On Wed 15-06-16 16:51:06, Oleg Nesterov wrote:
> On 06/09, Michal Hocko wrote:
> >
> > --- a/include/linux/sched.h
> > +++ b/include/linux/sched.h
> > @@ -1883,6 +1883,32 @@ extern int arch_task_struct_size __read_mostly;
> > #define TNF_FAULT_LOCAL 0x08
> > #define TNF_MIGRATE_FAIL 0x10
> >
> > +static inline bool in_vfork(struct task_struct *tsk)
> > +{
> > + bool ret;
> > +
> > + /*
> > + * need RCU to access ->real_parent if CLONE_VM was used along with
> > + * CLONE_PARENT.
> > + *
> > + * We check real_parent->mm == tsk->mm because CLONE_VFORK does not
> > + * imply CLONE_VM
> > + *
> > + * CLONE_VFORK can be used with CLONE_PARENT/CLONE_THREAD and thus
> > + * ->real_parent is not necessarily the task doing vfork(), so in
> > + * theory we can't rely on task_lock() if we want to dereference it.
> > + *
> > + * And in this case we can't trust the real_parent->mm == tsk->mm
> > + * check, it can be false negative. But we do not care, if init or
> > + * another oom-unkillable task does this it should blame itself.
> > + */
> > + rcu_read_lock();
> > + ret = tsk->vfork_done && tsk->real_parent->mm == tsk->mm;
> > + rcu_read_unlock();
> > +
> > + return ret;
> > +}
>
> ACK, but why sched.h ? It has a single caller in oom_kill.c.
It felt like generally reusable helper to get subtle details about the
vfork right.
--
Michal Hocko
SUSE Labs
