[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <d5f0dea1b9ce5f8d2187875adb1d73e747e21916.camel@redhat.com>
Date: Thu, 10 Nov 2022 14:06:47 +0100
From: Paolo Abeni <pabeni@...hat.com>
To: Siddharth Vadapalli <s-vadapalli@...com>, davem@...emloft.net,
edumazet@...gle.com, kuba@...nel.org, robh+dt@...nel.org,
krzysztof.kozlowski@...aro.org, krzysztof.kozlowski+dt@...aro.org,
linux@...linux.org.uk, vladimir.oltean@....com, vigneshr@...com,
nsekhar@...com
Cc: netdev@...r.kernel.org, devicetree@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
srk@...com
Subject: Re: [PATCH net-next v5 3/3] net: ethernet: ti: am65-cpsw: Add
support for SERDES configuration
hello,
On Wed, 2022-11-09 at 09:52 +0530, Siddharth Vadapalli wrote:
[...]
> +static void am65_cpsw_disable_serdes_phy(struct am65_cpsw_common *common)
> +{
> + struct device_node *node, *port_np;
> + struct device *dev = common->dev;
> + const char *name = "serdes-phy";
> + struct phy *phy;
> +
> + node = of_get_child_by_name(dev->of_node, "ethernet-ports");
> +
> + for_each_child_of_node(node, port_np) {
> + phy = devm_of_phy_get(dev, port_np, name);
The above will try to allocate some memory and can fail. Even if the
the following code will handle a NULL ptr, the phy will not be
disabled.
I think it's better if you cache the serdes phy ptr in
am65_cpsw_init_serdes_phy() and you use such reference here, without
resorting to devm_of_phy_get().
Cheers,
Paolo
Powered by blists - more mailing lists