| 摘要 |
PURPOSE:To execute 63-bit integer type division without using a floating decimal point instruction by executing sorting by means of the number of the effective digits of divisors, in the case of the divisor at >= n bits, shifting a divident and the divisor, obtaining a value whose number of the effective digits is <=n bits, and obtaining a quoitient. CONSTITUTION:A computer 10 has plural registers respectively holding data of n bits, and holds a binary integer division instruction executing mechanism set in the registers. When a divident X is divided by a divisor Y so that the quoitient may be expressed in the range of 2n bits by the computer 10 whose one word consists of n bits, first the number of the effective digits is obtained. By the number of the effective digits, a division error or a clear quoitient can be obtained. Further by paying attention to the number of the effective digits of the divisor Y, sorting is executed, the division instruction by means of the divisor of n bitz is used, and the division is executed. Thus the 63-bit integer type division in a FORTRAN program, etc., can be executed without using the floating decimal point instruction. |
