[PATCH] regulator/scmi: Add missing of_node_get()

[Liang He]

At 2022-06-22 16:04:08, "Cristian Marussi" <cristian.marussi at arm.com> wrote:
>On Wed, Jun 22, 2022 at 11:48:16AM +0800, Liang He wrote:
>> In scmi_regulator_probe(), of_find_node_by_name() will decrease
>> the refcount of its first argument and we need a of_node_get()
>> to keep reference balance.
>> 
>> Signed-off-by: Liang He <windhl at 126.com>
>
>Hi Liang,
>
>good catch.
>

Thanks.

>I saw many other drivers that had this same issue fixed, given it is
>not so obvious, had also added some explaining comment like:
>
>/* Balance of_node_put() done by of_find_node_by_name() */
>
>so maybe it is a good idea to add something like that, but it depends at
>the end on how Mark prefers really. (given it is also in the commit log
>indeed...)
>
>Instead, more certainly, I would add also a Fixes: tag to you patch.
>
>Other than this:
>

Thanks, Cristian.

In fact, I know little aboult the rule of  how to add 'Fixes: tag'.

Can you tell me about that  and I will certainly add 'Fixes: tag' in future.

>Reviewed-by: Cristian Marussi <cristian.marussi at arm.com>
>
>Thanks,
>Cristian
>
>> ---
>>  drivers/regulator/scmi-regulator.c | 1 +
>>  1 file changed, 1 insertion(+)
>> 
>> diff --git a/drivers/regulator/scmi-regulator.c b/drivers/regulator/scmi-regulator.c
>> index 41ae7ac27ff6..b9918f4fd241 100644
>> --- a/drivers/regulator/scmi-regulator.c
>> +++ b/drivers/regulator/scmi-regulator.c
>> @@ -343,6 +343,7 @@ static int scmi_regulator_probe(struct scmi_device *sdev)
>>  	 * plausible SCMI Voltage Domain number, all belonging to this SCMI
>>  	 * platform instance node (handle->dev->of_node).
>>  	 */
>> +	of_node_get(handle->dev->of_node);
>>  	np = of_find_node_by_name(handle->dev->of_node, "regulators");
>>  	for_each_child_of_node(np, child) {
>>  		ret = process_scmi_regulator_of_node(sdev, ph, child, rinfo);
>> -- 
>> 2.25.1
>>