"md5 -t" as indicator of CPU speed

[marius]

On 6/16/06, Peter Hessler <phessler at theapt.org> wrote:
> It depends on a few things.  Do you have any flags for apmd on start
> up?  Did you run apm with any flags?  Its also entirely possible that
> your system cannot handle speed changes (Support for *many* cpus was
> added recently to -current).  A full dmesg would tell us that.

Hi Peter,

I didn't include a dmesg or my test results on purpose. I was just
curious about the reliability of "md5 -t" in relatively comparing CPU
speeds. I want to get a little further along in my investigation
before asking questions about apm.

For those thaty are interested, I am running -current. I tested
Dimitri's latest SpeedStep patch which did not help even though my
system is supposed to support it (the dmesg is at
http://marc.theaimsgroup.com/?l=openbsd-tech&m=115005130301103&w=2 ).

To make a long story short, Dimitri and miod@ gave me a few pointers
(off-list) as to where to look, and I'm trying to narrow down
specifically where the problem is. As far as I can tell ichpcib(4) on
my laptop is trying to change the powerstate depending on the system
load (the state variable toggles between 0 and 1 depending on whether
setperf is over 50), and although the reported hw.cpuspeed goes down
when the state changes to 1, the speed does not appear to really
change (and as I mentioned before, hw.cpuspeed does not go up when
state changes to 0).

> I ran `md5 -t` with automatic mode disabled, and manually triggered
> each CPU speed jump.  Here are my results.
>
> # apm -L (hw.setperf=0)
> Time   = 1.095103 seconds
> Time   = 1.096470 seconds
> Time   = 1.097863 seconds
> Time   = 1.094756 seconds
> Time   = 1.101693 seconds
> # apm -H (hw.setperf=100)
> Time   = 0.430246 seconds
> Time   = 0.430267 seconds
> Time   = 0.430384 seconds
> Time   = 0.433962 seconds
> Time   = 0.432229 seconds
>
>
> That indicates that `md5 -t` is a decent way to compare cpu speed.  And
> by 'decent' I mean "just as bad as the others".

This is all I needed to know. Unfortunately neither of those apm
commands appears to make a difference on the "md5 -t" test for me. Now
I can get back to wrestling with ichpcib.c and figure out why.

Thanks, Marius