Re: Timing Functionality on MPC555

Hi Randall, Jay, and Alex,

Thank you all for taking some time to reply to my query - I had a lot
to work with today.

I think that using the decrementer would probably work fine - 32 bits
is probably fine for most anything this would be used for.  On the
other hand, I get the impression that it will cause an interrupt,
which isn't really useful for me (nor do I want to bother with
masking it.

The hardware watchpoints seems like a little bit of overkill.  I also
have the problem of not having any sort of scope that could measure
the time difference on those signals.

So, I'm using the TB register.  I managed to figure out how to inline
my assembly code with GNU so that I can read these.  I also like this
cause I could conceively take multiple timings at once (yes, I know
that my code has it's own overhead, but it should be small amount, and
I could estimate it if needed).

I'm still trying to figure out how to verify the frequency of TMBCLK
from system settings (observations suggest the TB is running at 2.5
ticks/microsecond), so if anyone wants to chime in I wouldn't mind (or
else hopefully I'll figure it out Monday).

Thanks again, everyone have a good weekend.

-Eric Castle
Emmeskay, Inc.

--- In [email protected], "Randall Young" <ryoung@n...> wrote:
> > I want to measure how long certain functions are taking in my
code, so
> > I want a way to track time rather accurately (precision on the
order
> > of micro-seconds, hopefully).  I was wondering what I should be
using
> > just to get two values - a starting time, and an ending time.  I
was
> > hoping to read directly from the real-time clock, but that seems
to
> > give garbage as output.
> >
> > Anyone know what I should be looking at?
> 
> I'd suggest using the time base registers.  Takes (2) 32 bit reads
to get
> the whole thing, so the code goes something like :
> 
>     mftbu   r5          # load upper word of time base
>     mftb    r4          # load lower word of time base
>     mftbu   r3          # re-read upper word
>     cmpw    r3,r5       # check that upper word didnt change
>     beq     got_time    # If not, branch
>     mftb    r4          # It changed, have to get new LSW to match
> got_time:               # (result in r3:r4)
> 
> As I recall, the resolution depends on your CPU clock and some
prescaler
> options on the MPC555 ... should be microseconds or better.  You'll
also
> have to enable the time base, which as I recall also enables the
decrementer
> and it's exception.
> 
> Randall Young
> Senior Software Engineer
> NavCom Technology, Inc.
> A John Deere Company





-----------------------------------------------------------
To learn more about Freescale Microcontrollers, please visit
http://www.freescale.com/mcu


 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/MPC500/

<*> To unsubscribe from this group, send an email to:
    [email protected]

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/