Re: [PATCH v2 3/9] userfaultfd/shmem: support minor fault registration for shmem

[Peter Xu <peterx@redhat.com>]

On Wed, Apr 14, 2021 at 12:36:13AM -0700, Hugh Dickins wrote:
> On Mon, 12 Apr 2021, Axel Rasmussen wrote:
> 
> > This patch allows shmem-backed VMAs to be registered for minor faults.
> > Minor faults are appropriately relayed to userspace in the fault path,
> > for VMAs with the relevant flag.
> > 
> > This commit doesn't hook up the UFFDIO_CONTINUE ioctl for shmem-backed
> > minor faults, though, so userspace doesn't yet have a way to resolve
> > such faults.
> 
> This is a very odd way to divide up the series: an "Intermission"
> half way through the implementation of MINOR/CONTINUE: this 3/9
> makes little sense without the 4/9 to mm/userfaultfd.c which follows.
> 
> But, having said that, I won't object and Peter did not object, and
> I don't know of anyone else looking here: it will only give each of
> us more trouble to insist on repartitioning the series, and it's the
> end state that's far more important to me and to all of us.

Agreed, ideally it should be after patch 4 since this patch enables the
feature already.

> 
> And I'll even seize on it, to give myself an intermission after
> this one, until tomorrow (when I'll look at 4/9 and 9/9 - but
> shall not look at the selftests ones at all).
> 
> Most of this is okay, except the mm/shmem.c part; and I've just now
> realized that somewhere (whether in this patch or separately) there
> needs to be an update to Documentation/admin-guide/mm/userfaultfd.rst
> (admin-guide? how weird, but not this series' business to correct).

(maybe some dir "devel" would suite better?  But I do also see soft-dirty.rst,
 idle_page_tracking.rst,..)

[...]

> >  static int shmem_getpage_gfp(struct inode *inode, pgoff_t index,
> > @@ -1820,6 +1820,14 @@ static int shmem_getpage_gfp(struct inode *inode, pgoff_t index,
> >  
> >  	page = pagecache_get_page(mapping, index,
> >  					FGP_ENTRY | FGP_HEAD | FGP_LOCK, 0);
> > +
> > +	if (page && vma && userfaultfd_minor(vma)) {
> > +		unlock_page(page);
> > +		put_page(page);
> > +		*fault_type = handle_userfault(vmf, VM_UFFD_MINOR);
> > +		return 0;
> > +	}
> > +
> 
> Okay, Peter persuaded you to move that up here: where indeed it
> does look better than the earlier "swapped" version.
> 
> But will crash on swap as it's currently written: it needs to say
> 		if (!xa_is_value(page)) {
> 			unlock_page(page);
> 			put_page(page);
> 		}

And this is definitely true...  Thanks,

> 
> I did say before that it's more robust to return from the swap
> case after doing the shmem_swapin_page(). But I might be slowly
> realizing that the ioctl to add the pte (in 4/9) will do its
> shmem_getpage_gfp(), and that will bring in the swap if user
> did not already do so: so I was wrong to claim more robustness
> the other way, this placement should be fine. I think.
> 
> >  	if (xa_is_value(page)) {
> >  		error = shmem_swapin_page(inode, index, &page,
> >  					  sgp, gfp, vma, fault_type);
> > -- 
> > 2.31.1.295.g9ea45b61b8-goog
> 

-- 
Peter Xu