[Simh] Excessive paging of the real machine running SIMH - VAX

[Villy Madsen]

On my system
it always happens on the system drive at startup - right after the bit map message ....

most of the time it also happens once the data drive is mounted  and if not, it'll happen as soon as anything writes to the data drive.

some times it happens when the data drive and/or the system drive is dismounted - the console hangs at the vms message "use console to halt the cpu (or some such wording)" until all of the W/XP io is complete

Villy

----- Original Message -----
From: "Hittner, David T." <david.hittner at ngc.com>
Date: Tuesday, July 17, 2007 6:15 am
Subject: Re: [Simh] Excessive paging of the real machine running SIMH - VAX
To: Richard Davis <richard_a_davis_2 at btinternet.com>, simh at trailing-edge.com

> Perhaps _flag was never initialized correctly, and the default
> (uninitialized) values sometimes work and sometimes don't.
>  
> Dave
> 
> ________________________________
> 
> From: simh-bounces at trailing-edge.com
> [mailto:simh-bounces at trailing-edge.com] On Behalf Of Richard Davis
> Sent: Tuesday, July 17, 2007 8:12 AM
> To: simh at trailing-edge.com
> Subject: [Simh] Excessive paging of the real machine running 
> SIMH - VAX
> 
> 
> 
> Villy,
> 
>  
> 
> The issue only occurs on my system when first starting the 
> system, i.e.
> when the disks are first mounted, my disks are 1Gb and 2Gb
> respectively.. 
> 
> A stack trace of the executable when the problem occurs shows the
> following call trace:
> 
>  
> 
> sim_process_event
> 
> rq_svc
> 
> sim_fseek
> 
> fsetpos
> 
> _fseeki64
> 
> _fseeki64_nolock
> 
> _flush
> 
> _write
> 
>  
> 
> As mentioned before, the strange value in the _flag member of 
> the stream
> causes the _flush routine to do a complete flush of the disk file.
> 
>  
> 
> I modified the SIMH sim_fopen routine to use the _fsopen call (rather
> than fopen) which allowed me to deny all sharing.. the problem still
> occurs, so It's nothing to do with sharing!
> 
>  
> 
> I need to understand the workings of SIMH a bit better, and need 
> to find
> out how things are put onto the event queue (the _flag member is 
> already'corrupted' when the event gets processed in the 
> sim_process_eventroutine.
> 
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.trailing-edge.com/pipermail/simh/attachments/20070717/e90a0ffd/attachment-0003.html>