Re: [RFC v2 PATCH 2/2] mm: mmap: zap pages with read mmap_sem for large mapping

[Yang Shi]

On 6/20/18 12:17 AM, Michal Hocko wrote:
> On Tue 19-06-18 14:13:05, Yang Shi wrote:
> > On 6/19/18 3:02 AM, Peter Zijlstra wrote:
> [...]
> > > Hold up, two things: you having to copy most of do_munmap() didn't seem
> > > to suggest a helper function? And second, since when are we allowed to
> > Yes, they will be extracted into a helper function in the next version.
> >
> > May bad, I don't think it is allowed. We could reform this to:
> >
> > acquire write mmap_sem
> > vma lookup (split vmas)
> > release write mmap_sem
> >
> > acquire read mmap_sem
> > zap pages
> > release read mmap_sem
> >
> > I'm supposed this is safe as what Michal said before.
> I didn't get to read your patches carefully yet but I am wondering why
> do you need to split in the first place. Why cannot you simply unmap the
> range (madvise(DONTNEED)) under the read lock and then take the lock for
> write to finish the rest?

Yes, we can. I just thought splitting vma up-front sounds more straight 
forward. But, I neglected the write mmap_sem issue. Will move the vma 
split into later write mmap_sem in the next version.

Thanks,
Yang