[PATCH v5 for-4.13 2/7] mlx5e: don't assume anything on the irq affinity mappings of the device

[Christoph Hellwig]

On Thu, Jun 15, 2017 at 04:33:09PM +0300, Sagi Grimberg wrote:
> mlx5e currently assumes that irq affinity is really spread first
> irq vectors across device home node cpus, with the new generic affinity
> mappings this is no longer the case, hence mlxe should not rely on
> this anymore.

Looks fine, but the explanation sounds a bit short - only spreading
the vectors of the single node sounds rather odd, so there needs to
be an explanation of why this was done before and isn't valid anymore.