> > For me struct dp_netdev_pmd_thread is *the* the PMD thread context in
> dpif-netdev. I don't really see the benefit of creating a sub-struct
> dp_netdev_pmd_thread_ctx and moving certain data into that. Can you
> explain your motivation?
> 
> Hello Jan,
> 
> IMHO all other fields in struct dp_netdev_pmd_thread has direct relation
> with the
> thread itself (like core_id, need_reload) or they are it's direct accessories
> (like mutexes, stats, caches, lists of polled ports). From the other hand,
> time and
> last_cycles has no actual relation with thread and can be moved out of
> struct
> dp_netdev_pmd_thread wihtout any logical issues. These fields are
> characteristics
> of the outside environment. For example, 'emc_insert_min' which I'm
> introducing in
> the next patch is actually the characteristics of the last received packet
> (maybe port)
> but not the thread itself.
> 

I see your point. But I am afraid that such classification will continue to be 
fairly subjective and there is little technical or readability gain, if any, by 
doing this. Or are we going to pass/copy/zero the entire sub-struct outside the 
context of the pmd struct at some point? Perhaps it would be enough to group 
these fields in struct dp_netdev_pmd_thread and head the section with a comment?

BR, Jan
_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to