From: "NeilBrown" <email@example.com>
To: "Dave Chinner" <firstname.lastname@example.org>
Cc: "James Yonan" <email@example.com>,
"Amir Goldstein" <firstname.lastname@example.org>,
"Al Viro" <email@example.com>,
Subject: Re: [PATCH] namei: implemented RENAME_NEWER flag for renameat2() conditional replace
Date: Wed, 29 Jun 2022 12:07:04 +1000 [thread overview]
Message-ID: <firstname.lastname@example.org> (raw)
On Wed, 29 Jun 2022, Dave Chinner wrote:
> On Tue, Jun 28, 2022 at 05:19:12PM -0600, James Yonan wrote:
> > On 6/28/22 12:34, Amir Goldstein wrote:
> > > On Tue, Jun 28, 2022 at 8:44 PM Al Viro <email@example.com> wrote:
> > > > On Mon, Jun 27, 2022 at 04:11:07PM -0600, James Yonan wrote:
> > > >
> > > > > && d_is_positive(new_dentry)
> > > > > && timespec64_compare(&d_backing_inode(old_dentry)->i_mtime,
> > > > > &d_backing_inode(new_dentry)->i_mtime) <= 0)
> > > > > goto exit5;
> > > > >
> > > > > It's pretty cool in a way that a new atomic file operation can even be
> > > > > implemented in just 5 lines of code, and it's thanks to the existing
> > > > > locking infrastructure around file rename/move that these operations
> > > > > become almost trivial. Unfortunately, every fs must approve a new
> > > > > renameat2() flag, so it bloats the patch a bit.
> > > > How is it atomic and what's to stabilize ->i_mtime in that test?
> > > > Confused...
> > > Good point.
> > > RENAME_EXCHANGE_WITH_NEWER would have been better
> > > in that regard.
> > >
> > > And you'd have to check in vfs_rename() after lock_two_nondirectories()
> > So I mean atomic in the sense that you are comparing the old and new mtimes
> > inside the lock_rename/unlock_rename critical section in do_renameat2(), so
> mtime is not stable during rename, even with the inode locked. e.g. a
> write page fault occurring concurrently with rename will change
> mtime, and so which inode is "newer" can change during the rename
I don't think that is really important for the proposed use case.
In any case where you might be using this new rename flag, the target
file wouldn't be open for write, so the mtime wouldn't change.
The atomicity is really wanted to make sure the file at the destination
name is still the one that was expected (I think).
So I think it would be reasonable for the rename to fail if the target
file (or even "either file") is open for write. Can that change while
the inode is locked?
It would be nice if renameat2 took a third fd so we could say "only
rename if <this> is the target file", but it doesn't.
next prev parent reply other threads:[~2022-06-29 2:07 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-06-27 22:11 [PATCH] namei: implemented RENAME_NEWER flag for renameat2() conditional replace James Yonan
2022-06-28 9:46 ` Amir Goldstein
2022-06-28 21:56 ` James Yonan
2022-06-29 5:15 ` Amir Goldstein
2022-06-30 16:18 ` James Yonan
2022-06-28 17:34 ` Al Viro
2022-06-28 18:34 ` Amir Goldstein
2022-06-28 23:19 ` James Yonan
2022-06-29 1:43 ` Dave Chinner
2022-06-29 2:07 ` NeilBrown [this message]
2022-06-29 2:35 ` Dave Chinner
2022-06-29 2:49 ` NeilBrown
2022-06-30 16:39 ` James Yonan
2022-07-01 9:23 ` [PATCH v2] namei: implemented RENAME_NEWER_MTIME " James Yonan
2022-07-01 10:34 ` Amir Goldstein
2022-07-01 20:06 ` James Yonan
2022-07-02 8:07 ` Dave Chinner
2022-07-05 13:30 ` [PATCH v3 1/2] RENAME_NEWER_MTIME is a new userspace-visible flag for renameat2(), and stands alongside existing flags including RENAME_NOREPLACE, RENAME_EXCHANGE, and RENAME_WHITEOUT James Yonan
2022-07-05 13:30 ` [PATCH v3 2/2] selftests: added a new target renameat2 James Yonan
2022-07-05 13:30 ` [PATCH man-pages] rename.2: document new renameat2() flag RENAME_NEWER_MTIME James Yonan
2022-07-05 14:03 ` [RESEND PATCH v3 1/2] namei: implemented RENAME_NEWER_MTIME flag for renameat2() conditional replace James Yonan
2022-07-11 19:13 ` [PATCH v4 " James Yonan
2022-07-11 19:13 ` [PATCH v4 2/2] selftests: added a new target renameat2 James Yonan
2022-07-11 23:10 ` [PATCH v4 1/2] namei: implemented RENAME_NEWER_MTIME flag for renameat2() conditional replace Dave Chinner
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:
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
* 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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.