Commit graph

1575 commits

Author SHA1 Message Date
Wunk
1d4d421097
shader_jit_a64: Optimize MOVA dest-enable (#7122)
Rather than branching the 3 cases of dest-enablement, just emit a single
move-and-sign-extend instruction for each case.

From this review:
https://github.com/citra-emu/citra/pull/7002#discussion_r1381560584
2023-11-07 11:46:40 -08:00
Wunk
8fe147b8f9
video_core: Use binary memory-literals for memory-sizes (#7127)
Replaces `... * 1024 * 1024` with `_MiB`/`_GiB` literals.
2023-11-06 23:38:54 +02:00
GPUCode
1f6393e7d5
video_core: Refactor GLSL fragment emitter (#7093)
* video_core: Refactor GLSL fragment emitter

* shader: Add back custom normal maps
2023-11-06 12:26:28 -08:00
Wunk
e13735b624
video_core: Implement an arm64 shader-jit backend (#7002)
* externals: Add oaksim submodule

Used for emitting ARM64 assembly

* common: Implement aarch64 ABI

Utilize oaknut to implement a stack frame.

* tests: Allow shader-jit tests for x64 and a64

Run the shader-jit tests for both x86_64 and arm64 targets

* video_core: Initialize arm64 shader-jit backend

Passes all current unit tests!

* shader_jit_a64: protect/unprotect memory when jit-ing

Required on MacOS. Memory needs to be fully unprotected and then
re-protected when writing or there will be memory access errors on
MacOS.

* shader_jit_a64: Fix ARM64-Imm overflow

These conditionals were throwing exceptions since the immediate values
were overflowing the available space in the `EOR` instructions. Instead
they are generated from `MOV` and then `EOR`-ed after.

* shader_jit_a64: Fix Geometry shader conditional

* shader_jit_a64: Replace `ADRL` with `MOVP2R`

Fixes some immediate-generation exceptions.

* common/aarch64: Fix CallFarFunction

* shader_jit_a64: Optimize `SantitizedMul`

Co-authored-by: merryhime <merryhime@users.noreply.github.com>

* shader_jit_a64: Fix address register offset behavior

Based on https://github.com/citra-emu/citra/pull/6942
Passes unit tests.

* shader_jit_a64: Fix `RET` address offset

A64 stack is 16-byte aligned rather than 8. So a direct port of the x64
code won't work. Fixes weird branches into invalid memory for any
shaders with subroutines.

* shader_jit_a64: Increase max program size

Tuned for A64 program size.

* shader_jit_a64: Use `UBFX` for extracting loop-state

Co-authored-by: JosJuice <JosJuice@users.noreply.github.com>

* shader_jit_a64: Optimize `SUB+CMP` to `SUBS`

Co-authored-by: JosJuice <JosJuice@users.noreply.github.com>

* shader_jit_a64: Optimize `CMP+B` to `CBNZ`

Co-authored-by: JosJuice <JosJuice@users.noreply.github.com>

* shader_jit_a64: Use `FMOV` for `ONE` vector

Co-authored-by: JosJuice <JosJuice@users.noreply.github.com>

* shader_jit_a64: Remove x86-specific documentation

* shader_jit_a64: Use `UBFX` to extract exponent

Co-authored-by: JosJuice <JosJuice@users.noreply.github.com>

* shader_jit_a64: Remove redundant MIN/MAX `SRC2`-NaN check

Special handling only needs to check SRC1 for NaN, not SRC2.
It would work as follows in the four possible cases:

No NaN: No special handling needed.
Only SRC1 is NaN: The special handling is triggered because SRC1 is NaN, and SRC2 is picked.
Only SRC2 is NaN: FMAX automatically picks SRC2 because it always picks the NaN if there is one.
Both SRC1 and SRC2 are NaN: The special handling is triggered because SRC1 is NaN, and SRC2 is picked.

Co-authored-by: JosJuice <JosJuice@users.noreply.github.com>

* shader_jit/tests:: Add catch-stringifier for vec2f/vec3f

* shader_jit/tests: Add Dest Mask unit test

* shader_jit_a64: Fix Dest-Mask `BSL` operand order

Passes the dest-mask unit tests now.

* shader_jit_a64: Use `MOVI` for DestEnable mask

Accelerate certain cases of masking with MOVI as well

Co-authored-by: JosJuice <JosJuice@users.noreply.github.com>

* shader_jit/tests: Add source-swizzle unit test

This is not expansive. Generating all `4^4` cases seems to make Catch2
crash. So I've added some component-masking(non-reordering) tests based
on the Dest-Mask unit-test and some additional ones to test
broadcasts/splats and component re-ordering.

* shader_jit_a64: Fix swizzle index generation

This was still generating `SHUFPS` indices and not the ones that we wanted for the `TBL` instruction. Passes all unit tests now.

* shader_jit/tests: Add `ShaderSetup` constructor to `ShaderTest`

Rather than using the direct output of `CompileShaderSetup` allow a
`ShaderSetup` object to be passed in directly.  This enabled the ability
emit assembly that is not directly supported by nihstro.

* shader_jit/tests: Add `CALL` unit-test

Tests nested `CALL` instructions to eventually reach an `EX2`
instruction.

EX2 is picked in particular since it is implemented as an even deeper
dispatch and ensures subroutines are properly implemented between `CALL`
instructions and implementation-calls.

* shader_jit_a64: Fix nested `BL` subroutines

`lr` was getting writen over by nested calls to `BL`, causing undefined
behavior with mixtures of `CALL`, `EX2`, and `LG2` instructions.

Each usage of `BL` is now protected with a stach push/pop to preserve
and restore teh `lr` register to allow nested subroutines to work
properly.

* shader_jit/tests: Allocate generated tests on heap

Each of these generated shader-test objects were causing the stack to
overflow.  Allocate each of the generated tests on the heap and use
unique_ptr so they only exist within the life-time of the `REQUIRE`
statement.

* shader_jit_a64: Preserve `lr` register from external function calls

`EMIT` makes an external function call, and should be preserving `lr`

* shader_jit/tests: Add `MAD` unit-test

The Inline Asm version requires an upstream fix:
https://github.com/neobrain/nihstro/issues/68

Instead, the program code is manually configured and added.

* shader_jit/tests: Fix uninitialized instructions

These `union`-type instruction-types were uninitialized, causing tests
to indeterminantly fail at times.

* shader_jit_a64: Remove unneeded `MOV`

Residue from the direct-port of x64 code.

* shader_jit_a64: Use `std::array` for `instr_table`

Add some type-safety and const-correctness around this type as well.

* shader_jit_a64: Avoid c-style offset casting

Add some more const-correctness to this function as well.

* video_core: Add arch preprocessor comments

* common/aarch64: Use X16 as the veneer register

https://developer.arm.com/documentation/102374/0101/Procedure-Call-Standard

* shader_jit/tests: Add uniform reading unit-test

Particularly to ensure that addresses are being properly truncated

* common/aarch64: Use `X0` as `ABI_RETURN`

`X8` is used as the indirect return result value in the case that the
result is bigger than 128-bits. Principally `X0` is the general-case
return register though.

* common/aarch64: Add veneer register note

`LR` is generally overwritten by `BLR` anyways, and would also be a safe
veneer to utilize for far-calls.

* shader_jit_a64: Remove unneeded scratch register from `SanitizedMul`

* shader_jit_a64: Fix CALLU condition

Should be `EQ` not `NE`. Fixes the regression on Kid Icarus.
No known regressions anymore!

---------

Co-authored-by: merryhime <merryhime@users.noreply.github.com>
Co-authored-by: JosJuice <JosJuice@users.noreply.github.com>
2023-11-05 21:40:31 +01:00
Wunk
3218af38d0
renderer_vulkan: Add scissor and viewport to dynamic pipeline state (#7114)
Adds the current viewport and scissor to the dynamic pipeline state to
reduce redundant viewport/scissor assignments in the command buffer.
This greatly reduces the amount of API calls to `vkCmdSetViewport` and
`vkCmdSetScissor` by only emitting the API call when the state actually
changes.
2023-11-05 12:26:09 -08:00
Wunk
1cf64ffaef
vk_stream_buf: Allow dedicated allocations (#7103)
* vk_stream_buf: Avoid protected memory heaps

* Add an "Exclude" argument when finding a memory-type that avoids
  `VK_MEMORY_PROPERTY_PROTECTED_BIT` by default

* vk_stream_buf: Utilize dedicated allocations when preferred by driver

`VK_KHR_dedicated_allocation` is part of the core Vulkan 1.1
specification and should be utilized when `prefersDedicatedAllocation`
is set.
2023-11-05 12:25:59 -08:00
Wunk
b10f3d96f5
command_processor: Fix out-of-bounds float-uniform access (#7111)
Addresses:
https://github.com/citra-emu/citra/issues/6696
https://github.com/citra-emu/citra/issues/6871
2023-11-03 03:35:52 -07:00
Wunk
ac9d72a95c
vk_texture_runtime: Fix debug scope label lambda-capture (#7102)
`DebugScope` was capturing a `string_view` in a lambda which is only
valid during the scope of this ctor. When the lambda gets invoked at a
later time, it will read undefined garbage. The lambda needs to make a
deep copy of this `string_view` into a `string` so that it is valid by
the time the scheduler invokes this lambda.
2023-11-01 21:30:54 +01:00
GPUCode
36146459f8
renderer_vulkan: Fix screenshots under NVIDIA vulkan (#7082) 2023-10-22 22:53:14 +03:00
GPUCode
ef43776c7b
shader: Fix address register offset behavior in x64 Jit (#6942)
* shader: Fix address register offset behavior in x64 Jit

* shader: Remove redundant jump

* tests: Add address register tests

* shader: Remove additional pre-multiplications by 16

* tests: Add catch-stringifier for vec4f

* tests: Format
2023-10-18 19:41:36 +03:00
Steveice10
4c59443ed2
common: Add more robust ZSTD handling. (#7071) 2023-10-15 14:08:29 -07:00
GPUCode
40ba5226c6
vk_instance: Perform vulkan logging as early as possible (#7058) 2023-10-11 15:11:43 -07:00
Steveice10
4220f69c06
renderer/opengl: Deduce GLES from actual driver in use. (#7056) 2023-10-10 12:52:47 -07:00
SachinVin
053b3618dc
Android: gl_shader_decompiler.cpp: get rid of the not-so-accurate multiplication (#7011) 2023-10-05 10:28:26 -07:00
Steveice10
50f22d1f59
video_core: Abstract shader generators. (#6990)
* video_core: Abstract shader generators.

* shader: Extract common generator structures and move generators to specific namespaces.

* shader: Minor fixes and clean-up.
2023-09-30 02:06:06 -07:00
Vitor K
a35f8cbb78
fix include/namespace related compilation errors (#7019)
a user on discord reported compilation errors when trying to compile
on Linux with GCC 13 and Clang 16.
2023-09-28 18:36:50 +05:30
Vitor K
6cfb8e02a8
clang format (#7017) 2023-09-27 13:42:19 -03:00
Steveice10
327a00b4d5
externals: Update fmt. (#7015) 2023-09-27 03:29:13 -07:00
GPUCode
542209c993
video_core: Initialize unrefered attributes to 1.f (#6966) 2023-09-16 14:46:44 -07:00
GPUCode
d2d37411bc
vk_platform: Check if vkEnumerateInstanceVersion exists before calling it (#6982)
Co-authored-by: Vitor K <vitor-kiguchi@hotmail.com>
2023-09-15 16:00:10 -07:00
GPUCode
30fcdc5474
renderer_vulkan: Misc fixes (#6974)
* vk_platform: Check if library was loaded

* pica_to_vk: Dont crash on unknow blend equation
2023-09-15 00:21:12 +03:00
GPUCode
dfa2fd0e0d
Add vulkan backend (#6512)
* code: Prepare frontend for vulkan support

* citra_qt: Add vulkan options to the GUI

* vk_instance: Collect tooling info

* renderer_vulkan: Add vulkan backend

* qt: Fix fullscreen and resize issues on macOS. (#47)

* qt: Fix bugged macOS full screen transition.

* renderer/vulkan: Fix swapchain recreation destroying in-use semaphore.

* renderer/vulkan: Make gl_Position invariant. (#48)

This fixes an issue with black artifacts in Pokemon games on Apple GPUs.
If the vertex calculations differ slightly between render passes, it can
cause parts of model faces to fail depth test.

* vk_renderpass_cache: Bump pixel format count

* android: Custom driver code

* vk_instance: Set moltenvk configuration

* rasterizer_cache: Proper surface unregister

* citra_qt: Fix invalid characters

* vk_rasterizer: Correct special unbind

* android: Allow async presentation toggle

* vk_graphics_pipeline: Fix async shader compilation

* We were actually waiting for the pipelines regardless of the setting, oops

* vk_rasterizer: More robust attribute loading

* android: Move PollEvents to OpenGL window

* Vulkan does not need this and it causes problems

* vk_instance: Enable robust buffer access

* Improves stability on mali devices

* vk_renderpass_cache: Bring back renderpass flushing

* externals: Update vulkan-headers

* gl_rasterizer: Separable shaders for everyone

* vk_blit_helper: Corect depth to color convertion

* renderer_vulkan: Implement reinterpretation with copy

* Allows reinterpreteration with simply copy on AMD

* vk_graphics_pipeline: Only fast compile if no shaders are pending

* With this shaders weren't being compiled in parallel

* vk_swapchain: Ensure vsync doesn't lock framerate

* vk_present_window: Match guest swapchain size to vulkan image count

* Less latency and fixes crashes that were caused by images being deleted before free

* vk_instance: Blacklist VK_EXT_pipeline_creation_cache_control with nvidia gpus

* Resolves crashes when async shader compilation is enabled

* vk_rasterizer: Bump async threshold to 6

* Many games have fullscreen quads with 6 vertices. Fixes pokemon textures missing with async shaders

* android: More robust surface recreation

* renderer_vulkan: Fix dynamic state being lost

* vk_pipeline_cache: Skip cache save when no pipeline cache exists

* This is the cache when loading a save state

* sdl: Fix surface initialization on macOS. (#49)

* sdl: Fix surface initialization on macOS.

* sdl: Fix render window events not being handled under Vulkan.

* renderer/vulkan: Fix binding/unbinding of shadow rendering buffer.

* vk_stream_buffer: Respect non coherent access alignment

* Required by nvidia GPUs on MacOS

* renderer/vulkan: Support VK_EXT_fragment_shader_interlock for shadow rendering. (#51)

* renderer_vulkan: Port some recent shader fixes

* vk_pipeline_cache: Improve shadow detection

* vk_swapchain: Add missing check

* renderer_vulkan: Fix hybrid screen

* Revert "gl_rasterizer: Separable shaders for everyone"

Causes crashes on mali GPUs, will need separate PR

This reverts commit d22d556d30.

* renderer_vulkan: Fix flipped screenshot

---------

Co-authored-by: Steveice10 <1269164+Steveice10@users.noreply.github.com>
2023-09-13 01:28:50 +03:00
GPUCode
9be4fd8b91
gl_rasterizer: Bind null surface when unit is disabled (#6967) 2023-09-11 19:34:41 +03:00
Vitor K
bbb47cd753
gles: fix crashing on mali gpus (#6956)
* gles: reorder framebuffer fetch extension directives

* Address review comments
2023-09-11 00:30:30 +03:00
GPUCode
0b0d3a4ac3
gpu: Correct display transfer output with vertical flip+crop lines (#6952) 2023-09-11 00:30:06 +03:00
GPUCode
6aa31d6ec2
video_core: Use epsilons for clip planes (#6945)
* video_core: Use epsilons for clip planes

* video_core: Add comments
2023-09-11 00:29:55 +03:00
GPUCode
928f352c94
renderer_gl: Use explicit bindings (#6940)
* renderer_gl: Use explicit bindings

* gl_state: Match shadow order with vulkan
2023-09-07 21:42:22 +03:00
GPUCode
cab0ad50f0
gl_shader_decompiler: Fix min/max NaN edge case (#6935) 2023-08-31 13:37:53 -07:00
Steveice10
5ad58e0605
shader: Fix shadow_texture_orthographic not being set correctly. (#6937) 2023-08-31 13:37:39 -07:00
GPUCode
1159e4d928
video_core: Take factors into account with min/max blending functions (#6925)
* sw_framebuffer: Take factors into account for min/max blending

* renderer_gl: Take factors into account for min/max blending

* Address review comments

* gl_shader_gen: Fix frambuffer fetch on qcom and mali

* renderer_opengl: Add fallback path for mesa

* gl_shader_gen: Avoid emitting blend emulation if minmax_factor is present
2023-08-30 21:26:28 +03:00
Steveice10
81a5e2355a
shader: Fix address register offset behavior in GLSL. (#6920) 2023-08-28 13:23:59 -07:00
GPUCode
d1f600601d
renderer_software: Multi-thread processing (#6698)
* renderer_software: Multi-thread processing

* Doubles the performance in most cases

* renderer_software: Move memory access out of the raster loop

* Profiling shows this has a significant impact
2023-08-28 01:09:23 -07:00
GPUCode
8b218e1b7d
tev: Use primary color for previous source in first stage (#6921) 2023-08-27 23:59:32 -07:00
Steveice10
df57012c50
video_core: Correct shader interpeter address register handling. (#6910) 2023-08-27 01:49:31 +03:00
GPUCode
61cf550d0c
video_core: Avoid setting alpha tev with Dot3_RGBA8 color op (#6907)
* Further reduces unnecessary shader regenerations
2023-08-27 01:48:45 +03:00
Steveice10
f2e0748a22
build: Enable link time optimization in release builds. (#6887) 2023-08-26 11:15:13 -07:00
GPUCode
a94297922b
shader_interpreter: Fix control flow edge cases and implement break/breakc (#6844) 2023-08-21 18:12:46 -07:00
m4xw
e129f07047
[ShaderGen] Avoid setting lighting states if components aren't enabled (#6895)
* [ShaderGen] Avoid setting lighting states if components aren't enabled

* [Squash] clang-format
2023-08-20 21:12:38 -07:00
Steveice10
6ddf4b241f
renderer/vulkan: Emulate custom border colors in shaders when unavailable. (#6878) 2023-08-17 22:22:25 +02:00
Vitor K
f3d92dd3b8
Fix post processing shaders crashing on Intel GPUs (#6890) 2023-08-17 22:17:24 +02:00
Vitor K
1a6860f35c
opengl: disable ClearTexture on Intel drivers (#6879) 2023-08-17 22:16:51 +02:00
GPUCode
bc0bf4d3d2
(gl/vk)_shader_gen: Use floor instead of int cast (#6885) 2023-08-17 22:16:28 +02:00
Steveice10
6a1fd38063
build: Use new glslang tool name. (#6889) 2023-08-13 15:45:22 -07:00
GPUCode
f76165d848
frame_dumper: Use jthread (#6855) 2023-08-07 18:21:28 -07:00
GPUCode
ac78b74c45
gl_texture_runtime: Implement ARB_clear_texture (#6850) 2023-08-08 01:10:28 +03:00
GPUCode
a478bedb12
rasterizer_cache: Bring back custom format parameter (#6873) 2023-08-08 01:10:11 +03:00
GPUCode
5b52849f90
rasterizer_cache: Allow custom recycle (#6851)
* surface_params: Allow custom surface recycling

* rasterizer_cache: Cleanup
2023-08-06 22:59:54 +03:00
GPUCode
ebac2e4978
custom_tex_manager: Allow old hash in more cases (#6843) 2023-08-03 00:51:10 +03:00
GPUCode
6f7612f73d
Res cache fixes (#6838)
* rasterizer_cache: Dont consider res_scale during recycle

* rasterizer_cache: Switch to plain erase loop

* rasterizer_cache: Fix crash due to memory corruption
2023-08-02 21:38:45 +03:00
GPUCode
88ea66053e
Miscallenious fixes to gl backend and qt frontend (#6834)
* renderer_gl: Make rasterizer normal class member

* It doesn't need to be heap allocated anymore

* gl_rasterizer: Remove default_texture

* It's unused

* gl_rasterizer: General cleanup

* gl_rasterizer: Lower case lambdas

* Match style with review comments from vulkan backend

* rasterizer_cache: Prevent memory leak

* Since the switch from shared_ptr these surfaces were no longer being destroyed properly. Use our garbage collector for that purpose to destroy it safely for both backends

* rasterizer_cache: Make temp copy of old surface

* The custom surface would override the memory region of the old region resulting in garbage data, this ensures the custom surface is constructed correctly

* citra_qt: Manually create dialog tabs

* Allows for custom constructors which is very useful. While at it, global state is now eliminated from configuration

* citra_qt: Eliminate global system usage

* core: Remove global system usage in memory and HIO

* citra_qt: Use qOverload

* tests: Run clang format

* gl_texture_runtime: Fix surface scaling
2023-08-02 01:40:39 +03:00