> note that this sequence is only used by a few platforms, seems a > little over half check against the p_emul->e_nsysent. Yes, some platforms made an effort a few years ago to shave some cycles off ... at the expense of correctness. seems wrong to me.