Re: [PATCH RFC 1/3] cpumask: Honor irq_default_affinity in cpumask_local_spread()

On Thu, 2026-08-20 at 18:09 +0200, Sebastian Andrzej Siewior wrote:
> On 2026-08-19 14:38:22 [-0400], Yury Norov wrote:
> > > --- a/lib/cpumask.c
> > > +++ b/lib/cpumask.c
> > > @@ -81,8 +82,9 @@ void __init free_bootmem_cpumask_var(cpumask_var_t mask)
> > >   * @i: index number
> > >   * @node: local numa_node
> > >   *
> > > - * Return: online CPU according to a numa aware policy; local cpus are returned
> > > - * first, followed by non-local ones, then it wraps around.
> > > + * Return: online CPU according to the default IRQ affinity and a numa aware
> > > + * policy; local cpus are returned first, followed by non-local ones, then it
> > > + * wraps around.
> > >   *
> > >   * For those who wants to enumerate all CPUs based on their NUMA distances,
> > >   * i.e. call this function in a loop, like:
> > > @@ -110,9 +112,9 @@ unsigned int cpumask_local_spread(unsigned int i, int node)
> > 
> > Please don't touch this function. There's ~40 users, and we don't want
> > to inspect every caller for their intention.
> 
> While this might be okay for the NICs, the usage of PMUs shouldn't be
> changed. So yes, Yury's suggestion to create your own function and then
> move the drivers one by one makes sense.
> You might also want to verify that your setup/ design works on some
> "newer" multi queue NICs even after the number of queues is reconfigured
> at runtime.
> 
> There is also netif_set_affinity_auto().
> 

Thanks for this pointer. 4 users and yet another implementation that
ignores the current userspace settings...

I'll put this on my "to be revisited" list. Thanks!