[vmips] Linux BogoMIPS

[Brian R. Gaeke]

And then spake Cable Guy, as follows:
> I see the same 1.05 BogoMIPS value you had.  This is true no matter
> what hardware I run VMIPS on.  I've run it on a pokey old PII 400 MHz
> and a new P4 3.2 GHz box and the BoboMIPS number is the same.  Is this
> due to something hard-coded in the RTC part of VMIPS?  In real time it
> seems to run about 3 times faster on the P4, but I expected it to be
> faster than that compared to the P2.

I'll attempt to respond to the rest of your emails maybe next
week (sorry, it's finals week here at UIUC, and I'm swamped!)

The short answer to this part is, unless "-o realtime" is turned on
-- at a significant performance penalty, currently -- the passage of
time in VMIPS is fixed to a certain number N of nanoseconds per
instruction. You can change this number using "-o clockspeed=N". The
downside is that this may have no correlation to how fast your machine
actually runs VMIPS.

For more information, check out the descriptions of the "clockspeed",
"realtime" and "timeratio" options in the "Customizing" section of the
VMIPS manual:

	http://www.dgate.org/vmips/doc/vmips.html#Customizing

-Brian

-- 
Brian R. Gaeke, brg at dgate.org -- GnuPG encrypted mail gleefully accepted