linux-kernel - Re: [GIT PULL rcu/next] rcu commits for 2.6.40

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20110516220710.GA6139@linux.vnet.ibm.com>
Date:	Mon, 16 May 2011 15:07:10 -0700
From:	"Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>
To:	Ingo Molnar <mingo@...e.hu>
Cc:	Yinghai Lu <yinghai@...nel.org>, linux-kernel@...r.kernel.org
Subject: Re: [GIT PULL rcu/next] rcu commits for 2.6.40

On Mon, May 16, 2011 at 02:33:31PM -0700, Paul E. McKenney wrote:
> On Mon, May 16, 2011 at 04:30:05PM +0200, Ingo Molnar wrote:
> > 
> > * Ingo Molnar <mingo@...e.hu> wrote:
> > 
> > > > I'll try it in any case.
> > > 
> > > oh, this was a new iteration, mea culpa!
> > > 
> > > And yes, it solves all problems for me as well. Mind pushing it as a fix? :-)
> > 
> > FYI, i am also getting the warning below with a defconfig.
> 
> Yep, this is me whining that rcu_enter_nohz() was called before the
> irq nesting count (->dynticks_nesting) had dropped to zero.  In other
> words, either I am mis-counting interrupts or we entered some irq
> handler and then somehow never left it, but still managed to get
> to process-level execution.  Most likely that I am miscounting
> somehow.
> 
> I suppose that I could remove the WARN_ON_ONCE() and pretend that all
> was well, but...  ;-)

Actually, your test did prove something very interesting.  I have
been suspecting NMIs, but since the patch you just tested commented
out rcu_nmi_enter() and rcu_nmi_exit() completely, we know that there
is some bug outside of these two functions.  So if I am miscounting
nesting levels, I must be doing soe in rcu_enter_nohz(), rcu_exit_nohz(),
rcu_irq_enter(), or rcu_irq_exit().  Or some combination of those four.

							Thanx, Paul

> > 	Ingo
> > 
> > initcall init_per_zone_wmark_min+0x0/0x5b returned 0 after 41 usecs
> > calling  kswapd_init+0x0/0x1d @ 1
> > ------------[ cut here ]------------
> > WARNING: at kernel/rcutree.c:364 rcu_enter_nohz+0x4f/0x60()
> > Hardware name: System Product Name
> > Modules linked in:
> > Pid: 0, comm: swapper Not tainted 2.6.39-rc7-tip-03260-gb177656-dirty #126707
> > Call Trace:
> >  [<c13974b6>] ? printk+0x18/0x1a
> >  [<c1038f0d>] warn_slowpath_common+0x6d/0xa0
> >  [<c1089bdf>] ? rcu_enter_nohz+0x4f/0x60
> >  [<c1089bdf>] ? rcu_enter_nohz+0x4f/0x60
> >  [<c1038f5d>] warn_slowpath_null+0x1d/0x20
> >  [<c1089bdf>] rcu_enter_nohz+0x4f/0x60
> >  [<c10639ca>] tick_nohz_stop_sched_tick+0x22a/0x470
> >  [<c1001685>] cpu_idle+0x65/0xe0
> >  [<c137d789>] rest_init+0xa1/0xa8
> >  [<c137d6e8>] ? reciprocal_value+0x48/0x48
> >  [<c156a6ef>] start_kernel+0x303/0x30b
> >  [<c156a1fd>] ? obsolete_checksetup+0x95/0x95
> >  [<c156a067>] i386_start_kernel+0x67/0x6d
> > ---[ end trace fe4ebffb2b8ff187 ]---
> > initcall kswapd_init+0x0/0x1d returned 0 after 70614 usecs
> > calling  extfrag_debug_init+0x0/0x78 @ 1
> > initcall extfrag_debug_init+0x0/0x78 returned 0 after 79 usecs
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@...r.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/