Vulkan: Migrate buffers between memory types to improve GPU performance (#4540)

* Initial implementation of migration between memory heaps

- Missing OOM handling
- Missing `_map` data safety when remapping
  - Copy may not have completed yet (needs some kind of fence)
  - Map may be unmapped before it is done being used. (needs scoped access)
- SSBO accesses are all "writes" - maybe pass info in another way.
- Missing keeping map type when resizing buffers (should this be done?)

* Ensure migrated data is in place before flushing.

* Fix issue where old waitable would be signalled.

- There is a real issue where existing Auto<> references need to be replaced.

* Swap bound Auto<> instances when swapping buffer backing

* Fix conversion buffers

* Don't try move buffers if the host has shared memory.

* Make GPU methods return PinnedSpan with scope

* Storage Hint

* Fix stupidity

* Fix rebase

* Tweak rules

Attempt to sidestep BOTW slowdown

* Remove line

* Migrate only when command buffers flush

* Change backing swap log to debug

* Address some feedback

* Disallow backing swap when the flush lock is held by the current thread

* Make PinnedSpan from ReadOnlySpan explicitly unsafe

* Fix some small issues

- Index buffer swap fixed
- Allocate DeviceLocal buffers using a separate block list to images.

* Remove alternative flags

* Address feedback

7 files changed, 25 insertions, 39 deletions

diff --git a/Ryujinx.Graphics.GAL/Multithreading/Commands/Buffer/BufferGetDataCommand.cs b/Ryujinx.Graphics.GAL/Multithreading/Commands/Buffer/BufferGetDataCommand.cs
index d3a255e7..031c6153 100644
--- a/Ryujinx.Graphics.GAL/Multithreading/Commands/Buffer/BufferGetDataCommand.cs
+++ b/Ryujinx.Graphics.GAL/Multithreading/Commands/Buffer/BufferGetDataCommand.cs
@@ -21,9 +21,9 @@ namespace Ryujinx.Graphics.GAL.Multithreading.Commands.Buffer
 
         public static void Run(ref BufferGetDataCommand command, ThreadedRenderer threaded, IRenderer renderer)
         {
-            ReadOnlySpan<byte> result = renderer.GetBufferData(threaded.Buffers.MapBuffer(command._buffer), command._offset, command._size);
+            PinnedSpan<byte> result = renderer.GetBufferData(threaded.Buffers.MapBuffer(command._buffer), command._offset, command._size);
 
-            command._result.Get(threaded).Result = new PinnedSpan<byte>(result);
+            command._result.Get(threaded).Result = result;
         }
     }
 }
diff --git a/Ryujinx.Graphics.GAL/Multithreading/Commands/Renderer/CreateBufferCommand.cs b/Ryujinx.Graphics.GAL/Multithreading/Commands/Renderer/CreateBufferCommand.cs
index 4f01dea2..b36d8bbe 100644
--- a/Ryujinx.Graphics.GAL/Multithreading/Commands/Renderer/CreateBufferCommand.cs
+++ b/Ryujinx.Graphics.GAL/Multithreading/Commands/Renderer/CreateBufferCommand.cs
@@ -5,16 +5,25 @@
         public CommandType CommandType => CommandType.CreateBuffer;
         private BufferHandle _threadedHandle;
         private int _size;
+        private BufferHandle _storageHint;
 
-        public void Set(BufferHandle threadedHandle, int size)
+        public void Set(BufferHandle threadedHandle, int size, BufferHandle storageHint)
         {
             _threadedHandle = threadedHandle;
             _size = size;
+            _storageHint = storageHint;
         }
 
         public static void Run(ref CreateBufferCommand command, ThreadedRenderer threaded, IRenderer renderer)
         {
-            threaded.Buffers.AssignBuffer(command._threadedHandle, renderer.CreateBuffer(command._size));
+            BufferHandle hint = BufferHandle.Null;
+
+            if (command._storageHint != BufferHandle.Null)
+            {
+                hint = threaded.Buffers.MapBuffer(command._storageHint);
+            }
+
+            threaded.Buffers.AssignBuffer(command._threadedHandle, renderer.CreateBuffer(command._size, hint));
         }
     }
 }
diff --git a/Ryujinx.Graphics.GAL/Multithreading/Commands/Texture/TextureGetDataCommand.cs b/Ryujinx.Graphics.GAL/Multithreading/Commands/Texture/TextureGetDataCommand.cs
index 1f519ccd..91320d45 100644
--- a/Ryujinx.Graphics.GAL/Multithreading/Commands/Texture/TextureGetDataCommand.cs
+++ b/Ryujinx.Graphics.GAL/Multithreading/Commands/Texture/TextureGetDataCommand.cs
@@ -18,9 +18,9 @@ namespace Ryujinx.Graphics.GAL.Multithreading.Commands.Texture
 
         public static void Run(ref TextureGetDataCommand command, ThreadedRenderer threaded, IRenderer renderer)
         {
-            ReadOnlySpan<byte> result = command._texture.Get(threaded).Base.GetData();
+            PinnedSpan<byte> result = command._texture.Get(threaded).Base.GetData();
 
-            command._result.Get(threaded).Result = new PinnedSpan<byte>(result);
+            command._result.Get(threaded).Result = result;
         }
     }
 }
diff --git a/Ryujinx.Graphics.GAL/Multithreading/Commands/Texture/TextureGetDataSliceCommand.cs b/Ryujinx.Graphics.GAL/Multithreading/Commands/Texture/TextureGetDataSliceCommand.cs
index 5ac05971..ec06cc4d 100644
--- a/Ryujinx.Graphics.GAL/Multithreading/Commands/Texture/TextureGetDataSliceCommand.cs
+++ b/Ryujinx.Graphics.GAL/Multithreading/Commands/Texture/TextureGetDataSliceCommand.cs
@@ -22,9 +22,9 @@ namespace Ryujinx.Graphics.GAL.Multithreading.Commands.Texture
 
         public static void Run(ref TextureGetDataSliceCommand command, ThreadedRenderer threaded, IRenderer renderer)
         {
-            ReadOnlySpan<byte> result = command._texture.Get(threaded).Base.GetData(command._layer, command._level);
+            PinnedSpan<byte> result = command._texture.Get(threaded).Base.GetData(command._layer, command._level);
 
-            command._result.Get(threaded).Result = new PinnedSpan<byte>(result);
+            command._result.Get(threaded).Result = result;
         }
     }
 }
diff --git a/Ryujinx.Graphics.GAL/Multithreading/Model/PinnedSpan.cs b/Ryujinx.Graphics.GAL/Multithreading/Model/PinnedSpan.cs
deleted file mode 100644
index 16e148c2..00000000
--- a/Ryujinx.Graphics.GAL/Multithreading/Model/PinnedSpan.cs
+++ /dev/null
@@ -1,23 +0,0 @@
-using System;
-using System.Runtime.CompilerServices;
-using System.Runtime.InteropServices;
-
-namespace Ryujinx.Graphics.GAL.Multithreading.Model
-{
-    unsafe struct PinnedSpan<T> where T : unmanaged
-    {
-        private void* _ptr;
-        private int _size;
-
-        public PinnedSpan(ReadOnlySpan<T> span)
-        {
-            _ptr = Unsafe.AsPointer(ref MemoryMarshal.GetReference(span));
-            _size = span.Length;
-        }
-
-        public ReadOnlySpan<T> Get()
-        {
-            return new ReadOnlySpan<T>(_ptr, _size * Unsafe.SizeOf<T>());
-        }
-    }
-}
diff --git a/Ryujinx.Graphics.GAL/Multithreading/Resources/ThreadedTexture.cs b/Ryujinx.Graphics.GAL/Multithreading/Resources/ThreadedTexture.cs
index 1267ab79..ee1cfa29 100644
--- a/Ryujinx.Graphics.GAL/Multithreading/Resources/ThreadedTexture.cs
+++ b/Ryujinx.Graphics.GAL/Multithreading/Resources/ThreadedTexture.cs
@@ -72,7 +72,7 @@ namespace Ryujinx.Graphics.GAL.Multithreading.Resources
             return newTex;
         }
 
-        public ReadOnlySpan<byte> GetData()
+        public PinnedSpan<byte> GetData()
         {
             if (_renderer.IsGpuThread())
             {
@@ -80,7 +80,7 @@ namespace Ryujinx.Graphics.GAL.Multithreading.Resources
                 _renderer.New<TextureGetDataCommand>().Set(Ref(this), Ref(box));
                 _renderer.InvokeCommand();
 
-                return box.Result.Get();
+                return box.Result;
             }
             else
             {
@@ -90,7 +90,7 @@ namespace Ryujinx.Graphics.GAL.Multithreading.Resources
             }
         }
 
-        public ReadOnlySpan<byte> GetData(int layer, int level)
+        public PinnedSpan<byte> GetData(int layer, int level)
         {
             if (_renderer.IsGpuThread())
             {
@@ -98,7 +98,7 @@ namespace Ryujinx.Graphics.GAL.Multithreading.Resources
                 _renderer.New<TextureGetDataSliceCommand>().Set(Ref(this), Ref(box), layer, level);
                 _renderer.InvokeCommand();
 
-                return box.Result.Get();
+                return box.Result;
             }
             else
             {
diff --git a/Ryujinx.Graphics.GAL/Multithreading/ThreadedRenderer.cs b/Ryujinx.Graphics.GAL/Multithreading/ThreadedRenderer.cs
index 74326f1d..2148f43f 100644
--- a/Ryujinx.Graphics.GAL/Multithreading/ThreadedRenderer.cs
+++ b/Ryujinx.Graphics.GAL/Multithreading/ThreadedRenderer.cs
@@ -265,10 +265,10 @@ namespace Ryujinx.Graphics.GAL.Multithreading
             }
         }
 
-        public BufferHandle CreateBuffer(int size)
+        public BufferHandle CreateBuffer(int size, BufferHandle storageHint)
         {
             BufferHandle handle = Buffers.CreateBufferHandle();
-            New<CreateBufferCommand>().Set(handle, size);
+            New<CreateBufferCommand>().Set(handle, size, storageHint);
             QueueCommand();
 
             return handle;
@@ -329,7 +329,7 @@ namespace Ryujinx.Graphics.GAL.Multithreading
             QueueCommand();
         }
 
-        public ReadOnlySpan<byte> GetBufferData(BufferHandle buffer, int offset, int size)
+        public PinnedSpan<byte> GetBufferData(BufferHandle buffer, int offset, int size)
         {
             if (IsGpuThread())
             {
@@ -337,7 +337,7 @@ namespace Ryujinx.Graphics.GAL.Multithreading
                 New<BufferGetDataCommand>().Set(buffer, offset, size, Ref(box));
                 InvokeCommand();
 
-                return box.Result.Get();
+                return box.Result;
             }
             else
             {