On Tue, Mar 26, 2019 at 10:22:40PM +0100, Thomas Gleixner wrote:
> On Wed, 27 Feb 2019, Ricardo Neri wrote:
> >
> > -struct cpumask watchdog_allowed_mask __read_mostly;
> > +static struct cpumask watchdog_allowed_mask __read_mostly;
>
> That hunk is correct.
I'll send a separate patch with this hunk only.
>
> > struct cpumask watchdog_cpumask __read_mostly;
> > unsigned long *watchdog_cpumask_bits = cpumask_bits(&watchdog_cpumask);
> > @@ -92,6 +92,11 @@ static int __init
> > hardlockup_all_cpu_backtrace_setup(char *str)
> > }
> > __setup("hardlockup_all_cpu_backtrace=",
> > hardlockup_all_cpu_backtrace_setup);
> > # endif /* CONFIG_SMP */
> > +
> > +struct cpumask *watchdog_get_allowed_cpumask(void)
> > +{
> > + return &watchdog_allowed_mask;
> > +}
>
> That part is pointless as I showed you in the other reply. You don't need
> that mask in your code at all.
Yes. I'll remove this hunk and follow the approach you suggested in your
reply to patch 11.
Thanks and BR,
Ricardo
>
> Thanks,
>
> tglx
>
>
