Hi!

> Introduce the lm3697 LED driver for
> backlighting and display.
> 
> Datasheet location:
> http://www.ti.com/lit/ds/symlink/lm3697.pdf
> 
> Signed-off-by: Dan Murphy <dmurphy@ti.com>

> +
> +#define LM3697_HVLED1_2_3_A		0
> +#define LM3697_HVLED1_B_HVLED2_3_A	1
> +#define LM3697_HVLED2_B_HVLED1_3_A	2
> +#define LM3697_HVLED1_2_B_HVLED3_A	3
> +#define LM3697_HVLED3_B_HVLED1_2_A	4
> +#define LM3697_HVLED1_3_B_HVLED2_A	5
> +#define LM3697_HVLED1_A_HVLED2_3_B	6
> +#define LM3697_HVLED1_2_3_B		7

That's rather long and verbose way to describe a bitmap, right?

> +static const struct regmap_config lm3697_regmap_config = {
> +	.reg_bits = 8,
> +	.val_bits = 8,
> +
> +	.max_register = LM3697_CTRL_ENABLE,
> +	.reg_defaults = lm3697_reg_defs,
> +	.num_reg_defaults = ARRAY_SIZE(lm3697_reg_defs),
> +	.cache_type = REGCACHE_RBTREE,
> +};

Is rbtree good idea? You don't have that many registers.

> +static int lm3697_init(struct lm3697 *priv)
> +{
> +	int ret;
> +
....
> +		regmap_write(priv->regmap, LM3697_RESET, LM3697_SW_RESET);

No error checking required here?

> +	if (priv->control_bank_config < LM3697_HVLED1_2_3_A ||
> +	    priv->control_bank_config > LM3697_HVLED1_2_3_B) {
> +		dev_err(&priv->client->dev, "Control bank configuration is out of range\n");
> +		ret = -EINVAL;
> +		goto out;
> +	}
> +
> +	device_for_each_child_node(priv->dev, child) {
> +		led = &priv->leds[i];
> +
> +		ret = fwnode_property_read_u32(child, "reg", &led->control_bank);
> +		if (ret) {
> +			dev_err(&priv->client->dev, "reg DT property missing\n");
> +			goto child_out;
> +		}
> +
> +		fwnode_property_read_string(child, "linux,default-trigger",
> +					    &led->led_dev.default_trigger);
> +
> +		ret = fwnode_property_read_string(child, "label", &name);
> +		if (ret)
> +			snprintf(led->label, sizeof(led->label),
> +				"%s::", priv->client->name);
> +		else
> +			snprintf(led->label, sizeof(led->label),
> +				 "%s:%s", priv->client->name, name);
> +
> +
> +		led->priv = priv;
> +		led->led_dev.name = led->label;
> +		led->led_dev.brightness_set_blocking = lm3697_brightness_set;
> +
> +		ret = devm_led_classdev_register(priv->dev, &led->led_dev);
> +		if (ret) {
> +			dev_err(&priv->client->dev, "led register err: %d\n", ret);
> +			goto child_out;
> +		}
> +
> +		if (priv->control_bank_config == LM3697_HVLED1_2_3_A ||
> +		    priv->control_bank_config == LM3697_HVLED1_2_3_B)
> +			goto child_out;

This checks if we have just one bank, I see it. Should it also check
the led actually uses the correct bank?

> +		i++;
> +		fwnode_handle_put(child);
> +	}
> +
> +child_out:
> +	fwnode_handle_put(child);

Is not the fwnode_handle_put() done twice for non-error case?

> +	ret = lm3697_init(led);
> +	if (ret)
> +		return ret;
> +
> +	return ret;
> +}

The if is not needed here.

> +static int lm3697_remove(struct i2c_client *client)
> +{
> +	struct lm3697 *led = i2c_get_clientdata(client);
> +	int ret;
> +
> +	ret = regmap_update_bits(led->regmap, LM3697_CTRL_ENABLE,
> +				 LM3697_CTRL_A_B_EN, 0);
> +	if (ret) {
> +		dev_err(&led->client->dev, "Failed to disable regulator\n");
> +		return ret;

Misleading, this does nothing with regulators.

									Pavel

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html