Fix attribution for Windows ARM64 support (origin: PR #99 by Demonese)

Author: JkarVN

CREDITS.md

@@ -24,12 +24,13 @@ This modernized fork of `pcg-cpp` by **Total-Random** integrates several critica
 - **Author:** [Timo Alho](https://github.com/tialho)
 - **Description:** Resolves various GCC warnings (clz/ctz truncation) when building with `-Wall`.
 
-## Total-Random Improvements
-
 ### 5. Native Windows ARM64 Support
-- **Author:** [Total-Random](https://github.com/Total-Random)
+- **Origin:** [imneme/pcg-cpp PR #99](https://github.com/imneme/pcg-cpp/pull/99)
+- **Author:** [Demonese](https://github.com/Demonese)
 - **Description:** Added native support for ARM64 on MSVC using `__umulh` for efficient 128-bit multiplication.
 
+## Total-Random Improvements
+
 ### 6. Modern CMake Build System
 - **Author:** [Total-Random](https://github.com/Total-Random)
 - **Description:** Comprehensive CMake integration with `find_package` support and automated testing via `ctest`.

README.md

@@ -13,10 +13,10 @@ This is the **maintained** version of the PCG C++ library, hosted by the **Total
 
 The original repository by Melissa O'Neill (`imneme/pcg-cpp`) has been a cornerstone of the C++ ecosystem but hasn't seen updates in several years. **Total-Random** exists to keep such vital legacy libraries alive by integrating community fixes, supporting new architectures, and fixing build breakages on modern IDEs.
 
-### Key Improvements in this Fork:
-- **Windows ARM64 Support**: Native support for MSVC on ARM64 architectures (Surface Pro, Apple Silicon via VM, etc.) using `__umulh` intrinsics.
+### Key Improvements and Integrated Fixes:
+- **Windows ARM64 Support**: Integrated native support for MSVC on ARM64 architectures (Surface Pro, Apple Silicon via VM, etc.) using `__umulh` intrinsics.
 - **Improved MSVC Compatibility**: Fixed ambiguous operator errors (`C2678`) in `set_stream` and `operator>>`.
-- **Optimized `unxorshift`**: Implemented an optimized version of the inverse xorshift operation, improving performance for large integer types.
+- **Optimized `unxorshift`**: Integrated an optimized version of the inverse xorshift operation for better performance.
 - **Consistent Typing**: Unified integer type handling across different platforms to avoid compiler warnings and errors.
 
 Detailed information about integrated community fixes and contributors can be found in [CREDITS.md](CREDITS.md).

include/pcg_uint128.hpp

@@ -737,6 +737,7 @@ uint_x4<UInt,UIntX2> operator*(const uint_x4<UInt,UIntX2>& a,
 #if defined(_M_AMD64) || defined(_M_IX86)
 #pragma intrinsic(_umul128)
 #elif defined(_M_ARM64)
+// This fix is from imneme/pcg-cpp PR #99 by Demonese
 #pragma intrinsic(__umulh)
 #else
 #error Unsupported architecture