摘要 |
In a processor where separate integer and floating point units are utilized, conditions generated in the integer unit are transferred and made compatible for use in the floating point unit by floating point conditional branch and move operations. Conversely, conditions generated in the floating point unit are transferred and made compatible for use in the integer unit by integer conditional branch and move operations. By providing semantic compatibility of conditions with conditional operations in both integer and floating point units, conditions can be generated in one numeric unit and operated in the other.
|