v0.4.4

• Add AVX10/256 kernels for Affine and hashing
• Add RISC-V Zbc CRC acceleration
• Add non-Apple version of CLMul SHA3 kernel (mostly useful for Qualcomm chips)
• Minor performance tweak to MD5 BMI1 kernel and ARM CLMul kernels
• RISC-V compile fixes