Re: [PATCH V5 2/6] mm: don't assume anonymous pages have SwapBacked flag

[Michal Hocko]

On Mon 27-02-17 08:10:24, Shaohua Li wrote:
> On Mon, Feb 27, 2017 at 03:35:34PM +0100, Michal Hocko wrote:
> > On Fri 24-02-17 13:31:45, Shaohua Li wrote:
> > > There are a few places the code assumes anonymous pages should have
> > > SwapBacked flag set. MADV_FREE pages are anonymous pages but we are
> > > going to add them to LRU_INACTIVE_FILE list and clear SwapBacked flag
> > > for them. The assumption doesn't hold any more, so fix them.
> > > 
> > > Cc: Michal Hocko <mhocko@xxxxxxxx>
> > > Cc: Minchan Kim <minchan@xxxxxxxxxx>
> > > Cc: Hugh Dickins <hughd@xxxxxxxxxx>
> > > Cc: Rik van Riel <riel@xxxxxxxxxx>
> > > Cc: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
> > > Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> > > Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx>
> > > Signed-off-by: Shaohua Li <shli@xxxxxx>

Anyway, feel free to add
Acked-by: Michal Hocko <mhocko@xxxxxxxx>

> > 
> > Looks good to me.
> > [...]
> > > index 96eb85c..c621088 100644
> > > --- a/mm/rmap.c
> > > +++ b/mm/rmap.c
> > > @@ -1416,7 +1416,8 @@ static int try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
> > >  			 * Store the swap location in the pte.
> > >  			 * See handle_pte_fault() ...
> > >  			 */
> > > -			VM_BUG_ON_PAGE(!PageSwapCache(page), page);
> > > +			VM_BUG_ON_PAGE(!PageSwapCache(page) && PageSwapBacked(page),
> > > +				page);
> > 
> > just this part makes me scratch my head. I really do not understand what
> > kind of problem it tries to prevent from, maybe I am missing something
> > obvoious...
> 
> Just check a page which isn't lazyfree but wrongly enters here without swap
> entry. Or maybe you suggest we delete this statement?

Ohh, I figured out when seeing later patch in the series, I then wanted
to get back to this one but forgot... This on its own didn't really tell
me much. Maybe a comment would be helpful or even drop the VM_BUG_ON
altogether.
-- 
Michal Hocko
SUSE Labs