Commit | Line | Data |
---|---|---|
f7ab97f7 OH |
1 | ============================================================================ |
2 | ||
3 | can.txt | |
4 | ||
5 | Readme file for the Controller Area Network Protocol Family (aka Socket CAN) | |
6 | ||
7 | This file contains | |
8 | ||
9 | 1 Overview / What is Socket CAN | |
10 | ||
11 | 2 Motivation / Why using the socket API | |
12 | ||
13 | 3 Socket CAN concept | |
14 | 3.1 receive lists | |
15 | 3.2 local loopback of sent frames | |
16 | 3.3 network security issues (capabilities) | |
17 | 3.4 network problem notifications | |
18 | ||
19 | 4 How to use Socket CAN | |
20 | 4.1 RAW protocol sockets with can_filters (SOCK_RAW) | |
21 | 4.1.1 RAW socket option CAN_RAW_FILTER | |
22 | 4.1.2 RAW socket option CAN_RAW_ERR_FILTER | |
23 | 4.1.3 RAW socket option CAN_RAW_LOOPBACK | |
24 | 4.1.4 RAW socket option CAN_RAW_RECV_OWN_MSGS | |
1e55659c | 25 | 4.1.5 RAW socket returned message flags |
f7ab97f7 OH |
26 | 4.2 Broadcast Manager protocol sockets (SOCK_DGRAM) |
27 | 4.3 connected transport protocols (SOCK_SEQPACKET) | |
28 | 4.4 unconnected transport protocols (SOCK_DGRAM) | |
29 | ||
30 | 5 Socket CAN core module | |
31 | 5.1 can.ko module params | |
32 | 5.2 procfs content | |
33 | 5.3 writing own CAN protocol modules | |
34 | ||
35 | 6 CAN network drivers | |
36 | 6.1 general settings | |
37 | 6.2 local loopback of sent frames | |
38 | 6.3 CAN controller hardware filters | |
e5d23048 | 39 | 6.4 The virtual CAN driver (vcan) |
e20dad96 WG |
40 | 6.5 The CAN network device driver interface |
41 | 6.5.1 Netlink interface to set/get devices properties | |
42 | 6.5.2 Setting the CAN bit-timing | |
43 | 6.5.3 Starting and stopping the CAN network device | |
44 | 6.6 supported CAN hardware | |
f7ab97f7 | 45 | |
e20dad96 WG |
46 | 7 Socket CAN resources |
47 | ||
48 | 8 Credits | |
f7ab97f7 OH |
49 | |
50 | ============================================================================ | |
51 | ||
52 | 1. Overview / What is Socket CAN | |
53 | -------------------------------- | |
54 | ||
55 | The socketcan package is an implementation of CAN protocols | |
56 | (Controller Area Network) for Linux. CAN is a networking technology | |
57 | which has widespread use in automation, embedded devices, and | |
58 | automotive fields. While there have been other CAN implementations | |
59 | for Linux based on character devices, Socket CAN uses the Berkeley | |
60 | socket API, the Linux network stack and implements the CAN device | |
61 | drivers as network interfaces. The CAN socket API has been designed | |
62 | as similar as possible to the TCP/IP protocols to allow programmers, | |
63 | familiar with network programming, to easily learn how to use CAN | |
64 | sockets. | |
65 | ||
66 | 2. Motivation / Why using the socket API | |
67 | ---------------------------------------- | |
68 | ||
69 | There have been CAN implementations for Linux before Socket CAN so the | |
70 | question arises, why we have started another project. Most existing | |
71 | implementations come as a device driver for some CAN hardware, they | |
72 | are based on character devices and provide comparatively little | |
73 | functionality. Usually, there is only a hardware-specific device | |
74 | driver which provides a character device interface to send and | |
75 | receive raw CAN frames, directly to/from the controller hardware. | |
76 | Queueing of frames and higher-level transport protocols like ISO-TP | |
77 | have to be implemented in user space applications. Also, most | |
78 | character-device implementations support only one single process to | |
79 | open the device at a time, similar to a serial interface. Exchanging | |
80 | the CAN controller requires employment of another device driver and | |
81 | often the need for adaption of large parts of the application to the | |
82 | new driver's API. | |
83 | ||
84 | Socket CAN was designed to overcome all of these limitations. A new | |
85 | protocol family has been implemented which provides a socket interface | |
86 | to user space applications and which builds upon the Linux network | |
87 | layer, so to use all of the provided queueing functionality. A device | |
88 | driver for CAN controller hardware registers itself with the Linux | |
89 | network layer as a network device, so that CAN frames from the | |
90 | controller can be passed up to the network layer and on to the CAN | |
91 | protocol family module and also vice-versa. Also, the protocol family | |
92 | module provides an API for transport protocol modules to register, so | |
93 | that any number of transport protocols can be loaded or unloaded | |
94 | dynamically. In fact, the can core module alone does not provide any | |
95 | protocol and cannot be used without loading at least one additional | |
96 | protocol module. Multiple sockets can be opened at the same time, | |
97 | on different or the same protocol module and they can listen/send | |
98 | frames on different or the same CAN IDs. Several sockets listening on | |
99 | the same interface for frames with the same CAN ID are all passed the | |
100 | same received matching CAN frames. An application wishing to | |
101 | communicate using a specific transport protocol, e.g. ISO-TP, just | |
102 | selects that protocol when opening the socket, and then can read and | |
103 | write application data byte streams, without having to deal with | |
104 | CAN-IDs, frames, etc. | |
105 | ||
106 | Similar functionality visible from user-space could be provided by a | |
107 | character device, too, but this would lead to a technically inelegant | |
108 | solution for a couple of reasons: | |
109 | ||
110 | * Intricate usage. Instead of passing a protocol argument to | |
111 | socket(2) and using bind(2) to select a CAN interface and CAN ID, an | |
112 | application would have to do all these operations using ioctl(2)s. | |
113 | ||
114 | * Code duplication. A character device cannot make use of the Linux | |
115 | network queueing code, so all that code would have to be duplicated | |
116 | for CAN networking. | |
117 | ||
118 | * Abstraction. In most existing character-device implementations, the | |
119 | hardware-specific device driver for a CAN controller directly | |
120 | provides the character device for the application to work with. | |
121 | This is at least very unusual in Unix systems for both, char and | |
122 | block devices. For example you don't have a character device for a | |
123 | certain UART of a serial interface, a certain sound chip in your | |
124 | computer, a SCSI or IDE controller providing access to your hard | |
125 | disk or tape streamer device. Instead, you have abstraction layers | |
126 | which provide a unified character or block device interface to the | |
127 | application on the one hand, and a interface for hardware-specific | |
128 | device drivers on the other hand. These abstractions are provided | |
129 | by subsystems like the tty layer, the audio subsystem or the SCSI | |
130 | and IDE subsystems for the devices mentioned above. | |
131 | ||
132 | The easiest way to implement a CAN device driver is as a character | |
133 | device without such a (complete) abstraction layer, as is done by most | |
134 | existing drivers. The right way, however, would be to add such a | |
135 | layer with all the functionality like registering for certain CAN | |
136 | IDs, supporting several open file descriptors and (de)multiplexing | |
137 | CAN frames between them, (sophisticated) queueing of CAN frames, and | |
138 | providing an API for device drivers to register with. However, then | |
139 | it would be no more difficult, or may be even easier, to use the | |
140 | networking framework provided by the Linux kernel, and this is what | |
141 | Socket CAN does. | |
142 | ||
143 | The use of the networking framework of the Linux kernel is just the | |
144 | natural and most appropriate way to implement CAN for Linux. | |
145 | ||
146 | 3. Socket CAN concept | |
147 | --------------------- | |
148 | ||
149 | As described in chapter 2 it is the main goal of Socket CAN to | |
150 | provide a socket interface to user space applications which builds | |
151 | upon the Linux network layer. In contrast to the commonly known | |
152 | TCP/IP and ethernet networking, the CAN bus is a broadcast-only(!) | |
153 | medium that has no MAC-layer addressing like ethernet. The CAN-identifier | |
154 | (can_id) is used for arbitration on the CAN-bus. Therefore the CAN-IDs | |
155 | have to be chosen uniquely on the bus. When designing a CAN-ECU | |
156 | network the CAN-IDs are mapped to be sent by a specific ECU. | |
157 | For this reason a CAN-ID can be treated best as a kind of source address. | |
158 | ||
159 | 3.1 receive lists | |
160 | ||
161 | The network transparent access of multiple applications leads to the | |
162 | problem that different applications may be interested in the same | |
163 | CAN-IDs from the same CAN network interface. The Socket CAN core | |
164 | module - which implements the protocol family CAN - provides several | |
165 | high efficient receive lists for this reason. If e.g. a user space | |
166 | application opens a CAN RAW socket, the raw protocol module itself | |
167 | requests the (range of) CAN-IDs from the Socket CAN core that are | |
168 | requested by the user. The subscription and unsubscription of | |
169 | CAN-IDs can be done for specific CAN interfaces or for all(!) known | |
170 | CAN interfaces with the can_rx_(un)register() functions provided to | |
171 | CAN protocol modules by the SocketCAN core (see chapter 5). | |
172 | To optimize the CPU usage at runtime the receive lists are split up | |
173 | into several specific lists per device that match the requested | |
174 | filter complexity for a given use-case. | |
175 | ||
176 | 3.2 local loopback of sent frames | |
177 | ||
178 | As known from other networking concepts the data exchanging | |
179 | applications may run on the same or different nodes without any | |
180 | change (except for the according addressing information): | |
181 | ||
182 | ___ ___ ___ _______ ___ | |
183 | | _ | | _ | | _ | | _ _ | | _ | | |
184 | ||A|| ||B|| ||C|| ||A| |B|| ||C|| | |
185 | |___| |___| |___| |_______| |___| | |
186 | | | | | | | |
187 | -----------------(1)- CAN bus -(2)--------------- | |
188 | ||
189 | To ensure that application A receives the same information in the | |
190 | example (2) as it would receive in example (1) there is need for | |
191 | some kind of local loopback of the sent CAN frames on the appropriate | |
192 | node. | |
193 | ||
194 | The Linux network devices (by default) just can handle the | |
195 | transmission and reception of media dependent frames. Due to the | |
d9195881 | 196 | arbitration on the CAN bus the transmission of a low prio CAN-ID |
f7ab97f7 OH |
197 | may be delayed by the reception of a high prio CAN frame. To |
198 | reflect the correct* traffic on the node the loopback of the sent | |
199 | data has to be performed right after a successful transmission. If | |
200 | the CAN network interface is not capable of performing the loopback for | |
201 | some reason the SocketCAN core can do this task as a fallback solution. | |
202 | See chapter 6.2 for details (recommended). | |
203 | ||
204 | The loopback functionality is enabled by default to reflect standard | |
205 | networking behaviour for CAN applications. Due to some requests from | |
206 | the RT-SocketCAN group the loopback optionally may be disabled for each | |
207 | separate socket. See sockopts from the CAN RAW sockets in chapter 4.1. | |
208 | ||
209 | * = you really like to have this when you're running analyser tools | |
210 | like 'candump' or 'cansniffer' on the (same) node. | |
211 | ||
212 | 3.3 network security issues (capabilities) | |
213 | ||
214 | The Controller Area Network is a local field bus transmitting only | |
215 | broadcast messages without any routing and security concepts. | |
216 | In the majority of cases the user application has to deal with | |
217 | raw CAN frames. Therefore it might be reasonable NOT to restrict | |
218 | the CAN access only to the user root, as known from other networks. | |
219 | Since the currently implemented CAN_RAW and CAN_BCM sockets can only | |
220 | send and receive frames to/from CAN interfaces it does not affect | |
221 | security of others networks to allow all users to access the CAN. | |
222 | To enable non-root users to access CAN_RAW and CAN_BCM protocol | |
223 | sockets the Kconfig options CAN_RAW_USER and/or CAN_BCM_USER may be | |
224 | selected at kernel compile time. | |
225 | ||
226 | 3.4 network problem notifications | |
227 | ||
228 | The use of the CAN bus may lead to several problems on the physical | |
229 | and media access control layer. Detecting and logging of these lower | |
230 | layer problems is a vital requirement for CAN users to identify | |
231 | hardware issues on the physical transceiver layer as well as | |
232 | arbitration problems and error frames caused by the different | |
233 | ECUs. The occurrence of detected errors are important for diagnosis | |
234 | and have to be logged together with the exact timestamp. For this | |
235 | reason the CAN interface driver can generate so called Error Frames | |
236 | that can optionally be passed to the user application in the same | |
237 | way as other CAN frames. Whenever an error on the physical layer | |
238 | or the MAC layer is detected (e.g. by the CAN controller) the driver | |
239 | creates an appropriate error frame. Error frames can be requested by | |
240 | the user application using the common CAN filter mechanisms. Inside | |
241 | this filter definition the (interested) type of errors may be | |
242 | selected. The reception of error frames is disabled by default. | |
e20dad96 WG |
243 | The format of the CAN error frame is briefly decribed in the Linux |
244 | header file "include/linux/can/error.h". | |
f7ab97f7 OH |
245 | |
246 | 4. How to use Socket CAN | |
247 | ------------------------ | |
248 | ||
249 | Like TCP/IP, you first need to open a socket for communicating over a | |
250 | CAN network. Since Socket CAN implements a new protocol family, you | |
251 | need to pass PF_CAN as the first argument to the socket(2) system | |
252 | call. Currently, there are two CAN protocols to choose from, the raw | |
253 | socket protocol and the broadcast manager (BCM). So to open a socket, | |
254 | you would write | |
255 | ||
256 | s = socket(PF_CAN, SOCK_RAW, CAN_RAW); | |
257 | ||
258 | and | |
259 | ||
260 | s = socket(PF_CAN, SOCK_DGRAM, CAN_BCM); | |
261 | ||
262 | respectively. After the successful creation of the socket, you would | |
263 | normally use the bind(2) system call to bind the socket to a CAN | |
264 | interface (which is different from TCP/IP due to different addressing | |
265 | - see chapter 3). After binding (CAN_RAW) or connecting (CAN_BCM) | |
266 | the socket, you can read(2) and write(2) from/to the socket or use | |
267 | send(2), sendto(2), sendmsg(2) and the recv* counterpart operations | |
268 | on the socket as usual. There are also CAN specific socket options | |
269 | described below. | |
270 | ||
271 | The basic CAN frame structure and the sockaddr structure are defined | |
272 | in include/linux/can.h: | |
273 | ||
274 | struct can_frame { | |
275 | canid_t can_id; /* 32 bit CAN_ID + EFF/RTR/ERR flags */ | |
276 | __u8 can_dlc; /* data length code: 0 .. 8 */ | |
277 | __u8 data[8] __attribute__((aligned(8))); | |
278 | }; | |
279 | ||
280 | The alignment of the (linear) payload data[] to a 64bit boundary | |
281 | allows the user to define own structs and unions to easily access the | |
282 | CAN payload. There is no given byteorder on the CAN bus by | |
283 | default. A read(2) system call on a CAN_RAW socket transfers a | |
284 | struct can_frame to the user space. | |
285 | ||
286 | The sockaddr_can structure has an interface index like the | |
287 | PF_PACKET socket, that also binds to a specific interface: | |
288 | ||
289 | struct sockaddr_can { | |
290 | sa_family_t can_family; | |
291 | int can_ifindex; | |
292 | union { | |
56690c21 OH |
293 | /* transport protocol class address info (e.g. ISOTP) */ |
294 | struct { canid_t rx_id, tx_id; } tp; | |
295 | ||
296 | /* reserved for future CAN protocols address information */ | |
f7ab97f7 OH |
297 | } can_addr; |
298 | }; | |
299 | ||
300 | To determine the interface index an appropriate ioctl() has to | |
301 | be used (example for CAN_RAW sockets without error checking): | |
302 | ||
303 | int s; | |
304 | struct sockaddr_can addr; | |
305 | struct ifreq ifr; | |
306 | ||
307 | s = socket(PF_CAN, SOCK_RAW, CAN_RAW); | |
308 | ||
309 | strcpy(ifr.ifr_name, "can0" ); | |
310 | ioctl(s, SIOCGIFINDEX, &ifr); | |
311 | ||
312 | addr.can_family = AF_CAN; | |
313 | addr.can_ifindex = ifr.ifr_ifindex; | |
314 | ||
315 | bind(s, (struct sockaddr *)&addr, sizeof(addr)); | |
316 | ||
317 | (..) | |
318 | ||
319 | To bind a socket to all(!) CAN interfaces the interface index must | |
320 | be 0 (zero). In this case the socket receives CAN frames from every | |
321 | enabled CAN interface. To determine the originating CAN interface | |
322 | the system call recvfrom(2) may be used instead of read(2). To send | |
323 | on a socket that is bound to 'any' interface sendto(2) is needed to | |
324 | specify the outgoing interface. | |
325 | ||
326 | Reading CAN frames from a bound CAN_RAW socket (see above) consists | |
327 | of reading a struct can_frame: | |
328 | ||
329 | struct can_frame frame; | |
330 | ||
331 | nbytes = read(s, &frame, sizeof(struct can_frame)); | |
332 | ||
333 | if (nbytes < 0) { | |
334 | perror("can raw socket read"); | |
335 | return 1; | |
336 | } | |
337 | ||
19f59460 | 338 | /* paranoid check ... */ |
f7ab97f7 OH |
339 | if (nbytes < sizeof(struct can_frame)) { |
340 | fprintf(stderr, "read: incomplete CAN frame\n"); | |
341 | return 1; | |
342 | } | |
343 | ||
344 | /* do something with the received CAN frame */ | |
345 | ||
346 | Writing CAN frames can be done similarly, with the write(2) system call: | |
347 | ||
348 | nbytes = write(s, &frame, sizeof(struct can_frame)); | |
349 | ||
350 | When the CAN interface is bound to 'any' existing CAN interface | |
351 | (addr.can_ifindex = 0) it is recommended to use recvfrom(2) if the | |
352 | information about the originating CAN interface is needed: | |
353 | ||
354 | struct sockaddr_can addr; | |
355 | struct ifreq ifr; | |
356 | socklen_t len = sizeof(addr); | |
357 | struct can_frame frame; | |
358 | ||
359 | nbytes = recvfrom(s, &frame, sizeof(struct can_frame), | |
360 | 0, (struct sockaddr*)&addr, &len); | |
361 | ||
362 | /* get interface name of the received CAN frame */ | |
363 | ifr.ifr_ifindex = addr.can_ifindex; | |
364 | ioctl(s, SIOCGIFNAME, &ifr); | |
365 | printf("Received a CAN frame from interface %s", ifr.ifr_name); | |
366 | ||
367 | To write CAN frames on sockets bound to 'any' CAN interface the | |
368 | outgoing interface has to be defined certainly. | |
369 | ||
370 | strcpy(ifr.ifr_name, "can0"); | |
371 | ioctl(s, SIOCGIFINDEX, &ifr); | |
372 | addr.can_ifindex = ifr.ifr_ifindex; | |
373 | addr.can_family = AF_CAN; | |
374 | ||
375 | nbytes = sendto(s, &frame, sizeof(struct can_frame), | |
376 | 0, (struct sockaddr*)&addr, sizeof(addr)); | |
377 | ||
378 | 4.1 RAW protocol sockets with can_filters (SOCK_RAW) | |
379 | ||
380 | Using CAN_RAW sockets is extensively comparable to the commonly | |
381 | known access to CAN character devices. To meet the new possibilities | |
382 | provided by the multi user SocketCAN approach, some reasonable | |
383 | defaults are set at RAW socket binding time: | |
384 | ||
385 | - The filters are set to exactly one filter receiving everything | |
386 | - The socket only receives valid data frames (=> no error frames) | |
387 | - The loopback of sent CAN frames is enabled (see chapter 3.2) | |
388 | - The socket does not receive its own sent frames (in loopback mode) | |
389 | ||
390 | These default settings may be changed before or after binding the socket. | |
391 | To use the referenced definitions of the socket options for CAN_RAW | |
392 | sockets, include <linux/can/raw.h>. | |
393 | ||
394 | 4.1.1 RAW socket option CAN_RAW_FILTER | |
395 | ||
396 | The reception of CAN frames using CAN_RAW sockets can be controlled | |
397 | by defining 0 .. n filters with the CAN_RAW_FILTER socket option. | |
398 | ||
399 | The CAN filter structure is defined in include/linux/can.h: | |
400 | ||
401 | struct can_filter { | |
402 | canid_t can_id; | |
403 | canid_t can_mask; | |
404 | }; | |
405 | ||
406 | A filter matches, when | |
407 | ||
408 | <received_can_id> & mask == can_id & mask | |
409 | ||
410 | which is analogous to known CAN controllers hardware filter semantics. | |
411 | The filter can be inverted in this semantic, when the CAN_INV_FILTER | |
412 | bit is set in can_id element of the can_filter structure. In | |
413 | contrast to CAN controller hardware filters the user may set 0 .. n | |
414 | receive filters for each open socket separately: | |
415 | ||
416 | struct can_filter rfilter[2]; | |
417 | ||
418 | rfilter[0].can_id = 0x123; | |
419 | rfilter[0].can_mask = CAN_SFF_MASK; | |
420 | rfilter[1].can_id = 0x200; | |
421 | rfilter[1].can_mask = 0x700; | |
422 | ||
423 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_FILTER, &rfilter, sizeof(rfilter)); | |
424 | ||
425 | To disable the reception of CAN frames on the selected CAN_RAW socket: | |
426 | ||
427 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_FILTER, NULL, 0); | |
428 | ||
429 | To set the filters to zero filters is quite obsolete as not read | |
430 | data causes the raw socket to discard the received CAN frames. But | |
431 | having this 'send only' use-case we may remove the receive list in the | |
432 | Kernel to save a little (really a very little!) CPU usage. | |
433 | ||
434 | 4.1.2 RAW socket option CAN_RAW_ERR_FILTER | |
435 | ||
436 | As described in chapter 3.4 the CAN interface driver can generate so | |
437 | called Error Frames that can optionally be passed to the user | |
438 | application in the same way as other CAN frames. The possible | |
439 | errors are divided into different error classes that may be filtered | |
440 | using the appropriate error mask. To register for every possible | |
441 | error condition CAN_ERR_MASK can be used as value for the error mask. | |
442 | The values for the error mask are defined in linux/can/error.h . | |
443 | ||
444 | can_err_mask_t err_mask = ( CAN_ERR_TX_TIMEOUT | CAN_ERR_BUSOFF ); | |
445 | ||
446 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_ERR_FILTER, | |
447 | &err_mask, sizeof(err_mask)); | |
448 | ||
449 | 4.1.3 RAW socket option CAN_RAW_LOOPBACK | |
450 | ||
451 | To meet multi user needs the local loopback is enabled by default | |
452 | (see chapter 3.2 for details). But in some embedded use-cases | |
453 | (e.g. when only one application uses the CAN bus) this loopback | |
454 | functionality can be disabled (separately for each socket): | |
455 | ||
456 | int loopback = 0; /* 0 = disabled, 1 = enabled (default) */ | |
457 | ||
458 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_LOOPBACK, &loopback, sizeof(loopback)); | |
459 | ||
460 | 4.1.4 RAW socket option CAN_RAW_RECV_OWN_MSGS | |
461 | ||
462 | When the local loopback is enabled, all the sent CAN frames are | |
463 | looped back to the open CAN sockets that registered for the CAN | |
464 | frames' CAN-ID on this given interface to meet the multi user | |
465 | needs. The reception of the CAN frames on the same socket that was | |
466 | sending the CAN frame is assumed to be unwanted and therefore | |
467 | disabled by default. This default behaviour may be changed on | |
468 | demand: | |
469 | ||
470 | int recv_own_msgs = 1; /* 0 = disabled (default), 1 = enabled */ | |
471 | ||
472 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_RECV_OWN_MSGS, | |
473 | &recv_own_msgs, sizeof(recv_own_msgs)); | |
474 | ||
1e55659c OH |
475 | 4.1.5 RAW socket returned message flags |
476 | ||
477 | When using recvmsg() call, the msg->msg_flags may contain following flags: | |
478 | ||
479 | MSG_DONTROUTE: set when the received frame was created on the local host. | |
480 | ||
481 | MSG_CONFIRM: set when the frame was sent via the socket it is received on. | |
482 | This flag can be interpreted as a 'transmission confirmation' when the | |
483 | CAN driver supports the echo of frames on driver level, see 3.2 and 6.2. | |
484 | In order to receive such messages, CAN_RAW_RECV_OWN_MSGS must be set. | |
485 | ||
f7ab97f7 OH |
486 | 4.2 Broadcast Manager protocol sockets (SOCK_DGRAM) |
487 | 4.3 connected transport protocols (SOCK_SEQPACKET) | |
488 | 4.4 unconnected transport protocols (SOCK_DGRAM) | |
489 | ||
490 | ||
491 | 5. Socket CAN core module | |
492 | ------------------------- | |
493 | ||
494 | The Socket CAN core module implements the protocol family | |
495 | PF_CAN. CAN protocol modules are loaded by the core module at | |
496 | runtime. The core module provides an interface for CAN protocol | |
497 | modules to subscribe needed CAN IDs (see chapter 3.1). | |
498 | ||
499 | 5.1 can.ko module params | |
500 | ||
501 | - stats_timer: To calculate the Socket CAN core statistics | |
502 | (e.g. current/maximum frames per second) this 1 second timer is | |
503 | invoked at can.ko module start time by default. This timer can be | |
d9195881 | 504 | disabled by using stattimer=0 on the module commandline. |
f7ab97f7 OH |
505 | |
506 | - debug: (removed since SocketCAN SVN r546) | |
507 | ||
508 | 5.2 procfs content | |
509 | ||
510 | As described in chapter 3.1 the Socket CAN core uses several filter | |
511 | lists to deliver received CAN frames to CAN protocol modules. These | |
512 | receive lists, their filters and the count of filter matches can be | |
513 | checked in the appropriate receive list. All entries contain the | |
514 | device and a protocol module identifier: | |
515 | ||
516 | foo@bar:~$ cat /proc/net/can/rcvlist_all | |
517 | ||
518 | receive list 'rx_all': | |
519 | (vcan3: no entry) | |
520 | (vcan2: no entry) | |
521 | (vcan1: no entry) | |
522 | device can_id can_mask function userdata matches ident | |
523 | vcan0 000 00000000 f88e6370 f6c6f400 0 raw | |
524 | (any: no entry) | |
525 | ||
526 | In this example an application requests any CAN traffic from vcan0. | |
527 | ||
528 | rcvlist_all - list for unfiltered entries (no filter operations) | |
529 | rcvlist_eff - list for single extended frame (EFF) entries | |
530 | rcvlist_err - list for error frames masks | |
531 | rcvlist_fil - list for mask/value filters | |
532 | rcvlist_inv - list for mask/value filters (inverse semantic) | |
533 | rcvlist_sff - list for single standard frame (SFF) entries | |
534 | ||
535 | Additional procfs files in /proc/net/can | |
536 | ||
537 | stats - Socket CAN core statistics (rx/tx frames, match ratios, ...) | |
538 | reset_stats - manual statistic reset | |
539 | version - prints the Socket CAN core version and the ABI version | |
540 | ||
541 | 5.3 writing own CAN protocol modules | |
542 | ||
543 | To implement a new protocol in the protocol family PF_CAN a new | |
544 | protocol has to be defined in include/linux/can.h . | |
545 | The prototypes and definitions to use the Socket CAN core can be | |
546 | accessed by including include/linux/can/core.h . | |
547 | In addition to functions that register the CAN protocol and the | |
548 | CAN device notifier chain there are functions to subscribe CAN | |
549 | frames received by CAN interfaces and to send CAN frames: | |
550 | ||
551 | can_rx_register - subscribe CAN frames from a specific interface | |
552 | can_rx_unregister - unsubscribe CAN frames from a specific interface | |
553 | can_send - transmit a CAN frame (optional with local loopback) | |
554 | ||
555 | For details see the kerneldoc documentation in net/can/af_can.c or | |
556 | the source code of net/can/raw.c or net/can/bcm.c . | |
557 | ||
558 | 6. CAN network drivers | |
559 | ---------------------- | |
560 | ||
561 | Writing a CAN network device driver is much easier than writing a | |
562 | CAN character device driver. Similar to other known network device | |
563 | drivers you mainly have to deal with: | |
564 | ||
565 | - TX: Put the CAN frame from the socket buffer to the CAN controller. | |
566 | - RX: Put the CAN frame from the CAN controller to the socket buffer. | |
567 | ||
568 | See e.g. at Documentation/networking/netdevices.txt . The differences | |
569 | for writing CAN network device driver are described below: | |
570 | ||
571 | 6.1 general settings | |
572 | ||
573 | dev->type = ARPHRD_CAN; /* the netdevice hardware type */ | |
574 | dev->flags = IFF_NOARP; /* CAN has no arp */ | |
575 | ||
576 | dev->mtu = sizeof(struct can_frame); | |
577 | ||
578 | The struct can_frame is the payload of each socket buffer in the | |
579 | protocol family PF_CAN. | |
580 | ||
581 | 6.2 local loopback of sent frames | |
582 | ||
583 | As described in chapter 3.2 the CAN network device driver should | |
584 | support a local loopback functionality similar to the local echo | |
585 | e.g. of tty devices. In this case the driver flag IFF_ECHO has to be | |
586 | set to prevent the PF_CAN core from locally echoing sent frames | |
587 | (aka loopback) as fallback solution: | |
588 | ||
589 | dev->flags = (IFF_NOARP | IFF_ECHO); | |
590 | ||
591 | 6.3 CAN controller hardware filters | |
592 | ||
593 | To reduce the interrupt load on deep embedded systems some CAN | |
594 | controllers support the filtering of CAN IDs or ranges of CAN IDs. | |
595 | These hardware filter capabilities vary from controller to | |
596 | controller and have to be identified as not feasible in a multi-user | |
597 | networking approach. The use of the very controller specific | |
598 | hardware filters could make sense in a very dedicated use-case, as a | |
599 | filter on driver level would affect all users in the multi-user | |
600 | system. The high efficient filter sets inside the PF_CAN core allow | |
601 | to set different multiple filters for each socket separately. | |
602 | Therefore the use of hardware filters goes to the category 'handmade | |
603 | tuning on deep embedded systems'. The author is running a MPC603e | |
604 | @133MHz with four SJA1000 CAN controllers from 2002 under heavy bus | |
605 | load without any problems ... | |
606 | ||
e5d23048 OH |
607 | 6.4 The virtual CAN driver (vcan) |
608 | ||
609 | Similar to the network loopback devices, vcan offers a virtual local | |
610 | CAN interface. A full qualified address on CAN consists of | |
611 | ||
612 | - a unique CAN Identifier (CAN ID) | |
613 | - the CAN bus this CAN ID is transmitted on (e.g. can0) | |
614 | ||
615 | so in common use cases more than one virtual CAN interface is needed. | |
616 | ||
617 | The virtual CAN interfaces allow the transmission and reception of CAN | |
618 | frames without real CAN controller hardware. Virtual CAN network | |
619 | devices are usually named 'vcanX', like vcan0 vcan1 vcan2 ... | |
620 | When compiled as a module the virtual CAN driver module is called vcan.ko | |
621 | ||
622 | Since Linux Kernel version 2.6.24 the vcan driver supports the Kernel | |
623 | netlink interface to create vcan network devices. The creation and | |
624 | removal of vcan network devices can be managed with the ip(8) tool: | |
625 | ||
626 | - Create a virtual CAN network interface: | |
e20dad96 | 627 | $ ip link add type vcan |
e5d23048 OH |
628 | |
629 | - Create a virtual CAN network interface with a specific name 'vcan42': | |
e20dad96 | 630 | $ ip link add dev vcan42 type vcan |
e5d23048 OH |
631 | |
632 | - Remove a (virtual CAN) network interface 'vcan42': | |
e20dad96 WG |
633 | $ ip link del vcan42 |
634 | ||
635 | 6.5 The CAN network device driver interface | |
636 | ||
637 | The CAN network device driver interface provides a generic interface | |
638 | to setup, configure and monitor CAN network devices. The user can then | |
639 | configure the CAN device, like setting the bit-timing parameters, via | |
640 | the netlink interface using the program "ip" from the "IPROUTE2" | |
641 | utility suite. The following chapter describes briefly how to use it. | |
642 | Furthermore, the interface uses a common data structure and exports a | |
643 | set of common functions, which all real CAN network device drivers | |
644 | should use. Please have a look to the SJA1000 or MSCAN driver to | |
645 | understand how to use them. The name of the module is can-dev.ko. | |
646 | ||
647 | 6.5.1 Netlink interface to set/get devices properties | |
648 | ||
649 | The CAN device must be configured via netlink interface. The supported | |
650 | netlink message types are defined and briefly described in | |
651 | "include/linux/can/netlink.h". CAN link support for the program "ip" | |
652 | of the IPROUTE2 utility suite is avaiable and it can be used as shown | |
653 | below: | |
654 | ||
655 | - Setting CAN device properties: | |
656 | ||
657 | $ ip link set can0 type can help | |
658 | Usage: ip link set DEVICE type can | |
659 | [ bitrate BITRATE [ sample-point SAMPLE-POINT] ] | | |
660 | [ tq TQ prop-seg PROP_SEG phase-seg1 PHASE-SEG1 | |
661 | phase-seg2 PHASE-SEG2 [ sjw SJW ] ] | |
662 | ||
663 | [ loopback { on | off } ] | |
664 | [ listen-only { on | off } ] | |
665 | [ triple-sampling { on | off } ] | |
666 | ||
667 | [ restart-ms TIME-MS ] | |
668 | [ restart ] | |
669 | ||
670 | Where: BITRATE := { 1..1000000 } | |
671 | SAMPLE-POINT := { 0.000..0.999 } | |
672 | TQ := { NUMBER } | |
673 | PROP-SEG := { 1..8 } | |
674 | PHASE-SEG1 := { 1..8 } | |
675 | PHASE-SEG2 := { 1..8 } | |
676 | SJW := { 1..4 } | |
677 | RESTART-MS := { 0 | NUMBER } | |
678 | ||
679 | - Display CAN device details and statistics: | |
680 | ||
681 | $ ip -details -statistics link show can0 | |
682 | 2: can0: <NOARP,UP,LOWER_UP,ECHO> mtu 16 qdisc pfifo_fast state UP qlen 10 | |
683 | link/can | |
684 | can <TRIPLE-SAMPLING> state ERROR-ACTIVE restart-ms 100 | |
685 | bitrate 125000 sample_point 0.875 | |
686 | tq 125 prop-seg 6 phase-seg1 7 phase-seg2 2 sjw 1 | |
687 | sja1000: tseg1 1..16 tseg2 1..8 sjw 1..4 brp 1..64 brp-inc 1 | |
688 | clock 8000000 | |
689 | re-started bus-errors arbit-lost error-warn error-pass bus-off | |
690 | 41 17457 0 41 42 41 | |
691 | RX: bytes packets errors dropped overrun mcast | |
692 | 140859 17608 17457 0 0 0 | |
693 | TX: bytes packets errors dropped carrier collsns | |
694 | 861 112 0 41 0 0 | |
695 | ||
696 | More info to the above output: | |
697 | ||
698 | "<TRIPLE-SAMPLING>" | |
699 | Shows the list of selected CAN controller modes: LOOPBACK, | |
700 | LISTEN-ONLY, or TRIPLE-SAMPLING. | |
701 | ||
702 | "state ERROR-ACTIVE" | |
703 | The current state of the CAN controller: "ERROR-ACTIVE", | |
704 | "ERROR-WARNING", "ERROR-PASSIVE", "BUS-OFF" or "STOPPED" | |
705 | ||
706 | "restart-ms 100" | |
707 | Automatic restart delay time. If set to a non-zero value, a | |
708 | restart of the CAN controller will be triggered automatically | |
709 | in case of a bus-off condition after the specified delay time | |
710 | in milliseconds. By default it's off. | |
711 | ||
712 | "bitrate 125000 sample_point 0.875" | |
713 | Shows the real bit-rate in bits/sec and the sample-point in the | |
714 | range 0.000..0.999. If the calculation of bit-timing parameters | |
715 | is enabled in the kernel (CONFIG_CAN_CALC_BITTIMING=y), the | |
716 | bit-timing can be defined by setting the "bitrate" argument. | |
717 | Optionally the "sample-point" can be specified. By default it's | |
718 | 0.000 assuming CIA-recommended sample-points. | |
719 | ||
720 | "tq 125 prop-seg 6 phase-seg1 7 phase-seg2 2 sjw 1" | |
721 | Shows the time quanta in ns, propagation segment, phase buffer | |
722 | segment 1 and 2 and the synchronisation jump width in units of | |
723 | tq. They allow to define the CAN bit-timing in a hardware | |
724 | independent format as proposed by the Bosch CAN 2.0 spec (see | |
725 | chapter 8 of http://www.semiconductors.bosch.de/pdf/can2spec.pdf). | |
726 | ||
727 | "sja1000: tseg1 1..16 tseg2 1..8 sjw 1..4 brp 1..64 brp-inc 1 | |
728 | clock 8000000" | |
729 | Shows the bit-timing constants of the CAN controller, here the | |
730 | "sja1000". The minimum and maximum values of the time segment 1 | |
731 | and 2, the synchronisation jump width in units of tq, the | |
732 | bitrate pre-scaler and the CAN system clock frequency in Hz. | |
733 | These constants could be used for user-defined (non-standard) | |
734 | bit-timing calculation algorithms in user-space. | |
735 | ||
736 | "re-started bus-errors arbit-lost error-warn error-pass bus-off" | |
737 | Shows the number of restarts, bus and arbitration lost errors, | |
738 | and the state changes to the error-warning, error-passive and | |
739 | bus-off state. RX overrun errors are listed in the "overrun" | |
740 | field of the standard network statistics. | |
741 | ||
742 | 6.5.2 Setting the CAN bit-timing | |
743 | ||
744 | The CAN bit-timing parameters can always be defined in a hardware | |
745 | independent format as proposed in the Bosch CAN 2.0 specification | |
746 | specifying the arguments "tq", "prop_seg", "phase_seg1", "phase_seg2" | |
747 | and "sjw": | |
748 | ||
749 | $ ip link set canX type can tq 125 prop-seg 6 \ | |
750 | phase-seg1 7 phase-seg2 2 sjw 1 | |
751 | ||
752 | If the kernel option CONFIG_CAN_CALC_BITTIMING is enabled, CIA | |
753 | recommended CAN bit-timing parameters will be calculated if the bit- | |
754 | rate is specified with the argument "bitrate": | |
755 | ||
756 | $ ip link set canX type can bitrate 125000 | |
757 | ||
758 | Note that this works fine for the most common CAN controllers with | |
759 | standard bit-rates but may *fail* for exotic bit-rates or CAN system | |
760 | clock frequencies. Disabling CONFIG_CAN_CALC_BITTIMING saves some | |
761 | space and allows user-space tools to solely determine and set the | |
762 | bit-timing parameters. The CAN controller specific bit-timing | |
763 | constants can be used for that purpose. They are listed by the | |
764 | following command: | |
765 | ||
766 | $ ip -details link show can0 | |
767 | ... | |
768 | sja1000: clock 8000000 tseg1 1..16 tseg2 1..8 sjw 1..4 brp 1..64 brp-inc 1 | |
769 | ||
770 | 6.5.3 Starting and stopping the CAN network device | |
771 | ||
772 | A CAN network device is started or stopped as usual with the command | |
773 | "ifconfig canX up/down" or "ip link set canX up/down". Be aware that | |
774 | you *must* define proper bit-timing parameters for real CAN devices | |
775 | before you can start it to avoid error-prone default settings: | |
776 | ||
777 | $ ip link set canX up type can bitrate 125000 | |
778 | ||
779 | A device may enter the "bus-off" state if too much errors occurred on | |
780 | the CAN bus. Then no more messages are received or sent. An automatic | |
781 | bus-off recovery can be enabled by setting the "restart-ms" to a | |
782 | non-zero value, e.g.: | |
783 | ||
784 | $ ip link set canX type can restart-ms 100 | |
785 | ||
786 | Alternatively, the application may realize the "bus-off" condition | |
787 | by monitoring CAN error frames and do a restart when appropriate with | |
788 | the command: | |
789 | ||
790 | $ ip link set canX type can restart | |
791 | ||
792 | Note that a restart will also create a CAN error frame (see also | |
793 | chapter 3.4). | |
f7ab97f7 | 794 | |
e20dad96 | 795 | 6.6 Supported CAN hardware |
f7ab97f7 | 796 | |
e20dad96 WG |
797 | Please check the "Kconfig" file in "drivers/net/can" to get an actual |
798 | list of the support CAN hardware. On the Socket CAN project website | |
799 | (see chapter 7) there might be further drivers available, also for | |
800 | older kernel versions. | |
f7ab97f7 | 801 | |
e20dad96 WG |
802 | 7. Socket CAN resources |
803 | ----------------------- | |
f7ab97f7 | 804 | |
e20dad96 WG |
805 | You can find further resources for Socket CAN like user space tools, |
806 | support for old kernel versions, more drivers, mailing lists, etc. | |
807 | at the BerliOS OSS project website for Socket CAN: | |
f7ab97f7 | 808 | |
e20dad96 | 809 | http://developer.berlios.de/projects/socketcan |
f7ab97f7 | 810 | |
e20dad96 WG |
811 | If you have questions, bug fixes, etc., don't hesitate to post them to |
812 | the Socketcan-Users mailing list. But please search the archives first. | |
f7ab97f7 | 813 | |
e20dad96 | 814 | 8. Credits |
f7ab97f7 OH |
815 | ---------- |
816 | ||
e20dad96 | 817 | Oliver Hartkopp (PF_CAN core, filters, drivers, bcm, SJA1000 driver) |
f7ab97f7 OH |
818 | Urs Thuermann (PF_CAN core, kernel integration, socket interfaces, raw, vcan) |
819 | Jan Kizka (RT-SocketCAN core, Socket-API reconciliation) | |
e20dad96 WG |
820 | Wolfgang Grandegger (RT-SocketCAN core & drivers, Raw Socket-API reviews, |
821 | CAN device driver interface, MSCAN driver) | |
f7ab97f7 OH |
822 | Robert Schwebel (design reviews, PTXdist integration) |
823 | Marc Kleine-Budde (design reviews, Kernel 2.6 cleanups, drivers) | |
824 | Benedikt Spranger (reviews) | |
825 | Thomas Gleixner (LKML reviews, coding style, posting hints) | |
e20dad96 | 826 | Andrey Volkov (kernel subtree structure, ioctls, MSCAN driver) |
f7ab97f7 OH |
827 | Matthias Brukner (first SJA1000 CAN netdevice implementation Q2/2003) |
828 | Klaus Hitschler (PEAK driver integration) | |
829 | Uwe Koppe (CAN netdevices with PF_PACKET approach) | |
830 | Michael Schulze (driver layer loopback requirement, RT CAN drivers review) | |
e20dad96 WG |
831 | Pavel Pisa (Bit-timing calculation) |
832 | Sascha Hauer (SJA1000 platform driver) | |
833 | Sebastian Haas (SJA1000 EMS PCI driver) | |
834 | Markus Plessing (SJA1000 EMS PCI driver) | |
835 | Per Dalen (SJA1000 Kvaser PCI driver) | |
836 | Sam Ravnborg (reviews, coding style, kbuild help) |