revert [shader_recompiler] handle dynamic texture descriptor strides (#3898)
this fixes dynamic texture descriptors that are not laid out as simple 8-byte entries
tested on steam deck/amd
---------------
Performance and graphical issues started from this point and forward, testing is required.
notes
- DynamicDescriptorSizeShift called twice because i moved it away from the struct but doing it this way keeps the patch just in this single file than adding a new derived field in the shared struct (i also think its just a cheap recomputation anyways)
- removed cbuf scanning because i figured out how to do a bounds check statically
credits:
- Mythrax <mythrax@mytrax-rs.org> (identified the 1024 descriptor cap fix in #3897)
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3898
Reviewed-by: crueter <crueter@eden-emu.dev>
Amaterasu from Discord asked about why nvnWindowGetNumActiveTextures returns 0 - after researching the console output it was asking for `DefaultDataSpace` (enum val 12) after some tests with the mod and debugging the console, this is my one-liner solution.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3956
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
- testriswitch submits buffers with a fence id of -1, just skip them instead of trying to process them?
- apm:u, which is removed, but hey, backwards compat never hurted
- another instance of shared_memory crashing NPad
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3983
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Anniversary is over!
May the next one be as silly :3
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3990
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
This fixes a rare visual corruption that can occur under specific conditions depending on the hardware used.
This bug is known to affect the loading screens in The Legend of Zelda: Tears of the Kingdom.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3986
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
This fixes a deadlock issue that prevents any game using OpenGL + GPU Accurate from booting when Sync Memory Operations is enabled.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3987
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
tldr: we (me and maran) found the missing piece to make delay_fence and should_flush coop and cover all remaining flicker issues tested so far.
the lil change is in src/video_core/fence_manager.h
the rest is toggle stuff.
END
(credits rolls up)
post credits scene for the toggle drama:
i've been forced to use custom builds for around 1 year now, coz gpu mode above fast was pulling performance down yet not covering flicker entirely in most games tested. and we know that it's a mess to rely on gpu mode as it brings along a bunch of undesired stuff.
i need this toggle. and i've seem far less relevant before, so hey, gimme this one so i can call it mine :3
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4004
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
This PR reloads Android configuration when starting a game, so Eden uses the latest config.ini state for the emulation session.
On some Android launchers, Eden can already be running in the background before a game is launched. In that case, config.ini has already been read, so configuration changes made outside of Eden's settings flow while the process is alive are not picked up by the next emulation session.
This improves launch/config behavior for external launcher workflows.
Co-authored-by: FL-gs <efge.crea@gmail.com>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3965
Reviewed-by: crueter <crueter@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
traditionally, when doing jthread:
```
jthread() calls function parameter operator()() with args
function operator()() calls the code within
code within is, say { ThreadMain(); }
3 calls because why not
```
now this just makes it be 2 calls, mainly benefits non-LTO builds
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3970
Reviewed-by: crueter <crueter@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Add `[[maybe_unused]]` to unused constants in the ARM NCE backend to suppress compiler warnings.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4002
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Adding 'api.epicgames.dev' to the banlist, as instructed by lizzie.
MCI insists on connection and randomly crashes at Network::Poll(...).
Airplane mode would do, but then the DLC gets unavailable.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3999
Reviewed-by: Maufeat <sahyno1996@gmail.com>
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
bans more NS domains making hb dns check pass
also fixes bogus "SCTP is not implemented" assertion for translate protocol
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3955
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: crueter <crueter@eden-emu.dev>
Removes Eden's own feature detection for SSE, SSE2, FMA4 and AVX_VNNI, in order to clean up code and make the function slightly lighter. They are all unused and I have specific reasons for their removal:
- SSE/SSE2 are always true for x64
- FMA4 is a deprecated AMD exclusive extension, replaced by the regular FMA and was logged here as "FMA" for no reason
- AVX_VNNI simply does not seem applicable for Eden (but can still be checked for through Xbyak's feature detection if necessary)
Co-authored-by: RedBlackAka <140876408+RedBlackAka@users.noreply.github.com>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3979
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: crueter <crueter@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Ported from QML branch.
Main "big" change is that EmuThread is now a shared state in QtCommon,
not individually managed/passed around by GRenderWindow and MainWindow.
Signed-off-by: crueter <crueter@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3916
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
hbloader is required to launch a bunch of shit (particularly hbmenu)
however due to some restrictive metadata validation, it doesn't get recognized as an NSP file
programs like hbmenu require hbloader (since it feeds them data like ConfigEntries and whatnot), launching hbmenu as-is standalone will result in a launchInit() error
adds an option to launch hbmenu via command line
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3981
Reviewed-by: Maufeat <sahyno1996@gmail.com>
Reviewed-by: crueter <crueter@eden-emu.dev>
Ideally this should increase perf when doing write/read operations
not by a lot, just enough for it not to suck so much as it does right now
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3972
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: Maufeat <sahyno1996@gmail.com>
Since the launch of the steam controller I think it's only best to push towards updating to SDL3 allowing for a wider range of controller support
I went ahead and started on getting it working. Everything here should be functional, I've personally tested it all on Arch Linux. Still untested on windows, so looking for feedback on that
Any feedback and help would be appreciated!
Main changes:
- Bump everything to SDL3
- Handle SDL3 audio and input
- Add steam controller support, including HD Rumble
- Improved battery reporting via the status icon by using real % rather than state alone
Co-authored-by: crueter <crueter@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3952
Reviewed-by: crueter <crueter@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
On NetBSD it stupidly aligns the stack acounting for the amd64 redzone.
This isn't... required, to put it lightly.
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3966
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
this makes it so VK and OGL backends map the NVIDIA's ETC2 into VK_FORMAT_ETC-whatever and GL_ETC-whatever remaps, instead of using the default fallback for AR8G8B8. in short, just make the ETC2 textures be submitted as ETC2 instead of being submit as A8R8G8B8.
Signed-off-by: lizzie lizzie@eden-emu.dev
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3237
Reviewed-by: Ghost <>
Reviewed-by: crueter <crueter@eden-emu.dev>
the nominal std::unordered_map<> isn't enough to warrant it's continued usage in xbyak internal structures, thus using ankerl should greatly remove a lot of indirection/stdc++ specific overhead from the usually poorly performant std::unordered_map
Both dynarmic and macroHLE should benefit greatly from a less-stupid unordered_dense
This should speedup both CPU and shader compilation latency (NOT BY A GREAT MARGIN) just enough to make loading zones in ToTK less horrific
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3716
Reviewed-by: crueter <crueter@eden-emu.dev>
so basically each construction of HLEContext and whatever would result in a heap allocation (atleast 1)
so what if instead of that we did a memset() at ctor time and we avoided heap allocations altogether?
reminder that std::vector<> CAN do small object optimisation but it's not guaranteed
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3605
Reviewed-by: crueter <crueter@eden-emu.dev>
also devirtualizes manually since compiler doesn't do it with LTO
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3864
Reviewed-by: crueter <crueter@eden-emu.dev>
inputs shouldnt be that critical to require a full mutex of them
this relies on CPU guaranteeing u32/u16/u8 atomic load/stores for EmulatedController fields, which works on x86_64 but may not have the same behaviour on other architectures - thats why i wrap them in `std::atomic<>`
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3866
Reviewed-by: crueter <crueter@eden-emu.dev>
sounds like word salad but let me say:
- std::map<> created a static ctor for EVERY SINGLE ZONEINFO
- fuck that, instead lets just use a raw array and construct things statically
- works the same except with less baggage carried around (+ less heap allocations!!!)
this should help reduce codesize due to the aforementioned global ctor/dtor
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3919
Reviewed-by: crueter <crueter@eden-emu.dev>
