linux-kernel - Re: Problems with hard irq? (inconsistent lock state)

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:	Wed, 4 Oct 2006 07:43:08 +0200
From:	Jarek Poplawski <jarkao2@...pl>
To:	"Axel C\. Voigt" <Axel.Voigt@...motec.com>
Cc:	linux-kernel@...r.kernel.org, David Kubicek <dave@....cz>,
	Greg Kroah-Hartman <gregkh@...e.de>
Subject: Re: Problems with hard irq? (inconsistent lock state)

On 29-09-2006 13:45, Axel C. Voigt wrote:
>  
> Hello all,
> 
> today I received the following stack backtrace using debian-2.6.18 with our communications driver accessing a ACM device. This happened when removing (powering off and/or on) the mobile phone (nokia 6630) at /dev/ttyACMx
> 
> Are someone able to get a hint for me?
> 
> --schnipp--
> Sep 29 13:29:53 mcs70 kernel:
> Sep 29 13:29:53 mcs70 kernel: =================================
> Sep 29 13:29:53 mcs70 kernel: [ INFO: inconsistent lock state ]
> Sep 29 13:29:53 mcs70 kernel: ---------------------------------
> Sep 29 13:29:53 mcs70 kernel: inconsistent {hardirq-on-W} -> {in-hardirq-W} usage.
> Sep 29 13:29:53 mcs70 kernel: startDV24/3864 [HC1[1]:SC0[0]:HE0:SE1] takes:
> Sep 29 13:29:53 mcs70 kernel: (&acm->read_lock){++..}, at: [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm]
> Sep 29 13:29:53 mcs70 kernel: {hardirq-on-W} state was registered at:
> Sep 29 13:29:53 mcs70 kernel: [<c01321f3>] lock_acquire+0x56/0x73
> Sep 29 13:29:53 mcs70 kernel: [<c02ce42a>] _spin_lock+0x1a/0x25
> Sep 29 13:29:53 mcs70 kernel: [<e08953a8>] acm_rx_tasklet+0x52/0x2be [cdc_acm]
> Sep 29 13:29:53 mcs70 kernel: [<c011f868>] tasklet_action+0x6d/0xd7
> Sep 29 13:29:53 mcs70 kernel: [<c011f526>] __do_softirq+0x79/0xf2
...
> Sep 29 13:29:53 mcs70 kernel: stack backtrace:
> Sep 29 13:29:53 mcs70 kernel: [<c010376e>] show_trace+0x16/0x18
> Sep 29 13:29:53 mcs70 kernel: [<c010383c>] dump_stack+0x19/0x1b
> Sep 29 13:29:53 mcs70 kernel: [<c0130ad8>] print_usage_bug+0x1e1/0x1eb
> Sep 29 13:29:53 mcs70 kernel: [<c0130b86>] mark_lock+0xa4/0x4d9
> Sep 29 13:29:53 mcs70 kernel: [<c0131823>] __lock_acquire+0x41a/0x841
> Sep 29 13:29:53 mcs70 kernel: [<c01321f3>] lock_acquire+0x56/0x73
> Sep 29 13:29:53 mcs70 kernel: [<c02ce42a>] _spin_lock+0x1a/0x25
> Sep 29 13:29:53 mcs70 kernel: [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm]
...
> Sep 29 13:29:53 mcs70 kernel: [<c0104b26>] do_IRQ+0x4e/0x5f
> Sep 29 13:29:53 mcs70 kernel: [<c0103339>] common_interrupt+0x25/0x2c
... 

It looks in drivers/usb/class/cdc-acm.c acm_rx_tasklet could be preempted
with acm->read_lock by acm_read_bulk which uses the same lock from hardirq
context.

So probably spin_lock_irqsave is needed.  

> Mit freundlichen Grüssen/ with kind regards,
>  
> Axel C. Voigt

Serdecznie pozdrawiam/ cheers,

Jarek P.
>  
-
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/