Repository navigation
What's Changed
- Benchmark FirebladeMath's own implementations and gate allocations in CI by @ctreffs in #75
- Speed up the portable 4x4 inverse, adjugate and quaternion multiply by @ctreffs in #76
- Compare every simd-backed function against Apple's simd and fix the differences found by @ctreffs in #77
- Check every public function against cross-platform reference values by @ctreffs in #78
Full Changelog: 1.3.0...1.4.0