[PATCH 3/3] Signal semantics for pid namespaces

[Oleg Nesterov]

On 09/03, sukadev at us.ibm.com wrote:
>
> Oleg Nesterov [oleg at tv-sign.ru] wrote:
> | On 08/31, sukadev at us.ibm.com wrote:
> | >
> | > @@ -48,7 +49,7 @@ static int sig_init_ignore(struct task_s
> | >  	if (likely(!is_container_init(tsk->group_leader)))
> | >  		return 0;
> | >  
> | > -	if (!in_interrupt())
> | > +	if (is_current_in_ancestor_pid_ns(tsk) && !in_interrupt())
> | >  		return 0;
> | 
> | We should return 1 in that case, afaics the logic is wrongly reversed.
> 
> Hmm. My unit tests worked as I thought they should :-)
> 
> return 1 implies we "ignore the signal" right ?

Oops.

> If the signal is from an ancestor namespace, and we are not in interrupt
> context, we don't want to ignore the signal. no ?

You are right of course, sorry ;)

Oleg.