Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.1 6/24/83; site usl.UUCP Path: utzoo!watmath!clyde!bonnie!akgua!akgub!usl!sigma From: sigma@usl.UUCP (Spyridon Triantafyllopoulos) Newsgroups: net.micro.pc Subject: Re: IBM 370/XT Message-ID: <313@usl.UUCP> Date: Sat, 16-Feb-85 21:10:03 EST Article-I.D.: usl.313 Posted: Sat Feb 16 21:10:03 1985 Date-Received: Mon, 18-Feb-85 05:30:22 EST Organization: USL, Lafayette, LA Lines: 31 > From: ptw@encore.UUCP (P. Tucker Withington) > Subject: How did they really make the XT/370? > > A while back there was talk that the XT/370 is based on a chip that IBM got > Motorola to make for them that was basically a 68000 with "new microcode". > I guess I can believe this, but are there any technical papers out there that > either confirm or "tell the whole story"? It uses 2 MC 68000 and an Intel 8087 for the S/370 operations. One 68000 performs the instruction fetch, decode and emulation of most fixed point 370 instructions. It also uses its registers and Program counter to emulate the 370's GPR's and PSW. The second 68000 manipulates the page table, processes exceptions (invalid opcodes, zerodivide,etc) and emulates system-related 370 instructions. This includes SVC calls and the DIAGNOSE instruction (that passes requests to PC-DOS). When accessed by the processor card, the first 480KB appear as locations 0-480K of S370 memory. The remaining 32K store the Motorola 68000 microcode used to emulate the 370 functions and interface operations w/PC-DOS as carried out by the other 68000 Byte magazine, special IBM issue, October 84, reprinted WithOut Permission and/or Persuation. (about as much 370 stuff as you will find--not much). Hope this helps... I have some experience with the beast... If (more_questions) then write. Have Fun... -- Spiros Triantafyllopoulos /* In the deaf man's door, -- USL Computer Science Dept. there is a bell with a light */ -- {ut-sally, akgua}!usl!sigma -- Old Greek Proverb ---