Re: [PATCH v7 19/21] timer: Implement the hierarchical pull model

[Anna-Maria Behnsen <anna-maria@linutronix.de>]

[-- Attachment #1: Type: text/plain, Size: 3165 bytes --]

On Wed, 7 Jun 2023, Frederic Weisbecker wrote:

> Le Wed, May 24, 2023 at 09:06:27AM +0200, Anna-Maria Behnsen a écrit :
> > +/*
> > + * Returns true, if there is nothing to be propagated to the next level
> > + *
> > + * @data->nextexp is reset to KTIME_MAX; it is reused for first global
> > + * event which needs to be handled by migrator (in toplevel group)
> > + *
> > + * This is the only place where group event expiry value is set.
> > + */
> > +static bool tmigr_update_events(struct tmigr_group *group,
> > +				struct tmigr_group *child,
> > +				struct tmigr_walk *data)
> > +{
> > +	struct tmigr_event *evt, *first_childevt;
> > +	bool walk_done, remote = data->remote;
> > +	u64 nextexp;
> > +
> > +	if (child) {
> > +		if (data->childstate.active)
> > +			return true;
> > +
> > +		raw_spin_lock(&child->lock);
> > +		raw_spin_lock_nested(&group->lock, SINGLE_DEPTH_NESTING);
> > +
> > +		first_childevt = tmigr_next_groupevt(child);
> > +		nextexp = child->next_expiry;
> > +		evt = &child->groupevt;
> > +	} else {
> > +		nextexp = data->nextexp;
> > +
> > +		/*
> > +		 * Set @data->nextexp to KTIME_MAX; it is reused for first
> > +		 * global event which needs to be handled by migrator (in
> > +		 * toplevel group)
> > +		 */
> > +		data->nextexp = KTIME_MAX;
> > +
> > +		first_childevt = evt = data->evt;
> > +
> > +		/*
> > +		 * Walking the hierarchy is required in any case, when a
> > +		 * remote expiry was done before.
> 
> You can probably remove that comma because it feels like breaking the condition link.
> 
> > +		 * This ensures to not lost
> 
> lose
> 
> > +		 * already queued events in non active groups (see section
> > +		 * "Required event and timerqueue update after remote
> > +		 * expiry" in documentation at the top).
> > +		 */
> > +		if (evt->ignore && !remote)
> 
> It looks like in the case of !remote, this branch will never end up
> stopping the propagation up because either:
> 
> * We come here from tmigr_inactive_up() which takes care of the propagation.
> 
> or
> 
> * We come here from tmigr_new_timer() and ->ignore can't be set.
> 
> If I'm right, can we add a comment about that so that the poor reviewer
> doesn't stutter on that for too long?
> 

Right. It will never stop the propagation - but the condition could be
fulfilled when call site is tmigr_inactive_up(). My proposal for expanding
the comment is the following:

	/*
	 * Walking the hierarchy is required in any case when a
	 * remote expiry was done before. This ensures to not lose
	 * already queued events in non active groups (see section
	 * "Required event and timerqueue update after remote
	 * expiry" in documentation at the top).
	 *
	 * The two call sites which are executed without a remote expiry
	 * before, are not prevented from propagating changes through
	 * the hierarchy by the return:
	 *  - When entering this path by tmigr_new_timer(), evt->ignore
	 *    is never set.
	 *  - tmigr_inactive_up() takes care of propagation by itself
	 *    and ignores return value. But an immediate return is
	 *    required because nothing has to be done in this level as
	 *    event could be ignored.
	 */

Thanks,

	Anna-Maria