Re: [PATCH V2 5/5] gpio/gpio-stmpe: ADD support for stmpe variant801

[Viresh Kumar]

On 11/18/2011 5:29 PM, Rabin Vincent wrote:
> On Thu, Nov 17, 2011 at 11:02, Viresh Kumar <viresh.kumar@xxxxxx> wrote:
>> diff --git a/drivers/gpio/gpio-stmpe.c b/drivers/gpio/gpio-stmpe.c
>> index 4c980b5..000b019 100644
>> --- a/drivers/gpio/gpio-stmpe.c
>> +++ b/drivers/gpio/gpio-stmpe.c
>> @@ -65,7 +65,15 @@ static void stmpe_gpio_set(struct gpio_chip *chip, unsigned offset, int val)
>> +       if (!val && (stmpe->regs[STMPE_IDX_GPSR_LSB] ==
>> +                               stmpe->regs[STMPE_IDX_GPCR_LSB]))
>> +               stmpe_set_bits(stmpe, reg, mask, ~mask);
>> +       else
>> +               stmpe_set_bits(stmpe, reg, mask, mask);
>>  }
> 
> This code,
> 
>  (1) for 801, when clearing one GPIO, sets all the others.

I assumed stmpe_set_bits will only affect bits which are 1
in mask and i was wrong. :(

>  (2) for other devices, adds an an unnecessary read (within stmpe_set_bits()),
>      which wasn't there before.
> 

Correct.

> Please rework to something like:
> 
> 	if (stmpe->regs[...)
> 		stmpe_set_bits(stmpe, reg, mask, val ? mask : 0);
> 	else
> 		stmpe_reg_write(stmpe, reg, mask);
> 

Sure.

>>
>>  static int stmpe_gpio_direction_output(struct gpio_chip *chip,
>> @@ -125,10 +133,19 @@ static struct gpio_chip template_chip = {
>>  static int stmpe_gpio_irq_set_type(struct irq_data *d, unsigned int type)
>>  {
>>        struct stmpe_gpio *stmpe_gpio = irq_data_get_irq_chip_data(d);
>> +       struct stmpe *stmpe = stmpe_gpio->stmpe;
>>        int offset = d->irq - stmpe_gpio->irq_base;
>>        int regoffset = offset / 8;
>>        int mask = 1 << (offset % 8);
>>
>> +       /* STMPE801 doesn't have RE and FE registers */
>> +       if (stmpe->partnum == STMPE801) {
>> +               if (type == IRQ_TYPE_LEVEL_LOW || type == IRQ_TYPE_LEVEL_HIGH)
>> +                       return 0;
> 
> This looks wrong.  From the datasheet I see that it supports edges only,
> so perhaps you meant to say IRQ_TYPE_EDGE_* above.
> 

I meant to say IRQ_TYPE_LEVEL* only, but i was wrong. I didn't read the manual
correctly. :(

> In that case please reorganize this to add the return 0 after the
> existing check which excludes levels (below).
> 

Sure

-- 
viresh
--
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/