Re: [PATCH 2/2] net/mlx5e: Leverage sched_numa_hop_mask()

From: Jakub Kicinski <kuba@kernel.org>
To: Tariq Toukan <ttoukan.linux@gmail.com>
Cc: Valentin Schneider <vschneid@redhat.com>,
	netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
	Tariq Toukan <tariqt@nvidia.com>,
	"David S. Miller" <davem@davemloft.net>,
	Saeed Mahameed <saeedm@nvidia.com>,
	Ingo Molnar <mingo@redhat.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Juri Lelli <juri.lelli@redhat.com>,
	Eric Dumazet <edumazet@google.com>,
	Paolo Abeni <pabeni@redhat.com>, Gal Pressman <gal@nvidia.com>,
	Vincent Guittot <vincent.guittot@linaro.org>
Subject: Re: [PATCH 2/2] net/mlx5e: Leverage sched_numa_hop_mask()
Date: Wed, 10 Aug 2022 10:42:09 -0700	[thread overview]
Message-ID: <20220810104209.36961cc1@kernel.org> (raw)
In-Reply-To: <8448dade-a64a-0b6b-1ed0-dd164917eedf@gmail.com>

On Wed, 10 Aug 2022 15:57:33 +0300 Tariq Toukan wrote:
> > +		for_each_cpu(cpu, mask) {
> > +			cpus[i] = cpu;
> > +			if (++i == ncomp_eqs)
> > +				goto spread_done;
> > +		}
> > +	}
> > +spread_done:
> > +	rcu_read_unlock();
> >   	ret = mlx5_irqs_request_vectors(dev, cpus, ncomp_eqs, table->comp_irqs);
> >   	kfree(cpus);
> >   	if (ret < 0)  
> 
> This logic is typical. Other drivers would also want to use it.
> It must be introduced as a service/API function, if not by the sched 
> topology, then at least by the networking subsystem.
> Jakub, WDYT?

Agreed, no preference where the helper would live tho.