Re: [PATCH 1/5] locking: Introduce range reader/writer lock

[Peter Zijlstra]

On Mon, Mar 06, 2017 at 09:03:26PM -0800, Davidlohr Bueso wrote:
> +static __always_inline int
> +__range_read_lock_common(struct range_rwlock_tree *tree,
> +			 struct range_rwlock *lock, long state)
> +{
> +	struct interval_tree_node *node;
> +	unsigned long flags;
> +
> +	spin_lock_irqsave(&tree->lock, flags);
> +	lock->reader = true;
> +
> +	if (!__range_intersects_intree(tree, lock))
> +		goto insert;
> +
> +	node = interval_tree_iter_first(&tree->root, lock->node.start,
> +					 lock->node.last);
> +	while (node) {

	for (node = interval_tree_iter_first(); node;
	     node = interval_tree_iter_next()) {

Or some interval_tree_for_each helper?

> +		struct range_rwlock *blocked_lock;
> +		blocked_lock = range_entry(node, struct range_rwlock, node);
> +
> +		if (!blocked_lock->reader)
> +			lock->blocking_ranges++;

Can this @blocked range now go EINTR and remove itself from
wait_for_ranges()? If so, who will decrement our blocking_ranges?

> +		node = interval_tree_iter_next(node, lock->node.start,
> +					       lock->node.last);
> +	}
> +insert:
> +	__range_tree_insert(tree, lock);
> +
> +	lock->task = current;
> +	spin_unlock_irqrestore(&tree->lock, flags);
> +
> +	return wait_for_ranges(tree, lock, state);
> +}