Re: [PATCH 1/1] input: mma8450: Add chip ID check in probe

[Dmitry Torokhov <dmitry.torokhov@gmail.com>]

On Mon, Dec 16, 2024 at 11:30:15AM -0500, Frank Li wrote:
> On Sat, Dec 14, 2024 at 12:02:56AM +0100, Uwe Kleine-König wrote:
> > Hello Frank,
> >
> > On Fri, Dec 13, 2024 at 05:23:09PM -0500, Frank Li wrote:
> > > From: Luwei Zhou <b45643@freescale.com>
> > >
> > > Prevent continuous polling error logs by adding a chip ID check in the
> > > probe  function. This ensures the driver only proceeds when the mma8450 is
> > > present, avoiding issues in scenarios like missing add-on cards.
> > >
> > > Signed-off-by: Luwei Zhou <b45643@freescale.com>
> > > Signed-off-by: Fugang Duan <B38611@freescale.com>
> > > Signed-off-by: Vipul Kumar <vipul_kumar@mentor.com>
> > > Signed-off-by: Dong Aisheng <aisheng.dong@nxp.com>
> > > Signed-off-by: Frank Li <Frank.Li@nxp.com>
> > > ---
> > >  drivers/input/misc/mma8450.c | 16 +++++++++++++++-
> > >  1 file changed, 15 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/input/misc/mma8450.c b/drivers/input/misc/mma8450.c
> > > index 08412239b8e69..da941748ed29b 100644
> > > --- a/drivers/input/misc/mma8450.c
> > > +++ b/drivers/input/misc/mma8450.c
> > > @@ -38,6 +38,8 @@
> > >
> > >  #define MMA8450_CTRL_REG1	0x38
> > >  #define MMA8450_CTRL_REG2	0x39
> > > +#define MMA8450_ID		0xc6
> > > +#define MMA8450_WHO_AM_I	0x0f
> > >
> > >  static int mma8450_read(struct i2c_client *c, unsigned int off)
> > >  {
> > > @@ -148,8 +150,20 @@ static void mma8450_close(struct input_dev *input)
> > >   */
> > >  static int mma8450_probe(struct i2c_client *c)
> > >  {
> > > +	struct i2c_adapter *adapter = to_i2c_adapter(c->dev.parent);
> >
> > +	struct i2c_adapter *adapter = c->adapter;
> >
> > >  	struct input_dev *input;
> > > -	int err;
> > > +	int err, client_id;
> > > +
> > > +	err = i2c_check_functionality(adapter,
> > > +				      I2C_FUNC_SMBUS_BYTE | I2C_FUNC_SMBUS_BYTE_DATA);
> > > +	if (!err)
> > > +		return err;
> >
> > How unusual. I would have expected no ! here.
> 
> yes, it should be better
> 
> 	if (!i2c_check_functionality())
> 		....
> 
> i2c_check_functionality() return 1 if adapter supports everything we need.

Yes, it would be much better, otherwise in the case when an adapter does
not support the required functionality your code returns 0 which signals
successful probing.

Thanks.

-- 
Dmitry