Based on:
- par2cmdline commit ff990c0 (0.8.2-dev)
- ParPar v0.4.2
Changes:
- Remove OpenMP dependency, enabling fully static Windows/MacOS builds
- Add SIMD/multi-threading acceleration to matrix inversion
- Add Apple M1 optimised implementation of the CLMul GF16 kernel
- Add RISC-V Vector implementation of the Shuffle128 GF16 kernel