in sirfsoc gpio probe(), we create 5 irq_domains for 5 gpio banks. but
in irq_create_of_mapping() of irqchip core level, irq_find_host() can
only return the 1st irq_domain attached the pinctrl dt device node as
we can see from the codes:

unsigned int irq_create_of_mapping(struct device_node *controller,
				   const u32 *intspec, unsigned int intsize)
{
	struct irq_domain *domain;
	...
	domain = controller ? irq_find_host(controller) : irq_default_domain;
}

struct irq_domain *irq_find_host(struct device_node *node)
{
	struct irq_domain *h, *found = NULL;
	int rc;

	/* We might want to match the legacy controller last since
	 * it might potentially be set to match all interrupts in
	 * the absence of a device node. This isn't a problem so far
	 * yet though...
	 */
	mutex_lock(&irq_domain_mutex);
	list_for_each_entry(h, &irq_domain_list, link) {
		if (h->ops->match)
			rc = h->ops->match(h, node);
		else
			rc = (h->of_node != NULL) && (h->of_node == node);

		if (rc) {
			found = h;
			break;
		}
	}
	mutex_unlock(&irq_domain_mutex);
	return found;
}

for sirfsoc, the 1st irq_domain attached to the device_node(controller) only
can do linear for the 1st 32 gpios. so for devices who use gpio hwirq above
32 and put the information in dt like:
                                tangoc-ts@5c{
                                        compatible = "pixcir,tangoc-ts";
+                                       interrupt-parent = <&gpio>;
+                                       interrupts = <34 0>;
                                };

we will fail to get the virq for these devices as hwirq will be bigger than
domain->revmap_data.linear.size in:
unsigned int irq_linear_revmap(struct irq_domain *domain,
			       irq_hw_number_t hwirq)
{

	/* Check revmap bounds; complain if exceeded */
	if (WARN_ON(hwirq >= domain->revmap_data.linear.size))
		return 0;

	return domain->revmap_data.linear.revmap[hwirq];
}

this patch drops redundant irq_domain and keep only one to fix the problem.
Signed-off-by: Barry Song <Baohua.Song@csr.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>