Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756059AbbGPTbv (ORCPT ); Thu, 16 Jul 2015 15:31:51 -0400 Received: from saturn.retrosnub.co.uk ([178.18.118.26]:51824 "EHLO saturn.retrosnub.co.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751637AbbGPTbs (ORCPT ); Thu, 16 Jul 2015 15:31:48 -0400 User-Agent: K-9 Mail for Android In-Reply-To: <1436357088-30743-2-git-send-email-daniel.baluta@intel.com> References: <1436357088-30743-1-git-send-email-daniel.baluta@intel.com> <1436357088-30743-2-git-send-email-daniel.baluta@intel.com> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=UTF-8 Subject: Re: [PATCH] DocBook: Add initial documentation for IIO From: Jonathan Cameron Date: Thu, 16 Jul 2015 20:31:29 +0100 To: Daniel Baluta , Daniel Baluta CC: pmeerw@pmeerw.net, knaack.h@gmx.de, lars@metafoo.de, linux-kernel@vger.kernel.org, linux-iio@vger.kernel.org, pmeerw@pmeerw.net, knaack.h@gmx.de, lars@metafoo.de, linux-kernel@vger.kernel.org, linux-iio@vger.kernel.org Message-ID: <0C20DA6F-E21C-495B-A31B-4EC0DC66EF7A@kernel.org> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 25856 Lines: 761 On 8 July 2015 13:04:48 BST, Daniel Baluta wrote: >This is intended to help developers faster find their way >inside the Industrial I/O core and reduce time spent on IIO >drivers development. > >Signed-off-by: Daniel Baluta >--- > Documentation/DocBook/Makefile | 2 +- >Documentation/DocBook/iio.tmpl | 593 >+++++++++++++++++++++++++++++++++++++++++ > 2 files changed, 594 insertions(+), 1 deletion(-) > create mode 100644 Documentation/DocBook/iio.tmpl > >diff --git a/Documentation/DocBook/Makefile >b/Documentation/DocBook/Makefile >index b6a6a2e..9e08606 100644 >--- a/Documentation/DocBook/Makefile >+++ b/Documentation/DocBook/Makefile >@@ -15,7 +15,7 @@ DOCBOOKS := z8530book.xml device-drivers.xml \ > 80211.xml debugobjects.xml sh.xml regulator.xml \ > alsa-driver-api.xml writing-an-alsa-driver.xml \ > tracepoint.xml drm.xml media_api.xml w1.xml \ >- writing_musb_glue_layer.xml crypto-API.xml >+ writing_musb_glue_layer.xml crypto-API.xml iio.xml > > include Documentation/DocBook/media/Makefile > >diff --git a/Documentation/DocBook/iio.tmpl >b/Documentation/DocBook/iio.tmpl >new file mode 100644 >index 0000000..417bb26 >--- /dev/null >+++ b/Documentation/DocBook/iio.tmpl >@@ -0,0 +1,593 @@ >+ >++ "http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd" []> >+ >+ >+ >+ Industrial I/O driver developer's guide >+ >+ >+ >+ Daniel >+ Baluta >+ >+
>+ daniel.baluta@intel.com >+
>+
>+
>+
>+ >+ >+ 2015 >+ Intel Corporation >+ >+ >+ >+ >+ This documentation is free software; you can redistribute >+ it and/or modify it under the terms of the GNU General Public >+ License version 2. >+ >+ >+ >+ This program is distributed in the hope that it will be >+ useful, but WITHOUT ANY WARRANTY; without even the implied >+ warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR >PURPOSE. >+ For more details see the file COPYING in the source >+ distribution of Linux. >+ >+ >+
>+ >+ >+ >+ >+ Introduction >+ >+ The main purpose of the Industrial I/O subsystem (IIO) is to >provide >+ support for devices that in some sense are analog to digital >converts >+ (ADCs). As many actual devices combine some ADCs with digital to >analog >+ converters (DACs), that functionality is also supported. I wonder if we now want to treat DACs at the same level in the docs as ADCs. Right now our support is simpler but there are patches adding full buffered support to output devices as well (Lars?) Mind you we can always change the emphasis whenever we like! > The aim >is to >+ fill the gap between the somewhat similar hwmon and input >subsystems. >+ Hwmon is very much directed at low sample rate sensors used in >+ applications such as fan speed control and temperature >measurement. Input >+ is, as its name suggests, focused on human interaction input >devices >+ (keyboard, mouse, touchscreen). In some cases there is >considerable >+ overlap between these and IIO. >+ >+ >+ Devices that fall into this category are: Include (keep it so we don't have to always remember to update this!) >+ >+ >+ >+ analog to digital converters (ADCs) >+ >+ >+ accelerometers >+ >+ >+ capacitance to digital converters (CDCs) >+ >+ >+ digital to analog converters (DACs) >+ >+ >+ gyroscopes >+ >+ >+ inertial measurement units (IMUs) >+ >+ >+ color and light sensors >+ >+ >+ magnetometers >+ >+ >+ pressure sensors >+ >+ >+ proximity sensors >+ >+ >+ temperature sensors >+ >+ >+ Usually these sensors are connected via SPI or I2C. Maybe add something about SOC integrated parts as well? > It is also a >common >+ use case to have combo functionality (e.g. light plus proximity >sensor). >+ >+ >+ >+ Industrial I/O core >+ >+ The Industrial I/O core offers: >+ >+ >+ a unified framework for writing drivers for many different >types of >+ embedded sensors. >+ >+ >+ a standard interface to user space applications manipulating >sensors. >+ >+ >+ The implementation can be found under >+ drivers/iio/industrialio-* >+ >+ >+ Industrial I/O devices >+ >+!Finclude/linux/iio/iio.h iio_dev >+!Fdrivers/iio/industrialio-core.c iio_device_alloc >+!Fdrivers/iio/industrialio-core.c iio_device_free >+!Fdrivers/iio/industrialio-core.c iio_device_register >+!Fdrivers/iio/industrialio-core.c iio_device_unregister >+ >+ >+ An IIO device usually corresponds to a single hardware sensor >and it >+ provides all the information needed by a driver handling a >device. Could clarify when it doesn't somewhere? Maybe that is later... >+ Let's first have a look at the functionality embedded in an IIO >+ device then we will show how a device driver makes use of an IIO >+ device. >+ >+ >+ There are two ways for a user space application to interact >+ with an IIO driver. >+ >+ >+ /sys/bus/iio/iio:deviceX/, this >+ represents a hardware sensor and groups together the data >+ channels of the same chip. >+ >+ >+ /dev/iio:deviceX, character device node >+ interface used for faster data transfer and for events >information >+ retrieval. >+ >+ >+ >+ A typical IIO driver will register itself as an I2C or SPI driver >and will >+ create two routines, probe and >remove >+ . At probe: >+ >+ call iio_device_alloc, which >allocates memory >+ for an IIO device. >+ >+ initialize IIO device fields with driver specific >information >+ (e.g. device name, device channels). >+ >+ call iio_device_register, this >registers the >+ device with the IIO core. After this call the device is ready to >accept >+ requests from user space applications. >+ >+ >+ At remove, we free the resources allocated >in >+ probe in reverse order: >+ >+ iio_device_unregister, unregister >the device >+ from the IIO core. >+ >+ iio_device_free, free the memory >allocated >+ for the IIO device. >+ >+ >+ >+ IIO device sysfs interface >+ >+ Attributes are sysfs files used to expose chip info and also >allowing >+ applications to set various configuration parameters. Common >+ attributes are: >+ >+ >/sys/bus/iio/iio:deviceX/name, >+ description of the physical chip for device X. >+ >+ /sys/bus/iio/iio:deviceX/dev, >+ shows the major:minor pair associated with >+ /dev/iio:deviceX node. >+ >+ >/sys/bus/iio:deviceX/sampling_frequency_available >+ available discrete set of sampling frequency >values >+ for device X. >+ >+ >+ Available standard attributes for IIO devices are described in >the >+ Documentation/ABI/testing/sysfs-bus-iio >file >+ in the Linux kernel sources. >+ >+ >+ IIO device channels >+!Finclude/linux/iio/iio.h iio_chan_spec structure. >+ >+ An IIO device channel is a representation of a data channel. Maybe define data channel? >An >+ IIO device can have one or multiple channels. For example: >+ >+ >+ a thermometer sensor has one channel representing the >+ temperature measurement. >+ >+ >+ a light sensor has With instead of has (because its an example) >two channels indicating the measurements >in >+ the visible and infrared spectrum. >+ >+ >+ an accelerometer can have up two to 3 channels representing >+ acceleration on X, Y and Z axes. >+ >+ >+ An IIO channel is described by the struct iio_chan_spec >+ . A thermometer driver for the temperature sensor in the >+ example above would have to describe its channel as follows: >+ >+ static const struct iio_chan_spec temp_channel[] = { >+ { >+ .type = IIO_TEMP, >+ .info_mask_separate = BIT(IIO_CHAN_INFO_PROCESSED), >+ }, >+ }; >+ >+ >+ When there are multiple channels per device we need to set the > >+ .modified field of iio_chan_spec to 1. For >example, >+ a light sensor can have two channels one, for infrared light and >one for >+ both infrared and visible light. Could also be indexed. Perhaps something describing that modifiers are to indicate a physically unique characteristic of the channel such as its direction or spectral response. If the channel is simply another instance then indexed is more appropriate. >+ >+ static const struct iio_chan_spec light_channels[] = { >+ { >+ .type = IIO_INTENSITY, >+ .modified = 1, >+ .channel2 = IIO_MOD_LIGHT_IR, >+ .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), >+ .info_mask_shared = BIT(IIO_CHAN_INFO_SAMP_FREQ), >+ }, >+ { >+ .type = IIO_INTENSITY, >+ .modified = 1, >+ .channel2 = IIO_MOD_LIGHT_BOTH, >+ .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), >+ .info_mask_shared = BIT(IIO_CHAN_INFO_SAMP_FREQ), >+ }, >+ } >+ >+ >+ This channel's definition will generate two separate sysfs files >+ for raw data retrieval: >+ >+ >+ >/sys/bus/iio/iio:deviceX/in_intensity_ir_raw >+ >+ >+ >/sys/bus/iio/iio:deviceX/in_intensity_both_raw >+ >+ >+ and one shared sysfs file for sampling frequency: >+ >+ >+ >/sys/bus/iio/iio:deviceX/in_intensity_sampling_frequency. >+ >+ >+ >+ >+ >+ + >+ Industrial I/O buffers >+!Finclude/linux/iio/buffer.h iio_buffer >+!Edrivers/iio/industrialio-buffer.c >+ >+ >+ The Industrial I/O core offers a way for continuous data capture >+ based on a trigger source. Multiple data channels can be read at >once >+ from /dev/iio:deviceX character device node, >+ thus reducing the CPU load. >+ Why reduced load? Also perhaps mention pseudo scans? (Sort of all at the same time) >+ >+ >+ IIO buffer sysfs interface >+ >+ An IIO buffer has an associated attributes directory under > >+ /sys/bus/iio/iio:deviceX/buffer/. Here are the >existing >+ attributes: >+ >+ >+ length, number of data samples contained by >the >+ buffer. >+ >+ >+ enable, activate buffer capture. >+ >+ >+ >+ In order to be useful, a buffer needs to have an associated >+ trigger. Future chapters will add details about triggers and how >+ drivers use triggers to start data capture, moving samples from >device >+ registers to buffer storage and then upward to user space >applications. >+ >+ >+ IIO buffer setup >+ The important bits for selecting data channels >+ configuration are exposed to userspace applications via the > >+ /sys/bus/iio/iio:deviceX/scan_elements/ directory. >This >+ file contains attributes of the following form: >+ >+ enable, used for enabling a >channel. >+ If and only if his attribute is non zero, thena Then a > triggered >capture >+ will contain data sample for this channel. >+ >+ type, description of the scan >element >+ data storage within the buffer and hence the form in which it >is >+ read from user space. Format is >+ [be|le]:[s|u]bits/storagebits[>>shift] . >+ >+ be or le >specifies >+ big or little endian. >+ >+ >+ s or u specifies if >+ signed (2's complement) or unsigned. >+ >+ bits is the number of bits of data >+ >+ storagebits is the space (after padding) that it >occupies in the >+ buffer. >+ >+ >+ shift if specified, is the shift that >needs >+ to be a applied prior to masking out unused bits >+ >+ >+ >+ >+ >+ For implementing buffer support a driver should initialize the >following >+ fields in iio_chan_spec definition: >+ >+ >+ struct iio_chan_spec { >+ /* other members */ >+ int scan_index >+ struct { >+ u8 realbits; >+ u8 storagebits; >+ u8 shift; >+ enum iio_endian endianness; >+ } scan_type; >+ } >+ >+ Here is what a 3-axis, 12 bits accelerometer channels definition >+ looks like with buffer support: >+ >+ >+ struct struct iio_chan_spec accel_channels[] = { >+ { >+ .type = IIO_ACCEL, >+ .modified = 1, >+ .channel2 = IIO_MOD_X, >+ /* other stuff here */ >+ .scan_index = 0, >+ .scan_type = { >+ .sign = 's', >+ .realbits = 12, >+ .storgebits = 16, >+ .shift = 4, >+ .endianess = IIO_CPU, >+ }, >+ } >+ /* similar for Y and Z axis */ >+ } >+ >+ Here scan_index is used for ordering data >samples >+ (scans) when read from buffer. >+ >+ >+ >+ >+ Industrial I/O triggers >+!Finclude/linux/iio/trigger.h iio_trigger >+!Edrivers/iio/industrialio-trigger.c >+ >+ In many situations it is useful for a driver to be able to >+ capture data based on some external event (trigger) as opposed >+ to periodically polling for data. An IIO trigger can be provided >+ by a device driver that also has an IIO device based on hardware >+ generated events (e.g. data ready or threshold exceeded) or >+ provided by a separate driver from an independent interrupt >+ source (e.g. GPIO line connected to some external system, timer >+ interrupt or user space reading a specific file in sysfs). A >+ trigger may initialize data capture for a number of sensors and >+ also it may be completely unrelated to the sensor itself. >+ >+ >+ IIO trigger sysfs interface > >+ There are two locations in sysfs related to triggers: >+ >+ /sys/bus/iio/devices/triggerX, >+ this file is created once an IIO triggered is registered >with >+ the IIO core and corresponds to trigger with index X. >Because >+ triggers can be very different depending on type there are >few >+ standard attributes that we can describe here: >+ >+ >+ name, trigger name that can be >later >+ used to for association with a device. >+ >+ >+ sampling_frequency, some timer based >+ triggers use this attribute to specify the frequency for >+ trigger calls. >+ >+ >+ >+ >+ >/sys/bus/iio/devices/iio:deviceX/trigger/, this >+ directory is created once the device supports a triggered >+ buffer. We can associate a trigger with our device by >writing >+ trigger's name in thecurrent_trigger >file. >+ >+ >+ >+ >+ IIO trigger setup >+ >+ >+ Let's see a simple example of how to setup a trigger to be used >+ by a driver. >+ >+ >+ struct iio_trigger_ops trigger_ops = { >+ .set_trigger_state = sample_trigger_state, >+ .validate_device = sample_validate_device, >+ } >+ >+ struct iio_trigger *trig; >+ >+ /* first, allocate memory for our trigger */ >+ trig = iio_trigger_alloc(dev, "trig-%s-%d", name, idx); >+ >+ /* setup trigger operations field */ >+ trig->ops = &trigger_ops; >+ >+ /* now register the trigger with the IIO core */ >+ iio_trigger_register(trig); >+ >+ >+ >+ >+ IIO trigger ops >+!Finclude/linux/iio/trigger.h iio_trigger_ops >+ >+ Notice that a trigger has a set of operations attached: >+ >+ >+ set_trigger_state, switch the trigger >on/off >+ on demand. >+ >+ >+ validate_device, function to validate >the >+ device when the current trigger gets changed. >+ >+ >+ >+ >+ >+ >+ Industrial I/O triggered buffers >+ >+ Now that we know what buffers and triggers are let's see how they >+ work together. >+ >+ IIO triggered buffer >setup >+!Edrivers/iio/industrialio-triggered-buffer.c >+!Finclude/linux/iio/iio.h iio_buffer_setup_ops >+ >+ >+ >+ A typical triggered buffer setup looks like this: >+ >+ const struct iio_buffer_setup_ops sensor_buffer_setup_ops = { >+ .preenable = sensor_buffer_preenable, >+ .postenable = sensor_buffer_postenable, >+ .postdisable = sensor_buffer_postdisable, >+ .predisable = sensor_buffer_predisable, >+ }; >+ >+ irqreturn_t sensor_iio_pollfunc(int irq, void *p) >+ { >+ pf->timestamp = iio_get_time_ns(); >+ return IRQ_WAKE_THREAD; >+ } >+ >+ irqreturn_t sensor_trigger_handler(int irq, void *p) >+ { >+ u16 buf[8]; >+ >+ /* read data for each active channel */ >+ for_each_set_bit(bit, active_scan_mask, masklength) >+ buf[i++] = sensor_get_data(bit) >+ >+ iio_push_to_buffers_with_timestamp(indio_dev, buffer, >timestamp); >+ >+ iio_trigger_notify_done(trigger); >+ } >+ >+ /* setup triggered buffer, usually in probe function */ >+ iio_triggered_buffer_setup(indio_dev, sensor_iio_polfunc, >+ sensor_trigger_handler, >+ sensor_buffer_setup_ops); >+ >+ >+ The important things to notice here are: >+ >+ iio_buffer_setup_ops, buffer setup >+ functions to be called at predefined points in buffer >configuration >+ sequence (e.g. before enable, after disable). If not specified, >IIO >+ core uses default iio_triggered_buffer_setup_ops. >+ >+ sensor_iio_pollfunc, function that >+ will be used as top half of poll function. It usually does little >+ processing (as it runs in interrupt context). Most common >operation >+ is recording of the current timestamp and for this reason one can >+ use the IIO core defined >iio_pollfunc_store_time >+ function. >+ >+ sensor_trigger_handler, function >that >+ will be used as bottom half of poll function. Maybe mention it is a thread? Here all the >+ processing takes place. It usually reads data from the device and >+ stores it in the internal buffer together with the timestamp >+ recorded in the top half. >+ >+ >+ >+ >+ >+ >+ Resources >+ IIO core may change during time so the best documentation to >read is the >+ source code. Hehe > There are several locations where you should look: >+ >+ >+ drivers/iio/, contains the IIO core >plus >+ and directories for each sensor type (e.g. accel, >magnetometer, >+ etc.) >+ >+ >+ include/linux/iio/, contains the header >+ files, nice to read for the internal kernel interfaces. >+ >+ >+ include/uapi/linux/iio/, contains file to >be >+ used by user space applications. >+ >+ >+ tools/iio/, contains tools for rapidly >+ testing buffers, events and device creation. >+ >+ >+ drivers/staging/iio/, contains code for >some >+ drivers or experimental features that are not yet mature >enough >+ to be moved out. >+ >+ >+ >+ Besides the code, there are some good online documentation >sources: >+ >+ >+ Industrial I/O >mailing >+ list >+ >+ >+ >+ Analog Device IIO wiki page >+ >+ >+ >+ Using the Linux IIO framework for SDR, Lars-Peter Clausen's >+ presentation at FOSDEM Cool. I had missed this one :) >+ >+ >+ >+ >+
>+ >+ Sorry for messy review. Ignoring my daughter down the park:) Anyhow, a great basis to build on. I like the example driven approach. Might have missed it but perhaps cross refer to the dummy driver? J I-- Sent from my Android device with K-9 Mail. Please excuse my brevity. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/