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

Jason Stevens neozeed at gmail.com
Wed Sep 24 20:45:58 EDT 2008


While far from the way 3.7-1 performs I did the following changes in
vax_cpu.c:
I changed the following two lines to read as follows

uint32 cpu_idle_ipl_mask = 0x8;                         /* idle if on IPL 3
*/
int32 cpu_idle_wait = 1000;                             /* for these cycles
*/

to
uint32 cpu_idle_ipl_mask = 0xB;                         /* idle if on IPL 3
*/
int32 cpu_idle_wait = 200;                             /* for these cycles
*/


The I altered the idle table

static struct os_idle os_tab[] = {
    { "VMS", 0x8 },
    { "NETBSD", 0x2 },
    { "ULTRIX", 0x2 },
    { "OPENBSD", 0x1 },
    { "32V", 0x1 },
    { NULL, 0 }
    };

I changed VMS back to being 0xB

I've now been running the VMS image from Gerry for a few minutes with it
bouncing between 40% and 2% utilization.. but it's keeping on the lowside.

    { "VMS", 0x8 },

I suppose as a fix there could be a LEGACYVMS that could set the IPL mask
back to 0xB and the idle timer to 200ms...

Anyways I hate to say it, but I'm a VMS n00b I only read email on it for
one semester then we got an RS/6000 and access to some Vaxen running
Ultrix....

I had to even look around for this gem:

@SYS$SYSTEM:SHUTDOWN


I'm not in the mood to fight with edt or FORTRAN... assuming those are in
the image... :)

Let me know if this IPL revert helped.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.trailing-edge.com/pipermail/simh/attachments/20080924/38709325/attachment-0003.html>


More information about the Simh mailing list