[PATCH 6/15] Make alloc_pid(), free_pid() and put_pid() work with struct upid

Oleg Nesterov oleg at tv-sign.ru
Sun Jul 29 03:16:51 PDT 2007


On 07/26, Pavel Emelyanov wrote:
>
> -struct pid *alloc_pid(void)
> +struct pid *alloc_pid(struct pid_namespace *ns)

Why? We have the only caller, copy_process(), ns == task_active_pid_ns()
always.

>  {
>  	struct pid *pid;
>  	enum pid_type type;
> -	int nr = -1;
> -	struct pid_namespace *ns;
> +	int i, nr;
> +	struct pid_namespace *tmp;
>  
> -	ns = task_active_pid_ns(current);
>  	pid = kmem_cache_alloc(ns->pid_cachep, GFP_KERNEL);
>  	if (!pid)
>  		goto out;
>  
> -	nr = alloc_pidmap(ns);
> -	if (nr < 0)
> -		goto out_free;
> +	tmp = ns;
> +	for (i = ns->level; i >= 0; i--) {
> +		nr = alloc_pidmap(tmp);
> +		if (nr < 0)
> +			goto out_free;
> +
> +		pid->numbers[i].nr = nr;
> +		pid->numbers[i].ns = tmp;
> +		tmp = tmp->parent;

Hm... There is no ->parent in "struct pid_namespace", and this
patch doesn't add it.

> +	if (ns != &init_pid_ns)
> +		get_pid_ns(ns);

Q: put_pid() checks "ns != &init_pid_ns" as well, this is just
an optimization, yes? Perhaps we can move this check into
get_pid_ns/put_pid_ns.

We are doing get_pid_ns() only for the "top namespace"... I guess
this can work if pid_namespace does get_pid_ns() on its ->parent.
This patch looks incomplete.

Oleg.



More information about the Containers mailing list