<p>The system uses both 15-bit and 31-bit logical addresses. Instructions are 16-bit; some require the 15-bit addresses and some require the 31-bit. The choice is specified by certain combinations of instruction bits 0, 5 and 10 to 15. Blocks of macro-instructions are pre-fetched into instruction cache logic (120). Macro-instruction decode logic (121) decodes the macro-instructions and produces address descriptor bits which specify the required address size. Displacement logic (122) responds to the descriptor bits to convert 15-bit logical addresses to 31-bit addresses. The 31-bit addresses, whether of this size ab initio or produced by conversion from 15-bit addresses are converted into physical addresses by an address translation unit.</p>
申请公布号
EP0149858(B1)
申请公布日期
1991.03.06
申请号
EP19840201484
申请日期
1981.04.27
申请人
DATA GENERAL CORPORATION
发明人
ALSING, CARL J.;HENRY, CARL;HOLBERGER, KENNETH D.;HOLLAND, CHARLES J.;STAUDAHER, STEVEN M.;VERES, JAMES E.;WALLACH, STEVEN;ZIEGLER, MICHAEL L.