Re: [RFCv2 08/16] KVM: Use GUP instead of copy_from/to_user() to access guest memory

[John Hubbard]

On 10/25/20 9:21 PM, Matthew Wilcox wrote:
> On Thu, Oct 22, 2020 at 12:58:14PM -0700, John Hubbard wrote:
> > On 10/22/20 4:49 AM, Matthew Wilcox wrote:
> > > On Tue, Oct 20, 2020 at 01:25:59AM -0700, John Hubbard wrote:
> > > > Should copy_to_guest() use pin_user_pages_unlocked() instead of gup_unlocked?
> > > > We wrote a  "Case 5" in Documentation/core-api/pin_user_pages.rst, just for this
> > > > situation, I think:
> > > >
> > > >
> > > > CASE 5: Pinning in order to write to the data within the page
> > > > -------------------------------------------------------------
> > > > Even though neither DMA nor Direct IO is involved, just a simple case of "pin,
> > > > write to a page's data, unpin" can cause a problem. Case 5 may be considered a
> > > > superset of Case 1, plus Case 2, plus anything that invokes that pattern. In
> > > > other words, if the code is neither Case 1 nor Case 2, it may still require
> > > > FOLL_PIN, for patterns like this:
> > > >
> > > > Correct (uses FOLL_PIN calls):
> > > >       pin_user_pages()
> > > >       write to the data within the pages
> > > >       unpin_user_pages()
> > >
> > > Case 5 is crap though.  That bug should have been fixed by getting
> > > the locking right.  ie:
> > >
> > > 	get_user_pages_fast();
> > > 	lock_page();
> > > 	kmap();
> > > 	set_bit();
> > > 	kunmap();
> > > 	set_page_dirty()
> > > 	unlock_page();
> > >
> > > I should have vetoed that patch at the time, but I was busy with other things.
> >
> > It does seem like lock_page() is better, for now at least, because it
> > forces the kind of synchronization with file system writeback that is
> > still yet to be implemented for pin_user_pages().
> >
> > Long term though, Case 5 provides an alternative way to do this
> > pattern--without using lock_page(). Also, note that Case 5, *in
> > general*, need not be done page-at-a-time, unlike the lock_page()
> > approach. Therefore, Case 5 might potentially help at some call sites,
> > either for deadlock avoidance or performance improvements.
> >
> > In other words, once the other half of the pin_user_pages() plan is
> > implemented, either of these approaches should work.
> >
> > Or, are you thinking that there is never a situation in which Case 5 is
> > valid?
>
> I don't think the page pinning approach is ever valid.  For file

Could you qualify that? Surely you don't mean that the entire pin_user_pages
story is a waste of time--I would have expected you to make more noise
earlier if you thought that, yes?

> mappings, the infiniband registration should take a lease on the inode,
> blocking truncation.  DAX shouldn't be using struct pages at all, so
> there shouldn't be anything there to pin.
>
> It's been five years since DAX was merged, and page pinning still
> doesn't work.  How much longer before the people who are pushing it
> realise that it's fundamentally flawed?

Is this a separate rant about *only* DAX, or is general RDMA in your sights
too? :)



thanks,
--
John Hubbard
NVIDIA