by Jonathan Cameron

[permalink] [raw]

Subject: Re: [PATCH v3 07/13] iio: adc: ad7091r: Set device mode through chip_info callback

On Thu, 7 Dec 2023 15:41:03 -0300
Marcelo Schmitt <[email protected]> wrote:

> AD7091R-5 devices have a few modes of operation (sample, command,
> autocycle) which are set by writing to configuration register fields.
> Follow up patches will add support for AD7091R-2/-4/-8 which don't have
> those operation modes nor the register fields for setting them.
> Make ad7091r_set_mode() a callback function of AD7091R chip_info struct
> so the base driver can appropriately handle each design without having
> to check which actual chip is connected.
>
> Signed-off-by: Marcelo Schmitt <[email protected]>

Hi Marcelo,
> +
> #define AD7091R_CHANNEL(idx, bits, ev, num_ev) { \
> .type = IIO_VOLTAGE, \
> .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \
> @@ -52,6 +60,7 @@ struct ad7091r_chip_info {
> unsigned int num_channels;
> const struct iio_chan_spec *channels;
> unsigned int vref_mV;
> + int (*ad7091r_set_mode)(struct ad7091r_state *st, enum ad7091r_mode mode);
Given it's embedded in a driver specific structure, I'm not seeing a clear
reason to prefix the callback with ad7091r. set_mode is probably enough.
Same for the ones introduced in later patches.

2023-12-10 19:54:46

by Marcelo Schmitt

[permalink] [raw]

Subject: Re: [PATCH v3 12/13] iio: adc: Add support for AD7091R-8

Hi Jonathan,

Thank you for all comments to this set.
Will do the changes and send v4.

Thanks,
Marcelo

On 12/10, Jonathan Cameron wrote:
> On Thu, 7 Dec 2023 15:42:56 -0300
> Marcelo Schmitt <[email protected]> wrote:
>
> > Add support for Analog Devices AD7091R-2, AD7091R-4, and AD7091R-8
> > low power 12-Bit SAR ADCs with SPI interface.
> > Extend ad7091r-base driver so it can be used by the AD7091R-8 driver.
> >
> > Signed-off-by: Marcelo Schmitt <[email protected]>
>
> A few trivial things inline.
> Otherwise looks pretty good to me.
>
> Jonathan
> > diff --git a/drivers/iio/adc/ad7091r8.c b/drivers/iio/adc/ad7091r8.c
> > new file mode 100644
> > index 000000000000..8dc0f784913b
> > --- /dev/null
> > +++ b/drivers/iio/adc/ad7091r8.c
> > @@ -0,0 +1,261 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * Analog Devices AD7091R8 12-bit SAR ADC driver
> > + *
> > + * Copyright 2023 Analog Devices Inc.
> > + */
> > +
> > +#include <linux/bitfield.h>
> > +#include <linux/iio/iio.h>
> > +#include <linux/module.h>
> > +#include <linux/regmap.h>
> > +#include <linux/gpio/consumer.h>
> > +#include <linux/spi/spi.h>
> > +
> > +#include "ad7091r-base.h"
> > +
> > +#define AD7091R8_REG_ADDR_MSK GENMASK(15, 11)
> > +#define AD7091R8_RD_WR_FLAG_MSK BIT(10)
> > +#define AD7091R8_REG_DATA_MSK GENMASK(9, 0)
> > +
> > +#define AD7091R2_DEV_NAME "ad7091r-2"
> > +#define AD7091R4_DEV_NAME "ad7091r-4"
> > +#define AD7091R8_DEV_NAME "ad7091r-8"
> Not seeing any advantage in these macros. It will be more readable to just
> have the strings inline where the macros are currently used.
>
> > +static int ad7091r8_gpio_setup(struct ad7091r_state *st)
> > +{
> > + st->convst_gpio = devm_gpiod_get(st->dev, "adi,conversion-start",
> > + GPIOD_OUT_LOW);
> > + if (IS_ERR(st->convst_gpio))
> > + return dev_err_probe(st->dev, PTR_ERR(st->convst_gpio),
> > + "Error getting convst GPIO\n");
> > +
> > + st->reset_gpio = devm_gpiod_get_optional(st->dev, "reset",
> > + GPIOD_OUT_HIGH);
> > + if (IS_ERR(st->reset_gpio))
> > + return PTR_ERR(st->reset_gpio);
> Maybe a dev_err_probe() here as well both for consistency and for the
> debug info that gets stashed if it's EPROBE_DEFER
> > +
> > + if (st->reset_gpio) {
> > + fsleep(20);
> > + gpiod_set_value_cansleep(st->reset_gpio, 0);
> > + }
> > +
> > + return 0;
> > +}
>
> > +
> > +static struct ad7091r_init_info ad7091r2_init_info = {
> > + .info_no_irq = AD7091R_SPI_CHIP_INFO(2),
> > + .regmap_config = &ad7091r2_reg_conf,
> > + .ad7091r_regmap_init = &ad7091r8_regmap_init,
> > + .ad7091r_setup = &ad7091r8_gpio_setup
> > +};
> > +
> > +static struct ad7091r_init_info ad7091r4_init_info = {
> > + .irq_info = AD7091R_SPI_CHIP_INFO_IRQ(4),
> > + .info_no_irq = AD7091R_SPI_CHIP_INFO(4),
> > + .regmap_config = &ad7091r4_reg_conf,
> > + .ad7091r_regmap_init = &ad7091r8_regmap_init,
> > + .ad7091r_setup = &ad7091r8_gpio_setup
> > +};
> > +
> > +static struct ad7091r_init_info ad7091r8_init_info = {
> > + .irq_info = AD7091R_SPI_CHIP_INFO_IRQ(8),
> > + .info_no_irq = AD7091R_SPI_CHIP_INFO(8),
> > + .regmap_config = &ad7091r8_reg_conf,
> > + .ad7091r_regmap_init = &ad7091r8_regmap_init,
> > + .ad7091r_setup = &ad7091r8_gpio_setup
> > +};
> > +
> > +static int ad7091r8_spi_probe(struct spi_device *spi)
> > +{
> > + const struct spi_device_id *id = spi_get_device_id(spi);
> > + const struct ad7091r_init_info *init_info;
> > +
> > + init_info = spi_get_device_match_data(spi);
> > + if (!init_info)
> > + return -EINVAL;
> > +
> > + return ad7091r_probe(&spi->dev, id->name, init_info, NULL, spi->irq);
> id->name isn't generally a good idea because we end up with lots of odd corner
> cases if the of_device_id and spi_device_id tables get out of sync - which
> can happen if fallback compatibles get used.
>
> Normal way round this is just put the naming of the device in the
> info structure. Costs a little storage, but makes the code simpler
> and less probe to odd corner cases. Also, I think you already have it
> in there!
>
> > +}
> > +
> > +static const struct of_device_id ad7091r8_of_match[] = {
> > + { .compatible = "adi,ad7091r2", .data = &ad7091r2_init_info },
> > + { .compatible = "adi,ad7091r4", .data = &ad7091r4_init_info },
> > + { .compatible = "adi,ad7091r8", .data = &ad7091r8_init_info },
> > + { }
> > +};
> > +MODULE_DEVICE_TABLE(of, ad7091r8_of_match);
> > +
> > +static const struct spi_device_id ad7091r8_spi_id[] = {
> > + { "ad7091r2", (kernel_ulong_t)&ad7091r2_init_info },
> > + { "ad7091r4", (kernel_ulong_t)&ad7091r4_init_info },
> > + { "ad7091r8", (kernel_ulong_t)&ad7091r8_init_info },
> > + {}
> Trivial but be consistent on spacing for these terminators. I like a space, so
> { } but I don't mind if an author prefers {} as long as they are consistent!
>
> > +};
> > +MODULE_DEVICE_TABLE(spi, ad7091r8_spi_id);
>