for_each_matching_node performs an of_node_get on each iteration, so a break out of the loop requires an of_node_put. A simplified version of the semantic patch that fixes this problem is as follows (http://coccinelle.lip6.fr): // <smpl> @@ local idexpression n; expression e; identifier l; @@ for_each_matching_node(n,...) { ... ( of_node_put(n); | e = n | + of_node_put(n); ? goto l; ) ... } ... l: ... when != n // </smpl> Besides the issue found by the semantic patch, this code also stores the device_node value in a list, which requires an of_node_get, and then cleans up the list on exit from the function, which requires an of_node_put. Signed-off-by: Julia Lawall <Julia.Lawall@xxxxxxx> --- drivers/of/irq.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/of/irq.c b/drivers/of/irq.c index 171798d..902b89b 100644 --- a/drivers/of/irq.c +++ b/drivers/of/irq.c @@ -501,10 +501,12 @@ void __init of_irq_init(const struct of_device_id *matches) * pointer, interrupt-parent device_node etc. */ desc = kzalloc(sizeof(*desc), GFP_KERNEL); - if (WARN_ON(!desc)) + if (WARN_ON(!desc)) { + of_node_put(np); goto err; + } - desc->dev = np; + desc->dev = of_node_get(np); desc->interrupt_parent = of_irq_find_parent(np); if (desc->interrupt_parent == np) desc->interrupt_parent = NULL; @@ -575,6 +577,7 @@ void __init of_irq_init(const struct of_device_id *matches) err: list_for_each_entry_safe(desc, temp_desc, &intc_desc_list, list) { list_del(&desc->list); + of_node_put(desc->dev); kfree(desc); } } -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
