Re: [RFC PATCH v3 19/37] irqchip: Add irq-kvx-apic-gic driver

[Krzysztof Kozlowski]

On 22/07/2024 11:41, ysionneau@xxxxxxxxxxxxx wrote:
> From: Yann Sionneau <ysionneau@xxxxxxxxxxxxx>
> 

...

> +
> +static int __init kvx_init_apic_gic(struct device_node *node,
> +				    struct device_node *parent)
> +{
> +	struct kvx_apic_gic *gic;
> +	int ret;
> +	unsigned int irq;
> +
> +	if (!parent) {
> +		pr_err("kvx apic gic does not have parent\n");

How is this possible? Aren't you controlling the code being executed?

> +		return -EINVAL;
> +	}
> +
> +	gic = kzalloc(sizeof(*gic), GFP_KERNEL);
> +	if (!gic)
> +		return -ENOMEM;
> +
> +	if (of_property_read_u32(node, "kalray,intc-nr-irqs",
> +						&gic->input_nr_irqs))

There is no such property. Also, there shouldn't be anyway...

> +		gic->input_nr_irqs = KVX_GIC_INPUT_IT_COUNT;
> +
> +	if (WARN_ON(gic->input_nr_irqs > KVX_GIC_INPUT_IT_COUNT)) {

Why? Please, drop all these WARN_ON from here and other patches. WARN_ON
is for cases which cannot happen, as it might panic entire system.

Instead, handle the case properly.

> +		ret = -EINVAL;
> +		goto err_kfree;
> +	}
> +
> +	gic->base = of_io_request_and_map(node, 0, node->name);
> +	if (!gic->base) {
> +		ret = -EINVAL;
> +		goto err_kfree;
> +	}
> +
> +	raw_spin_lock_init(&gic->lock);
> +	apic_gic_init(gic);


Best regards,
Krzysztof