Re: [PATCH v2] mm: fix missing wake-up event for FSDAX pages

[Muchun Song <songmuchun@bytedance.com>]

On Tue, Jul 05, 2022 at 08:00:42PM -0700, Andrew Morton wrote:
> On Wed, 6 Jul 2022 10:47:32 +0800 Muchun Song <songmuchun@bytedance.com> wrote:
> 
> > > If this wakeup is not one of these, then are there reports from the
> > > softlockup detector?
> > > 
> > > Do we have reports of processes permanently stuck in D state?
> > >
> > 
> > No. The task is in an TASK_INTERRUPTIBLE state (see __fuse_dax_break_layouts). 
> > The hung task reporter only reports D task (TASK_UNINTERRUPTIBLE).
> 
> Thanks, I updated the changelog a bit.
> 
> : FSDAX page refcounts are 1-based, rather than 0-based: if refcount is
> : 1, then the page is freed.  The FSDAX pages can be pinned through GUP,
> : then they will be unpinned via unpin_user_page() using a folio variant
> : to put the page, however, folio variants did not consider this special
> : case, the result will be to miss a wakeup event (like the user of
> : __fuse_dax_break_layouts()).  This results in a task being permanently
> : stuck in TASK_INTERRUPTIBLE state.
> : 
> : Since FSDAX pages are only possibly obtained by GUP users, so fix GUP
> : instead of folio_put() to lower overhead.
> 
> I believe these details are helpful for -stable maintainers who are
> wondering why they were sent stuff.  Also for maintainers of
> downstreeam older kernels who are scratching heads over some user bug
> report, trying to find a patch which might fix it - for this they want
> to see a description of the user-visible effects, for matching with
> that bug report.
>

Thanks Andrew, it's really helpful.