Re: lockdep: possible fasync_lock/f_lock irq lock inversion dependency

[Li Zefan]

Wu Fengguang wrote:
> Hello,
> 
> I get a lockdep warning on HP 6910p:
> 

I saw this too, immediately when I opened Konsole after system boot,
and I was using 2.6.29-rc6-mm1.

> [  313.825353]
> [  313.825356] =========================================================
> [  313.825375] [ INFO: possible irq lock inversion dependency detected ]
> [  313.825385] 2.6.29-rc5-next-20090220 #62
> [  313.825393] ---------------------------------------------------------
> [  313.825401] swapper/0 just changed the state of lock:
> [  313.825409]  (fasync_lock){.-....}, at: [<ffffffff810ec9ae>] kill_fasync+0x2e/0x60
> [  313.825443] but this lock took another, HARDIRQ-READ-irq-unsafe lock in the past:
> [  313.825451]  (&f->f_lock){+.+...}
> [  313.825464]
> [  313.825465] and interrupts could create inverse lock ordering between them.
> [  313.825468]
> [  313.825484]
> [  313.825486] other info that might help us debug this:
> [  313.825497] 4 locks held by swapper/0:
> [  313.825504]  #0:  (&serio->lock){-.-...}, at: [<ffffffff8134afa1>] serio_interrupt+0x31/0xa0
> [  313.825539]  #1:  (&dev->event_lock){-.-...}, at: [<ffffffff8134fffc>] input_event+0x5c/0xa0
> [  313.825570]  #2:  (rcu_read_lock){.+.+..}, at: [<ffffffff8134e570>] input_pass_event+0x0/0xf0
> [  313.825600]  #3:  (rcu_read_lock){.+.+..}, at: [<ffffffff813538d0>] evdev_event+0x0/0x120
> [  313.825630]

--
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/