From: Jonathan Cameron <jic23@kernel.org>
To: Antoniu Miclaus <antoniu.miclaus@analog.com>
Cc: <robh@kernel.org>, <conor+dt@kernel.org>,
<linux-iio@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
<devicetree@vger.kernel.org>
Subject: Re: [PATCH v7 4/6] iio: adc: add ade9000 support
Date: Sat, 13 Sep 2025 13:46:21 +0100 [thread overview]
Message-ID: <20250913134621.226ed6df@jic23-huawei> (raw)
In-Reply-To: <20250908073531.3639-5-antoniu.miclaus@analog.com>
On Mon, 8 Sep 2025 07:35:24 +0000
Antoniu Miclaus <antoniu.miclaus@analog.com> wrote:
> Add driver support for the ade9000. highly accurate,
> fully integrated, multiphase energy and power quality
> monitoring device.
>
> Signed-off-by: Antoniu Miclaus <antoniu.miclaus@analog.com>
So, I made a few comments this time that I should probably have
noticed in earlier versions. I'm not particularly keen to delay
an initial merge of the driver for them but I would like you consider
the suggestion of a perphase structure look up table to
reduce duplication as a potential follow up patch.
Jonathan
> diff --git a/drivers/iio/adc/ade9000.c b/drivers/iio/adc/ade9000.c
> new file mode 100644
> index 000000000000..a3be4bd24686
> --- /dev/null
> +++ b/drivers/iio/adc/ade9000.c
> +/* Voltage events (zero crossing on instantaneous voltage) */
> +static const struct iio_event_spec ade9000_voltage_events[] = {
> + {
> + /* Zero crossing detection - datasheet: ZXV interrupts */
> + .type = IIO_EV_TYPE_THRESH,
> + .dir = IIO_EV_DIR_EITHER,
> + .mask_separate = BIT(IIO_EV_INFO_ENABLE),
Perhaps this event might be easier to understand if we provide a value that
is hard coded to 0?
> + },
> +};
> +
> +/* Current events (zero crossing on instantaneous current) */
> +static const struct iio_event_spec ade9000_current_events[] = {
> + {
> + /* Zero crossing detection - datasheet: ZXI interrupts */
> + .type = IIO_EV_TYPE_THRESH,
> + .dir = IIO_EV_DIR_EITHER,
> + .mask_separate = BIT(IIO_EV_INFO_ENABLE),
As above.
> + },
> +};
> +static int ade9000_iio_push_streaming(struct iio_dev *indio_dev)
> +{
> + struct ade9000_state *st = iio_priv(indio_dev);
> + struct device *dev = &st->spi->dev;
> + u32 current_page, i;
> + int ret;
> +
> + guard(mutex)(&st->lock);
> +
> + ret = spi_sync(st->spi, &st->spi_msg);
> + if (ret) {
> + dev_err_ratelimited(dev, "SPI fail in trigger handler\n");
> + return ret;
> + }
> +
> + /* In streaming mode, only half the buffer is filled per interrupt */
> + for (i = 0; i < st->wfb_nr_samples / 2; i += st->wfb_nr_activ_chan)
> + iio_push_to_buffers(indio_dev, &st->rx_buff.word[i]);
> +
> + ret = regmap_read(st->regmap, ADE9000_REG_WFB_PG_IRQEN, ¤t_page);
> + if (ret) {
> + dev_err_ratelimited(dev, "IRQ0 WFB read fail\n");
> + return ret;
> + }
> +
> + if (current_page & ADE9000_MIDDLE_PAGE_BIT) {
> + ret = regmap_write(st->regmap, ADE9000_REG_WFB_PG_IRQEN,
> + ADE9000_LAST_PAGE_BIT);
> + if (ret) {
> + dev_err_ratelimited(dev, "IRQ0 WFB write fail\n");
> + return ret;
> + }
> +
> + ade9000_configure_scan(indio_dev,
> + ADE9000_REG_WF_HALF_BUFF);
> + } else {
> + ret = regmap_write(st->regmap, ADE9000_REG_WFB_PG_IRQEN,
> + ADE9000_MIDDLE_PAGE_BIT);
> + if (ret) {
> + dev_err_ratelimited(dev, "IRQ0 WFB write fail");
> + return IRQ_HANDLED;
> + }
> +
> + ade9000_configure_scan(indio_dev,
> + ADE9000_REG_WF_BUFF);
Excessive wrap. Fits on one line. I'll tidy this up if I pick up the patch.
> + }
> +
> + return 0;
> +}
> +
> +static int ade9000_read_event_config(struct iio_dev *indio_dev,
> + const struct iio_chan_spec *chan,
> + enum iio_event_type type,
> + enum iio_event_direction dir)
> +{
> + struct ade9000_state *st = iio_priv(indio_dev);
> + u32 interrupts1;
> + int ret;
> +
> + /* All events use MASK1 register */
> + ret = regmap_read(st->regmap, ADE9000_REG_MASK1, &interrupts1);
> + if (ret)
> + return ret;
> +
> + switch (chan->channel) {
> + case ADE9000_PHASE_A_NR:
> + if (chan->type == IIO_VOLTAGE && dir == IIO_EV_DIR_EITHER)
> + return !!(interrupts1 & ADE9000_ST1_ZXVA_BIT);
As below. If you had a look up table with appropriate per phase structures
the 3 repeats here could basically be change from code to data which is
generally a good idea and is easier to read.
> + else if (chan->type == IIO_CURRENT && dir == IIO_EV_DIR_EITHER)
> + return !!(interrupts1 & ADE9000_ST1_ZXIA_BIT);
> + else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_RISING)
> + return !!(interrupts1 & ADE9000_ST1_SWELLA_BIT);
> + else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_FALLING)
> + return !!(interrupts1 & ADE9000_ST1_DIPA_BIT);
> + dev_err_ratelimited(&indio_dev->dev, "Invalid channel type %d or direction %d for phase A\n", chan->type, dir);
> + return -EINVAL;
> + case ADE9000_PHASE_B_NR:
> + if (chan->type == IIO_VOLTAGE && dir == IIO_EV_DIR_EITHER)
> + return !!(interrupts1 & ADE9000_ST1_ZXVB_BIT);
> + else if (chan->type == IIO_CURRENT && dir == IIO_EV_DIR_EITHER)
> + return !!(interrupts1 & ADE9000_ST1_ZXIB_BIT);
> + else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_RISING)
> + return !!(interrupts1 & ADE9000_ST1_SWELLB_BIT);
> + else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_FALLING)
> + return !!(interrupts1 & ADE9000_ST1_DIPB_BIT);
> + dev_err_ratelimited(&indio_dev->dev, "Invalid channel type %d or direction %d for phase B\n", chan->type, dir);
> + return -EINVAL;
> + case ADE9000_PHASE_C_NR:
> + if (chan->type == IIO_VOLTAGE && dir == IIO_EV_DIR_EITHER)
> + return !!(interrupts1 & ADE9000_ST1_ZXVC_BIT);
> + else if (chan->type == IIO_CURRENT && dir == IIO_EV_DIR_EITHER)
> + return !!(interrupts1 & ADE9000_ST1_ZXIC_BIT);
> + else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_RISING)
> + return !!(interrupts1 & ADE9000_ST1_SWELLC_BIT);
> + else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_FALLING)
> + return !!(interrupts1 & ADE9000_ST1_DIPC_BIT);
> + dev_err_ratelimited(&indio_dev->dev, "Invalid channel type %d or direction %d for phase C\n", chan->type, dir);
> + return -EINVAL;
> + default:
> + return -EINVAL;
> + }
> +}
> +
> +static int ade9000_write_event_config(struct iio_dev *indio_dev,
> + const struct iio_chan_spec *chan,
> + enum iio_event_type type,
> + enum iio_event_direction dir,
> + bool state)
> +{
> + struct ade9000_state *st = iio_priv(indio_dev);
> + u32 bit_mask;
> + int ret;
> +
> + /* Clear all pending events in STATUS1 register (write 1 to clear) */
> + ret = regmap_write(st->regmap, ADE9000_REG_STATUS1, GENMASK(31, 0));
> + if (ret)
> + return ret;
> +
> + /* Determine which interrupt bit to enable/disable */
> + switch (chan->channel) {
> + case ADE9000_PHASE_A_NR:
> + if (chan->type == IIO_VOLTAGE && dir == IIO_EV_DIR_EITHER) {
Perhaps consider per phase look up tables?
> + bit_mask = ADE9000_ST1_ZXVA_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_ZXVA_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_ZXVA_BIT;
> + } else if (chan->type == IIO_CURRENT && dir == IIO_EV_DIR_EITHER) {
> + bit_mask = ADE9000_ST1_ZXIA_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_ZXIA_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_ZXIA_BIT;
> + } else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_RISING) {
> + bit_mask = ADE9000_ST1_SWELLA_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_SWELL_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_SWELL_BIT;
> + } else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_FALLING) {
> + bit_mask = ADE9000_ST1_DIPA_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_DIP_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_DIP_BIT;
> + } else {
> + dev_err_ratelimited(&indio_dev->dev, "Invalid channel type %d or direction %d for phase A\n",
> + chan->type, dir);
> + return -EINVAL;
> + }
> + break;
> + case ADE9000_PHASE_B_NR:
> + if (chan->type == IIO_VOLTAGE && dir == IIO_EV_DIR_EITHER) {
> + bit_mask = ADE9000_ST1_ZXVB_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_ZXVB_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_ZXVB_BIT;
> + } else if (chan->type == IIO_CURRENT && dir == IIO_EV_DIR_EITHER) {
> + bit_mask = ADE9000_ST1_ZXIB_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_ZXIB_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_ZXIB_BIT;
> + } else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_RISING) {
> + bit_mask = ADE9000_ST1_SWELLB_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_SWELL_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_SWELL_BIT;
> + } else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_FALLING) {
> + bit_mask = ADE9000_ST1_DIPB_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_DIP_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_DIP_BIT;
> + } else {
> + dev_err_ratelimited(&indio_dev->dev, "Invalid channel type %d or direction %d for phase B\n",
> + chan->type, dir);
> + return -EINVAL;
> + }
> + break;
> + case ADE9000_PHASE_C_NR:
> + if (chan->type == IIO_VOLTAGE && dir == IIO_EV_DIR_EITHER) {
> + bit_mask = ADE9000_ST1_ZXVC_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_ZXVC_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_ZXVC_BIT;
> + } else if (chan->type == IIO_CURRENT && dir == IIO_EV_DIR_EITHER) {
> + bit_mask = ADE9000_ST1_ZXIC_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_ZXIC_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_ZXIC_BIT;
> + } else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_RISING) {
> + bit_mask = ADE9000_ST1_SWELLC_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_SWELL_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_SWELL_BIT;
> + } else if (chan->type == IIO_ALTVOLTAGE && dir == IIO_EV_DIR_FALLING) {
> + bit_mask = ADE9000_ST1_DIPC_BIT;
> + if (state)
> + st->wfb_trg |= ADE9000_WFB_TRG_DIP_BIT;
> + else
> + st->wfb_trg &= ~ADE9000_WFB_TRG_DIP_BIT;
> + } else {
> + dev_err_ratelimited(&indio_dev->dev, "Invalid channel type %d or direction %d for phase C\n",
> + chan->type, dir);
> + return -EINVAL;
> + }
> + break;
> + default:
> + return -EINVAL;
> + }
> +
> + /* Set bits if enabling event, clear bits if disabling */
> + return regmap_assign_bits(st->regmap, ADE9000_REG_MASK1, bit_mask, state ? bit_mask : 0);
> +}
> +static int ade9000_setup(struct ade9000_state *st)
> +{
> + struct device *dev = &st->spi->dev;
> + int ret;
> +
> + ret = regmap_multi_reg_write(st->regmap, ade9000_initialization_sequence,
> + ARRAY_SIZE(ade9000_initialization_sequence));
Really trivial and I definitely not worth a respin but does feel like some
abbreviation would have worked here! _init_seq maybe?
next prev parent reply other threads:[~2025-09-13 12:46 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-09-08 7:35 [PATCH v7 0/6] iio: adc: add support for ADE9000 Energy Monitoring IC Antoniu Miclaus
2025-09-08 7:35 ` [PATCH v7 1/6] iio: add IIO_ALTCURRENT channel type Antoniu Miclaus
2025-09-08 7:35 ` [PATCH v7 2/6] iio: add power and energy measurement modifiers Antoniu Miclaus
2025-09-08 7:35 ` [PATCH v7 3/6] dt-bindings: iio: adc: add ade9000 Antoniu Miclaus
2025-09-08 18:27 ` Conor Dooley
2025-09-08 7:35 ` [PATCH v7 4/6] iio: adc: add ade9000 support Antoniu Miclaus
2025-09-13 12:46 ` Jonathan Cameron [this message]
2025-09-13 13:00 ` Jonathan Cameron
2025-09-08 7:35 ` [PATCH v7 5/6] docs: iio: add documentation for ade9000 driver Antoniu Miclaus
2025-09-13 12:46 ` Jonathan Cameron
2025-09-13 13:20 ` Jonathan Cameron
2025-09-08 7:35 ` [PATCH v7 6/6] Documentation: ABI: iio: add sinc4+lp Antoniu Miclaus
2025-09-13 13:22 ` [PATCH v7 0/6] iio: adc: add support for ADE9000 Energy Monitoring IC Jonathan Cameron
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20250913134621.226ed6df@jic23-huawei \
--to=jic23@kernel.org \
--cc=antoniu.miclaus@analog.com \
--cc=conor+dt@kernel.org \
--cc=devicetree@vger.kernel.org \
--cc=linux-iio@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=robh@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox