Add the missing of_node_put() to release the refcount incremented by of_find_matching_node().
Cc: stable@vger.kernel.org Fixes: 4266ab1a8ff5 ("irqchip/gic-v2m: Refactor to prepare for ACPI support") Signed-off-by: Ma Ke make24@iscas.ac.cn --- Changes in v2: - modified the patch according to suggestions. --- drivers/irqchip/irq-gic-v2m.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/irqchip/irq-gic-v2m.c b/drivers/irqchip/irq-gic-v2m.c index 51af63c046ed..d5988012eb40 100644 --- a/drivers/irqchip/irq-gic-v2m.c +++ b/drivers/irqchip/irq-gic-v2m.c @@ -407,12 +407,12 @@ static int __init gicv2m_of_init(struct fwnode_handle *parent_handle,
ret = gicv2m_init_one(&child->fwnode, spi_start, nr_spis, &res, 0); - if (ret) { - of_node_put(child); + if (ret) break; - } }
+ if (ret && child) + of_put_node(child); if (!ret) ret = gicv2m_allocate_domains(parent); if (ret)
On Tue, 20 Aug 2024 08:54:01 +0100, Ma Ke make24@iscas.ac.cn wrote:
Add the missing of_node_put() to release the refcount incremented by of_find_matching_node().
This doesn't reflect the patch anymore. Something like this should be closer to reality:
"We fail to perform an of_node_put() when of_address_to_resource() fails, leading to a refcount leak.
Address this by moving the error handling path outside of the loop and making it common to all failure modes."
With the commit message fixed:
Reviewed-by: Marc Zyngier maz@kernel.org
M.
linux-stable-mirror@lists.linaro.org
-
Ma Ke -
Marc Zyngier