Message-ID: <20210724190601.084d43e7@jic23-huawei>
Date:   Sat, 24 Jul 2021 19:06:01 +0100
From:   Jonathan Cameron <jic23@...nel.org>
To:     Lad Prabhakar <prabhakar.mahadev-lad.rj@...renesas.com>
Cc:     Geert Uytterhoeven <geert+renesas@...der.be>,
        Rob Herring <robh+dt@...nel.org>,
        Lars-Peter Clausen <lars@...afoo.de>,
        Michael Turquette <mturquette@...libre.com>,
        Magnus Damm <magnus.damm@...il.com>,
        Stephen Boyd <sboyd@...nel.org>,
        Philipp Zabel <p.zabel@...gutronix.de>,
        Alexandru Ardelean <aardelean@...iqon.com>,
        linux-iio@...r.kernel.org, devicetree@...r.kernel.org,
        linux-clk@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-renesas-soc@...r.kernel.org,
        Prabhakar <prabhakar.csengg@...il.com>,
        Biju Das <biju.das.jz@...renesas.com>
Subject: Re: [PATCH v2 2/4] iio: adc: Add driver for Renesas RZ/G2L A/D
 converter

On Mon, 19 Jul 2021 09:58:38 +0100
Lad Prabhakar <prabhakar.mahadev-lad.rj@...renesas.com> wrote:

> Add ADC driver support for Renesas RZ/G2L A/D converter in SW
> trigger mode.
> 
> A/D Converter block is a successive approximation analog-to-digital
> converter with a 12-bit accuracy and supports a maximum of 8 input
> channels.
> 
> Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@...renesas.com>
> Reviewed-by: Biju Das <biju.das.jz@...renesas.com>

Hi Lad,

A few additional comments inline.

Thanks,

Jonathan

> ---
>  MAINTAINERS                 |   8 +
>  drivers/iio/adc/Kconfig     |  10 +
>  drivers/iio/adc/Makefile    |   1 +
>  drivers/iio/adc/rzg2l_adc.c | 545 ++++++++++++++++++++++++++++++++++++
>  4 files changed, 564 insertions(+)
>  create mode 100644 drivers/iio/adc/rzg2l_adc.c
> 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 6c8be735cc91..6a52f9f4604c 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -15839,6 +15839,14 @@ L:	linux-renesas-soc@...r.kernel.org
>  S:	Maintained
>  F:	drivers/phy/renesas/phy-rcar-gen3-usb*.c
>  
> +RENESAS RZ/G2L A/D DRIVER
> +M:	Lad Prabhakar <prabhakar.mahadev-lad.rj@...renesas.com>
> +L:	linux-iio@...r.kernel.org
> +L:	linux-renesas-soc@...r.kernel.org
> +S:	Supported
> +F:	Documentation/devicetree/bindings/iio/adc/renesas,rzg2l-adc.yaml
> +F:	drivers/iio/adc/rzg2l_adc.c
> +
>  RESET CONTROLLER FRAMEWORK
>  M:	Philipp Zabel <p.zabel@...gutronix.de>
>  S:	Maintained
> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
> index db0c8fb60515..af168e1c9fdb 100644
> --- a/drivers/iio/adc/Kconfig
> +++ b/drivers/iio/adc/Kconfig
> @@ -887,6 +887,16 @@ config ROCKCHIP_SARADC
>  	  To compile this driver as a module, choose M here: the
>  	  module will be called rockchip_saradc.
>  
> +config RZG2L_ADC
> +	tristate "Renesas RZ/G2L ADC driver"
> +	depends on ARCH_R9A07G044 || COMPILE_TEST
> +	help
> +	  Say yes here to build support for the ADC found in Renesas
> +	  RZ/G2L family.
> +
> +	  To compile this driver as a module, choose M here: the
> +	  module will be called rzg2l_adc.
> +
>  config SC27XX_ADC
>  	tristate "Spreadtrum SC27xx series PMICs ADC"
>  	depends on MFD_SC27XX_PMIC || COMPILE_TEST
> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
> index f70d877c555a..d68550f493e3 100644
> --- a/drivers/iio/adc/Makefile
> +++ b/drivers/iio/adc/Makefile
> @@ -82,6 +82,7 @@ obj-$(CONFIG_QCOM_PM8XXX_XOADC) += qcom-pm8xxx-xoadc.o
>  obj-$(CONFIG_RCAR_GYRO_ADC) += rcar-gyroadc.o
>  obj-$(CONFIG_RN5T618_ADC) += rn5t618-adc.o
>  obj-$(CONFIG_ROCKCHIP_SARADC) += rockchip_saradc.o
> +obj-$(CONFIG_RZG2L_ADC) += rzg2l_adc.o
>  obj-$(CONFIG_SC27XX_ADC) += sc27xx_adc.o
>  obj-$(CONFIG_SPEAR_ADC) += spear_adc.o
>  obj-$(CONFIG_STX104) += stx104.o
> diff --git a/drivers/iio/adc/rzg2l_adc.c b/drivers/iio/adc/rzg2l_adc.c
> new file mode 100644
> index 000000000000..033cdec9e976
> --- /dev/null
> +++ b/drivers/iio/adc/rzg2l_adc.c
> @@ -0,0 +1,545 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * RZ/G2L A/D Converter driver
> + *
> + *  Copyright (c) 2021 Renesas Electronics Europe GmbH
> + *
> + * Author: Lad Prabhakar <prabhakar.mahadev-lad.rj@...renesas.com>
> + */
> +
> +#include <linux/clk.h>
> +#include <linux/completion.h>
> +#include <linux/delay.h>
> +#include <linux/iio/iio.h>
> +#include <linux/interrupt.h>
> +#include <linux/io.h>
> +#include <linux/module.h>
> +#include <linux/of.h>
> +#include <linux/of_device.h>

Add mod_devicetable.h and drop the of specific headers
as nothing in them used directly as far as I can tell.

> +#include <linux/platform_device.h>
> +#include <linux/pm_runtime.h>
> +#include <linux/reset.h>
> +
> +#define DRIVER_NAME		"rzg2l-adc"
> +
> +#define RZG2L_ADM(n)			((n) * 0x4)
> +#define RZG2L_ADM0_ADCE			BIT(0)
> +#define RZG2L_ADM0_ADBSY		BIT(1)
> +#define RZG2L_ADM0_PWDWNB		BIT(2)
> +#define RZG2L_ADM0_SRESB		BIT(15)
> +#define RZG2L_ADM1_TRG			BIT(0)
> +#define RZG2L_ADM1_MS			BIT(2)
> +#define RZG2L_ADM1_BS			BIT(4)
> +#define RZG2L_ADM1_EGA_CLEAR		~GENMASK(13, 12)
> +#define RZG2L_ADM2_CHSEL_CLEAR		~GENMASK(7, 0)

I'd rather see these defined as _MASK and the
inversion made explicit inline.

> +#define RZG2L_ADM3_ADSMP		0x578
> +#define RZG2L_ADM3_ADCMP		(0xe << 16)
> +#define RZG2L_ADM3_ADIL_CLEAR		~GENMASK(31, 24)
> +
> +#define RZG2L_ADINT			0x20
> +#define RZG2L_ADINT_CH_CLEAR		~GENMASK(7, 0)
> +#define RZG2L_ADINT_CSEEN		BIT(16)
> +#define RZG2L_ADINT_INTS		BIT(31)
> +#define RZG2L_ADSTS			0x24
> +#define RZG2L_ADINT_INTST_MASK		GENMASK(7, 0)
> +#define RZG2L_ADSTS_CSEST		BIT(16)
> +#define RZG2L_ADIVC			0x28
> +#define RZG2L_ADIVC_DIVADC_CLEAR	~GENMASK(8, 0)
> +#define RZG2L_ADIVC_DIVADC_4		0x4

Perhaps adopt the _MASK and FIELD_PREP() approach to setting
particular values.  That way it is obvious at the call site that
DVADC_4 is a value within the DIVADC_MASK covered field.

> +#define RZG2L_ADFIL			0x2c
> +#define RZG2L_ADCR(n)			(0x30 + ((n) * 0x4))
> +#define RZG2L_ADCR_AD_MASK		GENMASK(11, 0)
> +
> +#define RZG2L_ADC_MAX_CHANNELS		8
> +#define RZG2L_ADC_CHN_MASK		0x7
> +#define RZG2L_ADC_TIMEOUT		usecs_to_jiffies(1 * 4)
> +
> +struct rzg2l_adc_data {
> +	const struct iio_chan_spec *channels;
> +	u8 num_channels;
> +};
> +
> +struct rzg2l_adc {
> +	void __iomem *base;
> +	struct clk *pclk;
> +	struct clk *adclk;
> +	struct reset_control *presetn;
> +	struct reset_control *adrstn;
> +	struct completion completion;
> +	const struct rzg2l_adc_data *data;
> +	u16 last_val[RZG2L_ADC_MAX_CHANNELS];
> +};
> +
> +static const char * const rzg2l_adc_channel_name[] = {
> +	"adc0",
Is it useful to expose these as labels to userspace?
Seems unnecessary given they map directly to the channel
numbers.

> +	"adc1",
> +	"adc2",
> +	"adc3",
> +	"adc4",
> +	"adc5",
> +	"adc6",
> +	"adc7",
> +};
> +
> +static unsigned int rzg2l_adc_readl(struct rzg2l_adc *adc, u32 reg)
> +{
> +	return readl(adc->base + reg);
> +}
> +
> +static void rzg2l_adc_writel(struct rzg2l_adc *adc, unsigned int reg, u32 val)
> +{
> +	writel(val, adc->base + reg);
> +}
> +
> +static void rzg2l_adc_pwr(struct rzg2l_adc *adc, bool on)
> +{
> +	u32 reg;
> +
> +	reg = rzg2l_adc_readl(adc, RZG2L_ADM(0));
> +	if (on)
> +		reg |= RZG2L_ADM0_PWDWNB;
> +	else
> +		reg &= ~RZG2L_ADM0_PWDWNB;
> +	rzg2l_adc_writel(adc, RZG2L_ADM(0), reg);
> +	udelay(2);
> +}
> +
> +static void rzg2l_adc_start_stop(struct rzg2l_adc *adc, bool start)
> +{
> +	int timeout = 5;
> +	u32 reg;
> +
> +	/* stop A/D conversion */
> +	reg = rzg2l_adc_readl(adc, RZG2L_ADM(0));
> +	if (start)
> +		reg |= RZG2L_ADM0_ADCE;
> +	else
> +		reg &= ~RZG2L_ADM0_ADCE;
> +	rzg2l_adc_writel(adc, RZG2L_ADM(0), reg);
> +
> +	if (start)
> +		return;
> +
> +	do {
> +		usleep_range(100, 200);
> +		reg = rzg2l_adc_readl(adc, RZG2L_ADM(0));
> +		timeout--;
> +		if (!timeout) {
> +			pr_err("%s stopping ADC timed out\n", __func__);
> +			break;
> +		}
> +	} while (((reg & RZG2L_ADM0_ADBSY) || (reg & RZG2L_ADM0_ADCE)));
> +}
> +
> +static void rzg2l_set_trigger(struct rzg2l_adc *adc)
> +{
> +	u32 reg;
> +
> +	/* SW trigger */
> +	reg = rzg2l_adc_readl(adc, RZG2L_ADM(1));
> +	reg &= RZG2L_ADM1_EGA_CLEAR;
> +	reg &= ~RZG2L_ADM1_BS;
> +	reg |= RZG2L_ADM1_MS;
> +	reg &= ~RZG2L_ADM1_TRG;
> +	rzg2l_adc_writel(adc, RZG2L_ADM(1), reg);
> +}
> +
> +static int rzg2l_adc_conversion_setup(struct rzg2l_adc *adc, u8 ch)
> +{
> +	u32 reg;
> +
> +	if (rzg2l_adc_readl(adc, RZG2L_ADM(0)) & RZG2L_ADM0_ADBSY)
> +		return -EBUSY;
> +
> +	rzg2l_set_trigger(adc);
> +
> +	/* select channel */
> +	reg = rzg2l_adc_readl(adc, RZG2L_ADM(2));
> +	reg &= RZG2L_ADM2_CHSEL_CLEAR;
> +	reg |= BIT(ch);
> +	rzg2l_adc_writel(adc, RZG2L_ADM(2), reg);
> +
> +	reg = rzg2l_adc_readl(adc, RZG2L_ADM(3));
> +	reg &= RZG2L_ADM3_ADIL_CLEAR;
> +	reg |= RZG2L_ADM3_ADCMP;
> +	reg |= RZG2L_ADM3_ADSMP;

No loss in readability in combining the two lines above and shorter
code is always good.  Having the mask on a separate line makes
sense but the |= pair doesn't.

What is a bit unusual here is you clear some bits then write different
ones.  That is presumably relying on the fact that the ADCCMP part
is always set to the same value.  That seems unwise to assume from
a long term maintainability point of view.

> +	rzg2l_adc_writel(adc, RZG2L_ADM(3), reg);
> +
> +	reg = rzg2l_adc_readl(adc, RZG2L_ADIVC);
> +	reg &= RZG2L_ADIVC_DIVADC_CLEAR;
> +	reg |= RZG2L_ADIVC_DIVADC_4;
> +	rzg2l_adc_writel(adc, RZG2L_ADIVC, reg);
> +
> +	reg = rzg2l_adc_readl(adc, RZG2L_ADINT);
> +	reg &= ~RZG2L_ADINT_INTS;
> +	reg &= RZG2L_ADINT_CH_CLEAR;
> +	reg |= RZG2L_ADINT_CSEEN;
> +	reg |= BIT(ch);
> +	rzg2l_adc_writel(adc, RZG2L_ADINT, reg);
> +
> +	return 0;
> +}
> +
> +static int rzg2l_adc_set_power(struct iio_dev *indio_dev, bool on)
> +{
> +	struct device *dev = indio_dev->dev.parent;
> +
> +	if (on)
> +		return pm_runtime_resume_and_get(dev);
> +
> +	return pm_runtime_put_sync(dev);
> +}
> +
> +static int rzg2l_adc_conversion(struct iio_dev *indio_dev, struct rzg2l_adc *adc, u8 ch)
> +{
> +	int ret;
> +
> +	ret = rzg2l_adc_set_power(indio_dev, true);
> +	if (ret)
> +		return ret;
> +
> +	ret = rzg2l_adc_conversion_setup(adc, ch);
> +	if (ret) {
> +		rzg2l_adc_set_power(indio_dev, false);
> +		return ret;
> +	}
> +
> +	reinit_completion(&adc->completion);
> +
> +	rzg2l_adc_start_stop(adc, true);
> +
> +	if (!wait_for_completion_timeout(&adc->completion, RZG2L_ADC_TIMEOUT)) {
> +		rzg2l_adc_writel(adc, RZG2L_ADINT,
> +				 rzg2l_adc_readl(adc, RZG2L_ADINT) & RZG2L_ADINT_CH_CLEAR);
> +		rzg2l_adc_start_stop(adc, false);
> +		rzg2l_adc_set_power(indio_dev, false);
> +		return -ETIMEDOUT;
> +	}
> +
> +	return rzg2l_adc_set_power(indio_dev, false);
> +}
> +
> +static int rzg2l_adc_read_raw(struct iio_dev *indio_dev,
> +			      struct iio_chan_spec const *chan,
> +			      int *val, int *val2, long mask)
> +{
> +	struct rzg2l_adc *adc = iio_priv(indio_dev);
> +	int ret;
> +	u8 ch;
> +
> +	switch (mask) {
> +	case IIO_CHAN_INFO_RAW:
> +		if (chan->type != IIO_VOLTAGE)
> +			return -EINVAL;
> +
> +		ret = iio_device_claim_direct_mode(indio_dev);
> +		if (ret)
> +			return ret;
> +
> +		ch = chan->channel & RZG2L_ADC_CHN_MASK;
> +		ret = rzg2l_adc_conversion(indio_dev, adc, ch);
> +		if (ret) {
> +			iio_device_release_direct_mode(indio_dev);
> +			return ret;
> +		}
> +		*val = adc->last_val[ch];
> +		iio_device_release_direct_mode(indio_dev);
> +
> +		return IIO_VAL_INT;
> +
> +	default:
> +		return -EINVAL;
> +	}
> +}
> +
> +static int rzg2l_adc_read_label(struct iio_dev *iio_dev,
> +				const struct iio_chan_spec *chan,
> +				char *label)
> +{
> +	if (chan->channel >= RZG2L_ADC_MAX_CHANNELS)
> +		return -EINVAL;
> +
> +	return snprintf(label, PAGE_SIZE, "%s\n", rzg2l_adc_channel_name[chan->channel]);

sysfs_emit()?.

> +}
> +
> +static const struct iio_info rzg2l_adc_iio_info = {
> +	.read_raw = rzg2l_adc_read_raw,
> +	.read_label = rzg2l_adc_read_label,
> +};
> +
> +static irqreturn_t rzg2l_adc_isr(int irq, void *dev_id)
> +{
> +	struct rzg2l_adc *adc = (struct rzg2l_adc *)dev_id;
> +	u8 intst;
> +	u32 reg;
> +	u8 i;
> +
> +	reg = rzg2l_adc_readl(adc, RZG2L_ADSTS);
> +	if (reg & RZG2L_ADSTS_CSEST) {

Probably good to add a brief comment on what is going on in this check.
Perhaps even an error print if it's an error path.

> +		rzg2l_adc_writel(adc, RZG2L_ADSTS, reg);
> +		return IRQ_HANDLED;
> +	}
> +
> +	intst = reg & RZG2L_ADINT_INTST_MASK;
> +	if (!intst)

I'm guessing this means spurious interrupt in which case IRQ_NONE
so the core irq code knows about it.

> +		return IRQ_HANDLED;
> +
> +	for (i = 0; i < RZG2L_ADC_MAX_CHANNELS; i++) {
> +		if (intst & BIT(i))

for_each_bit_set() and make intst a long so the types are right for
that macro


> +			adc->last_val[i] = rzg2l_adc_readl(adc, RZG2L_ADCR(i)) & RZG2L_ADCR_AD_MASK;
> +	}
> +
> +	rzg2l_adc_writel(adc, RZG2L_ADSTS, reg);
> +
> +	complete(&adc->completion);
> +
> +	return IRQ_HANDLED;
> +}
> +
> +static int rzg2l_adc_parse_of(struct platform_device *pdev, struct rzg2l_adc *adc)
> +{
> +	struct iio_chan_spec *chan_array;
> +	struct fwnode_handle *fwnode;
> +	struct rzg2l_adc_data *data;
> +	unsigned int channel;
> +	int num_channels;
> +	int ret;
> +	u8 i;
> +
> +	data = devm_kzalloc(&pdev->dev, sizeof(*data), GFP_KERNEL);
> +	if (!data)
> +		return -ENOMEM;
> +
> +	num_channels = device_get_child_node_count(&pdev->dev);
> +	if (!num_channels) {
> +		dev_err(&pdev->dev, "no channel children\n");
> +		return -ENODEV;
> +	}
> +
> +	if (num_channels > RZG2L_ADC_MAX_CHANNELS) {
> +		dev_err(&pdev->dev, "num of channel children out of range\n");
> +		return -EINVAL;
> +	}
> +
> +	chan_array = devm_kcalloc(&pdev->dev, num_channels, sizeof(*chan_array),
> +				  GFP_KERNEL);
> +	if (!chan_array)
> +		return -ENOMEM;
> +
> +	i = 0;
> +	device_for_each_child_node(&pdev->dev, fwnode) {
> +		ret = fwnode_property_read_u32(fwnode, "reg", &channel);
> +		if (ret)
> +			return ret;
> +
> +		if (channel >= RZG2L_ADC_MAX_CHANNELS)
> +			return -EINVAL;
> +
> +		chan_array[i].type = IIO_VOLTAGE;
> +		chan_array[i].indexed = 1;
> +		chan_array[i].channel = channel;
> +		chan_array[i].info_mask_separate = BIT(IIO_CHAN_INFO_RAW);
> +		chan_array[i].datasheet_name = rzg2l_adc_channel_name[channel];
> +		i++;
> +	}
> +
> +	data->num_channels = num_channels;
> +	data->channels = chan_array;
> +	adc->data = data;
> +
> +	return 0;
> +}
> +
> +static int rzg2l_adc_sw_reset(struct rzg2l_adc *adc)
> +{
> +	int timeout = 5;
> +	u32 val;
> +
> +	val = rzg2l_adc_readl(adc, RZG2L_ADM(0));
> +	val |= RZG2L_ADM0_SRESB;
> +	rzg2l_adc_writel(adc, RZG2L_ADM(0), val);
> +
> +	while (!(rzg2l_adc_readl(adc, RZG2L_ADM(0)) & RZG2L_ADM0_SRESB)) {
> +		if (!timeout)
> +			return -EBUSY;
> +		timeout--;
> +		usleep_range(100, 200);
> +	}
> +
> +	return 0;
> +}
> +
> +static int rzg2l_adc_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct iio_dev *indio_dev;
> +	struct rzg2l_adc *adc;
> +	int ret;
> +	int irq;
> +
> +	indio_dev = devm_iio_device_alloc(dev, sizeof(*adc));
> +	if (!indio_dev)
> +		return -ENOMEM;
> +
> +	adc = iio_priv(indio_dev);
> +
> +	ret = rzg2l_adc_parse_of(pdev, adc);

*_parse_properties() or *_parse_firmware()
as it's not of specific now.

> +	if (ret)
> +		return ret;
> +
> +	adc->base = devm_platform_ioremap_resource(pdev, 0);
> +	if (IS_ERR(adc->base))
> +		return PTR_ERR(adc->base);
> +
> +	irq = platform_get_irq(pdev, 0);
> +	if (irq < 0) {
> +		dev_err(dev, "no irq resource\n");
> +		return irq;
> +	}
> +
> +	adc->pclk = devm_clk_get(dev, "pclk");
> +	if (IS_ERR(adc->pclk)) {
> +		dev_err(dev, "Failed to get pclk");
> +		return PTR_ERR(adc->pclk);
> +	}
> +
> +	adc->adclk = devm_clk_get(dev, "adclk");
> +	if (IS_ERR(adc->adclk)) {
> +		dev_err(dev, "Failed to get adclk");
> +		return PTR_ERR(adc->adclk);
> +	}
> +
> +	adc->adrstn = devm_reset_control_get_exclusive(dev, "adrst-n");
> +	if (IS_ERR(adc->adrstn)) {
> +		dev_err(dev, "failed to get adrstn\n");
> +		return PTR_ERR(adc->adrstn);
> +	}
> +
> +	adc->presetn = devm_reset_control_get_exclusive(dev, "presetn");
> +	if (IS_ERR(adc->presetn)) {
> +		dev_err(dev, "failed to get presetn\n");
> +		return PTR_ERR(adc->presetn);
> +	}
> +
> +	ret = reset_control_deassert(adc->adrstn);
> +	if (ret)
> +		return ret;

devm_add_action_or_reset() here to make the reasserting of these part
of the automated cleanup.

> +
> +	ret = reset_control_deassert(adc->presetn);
> +	if (ret)
> +		goto assert_adrstn;
> +
> +	ret = clk_prepare_enable(adc->pclk);
> +	if (ret)
> +		goto assert_presetn;
> +
> +	ret = rzg2l_adc_sw_reset(adc);
> +	if (ret)
> +		goto unprepare_pclk;
> +
> +	clk_disable_unprepare(adc->pclk);

As below, reorder as:

	ret = rzg2l_adc_sw_reset(adc);
	clk_disable_unprepare(adc->pclk);
	if (ret)
		goto assert_preset_n;

> +
> +	init_completion(&adc->completion);
> +
> +	platform_set_drvdata(pdev, indio_dev);
> +
> +	ret = devm_request_irq(dev, irq, rzg2l_adc_isr,
> +			       0, dev_name(dev), adc);

Another aspect of the comment below about mixing and matching between
devm and not.  If the order of remove is not that opposite of probe
then there should be a comment saying why.  There are far too many
race conditions introduced by people doing things in a different order
without carefully considering why.

> +	if (ret < 0)
> +		goto assert_presetn;
> +
> +	indio_dev->name = DRIVER_NAME;
> +	indio_dev->info = &rzg2l_adc_iio_info;
> +	indio_dev->modes = INDIO_DIRECT_MODE;
> +	indio_dev->channels = adc->data->channels;
> +	indio_dev->num_channels = adc->data->num_channels;
> +
> +	pm_runtime_enable(dev);
> +
> +	ret = devm_iio_device_register(dev, indio_dev);
> +	if (ret)
> +		goto err_iio_device_register;
> +
> +	return 0;
> +
> +unprepare_pclk:
> +	clk_disable_unprepare(adc->pclk);

Bad sign that you have this here and not in remove..
When you have cases like this, it's normally better to
either factor out the relevant code into a function
that lets you do this call whether or not there is an
error, or to just move this up to the relevant if statements.


> +err_iio_device_register:
> +	pm_runtime_disable(dev);
> +assert_presetn:
> +	reset_control_assert(adc->presetn);
> +assert_adrstn:
> +	reset_control_assert(adc->adrstn);
> +	return ret;
> +}
> +
> +static int rzg2l_adc_remove(struct platform_device *pdev)
> +{
> +	struct iio_dev *indio_dev = platform_get_drvdata(pdev);
> +	struct rzg2l_adc *adc = iio_priv(indio_dev);
> +
> +	reset_control_assert(adc->presetn);
> +	reset_control_assert(adc->adrstn);
> +	pm_runtime_disable(indio_dev->dev.parent);

You can't mix and match devm managed and unmanaged like this.
The userspace interfce is still exposed after you've put the
device into reset which is unlikely to be a good idea.

One option is to use devm_add_action_or_reset() to make
this stuff happen in the managed flow, which happens after
remove is done.  That would let you drop remove() entirely.


> +
> +	return 0;
> +}
> +
> +static const struct of_device_id rzg2l_adc_match[] = {
> +	{ .compatible = "renesas,rzg2l-adc",},
> +	{ /* sentinel */ }
> +};
> +MODULE_DEVICE_TABLE(of, rzg2l_adc_match);
> +
> +static int __maybe_unused rzg2l_adc_pm_runtime_suspend(struct device *dev)
> +{
> +	struct iio_dev *indio_dev = dev_get_drvdata(dev);
> +	struct rzg2l_adc *adc = iio_priv(indio_dev);
> +
> +	rzg2l_adc_pwr(adc, false);
> +	clk_disable_unprepare(adc->adclk);
> +	clk_disable_unprepare(adc->pclk);
> +
> +	return 0;
> +}
> +
> +static int __maybe_unused rzg2l_adc_pm_runtime_resume(struct device *dev)
> +{
> +	struct iio_dev *indio_dev = dev_get_drvdata(dev);
> +	struct rzg2l_adc *adc = iio_priv(indio_dev);
> +	int ret;
> +
> +	ret = clk_prepare_enable(adc->pclk);
> +	if (ret)
> +		return ret;
> +
> +	ret = clk_prepare_enable(adc->adclk);
> +	if (ret)
> +		return ret;
> +
> +	rzg2l_adc_pwr(adc, true);
> +
> +	return 0;
> +}
> +
> +static const struct dev_pm_ops rzg2l_adc_pm_ops = {
> +	SET_RUNTIME_PM_OPS(rzg2l_adc_pm_runtime_suspend,
> +			   rzg2l_adc_pm_runtime_resume,
> +			   NULL)
> +};
> +
> +static struct platform_driver rzg2l_adc_driver = {
> +	.probe		= rzg2l_adc_probe,
> +	.remove		= rzg2l_adc_remove,
> +	.driver		= {
> +		.name		= DRIVER_NAME,
> +		.of_match_table = rzg2l_adc_match,
> +		.pm		= &rzg2l_adc_pm_ops,
> +	},
> +};
> +
> +module_platform_driver(rzg2l_adc_driver);
> +
> +MODULE_AUTHOR("Lad Prabhakar <prabhakar.mahadev-lad.rj@...renesas.com>");
> +MODULE_DESCRIPTION("Renesas RZ/G2L ADC driver");
> +MODULE_LICENSE("GPL v2");

Powered by blists - more mailing lists