Re: [PATCH v1 1/5] lib/test_cpumask: drop cpu_possible_mask full test

[David Gow <davidgow@google.com>]

On Wed, Aug 10, 2022 at 4:45 PM Sander Vanheule <sander@svanheule.net> wrote:
>
> Hi David,
>
> On Wed, 2022-08-10 at 12:06 +0800, David Gow wrote:
> > On Wed, Aug 10, 2022 at 2:09 AM Sander Vanheule <sander@svanheule.net> wrote:
> > >
> > > cpu_possible_mask is not necessarily completely filled.  That means
> > > running a check on cpumask_full() doesn't make sense, so drop the test.
> > >
> > > Fixes: c41e8866c28c ("lib/test: introduce cpumask KUnit test suite")
> > > Link: https://lore.kernel.org/lkml/346cb279-8e75-24b0-7d12-9803f2b41c73@riseup.net/
> > > Reported-by: Maíra Canal <mairacanal@riseup.net>
> > > Signed-off-by: Sander Vanheule <sander@svanheule.net>
> > > Cc: David Gow <davidgow@google.com>
> > > ---
> >
> > Looks good to me. It'd maybe be worth noting _why_  cpu_possible_mask
> > is not always filled (i.e., that the number of available CPUs might
> > not match the maximum number of CPUs the kernel is built to support),
> > but it's probably not worth doing a new version of the patch series
> > just for that.
> >
> > Reviewed-by: David Gow <davidgow@google.com>
>
> Thanks for the reviews!
>
> Perhaps the commit message could be replaced by:
>
> "When the number of CPUs that can possibly be brought online is known at boot time, e.g. when
> HOTPLUG is disabled, nr_cpu_ids may be smaller than NR_CPUS. In that case, cpu_possible_mask would
> not be completely filled, and cpumask_full(cpu_possible_mask) may return false for valid system
> configurations."
>

Sounds good to me! Thanks!

-- David