Re: [PATCH] iio: hmc: fix a potential NULL pointer dereference

[Jonathan Cameron <jic23@kernel.org>]

On Sat, 9 Mar 2019 14:42:09 +0100
Tomasz Duszynski <tduszyns@gmail.com> wrote:

> On Fri, Mar 08, 2019 at 11:15:32PM -0600, Kangjie Lu wrote:
> > devm_regmap_init_i2c may fail and return NULL. The fix returns
> > the error when it fails.
> >
> > Signed-off-by: Kangjie Lu <kjlu@umn.edu>
> > ---
> >  drivers/iio/magnetometer/hmc5843_i2c.c | 7 ++++++-
> >  1 file changed, 6 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/iio/magnetometer/hmc5843_i2c.c b/drivers/iio/magnetometer/hmc5843_i2c.c
> > index 3de7f4426ac4..c0cd0823f8d5 100644
> > --- a/drivers/iio/magnetometer/hmc5843_i2c.c
> > +++ b/drivers/iio/magnetometer/hmc5843_i2c.c
> > @@ -58,8 +58,13 @@ static const struct regmap_config hmc5843_i2c_regmap_config = {
> >  static int hmc5843_i2c_probe(struct i2c_client *cli,
> >  			     const struct i2c_device_id *id)
> >  {
> > +	struct regmap *devm_regmap = devm_regmap_init_i2c(cli,
> > +			&hmc5843_i2c_regmap_config);
> > +	if (IS_ERR(devm_regmap))
> > +		return PTR_ERR(devm_regmap);  
> 
> This fixes only one part of the problem leaving identical spi issue
> behind. I guess this check should be part of *common_probe().
It could be, but that somewhat hides the error checking.

I'd prefer to see it fixed like this, but in both here and the spi
file.  Both this and spi in one patch would be great!

Thanks,

Jonathan

> 
> > +
> >  	return hmc5843_common_probe(&cli->dev,
> > -			devm_regmap_init_i2c(cli, &hmc5843_i2c_regmap_config),
> > +			devm_regmap,
> >  			id->driver_data, id->name);
> >  }
> >
> > --
> > 2.17.1
> >