aboutsummaryrefslogtreecommitdiff
path: root/Ryujinx.Graphics.Nvdec.Vp9/Dsp/Convolve.cs
diff options
context:
space:
mode:
authorgdkchan <gab.dark.100@gmail.com>2020-07-12 00:07:01 -0300
committerGitHub <noreply@github.com>2020-07-12 05:07:01 +0200
commit4d02a2d2c0451b4de1f6de3bbce54c457cacebe2 (patch)
tree120fe4fb8cfa1ac1c6ef4c97d92be47e955e8c0f /Ryujinx.Graphics.Nvdec.Vp9/Dsp/Convolve.cs
parent38b26cf4242999fa7d8c550993ac0940cd03d55f (diff)
New NVDEC and VIC implementation (#1384)
* Initial NVDEC and VIC implementation * Update FFmpeg.AutoGen to 4.3.0 * Add nvdec dependencies for Windows * Unify some VP9 structures * Rename VP9 structure fields * Improvements to Video API * XML docs for Common.Memory * Remove now unused or redundant overloads from MemoryAccessor * NVDEC UV surface read/write scalar paths * Add FIXME comments about hacky things/stuff that will need to be fixed in the future * Cleaned up VP9 memory allocation * Remove some debug logs * Rename some VP9 structs * Remove unused struct * No need to compile Ryujinx.Graphics.Host1x with unsafe anymore * Name AsyncWorkQueue threads to make debugging easier * Make Vp9PictureInfo a ref struct * LayoutConverter no longer needs the depth argument (broken by rebase) * Pooling of VP9 buffers, plus fix a memory leak on VP9 * Really wish VS could rename projects properly... * Address feedback * Remove using * Catch OperationCanceledException * Add licensing informations * Add THIRDPARTY.md to release too Co-authored-by: Thog <me@thog.eu>
Diffstat (limited to 'Ryujinx.Graphics.Nvdec.Vp9/Dsp/Convolve.cs')
-rw-r--r--Ryujinx.Graphics.Nvdec.Vp9/Dsp/Convolve.cs949
1 files changed, 949 insertions, 0 deletions
diff --git a/Ryujinx.Graphics.Nvdec.Vp9/Dsp/Convolve.cs b/Ryujinx.Graphics.Nvdec.Vp9/Dsp/Convolve.cs
new file mode 100644
index 00000000..b74c33dc
--- /dev/null
+++ b/Ryujinx.Graphics.Nvdec.Vp9/Dsp/Convolve.cs
@@ -0,0 +1,949 @@
+using Ryujinx.Common.Memory;
+using Ryujinx.Graphics.Nvdec.Vp9.Common;
+using System.Diagnostics;
+using System.Runtime.CompilerServices;
+using System.Runtime.InteropServices;
+using System.Runtime.Intrinsics;
+using System.Runtime.Intrinsics.X86;
+using static Ryujinx.Graphics.Nvdec.Vp9.Dsp.Filter;
+
+namespace Ryujinx.Graphics.Nvdec.Vp9.Dsp
+{
+ internal static class Convolve
+ {
+ private const bool UseIntrinsics = true;
+
+ [MethodImpl(MethodImplOptions.AggressiveInlining)]
+ private static Vector128<int> MultiplyAddAdjacent(
+ Vector128<short> vsrc0,
+ Vector128<short> vsrc1,
+ Vector128<short> vsrc2,
+ Vector128<short> vsrc3,
+ Vector128<short> vfilter,
+ Vector128<int> zero)
+ {
+ // < sumN, sumN, sumN, sumN >
+ Vector128<int> sum0 = Sse2.MultiplyAddAdjacent(vsrc0, vfilter);
+ Vector128<int> sum1 = Sse2.MultiplyAddAdjacent(vsrc1, vfilter);
+ Vector128<int> sum2 = Sse2.MultiplyAddAdjacent(vsrc2, vfilter);
+ Vector128<int> sum3 = Sse2.MultiplyAddAdjacent(vsrc3, vfilter);
+
+ // < 0, 0, sumN, sumN >
+ sum0 = Ssse3.HorizontalAdd(sum0, zero);
+ sum1 = Ssse3.HorizontalAdd(sum1, zero);
+ sum2 = Ssse3.HorizontalAdd(sum2, zero);
+ sum3 = Ssse3.HorizontalAdd(sum3, zero);
+
+ // < 0, 0, 0, sumN >
+ sum0 = Ssse3.HorizontalAdd(sum0, zero);
+ sum1 = Ssse3.HorizontalAdd(sum1, zero);
+ sum2 = Ssse3.HorizontalAdd(sum2, zero);
+ sum3 = Ssse3.HorizontalAdd(sum3, zero);
+
+ // < 0, 0, sum1, sum0 >
+ Vector128<int> sum01 = Sse2.UnpackLow(sum0, sum1);
+
+ // < 0, 0, sum3, sum2 >
+ Vector128<int> sum23 = Sse2.UnpackLow(sum2, sum3);
+
+ // < sum3, sum2, sum1, sum0 >
+ return Sse.MoveLowToHigh(sum01.AsSingle(), sum23.AsSingle()).AsInt32();
+ }
+
+ [MethodImpl(MethodImplOptions.AggressiveInlining)]
+ private static Vector128<int> RoundShift(Vector128<int> value, Vector128<int> const64)
+ {
+ return Sse2.ShiftRightArithmetic(Sse2.Add(value, const64), FilterBits);
+ }
+
+ [MethodImpl(MethodImplOptions.AggressiveInlining)]
+ private static Vector128<byte> PackUnsignedSaturate(Vector128<int> value, Vector128<int> zero)
+ {
+ return Sse2.PackUnsignedSaturate(Sse41.PackUnsignedSaturate(value, zero).AsInt16(), zero.AsInt16());
+ }
+
+ [MethodImpl(MethodImplOptions.AggressiveInlining)]
+ private static unsafe void ConvolveHorizSse41(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] xFilters,
+ int x0Q4,
+ int w,
+ int h)
+ {
+ Vector128<int> zero = Vector128<int>.Zero;
+ Vector128<int> const64 = Vector128.Create(64);
+
+ ulong x, y;
+ src -= SubpelTaps / 2 - 1;
+
+ fixed (Array8<short>* xFilter = xFilters)
+ {
+ Vector128<short> vfilter = Sse2.LoadVector128((short*)xFilter + (uint)(x0Q4 & SubpelMask) * 8);
+
+ for (y = 0; y < (uint)h; ++y)
+ {
+ ulong srcOffset = (uint)x0Q4 >> SubpelBits;
+ for (x = 0; x < (uint)w; x += 4)
+ {
+ Vector128<short> vsrc0 = Sse41.ConvertToVector128Int16(&src[srcOffset + x]);
+ Vector128<short> vsrc1 = Sse41.ConvertToVector128Int16(&src[srcOffset + x + 1]);
+ Vector128<short> vsrc2 = Sse41.ConvertToVector128Int16(&src[srcOffset + x + 2]);
+ Vector128<short> vsrc3 = Sse41.ConvertToVector128Int16(&src[srcOffset + x + 3]);
+
+ Vector128<int> sum0123 = MultiplyAddAdjacent(vsrc0, vsrc1, vsrc2, vsrc3, vfilter, zero);
+
+ Sse.StoreScalar((float*)&dst[x], PackUnsignedSaturate(RoundShift(sum0123, const64), zero).AsSingle());
+ }
+ src += srcStride;
+ dst += dstStride;
+ }
+ }
+ }
+
+ [MethodImpl(MethodImplOptions.AggressiveInlining)]
+ private static unsafe void ConvolveHoriz(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] xFilters,
+ int x0Q4,
+ int xStepQ4,
+ int w,
+ int h)
+ {
+ if (Sse41.IsSupported && UseIntrinsics && xStepQ4 == 1 << SubpelBits)
+ {
+ ConvolveHorizSse41(src, srcStride, dst, dstStride, xFilters, x0Q4, w, h);
+ return;
+ }
+
+ int x, y;
+ src -= SubpelTaps / 2 - 1;
+
+ for (y = 0; y < h; ++y)
+ {
+ int xQ4 = x0Q4;
+ for (x = 0; x < w; ++x)
+ {
+ byte* srcX = &src[xQ4 >> SubpelBits];
+ ref Array8<short> xFilter = ref xFilters[xQ4 & SubpelMask];
+ int k, sum = 0;
+ for (k = 0; k < SubpelTaps; ++k)
+ {
+ sum += srcX[k] * xFilter[k];
+ }
+
+ dst[x] = BitUtils.ClipPixel(BitUtils.RoundPowerOfTwo(sum, FilterBits));
+ xQ4 += xStepQ4;
+ }
+ src += srcStride;
+ dst += dstStride;
+ }
+ }
+
+ private static unsafe void ConvolveAvgHoriz(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] xFilters,
+ int x0Q4,
+ int xStepQ4,
+ int w,
+ int h)
+ {
+ int x, y;
+ src -= SubpelTaps / 2 - 1;
+
+ for (y = 0; y < h; ++y)
+ {
+ int xQ4 = x0Q4;
+ for (x = 0; x < w; ++x)
+ {
+ byte* srcX = &src[xQ4 >> SubpelBits];
+ ref Array8<short> xFilter = ref xFilters[xQ4 & SubpelMask];
+ int k, sum = 0;
+ for (k = 0; k < SubpelTaps; ++k)
+ {
+ sum += srcX[k] * xFilter[k];
+ }
+
+ dst[x] = (byte)BitUtils.RoundPowerOfTwo(dst[x] + BitUtils.ClipPixel(BitUtils.RoundPowerOfTwo(sum, FilterBits)), 1);
+ xQ4 += xStepQ4;
+ }
+ src += srcStride;
+ dst += dstStride;
+ }
+ }
+
+ [MethodImpl(MethodImplOptions.AggressiveInlining)]
+ private static unsafe void ConvolveVertAvx2(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] yFilters,
+ int y0Q4,
+ int w,
+ int h)
+ {
+ Vector128<int> zero = Vector128<int>.Zero;
+ Vector128<int> const64 = Vector128.Create(64);
+ Vector256<int> indices = Vector256.Create(
+ 0,
+ srcStride,
+ srcStride * 2,
+ srcStride * 3,
+ srcStride * 4,
+ srcStride * 5,
+ srcStride * 6,
+ srcStride * 7);
+
+ ulong x, y;
+ src -= srcStride * (SubpelTaps / 2 - 1);
+
+ fixed (Array8<short>* yFilter = yFilters)
+ {
+ Vector128<short> vfilter = Sse2.LoadVector128((short*)yFilter + (uint)(y0Q4 & SubpelMask) * 8);
+
+ ulong srcBaseY = (uint)y0Q4 >> SubpelBits;
+ for (y = 0; y < (uint)h; ++y)
+ {
+ ulong srcOffset = (srcBaseY + y) * (uint)srcStride;
+ for (x = 0; x < (uint)w; x += 4)
+ {
+ Vector256<int> vsrc = Avx2.GatherVector256((uint*)&src[srcOffset + x], indices, 1).AsInt32();
+
+ Vector128<int> vsrcL = vsrc.GetLower();
+ Vector128<int> vsrcH = vsrc.GetUpper();
+
+ Vector128<byte> vsrcUnpck11 = Sse2.UnpackLow(vsrcL.AsByte(), vsrcH.AsByte());
+ Vector128<byte> vsrcUnpck12 = Sse2.UnpackHigh(vsrcL.AsByte(), vsrcH.AsByte());
+
+ Vector128<byte> vsrcUnpck21 = Sse2.UnpackLow(vsrcUnpck11, vsrcUnpck12);
+ Vector128<byte> vsrcUnpck22 = Sse2.UnpackHigh(vsrcUnpck11, vsrcUnpck12);
+
+ Vector128<byte> vsrc01 = Sse2.UnpackLow(vsrcUnpck21, vsrcUnpck22);
+ Vector128<byte> vsrc23 = Sse2.UnpackHigh(vsrcUnpck21, vsrcUnpck22);
+
+ Vector128<byte> vsrc11 = Sse.MoveHighToLow(vsrc01.AsSingle(), vsrc01.AsSingle()).AsByte();
+ Vector128<byte> vsrc33 = Sse.MoveHighToLow(vsrc23.AsSingle(), vsrc23.AsSingle()).AsByte();
+
+ Vector128<short> vsrc0 = Sse41.ConvertToVector128Int16(vsrc01);
+ Vector128<short> vsrc1 = Sse41.ConvertToVector128Int16(vsrc11);
+ Vector128<short> vsrc2 = Sse41.ConvertToVector128Int16(vsrc23);
+ Vector128<short> vsrc3 = Sse41.ConvertToVector128Int16(vsrc33);
+
+ Vector128<int> sum0123 = MultiplyAddAdjacent(vsrc0, vsrc1, vsrc2, vsrc3, vfilter, zero);
+
+ Sse.StoreScalar((float*)&dst[x], PackUnsignedSaturate(RoundShift(sum0123, const64), zero).AsSingle());
+ }
+ dst += dstStride;
+ }
+ }
+ }
+
+ [MethodImpl(MethodImplOptions.AggressiveInlining)]
+ private static unsafe void ConvolveVert(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] yFilters,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ if (Avx2.IsSupported && UseIntrinsics && yStepQ4 == 1 << SubpelBits)
+ {
+ ConvolveVertAvx2(src, srcStride, dst, dstStride, yFilters, y0Q4, w, h);
+ return;
+ }
+
+ int x, y;
+ src -= srcStride * (SubpelTaps / 2 - 1);
+
+ for (x = 0; x < w; ++x)
+ {
+ int yQ4 = y0Q4;
+ for (y = 0; y < h; ++y)
+ {
+ byte* srcY = &src[(yQ4 >> SubpelBits) * srcStride];
+ ref Array8<short> yFilter = ref yFilters[yQ4 & SubpelMask];
+ int k, sum = 0;
+ for (k = 0; k < SubpelTaps; ++k)
+ {
+ sum += srcY[k * srcStride] * yFilter[k];
+ }
+
+ dst[y * dstStride] = BitUtils.ClipPixel(BitUtils.RoundPowerOfTwo(sum, FilterBits));
+ yQ4 += yStepQ4;
+ }
+ ++src;
+ ++dst;
+ }
+ }
+
+ private static unsafe void ConvolveAvgVert(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] yFilters,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ int x, y;
+ src -= srcStride * (SubpelTaps / 2 - 1);
+
+ for (x = 0; x < w; ++x)
+ {
+ int yQ4 = y0Q4;
+ for (y = 0; y < h; ++y)
+ {
+ byte* srcY = &src[(yQ4 >> SubpelBits) * srcStride];
+ ref Array8<short> yFilter = ref yFilters[yQ4 & SubpelMask];
+ int k, sum = 0;
+ for (k = 0; k < SubpelTaps; ++k)
+ {
+ sum += srcY[k * srcStride] * yFilter[k];
+ }
+
+ dst[y * dstStride] = (byte)BitUtils.RoundPowerOfTwo(
+ dst[y * dstStride] + BitUtils.ClipPixel(BitUtils.RoundPowerOfTwo(sum, FilterBits)), 1);
+ yQ4 += yStepQ4;
+ }
+ ++src;
+ ++dst;
+ }
+ }
+
+ public static unsafe void Convolve8Horiz(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ ConvolveHoriz(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, w, h);
+ }
+
+ public static unsafe void Convolve8AvgHoriz(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ ConvolveAvgHoriz(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, w, h);
+ }
+
+ public static unsafe void Convolve8Vert(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ ConvolveVert(src, srcStride, dst, dstStride, filter, y0Q4, yStepQ4, w, h);
+ }
+
+ public static unsafe void Convolve8AvgVert(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ ConvolveAvgVert(src, srcStride, dst, dstStride, filter, y0Q4, yStepQ4, w, h);
+ }
+
+ [StructLayout(LayoutKind.Sequential, Size = 64 * 135)]
+ struct Temp
+ {
+ }
+
+ public static unsafe void Convolve8(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ // Note: Fixed size intermediate buffer, temp, places limits on parameters.
+ // 2d filtering proceeds in 2 steps:
+ // (1) Interpolate horizontally into an intermediate buffer, temp.
+ // (2) Interpolate temp vertically to derive the sub-pixel result.
+ // Deriving the maximum number of rows in the temp buffer (135):
+ // --Smallest scaling factor is x1/2 ==> yStepQ4 = 32 (Normative).
+ // --Largest block size is 64x64 pixels.
+ // --64 rows in the downscaled frame span a distance of (64 - 1) * 32 in the
+ // original frame (in 1/16th pixel units).
+ // --Must round-up because block may be located at sub-pixel position.
+ // --Require an additional SubpelTaps rows for the 8-tap filter tails.
+ // --((64 - 1) * 32 + 15) >> 4 + 8 = 135.
+ // When calling in frame scaling function, the smallest scaling factor is x1/4
+ // ==> yStepQ4 = 64. Since w and h are at most 16, the temp buffer is still
+ // big enough.
+ Temp tempStruct;
+ byte* temp = (byte*)Unsafe.AsPointer(ref tempStruct); // Avoid zero initialization.
+ int intermediateHeight = (((h - 1) * yStepQ4 + y0Q4) >> SubpelBits) + SubpelTaps;
+
+ Debug.Assert(w <= 64);
+ Debug.Assert(h <= 64);
+ Debug.Assert(yStepQ4 <= 32 || (yStepQ4 <= 64 && h <= 32));
+ Debug.Assert(xStepQ4 <= 64);
+
+ ConvolveHoriz(src - srcStride * (SubpelTaps / 2 - 1), srcStride, temp, 64, filter, x0Q4, xStepQ4, w, intermediateHeight);
+ ConvolveVert(temp + 64 * (SubpelTaps / 2 - 1), 64, dst, dstStride, filter, y0Q4, yStepQ4, w, h);
+ }
+
+ public static unsafe void Convolve8Avg(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ // Fixed size intermediate buffer places limits on parameters.
+ byte* temp = stackalloc byte[64 * 64];
+ Debug.Assert(w <= 64);
+ Debug.Assert(h <= 64);
+
+ Convolve8(src, srcStride, temp, 64, filter, x0Q4, xStepQ4, y0Q4, yStepQ4, w, h);
+ ConvolveAvg(temp, 64, dst, dstStride, null, 0, 0, 0, 0, w, h);
+ }
+
+ public static unsafe void ConvolveCopy(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ int r;
+
+ for (r = h; r > 0; --r)
+ {
+ MemoryUtil.Copy(dst, src, w);
+ src += srcStride;
+ dst += dstStride;
+ }
+ }
+
+ public static unsafe void ConvolveAvg(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ int x, y;
+
+ for (y = 0; y < h; ++y)
+ {
+ for (x = 0; x < w; ++x)
+ {
+ dst[x] = (byte)BitUtils.RoundPowerOfTwo(dst[x] + src[x], 1);
+ }
+
+ src += srcStride;
+ dst += dstStride;
+ }
+ }
+
+ public static unsafe void ScaledHoriz(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ Convolve8Horiz(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, y0Q4, yStepQ4, w, h);
+ }
+
+ public static unsafe void ScaledVert(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ Convolve8Vert(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, y0Q4, yStepQ4, w, h);
+ }
+
+ public static unsafe void Scaled2D(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ Convolve8(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, y0Q4, yStepQ4, w, h);
+ }
+
+ public static unsafe void ScaledAvgHoriz(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ Convolve8AvgHoriz(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, y0Q4, yStepQ4, w, h);
+ }
+
+ public static unsafe void ScaledAvgVert(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ Convolve8AvgVert(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, y0Q4, yStepQ4, w, h);
+ }
+
+ public static unsafe void ScaledAvg2D(
+ byte* src,
+ int srcStride,
+ byte* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h)
+ {
+ Convolve8Avg(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, y0Q4, yStepQ4, w, h);
+ }
+
+ private static unsafe void HighbdConvolveHoriz(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] xFilters,
+ int x0Q4,
+ int xStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ int x, y;
+ src -= SubpelTaps / 2 - 1;
+
+ for (y = 0; y < h; ++y)
+ {
+ int xQ4 = x0Q4;
+ for (x = 0; x < w; ++x)
+ {
+ ushort* srcX = &src[xQ4 >> SubpelBits];
+ ref Array8<short> xFilter = ref xFilters[xQ4 & SubpelMask];
+ int k, sum = 0;
+ for (k = 0; k < SubpelTaps; ++k)
+ {
+ sum += srcX[k] * xFilter[k];
+ }
+
+ dst[x] = BitUtils.ClipPixelHighbd(BitUtils.RoundPowerOfTwo(sum, FilterBits), bd);
+ xQ4 += xStepQ4;
+ }
+ src += srcStride;
+ dst += dstStride;
+ }
+ }
+
+ private static unsafe void HighbdConvolveAvgHoriz(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] xFilters,
+ int x0Q4,
+ int xStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ int x, y;
+ src -= SubpelTaps / 2 - 1;
+
+ for (y = 0; y < h; ++y)
+ {
+ int xQ4 = x0Q4;
+ for (x = 0; x < w; ++x)
+ {
+ ushort* srcX = &src[xQ4 >> SubpelBits];
+ ref Array8<short> xFilter = ref xFilters[xQ4 & SubpelMask];
+ int k, sum = 0;
+ for (k = 0; k < SubpelTaps; ++k)
+ {
+ sum += srcX[k] * xFilter[k];
+ }
+
+ dst[x] = (ushort)BitUtils.RoundPowerOfTwo(dst[x] + BitUtils.ClipPixelHighbd(BitUtils.RoundPowerOfTwo(sum, FilterBits), bd), 1);
+ xQ4 += xStepQ4;
+ }
+ src += srcStride;
+ dst += dstStride;
+ }
+ }
+
+ private static unsafe void HighbdConvolveVert(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] yFilters,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ int x, y;
+ src -= srcStride * (SubpelTaps / 2 - 1);
+
+ for (x = 0; x < w; ++x)
+ {
+ int yQ4 = y0Q4;
+ for (y = 0; y < h; ++y)
+ {
+ ushort* srcY = &src[(yQ4 >> SubpelBits) * srcStride];
+ ref Array8<short> yFilter = ref yFilters[yQ4 & SubpelMask];
+ int k, sum = 0;
+ for (k = 0; k < SubpelTaps; ++k)
+ {
+ sum += srcY[k * srcStride] * yFilter[k];
+ }
+
+ dst[y * dstStride] = BitUtils.ClipPixelHighbd(BitUtils.RoundPowerOfTwo(sum, FilterBits), bd);
+ yQ4 += yStepQ4;
+ }
+ ++src;
+ ++dst;
+ }
+ }
+
+ private static unsafe void HighConvolveAvgVert(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] yFilters,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ int x, y;
+ src -= srcStride * (SubpelTaps / 2 - 1);
+
+ for (x = 0; x < w; ++x)
+ {
+ int yQ4 = y0Q4;
+ for (y = 0; y < h; ++y)
+ {
+ ushort* srcY = &src[(yQ4 >> SubpelBits) * srcStride];
+ ref Array8<short> yFilter = ref yFilters[yQ4 & SubpelMask];
+ int k, sum = 0;
+ for (k = 0; k < SubpelTaps; ++k)
+ {
+ sum += srcY[k * srcStride] * yFilter[k];
+ }
+
+ dst[y * dstStride] = (ushort)BitUtils.RoundPowerOfTwo(
+ dst[y * dstStride] + BitUtils.ClipPixelHighbd(BitUtils.RoundPowerOfTwo(sum, FilterBits), bd), 1);
+ yQ4 += yStepQ4;
+ }
+ ++src;
+ ++dst;
+ }
+ }
+
+ private static unsafe void HighbdConvolve(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ // Note: Fixed size intermediate buffer, temp, places limits on parameters.
+ // 2d filtering proceeds in 2 steps:
+ // (1) Interpolate horizontally into an intermediate buffer, temp.
+ // (2) Interpolate temp vertically to derive the sub-pixel result.
+ // Deriving the maximum number of rows in the temp buffer (135):
+ // --Smallest scaling factor is x1/2 ==> yStepQ4 = 32 (Normative).
+ // --Largest block size is 64x64 pixels.
+ // --64 rows in the downscaled frame span a distance of (64 - 1) * 32 in the
+ // original frame (in 1/16th pixel units).
+ // --Must round-up because block may be located at sub-pixel position.
+ // --Require an additional SubpelTaps rows for the 8-tap filter tails.
+ // --((64 - 1) * 32 + 15) >> 4 + 8 = 135.
+ ushort* temp = stackalloc ushort[64 * 135];
+ int intermediateHeight = (((h - 1) * yStepQ4 + y0Q4) >> SubpelBits) + SubpelTaps;
+
+ Debug.Assert(w <= 64);
+ Debug.Assert(h <= 64);
+ Debug.Assert(yStepQ4 <= 32);
+ Debug.Assert(xStepQ4 <= 32);
+
+ HighbdConvolveHoriz(src - srcStride * (SubpelTaps / 2 - 1), srcStride, temp, 64, filter, x0Q4, xStepQ4, w, intermediateHeight, bd);
+ HighbdConvolveVert(temp + 64 * (SubpelTaps / 2 - 1), 64, dst, dstStride, filter, y0Q4, yStepQ4, w, h, bd);
+ }
+
+ public static unsafe void HighbdConvolve8Horiz(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ HighbdConvolveHoriz(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, w, h, bd);
+ }
+
+ public static unsafe void HighbdConvolve8AvgHoriz(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ HighbdConvolveAvgHoriz(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, w, h, bd);
+ }
+
+ public static unsafe void HighbdConvolve8Vert(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ HighbdConvolveVert(src, srcStride, dst, dstStride, filter, y0Q4, yStepQ4, w, h, bd);
+ }
+
+ public static unsafe void HighbdConvolve8AvgVert(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ HighConvolveAvgVert(src, srcStride, dst, dstStride, filter, y0Q4, yStepQ4, w, h, bd);
+ }
+
+ public static unsafe void HighbdConvolve8(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ HighbdConvolve(src, srcStride, dst, dstStride, filter, x0Q4, xStepQ4, y0Q4, yStepQ4, w, h, bd);
+ }
+
+ public static unsafe void HighbdConvolve8Avg(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ // Fixed size intermediate buffer places limits on parameters.
+ ushort* temp = stackalloc ushort[64 * 64];
+ Debug.Assert(w <= 64);
+ Debug.Assert(h <= 64);
+
+ HighbdConvolve8(src, srcStride, temp, 64, filter, x0Q4, xStepQ4, y0Q4, yStepQ4, w, h, bd);
+ HighbdConvolveAvg(temp, 64, dst, dstStride, null, 0, 0, 0, 0, w, h, bd);
+ }
+
+ public static unsafe void HighbdConvolveCopy(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ int r;
+
+ for (r = h; r > 0; --r)
+ {
+ MemoryUtil.Copy(dst, src, w);
+ src += srcStride;
+ dst += dstStride;
+ }
+ }
+
+ public static unsafe void HighbdConvolveAvg(
+ ushort* src,
+ int srcStride,
+ ushort* dst,
+ int dstStride,
+ Array8<short>[] filter,
+ int x0Q4,
+ int xStepQ4,
+ int y0Q4,
+ int yStepQ4,
+ int w,
+ int h,
+ int bd)
+ {
+ int x, y;
+
+ for (y = 0; y < h; ++y)
+ {
+ for (x = 0; x < w; ++x)
+ {
+ dst[x] = (ushort)BitUtils.RoundPowerOfTwo(dst[x] + src[x], 1);
+ }
+
+ src += srcStride;
+ dst += dstStride;
+ }
+ }
+ }
+}