iio: adc: mxs-lradc-adc: Get rid of OF specifics
Subsystems (2)
PATCH v2 1/1 iio: adc: mxs-lradc-adc: Get rid of OF specifics
First of all, the additional conversion from vIRQ, and this is exactly what is returned by platform_get_irq_byname(), to vIRQ is not needed. Hence, drop no-op call to irq_of_parse_and_map().
Second, assign the firmware node instead of of_node.
---
v2: added property.h (Fabio), added tag (Fabio)
drivers/iio/adc/mxs-lradc-adc.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-)
drivers/iio/adc/mxs-lradc-adc.c +4 -6
@@ -17,8 +17,8 @@#include <linux/mfd/core.h> #include <linux/mfd/mxs-lradc.h> #include <linux/module.h> -#include <linux/of_irq.h> #include <linux/platform_device.h> +#include <linux/property.h> #include <linux/sysfs.h> #include <linux/iio/buffer.h>
@@ -692,7 +692,7 @@ static int mxs_lradc_adc_probe(struct platform_device *pdev)struct mxs_lradc_adc *adc;struct iio_dev *iio;struct resource *iores; -int ret, irq, virq, i, s, n; +int ret, irq, i, s, n;u64 scale_uv;const char **irq_name;
@@ -721,7 +721,7 @@ static int mxs_lradc_adc_probe(struct platform_device *pdev)platform_set_drvdata (pdev, iio); iio->name = pdev->name; - iio->dev.of_node = dev->parent->of_node; +device_set_node (&iio->dev,dev_fwnode (dev->parent)); iio->info = &mxs_lradc_adc_iio_info; iio->modes =INDIO_DIRECT_MODE ; iio->masklength =LRADC_MAX_TOTAL_CHANS ;
@@ -747,9 +747,7 @@ static int mxs_lradc_adc_probe(struct platform_device *pdev)if (irq <0 )return irq; - virq =irq_of_parse_and_map (dev->parent->of_node, irq); - - ret =devm_request_irq (dev, virq, mxs_lradc_adc_handle_irq, + ret =devm_request_irq (dev, irq, mxs_lradc_adc_handle_irq,0 , irq_name[i], iio);if (ret)return ret;
Signature · 1 line
2.35.1
Quoted 3 lines · Andy Shevchenko
First of all, the additional conversion from vIRQ, and this is exactly what is returned by
platform_get_irq_byname(), to vIRQ is not needed. Hence, drop no-op call toirq_of_parse_and_map().
Actually this patch is not correct. The MFD supplies IRQ lines without any domain behind it, because of that the supplied list of IRQs is abstract to the hardware and has no meaning in the Linux kernel. I dunno how it's supposed to work (perhaps it involves the GIC driver hooks for IMX23/28 platforms).
Signature · 2 lines
With Best Regards, Andy Shevchenko