On Tue, 2 Oct 2012, Paul E. McKenney wrote:
> > 1331e7a1bbe1f11b19c4327ba0853bee2a606543 is the first bad commit
> > commit 1331e7a1bbe1f11b19c4327ba0853bee2a606543
> > Author: Paul E. McKenney <paul.mcken...@linaro.org>
> > Date: Thu Aug 2 17:43:50 2012 -0700
> >
> > rcu: Remove _rcu_barrier() dependency on __stop_machine()
> >
> > Currently, _rcu_barrier() relies on preempt_disable() to prevent
> > any CPU from going offline, which in turn depends on CPU hotplug's
> > use of __stop_machine().
> >
> > This patch therefore makes _rcu_barrier() use get_online_cpus() to
> > block CPU-hotplug operations. This has the added benefit of removing
> > the need for _rcu_barrier() to adopt callbacks: Because CPU-hotplug
> > operations are excluded, there can be no callbacks to adopt. This
> > commit simplifies the code accordingly.
> >
> > Signed-off-by: Paul E. McKenney <paul.mcken...@linaro.org>
> > Signed-off-by: Paul E. McKenney <paul...@linux.vnet.ibm.com>
> > Reviewed-by: Josh Triplett <j...@joshtriplett.org>
> > ==
> >
> > is causing lockdep to complain (see the full trace below). I haven't yet
> > had time to analyze what exactly is happening, and probably will not have
> > time to do so until tomorrow, so just sending this as a heads-up in case
> > anyone sees the culprit immediately.
>
> Hmmm... Does the following patch help? It swaps the order in which
> rcu_barrier() acquires the hotplug and rcu_barrier locks.
It changed the report slightly (see for example the change in possible
unsafe locking scenario, rcu_sched_state.barrier_mutex vanished and it's
now directly about cpu_hotplug.lock). With the patch applied I get
======================================================
[ INFO: possible circular locking dependency detected ]
3.6.0-03888-g3f99f3b #145 Not tainted
-------------------------------------------------------
kworker/u:3/43 is trying to acquire lock:
(cpu_hotplug.lock){+.+.+.}, at: [<ffffffff81049287>] get_online_cpus+0x37/0x50
but task is already holding lock:
(slab_mutex){+.+.+.}, at: [<ffffffff81178175>] kmem_cache_destroy+0x45/0xe0
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (slab_mutex){+.+.+.}:
[<ffffffff810aeb22>] validate_chain+0x632/0x720
[<ffffffff810aef69>] __lock_acquire+0x359/0x580
[<ffffffff810af2b1>] lock_acquire+0x121/0x190
[<ffffffff8156130c>] __mutex_lock_common+0x5c/0x450
[<ffffffff8156182e>] mutex_lock_nested+0x3e/0x50
[<ffffffff8155cafa>] cpuup_callback+0x2f/0xbe
[<ffffffff81568bc3>] notifier_call_chain+0x93/0x140
[<ffffffff81077289>] __raw_notifier_call_chain+0x9/0x10
[<ffffffff8155b1ac>] _cpu_up+0xc9/0x162
[<ffffffff8155b301>] cpu_up+0xbc/0x11b
[<ffffffff81ae1793>] smp_init+0x6b/0x9f
[<ffffffff81ac57d6>] kernel_init+0x147/0x1dc
[<ffffffff8156eca4>] kernel_thread_helper+0x4/0x10
-> #0 (cpu_hotplug.lock){+.+.+.}:
[<ffffffff810ae48e>] check_prev_add+0x3de/0x440
[<ffffffff810aeb22>] validate_chain+0x632/0x720
[<ffffffff810aef69>] __lock_acquire+0x359/0x580
[<ffffffff810af2b1>] lock_acquire+0x121/0x190
[<ffffffff8156130c>] __mutex_lock_common+0x5c/0x450
[<ffffffff8156182e>] mutex_lock_nested+0x3e/0x50
[<ffffffff81049287>] get_online_cpus+0x37/0x50
[<ffffffff810f3a92>] _rcu_barrier+0x22/0x1f0
[<ffffffff810f3c70>] rcu_barrier_sched+0x10/0x20
[<ffffffff810f3c89>] rcu_barrier+0x9/0x10
[<ffffffff81178201>] kmem_cache_destroy+0xd1/0xe0
[<ffffffffa0488154>] nf_conntrack_cleanup_net+0xe4/0x110 [nf_conntrack]
[<ffffffffa04881aa>] nf_conntrack_cleanup+0x2a/0x70 [nf_conntrack]
[<ffffffffa04892ce>] nf_conntrack_net_exit+0x5e/0x80 [nf_conntrack]
[<ffffffff81458629>] ops_exit_list+0x39/0x60
[<ffffffff81458c5b>] cleanup_net+0xfb/0x1b0
[<ffffffff810691eb>] process_one_work+0x26b/0x4c0
[<ffffffff8106a03e>] worker_thread+0x12e/0x320
[<ffffffff8106f86e>] kthread+0xde/0xf0
[<ffffffff8156eca4>] kernel_thread_helper+0x4/0x10
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(slab_mutex);
lock(cpu_hotplug.lock);
lock(slab_mutex);
lock(cpu_hotplug.lock);
*** DEADLOCK ***
4 locks held by kworker/u:3/43:
#0: (netns){.+.+.+}, at: [<ffffffff81069122>] process_one_work+0x1a2/0x4c0
#1: (net_cleanup_work){+.+.+.}, at: [<ffffffff81069122>]
process_one_work+0x1a2/0x4c0
#2: (net_mutex){+.+.+.}, at: [<ffffffff81458be0>] cleanup_net+0x80/0x1b0
#3: (slab_mutex){+.+.+.}, at: [<ffffffff81178175>]
kmem_cache_destroy+0x45/0xe0
stack backtrace:
Pid: 43, comm: kworker/u:3 Not tainted 3.6.0-03888-g3f99f3b #145
Call Trace:
[<ffffffff810ac5cf>] print_circular_bug+0x10f/0x120
[<ffffffff810ae48e>] check_prev_add+0x3de/0x440
[<ffffffff810aeb22>] validate_chain+0x632/0x720
[<ffffffff810aef69>] __lock_acquire+0x359/0x580
[<ffffffff810af2b1>] lock_acquire+0x121/0x190
[<ffffffff81049287>] ? get_online_cpus+0x37/0x50
[<ffffffff8156130c>] __mutex_lock_common+0x5c/0x450
[<ffffffff81049287>] ? get_online_cpus+0x37/0x50
[<ffffffff810ada40>] ? mark_held_locks+0x80/0x120
[<ffffffff81049287>] ? get_online_cpus+0x37/0x50
[<ffffffff8156182e>] mutex_lock_nested+0x3e/0x50
[<ffffffff81049287>] get_online_cpus+0x37/0x50
[<ffffffff810f3a92>] _rcu_barrier+0x22/0x1f0
[<ffffffff810f3c70>] rcu_barrier_sched+0x10/0x20
[<ffffffff810f3c89>] rcu_barrier+0x9/0x10
[<ffffffff81178201>] kmem_cache_destroy+0xd1/0xe0
[<ffffffffa0488154>] nf_conntrack_cleanup_net+0xe4/0x110 [nf_conntrack]
[<ffffffffa04881aa>] nf_conntrack_cleanup+0x2a/0x70 [nf_conntrack]
[<ffffffffa04892ce>] nf_conntrack_net_exit+0x5e/0x80 [nf_conntrack]
[<ffffffff81458629>] ops_exit_list+0x39/0x60
[<ffffffff81458c5b>] cleanup_net+0xfb/0x1b0
[<ffffffff810691eb>] process_one_work+0x26b/0x4c0
[<ffffffff81069122>] ? process_one_work+0x1a2/0x4c0
[<ffffffff81069f69>] ? worker_thread+0x59/0x320
[<ffffffff81458b60>] ? net_drop_ns+0x40/0x40
[<ffffffff8106a03e>] worker_thread+0x12e/0x320
[<ffffffff81069f10>] ? manage_workers+0x1a0/0x1a0
[<ffffffff8106f86e>] kthread+0xde/0xf0
[<ffffffff8156eca4>] kernel_thread_helper+0x4/0x10
[<ffffffff81564b33>] ? retint_restore_args+0x13/0x13
[<ffffffff8106f790>] ? __init_kthread_worker+0x70/0x70
[<ffffffff8156eca0>] ? gs_change+0x13/0x13
--
Jiri Kosina
SUSE Labs
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
