[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <5067db4b-33fc-5737-8c05-5ab41c50f25a@gmail.com>
Date: Wed, 3 Apr 2019 22:59:46 +0200
From: Heiner Kallweit <hkallweit1@...il.com>
To: Andrew Lunn <andrew@...n.ch>
Cc: Florian Fainelli <f.fainelli@...il.com>,
David Miller <davem@...emloft.net>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: Re: [PATCH net-next 1/2] net: phy: allow a PHY driver to define
neither features nor get_features
On 03.04.2019 22:46, Andrew Lunn wrote:
>> @@ -2218,9 +2223,9 @@ int phy_driver_register(struct phy_driver *new_driver, struct module *owner)
>> /* Either the features are hard coded, or dynamically
>> * determine. It cannot be both or neither
>> */
>
> Hi Heiner
>
> The comment needs updating to match the code.
>
Indeed, I have to fix this.
>> - if (WARN_ON((!new_driver->features && !new_driver->get_features) ||
>> - (new_driver->features && new_driver->get_features))) {
>> - pr_err("%s: Driver features are missing\n", new_driver->name);
>> + if (WARN_ON(new_driver->features && new_driver->get_features)) {
>> + pr_err("%s: features and get_features must not both be set\n",
>> + new_driver->name);
>> return -EINVAL;
>
> Andrew
>
Powered by blists - more mailing lists