Re: [PATCH v2 1/3] hwmon: (ltc2992) Fix memory leak in ltc2992_parse_dt()
From: Guenter Roeck
Date: Wed May 29 2024 - 18:35:22 EST
On Thu, May 23, 2024 at 05:47:14PM +0200, Javier Carrasco wrote:
> A new error path was added to the fwnode_for_each_available_node() loop
> in ltc2992_parse_dt(), which leads to an early return that requires a
> call to fwnode_handle_put() to avoid a memory leak in that case.
>
> Add the missing fwnode_handle_put() in the error path from a zero value
> shunt resistor.
>
> Cc: stable@xxxxxxxxxxxxxxx
> Fixes: 10b029020487 ("hwmon: (ltc2992) Avoid division by zero")
> Signed-off-by: Javier Carrasco <javier.carrasco.cruz@xxxxxxxxx>
Applied.
Thanks,
Guenter
> ---
> drivers/hwmon/ltc2992.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/hwmon/ltc2992.c b/drivers/hwmon/ltc2992.c
> index 229aed15d5ca..d4a93223cd3b 100644
> --- a/drivers/hwmon/ltc2992.c
> +++ b/drivers/hwmon/ltc2992.c
> @@ -876,9 +876,11 @@ static int ltc2992_parse_dt(struct ltc2992_state *st)
>
> ret = fwnode_property_read_u32(child, "shunt-resistor-micro-ohms", &val);
> if (!ret) {
> - if (!val)
> + if (!val) {
> + fwnode_handle_put(child);
> return dev_err_probe(&st->client->dev, -EINVAL,
> "shunt resistor value cannot be zero\n");
> + }
> st->r_sense_uohm[addr] = val;
> }
> }