From mboxrd@z Thu Jan  1 00:00:00 1970
From: tglx@linutronix.de (Thomas Gleixner)
Date: Tue, 13 Sep 2016 17:21:26 +0200 (CEST)
Subject: [PATCH v5 2/9] drivers: irqchip: Add STM32 external interrupts
 support
In-Reply-To: <1473432124-6784-3-git-send-email-alexandre.torgue@st.com>
References: <1473432124-6784-1-git-send-email-alexandre.torgue@st.com>
 <1473432124-6784-3-git-send-email-alexandre.torgue@st.com>
Message-ID: <alpine.DEB.2.20.1609131713490.6233@nanos>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On Fri, 9 Sep 2016, Alexandre TORGUE wrote:
> +static void stm32_exti_free(struct irq_domain *d, unsigned int virq,
> +			    unsigned int nr_irqs)
> +{
> +	struct irq_data *data = irq_get_irq_data(virq);
> +
> +	irq_gc_mask_clr_bit(data->parent_data);

I have a hard time to understand this. The irq domain is not hierarchical.

> +	irq_domain_reset_irq_data(data);

> +	domain = irq_domain_add_linear(node, nr_exti,
> +				       &irq_exti_domain_ops, NULL);

It's a simple linear domain. So how can data->parent_data be a valid
irq_data pointer? Answer: It can't!

But it doesn't blow up in your face simply because the alloc/free callbacks
are never invoked for simple non hierarchical domains. So you should have
removed that stuff after copying some other irqchip driver.

Thanks,

	tglx