[MAGEEC] [BEEBS] Running Measurements

George Field george.field at bristol.ac.uk
Tue Jul 29 12:33:51 BST 2014


Update - fixed the issue of no measurements. That was a product of my own
stupidity, and a wiring problem. Make check is now running, although if
someone has a shrimp setup, can you tell me if the following measurements
look sensible?

name,energy,time,avg_power,avg_current,avg_voltage
> 2dfir,0.005749,0.062321,0.092251,0.017585,5.246109
> aha-compress,0.005654,0.062333,0.090698,0.017273,5.250885
> aha-mont64,0.005466,0.062333,0.087688,0.016708,5.248163
> blowfish,0.005539,0.062321,0.088882,0.016920,5.252919
>

Thanks,
George


On 29 July 2014 11:54, George Field <george.field at bristol.ac.uk> wrote:

> Hi all,
>
> I'm currently struggling getting DejaGNU to execute the benchmarks in
> BEEBS (currently branching from the latest beebsv2, 9b298d6) on a shrimp.
> I have platformrun working, albeit with errors about multiple measurements,
> and thus rather small energy results. Make check however appears to be
> timing out and giving 0 readings for everything.
>
> I'm invoking it as:
> AVRDUDE='/usr/bin/avrdude' AVRDUDE_FLAGS='-carduino -patmega328p
> -P/dev/ttyUSB0 -D' ENERGYTOOL_SERIAL='MSP0' ENERGYTOOL_PIN='PA0'
> ENERGYTOOL_POINT='2' TIMEOUT=120 ENERGY_DATA='tmpenergy.csv' make
> RUNTESTFLAGS="--target_board=avr-avrdude measure.exp" check
>
> I am seeing a "WARNING: Couldn't find tool init file" when running the
> above, just prior to the tests beginning. Could this be related?
>
> I've also switched to using a breadboard based shrimp, to rule out my
> admittedly shoddy soldering.
>
> Any help would be appreciated!
>
> George
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mageec.org/pipermail/mageec/attachments/20140729/0de5fd29/attachment.html>


More information about the mageec mailing list