Hi Shally,

> [Shally] That's the intention but until then I assume we agree to current 
> proposal?!
> If yes, then just to align and close on our discussion on this here’s the 
> summary :
> - There's no differentiation to device capability and qp capability.  If PMD 
> shows in its feature flag that it
> supports both sym and asym then it must support those on all its qps.
> - if PMD support both but internally has hw with dedicated qp for each 
> service then it *can* split itself
> into two: symmetric only and asymmetric only PMD instances.
> - Currently we don’t see a requirement to dedicate qp to a service, thus no 
> need for
> max_nb_sym/asym_qp or service_type info during qp_setup. However this is open 
> for review and
> discussion and can be added later, if any requirement is identified for same.
> Is that correct?
[Fiona] Yes.

Reply via email to