Commit Graph

4754 Commits

Author SHA1 Message Date
Scott Mansell 154cb4f722 Introduce an Event system to VideoCommon
A lot of the remaining complexity in Renderer is the massive Swap function
which tries to handle a bunch of FrameBegin/FrameEnd events.

Rather than create a new place for it. This event system will try
to distribute it all over the place
2023-01-31 19:41:24 +13:00
Scott Mansell d6cd8de1a7 Delete unused EFBHasAlphaChannel function 2023-01-31 19:41:24 +13:00
Scott Mansell 9d125a6e43 Move ConfigChanged out of RenderBase
There is this nice VideoConfig file that's perfect for it
2023-01-31 19:41:24 +13:00
Scott Mansell ca5ec13e13 Move GraphicsMod out of RenderBase 2023-01-31 19:41:24 +13:00
Scott Mansell 55d15bdd6e Move utiltily drawing out of RenderBase 2023-01-31 19:41:24 +13:00
Scott Mansell 99d3e489ea Move BoundingBox out of RenderBase
They were essentially just pass-though methods
2023-01-31 19:41:24 +13:00
Scott Mansell 8bc8e43dd6 Add virtual Initialize() to PerfQueryBase
Both DX12 and Vulkan already had one.
2023-01-31 19:41:24 +13:00
Scott Mansell 3c73707dfe Fix dead code. 2023-01-31 19:41:24 +13:00
Scott Mansell b007b8e104 Replace BeginUI/EndUI 2023-01-31 19:41:24 +13:00
Scott Mansell 18c799f0b6 Present: Set surface info before initializing ImGui
Otherwise you get subtle bugs in vulkan that take hours to track down.
2023-01-31 19:41:24 +13:00
Scott Mansell f158ff300b Handle VideoSoftware's present fallback better
Not a good idea to abuse bSupportsPostProcessing
2023-01-31 19:41:24 +13:00
Scott Mansell 26e00c3bb4 Fix warning about using & with bools 2023-01-31 19:41:23 +13:00
Scott Mansell 74d3b3443a Fix warning about compare sign mismatch 2023-01-31 19:41:23 +13:00
Scott Mansell 8f5b196019 Minimise include polution from RenderState 2023-01-31 19:41:23 +13:00
Scott Mansell 2a2014af09 Implement AbstractGfx for Vulkan 2023-01-31 19:41:23 +13:00
Scott Mansell 58b70b2fb2 Don't set common globals from Video Backends 2023-01-31 19:41:23 +13:00
Scott Mansell d37f83ffeb Implement AbstractGfx for Software & Null 2023-01-31 19:41:23 +13:00
Scott Mansell f0336a3129 Implement AbstractGfx for OpenGL
Mostly involves moving contents of OGLRender
to OGLGfx and OGLConfig
2023-01-31 18:46:04 +13:00
Scott Mansell 8a23629345 Split AbstractGfx out of Renderer
Almost all the virtual functions in Renderer are part of dolphin's
"graphics api abstraction layer", which has slowly formed over the
last decade or two.

Most of the work was done previously with the introduction of the
various "AbstractX" classes, associated with texture cache cleanups
and implementation of newer graphics APIs (Direct3D 12, Vulkan, Metal).
We are simply taking the last step and yeeting these functions out
of Renderer.

This "AbstractGfx" class is now completely agnostic of any details
from the flipper/hollywood GPU we are emulating, though somewhat
specialized.

(Will not build, this commit only contains changes outside VideoBackends)
2023-01-31 18:46:02 +13:00
Scott Mansell e57eb04ed3 Rename FrameDump.{cpp,h} to FrameDumpFFMpeg.{cpp,h}
The name kind of conflicts with a new FrameDumper class
2023-01-31 18:45:29 +13:00
Scott Mansell 0d4537d60f Move Presenting, Dumping and ImGui out of Renderer 2023-01-31 18:45:21 +13:00
Scott Mansell c38c76abad Refactor to remove virtual from RenderXFBToScreen
Slightly simplifies the upcoming refactor
2023-01-31 18:29:47 +13:00
Scott Mansell a01d5283ec TextureCache: Add content locking
Texture cache occasionally mutates textures for efficiency.
Which is awkward if we want to borrow those textures from texture cache
to do something else, such as a graphics debugger, or async presentation
on another thread.

Content locking provides a way to signal that the contents of a texture
cache entry should not change. Texture cache will be forced to use
alternative strategies.
2023-01-31 18:29:47 +13:00
Scott Mansell 606c18210d TextureCache: Refactor with smart pointers
The whole ownership model was getting a bit of a mess, with a some
of special cases to deal with. And I'm planning to make it even more
complex in the future.
So here is some upfront work to convert it over to reference counted
pointers.
2023-01-31 18:29:47 +13:00
Scott Mansell c1fd4a2013 Move TCacheEntry out of TextureCacheBase
Allows for fowards declaration
2023-01-31 18:29:47 +13:00
TellowKrinkle 600ad5f498 VideoCommon: Better logic op invert approximation 2023-01-30 14:04:37 -06:00
Pierre Bourdon a2f4606d33
Merge pull request #11205 from TellowKrinkle/AutoPresent
VideoBackends:Metal: Default to presentDrawable when vsync is on
2023-01-30 09:01:18 +01:00
Scott Mansell 6d7303993c
Merge pull request #11509 from Sam-Belliveau/improved-speed-counter
Tie Speed to CPU Speed and not VPS
2023-01-30 20:41:03 +13:00
TellowKrinkle 1119a9ba32 VideoCommon: Don't create pipelines with no render targets
Some backends don't like it when you do that
2023-01-29 14:35:23 -06:00
Sam Belliveau ebf114aef5 Tie Speed to CPU Speed and not VPS 2023-01-29 13:33:48 -05:00
iwubcode 7bea39b39e VideoBackends: add a way to load data into a specific level AND layer, default to layer 0 2023-01-27 18:46:53 -06:00
Admiral H. Curtiss 0a343007cb
PowerPC: Parametrize LR macro. 2023-01-27 15:22:43 +01:00
Admiral H. Curtiss be8d0b76ca
PowerPC: Remove PC macro. 2023-01-27 15:22:41 +01:00
Pierre Bourdon 9c9310bf44
Merge pull request #11208 from TellowKrinkle/CPUCull
Cull vertices on the CPU
2023-01-26 23:15:23 +01:00
TellowKrinkle a0e6d7c173 VideoCommon: Explicitly disable fp_contract for CPUCull 2023-01-25 02:21:56 -06:00
TellowKrinkle 1be0149146 VideoCommon: Cull vertices on the CPU 2023-01-25 02:21:56 -06:00
TellowKrinkle b170ef9651 VideoCommon: Add class for quickly transforming and culling vertices on the CPU 2023-01-25 02:21:56 -06:00
Pierre Bourdon 8d5edb13a9
Merge pull request #11367 from Sam-Belliveau/lagbegone
VideoCommon: VI Skip
2023-01-19 01:20:01 +01:00
Pierre Bourdon a7d1683149
Merge pull request #11432 from Pokechu22/generic-vertex-loader-test-error
VertexLoader: Fix loading tangent/binormal caches with NormalIndex3
2023-01-17 23:30:29 +01:00
Sam Belliveau 23ad07b368 Add VISkip 2023-01-14 01:56:37 -05:00
Pokechu22 3910bdd68b VertexLoader: Don't write position_cache if vertex is skipped
This is the behavior in the x64 and ARM64 vertex loaders. I don't know if it makes sense (the whole skipped vertex system seems jank, but several games behave incorrectly without it).
2023-01-13 15:38:00 -08:00
Pokechu22 16c0593a52 VertexLoader: Fix loading tangent/binormal caches with NormalIndex3 2023-01-13 15:38:00 -08:00
Pokechu22 2d53b73643 VertexLoaderTester: Add assertions for position/binormal/tangent caches 2023-01-13 15:38:00 -08:00
Pokechu22 c681d96d46 VertexLoaderTester: Use asserts instead of logs
Logs don't show up in unit tests, and since this is debugging functionality (though not enabled for tests by default) it's better to do it this way.
2023-01-13 15:38:00 -08:00
Sam Belliveau bc1cc9eeb4 CoreTiming: Throttle Before Every Event Using Chrono 2023-01-06 17:21:17 -05:00
Sam Belliveau 588a72a4fc
PerformanceTracker: Add ownership of m_log_name. 2023-01-06 20:27:25 +01:00
Sam Belliveau bc46089ab0
PerformanceTracker: Use shared_mutex instead of mutex so multiple threads can read at the same time. 2023-01-06 20:27:25 +01:00
Sam Belliveau 9143eb00fb
PerformanceMetrics: Fix Line Width Issue on Non HiDPI Screens 2023-01-06 20:27:17 +01:00
Mai cc14d60bbb
Merge pull request #11402 from Pokechu22/too-many-indices
IndexGenerator: Fix off-by-one in GetRemainingIndices
2023-01-05 15:15:09 +00:00
Pokechu22 cefcd9c93c IndexGenerator: Fix off-by-one in GetRemainingIndices
Fixes https://bugs.dolphin-emu.org/issues/13136.
2023-01-04 10:42:00 -08:00
Pokechu22 6c58ba353c IndexGenerator: Add assertion for overflow in GetRemainingIndices
This assertion is currently triggered by Pocoyo Racing (https://bugs.dolphin-emu.org/issues/13136).
2023-01-04 10:41:14 -08:00
Admiral H. Curtiss 2fdaf0a86e
HW/ProcessorInterface: Rename member variables to fit naming convention. 2023-01-04 03:17:26 +01:00
Admiral H. Curtiss 74e1577a2c
HW/ProcessorInterface: Refactor to class, move to Core::System. 2023-01-04 03:00:10 +01:00
Pokechu22 d91f340c86 VertexManagerBase: Move free space check to after the buffer is reset
Fixes incorrect logspam when the buffer needed to be reset on flushes (which we already were doing, but 52feed04db moved it to after the check was made). This is https://bugs.dolphin-emu.org/issues/10312.

I also converted it to an assert, as if this does happen, things are going to render incorrectly, so we want to make it obvious.
2023-01-03 17:06:51 -08:00
Admiral H. Curtiss ed9915308b
Merge pull request #11386 from Pokechu22/manual-texture-sampling-stereoscopic-layer-bounds-check
PixelShaderGen: Clamp texture layer when using manual texture sampling with stereoscopic 3D
2023-01-03 18:07:06 +01:00
Admiral H. Curtiss cb0a603c72
VideoCommon: De-globalize GeometryShaderManager class. 2022-12-29 15:33:19 +01:00
Admiral H. Curtiss 7e55cc9e90
VideoCommon/VertexShaderManager: Rename private variables. 2022-12-28 16:00:15 +01:00
Admiral H. Curtiss 50625728e0
VideoCommon: De-globalize VertexShaderManager class. 2022-12-28 15:52:29 +01:00
Pokechu22 f3df3a7727 PixelShaderGen: Clamp texture layer when using manual texture sampling with stereoscopic 3D
Otherwise, texelFetch() will use an out-of-bounds layer for game textures (that have 1 layer; EFB copies have 2 layers in stereoscopic 3D mode), which is undefined behavior (often resulting in a black image). The fast texture sampling path uses texture(), which always clamps (see https://www.khronos.org/opengl/wiki/Array_Texture#Access_in_shaders), so it was unaffected by this difference.
2022-12-27 13:45:13 -08:00
Admiral H. Curtiss 4549fb4acb
VideoCommon/PixelShaderManager: Rename s_ variables. 2022-12-27 20:20:27 +01:00
Admiral H. Curtiss 725bd64ec2
VideoCommon: De-globalize PixelShaderManager class. 2022-12-27 20:13:24 +01:00
Sam Belliveau 6052a5ca9b Fix Speed Counter Sample Size 2022-12-26 16:47:12 -05:00
Sam Belliveau e1d078ada6 Remove Redundant Access Specifiers 2022-12-23 20:13:01 -05:00
Sam Belliveau 673f81c18a New FrameTime/VBlank Analyzer + Graph 2022-12-23 19:52:53 -05:00
Minty-Meeo 05bebee802 Replace BitUtils with C++20: Counting Zeroes
With the upgrade to C++20, std::countl_zero and std::countr_zero can replace these home-spun implementations from the BitUtil.h library.
2022-12-21 04:17:00 -06:00
JosJuice 547d956278 Common: Use C++20 <bit> header in BitSet.h 2022-12-18 16:51:32 +01:00
Admiral H. Curtiss c486baffe6
VideoCommon/PixelEngine: Pass Core::System to methods. 2022-12-11 21:57:30 +01:00
Admiral H. Curtiss ec8aaf1f30
VideoCommon/PixelEngine: Refactor to class, move to Core::System. 2022-12-11 21:57:19 +01:00
Admiral H. Curtiss ceae4242fc
VideoCommon/Fifo: Pass Core::System to methods. 2022-12-10 17:16:26 +01:00
Admiral H. Curtiss 5624dd6d39
VideoCommon/Fifo: Refactor to class, move to Core::System. 2022-12-10 17:16:19 +01:00
Admiral H. Curtiss 8a3b8a925e
Core: Add option to force linear texture filtering. 2022-12-09 02:02:16 +01:00
Admiral H. Curtiss 839db591d9
HW/Memmap: Refactor Memory to class, move to Core::System. 2022-12-03 13:27:02 +01:00
Admiral H. Curtiss 6941d2e7e6
VideoCommon/CommandProcessor: Refactor to class, move to Core::System. 2022-11-29 08:15:01 +01:00
Admiral H. Curtiss c9558ecb4c
CoreTiming: Refactor to class. 2022-11-27 03:47:12 +01:00
Mai 7be4c90f67
Merge pull request #11166 from AdmiralCurtiss/mmio-system
HW: Pass System to MMIO handlers.
2022-11-24 00:58:45 +00:00
Pokechu22 cc5640245c Fix build errors related to formatting non-scoped enums 2022-11-23 13:45:43 -08:00
Admiral H. Curtiss 545fee9c94
Core/CommandProcessor: Reformat single/dual core dependent MMIO handlers. 2022-11-23 21:58:18 +01:00
Admiral H. Curtiss 0a6fdb9c13
HW: Pass System to MMIO handlers. 2022-11-23 05:52:21 +01:00
Mai a47ed2124f
Merge pull request #11253 from AdmiralCurtiss/core-timing-events-pass-system
CoreTiming: Pass Core::System to Events.
2022-11-23 04:46:11 +00:00
Mai cb232155d9
Merge pull request #11298 from Pokechu22/vertexloader-no-DataReader
VertexLoader: Eliminate use of DataReader
2022-11-23 04:40:53 +00:00
Pokechu22 e13094766d Fix VertexLoader.cpp using the wrong size for texture matrix indices without a corresponding texture
This regressed in 0a906f553f, I think (though I haven't confirmed it). Mario Tennis and Luigi's Mansion both use these for some reason (as far as I can tell, the data isn't actually used; it's just extra data included for no reason)
2022-11-22 17:49:54 -08:00
Pokechu22 0bcd3c79bb VertexLoader: Eliminate use of DataReader
DataReader is generally jank - it has a start and end pointer, but the end pointer is generally not used, and all of the vertex loaders mostly bypassed it anyways.

Wrapper code (the vertex loaer test, as well as Fifo.cpp and OpcodeDecoding.cpp) still uses it, as does the software vertex loader (which is not a subclass of VertexLoader). These can probably be eliminated later.
2022-11-22 17:17:11 -08:00
Sam Belliveau edb2c90b38
Improve FPS/VPS Counting and Revamp Appearance 2022-11-23 00:11:47 +01:00
Pokechu22 8f5a58f8be
Merge pull request #11285 from K0bin/cp-checks
VideoCommon:VertexLoaderManager: Reduce CPConfig checks
2022-11-21 15:35:06 -08:00
Robin Kertels f179fd9c33
VideoCommon:VertexLoaderManager: Reduce CPConfig checks
A bit of a micro optimization:
CheckCPConfiguration is called 350 times instead of 35k times.
2022-11-18 01:42:30 +01:00
Pokechu22 1fdf15a1be VideoCommon/FramebufferManager: Rename m_efb_cache_tiles_wide to m_efb_cache_tile_row_stride
This also initializes it in the non-tiled case just for consistency, although it's not actually read in that case.
2022-11-15 16:04:41 -08:00
Pokechu22 55e1c3b51d VideoCommon/FramebufferManager: use !IsUsingTiledEFBCache() in GetEFBCacheTileRect
This is for readability only, and should behave identically.
2022-11-15 15:48:56 -08:00
Pokechu22 90d05fe37e VideoCommon/FramebufferManager: Always update frame access mask on depth peeks
0e02ddcf52 removed separate logic for tiled versus non-tiled EFB peek caches, and as part of that made it so that color peeks updated the frame access mask even when a non-tiled cache is in use. However, the same change was not made for depth peeks. I'm not sure if this affected anything in practice.
2022-11-15 15:34:03 -08:00
JMC47 8a1c28be63
Merge pull request #11172 from K0bin/efb-refresh
VideoCommon:FramebufferManager: Mark cache as valid after refresh
2022-11-15 16:33:46 -05:00
TellowKrinkle 37a51f1d09 VideoCommon: Add an option to disable mipmaps
Needed by M1 fifoci to work around a minor non-determinism bug
2022-11-07 22:18:56 -06:00
Admiral H. Curtiss a98df18947
VideoCommon: Hold ImGui lock while initializing and shutting down ImGui. 2022-11-06 21:48:55 +01:00
Admiral H. Curtiss c931529e7a
Merge pull request #11225 from iwubcode/graphics-mod-textureload-callable
VideoCommon: call texture load graphics mod hook when textures are loaded
2022-11-06 18:58:51 +01:00
Admiral H. Curtiss 1348337277
Merge pull request #11246 from tellowkrinkle/PVDMemset
VideoCommon: Zero PortableVertexDeclarations on initialization
2022-11-06 18:54:21 +01:00
Admiral H. Curtiss a36a5c1308
CoreTiming: Pass Core::System to Events. 2022-11-06 17:54:58 +01:00
TellowKrinkle d9ed9ae331 VideoCommon: Zero PortableVertexDeclarations on initialization
Fixes an issue where the default initializer wouldn't initialize padding, making for fun non-determinism
2022-11-05 02:02:53 -05:00
Robin Kertels 0e02ddcf52
VideoCommon:FramebufferManager: Cleanup 2022-11-04 01:18:23 +01:00
Robin Kertels 371935d61e
VideoCommon:FramebufferManager: Mark cache as valid after refresh
Otherwise we might never hit the early return
if either depth or color doesnt have any active
tiles.
2022-11-04 01:18:23 +01:00
Pokechu22 fe559f3ed3 VideoCommon/Statistics: Require semicolons after statistics macros
This is clearer and reduces IntelliSense problems.
2022-10-29 15:39:41 -07:00
iwubcode ea3e133200 VideoCommon: call texture load graphics mod hook when Dolphin loads a texture 2022-10-28 19:24:43 -05:00
Pokechu22 581a575042 VertexLoader: Remove "too many initializer values" workaround functions
I originally added these in 2b1d1038a6, for both the TPipelineFunction and the size. The size was moved into the header in fdcd2b7d00 (making the size functions obsolete), but it seems that the functions themselves are no longer needed now.

I think I didn't use this approach before because it would have required ComponentFormatTable and ComponentCountRow to be templated, which would end up resulting in lines that were too long and thus wrapped in awkward places. (I *think* they didn't get inferred properly.) Now that we only need TPipelineFunction, the templating is not needed, and this ends up being a more readable version of the version with the wrapper functions.
2022-10-25 15:29:09 -07:00