Refactored thread-local stack allocator in AllocationManager to use ThreadLocalStackPool, removing global stack pointer arrays and locks. In FreeList, replaced fixed-size cache array and maxConcurrencyLevel with a dynamic linked-list system using SharedState and CacheReclaimer for thread cache lifecycle management. Block headers now store cache pointers instead of indices. Updated allocation/free logic and tests accordingly. Bumped assembly version to 3.1.3.
Updated Add/Remove/Enqueue/Push/etc. methods in core unsafe collections to accept parameters as 'scoped in T' for improved performance and safety. Bumped assembly versions in both csproj files.
Removed unmanaged struct requirement from SPMD job wrappers and extension methods, allowing managed types. Updated all wrappers and extension methods to require only the interface constraint. Refactored SPMD test jobs to use safe ref-based Store overloads. Improved README and docs with clearer debug/mimalloc instructions and a better SPMD example. Cleaned up Program.cs by removing obsolete experimental code. Enhanced math precision in GGXMipGenerationBenchmark. Updated T4 template to generate new constraints and APIs.
- Add static AllBitsSet property to ISPMDLane and implement in ScalarLane and WideLane
- Refactor WideLane shuffle table pointers and update usages
- Improve pointer safety and mask handling in CompressStore, Gather, and MaskLoad
- Enhance Sin, Cos, SinCos with fast-math and hardware fallback
- Add Newton-Raphson refinement for reciprocal/sqrt when not fast-math
- Optimize MathV.Vector vector loading (struct init, pointer ops)
- Update project file: version 1.3.4, content packaging, AOT settings
- Minor code cleanup and naming consistency fixes
- Require TLane : unmanaged, ISPMDLane for stricter type safety and direct memory ops
- Refactor GatherVectorN and WideLane<T>.Gather to use Unsafe.SkipInit and direct assignment, removing stackalloc and TLane.Load for better SIMD performance
- Use Vector.Sum in WideLane<T>.ReduceAdd
- Add/improve XML docs for ReduceAdd/ReduceMax/ReduceMin
- Update test project for AOT, AVX2, speed optimization, and disable reflection
- Tweak GGXMipGenerationBenchmark and Program.cs for improved benchmarking and output
- Added MaskLoad, Gather, and reduction methods (ReduceAdd, ReduceMax, ReduceMin) to ISPMDLane<TSelf, TNumber> with XML docs
- Changed SinCos to use out parameters instead of tuple return
- Implemented reductions in ScalarLane and WideLane (loop-based, TODO: SIMD)
- Added GetUnsafePtr to ISPMDLane
- Extended MathV to support Sin, Cos, SinCos, Tan, Asin, Acos, Atan, Atan2 for Vector2/3/4
- Improved WideLane.Sequence to use best vector type
- Updated GGX mip generation for new SinCos signature
- Bumped version to 1.3.2
- Enabled PNG dumping in GGX benchmark
Replaced struct with generic SPMD version for SIMD, added type aliases (commented), optimized RadicalInverse_VdC, and adjusted SampleEquirectangularMap for better performance and code separation.
- Replaced SIMD-based Sin/Cos/SinCos in WideLane with generic polynomial approximations for hardware independence.
- Updated ScalarLane Cast to use CreateTruncating.
- Applied AggressiveOptimization to key GGX methods; improved luma calculation and radical inverse LUT handling.
- Enhanced GGX benchmark setup, cleanup, and timing logic.
- Bumped project version to 1.3.1.
Updated AssemblyVersion in Jobs and LowLevel projects. Cleaned up Jobs csproj by removing unused properties, project reference, and content file inclusion. Deleted the Wrapper<T> disposable struct wrapper from Ptr.cs.
Relaxed generic constraints for job scheduling/execution to allow reference types (removed struct requirement). Updated IJob, IJobParallelFor, and IJobParallel extension methods to support both value and reference types, introducing RunRef for struct-specific overloads. Adjusted JobExecutor and JobScheduler to match new constraints. Bumped assembly version to 3.1.1. Added Value property to Wrapper<T> for ref access and inlined Get(). Changed GGXMipGenerationJob sample count to linear roughness. Removed unused usings in JobInfo.cs.
- Rename ISPMD interfaces to ISPMDLane for clarity
- Add gather and mask load methods to ISPMDLane, implement for ScalarLane and WideLane
- Add GetUnsafePtr() for direct pointer access
- Update MathV and vector types to use new interface and gather methods
- Update SPMD job interfaces and implementations to ISPMDLane
- Improve hash codes, range checks, and safety checks in vector types
- Update codegen templates for new interface/methods
- Refactor SPMD jobs to use gather methods for efficient vectorized access
Major refactor of job system and memory allocator:
- Replaced threadIndex with preferLocal for scheduling
- Switched local queues to SPMCQueue for better performance
- Introduced lock-free JobEdge pool for dependencies
- Removed remainingBatches; use ref counting for completion
- Updated all scheduling APIs and tests to new model
- Optimized FreeList struct sizes and block management
- Added allocation benchmarks
- Disabled OwnershipTransferAnalyzer temporarily
- Bumped assembly versions
Replaces unsafe pointer-based job data with JobDataPool<T> and ConcurrentSlotMap<T> for safer, type-safe management. JobInfo now references job data by (dataID, dataGeneration). JobExecutor and JobScheduler updated to use the new pool-based approach, requiring T : struct. Removed FreeList and pointer logic. WorkerThread now uses reference counting to prevent use-after-free. Updated all scheduling APIs and benchmarks to match new signatures. Improved documentation and inlining. Bumped assembly version to 3.0.0 due to breaking changes.
- Removed IJobScheduler interface; merged logic into JobScheduler
- Changed scheduling APIs to accept multiple dependencies (ReadOnlySpan)
- Moved WaitItem classes to JobScheduler.cs and updated types
- Updated JobExecutionContext to use JobScheduler and added docs
- Renamed AllocationManagerInitOpts to AllocationManagerDesc (required props)
- Added thread-safe TotalAllocatedMemory property to AllocationManager
- Refactored Integer to Data in benchmarks; updated usage
- Updated tests and improved documentation throughout
Major refactor:
- Add job priority tiers and async wait APIs to IJobScheduler
- Implement priority-based job queues and scheduling logic
- Introduce UnsafeParallelHashMap and refactor UnsafeParallelQueue
- Refactor UnsafeSlotMap to chunked storage for scalability
- Update SlotMap/ConcurrentSlotMap for consistency and perf
- Add new benchmarks and unit tests for parallel collections
- Misc: add MemoryUtility.AlignUp, version bumps, test improvements, bug fixes
- Refactored JobExecutionContext to use init-only properties and added State and SelfHandle for richer context.
- Updated JobInfo field layout for clarity and memory alignment.
- JobScheduler now accepts and exposes an optional state object, passed to each job context.
- Improved memory management by using _freeList for allocations and disposal.
- WorkerThread and benchmarks updated to use new JobExecutionContext pattern.
- Bumped version to 1.6.1 and performed minor code cleanup.
The `AllocationHandle` struct has been updated to replace
`public static readonly` fields with `public static` properties
using lambda expressions. The `Temp`, `FreeList`, and `Persistent`
fields are now properties, dynamically retrieving values from
their respective allocators. This change improves flexibility
and ensures the properties reflect the current state of the
underlying allocators.
Updated `JobScheduler` to enhance dependency tracking by counting valid dependencies upfront and dynamically adjusting counts using `Interlocked` operations. Improved job enqueueing logic to ensure jobs are only enqueued when all dependencies are met.
Replaced `Interlocked.Increment` with `Interlocked.Add` for batch updates to `_totalJobCount`, improving performance. Adjusted `VirtualStack` cleanup to use the correct size variable for memory deallocation.
Simplified `JobDispatchingJob` API by removing `ctx.ThreadIndex` parameter. Updated `TestJobSystem` to pass job handles as dependencies for proper execution order.
Incremented assembly version to 1.5.9 to reflect these changes.
Replaced the deprecated Allocator API with the new AllocationHandle API across the codebase. Updated constructors, methods, and tests to use AllocationHandle for memory management. Marked Allocator-based methods as [Obsolete] and provided alternatives.
Added OwnershipTransferAnalyzer to detect ownership transfer issues and introduced OwnershipTransferAttribute for marking parameters. Enhanced DefensiveCopyAnalyzer with additional checks for readonly and ValueType instances.
Refactored internal memory management in AllocationManager and updated benchmarks, utilities, and documentation to reflect the changes.
BREAKING CHANGE: Deprecated Allocator API in favor of AllocationHandle. Updated constructors and methods to use AllocationHandle. Users must migrate to the new API.
Refactored memory management by removing safety checks and introducing `MemoryHandle` for centralized tracking. Simplified allocation logic across allocators and enhanced `Dispose` methods for better resource cleanup.
Added `UnsafeChunkedQueue<T>`, a lock-free, dynamically resizing queue with chunk-based memory management, supporting parallel producers and consumers.
Updated unit tests to validate new queue functionality and ensure compatibility with refactored memory logic. Incremented assembly version to 1.6.12.
BREAKING CHANGE: Removed `#if MHP_ENABLE_SAFETY_CHECKS` blocks, altering memory validation behavior.
Refactor all unsafe collection enumerators to use ref struct types,
removing support for boxing and standard .NET enumeration interfaces.
GetEnumerator methods now return stack-only, more efficient enumerators
with [UnscopedRef] and inlining attributes. IEnumerable<T> and
IEnumerable implementations are removed from affected types. Interfaces
now require unmanaged types. Also includes minor doc and bug fixes.
BREAKING CHANGE: Enumerators are no longer compatible with LINQ, and collections no longer implement IEnumerable/IEnumerator.
Refactored UnsafeBitSet.Iterator to store a pointer to the bitset and accept a start index in its constructor. Updated GetIterator to support an optional start parameter and return an iterator using a pointer to the current instance. Revised Program.cs to demonstrate new usage and removed unrelated code. Bumped assembly version to 1.6.11.
Fixed conditional selection logic in quaternion and SVD math functions by swapping select argument order for correctness. Fixed LookRotationSafe and normalizesafe to return valid quaternions. Corrected SVD helper functions for proper value swapping and safe reciprocal. Added unit tests for matrix, reflection, projection, refraction, quaternion normalization, LookRotationSafe, and SVD operations. Incremented project version to 1.3.3. Minor formatting and using directive updates.
Added TestLookRotation method to TestQuaternion class. This test ensures quaternion.LookRotation produces a normalized quaternion and correctly rotates the forward and up vectors as expected.
Added [NumericConvertable] attributes for scalar type conversions to all vector structs. Fixed bitwise mask calculation in quaternion.cs for improved correctness. Updated Program.cs with quaternion test code and bumped project version to 1.3.2.
Added a TLSF (Two-Level Segregated Fit) memory allocator with O(1) allocation, free, and reallocation in `TLSF.cs`, plus comprehensive unit tests. Introduced a cache-line-padded `SeqLock` synchronization primitive. Refactored vector extension code for conciseness and fixed its usage to `extension(ref ...)`. Updated namespaces, removed unused code, and improved assertion and diagnostics. Updated NuGet dependencies and project files.
Refactor memory allocation system to use generic MemoryPool<TAllocator, TOpts> for arena, stack, and free list allocators, replacing custom allocator structs. Introduce MemoryBlock as a safer, more robust replacement for UnTypedArray. Improve thread safety, safety checks, and documentation. Reorder and clarify Allocator enum. Add comprehensive unit tests for all allocators and pointer assertion utilities. Update project to enable safety checks in Debug builds. Remove obsolete interfaces and ensure consistent deallocation with MemoryUtility.Free.
BREAKING CHANGE: Custom allocator structs are removed and replaced with MemoryPool-based abstraction. UnTypedArray is replaced by MemoryBlock. Allocator enum order and semantics are changed. Public API changes may require code updates.
Introduce a unified Reallocate method to all memory allocator types (Arena, Stack, FreeList, VirtualArena, VirtualStack, DynamicArena) and require it in the IMemoryAllocator interface. This enables efficient resizing of memory blocks, with fast-path optimizations for stack-like allocators.
Update AllocationManager and MemoryPool to use the new Reallocate method, simplifying and optimizing memory resizing logic. Add public properties for buffer pointers, sizes, and offsets to allocator structs for easier diagnostics.
Set FreeList's default concurrency level to 1 and make its allocation method return null on dispose instead of throwing. Clean up vector types for formatting, fix UnsafeList's RemoveRangeSwapBack logic, and simplify RemoveAtSwapBack.
Simplify Program.cs to only run SPMDBenchmark. Add new unit tests for FixedString, UnsafeList, UnsafeHashMap, and UnsafeHashSet. Apply minor test code cleanups for consistency in TestUnsafeQueue.
BREAKING CHANGE: IMemoryAllocator now requires a Reallocate method, and allocator APIs have changed accordingly.
Refactored UnsafeQueue<T>.Resize to prevent shrinking below
current count and handle wrap-around copying correctly.
Exposed AllocationHandle in UnsafeArray<T> via a new property.
Bumped project version to 1.6.10.
Added TestUnsafeQueue class to cover core queue operations.
Added scalar operator overloads for Vector types, fixed pointer math in Store methods, and improved enumerator and memory management. Updated test setup and removed allocation leak tests.
- Added left-hand scalar operator overloads for Vector2/3/4.
- Fixed pointer arithmetic in Store and GetUnsafePtr methods.
- Marked SetValue as readonly in UnsafeSparseSet.
- Improved enumerator initialization/reset for slot map and sparse set.
- Updated test projects' AssemblyVersion.
- Removed TestAllocationManager and added global AllocationManager setup/teardown.
- Updated TestConcurrentSlotMap for thread safety and correct cancellation.
- Minor formatting and parameter improvements.
Stack and VirtualStack Free methods now check pointer bounds and update offsets only when appropriate, improving safety. Program.cs refactored to use MemoryPool with VirtualStack and new test methods. Assembly version bumped to 1.6.7.
Added debug-only checks in Dispose methods of HashMapHelper, UnTypedArray, and UnsafeArray to log a message (with optional stack trace) if Dispose is called on an uninitialized or already disposed instance, unless the buffer is null. Incremented assembly version to 1.6.6.
Enhanced Dispose() in HashMapHelper, UnTypedArray, and UnsafeArray<T> to provide more informative debug messages when called on uninitialized or already disposed objects. When MHP_ENABLE_STACKTRACE is defined, the debug output now includes a stack trace with file, type, method, and line number details.
Also incremented project version to 1.6.5.
Add AllocationManager.GetTotalAllocatedMemory() to track total allocations when safety checks are enabled. Improve diagnostics by calling Debug.Fail in Dispose methods of HashMapHelper, UnTypedArray, and UnsafeArray when disposing uninitialized or already disposed arrays. Remove AddRange(ReadOnlyUnsafeCollection<T>) from UnsafeList<T>. Increment assembly version to 1.6.4. Ensure arrcpy is disposed in Program.cs.
Updated job state transition in WorkerThread to allow jobs already in the Running state to proceed, preventing unnecessary skipping. Also incremented project version to 1.5.8.
This change ensures jobs are not skipped if their state is already Running, improving reliability in job execution.
Replaces Unsafe.Copy/NativeMemory.Copy with direct pointer assignment for job data in JobScheduler, improving performance and code clarity. Adds a static Default property to AllocationManagerInitOpts for easier initialization. Updates test setup to use the new default options. Bumps assembly version to 1.5.7.
Updated FreeList allocator to guarantee at least one block is created per chunk by using Math.Max(1u, ...), and capped the number of blocks at 256. This prevents blocksToCreate from being zero and improves allocator robustness.
Bumped assembly version to 1.6.3.
- JobScheduler now explicitly frees unmanaged memory in _jobInfoPool before disposal.
- Removed the unused TempJobAllocator struct and implementation.
- Refactored AllocationManager to use conditional compilation for safety checks, making MemoryHandle usage conditional.
- Improved documentation in AllocationManager for clarity on allocation size and safety check behavior.
- Added UnsafeSetCount method to UnsafeList<T> for direct count manipulation with validation.
- Bumped AssemblyVersion in Jobs and LowLevel projects.
