linux/drivers/iio/industrialio-core.c

2033 lines
52 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-only
/* The industrial I/O core
*
* Copyright (c) 2008 Jonathan Cameron
*
* Based on elements of hwmon and input subsystems.
*/
#define pr_fmt(fmt) "iio-core: " fmt
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/idr.h>
#include <linux/kdev_t.h>
#include <linux/err.h>
#include <linux/device.h>
#include <linux/fs.h>
#include <linux/poll.h>
#include <linux/property.h>
#include <linux/sched.h>
#include <linux/wait.h>
#include <linux/cdev.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/slab.h>
#include <linux/anon_inodes.h>
#include <linux/debugfs.h>
#include <linux/mutex.h>
#include <linux/iio/iio.h>
iio: core: wrap IIO device into an iio_dev_opaque object There are plenty of bad designs we want to discourage or not have to review manually usually about accessing private (marked as [INTERN]) fields of 'struct iio_dev'. Sometimes users copy drivers that are not always the best examples. A better idea is to hide those fields into the framework. For 'struct iio_dev' this is a 'struct iio_dev_opaque' which wraps a public 'struct iio_dev' object. In the next series, some fields will be moved to this new struct, each with it's own rework. This rework will not be complete-able for a while, as many fields need some drivers to be reworked in order to finalize them (e.g. 'indio_dev->mlock'). But some fields can already be moved, and in time, all of them may get there (in the 'struct iio_dev_opaque' object). Since a lot of drivers also call 'iio_priv()', in order to preserve fast-paths (where this matters), the public iio_dev object will have a 'priv' field that will have the pointer to the private information already computed. The reference returned by this field should be guaranteed to be cacheline aligned. The opaque parts will be moved into the 'include/linux/iio/iio-opaque.h' header. Should the hidden information be required for some debugging or some special needs, it can be made available via this header. Otherwise, only the IIO core files should include this file. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-06-30 04:57:03 +00:00
#include <linux/iio/iio-opaque.h>
#include "iio_core.h"
#include "iio_core_trigger.h"
#include <linux/iio/sysfs.h>
#include <linux/iio/events.h>
#include <linux/iio/buffer.h>
#include <linux/iio/buffer_impl.h>
/* IDA to assign each registered device a unique id */
static DEFINE_IDA(iio_ida);
static dev_t iio_devt;
#define IIO_DEV_MAX 256
struct bus_type iio_bus_type = {
.name = "iio",
};
EXPORT_SYMBOL(iio_bus_type);
static struct dentry *iio_debugfs_dentry;
static const char * const iio_direction[] = {
[0] = "in",
[1] = "out",
};
static const char * const iio_chan_type_name_spec[] = {
[IIO_VOLTAGE] = "voltage",
[IIO_CURRENT] = "current",
[IIO_POWER] = "power",
[IIO_ACCEL] = "accel",
[IIO_ANGL_VEL] = "anglvel",
[IIO_MAGN] = "magn",
[IIO_LIGHT] = "illuminance",
[IIO_INTENSITY] = "intensity",
[IIO_PROXIMITY] = "proximity",
[IIO_TEMP] = "temp",
[IIO_INCLI] = "incli",
[IIO_ROT] = "rot",
[IIO_ANGL] = "angl",
[IIO_TIMESTAMP] = "timestamp",
[IIO_CAPACITANCE] = "capacitance",
[IIO_ALTVOLTAGE] = "altvoltage",
[IIO_CCT] = "cct",
[IIO_PRESSURE] = "pressure",
[IIO_HUMIDITYRELATIVE] = "humidityrelative",
[IIO_ACTIVITY] = "activity",
[IIO_STEPS] = "steps",
[IIO_ENERGY] = "energy",
[IIO_DISTANCE] = "distance",
[IIO_VELOCITY] = "velocity",
[IIO_CONCENTRATION] = "concentration",
[IIO_RESISTANCE] = "resistance",
[IIO_PH] = "ph",
[IIO_UVINDEX] = "uvindex",
[IIO_ELECTRICALCONDUCTIVITY] = "electricalconductivity",
[IIO_COUNT] = "count",
[IIO_INDEX] = "index",
[IIO_GRAVITY] = "gravity",
[IIO_POSITIONRELATIVE] = "positionrelative",
[IIO_PHASE] = "phase",
[IIO_MASSCONCENTRATION] = "massconcentration",
};
static const char * const iio_modifier_names[] = {
[IIO_MOD_X] = "x",
[IIO_MOD_Y] = "y",
[IIO_MOD_Z] = "z",
[IIO_MOD_X_AND_Y] = "x&y",
[IIO_MOD_X_AND_Z] = "x&z",
[IIO_MOD_Y_AND_Z] = "y&z",
[IIO_MOD_X_AND_Y_AND_Z] = "x&y&z",
[IIO_MOD_X_OR_Y] = "x|y",
[IIO_MOD_X_OR_Z] = "x|z",
[IIO_MOD_Y_OR_Z] = "y|z",
[IIO_MOD_X_OR_Y_OR_Z] = "x|y|z",
[IIO_MOD_ROOT_SUM_SQUARED_X_Y] = "sqrt(x^2+y^2)",
[IIO_MOD_SUM_SQUARED_X_Y_Z] = "x^2+y^2+z^2",
[IIO_MOD_LIGHT_BOTH] = "both",
[IIO_MOD_LIGHT_IR] = "ir",
[IIO_MOD_LIGHT_CLEAR] = "clear",
[IIO_MOD_LIGHT_RED] = "red",
[IIO_MOD_LIGHT_GREEN] = "green",
[IIO_MOD_LIGHT_BLUE] = "blue",
[IIO_MOD_LIGHT_UV] = "uv",
[IIO_MOD_LIGHT_DUV] = "duv",
[IIO_MOD_QUATERNION] = "quaternion",
[IIO_MOD_TEMP_AMBIENT] = "ambient",
[IIO_MOD_TEMP_OBJECT] = "object",
[IIO_MOD_NORTH_MAGN] = "from_north_magnetic",
[IIO_MOD_NORTH_TRUE] = "from_north_true",
[IIO_MOD_NORTH_MAGN_TILT_COMP] = "from_north_magnetic_tilt_comp",
[IIO_MOD_NORTH_TRUE_TILT_COMP] = "from_north_true_tilt_comp",
[IIO_MOD_RUNNING] = "running",
[IIO_MOD_JOGGING] = "jogging",
[IIO_MOD_WALKING] = "walking",
[IIO_MOD_STILL] = "still",
[IIO_MOD_ROOT_SUM_SQUARED_X_Y_Z] = "sqrt(x^2+y^2+z^2)",
[IIO_MOD_I] = "i",
[IIO_MOD_Q] = "q",
[IIO_MOD_CO2] = "co2",
[IIO_MOD_VOC] = "voc",
[IIO_MOD_PM1] = "pm1",
[IIO_MOD_PM2P5] = "pm2p5",
[IIO_MOD_PM4] = "pm4",
[IIO_MOD_PM10] = "pm10",
[IIO_MOD_ETHANOL] = "ethanol",
[IIO_MOD_H2] = "h2",
[IIO_MOD_O2] = "o2",
};
/* relies on pairs of these shared then separate */
static const char * const iio_chan_info_postfix[] = {
[IIO_CHAN_INFO_RAW] = "raw",
[IIO_CHAN_INFO_PROCESSED] = "input",
[IIO_CHAN_INFO_SCALE] = "scale",
[IIO_CHAN_INFO_OFFSET] = "offset",
[IIO_CHAN_INFO_CALIBSCALE] = "calibscale",
[IIO_CHAN_INFO_CALIBBIAS] = "calibbias",
[IIO_CHAN_INFO_PEAK] = "peak_raw",
[IIO_CHAN_INFO_PEAK_SCALE] = "peak_scale",
[IIO_CHAN_INFO_QUADRATURE_CORRECTION_RAW] = "quadrature_correction_raw",
[IIO_CHAN_INFO_AVERAGE_RAW] = "mean_raw",
[IIO_CHAN_INFO_LOW_PASS_FILTER_3DB_FREQUENCY]
= "filter_low_pass_3db_frequency",
[IIO_CHAN_INFO_HIGH_PASS_FILTER_3DB_FREQUENCY]
= "filter_high_pass_3db_frequency",
[IIO_CHAN_INFO_SAMP_FREQ] = "sampling_frequency",
[IIO_CHAN_INFO_FREQUENCY] = "frequency",
[IIO_CHAN_INFO_PHASE] = "phase",
[IIO_CHAN_INFO_HARDWAREGAIN] = "hardwaregain",
[IIO_CHAN_INFO_HYSTERESIS] = "hysteresis",
[IIO_CHAN_INFO_HYSTERESIS_RELATIVE] = "hysteresis_relative",
iio: Add INT_TIME (integration time) channel info attribute Integration time is in seconds; it controls the measurement time and influences the gain of a sensor. There are two typical ways that scaling is implemented in a device: 1) input amplifier, 2) reference to the ADC is changed. These both result in the accuracy of the ADC varying (by applying its sampling over a more relevant range). Integration time is a way of dealing with noise inherent in the analog sensor itself. In the case of a light sensor, a mixture of photon noise and device specific noise. Photon noise is dealt with by either improving the efficiency of the sensor, (more photons actually captured) which is not easily varied dynamically, or by integrating the measurement over a longer time period. Note that this can also be thought of as an averaging of a number of individual samples and is infact sometimes implemented this way. Altering integration time implies that the duration of a measurement changes, a fact the device's user may be interested in. Hence it makes sense to distinguish between integration time and simple scale. In some devices both types of control are present and whilst they will have similar effects on the amplitude of the reading, their effect on the noise of the measurements will differ considerably. Used by adjd_s311, tsl4531, tcs3472 The following drivers have similar controls (and could be adapted): * tsl2563 (integration time is controlled via CALIBSCALE among other things) * tsl2583 (has integration_time device_attr, but driver doesn't use channels yet) * tsl2x7x (has integration_time attr) Signed-off-by: Peter Meerwald <pmeerw@pmeerw.net> Cc: Jon Brenner <jon.brenner@ams.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
2013-09-08 15:20:00 +00:00
[IIO_CHAN_INFO_INT_TIME] = "integration_time",
[IIO_CHAN_INFO_ENABLE] = "en",
[IIO_CHAN_INFO_CALIBHEIGHT] = "calibheight",
[IIO_CHAN_INFO_CALIBWEIGHT] = "calibweight",
[IIO_CHAN_INFO_DEBOUNCE_COUNT] = "debounce_count",
[IIO_CHAN_INFO_DEBOUNCE_TIME] = "debounce_time",
[IIO_CHAN_INFO_CALIBEMISSIVITY] = "calibemissivity",
[IIO_CHAN_INFO_OVERSAMPLING_RATIO] = "oversampling_ratio",
[IIO_CHAN_INFO_THERMOCOUPLE_TYPE] = "thermocouple_type",
[IIO_CHAN_INFO_CALIBAMBIENT] = "calibambient",
};
/**
* iio_device_id() - query the unique ID for the device
* @indio_dev: Device structure whose ID is being queried
*
* The IIO device ID is a unique index used for example for the naming
* of the character device /dev/iio\:device[ID]
*/
int iio_device_id(struct iio_dev *indio_dev)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
return iio_dev_opaque->id;
}
EXPORT_SYMBOL_GPL(iio_device_id);
/**
* iio_sysfs_match_string_with_gaps - matches given string in an array with gaps
* @array: array of strings
* @n: number of strings in the array
* @str: string to match with
*
* Returns index of @str in the @array or -EINVAL, similar to match_string().
* Uses sysfs_streq instead of strcmp for matching.
*
* This routine will look for a string in an array of strings.
* The search will continue until the element is found or the n-th element
* is reached, regardless of any NULL elements in the array.
*/
static int iio_sysfs_match_string_with_gaps(const char * const *array, size_t n,
const char *str)
{
const char *item;
int index;
for (index = 0; index < n; index++) {
item = array[index];
if (!item)
continue;
if (sysfs_streq(item, str))
return index;
}
return -EINVAL;
}
#if defined(CONFIG_DEBUG_FS)
/*
* There's also a CONFIG_DEBUG_FS guard in include/linux/iio/iio.h for
* iio_get_debugfs_dentry() to make it inline if CONFIG_DEBUG_FS is undefined
*/
struct dentry *iio_get_debugfs_dentry(struct iio_dev *indio_dev)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
return iio_dev_opaque->debugfs_dentry;
}
EXPORT_SYMBOL_GPL(iio_get_debugfs_dentry);
#endif
/**
* iio_find_channel_from_si() - get channel from its scan index
* @indio_dev: device
* @si: scan index to match
*/
const struct iio_chan_spec
*iio_find_channel_from_si(struct iio_dev *indio_dev, int si)
{
int i;
for (i = 0; i < indio_dev->num_channels; i++)
if (indio_dev->channels[i].scan_index == si)
return &indio_dev->channels[i];
return NULL;
}
/* This turns up an awful lot */
ssize_t iio_read_const_attr(struct device *dev,
struct device_attribute *attr,
char *buf)
{
return sysfs_emit(buf, "%s\n", to_iio_const_attr(attr)->string);
}
EXPORT_SYMBOL(iio_read_const_attr);
/**
* iio_device_set_clock() - Set current timestamping clock for the device
* @indio_dev: IIO device structure containing the device
* @clock_id: timestamping clock posix identifier to set.
*/
int iio_device_set_clock(struct iio_dev *indio_dev, clockid_t clock_id)
{
int ret;
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
const struct iio_event_interface *ev_int = iio_dev_opaque->event_interface;
ret = mutex_lock_interruptible(&indio_dev->mlock);
if (ret)
return ret;
if ((ev_int && iio_event_enabled(ev_int)) ||
iio_buffer_enabled(indio_dev)) {
mutex_unlock(&indio_dev->mlock);
return -EBUSY;
}
iio_dev_opaque->clock_id = clock_id;
mutex_unlock(&indio_dev->mlock);
return 0;
}
EXPORT_SYMBOL(iio_device_set_clock);
/**
* iio_device_get_clock() - Retrieve current timestamping clock for the device
* @indio_dev: IIO device structure containing the device
*/
clockid_t iio_device_get_clock(const struct iio_dev *indio_dev)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
return iio_dev_opaque->clock_id;
}
EXPORT_SYMBOL(iio_device_get_clock);
/**
* iio_get_time_ns() - utility function to get a time stamp for events etc
* @indio_dev: device
*/
s64 iio_get_time_ns(const struct iio_dev *indio_dev)
{
struct timespec64 tp;
switch (iio_device_get_clock(indio_dev)) {
case CLOCK_REALTIME:
return ktime_get_real_ns();
case CLOCK_MONOTONIC:
return ktime_get_ns();
case CLOCK_MONOTONIC_RAW:
return ktime_get_raw_ns();
case CLOCK_REALTIME_COARSE:
return ktime_to_ns(ktime_get_coarse_real());
case CLOCK_MONOTONIC_COARSE:
ktime_get_coarse_ts64(&tp);
return timespec64_to_ns(&tp);
case CLOCK_BOOTTIME:
return ktime_get_boottime_ns();
case CLOCK_TAI:
return ktime_get_clocktai_ns();
default:
BUG();
}
}
EXPORT_SYMBOL(iio_get_time_ns);
/**
* iio_get_time_res() - utility function to get time stamp clock resolution in
* nano seconds.
* @indio_dev: device
*/
unsigned int iio_get_time_res(const struct iio_dev *indio_dev)
{
switch (iio_device_get_clock(indio_dev)) {
case CLOCK_REALTIME:
case CLOCK_MONOTONIC:
case CLOCK_MONOTONIC_RAW:
case CLOCK_BOOTTIME:
case CLOCK_TAI:
return hrtimer_resolution;
case CLOCK_REALTIME_COARSE:
case CLOCK_MONOTONIC_COARSE:
return LOW_RES_NSEC;
default:
BUG();
}
}
EXPORT_SYMBOL(iio_get_time_res);
static int __init iio_init(void)
{
int ret;
/* Register sysfs bus */
ret = bus_register(&iio_bus_type);
if (ret < 0) {
pr_err("could not register bus type\n");
goto error_nothing;
}
ret = alloc_chrdev_region(&iio_devt, 0, IIO_DEV_MAX, "iio");
if (ret < 0) {
pr_err("failed to allocate char dev region\n");
goto error_unregister_bus_type;
}
iio_debugfs_dentry = debugfs_create_dir("iio", NULL);
return 0;
error_unregister_bus_type:
bus_unregister(&iio_bus_type);
error_nothing:
return ret;
}
static void __exit iio_exit(void)
{
if (iio_devt)
unregister_chrdev_region(iio_devt, IIO_DEV_MAX);
bus_unregister(&iio_bus_type);
debugfs_remove(iio_debugfs_dentry);
}
#if defined(CONFIG_DEBUG_FS)
static ssize_t iio_debugfs_read_reg(struct file *file, char __user *userbuf,
size_t count, loff_t *ppos)
{
struct iio_dev *indio_dev = file->private_data;
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
unsigned val = 0;
int ret;
if (*ppos > 0)
return simple_read_from_buffer(userbuf, count, ppos,
iio_dev_opaque->read_buf,
iio_dev_opaque->read_buf_len);
ret = indio_dev->info->debugfs_reg_access(indio_dev,
iio_dev_opaque->cached_reg_addr,
0, &val);
if (ret) {
dev_err(indio_dev->dev.parent, "%s: read failed\n", __func__);
return ret;
}
iio_dev_opaque->read_buf_len = snprintf(iio_dev_opaque->read_buf,
sizeof(iio_dev_opaque->read_buf),
"0x%X\n", val);
return simple_read_from_buffer(userbuf, count, ppos,
iio_dev_opaque->read_buf,
iio_dev_opaque->read_buf_len);
}
static ssize_t iio_debugfs_write_reg(struct file *file,
const char __user *userbuf, size_t count, loff_t *ppos)
{
struct iio_dev *indio_dev = file->private_data;
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
unsigned reg, val;
char buf[80];
int ret;
count = min_t(size_t, count, (sizeof(buf)-1));
if (copy_from_user(buf, userbuf, count))
return -EFAULT;
buf[count] = 0;
ret = sscanf(buf, "%i %i", &reg, &val);
switch (ret) {
case 1:
iio_dev_opaque->cached_reg_addr = reg;
break;
case 2:
iio_dev_opaque->cached_reg_addr = reg;
ret = indio_dev->info->debugfs_reg_access(indio_dev, reg,
val, NULL);
if (ret) {
dev_err(indio_dev->dev.parent, "%s: write failed\n",
__func__);
return ret;
}
break;
default:
return -EINVAL;
}
return count;
}
static const struct file_operations iio_debugfs_reg_fops = {
.open = simple_open,
.read = iio_debugfs_read_reg,
.write = iio_debugfs_write_reg,
};
static void iio_device_unregister_debugfs(struct iio_dev *indio_dev)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
debugfs_remove_recursive(iio_dev_opaque->debugfs_dentry);
}
static void iio_device_register_debugfs(struct iio_dev *indio_dev)
{
struct iio_dev_opaque *iio_dev_opaque;
if (indio_dev->info->debugfs_reg_access == NULL)
return;
if (!iio_debugfs_dentry)
return;
iio_dev_opaque = to_iio_dev_opaque(indio_dev);
iio_dev_opaque->debugfs_dentry =
debugfs_create_dir(dev_name(&indio_dev->dev),
iio_debugfs_dentry);
debugfs_create_file("direct_reg_access", 0644,
iio_dev_opaque->debugfs_dentry, indio_dev,
&iio_debugfs_reg_fops);
}
#else
static void iio_device_register_debugfs(struct iio_dev *indio_dev)
{
}
static void iio_device_unregister_debugfs(struct iio_dev *indio_dev)
{
}
#endif /* CONFIG_DEBUG_FS */
static ssize_t iio_read_channel_ext_info(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct iio_dev *indio_dev = dev_to_iio_dev(dev);
struct iio_dev_attr *this_attr = to_iio_dev_attr(attr);
const struct iio_chan_spec_ext_info *ext_info;
ext_info = &this_attr->c->ext_info[this_attr->address];
return ext_info->read(indio_dev, ext_info->private, this_attr->c, buf);
}
static ssize_t iio_write_channel_ext_info(struct device *dev,
struct device_attribute *attr,
const char *buf,
size_t len)
{
struct iio_dev *indio_dev = dev_to_iio_dev(dev);
struct iio_dev_attr *this_attr = to_iio_dev_attr(attr);
const struct iio_chan_spec_ext_info *ext_info;
ext_info = &this_attr->c->ext_info[this_attr->address];
return ext_info->write(indio_dev, ext_info->private,
this_attr->c, buf, len);
}
ssize_t iio_enum_available_read(struct iio_dev *indio_dev,
uintptr_t priv, const struct iio_chan_spec *chan, char *buf)
{
const struct iio_enum *e = (const struct iio_enum *)priv;
unsigned int i;
size_t len = 0;
if (!e->num_items)
return 0;
for (i = 0; i < e->num_items; ++i) {
if (!e->items[i])
continue;
len += sysfs_emit_at(buf, len, "%s ", e->items[i]);
}
/* replace last space with a newline */
buf[len - 1] = '\n';
return len;
}
EXPORT_SYMBOL_GPL(iio_enum_available_read);
ssize_t iio_enum_read(struct iio_dev *indio_dev,
uintptr_t priv, const struct iio_chan_spec *chan, char *buf)
{
const struct iio_enum *e = (const struct iio_enum *)priv;
int i;
if (!e->get)
return -EINVAL;
i = e->get(indio_dev, chan);
if (i < 0)
return i;
else if (i >= e->num_items || !e->items[i])
return -EINVAL;
return sysfs_emit(buf, "%s\n", e->items[i]);
}
EXPORT_SYMBOL_GPL(iio_enum_read);
ssize_t iio_enum_write(struct iio_dev *indio_dev,
uintptr_t priv, const struct iio_chan_spec *chan, const char *buf,
size_t len)
{
const struct iio_enum *e = (const struct iio_enum *)priv;
int ret;
if (!e->set)
return -EINVAL;
ret = iio_sysfs_match_string_with_gaps(e->items, e->num_items, buf);
if (ret < 0)
return ret;
ret = e->set(indio_dev, chan, ret);
return ret ? ret : len;
}
EXPORT_SYMBOL_GPL(iio_enum_write);
static const struct iio_mount_matrix iio_mount_idmatrix = {
.rotation = {
"1", "0", "0",
"0", "1", "0",
"0", "0", "1"
}
};
static int iio_setup_mount_idmatrix(const struct device *dev,
struct iio_mount_matrix *matrix)
{
*matrix = iio_mount_idmatrix;
dev_info(dev, "mounting matrix not found: using identity...\n");
return 0;
}
ssize_t iio_show_mount_matrix(struct iio_dev *indio_dev, uintptr_t priv,
const struct iio_chan_spec *chan, char *buf)
{
const struct iio_mount_matrix *mtx = ((iio_get_mount_matrix_t *)
priv)(indio_dev, chan);
if (IS_ERR(mtx))
return PTR_ERR(mtx);
if (!mtx)
mtx = &iio_mount_idmatrix;
return sysfs_emit(buf, "%s, %s, %s; %s, %s, %s; %s, %s, %s\n",
mtx->rotation[0], mtx->rotation[1], mtx->rotation[2],
mtx->rotation[3], mtx->rotation[4], mtx->rotation[5],
mtx->rotation[6], mtx->rotation[7], mtx->rotation[8]);
}
EXPORT_SYMBOL_GPL(iio_show_mount_matrix);
/**
* iio_read_mount_matrix() - retrieve iio device mounting matrix from
* device "mount-matrix" property
* @dev: device the mounting matrix property is assigned to
* @matrix: where to store retrieved matrix
*
* If device is assigned no mounting matrix property, a default 3x3 identity
* matrix will be filled in.
*
* Return: 0 if success, or a negative error code on failure.
*/
int iio_read_mount_matrix(struct device *dev, struct iio_mount_matrix *matrix)
{
size_t len = ARRAY_SIZE(iio_mount_idmatrix.rotation);
int err;
err = device_property_read_string_array(dev, "mount-matrix", matrix->rotation, len);
if (err == len)
return 0;
if (err >= 0)
/* Invalid number of matrix entries. */
return -EINVAL;
if (err != -EINVAL)
/* Invalid matrix declaration format. */
return err;
/* Matrix was not declared at all: fallback to identity. */
return iio_setup_mount_idmatrix(dev, matrix);
}
EXPORT_SYMBOL(iio_read_mount_matrix);
static ssize_t __iio_format_value(char *buf, size_t offset, unsigned int type,
int size, const int *vals)
{
int tmp0, tmp1;
s64 tmp2;
bool scale_db = false;
switch (type) {
case IIO_VAL_INT:
return sysfs_emit_at(buf, offset, "%d", vals[0]);
case IIO_VAL_INT_PLUS_MICRO_DB:
scale_db = true;
fallthrough;
case IIO_VAL_INT_PLUS_MICRO:
if (vals[1] < 0)
return sysfs_emit_at(buf, offset, "-%d.%06u%s",
abs(vals[0]), -vals[1],
scale_db ? " dB" : "");
else
return sysfs_emit_at(buf, offset, "%d.%06u%s", vals[0],
vals[1], scale_db ? " dB" : "");
case IIO_VAL_INT_PLUS_NANO:
if (vals[1] < 0)
return sysfs_emit_at(buf, offset, "-%d.%09u",
abs(vals[0]), -vals[1]);
else
return sysfs_emit_at(buf, offset, "%d.%09u", vals[0],
vals[1]);
case IIO_VAL_FRACTIONAL:
tmp2 = div_s64((s64)vals[0] * 1000000000LL, vals[1]);
tmp1 = vals[1];
tmp0 = (int)div_s64_rem(tmp2, 1000000000, &tmp1);
if ((tmp2 < 0) && (tmp0 == 0))
return sysfs_emit_at(buf, offset, "-0.%09u", abs(tmp1));
else
return sysfs_emit_at(buf, offset, "%d.%09u", tmp0,
abs(tmp1));
case IIO_VAL_FRACTIONAL_LOG2:
tmp2 = shift_right((s64)vals[0] * 1000000000LL, vals[1]);
tmp0 = (int)div_s64_rem(tmp2, 1000000000LL, &tmp1);
if (tmp0 == 0 && tmp2 < 0)
return sysfs_emit_at(buf, offset, "-0.%09u", abs(tmp1));
else
return sysfs_emit_at(buf, offset, "%d.%09u", tmp0,
abs(tmp1));
case IIO_VAL_INT_MULTIPLE:
{
int i;
int l = 0;
for (i = 0; i < size; ++i)
l += sysfs_emit_at(buf, offset + l, "%d ", vals[i]);
return l;
}
case IIO_VAL_CHAR:
return sysfs_emit_at(buf, offset, "%c", (char)vals[0]);
default:
return 0;
}
}
/**
* iio_format_value() - Formats a IIO value into its string representation
* @buf: The buffer to which the formatted value gets written
* which is assumed to be big enough (i.e. PAGE_SIZE).
* @type: One of the IIO_VAL_* constants. This decides how the val
* and val2 parameters are formatted.
* @size: Number of IIO value entries contained in vals
* @vals: Pointer to the values, exact meaning depends on the
* type parameter.
*
* Return: 0 by default, a negative number on failure or the
* total number of characters written for a type that belongs
* to the IIO_VAL_* constant.
*/
ssize_t iio_format_value(char *buf, unsigned int type, int size, int *vals)
{
ssize_t len;
len = __iio_format_value(buf, 0, type, size, vals);
if (len >= PAGE_SIZE - 1)
return -EFBIG;
return len + sysfs_emit_at(buf, len, "\n");
}
EXPORT_SYMBOL_GPL(iio_format_value);
static ssize_t iio_read_channel_label(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct iio_dev *indio_dev = dev_to_iio_dev(dev);
struct iio_dev_attr *this_attr = to_iio_dev_attr(attr);
if (!indio_dev->info->read_label)
return -EINVAL;
return indio_dev->info->read_label(indio_dev, this_attr->c, buf);
}
static ssize_t iio_read_channel_info(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct iio_dev *indio_dev = dev_to_iio_dev(dev);
struct iio_dev_attr *this_attr = to_iio_dev_attr(attr);
int vals[INDIO_MAX_RAW_ELEMENTS];
int ret;
int val_len = 2;
if (indio_dev->info->read_raw_multi)
ret = indio_dev->info->read_raw_multi(indio_dev, this_attr->c,
INDIO_MAX_RAW_ELEMENTS,
vals, &val_len,
this_attr->address);
else
ret = indio_dev->info->read_raw(indio_dev, this_attr->c,
&vals[0], &vals[1], this_attr->address);
if (ret < 0)
return ret;
return iio_format_value(buf, ret, val_len, vals);
}
static ssize_t iio_format_list(char *buf, const int *vals, int type, int length,
const char *prefix, const char *suffix)
{
ssize_t len;
int stride;
int i;
switch (type) {
case IIO_VAL_INT:
stride = 1;
break;
default:
stride = 2;
break;
}
len = sysfs_emit(buf, prefix);
for (i = 0; i <= length - stride; i += stride) {
if (i != 0) {
len += sysfs_emit_at(buf, len, " ");
if (len >= PAGE_SIZE)
return -EFBIG;
}
len += __iio_format_value(buf, len, type, stride, &vals[i]);
if (len >= PAGE_SIZE)
return -EFBIG;
}
len += sysfs_emit_at(buf, len, "%s\n", suffix);
return len;
}
static ssize_t iio_format_avail_list(char *buf, const int *vals,
int type, int length)
{
return iio_format_list(buf, vals, type, length, "", "");
}
static ssize_t iio_format_avail_range(char *buf, const int *vals, int type)
{
return iio_format_list(buf, vals, type, 3, "[", "]");
}
static ssize_t iio_read_channel_info_avail(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct iio_dev *indio_dev = dev_to_iio_dev(dev);
struct iio_dev_attr *this_attr = to_iio_dev_attr(attr);
const int *vals;
int ret;
int length;
int type;
ret = indio_dev->info->read_avail(indio_dev, this_attr->c,
&vals, &type, &length,
this_attr->address);
if (ret < 0)
return ret;
switch (ret) {
case IIO_AVAIL_LIST:
return iio_format_avail_list(buf, vals, type, length);
case IIO_AVAIL_RANGE:
return iio_format_avail_range(buf, vals, type);
default:
return -EINVAL;
}
}
/**
* __iio_str_to_fixpoint() - Parse a fixed-point number from a string
* @str: The string to parse
* @fract_mult: Multiplier for the first decimal place, should be a power of 10
* @integer: The integer part of the number
* @fract: The fractional part of the number
* @scale_db: True if this should parse as dB
*
* Returns 0 on success, or a negative error code if the string could not be
* parsed.
*/
static int __iio_str_to_fixpoint(const char *str, int fract_mult,
int *integer, int *fract, bool scale_db)
{
int i = 0, f = 0;
bool integer_part = true, negative = false;
if (fract_mult == 0) {
*fract = 0;
return kstrtoint(str, 0, integer);
}
if (str[0] == '-') {
negative = true;
str++;
} else if (str[0] == '+') {
str++;
}
while (*str) {
if ('0' <= *str && *str <= '9') {
if (integer_part) {
i = i * 10 + *str - '0';
} else {
f += fract_mult * (*str - '0');
fract_mult /= 10;
}
} else if (*str == '\n') {
if (*(str + 1) == '\0')
break;
else
return -EINVAL;
} else if (!strncmp(str, " dB", sizeof(" dB") - 1) && scale_db) {
/* Ignore the dB suffix */
str += sizeof(" dB") - 1;
continue;
} else if (!strncmp(str, "dB", sizeof("dB") - 1) && scale_db) {
/* Ignore the dB suffix */
str += sizeof("dB") - 1;
continue;
} else if (*str == '.' && integer_part) {
integer_part = false;
} else {
return -EINVAL;
}
str++;
}
if (negative) {
if (i)
i = -i;
else
f = -f;
}
*integer = i;
*fract = f;
return 0;
}
/**
* iio_str_to_fixpoint() - Parse a fixed-point number from a string
* @str: The string to parse
* @fract_mult: Multiplier for the first decimal place, should be a power of 10
* @integer: The integer part of the number
* @fract: The fractional part of the number
*
* Returns 0 on success, or a negative error code if the string could not be
* parsed.
*/
int iio_str_to_fixpoint(const char *str, int fract_mult,
int *integer, int *fract)
{
return __iio_str_to_fixpoint(str, fract_mult, integer, fract, false);
}
EXPORT_SYMBOL_GPL(iio_str_to_fixpoint);
static ssize_t iio_write_channel_info(struct device *dev,
struct device_attribute *attr,
const char *buf,
size_t len)
{
struct iio_dev *indio_dev = dev_to_iio_dev(dev);
struct iio_dev_attr *this_attr = to_iio_dev_attr(attr);
int ret, fract_mult = 100000;
int integer, fract = 0;
bool is_char = false;
bool scale_db = false;
/* Assumes decimal - precision based on number of digits */
if (!indio_dev->info->write_raw)
return -EINVAL;
if (indio_dev->info->write_raw_get_fmt)
switch (indio_dev->info->write_raw_get_fmt(indio_dev,
this_attr->c, this_attr->address)) {
case IIO_VAL_INT:
fract_mult = 0;
break;
case IIO_VAL_INT_PLUS_MICRO_DB:
scale_db = true;
fallthrough;
case IIO_VAL_INT_PLUS_MICRO:
fract_mult = 100000;
break;
case IIO_VAL_INT_PLUS_NANO:
fract_mult = 100000000;
break;
case IIO_VAL_CHAR:
is_char = true;
break;
default:
return -EINVAL;
}
if (is_char) {
char ch;
if (sscanf(buf, "%c", &ch) != 1)
return -EINVAL;
integer = ch;
} else {
ret = __iio_str_to_fixpoint(buf, fract_mult, &integer, &fract,
scale_db);
if (ret)
return ret;
}
ret = indio_dev->info->write_raw(indio_dev, this_attr->c,
integer, fract, this_attr->address);
if (ret)
return ret;
return len;
}
static
int __iio_device_attr_init(struct device_attribute *dev_attr,
const char *postfix,
struct iio_chan_spec const *chan,
ssize_t (*readfunc)(struct device *dev,
struct device_attribute *attr,
char *buf),
ssize_t (*writefunc)(struct device *dev,
struct device_attribute *attr,
const char *buf,
size_t len),
enum iio_shared_by shared_by)
{
int ret = 0;
char *name = NULL;
char *full_postfix;
sysfs_attr_init(&dev_attr->attr);
/* Build up postfix of <extend_name>_<modifier>_postfix */
if (chan->modified && (shared_by == IIO_SEPARATE)) {
if (chan->extend_name)
full_postfix = kasprintf(GFP_KERNEL, "%s_%s_%s",
iio_modifier_names[chan
->channel2],
chan->extend_name,
postfix);
else
full_postfix = kasprintf(GFP_KERNEL, "%s_%s",
iio_modifier_names[chan
->channel2],
postfix);
} else {
if (chan->extend_name == NULL || shared_by != IIO_SEPARATE)
full_postfix = kstrdup(postfix, GFP_KERNEL);
else
full_postfix = kasprintf(GFP_KERNEL,
"%s_%s",
chan->extend_name,
postfix);
}
if (full_postfix == NULL)
return -ENOMEM;
if (chan->differential) { /* Differential can not have modifier */
switch (shared_by) {
case IIO_SHARED_BY_ALL:
name = kasprintf(GFP_KERNEL, "%s", full_postfix);
break;
case IIO_SHARED_BY_DIR:
name = kasprintf(GFP_KERNEL, "%s_%s",
iio_direction[chan->output],
full_postfix);
break;
case IIO_SHARED_BY_TYPE:
name = kasprintf(GFP_KERNEL, "%s_%s-%s_%s",
iio_direction[chan->output],
iio_chan_type_name_spec[chan->type],
iio_chan_type_name_spec[chan->type],
full_postfix);
break;
case IIO_SEPARATE:
if (!chan->indexed) {
WARN(1, "Differential channels must be indexed\n");
ret = -EINVAL;
goto error_free_full_postfix;
}
name = kasprintf(GFP_KERNEL,
"%s_%s%d-%s%d_%s",
iio_direction[chan->output],
iio_chan_type_name_spec[chan->type],
chan->channel,
iio_chan_type_name_spec[chan->type],
chan->channel2,
full_postfix);
break;
}
} else { /* Single ended */
switch (shared_by) {
case IIO_SHARED_BY_ALL:
name = kasprintf(GFP_KERNEL, "%s", full_postfix);
break;
case IIO_SHARED_BY_DIR:
name = kasprintf(GFP_KERNEL, "%s_%s",
iio_direction[chan->output],
full_postfix);
break;
case IIO_SHARED_BY_TYPE:
name = kasprintf(GFP_KERNEL, "%s_%s_%s",
iio_direction[chan->output],
iio_chan_type_name_spec[chan->type],
full_postfix);
break;
case IIO_SEPARATE:
if (chan->indexed)
name = kasprintf(GFP_KERNEL, "%s_%s%d_%s",
iio_direction[chan->output],
iio_chan_type_name_spec[chan->type],
chan->channel,
full_postfix);
else
name = kasprintf(GFP_KERNEL, "%s_%s_%s",
iio_direction[chan->output],
iio_chan_type_name_spec[chan->type],
full_postfix);
break;
}
}
if (name == NULL) {
ret = -ENOMEM;
goto error_free_full_postfix;
}
dev_attr->attr.name = name;
if (readfunc) {
dev_attr->attr.mode |= S_IRUGO;
dev_attr->show = readfunc;
}
if (writefunc) {
dev_attr->attr.mode |= S_IWUSR;
dev_attr->store = writefunc;
}
error_free_full_postfix:
kfree(full_postfix);
return ret;
}
static void __iio_device_attr_deinit(struct device_attribute *dev_attr)
{
kfree(dev_attr->attr.name);
}
int __iio_add_chan_devattr(const char *postfix,
struct iio_chan_spec const *chan,
ssize_t (*readfunc)(struct device *dev,
struct device_attribute *attr,
char *buf),
ssize_t (*writefunc)(struct device *dev,
struct device_attribute *attr,
const char *buf,
size_t len),
u64 mask,
enum iio_shared_by shared_by,
struct device *dev,
struct iio_buffer *buffer,
struct list_head *attr_list)
{
int ret;
struct iio_dev_attr *iio_attr, *t;
iio_attr = kzalloc(sizeof(*iio_attr), GFP_KERNEL);
if (iio_attr == NULL)
return -ENOMEM;
ret = __iio_device_attr_init(&iio_attr->dev_attr,
postfix, chan,
readfunc, writefunc, shared_by);
if (ret)
goto error_iio_dev_attr_free;
iio_attr->c = chan;
iio_attr->address = mask;
iio_attr->buffer = buffer;
list_for_each_entry(t, attr_list, l)
if (strcmp(t->dev_attr.attr.name,
iio_attr->dev_attr.attr.name) == 0) {
if (shared_by == IIO_SEPARATE)
dev_err(dev, "tried to double register : %s\n",
t->dev_attr.attr.name);
ret = -EBUSY;
goto error_device_attr_deinit;
}
list_add(&iio_attr->l, attr_list);
return 0;
error_device_attr_deinit:
__iio_device_attr_deinit(&iio_attr->dev_attr);
error_iio_dev_attr_free:
kfree(iio_attr);
return ret;
}
static int iio_device_add_channel_label(struct iio_dev *indio_dev,
struct iio_chan_spec const *chan)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
int ret;
if (!indio_dev->info->read_label)
return 0;
ret = __iio_add_chan_devattr("label",
chan,
&iio_read_channel_label,
NULL,
0,
IIO_SEPARATE,
&indio_dev->dev,
NULL,
&iio_dev_opaque->channel_attr_list);
if (ret < 0)
return ret;
return 1;
}
static int iio_device_add_info_mask_type(struct iio_dev *indio_dev,
struct iio_chan_spec const *chan,
enum iio_shared_by shared_by,
const long *infomask)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
int i, ret, attrcount = 0;
for_each_set_bit(i, infomask, sizeof(*infomask)*8) {
if (i >= ARRAY_SIZE(iio_chan_info_postfix))
return -EINVAL;
ret = __iio_add_chan_devattr(iio_chan_info_postfix[i],
chan,
&iio_read_channel_info,
&iio_write_channel_info,
i,
shared_by,
&indio_dev->dev,
NULL,
&iio_dev_opaque->channel_attr_list);
if ((ret == -EBUSY) && (shared_by != IIO_SEPARATE))
continue;
else if (ret < 0)
return ret;
attrcount++;
}
staging:iio: Add extended IIO channel info Sometimes devices have per channel properties which either do not map nicely to the current channel info scheme (e.g. string properties) or are very device specific, so it does not make sense to add generic support for them. Currently drivers define these attributes by hand for each channel. Depending on the number of channels this can amount to quite a few lines of boilerplate code. Especially if a driver supports multiple variations of a chip with different numbers of channels. In this case it becomes necessary to have a individual attribute list per chip variation and also a individual iio_info struct. This patch introduces a new scheme for handling such per channel attributes called extended channel info attributes. A extended channel info attribute consist of a name, a flag whether it is shared and read and write callbacks. The read and write callbacks are similar to the {read,write}_raw callbacks and take a IIO device and a channel as their first parameters, but instead of pre-parsed integer values they directly get passed the raw string value, which has been written to the sysfs file. It is possible to assign a list of extended channel info attributes to a channel. For each extended channel info attribute the IIO core will create a new sysfs attribute conforming to the IIO channel naming spec for the channels type, similar as for normal info attributes. Read and write access to this sysfs attribute will be redirected to the extended channel info attributes read and write callbacks. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Acked-by: Jonathan Cameron <jic23@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-02-21 17:38:12 +00:00
return attrcount;
}
static int iio_device_add_info_mask_type_avail(struct iio_dev *indio_dev,
struct iio_chan_spec const *chan,
enum iio_shared_by shared_by,
const long *infomask)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
int i, ret, attrcount = 0;
char *avail_postfix;
for_each_set_bit(i, infomask, sizeof(*infomask) * 8) {
if (i >= ARRAY_SIZE(iio_chan_info_postfix))
return -EINVAL;
avail_postfix = kasprintf(GFP_KERNEL,
"%s_available",
iio_chan_info_postfix[i]);
if (!avail_postfix)
return -ENOMEM;
ret = __iio_add_chan_devattr(avail_postfix,
chan,
&iio_read_channel_info_avail,
NULL,
i,
shared_by,
&indio_dev->dev,
NULL,
&iio_dev_opaque->channel_attr_list);
kfree(avail_postfix);
if ((ret == -EBUSY) && (shared_by != IIO_SEPARATE))
continue;
else if (ret < 0)
return ret;
attrcount++;
}
return attrcount;
}
static int iio_device_add_channel_sysfs(struct iio_dev *indio_dev,
struct iio_chan_spec const *chan)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
int ret, attrcount = 0;
const struct iio_chan_spec_ext_info *ext_info;
if (chan->channel < 0)
return 0;
ret = iio_device_add_info_mask_type(indio_dev, chan,
IIO_SEPARATE,
&chan->info_mask_separate);
if (ret < 0)
return ret;
attrcount += ret;
ret = iio_device_add_info_mask_type_avail(indio_dev, chan,
IIO_SEPARATE,
&chan->
info_mask_separate_available);
if (ret < 0)
return ret;
attrcount += ret;
ret = iio_device_add_info_mask_type(indio_dev, chan,
IIO_SHARED_BY_TYPE,
&chan->info_mask_shared_by_type);
if (ret < 0)
return ret;
attrcount += ret;
ret = iio_device_add_info_mask_type_avail(indio_dev, chan,
IIO_SHARED_BY_TYPE,
&chan->
info_mask_shared_by_type_available);
if (ret < 0)
return ret;
attrcount += ret;
ret = iio_device_add_info_mask_type(indio_dev, chan,
IIO_SHARED_BY_DIR,
&chan->info_mask_shared_by_dir);
if (ret < 0)
return ret;
attrcount += ret;
ret = iio_device_add_info_mask_type_avail(indio_dev, chan,
IIO_SHARED_BY_DIR,
&chan->info_mask_shared_by_dir_available);
if (ret < 0)
return ret;
attrcount += ret;
ret = iio_device_add_info_mask_type(indio_dev, chan,
IIO_SHARED_BY_ALL,
&chan->info_mask_shared_by_all);
if (ret < 0)
return ret;
attrcount += ret;
ret = iio_device_add_info_mask_type_avail(indio_dev, chan,
IIO_SHARED_BY_ALL,
&chan->info_mask_shared_by_all_available);
if (ret < 0)
return ret;
attrcount += ret;
ret = iio_device_add_channel_label(indio_dev, chan);
if (ret < 0)
return ret;
attrcount += ret;
staging:iio: Add extended IIO channel info Sometimes devices have per channel properties which either do not map nicely to the current channel info scheme (e.g. string properties) or are very device specific, so it does not make sense to add generic support for them. Currently drivers define these attributes by hand for each channel. Depending on the number of channels this can amount to quite a few lines of boilerplate code. Especially if a driver supports multiple variations of a chip with different numbers of channels. In this case it becomes necessary to have a individual attribute list per chip variation and also a individual iio_info struct. This patch introduces a new scheme for handling such per channel attributes called extended channel info attributes. A extended channel info attribute consist of a name, a flag whether it is shared and read and write callbacks. The read and write callbacks are similar to the {read,write}_raw callbacks and take a IIO device and a channel as their first parameters, but instead of pre-parsed integer values they directly get passed the raw string value, which has been written to the sysfs file. It is possible to assign a list of extended channel info attributes to a channel. For each extended channel info attribute the IIO core will create a new sysfs attribute conforming to the IIO channel naming spec for the channels type, similar as for normal info attributes. Read and write access to this sysfs attribute will be redirected to the extended channel info attributes read and write callbacks. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Acked-by: Jonathan Cameron <jic23@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-02-21 17:38:12 +00:00
if (chan->ext_info) {
unsigned int i = 0;
for (ext_info = chan->ext_info; ext_info->name; ext_info++) {
ret = __iio_add_chan_devattr(ext_info->name,
chan,
ext_info->read ?
&iio_read_channel_ext_info : NULL,
ext_info->write ?
&iio_write_channel_ext_info : NULL,
i,
ext_info->shared,
&indio_dev->dev,
NULL,
&iio_dev_opaque->channel_attr_list);
staging:iio: Add extended IIO channel info Sometimes devices have per channel properties which either do not map nicely to the current channel info scheme (e.g. string properties) or are very device specific, so it does not make sense to add generic support for them. Currently drivers define these attributes by hand for each channel. Depending on the number of channels this can amount to quite a few lines of boilerplate code. Especially if a driver supports multiple variations of a chip with different numbers of channels. In this case it becomes necessary to have a individual attribute list per chip variation and also a individual iio_info struct. This patch introduces a new scheme for handling such per channel attributes called extended channel info attributes. A extended channel info attribute consist of a name, a flag whether it is shared and read and write callbacks. The read and write callbacks are similar to the {read,write}_raw callbacks and take a IIO device and a channel as their first parameters, but instead of pre-parsed integer values they directly get passed the raw string value, which has been written to the sysfs file. It is possible to assign a list of extended channel info attributes to a channel. For each extended channel info attribute the IIO core will create a new sysfs attribute conforming to the IIO channel naming spec for the channels type, similar as for normal info attributes. Read and write access to this sysfs attribute will be redirected to the extended channel info attributes read and write callbacks. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Acked-by: Jonathan Cameron <jic23@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-02-21 17:38:12 +00:00
i++;
if (ret == -EBUSY && ext_info->shared)
continue;
if (ret)
return ret;
staging:iio: Add extended IIO channel info Sometimes devices have per channel properties which either do not map nicely to the current channel info scheme (e.g. string properties) or are very device specific, so it does not make sense to add generic support for them. Currently drivers define these attributes by hand for each channel. Depending on the number of channels this can amount to quite a few lines of boilerplate code. Especially if a driver supports multiple variations of a chip with different numbers of channels. In this case it becomes necessary to have a individual attribute list per chip variation and also a individual iio_info struct. This patch introduces a new scheme for handling such per channel attributes called extended channel info attributes. A extended channel info attribute consist of a name, a flag whether it is shared and read and write callbacks. The read and write callbacks are similar to the {read,write}_raw callbacks and take a IIO device and a channel as their first parameters, but instead of pre-parsed integer values they directly get passed the raw string value, which has been written to the sysfs file. It is possible to assign a list of extended channel info attributes to a channel. For each extended channel info attribute the IIO core will create a new sysfs attribute conforming to the IIO channel naming spec for the channels type, similar as for normal info attributes. Read and write access to this sysfs attribute will be redirected to the extended channel info attributes read and write callbacks. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Acked-by: Jonathan Cameron <jic23@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-02-21 17:38:12 +00:00
attrcount++;
}
}
return attrcount;
}
/**
* iio_free_chan_devattr_list() - Free a list of IIO device attributes
* @attr_list: List of IIO device attributes
*
* This function frees the memory allocated for each of the IIO device
* attributes in the list.
*/
void iio_free_chan_devattr_list(struct list_head *attr_list)
{
struct iio_dev_attr *p, *n;
list_for_each_entry_safe(p, n, attr_list, l) {
kfree_const(p->dev_attr.attr.name);
list_del(&p->l);
kfree(p);
}
}
static ssize_t iio_show_dev_name(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct iio_dev *indio_dev = dev_to_iio_dev(dev);
return sysfs_emit(buf, "%s\n", indio_dev->name);
}
static DEVICE_ATTR(name, S_IRUGO, iio_show_dev_name, NULL);
static ssize_t iio_show_dev_label(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct iio_dev *indio_dev = dev_to_iio_dev(dev);
return sysfs_emit(buf, "%s\n", indio_dev->label);
}
static DEVICE_ATTR(label, S_IRUGO, iio_show_dev_label, NULL);
static ssize_t iio_show_timestamp_clock(struct device *dev,
struct device_attribute *attr,
char *buf)
{
const struct iio_dev *indio_dev = dev_to_iio_dev(dev);
const clockid_t clk = iio_device_get_clock(indio_dev);
const char *name;
ssize_t sz;
switch (clk) {
case CLOCK_REALTIME:
name = "realtime\n";
sz = sizeof("realtime\n");
break;
case CLOCK_MONOTONIC:
name = "monotonic\n";
sz = sizeof("monotonic\n");
break;
case CLOCK_MONOTONIC_RAW:
name = "monotonic_raw\n";
sz = sizeof("monotonic_raw\n");
break;
case CLOCK_REALTIME_COARSE:
name = "realtime_coarse\n";
sz = sizeof("realtime_coarse\n");
break;
case CLOCK_MONOTONIC_COARSE:
name = "monotonic_coarse\n";
sz = sizeof("monotonic_coarse\n");
break;
case CLOCK_BOOTTIME:
name = "boottime\n";
sz = sizeof("boottime\n");
break;
case CLOCK_TAI:
name = "tai\n";
sz = sizeof("tai\n");
break;
default:
BUG();
}
memcpy(buf, name, sz);
return sz;
}
static ssize_t iio_store_timestamp_clock(struct device *dev,
struct device_attribute *attr,
const char *buf, size_t len)
{
clockid_t clk;
int ret;
if (sysfs_streq(buf, "realtime"))
clk = CLOCK_REALTIME;
else if (sysfs_streq(buf, "monotonic"))
clk = CLOCK_MONOTONIC;
else if (sysfs_streq(buf, "monotonic_raw"))
clk = CLOCK_MONOTONIC_RAW;
else if (sysfs_streq(buf, "realtime_coarse"))
clk = CLOCK_REALTIME_COARSE;
else if (sysfs_streq(buf, "monotonic_coarse"))
clk = CLOCK_MONOTONIC_COARSE;
else if (sysfs_streq(buf, "boottime"))
clk = CLOCK_BOOTTIME;
else if (sysfs_streq(buf, "tai"))
clk = CLOCK_TAI;
else
return -EINVAL;
ret = iio_device_set_clock(dev_to_iio_dev(dev), clk);
if (ret)
return ret;
return len;
}
iio: core: rework iio device group creation Up until now, the device groups that an IIO device had were limited to 6. Two of these groups would account for buffer attributes (the buffer/ and scan_elements/ directories). Since we want to add multiple buffers per IIO device, this number may not be enough, when adding a second buffer. So, this change reallocates the groups array whenever an IIO device group is added, via a iio_device_register_sysfs_group() helper. This also means that the groups array should be assigned to 'indio_dev.dev.groups' really late, right before {cdev_}device_add() is called to do the entire setup. And we also must take care to free this array when the sysfs resources are being cleaned up. With this change we can also move the 'groups' & 'groupcounter' fields to the iio_dev_opaque object. Up until now, this didn't make a whole lot of sense (especially since we weren't sure how multibuffer support would look like in the end). But doing it now kills one birds with one stone. An alternative, would be to add a configurable Kconfig symbol CONFIG_IIO_MAX_BUFFERS_PER_DEVICE (or something like that) and compute a static maximum of the groups we can support per IIO device. But that would probably annoy a few people since that would make the system less configurable. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20210215104043.91251-11-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2021-02-15 10:40:29 +00:00
int iio_device_register_sysfs_group(struct iio_dev *indio_dev,
const struct attribute_group *group)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
const struct attribute_group **new, **old = iio_dev_opaque->groups;
unsigned int cnt = iio_dev_opaque->groupcounter;
new = krealloc(old, sizeof(*new) * (cnt + 2), GFP_KERNEL);
if (!new)
return -ENOMEM;
new[iio_dev_opaque->groupcounter++] = group;
new[iio_dev_opaque->groupcounter] = NULL;
iio_dev_opaque->groups = new;
return 0;
}
static DEVICE_ATTR(current_timestamp_clock, S_IRUGO | S_IWUSR,
iio_show_timestamp_clock, iio_store_timestamp_clock);
static int iio_device_register_sysfs(struct iio_dev *indio_dev)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
int i, ret = 0, attrcount, attrn, attrcount_orig = 0;
struct iio_dev_attr *p;
struct attribute **attr, *clk = NULL;
/* First count elements in any existing group */
if (indio_dev->info->attrs) {
attr = indio_dev->info->attrs->attrs;
while (*attr++ != NULL)
attrcount_orig++;
}
attrcount = attrcount_orig;
/*
* New channel registration method - relies on the fact a group does
* not need to be initialized if its name is NULL.
*/
if (indio_dev->channels)
for (i = 0; i < indio_dev->num_channels; i++) {
const struct iio_chan_spec *chan =
&indio_dev->channels[i];
if (chan->type == IIO_TIMESTAMP)
clk = &dev_attr_current_timestamp_clock.attr;
ret = iio_device_add_channel_sysfs(indio_dev, chan);
if (ret < 0)
goto error_clear_attrs;
attrcount += ret;
}
if (iio_dev_opaque->event_interface)
clk = &dev_attr_current_timestamp_clock.attr;
if (indio_dev->name)
attrcount++;
if (indio_dev->label)
attrcount++;
if (clk)
attrcount++;
iio_dev_opaque->chan_attr_group.attrs =
kcalloc(attrcount + 1,
sizeof(iio_dev_opaque->chan_attr_group.attrs[0]),
GFP_KERNEL);
if (iio_dev_opaque->chan_attr_group.attrs == NULL) {
ret = -ENOMEM;
goto error_clear_attrs;
}
/* Copy across original attributes */
if (indio_dev->info->attrs) {
memcpy(iio_dev_opaque->chan_attr_group.attrs,
indio_dev->info->attrs->attrs,
sizeof(iio_dev_opaque->chan_attr_group.attrs[0])
*attrcount_orig);
iio_dev_opaque->chan_attr_group.is_visible =
indio_dev->info->attrs->is_visible;
}
attrn = attrcount_orig;
/* Add all elements from the list. */
list_for_each_entry(p, &iio_dev_opaque->channel_attr_list, l)
iio_dev_opaque->chan_attr_group.attrs[attrn++] = &p->dev_attr.attr;
if (indio_dev->name)
iio_dev_opaque->chan_attr_group.attrs[attrn++] = &dev_attr_name.attr;
if (indio_dev->label)
iio_dev_opaque->chan_attr_group.attrs[attrn++] = &dev_attr_label.attr;
if (clk)
iio_dev_opaque->chan_attr_group.attrs[attrn++] = clk;
iio: core: rework iio device group creation Up until now, the device groups that an IIO device had were limited to 6. Two of these groups would account for buffer attributes (the buffer/ and scan_elements/ directories). Since we want to add multiple buffers per IIO device, this number may not be enough, when adding a second buffer. So, this change reallocates the groups array whenever an IIO device group is added, via a iio_device_register_sysfs_group() helper. This also means that the groups array should be assigned to 'indio_dev.dev.groups' really late, right before {cdev_}device_add() is called to do the entire setup. And we also must take care to free this array when the sysfs resources are being cleaned up. With this change we can also move the 'groups' & 'groupcounter' fields to the iio_dev_opaque object. Up until now, this didn't make a whole lot of sense (especially since we weren't sure how multibuffer support would look like in the end). But doing it now kills one birds with one stone. An alternative, would be to add a configurable Kconfig symbol CONFIG_IIO_MAX_BUFFERS_PER_DEVICE (or something like that) and compute a static maximum of the groups we can support per IIO device. But that would probably annoy a few people since that would make the system less configurable. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20210215104043.91251-11-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2021-02-15 10:40:29 +00:00
ret = iio_device_register_sysfs_group(indio_dev,
&iio_dev_opaque->chan_attr_group);
if (ret)
goto error_clear_attrs;
return 0;
error_clear_attrs:
iio_free_chan_devattr_list(&iio_dev_opaque->channel_attr_list);
return ret;
}
static void iio_device_unregister_sysfs(struct iio_dev *indio_dev)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
iio_free_chan_devattr_list(&iio_dev_opaque->channel_attr_list);
kfree(iio_dev_opaque->chan_attr_group.attrs);
iio_dev_opaque->chan_attr_group.attrs = NULL;
iio: core: rework iio device group creation Up until now, the device groups that an IIO device had were limited to 6. Two of these groups would account for buffer attributes (the buffer/ and scan_elements/ directories). Since we want to add multiple buffers per IIO device, this number may not be enough, when adding a second buffer. So, this change reallocates the groups array whenever an IIO device group is added, via a iio_device_register_sysfs_group() helper. This also means that the groups array should be assigned to 'indio_dev.dev.groups' really late, right before {cdev_}device_add() is called to do the entire setup. And we also must take care to free this array when the sysfs resources are being cleaned up. With this change we can also move the 'groups' & 'groupcounter' fields to the iio_dev_opaque object. Up until now, this didn't make a whole lot of sense (especially since we weren't sure how multibuffer support would look like in the end). But doing it now kills one birds with one stone. An alternative, would be to add a configurable Kconfig symbol CONFIG_IIO_MAX_BUFFERS_PER_DEVICE (or something like that) and compute a static maximum of the groups we can support per IIO device. But that would probably annoy a few people since that would make the system less configurable. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20210215104043.91251-11-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2021-02-15 10:40:29 +00:00
kfree(iio_dev_opaque->groups);
}
static void iio_dev_release(struct device *device)
{
struct iio_dev *indio_dev = dev_to_iio_dev(device);
iio: core: wrap IIO device into an iio_dev_opaque object There are plenty of bad designs we want to discourage or not have to review manually usually about accessing private (marked as [INTERN]) fields of 'struct iio_dev'. Sometimes users copy drivers that are not always the best examples. A better idea is to hide those fields into the framework. For 'struct iio_dev' this is a 'struct iio_dev_opaque' which wraps a public 'struct iio_dev' object. In the next series, some fields will be moved to this new struct, each with it's own rework. This rework will not be complete-able for a while, as many fields need some drivers to be reworked in order to finalize them (e.g. 'indio_dev->mlock'). But some fields can already be moved, and in time, all of them may get there (in the 'struct iio_dev_opaque' object). Since a lot of drivers also call 'iio_priv()', in order to preserve fast-paths (where this matters), the public iio_dev object will have a 'priv' field that will have the pointer to the private information already computed. The reference returned by this field should be guaranteed to be cacheline aligned. The opaque parts will be moved into the 'include/linux/iio/iio-opaque.h' header. Should the hidden information be required for some debugging or some special needs, it can be made available via this header. Otherwise, only the IIO core files should include this file. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-06-30 04:57:03 +00:00
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
if (indio_dev->modes & INDIO_ALL_TRIGGERED_MODES)
iio_device_unregister_trigger_consumer(indio_dev);
iio_device_unregister_eventset(indio_dev);
iio_device_unregister_sysfs(indio_dev);
iio: Fix potential use after free There is no guarantee that the last reference to the iio device has already been dropped when iio_device_free is called. This means that we can up calling iio_dev_release after iio_device_free which will lead to a use after free. As the general rule the struct containing the device should always be freed in the release callback. This is what this patch does, it moves freeing the iio device struct as well as releasing the idr reference to the release callback. To ensure that the device is not freed before calling iio_device_free the device_unregister call in iio_device_unregister is broken apart. iio_device_unregister will now only call device_del to remove the device from the system and iio_device_free will call put_device to drop the reference we obtained in iio_devce_alloc. We also have to take care that calling iio_device_free without having called iio_device_register still works (i.e. this can happen if something failed during device initialization). For this to work properly two minor changes were necessary: channel_attr_list needs to be initialized in iio_device_alloc and we have to check whether the chrdev has been registered before releasing it in iio_device_release. This change also brings iio_device_unregister and iio_device_free more in sync with iio_device_register and iio_device_alloc which call device_add and device_initialize respectively. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Acked-by: Jonathan Cameron <jic23@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-06-04 08:41:42 +00:00
iio_device_detach_buffers(indio_dev);
ida_simple_remove(&iio_ida, iio_dev_opaque->id);
iio: core: wrap IIO device into an iio_dev_opaque object There are plenty of bad designs we want to discourage or not have to review manually usually about accessing private (marked as [INTERN]) fields of 'struct iio_dev'. Sometimes users copy drivers that are not always the best examples. A better idea is to hide those fields into the framework. For 'struct iio_dev' this is a 'struct iio_dev_opaque' which wraps a public 'struct iio_dev' object. In the next series, some fields will be moved to this new struct, each with it's own rework. This rework will not be complete-able for a while, as many fields need some drivers to be reworked in order to finalize them (e.g. 'indio_dev->mlock'). But some fields can already be moved, and in time, all of them may get there (in the 'struct iio_dev_opaque' object). Since a lot of drivers also call 'iio_priv()', in order to preserve fast-paths (where this matters), the public iio_dev object will have a 'priv' field that will have the pointer to the private information already computed. The reference returned by this field should be guaranteed to be cacheline aligned. The opaque parts will be moved into the 'include/linux/iio/iio-opaque.h' header. Should the hidden information be required for some debugging or some special needs, it can be made available via this header. Otherwise, only the IIO core files should include this file. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-06-30 04:57:03 +00:00
kfree(iio_dev_opaque);
}
struct device_type iio_device_type = {
.name = "iio_device",
.release = iio_dev_release,
};
/**
* iio_device_alloc() - allocate an iio_dev from a driver
* @parent: Parent device.
* @sizeof_priv: Space to allocate for private structure.
**/
struct iio_dev *iio_device_alloc(struct device *parent, int sizeof_priv)
{
iio: core: wrap IIO device into an iio_dev_opaque object There are plenty of bad designs we want to discourage or not have to review manually usually about accessing private (marked as [INTERN]) fields of 'struct iio_dev'. Sometimes users copy drivers that are not always the best examples. A better idea is to hide those fields into the framework. For 'struct iio_dev' this is a 'struct iio_dev_opaque' which wraps a public 'struct iio_dev' object. In the next series, some fields will be moved to this new struct, each with it's own rework. This rework will not be complete-able for a while, as many fields need some drivers to be reworked in order to finalize them (e.g. 'indio_dev->mlock'). But some fields can already be moved, and in time, all of them may get there (in the 'struct iio_dev_opaque' object). Since a lot of drivers also call 'iio_priv()', in order to preserve fast-paths (where this matters), the public iio_dev object will have a 'priv' field that will have the pointer to the private information already computed. The reference returned by this field should be guaranteed to be cacheline aligned. The opaque parts will be moved into the 'include/linux/iio/iio-opaque.h' header. Should the hidden information be required for some debugging or some special needs, it can be made available via this header. Otherwise, only the IIO core files should include this file. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-06-30 04:57:03 +00:00
struct iio_dev_opaque *iio_dev_opaque;
struct iio_dev *indio_dev;
size_t alloc_size;
iio: core: wrap IIO device into an iio_dev_opaque object There are plenty of bad designs we want to discourage or not have to review manually usually about accessing private (marked as [INTERN]) fields of 'struct iio_dev'. Sometimes users copy drivers that are not always the best examples. A better idea is to hide those fields into the framework. For 'struct iio_dev' this is a 'struct iio_dev_opaque' which wraps a public 'struct iio_dev' object. In the next series, some fields will be moved to this new struct, each with it's own rework. This rework will not be complete-able for a while, as many fields need some drivers to be reworked in order to finalize them (e.g. 'indio_dev->mlock'). But some fields can already be moved, and in time, all of them may get there (in the 'struct iio_dev_opaque' object). Since a lot of drivers also call 'iio_priv()', in order to preserve fast-paths (where this matters), the public iio_dev object will have a 'priv' field that will have the pointer to the private information already computed. The reference returned by this field should be guaranteed to be cacheline aligned. The opaque parts will be moved into the 'include/linux/iio/iio-opaque.h' header. Should the hidden information be required for some debugging or some special needs, it can be made available via this header. Otherwise, only the IIO core files should include this file. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-06-30 04:57:03 +00:00
alloc_size = sizeof(struct iio_dev_opaque);
if (sizeof_priv) {
alloc_size = ALIGN(alloc_size, IIO_ALIGN);
alloc_size += sizeof_priv;
}
iio: core: wrap IIO device into an iio_dev_opaque object There are plenty of bad designs we want to discourage or not have to review manually usually about accessing private (marked as [INTERN]) fields of 'struct iio_dev'. Sometimes users copy drivers that are not always the best examples. A better idea is to hide those fields into the framework. For 'struct iio_dev' this is a 'struct iio_dev_opaque' which wraps a public 'struct iio_dev' object. In the next series, some fields will be moved to this new struct, each with it's own rework. This rework will not be complete-able for a while, as many fields need some drivers to be reworked in order to finalize them (e.g. 'indio_dev->mlock'). But some fields can already be moved, and in time, all of them may get there (in the 'struct iio_dev_opaque' object). Since a lot of drivers also call 'iio_priv()', in order to preserve fast-paths (where this matters), the public iio_dev object will have a 'priv' field that will have the pointer to the private information already computed. The reference returned by this field should be guaranteed to be cacheline aligned. The opaque parts will be moved into the 'include/linux/iio/iio-opaque.h' header. Should the hidden information be required for some debugging or some special needs, it can be made available via this header. Otherwise, only the IIO core files should include this file. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-06-30 04:57:03 +00:00
iio_dev_opaque = kzalloc(alloc_size, GFP_KERNEL);
if (!iio_dev_opaque)
return NULL;
indio_dev = &iio_dev_opaque->indio_dev;
indio_dev->priv = (char *)iio_dev_opaque +
iio: core: wrap IIO device into an iio_dev_opaque object There are plenty of bad designs we want to discourage or not have to review manually usually about accessing private (marked as [INTERN]) fields of 'struct iio_dev'. Sometimes users copy drivers that are not always the best examples. A better idea is to hide those fields into the framework. For 'struct iio_dev' this is a 'struct iio_dev_opaque' which wraps a public 'struct iio_dev' object. In the next series, some fields will be moved to this new struct, each with it's own rework. This rework will not be complete-able for a while, as many fields need some drivers to be reworked in order to finalize them (e.g. 'indio_dev->mlock'). But some fields can already be moved, and in time, all of them may get there (in the 'struct iio_dev_opaque' object). Since a lot of drivers also call 'iio_priv()', in order to preserve fast-paths (where this matters), the public iio_dev object will have a 'priv' field that will have the pointer to the private information already computed. The reference returned by this field should be guaranteed to be cacheline aligned. The opaque parts will be moved into the 'include/linux/iio/iio-opaque.h' header. Should the hidden information be required for some debugging or some special needs, it can be made available via this header. Otherwise, only the IIO core files should include this file. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-06-30 04:57:03 +00:00
ALIGN(sizeof(struct iio_dev_opaque), IIO_ALIGN);
indio_dev->dev.parent = parent;
indio_dev->dev.type = &iio_device_type;
indio_dev->dev.bus = &iio_bus_type;
device_initialize(&indio_dev->dev);
iio_device_set_drvdata(indio_dev, (void *)indio_dev);
mutex_init(&indio_dev->mlock);
mutex_init(&iio_dev_opaque->info_exist_lock);
INIT_LIST_HEAD(&iio_dev_opaque->channel_attr_list);
iio_dev_opaque->id = ida_simple_get(&iio_ida, 0, 0, GFP_KERNEL);
if (iio_dev_opaque->id < 0) {
/* cannot use a dev_err as the name isn't available */
pr_err("failed to get device id\n");
iio: core: wrap IIO device into an iio_dev_opaque object There are plenty of bad designs we want to discourage or not have to review manually usually about accessing private (marked as [INTERN]) fields of 'struct iio_dev'. Sometimes users copy drivers that are not always the best examples. A better idea is to hide those fields into the framework. For 'struct iio_dev' this is a 'struct iio_dev_opaque' which wraps a public 'struct iio_dev' object. In the next series, some fields will be moved to this new struct, each with it's own rework. This rework will not be complete-able for a while, as many fields need some drivers to be reworked in order to finalize them (e.g. 'indio_dev->mlock'). But some fields can already be moved, and in time, all of them may get there (in the 'struct iio_dev_opaque' object). Since a lot of drivers also call 'iio_priv()', in order to preserve fast-paths (where this matters), the public iio_dev object will have a 'priv' field that will have the pointer to the private information already computed. The reference returned by this field should be guaranteed to be cacheline aligned. The opaque parts will be moved into the 'include/linux/iio/iio-opaque.h' header. Should the hidden information be required for some debugging or some special needs, it can be made available via this header. Otherwise, only the IIO core files should include this file. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-06-30 04:57:03 +00:00
kfree(iio_dev_opaque);
return NULL;
}
dev_set_name(&indio_dev->dev, "iio:device%d", iio_dev_opaque->id);
INIT_LIST_HEAD(&iio_dev_opaque->buffer_list);
iio: core: centralize ioctl() calls to the main chardev The aim of this is to improve a bit the organization of ioctl() calls in IIO core. Currently the chardev is split across IIO core sub-modules/files. The main chardev has to be able to handle ioctl() calls, and if we need to add buffer ioctl() calls, this would complicate things. The 'industrialio-core.c' file will provide a 'iio_device_ioctl()' which will iterate over a list of ioctls registered with the IIO device. These can be event ioctl() or buffer ioctl() calls, or something else. Each ioctl() handler will have to return a IIO_IOCTL_UNHANDLED code (which is positive 1), if the ioctl() did not handle the call in any. This eliminates any potential ambiguities about negative error codes, which should fail the call altogether. If any ioctl() returns 0, it was considered that it was serviced successfully and the loop will exit. This change also moves the handling of the IIO_GET_EVENT_FD_IOCTL command inside 'industrialio-event.c', where this is better suited. This patch is a combination of 2 other patches from an older series: Patch 1: iio: core: add simple centralized mechanism for ioctl() handlers Link: https://lore.kernel.org/linux-iio/20200427131100.50845-6-alexandru.ardelean@analog.com/ Patch 2: iio: core: use new common ioctl() mechanism Link: https://lore.kernel.org/linux-iio/20200427131100.50845-7-alexandru.ardelean@analog.com/ Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20200924084155.99406-1-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-09-24 08:41:55 +00:00
INIT_LIST_HEAD(&iio_dev_opaque->ioctl_handlers);
return indio_dev;
}
EXPORT_SYMBOL(iio_device_alloc);
/**
* iio_device_free() - free an iio_dev from a driver
* @dev: the iio_dev associated with the device
**/
void iio_device_free(struct iio_dev *dev)
{
iio: Fix potential use after free There is no guarantee that the last reference to the iio device has already been dropped when iio_device_free is called. This means that we can up calling iio_dev_release after iio_device_free which will lead to a use after free. As the general rule the struct containing the device should always be freed in the release callback. This is what this patch does, it moves freeing the iio device struct as well as releasing the idr reference to the release callback. To ensure that the device is not freed before calling iio_device_free the device_unregister call in iio_device_unregister is broken apart. iio_device_unregister will now only call device_del to remove the device from the system and iio_device_free will call put_device to drop the reference we obtained in iio_devce_alloc. We also have to take care that calling iio_device_free without having called iio_device_register still works (i.e. this can happen if something failed during device initialization). For this to work properly two minor changes were necessary: channel_attr_list needs to be initialized in iio_device_alloc and we have to check whether the chrdev has been registered before releasing it in iio_device_release. This change also brings iio_device_unregister and iio_device_free more in sync with iio_device_register and iio_device_alloc which call device_add and device_initialize respectively. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Acked-by: Jonathan Cameron <jic23@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-06-04 08:41:42 +00:00
if (dev)
put_device(&dev->dev);
}
EXPORT_SYMBOL(iio_device_free);
static void devm_iio_device_release(void *iio_dev)
{
iio_device_free(iio_dev);
}
/**
* devm_iio_device_alloc - Resource-managed iio_device_alloc()
* @parent: Device to allocate iio_dev for, and parent for this IIO device
* @sizeof_priv: Space to allocate for private structure.
*
* Managed iio_device_alloc. iio_dev allocated with this function is
* automatically freed on driver detach.
*
* RETURNS:
* Pointer to allocated iio_dev on success, NULL on failure.
*/
struct iio_dev *devm_iio_device_alloc(struct device *parent, int sizeof_priv)
{
struct iio_dev *iio_dev;
int ret;
iio_dev = iio_device_alloc(parent, sizeof_priv);
if (!iio_dev)
return NULL;
ret = devm_add_action_or_reset(parent, devm_iio_device_release,
iio_dev);
if (ret)
return NULL;
return iio_dev;
}
EXPORT_SYMBOL_GPL(devm_iio_device_alloc);
/**
* iio_chrdev_open() - chrdev file open for buffer access and ioctls
* @inode: Inode structure for identifying the device in the file system
* @filp: File structure for iio device used to keep and later access
* private data
*
* Return: 0 on success or -EBUSY if the device is already opened
**/
static int iio_chrdev_open(struct inode *inode, struct file *filp)
{
struct iio_dev_opaque *iio_dev_opaque =
container_of(inode->i_cdev, struct iio_dev_opaque, chrdev);
struct iio_dev *indio_dev = &iio_dev_opaque->indio_dev;
struct iio_dev_buffer_pair *ib;
if (test_and_set_bit(IIO_BUSY_BIT_POS, &iio_dev_opaque->flags))
return -EBUSY;
iio_device_get(indio_dev);
ib = kmalloc(sizeof(*ib), GFP_KERNEL);
if (!ib) {
iio_device_put(indio_dev);
clear_bit(IIO_BUSY_BIT_POS, &iio_dev_opaque->flags);
return -ENOMEM;
}
ib->indio_dev = indio_dev;
ib->buffer = indio_dev->buffer;
filp->private_data = ib;
return 0;
}
/**
* iio_chrdev_release() - chrdev file close buffer access and ioctls
* @inode: Inode structure pointer for the char device
* @filp: File structure pointer for the char device
*
* Return: 0 for successful release
*/
static int iio_chrdev_release(struct inode *inode, struct file *filp)
{
struct iio_dev_buffer_pair *ib = filp->private_data;
struct iio_dev_opaque *iio_dev_opaque =
container_of(inode->i_cdev, struct iio_dev_opaque, chrdev);
struct iio_dev *indio_dev = &iio_dev_opaque->indio_dev;
kfree(ib);
clear_bit(IIO_BUSY_BIT_POS, &iio_dev_opaque->flags);
iio_device_put(indio_dev);
return 0;
}
iio: core: centralize ioctl() calls to the main chardev The aim of this is to improve a bit the organization of ioctl() calls in IIO core. Currently the chardev is split across IIO core sub-modules/files. The main chardev has to be able to handle ioctl() calls, and if we need to add buffer ioctl() calls, this would complicate things. The 'industrialio-core.c' file will provide a 'iio_device_ioctl()' which will iterate over a list of ioctls registered with the IIO device. These can be event ioctl() or buffer ioctl() calls, or something else. Each ioctl() handler will have to return a IIO_IOCTL_UNHANDLED code (which is positive 1), if the ioctl() did not handle the call in any. This eliminates any potential ambiguities about negative error codes, which should fail the call altogether. If any ioctl() returns 0, it was considered that it was serviced successfully and the loop will exit. This change also moves the handling of the IIO_GET_EVENT_FD_IOCTL command inside 'industrialio-event.c', where this is better suited. This patch is a combination of 2 other patches from an older series: Patch 1: iio: core: add simple centralized mechanism for ioctl() handlers Link: https://lore.kernel.org/linux-iio/20200427131100.50845-6-alexandru.ardelean@analog.com/ Patch 2: iio: core: use new common ioctl() mechanism Link: https://lore.kernel.org/linux-iio/20200427131100.50845-7-alexandru.ardelean@analog.com/ Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20200924084155.99406-1-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-09-24 08:41:55 +00:00
void iio_device_ioctl_handler_register(struct iio_dev *indio_dev,
struct iio_ioctl_handler *h)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
list_add_tail(&h->entry, &iio_dev_opaque->ioctl_handlers);
}
void iio_device_ioctl_handler_unregister(struct iio_ioctl_handler *h)
{
list_del(&h->entry);
}
static long iio_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
{
struct iio_dev_buffer_pair *ib = filp->private_data;
struct iio_dev *indio_dev = ib->indio_dev;
iio: core: centralize ioctl() calls to the main chardev The aim of this is to improve a bit the organization of ioctl() calls in IIO core. Currently the chardev is split across IIO core sub-modules/files. The main chardev has to be able to handle ioctl() calls, and if we need to add buffer ioctl() calls, this would complicate things. The 'industrialio-core.c' file will provide a 'iio_device_ioctl()' which will iterate over a list of ioctls registered with the IIO device. These can be event ioctl() or buffer ioctl() calls, or something else. Each ioctl() handler will have to return a IIO_IOCTL_UNHANDLED code (which is positive 1), if the ioctl() did not handle the call in any. This eliminates any potential ambiguities about negative error codes, which should fail the call altogether. If any ioctl() returns 0, it was considered that it was serviced successfully and the loop will exit. This change also moves the handling of the IIO_GET_EVENT_FD_IOCTL command inside 'industrialio-event.c', where this is better suited. This patch is a combination of 2 other patches from an older series: Patch 1: iio: core: add simple centralized mechanism for ioctl() handlers Link: https://lore.kernel.org/linux-iio/20200427131100.50845-6-alexandru.ardelean@analog.com/ Patch 2: iio: core: use new common ioctl() mechanism Link: https://lore.kernel.org/linux-iio/20200427131100.50845-7-alexandru.ardelean@analog.com/ Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20200924084155.99406-1-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-09-24 08:41:55 +00:00
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
struct iio_ioctl_handler *h;
int ret = -ENODEV;
mutex_lock(&iio_dev_opaque->info_exist_lock);
iio: core: centralize ioctl() calls to the main chardev The aim of this is to improve a bit the organization of ioctl() calls in IIO core. Currently the chardev is split across IIO core sub-modules/files. The main chardev has to be able to handle ioctl() calls, and if we need to add buffer ioctl() calls, this would complicate things. The 'industrialio-core.c' file will provide a 'iio_device_ioctl()' which will iterate over a list of ioctls registered with the IIO device. These can be event ioctl() or buffer ioctl() calls, or something else. Each ioctl() handler will have to return a IIO_IOCTL_UNHANDLED code (which is positive 1), if the ioctl() did not handle the call in any. This eliminates any potential ambiguities about negative error codes, which should fail the call altogether. If any ioctl() returns 0, it was considered that it was serviced successfully and the loop will exit. This change also moves the handling of the IIO_GET_EVENT_FD_IOCTL command inside 'industrialio-event.c', where this is better suited. This patch is a combination of 2 other patches from an older series: Patch 1: iio: core: add simple centralized mechanism for ioctl() handlers Link: https://lore.kernel.org/linux-iio/20200427131100.50845-6-alexandru.ardelean@analog.com/ Patch 2: iio: core: use new common ioctl() mechanism Link: https://lore.kernel.org/linux-iio/20200427131100.50845-7-alexandru.ardelean@analog.com/ Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20200924084155.99406-1-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-09-24 08:41:55 +00:00
/**
* The NULL check here is required to prevent crashing when a device
* is being removed while userspace would still have open file handles
* to try to access this device.
*/
if (!indio_dev->info)
iio: core: centralize ioctl() calls to the main chardev The aim of this is to improve a bit the organization of ioctl() calls in IIO core. Currently the chardev is split across IIO core sub-modules/files. The main chardev has to be able to handle ioctl() calls, and if we need to add buffer ioctl() calls, this would complicate things. The 'industrialio-core.c' file will provide a 'iio_device_ioctl()' which will iterate over a list of ioctls registered with the IIO device. These can be event ioctl() or buffer ioctl() calls, or something else. Each ioctl() handler will have to return a IIO_IOCTL_UNHANDLED code (which is positive 1), if the ioctl() did not handle the call in any. This eliminates any potential ambiguities about negative error codes, which should fail the call altogether. If any ioctl() returns 0, it was considered that it was serviced successfully and the loop will exit. This change also moves the handling of the IIO_GET_EVENT_FD_IOCTL command inside 'industrialio-event.c', where this is better suited. This patch is a combination of 2 other patches from an older series: Patch 1: iio: core: add simple centralized mechanism for ioctl() handlers Link: https://lore.kernel.org/linux-iio/20200427131100.50845-6-alexandru.ardelean@analog.com/ Patch 2: iio: core: use new common ioctl() mechanism Link: https://lore.kernel.org/linux-iio/20200427131100.50845-7-alexandru.ardelean@analog.com/ Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20200924084155.99406-1-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-09-24 08:41:55 +00:00
goto out_unlock;
list_for_each_entry(h, &iio_dev_opaque->ioctl_handlers, entry) {
ret = h->ioctl(indio_dev, filp, cmd, arg);
if (ret != IIO_IOCTL_UNHANDLED)
break;
}
iio: core: centralize ioctl() calls to the main chardev The aim of this is to improve a bit the organization of ioctl() calls in IIO core. Currently the chardev is split across IIO core sub-modules/files. The main chardev has to be able to handle ioctl() calls, and if we need to add buffer ioctl() calls, this would complicate things. The 'industrialio-core.c' file will provide a 'iio_device_ioctl()' which will iterate over a list of ioctls registered with the IIO device. These can be event ioctl() or buffer ioctl() calls, or something else. Each ioctl() handler will have to return a IIO_IOCTL_UNHANDLED code (which is positive 1), if the ioctl() did not handle the call in any. This eliminates any potential ambiguities about negative error codes, which should fail the call altogether. If any ioctl() returns 0, it was considered that it was serviced successfully and the loop will exit. This change also moves the handling of the IIO_GET_EVENT_FD_IOCTL command inside 'industrialio-event.c', where this is better suited. This patch is a combination of 2 other patches from an older series: Patch 1: iio: core: add simple centralized mechanism for ioctl() handlers Link: https://lore.kernel.org/linux-iio/20200427131100.50845-6-alexandru.ardelean@analog.com/ Patch 2: iio: core: use new common ioctl() mechanism Link: https://lore.kernel.org/linux-iio/20200427131100.50845-7-alexandru.ardelean@analog.com/ Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20200924084155.99406-1-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-09-24 08:41:55 +00:00
if (ret == IIO_IOCTL_UNHANDLED)
ret = -ENODEV;
iio: core: centralize ioctl() calls to the main chardev The aim of this is to improve a bit the organization of ioctl() calls in IIO core. Currently the chardev is split across IIO core sub-modules/files. The main chardev has to be able to handle ioctl() calls, and if we need to add buffer ioctl() calls, this would complicate things. The 'industrialio-core.c' file will provide a 'iio_device_ioctl()' which will iterate over a list of ioctls registered with the IIO device. These can be event ioctl() or buffer ioctl() calls, or something else. Each ioctl() handler will have to return a IIO_IOCTL_UNHANDLED code (which is positive 1), if the ioctl() did not handle the call in any. This eliminates any potential ambiguities about negative error codes, which should fail the call altogether. If any ioctl() returns 0, it was considered that it was serviced successfully and the loop will exit. This change also moves the handling of the IIO_GET_EVENT_FD_IOCTL command inside 'industrialio-event.c', where this is better suited. This patch is a combination of 2 other patches from an older series: Patch 1: iio: core: add simple centralized mechanism for ioctl() handlers Link: https://lore.kernel.org/linux-iio/20200427131100.50845-6-alexandru.ardelean@analog.com/ Patch 2: iio: core: use new common ioctl() mechanism Link: https://lore.kernel.org/linux-iio/20200427131100.50845-7-alexandru.ardelean@analog.com/ Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20200924084155.99406-1-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-09-24 08:41:55 +00:00
out_unlock:
mutex_unlock(&iio_dev_opaque->info_exist_lock);
iio: core: centralize ioctl() calls to the main chardev The aim of this is to improve a bit the organization of ioctl() calls in IIO core. Currently the chardev is split across IIO core sub-modules/files. The main chardev has to be able to handle ioctl() calls, and if we need to add buffer ioctl() calls, this would complicate things. The 'industrialio-core.c' file will provide a 'iio_device_ioctl()' which will iterate over a list of ioctls registered with the IIO device. These can be event ioctl() or buffer ioctl() calls, or something else. Each ioctl() handler will have to return a IIO_IOCTL_UNHANDLED code (which is positive 1), if the ioctl() did not handle the call in any. This eliminates any potential ambiguities about negative error codes, which should fail the call altogether. If any ioctl() returns 0, it was considered that it was serviced successfully and the loop will exit. This change also moves the handling of the IIO_GET_EVENT_FD_IOCTL command inside 'industrialio-event.c', where this is better suited. This patch is a combination of 2 other patches from an older series: Patch 1: iio: core: add simple centralized mechanism for ioctl() handlers Link: https://lore.kernel.org/linux-iio/20200427131100.50845-6-alexandru.ardelean@analog.com/ Patch 2: iio: core: use new common ioctl() mechanism Link: https://lore.kernel.org/linux-iio/20200427131100.50845-7-alexandru.ardelean@analog.com/ Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20200924084155.99406-1-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2020-09-24 08:41:55 +00:00
return ret;
}
static const struct file_operations iio_buffer_fileops = {
.owner = THIS_MODULE,
.llseek = noop_llseek,
.read = iio_buffer_read_outer_addr,
.poll = iio_buffer_poll_addr,
.unlocked_ioctl = iio_ioctl,
.compat_ioctl = compat_ptr_ioctl,
.open = iio_chrdev_open,
.release = iio_chrdev_release,
};
iio: core: register chardev only if needed We only need a chardev if we need to support buffers and/or events. With this change, a chardev will be created only if an IIO buffer is attached OR an event_interface is configured. Otherwise, no chardev will be created, and the IIO device will get registered with the 'device_add()' call. Quite a lot of IIO devices don't really need a chardev, so this is a minor improvement to the IIO core, as the IIO device will take up (slightly) fewer resources. In order to not create a chardev, we mostly just need to not initialize the indio_dev->dev.devt field. If that is un-initialized, cdev_device_add() behaves like device_add(). This change has a small chance of breaking some userspace ABI, because it removes un-needed chardevs. While these chardevs (that are being removed) have always been unusable, it is likely that some scripts may check their existence (for whatever logic). And we also hope that before opening these chardevs, userspace would have already checked for some pre-conditions to make sure that opening these chardevs makes sense. For the most part, there is also the hope that it would be easier to change userspace code than revert this. But in the case that reverting this is required, it should be easy enough to do it. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20210215104043.91251-9-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2021-02-15 10:40:27 +00:00
static const struct file_operations iio_event_fileops = {
.owner = THIS_MODULE,
.llseek = noop_llseek,
.unlocked_ioctl = iio_ioctl,
.compat_ioctl = compat_ptr_ioctl,
.open = iio_chrdev_open,
.release = iio_chrdev_release,
};
static int iio_check_unique_scan_index(struct iio_dev *indio_dev)
{
int i, j;
const struct iio_chan_spec *channels = indio_dev->channels;
if (!(indio_dev->modes & INDIO_ALL_BUFFER_MODES))
return 0;
for (i = 0; i < indio_dev->num_channels - 1; i++) {
if (channels[i].scan_index < 0)
continue;
for (j = i + 1; j < indio_dev->num_channels; j++)
if (channels[i].scan_index == channels[j].scan_index) {
dev_err(&indio_dev->dev,
"Duplicate scan index %d\n",
channels[i].scan_index);
return -EINVAL;
}
}
return 0;
}
static const struct iio_buffer_setup_ops noop_ring_setup_ops;
int __iio_device_register(struct iio_dev *indio_dev, struct module *this_mod)
{
iio: core: register chardev only if needed We only need a chardev if we need to support buffers and/or events. With this change, a chardev will be created only if an IIO buffer is attached OR an event_interface is configured. Otherwise, no chardev will be created, and the IIO device will get registered with the 'device_add()' call. Quite a lot of IIO devices don't really need a chardev, so this is a minor improvement to the IIO core, as the IIO device will take up (slightly) fewer resources. In order to not create a chardev, we mostly just need to not initialize the indio_dev->dev.devt field. If that is un-initialized, cdev_device_add() behaves like device_add(). This change has a small chance of breaking some userspace ABI, because it removes un-needed chardevs. While these chardevs (that are being removed) have always been unusable, it is likely that some scripts may check their existence (for whatever logic). And we also hope that before opening these chardevs, userspace would have already checked for some pre-conditions to make sure that opening these chardevs makes sense. For the most part, there is also the hope that it would be easier to change userspace code than revert this. But in the case that reverting this is required, it should be easy enough to do it. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20210215104043.91251-9-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2021-02-15 10:40:27 +00:00
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
const char *label;
int ret;
if (!indio_dev->info)
return -EINVAL;
iio_dev_opaque->driver_module = this_mod;
/* If the calling driver did not initialize of_node, do it here */
if (!indio_dev->dev.of_node && indio_dev->dev.parent)
indio_dev->dev.of_node = indio_dev->dev.parent->of_node;
label = of_get_property(indio_dev->dev.of_node, "label", NULL);
if (label)
indio_dev->label = label;
ret = iio_check_unique_scan_index(indio_dev);
if (ret < 0)
return ret;
iio_device_register_debugfs(indio_dev);
ret = iio_buffers_alloc_sysfs_and_mask(indio_dev);
if (ret) {
dev_err(indio_dev->dev.parent,
"Failed to create buffer sysfs interfaces\n");
goto error_unreg_debugfs;
}
ret = iio_device_register_sysfs(indio_dev);
if (ret) {
dev_err(indio_dev->dev.parent,
"Failed to register sysfs interfaces\n");
goto error_buffer_free_sysfs;
}
ret = iio_device_register_eventset(indio_dev);
if (ret) {
dev_err(indio_dev->dev.parent,
"Failed to register event set\n");
goto error_free_sysfs;
}
if (indio_dev->modes & INDIO_ALL_TRIGGERED_MODES)
iio_device_register_trigger_consumer(indio_dev);
if ((indio_dev->modes & INDIO_ALL_BUFFER_MODES) &&
indio_dev->setup_ops == NULL)
indio_dev->setup_ops = &noop_ring_setup_ops;
if (iio_dev_opaque->attached_buffers_cnt)
cdev_init(&iio_dev_opaque->chrdev, &iio_buffer_fileops);
iio: core: register chardev only if needed We only need a chardev if we need to support buffers and/or events. With this change, a chardev will be created only if an IIO buffer is attached OR an event_interface is configured. Otherwise, no chardev will be created, and the IIO device will get registered with the 'device_add()' call. Quite a lot of IIO devices don't really need a chardev, so this is a minor improvement to the IIO core, as the IIO device will take up (slightly) fewer resources. In order to not create a chardev, we mostly just need to not initialize the indio_dev->dev.devt field. If that is un-initialized, cdev_device_add() behaves like device_add(). This change has a small chance of breaking some userspace ABI, because it removes un-needed chardevs. While these chardevs (that are being removed) have always been unusable, it is likely that some scripts may check their existence (for whatever logic). And we also hope that before opening these chardevs, userspace would have already checked for some pre-conditions to make sure that opening these chardevs makes sense. For the most part, there is also the hope that it would be easier to change userspace code than revert this. But in the case that reverting this is required, it should be easy enough to do it. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20210215104043.91251-9-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2021-02-15 10:40:27 +00:00
else if (iio_dev_opaque->event_interface)
cdev_init(&iio_dev_opaque->chrdev, &iio_event_fileops);
if (iio_dev_opaque->attached_buffers_cnt || iio_dev_opaque->event_interface) {
indio_dev->dev.devt = MKDEV(MAJOR(iio_devt), iio_dev_opaque->id);
iio_dev_opaque->chrdev.owner = this_mod;
iio: core: register chardev only if needed We only need a chardev if we need to support buffers and/or events. With this change, a chardev will be created only if an IIO buffer is attached OR an event_interface is configured. Otherwise, no chardev will be created, and the IIO device will get registered with the 'device_add()' call. Quite a lot of IIO devices don't really need a chardev, so this is a minor improvement to the IIO core, as the IIO device will take up (slightly) fewer resources. In order to not create a chardev, we mostly just need to not initialize the indio_dev->dev.devt field. If that is un-initialized, cdev_device_add() behaves like device_add(). This change has a small chance of breaking some userspace ABI, because it removes un-needed chardevs. While these chardevs (that are being removed) have always been unusable, it is likely that some scripts may check their existence (for whatever logic). And we also hope that before opening these chardevs, userspace would have already checked for some pre-conditions to make sure that opening these chardevs makes sense. For the most part, there is also the hope that it would be easier to change userspace code than revert this. But in the case that reverting this is required, it should be easy enough to do it. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20210215104043.91251-9-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2021-02-15 10:40:27 +00:00
}
iio: core: rework iio device group creation Up until now, the device groups that an IIO device had were limited to 6. Two of these groups would account for buffer attributes (the buffer/ and scan_elements/ directories). Since we want to add multiple buffers per IIO device, this number may not be enough, when adding a second buffer. So, this change reallocates the groups array whenever an IIO device group is added, via a iio_device_register_sysfs_group() helper. This also means that the groups array should be assigned to 'indio_dev.dev.groups' really late, right before {cdev_}device_add() is called to do the entire setup. And we also must take care to free this array when the sysfs resources are being cleaned up. With this change we can also move the 'groups' & 'groupcounter' fields to the iio_dev_opaque object. Up until now, this didn't make a whole lot of sense (especially since we weren't sure how multibuffer support would look like in the end). But doing it now kills one birds with one stone. An alternative, would be to add a configurable Kconfig symbol CONFIG_IIO_MAX_BUFFERS_PER_DEVICE (or something like that) and compute a static maximum of the groups we can support per IIO device. But that would probably annoy a few people since that would make the system less configurable. Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Link: https://lore.kernel.org/r/20210215104043.91251-11-alexandru.ardelean@analog.com Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2021-02-15 10:40:29 +00:00
/* assign device groups now; they should be all registered now */
indio_dev->dev.groups = iio_dev_opaque->groups;
ret = cdev_device_add(&iio_dev_opaque->chrdev, &indio_dev->dev);
if (ret < 0)
goto error_unreg_eventset;
return 0;
error_unreg_eventset:
iio_device_unregister_eventset(indio_dev);
error_free_sysfs:
iio_device_unregister_sysfs(indio_dev);
error_buffer_free_sysfs:
iio_buffers_free_sysfs_and_mask(indio_dev);
error_unreg_debugfs:
iio_device_unregister_debugfs(indio_dev);
return ret;
}
EXPORT_SYMBOL(__iio_device_register);
/**
* iio_device_unregister() - unregister a device from the IIO subsystem
* @indio_dev: Device structure representing the device.
**/
void iio_device_unregister(struct iio_dev *indio_dev)
{
struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
cdev_device_del(&iio_dev_opaque->chrdev, &indio_dev->dev);
mutex_lock(&iio_dev_opaque->info_exist_lock);
iio: core: fix a possible circular locking dependency This fixes a possible circular locking dependency detected warning seen with: - CONFIG_PROVE_LOCKING=y - consumer/provider IIO devices (ex: "voltage-divider" consumer of "adc") When using the IIO consumer interface, e.g. iio_channel_get(), the consumer device will likely call iio_read_channel_raw() or similar that rely on 'info_exist_lock' mutex. typically: ... mutex_lock(&chan->indio_dev->info_exist_lock); if (chan->indio_dev->info == NULL) { ret = -ENODEV; goto err_unlock; } ret = do_some_ops() err_unlock: mutex_unlock(&chan->indio_dev->info_exist_lock); return ret; ... Same mutex is also hold in iio_device_unregister(). The following deadlock warning happens when: - the consumer device has called an API like iio_read_channel_raw() at least once. - the consumer driver is unregistered, removed (unbind from sysfs) ====================================================== WARNING: possible circular locking dependency detected 4.19.24 #577 Not tainted ------------------------------------------------------ sh/372 is trying to acquire lock: (kn->count#30){++++}, at: kernfs_remove_by_name_ns+0x3c/0x84 but task is already holding lock: (&dev->info_exist_lock){+.+.}, at: iio_device_unregister+0x18/0x60 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 (&dev->info_exist_lock){+.+.}: __mutex_lock+0x70/0xa3c mutex_lock_nested+0x1c/0x24 iio_read_channel_raw+0x1c/0x60 iio_read_channel_info+0xa8/0xb0 dev_attr_show+0x1c/0x48 sysfs_kf_seq_show+0x84/0xec seq_read+0x154/0x528 __vfs_read+0x2c/0x15c vfs_read+0x8c/0x110 ksys_read+0x4c/0xac ret_fast_syscall+0x0/0x28 0xbedefb60 -> #0 (kn->count#30){++++}: lock_acquire+0xd8/0x268 __kernfs_remove+0x288/0x374 kernfs_remove_by_name_ns+0x3c/0x84 remove_files+0x34/0x78 sysfs_remove_group+0x40/0x9c sysfs_remove_groups+0x24/0x34 device_remove_attrs+0x38/0x64 device_del+0x11c/0x360 cdev_device_del+0x14/0x2c iio_device_unregister+0x24/0x60 release_nodes+0x1bc/0x200 device_release_driver_internal+0x1a0/0x230 unbind_store+0x80/0x130 kernfs_fop_write+0x100/0x1e4 __vfs_write+0x2c/0x160 vfs_write+0xa4/0x17c ksys_write+0x4c/0xac ret_fast_syscall+0x0/0x28 0xbe906840 other info that might help us debug this: Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&dev->info_exist_lock); lock(kn->count#30); lock(&dev->info_exist_lock); lock(kn->count#30); *** DEADLOCK *** ... cdev_device_del() can be called without holding the lock. It should be safe as info_exist_lock prevents kernelspace consumers to use the exported routines during/after provider removal. cdev_device_del() is for userspace. Help to reproduce: See example: Documentation/devicetree/bindings/iio/afe/voltage-divider.txt sysv { compatible = "voltage-divider"; io-channels = <&adc 0>; output-ohms = <22>; full-ohms = <222>; }; First, go to iio:deviceX for the "voltage-divider", do one read: $ cd /sys/bus/iio/devices/iio:deviceX $ cat in_voltage0_raw Then, unbind the consumer driver. It triggers above deadlock warning. $ cd /sys/bus/platform/drivers/iio-rescale/ $ echo sysv > unbind Note I don't actually expect stable will pick this up all the way back into IIO being in staging, but if's probably valid that far back. Signed-off-by: Fabrice Gasnier <fabrice.gasnier@st.com> Fixes: ac917a81117c ("staging:iio:core set the iio_dev.info pointer to null on unregister") Cc: <Stable@vger.kernel.org> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
2019-03-25 13:01:23 +00:00
iio_device_unregister_debugfs(indio_dev);
iio_disable_all_buffers(indio_dev);
indio_dev->info = NULL;
iio_device_wakeup_eventset(indio_dev);
iio_buffer_wakeup_poll(indio_dev);
mutex_unlock(&iio_dev_opaque->info_exist_lock);
iio_buffers_free_sysfs_and_mask(indio_dev);
}
EXPORT_SYMBOL(iio_device_unregister);
static void devm_iio_device_unreg(void *indio_dev)
{
iio_device_unregister(indio_dev);
}
int __devm_iio_device_register(struct device *dev, struct iio_dev *indio_dev,
struct module *this_mod)
{
int ret;
ret = __iio_device_register(indio_dev, this_mod);
if (ret)
return ret;
return devm_add_action_or_reset(dev, devm_iio_device_unreg, indio_dev);
}
EXPORT_SYMBOL_GPL(__devm_iio_device_register);
/**
* iio_device_claim_direct_mode - Keep device in direct mode
* @indio_dev: the iio_dev associated with the device
*
* If the device is in direct mode it is guaranteed to stay
* that way until iio_device_release_direct_mode() is called.
*
* Use with iio_device_release_direct_mode()
*
* Returns: 0 on success, -EBUSY on failure
*/
int iio_device_claim_direct_mode(struct iio_dev *indio_dev)
{
mutex_lock(&indio_dev->mlock);
if (iio_buffer_enabled(indio_dev)) {
mutex_unlock(&indio_dev->mlock);
return -EBUSY;
}
return 0;
}
EXPORT_SYMBOL_GPL(iio_device_claim_direct_mode);
/**
* iio_device_release_direct_mode - releases claim on direct mode
* @indio_dev: the iio_dev associated with the device
*
* Release the claim. Device is no longer guaranteed to stay
* in direct mode.
*
* Use with iio_device_claim_direct_mode()
*/
void iio_device_release_direct_mode(struct iio_dev *indio_dev)
{
mutex_unlock(&indio_dev->mlock);
}
EXPORT_SYMBOL_GPL(iio_device_release_direct_mode);
subsys_initcall(iio_init);
module_exit(iio_exit);
MODULE_AUTHOR("Jonathan Cameron <jic23@kernel.org>");
MODULE_DESCRIPTION("Industrial I/O core");
MODULE_LICENSE("GPL");