Commit Graph

124 Commits

Author SHA1 Message Date
gdkchan
6b23a2c125 New shader translator implementation (#654)
* Start implementing a new shader translator

* Fix shift instructions and a typo

* Small refactoring on StructuredProgram, move RemovePhis method to a separate class

* Initial geometry shader support

* Implement TLD4

* Fix -- There's no negation on FMUL32I

* Add constant folding and algebraic simplification optimizations, nits

* Some leftovers from constant folding

* Avoid cast for constant assignments

* Add a branch elimination pass, and misc small fixes

* Remove redundant branches, add expression propagation and other improvements on the code

* Small leftovers -- add missing break and continue, remove unused properties, other improvements

* Add null check to handle empty block cases on block visitor

* Add HADD2 and HMUL2 half float shader instructions

* Optimize pack/unpack sequences, some fixes related to half float instructions

* Add TXQ, TLD, TLDS and TLD4S shader texture instructions, and some support for bindless textures, some refactoring on codegen

* Fix copy paste mistake that caused RZ to be ignored on the AST instruction

* Add workaround for conditional exit, and fix half float instruction with constant buffer

* Add missing 0.0 source for TLDS.LZ variants

* Simplify the switch for TLDS.LZ

* Texture instructions related fixes

* Implement the HFMA instruction, and some misc. fixes

* Enable constant folding on UnpackHalf2x16 instructions

* Refactor HFMA to use OpCode* for opcode decoding rather than on the helper methods

* Remove the old shader translator

* Remove ShaderDeclInfo and other unused things

* Add dual vertex shader support

* Add ShaderConfig, used to pass shader type and maximum cbuffer size

* Move and rename some instruction enums

* Move texture instructions into a separate file

* Move operand GetExpression and locals management to OperandManager

* Optimize opcode decoding using a simple list and binary search

* Add missing condition for do-while on goto elimination

* Misc. fixes on texture instructions

* Simplify TLDS switch

* Address PR feedback, and a nit
2019-04-18 09:57:08 +10:00
jduncanator
af65ed3930
Add missing TextureCubeMapArray texture type entry (#657)
* Add missing TextureCubeMapArray texture type entry

* Duplicate comment from other Create path
2019-04-11 23:19:49 +10:00
BaronKiko
1b2e430e88 Cubemap array support (#649) 2019-03-18 09:55:02 +11:00
BaronKiko
a2a4650108 Rebased, in theory (#610) 2019-03-04 13:51:23 +11:00
Alex Barney
1f554c1093 Do naming refactoring on Ryujinx.Graphics (#611)
* Renaming part 1

* Renaming part 2

* Renaming part 3

* Renaming part 4

* Renaming part 5

* Renaming part 6

* Renaming part 7

* Renaming part 8

* Renaming part 9

* Renaming part 10

* General cleanup

* Thought I got all of these

* Apply #595

* Additional renaming

* Tweaks from feedback

* Rename files
2019-03-04 12:45:25 +11:00
BaronKiko
0973daefa1 Fixed Scissor Test on Intel based GPUs (#595)
* Reworked scissor tests to remove fixme and handle issues with intel gpu's

* Error handling for scissor tests

* Disable strict opengl by default

* Reformatting for JD

* Updated scheme for new property in config

* Fixed typo

* Moved magic value to constant. I liked the magic :(

* Fixed ordering for undertale

* Fixed undertale bug

* Removed strict opengl in favour of required. With this an exception is no longer thrown, just a warning for required extensions

* Uses clamp instead of if's

* Removed evil tabs and no longer used include
2019-03-02 21:50:21 +11:00
Thomas Guillemard
884b4e5fd3 Initial non 2D textures support (#525)
* Initial non 2D textures support

- Shaders still need to be changed
- Some types aren't yet implemented

* Start implementing texture instructions suffixes

Fix wrong texture type with cube and TEXS

Also support array textures in TEX and TEX.B

Clean up TEX and TEXS coords managment

Fix TEXS.LL with non-2d textures

Implement TEX.AOFFI

Get the right arguments for TEX, TEXS and TLDS

Also, store suffix operands in appropriate values to support multiple
suffix combinaisons

* Support depth in read/writeTexture

Also support WrapR and detect mipmap

* Proper cube map textures support + fix TEXS.LZ

* Implement depth compare

* some code clean up

* Implement CubeMap textures in OGLTexture.Create

* Implement TLD4 and TLD4S

* Add Texture 1D support

* updates comments

* fix some code style issues

* Fix some nits + rename some things to be less confusing

* Remove GetSuffix local functions

* AOFFI => AOffI

* TextureType => GalTextureTarget

* finish renaming TextureType to TextureTarget

* Disable LL, LZ and LB support in the decompiler

This needs more work at the GL level (GLSL implementation should be
right)

* Revert "Disable LL, LZ and LB support in the decompiler"

This reverts commit 64536c3d9f673645faff3152838d1413c3203395.

* Fix TEXS ARRAY_2D index

* ImageFormat depth should be 1 for all image format

* Fix shader build issues with sampler1DShadow and texture

* Fix DC & AOFFI combinaison with TEX/TEXS

* Support AOFFI with TLD4 and TLD4S

* Fix shader compilation error for TLD4.AOFFI with no DC

* Fix binding isuses on the 2d copy engine

TODO: support 2d array copy

* Support 2D array copy operation in the 2D engine

This make every copy right in the GPU side.
Thie CPU copy probably needs to be updated

* Implement GetGpuSize + fix somes issues with 2d engine copies

TODO: mipmap level in it

* Don't throw an exception in the layer handling

* Fix because of rebase

* Reject 2d layers of non textures in 2d copy engine

* Add 3D textures and mipmap support on BlockLinearSwizzle

* Fix naming on new BitUtils methods

* gpu cache: Make sure to invalidate textures that doesn't have the same target

* Add the concept of layer count for array instead of using depth

Also cleanup GetGpuSize as Swizzle can compute the size with mipmap

* Support multi layer with mip map in ReadTexture

* Add more check for cache invalidation & remove cubemap and cubemap array code for now

Also fix compressed 2d array

* Fix texelFetchOffset shader build error

* Start looking into cube map again

Also add some way to log write in register in engines

* fix write register log levles

* Remove debug logs in WriteRegister

* Disable AOFFI support on non NVIDIA drivers

* Fix code align
2019-02-28 12:12:24 +11:00
gdkchan
9cbcbaa90c Fix XMAD shader instruction, gl_FrontFacing and enable face culling (#583)
* Fix XMAD shader instruction implementation

* Fix gl_FrontFacing constant value

* Enable face culling again

* Fix typo
2019-02-15 14:23:14 +11:00
BaronKiko
f5b4f6ccc4 Scissor test fix (#563)
* Handle negative viewport coordinates

* Disable scissor before framebuffer blit

* Comment to explain scissor disable will be reenabled if needed

* Comma and spelling mistake
2019-01-31 22:37:07 -03:00
gdkchan
e10ff17e2d
Initial support for shader half float instructions (#507) 2019-01-31 09:43:24 -03:00
BaronKiko
0cd5ba03fe Scissor test implementation. Partially stubbed until geometry shaders… (#556)
* Scissor test implementation. Partially stubbed until geometry shaders are fixed

* Apply to all viewports when geometry shaders are disabled.

* Also apply enable cap to all viewports when geometry shaders are disabled

* Added fixme as per suggestion

Co-Authored-By: BaronKiko <BaronKiko@users.noreply.github.com>

* Apparently no alignment needed here.

* Comment on new line

* Correct height calculation
2019-01-13 23:26:42 +02:00
Thomas Guillemard
b4d91402c6 Some improvements for nvnflinger (#555)
* Initial fixes for last release of libnx

For now, the framebuffer aren't okay but it will not crash/

* Improve code reaadability in NvFlinger parsing

* Make surfaces access more userfriendly

* Add ColorFormat

* Fix code style in ColorFormat.cs

* Add multiple framebuffer support in nvnflinger

This fix libnx console rendering

* Move ReadStruct/WriteStruct to Ryujinx.Common

* fix the last nit

* Fix inverted color for R5G6B5

Also add some other format that libnx might uses.

* Remove hardcoded BlockHeight in nvflinger
2019-01-05 22:26:16 +01:00
gdkchan
1fd1ec5664
Do not attempt to read unused attribute, fix attributes count (#520) 2018-12-26 17:29:50 -02:00
Thomas Guillemard
33e7c89822 Move MaxUboSize definition (#530)
* Move MaxUboSize definition

This fix a crash on Ryujinx.ShaderTools caused by the absence of an
OpenGL context.

* Use a constant for the value in ShaderTools

* Address comments
2018-12-17 22:32:12 -02:00
gdkchan
59964f667c
Add support for bigger UBOs, fix sRGB regression, small improvement t… (#503)
* Add support for bigger UBOs, fix sRGB regression, small improvement to the 2D copy engine

* Break into multiple lines

* Read fractions for source/step values on the 2d copy engine aswell

* Use fixed point math for more speed

* Fix reinterpret when texture sizes are different
2018-11-28 21:09:44 -02:00
gdkchan
00579927e4
Better process implementation (#491)
* Initial implementation of KProcess

* Some improvements to the memory manager, implement back guest stack trace printing

* Better GetInfo implementation, improve checking in some places with information from process capabilities

* Allow the cpu to read/write from the correct memory locations for accesses crossing a page boundary

* Change long -> ulong for address/size on memory related methods to avoid unnecessary casts

* Attempt at implementing ldr:ro with new KProcess

* Allow BSS with size 0 on ldr:ro

* Add checking for memory block slab heap usage, return errors if full, exit gracefully

* Use KMemoryBlockSize const from KMemoryManager

* Allow all methods to read from non-contiguous locations

* Fix for TransactParcelAuto

* Address PR feedback, additionally fix some small issues related to the KIP loader and implement SVCs GetProcessId, GetProcessList, GetSystemInfo, CreatePort and ManageNamedPort

* Fix wrong check for source pages count from page list on MapPhysicalMemory

* Fix some issues with UnloadNro on ldr:ro
2018-11-28 20:18:09 -02:00
HorrorTroll
e7fe7d7247 Add RGBA16 Unorm (#513) 2018-11-23 12:21:45 -02:00
HorrorTroll
0f5efe050f Add BGRA8 Srgb (#505) 2018-11-19 02:07:09 -02:00
gdkchan
b7613dd4b8 Enable SRGB framebuffers by default (#502) 2018-11-17 05:10:51 +01:00
gdkchan
d2bb458b51 Improved GPU command lists decoding (#499)
* Better implementation of the DMA pusher, misc fixes

* Remove some debug code

* Correct RGBX8 format

* Add support for linked Texture Sampler Control

* Attempt to fix upside down screen issue
2018-11-17 05:01:31 +01:00
gdkchan
85ffd76016 Force cache to remove entries when memory usage exceeds a given threshold (#492) 2018-11-14 23:22:02 +01:00
gdkchan
453543fb88 Initialize blend registers with default values (#497) 2018-11-14 22:50:31 +01:00
Ac_K
437962a65d Add BGR5A1 Image Format (#495)
* Add `BGR5A1` to ImageUtils.cs

Add `BGR5A1` Image Format.

* Add BGR5A1 to OGLEnumConverter.cs
2018-11-11 02:27:06 -02:00
gdkchan
44c1cf3fe5
Implment common and independent blend properly (fixes #458) (#485)
* Implment common and independent blend properly

* Nits
2018-11-01 01:22:24 -03:00
gdkchan
5a87e58183 Fix regression caused by wrong time delta calculation on cache deletion methods 2018-10-30 11:42:27 -03:00
jduncanator
c1b7340023 Timing: Optimize Timestamp Aquisition (#479)
* Timing: Optimize Timestamp Aquisition

Currently, we make use of Environment.TickCount in a number of places. This has some downsides, mainly being that the TickCount is a signed 32-bit integer, and has an effective limit of ~25 days before overflowing and wrapping around. Due to the signed-ness of the value, this also caused issues with negative numbers. This resolves these issues by using a 64-bit tick count obtained from Performance Counters (via the Stopwatch class). This has a beneficial side effect of being significantly more accurate than the TickCount.

* Timing: Rename ElapsedTicks to ElapsedMilliseconds and expose TicksPerX

* Timing: Some style changes

* Timing: Align static variable initialization
2018-10-28 19:31:13 -03:00
gdkchan
f0a49a1c94
Fix Color Mask values (#473)
* Fix color mask common, set default value on first color mask register

* Missing check

* Better exception messages

* Address PR feedback

* Add fixme as per review feedback
2018-10-25 18:30:09 -03:00
gdkchan
9ace6b9285
Fix for render target and a shader compilation issue (#471)
* Fix render target using possibly deleted or wrong handles

* Fix basic blocks with only a KIL instruction on the shader translator

* Formatting fix
2018-10-23 17:59:52 -03:00
gdkchan
044b84b078
Add depth range support on the GPU (#472)
* Add depth range support on the GPU

* Address PR feedback
2018-10-23 16:04:08 -03:00
gdkchan
0e1e094b7a
Improve texture tables (#457)
* Improve texture tables

* More renaming and other tweaks

* Minor tweaks
2018-10-17 18:02:23 -03:00
HorrorTroll
76330b10b4 Add G8R8Unorm, G8R8Snorm, B8G8R8A8, D24_S8 Uint, R8Uint, R32Uint (#426)
* Implement B5G6R5Unorm & BGR5A1Unorm

* Fix R8G8 Unorm to G8R8 Unorm

* Added back R8G8 Unorm

* Fix G8R8Unorm, add R8G8B8A8

* Add D24_S8 Uint

* Add R8Uint & R32Uint

* Another fixed

* Reverting back

* R8G8B8A8 change to B8G8R8A8

* Add G8R8 Snorm
2018-10-14 18:23:23 -03:00
gdkchan
72317d7777
Add support for saturation on some shader instructions, fix ReadTexture alignment and add ColorMask support (#451)
* Add support for saturation on some shader instructions, fix ReadTexture alignment

* Add ColorMask support, other tweaks
2018-10-13 23:54:14 -03:00
gdkchan
aa1cd849cf Quads, QuadStrip, const attributes and half-float attributes support (#447)
* Quads, QuadStrip and const attributes support

* Add support for half float attributes and fix texture pitch alignment

* Throw when an unsupported float type is used as const attribute aswell
2018-10-13 01:37:01 +00:00
ReinUsesLisp
2562ca6c3f Fix multiple rendertargets (#427)
* Simplify render target bindings

* Implement multiple viewports

* Pack glViewportIndexed calls into a single glViewportArray

* Use ARB_viewport_array when available

* Cache framebuffer attachments

* Use get accessors in OGLExtension

* Address feedback
2018-09-26 00:55:30 +02:00
ReinUsesLisp
47a62e826f Implement DepthWriteMask and add R16G16 (#425) 2018-09-19 22:02:11 -03:00
ReinUsesLisp
bed13f2022 General improvements for GpuResourceManager (#421)
* General improvements to GpuResourceManager

* Address feedback

* Address feedback
2018-09-19 18:26:49 -03:00
ReinUsesLisp
e04221b293 Fixup SSY (#424) 2018-09-19 17:11:05 -03:00
gdkchan
33e2810ef3
Fix d32s8 format on OGLEnumConverter (#420) 2018-09-18 13:27:12 -03:00
gdkchan
d4187aaa9d
Allow "reinterpretation" of framebuffer/zeta formats (#418)
* (Re)Implement format reinterpretation, other changes

* Implement writeback to guest memory, some refactoring

* More refactoring, implement reinterpretation the old way again

* Clean up

* Some fixes on M2MF (old Dma engine), added partial support for P2MF, fix conditional ssy, add Z24S8 zeta format, other fixes

* nit: Formatting

* Address PR feedback
2018-09-18 01:30:35 -03:00
HorrorTroll
8a78a703f2 Implement B5G6R5Unorm, BGR5A1Unorm, RGBA32Uint, R16Unorm and Z16 texture format. Fix BC6H_UF16 from Unorm to Sfloat (#417)
* Implement B5G6R5Unorm & BGR5A1Unorm

* Implement RGBA32Uint

* Implement R16Unorm & Z16 texture format

* Fix BC6H_UF16 from Unorm to Sfloat
2018-09-17 02:24:55 +02:00
ReinUsesLisp
e5917f8968 Fixup image error message and add G8R8 to size query (#410) 2018-09-11 12:48:13 -03:00
ReinUsesLisp
ce1d5be212 Move GPU emulation from Ryujinx.HLE to Ryujinx.Graphics and misc changes (#402)
* Move GPU LLE emulation from HLE to Graphics

* Graphics: Move Gal/Texture to Texture

* Remove Engines/ directory and namespace

* Use tables for image formats

* Abstract OpCode decoding

* Simplify image table

* Do not leak Read* symbols in TextureReader

* Fixups

* Rename IGalFrameBuffer -> IGalRenderTarget

* Remove MaxBpp hardcoded value

* Change yet again texture data and add G8R8 flipping

* Rename GalFrameBufferFormat to GalSurfaceFormat

* Unident EnsureSetup in ImageHandler

* Add IsCompressed

* Address some feedback
2018-09-08 14:51:50 -03:00
gdkchan
b549daed03
Fix exception throw message for unimpl texture formats (#398) 2018-09-01 18:51:51 -03:00
ReinUsesLisp
e72fd3f7a7 Shaders: Handle Ipa PASS argument as needed in Fragment Shaders (#392) 2018-09-01 18:44:19 -03:00
HorrorTroll
bf28d8f1aa Add BGRA8Unorm, BGRA8Srgb, ZF32_X24S8 texture format (#377)
* Add BGRA8Unorm, BGRA8Srgb, ZF32_X24S8 texture format

* Add BGRA8Unorm, BGRA8Srgb, ZF32_X24S8 texture format

* Revert "Add BGRA8Unorm, BGRA8Srgb, ZF32_X24S8 texture format"

This reverts commit aea5c9db3a.

* Conflicts fix

* Wrong fix

* E

* e
2018-09-01 18:25:49 -03:00
ReinUsesLisp
42dc925c3d Implement SSY/SYNC shader instructions (#382)
* Use a program counter to control shaders' flow

* Cleanup

* Implement SSY/SYNC

* Address feedback

* Fixup commentary

* Fixup Ssy instruction
2018-08-31 13:14:04 -03:00
ReinUsesLisp
7cb6fd8090 Fixup HasColor for depth/stencil values (#379) 2018-08-30 13:14:45 -03:00
HorrorTroll
27023e7afc Add R32_G32 texture format (#383) 2018-08-27 11:18:21 -03:00
ReinUsesLisp
43c4e7c78d Use mirrored texture wraps when available (#361) 2018-08-25 16:39:08 -03:00
ReinUsesLisp
a42ab2e40c Implement vertex instancing (#381) 2018-08-25 01:16:58 -03:00