On 09/22/2014 12:46 AM, Matt Ranostay wrote:

On Sun, Sep 21, 2014 at 2:58 AM, Daniel Mack [off-list ref] wrote:

quoted

On 09/21/2014 05:01 AM, Matt Ranostay wrote:

quoted

quoted

@@ -313,12 +307,16 @@ static int cap1106_i2c_probe(struct i2c_client *i2c_client,

 static const struct of_device_id cap1106_dt_ids[] = {
      { .compatible = "microchip,cap1106", },
+     { .compatible = "microchip,cap1126", },
+     { .compatible = "microchip,cap1188", },

Hmm, how can that work unless you set .data to the number of channels
here? Did you test that with a DT-enabled board?

Yes it was tested on a BBB.  The num_channels is set from cap1106_i2c_ids

Ah ok. I forgot there's this fallback to the i2c ids. What others driver
do is to use of_match_device() in the probe function, and then access
->data of the returned match.

But I'm fine with falling back to cap1106_i2c_ids unless anyone else has
objections.


Thanks,
Daniel