Re: [PATCH V5 6/6] proc: show MADV_FREE pages info in smaps

[Shaohua Li]

On Wed, Mar 01, 2017 at 06:49:56PM +0100, Michal Hocko wrote:
> On Wed 01-03-17 09:37:10, Shaohua Li wrote:
> > On Wed, Mar 01, 2017 at 02:36:24PM +0100, Michal Hocko wrote:
> > > On Fri 24-02-17 13:31:49, Shaohua Li wrote:
> > > > show MADV_FREE pages info of each vma in smaps. The interface is for
> > > > diganose or monitoring purpose, userspace could use it to understand
> > > > what happens in the application. Since userspace could dirty MADV_FREE
> > > > pages without notice from kernel, this interface is the only place we
> > > > can get accurate accounting info about MADV_FREE pages.
> > > 
> > > I have just got to test this patchset and noticed something that was a
> > > bit surprising
> > > 
> > > madvise(mmap(len), len, MADV_FREE)
> > > Size:             102400 kB
> > > Rss:              102400 kB
> > > Pss:              102400 kB
> > > Shared_Clean:          0 kB
> > > Shared_Dirty:          0 kB
> > > Private_Clean:    102400 kB
> > > Private_Dirty:         0 kB
> > > Referenced:            0 kB
> > > Anonymous:        102400 kB
> > > LazyFree:         102368 kB
> > > 
> > > It took me a some time to realize that LazyFree is not accurate because
> > > there are still pages on the per-cpu lru_lazyfree_pvecs. I believe this
> > > is an implementation detail which shouldn't be visible to the userspace.
> > > Should we simply drain the pagevec? A crude way would be to simply
> > > lru_add_drain_all after we are done with the given range. We can also
> > > make this lru_lazyfree_pvecs specific but I am not sure this is worth
> > > the additional code.
> > 
> > Minchan's original patch includes a drain of pvec. I discard it because I think
> > it's not worth the effort. There aren't too many memory in the per-cpu vecs.
> 
> but multiply that by the number of CPUs.
> 
> > Like what you said, I doubt this is noticeable to userspace.
> 
> maybe I wasn't clear enough. I've noticed and I expect others would as
> well. We really shouldn't leak implementation details like that. So I
> _believe_ this should be fixed. Draining all pagevecs is rather coarse
> but it is the simplest thing to do. If you do not want to fold this
> into the original patch I can send a standalone one. Or do you have any
> concerns about draining?

No, no objection at all. Just doubt it's worthy. Looks nobody complains similar
issue, For exmaple, deactivate_file_page does the similar thing, then the smaps
'Referenced' could be inaccurate.

Thanks,
Shaohua