Re: RCU qsmask !=0 warnings on large-SMP...

[Steffen Persvold]

On 1/26/2012 02:58, Paul E. McKenney wrote:
> On Wed, Jan 25, 2012 at 11:48:58PM +0100, Steffen Persvold wrote:
[]
> This looks like it will produce useful information, but I am not seeing
> output from it below.
>
> 							Thanx, Paul
>
> > This run it was CPU24 that triggered the issue :

This line is the printout for the root level :

> > [  231.572688] CPU 24, treason uncloaked, rsp @ ffffffff81a1cd80 (rcu_sched), rnp @ ffffffff81a1cd80(r) qsmask=0x1f, c=5132 g=5132 nc=5132 ng=5133 sc=5132 sg=5133 mc=5132 mg=5133

(this is the WARN_ON printout) :
> > [  231.576167] ------------[ cut here ]------------
> > [  231.576167] WARNING: at kernel/rcutree_plugin.h:1011 rcu_preempt_check_blocked_tasks+0x27/0x30()
> > [  231.576167] Hardware name: H8QI6
> > [  231.576167] Modules linked in: rcutorture
> > [  231.576167] Pid: 4603, comm: rcu_torture_rea Not tainted 3.2.1-numaconnect10+ #77
> > [  231.576167] Call Trace:
> > [  231.576167]<IRQ>   [<ffffffff810bb217>] ? rcu_preempt_check_blocked_tasks+0x27/0x30
> > [  231.576167]  [<ffffffff8106f47b>] warn_slowpath_common+0x8b/0xc0
> > [  231.576167]  [<ffffffff8106f4c5>] warn_slowpath_null+0x15/0x20
> > [  231.576167]  [<ffffffff810bb217>] rcu_preempt_check_blocked_tasks+0x27/0x30
> > [  231.576167]  [<ffffffff810bb330>] rcu_start_gp+0x110/0x1b0
> > [  231.576167]  [<ffffffff810bbf3b>] __rcu_process_callbacks+0x8b/0xd0
> > [  231.576167]  [<ffffffff810bc7a0>] rcu_process_callbacks+0x20/0x40
> > [  231.576167]  [<ffffffff8107580d>] __do_softirq+0x9d/0x140
> > [  231.576167]  [<ffffffff815d982c>] call_softirq+0x1c/0x30
> > [  231.576167]  [<ffffffff8103451a>] do_softirq+0x4a/0x80
> > [  231.576167]  [<ffffffff81075b83>] irq_exit+0x43/0x60
> > [  231.576167]  [<ffffffff8104aed5>] smp_apic_timer_interrupt+0x45/0x60
> > [  231.576167]  [<ffffffff815d834b>] apic_timer_interrupt+0x6b/0x70
> > [  231.576167]<EOI>   [<ffffffff81067aa9>] ? finish_task_switch+0x59/0xc0
> > [  231.576167]  [<ffffffff815d4d37>] __schedule+0x337/0x710
> > [  231.576167]  [<ffffffff81090425>] ? sched_clock_local+0x15/0x80
> > [  231.576167]  [<ffffffff8107b826>] ? lock_timer_base+0x36/0x70
> > [  231.576167]  [<ffffffff8107baa2>] ? mod_timer+0xf2/0x1d0
> > [  231.576167]  [<ffffffffa0001510>] ? rcu_torture_shuffle+0x80/0x80 [rcutorture]
> > [  231.576167]  [<ffffffff815d53ea>] schedule+0x3a/0x60
> > [  231.576167]  [<ffffffffa0001640>] rcu_torture_reader+0x130/0x230 [rcutorture]
> > [  231.576167]  [<ffffffffa0001dc0>] ? rcu_torture_writer+0x160/0x160 [rcutorture]
> > [  231.576167]  [<ffffffffa0001510>] ? rcu_torture_shuffle+0x80/0x80 [rcutorture]
> > [  231.576167]  [<ffffffff8108a726>] kthread+0x96/0xa0
> > [  231.576167]  [<ffffffff815d9734>] kernel_thread_helper+0x4/0x10
> > [  231.576167]  [<ffffffff8108a690>] ? kthread_stop+0x70/0x70
> > [  231.576167]  [<ffffffff815d9730>] ? gs_change+0xb/0xb
> > [  231.576167] ---[ end trace 828c8d7afbd02d1b ]---

I didn't include the leaf node printout, but the counters were 
indentical to the root printout (with the exception of the rnp address 
and qsmask of course).

Cheers,
--
Steffen Persvold, Chief Architect NumaChip
Numascale AS - www.numascale.com
Tel: +47 92 49 25 54 Skype: spersvold
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/