Re: [PATCH v2 5/6] flow: Add queue pair tracking to flow management

[David Gibson <david@gibson.dropbear.id.au>]

[-- Attachment #1: Type: text/plain, Size: 1434 bytes --]

On Thu, Nov 27, 2025 at 10:17:43AM +0100, Laurent Vivier wrote:
> On 11/26/25 04:41, David Gibson wrote:
> > > +/**
> > > + * flow_qp() - Get the queue pair for a flow
> > > + * @f:		Flow to query (may be NULL)
> > > + *
> > > + * Return: queue pair number for the flow, or 0 if flow is NULL or has no
> > > + *         valid queue pair assignment
> > > + */
> > > +unsigned int flow_qp(const struct flow_common *f)
> > > +{
> > > +	if (f == NULL || f->qpair == FLOW_QPAIR_INVALID)
> > Are there any instances where you actually want to pass a NULL flow to
> > this?
> > 
> > If you're going to return 0 anyway, why not just set f->qpair to 0 by
> > default, rather than using FLOW_QPAIR_INVALID?
> > 
> 
> In fact, in the multithread part I need to know if the queue pair has been set or not.
> I agree that we should avoid this if possible because it takes one slot in
> the array. And as we have 16 qpairs we must use 5 bit (we need 17 values)
> rather than 4 bit. I'm reworking the multithread part, I will see if I can
> remove this.

Ok.  It's not obvious to me why there would be a reason to explicitly
track "unset" rather than setting a default value early.  So, I guess
we'll see in the next spin.

-- 
David Gibson (he or they)	| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you, not the other way
				| around.
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]