ryujinx-mirror/Ryujinx.Graphics.Gpu/Engine/GPFifo/GPFifoProcessor.cs
mpnico 8e1adb95cf
Add support for HLE macros and accelerate MultiDrawElementsIndirectCount #2 (#2557)
* Add support for HLE macros and accelerate MultiDrawElementsIndirectCount

* Add missing barrier

* Fix index buffer count

* Add support check for each macro hle before use

* Add missing xml doc

Co-authored-by: gdkchan <gab.dark.100@gmail.com>
2021-08-26 23:50:28 +02:00

331 lines
12 KiB
C#

using Ryujinx.Graphics.Device;
using Ryujinx.Graphics.Gpu.Engine.Compute;
using Ryujinx.Graphics.Gpu.Engine.Dma;
using Ryujinx.Graphics.Gpu.Engine.InlineToMemory;
using Ryujinx.Graphics.Gpu.Engine.Threed;
using Ryujinx.Graphics.Gpu.Engine.Twod;
using Ryujinx.Graphics.Gpu.Memory;
using System;
using System.Runtime.CompilerServices;
namespace Ryujinx.Graphics.Gpu.Engine.GPFifo
{
/// <summary>
/// Represents a GPU General Purpose FIFO command processor.
/// </summary>
class GPFifoProcessor
{
private const int MacrosCount = 0x80;
private const int MacroIndexMask = MacrosCount - 1;
private const int LoadInlineDataMethodOffset = 0x6d;
private const int UniformBufferUpdateDataMethodOffset = 0x8e4;
private readonly GpuChannel _channel;
/// <summary>
/// Channel memory manager.
/// </summary>
public MemoryManager MemoryManager => _channel.MemoryManager;
/// <summary>
/// 3D Engine.
/// </summary>
public ThreedClass ThreedClass => _3dClass;
/// <summary>
/// Internal GPFIFO state.
/// </summary>
private struct DmaState
{
public int Method;
public int SubChannel;
public int MethodCount;
public bool NonIncrementing;
public bool IncrementOnce;
}
private DmaState _state;
private readonly ThreedClass _3dClass;
private readonly ComputeClass _computeClass;
private readonly InlineToMemoryClass _i2mClass;
private readonly TwodClass _2dClass;
private readonly DmaClass _dmaClass;
private readonly GPFifoClass _fifoClass;
/// <summary>
/// Creates a new instance of the GPU General Purpose FIFO command processor.
/// </summary>
/// <param name="context">GPU context</param>
/// <param name="channel">Channel that the GPFIFO processor belongs to</param>
public GPFifoProcessor(GpuContext context, GpuChannel channel)
{
_channel = channel;
_fifoClass = new GPFifoClass(context, this);
_3dClass = new ThreedClass(context, channel);
_computeClass = new ComputeClass(context, channel, _3dClass);
_i2mClass = new InlineToMemoryClass(context, channel);
_2dClass = new TwodClass(channel);
_dmaClass = new DmaClass(context, channel, _3dClass);
}
/// <summary>
/// Processes a command buffer.
/// </summary>
/// <param name="baseGpuVa">Base GPU virtual address of the command buffer</param>
/// <param name="commandBuffer">Command buffer</param>
public void Process(ulong baseGpuVa, ReadOnlySpan<int> commandBuffer)
{
for (int index = 0; index < commandBuffer.Length; index++)
{
int command = commandBuffer[index];
ulong gpuVa = baseGpuVa + (ulong)index * 4;
if (_state.MethodCount != 0)
{
if (TryFastI2mBufferUpdate(commandBuffer, ref index))
{
continue;
}
Send(gpuVa, _state.Method, command, _state.SubChannel, _state.MethodCount <= 1);
if (!_state.NonIncrementing)
{
_state.Method++;
}
if (_state.IncrementOnce)
{
_state.NonIncrementing = true;
}
_state.MethodCount--;
}
else
{
CompressedMethod meth = Unsafe.As<int, CompressedMethod>(ref command);
if (TryFastUniformBufferUpdate(meth, commandBuffer, index))
{
index += meth.MethodCount;
continue;
}
switch (meth.SecOp)
{
case SecOp.IncMethod:
case SecOp.NonIncMethod:
case SecOp.OneInc:
_state.Method = meth.MethodAddress;
_state.SubChannel = meth.MethodSubchannel;
_state.MethodCount = meth.MethodCount;
_state.IncrementOnce = meth.SecOp == SecOp.OneInc;
_state.NonIncrementing = meth.SecOp == SecOp.NonIncMethod;
break;
case SecOp.ImmdDataMethod:
Send(gpuVa, meth.MethodAddress, meth.ImmdData, meth.MethodSubchannel, true);
break;
}
}
}
_3dClass.FlushUboDirty();
}
/// <summary>
/// Tries to perform a fast Inline-to-Memory data update.
/// If successful, all data will be copied at once, and <see cref="DmaState.MethodCount"/>
/// command buffer entries will be consumed.
/// </summary>
/// <param name="commandBuffer">Command buffer where the data is contained</param>
/// <param name="offset">Offset at <paramref name="commandBuffer"/> where the data is located, auto-incremented on success</param>
/// <returns>True if the fast copy was successful, false otherwise</returns>
[MethodImpl(MethodImplOptions.AggressiveInlining)]
private bool TryFastI2mBufferUpdate(ReadOnlySpan<int> commandBuffer, ref int offset)
{
if (_state.Method == LoadInlineDataMethodOffset && _state.NonIncrementing && _state.SubChannel <= 2)
{
int availableCount = commandBuffer.Length - offset;
int consumeCount = Math.Min(_state.MethodCount, availableCount);
var data = commandBuffer.Slice(offset, consumeCount);
if (_state.SubChannel == 0)
{
_3dClass.LoadInlineData(data);
}
else if (_state.SubChannel == 1)
{
_computeClass.LoadInlineData(data);
}
else /* if (_state.SubChannel == 2) */
{
_i2mClass.LoadInlineData(data);
}
offset += consumeCount - 1;
_state.MethodCount -= consumeCount;
return true;
}
return false;
}
/// <summary>
/// Tries to perform a fast constant buffer data update.
/// If successful, all data will be copied at once, and <see cref="CompressedMethod.MethodCount"/> + 1
/// command buffer entries will be consumed.
/// </summary>
/// <param name="meth">Compressed method to be checked</param>
/// <param name="commandBuffer">Command buffer where <paramref name="meth"/> is contained</param>
/// <param name="offset">Offset at <paramref name="commandBuffer"/> where <paramref name="meth"/> is located</param>
/// <returns>True if the fast copy was successful, false otherwise</returns>
[MethodImpl(MethodImplOptions.AggressiveInlining)]
private bool TryFastUniformBufferUpdate(CompressedMethod meth, ReadOnlySpan<int> commandBuffer, int offset)
{
int availableCount = commandBuffer.Length - offset;
if (meth.MethodAddress == UniformBufferUpdateDataMethodOffset &&
meth.MethodCount < availableCount &&
meth.SecOp == SecOp.NonIncMethod)
{
_3dClass.ConstantBufferUpdate(commandBuffer.Slice(offset + 1, meth.MethodCount));
return true;
}
return false;
}
/// <summary>
/// Sends a uncompressed method for processing by the graphics pipeline.
/// </summary>
/// <param name="gpuVa">GPU virtual address where the command word is located</param>
/// <param name="meth">Method to be processed</param>
private void Send(ulong gpuVa, int offset, int argument, int subChannel, bool isLastCall)
{
if (offset < 0x60)
{
_fifoClass.Write(offset * 4, argument);
}
else if (offset < 0xe00)
{
offset *= 4;
switch (subChannel)
{
case 0:
_3dClass.Write(offset, argument);
break;
case 1:
_computeClass.Write(offset, argument);
break;
case 2:
_i2mClass.Write(offset, argument);
break;
case 3:
_2dClass.Write(offset, argument);
break;
case 4:
_dmaClass.Write(offset, argument);
break;
}
}
else
{
IDeviceState state = subChannel switch
{
0 => _3dClass,
3 => _2dClass,
_ => null
};
if (state != null)
{
int macroIndex = (offset >> 1) & MacroIndexMask;
if ((offset & 1) != 0)
{
_fifoClass.MmePushArgument(macroIndex, gpuVa, argument);
}
else
{
_fifoClass.MmeStart(macroIndex, argument);
}
if (isLastCall)
{
_fifoClass.CallMme(macroIndex, state);
_3dClass.PerformDeferredDraws();
}
}
}
}
/// <summary>
/// Writes data directly to the state of the specified class.
/// </summary>
/// <param name="classId">ID of the class to write the data into</param>
/// <param name="offset">State offset in bytes</param>
/// <param name="value">Value to be written</param>
public void Write(ClassId classId, int offset, int value)
{
switch (classId)
{
case ClassId.Threed:
_3dClass.Write(offset, value);
break;
case ClassId.Compute:
_computeClass.Write(offset, value);
break;
case ClassId.InlineToMemory:
_i2mClass.Write(offset, value);
break;
case ClassId.Twod:
_2dClass.Write(offset, value);
break;
case ClassId.Dma:
_dmaClass.Write(offset, value);
break;
case ClassId.GPFifo:
_fifoClass.Write(offset, value);
break;
}
}
/// <summary>
/// Sets the shadow ram control value of all sub-channels.
/// </summary>
/// <param name="control">New shadow ram control value</param>
public void SetShadowRamControl(int control)
{
_3dClass.SetShadowRamControl(control);
}
/// <summary>
/// Forces a full host state update by marking all state as modified,
/// and also requests all GPU resources in use to be rebound.
/// </summary>
public void ForceAllDirty()
{
_3dClass.ForceStateDirty();
_channel.BufferManager.Rebind();
_channel.TextureManager.Rebind();
}
/// <summary>
/// Perform any deferred draws.
/// </summary>
public void PerformDeferredDraws()
{
_3dClass.PerformDeferredDraws();
}
}
}