Implement DMA texture copy component shuffle (#4717)

* Implement DMA texture copy component shuffle

* Set UInt24 alignment to 1

Ryujinx.Audio/Renderer/Server/StateUpdater.cs

@@ -11,6 +11,7 @@ using Ryujinx.Audio.Renderer.Server.Voice;
 using Ryujinx.Audio.Renderer.Utils;
 using Ryujinx.Common.Logging;
 using System;
+using System.Buffers;
 using System.Diagnostics;
 using System.Runtime.CompilerServices;
 using System.Runtime.InteropServices;
@@ -149,12 +150,16 @@ namespace Ryujinx.Audio.Renderer.Server
                 state.InUse = false;
             }
 
+            Memory<VoiceUpdateState>[] voiceUpdateStatesArray = ArrayPool<Memory<VoiceUpdateState>>.Shared.Rent(Constants.VoiceChannelCountMax);
+            
+            Span<Memory<VoiceUpdateState>> voiceUpdateStates = voiceUpdateStatesArray.AsSpan(0, Constants.VoiceChannelCountMax);
+
             // Start processing
             for (int i = 0; i < context.GetCount(); i++)
             {
                 VoiceInParameter parameter = parameters[i];
 
-                Memory<VoiceUpdateState>[] voiceUpdateStates = new Memory<VoiceUpdateState>[Constants.VoiceChannelCountMax];
+                voiceUpdateStates.Fill(Memory<VoiceUpdateState>.Empty);
 
                 ref VoiceOutStatus outStatus = ref SpanIOHelper.GetWriteRef<VoiceOutStatus>(ref _output)[0];
 
@@ -197,6 +202,8 @@ namespace Ryujinx.Audio.Renderer.Server
                 }
             }
 
+            ArrayPool<Memory<VoiceUpdateState>>.Shared.Return(voiceUpdateStatesArray);
+
             int currentOutputSize = _output.Length;
 
             OutputHeader.VoicesSize = (uint)(Unsafe.SizeOf<VoiceOutStatus>() * context.GetCount());

Ryujinx.Audio/Renderer/Server/Voice/VoiceState.cs

@@ -378,7 +378,7 @@ namespace Ryujinx.Audio.Renderer.Server.Voice
         /// <param name="outStatus">The given user output.</param>
         /// <param name="parameter">The user parameter.</param>
         /// <param name="voiceUpdateStates">The voice states associated to the <see cref="VoiceState"/>.</param>
-        public void WriteOutStatus(ref VoiceOutStatus outStatus, ref VoiceInParameter parameter, Memory<VoiceUpdateState>[] voiceUpdateStates)
+        public void WriteOutStatus(ref VoiceOutStatus outStatus, ref VoiceInParameter parameter, ReadOnlySpan<Memory<VoiceUpdateState>> voiceUpdateStates)
         {
 #if DEBUG
             // Sanity check in debug mode of the internal state
@@ -424,7 +424,7 @@ namespace Ryujinx.Audio.Renderer.Server.Voice
         /// <param name="voiceUpdateStates">The voice states associated to the <see cref="VoiceState"/>.</param>
         /// <param name="mapper">The mapper to use.</param>
         /// <param name="behaviourContext">The behaviour context.</param>
-        public void UpdateWaveBuffers(out ErrorInfo[] errorInfos, ref VoiceInParameter parameter, Memory<VoiceUpdateState>[] voiceUpdateStates, ref PoolMapper mapper, ref BehaviourContext behaviourContext)
+        public void UpdateWaveBuffers(out ErrorInfo[] errorInfos, ref VoiceInParameter parameter, ReadOnlySpan<Memory<VoiceUpdateState>> voiceUpdateStates, ref PoolMapper mapper, ref BehaviourContext behaviourContext)
         {
             errorInfos = new ErrorInfo[Constants.VoiceWaveBufferCount * 2];
 

Ryujinx.Ava/UI/ViewModels/AmiiboWindowViewModel.cs

@@ -7,6 +7,7 @@ using Ryujinx.Ava.UI.Helpers;
 using Ryujinx.Ava.UI.Windows;
 using Ryujinx.Common;
 using Ryujinx.Common.Configuration;
+using Ryujinx.Common.Logging;
 using Ryujinx.Common.Utilities;
 using Ryujinx.Ui.Common.Models.Amiibo;
 using System;
@@ -48,7 +49,12 @@ namespace Ryujinx.Ava.UI.ViewModels
         public AmiiboWindowViewModel(StyleableWindow owner, string lastScannedAmiiboId, string titleId)
         {
             _owner = owner;
-            _httpClient = new HttpClient { Timeout = TimeSpan.FromMilliseconds(5000) };
+
+            _httpClient = new HttpClient
+            {
+                Timeout = TimeSpan.FromSeconds(30)
+            };
+
             LastScannedAmiiboId = lastScannedAmiiboId;
             TitleId = titleId;
 
@@ -89,9 +95,7 @@ namespace Ryujinx.Ava.UI.ViewModels
             {
                 _showAllAmiibo = value;
 
-#pragma warning disable 4014
                 ParseAmiiboData();
-#pragma warning restore 4014
 
                 OnPropertyChanged();
             }
@@ -203,8 +207,10 @@ namespace Ryujinx.Ava.UI.ViewModels
                 {
                     amiiboJsonString = await DownloadAmiiboJson();
                 }
-                catch
+                catch (Exception ex)
                 {
+                    Logger.Error?.Print(LogClass.Application, $"Failed to download amiibo data: {ex}");
+
                     ShowInfoDialog();
                 }
             }
@@ -369,8 +375,10 @@ namespace Ryujinx.Ava.UI.ViewModels
 
                 return false;
             }
-            catch
+            catch (Exception ex)
             {
+                Logger.Error?.Print(LogClass.Application, $"Failed to check for amiibo updates: {ex}");
+
                 ShowInfoDialog();
 
                 return false;
@@ -393,6 +401,8 @@ namespace Ryujinx.Ava.UI.ViewModels
                 return amiiboJsonString;
             }
 
+            Logger.Error?.Print(LogClass.Application, $"Failed to download amiibo data. Response status code: {response.StatusCode}");
+
             await ContentDialogHelper.CreateInfoDialog(LocaleManager.Instance[LocaleKeys.DialogAmiiboApiTitle],
                 LocaleManager.Instance[LocaleKeys.DialogAmiiboApiFailFetchMessage],
                 LocaleManager.Instance[LocaleKeys.InputDialogOk],
@@ -429,6 +439,10 @@ namespace Ryujinx.Ava.UI.ViewModels
                     AmiiboImage = bitmap.CreateScaledBitmap(new PixelSize(resizeWidth, resizeHeight));
                 }
             }
+            else
+            {
+                Logger.Error?.Print(LogClass.Application, $"Failed to get amiibo preview. Response status code: {response.StatusCode}");
+            }
         }
 
         private void ResetAmiiboPreview()

Ryujinx.Common/Memory/ByteMemoryPool.ByteMemoryPoolBuffer.cs

@@ -0,0 +1,51 @@
+using System;
+using System.Buffers;
+using System.Threading;
+
+namespace Ryujinx.Common.Memory
+{
+    public sealed partial class ByteMemoryPool
+    {
+        /// <summary>
+        /// Represents a <see cref="IMemoryOwner{Byte}"/> that wraps an array rented from
+        /// <see cref="ArrayPool{Byte}.Shared"/> and exposes it as <see cref="Memory{Byte}"/>
+        /// with a length of the requested size.
+        /// </summary>
+        private sealed class ByteMemoryPoolBuffer : IMemoryOwner<byte>
+        {
+            private byte[] _array;
+            private readonly int _length;
+
+            public ByteMemoryPoolBuffer(int length)
+            {
+                _array = ArrayPool<byte>.Shared.Rent(length);
+                _length = length;
+            }
+
+            /// <summary>
+            /// Returns a <see cref="Memory{Byte}"/> belonging to this owner.
+            /// </summary>
+            public Memory<byte> Memory
+            {
+                get
+                {
+                    byte[] array = _array;
+                    
+                    ObjectDisposedException.ThrowIf(array is null, this);
+
+                    return new Memory<byte>(array, 0, _length);
+                }
+            }
+
+            public void Dispose()
+            {
+                var array = Interlocked.Exchange(ref _array, null);
+                
+                if (array != null)
+                {
+                    ArrayPool<byte>.Shared.Return(array);
+                }
+            }
+        }
+    }
+}

Ryujinx.Common/Memory/ByteMemoryPool.cs

@@ -0,0 +1,108 @@
+using System;
+using System.Buffers;
+
+namespace Ryujinx.Common.Memory
+{
+    /// <summary>
+    /// Provides a pool of re-usable byte array instances.
+    /// </summary>
+    public sealed partial class ByteMemoryPool
+    {
+        private static readonly ByteMemoryPool _shared = new ByteMemoryPool();
+
+        /// <summary>
+        /// Constructs a <see cref="ByteMemoryPool"/> instance. Private to force access through
+        /// the <see cref="ByteMemoryPool.Shared"/> instance.
+        /// </summary>
+        private ByteMemoryPool()
+        {
+            // No implementation
+        }
+
+        /// <summary>
+        /// Retrieves a shared <see cref="ByteMemoryPool"/> instance.
+        /// </summary>
+        public static ByteMemoryPool Shared => _shared;
+
+        /// <summary>
+        /// Returns the maximum buffer size supported by this pool.
+        /// </summary>
+        public int MaxBufferSize => Array.MaxLength;
+
+        /// <summary>
+        /// Rents a byte memory buffer from <see cref="ArrayPool{Byte}.Shared"/>.
+        /// The buffer may contain data from a prior use.
+        /// </summary>
+        /// <param name="length">The buffer's required length in bytes</param>
+        /// <returns>A <see cref="IMemoryOwner{Byte}"/> wrapping the rented memory</returns>
+        /// <exception cref="ArgumentOutOfRangeException"></exception>
+        public IMemoryOwner<byte> Rent(long length)
+            => RentImpl(checked((int)length));
+
+        /// <summary>
+        /// Rents a byte memory buffer from <see cref="ArrayPool{Byte}.Shared"/>.
+        /// The buffer may contain data from a prior use.
+        /// </summary>
+        /// <param name="length">The buffer's required length in bytes</param>
+        /// <returns>A <see cref="IMemoryOwner{Byte}"/> wrapping the rented memory</returns>
+        /// <exception cref="ArgumentOutOfRangeException"></exception>
+        public IMemoryOwner<byte> Rent(ulong length)
+            => RentImpl(checked((int)length));
+
+        /// <summary>
+        /// Rents a byte memory buffer from <see cref="ArrayPool{Byte}.Shared"/>.
+        /// The buffer may contain data from a prior use.
+        /// </summary>
+        /// <param name="length">The buffer's required length in bytes</param>
+        /// <returns>A <see cref="IMemoryOwner{Byte}"/> wrapping the rented memory</returns>
+        /// <exception cref="ArgumentOutOfRangeException"></exception>
+        public IMemoryOwner<byte> Rent(int length)
+            => RentImpl(length);
+
+        /// <summary>
+        /// Rents a byte memory buffer from <see cref="ArrayPool{Byte}.Shared"/>.
+        /// The buffer's contents are cleared (set to all 0s) before returning.
+        /// </summary>
+        /// <param name="length">The buffer's required length in bytes</param>
+        /// <returns>A <see cref="IMemoryOwner{Byte}"/> wrapping the rented memory</returns>
+        /// <exception cref="ArgumentOutOfRangeException"></exception>
+        public IMemoryOwner<byte> RentCleared(long length)
+            => RentCleared(checked((int)length));
+
+        /// <summary>
+        /// Rents a byte memory buffer from <see cref="ArrayPool{Byte}.Shared"/>.
+        /// The buffer's contents are cleared (set to all 0s) before returning.
+        /// </summary>
+        /// <param name="length">The buffer's required length in bytes</param>
+        /// <returns>A <see cref="IMemoryOwner{Byte}"/> wrapping the rented memory</returns>
+        /// <exception cref="ArgumentOutOfRangeException"></exception>
+        public IMemoryOwner<byte> RentCleared(ulong length)
+            => RentCleared(checked((int)length));
+
+        /// <summary>
+        /// Rents a byte memory buffer from <see cref="ArrayPool{Byte}.Shared"/>.
+        /// The buffer's contents are cleared (set to all 0s) before returning.
+        /// </summary>
+        /// <param name="length">The buffer's required length in bytes</param>
+        /// <returns>A <see cref="IMemoryOwner{Byte}"/> wrapping the rented memory</returns>
+        /// <exception cref="ArgumentOutOfRangeException"></exception>
+        public IMemoryOwner<byte> RentCleared(int length)
+        {
+            var buffer = RentImpl(length);
+            
+            buffer.Memory.Span.Clear();
+            
+            return buffer;
+        }
+
+        private static ByteMemoryPoolBuffer RentImpl(int length)
+        {
+            if ((uint)length > Array.MaxLength)
+            {
+                throw new ArgumentOutOfRangeException(nameof(length), length, null);
+            }
+
+            return new ByteMemoryPoolBuffer(length);
+        }
+    }
+}

Ryujinx.Graphics.Gpu/Engine/Dma/DmaClass.cs

@@ -6,6 +6,7 @@ using Ryujinx.Graphics.Texture;
 using System;
 using System.Collections.Generic;
 using System.Runtime.CompilerServices;
+using System.Runtime.InteropServices;
 using System.Runtime.Intrinsics;
 
 namespace Ryujinx.Graphics.Gpu.Engine.Dma
@@ -32,6 +33,69 @@ namespace Ryujinx.Graphics.Gpu.Engine.Dma
             RemapEnable = 1 << 10
         }
 
+        /// <summary>
+        /// Texture parameters for copy.
+        /// </summary>
+        private struct TextureParams
+        {
+            /// <summary>
+            /// Copy region X coordinate.
+            /// </summary>
+            public readonly int RegionX;
+
+            /// <summary>
+            /// Copy region Y coordinate.
+            /// </summary>
+            public readonly int RegionY;
+
+            /// <summary>
+            /// Offset from the base pointer of the data in memory.
+            /// </summary>
+            public readonly int BaseOffset;
+
+            /// <summary>
+            /// Bytes per pixel.
+            /// </summary>
+            public readonly int Bpp;
+
+            /// <summary>
+            /// Whether the texture is linear. If false, the texture is block linear.
+            /// </summary>
+            public readonly bool Linear;
+
+            /// <summary>
+            /// Pixel offset from XYZ coordinates calculator.
+            /// </summary>
+            public readonly OffsetCalculator Calculator;
+
+            /// <summary>
+            /// Creates texture parameters.
+            /// </summary>
+            /// <param name="regionX">Copy region X coordinate</param>
+            /// <param name="regionY">Copy region Y coordinate</param>
+            /// <param name="baseOffset">Offset from the base pointer of the data in memory</param>
+            /// <param name="bpp">Bytes per pixel</param>
+            /// <param name="linear">Whether the texture is linear. If false, the texture is block linear</param>
+            /// <param name="calculator">Pixel offset from XYZ coordinates calculator</param>
+            public TextureParams(int regionX, int regionY, int baseOffset, int bpp, bool linear, OffsetCalculator calculator)
+            {
+                RegionX = regionX;
+                RegionY = regionY;
+                BaseOffset = baseOffset;
+                Bpp = bpp;
+                Linear = linear;
+                Calculator = calculator;
+            }
+        }
+
+        [StructLayout(LayoutKind.Sequential, Size = 3, Pack = 1)]
+        private struct UInt24
+        {
+            public byte Byte0;
+            public byte Byte1;
+            public byte Byte2;
+        }
+
         /// <summary>
         /// Creates a new instance of the DMA copy engine class.
         /// </summary>
@@ -154,8 +218,10 @@ namespace Ryujinx.Graphics.Gpu.Engine.Dma
             {
                 // Buffer to texture copy.
                 int componentSize = (int)_state.State.SetRemapComponentsComponentSize + 1;
-                int srcBpp = remap ? ((int)_state.State.SetRemapComponentsNumSrcComponents + 1) * componentSize : 1;
-                int dstBpp = remap ? ((int)_state.State.SetRemapComponentsNumDstComponents + 1) * componentSize : 1;
+                int srcComponents = (int)_state.State.SetRemapComponentsNumSrcComponents + 1;
+                int dstComponents = (int)_state.State.SetRemapComponentsNumDstComponents + 1;
+                int srcBpp = remap ? srcComponents * componentSize : 1;
+                int dstBpp = remap ? dstComponents * componentSize : 1;
 
                 var dst = Unsafe.As<uint, DmaTexture>(ref _state.State.SetDstBlockSize);
                 var src = Unsafe.As<uint, DmaTexture>(ref _state.State.SetSrcBlockSize);
@@ -274,63 +340,51 @@ namespace Ryujinx.Graphics.Gpu.Engine.Dma
                     }
                 }
 
-                unsafe bool Convert<T>(Span<byte> dstSpan, ReadOnlySpan<byte> srcSpan) where T : unmanaged
-                {
-                    if (srcLinear && dstLinear && srcBpp == dstBpp)
-                    {
-                        // Optimized path for purely linear copies - we don't need to calculate every single byte offset,
-                        // and we can make use of Span.CopyTo which is very very fast (even compared to pointers)
-                        for (int y = 0; y < yCount; y++)
-                        {
-                            srcCalculator.SetY(srcRegionY + y);
-                            dstCalculator.SetY(dstRegionY + y);
-                            int srcOffset = srcCalculator.GetOffset(srcRegionX);
-                            int dstOffset = dstCalculator.GetOffset(dstRegionX);
-                            srcSpan.Slice(srcOffset - srcBaseOffset, xCount * srcBpp)
-                                .CopyTo(dstSpan.Slice(dstOffset - dstBaseOffset, xCount * dstBpp));
-                        }
-                    }
-                    else
-                    {
-                        fixed (byte* dstPtr = dstSpan, srcPtr = srcSpan)
-                        {
-                            byte* dstBase = dstPtr - dstBaseOffset; // Layout offset is relative to the base, so we need to subtract the span's offset.
-                            byte* srcBase = srcPtr - srcBaseOffset;
-
-                            for (int y = 0; y < yCount; y++)
-                            {
-                                srcCalculator.SetY(srcRegionY + y);
-                                dstCalculator.SetY(dstRegionY + y);
-
-                                for (int x = 0; x < xCount; x++)
-                                {
-                                    int srcOffset = srcCalculator.GetOffset(srcRegionX + x);
-                                    int dstOffset = dstCalculator.GetOffset(dstRegionX + x);
-
-                                    *(T*)(dstBase + dstOffset) = *(T*)(srcBase + srcOffset);
-                                }
-                            }
-                        }
-                    }
-
-                    return true;
-                }
-
                 // OPT: This allocates a (potentially) huge temporary array and then copies an existing
                 // region of memory into it, data that might get overwritten entirely anyways. Ideally this should
                 // all be rewritten to use pooled arrays, but that gets complicated with packed data and strides
                 Span<byte> dstSpan = memoryManager.GetSpan(dstGpuVa + (ulong)dstBaseOffset, dstSize).ToArray();
 
-                bool _ = srcBpp switch
+                TextureParams srcParams = new TextureParams(srcRegionX, srcRegionY, srcBaseOffset, srcBpp, srcLinear, srcCalculator);
+                TextureParams dstParams = new TextureParams(dstRegionX, dstRegionY, dstBaseOffset, dstBpp, dstLinear, dstCalculator);
+
+                // If remapping is enabled, we always copy the components directly, in order.
+                // If it's enabled, but the mapping is just XYZW, we also copy them in order.
+                bool isIdentityRemap = !remap ||
+                    (_state.State.SetRemapComponentsDstX == SetRemapComponentsDst.SrcX &&
+                    (dstComponents < 2 || _state.State.SetRemapComponentsDstY == SetRemapComponentsDst.SrcY) &&
+                    (dstComponents < 3 || _state.State.SetRemapComponentsDstZ == SetRemapComponentsDst.SrcZ) &&
+                    (dstComponents < 4 || _state.State.SetRemapComponentsDstW == SetRemapComponentsDst.SrcW));
+
+                if (isIdentityRemap)
                 {
-                    1 => Convert<byte>(dstSpan, srcSpan),
-                    2 => Convert<ushort>(dstSpan, srcSpan),
-                    4 => Convert<uint>(dstSpan, srcSpan),
-                    8 => Convert<ulong>(dstSpan, srcSpan),
-                    12 => Convert<Bpp12Pixel>(dstSpan, srcSpan),
-                    16 => Convert<Vector128<byte>>(dstSpan, srcSpan),
-                    _ => throw new NotSupportedException($"Unable to copy ${srcBpp} bpp pixel format.")
-                };
+                    // The order of the components doesn't change, so we can just copy directly
+                    // (with layout conversion if necessary).
+
+                    switch (srcBpp)
+                    {
+                        case 1: Copy<byte>(dstSpan, srcSpan, dstParams, srcParams); break;
+                        case 2: Copy<ushort>(dstSpan, srcSpan, dstParams, srcParams); break;
+                        case 4: Copy<uint>(dstSpan, srcSpan, dstParams, srcParams); break;
+                        case 8: Copy<ulong>(dstSpan, srcSpan, dstParams, srcParams); break;
+                        case 12: Copy<Bpp12Pixel>(dstSpan, srcSpan, dstParams, srcParams); break;
+                        case 16: Copy<Vector128<byte>>(dstSpan, srcSpan, dstParams, srcParams); break;
+                        default: throw new NotSupportedException($"Unable to copy ${srcBpp} bpp pixel format.");
+                    }
+                }
+                else
+                {
+                    // The order or value of the components might change.
+
+                    switch (componentSize)
+                    {
+                        case 1: CopyShuffle<byte>(dstSpan, srcSpan, dstParams, srcParams); break;
+                        case 2: CopyShuffle<ushort>(dstSpan, srcSpan, dstParams, srcParams); break;
+                        case 3: CopyShuffle<UInt24>(dstSpan, srcSpan, dstParams, srcParams); break;
+                        case 4: CopyShuffle<uint>(dstSpan, srcSpan, dstParams, srcParams); break;
+                        default: throw new NotSupportedException($"Unable to copy ${componentSize} component size.");
+                    }
+                }
 
                 memoryManager.Write(dstGpuVa + (ulong)dstBaseOffset, dstSpan);
             }
@@ -372,6 +426,133 @@ namespace Ryujinx.Graphics.Gpu.Engine.Dma
             }
         }
 
+        /// <summary>
+        /// Copies data from one texture to another, while performing layout conversion if necessary.
+        /// </summary>
+        /// <typeparam name="T">Pixel type</typeparam>
+        /// <param name="dstSpan">Destination texture memory region</param>
+        /// <param name="srcSpan">Source texture memory region</param>
+        /// <param name="dst">Destination texture parameters</param>
+        /// <param name="src">Source texture parameters</param>
+        private unsafe void Copy<T>(Span<byte> dstSpan, ReadOnlySpan<byte> srcSpan, TextureParams dst, TextureParams src) where T : unmanaged
+        {
+            int xCount = (int)_state.State.LineLengthIn;
+            int yCount = (int)_state.State.LineCount;
+
+            if (src.Linear && dst.Linear && src.Bpp == dst.Bpp)
+            {
+                // Optimized path for purely linear copies - we don't need to calculate every single byte offset,
+                // and we can make use of Span.CopyTo which is very very fast (even compared to pointers)
+                for (int y = 0; y < yCount; y++)
+                {
+                    src.Calculator.SetY(src.RegionY + y);
+                    dst.Calculator.SetY(dst.RegionY + y);
+                    int srcOffset = src.Calculator.GetOffset(src.RegionX);
+                    int dstOffset = dst.Calculator.GetOffset(dst.RegionX);
+                    srcSpan.Slice(srcOffset - src.BaseOffset, xCount * src.Bpp)
+                        .CopyTo(dstSpan.Slice(dstOffset - dst.BaseOffset, xCount * dst.Bpp));
+                }
+            }
+            else
+            {
+                fixed (byte* dstPtr = dstSpan, srcPtr = srcSpan)
+                {
+                    byte* dstBase = dstPtr - dst.BaseOffset; // Layout offset is relative to the base, so we need to subtract the span's offset.
+                    byte* srcBase = srcPtr - src.BaseOffset;
+
+                    for (int y = 0; y < yCount; y++)
+                    {
+                        src.Calculator.SetY(src.RegionY + y);
+                        dst.Calculator.SetY(dst.RegionY + y);
+
+                        for (int x = 0; x < xCount; x++)
+                        {
+                            int srcOffset = src.Calculator.GetOffset(src.RegionX + x);
+                            int dstOffset = dst.Calculator.GetOffset(dst.RegionX + x);
+
+                            *(T*)(dstBase + dstOffset) = *(T*)(srcBase + srcOffset);
+                        }
+                    }
+                }
+            }
+        }
+
+        /// <summary>
+        /// Sets texture pixel data to a constant value, while performing layout conversion if necessary.
+        /// </summary>
+        /// <typeparam name="T">Pixel type</typeparam>
+        /// <param name="dstSpan">Destination texture memory region</param>
+        /// <param name="dst">Destination texture parameters</param>
+        /// <param name="fillValue">Constant pixel value to be set</param>
+        private unsafe void Fill<T>(Span<byte> dstSpan, TextureParams dst, T fillValue) where T : unmanaged
+        {
+            int xCount = (int)_state.State.LineLengthIn;
+            int yCount = (int)_state.State.LineCount;
+
+            fixed (byte* dstPtr = dstSpan)
+            {
+                byte* dstBase = dstPtr - dst.BaseOffset; // Layout offset is relative to the base, so we need to subtract the span's offset.
+
+                for (int y = 0; y < yCount; y++)
+                {
+                    dst.Calculator.SetY(dst.RegionY + y);
+
+                    for (int x = 0; x < xCount; x++)
+                    {
+                        int dstOffset = dst.Calculator.GetOffset(dst.RegionX + x);
+
+                        *(T*)(dstBase + dstOffset) = fillValue;
+                    }
+                }
+            }
+        }
+
+        /// <summary>
+        /// Copies data from one texture to another, while performing layout conversion and component shuffling if necessary.
+        /// </summary>
+        /// <typeparam name="T">Pixel type</typeparam>
+        /// <param name="dstSpan">Destination texture memory region</param>
+        /// <param name="srcSpan">Source texture memory region</param>
+        /// <param name="dst">Destination texture parameters</param>
+        /// <param name="src">Source texture parameters</param>
+        private void CopyShuffle<T>(Span<byte> dstSpan, ReadOnlySpan<byte> srcSpan, TextureParams dst, TextureParams src) where T : unmanaged
+        {
+            int dstComponents = (int)_state.State.SetRemapComponentsNumDstComponents + 1;
+
+            for (int i = 0; i < dstComponents; i++)
+            {
+                SetRemapComponentsDst componentsDst = i switch
+                {
+                    0 => _state.State.SetRemapComponentsDstX,
+                    1 => _state.State.SetRemapComponentsDstY,
+                    2 => _state.State.SetRemapComponentsDstZ,
+                    _ => _state.State.SetRemapComponentsDstW
+                };
+
+                switch (componentsDst)
+                {
+                    case SetRemapComponentsDst.SrcX:
+                        Copy<T>(dstSpan.Slice(Unsafe.SizeOf<T>() * i), srcSpan, dst, src);
+                        break;
+                    case SetRemapComponentsDst.SrcY:
+                        Copy<T>(dstSpan.Slice(Unsafe.SizeOf<T>() * i), srcSpan.Slice(Unsafe.SizeOf<T>()), dst, src);
+                        break;
+                    case SetRemapComponentsDst.SrcZ:
+                        Copy<T>(dstSpan.Slice(Unsafe.SizeOf<T>() * i), srcSpan.Slice(Unsafe.SizeOf<T>() * 2), dst, src);
+                        break;
+                    case SetRemapComponentsDst.SrcW:
+                        Copy<T>(dstSpan.Slice(Unsafe.SizeOf<T>() * i), srcSpan.Slice(Unsafe.SizeOf<T>() * 3), dst, src);
+                        break;
+                    case SetRemapComponentsDst.ConstA:
+                        Fill<T>(dstSpan.Slice(Unsafe.SizeOf<T>() * i), dst, Unsafe.As<uint, T>(ref _state.State.SetRemapConstA));
+                        break;
+                    case SetRemapComponentsDst.ConstB:
+                        Fill<T>(dstSpan.Slice(Unsafe.SizeOf<T>() * i), dst, Unsafe.As<uint, T>(ref _state.State.SetRemapConstB));
+                        break;
+                }
+            }
+        }
+
         /// <summary>
         /// Copies block linear data with block linear GOBs to a block linear destination with linear GOBs.
         /// </summary>

Ryujinx.Graphics.Gpu/Shader/DiskCache/DiskCacheHostStorage.cs

@@ -22,7 +22,7 @@ namespace Ryujinx.Graphics.Gpu.Shader.DiskCache
         private const ushort FileFormatVersionMajor = 1;
         private const ushort FileFormatVersionMinor = 2;
         private const uint FileFormatVersionPacked = ((uint)FileFormatVersionMajor << 16) | FileFormatVersionMinor;
-        private const uint CodeGenVersion = 4703;
+        private const uint CodeGenVersion = 4707;
 
         private const string SharedTocFileName = "shared.toc";
         private const string SharedDataFileName = "shared.data";

Ryujinx.Graphics.Gpu/Shader/ShaderCache.cs

@@ -449,7 +449,7 @@ namespace Ryujinx.Graphics.Gpu.Shader
                         if (translatorContexts[i] != null)
                         {
                             translatorContexts[i].SetGeometryShaderLayerInputAttribute(info.GpLayerInputAttribute);
-                            translatorContexts[i].SetLastInVertexPipeline(translatorContexts[5] != null);
+                            translatorContexts[i].SetLastInVertexPipeline();
                             break;
                         }
                     }

Ryujinx.Graphics.Shader/CodeGen/Glsl/Declarations.cs

@@ -569,7 +569,23 @@ namespace Ryujinx.Graphics.Shader.CodeGen.Glsl
 
             if (context.Config.TransformFeedbackEnabled && context.Config.Stage == ShaderStage.Fragment)
             {
-                for (int c = 0; c < 4; c++)
+                int attrOffset = AttributeConsts.UserAttributeBase + attr * 16;
+                int components = context.Info.GetTransformFeedbackOutputComponents(attrOffset);
+
+                if (components > 1)
+                {
+                    string type = components switch
+                    {
+                        2 => "vec2",
+                        3 => "vec3",
+                        4 => "vec4",
+                        _ => "float"
+                    };
+
+                    context.AppendLine($"layout (location = {attr}) in {type} {name};");
+                }
+
+                for (int c = components > 1 ? components : 0; c < 4; c++)
                 {
                     char swzMask = "xyzw"[c];
 
@@ -642,7 +658,7 @@ namespace Ryujinx.Graphics.Shader.CodeGen.Glsl
             if (context.Config.TransformFeedbackEnabled && context.Config.LastInVertexPipeline)
             {
                 int attrOffset = AttributeConsts.UserAttributeBase + attr * 16;
-                int components = context.Config.LastInPipeline ? context.Info.GetTransformFeedbackOutputComponents(attrOffset) : 1;
+                int components = context.Info.GetTransformFeedbackOutputComponents(attrOffset);
 
                 if (components > 1)
                 {
@@ -664,22 +680,20 @@ namespace Ryujinx.Graphics.Shader.CodeGen.Glsl
 
                     context.AppendLine($"layout (location = {attr}{xfb}) out {type} {name};");
                 }
-                else
+
+                for (int c = components > 1 ? components : 0; c < 4; c++)
                 {
-                    for (int c = 0; c < 4; c++)
+                    char swzMask = "xyzw"[c];
+
+                    string xfb = string.Empty;
+
+                    var tfOutput = context.Info.GetTransformFeedbackOutput(attrOffset + c * 4);
+                    if (tfOutput.Valid)
                     {
-                        char swzMask = "xyzw"[c];
-
-                        string xfb = string.Empty;
-
-                        var tfOutput = context.Info.GetTransformFeedbackOutput(attrOffset + c * 4);
-                        if (tfOutput.Valid)
-                        {
-                            xfb = $", xfb_buffer = {tfOutput.Buffer}, xfb_offset = {tfOutput.Offset}, xfb_stride = {tfOutput.Stride}";
-                        }
-
-                        context.AppendLine($"layout (location = {attr}, component = {c}{xfb}) out float {name}_{swzMask};");
+                        xfb = $", xfb_buffer = {tfOutput.Buffer}, xfb_offset = {tfOutput.Offset}, xfb_stride = {tfOutput.Stride}";
                     }
+
+                    context.AppendLine($"layout (location = {attr}, component = {c}{xfb}) out float {name}_{swzMask};");
                 }
             }
             else

Ryujinx.Graphics.Shader/CodeGen/Glsl/OperandManager.cs

@@ -220,7 +220,7 @@ namespace Ryujinx.Graphics.Shader.CodeGen.Glsl
                     ((config.LastInVertexPipeline && isOutAttr) ||
                     (config.Stage == ShaderStage.Fragment && !isOutAttr)))
                 {
-                    int components = config.LastInPipeline ? context.Info.GetTransformFeedbackOutputComponents(attrOffset) : 1;
+                    int components = context.Info.GetTransformFeedbackOutputComponents(attrOffset);
                     string name = components > 1 ? $"{prefix}{(value >> 4)}" : $"{prefix}{(value >> 4)}_{swzMask}";
 
                     if (AttributeInfo.IsArrayAttributeGlsl(config.Stage, isOutAttr))

Ryujinx.Graphics.Shader/CodeGen/Spirv/CodeGenContext.cs

@@ -341,7 +341,7 @@ namespace Ryujinx.Graphics.Shader.CodeGen.Spirv
                 attrOffset = attr;
                 type = elemType;
 
-                if (Config.LastInPipeline && isOutAttr)
+                if (isOutAttr)
                 {
                     int components = Info.GetTransformFeedbackOutputComponents(attr);
 

Ryujinx.Graphics.Shader/CodeGen/Spirv/Declarations.cs

@@ -673,7 +673,7 @@ namespace Ryujinx.Graphics.Shader.CodeGen.Spirv
             int components = 1;
             var type = attrInfo.Type & AggregateType.ElementTypeMask;
 
-            if (context.Config.LastInPipeline && isOutAttr)
+            if (isOutAttr)
             {
                 components = context.Info.GetTransformFeedbackOutputComponents(attr);
 

Ryujinx.Graphics.Shader/StructuredIr/StructuredProgram.cs

@@ -65,7 +65,7 @@ namespace Ryujinx.Graphics.Shader.StructuredIr
                 context.LeaveFunction();
             }
 
-            if (config.TransformFeedbackEnabled && config.LastInVertexPipeline)
+            if (config.TransformFeedbackEnabled && (config.LastInVertexPipeline || config.Stage == ShaderStage.Fragment))
             {
                 for (int tfbIndex = 0; tfbIndex < 4; tfbIndex++)
                 {

Ryujinx.Graphics.Shader/Translation/ShaderConfig.cs

@@ -17,7 +17,6 @@ namespace Ryujinx.Graphics.Shader.Translation
         public ShaderStage Stage { get; }
 
         public bool GpPassthrough { get; }
-        public bool LastInPipeline { get; private set; }
         public bool LastInVertexPipeline { get; private set; }
 
         public bool HasLayerInputAttribute { get; private set; }
@@ -145,7 +144,6 @@ namespace Ryujinx.Graphics.Shader.Translation
             OmapSampleMask           = header.OmapSampleMask;
             OmapDepth                = header.OmapDepth;
             TransformFeedbackEnabled = gpuAccessor.QueryTransformFeedbackEnabled();
-            LastInPipeline           = true;
             LastInVertexPipeline     = header.Stage < ShaderStage.Fragment;
         }
 
@@ -253,13 +251,8 @@ namespace Ryujinx.Graphics.Shader.Translation
             GpLayerInputAttribute = attr;
         }
 
-        public void SetLastInVertexPipeline(bool hasFragment)
+        public void SetLastInVertexPipeline()
         {
-            if (!hasFragment)
-            {
-                LastInPipeline = true;
-            }
-
             LastInVertexPipeline = true;
         }
 
@@ -331,8 +324,6 @@ namespace Ryujinx.Graphics.Shader.Translation
                 config._perPatchAttributeLocations = locationsMap;
             }
 
-            LastInPipeline = false;
-
             // We don't consider geometry shaders using the geometry shader passthrough feature
             // as being the last because when this feature is used, it can't actually modify any of the outputs,
             // so the stage that comes before it is the last one that can do modifications.

Ryujinx.Graphics.Shader/Translation/TranslatorContext.cs

@@ -143,9 +143,9 @@ namespace Ryujinx.Graphics.Shader.Translation
             _config.SetGeometryShaderLayerInputAttribute(attr);
         }
 
-        public void SetLastInVertexPipeline(bool hasFragment)
+        public void SetLastInVertexPipeline()
         {
-            _config.SetLastInVertexPipeline(hasFragment);
+            _config.SetLastInVertexPipeline();
         }
 
         public ShaderProgram Translate(TranslatorContext other = null)

Ryujinx.HLE/HOS/Ipc/IpcMessage.cs

@@ -27,98 +27,103 @@ namespace Ryujinx.HLE.HOS.Ipc
 
         public IpcMessage()
         {
-            PtrBuff      = new List<IpcPtrBuffDesc>();
-            SendBuff     = new List<IpcBuffDesc>();
-            ReceiveBuff  = new List<IpcBuffDesc>();
-            ExchangeBuff = new List<IpcBuffDesc>();
-            RecvListBuff = new List<IpcRecvListBuffDesc>();
+            PtrBuff      = new List<IpcPtrBuffDesc>(0);
+            SendBuff     = new List<IpcBuffDesc>(0);
+            ReceiveBuff  = new List<IpcBuffDesc>(0);
+            ExchangeBuff = new List<IpcBuffDesc>(0);
+            RecvListBuff = new List<IpcRecvListBuffDesc>(0);
 
-            ObjectIds = new List<int>();
+            ObjectIds = new List<int>(0);
         }
 
-        public IpcMessage(ReadOnlySpan<byte> data, long cmdPtr) : this()
+        public IpcMessage(ReadOnlySpan<byte> data, long cmdPtr)
         {
             using (RecyclableMemoryStream ms = MemoryStreamManager.Shared.GetStream(data))
             {
                 BinaryReader reader = new BinaryReader(ms);
 
-                Initialize(reader, cmdPtr);
-            }
-        }
+                int word0 = reader.ReadInt32();
+                int word1 = reader.ReadInt32();
 
-        private void Initialize(BinaryReader reader, long cmdPtr)
-        {
-            int word0 = reader.ReadInt32();
-            int word1 = reader.ReadInt32();
+                Type = (IpcMessageType)(word0 & 0xffff);
 
-            Type = (IpcMessageType)(word0 & 0xffff);
+                int  ptrBuffCount  = (word0 >> 16) & 0xf;
+                int  sendBuffCount = (word0 >> 20) & 0xf;
+                int  recvBuffCount = (word0 >> 24) & 0xf;
+                int  xchgBuffCount = (word0 >> 28) & 0xf;
 
-            int  ptrBuffCount  = (word0 >> 16) & 0xf;
-            int  sendBuffCount = (word0 >> 20) & 0xf;
-            int  recvBuffCount = (word0 >> 24) & 0xf;
-            int  xchgBuffCount = (word0 >> 28) & 0xf;
+                int  rawDataSize   = (word1 >> 0) & 0x3ff;
+                int  recvListFlags = (word1 >> 10) & 0xf;
+                bool hndDescEnable = ((word1 >> 31) & 0x1) != 0;
 
-            int  rawDataSize   =  (word1 >>  0) & 0x3ff;
-            int  recvListFlags =  (word1 >> 10) & 0xf;
-            bool hndDescEnable = ((word1 >> 31) & 0x1) != 0;
-
-            if (hndDescEnable)
-            {
-                HandleDesc = new IpcHandleDesc(reader);
-            }
-
-            for (int index = 0; index < ptrBuffCount; index++)
-            {
-                PtrBuff.Add(new IpcPtrBuffDesc(reader));
-            }
-
-            void ReadBuff(List<IpcBuffDesc> buff, int count)
-            {
-                for (int index = 0; index < count; index++)
+                if (hndDescEnable)
                 {
-                    buff.Add(new IpcBuffDesc(reader));
+                    HandleDesc = new IpcHandleDesc(reader);
                 }
-            }
 
-            ReadBuff(SendBuff,     sendBuffCount);
-            ReadBuff(ReceiveBuff,  recvBuffCount);
-            ReadBuff(ExchangeBuff, xchgBuffCount);
+                PtrBuff = new List<IpcPtrBuffDesc>(ptrBuffCount);
 
-            rawDataSize *= 4;
+                for (int index = 0; index < ptrBuffCount; index++)
+                {
+                    PtrBuff.Add(new IpcPtrBuffDesc(reader));
+                }
 
-            long recvListPos = reader.BaseStream.Position + rawDataSize;
+                static List<IpcBuffDesc> ReadBuff(BinaryReader reader, int count)
+                {
+                    List<IpcBuffDesc> buff = new List<IpcBuffDesc>(count);
+                    
+                    for (int index = 0; index < count; index++)
+                    {
+                        buff.Add(new IpcBuffDesc(reader));
+                    }
+                    
+                    return buff;
+                }
+
+                SendBuff = ReadBuff(reader, sendBuffCount);
+                ReceiveBuff = ReadBuff(reader, recvBuffCount);
+                ExchangeBuff = ReadBuff(reader, xchgBuffCount);
+
+                rawDataSize *= 4;
+
+                long recvListPos = reader.BaseStream.Position + rawDataSize;
 
             // Only CMIF has the padding requirements.
             if (Type < IpcMessageType.TipcCloseSession)
             {
                 long pad0 = GetPadSize16(reader.BaseStream.Position + cmdPtr);
 
-                if (rawDataSize != 0)
-                {
-                    rawDataSize -= (int)pad0;
+                    if (rawDataSize != 0)
+                    {
+                        rawDataSize -= (int)pad0;
+                    }
+
+                    reader.BaseStream.Seek(pad0, SeekOrigin.Current);
                 }
 
-                reader.BaseStream.Seek(pad0, SeekOrigin.Current);
-            }
+                int recvListCount = recvListFlags - 2;
 
-            int recvListCount = recvListFlags - 2;
+                if (recvListCount == 0)
+                {
+                    recvListCount = 1;
+                }
+                else if (recvListCount < 0)
+                {
+                    recvListCount = 0;
+                }
 
-            if (recvListCount == 0)
-            {
-                recvListCount = 1;
-            }
-            else if (recvListCount < 0)
-            {
-                recvListCount = 0;
-            }
+                RawData = reader.ReadBytes(rawDataSize);
 
-            RawData = reader.ReadBytes(rawDataSize);
+                reader.BaseStream.Seek(recvListPos, SeekOrigin.Begin);
 
-            reader.BaseStream.Seek(recvListPos, SeekOrigin.Begin);
+                RecvListBuff = new List<IpcRecvListBuffDesc>(recvListCount);
 
-            for (int index = 0; index < recvListCount; index++)
-            {
-                RecvListBuff.Add(new IpcRecvListBuffDesc(reader.ReadUInt64()));
+                for (int index = 0; index < recvListCount; index++)
+                {
+                    RecvListBuff.Add(new IpcRecvListBuffDesc(reader.ReadUInt64()));
+                }
+
+                ObjectIds = new List<int>(0);
             }
         }
 

Ryujinx.HLE/HOS/Kernel/Common/KTimeManager.cs

@@ -71,7 +71,13 @@ namespace Ryujinx.HLE.HOS.Kernel.Common
         {
             lock (_context.CriticalSection.Lock)
             {
-                _waitingObjects.RemoveAll(x => x.Object == schedulerObj);
+                for (int index = _waitingObjects.Count - 1; index >= 0; index--)
+                {
+                    if (_waitingObjects[index].Object == schedulerObj)
+                    {
+                        _waitingObjects.RemoveAt(index);
+                    }
+                }
             }
         }
 
@@ -105,16 +111,19 @@ namespace Ryujinx.HLE.HOS.Kernel.Common
                             }
                             else
                             {
-                                while (Interlocked.Read(ref _enforceWakeupFromSpinWait) != 1 && PerformanceCounter.ElapsedTicks <= next.TimePoint)
+                                while (Interlocked.Read(ref _enforceWakeupFromSpinWait) != 1 && PerformanceCounter.ElapsedTicks < next.TimePoint)
                                 {
+                                    // Our time is close - don't let SpinWait go off and potentially Thread.Sleep().
                                     if (spinWait.NextSpinWillYield)
                                     {
                                         Thread.Yield();
 
                                         spinWait.Reset();
                                     }
-
-                                    spinWait.SpinOnce();
+                                    else
+                                    {
+                                        spinWait.SpinOnce();
+                                    }
                                 }
 
                                 spinWait.Reset();

Ryujinx.HLE/HOS/Kernel/SupervisorCall/Syscall.cs

@@ -9,6 +9,7 @@ using Ryujinx.HLE.HOS.Kernel.Process;
 using Ryujinx.HLE.HOS.Kernel.Threading;
 using Ryujinx.Horizon.Common;
 using System;
+using System.Buffers;
 using System.Threading;
 
 namespace Ryujinx.HLE.HOS.Kernel.SupervisorCall
@@ -553,7 +554,9 @@ namespace Ryujinx.HLE.HOS.Kernel.SupervisorCall
 
             KProcess currentProcess = KernelStatic.GetCurrentProcess();
 
-            KSynchronizationObject[] syncObjs = handles.Length == 0 ? Array.Empty<KSynchronizationObject>() : new KSynchronizationObject[handles.Length];
+            KSynchronizationObject[] syncObjsArray = ArrayPool<KSynchronizationObject>.Shared.Rent(handles.Length);
+
+            Span<KSynchronizationObject> syncObjs = syncObjsArray.AsSpan(0, handles.Length);
 
             for (int index = 0; index < handles.Length; index++)
             {
@@ -606,6 +609,8 @@ namespace Ryujinx.HLE.HOS.Kernel.SupervisorCall
                 }
             }
 
+            ArrayPool<KSynchronizationObject>.Shared.Return(syncObjsArray);
+
             return result;
         }
 

Ryujinx.HLE/HOS/Kernel/Threading/KSynchronization.cs

@@ -1,6 +1,7 @@
 using Ryujinx.HLE.HOS.Kernel.Common;
 using Ryujinx.Horizon.Common;
 using System;
+using System.Buffers;
 using System.Collections.Generic;
 
 namespace Ryujinx.HLE.HOS.Kernel.Threading
@@ -59,7 +60,9 @@ namespace Ryujinx.HLE.HOS.Kernel.Threading
             }
             else
             {
-                LinkedListNode<KThread>[] syncNodes = syncObjs.Length == 0 ? Array.Empty<LinkedListNode<KThread>>() : new LinkedListNode<KThread>[syncObjs.Length];
+                LinkedListNode<KThread>[] syncNodesArray = ArrayPool<LinkedListNode<KThread>>.Shared.Rent(syncObjs.Length);
+                
+                Span<LinkedListNode<KThread>> syncNodes = syncNodesArray.AsSpan(0, syncObjs.Length);
 
                 for (int index = 0; index < syncObjs.Length; index++)
                 {
@@ -101,6 +104,8 @@ namespace Ryujinx.HLE.HOS.Kernel.Threading
                         handleIndex = index;
                     }
                 }
+
+                ArrayPool<LinkedListNode<KThread>>.Shared.Return(syncNodesArray);
             }
 
             _context.CriticalSection.Leave();

Ryujinx.HLE/HOS/Services/Audio/AudioRenderer/AudioRendererServer.cs

@@ -1,4 +1,5 @@
 using Ryujinx.Common.Logging;
+using Ryujinx.Common.Memory;
 using Ryujinx.HLE.HOS.Ipc;
 using Ryujinx.HLE.HOS.Kernel.Threading;
 using Ryujinx.Horizon.Common;
@@ -68,25 +69,29 @@ namespace Ryujinx.HLE.HOS.Services.Audio.AudioRenderer
 
             ReadOnlyMemory<byte> input = context.Memory.GetSpan(inputPosition, (int)inputSize).ToArray();
 
-            Memory<byte> output = new byte[outputSize];
-            Memory<byte> performanceOutput = new byte[performanceOutputSize];
-
-            using MemoryHandle outputHandle = output.Pin();
-            using MemoryHandle performanceOutputHandle = performanceOutput.Pin();
-
-            ResultCode result = _impl.RequestUpdate(output, performanceOutput, input);
-
-            if (result == ResultCode.Success)
+            using (IMemoryOwner<byte> outputOwner = ByteMemoryPool.Shared.RentCleared(outputSize))
+            using (IMemoryOwner<byte> performanceOutputOwner = ByteMemoryPool.Shared.RentCleared(performanceOutputSize))
             {
-                context.Memory.Write(outputPosition, output.Span);
-                context.Memory.Write(performanceOutputPosition, performanceOutput.Span);
-            }
-            else
-            {
-                Logger.Error?.Print(LogClass.ServiceAudio, $"Error while processing renderer update: 0x{(int)result:X}");
-            }
+                Memory<byte> output = outputOwner.Memory;
+                Memory<byte> performanceOutput = performanceOutputOwner.Memory;
 
-            return result;
+                using MemoryHandle outputHandle = output.Pin();
+                using MemoryHandle performanceOutputHandle = performanceOutput.Pin();
+
+                ResultCode result = _impl.RequestUpdate(output, performanceOutput, input);
+
+                if (result == ResultCode.Success)
+                {
+                    context.Memory.Write(outputPosition, output.Span);
+                    context.Memory.Write(performanceOutputPosition, performanceOutput.Span);
+                }
+                else
+                {
+                    Logger.Error?.Print(LogClass.ServiceAudio, $"Error while processing renderer update: 0x{(int)result:X}");
+                }
+
+                return result;
+            }
         }
 
         [CommandCmif(5)]

Ryujinx.HLE/HOS/Services/IpcService.cs

@@ -76,6 +76,8 @@ namespace Ryujinx.HLE.HOS.Services
 
                 context.RequestData.BaseStream.Seek(0x10 + dataPayloadSize, SeekOrigin.Begin);
 
+                context.Request.ObjectIds.EnsureCapacity(inputObjCount);
+
                 for (int index = 0; index < inputObjCount; index++)
                 {
                     context.Request.ObjectIds.Add(context.RequestData.ReadInt32());

Ryujinx.HLE/HOS/Services/ServerBase.cs

@@ -217,6 +217,8 @@ namespace Ryujinx.HLE.HOS.Services
 
             if (noReceive)
             {
+                response.PtrBuff.EnsureCapacity(request.RecvListBuff.Count);
+
                 for (int i = 0; i < request.RecvListBuff.Count; i++)
                 {
                     ulong size = (ulong)BinaryPrimitives.ReadInt16LittleEndian(request.RawData.AsSpan(sizesOffset + i * 2, 2));

Ryujinx.HLE/HOS/Services/SurfaceFlinger/IHOSBinderDriver.cs

@@ -1,7 +1,9 @@
-using Ryujinx.HLE.HOS.Ipc;
+using Ryujinx.Common.Memory;
+using Ryujinx.HLE.HOS.Ipc;
 using Ryujinx.HLE.HOS.Kernel.Threading;
 using Ryujinx.Horizon.Common;
 using System;
+using System.Buffers;
 
 namespace Ryujinx.HLE.HOS.Services.SurfaceFlinger
 {
@@ -83,16 +85,19 @@ namespace Ryujinx.HLE.HOS.Services.SurfaceFlinger
 
             ReadOnlySpan<byte> inputParcel = context.Memory.GetSpan(dataPos, (int)dataSize);
 
-            Span<byte> outputParcel = new Span<byte>(new byte[replySize]);
-
-            ResultCode result = OnTransact(binderId, code, flags, inputParcel, outputParcel);
-
-            if (result == ResultCode.Success)
+            using (IMemoryOwner<byte> outputParcelOwner = ByteMemoryPool.Shared.RentCleared(replySize))
             {
-                context.Memory.Write(replyPos, outputParcel);
-            }
+                Span<byte> outputParcel = outputParcelOwner.Memory.Span;
                 
-            return result;
+                ResultCode result = OnTransact(binderId, code, flags, inputParcel, outputParcel);
+
+                if (result == ResultCode.Success)
+                {
+                    context.Memory.Write(replyPos, outputParcel);
+                }
+
+                return result;
+            }
         }
 
         protected abstract ResultCode AdjustRefcount(int binderId, int addVal, int type);

Ryujinx.Input.SDL2/SDL2Gamepad.cs

@@ -12,17 +12,7 @@ namespace Ryujinx.Input.SDL2
     {
         private bool HasConfiguration => _configuration != null;
 
-        private class ButtonMappingEntry
-        {
-            public readonly GamepadButtonInputId To;
-            public readonly GamepadButtonInputId From;
-
-            public ButtonMappingEntry(GamepadButtonInputId to, GamepadButtonInputId from)
-            {
-                To = to;
-                From = from;
-            }
-        }
+        private record struct ButtonMappingEntry(GamepadButtonInputId To, GamepadButtonInputId From);
 
         private StandardControllerInputConfig _configuration;
 
@@ -85,7 +75,7 @@ namespace Ryujinx.Input.SDL2
         public SDL2Gamepad(IntPtr gamepadHandle, string driverId)
         {
             _gamepadHandle = gamepadHandle;
-            _buttonsUserMapping = new List<ButtonMappingEntry>();
+            _buttonsUserMapping = new List<ButtonMappingEntry>(20);
 
             Name = SDL_GameControllerName(_gamepadHandle);
             Id = driverId;

Ryujinx/Program.cs

@@ -96,6 +96,8 @@ namespace Ryujinx
             // Delete backup files after updating.
             Task.Run(Updater.CleanupUpdate);
 
+            Console.Title = $"Ryujinx Console {Version}";
+
             // NOTE: GTK3 doesn't init X11 in a multi threaded way.
             // This ends up causing race condition and abort of XCB when a context is created by SPB (even if SPB do call XInitThreads).
             if (OperatingSystem.IsLinux())

Ryujinx/Ui/Windows/AmiiboWindow.cs

@@ -1,6 +1,7 @@
 using Gtk;
 using Ryujinx.Common;
 using Ryujinx.Common.Configuration;
+using Ryujinx.Common.Logging;
 using Ryujinx.Common.Utilities;
 using Ryujinx.Ui.Common.Configuration;
 using Ryujinx.Ui.Common.Models.Amiibo;
@@ -12,7 +13,6 @@ using System.Linq;
 using System.Net.Http;
 using System.Reflection;
 using System.Text;
-using System.Text.Json;
 using System.Threading.Tasks;
 using AmiiboApi = Ryujinx.Ui.Common.Models.Amiibo.AmiiboApi;
 using AmiiboJsonSerializerContext = Ryujinx.Ui.Common.Models.Amiibo.AmiiboJsonSerializerContext;
@@ -57,7 +57,7 @@ namespace Ryujinx.Ui.Windows
 
             _httpClient = new HttpClient()
             {
-                Timeout = TimeSpan.FromMilliseconds(5000)
+                Timeout = TimeSpan.FromSeconds(30)
             };
 
             Directory.CreateDirectory(System.IO.Path.Join(AppDataManager.BaseDirPath, "system", "amiibo"));
@@ -93,8 +93,10 @@ namespace Ryujinx.Ui.Windows
                 {
                     amiiboJsonString = await DownloadAmiiboJson();
                 }
-                catch
+                catch (Exception ex)
                 {
+                    Logger.Error?.Print(LogClass.Application, $"Failed to download amiibo data: {ex}");
+
                     ShowInfoDialog();
 
                     Close();
@@ -183,8 +185,10 @@ namespace Ryujinx.Ui.Windows
 
                 return false;
             }
-            catch
+            catch (Exception ex)
             {
+                Logger.Error?.Print(LogClass.Application, $"Failed to check for amiibo updates: {ex}");
+
                 ShowInfoDialog();
 
                 return false;
@@ -208,6 +212,8 @@ namespace Ryujinx.Ui.Windows
             }
             else
             {
+                Logger.Error?.Print(LogClass.Application, $"Failed to download amiibo data. Response status code: {response.StatusCode}");
+
                 GtkDialog.CreateInfoDialog($"Amiibo API", "An error occured while fetching information from the API.");
 
                 Close();
@@ -233,6 +239,10 @@ namespace Ryujinx.Ui.Windows
 
                 _amiiboImage.Pixbuf = amiiboPreview.ScaleSimple(resizeWidth, resizeHeight, Gdk.InterpType.Bilinear);
             }
+            else
+            {
+                Logger.Error?.Print(LogClass.Application, $"Failed to get amiibo preview. Response status code: {response.StatusCode}");
+            }
         }
 
         private void ShowInfoDialog()