On Fri, 2021-07-09 at 21:28 +0200, Thomas Gleixner wrote:
> On Fri, Jul 09 2021 at 07:21, Mike Galbraith wrote:
> > static bool has_cpu_slab(int cpu, struct kmem_cache *s)
> > @@ -2632,8 +2636,11 @@ static int slub_cpu_dead(unsigned int cp
> > struct kmem_cache *s;
> >
> > mutex_lock(&slab_mutex);
> > - list_for_each_entry(s, &slab_caches, list)
> > + list_for_each_entry(s, &slab_caches, list) {
> > + local_lock(&s->cpu_slab->lock);
>
> This one is odd. It locks the cpu_slab lock of the CPU which runs
> thiscallback and then flushes the slab of the dead CPU.
That spot I put back only because it used to exist, ergo I documented
it as an afterthought. Yeah, it clearly has nada to do with the
explosions, those were stopped by the other two as previously reported.
-Mike
