Hi,
On 12-09-16 07:32, Peter Meerwald-Stadler wrote:
+static const struct iio_chan_spec mc3230_channels[] = {
+ MC3230_CHANNEL(MC3230_REG_XOUT, X),
+ MC3230_CHANNEL(MC3230_REG_YOUT, Y),
+ MC3230_CHANNEL(MC3230_REG_ZOUT, Z),
+};
+
+struct mc3230_data {
+ struct i2c_client *client;
+};
probably don't need mc3230_data (or the client pointer within), not worth
changing as likely stuff will be added in the future
I was thinking the same, but since iio does not use
set / get private_data but instead allocs private data
as part of devm_iio_device_alloc, this seemed cleaner then doing:
devm_iio_device_alloc(&client->dev, sizeof(struct i2c_client *));
And then doing somewhat ugly casts to retrieve the client pointer again.
+static int mc3230_set_opcon(struct mc3230_data *data, int opcon)
+{
+ int ret;
+ struct i2c_client *client = data->client;
+
guess we want locking here
This register only gets touched by this function, and this
function only gets called:
-pre iio_device_register
-post iio_device_unregister
-on suspend/resume
So I do not see what it could race with.
I didn't check where set_opcon() is actually called, seems safe
I think there is a theoretic race in _remove() as _unregister() retval is
not checked, no need to change
OK, I'll do a v2 adding the missing static then.
Regards,
Hans
--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
