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>

2 files changed, 130 insertions, 0 deletions

diff --git a/Ryujinx.Graphics.Nvdec/Types/H264/PictureInfo.cs b/Ryujinx.Graphics.Nvdec/Types/H264/PictureInfo.cs
new file mode 100644
index 00000000..92767e35
--- /dev/null
+++ b/Ryujinx.Graphics.Nvdec/Types/H264/PictureInfo.cs
@@ -0,0 +1,120 @@
+using Ryujinx.Common.Memory;
+using Ryujinx.Graphics.Video;
+
+namespace Ryujinx.Graphics.Nvdec.Types.H264
+{
+    struct PictureInfo
+    {
+        Array18<uint> Unknown0;
+        public uint BitstreamSize;
+        public uint NumSlices;
+        public uint Unknown50;
+        public uint Unknown54;
+        public uint Log2MaxPicOrderCntLsbMinus4;
+        public uint DeltaPicOrderAlwaysZeroFlag;
+        public uint FrameMbsOnlyFlag;
+        public uint PicWidthInMbs;
+        public uint PicHeightInMbs;
+        public uint BlockLayout; // Not supported on T210
+        public uint EntropyCodingModeFlag;
+        public uint PicOrderPresentFlag;
+        public uint NumRefIdxL0ActiveMinus1;
+        public uint NumRefIdxL1ActiveMinus1;
+        public uint DeblockingFilterControlPresentFlag;
+        public uint RedundantPicCntPresentFlag;
+        public uint Transform8x8ModeFlag;
+        public uint LumaPitch;
+        public uint ChromaPitch;
+        public uint Unknown94;
+        public uint LumaSecondFieldOffset;
+        public uint Unknown9C;
+        public uint UnknownA0;
+        public uint ChromaSecondFieldOffset;
+        public uint UnknownA8;
+        public uint UnknownAC;
+        public ulong Flags;
+        public Array2<int> FieldOrderCnt;
+        public Array16<ReferenceFrame> RefFrames;
+        public Array6<Array16<byte>> ScalingLists4x4;
+        public Array2<Array64<byte>> ScalingLists8x8;
+        public byte MvcextNumInterViewRefsL0;
+        public byte MvcextNumInterViewRefsL1;
+        public ushort Padding2A2;
+        public uint Unknown2A4;
+        public uint Unknown2A8;
+        public uint Unknown2AC;
+        public Array16<byte> MvcextViewRefMasksL0;
+        public Array16<byte> MvcextViewRefMasksL1;
+        public uint Flags2;
+        public Array10<uint> Unknown2D4;
+
+        public bool MbAdaptiveFrameFieldFlag => (Flags & (1 << 0)) != 0;
+        public bool Direct8x8InferenceFlag => (Flags & (1 << 1)) != 0;
+        public bool WeightedPredFlag => (Flags & (1 << 2)) != 0;
+        public bool ConstrainedIntraPredFlag => (Flags & (1 << 3)) != 0;
+        public bool IsReference => (Flags & (1 << 4)) != 0;
+        public bool FieldPicFlag => (Flags & (1 << 5)) != 0;
+        public bool BottomFieldFlag => (Flags & (1 << 6)) != 0;
+        public uint Log2MaxFrameNumMinus4 => (uint)(Flags >> 8) & 0xf;
+        public ushort ChromaFormatIdc => (ushort)((Flags >> 12) & 3);
+        public uint PicOrderCntType => (uint)(Flags >> 14) & 3;
+        public int PicInitQpMinus26 => ExtractSx(Flags, 16, 6);
+        public int ChromaQpIndexOffset => ExtractSx(Flags, 22, 5);
+        public int SecondChromaQpIndexOffset => ExtractSx(Flags, 27, 5);
+        public uint WeightedBipredIdc => (uint)(Flags >> 32) & 3;
+        public uint LumaOutputSurfaceIndex => (uint)(Flags >> 34) & 0x7f;
+        public uint ChromaOutputSurfaceIndex => (uint)(Flags >> 41) & 0x1f;
+        public ushort FrameNum => (ushort)(Flags >> 46);
+        public bool QpprimeYZeroTransformBypassFlag => (Flags2 & (1 << 1)) != 0;
+
+        private static int ExtractSx(ulong packed, int lsb, int length)
+        {
+            return (int)((long)packed << (64 - (lsb + length)) >> (64 - length));
+        }
+
+        public H264PictureInfo Convert()
+        {
+            return new H264PictureInfo()
+            {
+                FieldOrderCnt = FieldOrderCnt,
+                IsReference = IsReference,
+                ChromaFormatIdc = ChromaFormatIdc,
+                FrameNum = FrameNum,
+                FieldPicFlag = FieldPicFlag,
+                BottomFieldFlag = BottomFieldFlag,
+                NumRefFrames = 0,
+                MbAdaptiveFrameFieldFlag = MbAdaptiveFrameFieldFlag,
+                ConstrainedIntraPredFlag = ConstrainedIntraPredFlag,
+                WeightedPredFlag = WeightedPredFlag,
+                WeightedBipredIdc = WeightedBipredIdc,
+                FrameMbsOnlyFlag = FrameMbsOnlyFlag != 0,
+                Transform8x8ModeFlag = Transform8x8ModeFlag != 0,
+                ChromaQpIndexOffset = ChromaQpIndexOffset,
+                SecondChromaQpIndexOffset = SecondChromaQpIndexOffset,
+                PicInitQpMinus26 = PicInitQpMinus26,
+                NumRefIdxL0ActiveMinus1 = NumRefIdxL0ActiveMinus1,
+                NumRefIdxL1ActiveMinus1 = NumRefIdxL1ActiveMinus1,
+                Log2MaxFrameNumMinus4 = Log2MaxFrameNumMinus4,
+                PicOrderCntType = PicOrderCntType,
+                Log2MaxPicOrderCntLsbMinus4 = Log2MaxPicOrderCntLsbMinus4,
+                DeltaPicOrderAlwaysZeroFlag = DeltaPicOrderAlwaysZeroFlag != 0,
+                Direct8x8InferenceFlag = Direct8x8InferenceFlag,
+                EntropyCodingModeFlag = EntropyCodingModeFlag != 0,
+                PicOrderPresentFlag = PicOrderPresentFlag != 0,
+                DeblockingFilterControlPresentFlag = DeblockingFilterControlPresentFlag != 0,
+                RedundantPicCntPresentFlag = RedundantPicCntPresentFlag != 0,
+                NumSliceGroupsMinus1 = 0,
+                SliceGroupMapType = 0,
+                SliceGroupChangeRateMinus1 = 0,
+                FmoAsoEnable = false,
+                ScalingMatrixPresent = true,
+                ScalingLists4x4 = ScalingLists4x4,
+                ScalingLists8x8 = ScalingLists8x8,
+                FrameType = 0,
+                PicWidthInMbsMinus1 = PicWidthInMbs - 1,
+                PicHeightInMapUnitsMinus1 = (PicHeightInMbs >> (FrameMbsOnlyFlag != 0 ? 0 : 1)) - 1,
+                QpprimeYZeroTransformBypassFlag = QpprimeYZeroTransformBypassFlag
+            };
+        }
+    }
+}
diff --git a/Ryujinx.Graphics.Nvdec/Types/H264/ReferenceFrame.cs b/Ryujinx.Graphics.Nvdec/Types/H264/ReferenceFrame.cs
new file mode 100644
index 00000000..5db311ae
--- /dev/null
+++ b/Ryujinx.Graphics.Nvdec/Types/H264/ReferenceFrame.cs
@@ -0,0 +1,10 @@
+namespace Ryujinx.Graphics.Nvdec.Types.H264
+{
+    struct ReferenceFrame
+    {
+        public uint Unknown0;
+        public uint Unknown4;
+        public uint Unknown8;
+        public uint UnknownC;
+    }
+}