[PPL-devel] [GIT] ppl/ppl(floating_point): Using uint16_t instead of uint32_t in float_ieee754_half.