FreeRDP/libfreerdp/primitives/prim_YUV.c
2024-08-29 15:34:22 +02:00

1885 lines
51 KiB
C

/**
* FreeRDP: A Remote Desktop Protocol Implementation
* Generic YUV/RGB conversion operations
*
* Copyright 2014 Marc-Andre Moreau <marcandre.moreau@gmail.com>
* Copyright 2015-2017 Armin Novak <armin.novak@thincast.com>
* Copyright 2015-2017 Norbert Federa <norbert.federa@thincast.com>
* Copyright 2015-2017 Vic Lee
* Copyright 2015-2017 Thincast Technologies GmbH
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include <winpr/wtypes.h>
#include <freerdp/config.h>
#include <freerdp/types.h>
#include <freerdp/primitives.h>
#include <freerdp/codec/color.h>
#include "prim_internal.h"
#include "prim_YUV.h"
static pstatus_t general_LumaToYUV444(const BYTE* const WINPR_RESTRICT pSrcRaw[3],
const UINT32 srcStep[3], BYTE* WINPR_RESTRICT pDstRaw[3],
const UINT32 dstStep[3],
const RECTANGLE_16* WINPR_RESTRICT roi)
{
const UINT32 nWidth = roi->right - roi->left;
const UINT32 nHeight = roi->bottom - roi->top;
const UINT32 halfWidth = (nWidth + 1) / 2;
const UINT32 halfHeight = (nHeight + 1) / 2;
const UINT32 oddY = 1;
const UINT32 evenY = 0;
const UINT32 oddX = 1;
const UINT32 evenX = 0;
const BYTE* pSrc[3] = { pSrcRaw[0] + 1ULL * roi->top * srcStep[0] + roi->left,
pSrcRaw[1] + 1ULL * roi->top / 2 * srcStep[1] + roi->left / 2,
pSrcRaw[2] + 1ULL * roi->top / 2 * srcStep[2] + roi->left / 2 };
BYTE* pDst[3] = { pDstRaw[0] + 1ULL * roi->top * dstStep[0] + roi->left,
pDstRaw[1] + 1ULL * roi->top * dstStep[1] + roi->left,
pDstRaw[2] + 1ULL * roi->top * dstStep[2] + roi->left };
/* Y data is already here... */
/* B1 */
for (size_t y = 0; y < nHeight; y++)
{
const BYTE* Ym = pSrc[0] + y * srcStep[0];
BYTE* pY = pDst[0] + dstStep[0] * y;
memcpy(pY, Ym, nWidth);
}
/* The first half of U, V are already here part of this frame. */
/* B2 and B3 */
for (size_t y = 0; y < halfHeight; y++)
{
const UINT32 val2y = (2ULL * y + evenY);
const UINT32 val2y1 = val2y + oddY;
const BYTE* Um = pSrc[1] + y * srcStep[1];
const BYTE* Vm = pSrc[2] + y * srcStep[2];
BYTE* pU = pDst[1] + 1ULL * dstStep[1] * val2y;
BYTE* pV = pDst[2] + 1ULL * dstStep[2] * val2y;
BYTE* pU1 = pDst[1] + 1ULL * dstStep[1] * val2y1;
BYTE* pV1 = pDst[2] + 1ULL * dstStep[2] * val2y1;
for (size_t x = 0; x < halfWidth; x++)
{
const UINT32 val2x = 2UL * x + evenX;
const UINT32 val2x1 = val2x + oddX;
pU[val2x] = Um[x];
pV[val2x] = Vm[x];
pU[val2x1] = Um[x];
pV[val2x1] = Vm[x];
pU1[val2x] = Um[x];
pV1[val2x] = Vm[x];
pU1[val2x1] = Um[x];
pV1[val2x1] = Vm[x];
}
}
return PRIMITIVES_SUCCESS;
}
static pstatus_t general_ChromaFilter(BYTE* WINPR_RESTRICT pDst[3], const UINT32 dstStep[3],
const RECTANGLE_16* WINPR_RESTRICT roi)
{
const UINT32 oddY = 1;
const UINT32 evenY = 0;
const UINT32 nWidth = roi->right - roi->left;
const UINT32 nHeight = roi->bottom - roi->top;
const UINT32 halfHeight = (nHeight + 1) / 2;
const UINT32 halfWidth = (nWidth + 1) / 2;
/* Filter */
for (UINT32 y = roi->top; y < halfHeight + roi->top; y++)
{
const UINT32 val2y = (y * 2 + evenY);
const UINT32 val2y1 = val2y + oddY;
BYTE* pU1 = pDst[1] + 1ULL * dstStep[1] * val2y1;
BYTE* pV1 = pDst[2] + 1ULL * dstStep[2] * val2y1;
BYTE* pU = pDst[1] + 1ULL * dstStep[1] * val2y;
BYTE* pV = pDst[2] + 1ULL * dstStep[2] * val2y;
if (val2y1 > nHeight)
continue;
for (UINT32 x = roi->left; x < halfWidth + roi->left; x++)
{
const UINT32 val2x = (x * 2);
const UINT32 val2x1 = val2x + 1;
const BYTE inU = pU[val2x];
const BYTE inV = pV[val2x];
const INT32 up = inU * 4;
const INT32 vp = inV * 4;
INT32 u2020 = 0;
INT32 v2020 = 0;
if (val2x1 > nWidth)
continue;
u2020 = up - pU[val2x1] - pU1[val2x] - pU1[val2x1];
v2020 = vp - pV[val2x1] - pV1[val2x] - pV1[val2x1];
pU[val2x] = CONDITIONAL_CLIP(u2020, inU);
pV[val2x] = CONDITIONAL_CLIP(v2020, inV);
}
}
return PRIMITIVES_SUCCESS;
}
static pstatus_t general_ChromaV1ToYUV444(const BYTE* const WINPR_RESTRICT pSrcRaw[3],
const UINT32 srcStep[3], BYTE* WINPR_RESTRICT pDstRaw[3],
const UINT32 dstStep[3],
const RECTANGLE_16* WINPR_RESTRICT roi)
{
const UINT32 mod = 16;
UINT32 uY = 0;
UINT32 vY = 0;
const UINT32 nWidth = roi->right - roi->left;
const UINT32 nHeight = roi->bottom - roi->top;
const UINT32 halfWidth = (nWidth) / 2;
const UINT32 halfHeight = (nHeight) / 2;
const UINT32 oddY = 1;
const UINT32 evenY = 0;
const UINT32 oddX = 1;
/* The auxilary frame is aligned to multiples of 16x16.
* We need the padded height for B4 and B5 conversion. */
const UINT32 padHeigth = nHeight + 16 - nHeight % 16;
const BYTE* pSrc[3] = { pSrcRaw[0] + 1ULL * roi->top * srcStep[0] + roi->left,
pSrcRaw[1] + 1ULL * roi->top / 2 * srcStep[1] + roi->left / 2,
pSrcRaw[2] + 1ULL * roi->top / 2 * srcStep[2] + roi->left / 2 };
BYTE* pDst[3] = { pDstRaw[0] + 1ULL * roi->top * dstStep[0] + roi->left,
pDstRaw[1] + 1ULL * roi->top * dstStep[1] + roi->left,
pDstRaw[2] + 1ULL * roi->top * dstStep[2] + roi->left };
/* The second half of U and V is a bit more tricky... */
/* B4 and B5 */
for (size_t y = 0; y < padHeigth; y++)
{
const BYTE* Ya = pSrc[0] + y * srcStep[0];
BYTE* pX = NULL;
if ((y) % mod < (mod + 1) / 2)
{
const size_t pos = (2 * uY++ + oddY);
if (pos >= nHeight)
continue;
pX = pDst[1] + dstStep[1] * pos;
}
else
{
const size_t pos = (2 * vY++ + oddY);
if (pos >= nHeight)
continue;
pX = pDst[2] + dstStep[2] * pos;
}
memcpy(pX, Ya, nWidth);
}
/* B6 and B7 */
for (size_t y = 0; y < halfHeight; y++)
{
const UINT32 val2y = (y * 2ULL + evenY);
const BYTE* Ua = pSrc[1] + y * srcStep[1];
const BYTE* Va = pSrc[2] + y * srcStep[2];
BYTE* pU = pDst[1] + 1ULL * dstStep[1] * val2y;
BYTE* pV = pDst[2] + 1ULL * dstStep[2] * val2y;
for (size_t x = 0; x < halfWidth; x++)
{
const UINT32 val2x1 = (x * 2 + oddX);
pU[val2x1] = Ua[x];
pV[val2x1] = Va[x];
}
}
/* Filter */
return general_ChromaFilter(pDst, dstStep, roi);
}
static pstatus_t general_ChromaV2ToYUV444(const BYTE* const WINPR_RESTRICT pSrc[3],
const UINT32 srcStep[3], UINT32 nTotalWidth,
UINT32 nTotalHeight, BYTE* WINPR_RESTRICT pDst[3],
const UINT32 dstStep[3],
const RECTANGLE_16* WINPR_RESTRICT roi)
{
const UINT32 nWidth = roi->right - roi->left;
const UINT32 nHeight = roi->bottom - roi->top;
const UINT32 halfWidth = (nWidth + 1) / 2;
const UINT32 halfHeight = (nHeight + 1) / 2;
const UINT32 quaterWidth = (nWidth + 3) / 4;
/* B4 and B5: odd UV values for width/2, height */
for (UINT32 y = 0; y < nHeight; y++)
{
const UINT32 yTop = y + roi->top;
const BYTE* pYaU = pSrc[0] + 1ULL * srcStep[0] * yTop + roi->left / 2;
const BYTE* pYaV = pYaU + nTotalWidth / 2;
BYTE* pU = pDst[1] + 1ULL * dstStep[1] * yTop + roi->left;
BYTE* pV = pDst[2] + 1ULL * dstStep[2] * yTop + roi->left;
for (size_t x = 0; x < halfWidth; x++)
{
const UINT32 odd = 2 * x + 1;
pU[odd] = *pYaU++;
pV[odd] = *pYaV++;
}
}
/* B6 - B9 */
for (size_t y = 0; y < halfHeight; y++)
{
const BYTE* pUaU = pSrc[1] + srcStep[1] * (y + roi->top / 2) + roi->left / 4;
const BYTE* pUaV = pUaU + nTotalWidth / 4;
const BYTE* pVaU = pSrc[2] + srcStep[2] * (y + roi->top / 2) + roi->left / 4;
const BYTE* pVaV = pVaU + nTotalWidth / 4;
BYTE* pU = pDst[1] + 1ULL * dstStep[1] * (2ULL * y + 1 + roi->top) + roi->left;
BYTE* pV = pDst[2] + 1ULL * dstStep[2] * (2ULL * y + 1 + roi->top) + roi->left;
for (size_t x = 0; x < quaterWidth; x++)
{
pU[4 * x + 0] = *pUaU++;
pV[4 * x + 0] = *pUaV++;
pU[4 * x + 2] = *pVaU++;
pV[4 * x + 2] = *pVaV++;
}
}
return general_ChromaFilter(pDst, dstStep, roi);
}
static pstatus_t general_YUV420CombineToYUV444(avc444_frame_type type,
const BYTE* const WINPR_RESTRICT pSrc[3],
const UINT32 srcStep[3], UINT32 nWidth,
UINT32 nHeight, BYTE* WINPR_RESTRICT pDst[3],
const UINT32 dstStep[3],
const RECTANGLE_16* WINPR_RESTRICT roi)
{
if (!pSrc || !pSrc[0] || !pSrc[1] || !pSrc[2])
return -1;
if (!pDst || !pDst[0] || !pDst[1] || !pDst[2])
return -1;
if (!roi)
return -1;
switch (type)
{
case AVC444_LUMA:
return general_LumaToYUV444(pSrc, srcStep, pDst, dstStep, roi);
case AVC444_CHROMAv1:
return general_ChromaV1ToYUV444(pSrc, srcStep, pDst, dstStep, roi);
case AVC444_CHROMAv2:
return general_ChromaV2ToYUV444(pSrc, srcStep, nWidth, nHeight, pDst, dstStep, roi);
default:
return -1;
}
}
static pstatus_t
general_YUV444SplitToYUV420(const BYTE* const WINPR_RESTRICT pSrc[3], const UINT32 srcStep[3],
BYTE* WINPR_RESTRICT pMainDst[3], const UINT32 dstMainStep[3],
BYTE* WINPR_RESTRICT pAuxDst[3], const UINT32 dstAuxStep[3],
const prim_size_t* WINPR_RESTRICT roi)
{
UINT32 uY = 0;
UINT32 vY = 0;
UINT32 halfWidth = 0;
UINT32 halfHeight = 0;
/* The auxilary frame is aligned to multiples of 16x16.
* We need the padded height for B4 and B5 conversion. */
const UINT32 padHeigth = roi->height + 16 - roi->height % 16;
halfWidth = (roi->width + 1) / 2;
halfHeight = (roi->height + 1) / 2;
/* B1 */
for (size_t y = 0; y < roi->height; y++)
{
const BYTE* pSrcY = pSrc[0] + y * srcStep[0];
BYTE* pY = pMainDst[0] + y * dstMainStep[0];
memcpy(pY, pSrcY, roi->width);
}
/* B2 and B3 */
for (size_t y = 0; y < halfHeight; y++)
{
const BYTE* pSrcU = pSrc[1] + 2ULL * y * srcStep[1];
const BYTE* pSrcV = pSrc[2] + 2ULL * y * srcStep[2];
const BYTE* pSrcU1 = pSrc[1] + (2ULL * y + 1ULL) * srcStep[1];
const BYTE* pSrcV1 = pSrc[2] + (2ULL * y + 1ULL) * srcStep[2];
BYTE* pU = pMainDst[1] + y * dstMainStep[1];
BYTE* pV = pMainDst[2] + y * dstMainStep[2];
for (size_t x = 0; x < halfWidth; x++)
{
/* Filter */
const INT32 u = pSrcU[2 * x] + pSrcU[2 * x + 1] + pSrcU1[2 * x] + pSrcU1[2 * x + 1];
const INT32 v = pSrcV[2 * x] + pSrcV[2 * x + 1] + pSrcV1[2 * x] + pSrcV1[2 * x + 1];
pU[x] = CLIP(u / 4L);
pV[x] = CLIP(v / 4L);
}
}
/* B4 and B5 */
for (size_t y = 0; y < padHeigth; y++)
{
BYTE* pY = pAuxDst[0] + y * dstAuxStep[0];
if (y % 16 < 8)
{
const size_t pos = (2 * uY++ + 1);
const BYTE* pSrcU = pSrc[1] + pos * srcStep[1];
if (pos >= roi->height)
continue;
memcpy(pY, pSrcU, roi->width);
}
else
{
const size_t pos = (2 * vY++ + 1);
const BYTE* pSrcV = pSrc[2] + pos * srcStep[2];
if (pos >= roi->height)
continue;
memcpy(pY, pSrcV, roi->width);
}
}
/* B6 and B7 */
for (size_t y = 0; y < halfHeight; y++)
{
const BYTE* pSrcU = pSrc[1] + 2 * y * srcStep[1];
const BYTE* pSrcV = pSrc[2] + 2 * y * srcStep[2];
BYTE* pU = pAuxDst[1] + y * dstAuxStep[1];
BYTE* pV = pAuxDst[2] + y * dstAuxStep[2];
for (size_t x = 0; x < halfWidth; x++)
{
pU[x] = pSrcU[2 * x + 1];
pV[x] = pSrcV[2 * x + 1];
}
}
return PRIMITIVES_SUCCESS;
}
static pstatus_t general_YUV444ToRGB_8u_P3AC4R_general(const BYTE* const WINPR_RESTRICT pSrc[3],
const UINT32 srcStep[3],
BYTE* WINPR_RESTRICT pDst, UINT32 dstStep,
UINT32 DstFormat,
const prim_size_t* WINPR_RESTRICT roi)
{
const DWORD formatSize = FreeRDPGetBytesPerPixel(DstFormat);
fkt_writePixel writePixel = getPixelWriteFunction(DstFormat, FALSE);
WINPR_ASSERT(pSrc);
WINPR_ASSERT(pDst);
WINPR_ASSERT(roi);
const UINT32 nWidth = roi->width;
const UINT32 nHeight = roi->height;
for (size_t y = 0; y < nHeight; y++)
{
const BYTE* pY = pSrc[0] + y * srcStep[0];
const BYTE* pU = pSrc[1] + y * srcStep[1];
const BYTE* pV = pSrc[2] + y * srcStep[2];
BYTE* pRGB = pDst + y * dstStep;
for (size_t x = 0; x < nWidth; x++)
{
const BYTE Y = pY[x];
const BYTE U = pU[x];
const BYTE V = pV[x];
const BYTE r = YUV2R(Y, U, V);
const BYTE g = YUV2G(Y, U, V);
const BYTE b = YUV2B(Y, U, V);
pRGB = writePixel(pRGB, formatSize, DstFormat, r, g, b, 0);
}
}
return PRIMITIVES_SUCCESS;
}
static pstatus_t general_YUV444ToRGB_8u_P3AC4R_BGRX(const BYTE* const WINPR_RESTRICT pSrc[3],
const UINT32 srcStep[3],
BYTE* WINPR_RESTRICT pDst, UINT32 dstStep,
UINT32 DstFormat,
const prim_size_t* WINPR_RESTRICT roi)
{
const DWORD formatSize = FreeRDPGetBytesPerPixel(DstFormat);
WINPR_ASSERT(pSrc);
WINPR_ASSERT(pDst);
WINPR_ASSERT(roi);
const UINT32 nWidth = roi->width;
const UINT32 nHeight = roi->height;
for (size_t y = 0; y < nHeight; y++)
{
const BYTE* pY = pSrc[0] + y * srcStep[0];
const BYTE* pU = pSrc[1] + y * srcStep[1];
const BYTE* pV = pSrc[2] + y * srcStep[2];
BYTE* pRGB = pDst + y * dstStep;
for (size_t x = 0; x < nWidth; x++)
{
const BYTE Y = pY[x];
const BYTE U = pU[x];
const BYTE V = pV[x];
const BYTE r = YUV2R(Y, U, V);
const BYTE g = YUV2G(Y, U, V);
const BYTE b = YUV2B(Y, U, V);
pRGB = writePixelBGRX(pRGB, formatSize, DstFormat, r, g, b, 0);
}
}
return PRIMITIVES_SUCCESS;
}
static pstatus_t general_YUV444ToRGB_8u_P3AC4R(const BYTE* const WINPR_RESTRICT pSrc[3],
const UINT32 srcStep[3], BYTE* WINPR_RESTRICT pDst,
UINT32 dstStep, UINT32 DstFormat,
const prim_size_t* WINPR_RESTRICT roi)
{
switch (DstFormat)
{
case PIXEL_FORMAT_BGRA32:
case PIXEL_FORMAT_BGRX32:
return general_YUV444ToRGB_8u_P3AC4R_BGRX(pSrc, srcStep, pDst, dstStep, DstFormat, roi);
default:
return general_YUV444ToRGB_8u_P3AC4R_general(pSrc, srcStep, pDst, dstStep, DstFormat,
roi);
}
}
/**
* | R | ( | 256 0 403 | | Y | )
* | G | = ( | 256 -48 -120 | | U - 128 | ) >> 8
* | B | ( | 256 475 0 | | V - 128 | )
*/
static pstatus_t general_YUV420ToRGB_8u_P3AC4R(const BYTE* const WINPR_RESTRICT pSrc[3],
const UINT32 srcStep[3], BYTE* WINPR_RESTRICT pDst,
UINT32 dstStep, UINT32 DstFormat,
const prim_size_t* WINPR_RESTRICT roi)
{
UINT32 dstPad = 0;
UINT32 srcPad[3];
BYTE Y = 0;
BYTE U = 0;
BYTE V = 0;
UINT32 halfWidth = 0;
UINT32 halfHeight = 0;
const BYTE* pY = NULL;
const BYTE* pU = NULL;
const BYTE* pV = NULL;
BYTE* pRGB = pDst;
UINT32 nWidth = 0;
UINT32 nHeight = 0;
UINT32 lastRow = 0;
UINT32 lastCol = 0;
const DWORD formatSize = FreeRDPGetBytesPerPixel(DstFormat);
fkt_writePixel writePixel = getPixelWriteFunction(DstFormat, FALSE);
pY = pSrc[0];
pU = pSrc[1];
pV = pSrc[2];
lastCol = roi->width & 0x01;
lastRow = roi->height & 0x01;
nWidth = (roi->width + 1) & ~0x0001;
nHeight = (roi->height + 1) & ~0x0001;
halfWidth = nWidth / 2;
halfHeight = nHeight / 2;
srcPad[0] = (srcStep[0] - nWidth);
srcPad[1] = (srcStep[1] - halfWidth);
srcPad[2] = (srcStep[2] - halfWidth);
dstPad = (dstStep - (nWidth * 4));
for (UINT32 y = 0; y < halfHeight;)
{
if (++y == halfHeight)
lastRow <<= 1;
for (UINT32 x = 0; x < halfWidth;)
{
BYTE r = 0;
BYTE g = 0;
BYTE b = 0;
if (++x == halfWidth)
lastCol <<= 1;
U = *pU++;
V = *pV++;
/* 1st pixel */
Y = *pY++;
r = YUV2R(Y, U, V);
g = YUV2G(Y, U, V);
b = YUV2B(Y, U, V);
pRGB = writePixel(pRGB, formatSize, DstFormat, r, g, b, 0);
/* 2nd pixel */
if (!(lastCol & 0x02))
{
Y = *pY++;
r = YUV2R(Y, U, V);
g = YUV2G(Y, U, V);
b = YUV2B(Y, U, V);
pRGB = writePixel(pRGB, formatSize, DstFormat, r, g, b, 0);
}
else
{
pY++;
pRGB += formatSize;
lastCol >>= 1;
}
}
pY += srcPad[0];
pU -= halfWidth;
pV -= halfWidth;
pRGB += dstPad;
if (lastRow & 0x02)
break;
for (UINT32 x = 0; x < halfWidth;)
{
BYTE r = 0;
BYTE g = 0;
BYTE b = 0;
if (++x == halfWidth)
lastCol <<= 1;
U = *pU++;
V = *pV++;
/* 3rd pixel */
Y = *pY++;
r = YUV2R(Y, U, V);
g = YUV2G(Y, U, V);
b = YUV2B(Y, U, V);
pRGB = writePixel(pRGB, formatSize, DstFormat, r, g, b, 0);
/* 4th pixel */
if (!(lastCol & 0x02))
{
Y = *pY++;
r = YUV2R(Y, U, V);
g = YUV2G(Y, U, V);
b = YUV2B(Y, U, V);
pRGB = writePixel(pRGB, formatSize, DstFormat, r, g, b, 0);
}
else
{
pY++;
pRGB += formatSize;
lastCol >>= 1;
}
}
pY += srcPad[0];
pU += srcPad[1];
pV += srcPad[2];
pRGB += dstPad;
}
return PRIMITIVES_SUCCESS;
}
/**
* | Y | ( | 54 183 18 | | R | ) | 0 |
* | U | = ( | -29 -99 128 | | G | ) >> 8 + | 128 |
* | V | ( | 128 -116 -12 | | B | ) | 128 |
*/
static INLINE BYTE RGB2Y(BYTE R, BYTE G, BYTE B)
{
return (54 * R + 183 * G + 18 * B) >> 8;
}
static INLINE BYTE RGB2U(BYTE R, BYTE G, BYTE B)
{
return ((-29 * R - 99 * G + 128 * B) >> 8) + 128;
}
static INLINE BYTE RGB2V(INT32 R, INT32 G, INT32 B)
{
return ((128 * R - 116 * G - 12 * B) >> 8) + 128;
}
static pstatus_t general_RGBToYUV444_8u_P3AC4R(const BYTE* WINPR_RESTRICT pSrc, UINT32 SrcFormat,
const UINT32 srcStep, BYTE* WINPR_RESTRICT pDst[3],
const UINT32 dstStep[3],
const prim_size_t* WINPR_RESTRICT roi)
{
const UINT32 bpp = FreeRDPGetBytesPerPixel(SrcFormat);
UINT32 nWidth = 0;
UINT32 nHeight = 0;
nWidth = roi->width;
nHeight = roi->height;
for (size_t y = 0; y < nHeight; y++)
{
const BYTE* pRGB = pSrc + y * srcStep;
BYTE* pY = pDst[0] + y * dstStep[0];
BYTE* pU = pDst[1] + y * dstStep[1];
BYTE* pV = pDst[2] + y * dstStep[2];
for (size_t x = 0; x < nWidth; x++)
{
BYTE B = 0;
BYTE G = 0;
BYTE R = 0;
const UINT32 color = FreeRDPReadColor(&pRGB[x * bpp], SrcFormat);
FreeRDPSplitColor(color, SrcFormat, &R, &G, &B, NULL, NULL);
pY[x] = RGB2Y(R, G, B);
pU[x] = RGB2U(R, G, B);
pV[x] = RGB2V(R, G, B);
}
}
return PRIMITIVES_SUCCESS;
}
static INLINE pstatus_t general_RGBToYUV420_BGRX(const BYTE* WINPR_RESTRICT pSrc, UINT32 srcStep,
BYTE* WINPR_RESTRICT pDst[3],
const UINT32 dstStep[3],
const prim_size_t* WINPR_RESTRICT roi)
{
size_t i = 0;
size_t x1 = 0;
size_t x2 = 4;
size_t x3 = srcStep;
size_t x4 = srcStep + 4;
size_t y1 = 0;
size_t y2 = 1;
size_t y3 = dstStep[0];
size_t y4 = dstStep[0] + 1;
UINT32 max_x = roi->width - 1;
UINT32 max_y = roi->height - 1;
for (size_t y = i = 0; y < roi->height; y += 2, i++)
{
const BYTE* src = pSrc + y * srcStep;
BYTE* ydst = pDst[0] + y * dstStep[0];
BYTE* udst = pDst[1] + i * dstStep[1];
BYTE* vdst = pDst[2] + i * dstStep[2];
for (size_t x = 0; x < roi->width; x += 2)
{
BYTE R = 0;
BYTE G = 0;
BYTE B = 0;
INT32 Ra = 0;
INT32 Ga = 0;
INT32 Ba = 0;
/* row 1, pixel 1 */
Ba = B = *(src + x1 + 0);
Ga = G = *(src + x1 + 1);
Ra = R = *(src + x1 + 2);
ydst[y1] = RGB2Y(R, G, B);
if (x < max_x)
{
/* row 1, pixel 2 */
Ba += B = *(src + x2 + 0);
Ga += G = *(src + x2 + 1);
Ra += R = *(src + x2 + 2);
ydst[y2] = RGB2Y(R, G, B);
}
if (y < max_y)
{
/* row 2, pixel 1 */
Ba += B = *(src + x3 + 0);
Ga += G = *(src + x3 + 1);
Ra += R = *(src + x3 + 2);
ydst[y3] = RGB2Y(R, G, B);
if (x < max_x)
{
/* row 2, pixel 2 */
Ba += B = *(src + x4 + 0);
Ga += G = *(src + x4 + 1);
Ra += R = *(src + x4 + 2);
ydst[y4] = RGB2Y(R, G, B);
}
}
Ba >>= 2;
Ga >>= 2;
Ra >>= 2;
*udst++ = RGB2U(Ra, Ga, Ba);
*vdst++ = RGB2V(Ra, Ga, Ba);
ydst += 2;
src += 8;
}
}
return PRIMITIVES_SUCCESS;
}
static INLINE pstatus_t general_RGBToYUV420_RGBX(const BYTE* WINPR_RESTRICT pSrc, UINT32 srcStep,
BYTE* WINPR_RESTRICT pDst[3],
const UINT32 dstStep[3],
const prim_size_t* WINPR_RESTRICT roi)
{
size_t x1 = 0;
size_t x2 = 4;
size_t x3 = srcStep;
size_t x4 = srcStep + 4;
size_t y1 = 0;
size_t y2 = 1;
size_t y3 = dstStep[0];
size_t y4 = dstStep[0] + 1;
UINT32 max_x = roi->width - 1;
UINT32 max_y = roi->height - 1;
for (size_t y = 0, i = 0; y < roi->height; y += 2, i++)
{
const BYTE* src = pSrc + y * srcStep;
BYTE* ydst = pDst[0] + y * dstStep[0];
BYTE* udst = pDst[1] + i * dstStep[1];
BYTE* vdst = pDst[2] + i * dstStep[2];
for (UINT32 x = 0; x < roi->width; x += 2)
{
BYTE R = 0;
BYTE G = 0;
BYTE B = 0;
INT32 Ra = 0;
INT32 Ga = 0;
INT32 Ba = 0;
/* row 1, pixel 1 */
Ra = R = *(src + x1 + 0);
Ga = G = *(src + x1 + 1);
Ba = B = *(src + x1 + 2);
ydst[y1] = RGB2Y(R, G, B);
if (x < max_x)
{
/* row 1, pixel 2 */
Ra += R = *(src + x2 + 0);
Ga += G = *(src + x2 + 1);
Ba += B = *(src + x2 + 2);
ydst[y2] = RGB2Y(R, G, B);
}
if (y < max_y)
{
/* row 2, pixel 1 */
Ra += R = *(src + x3 + 0);
Ga += G = *(src + x3 + 1);
Ba += B = *(src + x3 + 2);
ydst[y3] = RGB2Y(R, G, B);
if (x < max_x)
{
/* row 2, pixel 2 */
Ra += R = *(src + x4 + 0);
Ga += G = *(src + x4 + 1);
Ba += B = *(src + x4 + 2);
ydst[y4] = RGB2Y(R, G, B);
}
}
Ba >>= 2;
Ga >>= 2;
Ra >>= 2;
*udst++ = RGB2U(Ra, Ga, Ba);
*vdst++ = RGB2V(Ra, Ga, Ba);
ydst += 2;
src += 8;
}
}
return PRIMITIVES_SUCCESS;
}
static INLINE pstatus_t general_RGBToYUV420_ANY(const BYTE* WINPR_RESTRICT pSrc, UINT32 srcFormat,
UINT32 srcStep, BYTE* WINPR_RESTRICT pDst[3],
const UINT32 dstStep[3],
const prim_size_t* WINPR_RESTRICT roi)
{
const UINT32 bpp = FreeRDPGetBytesPerPixel(srcFormat);
size_t x1 = 0;
size_t x2 = bpp;
size_t x3 = srcStep;
size_t x4 = srcStep + bpp;
size_t y1 = 0;
size_t y2 = 1;
size_t y3 = dstStep[0];
size_t y4 = dstStep[0] + 1;
UINT32 max_x = roi->width - 1;
UINT32 max_y = roi->height - 1;
for (size_t y = 0, i = 0; y < roi->height; y += 2, i++)
{
const BYTE* src = pSrc + y * srcStep;
BYTE* ydst = pDst[0] + y * dstStep[0];
BYTE* udst = pDst[1] + i * dstStep[1];
BYTE* vdst = pDst[2] + i * dstStep[2];
for (size_t x = 0; x < roi->width; x += 2)
{
BYTE R = 0;
BYTE G = 0;
BYTE B = 0;
INT32 Ra = 0;
INT32 Ga = 0;
INT32 Ba = 0;
UINT32 color = 0;
/* row 1, pixel 1 */
color = FreeRDPReadColor(src + x1, srcFormat);
FreeRDPSplitColor(color, srcFormat, &R, &G, &B, NULL, NULL);
Ra = R;
Ga = G;
Ba = B;
ydst[y1] = RGB2Y(R, G, B);
if (x < max_x)
{
/* row 1, pixel 2 */
color = FreeRDPReadColor(src + x2, srcFormat);
FreeRDPSplitColor(color, srcFormat, &R, &G, &B, NULL, NULL);
Ra += R;
Ga += G;
Ba += B;
ydst[y2] = RGB2Y(R, G, B);
}
if (y < max_y)
{
/* row 2, pixel 1 */
color = FreeRDPReadColor(src + x3, srcFormat);
FreeRDPSplitColor(color, srcFormat, &R, &G, &B, NULL, NULL);
Ra += R;
Ga += G;
Ba += B;
ydst[y3] = RGB2Y(R, G, B);
if (x < max_x)
{
/* row 2, pixel 2 */
color = FreeRDPReadColor(src + x4, srcFormat);
FreeRDPSplitColor(color, srcFormat, &R, &G, &B, NULL, NULL);
Ra += R;
Ga += G;
Ba += B;
ydst[y4] = RGB2Y(R, G, B);
}
}
Ra >>= 2;
Ga >>= 2;
Ba >>= 2;
*udst++ = RGB2U(Ra, Ga, Ba);
*vdst++ = RGB2V(Ra, Ga, Ba);
ydst += 2;
src += 2ULL * bpp;
}
}
return PRIMITIVES_SUCCESS;
}
static pstatus_t general_RGBToYUV420_8u_P3AC4R(const BYTE* WINPR_RESTRICT pSrc, UINT32 srcFormat,
UINT32 srcStep, BYTE* WINPR_RESTRICT pDst[3],
const UINT32 dstStep[3],
const prim_size_t* WINPR_RESTRICT roi)
{
switch (srcFormat)
{
case PIXEL_FORMAT_BGRA32:
case PIXEL_FORMAT_BGRX32:
return general_RGBToYUV420_BGRX(pSrc, srcStep, pDst, dstStep, roi);
case PIXEL_FORMAT_RGBA32:
case PIXEL_FORMAT_RGBX32:
return general_RGBToYUV420_RGBX(pSrc, srcStep, pDst, dstStep, roi);
default:
return general_RGBToYUV420_ANY(pSrc, srcFormat, srcStep, pDst, dstStep, roi);
}
}
static INLINE void general_RGBToAVC444YUV_BGRX_DOUBLE_ROW(
const BYTE* WINPR_RESTRICT srcEven, const BYTE* WINPR_RESTRICT srcOdd,
BYTE* WINPR_RESTRICT b1Even, BYTE* WINPR_RESTRICT b1Odd, BYTE* WINPR_RESTRICT b2,
BYTE* WINPR_RESTRICT b3, BYTE* WINPR_RESTRICT b4, BYTE* WINPR_RESTRICT b5,
BYTE* WINPR_RESTRICT b6, BYTE* WINPR_RESTRICT b7, UINT32 width)
{
for (UINT32 x = 0; x < width; x += 2)
{
const BOOL lastX = (x + 1) >= width;
BYTE Y1e = 0;
BYTE Y2e = 0;
BYTE U1e = 0;
BYTE V1e = 0;
BYTE U2e = 0;
BYTE V2e = 0;
BYTE Y1o = 0;
BYTE Y2o = 0;
BYTE U1o = 0;
BYTE V1o = 0;
BYTE U2o = 0;
BYTE V2o = 0;
/* Read 4 pixels, 2 from even, 2 from odd lines */
{
const BYTE b = *srcEven++;
const BYTE g = *srcEven++;
const BYTE r = *srcEven++;
srcEven++;
Y1e = Y2e = Y1o = Y2o = RGB2Y(r, g, b);
U1e = U2e = U1o = U2o = RGB2U(r, g, b);
V1e = V2e = V1o = V2o = RGB2V(r, g, b);
}
if (!lastX)
{
const BYTE b = *srcEven++;
const BYTE g = *srcEven++;
const BYTE r = *srcEven++;
srcEven++;
Y2e = RGB2Y(r, g, b);
U2e = RGB2U(r, g, b);
V2e = RGB2V(r, g, b);
}
if (b1Odd)
{
const BYTE b = *srcOdd++;
const BYTE g = *srcOdd++;
const BYTE r = *srcOdd++;
srcOdd++;
Y1o = Y2o = RGB2Y(r, g, b);
U1o = U2o = RGB2U(r, g, b);
V1o = V2o = RGB2V(r, g, b);
}
if (b1Odd && !lastX)
{
const BYTE b = *srcOdd++;
const BYTE g = *srcOdd++;
const BYTE r = *srcOdd++;
srcOdd++;
Y2o = RGB2Y(r, g, b);
U2o = RGB2U(r, g, b);
V2o = RGB2V(r, g, b);
}
/* We have 4 Y pixels, so store them. */
*b1Even++ = Y1e;
*b1Even++ = Y2e;
if (b1Odd)
{
*b1Odd++ = Y1o;
*b1Odd++ = Y2o;
}
/* 2x 2y pixel in luma UV plane use averaging
*/
{
const BYTE Uavg = ((UINT16)U1e + (UINT16)U2e + (UINT16)U1o + (UINT16)U2o) / 4;
const BYTE Vavg = ((UINT16)V1e + (UINT16)V2e + (UINT16)V1o + (UINT16)V2o) / 4;
*b2++ = Uavg;
*b3++ = Vavg;
}
/* UV from 2x, 2y+1 */
if (b1Odd)
{
*b4++ = U1o;
*b5++ = V1o;
if (!lastX)
{
*b4++ = U2o;
*b5++ = V2o;
}
}
/* UV from 2x+1, 2y */
if (!lastX)
{
*b6++ = U2e;
*b7++ = V2e;
}
}
}
static INLINE pstatus_t general_RGBToAVC444YUV_BGRX(const BYTE* WINPR_RESTRICT pSrc, UINT32 srcStep,
BYTE* WINPR_RESTRICT pDst1[3],
const UINT32 dst1Step[3],
BYTE* WINPR_RESTRICT pDst2[3],
const UINT32 dst2Step[3],
const prim_size_t* WINPR_RESTRICT roi)
{
/**
* Note:
* Read information in function general_RGBToAVC444YUV_ANY below !
*/
const BYTE* pMaxSrc = pSrc + 1ULL * (roi->height - 1) * srcStep;
for (size_t y = 0; y < roi->height; y += 2)
{
const BOOL last = (y >= (roi->height - 1));
const BYTE* srcEven = y < roi->height ? pSrc + y * srcStep : pMaxSrc;
const BYTE* srcOdd = !last ? pSrc + (y + 1) * srcStep : pMaxSrc;
const UINT32 i = y >> 1;
const UINT32 n = (i & ~7) + i;
BYTE* b1Even = pDst1[0] + y * dst1Step[0];
BYTE* b1Odd = !last ? (b1Even + dst1Step[0]) : NULL;
BYTE* b2 = pDst1[1] + (y / 2) * dst1Step[1];
BYTE* b3 = pDst1[2] + (y / 2) * dst1Step[2];
BYTE* b4 = pDst2[0] + 1ULL * dst2Step[0] * n;
BYTE* b5 = b4 + 8ULL * dst2Step[0];
BYTE* b6 = pDst2[1] + (y / 2) * dst2Step[1];
BYTE* b7 = pDst2[2] + (y / 2) * dst2Step[2];
general_RGBToAVC444YUV_BGRX_DOUBLE_ROW(srcEven, srcOdd, b1Even, b1Odd, b2, b3, b4, b5, b6,
b7, roi->width);
}
return PRIMITIVES_SUCCESS;
}
static INLINE void general_RGBToAVC444YUV_RGBX_DOUBLE_ROW(
const BYTE* WINPR_RESTRICT srcEven, const BYTE* WINPR_RESTRICT srcOdd,
BYTE* WINPR_RESTRICT b1Even, BYTE* WINPR_RESTRICT b1Odd, BYTE* WINPR_RESTRICT b2,
BYTE* WINPR_RESTRICT b3, BYTE* WINPR_RESTRICT b4, BYTE* WINPR_RESTRICT b5,
BYTE* WINPR_RESTRICT b6, BYTE* WINPR_RESTRICT b7, UINT32 width)
{
for (UINT32 x = 0; x < width; x += 2)
{
const BOOL lastX = (x + 1) >= width;
BYTE Y1e = 0;
BYTE Y2e = 0;
BYTE U1e = 0;
BYTE V1e = 0;
BYTE U2e = 0;
BYTE V2e = 0;
BYTE Y1o = 0;
BYTE Y2o = 0;
BYTE U1o = 0;
BYTE V1o = 0;
BYTE U2o = 0;
BYTE V2o = 0;
/* Read 4 pixels, 2 from even, 2 from odd lines */
{
const BYTE r = *srcEven++;
const BYTE g = *srcEven++;
const BYTE b = *srcEven++;
srcEven++;
Y1e = Y2e = Y1o = Y2o = RGB2Y(r, g, b);
U1e = U2e = U1o = U2o = RGB2U(r, g, b);
V1e = V2e = V1o = V2o = RGB2V(r, g, b);
}
if (!lastX)
{
const BYTE r = *srcEven++;
const BYTE g = *srcEven++;
const BYTE b = *srcEven++;
srcEven++;
Y2e = RGB2Y(r, g, b);
U2e = RGB2U(r, g, b);
V2e = RGB2V(r, g, b);
}
if (b1Odd)
{
const BYTE r = *srcOdd++;
const BYTE g = *srcOdd++;
const BYTE b = *srcOdd++;
srcOdd++;
Y1o = Y2o = RGB2Y(r, g, b);
U1o = U2o = RGB2U(r, g, b);
V1o = V2o = RGB2V(r, g, b);
}
if (b1Odd && !lastX)
{
const BYTE r = *srcOdd++;
const BYTE g = *srcOdd++;
const BYTE b = *srcOdd++;
srcOdd++;
Y2o = RGB2Y(r, g, b);
U2o = RGB2U(r, g, b);
V2o = RGB2V(r, g, b);
}
/* We have 4 Y pixels, so store them. */
*b1Even++ = Y1e;
*b1Even++ = Y2e;
if (b1Odd)
{
*b1Odd++ = Y1o;
*b1Odd++ = Y2o;
}
/* 2x 2y pixel in luma UV plane use averaging
*/
{
const BYTE Uavg = ((UINT16)U1e + (UINT16)U2e + (UINT16)U1o + (UINT16)U2o) / 4;
const BYTE Vavg = ((UINT16)V1e + (UINT16)V2e + (UINT16)V1o + (UINT16)V2o) / 4;
*b2++ = Uavg;
*b3++ = Vavg;
}
/* UV from 2x, 2y+1 */
if (b1Odd)
{
*b4++ = U1o;
*b5++ = V1o;
if (!lastX)
{
*b4++ = U2o;
*b5++ = V2o;
}
}
/* UV from 2x+1, 2y */
if (!lastX)
{
*b6++ = U2e;
*b7++ = V2e;
}
}
}
static INLINE pstatus_t general_RGBToAVC444YUV_RGBX(const BYTE* WINPR_RESTRICT pSrc, UINT32 srcStep,
BYTE* WINPR_RESTRICT pDst1[3],
const UINT32 dst1Step[3],
BYTE* WINPR_RESTRICT pDst2[3],
const UINT32 dst2Step[3],
const prim_size_t* WINPR_RESTRICT roi)
{
/**
* Note:
* Read information in function general_RGBToAVC444YUV_ANY below !
*/
const BYTE* pMaxSrc = pSrc + 1ULL * (roi->height - 1) * srcStep;
for (size_t y = 0; y < roi->height; y += 2)
{
const BOOL last = (y >= (roi->height - 1));
const BYTE* srcEven = y < roi->height ? pSrc + y * srcStep : pMaxSrc;
const BYTE* srcOdd = !last ? pSrc + (y + 1) * srcStep : pMaxSrc;
const UINT32 i = y >> 1;
const UINT32 n = (i & ~7) + i;
BYTE* b1Even = pDst1[0] + y * dst1Step[0];
BYTE* b1Odd = !last ? (b1Even + dst1Step[0]) : NULL;
BYTE* b2 = pDst1[1] + (y / 2) * dst1Step[1];
BYTE* b3 = pDst1[2] + (y / 2) * dst1Step[2];
BYTE* b4 = pDst2[0] + 1ULL * dst2Step[0] * n;
BYTE* b5 = b4 + 8ULL * dst2Step[0];
BYTE* b6 = pDst2[1] + (y / 2) * dst2Step[1];
BYTE* b7 = pDst2[2] + (y / 2) * dst2Step[2];
general_RGBToAVC444YUV_RGBX_DOUBLE_ROW(srcEven, srcOdd, b1Even, b1Odd, b2, b3, b4, b5, b6,
b7, roi->width);
}
return PRIMITIVES_SUCCESS;
}
static INLINE void general_RGBToAVC444YUV_ANY_DOUBLE_ROW(
const BYTE* WINPR_RESTRICT srcEven, const BYTE* WINPR_RESTRICT srcOdd, UINT32 srcFormat,
BYTE* WINPR_RESTRICT b1Even, BYTE* WINPR_RESTRICT b1Odd, BYTE* WINPR_RESTRICT b2,
BYTE* WINPR_RESTRICT b3, BYTE* WINPR_RESTRICT b4, BYTE* WINPR_RESTRICT b5,
BYTE* WINPR_RESTRICT b6, BYTE* WINPR_RESTRICT b7, UINT32 width)
{
const UINT32 bpp = FreeRDPGetBytesPerPixel(srcFormat);
for (UINT32 x = 0; x < width; x += 2)
{
const BOOL lastX = (x + 1) >= width;
BYTE Y1e = 0;
BYTE Y2e = 0;
BYTE U1e = 0;
BYTE V1e = 0;
BYTE U2e = 0;
BYTE V2e = 0;
BYTE Y1o = 0;
BYTE Y2o = 0;
BYTE U1o = 0;
BYTE V1o = 0;
BYTE U2o = 0;
BYTE V2o = 0;
/* Read 4 pixels, 2 from even, 2 from odd lines */
{
BYTE r = 0;
BYTE g = 0;
BYTE b = 0;
const UINT32 color = FreeRDPReadColor(srcEven, srcFormat);
srcEven += bpp;
FreeRDPSplitColor(color, srcFormat, &r, &g, &b, NULL, NULL);
Y1e = Y2e = Y1o = Y2o = RGB2Y(r, g, b);
U1e = U2e = U1o = U2o = RGB2U(r, g, b);
V1e = V2e = V1o = V2o = RGB2V(r, g, b);
}
if (!lastX)
{
BYTE r = 0;
BYTE g = 0;
BYTE b = 0;
const UINT32 color = FreeRDPReadColor(srcEven, srcFormat);
srcEven += bpp;
FreeRDPSplitColor(color, srcFormat, &r, &g, &b, NULL, NULL);
Y2e = RGB2Y(r, g, b);
U2e = RGB2U(r, g, b);
V2e = RGB2V(r, g, b);
}
if (b1Odd)
{
BYTE r = 0;
BYTE g = 0;
BYTE b = 0;
const UINT32 color = FreeRDPReadColor(srcOdd, srcFormat);
srcOdd += bpp;
FreeRDPSplitColor(color, srcFormat, &r, &g, &b, NULL, NULL);
Y1o = Y2o = RGB2Y(r, g, b);
U1o = U2o = RGB2U(r, g, b);
V1o = V2o = RGB2V(r, g, b);
}
if (b1Odd && !lastX)
{
BYTE r = 0;
BYTE g = 0;
BYTE b = 0;
const UINT32 color = FreeRDPReadColor(srcOdd, srcFormat);
srcOdd += bpp;
FreeRDPSplitColor(color, srcFormat, &r, &g, &b, NULL, NULL);
Y2o = RGB2Y(r, g, b);
U2o = RGB2U(r, g, b);
V2o = RGB2V(r, g, b);
}
/* We have 4 Y pixels, so store them. */
*b1Even++ = Y1e;
*b1Even++ = Y2e;
if (b1Odd)
{
*b1Odd++ = Y1o;
*b1Odd++ = Y2o;
}
/* 2x 2y pixel in luma UV plane use averaging
*/
{
const BYTE Uavg = ((UINT16)U1e + (UINT16)U2e + (UINT16)U1o + (UINT16)U2o) / 4;
const BYTE Vavg = ((UINT16)V1e + (UINT16)V2e + (UINT16)V1o + (UINT16)V2o) / 4;
*b2++ = Uavg;
*b3++ = Vavg;
}
/* UV from 2x, 2y+1 */
if (b1Odd)
{
*b4++ = U1o;
*b5++ = V1o;
if (!lastX)
{
*b4++ = U2o;
*b5++ = V2o;
}
}
/* UV from 2x+1, 2y */
if (!lastX)
{
*b6++ = U2e;
*b7++ = V2e;
}
}
}
static INLINE pstatus_t general_RGBToAVC444YUV_ANY(
const BYTE* WINPR_RESTRICT pSrc, UINT32 srcFormat, UINT32 srcStep,
BYTE* WINPR_RESTRICT pDst1[3], const UINT32 dst1Step[3], BYTE* WINPR_RESTRICT pDst2[3],
const UINT32 dst2Step[3], const prim_size_t* WINPR_RESTRICT roi)
{
/**
* Note: According to [MS-RDPEGFX 2.2.4.4 RFX_AVC420_BITMAP_STREAM] the
* width and height of the MPEG-4 AVC/H.264 codec bitstream MUST be aligned
* to a multiple of 16.
* Hence the passed destination YUV420/CHROMA420 buffers must have been
* allocated accordingly !!
*/
/**
* [MS-RDPEGFX 3.3.8.3.2 YUV420p Stream Combination] defines the following "Bx areas":
*
* YUV420 frame (main view):
* B1: From Y444 all pixels
* B2: From U444 all pixels in even rows with even columns
* B3: From V444 all pixels in even rows with even columns
*
* Chroma420 frame (auxillary view):
* B45: From U444 and V444 all pixels from all odd rows
* (The odd U444 and V444 rows must be interleaved in 8-line blocks in B45 !!!)
* B6: From U444 all pixels in even rows with odd columns
* B7: From V444 all pixels in even rows with odd columns
*
* Microsoft's horrible unclear description in MS-RDPEGFX translated to pseudo code looks like
* this:
*
* for (y = 0; y < fullHeight; y++)
* {
* for (x = 0; x < fullWidth; x++)
* {
* B1[x,y] = Y444[x,y];
* }
* }
*
* for (y = 0; y < halfHeight; y++)
* {
* for (x = 0; x < halfWidth; x++)
* {
* B2[x,y] = U444[2 * x, 2 * y];
* B3[x,y] = V444[2 * x, 2 * y];
* B6[x,y] = U444[2 * x + 1, 2 * y];
* B7[x,y] = V444[2 * x + 1, 2 * y];
* }
* }
*
* for (y = 0; y < halfHeight; y++)
* {
* yU = (y / 8) * 16; // identify first row of correct 8-line U block in B45
* yU += (y % 8); // add offset rows in destination block
* yV = yU + 8; // the corresponding v line is always 8 rows ahead
*
* for (x = 0; x < fullWidth; x++)
* {
* B45[x,yU] = U444[x, 2 * y + 1];
* B45[x,yV] = V444[x, 2 * y + 1];
* }
* }
*
*/
const BYTE* pMaxSrc = pSrc + 1ULL * (roi->height - 1) * srcStep;
for (size_t y = 0; y < roi->height; y += 2)
{
const BOOL last = (y >= (roi->height - 1));
const BYTE* srcEven = y < roi->height ? pSrc + y * srcStep : pMaxSrc;
const BYTE* srcOdd = !last ? pSrc + (y + 1) * srcStep : pMaxSrc;
const UINT32 i = y >> 1;
const UINT32 n = (i & ~7) + i;
BYTE* b1Even = pDst1[0] + y * dst1Step[0];
BYTE* b1Odd = !last ? (b1Even + dst1Step[0]) : NULL;
BYTE* b2 = pDst1[1] + (y / 2) * dst1Step[1];
BYTE* b3 = pDst1[2] + (y / 2) * dst1Step[2];
BYTE* b4 = pDst2[0] + 1ULL * dst2Step[0] * n;
BYTE* b5 = b4 + 8ULL * dst2Step[0];
BYTE* b6 = pDst2[1] + (y / 2) * dst2Step[1];
BYTE* b7 = pDst2[2] + (y / 2) * dst2Step[2];
general_RGBToAVC444YUV_ANY_DOUBLE_ROW(srcEven, srcOdd, srcFormat, b1Even, b1Odd, b2, b3, b4,
b5, b6, b7, roi->width);
}
return PRIMITIVES_SUCCESS;
}
static INLINE pstatus_t general_RGBToAVC444YUV(const BYTE* WINPR_RESTRICT pSrc, UINT32 srcFormat,
UINT32 srcStep, BYTE* WINPR_RESTRICT pDst1[3],
const UINT32 dst1Step[3],
BYTE* WINPR_RESTRICT pDst2[3],
const UINT32 dst2Step[3],
const prim_size_t* WINPR_RESTRICT roi)
{
if (!pSrc || !pDst1 || !dst1Step || !pDst2 || !dst2Step)
return -1;
if (!pDst1[0] || !pDst1[1] || !pDst1[2])
return -1;
if (!dst1Step[0] || !dst1Step[1] || !dst1Step[2])
return -1;
if (!pDst2[0] || !pDst2[1] || !pDst2[2])
return -1;
if (!dst2Step[0] || !dst2Step[1] || !dst2Step[2])
return -1;
switch (srcFormat)
{
case PIXEL_FORMAT_BGRA32:
case PIXEL_FORMAT_BGRX32:
return general_RGBToAVC444YUV_BGRX(pSrc, srcStep, pDst1, dst1Step, pDst2, dst2Step,
roi);
case PIXEL_FORMAT_RGBA32:
case PIXEL_FORMAT_RGBX32:
return general_RGBToAVC444YUV_RGBX(pSrc, srcStep, pDst1, dst1Step, pDst2, dst2Step,
roi);
default:
return general_RGBToAVC444YUV_ANY(pSrc, srcFormat, srcStep, pDst1, dst1Step, pDst2,
dst2Step, roi);
}
return !PRIMITIVES_SUCCESS;
}
static INLINE void general_RGBToAVC444YUVv2_ANY_DOUBLE_ROW(
const BYTE* WINPR_RESTRICT srcEven, const BYTE* WINPR_RESTRICT srcOdd, UINT32 srcFormat,
BYTE* WINPR_RESTRICT yLumaDstEven, BYTE* WINPR_RESTRICT yLumaDstOdd,
BYTE* WINPR_RESTRICT uLumaDst, BYTE* WINPR_RESTRICT vLumaDst,
BYTE* WINPR_RESTRICT yEvenChromaDst1, BYTE* WINPR_RESTRICT yEvenChromaDst2,
BYTE* WINPR_RESTRICT yOddChromaDst1, BYTE* WINPR_RESTRICT yOddChromaDst2,
BYTE* WINPR_RESTRICT uChromaDst1, BYTE* WINPR_RESTRICT uChromaDst2,
BYTE* WINPR_RESTRICT vChromaDst1, BYTE* WINPR_RESTRICT vChromaDst2, UINT32 width)
{
const UINT32 bpp = FreeRDPGetBytesPerPixel(srcFormat);
for (UINT32 x = 0; x < width; x += 2)
{
BYTE Ya = 0;
BYTE Ua = 0;
BYTE Va = 0;
BYTE Yb = 0;
BYTE Ub = 0;
BYTE Vb = 0;
BYTE Yc = 0;
BYTE Uc = 0;
BYTE Vc = 0;
BYTE Yd = 0;
BYTE Ud = 0;
BYTE Vd = 0;
{
BYTE b = 0;
BYTE g = 0;
BYTE r = 0;
const UINT32 color = FreeRDPReadColor(srcEven, srcFormat);
srcEven += bpp;
FreeRDPSplitColor(color, srcFormat, &r, &g, &b, NULL, NULL);
Ya = RGB2Y(r, g, b);
Ua = RGB2U(r, g, b);
Va = RGB2V(r, g, b);
}
if (x < width - 1)
{
BYTE b = 0;
BYTE g = 0;
BYTE r = 0;
const UINT32 color = FreeRDPReadColor(srcEven, srcFormat);
srcEven += bpp;
FreeRDPSplitColor(color, srcFormat, &r, &g, &b, NULL, NULL);
Yb = RGB2Y(r, g, b);
Ub = RGB2U(r, g, b);
Vb = RGB2V(r, g, b);
}
else
{
Yb = Ya;
Ub = Ua;
Vb = Va;
}
if (srcOdd)
{
BYTE b = 0;
BYTE g = 0;
BYTE r = 0;
const UINT32 color = FreeRDPReadColor(srcOdd, srcFormat);
srcOdd += bpp;
FreeRDPSplitColor(color, srcFormat, &r, &g, &b, NULL, NULL);
Yc = RGB2Y(r, g, b);
Uc = RGB2U(r, g, b);
Vc = RGB2V(r, g, b);
}
else
{
Yc = Ya;
Uc = Ua;
Vc = Va;
}
if (srcOdd && (x < width - 1))
{
BYTE b = 0;
BYTE g = 0;
BYTE r = 0;
const UINT32 color = FreeRDPReadColor(srcOdd, srcFormat);
srcOdd += bpp;
FreeRDPSplitColor(color, srcFormat, &r, &g, &b, NULL, NULL);
Yd = RGB2Y(r, g, b);
Ud = RGB2U(r, g, b);
Vd = RGB2V(r, g, b);
}
else
{
Yd = Ya;
Ud = Ua;
Vd = Va;
}
/* Y [b1] */
*yLumaDstEven++ = Ya;
if (x < width - 1)
*yLumaDstEven++ = Yb;
if (srcOdd)
*yLumaDstOdd++ = Yc;
if (srcOdd && (x < width - 1))
*yLumaDstOdd++ = Yd;
/* 2x 2y [b2,b3] */
*uLumaDst++ = (Ua + Ub + Uc + Ud) / 4;
*vLumaDst++ = (Va + Vb + Vc + Vd) / 4;
/* 2x+1, y [b4,b5] even */
if (x < width - 1)
{
*yEvenChromaDst1++ = Ub;
*yEvenChromaDst2++ = Vb;
}
if (srcOdd)
{
/* 2x+1, y [b4,b5] odd */
if (x < width - 1)
{
*yOddChromaDst1++ = Ud;
*yOddChromaDst2++ = Vd;
}
/* 4x 2y+1 [b6, b7] */
if (x % 4 == 0)
{
*uChromaDst1++ = Uc;
*uChromaDst2++ = Vc;
}
/* 4x+2 2y+1 [b8, b9] */
else
{
*vChromaDst1++ = Uc;
*vChromaDst2++ = Vc;
}
}
}
}
static INLINE pstatus_t general_RGBToAVC444YUVv2_ANY(
const BYTE* WINPR_RESTRICT pSrc, UINT32 srcFormat, UINT32 srcStep,
BYTE* WINPR_RESTRICT pDst1[3], const UINT32 dst1Step[3], BYTE* WINPR_RESTRICT pDst2[3],
const UINT32 dst2Step[3], const prim_size_t* WINPR_RESTRICT roi)
{
/**
* Note: According to [MS-RDPEGFX 2.2.4.4 RFX_AVC420_BITMAP_STREAM] the
* width and height of the MPEG-4 AVC/H.264 codec bitstream MUST be aligned
* to a multiple of 16.
* Hence the passed destination YUV420/CHROMA420 buffers must have been
* allocated accordingly !!
*/
/**
* [MS-RDPEGFX 3.3.8.3.3 YUV420p Stream Combination for YUV444v2 mode] defines the following "Bx
* areas":
*
* YUV420 frame (main view):
* B1: From Y444 all pixels
* B2: From U444 all pixels in even rows with even rows and columns
* B3: From V444 all pixels in even rows with even rows and columns
*
* Chroma420 frame (auxillary view):
* B45: From U444 and V444 all pixels from all odd columns
* B67: From U444 and V444 every 4th pixel in odd rows
* B89: From U444 and V444 every 4th pixel (initial offset of 2) in odd rows
*
* Chroma Bxy areas correspond to the left and right half of the YUV420 plane.
* for (y = 0; y < fullHeight; y++)
* {
* for (x = 0; x < fullWidth; x++)
* {
* B1[x,y] = Y444[x,y];
* }
*
* for (x = 0; x < halfWidth; x++)
* {
* B4[x,y] = U444[2 * x, 2 * y];
* B5[x,y] = V444[2 * x, 2 * y];
* }
* }
*
* for (y = 0; y < halfHeight; y++)
* {
* for (x = 0; x < halfWidth; x++)
* {
* B2[x,y] = U444[2 * x, 2 * y];
* B3[x,y] = V444[2 * x, 2 * y];
* B6[x,y] = U444[4 * x, 2 * y + 1];
* B7[x,y] = V444[4 * x, 2 * y + 1];
* B8[x,y] = V444[4 * x + 2, 2 * y + 1];
* B9[x,y] = V444[4 * x + 2, 2 * y] + 1;
* }
* }
*
*/
if (roi->height < 1 || roi->width < 1)
return !PRIMITIVES_SUCCESS;
for (size_t y = 0; y < roi->height; y += 2)
{
const BYTE* srcEven = (pSrc + y * srcStep);
const BYTE* srcOdd = (y < roi->height - 1) ? (srcEven + srcStep) : NULL;
BYTE* dstLumaYEven = (pDst1[0] + y * dst1Step[0]);
BYTE* dstLumaYOdd = (dstLumaYEven + dst1Step[0]);
BYTE* dstLumaU = (pDst1[1] + (y / 2) * dst1Step[1]);
BYTE* dstLumaV = (pDst1[2] + (y / 2) * dst1Step[2]);
BYTE* dstEvenChromaY1 = (pDst2[0] + y * dst2Step[0]);
BYTE* dstEvenChromaY2 = dstEvenChromaY1 + roi->width / 2;
BYTE* dstOddChromaY1 = dstEvenChromaY1 + dst2Step[0];
BYTE* dstOddChromaY2 = dstEvenChromaY2 + dst2Step[0];
BYTE* dstChromaU1 = (pDst2[1] + (y / 2) * dst2Step[1]);
BYTE* dstChromaV1 = (pDst2[2] + (y / 2) * dst2Step[2]);
BYTE* dstChromaU2 = dstChromaU1 + roi->width / 4;
BYTE* dstChromaV2 = dstChromaV1 + roi->width / 4;
general_RGBToAVC444YUVv2_ANY_DOUBLE_ROW(
srcEven, srcOdd, srcFormat, dstLumaYEven, dstLumaYOdd, dstLumaU, dstLumaV,
dstEvenChromaY1, dstEvenChromaY2, dstOddChromaY1, dstOddChromaY2, dstChromaU1,
dstChromaU2, dstChromaV1, dstChromaV2, roi->width);
}
return PRIMITIVES_SUCCESS;
}
static INLINE void general_RGBToAVC444YUVv2_BGRX_DOUBLE_ROW(
const BYTE* WINPR_RESTRICT srcEven, const BYTE* WINPR_RESTRICT srcOdd,
BYTE* WINPR_RESTRICT yLumaDstEven, BYTE* WINPR_RESTRICT yLumaDstOdd,
BYTE* WINPR_RESTRICT uLumaDst, BYTE* WINPR_RESTRICT vLumaDst,
BYTE* WINPR_RESTRICT yEvenChromaDst1, BYTE* WINPR_RESTRICT yEvenChromaDst2,
BYTE* WINPR_RESTRICT yOddChromaDst1, BYTE* WINPR_RESTRICT yOddChromaDst2,
BYTE* WINPR_RESTRICT uChromaDst1, BYTE* WINPR_RESTRICT uChromaDst2,
BYTE* WINPR_RESTRICT vChromaDst1, BYTE* WINPR_RESTRICT vChromaDst2, UINT32 width)
{
for (UINT32 x = 0; x < width; x += 2)
{
BYTE Ya = 0;
BYTE Ua = 0;
BYTE Va = 0;
BYTE Yb = 0;
BYTE Ub = 0;
BYTE Vb = 0;
BYTE Yc = 0;
BYTE Uc = 0;
BYTE Vc = 0;
BYTE Yd = 0;
BYTE Ud = 0;
BYTE Vd = 0;
{
const BYTE b = *srcEven++;
const BYTE g = *srcEven++;
const BYTE r = *srcEven++;
srcEven++;
Ya = RGB2Y(r, g, b);
Ua = RGB2U(r, g, b);
Va = RGB2V(r, g, b);
}
if (x < width - 1)
{
const BYTE b = *srcEven++;
const BYTE g = *srcEven++;
const BYTE r = *srcEven++;
srcEven++;
Yb = RGB2Y(r, g, b);
Ub = RGB2U(r, g, b);
Vb = RGB2V(r, g, b);
}
else
{
Yb = Ya;
Ub = Ua;
Vb = Va;
}
if (srcOdd)
{
const BYTE b = *srcOdd++;
const BYTE g = *srcOdd++;
const BYTE r = *srcOdd++;
srcOdd++;
Yc = RGB2Y(r, g, b);
Uc = RGB2U(r, g, b);
Vc = RGB2V(r, g, b);
}
else
{
Yc = Ya;
Uc = Ua;
Vc = Va;
}
if (srcOdd && (x < width - 1))
{
const BYTE b = *srcOdd++;
const BYTE g = *srcOdd++;
const BYTE r = *srcOdd++;
srcOdd++;
Yd = RGB2Y(r, g, b);
Ud = RGB2U(r, g, b);
Vd = RGB2V(r, g, b);
}
else
{
Yd = Ya;
Ud = Ua;
Vd = Va;
}
/* Y [b1] */
*yLumaDstEven++ = Ya;
if (x < width - 1)
*yLumaDstEven++ = Yb;
if (srcOdd)
*yLumaDstOdd++ = Yc;
if (srcOdd && (x < width - 1))
*yLumaDstOdd++ = Yd;
/* 2x 2y [b2,b3] */
*uLumaDst++ = (Ua + Ub + Uc + Ud) / 4;
*vLumaDst++ = (Va + Vb + Vc + Vd) / 4;
/* 2x+1, y [b4,b5] even */
if (x < width - 1)
{
*yEvenChromaDst1++ = Ub;
*yEvenChromaDst2++ = Vb;
}
if (srcOdd)
{
/* 2x+1, y [b4,b5] odd */
if (x < width - 1)
{
*yOddChromaDst1++ = Ud;
*yOddChromaDst2++ = Vd;
}
/* 4x 2y+1 [b6, b7] */
if (x % 4 == 0)
{
*uChromaDst1++ = Uc;
*uChromaDst2++ = Vc;
}
/* 4x+2 2y+1 [b8, b9] */
else
{
*vChromaDst1++ = Uc;
*vChromaDst2++ = Vc;
}
}
}
}
static INLINE pstatus_t general_RGBToAVC444YUVv2_BGRX(const BYTE* WINPR_RESTRICT pSrc,
UINT32 srcStep, BYTE* WINPR_RESTRICT pDst1[3],
const UINT32 dst1Step[3],
BYTE* WINPR_RESTRICT pDst2[3],
const UINT32 dst2Step[3],
const prim_size_t* WINPR_RESTRICT roi)
{
if (roi->height < 1 || roi->width < 1)
return !PRIMITIVES_SUCCESS;
for (size_t y = 0; y < roi->height; y += 2)
{
const BYTE* srcEven = (pSrc + y * srcStep);
const BYTE* srcOdd = (y < roi->height - 1) ? (srcEven + srcStep) : NULL;
BYTE* dstLumaYEven = (pDst1[0] + y * dst1Step[0]);
BYTE* dstLumaYOdd = (dstLumaYEven + dst1Step[0]);
BYTE* dstLumaU = (pDst1[1] + (y / 2) * dst1Step[1]);
BYTE* dstLumaV = (pDst1[2] + (y / 2) * dst1Step[2]);
BYTE* dstEvenChromaY1 = (pDst2[0] + y * dst2Step[0]);
BYTE* dstEvenChromaY2 = dstEvenChromaY1 + roi->width / 2;
BYTE* dstOddChromaY1 = dstEvenChromaY1 + dst2Step[0];
BYTE* dstOddChromaY2 = dstEvenChromaY2 + dst2Step[0];
BYTE* dstChromaU1 = (pDst2[1] + (y / 2) * dst2Step[1]);
BYTE* dstChromaV1 = (pDst2[2] + (y / 2) * dst2Step[2]);
BYTE* dstChromaU2 = dstChromaU1 + roi->width / 4;
BYTE* dstChromaV2 = dstChromaV1 + roi->width / 4;
general_RGBToAVC444YUVv2_BGRX_DOUBLE_ROW(
srcEven, srcOdd, dstLumaYEven, dstLumaYOdd, dstLumaU, dstLumaV, dstEvenChromaY1,
dstEvenChromaY2, dstOddChromaY1, dstOddChromaY2, dstChromaU1, dstChromaU2, dstChromaV1,
dstChromaV2, roi->width);
}
return PRIMITIVES_SUCCESS;
}
static INLINE pstatus_t general_RGBToAVC444YUVv2(const BYTE* WINPR_RESTRICT pSrc, UINT32 srcFormat,
UINT32 srcStep, BYTE* WINPR_RESTRICT pDst1[3],
const UINT32 dst1Step[3],
BYTE* WINPR_RESTRICT pDst2[3],
const UINT32 dst2Step[3],
const prim_size_t* WINPR_RESTRICT roi)
{
switch (srcFormat)
{
case PIXEL_FORMAT_BGRA32:
case PIXEL_FORMAT_BGRX32:
return general_RGBToAVC444YUVv2_BGRX(pSrc, srcStep, pDst1, dst1Step, pDst2, dst2Step,
roi);
default:
return general_RGBToAVC444YUVv2_ANY(pSrc, srcFormat, srcStep, pDst1, dst1Step, pDst2,
dst2Step, roi);
}
return !PRIMITIVES_SUCCESS;
}
void primitives_init_YUV(primitives_t* WINPR_RESTRICT prims)
{
prims->YUV420ToRGB_8u_P3AC4R = general_YUV420ToRGB_8u_P3AC4R;
prims->YUV444ToRGB_8u_P3AC4R = general_YUV444ToRGB_8u_P3AC4R;
prims->RGBToYUV420_8u_P3AC4R = general_RGBToYUV420_8u_P3AC4R;
prims->RGBToYUV444_8u_P3AC4R = general_RGBToYUV444_8u_P3AC4R;
prims->YUV420CombineToYUV444 = general_YUV420CombineToYUV444;
prims->YUV444SplitToYUV420 = general_YUV444SplitToYUV420;
prims->RGBToAVC444YUV = general_RGBToAVC444YUV;
prims->RGBToAVC444YUVv2 = general_RGBToAVC444YUVv2;
}
void primitives_init_YUV_opt(primitives_t* WINPR_RESTRICT prims)
{
primitives_init_YUV_ssse3(prims);
primitives_init_YUV_neon(prims);
}