1 ===================================
2 SocketCAN - Controller Area Network
3 ===================================
5 Overview / What is SocketCAN
6 ============================
8 The socketcan package is an implementation of CAN protocols
9 (Controller Area Network) for Linux. CAN is a networking technology
10 which has widespread use in automation, embedded devices, and
11 automotive fields. While there have been other CAN implementations
12 for Linux based on character devices, SocketCAN uses the Berkeley
13 socket API, the Linux network stack and implements the CAN device
14 drivers as network interfaces. The CAN socket API has been designed
15 as similar as possible to the TCP/IP protocols to allow programmers,
16 familiar with network programming, to easily learn how to use CAN
20 .. _socketcan-motivation:
22 Motivation / Why Using the Socket API
23 =====================================
25 There have been CAN implementations for Linux before SocketCAN so the
26 question arises, why we have started another project. Most existing
27 implementations come as a device driver for some CAN hardware, they
28 are based on character devices and provide comparatively little
29 functionality. Usually, there is only a hardware-specific device
30 driver which provides a character device interface to send and
31 receive raw CAN frames, directly to/from the controller hardware.
32 Queueing of frames and higher-level transport protocols like ISO-TP
33 have to be implemented in user space applications. Also, most
34 character-device implementations support only one single process to
35 open the device at a time, similar to a serial interface. Exchanging
36 the CAN controller requires employment of another device driver and
37 often the need for adaption of large parts of the application to the
40 SocketCAN was designed to overcome all of these limitations. A new
41 protocol family has been implemented which provides a socket interface
42 to user space applications and which builds upon the Linux network
43 layer, enabling use all of the provided queueing functionality. A device
44 driver for CAN controller hardware registers itself with the Linux
45 network layer as a network device, so that CAN frames from the
46 controller can be passed up to the network layer and on to the CAN
47 protocol family module and also vice-versa. Also, the protocol family
48 module provides an API for transport protocol modules to register, so
49 that any number of transport protocols can be loaded or unloaded
50 dynamically. In fact, the can core module alone does not provide any
51 protocol and cannot be used without loading at least one additional
52 protocol module. Multiple sockets can be opened at the same time,
53 on different or the same protocol module and they can listen/send
54 frames on different or the same CAN IDs. Several sockets listening on
55 the same interface for frames with the same CAN ID are all passed the
56 same received matching CAN frames. An application wishing to
57 communicate using a specific transport protocol, e.g. ISO-TP, just
58 selects that protocol when opening the socket, and then can read and
59 write application data byte streams, without having to deal with
62 Similar functionality visible from user-space could be provided by a
63 character device, too, but this would lead to a technically inelegant
64 solution for a couple of reasons:
66 * **Intricate usage:** Instead of passing a protocol argument to
67 socket(2) and using bind(2) to select a CAN interface and CAN ID, an
68 application would have to do all these operations using ioctl(2)s.
70 * **Code duplication:** A character device cannot make use of the Linux
71 network queueing code, so all that code would have to be duplicated
74 * **Abstraction:** In most existing character-device implementations, the
75 hardware-specific device driver for a CAN controller directly
76 provides the character device for the application to work with.
77 This is at least very unusual in Unix systems for both, char and
78 block devices. For example you don't have a character device for a
79 certain UART of a serial interface, a certain sound chip in your
80 computer, a SCSI or IDE controller providing access to your hard
81 disk or tape streamer device. Instead, you have abstraction layers
82 which provide a unified character or block device interface to the
83 application on the one hand, and a interface for hardware-specific
84 device drivers on the other hand. These abstractions are provided
85 by subsystems like the tty layer, the audio subsystem or the SCSI
86 and IDE subsystems for the devices mentioned above.
88 The easiest way to implement a CAN device driver is as a character
89 device without such a (complete) abstraction layer, as is done by most
90 existing drivers. The right way, however, would be to add such a
91 layer with all the functionality like registering for certain CAN
92 IDs, supporting several open file descriptors and (de)multiplexing
93 CAN frames between them, (sophisticated) queueing of CAN frames, and
94 providing an API for device drivers to register with. However, then
95 it would be no more difficult, or may be even easier, to use the
96 networking framework provided by the Linux kernel, and this is what
99 The use of the networking framework of the Linux kernel is just the
100 natural and most appropriate way to implement CAN for Linux.
103 .. _socketcan-concept:
108 As described in :ref:`socketcan-motivation` the main goal of SocketCAN is to
109 provide a socket interface to user space applications which builds
110 upon the Linux network layer. In contrast to the commonly known
111 TCP/IP and ethernet networking, the CAN bus is a broadcast-only(!)
112 medium that has no MAC-layer addressing like ethernet. The CAN-identifier
113 (can_id) is used for arbitration on the CAN-bus. Therefore the CAN-IDs
114 have to be chosen uniquely on the bus. When designing a CAN-ECU
115 network the CAN-IDs are mapped to be sent by a specific ECU.
116 For this reason a CAN-ID can be treated best as a kind of source address.
119 .. _socketcan-receive-lists:
124 The network transparent access of multiple applications leads to the
125 problem that different applications may be interested in the same
126 CAN-IDs from the same CAN network interface. The SocketCAN core
127 module - which implements the protocol family CAN - provides several
128 high efficient receive lists for this reason. If e.g. a user space
129 application opens a CAN RAW socket, the raw protocol module itself
130 requests the (range of) CAN-IDs from the SocketCAN core that are
131 requested by the user. The subscription and unsubscription of
132 CAN-IDs can be done for specific CAN interfaces or for all(!) known
133 CAN interfaces with the can_rx_(un)register() functions provided to
134 CAN protocol modules by the SocketCAN core (see :ref:`socketcan-core-module`).
135 To optimize the CPU usage at runtime the receive lists are split up
136 into several specific lists per device that match the requested
137 filter complexity for a given use-case.
140 .. _socketcan-local-loopback1:
142 Local Loopback of Sent Frames
143 -----------------------------
145 As known from other networking concepts the data exchanging
146 applications may run on the same or different nodes without any
147 change (except for the according addressing information):
151 ___ ___ ___ _______ ___
152 | _ | | _ | | _ | | _ _ | | _ |
153 ||A|| ||B|| ||C|| ||A| |B|| ||C||
154 |___| |___| |___| |_______| |___|
156 -----------------(1)- CAN bus -(2)---------------
158 To ensure that application A receives the same information in the
159 example (2) as it would receive in example (1) there is need for
160 some kind of local loopback of the sent CAN frames on the appropriate
163 The Linux network devices (by default) just can handle the
164 transmission and reception of media dependent frames. Due to the
165 arbitration on the CAN bus the transmission of a low prio CAN-ID
166 may be delayed by the reception of a high prio CAN frame. To
167 reflect the correct [#f1]_ traffic on the node the loopback of the sent
168 data has to be performed right after a successful transmission. If
169 the CAN network interface is not capable of performing the loopback for
170 some reason the SocketCAN core can do this task as a fallback solution.
171 See :ref:`socketcan-local-loopback2` for details (recommended).
173 The loopback functionality is enabled by default to reflect standard
174 networking behaviour for CAN applications. Due to some requests from
175 the RT-SocketCAN group the loopback optionally may be disabled for each
176 separate socket. See sockopts from the CAN RAW sockets in :ref:`socketcan-raw-sockets`.
178 .. [#f1] you really like to have this when you're running analyser
179 tools like 'candump' or 'cansniffer' on the (same) node.
182 .. _socketcan-network-problem-notifications:
184 Network Problem Notifications
185 -----------------------------
187 The use of the CAN bus may lead to several problems on the physical
188 and media access control layer. Detecting and logging of these lower
189 layer problems is a vital requirement for CAN users to identify
190 hardware issues on the physical transceiver layer as well as
191 arbitration problems and error frames caused by the different
192 ECUs. The occurrence of detected errors are important for diagnosis
193 and have to be logged together with the exact timestamp. For this
194 reason the CAN interface driver can generate so called Error Message
195 Frames that can optionally be passed to the user application in the
196 same way as other CAN frames. Whenever an error on the physical layer
197 or the MAC layer is detected (e.g. by the CAN controller) the driver
198 creates an appropriate error message frame. Error messages frames can
199 be requested by the user application using the common CAN filter
200 mechanisms. Inside this filter definition the (interested) type of
201 errors may be selected. The reception of error messages is disabled
202 by default. The format of the CAN error message frame is briefly
203 described in the Linux header file "include/uapi/linux/can/error.h".
209 Like TCP/IP, you first need to open a socket for communicating over a
210 CAN network. Since SocketCAN implements a new protocol family, you
211 need to pass PF_CAN as the first argument to the socket(2) system
212 call. Currently, there are two CAN protocols to choose from, the raw
213 socket protocol and the broadcast manager (BCM). So to open a socket,
216 s = socket(PF_CAN, SOCK_RAW, CAN_RAW);
220 s = socket(PF_CAN, SOCK_DGRAM, CAN_BCM);
222 respectively. After the successful creation of the socket, you would
223 normally use the bind(2) system call to bind the socket to a CAN
224 interface (which is different from TCP/IP due to different addressing
225 - see :ref:`socketcan-concept`). After binding (CAN_RAW) or connecting (CAN_BCM)
226 the socket, you can read(2) and write(2) from/to the socket or use
227 send(2), sendto(2), sendmsg(2) and the recv* counterpart operations
228 on the socket as usual. There are also CAN specific socket options
231 The Classical CAN frame structure (aka CAN 2.0B), the CAN FD frame structure
232 and the sockaddr structure are defined in include/linux/can.h:
237 canid_t can_id; /* 32 bit CAN_ID + EFF/RTR/ERR flags */
239 /* CAN frame payload length in byte (0 .. CAN_MAX_DLEN)
240 * was previously named can_dlc so we need to carry that
241 * name for legacy support
244 __u8 can_dlc; /* deprecated */
246 __u8 __pad; /* padding */
247 __u8 __res0; /* reserved / padding */
248 __u8 len8_dlc; /* optional DLC for 8 byte payload length (9 .. 15) */
249 __u8 data[8] __attribute__((aligned(8)));
252 Remark: The len element contains the payload length in bytes and should be
253 used instead of can_dlc. The deprecated can_dlc was misleadingly named as
254 it always contained the plain payload length in bytes and not the so called
255 'data length code' (DLC).
257 To pass the raw DLC from/to a Classical CAN network device the len8_dlc
258 element can contain values 9 .. 15 when the len element is 8 (the real
259 payload length for all DLC values greater or equal to 8).
261 The alignment of the (linear) payload data[] to a 64bit boundary
262 allows the user to define their own structs and unions to easily access
263 the CAN payload. There is no given byteorder on the CAN bus by
264 default. A read(2) system call on a CAN_RAW socket transfers a
265 struct can_frame to the user space.
267 The sockaddr_can structure has an interface index like the
268 PF_PACKET socket, that also binds to a specific interface:
272 struct sockaddr_can {
273 sa_family_t can_family;
276 /* transport protocol class address info (e.g. ISOTP) */
277 struct { canid_t rx_id, tx_id; } tp;
279 /* J1939 address information */
281 /* 8 byte name when using dynamic addressing */
285 * 8 bit: PS in PDU2 case, else 0
296 /* reserved for future CAN protocols address information */
300 To determine the interface index an appropriate ioctl() has to
301 be used (example for CAN_RAW sockets without error checking):
306 struct sockaddr_can addr;
309 s = socket(PF_CAN, SOCK_RAW, CAN_RAW);
311 strcpy(ifr.ifr_name, "can0" );
312 ioctl(s, SIOCGIFINDEX, &ifr);
314 addr.can_family = AF_CAN;
315 addr.can_ifindex = ifr.ifr_ifindex;
317 bind(s, (struct sockaddr *)&addr, sizeof(addr));
321 To bind a socket to all(!) CAN interfaces the interface index must
322 be 0 (zero). In this case the socket receives CAN frames from every
323 enabled CAN interface. To determine the originating CAN interface
324 the system call recvfrom(2) may be used instead of read(2). To send
325 on a socket that is bound to 'any' interface sendto(2) is needed to
326 specify the outgoing interface.
328 Reading CAN frames from a bound CAN_RAW socket (see above) consists
329 of reading a struct can_frame:
333 struct can_frame frame;
335 nbytes = read(s, &frame, sizeof(struct can_frame));
338 perror("can raw socket read");
342 /* paranoid check ... */
343 if (nbytes < sizeof(struct can_frame)) {
344 fprintf(stderr, "read: incomplete CAN frame\n");
348 /* do something with the received CAN frame */
350 Writing CAN frames can be done similarly, with the write(2) system call::
352 nbytes = write(s, &frame, sizeof(struct can_frame));
354 When the CAN interface is bound to 'any' existing CAN interface
355 (addr.can_ifindex = 0) it is recommended to use recvfrom(2) if the
356 information about the originating CAN interface is needed:
360 struct sockaddr_can addr;
362 socklen_t len = sizeof(addr);
363 struct can_frame frame;
365 nbytes = recvfrom(s, &frame, sizeof(struct can_frame),
366 0, (struct sockaddr*)&addr, &len);
368 /* get interface name of the received CAN frame */
369 ifr.ifr_ifindex = addr.can_ifindex;
370 ioctl(s, SIOCGIFNAME, &ifr);
371 printf("Received a CAN frame from interface %s", ifr.ifr_name);
373 To write CAN frames on sockets bound to 'any' CAN interface the
374 outgoing interface has to be defined certainly:
378 strcpy(ifr.ifr_name, "can0");
379 ioctl(s, SIOCGIFINDEX, &ifr);
380 addr.can_ifindex = ifr.ifr_ifindex;
381 addr.can_family = AF_CAN;
383 nbytes = sendto(s, &frame, sizeof(struct can_frame),
384 0, (struct sockaddr*)&addr, sizeof(addr));
386 An accurate timestamp can be obtained with an ioctl(2) call after reading
387 a message from the socket:
392 ioctl(s, SIOCGSTAMP, &tv);
394 The timestamp has a resolution of one microsecond and is set automatically
395 at the reception of a CAN frame.
397 Remark about CAN FD (flexible data rate) support:
399 Generally the handling of CAN FD is very similar to the formerly described
400 examples. The new CAN FD capable CAN controllers support two different
401 bitrates for the arbitration phase and the payload phase of the CAN FD frame
402 and up to 64 bytes of payload. This extended payload length breaks all the
403 kernel interfaces (ABI) which heavily rely on the CAN frame with fixed eight
404 bytes of payload (struct can_frame) like the CAN_RAW socket. Therefore e.g.
405 the CAN_RAW socket supports a new socket option CAN_RAW_FD_FRAMES that
406 switches the socket into a mode that allows the handling of CAN FD frames
407 and Classical CAN frames simultaneously (see :ref:`socketcan-rawfd`).
409 The struct canfd_frame is defined in include/linux/can.h:
414 canid_t can_id; /* 32 bit CAN_ID + EFF/RTR/ERR flags */
415 __u8 len; /* frame payload length in byte (0 .. 64) */
416 __u8 flags; /* additional flags for CAN FD */
417 __u8 __res0; /* reserved / padding */
418 __u8 __res1; /* reserved / padding */
419 __u8 data[64] __attribute__((aligned(8)));
422 The struct canfd_frame and the existing struct can_frame have the can_id,
423 the payload length and the payload data at the same offset inside their
424 structures. This allows to handle the different structures very similar.
425 When the content of a struct can_frame is copied into a struct canfd_frame
426 all structure elements can be used as-is - only the data[] becomes extended.
428 When introducing the struct canfd_frame it turned out that the data length
429 code (DLC) of the struct can_frame was used as a length information as the
430 length and the DLC has a 1:1 mapping in the range of 0 .. 8. To preserve
431 the easy handling of the length information the canfd_frame.len element
432 contains a plain length value from 0 .. 64. So both canfd_frame.len and
433 can_frame.len are equal and contain a length information and no DLC.
434 For details about the distinction of CAN and CAN FD capable devices and
435 the mapping to the bus-relevant data length code (DLC), see :ref:`socketcan-can-fd-driver`.
437 The length of the two CAN(FD) frame structures define the maximum transfer
438 unit (MTU) of the CAN(FD) network interface and skbuff data length. Two
439 definitions are specified for CAN specific MTUs in include/linux/can.h:
443 #define CAN_MTU (sizeof(struct can_frame)) == 16 => Classical CAN frame
444 #define CANFD_MTU (sizeof(struct canfd_frame)) == 72 => CAN FD frame
447 Returned Message Flags
448 ----------------------
450 When using the system call recvmsg(2) on a RAW or a BCM socket, the
451 msg->msg_flags field may contain the following flags:
454 set when the received frame was created on the local host.
457 set when the frame was sent via the socket it is received on.
458 This flag can be interpreted as a 'transmission confirmation' when the
459 CAN driver supports the echo of frames on driver level, see
460 :ref:`socketcan-local-loopback1` and :ref:`socketcan-local-loopback2`.
461 (Note: In order to receive such messages on a RAW socket,
462 CAN_RAW_RECV_OWN_MSGS must be set.)
465 .. _socketcan-raw-sockets:
467 RAW Protocol Sockets with can_filters (SOCK_RAW)
468 ------------------------------------------------
470 Using CAN_RAW sockets is extensively comparable to the commonly
471 known access to CAN character devices. To meet the new possibilities
472 provided by the multi user SocketCAN approach, some reasonable
473 defaults are set at RAW socket binding time:
475 - The filters are set to exactly one filter receiving everything
476 - The socket only receives valid data frames (=> no error message frames)
477 - The loopback of sent CAN frames is enabled (see :ref:`socketcan-local-loopback2`)
478 - The socket does not receive its own sent frames (in loopback mode)
480 These default settings may be changed before or after binding the socket.
481 To use the referenced definitions of the socket options for CAN_RAW
482 sockets, include <linux/can/raw.h>.
485 .. _socketcan-rawfilter:
487 RAW socket option CAN_RAW_FILTER
488 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
490 The reception of CAN frames using CAN_RAW sockets can be controlled
491 by defining 0 .. n filters with the CAN_RAW_FILTER socket option.
493 The CAN filter structure is defined in include/linux/can.h:
502 A filter matches, when:
506 <received_can_id> & mask == can_id & mask
508 which is analogous to known CAN controllers hardware filter semantics.
509 The filter can be inverted in this semantic, when the CAN_INV_FILTER
510 bit is set in can_id element of the can_filter structure. In
511 contrast to CAN controller hardware filters the user may set 0 .. n
512 receive filters for each open socket separately:
516 struct can_filter rfilter[2];
518 rfilter[0].can_id = 0x123;
519 rfilter[0].can_mask = CAN_SFF_MASK;
520 rfilter[1].can_id = 0x200;
521 rfilter[1].can_mask = 0x700;
523 setsockopt(s, SOL_CAN_RAW, CAN_RAW_FILTER, &rfilter, sizeof(rfilter));
525 To disable the reception of CAN frames on the selected CAN_RAW socket:
529 setsockopt(s, SOL_CAN_RAW, CAN_RAW_FILTER, NULL, 0);
531 To set the filters to zero filters is quite obsolete as to not read
532 data causes the raw socket to discard the received CAN frames. But
533 having this 'send only' use-case we may remove the receive list in the
534 Kernel to save a little (really a very little!) CPU usage.
536 CAN Filter Usage Optimisation
537 .............................
539 The CAN filters are processed in per-device filter lists at CAN frame
540 reception time. To reduce the number of checks that need to be performed
541 while walking through the filter lists the CAN core provides an optimized
542 filter handling when the filter subscription focusses on a single CAN ID.
544 For the possible 2048 SFF CAN identifiers the identifier is used as an index
545 to access the corresponding subscription list without any further checks.
546 For the 2^29 possible EFF CAN identifiers a 10 bit XOR folding is used as
547 hash function to retrieve the EFF table index.
549 To benefit from the optimized filters for single CAN identifiers the
550 CAN_SFF_MASK or CAN_EFF_MASK have to be set into can_filter.mask together
551 with set CAN_EFF_FLAG and CAN_RTR_FLAG bits. A set CAN_EFF_FLAG bit in the
552 can_filter.mask makes clear that it matters whether a SFF or EFF CAN ID is
553 subscribed. E.g. in the example from above:
557 rfilter[0].can_id = 0x123;
558 rfilter[0].can_mask = CAN_SFF_MASK;
560 both SFF frames with CAN ID 0x123 and EFF frames with 0xXXXXX123 can pass.
562 To filter for only 0x123 (SFF) and 0x12345678 (EFF) CAN identifiers the
563 filter has to be defined in this way to benefit from the optimized filters:
567 struct can_filter rfilter[2];
569 rfilter[0].can_id = 0x123;
570 rfilter[0].can_mask = (CAN_EFF_FLAG | CAN_RTR_FLAG | CAN_SFF_MASK);
571 rfilter[1].can_id = 0x12345678 | CAN_EFF_FLAG;
572 rfilter[1].can_mask = (CAN_EFF_FLAG | CAN_RTR_FLAG | CAN_EFF_MASK);
574 setsockopt(s, SOL_CAN_RAW, CAN_RAW_FILTER, &rfilter, sizeof(rfilter));
577 RAW Socket Option CAN_RAW_ERR_FILTER
578 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
580 As described in :ref:`socketcan-network-problem-notifications` the CAN interface driver can generate so
581 called Error Message Frames that can optionally be passed to the user
582 application in the same way as other CAN frames. The possible
583 errors are divided into different error classes that may be filtered
584 using the appropriate error mask. To register for every possible
585 error condition CAN_ERR_MASK can be used as value for the error mask.
586 The values for the error mask are defined in linux/can/error.h:
590 can_err_mask_t err_mask = ( CAN_ERR_TX_TIMEOUT | CAN_ERR_BUSOFF );
592 setsockopt(s, SOL_CAN_RAW, CAN_RAW_ERR_FILTER,
593 &err_mask, sizeof(err_mask));
596 RAW Socket Option CAN_RAW_LOOPBACK
597 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
599 To meet multi user needs the local loopback is enabled by default
600 (see :ref:`socketcan-local-loopback1` for details). But in some embedded use-cases
601 (e.g. when only one application uses the CAN bus) this loopback
602 functionality can be disabled (separately for each socket):
606 int loopback = 0; /* 0 = disabled, 1 = enabled (default) */
608 setsockopt(s, SOL_CAN_RAW, CAN_RAW_LOOPBACK, &loopback, sizeof(loopback));
611 RAW socket option CAN_RAW_RECV_OWN_MSGS
612 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
614 When the local loopback is enabled, all the sent CAN frames are
615 looped back to the open CAN sockets that registered for the CAN
616 frames' CAN-ID on this given interface to meet the multi user
617 needs. The reception of the CAN frames on the same socket that was
618 sending the CAN frame is assumed to be unwanted and therefore
619 disabled by default. This default behaviour may be changed on
624 int recv_own_msgs = 1; /* 0 = disabled (default), 1 = enabled */
626 setsockopt(s, SOL_CAN_RAW, CAN_RAW_RECV_OWN_MSGS,
627 &recv_own_msgs, sizeof(recv_own_msgs));
629 Note that reception of a socket's own CAN frames are subject to the same
630 filtering as other CAN frames (see :ref:`socketcan-rawfilter`).
634 RAW Socket Option CAN_RAW_FD_FRAMES
635 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
637 CAN FD support in CAN_RAW sockets can be enabled with a new socket option
638 CAN_RAW_FD_FRAMES which is off by default. When the new socket option is
639 not supported by the CAN_RAW socket (e.g. on older kernels), switching the
640 CAN_RAW_FD_FRAMES option returns the error -ENOPROTOOPT.
642 Once CAN_RAW_FD_FRAMES is enabled the application can send both CAN frames
643 and CAN FD frames. OTOH the application has to handle CAN and CAN FD frames
644 when reading from the socket:
648 CAN_RAW_FD_FRAMES enabled: CAN_MTU and CANFD_MTU are allowed
649 CAN_RAW_FD_FRAMES disabled: only CAN_MTU is allowed (default)
655 [ remember: CANFD_MTU == sizeof(struct canfd_frame) ]
657 struct canfd_frame cfd;
659 nbytes = read(s, &cfd, CANFD_MTU);
661 if (nbytes == CANFD_MTU) {
662 printf("got CAN FD frame with length %d\n", cfd.len);
663 /* cfd.flags contains valid data */
664 } else if (nbytes == CAN_MTU) {
665 printf("got Classical CAN frame with length %d\n", cfd.len);
666 /* cfd.flags is undefined */
668 fprintf(stderr, "read: invalid CAN(FD) frame\n");
672 /* the content can be handled independently from the received MTU size */
674 printf("can_id: %X data length: %d data: ", cfd.can_id, cfd.len);
675 for (i = 0; i < cfd.len; i++)
676 printf("%02X ", cfd.data[i]);
678 When reading with size CANFD_MTU only returns CAN_MTU bytes that have
679 been received from the socket a Classical CAN frame has been read into the
680 provided CAN FD structure. Note that the canfd_frame.flags data field is
681 not specified in the struct can_frame and therefore it is only valid in
682 CANFD_MTU sized CAN FD frames.
684 Implementation hint for new CAN applications:
686 To build a CAN FD aware application use struct canfd_frame as basic CAN
687 data structure for CAN_RAW based applications. When the application is
688 executed on an older Linux kernel and switching the CAN_RAW_FD_FRAMES
689 socket option returns an error: No problem. You'll get Classical CAN frames
690 or CAN FD frames and can process them the same way.
692 When sending to CAN devices make sure that the device is capable to handle
693 CAN FD frames by checking if the device maximum transfer unit is CANFD_MTU.
694 The CAN device MTU can be retrieved e.g. with a SIOCGIFMTU ioctl() syscall.
697 RAW socket option CAN_RAW_JOIN_FILTERS
698 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
700 The CAN_RAW socket can set multiple CAN identifier specific filters that
701 lead to multiple filters in the af_can.c filter processing. These filters
702 are indenpendent from each other which leads to logical OR'ed filters when
703 applied (see :ref:`socketcan-rawfilter`).
705 This socket option joines the given CAN filters in the way that only CAN
706 frames are passed to user space that matched *all* given CAN filters. The
707 semantic for the applied filters is therefore changed to a logical AND.
709 This is useful especially when the filterset is a combination of filters
710 where the CAN_INV_FILTER flag is set in order to notch single CAN IDs or
711 CAN ID ranges from the incoming traffic.
714 Broadcast Manager Protocol Sockets (SOCK_DGRAM)
715 -----------------------------------------------
717 The Broadcast Manager protocol provides a command based configuration
718 interface to filter and send (e.g. cyclic) CAN messages in kernel space.
720 Receive filters can be used to down sample frequent messages; detect events
721 such as message contents changes, packet length changes, and do time-out
722 monitoring of received messages.
724 Periodic transmission tasks of CAN frames or a sequence of CAN frames can be
725 created and modified at runtime; both the message content and the two
726 possible transmit intervals can be altered.
728 A BCM socket is not intended for sending individual CAN frames using the
729 struct can_frame as known from the CAN_RAW socket. Instead a special BCM
730 configuration message is defined. The basic BCM configuration message used
731 to communicate with the broadcast manager and the available operations are
732 defined in the linux/can/bcm.h include. The BCM message consists of a
733 message header with a command ('opcode') followed by zero or more CAN frames.
734 The broadcast manager sends responses to user space in the same form:
738 struct bcm_msg_head {
739 __u32 opcode; /* command */
740 __u32 flags; /* special flags */
741 __u32 count; /* run 'count' times with ival1 */
742 struct timeval ival1, ival2; /* count and subsequent interval */
743 canid_t can_id; /* unique can_id for task */
744 __u32 nframes; /* number of can_frames following */
745 struct can_frame frames[0];
748 The aligned payload 'frames' uses the same basic CAN frame structure defined
749 at the beginning of :ref:`socketcan-rawfd` and in the include/linux/can.h include. All
750 messages to the broadcast manager from user space have this structure.
752 Note a CAN_BCM socket must be connected instead of bound after socket
753 creation (example without error checking):
758 struct sockaddr_can addr;
761 s = socket(PF_CAN, SOCK_DGRAM, CAN_BCM);
763 strcpy(ifr.ifr_name, "can0");
764 ioctl(s, SIOCGIFINDEX, &ifr);
766 addr.can_family = AF_CAN;
767 addr.can_ifindex = ifr.ifr_ifindex;
769 connect(s, (struct sockaddr *)&addr, sizeof(addr));
773 The broadcast manager socket is able to handle any number of in flight
774 transmissions or receive filters concurrently. The different RX/TX jobs are
775 distinguished by the unique can_id in each BCM message. However additional
776 CAN_BCM sockets are recommended to communicate on multiple CAN interfaces.
777 When the broadcast manager socket is bound to 'any' CAN interface (=> the
778 interface index is set to zero) the configured receive filters apply to any
779 CAN interface unless the sendto() syscall is used to overrule the 'any' CAN
780 interface index. When using recvfrom() instead of read() to retrieve BCM
781 socket messages the originating CAN interface is provided in can_ifindex.
784 Broadcast Manager Operations
785 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
787 The opcode defines the operation for the broadcast manager to carry out,
788 or details the broadcast managers response to several events, including
791 Transmit Operations (user space to broadcast manager):
794 Create (cyclic) transmission task.
797 Remove (cyclic) transmission task, requires only can_id.
800 Read properties of (cyclic) transmission task for can_id.
805 Transmit Responses (broadcast manager to user space):
808 Reply to TX_READ request (transmission task configuration).
811 Notification when counter finishes sending at initial interval
812 'ival1'. Requires the TX_COUNTEVT flag to be set at TX_SETUP.
814 Receive Operations (user space to broadcast manager):
817 Create RX content filter subscription.
820 Remove RX content filter subscription, requires only can_id.
823 Read properties of RX content filter subscription for can_id.
825 Receive Responses (broadcast manager to user space):
828 Reply to RX_READ request (filter task configuration).
831 Cyclic message is detected to be absent (timer ival1 expired).
834 BCM message with updated CAN frame (detected content change).
835 Sent on first message received or on receipt of revised CAN messages.
838 Broadcast Manager Message Flags
839 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
841 When sending a message to the broadcast manager the 'flags' element may
842 contain the following flag definitions which influence the behaviour:
845 Set the values of ival1, ival2 and count
848 Start the timer with the actual values of ival1, ival2
849 and count. Starting the timer leads simultaneously to emit a CAN frame.
852 Create the message TX_EXPIRED when count expires
855 A change of data by the process is emitted immediately.
858 Copies the can_id from the message header to each
859 subsequent frame in frames. This is intended as usage simplification. For
860 TX tasks the unique can_id from the message header may differ from the
861 can_id(s) stored for transmission in the subsequent struct can_frame(s).
864 Filter by can_id alone, no frames required (nframes=0).
867 A change of the DLC leads to an RX_CHANGED.
870 Prevent automatically starting the timeout monitor.
873 If passed at RX_SETUP and a receive timeout occurred, a
874 RX_CHANGED message will be generated when the (cyclic) receive restarts.
877 Reset the index for the multiple frame transmission.
880 Send reply for RTR-request (placed in op->frames[0]).
883 The CAN frames following the bcm_msg_head are struct canfd_frame's
885 Broadcast Manager Transmission Timers
886 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
888 Periodic transmission configurations may use up to two interval timers.
889 In this case the BCM sends a number of messages ('count') at an interval
890 'ival1', then continuing to send at another given interval 'ival2'. When
891 only one timer is needed 'count' is set to zero and only 'ival2' is used.
892 When SET_TIMER and START_TIMER flag were set the timers are activated.
893 The timer values can be altered at runtime when only SET_TIMER is set.
896 Broadcast Manager message sequence transmission
897 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
899 Up to 256 CAN frames can be transmitted in a sequence in the case of a cyclic
900 TX task configuration. The number of CAN frames is provided in the 'nframes'
901 element of the BCM message head. The defined number of CAN frames are added
902 as array to the TX_SETUP BCM configuration message:
906 /* create a struct to set up a sequence of four CAN frames */
908 struct bcm_msg_head msg_head;
909 struct can_frame frame[4];
913 mytxmsg.msg_head.nframes = 4;
916 write(s, &mytxmsg, sizeof(mytxmsg));
918 With every transmission the index in the array of CAN frames is increased
919 and set to zero at index overflow.
922 Broadcast Manager Receive Filter Timers
923 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
925 The timer values ival1 or ival2 may be set to non-zero values at RX_SETUP.
926 When the SET_TIMER flag is set the timers are enabled:
929 Send RX_TIMEOUT when a received message is not received again within
930 the given time. When START_TIMER is set at RX_SETUP the timeout detection
931 is activated directly - even without a former CAN frame reception.
934 Throttle the received message rate down to the value of ival2. This
935 is useful to reduce messages for the application when the signal inside the
936 CAN frame is stateless as state changes within the ival2 period may get
939 Broadcast Manager Multiplex Message Receive Filter
940 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
942 To filter for content changes in multiplex message sequences an array of more
943 than one CAN frames can be passed in a RX_SETUP configuration message. The
944 data bytes of the first CAN frame contain the mask of relevant bits that
945 have to match in the subsequent CAN frames with the received CAN frame.
946 If one of the subsequent CAN frames is matching the bits in that frame data
947 mark the relevant content to be compared with the previous received content.
948 Up to 257 CAN frames (multiplex filter bit mask CAN frame plus 256 CAN
949 filters) can be added as array to the TX_SETUP BCM configuration message:
953 /* usually used to clear CAN frame data[] - beware of endian problems! */
954 #define U64_DATA(p) (*(unsigned long long*)(p)->data)
957 struct bcm_msg_head msg_head;
958 struct can_frame frame[5];
961 msg.msg_head.opcode = RX_SETUP;
962 msg.msg_head.can_id = 0x42;
963 msg.msg_head.flags = 0;
964 msg.msg_head.nframes = 5;
965 U64_DATA(&msg.frame[0]) = 0xFF00000000000000ULL; /* MUX mask */
966 U64_DATA(&msg.frame[1]) = 0x01000000000000FFULL; /* data mask (MUX 0x01) */
967 U64_DATA(&msg.frame[2]) = 0x0200FFFF000000FFULL; /* data mask (MUX 0x02) */
968 U64_DATA(&msg.frame[3]) = 0x330000FFFFFF0003ULL; /* data mask (MUX 0x33) */
969 U64_DATA(&msg.frame[4]) = 0x4F07FC0FF0000000ULL; /* data mask (MUX 0x4F) */
971 write(s, &msg, sizeof(msg));
974 Broadcast Manager CAN FD Support
975 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
977 The programming API of the CAN_BCM depends on struct can_frame which is
978 given as array directly behind the bcm_msg_head structure. To follow this
979 schema for the CAN FD frames a new flag 'CAN_FD_FRAME' in the bcm_msg_head
980 flags indicates that the concatenated CAN frame structures behind the
981 bcm_msg_head are defined as struct canfd_frame:
986 struct bcm_msg_head msg_head;
987 struct canfd_frame frame[5];
990 msg.msg_head.opcode = RX_SETUP;
991 msg.msg_head.can_id = 0x42;
992 msg.msg_head.flags = CAN_FD_FRAME;
993 msg.msg_head.nframes = 5;
996 When using CAN FD frames for multiplex filtering the MUX mask is still
997 expected in the first 64 bit of the struct canfd_frame data section.
1000 Connected Transport Protocols (SOCK_SEQPACKET)
1001 ----------------------------------------------
1006 Unconnected Transport Protocols (SOCK_DGRAM)
1007 --------------------------------------------
1012 .. _socketcan-core-module:
1014 SocketCAN Core Module
1015 =====================
1017 The SocketCAN core module implements the protocol family
1018 PF_CAN. CAN protocol modules are loaded by the core module at
1019 runtime. The core module provides an interface for CAN protocol
1020 modules to subscribe needed CAN IDs (see :ref:`socketcan-receive-lists`).
1023 can.ko Module Params
1024 --------------------
1027 To calculate the SocketCAN core statistics
1028 (e.g. current/maximum frames per second) this 1 second timer is
1029 invoked at can.ko module start time by default. This timer can be
1030 disabled by using stattimer=0 on the module commandline.
1033 (removed since SocketCAN SVN r546)
1039 As described in :ref:`socketcan-receive-lists` the SocketCAN core uses several filter
1040 lists to deliver received CAN frames to CAN protocol modules. These
1041 receive lists, their filters and the count of filter matches can be
1042 checked in the appropriate receive list. All entries contain the
1043 device and a protocol module identifier::
1045 foo@bar:~$ cat /proc/net/can/rcvlist_all
1047 receive list 'rx_all':
1051 device can_id can_mask function userdata matches ident
1052 vcan0 000 00000000 f88e6370 f6c6f400 0 raw
1055 In this example an application requests any CAN traffic from vcan0::
1057 rcvlist_all - list for unfiltered entries (no filter operations)
1058 rcvlist_eff - list for single extended frame (EFF) entries
1059 rcvlist_err - list for error message frames masks
1060 rcvlist_fil - list for mask/value filters
1061 rcvlist_inv - list for mask/value filters (inverse semantic)
1062 rcvlist_sff - list for single standard frame (SFF) entries
1064 Additional procfs files in /proc/net/can::
1066 stats - SocketCAN core statistics (rx/tx frames, match ratios, ...)
1067 reset_stats - manual statistic reset
1068 version - prints SocketCAN core and ABI version (removed in Linux 5.10)
1071 Writing Own CAN Protocol Modules
1072 --------------------------------
1074 To implement a new protocol in the protocol family PF_CAN a new
1075 protocol has to be defined in include/linux/can.h .
1076 The prototypes and definitions to use the SocketCAN core can be
1077 accessed by including include/linux/can/core.h .
1078 In addition to functions that register the CAN protocol and the
1079 CAN device notifier chain there are functions to subscribe CAN
1080 frames received by CAN interfaces and to send CAN frames::
1082 can_rx_register - subscribe CAN frames from a specific interface
1083 can_rx_unregister - unsubscribe CAN frames from a specific interface
1084 can_send - transmit a CAN frame (optional with local loopback)
1086 For details see the kerneldoc documentation in net/can/af_can.c or
1087 the source code of net/can/raw.c or net/can/bcm.c .
1093 Writing a CAN network device driver is much easier than writing a
1094 CAN character device driver. Similar to other known network device
1095 drivers you mainly have to deal with:
1097 - TX: Put the CAN frame from the socket buffer to the CAN controller.
1098 - RX: Put the CAN frame from the CAN controller to the socket buffer.
1100 See e.g. at Documentation/networking/netdevices.rst . The differences
1101 for writing CAN network device driver are described below:
1109 dev->type = ARPHRD_CAN; /* the netdevice hardware type */
1110 dev->flags = IFF_NOARP; /* CAN has no arp */
1112 dev->mtu = CAN_MTU; /* sizeof(struct can_frame) -> Classical CAN interface */
1114 or alternative, when the controller supports CAN with flexible data rate:
1115 dev->mtu = CANFD_MTU; /* sizeof(struct canfd_frame) -> CAN FD interface */
1117 The struct can_frame or struct canfd_frame is the payload of each socket
1118 buffer (skbuff) in the protocol family PF_CAN.
1121 .. _socketcan-local-loopback2:
1123 Local Loopback of Sent Frames
1124 -----------------------------
1126 As described in :ref:`socketcan-local-loopback1` the CAN network device driver should
1127 support a local loopback functionality similar to the local echo
1128 e.g. of tty devices. In this case the driver flag IFF_ECHO has to be
1129 set to prevent the PF_CAN core from locally echoing sent frames
1130 (aka loopback) as fallback solution::
1132 dev->flags = (IFF_NOARP | IFF_ECHO);
1135 CAN Controller Hardware Filters
1136 -------------------------------
1138 To reduce the interrupt load on deep embedded systems some CAN
1139 controllers support the filtering of CAN IDs or ranges of CAN IDs.
1140 These hardware filter capabilities vary from controller to
1141 controller and have to be identified as not feasible in a multi-user
1142 networking approach. The use of the very controller specific
1143 hardware filters could make sense in a very dedicated use-case, as a
1144 filter on driver level would affect all users in the multi-user
1145 system. The high efficient filter sets inside the PF_CAN core allow
1146 to set different multiple filters for each socket separately.
1147 Therefore the use of hardware filters goes to the category 'handmade
1148 tuning on deep embedded systems'. The author is running a MPC603e
1149 @133MHz with four SJA1000 CAN controllers from 2002 under heavy bus
1150 load without any problems ...
1153 Switchable Termination Resistors
1154 --------------------------------
1156 CAN bus requires a specific impedance across the differential pair,
1157 typically provided by two 120Ohm resistors on the farthest nodes of
1158 the bus. Some CAN controllers support activating / deactivating a
1159 termination resistor(s) to provide the correct impedance.
1161 Query the available resistances::
1163 $ ip -details link show can0
1165 termination 120 [ 0, 120 ]
1167 Activate the terminating resistor::
1169 $ ip link set dev can0 type can termination 120
1171 Deactivate the terminating resistor::
1173 $ ip link set dev can0 type can termination 0
1175 To enable termination resistor support to a can-controller, either
1176 implement in the controller's struct can-priv::
1179 termination_const_cnt
1182 or add gpio control with the device tree entries from
1183 Documentation/devicetree/bindings/net/can/can-controller.yaml
1186 The Virtual CAN Driver (vcan)
1187 -----------------------------
1189 Similar to the network loopback devices, vcan offers a virtual local
1190 CAN interface. A full qualified address on CAN consists of
1192 - a unique CAN Identifier (CAN ID)
1193 - the CAN bus this CAN ID is transmitted on (e.g. can0)
1195 so in common use cases more than one virtual CAN interface is needed.
1197 The virtual CAN interfaces allow the transmission and reception of CAN
1198 frames without real CAN controller hardware. Virtual CAN network
1199 devices are usually named 'vcanX', like vcan0 vcan1 vcan2 ...
1200 When compiled as a module the virtual CAN driver module is called vcan.ko
1202 Since Linux Kernel version 2.6.24 the vcan driver supports the Kernel
1203 netlink interface to create vcan network devices. The creation and
1204 removal of vcan network devices can be managed with the ip(8) tool::
1206 - Create a virtual CAN network interface:
1207 $ ip link add type vcan
1209 - Create a virtual CAN network interface with a specific name 'vcan42':
1210 $ ip link add dev vcan42 type vcan
1212 - Remove a (virtual CAN) network interface 'vcan42':
1213 $ ip link del vcan42
1216 The CAN Network Device Driver Interface
1217 ---------------------------------------
1219 The CAN network device driver interface provides a generic interface
1220 to setup, configure and monitor CAN network devices. The user can then
1221 configure the CAN device, like setting the bit-timing parameters, via
1222 the netlink interface using the program "ip" from the "IPROUTE2"
1223 utility suite. The following chapter describes briefly how to use it.
1224 Furthermore, the interface uses a common data structure and exports a
1225 set of common functions, which all real CAN network device drivers
1226 should use. Please have a look to the SJA1000 or MSCAN driver to
1227 understand how to use them. The name of the module is can-dev.ko.
1230 Netlink interface to set/get devices properties
1231 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1233 The CAN device must be configured via netlink interface. The supported
1234 netlink message types are defined and briefly described in
1235 "include/linux/can/netlink.h". CAN link support for the program "ip"
1236 of the IPROUTE2 utility suite is available and it can be used as shown
1239 Setting CAN device properties::
1241 $ ip link set can0 type can help
1242 Usage: ip link set DEVICE type can
1243 [ bitrate BITRATE [ sample-point SAMPLE-POINT] ] |
1244 [ tq TQ prop-seg PROP_SEG phase-seg1 PHASE-SEG1
1245 phase-seg2 PHASE-SEG2 [ sjw SJW ] ]
1247 [ dbitrate BITRATE [ dsample-point SAMPLE-POINT] ] |
1248 [ dtq TQ dprop-seg PROP_SEG dphase-seg1 PHASE-SEG1
1249 dphase-seg2 PHASE-SEG2 [ dsjw SJW ] ]
1251 [ loopback { on | off } ]
1252 [ listen-only { on | off } ]
1253 [ triple-sampling { on | off } ]
1254 [ one-shot { on | off } ]
1255 [ berr-reporting { on | off } ]
1257 [ fd-non-iso { on | off } ]
1258 [ presume-ack { on | off } ]
1259 [ cc-len8-dlc { on | off } ]
1261 [ restart-ms TIME-MS ]
1264 Where: BITRATE := { 1..1000000 }
1265 SAMPLE-POINT := { 0.000..0.999 }
1267 PROP-SEG := { 1..8 }
1268 PHASE-SEG1 := { 1..8 }
1269 PHASE-SEG2 := { 1..8 }
1271 RESTART-MS := { 0 | NUMBER }
1273 Display CAN device details and statistics::
1275 $ ip -details -statistics link show can0
1276 2: can0: <NOARP,UP,LOWER_UP,ECHO> mtu 16 qdisc pfifo_fast state UP qlen 10
1278 can <TRIPLE-SAMPLING> state ERROR-ACTIVE restart-ms 100
1279 bitrate 125000 sample_point 0.875
1280 tq 125 prop-seg 6 phase-seg1 7 phase-seg2 2 sjw 1
1281 sja1000: tseg1 1..16 tseg2 1..8 sjw 1..4 brp 1..64 brp-inc 1
1283 re-started bus-errors arbit-lost error-warn error-pass bus-off
1285 RX: bytes packets errors dropped overrun mcast
1286 140859 17608 17457 0 0 0
1287 TX: bytes packets errors dropped carrier collsns
1290 More info to the above output:
1293 Shows the list of selected CAN controller modes: LOOPBACK,
1294 LISTEN-ONLY, or TRIPLE-SAMPLING.
1296 "state ERROR-ACTIVE"
1297 The current state of the CAN controller: "ERROR-ACTIVE",
1298 "ERROR-WARNING", "ERROR-PASSIVE", "BUS-OFF" or "STOPPED"
1301 Automatic restart delay time. If set to a non-zero value, a
1302 restart of the CAN controller will be triggered automatically
1303 in case of a bus-off condition after the specified delay time
1304 in milliseconds. By default it's off.
1306 "bitrate 125000 sample-point 0.875"
1307 Shows the real bit-rate in bits/sec and the sample-point in the
1308 range 0.000..0.999. If the calculation of bit-timing parameters
1309 is enabled in the kernel (CONFIG_CAN_CALC_BITTIMING=y), the
1310 bit-timing can be defined by setting the "bitrate" argument.
1311 Optionally the "sample-point" can be specified. By default it's
1312 0.000 assuming CIA-recommended sample-points.
1314 "tq 125 prop-seg 6 phase-seg1 7 phase-seg2 2 sjw 1"
1315 Shows the time quanta in ns, propagation segment, phase buffer
1316 segment 1 and 2 and the synchronisation jump width in units of
1317 tq. They allow to define the CAN bit-timing in a hardware
1318 independent format as proposed by the Bosch CAN 2.0 spec (see
1319 chapter 8 of http://www.semiconductors.bosch.de/pdf/can2spec.pdf).
1321 "sja1000: tseg1 1..16 tseg2 1..8 sjw 1..4 brp 1..64 brp-inc 1 clock 8000000"
1322 Shows the bit-timing constants of the CAN controller, here the
1323 "sja1000". The minimum and maximum values of the time segment 1
1324 and 2, the synchronisation jump width in units of tq, the
1325 bitrate pre-scaler and the CAN system clock frequency in Hz.
1326 These constants could be used for user-defined (non-standard)
1327 bit-timing calculation algorithms in user-space.
1329 "re-started bus-errors arbit-lost error-warn error-pass bus-off"
1330 Shows the number of restarts, bus and arbitration lost errors,
1331 and the state changes to the error-warning, error-passive and
1332 bus-off state. RX overrun errors are listed in the "overrun"
1333 field of the standard network statistics.
1335 Setting the CAN Bit-Timing
1336 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1338 The CAN bit-timing parameters can always be defined in a hardware
1339 independent format as proposed in the Bosch CAN 2.0 specification
1340 specifying the arguments "tq", "prop_seg", "phase_seg1", "phase_seg2"
1343 $ ip link set canX type can tq 125 prop-seg 6 \
1344 phase-seg1 7 phase-seg2 2 sjw 1
1346 If the kernel option CONFIG_CAN_CALC_BITTIMING is enabled, CIA
1347 recommended CAN bit-timing parameters will be calculated if the bit-
1348 rate is specified with the argument "bitrate"::
1350 $ ip link set canX type can bitrate 125000
1352 Note that this works fine for the most common CAN controllers with
1353 standard bit-rates but may *fail* for exotic bit-rates or CAN system
1354 clock frequencies. Disabling CONFIG_CAN_CALC_BITTIMING saves some
1355 space and allows user-space tools to solely determine and set the
1356 bit-timing parameters. The CAN controller specific bit-timing
1357 constants can be used for that purpose. They are listed by the
1360 $ ip -details link show can0
1362 sja1000: clock 8000000 tseg1 1..16 tseg2 1..8 sjw 1..4 brp 1..64 brp-inc 1
1365 Starting and Stopping the CAN Network Device
1366 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1368 A CAN network device is started or stopped as usual with the command
1369 "ifconfig canX up/down" or "ip link set canX up/down". Be aware that
1370 you *must* define proper bit-timing parameters for real CAN devices
1371 before you can start it to avoid error-prone default settings::
1373 $ ip link set canX up type can bitrate 125000
1375 A device may enter the "bus-off" state if too many errors occurred on
1376 the CAN bus. Then no more messages are received or sent. An automatic
1377 bus-off recovery can be enabled by setting the "restart-ms" to a
1378 non-zero value, e.g.::
1380 $ ip link set canX type can restart-ms 100
1382 Alternatively, the application may realize the "bus-off" condition
1383 by monitoring CAN error message frames and do a restart when
1384 appropriate with the command::
1386 $ ip link set canX type can restart
1388 Note that a restart will also create a CAN error message frame (see
1389 also :ref:`socketcan-network-problem-notifications`).
1392 .. _socketcan-can-fd-driver:
1394 CAN FD (Flexible Data Rate) Driver Support
1395 ------------------------------------------
1397 CAN FD capable CAN controllers support two different bitrates for the
1398 arbitration phase and the payload phase of the CAN FD frame. Therefore a
1399 second bit timing has to be specified in order to enable the CAN FD bitrate.
1401 Additionally CAN FD capable CAN controllers support up to 64 bytes of
1402 payload. The representation of this length in can_frame.len and
1403 canfd_frame.len for userspace applications and inside the Linux network
1404 layer is a plain value from 0 .. 64 instead of the CAN 'data length code'.
1405 The data length code was a 1:1 mapping to the payload length in the Classical
1406 CAN frames anyway. The payload length to the bus-relevant DLC mapping is
1407 only performed inside the CAN drivers, preferably with the helper
1408 functions can_fd_dlc2len() and can_fd_len2dlc().
1410 The CAN netdevice driver capabilities can be distinguished by the network
1411 devices maximum transfer unit (MTU)::
1413 MTU = 16 (CAN_MTU) => sizeof(struct can_frame) => Classical CAN device
1414 MTU = 72 (CANFD_MTU) => sizeof(struct canfd_frame) => CAN FD capable device
1416 The CAN device MTU can be retrieved e.g. with a SIOCGIFMTU ioctl() syscall.
1417 N.B. CAN FD capable devices can also handle and send Classical CAN frames.
1419 When configuring CAN FD capable CAN controllers an additional 'data' bitrate
1420 has to be set. This bitrate for the data phase of the CAN FD frame has to be
1421 at least the bitrate which was configured for the arbitration phase. This
1422 second bitrate is specified analogue to the first bitrate but the bitrate
1423 setting keywords for the 'data' bitrate start with 'd' e.g. dbitrate,
1424 dsample-point, dsjw or dtq and similar settings. When a data bitrate is set
1425 within the configuration process the controller option "fd on" can be
1426 specified to enable the CAN FD mode in the CAN controller. This controller
1427 option also switches the device MTU to 72 (CANFD_MTU).
1429 The first CAN FD specification presented as whitepaper at the International
1430 CAN Conference 2012 needed to be improved for data integrity reasons.
1431 Therefore two CAN FD implementations have to be distinguished today:
1433 - ISO compliant: The ISO 11898-1:2015 CAN FD implementation (default)
1434 - non-ISO compliant: The CAN FD implementation following the 2012 whitepaper
1436 Finally there are three types of CAN FD controllers:
1438 1. ISO compliant (fixed)
1439 2. non-ISO compliant (fixed, like the M_CAN IP core v3.0.1 in m_can.c)
1440 3. ISO/non-ISO CAN FD controllers (switchable, like the PEAK PCAN-USB FD)
1442 The current ISO/non-ISO mode is announced by the CAN controller driver via
1443 netlink and displayed by the 'ip' tool (controller option FD-NON-ISO).
1444 The ISO/non-ISO-mode can be altered by setting 'fd-non-iso {on|off}' for
1445 switchable CAN FD controllers only.
1447 Example configuring 500 kbit/s arbitration bitrate and 4 Mbit/s data bitrate::
1449 $ ip link set can0 up type can bitrate 500000 sample-point 0.75 \
1450 dbitrate 4000000 dsample-point 0.8 fd on
1451 $ ip -details link show can0
1452 5: can0: <NOARP,UP,LOWER_UP,ECHO> mtu 72 qdisc pfifo_fast state UNKNOWN \
1453 mode DEFAULT group default qlen 10
1454 link/can promiscuity 0
1455 can <FD> state ERROR-ACTIVE (berr-counter tx 0 rx 0) restart-ms 0
1456 bitrate 500000 sample-point 0.750
1457 tq 50 prop-seg 14 phase-seg1 15 phase-seg2 10 sjw 1
1458 pcan_usb_pro_fd: tseg1 1..64 tseg2 1..16 sjw 1..16 brp 1..1024 \
1460 dbitrate 4000000 dsample-point 0.800
1461 dtq 12 dprop-seg 7 dphase-seg1 8 dphase-seg2 4 dsjw 1
1462 pcan_usb_pro_fd: dtseg1 1..16 dtseg2 1..8 dsjw 1..4 dbrp 1..1024 \
1466 Example when 'fd-non-iso on' is added on this switchable CAN FD adapter::
1468 can <FD,FD-NON-ISO> state ERROR-ACTIVE (berr-counter tx 0 rx 0) restart-ms 0
1471 Supported CAN Hardware
1472 ----------------------
1474 Please check the "Kconfig" file in "drivers/net/can" to get an actual
1475 list of the support CAN hardware. On the SocketCAN project website
1476 (see :ref:`socketcan-resources`) there might be further drivers available, also for
1477 older kernel versions.
1480 .. _socketcan-resources:
1485 The Linux CAN / SocketCAN project resources (project site / mailing list)
1486 are referenced in the MAINTAINERS file in the Linux source tree.
1487 Search for CAN NETWORK [LAYERS|DRIVERS].
1492 - Oliver Hartkopp (PF_CAN core, filters, drivers, bcm, SJA1000 driver)
1493 - Urs Thuermann (PF_CAN core, kernel integration, socket interfaces, raw, vcan)
1494 - Jan Kizka (RT-SocketCAN core, Socket-API reconciliation)
1495 - Wolfgang Grandegger (RT-SocketCAN core & drivers, Raw Socket-API reviews, CAN device driver interface, MSCAN driver)
1496 - Robert Schwebel (design reviews, PTXdist integration)
1497 - Marc Kleine-Budde (design reviews, Kernel 2.6 cleanups, drivers)
1498 - Benedikt Spranger (reviews)
1499 - Thomas Gleixner (LKML reviews, coding style, posting hints)
1500 - Andrey Volkov (kernel subtree structure, ioctls, MSCAN driver)
1501 - Matthias Brukner (first SJA1000 CAN netdevice implementation Q2/2003)
1502 - Klaus Hitschler (PEAK driver integration)
1503 - Uwe Koppe (CAN netdevices with PF_PACKET approach)
1504 - Michael Schulze (driver layer loopback requirement, RT CAN drivers review)
1505 - Pavel Pisa (Bit-timing calculation)
1506 - Sascha Hauer (SJA1000 platform driver)
1507 - Sebastian Haas (SJA1000 EMS PCI driver)
1508 - Markus Plessing (SJA1000 EMS PCI driver)
1509 - Per Dalen (SJA1000 Kvaser PCI driver)
1510 - Sam Ravnborg (reviews, coding style, kbuild help)