Re: [PATCH] break_lock forever broken

[Ingo Molnar <mingo@elte.hu>]

* Hugh Dickins <hugh@veritas.com> wrote:

> @@ -187,6 +187,8 @@ void __lockfunc _##op##_lock(locktype##_
>  			cpu_relax();					\
>  		preempt_disable();					\
>  	}								\
> +	if ((lock)->break_lock)						\
> +		(lock)->break_lock = 0;					\

while writing the ->break_lock feature i intentionally avoided overhead
in the spinlock fastpath. A better solution for the bug you noticed is
to clear the break_lock flag in places that use need_lock_break()
explicitly.

One robust way for that seems to be to make the need_lock_break() macro
clear the flag if it sees it set, and to make all the other (internal)
users use __need_lock_break() that doesnt clear the flag. I'll cook up a
patch for this.

	Ingo