wrp_mul_u32
wrp_mul_u32(nfp_prog, dst_hi, dst_lo, multiplicand, multiplier,
wrp_mul_u32(nfp_prog, imm_both(nfp_prog), reg_none(), dst_a,
wrp_mul_u32(nfp_prog, dst_both, reg_none(), dst_a, magic, true);