From: Bryan Freed <bfreed@chromium.org>
To: linux-kernel@vger.kernel.org
Cc: jic23@cam.ac.uk, gregkh@suse.de, rklein@nvidia.com,
linux-iio@vger.kernel.org, Bryan Freed <bfreed@chromium.org>
Subject: [PATCH v2 3/4] staging:iio:light:isl29018: Convert some of the isl29018 driver to the new iio_chan_spec framework.
Date: Thu, 7 Jul 2011 12:01:56 -0700 [thread overview]
Message-ID: <1310065317-3872-4-git-send-email-bfreed@chromium.org> (raw)
In-Reply-To: <1310065317-3872-1-git-send-email-bfreed@chromium.org>
Remove the driver's get_sensor_data() interfaces and replace them with
iio_chan_spec channels. This converts 4 files to the new framework.
Driver ABI change: The intensity_infrared_raw file is now intensity_ir_raw.
Signed-off-by: Bryan Freed <bfreed@chromium.org>
---
drivers/staging/iio/light/isl29018.c | 168 ++++++++++++++++------------------
1 files changed, 78 insertions(+), 90 deletions(-)
diff --git a/drivers/staging/iio/light/isl29018.c b/drivers/staging/iio/light/isl29018.c
index cc6d718..070e94d 100644
--- a/drivers/staging/iio/light/isl29018.c
+++ b/drivers/staging/iio/light/isl29018.c
@@ -224,74 +224,7 @@ static int isl29018_read_proximity_ir(struct i2c_client *client, int scheme,
return 0;
}
-static ssize_t get_sensor_data(struct device *dev, char *buf, int mode)
-{
- struct iio_dev *indio_dev = dev_get_drvdata(dev);
- struct isl29018_chip *chip = iio_priv(indio_dev);
- struct i2c_client *client = chip->client;
- int value = 0;
- int status;
-
- mutex_lock(&chip->lock);
- switch (mode) {
- case COMMMAND1_OPMODE_PROX_ONCE:
- status = isl29018_read_proximity_ir(client,
- chip->prox_scheme, &value);
- break;
-
- case COMMMAND1_OPMODE_ALS_ONCE:
- status = isl29018_read_lux(client, &value);
- break;
-
- case COMMMAND1_OPMODE_IR_ONCE:
- status = isl29018_read_ir(client, &value);
- break;
-
- default:
- dev_err(&client->dev, "Mode %d is not supported\n", mode);
- mutex_unlock(&chip->lock);
- return -EBUSY;
- }
- if (status < 0) {
- dev_err(&client->dev, "Error in Reading data");
- mutex_unlock(&chip->lock);
- return status;
- }
-
- mutex_unlock(&chip->lock);
-
- return sprintf(buf, "%d\n", value);
-}
-
/* Sysfs interface */
-/* lux_scale */
-static ssize_t show_lux_scale(struct device *dev,
- struct device_attribute *attr, char *buf)
-{
- struct iio_dev *indio_dev = dev_get_drvdata(dev);
- struct isl29018_chip *chip = indio_dev->dev_data;
-
- return sprintf(buf, "%d\n", chip->lux_scale);
-}
-
-static ssize_t store_lux_scale(struct device *dev,
- struct device_attribute *attr, const char *buf, size_t count)
-{
- struct iio_dev *indio_dev = dev_get_drvdata(dev);
- struct isl29018_chip *chip = indio_dev->dev_data;
- unsigned long lval;
-
- lval = simple_strtoul(buf, NULL, 10);
- if (lval == 0)
- return -EINVAL;
-
- mutex_lock(&chip->lock);
- chip->lux_scale = lval;
- mutex_unlock(&chip->lock);
-
- return count;
-}
-
/* range */
static ssize_t show_range(struct device *dev,
struct device_attribute *attr, char *buf)
@@ -409,27 +342,87 @@ static ssize_t store_prox_infrared_supression(struct device *dev,
return count;
}
-/* Read lux */
-static ssize_t show_lux(struct device *dev,
- struct device_attribute *devattr, char *buf)
+/* Channel IO */
+static int isl29018_write_raw(struct iio_dev *indio_dev,
+ struct iio_chan_spec const *chan,
+ int val,
+ int val2,
+ long mask)
{
- return get_sensor_data(dev, buf, COMMMAND1_OPMODE_ALS_ONCE);
-}
+ struct isl29018_chip *chip = iio_priv(indio_dev);
+ int ret = -EINVAL;
-/* Read ir */
-static ssize_t show_ir(struct device *dev,
- struct device_attribute *devattr, char *buf)
-{
- return get_sensor_data(dev, buf, COMMMAND1_OPMODE_IR_ONCE);
+ mutex_lock(&chip->lock);
+ if (mask == (1 << IIO_CHAN_INFO_CALIBSCALE_SEPARATE) &&
+ chan->type == IIO_LIGHT) {
+ chip->lux_scale = val;
+ ret = 0;
+ }
+ mutex_unlock(&chip->lock);
+
+ return 0;
}
-/* Read nearest ir */
-static ssize_t show_proxim_ir(struct device *dev,
- struct device_attribute *devattr, char *buf)
+static int isl29018_read_raw(struct iio_dev *indio_dev,
+ struct iio_chan_spec const *chan,
+ int *val,
+ int *val2,
+ long mask)
{
- return get_sensor_data(dev, buf, COMMMAND1_OPMODE_PROX_ONCE);
+ int ret = -EINVAL;
+ struct isl29018_chip *chip = iio_priv(indio_dev);
+ struct i2c_client *client = chip->client;
+
+ mutex_lock(&chip->lock);
+ switch (mask) {
+ case 0:
+ switch (chan->type) {
+ case IIO_LIGHT:
+ ret = isl29018_read_lux(client, val);
+ break;
+ case IIO_INTENSITY:
+ ret = isl29018_read_ir(client, val);
+ break;
+ case IIO_PROXIMITY:
+ ret = isl29018_read_proximity_ir(client,
+ chip->prox_scheme, val);
+ break;
+ default:
+ break;
+ }
+ if (!ret)
+ ret = IIO_VAL_INT;
+ break;
+ case (1 << IIO_CHAN_INFO_CALIBSCALE_SEPARATE):
+ if (chan->type == IIO_LIGHT) {
+ *val = chip->lux_scale;
+ ret = IIO_VAL_INT;
+ }
+ break;
+ default:
+ break;
+ }
+ mutex_unlock(&chip->lock);
+ return ret;
}
+static const struct iio_chan_spec isl29018_channels[] = {
+ {
+ .type = IIO_LIGHT,
+ .indexed = 1,
+ .channel = 0,
+ .processed_val = 1,
+ .info_mask = (1 << IIO_CHAN_INFO_CALIBSCALE_SEPARATE),
+ }, {
+ .type = IIO_INTENSITY,
+ .modified = 1,
+ .channel2 = IIO_MOD_LIGHT_IR,
+ }, {
+ /* Unindexed in current ABI. But perhaps it should be. */
+ .type = IIO_PROXIMITY,
+ }
+};
+
static IIO_DEVICE_ATTR(range, S_IRUGO | S_IWUSR, show_range, store_range, 0);
static IIO_CONST_ATTR(range_available, "1000 4000 16000 64000");
static IIO_CONST_ATTR(adc_resolution_available, "4 8 12 16");
@@ -439,11 +432,6 @@ static IIO_DEVICE_ATTR(proximity_on_chip_ambient_infrared_supression,
S_IRUGO | S_IWUSR,
show_prox_infrared_supression,
store_prox_infrared_supression, 0);
-static IIO_DEVICE_ATTR(illuminance0_input, S_IRUGO, show_lux, NULL, 0);
-static IIO_DEVICE_ATTR(illuminance0_calibscale, S_IRUGO | S_IWUSR,
- show_lux_scale, store_lux_scale, 0);
-static IIO_DEVICE_ATTR(intensity_infrared_raw, S_IRUGO, show_ir, NULL, 0);
-static IIO_DEVICE_ATTR(proximity_raw, S_IRUGO, show_proxim_ir, NULL, 0);
#define ISL29018_DEV_ATTR(name) (&iio_dev_attr_##name.dev_attr.attr)
#define ISL29018_CONST_ATTR(name) (&iio_const_attr_##name.dev_attr.attr)
@@ -453,10 +441,6 @@ static struct attribute *isl29018_attributes[] = {
ISL29018_DEV_ATTR(adc_resolution),
ISL29018_CONST_ATTR(adc_resolution_available),
ISL29018_DEV_ATTR(proximity_on_chip_ambient_infrared_supression),
- ISL29018_DEV_ATTR(illuminance0_input),
- ISL29018_DEV_ATTR(illuminance0_calibscale),
- ISL29018_DEV_ATTR(intensity_infrared_raw),
- ISL29018_DEV_ATTR(proximity_raw),
NULL
};
@@ -489,6 +473,8 @@ static int isl29018_chip_init(struct i2c_client *client)
static const struct iio_info isl29108_info = {
.attrs = &isl29108_group,
.driver_module = THIS_MODULE,
+ .read_raw = &isl29018_read_raw,
+ .write_raw = &isl29018_write_raw,
};
static int __devinit isl29018_probe(struct i2c_client *client,
@@ -520,6 +506,8 @@ static int __devinit isl29018_probe(struct i2c_client *client,
goto exit_iio_free;
indio_dev->info = &isl29108_info;
+ indio_dev->channels = isl29018_channels;
+ indio_dev->num_channels = ARRAY_SIZE(isl29018_channels);
indio_dev->name = id->name;
indio_dev->dev.parent = &client->dev;
indio_dev->modes = INDIO_DIRECT_MODE;
--
1.7.3.1
next prev parent reply other threads:[~2011-07-07 19:02 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-07-07 19:01 [PATCH 0/4] Apply these staging:iio patches to staging-next Bryan Freed
2011-07-07 19:01 ` [PATCH v2 1/4] staging:iio: Reorder channel type strings to match the iio.h enums Bryan Freed
2011-07-07 19:01 ` [PATCH v2 2/4] staging:iio: Add an iio channel type string to support proximity sensors Bryan Freed
2011-07-07 19:01 ` Bryan Freed [this message]
2011-07-07 19:01 ` [PATCH v1 4/4] staging:iio:light:isl29018: Fix the "Init of isl29018 fails" failure Bryan Freed
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=1310065317-3872-4-git-send-email-bfreed@chromium.org \
--to=bfreed@chromium.org \
--cc=gregkh@suse.de \
--cc=jic23@cam.ac.uk \
--cc=linux-iio@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=rklein@nvidia.com \
/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
all inboxes | Powered by JetHome®