netdev - Re: Hang deconfiguring network interface (in shutdown) on 3.3-rc1

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]

Date:	Sat, 28 Jan 2012 15:23:37 -0800
From:	"Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>
To:	James Bottomley <James.Bottomley@...senpartnership.com>
Cc:	Parisc List <linux-parisc@...r.kernel.org>,
	linux-kernel <linux-kernel@...r.kernel.org>,
	netdev <netdev@...r.kernel.org>,
	Michael Chan <mchan@...adcom.com>
Subject: Re: Hang deconfiguring network interface (in shutdown) on 3.3-rc1

On Sat, Jan 28, 2012 at 01:56:05PM -0600, James Bottomley wrote:
> It looks like it might be a tg3 or RCU issue.  When I shut down my
> parisc SMP 4 way system, I get an immediate hang here
> 
> Deconfiguring network interfaces...Internet Systems Consortium DHCP
> Client 4.1.1-P1
> Copyright 2004-2010 Internet Systems Consortium.
> All rights reserved.
> For info, please visit https://www.isc.org/software/dhcp/
> 
> Listening on LPF/eth0/00:30:6e:4b:15:59
> Sending on   LPF/eth0/00:30:6e:4b:15:59
> Sending on   Socket/fallback
> DHCPRELEASE on eth0 to 153.66.140.171 port 67
> 
> Followed some seconds later by
> 
> [ 5714.268000] INFO: rcu_sched detected stall on CPU 3 (t=15000 jiffies)
> [ 5714.268000] Backtrace:
> [ 5714.268000]  [<000000004011fdd4>] show_stack+0x14/0x20
> [ 5714.268000]  [<000000004011fdf8>] dump_stack+0x18/0x28
> [ 5714.268000]  [<00000000401c1fec>] __rcu_pending+0xcc/0x5c8
> [ 5714.276000]  [<00000000401c2d60>] rcu_check_callbacks+0x80/0xf8
> [ 5714.276000] INFO: rcu_sched detected stalls on CPUs/tasks: { 3}
> (detected by 2, t=15002 jiffies)
> [ 5714.276000] Backtrace:
> [ 5714.276000]  [<000000004011fdd4>] show_stack+0x14/0x20
> [ 5714.276000]  [<000000004011fdf8>] dump_stack+0x18/0x28
> [ 5714.276000]  [<00000000401c2484>] __rcu_pending+0x564/0x5c8
> [ 5714.276000]  [<00000000401c2d60>] rcu_check_callbacks+0x80/0xf8
> [ 5714.276000]  [<0000000040155dc8>] update_process_times+0x68/0xd8
> [ 5714.276000]  [<0000000040121378>] timer_interrupt+0x1c0/0x220
> [ 5714.276000]  [<00000000401b9cfc>] handle_irq_event_percpu+0xa4/0x2a0
> [ 5714.276000]  [<00000000401be58c>] handle_percpu_irq+0x9c/0xd0
> [ 5714.276000]  [<00000000401b9500>] generic_handle_irq+0x48/0x60
> [ 5714.276000]  [<0000000040121a50>] do_cpu_irq_mask+0x1b8/0x2a8
> [ 5714.276000]  [<0000000040105074>] intr_return+0x0/0x4
> [ 5714.276000]  [<0000000040105074>] intr_return+0x0/0x4
> [ 5714.276000]  [<00000000401296dc>] cpu_idle+0x74/0x80
> [ 5714.276000]  [<000000004078e1d0>] smp_callin+0x150/0x1a0
> [ 5714.276000] 
> [ 5714.348000]  [<0000000040155dc8>] update_process_times+0x68/0xd8
> [ 5714.348000]  [<0000000040121378>] timer_interrupt+0x1c0/0x220
> [ 5714.356000]  [<00000000401b9cfc>] handle_irq_event_percpu+0xa4/0x2a0
> [ 5714.364000]  [<00000000401be58c>] handle_percpu_irq+0x9c/0xd0
> [ 5714.364000]  [<00000000401b9500>] generic_handle_irq+0x48/0x60
> [ 5714.372000]  [<0000000040121a50>] do_cpu_irq_mask+0x1b8/0x2a8
> [ 5714.372000]  [<0000000040105074>] intr_return+0x0/0x4
> [ 5714.380000] 
> 
> This didn't happen in 3.2
> 
> Sysrq still works and sysrq-T shows ifconfig stuck:
> 
> 
> [ 6030.376000] ifconfig        R  running task        0  1470   1452
> 0x00000014
> [ 6030.376000] Backtrace:
> [ 6030.376000]  [<000000004017c6c8>] scheduler_tick+0x180/0x1a0
> [ 6030.376000]  [<0000000040155e1c>] update_process_times+0xbc/0xd8
> [ 6030.376000]  [<0000000040121378>] timer_interrupt+0x1c0/0x220
> [ 6030.376000]  [<00000000401b9d54>] handle_irq_event_percpu+0xfc/0x2a0
> [ 6030.376000]  [<0000000040105074>] intr_return+0x0/0x4
> [ 6030.376000]  [<000000004011c638>] _raw_spin_lock_bh+0x30/0x40
> [ 6030.376000]  [<000000004011c620>] _raw_spin_lock_bh+0x18/0x40
> [ 6030.376000]  [<000000001c6e3a64>] tg3_chip_reset+0x9c4/0x1328 [tg3]
> [ 6030.376000]  [<000000001c6eca9c>] tg3_halt+0xdc/0x1d8 [tg3]
> [ 6030.376000]  [<000000001c6f9964>] tg3_close+0x194/0x3f0 [tg3]
> [ 6030.376000]  [<0000000040411518>] __dev_close_many+0x100/0x178
> [ 6030.376000]  [<0000000040415130>] __dev_close+0x30/0x50
> [ 6030.376000]  [<000000004040ebf8>] __dev_change_flags+0xb0/0x1d0
> [ 6030.376000]  [<00000000404113b0>] dev_change_flags+0x28/0x90
> [ 6030.376000]  [<0000000040488d70>] devinet_ioctl+0x748/0x898
> [ 6030.376000]  [<000000004048a5f4>] inet_ioctl+0x204/0x228

If ifconfig is spinning in the kernel with preemption disabled, the RCU
CPU stall warning is expected behavior.  That said, judging from the
stack traces ifconfig was not running on CPU 3.  But I have seen similar
stack traces when someone forgets to drop a lock.

Do multiple sysrq-T commmands get the same picture, that ifconfig
is spinning on a lock in tg3_chip_reset()?  If so, does this reproduce
with lockdep enabled?

							Thanx, Paul

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html