Re: [1/2] ARM: irqchip: stm32: Optimizes and cleans up stm32-exti irq_domain

From: RadosÅaw Pietrzyk
Date: Fri Feb 23 2018 - 03:34:35 EST


Hi Ludovic,
Please check latest v2 patches series and let me know if it works on
stm32h7 as it does on stm32f4

2018-02-22 12:01 GMT+01:00 RadosÅaw Pietrzyk <radoslaw.pietrzyk@xxxxxxxxx>:
> Hi Ludovic,
> I have tested on stm32f429-disco and it works without a problem.
> handle_edge_irq is not used in current implementation because it is
> substituted by handle_simple_irq in an alloc callback either way which
> causes that irq_ack callback is not invoked as well (acking is done in
> chained handler).
>
> 2018-02-22 11:55 GMT+01:00 Ludovic BARRE <ludovic.barre@xxxxxx>:
>> Hi Radek
>>
>> I've tested your change on stm32h743i-eval.dts board
>> and gpio-keys tests are not functional.
>> No interrupt occurs when we push or release button
>> (cat /proc/interrupt).
>>
>> Board: stm32h743i-eval.dts
>> Test description: gpio-keys (gpioc 13 => exti 13), with gpio or interrupt
>> property.
>>
>> comment below
>>
>>
>> On 02/21/2018 02:50 PM, radek wrote:
>>>
>>> - allocates generic chip with handle_simple_irq
>>> which simplifies irq_domain_ops.alloc callback
>>> - removes ack register from generic chip which is not used for
>>> handle_simple_irq as acking is done in a chained handler
>>> - removes unneeded irq_domain_ops.xlate callback
>>>
>>> Signed-off-by: Radoslaw Pietrzyk <radoslaw.pietrzyk@xxxxxxxxx>
>>> ---
>>> drivers/irqchip/irq-stm32-exti.c | 10 +---------
>>> 1 file changed, 1 insertion(+), 9 deletions(-)
>>>
>>> diff --git a/drivers/irqchip/irq-stm32-exti.c
>>> b/drivers/irqchip/irq-stm32-exti.c
>>> index 36f0fbe..42e74e3 100644
>>> --- a/drivers/irqchip/irq-stm32-exti.c
>>> +++ b/drivers/irqchip/irq-stm32-exti.c
>>> @@ -176,16 +176,12 @@ static int stm32_irq_set_wake(struct irq_data *data,
>>> unsigned int on)
>>> static int stm32_exti_alloc(struct irq_domain *d, unsigned int virq,
>>> unsigned int nr_irqs, void *data)
>>> {
>>> - struct irq_chip_generic *gc;
>>> struct irq_fwspec *fwspec = data;
>>> irq_hw_number_t hwirq;
>>> hwirq = fwspec->param[0];
>>> - gc = irq_get_domain_generic_chip(d, hwirq);
>>> irq_map_generic_chip(d, virq, hwirq);
>>> - irq_domain_set_info(d, virq, hwirq, &gc->chip_types->chip, gc,
>>> - handle_simple_irq, NULL, NULL);
>>> return 0;
>>> }
>>> @@ -200,7 +196,6 @@ static void stm32_exti_free(struct irq_domain *d,
>>> unsigned int virq,
>>> struct irq_domain_ops irq_exti_domain_ops = {
>>> .map = irq_map_generic_chip,
>>> - .xlate = irq_domain_xlate_onetwocell,
>>> .alloc = stm32_exti_alloc,
>>> .free = stm32_exti_free,
>>> };
>>> @@ -231,7 +226,7 @@ __init stm32_exti_init(const struct stm32_exti_bank
>>> **stm32_exti_banks,
>>> }
>>> ret = irq_alloc_domain_generic_chips(domain, IRQS_PER_BANK, 1,
>>> "exti",
>>> - handle_edge_irq, clr, 0, 0);
>>> + handle_simple_irq, clr, 0,
>>> 0);
>>
>>
>> EXTI hardware block is trigged on pulse event rising or failing edge.
>> Why do you change to handle_simple_irq ?
>>
>>> if (ret) {
>>> pr_err("%pOF: Could not allocate generic interrupt
>>> chip.\n",
>>> node);
>>> @@ -246,13 +241,10 @@ __init stm32_exti_init(const struct stm32_exti_bank
>>> **stm32_exti_banks,
>>> gc->reg_base = base;
>>> gc->chip_types->type = IRQ_TYPE_EDGE_BOTH;
>>> - gc->chip_types->chip.irq_ack = irq_gc_ack_set_bit;
>>> gc->chip_types->chip.irq_mask = irq_gc_mask_clr_bit;
>>> gc->chip_types->chip.irq_unmask = irq_gc_mask_set_bit;
>>> gc->chip_types->chip.irq_set_type = stm32_irq_set_type;
>>> gc->chip_types->chip.irq_set_wake = stm32_irq_set_wake;
>>> - gc->chip_types->regs.ack = stm32_bank->pr_ofst;
>>> - gc->chip_types->regs.mask = stm32_bank->imr_ofst;
>>> gc->private = (void *)stm32_bank;
>>> /* Determine number of irqs supported */
>>>
>>
>> BR
>> Ludo