drashna/ryujinx
1
0
Fork 0
Code Activity Actions

Initial support for GPU channels (#2372)

Browse source 
* Ground work for separate GPU channels

* Rename TextureManager to TextureCache

* Decouple texture bindings management from the texture cache

* Rename BufferManager to BufferCache

* Decouple buffer bindings management from the buffer cache

* More comments and proper disposal

* PR feedback

* Force host state update on channel switch

* Typo

* PR feedback

* Missing using
This commit is contained in:
gdkchan 2021-06-23 20:51:41 -03:00 • committed by GitHub
parent 12a7a2ead8
commit a10b2c5ff2
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
24 changed files with 1745 additions and 1456 deletions
Download patch file Download diff file Expand all files Collapse all files

438
Ryujinx.Graphics.Gpu/Memory/BufferManager.cs
View file

@ -1,9 +1,7 @@
using Ryujinx.Common;
using Ryujinx.Common;
using Ryujinx.Graphics.GAL;
using Ryujinx.Graphics.Gpu.Image;
using Ryujinx.Graphics.Gpu.State;
using Ryujinx.Graphics.Shader;
using Ryujinx.Memory.Range;
using System;
using System.Collections.Generic;
using System.Collections.ObjectModel;
@ -14,26 +12,16 @@ namespace Ryujinx.Graphics.Gpu.Memory
/// <summary>
/// Buffer manager.
/// </summary>
class BufferManager
class BufferManager : IDisposable
{
private const int StackToHeapThreshold = 16;
private const int OverlapsBufferInitialCapacity = 10;
private const int OverlapsBufferMaxCapacity = 10000;
private const ulong BufferAlignmentSize = 0x1000;
private const ulong BufferAlignmentMask = BufferAlignmentSize - 1;
private GpuContext _context;
private RangeList<Buffer> _buffers;
private Buffer[] _bufferOverlaps;
private readonly GpuContext _context;
private IndexBuffer _indexBuffer;
private VertexBuffer[] _vertexBuffers;
private BufferBounds[] _transformFeedbackBuffers;
private List<BufferTextureBinding> _bufferTextures;
private readonly VertexBuffer[] _vertexBuffers;
private readonly BufferBounds[] _transformFeedbackBuffers;
private readonly List<BufferTextureBinding> _bufferTextures;
/// <summary>
/// Holds shader stage buffer state and binding information.
@ -94,10 +82,10 @@ namespace Ryujinx.Graphics.Gpu.Memory
}
}
private BuffersPerStage _cpStorageBuffers;
private BuffersPerStage _cpUniformBuffers;
private BuffersPerStage[] _gpStorageBuffers;
private BuffersPerStage[] _gpUniformBuffers;
private readonly BuffersPerStage _cpStorageBuffers;
private readonly BuffersPerStage _cpUniformBuffers;
private readonly BuffersPerStage[] _gpStorageBuffers;
private readonly BuffersPerStage[] _gpUniformBuffers;
private int _cpStorageBufferBindings;
private int _cpUniformBufferBindings;
@ -114,20 +102,14 @@ namespace Ryujinx.Graphics.Gpu.Memory
private bool _rebind;
private Dictionary<ulong, BufferCacheEntry> _dirtyCache;
/// <summary>
/// Creates a new instance of the buffer manager.
/// </summary>
/// <param name="context">The GPU context that the buffer manager belongs to</param>
/// <param name="context">GPU context that the buffer manager belongs to</param>
public BufferManager(GpuContext context)
{
_context = context;
_buffers = new RangeList<Buffer>();
_bufferOverlaps = new Buffer[OverlapsBufferInitialCapacity];
_vertexBuffers = new VertexBuffer[Constants.TotalVertexBuffers];
_transformFeedbackBuffers = new BufferBounds[Constants.TotalTransformFeedbackBuffers];
@ -146,9 +128,10 @@ namespace Ryujinx.Graphics.Gpu.Memory
_bufferTextures = new List<BufferTextureBinding>();
_dirtyCache = new Dictionary<ulong, BufferCacheEntry>();
context.Methods.BufferCache.NotifyBuffersModified += Rebind;
}
/// <summary>
/// Sets the memory range with the index buffer data, to be used for subsequent draw calls.
/// </summary>
@ -157,11 +140,11 @@ namespace Ryujinx.Graphics.Gpu.Memory
/// <param name="type">Type of each index buffer element</param>
public void SetIndexBuffer(ulong gpuVa, ulong size, IndexType type)
{
ulong address = TranslateAndCreateBuffer(gpuVa, size);
ulong address = _context.Methods.BufferCache.TranslateAndCreateBuffer(gpuVa, size);
_indexBuffer.Address = address;
_indexBuffer.Size = size;
_indexBuffer.Type = type;
_indexBuffer.Size = size;
_indexBuffer.Type = type;
_indexBufferDirty = true;
}
@ -188,11 +171,11 @@ namespace Ryujinx.Graphics.Gpu.Memory
/// <param name="divisor">Vertex divisor of the buffer, for instanced draws</param>
public void SetVertexBuffer(int index, ulong gpuVa, ulong size, int stride, int divisor)
{
ulong address = TranslateAndCreateBuffer(gpuVa, size);
ulong address = _context.Methods.BufferCache.TranslateAndCreateBuffer(gpuVa, size);
_vertexBuffers[index].Address = address;
_vertexBuffers[index].Size = size;
_vertexBuffers[index].Stride = stride;
_vertexBuffers[index].Size = size;
_vertexBuffers[index].Stride = stride;
_vertexBuffers[index].Divisor = divisor;
_vertexBuffersDirty = true;
@ -216,7 +199,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
/// <param name="size">Size in bytes of the transform feedback buffer</param>
public void SetTransformFeedbackBuffer(int index, ulong gpuVa, ulong size)
{
ulong address = TranslateAndCreateBuffer(gpuVa, size);
ulong address = _context.Methods.BufferCache.TranslateAndCreateBuffer(gpuVa, size);
_transformFeedbackBuffers[index] = new BufferBounds(address, size);
_transformFeedbackBuffersDirty = true;
@ -236,7 +219,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
gpuVa = BitUtils.AlignDown(gpuVa, _context.Capabilities.StorageBufferOffsetAlignment);
ulong address = TranslateAndCreateBuffer(gpuVa, size);
ulong address = _context.Methods.BufferCache.TranslateAndCreateBuffer(gpuVa, size);
_cpStorageBuffers.SetBounds(index, address, size, flags);
}
@ -256,10 +239,10 @@ namespace Ryujinx.Graphics.Gpu.Memory
gpuVa = BitUtils.AlignDown(gpuVa, _context.Capabilities.StorageBufferOffsetAlignment);
ulong address = TranslateAndCreateBuffer(gpuVa, size);
ulong address = _context.Methods.BufferCache.TranslateAndCreateBuffer(gpuVa, size);
if (_gpStorageBuffers[stage].Buffers[index].Address != address ||
_gpStorageBuffers[stage].Buffers[index].Size != size)
_gpStorageBuffers[stage].Buffers[index].Size != size)
{
_gpStorageBuffersDirty = true;
}
@ -276,7 +259,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
/// <param name="size">Size in bytes of the storage buffer</param>
public void SetComputeUniformBuffer(int index, ulong gpuVa, ulong size)
{
ulong address = TranslateAndCreateBuffer(gpuVa, size);
ulong address = _context.Methods.BufferCache.TranslateAndCreateBuffer(gpuVa, size);
_cpUniformBuffers.SetBounds(index, address, size);
}
@ -291,7 +274,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
/// <param name="size">Size in bytes of the storage buffer</param>
public void SetGraphicsUniformBuffer(int stage, int index, ulong gpuVa, ulong size)
{
ulong address = TranslateAndCreateBuffer(gpuVa, size);
ulong address = _context.Methods.BufferCache.TranslateAndCreateBuffer(gpuVa, size);
_gpUniformBuffers[stage].SetBounds(index, address, size);
_gpUniformBuffersDirty = true;
@ -397,191 +380,6 @@ namespace Ryujinx.Graphics.Gpu.Memory
return mask;
}
/// <summary>
/// Handles removal of buffers written to a memory region being unmapped.
/// </summary>
/// <param name="sender">Sender object</param>
/// <param name="e">Event arguments</param>
public void MemoryUnmappedHandler(object sender, UnmapEventArgs e)
{
Buffer[] overlaps = new Buffer[10];
int overlapCount;
ulong address = _context.MemoryManager.Translate(e.Address);
ulong size = e.Size;
lock (_buffers)
{
overlapCount = _buffers.FindOverlaps(address, size, ref overlaps);
}
for (int i = 0; i < overlapCount; i++)
{
overlaps[i].Unmapped(address, size);
}
}
/// <summary>
/// Performs address translation of the GPU virtual address, and creates a
/// new buffer, if needed, for the specified range.
/// </summary>
/// <param name="gpuVa">Start GPU virtual address of the buffer</param>
/// <param name="size">Size in bytes of the buffer</param>
/// <returns>CPU virtual address of the buffer, after address translation</returns>
private ulong TranslateAndCreateBuffer(ulong gpuVa, ulong size)
{
if (gpuVa == 0)
{
return 0;
}
ulong address = _context.MemoryManager.Translate(gpuVa);
if (address == MemoryManager.PteUnmapped)
{
return 0;
}
CreateBuffer(address, size);
return address;
}
/// <summary>
/// Creates a new buffer for the specified range, if it does not yet exist.
/// This can be used to ensure the existance of a buffer.
/// </summary>
/// <param name="address">Address of the buffer in memory</param>
/// <param name="size">Size of the buffer in bytes</param>
public void CreateBuffer(ulong address, ulong size)
{
ulong endAddress = address + size;
ulong alignedAddress = address & ~BufferAlignmentMask;
ulong alignedEndAddress = (endAddress + BufferAlignmentMask) & ~BufferAlignmentMask;
// The buffer must have the size of at least one page.
if (alignedEndAddress == alignedAddress)
{
alignedEndAddress += BufferAlignmentSize;
}
CreateBufferAligned(alignedAddress, alignedEndAddress - alignedAddress);
}
/// <summary>
/// Performs address translation of the GPU virtual address, and attempts to force
/// the buffer in the region as dirty.
/// The buffer lookup for this function is cached in a dictionary for quick access, which
/// accelerates common UBO updates.
/// </summary>
/// <param name="gpuVa">Start GPU virtual address of the buffer</param>
/// <param name="size">Size in bytes of the buffer</param>
public void ForceDirty(ulong gpuVa, ulong size)
{
BufferCacheEntry result;
if (!_dirtyCache.TryGetValue(gpuVa, out result) || result.EndGpuAddress < gpuVa + size || result.UnmappedSequence != result.Buffer.UnmappedSequence)
{
ulong address = TranslateAndCreateBuffer(gpuVa, size);
result = new BufferCacheEntry(address, gpuVa, GetBuffer(address, size));
_dirtyCache[gpuVa] = result;
}
result.Buffer.ForceDirty(result.Address, size);
}
/// <summary>
/// Creates a new buffer for the specified range, if needed.
/// If a buffer where this range can be fully contained already exists,
/// then the creation of a new buffer is not necessary.
/// </summary>
/// <param name="address">Address of the buffer in guest memory</param>
/// <param name="size">Size in bytes of the buffer</param>
private void CreateBufferAligned(ulong address, ulong size)
{
int overlapsCount;
lock (_buffers)
{
overlapsCount = _buffers.FindOverlapsNonOverlapping(address, size, ref _bufferOverlaps);
}
if (overlapsCount != 0)
{
// The buffer already exists. We can just return the existing buffer
// if the buffer we need is fully contained inside the overlapping buffer.
// Otherwise, we must delete the overlapping buffers and create a bigger buffer
// that fits all the data we need. We also need to copy the contents from the
// old buffer(s) to the new buffer.
ulong endAddress = address + size;
if (_bufferOverlaps[0].Address > address || _bufferOverlaps[0].EndAddress < endAddress)
{
for (int index = 0; index < overlapsCount; index++)
{
Buffer buffer = _bufferOverlaps[index];
address = Math.Min(address, buffer.Address);
endAddress = Math.Max(endAddress, buffer.EndAddress);
lock (_buffers)
{
_buffers.Remove(buffer);
}
}
Buffer newBuffer = new Buffer(_context, address, endAddress - address, _bufferOverlaps.Take(overlapsCount));
lock (_buffers)
{
_buffers.Add(newBuffer);
}
for (int index = 0; index < overlapsCount; index++)
{
Buffer buffer = _bufferOverlaps[index];
int dstOffset = (int)(buffer.Address - newBuffer.Address);
buffer.CopyTo(newBuffer, dstOffset);
newBuffer.InheritModifiedRanges(buffer);
buffer.DisposeData();
}
newBuffer.SynchronizeMemory(address, endAddress - address);
// Existing buffers were modified, we need to rebind everything.
_rebind = true;
}
}
else
{
// No overlap, just create a new buffer.
Buffer buffer = new Buffer(_context, address, size);
lock (_buffers)
{
_buffers.Add(buffer);
}
}
ShrinkOverlapsBufferIfNeeded();
}
/// <summary>
/// Resizes the temporary buffer used for range list intersection results, if it has grown too much.
/// </summary>
private void ShrinkOverlapsBufferIfNeeded()
{
if (_bufferOverlaps.Length > OverlapsBufferMaxCapacity)
{
Array.Resize(ref _bufferOverlaps, OverlapsBufferMaxCapacity);
}
}
/// <summary>
/// Gets the address of the compute uniform buffer currently bound at the given index.
@ -624,7 +422,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
{
// The storage buffer size is not reliable (it might be lower than the actual size),
// so we bind the entire buffer to allow otherwise out of range accesses to work.
sRanges[bindingInfo.Binding] = GetBufferRangeTillEnd(
sRanges[bindingInfo.Binding] = _context.Methods.BufferCache.GetBufferRangeTillEnd(
bounds.Address,
bounds.Size,
bounds.Flags.HasFlag(BufferUsageFlags.Write));
@ -645,7 +443,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
if (bounds.Address != 0)
{
uRanges[bindingInfo.Binding] = GetBufferRange(bounds.Address, bounds.Size);
uRanges[bindingInfo.Binding] = _context.Methods.BufferCache.GetBufferRange(bounds.Address, bounds.Size);
}
}
@ -654,7 +452,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
CommitBufferTextureBindings();
// Force rebind after doing compute work.
_rebind = true;
Rebind();
}
/// <summary>
@ -666,7 +464,9 @@ namespace Ryujinx.Graphics.Gpu.Memory
{
foreach (var binding in _bufferTextures)
{
binding.Texture.SetStorage(GetBufferRange(binding.Address, binding.Size, binding.BindingInfo.Flags.HasFlag(TextureUsageFlags.ImageStore)));
var isStore = binding.BindingInfo.Flags.HasFlag(TextureUsageFlags.ImageStore);
var range = _context.Methods.BufferCache.GetBufferRange(binding.Address, binding.Size, isStore);
binding.Texture.SetStorage(range);
// The texture must be rebound to use the new storage if it was updated.
@ -696,14 +496,14 @@ namespace Ryujinx.Graphics.Gpu.Memory
if (_indexBuffer.Address != 0)
{
BufferRange buffer = GetBufferRange(_indexBuffer.Address, _indexBuffer.Size);
BufferRange buffer = _context.Methods.BufferCache.GetBufferRange(_indexBuffer.Address, _indexBuffer.Size);
_context.Renderer.Pipeline.SetIndexBuffer(buffer, _indexBuffer.Type);
}
}
else if (_indexBuffer.Address != 0)
{
SynchronizeBufferRange(_indexBuffer.Address, _indexBuffer.Size);
_context.Methods.BufferCache.SynchronizeBufferRange(_indexBuffer.Address, _indexBuffer.Size);
}
uint vbEnableMask = _vertexBuffersEnableMask;
@ -723,7 +523,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
continue;
}
BufferRange buffer = GetBufferRange(vb.Address, vb.Size);
BufferRange buffer = _context.Methods.BufferCache.GetBufferRange(vb.Address, vb.Size);
vertexBuffers[index] = new VertexBufferDescriptor(buffer, vb.Stride, vb.Divisor);
}
@ -741,7 +541,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
continue;
}
SynchronizeBufferRange(vb.Address, vb.Size);
_context.Methods.BufferCache.SynchronizeBufferRange(vb.Address, vb.Size);
}
}
@ -761,7 +561,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
continue;
}
tfbs[index] = GetBufferRange(tfb.Address, tfb.Size);
tfbs[index] = _context.Methods.BufferCache.GetBufferRange(tfb.Address, tfb.Size);
}
_context.Renderer.Pipeline.SetTransformFeedbackBuffers(tfbs);
@ -777,7 +577,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
continue;
}
SynchronizeBufferRange(tfb.Address, tfb.Size);
_context.Methods.BufferCache.SynchronizeBufferRange(tfb.Address, tfb.Size);
}
}
@ -831,9 +631,10 @@ namespace Ryujinx.Graphics.Gpu.Memory
if (bounds.Address != 0)
{
var isWrite = bounds.Flags.HasFlag(BufferUsageFlags.Write);
ranges[bindingInfo.Binding] = isStorage
? GetBufferRangeTillEnd(bounds.Address, bounds.Size, bounds.Flags.HasFlag(BufferUsageFlags.Write))
: GetBufferRange(bounds.Address, bounds.Size, bounds.Flags.HasFlag(BufferUsageFlags.Write));
? _context.Methods.BufferCache.GetBufferRangeTillEnd(bounds.Address, bounds.Size, isWrite)
: _context.Methods.BufferCache.GetBufferRange(bounds.Address, bounds.Size, isWrite);
}
}
}
@ -869,7 +670,7 @@ namespace Ryujinx.Graphics.Gpu.Memory
continue;
}
SynchronizeBufferRange(bounds.Address, bounds.Size);
_context.Methods.BufferCache.SynchronizeBufferRange(bounds.Address, bounds.Size);
}
}
}
@ -885,167 +686,26 @@ namespace Ryujinx.Graphics.Gpu.Memory
/// <param name="isImage">Whether the binding is for an image or a sampler</param>
public void SetBufferTextureStorage(ITexture texture, ulong address, ulong size, TextureBindingInfo bindingInfo, Format format, bool isImage)
{
CreateBuffer(address, size);
_context.Methods.BufferCache.CreateBuffer(address, size);
_bufferTextures.Add(new BufferTextureBinding(texture, address, size, bindingInfo, format, isImage));
}
/// <summary>
/// Copy a buffer data from a given address to another.
/// Force all bound textures and images to be rebound the next time CommitBindings is called.
/// </summary>
/// <remarks>
/// This does a GPU side copy.
/// </remarks>
/// <param name="srcVa">GPU virtual address of the copy source</param>
/// <param name="dstVa">GPU virtual address of the copy destination</param>
/// <param name="size">Size in bytes of the copy</param>
public void CopyBuffer(GpuVa srcVa, GpuVa dstVa, ulong size)
public void Rebind()
{
ulong srcAddress = TranslateAndCreateBuffer(srcVa.Pack(), size);
ulong dstAddress = TranslateAndCreateBuffer(dstVa.Pack(), size);
Buffer srcBuffer = GetBuffer(srcAddress, size);
Buffer dstBuffer = GetBuffer(dstAddress, size);
int srcOffset = (int)(srcAddress - srcBuffer.Address);
int dstOffset = (int)(dstAddress - dstBuffer.Address);
_context.Renderer.Pipeline.CopyBuffer(
srcBuffer.Handle,
dstBuffer.Handle,
srcOffset,
dstOffset,
(int)size);
if (srcBuffer.IsModified(srcAddress, size))
{
dstBuffer.SignalModified(dstAddress, size);
}
else
{
// Optimization: If the data being copied is already in memory, then copy it directly instead of flushing from GPU.
dstBuffer.ClearModified(dstAddress, size);
_context.PhysicalMemory.WriteUntracked(dstAddress, _context.PhysicalMemory.GetSpan(srcAddress, (int)size));
}
_rebind = true;
}
/// <summary>
/// Clears a buffer at a given address with the specified value.
/// </summary>
/// <remarks>
/// Both the address and size must be aligned to 4 bytes.
/// </remarks>
/// <param name="gpuVa">GPU virtual address of the region to clear</param>
/// <param name="size">Number of bytes to clear</param>
/// <param name="value">Value to be written into the buffer</param>
public void ClearBuffer(GpuVa gpuVa, ulong size, uint value)
{
ulong address = TranslateAndCreateBuffer(gpuVa.Pack(), size);
Buffer buffer = GetBuffer(address, size);
int offset = (int)(address - buffer.Address);
_context.Renderer.Pipeline.ClearBuffer(buffer.Handle, offset, (int)size, value);
buffer.SignalModified(address, size);
}
/// <summary>
/// Gets a buffer sub-range starting at a given memory address.
/// </summary>
/// <param name="address">Start address of the memory range</param>
/// <param name="size">Size in bytes of the memory range</param>
/// <param name="write">Whether the buffer will be written to by this use</param>
/// <returns>The buffer sub-range starting at the given memory address</returns>
private BufferRange GetBufferRangeTillEnd(ulong address, ulong size, bool write = false)
{
return GetBuffer(address, size, write).GetRange(address);
}
/// <summary>
/// Gets a buffer sub-range for a given memory range.
/// </summary>
/// <param name="address">Start address of the memory range</param>
/// <param name="size">Size in bytes of the memory range</param>
/// <param name="write">Whether the buffer will be written to by this use</param>
/// <returns>The buffer sub-range for the given range</returns>
private BufferRange GetBufferRange(ulong address, ulong size, bool write = false)
{
return GetBuffer(address, size, write).GetRange(address, size);
}
/// <summary>
/// Gets a buffer for a given memory range.
/// A buffer overlapping with the specified range is assumed to already exist on the cache.
/// </summary>
/// <param name="address">Start address of the memory range</param>
/// <param name="size">Size in bytes of the memory range</param>
/// <param name="write">Whether the buffer will be written to by this use</param>
/// <returns>The buffer where the range is fully contained</returns>
private Buffer GetBuffer(ulong address, ulong size, bool write = false)
{
Buffer buffer;
if (size != 0)
{
lock (_buffers)
{
buffer = _buffers.FindFirstOverlap(address, size);
}
buffer.SynchronizeMemory(address, size);
if (write)
{
buffer.SignalModified(address, size);
}
}
else
{
lock (_buffers)
{
buffer = _buffers.FindFirstOverlap(address, 1);
}
}
return buffer;
}
/// <summary>
/// Performs guest to host memory synchronization of a given memory range.
/// </summary>
/// <param name="address">Start address of the memory range</param>
/// <param name="size">Size in bytes of the memory range</param>
private void SynchronizeBufferRange(ulong address, ulong size)
{
if (size != 0)
{
Buffer buffer;
lock (_buffers)
{
buffer = _buffers.FindFirstOverlap(address, size);
}
buffer.SynchronizeMemory(address, size);
}
}
/// <summary>
/// Disposes all buffers in the cache.
/// It's an error to use the buffer manager after disposal.
/// Disposes the buffer manager.
/// It is an error to use the buffer manager after disposal.
/// </summary>
public void Dispose()
{
lock (_buffers)
{
foreach (Buffer buffer in _buffers)
{
buffer.Dispose();
}
}
_context.Methods.BufferCache.NotifyBuffersModified -= Rebind;
}
}
}
}