2020-05-04 04:24:59 +02:00
|
|
|
|
using OpenTK.Graphics.OpenGL;
|
|
|
|
|
using Ryujinx.Graphics.GAL;
|
|
|
|
|
using System;
|
|
|
|
|
using System.Threading;
|
|
|
|
|
|
|
|
|
|
namespace Ryujinx.Graphics.OpenGL.Queries
|
|
|
|
|
{
|
|
|
|
|
class CounterQueueEvent : ICounterEvent
|
|
|
|
|
{
|
|
|
|
|
public event EventHandler<ulong> OnResult;
|
|
|
|
|
|
|
|
|
|
public QueryTarget Type { get; }
|
|
|
|
|
public bool ClearCounter { get; private set; }
|
|
|
|
|
public int Query => _counter.Query;
|
|
|
|
|
|
|
|
|
|
public bool Disposed { get; private set; }
|
|
|
|
|
public bool Invalid { get; set; }
|
|
|
|
|
|
2020-12-03 19:42:59 +01:00
|
|
|
|
public ulong DrawIndex { get; }
|
|
|
|
|
|
2020-05-04 04:24:59 +02:00
|
|
|
|
private CounterQueue _queue;
|
|
|
|
|
private BufferedQuery _counter;
|
|
|
|
|
|
Add a Multithreading layer for the GAL, multi-thread shader compilation at runtime (#2501)
* Initial Implementation
About as fast as nvidia GL multithreading, can be improved with faster command queuing.
* Struct based command list
Speeds up a bit. Still a lot of time lost to resource copy.
* Do shader init while the render thread is active.
* Introduce circular span pool V1
Ideally should be able to use structs instead of references for storing these spans on commands. Will try that next.
* Refactor SpanRef some more
Use a struct to represent SpanRef, rather than a reference.
* Flush buffers on background thread
* Use a span for UpdateRenderScale.
Much faster than copying the array.
* Calculate command size using reflection
* WIP parallel shaders
* Some minor optimisation
* Only 2 max refs per command now.
The command with 3 refs is gone. :relieved:
* Don't cast on the GPU side
* Remove redundant casts, force sync on window present
* Fix Shader Cache
* Fix host shader save.
* Fixup to work with new renderer stuff
* Make command Run static, use array of delegates as lookup
Profile says this takes less time than the previous way.
* Bring up to date
* Add settings toggle. Fix Muiltithreading Off mode.
* Fix warning.
* Release tracking lock for flushes
* Fix Conditional Render fast path with threaded gal
* Make handle iteration safe when releasing the lock
This is mostly temporary.
* Attempt to set backend threading on driver
Only really works on nvidia before launching a game.
* Fix race condition with BufferModifiedRangeList, exceptions in tracking actions
* Update buffer set commands
* Some cleanup
* Only use stutter workaround when using opengl renderer non-threaded
* Add host-conditional reservation of counter events
There has always been the possibility that conditional rendering could use a query object just as it is disposed by the counter queue. This change makes it so that when the host decides to use host conditional rendering, the query object is reserved so that it cannot be deleted. Counter events can optionally start reserved, as the threaded implementation can reserve them before the backend creates them, and there would otherwise be a short amount of time where the counter queue could dispose the event before a call to reserve it could be made.
* Address Feedback
* Make counter flush tracked again.
Hopefully does not cause any issues this time.
* Wait for FlushTo on the main queue thread.
Currently assumes only one thread will want to FlushTo (in this case, the GPU thread)
* Add SDL2 headless integration
* Add HLE macro commands.
Co-authored-by: Mary <mary@mary.zone>
2021-08-27 00:31:29 +02:00
|
|
|
|
private bool _hostAccessReserved = false;
|
|
|
|
|
private int _refCount = 1; // Starts with a reference from the counter queue.
|
|
|
|
|
|
2020-05-04 04:24:59 +02:00
|
|
|
|
private object _lock = new object();
|
Add a Multithreading layer for the GAL, multi-thread shader compilation at runtime (#2501)
* Initial Implementation
About as fast as nvidia GL multithreading, can be improved with faster command queuing.
* Struct based command list
Speeds up a bit. Still a lot of time lost to resource copy.
* Do shader init while the render thread is active.
* Introduce circular span pool V1
Ideally should be able to use structs instead of references for storing these spans on commands. Will try that next.
* Refactor SpanRef some more
Use a struct to represent SpanRef, rather than a reference.
* Flush buffers on background thread
* Use a span for UpdateRenderScale.
Much faster than copying the array.
* Calculate command size using reflection
* WIP parallel shaders
* Some minor optimisation
* Only 2 max refs per command now.
The command with 3 refs is gone. :relieved:
* Don't cast on the GPU side
* Remove redundant casts, force sync on window present
* Fix Shader Cache
* Fix host shader save.
* Fixup to work with new renderer stuff
* Make command Run static, use array of delegates as lookup
Profile says this takes less time than the previous way.
* Bring up to date
* Add settings toggle. Fix Muiltithreading Off mode.
* Fix warning.
* Release tracking lock for flushes
* Fix Conditional Render fast path with threaded gal
* Make handle iteration safe when releasing the lock
This is mostly temporary.
* Attempt to set backend threading on driver
Only really works on nvidia before launching a game.
* Fix race condition with BufferModifiedRangeList, exceptions in tracking actions
* Update buffer set commands
* Some cleanup
* Only use stutter workaround when using opengl renderer non-threaded
* Add host-conditional reservation of counter events
There has always been the possibility that conditional rendering could use a query object just as it is disposed by the counter queue. This change makes it so that when the host decides to use host conditional rendering, the query object is reserved so that it cannot be deleted. Counter events can optionally start reserved, as the threaded implementation can reserve them before the backend creates them, and there would otherwise be a short amount of time where the counter queue could dispose the event before a call to reserve it could be made.
* Address Feedback
* Make counter flush tracked again.
Hopefully does not cause any issues this time.
* Wait for FlushTo on the main queue thread.
Currently assumes only one thread will want to FlushTo (in this case, the GPU thread)
* Add SDL2 headless integration
* Add HLE macro commands.
Co-authored-by: Mary <mary@mary.zone>
2021-08-27 00:31:29 +02:00
|
|
|
|
private ulong _result = ulong.MaxValue;
|
2020-05-04 04:24:59 +02:00
|
|
|
|
|
2020-12-03 19:42:59 +01:00
|
|
|
|
public CounterQueueEvent(CounterQueue queue, QueryTarget type, ulong drawIndex)
|
2020-05-04 04:24:59 +02:00
|
|
|
|
{
|
|
|
|
|
_queue = queue;
|
|
|
|
|
|
|
|
|
|
_counter = queue.GetQueryObject();
|
|
|
|
|
Type = type;
|
|
|
|
|
|
2020-12-03 19:42:59 +01:00
|
|
|
|
DrawIndex = drawIndex;
|
|
|
|
|
|
2020-05-04 04:24:59 +02:00
|
|
|
|
_counter.Begin();
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
internal void Clear()
|
|
|
|
|
{
|
|
|
|
|
_counter.Reset();
|
|
|
|
|
ClearCounter = true;
|
|
|
|
|
}
|
|
|
|
|
|
2020-12-12 00:31:39 +01:00
|
|
|
|
internal void Complete(bool withResult)
|
2020-05-04 04:24:59 +02:00
|
|
|
|
{
|
2020-12-12 00:31:39 +01:00
|
|
|
|
_counter.End(withResult);
|
2020-05-04 04:24:59 +02:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
internal bool TryConsume(ref ulong result, bool block, AutoResetEvent wakeSignal = null)
|
|
|
|
|
{
|
|
|
|
|
lock (_lock)
|
|
|
|
|
{
|
|
|
|
|
if (Disposed)
|
|
|
|
|
{
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (ClearCounter || Type == QueryTarget.Timestamp)
|
|
|
|
|
{
|
|
|
|
|
result = 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
long queryResult;
|
|
|
|
|
|
|
|
|
|
if (block)
|
|
|
|
|
{
|
|
|
|
|
queryResult = _counter.AwaitResult(wakeSignal);
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
if (!_counter.TryGetResult(out queryResult))
|
|
|
|
|
{
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
result += (ulong)queryResult;
|
|
|
|
|
|
Add a Multithreading layer for the GAL, multi-thread shader compilation at runtime (#2501)
* Initial Implementation
About as fast as nvidia GL multithreading, can be improved with faster command queuing.
* Struct based command list
Speeds up a bit. Still a lot of time lost to resource copy.
* Do shader init while the render thread is active.
* Introduce circular span pool V1
Ideally should be able to use structs instead of references for storing these spans on commands. Will try that next.
* Refactor SpanRef some more
Use a struct to represent SpanRef, rather than a reference.
* Flush buffers on background thread
* Use a span for UpdateRenderScale.
Much faster than copying the array.
* Calculate command size using reflection
* WIP parallel shaders
* Some minor optimisation
* Only 2 max refs per command now.
The command with 3 refs is gone. :relieved:
* Don't cast on the GPU side
* Remove redundant casts, force sync on window present
* Fix Shader Cache
* Fix host shader save.
* Fixup to work with new renderer stuff
* Make command Run static, use array of delegates as lookup
Profile says this takes less time than the previous way.
* Bring up to date
* Add settings toggle. Fix Muiltithreading Off mode.
* Fix warning.
* Release tracking lock for flushes
* Fix Conditional Render fast path with threaded gal
* Make handle iteration safe when releasing the lock
This is mostly temporary.
* Attempt to set backend threading on driver
Only really works on nvidia before launching a game.
* Fix race condition with BufferModifiedRangeList, exceptions in tracking actions
* Update buffer set commands
* Some cleanup
* Only use stutter workaround when using opengl renderer non-threaded
* Add host-conditional reservation of counter events
There has always been the possibility that conditional rendering could use a query object just as it is disposed by the counter queue. This change makes it so that when the host decides to use host conditional rendering, the query object is reserved so that it cannot be deleted. Counter events can optionally start reserved, as the threaded implementation can reserve them before the backend creates them, and there would otherwise be a short amount of time where the counter queue could dispose the event before a call to reserve it could be made.
* Address Feedback
* Make counter flush tracked again.
Hopefully does not cause any issues this time.
* Wait for FlushTo on the main queue thread.
Currently assumes only one thread will want to FlushTo (in this case, the GPU thread)
* Add SDL2 headless integration
* Add HLE macro commands.
Co-authored-by: Mary <mary@mary.zone>
2021-08-27 00:31:29 +02:00
|
|
|
|
_result = result;
|
|
|
|
|
|
2020-05-04 04:24:59 +02:00
|
|
|
|
OnResult?.Invoke(this, result);
|
|
|
|
|
|
|
|
|
|
Dispose(); // Return the our resources to the pool.
|
|
|
|
|
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
public void Flush()
|
|
|
|
|
{
|
|
|
|
|
if (Disposed)
|
|
|
|
|
{
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// Tell the queue to process all events up to this one.
|
|
|
|
|
_queue.FlushTo(this);
|
|
|
|
|
}
|
|
|
|
|
|
Add a Multithreading layer for the GAL, multi-thread shader compilation at runtime (#2501)
* Initial Implementation
About as fast as nvidia GL multithreading, can be improved with faster command queuing.
* Struct based command list
Speeds up a bit. Still a lot of time lost to resource copy.
* Do shader init while the render thread is active.
* Introduce circular span pool V1
Ideally should be able to use structs instead of references for storing these spans on commands. Will try that next.
* Refactor SpanRef some more
Use a struct to represent SpanRef, rather than a reference.
* Flush buffers on background thread
* Use a span for UpdateRenderScale.
Much faster than copying the array.
* Calculate command size using reflection
* WIP parallel shaders
* Some minor optimisation
* Only 2 max refs per command now.
The command with 3 refs is gone. :relieved:
* Don't cast on the GPU side
* Remove redundant casts, force sync on window present
* Fix Shader Cache
* Fix host shader save.
* Fixup to work with new renderer stuff
* Make command Run static, use array of delegates as lookup
Profile says this takes less time than the previous way.
* Bring up to date
* Add settings toggle. Fix Muiltithreading Off mode.
* Fix warning.
* Release tracking lock for flushes
* Fix Conditional Render fast path with threaded gal
* Make handle iteration safe when releasing the lock
This is mostly temporary.
* Attempt to set backend threading on driver
Only really works on nvidia before launching a game.
* Fix race condition with BufferModifiedRangeList, exceptions in tracking actions
* Update buffer set commands
* Some cleanup
* Only use stutter workaround when using opengl renderer non-threaded
* Add host-conditional reservation of counter events
There has always been the possibility that conditional rendering could use a query object just as it is disposed by the counter queue. This change makes it so that when the host decides to use host conditional rendering, the query object is reserved so that it cannot be deleted. Counter events can optionally start reserved, as the threaded implementation can reserve them before the backend creates them, and there would otherwise be a short amount of time where the counter queue could dispose the event before a call to reserve it could be made.
* Address Feedback
* Make counter flush tracked again.
Hopefully does not cause any issues this time.
* Wait for FlushTo on the main queue thread.
Currently assumes only one thread will want to FlushTo (in this case, the GPU thread)
* Add SDL2 headless integration
* Add HLE macro commands.
Co-authored-by: Mary <mary@mary.zone>
2021-08-27 00:31:29 +02:00
|
|
|
|
public void DecrementRefCount()
|
|
|
|
|
{
|
|
|
|
|
if (Interlocked.Decrement(ref _refCount) == 0)
|
|
|
|
|
{
|
|
|
|
|
DisposeInternal();
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
public bool ReserveForHostAccess()
|
|
|
|
|
{
|
|
|
|
|
if (_hostAccessReserved)
|
|
|
|
|
{
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (IsValueAvailable())
|
|
|
|
|
{
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (Interlocked.Increment(ref _refCount) == 1)
|
|
|
|
|
{
|
|
|
|
|
Interlocked.Decrement(ref _refCount);
|
|
|
|
|
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
_hostAccessReserved = true;
|
|
|
|
|
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
public void ReleaseHostAccess()
|
|
|
|
|
{
|
|
|
|
|
_hostAccessReserved = false;
|
|
|
|
|
|
|
|
|
|
DecrementRefCount();
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
private void DisposeInternal()
|
|
|
|
|
{
|
|
|
|
|
_queue.ReturnQueryObject(_counter);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
private bool IsValueAvailable()
|
|
|
|
|
{
|
|
|
|
|
return _result != ulong.MaxValue || _counter.TryGetResult(out _);
|
|
|
|
|
}
|
|
|
|
|
|
2020-05-04 04:24:59 +02:00
|
|
|
|
public void Dispose()
|
|
|
|
|
{
|
|
|
|
|
Disposed = true;
|
Add a Multithreading layer for the GAL, multi-thread shader compilation at runtime (#2501)
* Initial Implementation
About as fast as nvidia GL multithreading, can be improved with faster command queuing.
* Struct based command list
Speeds up a bit. Still a lot of time lost to resource copy.
* Do shader init while the render thread is active.
* Introduce circular span pool V1
Ideally should be able to use structs instead of references for storing these spans on commands. Will try that next.
* Refactor SpanRef some more
Use a struct to represent SpanRef, rather than a reference.
* Flush buffers on background thread
* Use a span for UpdateRenderScale.
Much faster than copying the array.
* Calculate command size using reflection
* WIP parallel shaders
* Some minor optimisation
* Only 2 max refs per command now.
The command with 3 refs is gone. :relieved:
* Don't cast on the GPU side
* Remove redundant casts, force sync on window present
* Fix Shader Cache
* Fix host shader save.
* Fixup to work with new renderer stuff
* Make command Run static, use array of delegates as lookup
Profile says this takes less time than the previous way.
* Bring up to date
* Add settings toggle. Fix Muiltithreading Off mode.
* Fix warning.
* Release tracking lock for flushes
* Fix Conditional Render fast path with threaded gal
* Make handle iteration safe when releasing the lock
This is mostly temporary.
* Attempt to set backend threading on driver
Only really works on nvidia before launching a game.
* Fix race condition with BufferModifiedRangeList, exceptions in tracking actions
* Update buffer set commands
* Some cleanup
* Only use stutter workaround when using opengl renderer non-threaded
* Add host-conditional reservation of counter events
There has always been the possibility that conditional rendering could use a query object just as it is disposed by the counter queue. This change makes it so that when the host decides to use host conditional rendering, the query object is reserved so that it cannot be deleted. Counter events can optionally start reserved, as the threaded implementation can reserve them before the backend creates them, and there would otherwise be a short amount of time where the counter queue could dispose the event before a call to reserve it could be made.
* Address Feedback
* Make counter flush tracked again.
Hopefully does not cause any issues this time.
* Wait for FlushTo on the main queue thread.
Currently assumes only one thread will want to FlushTo (in this case, the GPU thread)
* Add SDL2 headless integration
* Add HLE macro commands.
Co-authored-by: Mary <mary@mary.zone>
2021-08-27 00:31:29 +02:00
|
|
|
|
|
|
|
|
|
DecrementRefCount();
|
2020-05-04 04:24:59 +02:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|