[Devel] pid_ns: Fix race between setns'ed fork() and zap_pid_ns_processes()

Submitted by Kirill Tkhai on May 15, 2017, 3:38 p.m.


Message ID 149486270990.7915.13458900181697971558.stgit@localhost.localdomain
State New
Series "pid_ns: Fix race between setns'ed fork() and zap_pid_ns_processes()"
Headers show

Commit Message

Kirill Tkhai May 15, 2017, 3:38 p.m.
This will go to mainstream:

Imagine we have a pid namespace and a task from its parent's pid_ns,
which made setns() to the pid namespace. The task is doing fork(),
while the pid namespace's child reaper is dying. We have the race
between them:

Task from parent pid_ns             Child reaper
copy_process()                      ..
alloc_pid()                       ..
..                                zap_pid_ns_processes()
..                                  disable_pid_allocation()
..                                  read_lock(&tasklist_lock)
..                                  iterate over pids in pid_ns
..                                    kill tasks linked to pids
..                                  read_unlock(&tasklist_lock)
write_lock_irq(&tasklist_lock);   ..
attach_pid(p, PIDTYPE_PID);       ..
..                                ..

So, just created task p won't receive SIGKILL signal,
and the pid namespace will be in contradictory state.
Only manual kill will help there, but does the userspace
care about this? I suppose, the most users just inject
a task into a pid namespace and wait a SIGCHLD from it.

The patch fixes the problem. It simply checks for
(pid_ns->nr_hashed & PIDNS_HASH_ADDING) in copy_process().
We do it under the tasklist_lock, and can't skip
PIDNS_HASH_ADDING as noted by Oleg:

"zap_pid_ns_processes() does disable_pid_allocation()
and then takes tasklist_lock to kill the whole namespace.
Given that copy_process() checks PIDNS_HASH_ADDING
under write_lock(tasklist) they can't race;
if copy_process() takes this lock first, the new child will
be killed, otherwise copy_process() can't miss
the change in ->nr_hashed."

If allocation is disabled, we just return -ENOMEM
like it's made for such cases in alloc_pid().

v2: Do not move disable_pid_allocation(), do not
introduce a new variable in copy_process() and simplify
the patch as suggested by Oleg Nesterov.
Account the problem with double irq enabling
found by Eric W. Biederman.

Fixes: c876ad768215 ("pidns: Stop pid allocation when init dies")
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
CC: Andrew Morton <akpm@linux-foundation.org>
CC: Ingo Molnar <mingo@kernel.org>
CC: Peter Zijlstra <peterz@infradead.org>
CC: Oleg Nesterov <oleg@redhat.com>
CC: Mike Rapoport <rppt@linux.vnet.ibm.com>
CC: Michal Hocko <mhocko@suse.com>
CC: Andy Lutomirski <luto@kernel.org>
CC: "Eric W. Biederman" <ebiederm@xmission.com>
CC: Andrei Vagin <avagin@openvz.org>
CC: Cyrill Gorcunov <gorcunov@openvz.org>
CC: Serge Hallyn <serge@hallyn.com>
Cc: stable@vger.kernel.org
Acked-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>

Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
 kernel/fork.c |    8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

Patch hide | download patch | download mbox

diff --git a/kernel/fork.c b/kernel/fork.c
index 24e178ffe68..0509a83bf72 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -1601,11 +1601,13 @@  static struct task_struct *copy_process(unsigned long clone_flags,
 	if (signal_pending(current)) {
-		spin_unlock(&current->sighand->siglock);
-		write_unlock_irq(&tasklist_lock);
 		retval = -ERESTARTNOINTR;
 		goto bad_fork_cancel_cgroup;
+	if (unlikely(!(ns_of_pid(pid)->nr_hashed & PIDNS_HASH_ADDING))) {
+		retval = -ENOMEM;
+		goto bad_fork_cancel_cgroup;
+	}
 	if (likely(p->pid)) {
 		ptrace_init_task(p, (clone_flags & CLONE_PTRACE) || trace);
@@ -1655,6 +1657,8 @@  static struct task_struct *copy_process(unsigned long clone_flags,
 	return p;
+	spin_unlock(&current->sighand->siglock);
+	write_unlock_irq(&tasklist_lock);
 	cgroup_cancel_fork(p, cgrp_ss_priv);
 	if (pid != &init_struct_pid)