From: Daniel Colascione <dancol@google.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Jonathan Kowalski <bl0pbl33p@gmail.com>,
Christian Brauner <christian@brauner.io>,
Jann Horn <jannh@google.com>, Aleksa Sarai <cyphar@cyphar.com>,
Andy Lutomirski <luto@amacapital.net>,
Andrew Lutomirski <luto@kernel.org>,
David Howells <dhowells@redhat.com>,
"Serge E. Hallyn" <serge@hallyn.com>,
Linux API <linux-api@vger.kernel.org>,
Linux List Kernel Mailing <linux-kernel@vger.kernel.org>,
Arnd Bergmann <arnd@arndb.de>,
"Eric W. Biederman" <ebiederm@xmission.com>,
Konstantin Khlebnikov <khlebnikov@yandex-team.ru>,
Kees Cook <keescook@chromium.org>,
Alexey Dobriyan <adobriyan@gmail.com>,
Thomas Gleixner <tglx@linutronix.de>,
Michael Kerrisk-manpages <mtk.manpages@gmail.com>,
"Dmitry V. Levin" <ldv@altlinux.org>,
Andrew Morton <akpm@linux-foundation.org>,
Oleg Nesterov <oleg@redhat.com>,
Nagarathnam Muthusamy <nagarathnam.muthusamy@oracle.com>,
Al Viro <viro@zeniv.linux.org.uk>,
Joel Fernandes <joel@joelfernandes.org>
Subject: Re: [PATCH v2 0/5] pid: add pidfd_open()
Date: Mon, 1 Apr 2019 15:34:05 -0700 [thread overview]
Message-ID: <CAKOZuettqvG0S=3XBjnBLZvMEQMPcQswwPXogSfxwJZA85fdBA@mail.gmail.com> (raw)
In-Reply-To: <CAHk-=wiDjdoS4erUTjWaHMY+ZT28omwA24pPMRoQEE-rtNcRpw@mail.gmail.com>
On Mon, Apr 1, 2019 at 3:13 PM Linus Torvalds
<torvalds@linux-foundation.org> wrote:
>
> On Mon, Apr 1, 2019 at 2:58 PM Jonathan Kowalski <bl0pbl33p@gmail.com> wrote:
> >
> > You mention the race about learning the PID, PID being recycled, and
> > pidfd_open getting the wrong reference.
> >
> > This exists with the /proc model to way. How do you propose to address this?
>
> Note that that race exists _regardless_ of any interfaces.
> pidfd_open() has the same race: any time you have a pid, the lifetime
> of it is only as long as the process existing.
>
> That's why we talked about the CLONE_PIDFD flag, which would return
> the pidfd itself when creating a new process. That's one truly
> race-free way to handle it.
Yes. Returning a pidfd from clone seems like a simple and robust approach.
> Or just do the fork(), and know that the pid won't be re-used until
> you've done the wait() for it, and block SIGCHLD until you've done the
> lookup.
That doesn't work when some other thread is running a waitpid(-1)
loop. I think it's important to create an interface that libraries can
use without global coordination.
> That said, in *practice*, you can probably use any of the racy "look
> up pidfd using pid" models, as long as you just verify the end result
> after you've opened it.
>
> That verification could be as simple as "look up the parent pid of the
> pidfd I got", if you know you created it with fork() (and you can
> obviously track what _other_ thread you yourself created, so you can
> verify whether it is yours or not).
>
> For example, using "openat(pidfd, "status", ..)", but also by just
> tracking what you've done waitpid() on (but you need to look out for
> subtle races with another thread being in the process of doing so).
>
> Or you can just say that as long as you got the pidfd quickly after
> the fork(), any pid wrapping attack is practically not possible even
> if it might be racy in theory.
I don't like ignoring races just because they're rare. The cost of
complete race freedom for the process interface is low considering the
work we're doing on pidfds anyway.
WARNING: multiple messages have this Message-ID (diff)
From: Daniel Colascione <dancol@google.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Jonathan Kowalski <bl0pbl33p@gmail.com>,
Christian Brauner <christian@brauner.io>,
Jann Horn <jannh@google.com>, Aleksa Sarai <cyphar@cyphar.com>,
Andy Lutomirski <luto@amacapital.net>,
Andrew Lutomirski <luto@kernel.org>,
David Howells <dhowells@redhat.com>,
"Serge E. Hallyn" <serge@hallyn.com>,
Linux API <linux-api@vger.kernel.org>,
Linux List Kernel Mailing <linux-kernel@vger.kernel.org>,
Arnd Bergmann <arnd@arndb.de>,
"Eric W. Biederman" <ebiederm@xmission.com>,
Konstantin Khlebnikov <khlebnikov@yandex-team.ru>,
Kees Cook <keescook@chromium.org>,
Alexey Dobriyan <adobriyan@gmail.com>,
Thomas Gleixner <tglx@linutronix.de>,
Michael Kerrisk-manpages <mtk.manpages@gmail.com>,
"Dmitry V. Levin" <ldv@altlinux.org>,
Andrew Morton <akpm@linux-foundation.>
Subject: Re: [PATCH v2 0/5] pid: add pidfd_open()
Date: Mon, 1 Apr 2019 15:34:05 -0700 [thread overview]
Message-ID: <CAKOZuettqvG0S=3XBjnBLZvMEQMPcQswwPXogSfxwJZA85fdBA@mail.gmail.com> (raw)
In-Reply-To: <CAHk-=wiDjdoS4erUTjWaHMY+ZT28omwA24pPMRoQEE-rtNcRpw@mail.gmail.com>
On Mon, Apr 1, 2019 at 3:13 PM Linus Torvalds
<torvalds@linux-foundation.org> wrote:
>
> On Mon, Apr 1, 2019 at 2:58 PM Jonathan Kowalski <bl0pbl33p@gmail.com> wrote:
> >
> > You mention the race about learning the PID, PID being recycled, and
> > pidfd_open getting the wrong reference.
> >
> > This exists with the /proc model to way. How do you propose to address this?
>
> Note that that race exists _regardless_ of any interfaces.
> pidfd_open() has the same race: any time you have a pid, the lifetime
> of it is only as long as the process existing.
>
> That's why we talked about the CLONE_PIDFD flag, which would return
> the pidfd itself when creating a new process. That's one truly
> race-free way to handle it.
Yes. Returning a pidfd from clone seems like a simple and robust approach.
> Or just do the fork(), and know that the pid won't be re-used until
> you've done the wait() for it, and block SIGCHLD until you've done the
> lookup.
That doesn't work when some other thread is running a waitpid(-1)
loop. I think it's important to create an interface that libraries can
use without global coordination.
> That said, in *practice*, you can probably use any of the racy "look
> up pidfd using pid" models, as long as you just verify the end result
> after you've opened it.
>
> That verification could be as simple as "look up the parent pid of the
> pidfd I got", if you know you created it with fork() (and you can
> obviously track what _other_ thread you yourself created, so you can
> verify whether it is yours or not).
>
> For example, using "openat(pidfd, "status", ..)", but also by just
> tracking what you've done waitpid() on (but you need to look out for
> subtle races with another thread being in the process of doing so).
>
> Or you can just say that as long as you got the pidfd quickly after
> the fork(), any pid wrapping attack is practically not possible even
> if it might be racy in theory.
I don't like ignoring races just because they're rare. The cost of
complete race freedom for the process interface is low considering the
work we're doing on pidfds anyway.
next prev parent reply other threads:[~2019-04-01 22:34 UTC|newest]
Thread overview: 158+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-03-29 15:54 [PATCH v2 0/5] pid: add pidfd_open() Christian Brauner
2019-03-29 15:54 ` [PATCH v2 1/5] Make anon_inodes unconditional Christian Brauner
2019-03-29 15:54 ` [PATCH v2 2/5] pid: add pidfd_open() Christian Brauner
2019-03-29 23:45 ` Jann Horn
2019-03-29 23:45 ` Jann Horn
2019-03-29 23:55 ` Christian Brauner
2019-03-29 23:55 ` Christian Brauner
2019-03-30 11:53 ` Jürg Billeter
2019-03-30 14:37 ` Christian Brauner
2019-03-30 14:51 ` Jonathan Kowalski
2019-03-30 14:51 ` Jonathan Kowalski
2019-03-29 15:54 ` [PATCH v2 3/5] signal: support pidfd_open() with pidfd_send_signal() Christian Brauner
2019-03-29 15:54 ` [PATCH v2 4/5] signal: PIDFD_SIGNAL_TID threads via pidfds Christian Brauner
2019-03-30 1:06 ` Jann Horn
2019-03-30 1:06 ` Jann Horn
2019-03-30 1:22 ` Christian Brauner
2019-03-30 1:22 ` Christian Brauner
2019-03-30 1:34 ` Christian Brauner
2019-03-30 1:34 ` Christian Brauner
2019-03-30 1:42 ` Christian Brauner
2019-03-30 1:42 ` Christian Brauner
2019-03-29 15:54 ` [PATCH v2 5/5] tests: add pidfd_open() tests Christian Brauner
2019-03-30 16:09 ` [PATCH v2 0/5] pid: add pidfd_open() Linus Torvalds
2019-03-30 16:09 ` Linus Torvalds
2019-03-30 16:11 ` Daniel Colascione
2019-03-30 16:11 ` Daniel Colascione
2019-03-30 16:16 ` Linus Torvalds
2019-03-30 16:16 ` Linus Torvalds
2019-03-30 16:18 ` Linus Torvalds
2019-03-30 16:18 ` Linus Torvalds
2019-03-31 1:07 ` Joel Fernandes
2019-03-31 1:07 ` Joel Fernandes
2019-03-31 2:34 ` Jann Horn
2019-03-31 2:34 ` Jann Horn
2019-03-31 4:08 ` Joel Fernandes
2019-03-31 4:08 ` Joel Fernandes
2019-03-31 4:46 ` Jann Horn
2019-03-31 4:46 ` Jann Horn
2019-03-31 14:52 ` Linus Torvalds
2019-03-31 14:52 ` Linus Torvalds
2019-03-31 15:05 ` Christian Brauner
2019-03-31 15:05 ` Christian Brauner
2019-03-31 15:21 ` Daniel Colascione
2019-03-31 15:21 ` Daniel Colascione
2019-03-31 15:33 ` Jonathan Kowalski
2019-03-31 15:33 ` Jonathan Kowalski
2019-03-30 16:19 ` Christian Brauner
2019-03-30 16:19 ` Christian Brauner
2019-03-30 16:24 ` Linus Torvalds
2019-03-30 16:24 ` Linus Torvalds
2019-03-30 16:34 ` Daniel Colascione
2019-03-30 16:34 ` Daniel Colascione
2019-03-30 16:38 ` Christian Brauner
2019-03-30 16:38 ` Christian Brauner
2019-03-30 17:04 ` Linus Torvalds
2019-03-30 17:04 ` Linus Torvalds
2019-03-30 17:12 ` Christian Brauner
2019-03-30 17:12 ` Christian Brauner
2019-03-30 17:24 ` Linus Torvalds
2019-03-30 17:24 ` Linus Torvalds
2019-03-30 17:37 ` Christian Brauner
2019-03-30 17:37 ` Christian Brauner
2019-03-30 17:50 ` Jonathan Kowalski
2019-03-30 17:50 ` Jonathan Kowalski
2019-03-30 17:52 ` Christian Brauner
2019-03-30 17:52 ` Christian Brauner
2019-03-30 17:59 ` Jonathan Kowalski
2019-03-30 17:59 ` Jonathan Kowalski
2019-03-30 18:02 ` Christian Brauner
2019-03-30 18:02 ` Christian Brauner
2019-03-30 18:00 ` Jann Horn
2019-03-30 18:00 ` Jann Horn
2019-03-31 20:09 ` Andy Lutomirski
2019-03-31 20:09 ` Andy Lutomirski
2019-03-31 21:03 ` Linus Torvalds
2019-03-31 21:03 ` Linus Torvalds
2019-03-31 21:10 ` Christian Brauner
2019-03-31 21:10 ` Christian Brauner
2019-03-31 21:17 ` Linus Torvalds
2019-03-31 21:17 ` Linus Torvalds
2019-03-31 22:03 ` Christian Brauner
2019-03-31 22:03 ` Christian Brauner
2019-03-31 22:16 ` Linus Torvalds
2019-03-31 22:16 ` Linus Torvalds
2019-03-31 22:33 ` Christian Brauner
2019-03-31 22:33 ` Christian Brauner
2019-04-01 0:52 ` Jann Horn
2019-04-01 0:52 ` Jann Horn
2019-04-01 8:47 ` Yann Droneaud
2019-04-01 8:47 ` Yann Droneaud
2019-04-01 10:03 ` Jonathan Kowalski
2019-04-01 10:03 ` Jonathan Kowalski
2019-03-31 23:40 ` Linus Torvalds
2019-03-31 23:40 ` Linus Torvalds
2019-04-01 0:09 ` Al Viro
2019-04-01 0:09 ` Al Viro
2019-04-01 0:18 ` Linus Torvalds
2019-04-01 0:18 ` Linus Torvalds
2019-04-01 0:21 ` Christian Brauner
2019-04-01 0:21 ` Christian Brauner
2019-04-01 6:37 ` Al Viro
2019-04-01 6:37 ` Al Viro
2019-04-01 6:41 ` Al Viro
2019-04-01 6:41 ` Al Viro
2019-03-31 22:03 ` Jonathan Kowalski
2019-03-31 22:03 ` Jonathan Kowalski
2019-04-01 2:13 ` Andy Lutomirski
2019-04-01 2:13 ` Andy Lutomirski
2019-04-01 11:40 ` Aleksa Sarai
2019-04-01 11:40 ` Aleksa Sarai
2019-04-01 15:36 ` Linus Torvalds
2019-04-01 15:36 ` Linus Torvalds
2019-04-01 15:47 ` Christian Brauner
2019-04-01 15:47 ` Christian Brauner
2019-04-01 15:55 ` Daniel Colascione
2019-04-01 15:55 ` Daniel Colascione
2019-04-01 16:01 ` Linus Torvalds
2019-04-01 16:01 ` Linus Torvalds
2019-04-01 16:13 ` Daniel Colascione
2019-04-01 16:13 ` Daniel Colascione
2019-04-01 19:42 ` Christian Brauner
2019-04-01 19:42 ` Christian Brauner
2019-04-01 21:30 ` Linus Torvalds
2019-04-01 21:30 ` Linus Torvalds
2019-04-01 21:58 ` Jonathan Kowalski
2019-04-01 21:58 ` Jonathan Kowalski
2019-04-01 22:13 ` Linus Torvalds
2019-04-01 22:13 ` Linus Torvalds
2019-04-01 22:34 ` Daniel Colascione [this message]
2019-04-01 22:34 ` Daniel Colascione
2019-04-01 16:07 ` Jonathan Kowalski
2019-04-01 16:07 ` Jonathan Kowalski
2019-04-01 16:15 ` Linus Torvalds
2019-04-01 16:15 ` Linus Torvalds
2019-04-01 16:27 ` Jonathan Kowalski
2019-04-01 16:27 ` Jonathan Kowalski
2019-04-01 16:21 ` Daniel Colascione
2019-04-01 16:21 ` Daniel Colascione
2019-04-01 16:29 ` Linus Torvalds
2019-04-01 16:29 ` Linus Torvalds
2019-04-01 16:45 ` Daniel Colascione
2019-04-01 16:45 ` Daniel Colascione
2019-04-01 17:00 ` David Laight
2019-04-01 17:00 ` David Laight
2019-04-01 17:32 ` Linus Torvalds
2019-04-01 17:32 ` Linus Torvalds
2019-04-02 11:03 ` Florian Weimer
2019-04-02 11:03 ` Florian Weimer
2019-04-01 16:10 ` Andy Lutomirski
2019-04-01 16:10 ` Andy Lutomirski
2019-04-01 12:04 ` Christian Brauner
2019-04-01 12:04 ` Christian Brauner
2019-04-01 13:43 ` Jann Horn
2019-04-01 13:43 ` Jann Horn
2019-03-31 21:19 ` Christian Brauner
2019-03-31 21:19 ` Christian Brauner
2019-03-30 16:37 ` Christian Brauner
2019-03-30 16:37 ` Christian Brauner
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='CAKOZuettqvG0S=3XBjnBLZvMEQMPcQswwPXogSfxwJZA85fdBA@mail.gmail.com' \
--to=dancol@google.com \
--cc=adobriyan@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=arnd@arndb.de \
--cc=bl0pbl33p@gmail.com \
--cc=christian@brauner.io \
--cc=cyphar@cyphar.com \
--cc=dhowells@redhat.com \
--cc=ebiederm@xmission.com \
--cc=jannh@google.com \
--cc=joel@joelfernandes.org \
--cc=keescook@chromium.org \
--cc=khlebnikov@yandex-team.ru \
--cc=ldv@altlinux.org \
--cc=linux-api@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@amacapital.net \
--cc=luto@kernel.org \
--cc=mtk.manpages@gmail.com \
--cc=nagarathnam.muthusamy@oracle.com \
--cc=oleg@redhat.com \
--cc=serge@hallyn.com \
--cc=tglx@linutronix.de \
--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 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.