Re: [PATCH v6] numa: make node_to_cpumask_map() NUMA_NO_NODE aware

[Michal Hocko]

On Wed 25-09-19 12:40:40, Peter Zijlstra wrote:
> On Tue, Sep 24, 2019 at 03:19:39PM +0200, Michal Hocko wrote:
> 
> > > The below would get rid of the PMU and workqueue warnings with no
> > > side-effects (the device isn't used for anything except sysfs).
> > 
> > Hardcoding to 0 is simply wrong, if the node0 is cpuless for example...
> 
> It doesn't matter.... that 0 is _never_ used. These are fake devices,
> and all we care about is getting rid of that error.

That is a very subtle and hard to review assumption. Even if this holds
now a future change might easily break this AFAIU. It also assumes that
you catch all such special devices.

I am sorry but I still do not understand why you consider this whack a
mole better then simply live with the fact that NUMA_NO_NODE is a
reality and that using the full cpu mask is a reasonable answer to that.
Anyway, I feel we are loop here so I will leave out the final decision
to you.

> If it makes you feel better we can make it -2 and have dev_to_node()
> WARN if it ever sees one.

That would help

-- 
Michal Hocko
SUSE Labs