Re: [PATCH] net: mdio-ipq4019: fix possible invalid pointer dereference

[Hui Tang]

On 2022/11/15 21:36, Andrew Lunn wrote:
> On Tue, Nov 15, 2022 at 12:50:28PM +0800, Hui Tang wrote:
> > priv->eth_ldo_rdy is saved the return value of devm_ioremap_resource(),
> > which !IS_ERR() should be used to check.
> >
> > Fixes: 23a890d493e3 ("net: mdio: Add the reset function for IPQ MDIO driver")
> > Signed-off-by: Hui Tang <tanghui20@xxxxxxxxxx>
> > ---
> >  drivers/net/mdio/mdio-ipq4019.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/net/mdio/mdio-ipq4019.c b/drivers/net/mdio/mdio-ipq4019.c
> > index 4eba5a91075c..d7a1f7c56f97 100644
> > --- a/drivers/net/mdio/mdio-ipq4019.c
> > +++ b/drivers/net/mdio/mdio-ipq4019.c
> > @@ -188,7 +188,7 @@ static int ipq_mdio_reset(struct mii_bus *bus)
> >  	/* To indicate CMN_PLL that ethernet_ldo has been ready if platform resource 1
> >  	 * is specified in the device tree.
> >  	 */
> > -	if (priv->eth_ldo_rdy) {
> > +	if (!IS_ERR(priv->eth_ldo_rdy)) {
> >  		val = readl(priv->eth_ldo_rdy);
> >  		val |= BIT(0);
> >  		writel(val, priv->eth_ldo_rdy);
>
> There is a general pattern in the kernel that optional things are set
> to NULL if the resource does not exist. Often there is a
> get_foo_optional() which will return a NULL point if the things does
> not exist, the thing if it does exist, or an error code if a real
> error happened.
>
> So please follow this patterns. Check the return value in
> ipq4019_mdio_probe(). Looking at __devm_ioremap_resource() i _think_
> it returns -ENOMEM if the resource does not exist? So maybe any other
> error is a real error, and should be reported, and -ENOMEM should
> result in eth_ldo_rdy set to NULL?

Thanks, I will resend it according to the style you said.