Re: [PATCH v4 3/3] MFD MAX8998/LP3974: Support Charger

[MyungJoo Ham]

On Tue, Jan 4, 2011 at 10:56 PM, Mark Brown
<broonie@xxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
> On Tue, Jan 04, 2011 at 02:17:41PM +0900, MyungJoo Ham wrote:
>> With the new regulator, "CHARGER", users can control charging
>> current and turn on and off the charger. Note that the charger
>> specification of LP3974 is different from that of MAX8998.
>>
>> driver/power/max8998.c supports power supply APIs for
>
> It's probably better to split the power supply driver into a separate
> patch as there should be no build time dependency between the two. ÂI've
> added Anton to the CCs as he is the drivers/power maintainer.
>
> One thing that jumps out here is that there's no regulator API usage at
> all in the power driver - the power driver just jumps straight in
> and updates registers when it needs to do anything. ÂAre you sure that
> the regulator API driver is needed at all, if it is I'd expect to see
> use of it in the power API.
>
>> +static const char * const manufacturers[] = {
>> + Â Â [TYPE_MAX8998] = "Maxim",
>> + Â Â [TYPE_LP3974] = "National Semiconductor",
>> + Â Â [TYPE_UNKNOWN] = "Unknown",
>
> Normally this refers to the battery rather than the chip.
>
>> +static void _max8998_update_eoc(struct platform_device *pdev)
>> +{
>> + Â Â struct max8998_battery_data *max8998 = platform_get_drvdata(pdev);
>> + Â Â struct i2c_client *i2c = max8998->iodev->i2c;
>> + Â Â int charge_current = 0;
>> + Â Â int target_eoc_ratio;
>> + Â Â u8 val;
>> +
>> + Â Â /* Nothing to do. User set EOC with % */
>> + Â Â if (max8998->eoc_in_mA == 0)
>> + Â Â Â Â Â Â return;
>> +
>> + Â Â /* Not initialized. */
>> + Â Â if (!charger_current_map_desc)
>> + Â Â Â Â Â Â return;
>
> This should be be driver data rather than a global for neatness (though
> in reality the chances of more than one of these chargers in a system
> are pretty low).

Ok. I'll make it possible to have max8998 and lp3974 at the same time
on a board. :)

>
>> --- a/drivers/regulator/max8998.c
>> +++ b/drivers/regulator/max8998.c
>> @@ -86,6 +86,13 @@ static const struct voltage_map_desc buck3_voltage_map_desc = {
>> Âstatic const struct voltage_map_desc buck4_voltage_map_desc = {
>> Â Â Â .min = 800, Â Â .step = 100, Â Â.max = 2300,
>> Â};
>> +static const int charger_current_map_desc_max8998[] = {
>> + Â Â 90, 380, 475, 550, 570, 600, 700, 800
>> +};
>> +static const int charger_current_map_desc_lp3974[] = {
>> + Â Â 100, 400, 450, 500, 550, 600, 700, 800
>> +};
>> +const int *charger_current_map_desc;
>
> Ah, this is exported from the regulator driver... ÂThat's slightly odd.
> For voltages we've an enumeration API for the supported settings, we
> probably ought to add that for current regulators too.
>
>> + Â Â dev_info(&rdev->dev, "charger current limit = %dmA (%xh)\n",
>> + Â Â Â Â Â Â Â Â Â Â chosen_current, chosen);
>
> dev_dbg() or remove it entirely please, otherwise it might get a bit
> noisy.

Ok.

>
>> +static struct regulator_ops max8998_charger_ops = {
>> +   .is_enabled       = max8998_ldo_is_enabled_negated,
>> + Â Â /* enable and disable are intentionally negated */
>> +   .enable         = max8998_ldo_disable,
>> +   .disable        Â= max8998_ldo_enable,
>
> That's really confusing... Âwhy?
>

I've negated max8998_charger_ops's enable/disable because other
max8998 registers use 1 for "ON" while this register use 0 for "ON".

-- 
MyungJoo Ham (íëì), Ph.D.
Mobile Software Platform Lab,
Digital Media and Communications (DMC) Business
Samsung Electronics
cell: 82-10-6714-2858
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/