Add macro-controlled SIMD support (SSE/NEON) to math library#64

Draft

Copilot wants to merge 9 commits into

mainfrom

copilot/improve-math-library-simd-support

Commits on Feb 16, 2026

Initial plan
Copilot
committed
Add SIMD support to math library with macro-controlled SSE/NEON paths

Copilot
and
bluesky013
committed
Add Newton-Raphson refinement to SIMD normalize for improved precision

Copilot
and
bluesky013
committed
Final: SIMD math library support complete

Copilot
and
bluesky013
committed
Remove codeql artifact and add to .gitignore

Copilot
and
bluesky013
committed
Split SIMD data types: extract SFloat4 abstraction from Vector4/Matrix4

Copilot
and
bluesky013
committed
Improve scalar fallback readability in SFloat4.h

Copilot
and
bluesky013
committed
Restore __m128/float32x4_t as union member in Vector4

Copilot
and
bluesky013
committed
Delete SFloat4 abstraction layer, use raw SIMD intrinsics directly

Copilot
and
bluesky013
committed