6.6-stable review patch. If anyone has any objections, please let me know.
------------------
From: Dan Carpenter dan.carpenter@linaro.org
[ Upstream commit 7dbc0d40d8347bd9de55c904f59ea44bcc8dedb7 ]
If irq_domain_translate_twocell() sets "hwirq" to >= MCHP_EIC_NIRQ (2) then it results in an out of bounds access.
The code checks for invalid values, but doesn't set the error code. Return -EINVAL in that case, instead of returning success.
Fixes: 00fa3461c86d ("irqchip/mchp-eic: Add support for the Microchip EIC") Signed-off-by: Dan Carpenter dan.carpenter@linaro.org Signed-off-by: Thomas Gleixner tglx@linutronix.de Reviewed-by: Claudiu Beznea claudiu.beznea@tuxon.dev Link: https://patch.msgid.link/aTfHmOz6IBpTIPU5@stanley.mountain Signed-off-by: Sasha Levin sashal@kernel.org --- drivers/irqchip/irq-mchp-eic.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/irqchip/irq-mchp-eic.c b/drivers/irqchip/irq-mchp-eic.c index 5dcd94c000a26..8a5baa0987a4b 100644 --- a/drivers/irqchip/irq-mchp-eic.c +++ b/drivers/irqchip/irq-mchp-eic.c @@ -166,7 +166,7 @@ static int mchp_eic_domain_alloc(struct irq_domain *domain, unsigned int virq,
ret = irq_domain_translate_twocell(domain, fwspec, &hwirq, &type); if (ret || hwirq >= MCHP_EIC_NIRQ) - return ret; + return ret ?: -EINVAL;
switch (type) { case IRQ_TYPE_EDGE_RISING: