All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jonathan Cameron <jic23@kernel.org>
To: Alexandru Ardelean <alexandru.ardelean@analog.com>
Cc: <linux-iio@vger.kernel.org>
Subject: Re: [PATCH] iio: dac: ad7303: replace mlock with own lock
Date: Sat, 5 Oct 2019 15:54:07 +0100	[thread overview]
Message-ID: <20191005155252.1bfa52da@archlinux> (raw)
In-Reply-To: <20190919131328.13633-1-alexandru.ardelean@analog.com>

On Thu, 19 Sep 2019 16:13:28 +0300
Alexandru Ardelean <alexandru.ardelean@analog.com> wrote:

> This change replaces indio_dev's mlock with the driver's own lock. The lock
> is mostly needed to protect state when changing the `config` & `dac_cache`
> fields. As such, the lock has been extended to covert the read of these
> fields.
This last bit would be needed if we had anywhere that needed to see a consistent
state across those two fields. I can't immediately see where that happens.
> 
> Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com>
See below.

> ---
>  drivers/iio/dac/ad7303.c | 21 +++++++++++++++------
>  1 file changed, 15 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/iio/dac/ad7303.c b/drivers/iio/dac/ad7303.c
> index 8de9f40226e6..f45352e80db7 100644
> --- a/drivers/iio/dac/ad7303.c
> +++ b/drivers/iio/dac/ad7303.c
> @@ -41,6 +41,7 @@ struct ad7303_state {
>  	struct regulator *vdd_reg;
>  	struct regulator *vref_reg;
>  
> +	struct mutex lock;
>  	/*
>  	 * DMA (thus cache coherency maintenance) requires the
>  	 * transfer buffers to live in their own cache lines.
> @@ -62,9 +63,13 @@ static ssize_t ad7303_read_dac_powerdown(struct iio_dev *indio_dev,
>  	uintptr_t private, const struct iio_chan_spec *chan, char *buf)
>  {
>  	struct ad7303_state *st = iio_priv(indio_dev);
> +	int val;
>  
> -	return sprintf(buf, "%d\n", (bool)(st->config &
> -		AD7303_CFG_POWER_DOWN(chan->channel)));
> +	mutex_lock(&st->lock);
> +	val = (bool)(st->config & AD7303_CFG_POWER_DOWN(chan->channel));
> +	mutex_unlock(&st->lock);

I think the read of st->config should be atomic anyway so you'll either
get the value before or after some concurrent write. That's true
even with the locks, so I don't htink they are needed here.

> +
> +	return sprintf(buf, "%d\n", val);
>  }
>  
>  static ssize_t ad7303_write_dac_powerdown(struct iio_dev *indio_dev,
> @@ -79,7 +84,7 @@ static ssize_t ad7303_write_dac_powerdown(struct iio_dev *indio_dev,
>  	if (ret)
>  		return ret;
>  
> -	mutex_lock(&indio_dev->mlock);
> +	mutex_lock(&st->lock);
>  
>  	if (pwr_down)
>  		st->config |= AD7303_CFG_POWER_DOWN(chan->channel);
> @@ -90,7 +95,7 @@ static ssize_t ad7303_write_dac_powerdown(struct iio_dev *indio_dev,
>  	 * mode, so just write one of the DAC channels again */
>  	ad7303_write(st, chan->channel, st->dac_cache[chan->channel]);
>  
> -	mutex_unlock(&indio_dev->mlock);
> +	mutex_unlock(&st->lock);
>  	return len;
>  }
>  
> @@ -116,7 +121,9 @@ static int ad7303_read_raw(struct iio_dev *indio_dev,
>  
>  	switch (info) {
>  	case IIO_CHAN_INFO_RAW:
> +		mutex_lock(&st->lock);
>  		*val = st->dac_cache[chan->channel];
> +		mutex_unlock(&st->lock);
>  		return IIO_VAL_INT;
>  	case IIO_CHAN_INFO_SCALE:
>  		vref_uv = ad7303_get_vref(st, chan);
> @@ -144,11 +151,11 @@ static int ad7303_write_raw(struct iio_dev *indio_dev,
>  		if (val >= (1 << chan->scan_type.realbits) || val < 0)
>  			return -EINVAL;
>  
> -		mutex_lock(&indio_dev->mlock);
> +		mutex_lock(&st->lock);
>  		ret = ad7303_write(st, chan->address, val);
>  		if (ret == 0)
>  			st->dac_cache[chan->channel] = val;
> -		mutex_unlock(&indio_dev->mlock);
> +		mutex_unlock(&st->lock);
>  		break;
>  	default:
>  		ret = -EINVAL;
> @@ -211,6 +218,8 @@ static int ad7303_probe(struct spi_device *spi)
>  
>  	st->spi = spi;
>  
> +	mutex_init(&st->lock);
> +
>  	st->vdd_reg = devm_regulator_get(&spi->dev, "Vdd");
>  	if (IS_ERR(st->vdd_reg))
>  		return PTR_ERR(st->vdd_reg);


  reply	other threads:[~2019-10-05 14:54 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-09-19 13:13 [PATCH] iio: dac: ad7303: replace mlock with own lock Alexandru Ardelean
2019-10-05 14:54 ` Jonathan Cameron [this message]
2019-10-07  8:26 ` [PATCH v2] " Alexandru Ardelean
2019-10-12 13:45   ` 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=20191005155252.1bfa52da@archlinux \
    --to=jic23@kernel.org \
    --cc=alexandru.ardelean@analog.com \
    --cc=linux-iio@vger.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.