Commit | Line | Data |
---|---|---|
f7ab97f7 OH |
1 | ============================================================================ |
2 | ||
3 | can.txt | |
4 | ||
f35f6c8f | 5 | Readme file for the Controller Area Network Protocol Family (aka SocketCAN) |
f7ab97f7 OH |
6 | |
7 | This file contains | |
8 | ||
f35f6c8f | 9 | 1 Overview / What is SocketCAN |
f7ab97f7 OH |
10 | |
11 | 2 Motivation / Why using the socket API | |
12 | ||
f35f6c8f | 13 | 3 SocketCAN concept |
f7ab97f7 OH |
14 | 3.1 receive lists |
15 | 3.2 local loopback of sent frames | |
f35f6c8f | 16 | 3.3 network problem notifications |
f7ab97f7 | 17 | |
f35f6c8f | 18 | 4 How to use SocketCAN |
f7ab97f7 OH |
19 | 4.1 RAW protocol sockets with can_filters (SOCK_RAW) |
20 | 4.1.1 RAW socket option CAN_RAW_FILTER | |
21 | 4.1.2 RAW socket option CAN_RAW_ERR_FILTER | |
22 | 4.1.3 RAW socket option CAN_RAW_LOOPBACK | |
23 | 4.1.4 RAW socket option CAN_RAW_RECV_OWN_MSGS | |
ea53fe0c OH |
24 | 4.1.5 RAW socket option CAN_RAW_FD_FRAMES |
25 | 4.1.6 RAW socket returned message flags | |
f7ab97f7 | 26 | 4.2 Broadcast Manager protocol sockets (SOCK_DGRAM) |
51b2f451 OH |
27 | 4.2.1 Broadcast Manager operations |
28 | 4.2.2 Broadcast Manager message flags | |
29 | 4.2.3 Broadcast Manager transmission timers | |
30 | 4.2.4 Broadcast Manager message sequence transmission | |
31 | 4.2.5 Broadcast Manager receive filter timers | |
32 | 4.2.6 Broadcast Manager multiplex message receive filter | |
f7ab97f7 OH |
33 | 4.3 connected transport protocols (SOCK_SEQPACKET) |
34 | 4.4 unconnected transport protocols (SOCK_DGRAM) | |
35 | ||
f35f6c8f | 36 | 5 SocketCAN core module |
f7ab97f7 OH |
37 | 5.1 can.ko module params |
38 | 5.2 procfs content | |
39 | 5.3 writing own CAN protocol modules | |
40 | ||
41 | 6 CAN network drivers | |
42 | 6.1 general settings | |
43 | 6.2 local loopback of sent frames | |
44 | 6.3 CAN controller hardware filters | |
e5d23048 | 45 | 6.4 The virtual CAN driver (vcan) |
e20dad96 WG |
46 | 6.5 The CAN network device driver interface |
47 | 6.5.1 Netlink interface to set/get devices properties | |
48 | 6.5.2 Setting the CAN bit-timing | |
49 | 6.5.3 Starting and stopping the CAN network device | |
ea53fe0c OH |
50 | 6.6 CAN FD (flexible data rate) driver support |
51 | 6.7 supported CAN hardware | |
f7ab97f7 | 52 | |
f35f6c8f | 53 | 7 SocketCAN resources |
e20dad96 WG |
54 | |
55 | 8 Credits | |
f7ab97f7 OH |
56 | |
57 | ============================================================================ | |
58 | ||
f35f6c8f | 59 | 1. Overview / What is SocketCAN |
f7ab97f7 OH |
60 | -------------------------------- |
61 | ||
62 | The socketcan package is an implementation of CAN protocols | |
63 | (Controller Area Network) for Linux. CAN is a networking technology | |
64 | which has widespread use in automation, embedded devices, and | |
65 | automotive fields. While there have been other CAN implementations | |
f35f6c8f | 66 | for Linux based on character devices, SocketCAN uses the Berkeley |
f7ab97f7 OH |
67 | socket API, the Linux network stack and implements the CAN device |
68 | drivers as network interfaces. The CAN socket API has been designed | |
69 | as similar as possible to the TCP/IP protocols to allow programmers, | |
70 | familiar with network programming, to easily learn how to use CAN | |
71 | sockets. | |
72 | ||
73 | 2. Motivation / Why using the socket API | |
74 | ---------------------------------------- | |
75 | ||
f35f6c8f | 76 | There have been CAN implementations for Linux before SocketCAN so the |
f7ab97f7 OH |
77 | question arises, why we have started another project. Most existing |
78 | implementations come as a device driver for some CAN hardware, they | |
79 | are based on character devices and provide comparatively little | |
80 | functionality. Usually, there is only a hardware-specific device | |
81 | driver which provides a character device interface to send and | |
82 | receive raw CAN frames, directly to/from the controller hardware. | |
83 | Queueing of frames and higher-level transport protocols like ISO-TP | |
84 | have to be implemented in user space applications. Also, most | |
85 | character-device implementations support only one single process to | |
86 | open the device at a time, similar to a serial interface. Exchanging | |
87 | the CAN controller requires employment of another device driver and | |
88 | often the need for adaption of large parts of the application to the | |
89 | new driver's API. | |
90 | ||
f35f6c8f | 91 | SocketCAN was designed to overcome all of these limitations. A new |
f7ab97f7 OH |
92 | protocol family has been implemented which provides a socket interface |
93 | to user space applications and which builds upon the Linux network | |
f35f6c8f | 94 | layer, enabling use all of the provided queueing functionality. A device |
f7ab97f7 OH |
95 | driver for CAN controller hardware registers itself with the Linux |
96 | network layer as a network device, so that CAN frames from the | |
97 | controller can be passed up to the network layer and on to the CAN | |
98 | protocol family module and also vice-versa. Also, the protocol family | |
99 | module provides an API for transport protocol modules to register, so | |
100 | that any number of transport protocols can be loaded or unloaded | |
101 | dynamically. In fact, the can core module alone does not provide any | |
102 | protocol and cannot be used without loading at least one additional | |
103 | protocol module. Multiple sockets can be opened at the same time, | |
104 | on different or the same protocol module and they can listen/send | |
105 | frames on different or the same CAN IDs. Several sockets listening on | |
106 | the same interface for frames with the same CAN ID are all passed the | |
107 | same received matching CAN frames. An application wishing to | |
108 | communicate using a specific transport protocol, e.g. ISO-TP, just | |
109 | selects that protocol when opening the socket, and then can read and | |
110 | write application data byte streams, without having to deal with | |
111 | CAN-IDs, frames, etc. | |
112 | ||
113 | Similar functionality visible from user-space could be provided by a | |
114 | character device, too, but this would lead to a technically inelegant | |
115 | solution for a couple of reasons: | |
116 | ||
117 | * Intricate usage. Instead of passing a protocol argument to | |
118 | socket(2) and using bind(2) to select a CAN interface and CAN ID, an | |
119 | application would have to do all these operations using ioctl(2)s. | |
120 | ||
121 | * Code duplication. A character device cannot make use of the Linux | |
122 | network queueing code, so all that code would have to be duplicated | |
123 | for CAN networking. | |
124 | ||
125 | * Abstraction. In most existing character-device implementations, the | |
126 | hardware-specific device driver for a CAN controller directly | |
127 | provides the character device for the application to work with. | |
128 | This is at least very unusual in Unix systems for both, char and | |
129 | block devices. For example you don't have a character device for a | |
130 | certain UART of a serial interface, a certain sound chip in your | |
131 | computer, a SCSI or IDE controller providing access to your hard | |
132 | disk or tape streamer device. Instead, you have abstraction layers | |
133 | which provide a unified character or block device interface to the | |
134 | application on the one hand, and a interface for hardware-specific | |
135 | device drivers on the other hand. These abstractions are provided | |
136 | by subsystems like the tty layer, the audio subsystem or the SCSI | |
137 | and IDE subsystems for the devices mentioned above. | |
138 | ||
139 | The easiest way to implement a CAN device driver is as a character | |
140 | device without such a (complete) abstraction layer, as is done by most | |
141 | existing drivers. The right way, however, would be to add such a | |
142 | layer with all the functionality like registering for certain CAN | |
143 | IDs, supporting several open file descriptors and (de)multiplexing | |
144 | CAN frames between them, (sophisticated) queueing of CAN frames, and | |
145 | providing an API for device drivers to register with. However, then | |
146 | it would be no more difficult, or may be even easier, to use the | |
147 | networking framework provided by the Linux kernel, and this is what | |
f35f6c8f | 148 | SocketCAN does. |
f7ab97f7 OH |
149 | |
150 | The use of the networking framework of the Linux kernel is just the | |
151 | natural and most appropriate way to implement CAN for Linux. | |
152 | ||
f35f6c8f | 153 | 3. SocketCAN concept |
f7ab97f7 OH |
154 | --------------------- |
155 | ||
f35f6c8f | 156 | As described in chapter 2 it is the main goal of SocketCAN to |
f7ab97f7 OH |
157 | provide a socket interface to user space applications which builds |
158 | upon the Linux network layer. In contrast to the commonly known | |
159 | TCP/IP and ethernet networking, the CAN bus is a broadcast-only(!) | |
160 | medium that has no MAC-layer addressing like ethernet. The CAN-identifier | |
161 | (can_id) is used for arbitration on the CAN-bus. Therefore the CAN-IDs | |
162 | have to be chosen uniquely on the bus. When designing a CAN-ECU | |
163 | network the CAN-IDs are mapped to be sent by a specific ECU. | |
164 | For this reason a CAN-ID can be treated best as a kind of source address. | |
165 | ||
166 | 3.1 receive lists | |
167 | ||
168 | The network transparent access of multiple applications leads to the | |
169 | problem that different applications may be interested in the same | |
f35f6c8f | 170 | CAN-IDs from the same CAN network interface. The SocketCAN core |
f7ab97f7 OH |
171 | module - which implements the protocol family CAN - provides several |
172 | high efficient receive lists for this reason. If e.g. a user space | |
173 | application opens a CAN RAW socket, the raw protocol module itself | |
f35f6c8f | 174 | requests the (range of) CAN-IDs from the SocketCAN core that are |
f7ab97f7 OH |
175 | requested by the user. The subscription and unsubscription of |
176 | CAN-IDs can be done for specific CAN interfaces or for all(!) known | |
177 | CAN interfaces with the can_rx_(un)register() functions provided to | |
178 | CAN protocol modules by the SocketCAN core (see chapter 5). | |
179 | To optimize the CPU usage at runtime the receive lists are split up | |
180 | into several specific lists per device that match the requested | |
181 | filter complexity for a given use-case. | |
182 | ||
183 | 3.2 local loopback of sent frames | |
184 | ||
185 | As known from other networking concepts the data exchanging | |
186 | applications may run on the same or different nodes without any | |
187 | change (except for the according addressing information): | |
188 | ||
189 | ___ ___ ___ _______ ___ | |
190 | | _ | | _ | | _ | | _ _ | | _ | | |
191 | ||A|| ||B|| ||C|| ||A| |B|| ||C|| | |
192 | |___| |___| |___| |_______| |___| | |
193 | | | | | | | |
194 | -----------------(1)- CAN bus -(2)--------------- | |
195 | ||
196 | To ensure that application A receives the same information in the | |
197 | example (2) as it would receive in example (1) there is need for | |
198 | some kind of local loopback of the sent CAN frames on the appropriate | |
199 | node. | |
200 | ||
201 | The Linux network devices (by default) just can handle the | |
202 | transmission and reception of media dependent frames. Due to the | |
d9195881 | 203 | arbitration on the CAN bus the transmission of a low prio CAN-ID |
f7ab97f7 OH |
204 | may be delayed by the reception of a high prio CAN frame. To |
205 | reflect the correct* traffic on the node the loopback of the sent | |
206 | data has to be performed right after a successful transmission. If | |
207 | the CAN network interface is not capable of performing the loopback for | |
208 | some reason the SocketCAN core can do this task as a fallback solution. | |
209 | See chapter 6.2 for details (recommended). | |
210 | ||
211 | The loopback functionality is enabled by default to reflect standard | |
212 | networking behaviour for CAN applications. Due to some requests from | |
213 | the RT-SocketCAN group the loopback optionally may be disabled for each | |
214 | separate socket. See sockopts from the CAN RAW sockets in chapter 4.1. | |
215 | ||
216 | * = you really like to have this when you're running analyser tools | |
217 | like 'candump' or 'cansniffer' on the (same) node. | |
218 | ||
f35f6c8f | 219 | 3.3 network problem notifications |
f7ab97f7 OH |
220 | |
221 | The use of the CAN bus may lead to several problems on the physical | |
222 | and media access control layer. Detecting and logging of these lower | |
223 | layer problems is a vital requirement for CAN users to identify | |
224 | hardware issues on the physical transceiver layer as well as | |
225 | arbitration problems and error frames caused by the different | |
226 | ECUs. The occurrence of detected errors are important for diagnosis | |
227 | and have to be logged together with the exact timestamp. For this | |
d6e640f9 OH |
228 | reason the CAN interface driver can generate so called Error Message |
229 | Frames that can optionally be passed to the user application in the | |
230 | same way as other CAN frames. Whenever an error on the physical layer | |
f7ab97f7 | 231 | or the MAC layer is detected (e.g. by the CAN controller) the driver |
d6e640f9 OH |
232 | creates an appropriate error message frame. Error messages frames can |
233 | be requested by the user application using the common CAN filter | |
234 | mechanisms. Inside this filter definition the (interested) type of | |
235 | errors may be selected. The reception of error messages is disabled | |
236 | by default. The format of the CAN error message frame is briefly | |
237 | described in the Linux header file "include/linux/can/error.h". | |
f7ab97f7 | 238 | |
f35f6c8f | 239 | 4. How to use SocketCAN |
f7ab97f7 OH |
240 | ------------------------ |
241 | ||
242 | Like TCP/IP, you first need to open a socket for communicating over a | |
f35f6c8f | 243 | CAN network. Since SocketCAN implements a new protocol family, you |
f7ab97f7 OH |
244 | need to pass PF_CAN as the first argument to the socket(2) system |
245 | call. Currently, there are two CAN protocols to choose from, the raw | |
246 | socket protocol and the broadcast manager (BCM). So to open a socket, | |
247 | you would write | |
248 | ||
249 | s = socket(PF_CAN, SOCK_RAW, CAN_RAW); | |
250 | ||
251 | and | |
252 | ||
253 | s = socket(PF_CAN, SOCK_DGRAM, CAN_BCM); | |
254 | ||
255 | respectively. After the successful creation of the socket, you would | |
256 | normally use the bind(2) system call to bind the socket to a CAN | |
257 | interface (which is different from TCP/IP due to different addressing | |
258 | - see chapter 3). After binding (CAN_RAW) or connecting (CAN_BCM) | |
259 | the socket, you can read(2) and write(2) from/to the socket or use | |
260 | send(2), sendto(2), sendmsg(2) and the recv* counterpart operations | |
261 | on the socket as usual. There are also CAN specific socket options | |
262 | described below. | |
263 | ||
264 | The basic CAN frame structure and the sockaddr structure are defined | |
265 | in include/linux/can.h: | |
266 | ||
267 | struct can_frame { | |
268 | canid_t can_id; /* 32 bit CAN_ID + EFF/RTR/ERR flags */ | |
ea53fe0c | 269 | __u8 can_dlc; /* frame payload length in byte (0 .. 8) */ |
f7ab97f7 OH |
270 | __u8 data[8] __attribute__((aligned(8))); |
271 | }; | |
272 | ||
273 | The alignment of the (linear) payload data[] to a 64bit boundary | |
f35f6c8f JW |
274 | allows the user to define their own structs and unions to easily access |
275 | the CAN payload. There is no given byteorder on the CAN bus by | |
f7ab97f7 OH |
276 | default. A read(2) system call on a CAN_RAW socket transfers a |
277 | struct can_frame to the user space. | |
278 | ||
279 | The sockaddr_can structure has an interface index like the | |
280 | PF_PACKET socket, that also binds to a specific interface: | |
281 | ||
282 | struct sockaddr_can { | |
283 | sa_family_t can_family; | |
284 | int can_ifindex; | |
285 | union { | |
56690c21 OH |
286 | /* transport protocol class address info (e.g. ISOTP) */ |
287 | struct { canid_t rx_id, tx_id; } tp; | |
288 | ||
289 | /* reserved for future CAN protocols address information */ | |
f7ab97f7 OH |
290 | } can_addr; |
291 | }; | |
292 | ||
293 | To determine the interface index an appropriate ioctl() has to | |
294 | be used (example for CAN_RAW sockets without error checking): | |
295 | ||
296 | int s; | |
297 | struct sockaddr_can addr; | |
298 | struct ifreq ifr; | |
299 | ||
300 | s = socket(PF_CAN, SOCK_RAW, CAN_RAW); | |
301 | ||
302 | strcpy(ifr.ifr_name, "can0" ); | |
303 | ioctl(s, SIOCGIFINDEX, &ifr); | |
304 | ||
305 | addr.can_family = AF_CAN; | |
306 | addr.can_ifindex = ifr.ifr_ifindex; | |
307 | ||
308 | bind(s, (struct sockaddr *)&addr, sizeof(addr)); | |
309 | ||
310 | (..) | |
311 | ||
312 | To bind a socket to all(!) CAN interfaces the interface index must | |
313 | be 0 (zero). In this case the socket receives CAN frames from every | |
314 | enabled CAN interface. To determine the originating CAN interface | |
315 | the system call recvfrom(2) may be used instead of read(2). To send | |
316 | on a socket that is bound to 'any' interface sendto(2) is needed to | |
317 | specify the outgoing interface. | |
318 | ||
319 | Reading CAN frames from a bound CAN_RAW socket (see above) consists | |
320 | of reading a struct can_frame: | |
321 | ||
322 | struct can_frame frame; | |
323 | ||
324 | nbytes = read(s, &frame, sizeof(struct can_frame)); | |
325 | ||
326 | if (nbytes < 0) { | |
327 | perror("can raw socket read"); | |
328 | return 1; | |
329 | } | |
330 | ||
19f59460 | 331 | /* paranoid check ... */ |
f7ab97f7 OH |
332 | if (nbytes < sizeof(struct can_frame)) { |
333 | fprintf(stderr, "read: incomplete CAN frame\n"); | |
334 | return 1; | |
335 | } | |
336 | ||
337 | /* do something with the received CAN frame */ | |
338 | ||
339 | Writing CAN frames can be done similarly, with the write(2) system call: | |
340 | ||
341 | nbytes = write(s, &frame, sizeof(struct can_frame)); | |
342 | ||
343 | When the CAN interface is bound to 'any' existing CAN interface | |
344 | (addr.can_ifindex = 0) it is recommended to use recvfrom(2) if the | |
345 | information about the originating CAN interface is needed: | |
346 | ||
347 | struct sockaddr_can addr; | |
348 | struct ifreq ifr; | |
349 | socklen_t len = sizeof(addr); | |
350 | struct can_frame frame; | |
351 | ||
352 | nbytes = recvfrom(s, &frame, sizeof(struct can_frame), | |
353 | 0, (struct sockaddr*)&addr, &len); | |
354 | ||
355 | /* get interface name of the received CAN frame */ | |
356 | ifr.ifr_ifindex = addr.can_ifindex; | |
357 | ioctl(s, SIOCGIFNAME, &ifr); | |
358 | printf("Received a CAN frame from interface %s", ifr.ifr_name); | |
359 | ||
360 | To write CAN frames on sockets bound to 'any' CAN interface the | |
361 | outgoing interface has to be defined certainly. | |
362 | ||
363 | strcpy(ifr.ifr_name, "can0"); | |
364 | ioctl(s, SIOCGIFINDEX, &ifr); | |
365 | addr.can_ifindex = ifr.ifr_ifindex; | |
366 | addr.can_family = AF_CAN; | |
367 | ||
368 | nbytes = sendto(s, &frame, sizeof(struct can_frame), | |
369 | 0, (struct sockaddr*)&addr, sizeof(addr)); | |
370 | ||
ea53fe0c OH |
371 | Remark about CAN FD (flexible data rate) support: |
372 | ||
373 | Generally the handling of CAN FD is very similar to the formerly described | |
374 | examples. The new CAN FD capable CAN controllers support two different | |
375 | bitrates for the arbitration phase and the payload phase of the CAN FD frame | |
376 | and up to 64 bytes of payload. This extended payload length breaks all the | |
377 | kernel interfaces (ABI) which heavily rely on the CAN frame with fixed eight | |
378 | bytes of payload (struct can_frame) like the CAN_RAW socket. Therefore e.g. | |
379 | the CAN_RAW socket supports a new socket option CAN_RAW_FD_FRAMES that | |
380 | switches the socket into a mode that allows the handling of CAN FD frames | |
381 | and (legacy) CAN frames simultaneously (see section 4.1.5). | |
382 | ||
383 | The struct canfd_frame is defined in include/linux/can.h: | |
384 | ||
385 | struct canfd_frame { | |
386 | canid_t can_id; /* 32 bit CAN_ID + EFF/RTR/ERR flags */ | |
387 | __u8 len; /* frame payload length in byte (0 .. 64) */ | |
388 | __u8 flags; /* additional flags for CAN FD */ | |
389 | __u8 __res0; /* reserved / padding */ | |
390 | __u8 __res1; /* reserved / padding */ | |
391 | __u8 data[64] __attribute__((aligned(8))); | |
392 | }; | |
393 | ||
394 | The struct canfd_frame and the existing struct can_frame have the can_id, | |
395 | the payload length and the payload data at the same offset inside their | |
396 | structures. This allows to handle the different structures very similar. | |
397 | When the content of a struct can_frame is copied into a struct canfd_frame | |
398 | all structure elements can be used as-is - only the data[] becomes extended. | |
399 | ||
400 | When introducing the struct canfd_frame it turned out that the data length | |
401 | code (DLC) of the struct can_frame was used as a length information as the | |
402 | length and the DLC has a 1:1 mapping in the range of 0 .. 8. To preserve | |
403 | the easy handling of the length information the canfd_frame.len element | |
404 | contains a plain length value from 0 .. 64. So both canfd_frame.len and | |
405 | can_frame.can_dlc are equal and contain a length information and no DLC. | |
406 | For details about the distinction of CAN and CAN FD capable devices and | |
407 | the mapping to the bus-relevant data length code (DLC), see chapter 6.6. | |
408 | ||
409 | The length of the two CAN(FD) frame structures define the maximum transfer | |
410 | unit (MTU) of the CAN(FD) network interface and skbuff data length. Two | |
411 | definitions are specified for CAN specific MTUs in include/linux/can.h : | |
412 | ||
413 | #define CAN_MTU (sizeof(struct can_frame)) == 16 => 'legacy' CAN frame | |
414 | #define CANFD_MTU (sizeof(struct canfd_frame)) == 72 => CAN FD frame | |
415 | ||
f7ab97f7 OH |
416 | 4.1 RAW protocol sockets with can_filters (SOCK_RAW) |
417 | ||
418 | Using CAN_RAW sockets is extensively comparable to the commonly | |
419 | known access to CAN character devices. To meet the new possibilities | |
420 | provided by the multi user SocketCAN approach, some reasonable | |
421 | defaults are set at RAW socket binding time: | |
422 | ||
423 | - The filters are set to exactly one filter receiving everything | |
d6e640f9 | 424 | - The socket only receives valid data frames (=> no error message frames) |
f7ab97f7 OH |
425 | - The loopback of sent CAN frames is enabled (see chapter 3.2) |
426 | - The socket does not receive its own sent frames (in loopback mode) | |
427 | ||
428 | These default settings may be changed before or after binding the socket. | |
429 | To use the referenced definitions of the socket options for CAN_RAW | |
430 | sockets, include <linux/can/raw.h>. | |
431 | ||
432 | 4.1.1 RAW socket option CAN_RAW_FILTER | |
433 | ||
434 | The reception of CAN frames using CAN_RAW sockets can be controlled | |
435 | by defining 0 .. n filters with the CAN_RAW_FILTER socket option. | |
436 | ||
437 | The CAN filter structure is defined in include/linux/can.h: | |
438 | ||
439 | struct can_filter { | |
440 | canid_t can_id; | |
441 | canid_t can_mask; | |
442 | }; | |
443 | ||
444 | A filter matches, when | |
445 | ||
446 | <received_can_id> & mask == can_id & mask | |
447 | ||
448 | which is analogous to known CAN controllers hardware filter semantics. | |
449 | The filter can be inverted in this semantic, when the CAN_INV_FILTER | |
450 | bit is set in can_id element of the can_filter structure. In | |
451 | contrast to CAN controller hardware filters the user may set 0 .. n | |
452 | receive filters for each open socket separately: | |
453 | ||
454 | struct can_filter rfilter[2]; | |
455 | ||
456 | rfilter[0].can_id = 0x123; | |
457 | rfilter[0].can_mask = CAN_SFF_MASK; | |
458 | rfilter[1].can_id = 0x200; | |
459 | rfilter[1].can_mask = 0x700; | |
460 | ||
461 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_FILTER, &rfilter, sizeof(rfilter)); | |
462 | ||
463 | To disable the reception of CAN frames on the selected CAN_RAW socket: | |
464 | ||
465 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_FILTER, NULL, 0); | |
466 | ||
f35f6c8f | 467 | To set the filters to zero filters is quite obsolete as to not read |
f7ab97f7 OH |
468 | data causes the raw socket to discard the received CAN frames. But |
469 | having this 'send only' use-case we may remove the receive list in the | |
470 | Kernel to save a little (really a very little!) CPU usage. | |
471 | ||
277bd320 OH |
472 | 4.1.1.1 CAN filter usage optimisation |
473 | ||
474 | The CAN filters are processed in per-device filter lists at CAN frame | |
475 | reception time. To reduce the number of checks that need to be performed | |
476 | while walking through the filter lists the CAN core provides an optimized | |
477 | filter handling when the filter subscription focusses on a single CAN ID. | |
478 | ||
479 | For the possible 2048 SFF CAN identifiers the identifier is used as an index | |
480 | to access the corresponding subscription list without any further checks. | |
481 | For the 2^29 possible EFF CAN identifiers a 10 bit XOR folding is used as | |
482 | hash function to retrieve the EFF table index. | |
483 | ||
484 | To benefit from the optimized filters for single CAN identifiers the | |
485 | CAN_SFF_MASK or CAN_EFF_MASK have to be set into can_filter.mask together | |
486 | with set CAN_EFF_FLAG and CAN_RTR_FLAG bits. A set CAN_EFF_FLAG bit in the | |
487 | can_filter.mask makes clear that it matters whether a SFF or EFF CAN ID is | |
488 | subscribed. E.g. in the example from above | |
489 | ||
490 | rfilter[0].can_id = 0x123; | |
491 | rfilter[0].can_mask = CAN_SFF_MASK; | |
492 | ||
493 | both SFF frames with CAN ID 0x123 and EFF frames with 0xXXXXX123 can pass. | |
494 | ||
495 | To filter for only 0x123 (SFF) and 0x12345678 (EFF) CAN identifiers the | |
496 | filter has to be defined in this way to benefit from the optimized filters: | |
497 | ||
498 | struct can_filter rfilter[2]; | |
499 | ||
500 | rfilter[0].can_id = 0x123; | |
501 | rfilter[0].can_mask = (CAN_EFF_FLAG | CAN_RTR_FLAG | CAN_SFF_MASK); | |
502 | rfilter[1].can_id = 0x12345678 | CAN_EFF_FLAG; | |
503 | rfilter[1].can_mask = (CAN_EFF_FLAG | CAN_RTR_FLAG | CAN_EFF_MASK); | |
504 | ||
505 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_FILTER, &rfilter, sizeof(rfilter)); | |
506 | ||
f7ab97f7 OH |
507 | 4.1.2 RAW socket option CAN_RAW_ERR_FILTER |
508 | ||
509 | As described in chapter 3.4 the CAN interface driver can generate so | |
d6e640f9 | 510 | called Error Message Frames that can optionally be passed to the user |
f7ab97f7 OH |
511 | application in the same way as other CAN frames. The possible |
512 | errors are divided into different error classes that may be filtered | |
513 | using the appropriate error mask. To register for every possible | |
514 | error condition CAN_ERR_MASK can be used as value for the error mask. | |
515 | The values for the error mask are defined in linux/can/error.h . | |
516 | ||
517 | can_err_mask_t err_mask = ( CAN_ERR_TX_TIMEOUT | CAN_ERR_BUSOFF ); | |
518 | ||
519 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_ERR_FILTER, | |
520 | &err_mask, sizeof(err_mask)); | |
521 | ||
522 | 4.1.3 RAW socket option CAN_RAW_LOOPBACK | |
523 | ||
524 | To meet multi user needs the local loopback is enabled by default | |
525 | (see chapter 3.2 for details). But in some embedded use-cases | |
526 | (e.g. when only one application uses the CAN bus) this loopback | |
527 | functionality can be disabled (separately for each socket): | |
528 | ||
529 | int loopback = 0; /* 0 = disabled, 1 = enabled (default) */ | |
530 | ||
531 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_LOOPBACK, &loopback, sizeof(loopback)); | |
532 | ||
533 | 4.1.4 RAW socket option CAN_RAW_RECV_OWN_MSGS | |
534 | ||
535 | When the local loopback is enabled, all the sent CAN frames are | |
536 | looped back to the open CAN sockets that registered for the CAN | |
537 | frames' CAN-ID on this given interface to meet the multi user | |
538 | needs. The reception of the CAN frames on the same socket that was | |
539 | sending the CAN frame is assumed to be unwanted and therefore | |
540 | disabled by default. This default behaviour may be changed on | |
541 | demand: | |
542 | ||
543 | int recv_own_msgs = 1; /* 0 = disabled (default), 1 = enabled */ | |
544 | ||
545 | setsockopt(s, SOL_CAN_RAW, CAN_RAW_RECV_OWN_MSGS, | |
546 | &recv_own_msgs, sizeof(recv_own_msgs)); | |
547 | ||
ea53fe0c OH |
548 | 4.1.5 RAW socket option CAN_RAW_FD_FRAMES |
549 | ||
550 | CAN FD support in CAN_RAW sockets can be enabled with a new socket option | |
551 | CAN_RAW_FD_FRAMES which is off by default. When the new socket option is | |
552 | not supported by the CAN_RAW socket (e.g. on older kernels), switching the | |
553 | CAN_RAW_FD_FRAMES option returns the error -ENOPROTOOPT. | |
554 | ||
555 | Once CAN_RAW_FD_FRAMES is enabled the application can send both CAN frames | |
556 | and CAN FD frames. OTOH the application has to handle CAN and CAN FD frames | |
557 | when reading from the socket. | |
558 | ||
559 | CAN_RAW_FD_FRAMES enabled: CAN_MTU and CANFD_MTU are allowed | |
560 | CAN_RAW_FD_FRAMES disabled: only CAN_MTU is allowed (default) | |
561 | ||
562 | Example: | |
563 | [ remember: CANFD_MTU == sizeof(struct canfd_frame) ] | |
564 | ||
565 | struct canfd_frame cfd; | |
566 | ||
567 | nbytes = read(s, &cfd, CANFD_MTU); | |
568 | ||
569 | if (nbytes == CANFD_MTU) { | |
570 | printf("got CAN FD frame with length %d\n", cfd.len); | |
571 | /* cfd.flags contains valid data */ | |
572 | } else if (nbytes == CAN_MTU) { | |
573 | printf("got legacy CAN frame with length %d\n", cfd.len); | |
574 | /* cfd.flags is undefined */ | |
575 | } else { | |
576 | fprintf(stderr, "read: invalid CAN(FD) frame\n"); | |
577 | return 1; | |
578 | } | |
579 | ||
580 | /* the content can be handled independently from the received MTU size */ | |
581 | ||
582 | printf("can_id: %X data length: %d data: ", cfd.can_id, cfd.len); | |
583 | for (i = 0; i < cfd.len; i++) | |
584 | printf("%02X ", cfd.data[i]); | |
585 | ||
586 | When reading with size CANFD_MTU only returns CAN_MTU bytes that have | |
587 | been received from the socket a legacy CAN frame has been read into the | |
588 | provided CAN FD structure. Note that the canfd_frame.flags data field is | |
589 | not specified in the struct can_frame and therefore it is only valid in | |
590 | CANFD_MTU sized CAN FD frames. | |
591 | ||
ea53fe0c OH |
592 | Implementation hint for new CAN applications: |
593 | ||
594 | To build a CAN FD aware application use struct canfd_frame as basic CAN | |
595 | data structure for CAN_RAW based applications. When the application is | |
596 | executed on an older Linux kernel and switching the CAN_RAW_FD_FRAMES | |
597 | socket option returns an error: No problem. You'll get legacy CAN frames | |
598 | or CAN FD frames and can process them the same way. | |
599 | ||
600 | When sending to CAN devices make sure that the device is capable to handle | |
601 | CAN FD frames by checking if the device maximum transfer unit is CANFD_MTU. | |
602 | The CAN device MTU can be retrieved e.g. with a SIOCGIFMTU ioctl() syscall. | |
603 | ||
604 | 4.1.6 RAW socket returned message flags | |
1e55659c OH |
605 | |
606 | When using recvmsg() call, the msg->msg_flags may contain following flags: | |
607 | ||
608 | MSG_DONTROUTE: set when the received frame was created on the local host. | |
609 | ||
610 | MSG_CONFIRM: set when the frame was sent via the socket it is received on. | |
611 | This flag can be interpreted as a 'transmission confirmation' when the | |
612 | CAN driver supports the echo of frames on driver level, see 3.2 and 6.2. | |
613 | In order to receive such messages, CAN_RAW_RECV_OWN_MSGS must be set. | |
614 | ||
f7ab97f7 | 615 | 4.2 Broadcast Manager protocol sockets (SOCK_DGRAM) |
51b2f451 OH |
616 | |
617 | The Broadcast Manager protocol provides a command based configuration | |
618 | interface to filter and send (e.g. cyclic) CAN messages in kernel space. | |
619 | ||
620 | Receive filters can be used to down sample frequent messages; detect events | |
621 | such as message contents changes, packet length changes, and do time-out | |
622 | monitoring of received messages. | |
623 | ||
624 | Periodic transmission tasks of CAN frames or a sequence of CAN frames can be | |
625 | created and modified at runtime; both the message content and the two | |
626 | possible transmit intervals can be altered. | |
627 | ||
628 | A BCM socket is not intended for sending individual CAN frames using the | |
629 | struct can_frame as known from the CAN_RAW socket. Instead a special BCM | |
630 | configuration message is defined. The basic BCM configuration message used | |
631 | to communicate with the broadcast manager and the available operations are | |
632 | defined in the linux/can/bcm.h include. The BCM message consists of a | |
633 | message header with a command ('opcode') followed by zero or more CAN frames. | |
634 | The broadcast manager sends responses to user space in the same form: | |
635 | ||
636 | struct bcm_msg_head { | |
637 | __u32 opcode; /* command */ | |
638 | __u32 flags; /* special flags */ | |
639 | __u32 count; /* run 'count' times with ival1 */ | |
640 | struct timeval ival1, ival2; /* count and subsequent interval */ | |
641 | canid_t can_id; /* unique can_id for task */ | |
642 | __u32 nframes; /* number of can_frames following */ | |
643 | struct can_frame frames[0]; | |
644 | }; | |
645 | ||
646 | The aligned payload 'frames' uses the same basic CAN frame structure defined | |
647 | at the beginning of section 4 and in the include/linux/can.h include. All | |
648 | messages to the broadcast manager from user space have this structure. | |
649 | ||
650 | Note a CAN_BCM socket must be connected instead of bound after socket | |
651 | creation (example without error checking): | |
652 | ||
653 | int s; | |
654 | struct sockaddr_can addr; | |
655 | struct ifreq ifr; | |
656 | ||
657 | s = socket(PF_CAN, SOCK_DGRAM, CAN_BCM); | |
658 | ||
659 | strcpy(ifr.ifr_name, "can0"); | |
660 | ioctl(s, SIOCGIFINDEX, &ifr); | |
661 | ||
662 | addr.can_family = AF_CAN; | |
663 | addr.can_ifindex = ifr.ifr_ifindex; | |
664 | ||
665 | connect(s, (struct sockaddr *)&addr, sizeof(addr)) | |
666 | ||
667 | (..) | |
668 | ||
669 | The broadcast manager socket is able to handle any number of in flight | |
670 | transmissions or receive filters concurrently. The different RX/TX jobs are | |
671 | distinguished by the unique can_id in each BCM message. However additional | |
672 | CAN_BCM sockets are recommended to communicate on multiple CAN interfaces. | |
673 | When the broadcast manager socket is bound to 'any' CAN interface (=> the | |
674 | interface index is set to zero) the configured receive filters apply to any | |
675 | CAN interface unless the sendto() syscall is used to overrule the 'any' CAN | |
676 | interface index. When using recvfrom() instead of read() to retrieve BCM | |
677 | socket messages the originating CAN interface is provided in can_ifindex. | |
678 | ||
679 | 4.2.1 Broadcast Manager operations | |
680 | ||
681 | The opcode defines the operation for the broadcast manager to carry out, | |
682 | or details the broadcast managers response to several events, including | |
683 | user requests. | |
684 | ||
685 | Transmit Operations (user space to broadcast manager): | |
686 | ||
687 | TX_SETUP: Create (cyclic) transmission task. | |
688 | ||
689 | TX_DELETE: Remove (cyclic) transmission task, requires only can_id. | |
690 | ||
691 | TX_READ: Read properties of (cyclic) transmission task for can_id. | |
692 | ||
693 | TX_SEND: Send one CAN frame. | |
694 | ||
695 | Transmit Responses (broadcast manager to user space): | |
696 | ||
697 | TX_STATUS: Reply to TX_READ request (transmission task configuration). | |
698 | ||
699 | TX_EXPIRED: Notification when counter finishes sending at initial interval | |
700 | 'ival1'. Requires the TX_COUNTEVT flag to be set at TX_SETUP. | |
701 | ||
702 | Receive Operations (user space to broadcast manager): | |
703 | ||
704 | RX_SETUP: Create RX content filter subscription. | |
705 | ||
706 | RX_DELETE: Remove RX content filter subscription, requires only can_id. | |
707 | ||
708 | RX_READ: Read properties of RX content filter subscription for can_id. | |
709 | ||
710 | Receive Responses (broadcast manager to user space): | |
711 | ||
712 | RX_STATUS: Reply to RX_READ request (filter task configuration). | |
713 | ||
714 | RX_TIMEOUT: Cyclic message is detected to be absent (timer ival1 expired). | |
715 | ||
716 | RX_CHANGED: BCM message with updated CAN frame (detected content change). | |
717 | Sent on first message received or on receipt of revised CAN messages. | |
718 | ||
719 | 4.2.2 Broadcast Manager message flags | |
720 | ||
721 | When sending a message to the broadcast manager the 'flags' element may | |
722 | contain the following flag definitions which influence the behaviour: | |
723 | ||
724 | SETTIMER: Set the values of ival1, ival2 and count | |
725 | ||
726 | STARTTIMER: Start the timer with the actual values of ival1, ival2 | |
727 | and count. Starting the timer leads simultaneously to emit a CAN frame. | |
728 | ||
729 | TX_COUNTEVT: Create the message TX_EXPIRED when count expires | |
730 | ||
731 | TX_ANNOUNCE: A change of data by the process is emitted immediately. | |
732 | ||
733 | TX_CP_CAN_ID: Copies the can_id from the message header to each | |
734 | subsequent frame in frames. This is intended as usage simplification. For | |
735 | TX tasks the unique can_id from the message header may differ from the | |
736 | can_id(s) stored for transmission in the subsequent struct can_frame(s). | |
737 | ||
738 | RX_FILTER_ID: Filter by can_id alone, no frames required (nframes=0). | |
739 | ||
740 | RX_CHECK_DLC: A change of the DLC leads to an RX_CHANGED. | |
741 | ||
742 | RX_NO_AUTOTIMER: Prevent automatically starting the timeout monitor. | |
743 | ||
c98be0c9 | 744 | RX_ANNOUNCE_RESUME: If passed at RX_SETUP and a receive timeout occurred, a |
51b2f451 OH |
745 | RX_CHANGED message will be generated when the (cyclic) receive restarts. |
746 | ||
747 | TX_RESET_MULTI_IDX: Reset the index for the multiple frame transmission. | |
748 | ||
749 | RX_RTR_FRAME: Send reply for RTR-request (placed in op->frames[0]). | |
750 | ||
751 | 4.2.3 Broadcast Manager transmission timers | |
752 | ||
753 | Periodic transmission configurations may use up to two interval timers. | |
754 | In this case the BCM sends a number of messages ('count') at an interval | |
755 | 'ival1', then continuing to send at another given interval 'ival2'. When | |
756 | only one timer is needed 'count' is set to zero and only 'ival2' is used. | |
757 | When SET_TIMER and START_TIMER flag were set the timers are activated. | |
758 | The timer values can be altered at runtime when only SET_TIMER is set. | |
759 | ||
760 | 4.2.4 Broadcast Manager message sequence transmission | |
761 | ||
762 | Up to 256 CAN frames can be transmitted in a sequence in the case of a cyclic | |
763 | TX task configuration. The number of CAN frames is provided in the 'nframes' | |
764 | element of the BCM message head. The defined number of CAN frames are added | |
765 | as array to the TX_SETUP BCM configuration message. | |
766 | ||
767 | /* create a struct to set up a sequence of four CAN frames */ | |
768 | struct { | |
769 | struct bcm_msg_head msg_head; | |
770 | struct can_frame frame[4]; | |
771 | } mytxmsg; | |
772 | ||
773 | (..) | |
774 | mytxmsg.nframes = 4; | |
775 | (..) | |
776 | ||
777 | write(s, &mytxmsg, sizeof(mytxmsg)); | |
778 | ||
779 | With every transmission the index in the array of CAN frames is increased | |
780 | and set to zero at index overflow. | |
781 | ||
782 | 4.2.5 Broadcast Manager receive filter timers | |
783 | ||
784 | The timer values ival1 or ival2 may be set to non-zero values at RX_SETUP. | |
785 | When the SET_TIMER flag is set the timers are enabled: | |
786 | ||
787 | ival1: Send RX_TIMEOUT when a received message is not received again within | |
788 | the given time. When START_TIMER is set at RX_SETUP the timeout detection | |
789 | is activated directly - even without a former CAN frame reception. | |
790 | ||
791 | ival2: Throttle the received message rate down to the value of ival2. This | |
792 | is useful to reduce messages for the application when the signal inside the | |
793 | CAN frame is stateless as state changes within the ival2 periode may get | |
794 | lost. | |
795 | ||
796 | 4.2.6 Broadcast Manager multiplex message receive filter | |
797 | ||
798 | To filter for content changes in multiplex message sequences an array of more | |
799 | than one CAN frames can be passed in a RX_SETUP configuration message. The | |
800 | data bytes of the first CAN frame contain the mask of relevant bits that | |
801 | have to match in the subsequent CAN frames with the received CAN frame. | |
802 | If one of the subsequent CAN frames is matching the bits in that frame data | |
803 | mark the relevant content to be compared with the previous received content. | |
804 | Up to 257 CAN frames (multiplex filter bit mask CAN frame plus 256 CAN | |
805 | filters) can be added as array to the TX_SETUP BCM configuration message. | |
806 | ||
807 | /* usually used to clear CAN frame data[] - beware of endian problems! */ | |
808 | #define U64_DATA(p) (*(unsigned long long*)(p)->data) | |
809 | ||
810 | struct { | |
811 | struct bcm_msg_head msg_head; | |
812 | struct can_frame frame[5]; | |
813 | } msg; | |
814 | ||
815 | msg.msg_head.opcode = RX_SETUP; | |
816 | msg.msg_head.can_id = 0x42; | |
817 | msg.msg_head.flags = 0; | |
818 | msg.msg_head.nframes = 5; | |
819 | U64_DATA(&msg.frame[0]) = 0xFF00000000000000ULL; /* MUX mask */ | |
820 | U64_DATA(&msg.frame[1]) = 0x01000000000000FFULL; /* data mask (MUX 0x01) */ | |
821 | U64_DATA(&msg.frame[2]) = 0x0200FFFF000000FFULL; /* data mask (MUX 0x02) */ | |
822 | U64_DATA(&msg.frame[3]) = 0x330000FFFFFF0003ULL; /* data mask (MUX 0x33) */ | |
823 | U64_DATA(&msg.frame[4]) = 0x4F07FC0FF0000000ULL; /* data mask (MUX 0x4F) */ | |
824 | ||
825 | write(s, &msg, sizeof(msg)); | |
826 | ||
f7ab97f7 OH |
827 | 4.3 connected transport protocols (SOCK_SEQPACKET) |
828 | 4.4 unconnected transport protocols (SOCK_DGRAM) | |
829 | ||
830 | ||
f35f6c8f | 831 | 5. SocketCAN core module |
f7ab97f7 OH |
832 | ------------------------- |
833 | ||
f35f6c8f | 834 | The SocketCAN core module implements the protocol family |
f7ab97f7 OH |
835 | PF_CAN. CAN protocol modules are loaded by the core module at |
836 | runtime. The core module provides an interface for CAN protocol | |
837 | modules to subscribe needed CAN IDs (see chapter 3.1). | |
838 | ||
839 | 5.1 can.ko module params | |
840 | ||
f35f6c8f | 841 | - stats_timer: To calculate the SocketCAN core statistics |
f7ab97f7 OH |
842 | (e.g. current/maximum frames per second) this 1 second timer is |
843 | invoked at can.ko module start time by default. This timer can be | |
d9195881 | 844 | disabled by using stattimer=0 on the module commandline. |
f7ab97f7 OH |
845 | |
846 | - debug: (removed since SocketCAN SVN r546) | |
847 | ||
848 | 5.2 procfs content | |
849 | ||
f35f6c8f | 850 | As described in chapter 3.1 the SocketCAN core uses several filter |
f7ab97f7 OH |
851 | lists to deliver received CAN frames to CAN protocol modules. These |
852 | receive lists, their filters and the count of filter matches can be | |
853 | checked in the appropriate receive list. All entries contain the | |
854 | device and a protocol module identifier: | |
855 | ||
856 | foo@bar:~$ cat /proc/net/can/rcvlist_all | |
857 | ||
858 | receive list 'rx_all': | |
859 | (vcan3: no entry) | |
860 | (vcan2: no entry) | |
861 | (vcan1: no entry) | |
862 | device can_id can_mask function userdata matches ident | |
863 | vcan0 000 00000000 f88e6370 f6c6f400 0 raw | |
864 | (any: no entry) | |
865 | ||
866 | In this example an application requests any CAN traffic from vcan0. | |
867 | ||
868 | rcvlist_all - list for unfiltered entries (no filter operations) | |
869 | rcvlist_eff - list for single extended frame (EFF) entries | |
d6e640f9 | 870 | rcvlist_err - list for error message frames masks |
f7ab97f7 OH |
871 | rcvlist_fil - list for mask/value filters |
872 | rcvlist_inv - list for mask/value filters (inverse semantic) | |
873 | rcvlist_sff - list for single standard frame (SFF) entries | |
874 | ||
875 | Additional procfs files in /proc/net/can | |
876 | ||
f35f6c8f | 877 | stats - SocketCAN core statistics (rx/tx frames, match ratios, ...) |
f7ab97f7 | 878 | reset_stats - manual statistic reset |
f35f6c8f | 879 | version - prints the SocketCAN core version and the ABI version |
f7ab97f7 OH |
880 | |
881 | 5.3 writing own CAN protocol modules | |
882 | ||
883 | To implement a new protocol in the protocol family PF_CAN a new | |
884 | protocol has to be defined in include/linux/can.h . | |
f35f6c8f | 885 | The prototypes and definitions to use the SocketCAN core can be |
f7ab97f7 OH |
886 | accessed by including include/linux/can/core.h . |
887 | In addition to functions that register the CAN protocol and the | |
888 | CAN device notifier chain there are functions to subscribe CAN | |
889 | frames received by CAN interfaces and to send CAN frames: | |
890 | ||
891 | can_rx_register - subscribe CAN frames from a specific interface | |
892 | can_rx_unregister - unsubscribe CAN frames from a specific interface | |
893 | can_send - transmit a CAN frame (optional with local loopback) | |
894 | ||
895 | For details see the kerneldoc documentation in net/can/af_can.c or | |
896 | the source code of net/can/raw.c or net/can/bcm.c . | |
897 | ||
898 | 6. CAN network drivers | |
899 | ---------------------- | |
900 | ||
901 | Writing a CAN network device driver is much easier than writing a | |
902 | CAN character device driver. Similar to other known network device | |
903 | drivers you mainly have to deal with: | |
904 | ||
905 | - TX: Put the CAN frame from the socket buffer to the CAN controller. | |
906 | - RX: Put the CAN frame from the CAN controller to the socket buffer. | |
907 | ||
908 | See e.g. at Documentation/networking/netdevices.txt . The differences | |
909 | for writing CAN network device driver are described below: | |
910 | ||
911 | 6.1 general settings | |
912 | ||
913 | dev->type = ARPHRD_CAN; /* the netdevice hardware type */ | |
914 | dev->flags = IFF_NOARP; /* CAN has no arp */ | |
915 | ||
ea53fe0c | 916 | dev->mtu = CAN_MTU; /* sizeof(struct can_frame) -> legacy CAN interface */ |
f7ab97f7 | 917 | |
ea53fe0c OH |
918 | or alternative, when the controller supports CAN with flexible data rate: |
919 | dev->mtu = CANFD_MTU; /* sizeof(struct canfd_frame) -> CAN FD interface */ | |
920 | ||
921 | The struct can_frame or struct canfd_frame is the payload of each socket | |
922 | buffer (skbuff) in the protocol family PF_CAN. | |
f7ab97f7 OH |
923 | |
924 | 6.2 local loopback of sent frames | |
925 | ||
926 | As described in chapter 3.2 the CAN network device driver should | |
927 | support a local loopback functionality similar to the local echo | |
928 | e.g. of tty devices. In this case the driver flag IFF_ECHO has to be | |
929 | set to prevent the PF_CAN core from locally echoing sent frames | |
930 | (aka loopback) as fallback solution: | |
931 | ||
932 | dev->flags = (IFF_NOARP | IFF_ECHO); | |
933 | ||
934 | 6.3 CAN controller hardware filters | |
935 | ||
936 | To reduce the interrupt load on deep embedded systems some CAN | |
937 | controllers support the filtering of CAN IDs or ranges of CAN IDs. | |
938 | These hardware filter capabilities vary from controller to | |
939 | controller and have to be identified as not feasible in a multi-user | |
940 | networking approach. The use of the very controller specific | |
941 | hardware filters could make sense in a very dedicated use-case, as a | |
942 | filter on driver level would affect all users in the multi-user | |
943 | system. The high efficient filter sets inside the PF_CAN core allow | |
944 | to set different multiple filters for each socket separately. | |
945 | Therefore the use of hardware filters goes to the category 'handmade | |
946 | tuning on deep embedded systems'. The author is running a MPC603e | |
947 | @133MHz with four SJA1000 CAN controllers from 2002 under heavy bus | |
948 | load without any problems ... | |
949 | ||
e5d23048 OH |
950 | 6.4 The virtual CAN driver (vcan) |
951 | ||
952 | Similar to the network loopback devices, vcan offers a virtual local | |
953 | CAN interface. A full qualified address on CAN consists of | |
954 | ||
955 | - a unique CAN Identifier (CAN ID) | |
956 | - the CAN bus this CAN ID is transmitted on (e.g. can0) | |
957 | ||
958 | so in common use cases more than one virtual CAN interface is needed. | |
959 | ||
960 | The virtual CAN interfaces allow the transmission and reception of CAN | |
961 | frames without real CAN controller hardware. Virtual CAN network | |
962 | devices are usually named 'vcanX', like vcan0 vcan1 vcan2 ... | |
963 | When compiled as a module the virtual CAN driver module is called vcan.ko | |
964 | ||
965 | Since Linux Kernel version 2.6.24 the vcan driver supports the Kernel | |
966 | netlink interface to create vcan network devices. The creation and | |
967 | removal of vcan network devices can be managed with the ip(8) tool: | |
968 | ||
969 | - Create a virtual CAN network interface: | |
e20dad96 | 970 | $ ip link add type vcan |
e5d23048 OH |
971 | |
972 | - Create a virtual CAN network interface with a specific name 'vcan42': | |
e20dad96 | 973 | $ ip link add dev vcan42 type vcan |
e5d23048 OH |
974 | |
975 | - Remove a (virtual CAN) network interface 'vcan42': | |
e20dad96 WG |
976 | $ ip link del vcan42 |
977 | ||
978 | 6.5 The CAN network device driver interface | |
979 | ||
980 | The CAN network device driver interface provides a generic interface | |
981 | to setup, configure and monitor CAN network devices. The user can then | |
982 | configure the CAN device, like setting the bit-timing parameters, via | |
983 | the netlink interface using the program "ip" from the "IPROUTE2" | |
984 | utility suite. The following chapter describes briefly how to use it. | |
985 | Furthermore, the interface uses a common data structure and exports a | |
986 | set of common functions, which all real CAN network device drivers | |
987 | should use. Please have a look to the SJA1000 or MSCAN driver to | |
988 | understand how to use them. The name of the module is can-dev.ko. | |
989 | ||
990 | 6.5.1 Netlink interface to set/get devices properties | |
991 | ||
992 | The CAN device must be configured via netlink interface. The supported | |
993 | netlink message types are defined and briefly described in | |
994 | "include/linux/can/netlink.h". CAN link support for the program "ip" | |
c94bed8e | 995 | of the IPROUTE2 utility suite is available and it can be used as shown |
e20dad96 WG |
996 | below: |
997 | ||
998 | - Setting CAN device properties: | |
999 | ||
1000 | $ ip link set can0 type can help | |
1001 | Usage: ip link set DEVICE type can | |
1002 | [ bitrate BITRATE [ sample-point SAMPLE-POINT] ] | | |
1003 | [ tq TQ prop-seg PROP_SEG phase-seg1 PHASE-SEG1 | |
1004 | phase-seg2 PHASE-SEG2 [ sjw SJW ] ] | |
1005 | ||
1006 | [ loopback { on | off } ] | |
1007 | [ listen-only { on | off } ] | |
1008 | [ triple-sampling { on | off } ] | |
1009 | ||
1010 | [ restart-ms TIME-MS ] | |
1011 | [ restart ] | |
1012 | ||
1013 | Where: BITRATE := { 1..1000000 } | |
1014 | SAMPLE-POINT := { 0.000..0.999 } | |
1015 | TQ := { NUMBER } | |
1016 | PROP-SEG := { 1..8 } | |
1017 | PHASE-SEG1 := { 1..8 } | |
1018 | PHASE-SEG2 := { 1..8 } | |
1019 | SJW := { 1..4 } | |
1020 | RESTART-MS := { 0 | NUMBER } | |
1021 | ||
1022 | - Display CAN device details and statistics: | |
1023 | ||
1024 | $ ip -details -statistics link show can0 | |
1025 | 2: can0: <NOARP,UP,LOWER_UP,ECHO> mtu 16 qdisc pfifo_fast state UP qlen 10 | |
1026 | link/can | |
1027 | can <TRIPLE-SAMPLING> state ERROR-ACTIVE restart-ms 100 | |
1028 | bitrate 125000 sample_point 0.875 | |
1029 | tq 125 prop-seg 6 phase-seg1 7 phase-seg2 2 sjw 1 | |
1030 | sja1000: tseg1 1..16 tseg2 1..8 sjw 1..4 brp 1..64 brp-inc 1 | |
1031 | clock 8000000 | |
1032 | re-started bus-errors arbit-lost error-warn error-pass bus-off | |
1033 | 41 17457 0 41 42 41 | |
1034 | RX: bytes packets errors dropped overrun mcast | |
1035 | 140859 17608 17457 0 0 0 | |
1036 | TX: bytes packets errors dropped carrier collsns | |
1037 | 861 112 0 41 0 0 | |
1038 | ||
1039 | More info to the above output: | |
1040 | ||
1041 | "<TRIPLE-SAMPLING>" | |
1042 | Shows the list of selected CAN controller modes: LOOPBACK, | |
1043 | LISTEN-ONLY, or TRIPLE-SAMPLING. | |
1044 | ||
1045 | "state ERROR-ACTIVE" | |
1046 | The current state of the CAN controller: "ERROR-ACTIVE", | |
1047 | "ERROR-WARNING", "ERROR-PASSIVE", "BUS-OFF" or "STOPPED" | |
1048 | ||
1049 | "restart-ms 100" | |
1050 | Automatic restart delay time. If set to a non-zero value, a | |
1051 | restart of the CAN controller will be triggered automatically | |
1052 | in case of a bus-off condition after the specified delay time | |
1053 | in milliseconds. By default it's off. | |
1054 | ||
5fb7639d | 1055 | "bitrate 125000 sample-point 0.875" |
e20dad96 WG |
1056 | Shows the real bit-rate in bits/sec and the sample-point in the |
1057 | range 0.000..0.999. If the calculation of bit-timing parameters | |
1058 | is enabled in the kernel (CONFIG_CAN_CALC_BITTIMING=y), the | |
1059 | bit-timing can be defined by setting the "bitrate" argument. | |
1060 | Optionally the "sample-point" can be specified. By default it's | |
1061 | 0.000 assuming CIA-recommended sample-points. | |
1062 | ||
1063 | "tq 125 prop-seg 6 phase-seg1 7 phase-seg2 2 sjw 1" | |
1064 | Shows the time quanta in ns, propagation segment, phase buffer | |
1065 | segment 1 and 2 and the synchronisation jump width in units of | |
1066 | tq. They allow to define the CAN bit-timing in a hardware | |
1067 | independent format as proposed by the Bosch CAN 2.0 spec (see | |
1068 | chapter 8 of http://www.semiconductors.bosch.de/pdf/can2spec.pdf). | |
1069 | ||
1070 | "sja1000: tseg1 1..16 tseg2 1..8 sjw 1..4 brp 1..64 brp-inc 1 | |
1071 | clock 8000000" | |
1072 | Shows the bit-timing constants of the CAN controller, here the | |
1073 | "sja1000". The minimum and maximum values of the time segment 1 | |
1074 | and 2, the synchronisation jump width in units of tq, the | |
1075 | bitrate pre-scaler and the CAN system clock frequency in Hz. | |
1076 | These constants could be used for user-defined (non-standard) | |
1077 | bit-timing calculation algorithms in user-space. | |
1078 | ||
1079 | "re-started bus-errors arbit-lost error-warn error-pass bus-off" | |
1080 | Shows the number of restarts, bus and arbitration lost errors, | |
1081 | and the state changes to the error-warning, error-passive and | |
1082 | bus-off state. RX overrun errors are listed in the "overrun" | |
1083 | field of the standard network statistics. | |
1084 | ||
1085 | 6.5.2 Setting the CAN bit-timing | |
1086 | ||
1087 | The CAN bit-timing parameters can always be defined in a hardware | |
1088 | independent format as proposed in the Bosch CAN 2.0 specification | |
1089 | specifying the arguments "tq", "prop_seg", "phase_seg1", "phase_seg2" | |
1090 | and "sjw": | |
1091 | ||
1092 | $ ip link set canX type can tq 125 prop-seg 6 \ | |
1093 | phase-seg1 7 phase-seg2 2 sjw 1 | |
1094 | ||
1095 | If the kernel option CONFIG_CAN_CALC_BITTIMING is enabled, CIA | |
1096 | recommended CAN bit-timing parameters will be calculated if the bit- | |
1097 | rate is specified with the argument "bitrate": | |
1098 | ||
1099 | $ ip link set canX type can bitrate 125000 | |
1100 | ||
1101 | Note that this works fine for the most common CAN controllers with | |
1102 | standard bit-rates but may *fail* for exotic bit-rates or CAN system | |
1103 | clock frequencies. Disabling CONFIG_CAN_CALC_BITTIMING saves some | |
1104 | space and allows user-space tools to solely determine and set the | |
1105 | bit-timing parameters. The CAN controller specific bit-timing | |
1106 | constants can be used for that purpose. They are listed by the | |
1107 | following command: | |
1108 | ||
1109 | $ ip -details link show can0 | |
1110 | ... | |
1111 | sja1000: clock 8000000 tseg1 1..16 tseg2 1..8 sjw 1..4 brp 1..64 brp-inc 1 | |
1112 | ||
1113 | 6.5.3 Starting and stopping the CAN network device | |
1114 | ||
1115 | A CAN network device is started or stopped as usual with the command | |
1116 | "ifconfig canX up/down" or "ip link set canX up/down". Be aware that | |
1117 | you *must* define proper bit-timing parameters for real CAN devices | |
1118 | before you can start it to avoid error-prone default settings: | |
1119 | ||
1120 | $ ip link set canX up type can bitrate 125000 | |
1121 | ||
f35f6c8f | 1122 | A device may enter the "bus-off" state if too many errors occurred on |
e20dad96 WG |
1123 | the CAN bus. Then no more messages are received or sent. An automatic |
1124 | bus-off recovery can be enabled by setting the "restart-ms" to a | |
1125 | non-zero value, e.g.: | |
1126 | ||
1127 | $ ip link set canX type can restart-ms 100 | |
1128 | ||
1129 | Alternatively, the application may realize the "bus-off" condition | |
d6e640f9 OH |
1130 | by monitoring CAN error message frames and do a restart when |
1131 | appropriate with the command: | |
e20dad96 WG |
1132 | |
1133 | $ ip link set canX type can restart | |
1134 | ||
d6e640f9 OH |
1135 | Note that a restart will also create a CAN error message frame (see |
1136 | also chapter 3.4). | |
f7ab97f7 | 1137 | |
ea53fe0c OH |
1138 | 6.6 CAN FD (flexible data rate) driver support |
1139 | ||
1140 | CAN FD capable CAN controllers support two different bitrates for the | |
1141 | arbitration phase and the payload phase of the CAN FD frame. Therefore a | |
f35f6c8f | 1142 | second bit timing has to be specified in order to enable the CAN FD bitrate. |
ea53fe0c OH |
1143 | |
1144 | Additionally CAN FD capable CAN controllers support up to 64 bytes of | |
1145 | payload. The representation of this length in can_frame.can_dlc and | |
1146 | canfd_frame.len for userspace applications and inside the Linux network | |
1147 | layer is a plain value from 0 .. 64 instead of the CAN 'data length code'. | |
1148 | The data length code was a 1:1 mapping to the payload length in the legacy | |
1149 | CAN frames anyway. The payload length to the bus-relevant DLC mapping is | |
1150 | only performed inside the CAN drivers, preferably with the helper | |
1151 | functions can_dlc2len() and can_len2dlc(). | |
1152 | ||
1153 | The CAN netdevice driver capabilities can be distinguished by the network | |
1154 | devices maximum transfer unit (MTU): | |
1155 | ||
1156 | MTU = 16 (CAN_MTU) => sizeof(struct can_frame) => 'legacy' CAN device | |
1157 | MTU = 72 (CANFD_MTU) => sizeof(struct canfd_frame) => CAN FD capable device | |
1158 | ||
1159 | The CAN device MTU can be retrieved e.g. with a SIOCGIFMTU ioctl() syscall. | |
1160 | N.B. CAN FD capable devices can also handle and send legacy CAN frames. | |
1161 | ||
1162 | FIXME: Add details about the CAN FD controller configuration when available. | |
1163 | ||
1164 | 6.7 Supported CAN hardware | |
f7ab97f7 | 1165 | |
e20dad96 | 1166 | Please check the "Kconfig" file in "drivers/net/can" to get an actual |
f35f6c8f | 1167 | list of the support CAN hardware. On the SocketCAN project website |
e20dad96 WG |
1168 | (see chapter 7) there might be further drivers available, also for |
1169 | older kernel versions. | |
f7ab97f7 | 1170 | |
f35f6c8f | 1171 | 7. SocketCAN resources |
e20dad96 | 1172 | ----------------------- |
f7ab97f7 | 1173 | |
f35f6c8f JW |
1174 | The Linux CAN / SocketCAN project ressources (project site / mailing list) |
1175 | are referenced in the MAINTAINERS file in the Linux source tree. | |
1176 | Search for CAN NETWORK [LAYERS|DRIVERS]. | |
f7ab97f7 | 1177 | |
e20dad96 | 1178 | 8. Credits |
f7ab97f7 OH |
1179 | ---------- |
1180 | ||
e20dad96 | 1181 | Oliver Hartkopp (PF_CAN core, filters, drivers, bcm, SJA1000 driver) |
f7ab97f7 OH |
1182 | Urs Thuermann (PF_CAN core, kernel integration, socket interfaces, raw, vcan) |
1183 | Jan Kizka (RT-SocketCAN core, Socket-API reconciliation) | |
e20dad96 WG |
1184 | Wolfgang Grandegger (RT-SocketCAN core & drivers, Raw Socket-API reviews, |
1185 | CAN device driver interface, MSCAN driver) | |
f7ab97f7 OH |
1186 | Robert Schwebel (design reviews, PTXdist integration) |
1187 | Marc Kleine-Budde (design reviews, Kernel 2.6 cleanups, drivers) | |
1188 | Benedikt Spranger (reviews) | |
1189 | Thomas Gleixner (LKML reviews, coding style, posting hints) | |
e20dad96 | 1190 | Andrey Volkov (kernel subtree structure, ioctls, MSCAN driver) |
f7ab97f7 OH |
1191 | Matthias Brukner (first SJA1000 CAN netdevice implementation Q2/2003) |
1192 | Klaus Hitschler (PEAK driver integration) | |
1193 | Uwe Koppe (CAN netdevices with PF_PACKET approach) | |
1194 | Michael Schulze (driver layer loopback requirement, RT CAN drivers review) | |
e20dad96 WG |
1195 | Pavel Pisa (Bit-timing calculation) |
1196 | Sascha Hauer (SJA1000 platform driver) | |
1197 | Sebastian Haas (SJA1000 EMS PCI driver) | |
1198 | Markus Plessing (SJA1000 EMS PCI driver) | |
1199 | Per Dalen (SJA1000 Kvaser PCI driver) | |
1200 | Sam Ravnborg (reviews, coding style, kbuild help) |