From: Christian Brauner <christian.brauner@ubuntu.com>
To: Kees Cook <keescook@chromium.org>
Cc: Bernd Edlinger <bernd.edlinger@hotmail.de>,
"Eric W. Biederman" <ebiederm@xmission.com>,
Jann Horn <jannh@google.com>, Jonathan Corbet <corbet@lwn.net>,
Alexander Viro <viro@zeniv.linux.org.uk>,
Andrew Morton <akpm@linux-foundation.org>,
Alexey Dobriyan <adobriyan@gmail.com>,
Thomas Gleixner <tglx@linutronix.de>,
Oleg Nesterov <oleg@redhat.com>,
Frederic Weisbecker <frederic@kernel.org>,
Andrei Vagin <avagin@gmail.com>, Ingo Molnar <mingo@kernel.org>,
"Peter Zijlstra (Intel)" <peterz@infradead.org>,
Yuyang Du <duyuyang@gmail.com>,
David Hildenbrand <david@redhat.com>,
Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
Anshuman Khandual <anshuman.khandual@arm.com>,
David Howells <dhowells@redhat.com>,
James Morris <jamorris@linux.microsoft.com>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Shakeel Butt <shakeelb@google.com>,
Jason Gunthorpe <jgg@ziepe.ca>,
Christian Kellner <christian@kellner.me>,
Andrea Arcangeli <aarcange@redhat.com>,
Aleksa Sarai <cyphar@cyphar.com>,
"Dmitry V. Levin" <ldv@altlinux.org>,
"linux-doc@vger.kernel.org" <linux-doc@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>,
"linux-mm@kvack.org" <linux-mm@kvack.org>,
"stable@vger.kernel.org" <stable@vger.kernel.org>
Subject: Re: [PATCHv4] exec: Fix a deadlock in ptrace
Date: Tue, 3 Mar 2020 09:58:02 +0100 [thread overview]
Message-ID: <20200303085802.eqn6jbhwxtmz4j2x@wittgenstein> (raw)
In-Reply-To: <202003021531.C77EF10@keescook>
On Mon, Mar 02, 2020 at 06:26:47PM -0800, Kees Cook wrote:
> On Mon, Mar 02, 2020 at 10:18:07PM +0000, Bernd Edlinger wrote:
> > This fixes a deadlock in the tracer when tracing a multi-threaded
> > application that calls execve while more than one thread are running.
> >
> > I observed that when running strace on the gcc test suite, it always
> > blocks after a while, when expect calls execve, because other threads
> > have to be terminated. They send ptrace events, but the strace is no
> > longer able to respond, since it is blocked in vm_access.
> >
> > The deadlock is always happening when strace needs to access the
> > tracees process mmap, while another thread in the tracee starts to
> > execve a child process, but that cannot continue until the
> > PTRACE_EVENT_EXIT is handled and the WIFEXITED event is received:
> >
> > strace D 0 30614 30584 0x00000000
> > Call Trace:
> > __schedule+0x3ce/0x6e0
> > schedule+0x5c/0xd0
> > schedule_preempt_disabled+0x15/0x20
> > __mutex_lock.isra.13+0x1ec/0x520
> > __mutex_lock_killable_slowpath+0x13/0x20
> > mutex_lock_killable+0x28/0x30
> > mm_access+0x27/0xa0
> > process_vm_rw_core.isra.3+0xff/0x550
> > process_vm_rw+0xdd/0xf0
> > __x64_sys_process_vm_readv+0x31/0x40
> > do_syscall_64+0x64/0x220
> > entry_SYSCALL_64_after_hwframe+0x44/0xa9
> >
> > expect D 0 31933 30876 0x80004003
> > Call Trace:
> > __schedule+0x3ce/0x6e0
> > schedule+0x5c/0xd0
> > flush_old_exec+0xc4/0x770
> > load_elf_binary+0x35a/0x16c0
> > search_binary_handler+0x97/0x1d0
> > __do_execve_file.isra.40+0x5d4/0x8a0
> > __x64_sys_execve+0x49/0x60
> > do_syscall_64+0x64/0x220
> > entry_SYSCALL_64_after_hwframe+0x44/0xa9
> >
> > The proposed solution is to take the cred_guard_mutex only
> > in a critical section at the beginning, and at the end of the
> > execve function, and let PTRACE_ATTACH fail with EAGAIN while
> > execve is not complete, but other functions like vm_access are
> > allowed to complete normally.
>
> Sorry to be bummer, but I don't think this will work. A few more things
> during the exec process depend on cred_guard_mutex being held.
>
> If I'm reading this patch correctly, this changes the lifetime of the
> cred_guard_mutex lock to be:
> - during prepare_bprm_creds()
> - from flush_old_exec() through install_exec_creds()
> Before, cred_guard_mutex was held from prepare_bprm_creds() through
> install_exec_creds().
>
> That means, for example, that check_unsafe_exec()'s documented invariant
> is violated:
> /*
> * determine how safe it is to execute the proposed program
> * - the caller must hold ->cred_guard_mutex to protect against
> * PTRACE_ATTACH or seccomp thread-sync
> */
> static void check_unsafe_exec(struct linux_binprm *bprm) ...
> which is looking at no_new_privs as well as other details, and making
> decisions about the bprm state from the current state.
>
> I think it also means that the potentially multiple invocations
> of bprm_fill_uid() (via prepare_binprm() via binfmt_script.c and
> binfmt_misc.c) would be changing bprm->cred details (uid, gid) without
> a lock (another place where current's no_new_privs is evaluated).
>
> Related, it also means that cred_guard_mutex is unheld for every
> invocation of search_binary_handler() (which can loop via the previously
> mentioned binfmt_script.c and binfmt_misc.c), if any of them have hidden
> dependencies on cred_guard_mutex. (Thought I only see bprm_fill_uid()
> currently.)
So one issue I see with having to reacquire the cred_guard_mutex might
be that this would allow tasks holding the cred_guard_mutex to block a
killed exec'ing task from exiting, right?
next prev parent reply other threads:[~2020-03-03 8:58 UTC|newest]
Thread overview: 203+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-01 11:27 [PATCH] exec: Fix a deadlock in ptrace Bernd Edlinger
2020-03-01 15:13 ` Aleksa Sarai
2020-03-01 15:58 ` Christian Brauner
2020-03-01 17:46 ` Bernd Edlinger
2020-03-01 18:20 ` Christian Brauner
2020-03-01 17:24 ` Bernd Edlinger
2020-03-01 18:21 ` Jann Horn
2020-03-01 18:52 ` Christian Brauner
2020-03-01 19:00 ` Bernd Edlinger
2020-03-01 20:00 ` Jann Horn
2020-03-01 20:34 ` [PATCHv2] " Bernd Edlinger
2020-03-02 6:38 ` Eric W. Biederman
2020-03-02 15:43 ` Bernd Edlinger
2020-03-02 15:57 ` Eric W. Biederman
2020-03-02 16:02 ` Bernd Edlinger
2020-03-02 16:17 ` Eric W. Biederman
2020-03-02 16:43 ` Jann Horn
2020-03-02 17:01 ` Bernd Edlinger
2020-03-02 17:37 ` Jann Horn
2020-03-02 17:42 ` christian
2020-03-02 18:08 ` Jann Horn
2020-03-02 20:10 ` [PATCHv3] " Bernd Edlinger
2020-03-02 20:28 ` Bernd Edlinger
2020-03-02 17:13 ` [PATCHv2] " Bernd Edlinger
2020-03-02 21:49 ` Eric W. Biederman
2020-03-02 22:00 ` Bernd Edlinger
2020-03-02 22:18 ` [PATCHv4] " Bernd Edlinger
2020-03-03 2:26 ` Kees Cook
2020-03-03 4:54 ` Bernd Edlinger
2020-03-03 5:29 ` Kees Cook
2020-03-03 8:08 ` Bernd Edlinger
2020-03-03 8:34 ` Christian Brauner
2020-03-03 8:43 ` Christian Brauner
2020-03-04 15:30 ` Christian Brauner
2020-03-03 8:58 ` Christian Brauner [this message]
2020-03-03 10:34 ` Bernd Edlinger
2020-03-03 11:23 ` Bernd Edlinger
2020-03-03 14:20 ` Christian Brauner
2020-03-03 13:02 ` [PATCHv5] " Bernd Edlinger
2020-03-03 15:18 ` Eric W. Biederman
2020-03-03 16:48 ` Bernd Edlinger
2020-03-03 17:01 ` Christian Brauner
2020-03-03 17:20 ` Christian Brauner
2020-03-03 20:08 ` Eric W. Biederman
2020-03-04 14:37 ` Bernd Edlinger
2020-03-04 16:33 ` Eric W. Biederman
2020-03-04 21:49 ` Bernd Edlinger
2020-03-04 21:56 ` [PATCHv6] " Bernd Edlinger
2020-03-05 18:36 ` Bernd Edlinger
2020-03-05 21:14 ` [PATCH 0/2] Infrastructure to allow fixing exec deadlocks Eric W. Biederman
2020-03-05 21:15 ` [PATCH 1/2] exec: Properly mark the point of no return Eric W. Biederman
2020-03-05 22:34 ` Bernd Edlinger
2020-03-06 5:19 ` Eric W. Biederman
2020-03-05 22:56 ` Bernd Edlinger
2020-03-06 5:09 ` Eric W. Biederman
2020-03-06 16:26 ` Bernd Edlinger
2020-03-06 17:16 ` Eric W. Biederman
2020-03-05 21:16 ` [PATCH 2/2] exec: Add a exec_update_mutex to replace cred_guard_mutex Eric W. Biederman
2020-03-05 21:51 ` Bernd Edlinger
2020-03-06 5:17 ` Eric W. Biederman
2020-03-06 11:46 ` Bernd Edlinger
2020-03-06 21:18 ` Eric W. Biederman
2020-03-06 19:16 ` Bernd Edlinger
2020-03-06 21:58 ` Eric W. Biederman
2020-03-06 22:29 ` Eric W. Biederman
2020-03-07 1:03 ` Eric W. Biederman
2020-03-08 12:58 ` [PATCH] exec: make de_thread alloc new signal struct earlier Bernd Edlinger
2020-03-08 18:12 ` Eric W. Biederman
2020-03-05 22:31 ` [PATCH 0/2] Infrastructure to allow fixing exec deadlocks Bernd Edlinger
2020-03-06 5:06 ` Eric W. Biederman
2020-03-08 21:34 ` [PATCH 0/5] " Eric W. Biederman
2020-03-08 21:35 ` [PATCH v2 1/5] exec: Only compute current once in flush_old_exec Eric W. Biederman
2020-03-09 13:56 ` Bernd Edlinger
2020-03-09 17:34 ` Eric W. Biederman
2020-03-09 17:56 ` Bernd Edlinger
2020-03-09 19:27 ` Bernd Edlinger
2020-03-10 20:17 ` Kees Cook
2020-03-10 21:12 ` Christian Brauner
2020-03-08 21:36 ` [PATCH v2 2/5] exec: Factor unshare_sighand out of de_thread and call it separately Eric W. Biederman
2020-03-09 19:28 ` Bernd Edlinger
2020-03-10 20:29 ` Kees Cook
2020-03-10 20:34 ` Bernd Edlinger
2020-03-10 20:57 ` Kees Cook
2020-03-10 21:21 ` Christian Brauner
2020-03-08 21:36 ` [PATCH v2 3/5] exec: Move cleanup of posix timers on exec out of de_thread Eric W. Biederman
2020-03-09 19:30 ` Bernd Edlinger
2020-03-09 19:59 ` Christian Brauner
2020-03-09 20:06 ` Eric W. Biederman
2020-03-09 20:17 ` Christian Brauner
2020-03-09 20:48 ` Eric W. Biederman
2020-03-10 8:55 ` Christian Brauner
2020-03-10 18:52 ` [PATCH] pidfd: Stop taking cred_guard_mutex Eric W. Biederman
2020-03-10 19:15 ` Christian Brauner
2020-03-10 19:16 ` Jann Horn
2020-03-10 19:27 ` Eric W. Biederman
2020-03-10 20:00 ` Jann Horn
2020-03-10 20:10 ` Jann Horn
2020-03-10 20:22 ` Bernd Edlinger
2020-03-11 6:11 ` Bernd Edlinger
2020-03-11 14:56 ` Jann Horn
2020-03-10 20:57 ` Eric W. Biederman
2020-03-10 21:29 ` Christian Brauner
2020-03-11 18:49 ` Kees Cook
2020-03-14 9:12 ` [PATCH] pidfd: Use new infrastructure to fix deadlocks in execve Bernd Edlinger
2020-03-10 20:16 ` [PATCH v2 3/5] exec: Move cleanup of posix timers on exec out of de_thread Kees Cook
2020-03-10 20:31 ` Kees Cook
2020-03-10 20:57 ` Jann Horn
2020-03-10 21:05 ` Eric W. Biederman
2020-03-10 21:22 ` Christian Brauner
2020-03-08 21:38 ` [PATCH v2 4/5] exec: Move exec_mmap right after de_thread in flush_old_exec Eric W. Biederman
2020-03-09 19:34 ` Bernd Edlinger
2020-03-09 19:45 ` Eric W. Biederman
2020-03-09 19:52 ` Bernd Edlinger
2020-03-09 19:58 ` Eric W. Biederman
2020-03-09 20:03 ` Bernd Edlinger
2020-03-09 20:35 ` Eric W. Biederman
2020-03-10 20:44 ` Kees Cook
2020-03-10 21:20 ` Eric W. Biederman
2020-03-10 20:47 ` Kees Cook
2020-03-10 21:09 ` Eric W. Biederman
2020-03-08 21:38 ` [PATCH v2 5/5] exec: Add a exec_update_mutex to replace cred_guard_mutex Eric W. Biederman
2020-03-09 13:45 ` Bernd Edlinger
2020-03-09 17:40 ` Eric W. Biederman
2020-03-09 18:01 ` Bernd Edlinger
2020-03-09 18:10 ` Eric W. Biederman
2020-03-09 18:24 ` Eric W. Biederman
2020-03-09 18:36 ` Eric W. Biederman
2020-03-09 18:47 ` Bernd Edlinger
2020-03-09 19:02 ` Eric W. Biederman
2020-03-09 19:24 ` Bernd Edlinger
2020-03-09 19:35 ` Eric W. Biederman
2020-03-09 19:39 ` Eric W. Biederman
2020-03-10 13:43 ` [PATCH 0/4] Use new infrastructure to fix deadlocks in execve Bernd Edlinger
2020-03-10 15:35 ` Eric W. Biederman
2020-03-10 17:44 ` [PATCH 0/4] Use new infrastructure in more simple cases Bernd Edlinger
2020-03-10 17:45 ` [PATCH 1/4] kernel/kcmp.c: Use new infrastructure to fix deadlocks in execve Bernd Edlinger
2020-03-10 19:01 ` Eric W. Biederman
2020-03-10 19:42 ` Bernd Edlinger
2020-03-10 17:45 ` [PATCH 2/4] proc: " Bernd Edlinger
2020-03-11 18:59 ` Kees Cook
2020-03-11 19:10 ` Kees Cook
2020-03-11 19:38 ` Bernd Edlinger
2020-03-10 17:45 ` [PATCH 3/4] proc: io_accounting: " Bernd Edlinger
2020-03-10 19:06 ` Eric W. Biederman
2020-03-10 20:19 ` Bernd Edlinger
2020-03-10 21:25 ` Eric W. Biederman
2020-03-11 19:08 ` Kees Cook
2020-03-11 19:48 ` Bernd Edlinger
2020-03-11 19:48 ` Eric W. Biederman
2020-03-10 17:45 ` [PATCH 4/4] perf: " Bernd Edlinger
2020-03-10 13:43 ` [PATCH 1/4] exec: Fix a deadlock in ptrace Bernd Edlinger
2020-03-10 15:13 ` Eric W. Biederman
2020-03-10 15:17 ` Bernd Edlinger
2020-03-10 21:00 ` Kees Cook
2020-03-10 13:44 ` [PATCH 2/4] selftests/ptrace: add test cases for dead-locks Bernd Edlinger
2020-03-10 21:36 ` Kees Cook
2020-03-10 22:41 ` Dmitry V. Levin
2020-03-10 13:44 ` [PATCH 3/4] mm: docs: Fix a comment in process_vm_rw_core Bernd Edlinger
2020-03-11 18:53 ` Kees Cook
2020-03-10 13:44 ` [PATCH 4/4] kernel: doc: remove outdated comment cred.c Bernd Edlinger
2020-03-11 18:54 ` Kees Cook
2020-03-09 19:33 ` [PATCH v2 5/5] exec: Add a exec_update_mutex to replace cred_guard_mutex Dmitry V. Levin
2020-03-09 19:42 ` Eric W. Biederman
2020-03-10 20:55 ` Kees Cook
2020-03-10 21:02 ` Eric W. Biederman
2020-03-10 21:21 ` Jann Horn
2020-03-10 21:30 ` Eric W. Biederman
2020-03-10 23:21 ` Jann Horn
2020-03-11 0:15 ` Eric W. Biederman
2020-03-11 6:33 ` Bernd Edlinger
2020-03-11 16:29 ` Eric W. Biederman
2020-03-11 13:18 ` Qian Cai
2020-03-12 10:27 ` Kirill Tkhai
2020-03-12 12:24 ` Eric W. Biederman
2020-03-12 13:45 ` Kirill Tkhai
2020-03-12 14:38 ` Eric W. Biederman
2020-03-12 15:23 ` Kirill Tkhai
2020-03-13 1:05 ` Bernd Edlinger
2020-03-13 9:13 ` Kirill Tkhai
2020-03-14 9:11 ` [PATCH v3 " Bernd Edlinger
2020-03-17 8:56 ` Kirill Tkhai
2020-03-17 21:53 ` Bernd Edlinger
2020-03-18 12:22 ` Kirill Tkhai
2020-03-18 20:06 ` Bernd Edlinger
2020-03-19 7:13 ` Kirill Tkhai
2020-03-19 7:19 ` Bernd Edlinger
2020-03-19 9:11 ` [PATCH v4 3/5] " Bernd Edlinger
2020-03-19 9:13 ` Bernd Edlinger
2020-03-19 9:19 ` Greg Kroah-Hartman
2020-03-19 9:20 ` Bernd Edlinger
2020-03-21 22:53 ` Bernd Edlinger
2020-03-14 9:12 ` [PATCH 0/2] exec: Fix dead-lock in de_thread with ptrace_attach Bernd Edlinger
2020-03-14 9:12 ` [PATCH 1/2] " Bernd Edlinger
2020-03-14 9:13 ` [PATCH 2/2] doc: Update documentation of ->exec_*_mutex Bernd Edlinger
2020-03-14 9:57 ` [PATCH v2 5/5] exec: Add a exec_update_mutex to replace cred_guard_mutex Bernd Edlinger
2020-03-14 10:02 ` Bernd Edlinger
2020-03-17 8:58 ` Kirill Tkhai
2020-03-09 13:58 ` [PATCH 0/5] Infrastructure to allow fixing exec deadlocks Bernd Edlinger
2020-03-03 16:50 ` [PATCHv5] exec: Fix a deadlock in ptrace Christian Brauner
2020-03-02 12:28 ` [PATCHv2] " Oleg Nesterov
2020-03-02 15:56 ` Bernd Edlinger
2020-03-02 7:47 ` [PATCH] " Christian Brauner
2020-03-02 7:48 ` 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=20200303085802.eqn6jbhwxtmz4j2x@wittgenstein \
--to=christian.brauner@ubuntu.com \
--cc=aarcange@redhat.com \
--cc=adobriyan@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=anshuman.khandual@arm.com \
--cc=avagin@gmail.com \
--cc=bernd.edlinger@hotmail.de \
--cc=bigeasy@linutronix.de \
--cc=christian@kellner.me \
--cc=corbet@lwn.net \
--cc=cyphar@cyphar.com \
--cc=david@redhat.com \
--cc=dhowells@redhat.com \
--cc=duyuyang@gmail.com \
--cc=ebiederm@xmission.com \
--cc=frederic@kernel.org \
--cc=gregkh@linuxfoundation.org \
--cc=jamorris@linux.microsoft.com \
--cc=jannh@google.com \
--cc=jgg@ziepe.ca \
--cc=keescook@chromium.org \
--cc=ldv@altlinux.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mingo@kernel.org \
--cc=oleg@redhat.com \
--cc=peterz@infradead.org \
--cc=shakeelb@google.com \
--cc=stable@vger.kernel.org \
--cc=tglx@linutronix.de \
--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).