Commit | Line | Data |
---|---|---|
43316044 WVS |
1 | The Linux WatchDog Timer Driver Core kernel API. |
2 | =============================================== | |
3 | Last reviewed: 22-Jul-2011 | |
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 device structure looks like this: | |
40 | ||
41 | struct watchdog_device { | |
42 | const struct watchdog_info *info; | |
43 | const struct watchdog_ops *ops; | |
2fa03560 | 44 | unsigned int bootstatus; |
014d694e | 45 | unsigned int timeout; |
43316044 WVS |
46 | void *driver_data; |
47 | unsigned long status; | |
48 | }; | |
49 | ||
50 | It contains following fields: | |
51 | * info: a pointer to a watchdog_info structure. This structure gives some | |
52 | additional information about the watchdog timer itself. (Like it's unique name) | |
53 | * ops: a pointer to the list of watchdog operations that the watchdog supports. | |
014d694e | 54 | * timeout: the watchdog timer's timeout value (in seconds). |
2fa03560 WVS |
55 | * bootstatus: status of the device after booting (reported with watchdog |
56 | WDIOF_* status bits). | |
43316044 WVS |
57 | * driver_data: a pointer to the drivers private data of a watchdog device. |
58 | This data should only be accessed via the watchdog_set_drvadata and | |
59 | watchdog_get_drvdata routines. | |
60 | * status: this field contains a number of status bits that give extra | |
234445b4 | 61 | information about the status of the device (Like: is the watchdog timer |
7e192b9c WVS |
62 | running/active, is the nowayout bit set, is the device opened via |
63 | the /dev/watchdog interface or not, ...). | |
43316044 WVS |
64 | |
65 | The list of watchdog operations is defined as: | |
66 | ||
67 | struct watchdog_ops { | |
68 | struct module *owner; | |
69 | /* mandatory operations */ | |
70 | int (*start)(struct watchdog_device *); | |
71 | int (*stop)(struct watchdog_device *); | |
72 | /* optional operations */ | |
73 | int (*ping)(struct watchdog_device *); | |
2fa03560 | 74 | unsigned int (*status)(struct watchdog_device *); |
014d694e | 75 | int (*set_timeout)(struct watchdog_device *, unsigned int); |
78d88fc0 | 76 | long (*ioctl)(struct watchdog_device *, unsigned int, unsigned long); |
43316044 WVS |
77 | }; |
78 | ||
79 | It is important that you first define the module owner of the watchdog timer | |
80 | driver's operations. This module owner will be used to lock the module when | |
81 | the watchdog is active. (This to avoid a system crash when you unload the | |
82 | module and /dev/watchdog is still open). | |
83 | Some operations are mandatory and some are optional. The mandatory operations | |
84 | are: | |
85 | * start: this is a pointer to the routine that starts the watchdog timer | |
86 | device. | |
87 | The routine needs a pointer to the watchdog timer device structure as a | |
88 | parameter. It returns zero on success or a negative errno code for failure. | |
89 | * stop: with this routine the watchdog timer device is being stopped. | |
90 | The routine needs a pointer to the watchdog timer device structure as a | |
91 | parameter. It returns zero on success or a negative errno code for failure. | |
92 | Some watchdog timer hardware can only be started and not be stopped. The | |
93 | driver supporting this hardware needs to make sure that a start and stop | |
94 | routine is being provided. This can be done by using a timer in the driver | |
95 | that regularly sends a keepalive ping to the watchdog timer hardware. | |
96 | ||
97 | Not all watchdog timer hardware supports the same functionality. That's why | |
98 | all other routines/operations are optional. They only need to be provided if | |
99 | they are supported. These optional routines/operations are: | |
100 | * ping: this is the routine that sends a keepalive ping to the watchdog timer | |
101 | hardware. | |
102 | The routine needs a pointer to the watchdog timer device structure as a | |
103 | parameter. It returns zero on success or a negative errno code for failure. | |
104 | Most hardware that does not support this as a separate function uses the | |
105 | start function to restart the watchdog timer hardware. And that's also what | |
106 | the watchdog timer driver core does: to send a keepalive ping to the watchdog | |
107 | timer hardware it will either use the ping operation (when available) or the | |
108 | start operation (when the ping operation is not available). | |
c2dc00e4 WVS |
109 | (Note: the WDIOC_KEEPALIVE ioctl call will only be active when the |
110 | WDIOF_KEEPALIVEPING bit has been set in the option field on the watchdog's | |
111 | info structure). | |
2fa03560 WVS |
112 | * status: this routine checks the status of the watchdog timer device. The |
113 | status of the device is reported with watchdog WDIOF_* status flags/bits. | |
014d694e WVS |
114 | * set_timeout: this routine checks and changes the timeout of the watchdog |
115 | timer device. It returns 0 on success, -EINVAL for "parameter out of range" | |
116 | and -EIO for "could not write value to the watchdog". On success the timeout | |
117 | value of the watchdog_device will be changed to the value that was just used | |
118 | to re-program the watchdog timer device. | |
119 | (Note: the WDIOF_SETTIMEOUT needs to be set in the options field of the | |
120 | watchdog's info structure). | |
78d88fc0 WVS |
121 | * ioctl: if this routine is present then it will be called first before we do |
122 | our own internal ioctl call handling. This routine should return -ENOIOCTLCMD | |
123 | if a command is not supported. The parameters that are passed to the ioctl | |
124 | call are: watchdog_device, cmd and arg. | |
43316044 WVS |
125 | |
126 | The status bits should (preferably) be set with the set_bit and clear_bit alike | |
127 | bit-operations. The status bits that are defined are: | |
234445b4 WVS |
128 | * WDOG_ACTIVE: this status bit indicates whether or not a watchdog timer device |
129 | is active or not. When the watchdog is active after booting, then you should | |
130 | set this status bit (Note: when you register the watchdog timer device with | |
131 | this bit set, then opening /dev/watchdog will skip the start operation) | |
43316044 WVS |
132 | * WDOG_DEV_OPEN: this status bit shows whether or not the watchdog device |
133 | was opened via /dev/watchdog. | |
134 | (This bit should only be used by the WatchDog Timer Driver Core). | |
017cf080 WVS |
135 | * WDOG_ALLOW_RELEASE: this bit stores whether or not the magic close character |
136 | has been sent (so that we can support the magic close feature). | |
137 | (This bit should only be used by the WatchDog Timer Driver Core). | |
7e192b9c WVS |
138 | * WDOG_NO_WAY_OUT: this bit stores the nowayout setting for the watchdog. |
139 | If this bit is set then the watchdog timer will not be able to stop. | |
017cf080 | 140 | |
7e192b9c WVS |
141 | Note: The WatchDog Timer Driver Core supports the magic close feature and |
142 | the nowayout feature. To use the magic close feature you must set the | |
143 | WDIOF_MAGICCLOSE bit in the options field of the watchdog's info structure. | |
144 | The nowayout feature will overrule the magic close feature. | |
43316044 WVS |
145 | |
146 | To get or set driver specific data the following two helper functions should be | |
147 | used: | |
148 | ||
149 | static inline void watchdog_set_drvdata(struct watchdog_device *wdd, void *data) | |
150 | static inline void *watchdog_get_drvdata(struct watchdog_device *wdd) | |
151 | ||
152 | The watchdog_set_drvdata function allows you to add driver specific data. The | |
153 | arguments of this function are the watchdog device where you want to add the | |
154 | driver specific data to and a pointer to the data itself. | |
155 | ||
156 | The watchdog_get_drvdata function allows you to retrieve driver specific data. | |
157 | The argument of this function is the watchdog device where you want to retrieve | |
158 | data from. The function retruns the pointer to the driver specific data. |