HECnet

hecnet@lists.dfupdate.se

15562 discussions

by tonyblews＠gmail.com

Hi, Despite me not doing anything different apart from rebooting my machine, TARDIS is back up. This has spurred me into action. What I'd like to do is run the bridge program on the same raspberry that the SIMH image runs on, but I have no idea how to do this. I seem to recall a post about doing it recently, but I can't recall how to view the list archives. Ideally I'd like to be the bridge, the TARDIS SIMH image and Linus DECnet on the same machine. Is this possible? Tony.

12 years, 7 months

New node, SIIRI - 47.113

by sampsa＠mac.com

Johnny, When you get a chance can you register this for me? sampsa

12 years, 7 months

KLH10 on Raspberri Pi / Nodelist update

by jg＠jordi.guillaumes.name

Jordi Guillaumes i Pons Barcelona - Catalunya - Europa El 04/12/2012, a les 22:55, "Boyanich, Alastair" <Alastair.Boyanich at au.fujitsu.com> va escriure: I've not messed around with TOPS-10/20 and only VMS on simh, but there you need specify which O/S is running inside the VM and specify it with something like: set cpu idle=VMS Well, KLH10 os a different beast. You have to patch the guest OS to do a write to an " idler" device so the simulator knows the guest is not doing anything and it can enter a wait state until next timer tick or interrups happens. This patch is included in the Panda monitor and it works as expected... ... Unless you happen to be using a "high performance" host system (he, and the Pi seems to fit that class lol...). Then you have to build the simulator to use a synchronous implementation for the timers and device interrupts (see https://groups.google.com/forum/m/?fromgroups#!msg/alt.sys.pdp10/vHcQ49tjMa…) and the idler simply does not work at all. Unfortunately it seem there is no fix for this problem (see https://groups.google.com/forum/m/?hl=en&fromgroups#!msg/alt.sys.pdp10/xndb…) so I will have to live with it. Not a big problem in the raspi anyway... (Apologies for the double post... The bus just caught a bump and my finger moved to the send button when I was completing the message :))

12 years, 7 months

KLH10 on Raspberri Pi / Nodelist update

by jg＠jordi.guillaumes.name

12 years, 7 months

KLH10 on Raspberri Pi / Nodelist update

by Paul_Koning＠Dell.com

On Dec 4, 2012, at 4:55 PM, Boyanich, Alastair wrote: Hi Jordi, I've not messed around with TOPS-10/20 and only VMS on simh, but there you need specify which O/S is running inside the VM and specify it with something like: set cpu idle=VMS If I get it wrong or omit it, then it will sit there wasting a huge amount of cycles on my alpha. Off the top of my head, I suspect simh is actually looking for instruction patterns for the idle loop and detecting based upon that, which would differ from O/S to O/S. Perhaps TOPS-10/20 requires something similar? Yes, that's what SIMH does. Depending on the processor type, this code may be rather fragile. On a PDP-11, most (though not all) idle loops have a WAIT in them somewhere. On VAX, there is no such thing, and the code looks for "scheduler idle loop" patterns. It's pretty ugly and your particular OS may not work well. For example, NetBSD 6 does not match any of the VAX idle patterns, and in fact it's not clear any idle handling can be done sanely there without help from the NetBSD code... The first step would be to make sure the idle checker is set for your OS. If that still doesn't work, you may need to read the emulation code to see what the checker actually does, to determine if it needs to be updated to handle the OS you're running. paul

12 years, 7 months

KLH10 on Raspberri Pi / Nodelist update

by Alastair.Boyanich＠au.fujitsu.com

Hi Jordi, I've not messed around with TOPS-10/20 and only VMS on simh, but there you need specify which O/S is running inside the VM and specify it with something like: set cpu idle=VMS If I get it wrong or omit it, then it will sit there wasting a huge amount of cycles on my alpha. Off the top of my head, I suspect simh is actually looking for instruction patterns for the idle loop and detecting based upon that, which would differ from O/S to O/S. Perhaps TOPS-10/20 requires something similar? Al. -----Original Message----- From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Jordi Guillaumes i Pons Sent: Monday, 3 December 2012 8:25 PM To: hecnet at Update.UU.SE Subject: [HECnet] KLH10 on Raspberri Pi / Nodelist update Hello, Please add to the HECnet nodelist: 7.79 BITXT0 TOPS-20(PANDA) - KLH10 - Raspbery Pi I am running this simulator in the same Pi which currently holds 5 more VMs. I will see if it keeps stable, and if it does I will run it permanently. Issues to solve - KLH10 does not idle in the Pi. I'll check the idler device code to see if I can guess what is happening. - I tried to give KLH10 its own bridge device, plugged into the rest of my network via VDE switches and plugs, but it does not work. I have lots of IPBLP bugchecks and IP networking is unusable (weird enough, DECNET and LAT seem to work). It works fine if I use the main host interface in non-exclusive mode (but then I don't get LAT). - KLH10 frontend can't load *.EXE files. - TOPS-10 does not run well. I used a x86 KLH to dump the load code in SAV format and then I can boot TOPS-10, but it hungs after a while. On the other hand, I'm thinking about building a blinkenlights device using the GPIO pins of the Rasp-pi basing it on the actual parallel port blinkenlights panel driven by the "lites" device... :) It could be another nice hobby project... Jordi Guillaumes i Pons jg at jordi.guillaumes.name HECnet: BITXOV::JGUILLAUMES

12 years, 7 months

Largest PATA drive that a DS10 can use?

by form＠pdp-11.org.ru

sampsa at mac.com wrote: Gentlemen, I've currently got 3 x 127 GB PATA drives on CHIMPY - what is the largest drive this machine (DS10) will be capable of using? Sampsa Just tested WD2500BB IDE drive (250Gb). It works just like 127Gb disk... --- Initializing... *** keyboard not plugged in... 2048 Meg of system memory probing hose 0, PCI probing PCI-to-ISA bridge, bus 1 bus 0, slot 9 -- ewa -- DE500-BA Network Controller bus 0, slot 11 -- ewb -- DE500-BA Network Controller bus 0, slot 13 -- dqa -- Acer Labs M1543C IDE bus 0, slot 13 -- dqb -- Acer Labs M1543C IDE bus 0, slot 14 -- vga -- S3 Trio64/Trio32 bus 0, slot 16 -- pka -- Adaptec AIC-7880 initializing GCT/FRU at 3ff40000 System Temperature is 36 degrees C AlphaServer DS10 617 MHz Console V7.3-1, Feb 27 2007 13:17:58 >>>sh dev dqa0.0.0.13.0 DQA0 WDC WD40 0BB-00DEA0 05.03E05 dqa1.1.0.13.0 DQA1 Maxtor 6 Y120P0 YAR41BW0 dqb0.0.1.13.0 DQB0 WDC WD25 00BB-22RDA0 20.00K20 dva0.0.0.0.0 DVA0 ewa0.0.0.9.0 EWA0 00-02-A5-C6-81-45 ewb0.0.0.11.0 EWB0 00-02-A5-C6-81-44 pka0.7.0.16.0 PKA0 SCSI Bus ID 7 >>> --- $ INI DQB0: ALPHATST $ MOU DQB0: ALPHATST %MOUNT-I-MOUNTED, ALPHATST mounted on _CTEPBA$DQB0: $ SH DEV/MO Device Device Error Volume Free Trans Mnt Name Status Count Label Blocks Count Cnt CTEPBA$DQA0: Mounted 0 ALPHASYS 72841968 396 1 CTEPBA$DQA1: Mounted 0 ALPHAUSR 219968272 1 1 CTEPBA$DQB0: Mounted alloc 0 ALPHATST 268433264 1 1 $

12 years, 7 months

PDT-11

by wilson＠dbit.com

I know PDT-11s are off-topic but they came up before: V6.1 of the Ersatz-11 PDP-11 emulator is now done, and the demo/hobby version is available here: http://www.dbit.com/demo.html This release adds emulation of the PDT-11/130 and PDT-11/150 microcomputers, as well as a rough cut of the Terak 8510/a microcomputer (but the Terak graphics require DOS, Windows, or OS/2 -- no Linux unless/until a better linker turns up). There's an appendix showing how to configure each. John Wilson D Bit

12 years, 7 months

KLH10 on Raspberri Pi / Nodelist update

by jg＠jordi.guillaumes.name

I have put the basic setup in BITXOV:: (default FAL directory), in the file TOPS20PI.ZIP The ZIP contains the compiled binaries and the setup files I use. If you drop the PANDA distribution RH.RP07.0 in the same directory you should be fine. Just follow the instructions in the PANDA docs to set up TOPS-20 networking and you will have IP connectivity. If you want DECNET you will have to uncomment a line in SYSTEM:7-1-CONFIG.CMD (near the end), add the DECNET address to the "NODE" line (it should read NODE NODNAM XX.XX). Add (or uncomment) these lines to SYSTEM:SYSTEM.CMD: NCP SET CIRCUIT NI-0-0 SERVICE ENABLED NCP SET CIRCUIT NI-0-0 STATE ON With all of this you should have both IP and DECNET running... El 03/12/2012, a les 10:25, sampsa at mac.com va escriure: Jordi, Can I get a copy of that setup? I've got a Pi sitting here doing nothing, a TOPS-20 machine sounds awesome. Sampsa On 3 Dec 2012, at 11:24, Jordi Guillaumes i Pons <jg at jordi.guillaumes.name> wrote: Hello, Please add to the HECnet nodelist: 7.79 BITXT0 TOPS-20(PANDA) - KLH10 - Raspbery Pi I am running this simulator in the same Pi which currently holds 5 more VMs. I will see if it keeps stable, and if it does I will run it permanently. Issues to solve - KLH10 does not idle in the Pi. I'll check the idler device code to see if I can guess what is happening. - I tried to give KLH10 its own bridge device, plugged into the rest of my network via VDE switches and plugs, but it does not work. I have lots of IPBLP bugchecks and IP networking is unusable (weird enough, DECNET and LAT seem to work). It works fine if I use the main host interface in non-exclusive mode (but then I don't get LAT). - KLH10 frontend can't load *.EXE files. - TOPS-10 does not run well. I used a x86 KLH to dump the load code in SAV format and then I can boot TOPS-10, but it hungs after a while. On the other hand, I'm thinking about building a blinkenlights device using the GPIO pins of the Rasp-pi basing it on the actual parallel port blinkenlights panel driven by the "lites" device... :) It could be another nice hobby project... Jordi Guillaumes i Pons jg at jordi.guillaumes.name HECnet: BITXOV::JGUILLAUMES Jordi Guillaumes i Pons jg at jordi.guillaumes.name HECnet: BITXOV::JGUILLAUMES

12 years, 7 months

3B1s and "old school" UNIX, was Re: [HECnet] US Seller with some interesting stuff

by phiber＠phiber.com

I used to use 3B1 and 7300 systems quite a bit. There was a CUI/GUI system called MGR, and they served quite well as UUCP mail-enabled workstations. Motorola 68k based. However, everything from the 3B2 on up was based on the Western Electric 32000-series processors. The Regional Bell's original Operations and Support Systems were largely PDP-11/45 and PDP-11/70 minis. These were transitioned for the most part to AT&T 3B20S ("Simplex" processor) servers towards the late 80's, the internal OSS applications these machines ran being designed by Bellcore. The RT version you're referring to is the AT&T 3B20D ("Duplex processor), which served as the processing core of the AT&T 5ESS (Electronic Switching System), at one time one of the two most powerful digital switching systems in the telephony world. The original real-time Unix developed for this platform (by AT&T) was called DMERT, for Duplex Multi Environment Real Time. This ultimately evolved into the 5E-2000 switch, with the OS evolving into AT&T UNIX RTR (Real Time Reliable). Not DECnet capable. The standard hard-wired networking method was AT&T Datakit VCS (Virtual Circuit System). Telenet X.25 hardware was also a common networking interface. By the way, I'm also from Queens. -Mark On Wed, Nov 28, 2012 at 7:19 AM, Gregg Levine <gregg.drwho8 at gmail.com> wrote: On Wed, Nov 28, 2012 at 9:20 AM, Dave McGuire <mcguire at neurotica.com> wrote: On 11/27/2012 10:25 PM, Boyanich, Alastair wrote: Wasn't there some bitsliced cpu variants ? Of the 7300/3B1? No, not that I'm aware of. They're call 68010-based. I know there was the AT&T UNIX PC which was a 68010 + custom external MMU, but there was also a proprietry cpu version I ran into the trap of in a past life. The "UNIX PC" is the 7300, which is nearly identical to the 3B1. The 3B1 has a slightly larger top cover to accommodate a full-height hard drive, while the 7300 will only hold a half-height drive. They are otherwise identical, using 68010s. I'd sure like to know more about the "proprietary cpu" version you're talking about. -Dave -- Dave McGuire, AK4HZ New Kensington, PA Hello! He's thinking of the WE designed processors that were used in the later units. They were not bit slice but were fabricated using the normal methods. Not surprisingly enough the devices could not even be sold separately. They were ran an appropriately written release of UNIX as native. One of the first applications for them and the later models was in running the first and second generation Electronic Switching Services otherwise known as exchanges. There's a whole article online someplace on the RT extensions that needed to be written out and added to UNIX for that application. ----- Gregg C Levine gregg.drwho8 at gmail.com "This signature fought the Time Wars, time and again."

12 years, 7 months

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

HECnet