Re: [PATCH 3/4] mm/hmm: HMM should have a callback before MM is destroyed

[Jerome Glisse]

On Tue, Mar 20, 2018 at 12:33:26PM +0100, Michal Hocko wrote:
> On Thu 15-03-18 15:48:29, Andrew Morton wrote:
> > On Thu, 15 Mar 2018 14:36:59 -0400 jglisse@xxxxxxxxxx wrote:
> > 
> > > From: Ralph Campbell <rcampbell@xxxxxxxxxx>
> > > 
> > > The hmm_mirror_register() function registers a callback for when
> > > the CPU pagetable is modified. Normally, the device driver will
> > > call hmm_mirror_unregister() when the process using the device is
> > > finished. However, if the process exits uncleanly, the struct_mm
> > > can be destroyed with no warning to the device driver.
> > 
> > The changelog doesn't tell us what the runtime effects of the bug are. 
> > This makes it hard for me to answer the "did Jerome consider doing
> > cc:stable" question.
> 
> There is no upstream user of this code IIRC, so does it make sense to
> mark anything for stable trees?

I am fine with dropping stable, distribution that care about out of tree
drivers can easily backport themself. I am hoping to have the nouveau
part upstream in 4.18/4.19 ...

Cheers,
Jérôme