Re: [PATCH v2 1/2] mm: Move page struct poisoning to CONFIG_DEBUG_VM_PAGE_INIT_POISON

[Michal Hocko]

On Thu 06-09-18 10:07:51, Dave Hansen wrote:
> On 09/06/2018 09:12 AM, Pasha Tatashin wrote:
> > 
> > I do not want to make this feature less tested. Poisoning memory allowed
> > us to catch corner case bugs like these:
> > 
> > ab1e8d8960b68f54af42b6484b5950bd13a4054b
> > mm: don't allow deferred pages with NEED_PER_CPU_KM
> > 
> > e181ae0c5db9544de9c53239eb22bc012ce75033
> > mm: zero unavailable pages before memmap init
> > 
> > And several more that were fixed by other people.
> 
> Just curious: were these found in the wild, or by a developer doing
> normal development having turned on lots of debug options?

Some of those were 0day AFAIR but my memory is quite dim. Pavel will
know better. The bottom line is, however, that those bugs depend on
strange or unexpected memory configurations or HW which is usually
deployed outside of developers machine pool. So more people have this
enabled the more likely we hit all those strange corner cases nobody
even thought of.

-- 
Michal Hocko
SUSE Labs