const int16_t *filter, |
|
|
|
|
const int32_t *filterPos, int filterSize) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
@ -104,9 +106,9 @@ static void hScale16To15_c(SwsContext *c, int16_t *dst, int dstW, const uint8_t |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
// bilinear / bicubic scaling
|
|
|
|
|
static void hScale8To15_c(SwsContext *c, int16_t *dst, int dstW, const uint8_t *src, |
|
|
|
|
const int16_t *filter, const int32_t *filterPos, |
|
|
|
|
int filterSize) |
|
|
|
|
static void hScale8To15_c(SwsContext *c, int16_t *dst, int dstW, |
|
|
|
|
const uint8_t *src, const int16_t *filter, |
|
|
|
|
const int32_t *filterPos, int filterSize) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
|
for (i = 0; i < dstW; i++) { |
|
|
|
@ -116,15 +118,13 @@ static void hScale8To15_c(SwsContext *c, int16_t *dst, int dstW, const uint8_t * |
|
|
|
|
for (j = 0; j < filterSize; j++) { |
|
|
|
|
val += ((int)src[srcPos + j]) * filter[filterSize * i + j]; |
|
|
|
|
} |
|
|
|
|
//filter += hFilterSize;
|
|
|
|
|
dst[i] = FFMIN(val >> 7, (1 << 15) - 1); // the cubic equation does overflow ...
|
|
|
|
|
//dst[i] = val>>7;
|
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void hScale8To19_c(SwsContext *c, int16_t *_dst, int dstW, const uint8_t *src, |
|
|
|
|
const int16_t *filter, const int32_t *filterPos, |
|
|
|
|
int filterSize) |
|
|
|
|
static void hScale8To19_c(SwsContext *c, int16_t *_dst, int dstW, |
|
|
|
|
const uint8_t *src, const int16_t *filter, |
|
|
|
|
const int32_t *filterPos, int filterSize) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
|
int32_t *dst = (int32_t *) _dst; |
|
|
|
@ -135,9 +135,7 @@ static void hScale8To19_c(SwsContext *c, int16_t *_dst, int dstW, const uint8_t |
|
|
|
|
for (j = 0; j < filterSize; j++) { |
|
|
|
|
val += ((int)src[srcPos + j]) * filter[filterSize * i + j]; |
|
|
|
|
} |
|
|
|
|
//filter += hFilterSize;
|
|
|
|
|
dst[i] = FFMIN(val >> 3, (1 << 19) - 1); // the cubic equation does overflow ...
|
|
|
|
|
//dst[i] = val>>7;
|
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
@ -151,6 +149,7 @@ static void chrRangeToJpeg_c(int16_t *dstU, int16_t *dstV, int width) |
|
|
|
|
dstV[i] = (FFMIN(dstV[i], 30775) * 4663 - 9289992) >> 12; // -264
|
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void chrRangeFromJpeg_c(int16_t *dstU, int16_t *dstV, int width) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
@ -159,12 +158,14 @@ static void chrRangeFromJpeg_c(int16_t *dstU, int16_t *dstV, int width) |
|
|
|
|
dstV[i] = (dstV[i] * 1799 + 4081085) >> 11; // 1469
|
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void lumRangeToJpeg_c(int16_t *dst, int width) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
|
for (i = 0; i < width; i++) |
|
|
|
|
dst[i] = (FFMIN(dst[i], 30189) * 19077 - 39057361) >> 14; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void lumRangeFromJpeg_c(int16_t *dst, int width) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
@ -182,6 +183,7 @@ static void chrRangeToJpeg16_c(int16_t *_dstU, int16_t *_dstV, int width) |
|
|
|
|
dstV[i] = (FFMIN(dstV[i], 30775 << 4) * 4663 - (9289992 << 4)) >> 12; // -264
|
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void chrRangeFromJpeg16_c(int16_t *_dstU, int16_t *_dstV, int width) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
@ -192,6 +194,7 @@ static void chrRangeFromJpeg16_c(int16_t *_dstU, int16_t *_dstV, int width) |
|
|
|
|
dstV[i] = (dstV[i] * 1799 + (4081085 << 4)) >> 11; // 1469
|
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void lumRangeToJpeg16_c(int16_t *_dst, int width) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
@ -199,6 +202,7 @@ static void lumRangeToJpeg16_c(int16_t *_dst, int width) |
|
|
|
|
for (i = 0; i < width; i++) |
|
|
|
|
dst[i] = (FFMIN(dst[i], 30189 << 4) * 4769 - (39057361 << 2)) >> 12; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static void lumRangeFromJpeg16_c(int16_t *_dst, int width) |
|
|
|
|
{ |
|
|
|
|
int i; |
|
|
|
@ -222,13 +226,16 @@ static void hyscale_fast_c(SwsContext *c, int16_t *dst, int dstWidth, |
|
|
|
|
|
|
|
|
|
// *** horizontal scale Y line to temp buffer
|
|
|
|
|
static av_always_inline void hyscale(SwsContext *c, int16_t *dst, int dstWidth, |
|
|
|
|
const uint8_t *src_in[4], int srcW, int xInc, |
|
|
|
|
const uint8_t *src_in[4], |
|
|
|
|
int srcW, int xInc, |
|
|
|
|
const int16_t *hLumFilter, |
|
|
|
|
const int32_t *hLumFilterPos, int hLumFilterSize, |
|
|
|
|
const int32_t *hLumFilterPos, |
|
|
|
|
int hLumFilterSize, |
|
|
|
|
uint8_t *formatConvBuffer, |
|
|
|
|
uint32_t *pal, int isAlpha) |
|
|
|
|
{ |
|
|
|
|
void (*toYV12)(uint8_t *, const uint8_t *, int, uint32_t *) = isAlpha ? c->alpToYV12 : c->lumToYV12; |
|
|
|
|
void (*toYV12)(uint8_t *, const uint8_t *, int, uint32_t *) = |
|
|
|
|
isAlpha ? c->alpToYV12 : c->lumToYV12; |
|
|
|
|
void (*convertRange)(int16_t *, int) = isAlpha ? NULL : c->lumConvertRange; |
|
|
|
|
const uint8_t *src = src_in[isAlpha ? 3 : 0]; |
|
|
|
|
|
|
|
|
@ -241,7 +248,8 @@ static av_always_inline void hyscale(SwsContext *c, int16_t *dst, int dstWidth, |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
if (!c->hyscale_fast) { |
|
|
|
|
c->hyScale(c, dst, dstWidth, src, hLumFilter, hLumFilterPos, hLumFilterSize); |
|
|
|
|
c->hyScale(c, dst, dstWidth, src, hLumFilter, |
|
|
|
|
hLumFilterPos, hLumFilterSize); |
|
|
|
|
} else { // fast bilinear upscale / crap downscale
|
|
|
|
|
c->hyscale_fast(c, dst, dstWidth, src, srcW, xInc); |
|
|
|
|
} |
|
|
|
@ -265,20 +273,25 @@ static void hcscale_fast_c(SwsContext *c, int16_t *dst1, int16_t *dst2, |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
static av_always_inline void hcscale(SwsContext *c, int16_t *dst1, int16_t *dst2, int dstWidth, |
|
|
|
|
static av_always_inline void hcscale(SwsContext *c, int16_t *dst1, |
|
|
|
|
int16_t *dst2, int dstWidth, |
|
|
|
|
const uint8_t *src_in[4], |
|
|
|
|
int srcW, int xInc, const int16_t *hChrFilter, |
|
|
|
|
const int32_t *hChrFilterPos, int hChrFilterSize, |
|
|
|
|
int srcW, int xInc, |
|
|
|
|
const int16_t *hChrFilter, |
|
|
|
|
const int32_t *hChrFilterPos, |
|
|
|
|
int hChrFilterSize, |
|
|
|
|
uint8_t *formatConvBuffer, uint32_t *pal) |
|
|
|
|
{ |
|
|
|
|
const uint8_t *src1 = src_in[1], *src2 = src_in[2]; |
|
|
|
|
if (c->chrToYV12) { |
|
|
|
|
uint8_t *buf2 = formatConvBuffer + FFALIGN(srcW * FFALIGN(c->srcBpc, 8) >> 3, 16); |
|
|
|
|
uint8_t *buf2 = formatConvBuffer + |
|
|
|
|
FFALIGN(srcW * FFALIGN(c->srcBpc, 8) >> 3, 16); |
|
|
|
|
c->chrToYV12(formatConvBuffer, buf2, src1, src2, srcW, pal); |
|
|
|
|
src1 = formatConvBuffer; |
|
|
|
|
src2 = buf2; |
|
|
|
|
} else if (c->readChrPlanar) { |
|
|
|
|
uint8_t *buf2 = formatConvBuffer + FFALIGN(srcW * FFALIGN(c->srcBpc, 8) >> 3, 16); |
|
|
|
|
uint8_t *buf2 = formatConvBuffer + |
|
|
|
|
FFALIGN(srcW * FFALIGN(c->srcBpc, 8) >> 3, 16); |
|
|
|
|
c->readChrPlanar(formatConvBuffer, buf2, src_in, srcW); |
|
|
|
|
src1 = formatConvBuffer; |
|
|
|
|
src2 = buf2; |
|
|
|
@ -296,13 +309,16 @@ static av_always_inline void hcscale(SwsContext *c, int16_t *dst1, int16_t *dst2 |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
#define DEBUG_SWSCALE_BUFFERS 0 |
|
|
|
|
#define DEBUG_BUFFERS(...) if (DEBUG_SWSCALE_BUFFERS) av_log(c, AV_LOG_DEBUG, __VA_ARGS__) |
|
|
|
|
#define DEBUG_BUFFERS(...) \ |
|
|
|
|
if (DEBUG_SWSCALE_BUFFERS) \
|
|
|
|
|
av_log(c, AV_LOG_DEBUG, __VA_ARGS__) |
|
|
|
|
|
|
|
|
|
static int swScale(SwsContext *c, const uint8_t *src[], |
|
|
|
|
int srcStride[], int srcSliceY, |
|
|
|
|
int srcSliceH, uint8_t *dst[], int dstStride[]) |
|
|
|
|
{ |
|
|
|
|
/* load a few things into local vars to make the code more readable? and faster */ |
|
|
|
|
/* load a few things into local vars to make the code more readable?
|
|
|
|
|
* and faster */ |
|
|
|
|
const int srcW = c->srcW; |
|
|
|
|
const int dstW = c->dstW; |
|
|
|
|
const int dstH = c->dstH; |
|
|
|
@ -333,9 +349,6 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
const int vLumBufSize = c->vLumBufSize; |
|
|
|
|
const int vChrBufSize = c->vChrBufSize; |
|
|
|
|
uint8_t *formatConvBuffer = c->formatConvBuffer; |
|
|
|
|
const int chrSrcSliceY= srcSliceY >> c->chrSrcVSubSample; |
|
|
|
|
const int chrSrcSliceH= -((-srcSliceH) >> c->chrSrcVSubSample); |
|
|
|
|
int lastDstY; |
|
|
|
|
uint32_t *pal = c->pal_yuv; |
|
|
|
|
yuv2planar1_fn yuv2plane1 = c->yuv2plane1; |
|
|
|
|
yuv2planarX_fn yuv2planeX = c->yuv2planeX; |
|
|
|
@ -343,7 +356,11 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
yuv2packed1_fn yuv2packed1 = c->yuv2packed1; |
|
|
|
|
yuv2packed2_fn yuv2packed2 = c->yuv2packed2; |
|
|
|
|
yuv2packedX_fn yuv2packedX = c->yuv2packedX; |
|
|
|
|
int should_dither = is9_OR_10BPS(c->srcFormat) || is16BPS(c->srcFormat); |
|
|
|
|
const int chrSrcSliceY = srcSliceY >> c->chrSrcVSubSample; |
|
|
|
|
const int chrSrcSliceH = -((-srcSliceH) >> c->chrSrcVSubSample); |
|
|
|
|
int should_dither = is9_OR_10BPS(c->srcFormat) || |
|
|
|
|
is16BPS(c->srcFormat); |
|
|
|
|
int lastDstY; |
|
|
|
|
|
|
|
|
|
/* vars which will change and which we need to store back in the context */ |
|
|
|
|
int dstY = c->dstY; |
|
|
|
@ -366,25 +383,29 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
srcStride[2] <<= c->vChrDrop; |
|
|
|
|
|
|
|
|
|
DEBUG_BUFFERS("swScale() %p[%d] %p[%d] %p[%d] %p[%d] -> %p[%d] %p[%d] %p[%d] %p[%d]\n", |
|
|
|
|
src[0], srcStride[0], src[1], srcStride[1], src[2], srcStride[2], src[3], srcStride[3], |
|
|
|
|
dst[0], dstStride[0], dst[1], dstStride[1], dst[2], dstStride[2], dst[3], dstStride[3]); |
|
|
|
|
src[0], srcStride[0], src[1], srcStride[1], |
|
|
|
|
src[2], srcStride[2], src[3], srcStride[3], |
|
|
|
|
dst[0], dstStride[0], dst[1], dstStride[1], |
|
|
|
|
dst[2], dstStride[2], dst[3], dstStride[3]); |
|
|
|
|
DEBUG_BUFFERS("srcSliceY: %d srcSliceH: %d dstY: %d dstH: %d\n", |
|
|
|
|
srcSliceY, srcSliceH, dstY, dstH); |
|
|
|
|
DEBUG_BUFFERS("vLumFilterSize: %d vLumBufSize: %d vChrFilterSize: %d vChrBufSize: %d\n", |
|
|
|
|
vLumFilterSize, vLumBufSize, vChrFilterSize, vChrBufSize); |
|
|
|
|
|
|
|
|
|
if (dstStride[0]%8 !=0 || dstStride[1]%8 !=0 || dstStride[2]%8 !=0 || dstStride[3]%8 != 0) { |
|
|
|
|
static int warnedAlready=0; //FIXME move this into the context perhaps
|
|
|
|
|
if (dstStride[0] % 8 != 0 || dstStride[1] % 8 != 0 || |
|
|
|
|
dstStride[2] % 8 != 0 || dstStride[3] % 8 != 0) { |
|
|
|
|
static int warnedAlready = 0; // FIXME maybe move this into the context
|
|
|
|
|
if (flags & SWS_PRINT_INFO && !warnedAlready) { |
|
|
|
|
av_log(c, AV_LOG_WARNING, "Warning: dstStride is not aligned!\n" |
|
|
|
|
av_log(c, AV_LOG_WARNING, |
|
|
|
|
"Warning: dstStride is not aligned!\n" |
|
|
|
|
" ->cannot do aligned memory accesses anymore\n"); |
|
|
|
|
warnedAlready = 1; |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
/* Note the user might start scaling the picture in the middle so this
|
|
|
|
|
will not get executed. This is not really intended but works |
|
|
|
|
currently, so people might do it. */ |
|
|
|
|
* will not get executed. This is not really intended but works |
|
|
|
|
* currently, so people might do it. */ |
|
|
|
|
if (srcSliceY == 0) { |
|
|
|
|
lumBufIndex = -1; |
|
|
|
|
chrBufIndex = -1; |
|
|
|
@ -407,9 +428,11 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
(CONFIG_SWSCALE_ALPHA && alpPixBuf) ? dst[3] + dstStride[3] * dstY : NULL, |
|
|
|
|
}; |
|
|
|
|
|
|
|
|
|
const int firstLumSrcY= FFMAX(1 - vLumFilterSize, vLumFilterPos[dstY]); //First line needed as input
|
|
|
|
|
// First line needed as input
|
|
|
|
|
const int firstLumSrcY = FFMAX(1 - vLumFilterSize, vLumFilterPos[dstY]); |
|
|
|
|
const int firstLumSrcY2 = FFMAX(1 - vLumFilterSize, vLumFilterPos[FFMIN(dstY | ((1 << c->chrDstVSubSample) - 1), dstH - 1)]); |
|
|
|
|
const int firstChrSrcY= FFMAX(1 - vChrFilterSize, vChrFilterPos[chrDstY]); //First line needed as input
|
|
|
|
|
// First line needed as input
|
|
|
|
|
const int firstChrSrcY = FFMAX(1 - vChrFilterSize, vChrFilterPos[chrDstY]); |
|
|
|
|
|
|
|
|
|
// Last line needed as input
|
|
|
|
|
int lastLumSrcY = FFMIN(c->srcH, firstLumSrcY + vLumFilterSize) - 1; |
|
|
|
@ -418,8 +441,10 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
int enough_lines; |
|
|
|
|
|
|
|
|
|
// handle holes (FAST_BILINEAR & weird filters)
|
|
|
|
|
if (firstLumSrcY > lastInLumBuf) lastInLumBuf= firstLumSrcY-1; |
|
|
|
|
if (firstChrSrcY > lastInChrBuf) lastInChrBuf= firstChrSrcY-1; |
|
|
|
|
if (firstLumSrcY > lastInLumBuf) |
|
|
|
|
lastInLumBuf = firstLumSrcY - 1; |
|
|
|
|
if (firstChrSrcY > lastInChrBuf) |
|
|
|
|
lastInChrBuf = firstChrSrcY - 1; |
|
|
|
|
assert(firstLumSrcY >= lastInLumBuf - vLumBufSize + 1); |
|
|
|
|
assert(firstChrSrcY >= lastInChrBuf - vChrBufSize + 1); |
|
|
|
|
|
|
|
|
@ -430,7 +455,8 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
firstChrSrcY, lastChrSrcY, lastInChrBuf); |
|
|
|
|
|
|
|
|
|
// Do we have enough lines in this slice to output the dstY line
|
|
|
|
|
enough_lines = lastLumSrcY2 < srcSliceY + srcSliceH && lastChrSrcY < -((-srcSliceY - srcSliceH)>>c->chrSrcVSubSample); |
|
|
|
|
enough_lines = lastLumSrcY2 < srcSliceY + srcSliceH && |
|
|
|
|
lastChrSrcY < -((-srcSliceY - srcSliceH) >> c->chrSrcVSubSample); |
|
|
|
|
|
|
|
|
|
if (!enough_lines) { |
|
|
|
|
lastLumSrcY = srcSliceY + srcSliceH - 1; |
|
|
|
@ -453,13 +479,11 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
assert(lastInLumBuf + 1 - srcSliceY >= 0); |
|
|
|
|
hyscale(c, lumPixBuf[lumBufIndex], dstW, src1, srcW, lumXInc, |
|
|
|
|
hLumFilter, hLumFilterPos, hLumFilterSize, |
|
|
|
|
formatConvBuffer, |
|
|
|
|
pal, 0); |
|
|
|
|
formatConvBuffer, pal, 0); |
|
|
|
|
if (CONFIG_SWSCALE_ALPHA && alpPixBuf) |
|
|
|
|
hyscale(c, alpPixBuf[lumBufIndex], dstW, src1, srcW, |
|
|
|
|
lumXInc, hLumFilter, hLumFilterPos, hLumFilterSize, |
|
|
|
|
formatConvBuffer, |
|
|
|
|
pal, 1); |
|
|
|
|
formatConvBuffer, pal, 1); |
|
|
|
|
lastInLumBuf++; |
|
|
|
|
DEBUG_BUFFERS("\t\tlumBufIndex %d: lastInLumBuf: %d\n", |
|
|
|
|
lumBufIndex, lastInLumBuf); |
|
|
|
@ -487,20 +511,24 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
chrBufIndex, lastInChrBuf); |
|
|
|
|
} |
|
|
|
|
// wrap buf index around to stay inside the ring buffer
|
|
|
|
|
if (lumBufIndex >= vLumBufSize) lumBufIndex-= vLumBufSize; |
|
|
|
|
if (chrBufIndex >= vChrBufSize) chrBufIndex-= vChrBufSize; |
|
|
|
|
if (lumBufIndex >= vLumBufSize) |
|
|
|
|
lumBufIndex -= vLumBufSize; |
|
|
|
|
if (chrBufIndex >= vChrBufSize) |
|
|
|
|
chrBufIndex -= vChrBufSize; |
|
|
|
|
if (!enough_lines) |
|
|
|
|
break; // we can't output a dstY line so let's try with the next slice
|
|
|
|
|
|
|
|
|
|
#if HAVE_MMX |
|
|
|
|
updateMMXDitherTables(c, dstY, lumBufIndex, chrBufIndex, lastInLumBuf, lastInChrBuf); |
|
|
|
|
updateMMXDitherTables(c, dstY, lumBufIndex, chrBufIndex, |
|
|
|
|
lastInLumBuf, lastInChrBuf); |
|
|
|
|
#endif |
|
|
|
|
if (should_dither) { |
|
|
|
|
c->chrDither8 = dither_8x8_128[chrDstY & 7]; |
|
|
|
|
c->lumDither8 = dither_8x8_128[dstY & 7]; |
|
|
|
|
} |
|
|
|
|
if (dstY >= dstH - 2) { |
|
|
|
|
// hmm looks like we can't use MMX here without overwriting this array's tail
|
|
|
|
|
/* hmm looks like we can't use MMX here without overwriting
|
|
|
|
|
* this array's tail */ |
|
|
|
|
ff_sws_init_output_funcs(c, &yuv2plane1, &yuv2planeX, &yuv2nv12cX, |
|
|
|
|
&yuv2packed1, &yuv2packed2, &yuv2packedX); |
|
|
|
|
} |
|
|
|
@ -509,11 +537,14 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
const int16_t **lumSrcPtr = (const int16_t **)lumPixBuf + lumBufIndex + firstLumSrcY - lastInLumBuf + vLumBufSize; |
|
|
|
|
const int16_t **chrUSrcPtr = (const int16_t **)chrUPixBuf + chrBufIndex + firstChrSrcY - lastInChrBuf + vChrBufSize; |
|
|
|
|
const int16_t **chrVSrcPtr = (const int16_t **)chrVPixBuf + chrBufIndex + firstChrSrcY - lastInChrBuf + vChrBufSize; |
|
|
|
|
const int16_t **alpSrcPtr= (CONFIG_SWSCALE_ALPHA && alpPixBuf) ? (const int16_t **) alpPixBuf + lumBufIndex + firstLumSrcY - lastInLumBuf + vLumBufSize : NULL; |
|
|
|
|
const int16_t **alpSrcPtr = (CONFIG_SWSCALE_ALPHA && alpPixBuf) ? |
|
|
|
|
(const int16_t **)alpPixBuf + lumBufIndex + firstLumSrcY - lastInLumBuf + vLumBufSize : NULL; |
|
|
|
|
|
|
|
|
|
if (firstLumSrcY < 0 || firstLumSrcY + vLumFilterSize > c->srcH) { |
|
|
|
|
const int16_t **tmpY = (const int16_t **) lumPixBuf + 2 * vLumBufSize; |
|
|
|
|
int neg = -firstLumSrcY, i, end = FFMIN(c->srcH - firstLumSrcY, vLumFilterSize); |
|
|
|
|
const int16_t **tmpY = (const int16_t **)lumPixBuf + |
|
|
|
|
2 * vLumBufSize; |
|
|
|
|
int neg = -firstLumSrcY, i; |
|
|
|
|
int end = FFMIN(c->srcH - firstLumSrcY, vLumFilterSize); |
|
|
|
|
for (i = 0; i < neg; i++) |
|
|
|
|
tmpY[i] = lumSrcPtr[neg]; |
|
|
|
|
for (; i < end; i++) |
|
|
|
@ -523,7 +554,8 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
lumSrcPtr = tmpY; |
|
|
|
|
|
|
|
|
|
if (alpSrcPtr) { |
|
|
|
|
const int16_t **tmpA = (const int16_t **) alpPixBuf + 2 * vLumBufSize; |
|
|
|
|
const int16_t **tmpA = (const int16_t **)alpPixBuf + |
|
|
|
|
2 * vLumBufSize; |
|
|
|
|
for (i = 0; i < neg; i++) |
|
|
|
|
tmpA[i] = alpSrcPtr[neg]; |
|
|
|
|
for (; i < end; i++) |
|
|
|
@ -533,10 +565,12 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
alpSrcPtr = tmpA; |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
if (firstChrSrcY < 0 || firstChrSrcY + vChrFilterSize > c->chrSrcH) { |
|
|
|
|
if (firstChrSrcY < 0 || |
|
|
|
|
firstChrSrcY + vChrFilterSize > c->chrSrcH) { |
|
|
|
|
const int16_t **tmpU = (const int16_t **)chrUPixBuf + 2 * vChrBufSize, |
|
|
|
|
**tmpV = (const int16_t **)chrVPixBuf + 2 * vChrBufSize; |
|
|
|
|
int neg = -firstChrSrcY, i, end = FFMIN(c->chrSrcH - firstChrSrcY, vChrFilterSize); |
|
|
|
|
int neg = -firstChrSrcY, i; |
|
|
|
|
int end = FFMIN(c->chrSrcH - firstChrSrcY, vChrFilterSize); |
|
|
|
|
for (i = 0; i < neg; i++) { |
|
|
|
|
tmpU[i] = chrUSrcPtr[neg]; |
|
|
|
|
tmpV[i] = chrVSrcPtr[neg]; |
|
|
|
@ -553,47 +587,57 @@ static int swScale(SwsContext *c, const uint8_t* src[], |
|
|
|
|
chrVSrcPtr = tmpV; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
if (isPlanarYUV(dstFormat) || (isGray(dstFormat) && !isALPHA(dstFormat))) { //YV12 like
|
|
|
|
|
if (isPlanarYUV(dstFormat) || |
|
|
|
|
(isGray(dstFormat) && !isALPHA(dstFormat))) { // YV12 like
|
|
|
|
|
const int chrSkipMask = (1 << c->chrDstVSubSample) - 1; |
|
|
|
|
|
|
|
|
|
if (vLumFilterSize == 1) { |
|
|
|
|
yuv2plane1(lumSrcPtr[0], dest[0], dstW, c->lumDither8, 0); |
|
|
|
|
} else { |
|
|
|
|
yuv2planeX(vLumFilter + dstY * vLumFilterSize, vLumFilterSize, |
|
|
|
|
lumSrcPtr, dest[0], dstW, c->lumDither8, 0); |
|
|
|
|
yuv2planeX(vLumFilter + dstY * vLumFilterSize, |
|
|
|
|
vLumFilterSize, lumSrcPtr, dest[0], |
|
|
|
|
dstW, c->lumDither8, 0); |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
if (!((dstY & chrSkipMask) || isGray(dstFormat))) { |
|
|
|
|
if (yuv2nv12cX) { |
|
|
|
|
yuv2nv12cX(c, vChrFilter + chrDstY * vChrFilterSize, vChrFilterSize, chrUSrcPtr, chrVSrcPtr, dest[1], chrDstW); |
|
|
|
|
yuv2nv12cX(c, vChrFilter + chrDstY * vChrFilterSize, |
|
|
|
|
vChrFilterSize, chrUSrcPtr, chrVSrcPtr, |
|
|
|
|
dest[1], chrDstW); |
|
|
|
|
} else if (vChrFilterSize == 1) { |
|
|
|
|
yuv2plane1(chrUSrcPtr[0], dest[1], chrDstW, c->chrDither8, 0); |
|
|
|
|
yuv2plane1(chrVSrcPtr[0], dest[2], chrDstW, c->chrDither8, 3); |
|
|
|
|
} else { |
|
|
|
|
yuv2planeX(vChrFilter + chrDstY * vChrFilterSize, vChrFilterSize, |
|
|
|
|
chrUSrcPtr, dest[1], chrDstW, c->chrDither8, 0); |
|
|
|
|
yuv2planeX(vChrFilter + chrDstY * vChrFilterSize, vChrFilterSize, |
|
|
|
|
chrVSrcPtr, dest[2], chrDstW, c->chrDither8, 3); |
|
|
|
|
yuv2planeX(vChrFilter + chrDstY * vChrFilterSize, |
|
|
|
|
vChrFilterSize, chrUSrcPtr, dest[1], |
|
|
|
|
chrDstW, c->chrDither8, 0); |
|
|
|
|
yuv2planeX(vChrFilter + chrDstY * vChrFilterSize, |
|
|
|
|
vChrFilterSize, chrVSrcPtr, dest[2], |
|
|
|
|
chrDstW, c->chrDither8, 3); |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
if (CONFIG_SWSCALE_ALPHA && alpPixBuf) { |
|
|
|
|
if (vLumFilterSize == 1) { |
|
|
|
|
yuv2plane1(alpSrcPtr[0], dest[3], dstW, c->lumDither8, 0); |
|
|
|
|
yuv2plane1(alpSrcPtr[0], dest[3], dstW, |
|
|
|
|
c->lumDither8, 0); |
|
|
|
|
} else { |
|
|
|
|
yuv2planeX(vLumFilter + dstY * vLumFilterSize, vLumFilterSize, |
|
|
|
|
alpSrcPtr, dest[3], dstW, c->lumDither8, 0); |
|
|
|
|
yuv2planeX(vLumFilter + dstY * vLumFilterSize, |
|
|
|
|
vLumFilterSize, alpSrcPtr, dest[3], |
|
|
|
|
dstW, c->lumDither8, 0); |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|
} else { |
|
|
|
|
assert(lumSrcPtr + vLumFilterSize - 1 < lumPixBuf + vLumBufSize * 2); |
|
|
|
|
assert(chrUSrcPtr + vChrFilterSize - 1 < chrUPixBuf + vChrBufSize * 2); |
|
|
|
|
if (c->yuv2packed1 && vLumFilterSize == 1 && vChrFilterSize <= 2) { //unscaled RGB
|
|
|
|
|
if (c->yuv2packed1 && vLumFilterSize == 1 && |
|
|
|
|
vChrFilterSize <= 2) { // unscaled RGB
|
|
|
|
|
int chrAlpha = vChrFilterSize == 1 ? 0 : vChrFilter[2 * dstY + 1]; |
|
|
|
|
yuv2packed1(c, *lumSrcPtr, chrUSrcPtr, chrVSrcPtr, |
|
|
|
|
alpPixBuf ? *alpSrcPtr : NULL, |
|
|
|
|
dest[0], dstW, chrAlpha, dstY); |
|
|
|
|
} else if (c->yuv2packed2 && vLumFilterSize == 2 && vChrFilterSize == 2) { //bilinear upscale RGB
|
|
|
|
|
} else if (c->yuv2packed2 && vLumFilterSize == 2 && |
|
|
|
|
vChrFilterSize == 2) { // bilinear upscale RGB
|
|
|
|
|
int lumAlpha = vLumFilter[2 * dstY + 1]; |
|
|
|
|
int chrAlpha = vChrFilter[2 * dstY + 1]; |
|
|
|
|
lumMmxFilter[2] = |
|
|
|
@ -654,7 +698,8 @@ static av_cold void sws_init_swScale_c(SwsContext *c) |
|
|
|
|
c->hyScale = c->hcScale = hScale8To19_c; |
|
|
|
|
} |
|
|
|
|
} else { |
|
|
|
|
c->hyScale = c->hcScale = c->dstBpc > 10 ? hScale16To19_c : hScale16To15_c; |
|
|
|
|
c->hyScale = c->hcScale = c->dstBpc > 10 ? hScale16To19_c |
|
|
|
|
: hScale16To15_c; |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
if (c->srcRange != c->dstRange && !isAnyRGB(c->dstFormat)) { |
|
|
|
|