Re: [dm-devel] [PATCH v5] fault-injection: introduce kvmalloc fallback options

From: Mikulas Patocka
Date: Fri Apr 27 2018 - 06:20:29 EST

On Fri, 27 Apr 2018, Michal Hocko wrote:

> On Thu 26-04-18 18:52:05, Mikulas Patocka wrote:
> >
> >
> > On Fri, 27 Apr 2018, Michael S. Tsirkin wrote:
> [...]
> > > But assuming it's important to control this kind of
> > > fault injection to be controlled from
> > > a dedicated menuconfig option, why not the rest of
> > > faults?
> >
> > The injected faults cause damage to the user, so there's no point to
> > enable them by default. vmalloc fallback should not cause any damage
> > (assuming that the code is correctly written).
> But you want to find those bugs which would BUG_ON easier, so there is a
> risk of harm IIUC

Yes, I want to harm them, but I only want to harm the users using the
debugging kernel. Testers should be "harmed" by crashes - so that the
users of production kernels are harmed less.

If someone hits this, he should report it, use the kernel parameter to
turn it off and continue with the testing.

> and this is not much different than other fault injecting paths.

Fault injections causes misbehavior even on completely bug-free code (for
example, syscalls randomly returning -ENOMEM). This won't cause
misbehavior on bug-free code.