Based on kernel version 4.3. Page generated on 2015-11-02 12:50 EST.
1 rfkill - RF kill switch support 2 =============================== 3 4 1. Introduction 5 2. Implementation details 6 3. Kernel API 7 4. Userspace support 8 9 10 1. Introduction 11 12 The rfkill subsystem provides a generic interface to disabling any radio 13 transmitter in the system. When a transmitter is blocked, it shall not 14 radiate any power. 15 16 The subsystem also provides the ability to react on button presses and 17 disable all transmitters of a certain type (or all). This is intended for 18 situations where transmitters need to be turned off, for example on 19 aircraft. 20 21 The rfkill subsystem has a concept of "hard" and "soft" block, which 22 differ little in their meaning (block == transmitters off) but rather in 23 whether they can be changed or not: 24 - hard block: read-only radio block that cannot be overridden by software 25 - soft block: writable radio block (need not be readable) that is set by 26 the system software. 27 28 The rfkill subsystem has two parameters, rfkill.default_state and 29 rfkill.master_switch_mode, which are documented in kernel-parameters.txt. 30 31 32 2. Implementation details 33 34 The rfkill subsystem is composed of three main components: 35 * the rfkill core, 36 * the deprecated rfkill-input module (an input layer handler, being 37 replaced by userspace policy code) and 38 * the rfkill drivers. 39 40 The rfkill core provides API for kernel drivers to register their radio 41 transmitter with the kernel, methods for turning it on and off and, letting 42 the system know about hardware-disabled states that may be implemented on 43 the device. 44 45 The rfkill core code also notifies userspace of state changes, and provides 46 ways for userspace to query the current states. See the "Userspace support" 47 section below. 48 49 When the device is hard-blocked (either by a call to rfkill_set_hw_state() 50 or from query_hw_block) set_block() will be invoked for additional software 51 block, but drivers can ignore the method call since they can use the return 52 value of the function rfkill_set_hw_state() to sync the software state 53 instead of keeping track of calls to set_block(). In fact, drivers should 54 use the return value of rfkill_set_hw_state() unless the hardware actually 55 keeps track of soft and hard block separately. 56 57 58 3. Kernel API 59 60 61 Drivers for radio transmitters normally implement an rfkill driver. 62 63 Platform drivers might implement input devices if the rfkill button is just 64 that, a button. If that button influences the hardware then you need to 65 implement an rfkill driver instead. This also applies if the platform provides 66 a way to turn on/off the transmitter(s). 67 68 For some platforms, it is possible that the hardware state changes during 69 suspend/hibernation, in which case it will be necessary to update the rfkill 70 core with the current state is at resume time. 71 72 To create an rfkill driver, driver's Kconfig needs to have 73 74 depends on RFKILL || !RFKILL 75 76 to ensure the driver cannot be built-in when rfkill is modular. The !RFKILL 77 case allows the driver to be built when rfkill is not configured, which 78 case all rfkill API can still be used but will be provided by static inlines 79 which compile to almost nothing. 80 81 Calling rfkill_set_hw_state() when a state change happens is required from 82 rfkill drivers that control devices that can be hard-blocked unless they also 83 assign the poll_hw_block() callback (then the rfkill core will poll the 84 device). Don't do this unless you cannot get the event in any other way. 85 86 87 88 5. Userspace support 89 90 The recommended userspace interface to use is /dev/rfkill, which is a misc 91 character device that allows userspace to obtain and set the state of rfkill 92 devices and sets of devices. It also notifies userspace about device addition 93 and removal. The API is a simple read/write API that is defined in 94 linux/rfkill.h, with one ioctl that allows turning off the deprecated input 95 handler in the kernel for the transition period. 96 97 Except for the one ioctl, communication with the kernel is done via read() 98 and write() of instances of 'struct rfkill_event'. In this structure, the 99 soft and hard block are properly separated (unlike sysfs, see below) and 100 userspace is able to get a consistent snapshot of all rfkill devices in the 101 system. Also, it is possible to switch all rfkill drivers (or all drivers of 102 a specified type) into a state which also updates the default state for 103 hotplugged devices. 104 105 After an application opens /dev/rfkill, it can read the current state of all 106 devices. Changes can be either obtained by either polling the descriptor for 107 hotplug or state change events or by listening for uevents emitted by the 108 rfkill core framework. 109 110 Additionally, each rfkill device is registered in sysfs and emits uevents. 111 112 rfkill devices issue uevents (with an action of "change"), with the following 113 environment variables set: 114 115 RFKILL_NAME 116 RFKILL_STATE 117 RFKILL_TYPE 118 119 The contents of these variables corresponds to the "name", "state" and 120 "type" sysfs files explained above. 121 122 123 For further details consult Documentation/ABI/stable/sysfs-class-rfkill.