[LITMUS^RT] "could not become RT task: Invalid argument" on C-EDF Plugin

[Glenn Elliott]

It's not that you have to specify a cluster.  You have to specify a CPU in the cluster that you want.  It's a bit kludgy.


On Aug 30, 2012, at 6:29 PM, Jeremy Erickson <jerickso at cs.unc.edu> wrote:

> Glenn pointed out to me offline that you have to specify the cluster for C-EDF.  Apparently this is the case even if there's only one cluster, which does make sense.  I should have thought of that, but the fact that it didn't consistently fail threw me off.
> 
> When I add the "-p" stuff to the rtspin commands, it works fine.
> 
> -Jeremy Erickson
> 
> On Thu, Aug 30, 2012 at 4:45 PM, Jeremy Erickson <jerickso at cs.unc.edu> wrote:
> I discovered what appears to be a bug in the C-EDF plugin, unless it is a bug in Chris's qcow2 image.  Even in 2012.2, when I run the attached script, I often get the error "could not become RT task: Invalid argument" for a random number of the tasks - sometimes just one, sometimes two, sometimes all three, sometimes none.  I reverted to Chris's original image and continued to reproduce the error, although I did not see it happen on my VirtualBox image on my MacBook Air.
> 
> Has the C-EDF plugin been maintained recently?  Can anyone reproduce this bug on a different machine?  Thanks!
> 
> In order to run the script, put it in /root, along with liblitmus and feather-trace-tools directories with compiled binaries.  Alternatively, you can modify the variable declarations at the top.
> 
> -Jeremy Erickson
> 
> _______________________________________________
> litmus-dev mailing list
> litmus-dev at lists.litmus-rt.org
> https://lists.litmus-rt.org/listinfo/litmus-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.litmus-rt.org/pipermail/litmus-dev/attachments/20120830/b29a0f4c/attachment.html>