[PATCH v3 1/2] mfd: add irq domain support for max8997 interrupts
Thomas Abraham
thomas.abraham at linaro.org
Tue Mar 13 02:05:53 EDT 2012
On 13 March 2012 08:58, Grant Likely <grant.likely at secretlab.ca> wrote:
> On Thu, 23 Feb 2012 18:08:07 +0530, Thomas Abraham <thomas.abraham at linaro.org> wrote:
>> Add irq domain support for max8997 interrupts. The reverse mapping method
>> used is linear mapping since the sub-drivers of max8997 such as regulator
>> and charger drivers can use the max8997 irq_domain to get the linux irq
>> number for max8997 interrupts. All uses of irq_base in platform data and
>> max8997 driver private data are removed.
>>
>> Cc: Grant Likely <grant.likely at secretlab.ca>
>> Cc: MyungJoo Ham <myungjoo.ham at samsung.com>
>> Signed-off-by: Thomas Abraham <thomas.abraham at linaro.org>
>> ---
>> arch/arm/mach-exynos/mach-nuri.c | 4 --
>> arch/arm/mach-exynos/mach-origen.c | 1 -
>> drivers/mfd/max8997-irq.c | 60 ++++++++++++++++++++--------------
>> drivers/mfd/max8997.c | 1 -
>> include/linux/mfd/max8997-private.h | 4 ++-
>> include/linux/mfd/max8997.h | 1 -
>> 6 files changed, 38 insertions(+), 33 deletions(-)
>
> Nice patch. Nice diffstat too. Some comments below, but I'm happy with where
> this is at.
>
> Acked-by: Grant Likely <grant.likely at secretlab.ca>
Thank you Grant for reviewing this patch.
>
>>
>> diff --git a/arch/arm/mach-exynos/mach-nuri.c b/arch/arm/mach-exynos/mach-nuri.c
>> index a6b9162..1e9ba12 100644
>> --- a/arch/arm/mach-exynos/mach-nuri.c
>> +++ b/arch/arm/mach-exynos/mach-nuri.c
>> @@ -1078,12 +1078,8 @@ static struct platform_device nuri_max8903_device = {
>> static void __init nuri_power_init(void)
>> {
>> int gpio;
>> - int irq_base = IRQ_GPIO_END + 1;
>> int ta_en = 0;
>>
>> - nuri_max8997_pdata.irq_base = irq_base;
>> - irq_base += MAX8997_IRQ_NR;
>> -
>> gpio = EXYNOS4_GPX0(7);
>> gpio_request(gpio, "AP_PMIC_IRQ");
>> s3c_gpio_cfgpin(gpio, S3C_GPIO_SFN(0xf));
>> diff --git a/arch/arm/mach-exynos/mach-origen.c b/arch/arm/mach-exynos/mach-origen.c
>> index 02c242e..328dadb 100644
>> --- a/arch/arm/mach-exynos/mach-origen.c
>> +++ b/arch/arm/mach-exynos/mach-origen.c
>> @@ -424,7 +424,6 @@ static struct max8997_platform_data __initdata origen_max8997_pdata = {
>> .buck1_gpiodvs = false,
>> .buck2_gpiodvs = false,
>> .buck5_gpiodvs = false,
>> - .irq_base = IRQ_GPIO_END + 1,
>>
>> .ignore_gpiodvs_side_effect = true,
>> .buck125_default_idx = 0x0,
>> diff --git a/drivers/mfd/max8997-irq.c b/drivers/mfd/max8997-irq.c
>> index 09274cf..3d8c9d1 100644
>> --- a/drivers/mfd/max8997-irq.c
>> +++ b/drivers/mfd/max8997-irq.c
>> @@ -142,7 +142,8 @@ static void max8997_irq_sync_unlock(struct irq_data *data)
>> static const inline struct max8997_irq_data *
>> irq_to_max8997_irq(struct max8997_dev *max8997, int irq)
>> {
>> - return &max8997_irqs[irq - max8997->irq_base];
>> + struct irq_data *data = irq_get_irq_data(irq);
>> + return &max8997_irqs[data->hwirq];
>
> Hahaha. The max8997_irqs table is kind of overdone when the group and mask values
> can be arithmetically derived from the hwirq number..
Yes, I agree. Each group could have maximum of 8 interrupts. Then,
data->hwirq % 8 will be the group number and data->hwirq & 7 would be
the mask. The enum max8997_irq can then be adjusted to accommodate
unused irq numbers.
>
> This change is fine, but the driver could use some refactoring. :-)
Ok.
>
>> }
>>
>> static void max8997_irq_mask(struct irq_data *data)
>> @@ -182,7 +183,7 @@ static irqreturn_t max8997_irq_thread(int irq, void *data)
>> u8 irq_reg[MAX8997_IRQ_GROUP_NR] = {};
>> u8 irq_src;
>> int ret;
>> - int i;
>> + int i, cur_irq;
>>
>> ret = max8997_read_reg(max8997->i2c, MAX8997_REG_INTSRC, &irq_src);
>> if (ret < 0) {
>> @@ -269,8 +270,10 @@ static irqreturn_t max8997_irq_thread(int irq, void *data)
>>
>> /* Report */
>> for (i = 0; i < MAX8997_IRQ_NR; i++) {
>> - if (irq_reg[max8997_irqs[i].group] & max8997_irqs[i].mask)
>> - handle_nested_irq(max8997->irq_base + i);
>> + if (irq_reg[max8997_irqs[i].group] & max8997_irqs[i].mask) {
>> + cur_irq = irq_find_mapping(max8997->irq_domain, i);
>> + handle_nested_irq(cur_irq);
>> + }
>
> Well, if you're using the linear mapping, then only actually allocated irqs will
> have a mapping set up. So the driver could simply do:
>
> for (i = 0; i < MAX8997_IRQ_NR; i++) {
> cur_irq = irq_find_mapping(max8997->irq_domain, i);
> if (cur_irq)
> handle_nested_irq(cur_irq);
> }
Thanks for the suggestion. I will include this change and repost this patch.
Regards,
Thomas.
[...]
More information about the linux-arm-kernel
mailing list