On 08/22, Andy Lutomirski wrote:
>
> nsproxy.pid_ns is *not* the task's pid namespace. The name should clarify
> that.
>
> This makes it more obvious that setns on a pid namespace is weird --
> it won't change the pid namespace shown in procfs.
>
> ...
>
> + * The pid namespace is an exception -- it's accessed using
> + * task_active_pid_ns. The pid namespace here is the
> + * namespace that children will use.
> + *
> * 'count' is the number of tasks holding a reference.
> * The count for each namespace, then, will be the number
> * of nsproxies pointing to it, not the number of tasks.
> @@ -27,7 +31,7 @@ struct nsproxy {
> struct uts_namespace *uts_ns;
> struct ipc_namespace *ipc_ns;
> struct mnt_namespace *mnt_ns;
> - struct pid_namespace *pid_ns;
> + struct pid_namespace *pid_ns_for_children;
> struct net *net_ns;
Personally I agree. ->pid_ns is "strange" and it makes sense to
document and make clear the fact that it became the implicit
argument for clone().
Oleg.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
