Re: mmotm 2010-04-05 - another RCU whinge (not network this time)

[Oleg Nesterov <oleg@redhat.com>]

On 04/09, Paul E. McKenney wrote:
>
> > Oleg, looks like proc-make-collect_sigign_sigcatch-rcu-safe.patch is the
> > offender here, it added the line that causes the whinge.
>
> If collect_sigign_sigcatch() is OK to call by updaters as well as
> readers, we need something like:
>
> 	struct sighand_struct *sighand;
> 	
> 	sighand = rcu_dereference_check(p->sighand,
> 					rcu_read_lock_held() ||
> 					lockdep_is_held(&???));
>
> Where the "???" is replaced with whichever of the two locks is protecting
> updates.  My guess would be the sighand lock, but I would not rely on
> my guesses in this case.  ;-)

Yes, it should be p->sighand->siglock.

Actually, I was going to change another caller, do_task_stat(), to call
collect_sigign_sigcatch() without ->siglock too, but now I am not sure
when/if this will happen.

OK, thanks, I'll send the patch to make rcu_dereference_check() happy.




While we are here... __exit_signal() does

	sighand = rcu_dereference_check(tsk->sighand,
					rcu_read_lock_held() ||
					lockdep_tasklist_lock_is_held());

What is the point? We know that the single caller must hold tasklist,
otherwise everything is broken. Perhaps it would be better to
use rcu_dereference_raw() ?

In fact, I don't really understand why __exit_signal() needs
rcu_dereference() at all.

Oleg.