Results 1 to 1 of 1
Hello all, I'm trying to compare the latencies of vanilla Linux, real-time patched Linux, Xenomai and RTAI, for an assignment. I have installed all of them on top of a ...
Enjoy an ad free experience by logging in. Not a member yet? Register.
- 12-16-2010 #1
- Join Date
- Dec 2010
Xenomai, RTAI and Cyclictest.
Hello all, I'm trying to compare the latencies of vanilla Linux, real-time patched Linux, Xenomai and RTAI, for an assignment. I have installed all of them on top of a Slackware 13.1 installation. To measure the latencies I use cyclictest on everything except RTAI.
My main question is if there is a way to run Cyclictest on RTAI, that would help me alot... When I asked this on the RTAI mailing list they said that the latency benchmark provided in the RTAI testsuite does the same thing as Cyclictest. When I run cyclictest (not on RTAI) I get an output like the first attachment and when I run the latency benchmark on RTAI I get an output like the second attachment
What I don't understand is how can latency be negative and why the maximum latency is so much bigger on RTAI when it's supposed to have better performance than the other real-time platforms. So how can I compare the two outputs correctly and come to a valid conclusion?
As you probably understand I'm a noob at this real-time stuff but I have this assignment that must be done and all documentation I have found is written from experts and targeted at experts... Any help would be very appreciated.