Re: [PATCH 2/4] RDMA/core: Introduce shared CQ pool API

[Leon Romanovsky <leon@kernel.org>]

On Mon, May 11, 2020 at 03:08:45PM +0300, Yamin Friedman wrote:
>
> On 5/11/2020 8:07 AM, Leon Romanovsky wrote:
> > On Sun, May 10, 2020 at 05:55:55PM +0300, Yamin Friedman wrote:
> > > Allow a ULP to ask the core to provide a completion queue based on a
> > > least-used search on a per-device CQ pools. The device CQ pools grow in a
> > > lazy fashion when more CQs are requested.

<...>

> > > +	while (!found) {
> > > +		spin_lock_irqsave(&dev->cq_pools_lock, flags);
> > > +		list_for_each_entry(cq, &dev->cq_pools[poll_ctx - 1],
> > > +				    pool_entry) {
> > > +			if (vector != cq->comp_vector)
> > > +				continue;
> > > +			if (cq->cqe_used + nr_cqe > cq->cqe)
> > > +				continue;
> > > +			if (found && cq->cqe_used >= found->cqe_used)
> > > +				continue;
> > > +			found = cq;
> > Don't you need to break from this loop at some point of time?
> In order to find the emptiest cq it loops over the entire list. If you think
> it is better just to find the first one with enough space I will break once
> one is found.

Does it matter to find emptiest CQ as long as you find one with enough cqe?
If yes, please document it, if no, this loop can be simplified.

Thanks