From: "J. R. Okajima" <hooanon05@yahoo.co.jp>
To: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Miklos Szeredi <miklos@szeredi.hu>,
Andrew Morton <akpm@linux-foundation.org>,
torvalds@linux-foundation.org, linux-fsdevel@vger.kernel.org,
linux-kernel@vger.kernel.org, hch@infradead.org,
apw@canonical.com, nbd@openwrt.org, neilb@suse.de,
jordipujolp@gmail.com, ezk@fsl.cs.sunysb.edu,
dhowells@redhat.com, sedat.dilek@googlemail.com,
mszeredi@suse.cz
Subject: Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)
Date: Fri, 15 Mar 2013 13:15:36 +0900 [thread overview]
Message-ID: <19058.1363320936@jrobl> (raw)
In-Reply-To: <20130315012541.GU21522@ZenIV.linux.org.uk>
Al Viro:
> Umm... I would prefer it to go through vfs.git, with serious modifications.
> I really don't like the idea of xattr-based whiteouts, for example.
I'd ask you how do you think another whiteout approach?
Here is a part of a posts which describes about aufs2 in 2009
<http://marc.info/?l=linux-kernel&m=123934927611907&w=2>
:::
+- whiteout is hardlinked in order to reduce the consumption of inodes
+ on branch
:::
+- kernel thread for removing the dir who has a plenty of whiteouts
:::
+The whiteout is for hiding files on lower branches. Also it is applied
+to stop readdir going lower branches.
+The latter case is called \[oq]opaque directory.\[cq] Any
+whiteout is an empty file, it means whiteout is just an mark.
+In the case of hiding lower files, the name of whiteout is
+\[oq]\*[AUFS_WH_PFX]<filename>.\[cq]
+And in the case of stopping readdir, the name is
+\[oq]\*[AUFS_WH_PFX]\*[AUFS_WH_PFX].opq\[cq] or
+\[oq]\*[AUFS_WH_PFX]__dir_opaque.\[cq] The name depends upon your compile
+configuration
+CONFIG_AUFS_COMPAT.
+.\" All of newly created or renamed directory will be opaque.
+All whiteouts are hardlinked,
+including \[oq]<writable branch top dir>/\*[AUFS_WH_BASE].\[cq]
:::
+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.
+
+Basically the whiteout represents "logical deletion" which stops aufs to
+lookup further, but also it represents "dir is opaque" which also stop
+lookup.
+
+In aufs, rmdir(2) and rename(2) for dir uses whiteout alternatively.
+In order to make several functions in a single systemcall to be
+revertible, aufs adopts an approach to rename a directory to a temporary
+unique whiteouted name.
+For example, in rename(2) dir where the target dir already existed, aufs
+renames the target dir to a temporary unique whiteouted name before the
+actual rename on a branch and then handles other actions (make it opaque,
+update the attributes, etc). If an error happens in these actions, aufs
+simply renames the whiteouted name back and returns an error. If all are
+succeeded, aufs registers a function to remove the whiteouted unique
+temporary name completely and asynchronously to the system global
+workqueue.
:::
Latest version is aufs3 (for linux-3.x series), there is no essential
changes about whiteout since aufs2.
J. R. Okajima
next prev parent reply other threads:[~2013-03-15 4:15 UTC|newest]
Thread overview: 79+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-03-13 14:16 [PATCH 0/9] overlay filesystem: request for inclusion (v17) Miklos Szeredi
2013-03-13 14:16 ` [PATCH 1/9] vfs: add i_op->dentry_open() Miklos Szeredi
2013-03-13 22:44 ` Andrew Morton
2013-03-14 11:15 ` Miklos Szeredi
2013-03-13 14:16 ` [PATCH 2/9] vfs: export do_splice_direct() to modules Miklos Szeredi
2013-03-13 22:45 ` Andrew Morton
2013-03-13 14:16 ` [PATCH 3/9] vfs: export __inode_permission() " Miklos Szeredi
2013-03-13 14:16 ` [PATCH 4/9] vfs: introduce clone_private_mount() Miklos Szeredi
2013-03-13 22:48 ` Andrew Morton
2013-03-14 13:28 ` Miklos Szeredi
2013-03-13 14:16 ` [PATCH 5/9] overlay filesystem Miklos Szeredi
2013-03-13 22:53 ` Andrew Morton
2013-03-13 14:16 ` [PATCH 6/9] overlayfs: add statfs support Miklos Szeredi
2013-03-13 14:16 ` [PATCH 7/9] overlayfs: implement show_options Miklos Szeredi
2013-03-13 14:16 ` [PATCH 8/9] overlay: overlay filesystem documentation Miklos Szeredi
2013-03-13 23:06 ` Andrew Morton
2013-03-14 13:35 ` Miklos Szeredi
2013-03-13 14:16 ` [PATCH 9/9] fs: limit filesystem stacking depth Miklos Szeredi
2013-03-13 14:31 ` [PATCH 0/9] overlay filesystem: request for inclusion (v17) Sedat Dilek
2013-03-13 15:13 ` Sedat Dilek
2013-03-13 15:18 ` Miklos Szeredi
2013-03-13 15:26 ` Sedat Dilek
2013-03-13 15:53 ` Sedat Dilek
2013-03-13 16:10 ` Sedat Dilek
2013-03-13 16:21 ` Miklos Szeredi
2013-03-13 16:35 ` Sedat Dilek
2013-03-13 16:51 ` Sedat Dilek
2013-03-13 18:12 ` Robin Holt
2013-03-13 18:37 ` Felix Fietkau
2013-03-13 19:10 ` Sedat Dilek
2013-03-13 19:54 ` Eric W. Biederman
2013-03-13 19:58 ` Linus Torvalds
2013-03-13 20:27 ` Sedat Dilek
[not found] ` <CAB3woddVfZ9PdYPpzidJLBMmUeRx0Rxgb5Pc8bTM9U-tkcS_uA@mail.gmail.com>
2013-03-13 20:32 ` Sedat Dilek
2013-03-13 20:36 ` Phillip Lougher
2013-03-13 23:08 ` Andrew Morton
2013-03-14 13:43 ` Miklos Szeredi
2013-03-15 1:25 ` Al Viro
2013-03-15 4:15 ` J. R. Okajima [this message]
2013-03-15 4:44 ` Al Viro
2013-03-15 5:09 ` J. R. Okajima
2013-03-15 5:13 ` Al Viro
2013-03-15 8:15 ` James Bottomley
2013-03-15 12:12 ` Al Viro
2013-03-15 18:57 ` J. R. Okajima
2013-03-15 19:26 ` Erez Zadok
2013-03-15 20:30 ` Al Viro
2013-03-16 13:55 ` J. R. Okajima
2013-03-15 19:11 ` Linus Torvalds
2013-03-16 13:57 ` J. R. Okajima
2013-03-17 13:06 ` [PATCH 2/9] vfs: export do_splice_direct() to modules David Howells
2013-03-18 2:31 ` Dave Chinner
2013-03-18 15:39 ` Jan Kara
2013-03-18 21:53 ` Al Viro
2013-03-18 23:01 ` Al Viro
2013-03-19 1:38 ` Al Viro
2013-03-19 9:00 ` J. R. Okajima
2013-03-19 10:29 ` Miklos Szeredi
2013-03-19 17:03 ` Al Viro
2013-03-19 18:32 ` Miklos Szeredi
2013-03-19 21:24 ` Al Viro
2013-03-20 9:15 ` Miklos Szeredi
2013-03-19 11:04 ` David Howells
2013-03-19 11:40 ` Miklos Szeredi
2013-03-19 20:54 ` Jan Kara
2013-03-19 20:25 ` Jan Kara
2013-03-19 21:38 ` Al Viro
2013-03-19 22:10 ` Al Viro
2013-03-20 2:33 ` Al Viro
2013-03-20 19:52 ` Jan Kara
2013-03-20 21:48 ` Al Viro
2013-03-20 22:19 ` Jan Kara
2013-03-20 12:30 ` David Howells
2013-03-22 17:37 ` J. R. Okajima
2013-03-22 18:11 ` Al Viro
2013-03-22 18:21 ` Al Viro
2013-03-23 2:49 ` J. R. Okajima
2013-03-23 4:41 ` Al Viro
2013-03-23 5:37 ` J. R. Okajima
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=19058.1363320936@jrobl \
--to=hooanon05@yahoo.co.jp \
--cc=akpm@linux-foundation.org \
--cc=apw@canonical.com \
--cc=dhowells@redhat.com \
--cc=ezk@fsl.cs.sunysb.edu \
--cc=hch@infradead.org \
--cc=jordipujolp@gmail.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=miklos@szeredi.hu \
--cc=mszeredi@suse.cz \
--cc=nbd@openwrt.org \
--cc=neilb@suse.de \
--cc=sedat.dilek@googlemail.com \
--cc=torvalds@linux-foundation.org \
--cc=viro@ZenIV.linux.org.uk \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).