That activity is being driven in musl directly. 1.2 has introduced optimizations for arm, aarch64, x86 and x86_64. there's more to optimize, but some are already there.