Re: [patch] scalable timers implementation, 2.4.16, 2.5.0

[Andi Kleen]

Rusty Russell <[e-mail]> writes:

> The deadlock you're referring to is, I assume, del_timer_sync() called
> inside the timer itself?  Can you think of any other dangerous cases?

Common seems to be: 

	CPU#0                CPU#1

                             timer fires
spinlock()                 
                             spinlock() - spinning
del_timer_sync() 

-> deadlock. 

-Andi