HECnet

hecnet@lists.dfupdate.se

6 participants
15552 discussions

by b4＠gewt.net

On Tue, 20 May 2014, Johnny Billquist wrote: pleasant. M (non +) runs great on F11-based systems anyway. I don't know (remember) what I might have said. But depending on different things, running M+ on an 11/23 or 11/24 can be a little painful. But running 11M in general is more painful than M+ anyway, so I would at least recommend trying M+. The baseline of M+ (4.6) doesn't even boot on non-split processors in simh. (That said, without split I/D-space, you'll have preciously little pool space, but that might not be a big issue for you right here.) I think that was why. Yeah. Without supervisor mode and split I/D-space, you will have very little system pool. If you are not running a bunch of things on the system, you should still be able to survive though. Oh, and my TCP/IP will not work on those systems. So long as I can still run DECnet i'll be fine. Johnny -- Cory Smelosky http://gewt.net Personal stuff http://gimme-sympathy.org Projects

10 years, 6 months

RSX-11 setback

by b4＠gewt.net

On Tue, 20 May 2014, Johnny Billquist wrote: I would not think so. If you have a device driver for something that there is no hardware for, the device should just become offline. That's what I was thinking...but for the sake of troubleshooting it stayed on the table. Not sure exactly what you configured, but maybe something wrong in all your answers... :-) I'll grab the log when back at my desk. That could help. :-) https://ghostbin.com/paste/zo8dc What garbage that might be left in memory should not be a problem. Memory always have content... Mmm. right. I'm out of ideas then. Johnny -- Cory Smelosky http://gewt.net Personal stuff http://gimme-sympathy.org Projects

10 years, 6 months

RSX-11 setback

by b4＠gewt.net

On Tue, 20 May 2014, Johnny Billquist wrote: The fact that you managed to boot the baseline system and run a sysgen suggests that things are running pretty well... Yeah. That's what I was thinking. Also possible the file I should've BOOted was named after the name I gave the system. No, the output file is RSX11M.SYS, which should have booted ok unless you actually configured it for hardware you do not have... Huh. I wonder why it failed then... Johnny -- Cory Smelosky http://gewt.net Personal stuff http://gimme-sympathy.org Projects

10 years, 6 months

RSTS/E is really picky

by bqt＠softjar.se

On 2014-05-20 21:40, Dave McGuire wrote: On 05/20/2014 10:55 AM, Johnny Billquist wrote: Back to trying RSX-11M 4.8. That's 100M and doubtful it'll complain. Unless you want to feel the pain (or the joy) of manually defining memory partitions, I'd suggest you to go PLUS :) Can't! No Split I&D. M+ do not need split I/D-space. But you do need 512K of memory. But the 11/23 and 11/24 are supported CPUs. Didn't you advise me against running M+ on an 11/24 for reasons of no split I/D space? It's not that it won't work, it's that it won't be pleasant. M (non +) runs great on F11-based systems anyway. I don't know (remember) what I might have said. But depending on different things, running M+ on an 11/23 or 11/24 can be a little painful. But running 11M in general is more painful than M+ anyway, so I would at least recommend trying M+. (That said, without split I/D-space, you'll have preciously little pool space, but that might not be a big issue for you right here.) I think that was why. Yeah. Without supervisor mode and split I/D-space, you will have very little system pool. If you are not running a bunch of things on the system, you should still be able to survive though. Oh, and my TCP/IP will not work on those systems. Johnny

10 years, 6 months

RSTS/E is really picky

by bqt＠softjar.se

On 2014-05-20 21:39, Dave McGuire wrote: On 05/20/2014 11:04 AM, Johnny Billquist wrote: M+ do not need split I/D-space. But you do need 512K of memory. But the 11/23 and 11/24 are supported CPUs. I have 1M of memory. Sure you're not confusing the 11/23 with the 11/23+? I was poking at 4.6M+'s SYSGEN and it wouldn't accept 11/23 as an answer. The difference between the 11/23 and 11/23+ is the amount of memory. The original 11/23 could not go above 256K... If you have more, then you actually have an 11/23+. In practice, though, the vast majority of systems out there that say "PDP-11/23" (not -PLUS) on the front are in fact capable of 22-bit addressing; you've gotta look at the etch rev of the CPU board. See my other email. Aha. Thanks for the information, Dave. I knew that very few 11/23 systems really were limited to 18 bits, but I didn't know exactly which ones. So even the M8186 boards are mostly 22-bit. Really nice to know. Johnny

10 years, 6 months

RSX-11 setback

by bqt＠softjar.se

On 2014-05-20 20:08, Cory Smelosky wrote: boo [1,54]rsx11m SYSTEM CRASH AT LOCATION 014402 REGISTERS R0=000002 R1=120550 R2=000000 R3=120546 R4=120744 R5=031610 SP=117772 PS=020010 SYSTEM STACK DUMP LOCATION CONTENTS 117772 000004 001344 That is a very weird and unexpected stack pointer. RSX normally have the stack in rather low addresses. Wonder if it's related to the tape controller that isn't actually technically present. I would not think so. If you have a device driver for something that there is no hardware for, the device should just become offline. Not sure exactly what you configured, but maybe something wrong in all your answers... :-) I'll grab the log when back at my desk. That could help. :-) After reboot: boo [1.54]rsx11m MCR -- Task not in system ins $boo boo [1,54]rsx11,\,\m [no output] I seem to be TERRIBLE and SYSGEN. Well, since the initial boot failed, any subsequent boots of that image I would expect to fail equally bad. It could have been unclean memory at some random address. This system also tends to randomly crash on first-time boots, too sometimes. A couple halts and it usually works then. What garbage that might be left in memory should not be a problem. Memory always have content... Johnny

10 years, 6 months

RSX-11 setback

by bqt＠softjar.se

On 2014-05-20 19:52, Cory Smelosky wrote: On Tue, 20 May 2014, Paul_Koning at Dell.com wrote: On May 20, 2014, at 1:47 PM, Cory Smelosky <b4 at gewt.net> wrote: Well, I'm just not having ANY luck with this... ... SYSTEM CRASH AT LOCATION 014402 REGISTERS R0=000002 R1=120550 R2=000000 R3=120546 R4=120744 R5=031610 SP=117772 PS=020010 ... I seem to be TERRIBLE and SYSGEN. Maybe you just have a bad drive, and RSX is unhappy just as RSTS was. That is possible. Although it was after the GEN finished. The fact that you managed to boot the baseline system and run a sysgen suggests that things are running pretty well... Also possible the file I should've BOOted was named after the name I gave the system. No, the output file is RSX11M.SYS, which should have booted ok unless you actually configured it for hardware you do not have... Johnny

10 years, 6 months

RSTS/E is really picky

by mcguire＠neurotica.com

On 05/20/2014 10:55 AM, Johnny Billquist wrote: Back to trying RSX-11M 4.8. That's 100M and doubtful it'll complain. Unless you want to feel the pain (or the joy) of manually defining memory partitions, I'd suggest you to go PLUS :) Can't! No Split I&D. M+ do not need split I/D-space. But you do need 512K of memory. But the 11/23 and 11/24 are supported CPUs. Didn't you advise me against running M+ on an 11/24 for reasons of no split I/D space? It's not that it won't work, it's that it won't be pleasant. M (non +) runs great on F11-based systems anyway. (That said, without split I/D-space, you'll have preciously little pool space, but that might not be a big issue for you right here.) I think that was why. -Dave -- Dave McGuire, AK4HZ New Kensington, PA

10 years, 6 months

RSTS/E is really picky

by mcguire＠neurotica.com

On 05/20/2014 11:04 AM, Johnny Billquist wrote: M+ do not need split I/D-space. But you do need 512K of memory. But the 11/23 and 11/24 are supported CPUs. I have 1M of memory. Sure you're not confusing the 11/23 with the 11/23+? I was poking at 4.6M+'s SYSGEN and it wouldn't accept 11/23 as an answer. The difference between the 11/23 and 11/23+ is the amount of memory. The original 11/23 could not go above 256K... If you have more, then you actually have an 11/23+. In practice, though, the vast majority of systems out there that say "PDP-11/23" (not -PLUS) on the front are in fact capable of 22-bit addressing; you've gotta look at the etch rev of the CPU board. See my other email. -Dave -- Dave McGuire, AK4HZ New Kensington, PA

10 years, 6 months

11/23 confusion, was Re: [HECnet] RSTS/E is really picky

by mcguire＠neurotica.com

On 05/20/2014 11:40 AM, Johnny Billquist wrote: My memory board is third-party. Looks like it's truncating my memory then. ;) Are you sure. Very few 11/23 systems actually exist. Most people really have 11/23+ systems, even if they are not aware of it. It's a dual-height board. I don't think it /can/ be an 11/23+! Doh! So you actually have a KDF11-AA (M8186). That sucks. I'm not sure if those were ever upgraded to 22-bit addressing, or if you need the KDF11-BA (M8189) for that. Oh well, so you might actually be stuck with 11M then. Ah, no. Nearly all dual-wide 11/23 CPUs (KDF11-AA, M8186) support 22-bit addressing. Etch revs A and B did not, but in practice there are very few of these out in the field. I have had probably a hundred KDF11-AA boards go through my hands over the years, and have easily thirty of them now, and have never seen a Rev A or Rev B board, nor have I ever known anyone who has. -Dave -- Dave McGuire, AK4HZ New Kensington, PA

10 years, 6 months

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

HECnet