MBUF Cluster Network freeze problem

[Eric Norum]

Antti P Miettinen <antti.p.miettinen at nokia.com> wrote:
>
> Yes - I'm aware of the network semaphore so I thought that there
> should not be any need for the RTEMS_NO_TIMESLICE flag for task
> creation. Is this just a leftover or a safeguard or really required?
>
> I dunno if this is in any way relevant in real life but I think that
> in principle time slicing should be allowed. Allowing time slicing
> might have an effect in some pathological situation.

What good would time-slicing do?  If  a task holds the network semaphore no  
other task can obtain the semaphore, so there would be no point in trying to  
run other network-related tasks by enabling time slice scheduling for the  
network tasks.
If you want to experiment you could try releasing the network semaphore,  
then yielding, then obtaining the semaphore again in the task which you  
suspect is hogging the CPU.  This of course adds overhead, but if you're  
running out of resources it might be worth a try.

---
Eric Norum
<e.norum at sk.sympatico.ca>
(306) 373-5820