From: Omar Sandoval <osandov@osandov.com>
To: Al Viro <viro@zeniv.linux.org.uk>, linux-fsdevel@vger.kernel.org
Cc: Linus Torvalds <torvalds@linux-foundation.org>,
linux-api@vger.kernel.org, kernel-team@fb.com,
Xi Wang <xi@cs.washington.edu>
Subject: [RFC PATCH v2 0/2] fs: add AT_REPLACE flag for linkat()
Date: Tue, 21 Mar 2017 07:51:46 -0700 [thread overview]
Message-ID: <cover.1490103963.git.osandov@fb.com> (raw)
From: Omar Sandoval <osandov@fb.com>
The goal of this series is to allow for updating a file atomically
in-place with an O_TMPFILE like so:
- open temporary file with O_TMPFILE
- write temporary file contents
- fsync temporary file
- atomically replace permanent location with the temporary file
This series implements atomic replace step with a new linkat(..., AT_REPLACE).
Al, I took a look at implementing this usecase with renameat2() after we
talked. Like you said, on the dcache side, "replace with tmpfile" looks
a lot like rename. However, on the filesystem side, trying to use rename
for this is much uglier than using link. We can't give i_op->rename() a
meaningful old_dir, and dealing with that new special case gets messy
fast.
So, here is the approach of having AT_REPLACE just unhash the replaced
dentry.
>From the original cover letter:
This is a proof-of-concept patch series implementing an AT_REPLACE flag for
linkat(2) which allows us to replace the target. This is a nice primitive on
its own, but it's most interesting when combined with O_TMPFILE, as it allows
you to do an atomic update of a file with an O_TMPFILE.
Patch 1 implements the VFS support for this flag. The implementation
resembles sys_renameat2(), and I took care to preserve all of the
original error cases and make the new error cases consistent with
rename.
Patch 2 adds support for AT_REPLACE to Btrfs. That's the codebase I'm
most familiar with so that's where I started, but it should be
straightforward to implement for other filesystems.
Cc: Xi Wang <xi@cs.washington.edu>
Omar Sandoval (2):
fs: add AT_REPLACE flag for linkat() which replaces the target
Btrfs: add support for linkat() AT_REPLACE
fs/btrfs/inode.c | 65 +++++++++++++++-
fs/ecryptfs/inode.c | 2 +-
fs/namei.c | 181 ++++++++++++++++++++++++++++++++++++---------
fs/nfsd/vfs.c | 2 +-
fs/overlayfs/overlayfs.h | 2 +-
include/linux/fs.h | 3 +-
include/uapi/linux/fcntl.h | 1 +
7 files changed, 211 insertions(+), 45 deletions(-)
--
2.12.0
next reply other threads:[~2017-03-21 14:52 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-03-21 14:51 Omar Sandoval [this message]
2017-03-21 14:51 ` [RFC PATCH v2 1/2] fs: add AT_REPLACE flag for linkat() which replaces the target Omar Sandoval
2017-03-21 16:30 ` Linus Torvalds
2017-03-21 18:44 ` Omar Sandoval
2017-03-21 19:02 ` Linus Torvalds
2017-03-21 20:37 ` Miklos Szeredi
2017-03-21 21:06 ` Linus Torvalds
2017-03-21 14:51 ` [RFC PATCH v2 2/2] Btrfs: add support for linkat() AT_REPLACE Omar Sandoval
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=cover.1490103963.git.osandov@fb.com \
--to=osandov@osandov.com \
--cc=kernel-team@fb.com \
--cc=linux-api@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=torvalds@linux-foundation.org \
--cc=viro@zeniv.linux.org.uk \
--cc=xi@cs.washington.edu \
/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).