Re: [PATCH] x86/irq: Cure live lock in irq_force_complete_move()

[Thomas Gleixner]

On Sat, 12 Mar 2016, Ingo Molnar wrote:
> Cool fix!! :-)

Not so cool, as it's not correct. Working on it ...
 
> How much time did it take for you to figure out this one?? ...

Harry provided backtraces of all cpus, so figuring out what was wrong was not
the problem, but wrapping the head around a fix still is one :)