lkml.org 
[lkml]   [2011]   [Jul]   [15]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: [PATCH 0/7] overlay filesystem: request for inclusion
Date
"J. R. Okajima" <hooanon05@yahoo.co.jp> writes:

> Miklos Szeredi:
>> st_nlink==1 for directories is widely accepted way of saying that the
>> number of subdirectories is unknown. Various filesystems already do
>> this, and versions of GNU utils that I have come across accept it.
>
> When the upperdir is tmpfs, the link count of the directories in it will
> not be 1, won't it?

In overlayfs directory will have a link count of one if it's a "merged"
directory, regardless of the filesystem types used. If it's not a
merged directory, then st_nlink will correctly refrect the number of
subdirs.

>> I hadn't thought about this case, so overlayfs would modify the lower
>> filesystem in that case, which is a no-no.
>>
>> Following patch fixes this and return -EROFS for the above case. I also
>> updated the non-standard section in the docs.
>
> Hmm, such changes to mnt_flags looks slightly rude to me. Do we have to
> consider about these?
> - there may exist files opened as RW on the lower.
> - when overlayfs is unmounted, it should restore the original mnt_flags.
> - (there may exist more...)
>
> If overlayfs doesn't expect the lower mounted as RW, then it might be
> better to reject it simply at mounting.

Overlayfs will create a private clone of both the lower and upper
mounts. The mnt_flags are only modified on the private clone, whose
sole user is overlayfs itself.

Thanks,
Miklos


\
 
 \ /
  Last update: 2011-07-15 15:09    [W:0.098 / U:8.868 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site