Re: futex breakage in 4.9 stable branch

[Mike Galbraith <efault@gmx.de>]

On Thu, 2021-03-04 at 19:47 +0100, Thomas Gleixner wrote:
> On Thu, Mar 04 2021 at 10:12, Mike Galbraith wrote:
> > On Mon, 2021-03-01 at 18:29 +0100, Ben Hutchings wrote:
> >
> > --- a/kernel/futex.c
> > +++ b/kernel/futex.c
> > @@ -874,8 +874,12 @@ static void free_pi_state(struct futex_p
> >  	 * and has cleaned up the pi_state already
> >  	 */
> >  	if (pi_state->owner) {
> > +		unsigned long flags;
> > +
> > +		raw_spin_lock_irqsave(&pi_state->pi_mutex.wait_lock, flags);
> >  		pi_state_update_owner(pi_state, NULL);
> >  		rt_mutex_proxy_unlock(&pi_state->pi_mutex);
> > +             raw_spin_unlock_irqrestore(&pi_state->pi_mutex.wait_lock, flags);
>
> This hunk is missing in 4.9 as well.
>
> >  	}
> >
> >  	if (current->pi_state_cache)
> > @@ -1406,7 +1410,7 @@ static int wake_futex_pi(u32 __user *uad
> >  	if (pi_state->owner != current)
> >  		return -EINVAL;
> >
> > -	raw_spin_lock(&pi_state->pi_mutex.wait_lock);
> > +	raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock);
> >  	new_owner = rt_mutex_next_owner(&pi_state->pi_mutex);
> >
> >  	/*
>
> That looks correct.

Thanks for the eyeballs.  FWIW, updated 4.4.259-rt214 on top of that is
a happy camper as well.

	-Mike