[Simh] Idle problem with VAX/VMS V4.7 on Simh V3.8-0

[Bill Gallagher]

The older versions of the VMS internals books have now appeaered on bitsavers.org.

http://www.mirrorservice.org/sites/www.bitsavers.org/pdf/dec/vax/vms/training/

Might help to find how it idles.

Regards,

B

> -----Original Message-----
> From: simh-bounces at trailing-edge.com
> [mailto:simh-bounces at trailing-edge.com]On Behalf Of HoffmanLabs
> Sent: 20 September 2008 18:23
> To: simh at trailing-edge.com
> Subject: Re: [Simh] Idle problem with VAX/VMS V4.7 on Simh V3.8-0
> 
> 
> gerry77 wrote:
> > I'm a member of the Italian Hobbyist DECnet Network [1] and 
> we are having some
> > problems running VAX/VMS V4.7 on Simh V3.8-0 owing to the 
> idle function being
> > not effective, maybe because of some quite big internal 
> engineering changes
> > among VAX/VMS releases before and after the V5.0 milestone.
> 
> The modular executive at V5 moved the idle loop from a 
> process context 
> to the system context, IIRC.
> 
> Try running a zero-priority (detached) process that's in the 
> instruction 
> size range to trigger the idle processing.  Possibly a 
> detached process 
> that's started up each time OpenVMS VAX boots. 
> 
> I'd probably start with some priority zero process that runs with a 
> deliberate infinite loop:
> 
> 10$: brb 10$
> 
> This as a potential workaround.  This would usurp the idle loop and 
> might well fool the simh idle loop detection logic into working.
> 
> If the workaround fails, then some idea of the size of the 
> idle loop on 
> V4 would have to be migrated over to simh to get the idle 
> loop detection 
> logic to work, or some idea of what the loop is doing that 
> confuses the 
> logic. 
> 
> Another option might be the introduction of a (patch) of a wait 
> instruction or XFC call or another notification or trap into the idle 
> loop to identify the idle loop and to trigger the transition 
> back to the 
> emulator, and a change to the emulator that fields this trap.   The 
> PDP-11 had something like this; there was a WAIT wait for interrupt 
> instruction, IIRC.  The idle loop patch would be easiest with 
> the fiche, 
> if that's around somewhere.  It's not nearly as modular as 
> updating the 
> logic in simh, though. 
> 
> 
> 
> _______________________________________________
> Simh mailing list
> Simh at trailing-edge.com
> http://mailman.trailing-edge.com/mailman/listinfo/simh
>