Messages in this thread | | | Subject | Re: rcu_read_lock/unlock protect find_task_by_vpid call in check_clock | From | Peter Zijlstra <> | Date | Thu, 25 Nov 2010 09:40:37 +0100 |
| |
On Wed, 2010-11-24 at 20:35 -0500, Dave Jones wrote: > =================================================== > [ INFO: suspicious rcu_dereference_check() usage. ] > --------------------------------------------------- > kernel/pid.c:419 invoked rcu_dereference_check() without protection! > > other info that might help us debug this: > > > rcu_scheduler_active = 1, debug_locks = 1 > 1 lock held by scrashme/13382: > #0: (tasklist_lock){.?.?..}, at: [<ffffffff8106ddea>] check_clock+0x46/0x9a > > stack backtrace: > Pid: 13382, comm: scrashme Not tainted 2.6.37-rc3+ #8 > Call Trace: > [<ffffffff8107cfe1>] lockdep_rcu_dereference+0x9d/0xa5 > [<ffffffff81069d08>] find_task_by_pid_ns+0x44/0x5d > [<ffffffff81069d43>] find_task_by_vpid+0x22/0x24 > [<ffffffff8106ddf2>] check_clock+0x4e/0x9a > [<ffffffff8106deac>] posix_cpu_clock_getres+0x16/0x41 > [<ffffffff8106be74>] sys_clock_getres+0x39/0xa0 > [<ffffffff81009cb2>] system_call_fastpath+0x16/0x1b > > Signed-off-by: Dave Jones <davej@redhat.com> > > diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c > index 6842eeb..4bef9aa 100644 > --- a/kernel/posix-cpu-timers.c > +++ b/kernel/posix-cpu-timers.c > @@ -38,11 +38,13 @@ static int check_clock(const clockid_t which_clock) > return 0; > > read_lock(&tasklist_lock); > + rcu_read_lock(); > p = find_task_by_vpid(pid); > if (!p || !(CPUCLOCK_PERTHREAD(which_clock) ? > same_thread_group(p, current) : thread_group_leader(p))) { > error = -EINVAL; > } > + rcu_read_unlock(); > read_unlock(&tasklist_lock); > > return error;
Pretty much the same comment as the other patch..
<copy/paste>
Do we still need the tasklist_lock in this case?
Also, why is that think complaining, surely the tasklist_lock pins any and all PID objects?
| |