MBUF Cluster Network freeze problem
joel.sherrill at OARcorp.com
Fri May 4 14:25:23 UTC 2001
Antti P Miettinen wrote:
> Have you tried playing with rtems_bsdnet_config.network_task_priority?
> What is the rationale of not having time slicing for the networking
> tasks? I know they should not stay busy for long periods so time
> slicing should not be _needed_ but if there _are_ several tasks
> competing for CPU I think the reasonable thing to do would be to share
> the CPU between them.
Timeslicing breaks the atomicity assumptions of the network stack.
The stack was designed to run in non-interruptable UNIX kernel space
so the tasks are non-preempt, non-timesliced.
Timeslicing implies that preemption is enabled. Otherwise, your
timeslice expires and the kernel is not allowed to eject the task.
> Is there a performance penalty for time slicing?
Except for adding context switches at the end of time slices, it
is not measurable.
> Or does the correct operation of the stack depend on network tasks not
> pre-empting each other?
> Antti P Miettinen
> Nokia Networks
Joel Sherrill, Ph.D. Director of Research & Development
joel at OARcorp.com On-Line Applications Research
Ask me about RTEMS: a free RTOS Huntsville AL 35805
Support Available (256) 722-9985
More information about the users