[Devel] Re: [PATCH 1/4] Virtualization/containers: introduction

[Kirill Korotaev]

>>@@ -1132,6 +1133,7 @@ static task_t *copy_process(unsigned lon
>>        p->ioprio = current->ioprio;
>> 
>>        SET_LINKS(p);
>>+       (void)get_container(p->container);
>>        if (unlikely(p->ptrace & PT_PTRACED))
>>                __ptrace_link(p, current->parent); 
> 
> 
> This entire patch looks nice and very straightforward, except for this
> bit. :)  The "(void)" bit isn't usual kernel coding style.  You can
> probably kill it.
it is to avoid warning message the value has no effect.

> BTW, why does get_container() return the container argument?
> get_task_struct(), for instance is just a do{}while(0) loop, so it
> doesn't have a return value.  Is there some magic later on in your patch
> set that utilizes this?
ok, I will remake it without a return value. not a real problem at all.

> One other really minor thing: I usually try to do is keep the !
> CONFIG_FOO functions static inlines, just like the full versions.  The
> advantage is that you get some compile-time type checking, even when
> your CONFIG option is off.
it is not always appropriate :( I try to follow this as well :)

Kirill