[dpdk-dev] [RFC] Service Cores concept

[Wiles, Keith]

> On May 16, 2017, at 3:11 PM, Thomas Monjalon <thomas at monjalon.net> wrote:
> 
> 03/05/2017 13:29, Harry van Haaren:
>> The concept is to allow a software function register itself with EAL as
>> a "service", which requires CPU time to perform its duties. Multiple
>> services can be registered in an application, if more than one service
>> exists. The application can retrieve a list of services, and decide how
>> many "service cores" to use. The number of service cores is removed
>> from the application usage, and they are mapped to services based on
>> an application supplied coremask.
>> 
>> The application now continues as normal, without having to manually
>> schedule and implement arbitration of CPU time for the SW services.
> 
> I think it should not be the DPDK responsibility to schedule threads.
> The mainloops and scheduling are application design choices.
> 
> If I understand well the idea of your proposal, it is a helper for
> the application to configure the thread scheduling of known services.
> So I think we could add interrupt processing and other thread creations
> in this concept.
> Could we also embed the rte_eal_mp_remote_launch() calls in this concept?

I did not really see the RFC as replacing the current design in DPDK for thread handling, but that maybe just me. What I saw was a design to create and destroy threads more dynamically or possible some type of thread pool design.

DPDK needs a thread handler anyway (as we have one now) and the current one seems to work, but it does not allow for someone to replace the thread handling with something different. What I would suggest is figure out how to pull DPDK current thread handling out into a plugin design, then someone can replace that plugin with his own thread handling scheme.

We need to keep DPDK stable, so adding the plugin support and pulling out the current design needs to be transparent to all current applications using the current APIs and methods.

Regards,
Keith