linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Christian Brauner <christian.brauner@ubuntu.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Miklos Szeredi <miklos@szeredi.hu>,
	linux-fsdevel <linux-fsdevel@vger.kernel.org>,
	David Howells <dhowells@redhat.com>,
	Al Viro <viro@zeniv.linux.org.uk>, Karel Zak <kzak@redhat.com>,
	Jeff Layton <jlayton@redhat.com>,
	Miklos Szeredi <mszeredi@redhat.com>,
	Nicolas Dichtel <nicolas.dichtel@6wind.com>,
	Christian Brauner <christian@brauner.io>,
	Lennart Poettering <lennart@poettering.net>,
	Linux API <linux-api@vger.kernel.org>,
	Ian Kent <raven@themaw.net>,
	LSM <linux-security-module@vger.kernel.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: file metadata via fs API (was: [GIT PULL] Filesystem Information)
Date: Tue, 11 Aug 2020 21:39:16 +0200	[thread overview]
Message-ID: <20200811193916.zcwebstmbyvushau@wittgenstein> (raw)
In-Reply-To: <CAHk-=wg_bfVf5eazwH2uXTG-auCYZUpq-xb1kDeNjY7yaXS7bw@mail.gmail.com>

On Tue, Aug 11, 2020 at 09:05:22AM -0700, Linus Torvalds wrote:
> On Tue, Aug 11, 2020 at 8:30 AM Miklos Szeredi <miklos@szeredi.hu> wrote:
> >
> > What's the disadvantage of doing it with a single lookup WITH an enabling flag?
> >
> > It's definitely not going to break anything, so no backward
> > compatibility issues whatsoever.
> 
> No backwards compatibility issues for existing programs, no.
> 
> But your suggestion is fundamentally ambiguous, and you most
> definitely *can* hit that if people start using this in new programs.
> 
> Where does that "unified" pathname come from? It will be generated
> from "base filename + metadata name" in user space, and
> 
>  (a) the base filename might have double or triple slashes in it for
> whatever reasons.
> 
> This is not some "made-up gotcha" thing - I see double slashes *all*
> the time when we have things like Makefiles doing
> 
>     srctree=../../src/
> 
> and then people do "$(srctree)/". If you haven't seen that kind of
> pattern where the pathname has two (or sometimes more!) slashes in the
> middle, you've led a very sheltered life.
> 
>  (b) even if the new user space were to think about that, and remove
> those (hah! when have you ever seen user space do that?), as Al
> mentioned, the user *filesystem* might have pathnames with double
> slashes as part of symlinks.
> 
> So now we'd have to make sure that when we traverse symlinks, that
> O_ALT gets cleared. Which means that it's not a unified namespace
> after all, because you can't make symlinks point to metadata.
> 
> Or we'd retroactively change the semantics of a symlink, and that _is_
> a backwards compatibility issue. Not with old software, no, but it
> changes the meaning of old symlinks!
> 
> So no, I don't think a unified namespace ends up working.
> 
> And I say that as somebody who actually loves the concept. Ask Al: I
> have a few times pushed for "let's allow directory behavior on regular
> files", so that you could do things like a tar-filesystem, and access
> the contents of a tar-file by just doing
> 
>     cat my-file.tar/inside/the/archive.c
> 
> or similar.
> 
> Al has convinced me it's a horrible idea (and there you have a
> non-ambiguous marker: the slash at the end of a pathname that
> otherwise looks and acts as a non-directory)
> 

Putting my kernel hat down, putting my userspace hat on.

I'm looking at this from a potential user of this interface.
I'm not a huge fan of the metadata fd approach I'd much rather have a
dedicated system call rather than opening a side-channel metadata fd
that I can read binary data from. Maybe I'm alone in this but I was
under the impression that other users including Ian, Lennart, and Karel
have said on-list in some form that they would prefer this approach.
There are even patches for systemd and libmount, I thought?

But if we want to go down a completely different route then I'd prefer
if this metadata fd with "special semantics" did not in any way alter
the meaning of regular paths. This has the potential to cause a lot of
churn for userspace. I think having to play concatenation games in
shared libraries for mount information is a bad plan in addition to all
the issues you raised here.

Christian

  parent reply	other threads:[~2020-08-11 19:39 UTC|newest]

Thread overview: 89+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-08-03 15:27 [GIT PULL] Mount notifications David Howells
2020-08-03 15:49 ` [GIT PULL] Filesystem Information David Howells
2020-08-03 16:42   ` Miklos Szeredi
2020-08-04  2:15     ` Ian Kent
2020-08-04 14:36       ` Miklos Szeredi
2020-08-05  1:33         ` Ian Kent
2020-08-05  8:00           ` Miklos Szeredi
2020-08-05 11:13             ` Ian Kent
2020-08-05  8:24         ` file metadata via fs API (was: [GIT PULL] Filesystem Information) Miklos Szeredi
2020-08-11 13:54           ` Miklos Szeredi
2020-08-11 14:08             ` Al Viro
2020-08-11 14:22               ` Miklos Szeredi
2020-08-11 14:31                 ` Al Viro
     [not found]                   ` <CAAgocE07=vVKpQhG+rjEGO=NEBKZ02gjg4TRPxECAc+RKrzn=Q@mail.gmail.com>
2020-08-11 14:36                     ` Al Viro
2020-08-11 14:36                   ` Miklos Szeredi
2020-08-11 14:42                     ` Al Viro
2020-08-11 14:47                       ` Miklos Szeredi
2020-08-11 15:20             ` Linus Torvalds
2020-08-11 15:30               ` Miklos Szeredi
2020-08-11 16:05                 ` Linus Torvalds
2020-08-11 18:49                   ` Miklos Szeredi
2020-08-11 19:31                     ` Lennart Poettering
2020-08-11 19:50                       ` Christian Brauner
2020-08-11 19:39                   ` Christian Brauner [this message]
2020-08-12  0:53                     ` Ian Kent
2020-08-11 15:39               ` Andy Lutomirski
2020-08-11 16:17                 ` Casey Schaufler
2020-08-11 16:30                   ` Linus Torvalds
2020-08-11 20:28                   ` Miklos Szeredi
2020-08-11 20:36                     ` Jann Horn
2020-08-11 20:56                       ` Miklos Szeredi
2020-08-11 21:17                         ` Andy Lutomirski
2020-08-11 21:18                         ` Linus Torvalds
2020-08-12  7:23                           ` Miklos Szeredi
2020-08-12 14:39                             ` Al Viro
2020-08-12 14:46                               ` Miklos Szeredi
2020-08-12 15:08                                 ` Al Viro
2020-08-12 15:13                                   ` Miklos Szeredi
2020-08-12 16:33                                     ` Al Viro
2020-08-12 17:16                                       ` Miklos Szeredi
2020-08-12 17:39                                         ` Al Viro
2020-08-12 18:33                                           ` Al Viro
2020-08-12 21:30                                             ` Al Viro
2020-08-18  9:41                                               ` Miklos Szeredi
2020-08-18  9:30                                             ` Miklos Szeredi
2020-08-12 15:22                                   ` David Howells
2020-08-11 21:20                     ` Al Viro
2020-08-11 21:35                     ` Casey Schaufler
2020-08-11 16:05               ` Al Viro
2020-08-11 16:09                 ` Linus Torvalds
2020-08-11 16:39                   ` Al Viro
2020-08-12 10:14               ` Karel Zak
2020-08-12 13:09                 ` Miklos Szeredi
2020-08-12 13:33                 ` David Howells
2020-08-12 13:54                   ` Miklos Szeredi
2020-08-12  0:05             ` David Howells
2020-08-12  7:55               ` Miklos Szeredi
2020-08-12  8:29               ` David Howells
2020-08-12  8:37                 ` Miklos Szeredi
2020-08-12  9:43                   ` file metadata via fs API Steven Whitehouse
2020-08-12 10:04                     ` Miklos Szeredi
2020-08-12 11:28                       ` Karel Zak
2020-08-12 12:43                         ` Miklos Szeredi
2020-08-13  8:52                           ` Karel Zak
2020-08-12 13:06                         ` David Howells
2020-08-13  1:01                           ` Ian Kent
2020-08-12 18:18               ` file metadata via fs API (was: [GIT PULL] Filesystem Information) Linus Torvalds
2020-08-12 19:34                 ` file metadata via fs API Steven Whitehouse
2020-08-12 19:50                   ` Linus Torvalds
2020-08-13  3:44                     ` Ian Kent
2020-08-13 10:36                     ` Karel Zak
2020-08-14  7:58                     ` Lennart Poettering
2020-08-17 11:32                     ` Steven Whitehouse
2020-08-17 17:15                       ` Linus Torvalds
2020-08-17 22:44                         ` Linus Torvalds
2020-08-18 12:50                           ` Miklos Szeredi
2020-08-18 18:51                             ` Linus Torvalds
2020-08-18 20:18                               ` Miklos Szeredi
2020-08-18 20:53                                 ` Linus Torvalds
2020-08-21 13:17                                   ` Miklos Szeredi
2020-08-19  2:29                               ` Al Viro
2020-08-13  3:53                 ` file metadata via fs API (was: [GIT PULL] Filesystem Information) Jeffrey E Altman
2020-08-14 17:05                   ` Linus Torvalds
2020-08-18 15:01                     ` Jeffrey E Altman
2020-08-14  8:06                 ` Lennart Poettering
2020-08-12 13:54             ` David Howells
2020-08-12 14:10               ` Miklos Szeredi
2020-08-12 14:23               ` David Howells
2020-08-03 22:48 ` [GIT PULL] Mount notifications Ian Kent

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=20200811193916.zcwebstmbyvushau@wittgenstein \
    --to=christian.brauner@ubuntu.com \
    --cc=christian@brauner.io \
    --cc=dhowells@redhat.com \
    --cc=jlayton@redhat.com \
    --cc=kzak@redhat.com \
    --cc=lennart@poettering.net \
    --cc=linux-api@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-security-module@vger.kernel.org \
    --cc=miklos@szeredi.hu \
    --cc=mszeredi@redhat.com \
    --cc=nicolas.dichtel@6wind.com \
    --cc=raven@themaw.net \
    --cc=torvalds@linux-foundation.org \
    --cc=viro@zeniv.linux.org.uk \
    --subject='Re: file metadata via fs API (was: [GIT PULL] Filesystem Information)' \
    /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

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).