lkml.org 
[lkml]   [2013]   [Mar]   [15]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: [PATCH 0/9] overlay filesystem: request for inclusion (v17)
Date

Al Viro:
> > +- whiteout is hardlinked in order to reduce the consumption of inodes
> > + on branch
>
> *blink* Whiteouts have no inodes at all. Filesystem has an additional
> kind of directory entries, recognizable as whiteouts. How they are
> done is up to filesystem in question.

"no inodes at all"?
Are you assuming the implementation in dcache only (with a new d_type
flag)? And it is up to the real fs (layer or branch) whether it consumes
inode or not?
If so, it has a big disadvantage for the layer-fs (or branch-fs) to have
to implement a new method for whiteout.

Overlayfs implements whiteout as symlink+xattr which consumes an
inode. And you don't like it, right?
What I showed is another generic approach without xattr where the new
method to whiteout is unnecessary.


> > +The whiteout in aufs is very similar to Unionfs's. That is represented
> > +by its filename. UnionMount takes an approach of a file mode, but I am
> > +afraid several utilities (find(1) or something) will have to support it.
>
> Why the devil should find(1) even see them?

It is the case when find(1) for the layer-fs/branch-fs directly.


J. R. Okajima


\
 
 \ /
  Last update: 2013-03-15 06:44    [W:0.086 / U:1.044 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site