[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1b0de79d-87c9-46df-8b2f-f40bf58108c5@kili.mountain>
Date: Thu, 25 May 2023 08:05:58 +0300
From: Dan Carpenter <dan.carpenter@...aro.org>
To: Christophe JAILLET <christophe.jaillet@...adoo.fr>
Cc: Wang Zhang <silver_code@...t.edu.cn>,
Peter Korsgaard <peter@...sgaard.com>,
Andrew Lunn <andrew@...n.ch>,
hust-os-kernel-patches@...glegroups.com, linux-i2c@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3] i2c: ocores: use devm_ managed clks
On Wed, May 24, 2023 at 09:21:56PM +0200, Christophe JAILLET wrote:
> > + if (ret) {
> > + dev_err(dev,
> > + "clk_prepare_enable failed: %d\n", ret);
> > + return ret;
> > }
> > + rate = clk_get_rate(i2c->clk) / 1000;
>
> Now (because of the devm_clk_get_optional_enabled()), i2c->clk can be NULL,
> so this would deference a NULL pointer.
>
No, it's fine. clk_get_rate() checks for NULL. When a function returns
a mix of error pointers and NULL, like devm_clk_get_optional_enabled(),
then all the functions like this must have NULL checks built it.
regards,
dan carpenter
Powered by blists - more mailing lists