[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