Documentation/watchdog/watchdog-kernel-api.txt

   1 The Linux WatchDog Timer Driver Core kernel API.
   2 ===============================================
   3 Last reviewed: 12-Feb-2013
   4
   5 Wim Van Sebroeck <wim@iguana.be>
   6
   7 Introduction
   8 ------------
   9 This document does not describe what a WatchDog Timer (WDT) Driver or Device is.
  10 It also does not describe the API which can be used by user space to communicate
  11 with a WatchDog Timer. If you want to know this then please read the following
  12 file: Documentation/watchdog/watchdog-api.txt .
  13
  14 So what does this document describe? It describes the API that can be used by
  15 WatchDog Timer Drivers that want to use the WatchDog Timer Driver Core
  16 Framework. This framework provides all interfacing towards user space so that
  17 the same code does not have to be reproduced each time. This also means that
  18 a watchdog timer driver then only needs to provide the different routines
  19 (operations) that control the watchdog timer (WDT).
  20
  21 The API
  22 -------
  23 Each watchdog timer driver that wants to use the WatchDog Timer Driver Core
  24 must #include <linux/watchdog.h> (you would have to do this anyway when
  25 writing a watchdog device driver). This include file contains following
  26 register/unregister routines:
  27
  28 extern int watchdog_register_device(struct watchdog_device *);
  29 extern void watchdog_unregister_device(struct watchdog_device *);
  30
  31 The watchdog_register_device routine registers a watchdog timer device.
  32 The parameter of this routine is a pointer to a watchdog_device structure.
  33 This routine returns zero on success and a negative errno code for failure.
  34
  35 The watchdog_unregister_device routine deregisters a registered watchdog timer
  36 device. The parameter of this routine is the pointer to the registered
  37 watchdog_device structure.
  38
  39 The watchdog subsystem includes an registration deferral mechanism,
  40 which allows you to register an watchdog as early as you wish during
  41 the boot process.
  42
  43 The watchdog device structure looks like this:
  44
  45 struct watchdog_device {
  46         int id;
  47         struct device *parent;
  48         const struct attribute_group **groups;
  49         const struct watchdog_info *info;
  50         const struct watchdog_ops *ops;
  51         unsigned int bootstatus;
  52         unsigned int timeout;
  53         unsigned int min_timeout;
  54         unsigned int max_timeout;
  55         unsigned int min_hw_heartbeat_ms;
  56         unsigned int max_hw_heartbeat_ms;
  57         struct notifier_block reboot_nb;
  58         struct notifier_block restart_nb;
  59         void *driver_data;
  60         struct watchdog_core_data *wd_data;
  61         unsigned long status;
  62         struct list_head deferred;
  63 };
  64
  65 It contains following fields:
  66 * id: set by watchdog_register_device, id 0 is special. It has both a
  67   /dev/watchdog0 cdev (dynamic major, minor 0) as well as the old
  68   /dev/watchdog miscdev. The id is set automatically when calling
  69   watchdog_register_device.
  70 * parent: set this to the parent device (or NULL) before calling
  71   watchdog_register_device.
  72 * groups: List of sysfs attribute groups to create when creating the watchdog
  73   device.
  74 * info: a pointer to a watchdog_info structure. This structure gives some
  75   additional information about the watchdog timer itself. (Like it's unique name)
  76 * ops: a pointer to the list of watchdog operations that the watchdog supports.
  77 * timeout: the watchdog timer's timeout value (in seconds).
  78   This is the time after which the system will reboot if user space does
  79   not send a heartbeat request if WDOG_ACTIVE is set.
  80 * min_timeout: the watchdog timer's minimum timeout value (in seconds).
  81   If set, the minimum configurable value for 'timeout'.
  82 * max_timeout: the watchdog timer's maximum timeout value (in seconds),
  83   as seen from userspace. If set, the maximum configurable value for
  84   'timeout'. Not used if max_hw_heartbeat_ms is non-zero.
  85 * min_hw_heartbeat_ms: Minimum time between heartbeats sent to the chip,
  86   in milli-seconds.
  87 * max_hw_heartbeat_ms: Maximum hardware heartbeat, in milli-seconds.
  88   If set, the infrastructure will send heartbeats to the watchdog driver
  89   if 'timeout' is larger than max_hw_heartbeat_ms, unless WDOG_ACTIVE
  90   is set and userspace failed to send a heartbeat for at least 'timeout'
  91   seconds. max_hw_heartbeat_ms must be set if a driver does not implement
  92   the stop function.
  93 * reboot_nb: notifier block that is registered for reboot notifications, for
  94   internal use only. If the driver calls watchdog_stop_on_reboot, watchdog core
  95   will stop the watchdog on such notifications.
  96 * restart_nb: notifier block that is registered for machine restart, for
  97   internal use only. If a watchdog is capable of restarting the machine, it
  98   should define ops->restart. Priority can be changed through
  99   watchdog_set_restart_priority.
 100 * bootstatus: status of the device after booting (reported with watchdog
 101   WDIOF_* status bits).
 102 * driver_data: a pointer to the drivers private data of a watchdog device.
 103   This data should only be accessed via the watchdog_set_drvdata and
 104   watchdog_get_drvdata routines.
 105 * wd_data: a pointer to watchdog core internal data.
 106 * status: this field contains a number of status bits that give extra
 107   information about the status of the device (Like: is the watchdog timer
 108   running/active, or is the nowayout bit set).
 109 * deferred: entry in wtd_deferred_reg_list which is used to
 110   register early initialized watchdogs.
 111
 112 The list of watchdog operations is defined as:
 113
 114 struct watchdog_ops {
 115         struct module *owner;
 116         /* mandatory operations */
 117         int (*start)(struct watchdog_device *);
 118         int (*stop)(struct watchdog_device *);
 119         /* optional operations */
 120         int (*ping)(struct watchdog_device *);
 121         unsigned int (*status)(struct watchdog_device *);
 122         int (*set_timeout)(struct watchdog_device *, unsigned int);
 123         unsigned int (*get_timeleft)(struct watchdog_device *);
 124         int (*restart)(struct watchdog_device *);
 125         void (*ref)(struct watchdog_device *) __deprecated;
 126         void (*unref)(struct watchdog_device *) __deprecated;
 127         long (*ioctl)(struct watchdog_device *, unsigned int, unsigned long);
 128 };
 129
 130 It is important that you first define the module owner of the watchdog timer
 131 driver's operations. This module owner will be used to lock the module when
 132 the watchdog is active. (This to avoid a system crash when you unload the
 133 module and /dev/watchdog is still open).
 134
 135 Some operations are mandatory and some are optional. The mandatory operations
 136 are:
 137 * start: this is a pointer to the routine that starts the watchdog timer
 138   device.
 139   The routine needs a pointer to the watchdog timer device structure as a
 140   parameter. It returns zero on success or a negative errno code for failure.
 141
 142 Not all watchdog timer hardware supports the same functionality. That's why
 143 all other routines/operations are optional. They only need to be provided if
 144 they are supported. These optional routines/operations are:
 145 * stop: with this routine the watchdog timer device is being stopped.
 146   The routine needs a pointer to the watchdog timer device structure as a
 147   parameter. It returns zero on success or a negative errno code for failure.
 148   Some watchdog timer hardware can only be started and not be stopped. A
 149   driver supporting such hardware does not have to implement the stop routine.
 150   If a driver has no stop function, the watchdog core will set WDOG_HW_RUNNING
 151   and start calling the driver's keepalive pings function after the watchdog
 152   device is closed.
 153   If a watchdog driver does not implement the stop function, it must set
 154   max_hw_heartbeat_ms.
 155 * ping: this is the routine that sends a keepalive ping to the watchdog timer
 156   hardware.
 157   The routine needs a pointer to the watchdog timer device structure as a
 158   parameter. It returns zero on success or a negative errno code for failure.
 159   Most hardware that does not support this as a separate function uses the
 160   start function to restart the watchdog timer hardware. And that's also what
 161   the watchdog timer driver core does: to send a keepalive ping to the watchdog
 162   timer hardware it will either use the ping operation (when available) or the
 163   start operation (when the ping operation is not available).
 164   (Note: the WDIOC_KEEPALIVE ioctl call will only be active when the
 165   WDIOF_KEEPALIVEPING bit has been set in the option field on the watchdog's
 166   info structure).
 167 * status: this routine checks the status of the watchdog timer device. The
 168   status of the device is reported with watchdog WDIOF_* status flags/bits.
 169 * set_timeout: this routine checks and changes the timeout of the watchdog
 170   timer device. It returns 0 on success, -EINVAL for "parameter out of range"
 171   and -EIO for "could not write value to the watchdog". On success this
 172   routine should set the timeout value of the watchdog_device to the
 173   achieved timeout value (which may be different from the requested one
 174   because the watchdog does not necessarily have a 1 second resolution).
 175   Drivers implementing max_hw_heartbeat_ms set the hardware watchdog heartbeat
 176   to the minimum of timeout and max_hw_heartbeat_ms. Those drivers set the
 177   timeout value of the watchdog_device either to the requested timeout value
 178   (if it is larger than max_hw_heartbeat_ms), or to the achieved timeout value.
 179   (Note: the WDIOF_SETTIMEOUT needs to be set in the options field of the
 180   watchdog's info structure).
 181   If the watchdog driver does not have to perform any action but setting the
 182   watchdog_device.timeout, this callback can be omitted.
 183   If set_timeout is not provided but, WDIOF_SETTIMEOUT is set, the watchdog
 184   infrastructure updates the timeout value of the watchdog_device internally
 185   to the requested value.
 186 * get_timeleft: this routines returns the time that's left before a reset.
 187 * restart: this routine restarts the machine. It returns 0 on success or a
 188   negative errno code for failure.
 189 * ioctl: if this routine is present then it will be called first before we do
 190   our own internal ioctl call handling. This routine should return -ENOIOCTLCMD
 191   if a command is not supported. The parameters that are passed to the ioctl
 192   call are: watchdog_device, cmd and arg.
 193
 194 The 'ref' and 'unref' operations are no longer used and deprecated.
 195
 196 The status bits should (preferably) be set with the set_bit and clear_bit alike
 197 bit-operations. The status bits that are defined are:
 198 * WDOG_ACTIVE: this status bit indicates whether or not a watchdog timer device
 199   is active or not from user perspective. User space is expected to send
 200   heartbeat requests to the driver while this flag is set.
 201 * WDOG_NO_WAY_OUT: this bit stores the nowayout setting for the watchdog.
 202   If this bit is set then the watchdog timer will not be able to stop.
 203 * WDOG_HW_RUNNING: Set by the watchdog driver if the hardware watchdog is
 204   running. The bit must be set if the watchdog timer hardware can not be
 205   stopped. The bit may also be set if the watchdog timer is running after
 206   booting, before the watchdog device is opened. If set, the watchdog
 207   infrastructure will send keepalives to the watchdog hardware while
 208   WDOG_ACTIVE is not set.
 209   Note: when you register the watchdog timer device with this bit set,
 210   then opening /dev/watchdog will skip the start operation but send a keepalive
 211   request instead.
 212
 213   To set the WDOG_NO_WAY_OUT status bit (before registering your watchdog
 214   timer device) you can either:
 215   * set it statically in your watchdog_device struct with
 216         .status = WATCHDOG_NOWAYOUT_INIT_STATUS,
 217     (this will set the value the same as CONFIG_WATCHDOG_NOWAYOUT) or
 218   * use the following helper function:
 219   static inline void watchdog_set_nowayout(struct watchdog_device *wdd, int nowayout)
 220
 221 Note: The WatchDog Timer Driver Core supports the magic close feature and
 222 the nowayout feature. To use the magic close feature you must set the
 223 WDIOF_MAGICCLOSE bit in the options field of the watchdog's info structure.
 224 The nowayout feature will overrule the magic close feature.
 225
 226 To get or set driver specific data the following two helper functions should be
 227 used:
 228
 229 static inline void watchdog_set_drvdata(struct watchdog_device *wdd, void *data)
 230 static inline void *watchdog_get_drvdata(struct watchdog_device *wdd)
 231
 232 The watchdog_set_drvdata function allows you to add driver specific data. The
 233 arguments of this function are the watchdog device where you want to add the
 234 driver specific data to and a pointer to the data itself.
 235
 236 The watchdog_get_drvdata function allows you to retrieve driver specific data.
 237 The argument of this function is the watchdog device where you want to retrieve
 238 data from. The function returns the pointer to the driver specific data.
 239
 240 To initialize the timeout field, the following function can be used:
 241
 242 extern int watchdog_init_timeout(struct watchdog_device *wdd,
 243                                   unsigned int timeout_parm, struct device *dev);
 244
 245 The watchdog_init_timeout function allows you to initialize the timeout field
 246 using the module timeout parameter or by retrieving the timeout-sec property from
 247 the device tree (if the module timeout parameter is invalid). Best practice is
 248 to set the default timeout value as timeout value in the watchdog_device and
 249 then use this function to set the user "preferred" timeout value.
 250 This routine returns zero on success and a negative errno code for failure.
 251
 252 To disable the watchdog on reboot, the user must call the following helper:
 253
 254 static inline void watchdog_stop_on_reboot(struct watchdog_device *wdd);
 255
 256 To change the priority of the restart handler the following helper should be
 257 used:
 258
 259 void watchdog_set_restart_priority(struct watchdog_device *wdd, int priority);
 260
 261 User should follow the following guidelines for setting the priority:
 262 * 0: should be called in last resort, has limited restart capabilities
 263 * 128: default restart handler, use if no other handler is expected to be
 264   available, and/or if restart is sufficient to restart the entire system
 265 * 255: highest priority, will preempt all other restart handlers