linux-kernel - Re: [PATCH] iio: dac: ti-dac5571: make vref regulator optional

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Sat, 25 Aug 2018 09:55:31 +0100
From:   Jonathan Cameron <jic23@...nel.org>
To:     Marcus Folkesson <marcus.folkesson@...il.com>
Cc:     Hartmut Knaack <knaack.h@....de>,
        Lars-Peter Clausen <lars@...afoo.de>,
        Peter Meerwald-Stadler <pmeerw@...erw.net>,
        linux-iio@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] iio: dac: ti-dac5571: make vref regulator optional

On Fri, 24 Aug 2018 22:24:47 +0200
Marcus Folkesson <marcus.folkesson@...il.com> wrote:

> The `vref` regulator is declared as optional in the device-tree binding,
> but the driver does require it.
> 
> Go for the device-tree binding and make the `vref` regulator optional.
> 
> Signed-off-by: Marcus Folkesson <marcus.folkesson@...il.com>
Hmm. 
If you do a voltage get on a stub regulator it will return an
error (IIRC).

So I am unconvinced this actually makes any real difference?

What am I missing?

Jonathan

> ---
>  drivers/iio/dac/ti-dac5571.c | 30 ++++++++++++++++++++++--------
>  1 file changed, 22 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/iio/dac/ti-dac5571.c b/drivers/iio/dac/ti-dac5571.c
> index f6dcd8bce2b0..bf21cc312096 100644
> --- a/drivers/iio/dac/ti-dac5571.c
> +++ b/drivers/iio/dac/ti-dac5571.c
> @@ -251,6 +251,9 @@ static int dac5571_read_raw(struct iio_dev *indio_dev,
>  		return IIO_VAL_INT;
>  
>  	case IIO_CHAN_INFO_SCALE:
> +		if (!data->vref)
> +			return -EOPNOTSUPP;
> +
>  		ret = regulator_get_voltage(data->vref);
>  		if (ret < 0)
>  			return ret;
> @@ -335,13 +338,21 @@ static int dac5571_probe(struct i2c_client *client,
>  	indio_dev->num_channels = spec->num_channels;
>  	data->spec = spec;
>  
> -	data->vref = devm_regulator_get(dev, "vref");
> -	if (IS_ERR(data->vref))
> -		return PTR_ERR(data->vref);
> +	data->vref = devm_regulator_get_optional(dev, "vref");
> +	if (IS_ERR(data->vref)) {
> +		if (PTR_ERR(data->vref) == -ENODEV) {
> +			data->vref = NULL;
> +		} else {
> +			dev_err(dev, "failed to get regulator (%ld)\n",
> +					PTR_ERR(data->vref));
> +			return PTR_ERR(data->vref);

That is as likely as not a deferred probe so it should not be
as noisy as this.  It may happen during a normal boot process a lot
of times.


> +		}
>  
> -	ret = regulator_enable(data->vref);
> -	if (ret < 0)
> -		return ret;
> +	} else {
> +		ret = regulator_enable(data->vref);
> +		if (ret)
> +			return ret;
> +	}
>  
>  	mutex_init(&data->lock);
>  
> @@ -373,7 +384,9 @@ static int dac5571_probe(struct i2c_client *client,
>  	return 0;
>  
>   err:
> -	regulator_disable(data->vref);
> +	if (data->vref)
> +		regulator_disable(data->vref);
> +
>  	return ret;
>  }
>  
> @@ -383,7 +396,8 @@ static int dac5571_remove(struct i2c_client *i2c)
>  	struct dac5571_data *data = iio_priv(indio_dev);
>  
>  	iio_device_unregister(indio_dev);
> -	regulator_disable(data->vref);
> +	if (data->vref)
> +		regulator_disable(data->vref);
>  
>  	return 0;
>  }