template <bool align> void BgrToBgra(const uint8_t * bgr, size_t width, size_t height, size_t bgrStride, uint8_t * bgra, size_t bgraStride, uint8_t alpha)
{
assert(width >= A);
if(align)
assert(Aligned(bgra) && Aligned(bgraStride) && Aligned(bgr) && Aligned(bgrStride));
size_t alignedWidth = AlignLo(width, A);
if(width == alignedWidth)
alignedWidth -= A;
const v128_u8 _alpha = SetU8(alpha);
for(size_t row = 0; row < height; ++row)
{
Loader<align> _bgr(bgr);
Storer<align> _bgra(bgra);
BgrToBgra<align, true>(_bgr, _alpha, _bgra);
for(size_t col = A; col < alignedWidth; col += A)
BgrToBgra<align, false>(_bgr, _alpha, _bgra);
Flush(_bgra);
if(width != alignedWidth)
{
Loader<false> _bgr(bgr + 3*(width - A));
Storer<false> _bgra(bgra + 4*(width - A));
BgrToBgra<false, true>(_bgr, _alpha, _bgra);
Flush(_bgra);
}
bgra += bgraStride;
bgr += bgrStride;
}
}
template <bool align> void Bgr48pToBgra32(const uint8_t * blue, size_t blueStride, size_t width, size_t height,
const uint8_t * green, size_t greenStride, const uint8_t * red, size_t redStride, uint8_t * bgra, size_t bgraStride, uint8_t alpha)
{
assert(width >= HA);
if(align)
{
assert(Aligned(blue) && Aligned(blueStride));
assert(Aligned(green) && Aligned(greenStride));
assert(Aligned(red) && Aligned(redStride));
assert(Aligned(bgra) && Aligned(bgraStride));
}
v128_u8 _alpha = SetU8(alpha);
size_t alignedWidth = AlignLo(width, HA);
for(size_t row = 0; row < height; ++row)
{
Storer<align> _bgra(bgra);
Bgr48pToBgra32<align, true>(blue, green, red, 0, _alpha, _bgra);
for(size_t col = HA; col < alignedWidth; col += HA)
Bgr48pToBgra32<align, false>(blue, green, red, col*2, _alpha, _bgra);
Flush(_bgra);
if(width != alignedWidth)
{
Storer<false> _bgra(bgra + (width - HA)*4);
Bgr48pToBgra32<false, true>(blue, green, red, (width - HA)*2, _alpha, _bgra);
Flush(_bgra);
}
blue += blueStride;
green += greenStride;
red += redStride;
bgra += bgraStride;
}
}
void BgraToBayer(const uint8_t * bgra, size_t width, size_t height, size_t bgraStride, uint8_t * bayer, size_t bayerStride)
{
assert(width >= A);
if(align)
assert(Aligned(bgra) && Aligned(bgraStride) && Aligned(bayer) && Aligned(bayerStride));
size_t alignedWidth = AlignLo(width, A);
const v128_u8 perm[4][2] =
{
{K8_PERM_GR, K8_PERM_BG},
{K8_PERM_GB, K8_PERM_RG},
{K8_PERM_RG, K8_PERM_GB},
{K8_PERM_BG, K8_PERM_GR}
};
for(size_t row = 0; row < height; row += 2)
{
Loader<align> _bgra0(bgra);
Storer<align> _bayer0(bayer);
BgraToBayer<format, 0, align, true>(_bgra0, perm, _bayer0);
for(size_t col = A; col < alignedWidth; col += A)
BgraToBayer<format, 0, align, false>(_bgra0, perm, _bayer0);
Flush(_bayer0);
if(width != alignedWidth)
{
Loader<false> _bgra(bgra + 4*(width - A));
Storer<false> _bayer(bayer + width - A);
BgraToBayer<format, 0, false, true>(_bgra, perm, _bayer);
Flush(_bayer);
}
bgra += bgraStride;
bayer += bayerStride;
Loader<align> _bgra1(bgra);
Storer<align> _bayer1(bayer);
BgraToBayer<format, 1, align, true>(_bgra1, perm, _bayer1);
for(size_t col = A; col < alignedWidth; col += A)
BgraToBayer<format, 1, align, false>(_bgra1, perm, _bayer1);
Flush(_bayer1);
if(width != alignedWidth)
{
Loader<false> _bgra(bgra + 4*(width - A));
Storer<false> _bayer(bayer + width - A);
BgraToBayer<format, 1, false, true>(_bgra, perm, _bayer);
Flush(_bayer);
}
bgra += bgraStride;
bayer += bayerStride;
}
}
