* Implement protecting multiple VMAs
A handful of games expect this to work, and updated versions of Grand Theft Auto V crash if it doesn't work.
* Clang
* memory: Consider flexible mappings as gpu accessible
Multiple guest apps do this with perfectly valid sharps in simple shaders. This needs some hw testing to see how it is handled but for now doesnt hurt to handle it
* memory: Clamp large buffers to mapped area
Sometimes huge buffers can be bound that start on some valid mapping but arent fully contained by it. It is not reasonable to expect the game needing all of the memory, so clamp the size to avoid the gpu tracking assert
* clang-format fix
---------
Co-authored-by: georgemoralis <giorgosmrls@gmail.com>
* Added all stubs + logging
* Added back memory and math functions from the original code + added some more math functions
* More string functions, snprintf, memmove and setjmp
* Add longjmp + clang
* gmtime, __cxa_atexit and log what functions some games use
* Add Mspace unreachable
* Renaming
* Take out mspace functions to their own file
* Factor out io to a new file
* Empty str and memory files
* Overloaded functions be like:
* str
* memory + math +clang
* clang...
* adjustments :D
* longjmp is questionable
---------
Co-authored-by: georgemoralis <giorgosmrls@gmail.com>
Unity, being the awful game engine it is, checks for a return value of zero to determine if sceKernelLoadStartModule failed. This results in it throwing an error code into sceKernelDlsym's handle parameter when the module it's searching for doesn't exist.
* Handle errors in sceKernelAllocateDirectMemory
* Improve accuracy of error cases
Some of our existing cases are normally EAGAIN returns.
* Improve logging on errors
* Clang
* TEMPORARY HACK FOR NBA TESTS
This will be removed before this PR is marked as ready, and is only here to make sure the other NBA games (and perhaps DOA3) work if some missing init behavior is handled.
* Revert "TEMPORARY HACK FOR NBA TESTS"
This reverts commit a0e27b0229.
* Change error message
* Fix event data for VideoOut events
Fix is based on some decompilation work shared by red_prig.
* Cleanup
* Oops
* Style fixes
* Clang
* Fix libSceVideoOut event idents
Based on some decompilation work, events coming from libSceVideoOut use a separate set of values for identifiers. These values are only converted to OrbisVideoOutEventId values during calls to sceVideoOutGetEventId.
For convenience, I've placed all relevant identifiers into a enum called OrbisVideoOutInternalEventId.
Thanks to @red_prig for the tips.
* Fix?
Seems like `static_cast<u32>(hint) & 0xFF == event.ident` here, and doing those right shifts on the event.ident winds up breaking stuff.
Without this change, the if always fails because event_id was getting set to 0 instead.
* Clang
* added support for loading keyboard config from file
* final minor update before pull request
* fix messing up the merge
* fix waitEvent to correctly handle mouse inputs
* add license
* Applied coding style fixes
* clang-format fucked up the .ini file
* actually fix clang changing ini syntax
use relative path for the ini file
* remove big commented out code blocks,
and fixed platform-dependent code
* fix windows hating me
* added mouse config option
* added toggle for mouse movement input (f7)
* fix license and style
* add numpad support i accidentally left out
* added support for mouse wheel (to buttons only)
* if keyboard config doesn't exist, autogenerate it
* added keybinds for "walk mode"
* Mouse movement input is now off by default
* code cleanup and misc fixes
* delete config file since it is now autogenerated
* F6 = F7 + F9
* added better mouse handling with config options
* Added capslock support
* fix clang-format
* Added support for mod key toggle key
* F6 and F7 are removed, F9 captures and enables the mouse
* Encapsulated globals and new classes in a new namespace
* Added mouse side button support
* Added per-game config
* relocated input parser to the new namespace
* changed parser parameters to make it possible to use it from the gui
* added home, end, pgup and pgdown
* Resolved merge conflict and refactored code
* Updated default keybindings
* Changed input handling to be single-threaded
* General code cleanup
* Start working on new backend
* Mouse polling, CMakeLists, and basic framework
* Output update handling, and reworked file creating, reading and parsing
* Parsing works now
* Single key button inputs work now
* Axis outputs work now
* Wheel works now (for me), l2/r2 handling improvements, and misc bugfixes
* Downgraded prints to log_debug, and implemented input hierarchy
* Implemented key toggle
* Added mouse parameter parsing
* clang-format
* Fixed clang and added a const keyword for mac
* Fix input hierarchy
* Fixed joysick halfmodes, and possibly the last update on input hierarchy
* clang-format
* Rewrote the default config to reflect new changes
* clang
* Update code style
* Updated sorting to accomodate for that one specific edge case
* Fix default config and the latest bug with input hiearchy
* Fix typo
* Temporarily added my GUI
* Update cmakelists
* Possible fix for Gravity Rush
* Update Help text, default config, and clang
* Updated README with the new keybind info
* okay so maybe the gravity rush fix might have slightly broken the joystick halfmode and key toggle
* Fixed mistakenly overwriting the last opened config with the default one if the GUI is opened multiple times in a session
* Updated Help descriptions and fixed mouse movement default parameters
* Fix crash if the Help dialog was opened a second time
If it's closed with the top right close button instead of clicking the Help button again, a required flag wasn't reset, making the next click on Help try to close a nonexistent window and segfault
* Added closing the config also closing the Help window, and fixed more segfaults due to mismatched flags
* Initial controller support
* clang and debug print cleanup
* Initial axis-to-button logic
* Updated Help text
* Added 'Reset to Default' button in GUI
* Minor text and description updates + fixed an issue with Help text box rendering
* Fix button-to-touchpad logic and l2/r2 handling, as they are both axes and buttons
The touchpad's button state was correctly handled, so games that use that were fine, but the touchDown flag was always set to true, so games that use this flag had problems, like Gravity Rush
* Fix merge conflict
* Clang
* Added back back button to touchpad binding
* Added touchpad button handling
* Added end-of-line comments and fixed some crashes happening with the VS debugger
* Apply recent changes from kbm-only
* Deadzone + initial directional axis-to-button mapping
* Added that one missing space in the README. Are you all happy now?
* Fixups from making everything use SDL
* Revert directional joystick code and fix a memory leak
* Change config directory name again to conform to project standards
* Clang
* Revert the old deeadzone code and properly add the new one
* Clang
* savedata: rewrite save memory functions to handle slot id & better backup management
* savedata: auto-restore backup if needed
* safe save backup shutdown
replaced exit by quick_exit
If the emulator code is above a 40-bit address, the embedded shaders need to use address-hi to work. Embedded VS shader already supplies it, PS shader should as well.
* Unmap memory in chunks if spanning over multiple VMAs
* clang
* Merge fixups
* Minor code style changes
* Update function declarations
---------
Co-authored-by: Marcin Mikołajczyk <marcinmikolajcz@gmail.com>
Should fix some `Region coalescing failed: Attempt to access invalid address.` crashes.
Co-authored-by: TheTurtle <47210458+raphaelthegreat@users.noreply.github.com>
* Error return on sceNpGetAccountIdA
Confirmed through hardware testing, this returns ORBIS_NP_ERROR_SIGNED_OUT on a signed out console.
Parameters are based on fpPS4 code.
* Fix compile
* Move errors to separate file
* Cleanup function headers
Swaps user_ids to use our OrbisUserServiceUserId type, and fixes parameter names to align with our coding standards.
* Add proper parameter checks
* Implement sceNpGetAccountId
This function takes an online_id, uses an NpManager function to get the user_id, then uses that user_id to perform the same internal functions as sceNpGetAccountIdA.
* Implement sceNpHasSignedUp
* Fix sceNpGetAccountId
Further hardware testing shows that these always write 0 to account_id when failing.
* Update np_manager.cpp
* presenter: render the game inside a ImGui window
* presenter: render the previous frame to keep the render rendering
* swapchain: fix swapchain image view format not being converted to unorm
* devtools: fix frame graph timing
* Auto-generate libSceHttp2
* Improved stub for sceHttp2Init
Needed for updated versions of Cyberpunk 2077.
Parameters are based on fpPS4, while the stub itself is based on similar stubs in our other networking libraries.
* Clang
I guess the line length calculations in the moduleGenerator are still not perfect?
* Auto-generate libSceSsl2 stubs
* Copy over sceSslInit stub
* Update CMakeLists.txt
* Swap to Lib_Ssl2 log category
* Fix compile
Since libSceSsl has many functions of the same name, these functions get treated as overloaded functions and break compiling.
* Clang
* devtools: fix popen in non-windows environment
* devtools: fix frame crash assertion when hidden
* devtools: add search to shader list
* devtools: add copy name to shader list
* devtools: frame dump: search by shader name
* handle getdents + fix condition + add info to description
* fix not handling dents errors
* to not overwrite it, only gather separate update entries when normal folder is done
* fix always setting entries to 0 and guest name including "UPDATE"
* reset indexes on completion
* don't use concat, fixes long standing bug
* make sce_module module loading take both paths into account
* coroutine code prettification
* asc queues submission refactoring
* better asc ring context handling
* final touches and review notes
* even more simplification for context saving
* shader_recompiler: Tessellation WIP
* fix compiler errors after merge
DONT MERGE set log file to /dev/null
DONT MERGE linux pthread bb fix
save work
DONT MERGE dump ir
save more work
fix mistake with ES shader
skip list
add input patch control points dynamic state
random stuff
* WIP Tessellation partial implementation. Squash commits
* test: make local/tcs use attr arrays
* attr arrays in TCS/TES
* dont define empty attr arrays
* switch to special opcodes for tess tcs/tes reads and tcs writes
* impl tcs/tes read attr insts
* rebase fix
* save some work
* save work probably broken and slow
* put Vertex LogicalStage after TCS and TES to fix bindings
* more refactors
* refactor pattern matching and optimize modulos (disabled)
* enable modulo opt
* copyright
* rebase fixes
* remove some prints
* remove some stuff
* Add TCS/TES support for shader patching and use LogicalStage
* refactor and handle wider DS instructions
* get rid of GetAttributes for special tess constants reads. Immediately replace some upon seeing readconstbuffer. Gets rid of some extra passes over IR
* stop relying on GNMX HsConstants struct. Change runtime_info.hs_info and some regs
* delete some more stuff
* update comments for current implementation
* some cleanup
* uint error
* more cleanup
* remove patch control points dynamic state (because runtime_info already depends on it)
* fix potential problem with determining passthrough
---------
Co-authored-by: IndecisiveTurtle <47210458+raphaelthegreat@users.noreply.github.com>
* texture_cache: Improve support for stencil reads
* libraries: Supress some spammy logs
* core: Support loading font libraries
* texture_cache: Remove assert
* core: Split error codes into separate files
* Reduces build times and is cleaner
* core: Bring structs and enums to codebase style
* core: More style changes
* Revert "semaphore: Use binary_semaphore instead of condvar"
This reverts commit 85dc57b868.
* semaphore: Use separate signaled flag to prevent races
* mutex: Few misc fixes
* condvar: Few misc fixes
* signals: Add thread name to unhandled signal message.
* Remove save migration code (#1621)
* Kernel Fixes (#1605)
* scePthreadSetprio Changes
FindThread uses posix error codes, so the function export should apply the ORBIS wrapper to convert these. Since it uses posix codes, I've also renamed the function to align with other posix functions. Lastly, this fixes a compile warning about ret sometimes not getting initialized.
* Implement posix_munmap
Used by Hatsune Miku Project Diva X during intros. May help with stability on Linux, probably won't change anything on Windows.
* Exports
Some missing function exports I've seen in my tests.
sceKernelAvailableFlexibleMemorySize export is used in Final Fantasy XV Episode Duscae
posix_pthread_setprio and posix_pthread_getschedparam are used by Spider-Man Miles Morales
scePthreadKeyDelete is used in UE4 games.
I've also added in a typo fix related to my previous PR.
* libScePosix export for posix_pthread_attr_setguardsize
Used in Hatsune Miku Project Diva X v1.02
* libraries: Add libSceRazorCpu HLE skeleton.
---------
Co-authored-by: ¥IGA <164882787+Xphalnos@users.noreply.github.com>
Co-authored-by: Stephen Miller <56742918+StevenMiller123@users.noreply.github.com>
* scePthreadSetprio Changes
FindThread uses posix error codes, so the function export should apply the ORBIS wrapper to convert these. Since it uses posix codes, I've also renamed the function to align with other posix functions. Lastly, this fixes a compile warning about ret sometimes not getting initialized.
* Implement posix_munmap
Used by Hatsune Miku Project Diva X during intros. May help with stability on Linux, probably won't change anything on Windows.
* Exports
Some missing function exports I've seen in my tests.
sceKernelAvailableFlexibleMemorySize export is used in Final Fantasy XV Episode Duscae
posix_pthread_setprio and posix_pthread_getschedparam are used by Spider-Man Miles Morales
scePthreadKeyDelete is used in UE4 games.
I've also added in a typo fix related to my previous PR.
* libScePosix export for posix_pthread_attr_setguardsize
Used in Hatsune Miku Project Diva X v1.02
* libkernel: Cleanup some function places
* kernel: Refactor thread functions
* kernel: It builds
* kernel: Fix a bunch of bugs, kernel thread heap
* kernel: File cleanup pt1
* File cleanup pt2
* File cleanup pt3
* File cleanup pt4
* kernel: Add missing funcs
* kernel: Add basic exceptions for linux
* gnmdriver: Add workload functions
* kernel: Fix new pthreads code on macOS. (#1441)
* kernel: Downgrade edeadlk to log
* gnmdriver: Add sceGnmSubmitCommandBuffersForWorkload
* exception: Add context register population for macOS. (#1444)
* kernel: Pthread rewrite touchups for Windows
* kernel: Multiplatform thread implementation
* mutex: Remove spamming log
* pthread_spec: Make assert into a log
* pthread_spec: Zero initialize array
* Attempt to fix non-Windows builds
* hotfix: change incorrect NID for scePthreadAttrSetaffinity
* scePthreadAttrSetaffinity implementation
* Attempt to fix Linux
* windows: Address a bunch of address space problems
* address_space: Fix unmap of region surrounded by placeholders
* libs: Reduce logging
* pthread: Implement condvar with waitable atomics and sleepqueue
* sleepq: Separate and make faster
* time: Remove delay execution
* Causes high cpu usage in Tohou Luna Nights
* kernel: Cleanup files again
* pthread: Add missing include
* semaphore: Use binary_semaphore instead of condvar
* Seems more reliable
* libraries/sysmodule: log module on `sceSysmoduleIsLoaded`
* libraries/kernel: implement `scePthreadSetPrio`
---------
Co-authored-by: squidbus <175574877+squidbus@users.noreply.github.com>
Co-authored-by: Daniel R. <47796739+polybiusproxy@users.noreply.github.com>
* `RendererVulkan` -> `Presenter`
* support for Video Out gamma setting
* sRGB hack removed
* added post process pass to presenter
* splash functionality restored
* Skip destruction of adaptive mutex initializers
Based around similar behaviors implemented in the more-kernel branch. Hatsune Miku Project Diva X needs this.
* Fix posix_lseek result overflow
Seen when testing Spider-Man Miles Morales on more-kernel.
* Add posix_fsync
Used by Spider-Man Miles Morales. I've added the normal posix error handling to this function, though I'm aware sceKernelFsync doesn't return any errors currently. This is for future proofing and accuracy, as the actual libkernel does the usual error handling too.
* Properly handle VirtualQuery calls on PoolReserved memory.
* Export posix_getpagesize under libkernel
Bloons TD 5 needs this.
* Clang
* ajm: handle single-frame decode jobs (+mp3 imrovements)
* disable breaking the loop in multi-frame if storage is insufficient
* simplified gapless decoding
* AJM: Added support for different PCM formats
* updated libatrac9 ref
* remove log
* Add support for non-interleaved flag
* Added support for output sideband format query
