Re: [PATCH v5 2/4] ovl: allocate anonymous devs for lowerdirs

[Vivek Goyal <vgoyal@redhat.com>]

On Wed, Nov 01, 2017 at 05:02:55PM +0200, Amir Goldstein wrote:
> On Wed, Nov 1, 2017 at 4:42 PM, Vivek Goyal <vgoyal@redhat.com> wrote:
> > On Mon, Oct 30, 2017 at 10:27:25PM +0200, Amir Goldstein wrote:
> >> From: Chandan Rajendra <chandan@linux.vnet.ibm.com>
> >>
> >> For stat(2) on lowerdir non-dir entries in non-samefs case, this commit
> >> provides unique values for st_dev. The unique values are obtained by
> >> allocating anonymous bdevs for each of the lowerdirs in the overlayfs
> >> instance.
> >
> > Hi Amir, Chandan,
> >
> > In the commit message, can we also mention what's the current behavior
> > and why this new behavior beneficial/desirable.
> >
> 
> This is the blurb from the uptodate patch on my branch:
> 
>      For non-samefs setup, to make sure that st_dev/st_ino pair
>     is unique across the system, we return a unique anonymous
>     st_dev for stat(2) of lower layer inode.
> 
> A bit fatter, but not fat enough...
> 
> Actually, it is not accurate, because st_dev/st_ino pair of pure
> upper is still same values as underlying inode for non-samefs so the
> values are not unique among all inodes in the system.

Hi Amir,

So as of now for non-samefs non-dir case we return st_dev/st_ino of
lower inode. And with this change we will return st_dev of overlayfs
while inode of lower, right?

What does unique mean in this context. IIUC, st_dev/st_inode of lower
will be unique in the system, isn't it. Which other inode can have
same st_dev/st_ino pair.

Or is it the case that if same inode is accessed through overlayfs, we
want to report a different st_dev.

> 
> I can't remember if there was a reason for not allocating anonymous bdev
> for upper

That's a good point.

> or if it just because we did not need it to guaranty uniqueness
> of st_dev/st_ino *among* overlay inodes

Even for lower, st_dev will be unique for different lower on non same-fs,
right. IOW, when it come to uniqueness of st_dev/st_ino pair, among
overlay inodes, lower and upper should have same requirements.

> while guarantying constant
> st_dev/st_ino across copy up.

Hmm..., I did not get this point. Over copy up, atleast st_ino will change
for non-samefs case. 

I will spend more time on patch.

Vivek

> 
> I will update commit message.
> 
> Thanks,
> Amir.