Re: [PATCH] net: rswitch: Fix a possible memory leak in rswitch_phy_device_init()

[Andrew Lunn]

On Mon, Sep 09, 2024 at 05:28:25PM +0800, Jinjie Ruan wrote:
> of_phy_find_device() calls bus_find_device(), which calls get_device()
> on the returned struct device * to increment the refcount. The current
> implementation does not decrement the refcount, which causes memory leak.
> 
> So add the missing phy_device_free() call to decrement the
> refcount via put_device() to balance the refcount.
> 
> Fixes: 0df024d0f1d3 ("net: renesas: rswitch: Add host_interfaces setting")
> Signed-off-by: Jinjie Ruan <ruanjinjie@xxxxxxxxxx>
> ---
>  drivers/net/ethernet/renesas/rswitch.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/net/ethernet/renesas/rswitch.c b/drivers/net/ethernet/renesas/rswitch.c
> index ff50e20856ec..69a67bd75f33 100644
> --- a/drivers/net/ethernet/renesas/rswitch.c
> +++ b/drivers/net/ethernet/renesas/rswitch.c
> @@ -1404,6 +1404,7 @@ static int rswitch_phy_device_init(struct rswitch_device *rdev)
>  		goto out;
>  	__set_bit(rdev->etha->phy_interface, phydev->host_interfaces);
>  	phydev->mac_managed_pm = true;
> +	phy_device_free(phydev);
>  
>  	phydev = of_phy_connect(rdev->ndev, phy, rswitch_adjust_link, 0,
>  				rdev->etha->phy_interface);

This has the same problem as discussed in

net: ethernet: nxp: Fix a possible memory leak in lpc_mii_probe()

I've not looked to see if there are more of these 'fixes'. If there
are, it would be good to self NACK them so they don't get accidentally
merged.

    Andrew

---
pw-bot: cr