[Simh] Simulator lockup when booting RSX-11/M+ V4.6 BL87 baseline system disk with insufficient memory

[Johnny Billquist]

On 2014-01-23 06:27, Christian Gauger-Cosgrove wrote:
> On 22 January 2014 11:30, Mark Pizzolato - Info Comm <Mark at infocomm.com> wrote:
>> Hi Christian,
>>
>> Thanks for this report.
>>
>> This does sound odd.
>>
>> I can't speak to what to expect when you configure a simulator with insufficient resources for the operating system to run, but your observation that Ctrl-E doesn't provide a way back to the "sim>" prompt suggests something strange in the simulator itself.
>>
>> Please create an issue at https://github.com/simh/simh/issues to report and track this problem.
>>
>> I doubt it has anything to do with the particular windows system you are running on, so you don't have to run back home to test.  The configuration file you're running with would be useful.
>>
>> Thanks.
>>
>> - Mark
> Well, actually I didn't expect the RSX-11/S baseline system to require
> more than 256KB of RAM, but I am apparently wrong. (I think it's
> RSX-11/S that is used for the baseline, I'm not sure. Wholly new to
> the world of RSX-11.) Trying it with 384KB of RAM worked fine (it
> didn't crash kill the sim). And the bug is reproducible on my home W7
> system as well.
>
> As a further test; I attempted to bootstrap the baseline system on a
> 256KB system, with the system console directed to a telnet port as
> opposed to running straight from the SIMH executable. While the RSX-11
> baseline still dies, the simulator itself doesn't. (I.e. I can send a
> Ctrl+E and the simulator does halt to the sim> prompt.)

As far as I could see in your initial post, you were booting 
RSX-11M-PLUS, not -11S/M.

There is a huge difference between these systems. -11M will boot fine in 
much less memory, and -11S does not even come with a bootable baseline, 
since you have to tailor it on another host system.

	Johnny

>
>
> I'll open a proper issue/trouble ticket on the GitHub shortly.
>
>
>
> On 22 January 2014 11:34, Johnny Billquist <bqt at softjar.se> wrote:
>> Sounds like a problem in simh, yes.
>>
>> But just to clarify one thing with RSX here. The problem is not about not
>> having enough memory to allocate a checkpoint space. What the error means is
>> that RSX do not have a checkpoint space.
>> Checkpoint space is the RSX term for disk space used to swap out running
>> tasks to, in order to free memory.
>>
>> But yes, the basic problem is that you have too little memory. The startup
>> script will setup a checkpoint space for you, but you don't even make it
>> that far... The basic tasks activated at startup is already requiring more
>> memory than you have.
>>
>>          Johnny
> I'm brand new to the RSX-11 game so I didn't know that. As I mentioned
> in my reply to Mark, I think the baseline install system is RSX-11/S,
> which I thought had very small memory requirements. Plus it seems a
> bit of a troll on the part of the RSX-11 developers, it won't run on
> an 18-bit system, but will run a 22-bit system with just 128KB more
> RAM; compare to RSTS/E 10.1 which works fine on a 256KB system.
>
>
> In any case, this appears to be a problem with SIMH itself, not so
> much with RSX-11 (even if it is a misconfiguration issue that causes
> RSX-11 to crash and show the bug).
>
>
> Regards,
> Christian
>