From vsocci@best.com Thu Jun 15 17:54:05 PDT 1995 Article: 1216 of alt.sys.pdp10 Path: news.u.washington.edu!news.uoregon.edu!usenet.eel.ufl.edu!hookup!news.sprintlink.net!simtel!agis!news1.best.com!vsocci From: vsocci@best.com (Vance Socci) Newsgroups: alt.sys.pdp10 Subject: Re: Emulator Date: Thu, 15 Jun 95 04:07:11 GMT Organization: Vcc Technical Services Lines: 58 Distribution: world Message-ID: <3romc4$bsv@news1.best.com> References: , <3rmgf3$ekr@sacsa3.mp.usbr.gov> NNTP-Posting-Host: vsocci.vip.best.com X-Newsreader: News Xpress Version 1.0 Beta #2.1 henrym@sacto.mp.usbr.gov wrote: >>Nope! But my housemate has a nonfunctioning Foonly F4 out in a storage >>shed (it no longer seems to be able to load microcode). Maybe that's >>a start. :-) > > > I dunno - did it have a properly working PCXT [BLT] ??? > >-HWM >---------- >Henry W. Miller >Assistant Systems and Network Manager >U.S. Bureau of Reclamation, Mid Pacific Region >2800 Cottage Way MP1130 >Sacramento, CA 95825 (916) 979-2373 Last time I saw one they did. The F-4 was an F-3 converted to be largely KL-10 compatible (with KI style paging, no extended addressing). The F-4 was originally created from the F-3 by Foonly, Inc. for Tymshare, Inc. and as far as I know Tymshare was the only place that bought any (but there may be other customers that I didn't know about). (I helped write and debug the microcode for the KI style paging, re-wrote the tape controller microcode to allow interrupts during tape transfers, and implemented the KL only user level instructions like ADJBP, DMOVE, DADD, etc.) The F-3 was originally designed to be the front end for the Super-Foonly, the F-1. I think only one Super Foonly ever existed, and was developed for and owned by III down in Los Angeles somewhere. The device controllers for the F-3 and F-4 were done in microcode, as well as the Tymnet base interface. They were unique to the F-4. What ever happened to Dave Poole? Time was when the PDP-10 designers at DEC would go out to visit Poole and (Phil) Petit to brainstorm before attempting the next PDP-10 design. The KL-10 cache was a variant on some ideas espoused by Poole and Petit. And yes, PXCT worked in all its flavors on the F-4. PXCT is not all that hard to understand once you understand what its trying to accomplish and the nature of its target instructions. For instructions that have only one operand, the only two bits that operate in PXCT are 1) the bit that says to calculate the effective address in the previous context; and 2) the bit that says to fetch or store the operand itself in the previous context. The other two bits are for specifying previous context for the second operand (BLT, PUSH, data in byte instructions) and for the effective address of byte instructions. These two bits also affect the context for EXTEND type instructions. - Vance /=======================================\ | Vance Socci vsocci@best.com | | "The worst secrets are those we keep | | from ourselves . . ." | | "I am not a number; I am a free man! | \=======================================/