[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20160627151008.GZ28202@sirena.org.uk>
Date: Mon, 27 Jun 2016 16:10:08 +0100
From: Mark Brown <broonie@...nel.org>
To: megous@...ous.com
Cc: dev@...ux-sunxi.org, linux-arm-kernel@...ts.infradead.org,
Liam Girdwood <lgirdwood@...il.com>,
open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2 04/14] regulator: SY8106A regulator driver
On Sat, Jun 25, 2016 at 05:45:01AM +0200, megous@...ous.com wrote:
> + config.init_data = of_get_regulator_init_data(dev, dev->of_node, &sy8106a_reg);
> + if (!config.init_data) {
> + return -EINVAL;
> + }
This is broken, constraints are entirely optional - the driver should
just instantiate no matter what (if anything) the constraints are. This
means people can read the current state even if there is no need for
runtime configuration.
> + dev_info(&i2c->dev, "SY8106A voltage at boot: %u mV\n", SY8106A_MIN_MV +
> + SY8106A_STEP_MV * (selector & SY8106A_REG_VOUT1_SEL_MASK));
This is just noise, remove it - if we want to list the voltage for
regulators at probe we should be doing it consistently for all
regulators in the core rather than in individual drivers.
> +/*
> + * This is useless for OF-enabled devices, but it is needed by I2C subsystem
> + */
> +static const struct i2c_device_id sy8106a_i2c_id[] = {
> + { },
> +};
> +MODULE_DEVICE_TABLE(i2c, sy8106a_i2c_id);
No, this is not the case - supporting DT does not mean that other
enumeration mechanisms can't be supported and there's no reason not to
list a sensible ID here.
Download attachment "signature.asc" of type "application/pgp-signature" (474 bytes)
Powered by blists - more mailing lists