0
0
Fork 0
This repository has been archived on 2024-10-12. You can view files and clone it, but cannot push or open issues or pull requests.
ryujinx-final/src/Ryujinx.Graphics.Vulkan/Queries/BufferedQuery.cs
riperiperi 492a046335
Vulkan: Buffer Mirrors for MacOS performance (#4899)
* Initial implementation of buffer mirrors

Generally slower right now, goal is to reduce render passes in games that do inline updates

Fix support buffer mirrors

Reintroduce vertex buffer mirror

Add storage buffer support

Optimisation part 1

More optimisation

Avoid useless data copies.

Remove unused cbIndex stuff

Properly set write flag for storage buffers.

Fix minor issues

Not sure why this was here.

Fix BufferRangeList

Fix some big issues

Align storage buffers rather than getting full buffer as a range

Improves mirrorability of read-only storage buffers

Increase staging buffer size, as it now contains mirrors

Fix some issues with buffers not updating

Fix buffer SetDataUnchecked offset for one of the paths when using mirrors

Fix buffer mirrors interaction with buffer textures

Fix mirror rebinding

Move GetBuffer calls on indirect draws before BeginRenderPass to avoid draws without render pass

Fix mirrors rebase

Fix rebase 2023

* Fix crash when using stale vertex buffer

Similar to `Get` with a size that's too large, just treat it as a clamp.

* Explicitly set support buffer as mirrorable

* Address feedback

* Remove unused fragment of MVK workaround

* Replace logging for staging buffer OOM

* Address format issues

* Address more format issues

* Mini cleanup

* Address more things

* Rename BufferRangeList

* Support bounding range for ClearMirrors and UploadPendingData

* Add maximum size for vertex buffer mirrors

* Enable index buffer mirrors

Enabled on all platforms for the IbStreamer.

* Feedback

* Remove mystery BufferCache change

Probably macos related?

* Fix mirrors not creating when staging buffer is empty.

* Change log level to debug
2023-08-14 14:18:47 -03:00

216 lines
6.7 KiB
C#

using Ryujinx.Common.Logging;
using Ryujinx.Graphics.GAL;
using Silk.NET.Vulkan;
using System;
using System.Runtime.InteropServices;
using System.Threading;
namespace Ryujinx.Graphics.Vulkan.Queries
{
class BufferedQuery : IDisposable
{
private const int MaxQueryRetries = 5000;
private const long DefaultValue = -1;
private const long DefaultValueInt = 0xFFFFFFFF;
private const ulong HighMask = 0xFFFFFFFF00000000;
private readonly Vk _api;
private readonly Device _device;
private readonly PipelineFull _pipeline;
private QueryPool _queryPool;
private readonly BufferHolder _buffer;
private readonly IntPtr _bufferMap;
private readonly CounterType _type;
private readonly bool _result32Bit;
private readonly bool _isSupported;
private readonly long _defaultValue;
private int? _resetSequence;
public unsafe BufferedQuery(VulkanRenderer gd, Device device, PipelineFull pipeline, CounterType type, bool result32Bit)
{
_api = gd.Api;
_device = device;
_pipeline = pipeline;
_type = type;
_result32Bit = result32Bit;
_isSupported = QueryTypeSupported(gd, type);
if (_isSupported)
{
QueryPipelineStatisticFlags flags = type == CounterType.PrimitivesGenerated ?
QueryPipelineStatisticFlags.GeometryShaderPrimitivesBit : 0;
var queryPoolCreateInfo = new QueryPoolCreateInfo
{
SType = StructureType.QueryPoolCreateInfo,
QueryCount = 1,
QueryType = GetQueryType(type),
PipelineStatistics = flags,
};
gd.Api.CreateQueryPool(device, queryPoolCreateInfo, null, out _queryPool).ThrowOnError();
}
var buffer = gd.BufferManager.Create(gd, sizeof(long), forConditionalRendering: true);
_bufferMap = buffer.Map(0, sizeof(long));
_defaultValue = result32Bit ? DefaultValueInt : DefaultValue;
Marshal.WriteInt64(_bufferMap, _defaultValue);
_buffer = buffer;
}
private static bool QueryTypeSupported(VulkanRenderer gd, CounterType type)
{
return type switch
{
CounterType.SamplesPassed => true,
CounterType.PrimitivesGenerated => gd.Capabilities.SupportsPipelineStatisticsQuery,
CounterType.TransformFeedbackPrimitivesWritten => gd.Capabilities.SupportsTransformFeedbackQueries,
_ => false,
};
}
private static QueryType GetQueryType(CounterType type)
{
return type switch
{
CounterType.SamplesPassed => QueryType.Occlusion,
CounterType.PrimitivesGenerated => QueryType.PipelineStatistics,
CounterType.TransformFeedbackPrimitivesWritten => QueryType.TransformFeedbackStreamExt,
_ => QueryType.Occlusion,
};
}
public Auto<DisposableBuffer> GetBuffer()
{
return _buffer.GetBuffer();
}
public void Reset()
{
End(false);
Begin(null);
}
public void Begin(int? resetSequence)
{
if (_isSupported)
{
bool needsReset = resetSequence == null || _resetSequence == null || resetSequence.Value != _resetSequence.Value;
bool isOcclusion = _type == CounterType.SamplesPassed;
_pipeline.BeginQuery(this, _queryPool, needsReset, isOcclusion, isOcclusion && resetSequence != null);
}
_resetSequence = null;
}
public void End(bool withResult)
{
if (_isSupported)
{
_pipeline.EndQuery(_queryPool);
}
if (withResult && _isSupported)
{
Marshal.WriteInt64(_bufferMap, _defaultValue);
_pipeline.CopyQueryResults(this);
}
else
{
// Dummy result, just return 0.
Marshal.WriteInt64(_bufferMap, 0);
}
}
private bool WaitingForValue(long data)
{
return data == _defaultValue ||
(!_result32Bit && ((ulong)data & HighMask) == ((ulong)_defaultValue & HighMask));
}
public bool TryGetResult(out long result)
{
result = Marshal.ReadInt64(_bufferMap);
return result != _defaultValue;
}
public long AwaitResult(AutoResetEvent wakeSignal = null)
{
long data = _defaultValue;
if (wakeSignal == null)
{
while (WaitingForValue(data))
{
data = Marshal.ReadInt64(_bufferMap);
}
}
else
{
int iterations = 0;
while (WaitingForValue(data) && iterations++ < MaxQueryRetries)
{
data = Marshal.ReadInt64(_bufferMap);
if (WaitingForValue(data))
{
wakeSignal.WaitOne(1);
}
}
if (iterations >= MaxQueryRetries)
{
Logger.Error?.Print(LogClass.Gpu, $"Error: Query result {_type} timed out. Took more than {MaxQueryRetries} tries.");
}
}
return data;
}
public void PoolReset(CommandBuffer cmd, int resetSequence)
{
if (_isSupported)
{
_api.CmdResetQueryPool(cmd, _queryPool, 0, 1);
}
_resetSequence = resetSequence;
}
public void PoolCopy(CommandBufferScoped cbs)
{
var buffer = _buffer.GetBuffer(cbs.CommandBuffer, true).Get(cbs, 0, sizeof(long), true).Value;
QueryResultFlags flags = QueryResultFlags.ResultWaitBit;
if (!_result32Bit)
{
flags |= QueryResultFlags.Result64Bit;
}
_api.CmdCopyQueryPoolResults(
cbs.CommandBuffer,
_queryPool,
0,
1,
buffer,
0,
(ulong)(_result32Bit ? sizeof(int) : sizeof(long)),
flags);
}
public unsafe void Dispose()
{
_buffer.Dispose();
if (_isSupported)
{
_api.DestroyQueryPool(_device, _queryPool, null);
}
_queryPool = default;
}
}
}