Pretty much self explanatory, own implementation based on LSFG-VK available with some additional reverse engineer work for the actual use of shaders inside Lossless.dll, we don't store/ bundle any tools that bypass security and self ownership: to use this feature is required to own your own copy of Lossless Scaling and I encourage to support developers behind this work.
Special thanks:
1.- THS - Original developer behind Lossless Scaling
2.- PancakeTAS - Developer behind LSFG-VK
3.- The Big Smolio The Sleeper (@gidoly)
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4263
This insufferable work tries to cover some holes on previous threading implementation from yuzu's team, starting with Windows and Linux reordering of priorities (NICE), reworks previous Android's threading and cpu affinity with adpf, adjust emulated clocks/gpu for better "accuracy" with their work, bumps android minSDK for all flavors, legacy will now work with AP 29 to cover A10 - A12, standard will reach A13 as base and finally the optimized build will come with API 35, mostly targeted on devices with A15 support and newer, NDK and AGP wasn't upgraded yet. The performance cost efficiency have been improved based on device power configuration; preventing overheating if certain devices tended to fall into NICE0 (not allocated threads priority, all task ran with higher priority, 11 tasks running within the limited 2 - 7 threads available on the most common configuration 1x3x4 or 1x4x3).
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4254
Adds a VK_EXT_descriptor_buffer, VK_KHR_device_buffer_address path alongside push descriptors and descriptor sets. The ring is chunked and sized from the device limits, so drivers with a small sampler range (QCOM reports 128 KiB) get more, chunks per frame instead of losing the feature. A pipeline uses one path or the other, never both, since some drivers do not support bufferless push descriptors. The chunk binding is only re-emitted when it changes, and a draw that repeats the previous payload reuses its allocation instead of rewriting it. This whole change will help to reduce the constant use of descriptors on the slower path, improves the "slow implementation" on QCOM drivers (A7xx and older) for push descriptors, improves latency when compiling pipelines. The implementation on indexing descriptors was also refined.
_Special Thanks_
1.- Meowly The Bindless Smol (@Gidoly)
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4251
This is an adjustment made due to problems caused by my ASTC changes some pr's earlier; added sanity to decoded blocks, removes the inconditional storage view for ASTC formats (verified against MaxwellToVk logic), added support to VK_EXT_astc_decode_mode which will ensure a better way to resolve RGFA16/RGFA32 on tilers, simplified ASTC HDR decoding into LDR valid formats.
This fixes some issues with SteamDeck, Fire Emblem: Engage washed colors and other problems related to ASTC.
_Special Thanks_
1.- Meowly The Smol (@Gidoly)
2.- @simply0001 for the actual idea on the redundant storage views on the transcoded ASTC path.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4250
Contains the minimal set of functionalities to allow tico installer succeed, and tico work normally EXCEPT for game launching (which me or someone else will investigate later)
First three commits are from PR 4012.
The other six, kinda dizzy to explain each one. All were implemented based on tico source, switchbrew and libnx.
Hopefully the commit messages will do.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4086
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: Maufeat <sahyno1996@gmail.com>
The approach on this PR moves around refactoring/ updating residual work; main issue was the whole logic behind msaa upload/download images with the constant image copy per pass required; which was gated via shaderStorageImageMultisample, which is not available on all the platforms supported including Android, making them to rely into a more heavier approach to convert msaa image copy into a non msaa image copy, not only losing precision in the conversion, but also creating bugs around the resolve of the image itself, if the supported path had to copy 2 times the same image to be actually presentable, when the fallback was enabled 3 copies were performed, not only decreasing the performance but also increasing race conditions due to the certain capabilities processed in this chain (storage_views). In order to resolve them more naturally, the constant copies were removed from the now "common" path along establishing a proper color resolve for Android and actually making them more direct.
Yet this opened a new pack of outdated handling on our current backend; since we now pass msaa images more directly (whenever a game request a msaa type of image/ which seems to be not common + each fragment resolves images on how it's actually required, adding missing depth/stencil), how we often load and store them becomes a critical issue, since our current configuration renderpass/framebuffer forces to always load and load clear whenever a new attachment it's passing through the renderpass, making this really heavier on memory bandwidth limited devices as SteamDeck and others that doesn't rely on the inmediate rendering mode (Tilers vs iGPU); to prevent over allocating data, I had to re-structure part of the current renderpass and framebuffer to actually reuse part of the data/ attachments called, with flags as DON'T_CARE, simplifying partially on how we actually call attachments including their barriers and conditions (occlusion, etc), the DON'T_CARE it's specifically useful to not only prevent calling new loads (which is a flush and heavier on tiler devices) and not only reusing the actual data store, which works on the future time, this saves memory bandwidth and gives space to actually avoid the constant pressure on SYSMEM/GMEM shader resolver path.
This maintenance also improved the current removal features logic on QCOM drivers, closing gaps between implementations on descriptors (VK_EXT_indexing_descriptor), adjusting the whole WMEL to avoid being so stricter if certain capabilities are missing to actually use this feature, leading to test and play more on how QCOM drivers are actually behaving; meanwhile the main objective is to open path for a robust and bigger implementation in the future, the changes made here also contains small bug fixes on SPIR-V logic, adding missing cases to resolve shaders (FP64 -> FP16), adding more robustness on how narrow features are actually being used and how to wired them whenever they're not available, meanwhile there are still some issues in regards on how Tegra vs Adreno can preserve unorms at FP32 and FP16, leaving better understanding on base for what VK_EXT_shader_float_controls is actually doing most of the times (flushes NaN's and Denorms equally).
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4212
This simplifies the GPU accuracy setting by removing the intermediate Balanced mode and setting High as the default on desktop platforms.
It introduces a dedicated setting for GPU fence behavior, allowing the synchronization policy to be configured independently of GPU accuracy.
The Vulkan buffer cache now tracks GPU recording timeline ticks and waits only when necessary, reducing unnecessary synchronization while maintaining correctness for hard-to-trace graphical bugs.
GPU buffer readback has also been refined to synchronize only the affected upload regions when needed, and default DMA behavior has been updated to align with the new GPU accuracy model.
### TL;DR
The fix for particles freezing and unfreezing in mid-air in `Super Mario Odyssey` has been improved, resulting in less of a performance hit.
This game requires the new `Enable GPU Buffer Readback` option to be enabled to fix this issue.
The vertex explosions that occurred in `Super Mario Bros. Wonder`, especially in World 4, have been completely eliminated. You can now enjoy a smooth experience without graphical glitches exploding across the screen.
This game requires the new `GPU Fence Behavior` option to be set to `Strict` to fully fix this issue.
The flickering issue inside certain Shrines in `The Legend of Zelda: Tears of the Kingdom` has also been fixed.
For now, this game requires the new `GPU Fence Behavior` option to be set to `Accurate` to fully fix this issue.
These options are intended to fix graphical bugs in games that require better synchronization behavior between CPU and GPU, so other games may be affected as well.
Co-authored-by: xbzk <xbzk@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4182
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
This is the first step into the conversion on full GPU video decoding for Android; currently due to the VIC structure, I couldn't set it on surface mode to prevent the latency when sending video decoded (which is processed by GPU now), because currently we convert YUV420 into the Nvidia's format each frame constantly on CPU, aside the requirements from other extensions to work + VIC rewrite, which is a work currently not planned to happen on this PR; the actual configuration for video decoding is ByteBuffer, using GPU to decode NVDEC data (all codecs supported, h264, vp8 and vp9) and send it to CPU for display purposes, which is more faster than relying purely on CPU for any drawing task, saving devices resources/ heating, the downside on this will be the slight latency when a new frame is displayed, which is gonna be a black frame for less than a second, nothing major to harm the experience rather than actually trying our best to take advantage on hardware accelerated.
Aside this, I also added a bunch of minor Vulkan fixes to grant drivers less thinkering when receiving spir-v instructions, meaning this has new bans for extensions on QCOM (following reported issues on other projects working around Adreno driver behavior), this more than providing performance aims to enhance the stability on the driver, performance it's gonna likely to be hit based on the UBO's (StorageBufferAccess) operations and SSBO (uniformStorageBufferAccess), there was an already existing path for the emulation which forces to wide them into 32bit packed operations. I also included some smaller changes/ bugs + VUID's fixes from earlier changes on my work.
Special Thanks:
-> Mr. Smoly Gidolard (@gidoly)
Sources:
1.- https://github.com/microsoft/DirectXShaderCompiler/issues/2842
2.- https://github.com/mlc-ai/web-llm/issues/836
3.- https://github.com/ggml-org/llama.cpp/issues/5186
4.- https://github.com/encounter/aurora/pull/202
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4191
Based on the register of the ASTC implementation from Ameerj (yuzu), currently our conversion from ASTC on Desktop tends to duplicates the unpacking of ASTC converted textures, leading to bad performance; since there are some graphical effects that actually access on the band of 32bits (RGAF32) than the common 16bits (RGAF16) from our ASTC path decoding method, which commonly uses ARGB8 (degradation -> LDR) transformation to reduce the memory cost per conversion; meanwhile it could actually provide initially better performance; the engine tended to get stalled when this writted stage effect got skipped; mostly because CPU tends to look for an unimplemented texture/ format query (from which doesn't exist on our MaxwellToVK); unlike what's commonly believed HDR was a format to write enviroment effects even before their appearance on video output (since Switch doesn't have HDR, only SDR on certain engines).
The test result on this PR showed games like Astral Chain and other heavy reliant on ASTC decoding improved their performance and stability by around 15%; meanwhile could it be tied to a hardware gain, it's clearly a gain for direct access on what's required on most games. This PR also contains few Vulkan bug fix I encountered along this implementation.
Special Thanks:
-> Mr. Smoly Gidolard (@gidoly)
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4190
Following the philosophy on the previous Vulkan maintenance PR's, this one is reviewing the video_core to resolve VUID's related to the missing handling/ safechecks (UBO's case on missing StorageBufferAccess 8/16 bits on QCOM drivers, resulting on bad compute pipeline compiled), wiring topology representative to EDS2 configuration (following #4117), refactored the sampled image access via new memeber function for type "typeless" which adds a handling on the integer mistmatch (uint to sint, float to uint) + fixing bugs on current texture sampling (int) including the depth/stencil path resolve, maintenance to previous changes on query cache (resolving bugs from #3853), reduced the amount of binding BindVertexBufferEXT2 when a new tick/ frame is presented (with this being a reason for performance reduction on games where vertex polutes the stage, like BOTW/TOTK where grass is draw with vertex); implemented color border swizzle and color write enable for future improvements on EDS3, added initial implementation on Synchronization2 starting a path to ensure an access to VK 1.3 features safely and other smaller fixes along the road.
Special Thanks:
-> Mr. Smolio Gidolard (@gidoly)
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4189
This minor change in src\video_core\renderer_vulkan\pipeline_helper.h allows renderdoc capture on mhr sunbreak.
Maybe it sanitizes some crashes on old vulkan GPUs.
Device log (nvidia kepler gpu vulkan 1.1.117):
[ 17.605262] Render.Vulkan <Info> video_core\vulkan_common\vulkan_device.cpp:1041:GetSuitability: Device doesn't support feature nullDescriptor
VUID-VkWriteDescriptorSet-descriptorType-02997
For sampled/combined/storage image descriptors, Khronos says imageView must be valid or VK_NULL_HANDLE
(https://docs.vulkan.org/refpages/latest/refpages/source/VkPhysicalDeviceRobustness2FeaturesKHR.html)
but then it says: if nullDescriptor is not enabled, imageView must not be VK_NULL_HANDLE.
(https://docs.vulkan.org/spec/latest/chapters/descriptorsets.html)
nullDescriptor is exactly the feature that allows descriptors to be written with null handles.
The fix relies on conditionally replacing the null sampled view with Eden’s dummy null-image-view object (ImageView constructor checks HasNullDescriptor which checks features.robustness2.nullDescriptor. cheap checks, and only for null handles, so no notable cost).
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4056
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
FIXES A REGRESSION FROM 4012.
TryFindSize's Mesa fallback matched the unconditional @PT EXIT (0xE30000000007000F) that nv50_ir emits at program end.
That exact word is also emitted by NVN mid-program for unconditional early-outs, so on retail titles (e.g. Super Mario 3D World) the scan truncated at the first early EXIT and dropped the rest of the shader, making textures vanish.
Mesa's terminal EXIT and NVN's mid-program EXIT are bit-identical, so no mask can separate them. Gate the heuristic on !is_proprietary_driver:
NVN binaries (bindless cbuf slot 2) keep self-branch-only sizing, while nouveau/Mesa homebrew (single terminal EXIT, no self-branch trailer) still uses the EXIT terminator. nv50_ir emits exactly one EXIT (early returns lower to BRA-to-exit, discard to DISCARD), so a single terminal match is correct for that path.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4151
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
EDS3 states were dynamic just because the driver supports them. But in EDS0-EDS2, we does not actually emit the EDS3 dynamic commands for them. So the pipeline builder skipped baking some fixed graphics state into the Vulkan pipeline, but the runtime also did not set that state dynamically. That leaves wrong Vulkan state during rendering causing rendering glitches in XC2.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4117
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Added an option to control the GPU buffer readback, as it causes issues if the hardware cannot keep up with the additional workload.
Some games require this to render certain effects properly.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4126
Inline index draws arrive as a batch but were processed one word at a time, so each index ran through the full per-method path and got appended a byte at a time.
Handle the whole batch in one pass instead, updating state once and appending all the indices together. Replay shadow control still goes word by word since it has to reload each one.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4083
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
5af7771f83-Bugfix: Made gpu_log_level global-only (was per-game switchable). Fixed Android non-determinism where a per-game profile silently overrode the global to Off and trapped GPULogger::Initialize() in a dead state, making shader dumps fail invisibly. Android per-game UI now hides the whole GPU logging block; Qt UI is untouched (global-only anyway).
bf4aabe8ab-Refactor/Cleanup: Removed gpu_logging_enabled master toggle as redundant with gpu_log_level == Off. Introduced GPU::Logging::IsActive() helper, replaced 14 call sites across vk_*.cpp. Refactored LogShaderCompilation() to be text-only and extracted SPIR-V dumping into a standalone GPU::Logging::DumpSpirvShader() free function. No singleton dependency, gated only by gpu_log_shader_dumps. Now gpu_log_level and gpu_log_shader_dumps are fully orthogonal. Cleaned up Android (BooleanSetting, SettingsItem, presenter, 7 locale string files).
865a1c5027-Refactor: Renamed dump_shaders → dump_guest_shaders to disambiguate from gpu_log_shader_dumps. Updated Qt label to "Dump Guest (Maxwell) Shaders" and rewrote the tooltip to mention .ash, the DumpDir/shaders/ location, and nvdisasm.
7cab456fdf-Feature: Added Qt UI control for GPU log level in the Logging session. Added gpu_log_shader_dumps checkbox to the Graphics column right below dump_guest_shaders.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4018
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
GetMacroAddress only reads a couple of indices per macro, but ProcessMacro was
then building a full std::vector GPUVAddr> with one push_back per parameter word
every submission. macro_segments already holds base, count per chunk, so
GetMacroAddress can just walk it instead, drops the per-word loop, a
.clear(), and a vector member.
Also makes it so it returns on the first match in the macro dispatch instead of running every
std::get_if check.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4067
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Prevent the CPU from getting too far ahead of the GPU by limiting pending flushes.
This fixes graphical issues in games that rely on better CPU and GPU synchronization in the Scheduler, such as the flickering bug inside shrines in The Legend of Zelda: Tears of the Kingdom.
This is enabled when the GPU is in Accurate mode.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4085
Reviewed-by: crueter <crueter@eden-emu.dev>
Reviewed-by: Shinmegumi <shinmegumi@eden-emu.dev>
The GL shader profile hardcoded max_user_clip_distances to 8. Query the
device limit like the Vulkan path already does (it reads
maxClipDistances), so we use what the host actually reports.
Clamp to Maxwell's NumClipDistances (8) since the guest never produces
more than that and the SPIR-V output array is sized for at most 8. So a
host reporting fewer than 8 is respected, and one reporting more can't
overrun anything.
Fixes#3910https://git.eden-emu.dev/eden-emu/eden/issues/3910
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4095
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
two main issues:
shader_environment.cpp
Support Mesa nv50_ir-compiled homebrew fragment shaders
Mesa nv50_ir (used by NZP-Switch via libdrm_nouveau) emits Maxwell SASS that differs from NVN in two ways our shader pipeline didn't account for, causing "infinite" loops during shader scanning.
How? TryFindSize() previously detected end-of-shader only via NVN's "BRA $-1" self-branch trailer (SELF_BRANCH_A/B). Mesa-compiled shaders end with an unconditional @PT EXIT and emit no trailer, so scanning ran past the shader and looped to MAXIMUM_SIZE.
Fix: Added a secondary match against @PT EXIT T (opcode 0xE30, predicate PT, flow T) as a fallback terminator. The mask is estrictive enough to reject predicated EXITs, conditional-flow EXITs, and sched control words, so well-formed NVN shaders see no ehavior change (their single @PT EXIT immediately precedes BRA $-1, and both detections return the same size).
load_store_attribute.cpp + attribute.h
IPA's is_perspective check only applied the ×position_w correction for IR::IsGeneric() attributes, and used a per-component SPH lookup. Two failure modes followed:
1. Mesa fragment shaders read varyings via legacy attribute slots
(ColorFrontDiffuse, FixedFncTexture, FogCoordinate) which are
remapped to generic varyings later by ConvertLegacyToGeneric. At
IPA-translation time they're still legacy, so IsGeneric() was
false and the inject was skipped — but the resulting GLSL/SPIR-V
varying was Smooth (perspective-correct), and the SASS still
issued its manual perspective dance via MUFU.RCP(ATTR_W)+FMUL,
compounding an extra ×clip_w factor. NZP fog (eye_z distance
varying) ended up as eye_z⁴×density² instead of eye_z²×density²,
saturating fog to white everywhere.
2. Even for legacy attrs that should inject, Mesa stores them in two formats selected by the IPA's interpolation_mode field:
Pass/Multiply/Constant → attr/w (perspective); inject ×position_w
Sc (ScreenLinear) → raw attr; do NOT inject
Unconditionally injecting for Sc dims vertex colors by 1/clip_w because there's no SASS multiplier to round-trip it back.
Fix: extend the IPA correction path to also fire on IsLegacyAttribute(), gated on interpolation_mode != Sc. While here, the generic-path check is widened from per-component to vector-level (first non-Unused PixelImap, with all-Unused fallthrough treated as Perspective) so it matches the GLSL interpolation qualifier picked by CollectInterpolationInfo — this fixes the same all-zeroed-imap case Mesa exposes for explicit generics. IsLegacyAttribute moved from translate_program.cpp's anonymous namespace into attribute.h next to IsGeneric so IPA can reach it.
Tested on NaziZombies:Portable (Switch homebrew): fog now matches real
hardware, vertex-colored geometry (hand, gun, decals) renders correctly.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4012
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
lizzie
CamilleLaVey
crueter
simply0001
PavelBARABANOV
xbzk
MaranBr
MaranBr
John
wildcard
BoiledElectricity