* Refactoring and optimization on CPU translation
* Remove now unused property
* Rename ilBlock -> block (local)
* Change equality comparison on RegisterMask for consistency
Co-Authored-By: gdkchan <gab.dark.100@gmail.com>
* Add back the aggressive inlining attribute to the Synchronize method
* Implement IEquatable on the Register struct
* Fix identation
* Profiler initial setup
* Capture actual timing data
* Profiling data dumped to file on close
* Support for multiple sessions under the same name
* Service profiling
* Sort output for easier read
* csv output
* Split session into 2 seperate values
* Refactor name to category
* Basic profiling window dummy. Toggle with F1 or set key with config
No actual data displayed yet, just a pretty triangle
* Simple font rendering
* Display some actual timing data
* Fix font bearing being ignored
* x bearing and advance. Fixed y bearing calc
* Different coloured lines to make reading easier
* Scrolling
* Multiple columns for name
* Column titles
* display in ms rather than ticks
* Bars to display times
* Sortable columns
* Regex filtering
* Better instant timing calculation
Fixed minor regex bug
* Better filtering
Better max value calculation
Skip some rendering to reduce profiler weight
* Variable update rate
* Show/hide inactive button
Some other touchups
* Add missing project reference
* Hide inactive and pause
* Fix viewport errors
* Update initial window position
* Variable name cleanup
* Disable timing dump by default
* Internal Profile refactor and cleanup
* Timing info cleanup
* Profile config cleanup
* Settings cleanup
* Button refactor
* Profile refactor
* Profile window cleanup
* Window manager refactor
* Font service cleanup
* Fixed bug in profiling method where method was called twice without profiling enabled
* Allow update rates of less than 1hz
* Stop using window.run because it's apparently not great for performance.
Some other performance things, should only draw a new frame when something has changed
* Improved time tracking to keep history
* Profile window was getting too long so I added regions and split bar rendering out into partial class
* Dummy graph view with button to toggle
* Realtime graphing initial commit
* Display totals on new bar
* Simple zooming support with arrow keys
* Limit graph zoom and label start and stop
* Added support for timing flags
* Stop data running away when paused and frame updated
* Manual step button
* Update at when flag issued (ie every frame)
* Removed useless finish profiling call
* Enable and disable profiling at compile time.
* Better plage for frame swap flag, also kept enough flags to cover larger time spans
* No more stopwatches created, uses PerformanceCounter now
* public and internal fields to props
* Move visible update to update rather than draw as it causes a lockup if called from draw
Also added profile window disposal so closing main window closes profiler too
* Fixed optimization settings for profiled builds
* Appveyer script guess to add profiling builds
* Quotes
* 1 less quote
* Maybe escape space?
* Specify config
* Different approach
* Fix file paths
* Fix another path
* Better artifact naming
* Missing -
* test string
* Removed for, to test
* readd for
* moved dashes around so artifacts can begin with letters
* quote env vars
* martix
* Removed configs
* Much more efficient capture, ConcurrentDictionary was causing too much overhead
* Skip repeating pixels during draw
* Stop ram usage getting too high. Compensating for cleanup doing more now
* Profile CPU, execute skipped because it's just too much work
* Fixed bug with skipping draws. Furthest needed to be reset every loop
* Less distracting colour for timing flags
* Removed profile method function. It just doesn't play nice with conditional compilation so best to remove it now before it's used a lot
* Null check for category, group and item
* Forgot to reset instant count/time
* Increment line when blank
* Fix threading conflict
Fixed instant count and time. Now accuratly represents the total time and count in the buffer
* Fixed bug in time rendering where times were being trimmed to an int.
Also added microsecond/millisecond formatting to reduce the number of decimal places needed
* Support for multiple profiling levels
* Sometimes it would have to wait a long time for lock to clear so moved it to a tryenter and skip if already locked
* Dumb bug regarding clearing of timestamps. Start is already removed so no need to add it to the start
* Optimisations in drawing routine:
Only calculate bar top and bottom once per bar rather than once per timestamp
Pre-calculate the right side of the graph as it was being calculated multiple times per bar
Skip rendering timestamps that occupy the same pixel space now uses the raw timestamp to decide. While technically not as accurate it's much easier as the right side of the bar doesn't have to be calculated for a skipped timestamp
* Couple alignment changes
* Custom equals overload for profile config. The default implpmentation was just too slow
* Bump cleanup thread priority. It clears the timer queue so it need to be run frequently
* Fixed bug with scrolling caused by recent rendering optimisations. Simply forgot to increment the line index on a skipped line
* Stopped blocking memory disposal so much. Also parralised(?) cleanup call
* Uses Arial for font.
* Enable AA
* Inital seperated config support
* Fix profile input from keyboard
* Check toggle visible key from profiler
* Can't use conditional here as _profileWindow doesn't exist it non-profiling build
* Removed junk from merge in sln
* Fromatting cleanup for review
* Fiked small bug caused by race condition
* Added multiple flags with colours
Added way to set max flags
* Fixed flag times
Dispays time flags in window
* Colors for text frame times
* enable and disable flags button added
better fix for race crash
* Re factored npad out
* Explicitly specified type in foreach
* Removed extra line
* Added s to fix nit
* Comment to clarify default time
* Another s nit
* Ordering nit
* Uses Interlocked.Increment over lock
* Unindented #if's and #regions
* Comment to clarify these are indexes in the list
* Uses iequatable over override equals to avoid conversion and checks at runtime
* Removed no longer used variable
* Add optimizations related to caller/callee saved registers, thread synchronization and disable tier 0
* Refactoring
* Add a config entry to enable or disable the reg load/store opt.
* Remove unnecessary register state stores for calls when the callee is know
* Rename IoType to VarType
* Enable tier 0 while fixing some perf issues related to tier 0
* Small tweak -- Compile before adding to the cache, to avoid lags
* Add required config entry
* Implement faster address translation and write tracking on the MMU
* Rename MemoryAlloc to MemoryManagement, and other nits
* Support multi-level page tables
* Fix typo
* Reword comment a bit
* Support scalar vector loads/stores on the memory fast path, and minor fixes
* Add missing cast
* Alignment
* Fix VirtualFree function signature
* Change MemoryProtection enum to uint aswell for consistency
* Add fixed-point variant of the UCVTF instruction
* Change encoding of some fixed-point instructions to not allow invalid encodings
* Fix Fcvtzu_Gp_Fixed encoding
* Add SCVTF (fixed-point GP to Scalar) instruction
* Simplify *Fixed encodings
* Implement ARM exclusive load/store with compare exchange insts, and enable multicore by default
* Fix comment typo
* Support Linux and OSX on MemoryAlloc and CompareExchange128, some cleanup
* Use intel syntax on assembly code
* Adjust identation
* Add CPUID check and fix exclusive reservation granule size
* Update schema multicore scheduling default value
* Make the cpu id check code lower case aswell
* Implement speculative translation on the cpu, and change the way how branches to unknown or untranslated addresses works
* Port t0opt changes and other cleanups
* Change namespace from translation related classes to ChocolArm64.Translation, other minor tweaks
* Fix typo
* Translate higher quality code for indirect jumps aswell, and on some cases that were missed when lower quality (tier 0) code was available
* Remove debug print
* Remove direct argument passing optimization, and enable tail calls for BR instructions
* Call delegates directly with Callvirt rather than calling Execute, do not emit calls for tier 0 code
* Remove unused property
* Rename argument on ArmSubroutine delegate
* Implement ARM32 memory instructions: LDM, LDR, LDRB, LDRD, LDRH, LDRSB, LDRSH, STM, STR, STRB, STRD, STRH (immediate and register + immediate variants), implement CMP (immediate and register shifted by immediate variants)
* Rename some opcode classes and flag masks for consistency
* Fix a few suboptimal ARM32 codegen issues, only loads should be considered on decoder when checking if Rt == PC, and only NZCV flags should be considered for comparison optimizations
* Take into account Rt2 for LDRD instructions aswell when checking if the instruction changes PC
* Re-align arm32 instructions on the opcode table
* Remove ARM32 interpreter and add ARM32 support on the translator
* Nits.
* Rename Cond -> Condition
* Align code again
* Rename Data to Alu
* Enable ARM32 support and handle undefined instructions
* Use the IsThumb method to check if its a thumb opcode
* Remove another 32-bits check
* Optimized memory modified check
This was initially in some cases more expensive than plainly sending the data. Now it should have way better performance.
* Small refactoring
* renamed InvalidAccessEventArgs
* Renamed PtPageBits
* Removed ValueRange(set)
They are currently unused and won't be likely to be used in the near future
* Fix and simplify TranslatorCache
* Fix some assignment alignments, remove some unused usings
* Changes to ILEmitter, separate it from ILEmitterCtx
* Rename ILEmitter to ILMethodBuilder
* Rename LdrLit and *_Fix opcodes
* Revert TranslatorCache impl to the more performant one, fix a few issues with it
* Allow EmitOpCode to be called even after everything has been emitted
* Make Emit and AdvanceOpCode private, simplify it a bit now that it starts emiting from the entry point
* Remove unneeded temp use
* Add missing exit call on TestExclusive
* Use better hash
* Implement the == and != operators
* 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
* 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
* Change naming convention for Ryujinx project
* Change naming convention for ChocolArm64 project
* Fix NaN
* Remove unneeded this. from Ryujinx project
* Adjust naming from new PRs
* Name changes based on feedback
* How did this get removed?
* Rebasing fix
* Change FP enum case
* Remove prefix from ChocolArm64 classes - Part 1
* Remove prefix from ChocolArm64 classes - Part 2
* Fix alignment from last commit's renaming
* Rename namespaces
* Rename stragglers
* Fix alignment
* Rename OpCode class
* Missed a few
* Adjust alignment
* 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
* Print stack trace on invalid memory accesses
* Rebased, change code region base address for 39-bits address space, print stack trace on break and undefined instructions too
* Remove unused tracing functionality from the CPU
* GetNsoExecutable -> GetExecutable
* Unsigned comparison
* Re-add cpu tracing
* Config change
* Remove cold methods from the translation cache on the cpu
* Replace lock with try lock, pass new ATranslatorCache instead of ATranslator
* Rebase fixups
* Started to rewrite the thread scheduler
* Add a single core-like scheduling mode, enabled by default
* Clear exclusive monitor on context switch
* Add SetThreadActivity, misc fixes
* Implement WaitForAddress and SignalToAddress svcs, misc fixes
* Misc fixes (on SetActivity and Arbiter), other tweaks
* Rebased
* Add missing null check
* Rename multicore key on config, fix UpdatePriorityInheritance
* Make scheduling data MLQs private
* nit: Ordering
