On Tuesday 19 Jun 2018 at 13:31:06 (+0200), Peter Zijlstra wrote:
> On Mon, May 21, 2018 at 03:24:58PM +0100, Quentin Perret wrote:
> > + read_lock_irqsave(&em_data_lock, flags);
> > + for_each_cpu(cpu, cpu_possible_mask) {
>
> I know we're likely to only use this on small systems, but this pattern
> is a very bad, Please look at alternatives.Ok, this isn't supposed to be called very often (only once, at boot time, for Arm for example), but I see your point.
