Messages in this thread | | | Date | Wed, 10 Dec 2014 20:23:40 +0100 | From | David Hildenbrand <> | Subject | Re: [PATCH v4] CPU hotplug: active_writer not woken up in some cases - deadlock |
| |
only thing that is bugging me is this part. Without the lock we can't > > guarantee that another get_online_cpus() just arrived and bumped the refcount > > to 0. > > > > Of course this only applies to misuse of put/get_online_cpus. > > > > We could hack some loop that tries to cmp_xchng with the old value until it > > fits to work around this, but wouldn't make the code any better readable. > > Well, we didn't have this diagnostic in the original, so one approach > would be to simply leave it out. Another approach would be to just > have the WARN_ON() without the attempt to fix it up. > > Thanx, Paul
Well, I think the WARN_ON would be sufficient for debugging purposes.
Thanks!
David
| |