Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20200612004644.255692-1-mike.kravetz@oracle.com>
 <20200612015842.GC23230@ZenIV.linux.org.uk> <b1756da5-4e91-298f-32f1-e5642a680cbf@oracle.com>
 <CAOQ4uxg=o2SVbfUiz0nOg-XHG8irvAsnXzFWjExjubk2v_6c_A@mail.gmail.com>
 <6e8924b0-bfc4-eaf5-1775-54f506cdf623@oracle.com> <CAJfpegsugobr8LnJ7e3D1+QFHCdYkW1swtSZ_hKouf_uhZreMg@mail.gmail.com>
In-Reply-To: <CAJfpegsugobr8LnJ7e3D1+QFHCdYkW1swtSZ_hKouf_uhZreMg@mail.gmail.com>
From:   Amir Goldstein <amir73il@gmail.com>
Date:   Mon, 15 Jun 2020 13:05:08 +0300
Message-ID: <CAOQ4uxgA+_4_UtVz17_eJL6m0CsDEVuiriBj1ZOkho+Ub1yuSA@mail.gmail.com>
Subject: Re: [PATCH v4 1/2] hugetlb: use f_mode & FMODE_HUGETLBFS to identify
 hugetlbfs files
To:     Miklos Szeredi <miklos@szeredi.hu>
Cc:     Mike Kravetz <mike.kravetz@oracle.com>,
        Al Viro <viro@zeniv.linux.org.uk>,
        Linux MM <linux-mm@kvack.org>,
        linux-fsdevel <linux-fsdevel@vger.kernel.org>,
        overlayfs <linux-unionfs@vger.kernel.org>,
        linux-kernel <linux-kernel@vger.kernel.org>,
        Matthew Wilcox <willy@infradead.org>,
        Colin Walters <walters@verbum.org>,
        Andrew Morton <akpm@linux-foundation.org>,
        syzbot <syzbot+d6ec23007e951dadf3de@syzkaller.appspotmail.com>,
        syzkaller-bugs <syzkaller-bugs@googlegroups.com>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, Jun 15, 2020 at 10:53 AM Miklos Szeredi <miklos@szeredi.hu> wrote:
>
> On Sat, Jun 13, 2020 at 9:12 PM Mike Kravetz <mike.kravetz@oracle.com> wrote:
> >
> > On 6/12/20 11:53 PM, Amir Goldstein wrote:
>
> > As a hugetlbfs developer, I do not know of a use case for interoperability
> > with overlayfs.  So yes, I am not too interested in making them work well
> > together.  However, if there was an actual use case I would be more than
> > happy to consider doing the work.  Just hate to put effort into fixing up
> > two 'special' filesystems for functionality that may not be used.
> >
> > I can't speak for overlayfs developers.
>
> As I said, I only know of tmpfs being upper layer as a valid use case.
>    Does that work with hugepages?  How would I go about testing that?

Simple, after enabling CONFIG_HUGETLBFS:

diff --git a/mount_union.py b/mount_union.py
index fae8899..4070c70 100644
--- a/mount_union.py
+++ b/mount_union.py
@@ -15,7 +15,7 @@ def mount_union(ctx):
         snapshot_mntroot = cfg.snapshot_mntroot()
         if cfg.should_mount_upper():
             system("mount " + upper_mntroot + " 2>/dev/null"
-                    " || mount -t tmpfs upper_layer " + upper_mntroot)
+                    " || mount -t hugetlbfs upper_layer " + upper_mntroot)
         layer_mntroot = upper_mntroot + "/" + ctx.curr_layer()
         upperdir = layer_mntroot + "/u"
         workdir = layer_mntroot + "/w"

It fails colossally, because hugetlbfs, does not have write_iter().
It is only meant as an interface to create named maps of huge pages.
So I don't really see the use case for using it as upper.

Thanks,
Amir.