I made a mistake assuming the device tree configuration for interrupt
triggering was somehow passed to the SPI device but it's not.
In the Motorola Linux kernel tree CPCAP PMIC is configured as a rising
edge triggered interrupt, but then then it's interrupt handler keeps
looping until the GPIO line goes down. So the CPCAP interrupt is clearly
a level interrupt and not an edge interrupt.
Earlier when I tried to configure it as level interrupt using the
device tree, I did not account that the triggering only gets passed
to the SPI core and it also needs to be specified in the CPCAP driver
when we do devm_regmap_add_irq_chip().
Fixes: 56e1d40d3bea ("mfd: cpcap: Add minimal support")
Cc: Charles Keepax <ckee...@opensource.wolfsonmicro.com>
Cc: Marcel Partap <mpar...@gmx.net>
Cc: Michael Scott <michael.sc...@linaro.org>
Cc: Sebastian Reichel <s...@kernel.org>
Signed-off-by: Tony Lindgren <t...@atomide.com>
---
drivers/mfd/motorola-cpcap.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/mfd/motorola-cpcap.c b/drivers/mfd/motorola-cpcap.c
--- a/drivers/mfd/motorola-cpcap.c
+++ b/drivers/mfd/motorola-cpcap.c
@@ -126,7 +126,7 @@ static int cpcap_init_irq_chip(struct cpcap_ddata *cpcap,
int irq_chip,
ret = devm_regmap_add_irq_chip(&cpcap->spi->dev, cpcap->regmap,
cpcap->spi->irq,
- IRQF_TRIGGER_RISING |
+ irq_get_trigger_type(cpcap->spi->irq) |
IRQF_SHARED, -1,
chip, &cpcap->irqdata[irq_chip]);
if (ret) {
--
2.12.2
