Now usagetracker is properly in place, indirect draw HLE can be used without requiring any hacks. Dirtiness is now ignored when fetching macro arguments, and it's now the duty of the HLE impls themselves to perform flushing if they require it.
Indirect draws are implemented by having the macro arguments overflow into a seperate GP Entry that points directly to the indirect argument buffer. To HLE indirect draws a buffer needs to be created from this pointer, and it cannot be dereferenced on the CPU at any point to avoid hitting traps.
Allowing for parallel execution of channels never really benefitted many games and prevented optimisations such as keeping frequently used resources always locked to avoid the constant overhead of locking on the hot path.
GPFIFO code is very high throughput due to the sheer number of commands used for rendering. Adjust some types and switch to a if statement with hints to slightly increase processing speed.
Entirely rewrites the engine and interconnect code to take advantage of the subpixel and OOB blit support offered by the blit helper shader. The interconnect code is also cleaned up significantly with the 'context' naming being dropped due to potential conflicts with the 'context' from context lock
The position at which we call submit is a significant factor in performance and we did so at the end of PBs (PushBuffers), this isn't optimal as there could be multiple PBs queued up that would benefit from being in the same submission. We now delay the submission of the workload till we run out of PBs.
The flush callbacks inside `CommandExecutor` weren't being called prior to submission as they should've been, this fixes that by calling them. It additionally removes the requirement to manually flush Maxwell3D at the end of `ChannelGpfifo` pushbuffers as it's a flush callback and will automatically be called by `Submit`.
Co-authored-by: Billy Laws <blaws05@gmail.com>
Any work that was done in a `ChannelGpfifo` pushbuffer needs to be submitted at the end of it, if it isn't done then the work might incorrectly be not done till the next submission. This commit fixes it by calling `CommandExecutor::Submit` at the end of a pushbuffer, submitting any buffers that would've been left over.
Co-authored-by: Billy Laws <blaws05@gmail.com>
We utilize `pthread_setname_np` to set the thread names but didn't check for any errors which resulted in the `Skyline-Choreographer` and `ChannelCmdFifo` not having proper names as they exceeded the 16 character limit on thread names for the pthread function. This has now been fixed by changing the names and introducing error checking to invocations of this function.
The Fermi 2D engine implements both image blit and resolve operations, supporting subpixel sampling with both linear and point filtering.
Resolve operations are performed by sampling from the center of each pixel in order to resolve the final image from the MSAA samples
MSAA images are stored in memory like regular images but each pixels dimensions are scaled: e.g for 2x2 MSAA
```
112233
112233
445566
445566
```
These would be sampled with both duDx and duDy as 2 (integer part), resolving to the following:
```
123
456
```
Blit operations are performed by sampling from the corner of each pixel, scaling the image as one would expect.
This implementation isn't fully complete as Vulkan blit doesn't support some combinations which Fermi does, most notably between colour and depth stencil. These will be implemented properly at a later date, likely after the texture manager rework.
Out of Bounds Blit, used by some OpenGL games is also missing since supporting it requires texture aliasing, this will also be supported after the texture manager rework.
Co-authored-by: Billy Laws <blaws05@gmail.com>
Skyline's `exception` class now stores a list of all stack frames during the invocation of the exception. These can later be parsed by the exception handler to generate a human-readable stack trace. To assist with more complete stack traces, `-fno-omit-frame-pointer` is now passed on debug builds which forces the inclusion of frames on function calls.
A lot of logs are incomplete due to being unable to flush inside the signal handler, now we flush after any exceptions so that there is a guarantee of any exceptions being logged as this is crucial for proper debugging.
The DMA engine is used to perform DMA buffer/texture copies directly on the GPU. It can deswizzle arbritary regions of input textures, perform component remapping and swizzle into output textures.
This impl only supports 1D buffer copies, 2D ones will come later.
We cannot ignore accesses from the host to a region protected by the NCE Memory Trapping API, there's often access to regions which have overlap with a protected region unintentionally and those accesses need to be handled correctly rather than leading to a crash. This is done by implementing an additional signal handler `NCE::HostSignalHandler` to lookup any potential traps on a `SIGSEGV` and handle them correctly or when there isn't a corresponding trap raise a `SIGTRAP` when debugger is connected or delegate to `signal::ExceptionalSignalHandler` when it isn't.
The Kepler compute engine is used to run compute jobs encapsulated in to QMDs on the GPU, this commit doesn't implement compute itself but adds the register and QMD structs that will be needed for it in the future.
Previously for methods with count > 1 the subchannel and engine would be looked up for each part of the method rather than only doing so at the start. Each call also needed to be looked up to see if it touched a macro or GPFIFO method. Fix this by doing checks outside of the main dispatch loop with templated helper lambdas to avoid needing to repeat lots of code. Maxwell3D is the only subchannel with a fast path for now but more can be added later if needed.
Fermi2D supports macros in addition to Maxwell3D, these both share code memory. To support this we rework the macro interpreter to support passing in a target engine and abstract the communications out into an interface that can be implemented by applicable engines.
```
GPFIFO <-> MME <-> Maxwell3D
^ ^---> Fermi2D
X------------> I2M
X------------> MaxwellComputeB
X--Flush-----> MaxwellDMA
```
These are used heavily in OpenGL games, which now, together with the
previous syncpoint changes, work perfectly. The actual implementation is
rather novel as rather than using a per-class state machine for all
methods we only use it for those that are known to be split across
GpEntry boundaries, as a result only a single bounds check is added to
the hot path of contiguous method execution and the performance loss is
negligible.
Allows the execution of multiple channels at the same time, with locking
being performed on the host GPU scheduler layer, address spaces can be
bound to one or more channels.
The unique pointer to a device in the map was simply reset rather than deleting the entry from the map, this resulted in the device not being properly closed and when the device was reopened then the `emplace` was a NOP as the entry already existed. This resulted in a `nullptr` dereference down the line when an application attempted to issue an IOCTL to a device that was previously closed and reopened. This is known to occur in Deko3D as it recreates the context when loading an example which includes closing and reopening devices.
Using a u32 for the loop index prevents masking on all increments,
giving a moderate performance increase.
Passing methods as u32 parameters and stopping subChannel being passed
gives quite a significant increase when combined with the inlining
allowed by subchannel based engine selection.
We decided to restructure Skyline to draw a layer of separation between guest and host GPU. We're reserving the `gpu` namespace and directory for purely host GPU and creating a new `soc` directory and namespace for emulation of parts of the X1 SoC which is currently limited to guest GPU but will be expanded to contain components like the audio DSP down the line.