With commit 33a9249571 ("slcand: daemonise only once setup is complete")
the daemonize syscall is moved behind serial tty setup.
This patch moves the missing assignment of the loop variable slcand_running
to get it into the code context of the while() loop.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Hardware timestamps have been introduced in Linux 2.6.30. Check for the
availability of SO_TIMESTAMPING and print an appropriate error message if
missing. Additionally fix some style issues and define the proper length
of ctrlmsg[].
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
GitHub user 'crossband' raised an issue regarding the strict-aliasing compiler
warning in his specific setup: https://github.com/linux-can/can-utils/issues/42
In fact memcpy() and memset() are a better solution than the former pointer
magics, so remove the issues and the compiler warning flag too.
Reported-by: crossband (https://github.com/crossband)
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Alan sent a pull request to update the README.md with some markdown tweaks
and a SocketCAN logo. Good idea!
After creating a SocketCAN logo PNG in the can-logos repository we can
reference the PNG inside the linux-can organization.
A license file needs to be added too - but this has to be discussed on the
mailing list first.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
The cmdlinename[i] output contains the filter definitions e.g.
'can0,300:700' but you want to see only 'can0' when printing on stderr.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
listen-only option (-l) has been added to slcan-attach in oct 2015.
It is now replicated in slcand
Signed-off-by: interlocuteur <gallard@mmic.net>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Upstream commit ba61a8d9d780980e8284355a0be750897e7af212
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit?id=ba61a8d9d780980e8284355a0be750897e7af212
The can subsystem communicates with user space using a bcm_msg_head
header, which contains two timestamps. This is problematic for
multiple reasons:
a) The structure layout is currently incompatible between 64-bit
user space and 32-bit user space, and cannot work in compat
mode (other than x32).
b) The timeval structure layout will change in 32-bit user
space when we fix the y2038 overflow problem by redefining
time_t to 64-bit, making new 32-bit user space incompatible
with the current kernel interface.
Cars last a long time and often use old kernels, so the actual
users of this code are the most likely ones to migrate to y2038
safe user space.
This tries to work around part of the problem by changing the
publicly visible user interface in the header, but not the binary
interface. Fortunately, the values passed around in the structure
are relative times and do not actually suffer from the y2038
overflow, so 32-bit is enough here.
We replace the use of 'struct timeval' with a newly defined
'struct bcm_timeval' that uses the exact same binary layout
as before and that still suffers from problem a) but not problem
b).
The downside of this approach is that any user space program
that currently assigns a timeval structure to these members
rather than writing the tv_sec/tv_usec portions individually
will suffer a compile-time error when built with an updated
kernel header. Fixing this error makes it work fine with old
and new headers though.
We could address problem a) by using '__u32' or 'int' members
rather than 'long', but that would have a more significant
downside in also breaking support for all existing 64-bit user
binaries that might be using this interface, which is likely
not acceptable.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Oliver Hartkopp <socketcan@hartkopp.net>
Cc: linux-can@vger.kernel.org
Cc: linux-api@vger.kernel.org
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Similar to referencing iptables rules by their line number this UID allows to
reference created routing jobs, e.g. to alter configured data modifications.
The UID is an optional non-zero value which can be provided at routing job
creation time. When the UID is set the UID replaces the data modification
configuration as job identification attribute e.g. at job removal time.
The UID option is provided by mainline Linux 4.2+
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
The CAN_RAW socket can set multiple CAN identifier specific filters that lead
to multiple filters in the af_can.c filter processing. These filters are
indenpendent from each other which leads to logical OR'ed filters when applied.
This socket option joines the given CAN filters in the way that only CAN frames
are passed to user space that matched *all* given CAN filters. The semantic for
the applied filters is therefore changed to a logical AND.
This is useful especially when the filterset is a combination of filters where
the CAN_INV_FILTER flag is set in order to notch single CAN IDs or CAN ID
ranges from the incoming traffic.
Example:
candump can0,100~7FF,200~7FF,400~7FF (logical OR'ed filters)
candump can0,100~7FF,200~7FF,400~7FF,J (logical AND'ed filters)
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Allow to specify the TX and RX padding content separately.
Due to the optional values it is possible to provide either TX or RX values
which enables and set the given values.
-p 55:AA => TX padding 0x55, RX padding 0xAA
-p 55 => TX padding 0x55
-p 55: => TX padding 0x55
-p :AA => RX padding 0xAA
Additionally the comment for the -P <mode> option was updated to point out the
checking mode of the padding in the received PDUs (rx padding).
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Enable to pass the optional second extended address with the '-x' option and
do not pollute the command line options with just another '-X' option.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
As requested by Laurent Vaudoit the extended address can be different in the
tx and he rx path:
(..) how can i have a segmented transfer like this:
0x6a7 0x55 0x10 0x08 ........
0x687 0xAA 0x30 0x00 0x00
0x6a7 0x55 0x21 .....
The connection i need is between two ECU, using IDs 0x6a7/687 and one has
adress extension 0x55, the other 0xAA (this adressing method is used on some
FIAT ECUs for example).
http://marc.info/?l=linux-can&m=140354647413513&w=2
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Allow PDU sizes greater than 4095 to be able to check the correct
implementation of the kernel socket API error handlings.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Inspired by the SLCAN rework from Andre Naujoks this patch replaces the
extensive use of sprintf() with simple and efficient ASCII helpers.
valgrind/kcachegrind reported a cyle estimation of about 5.3 million cycles
before and about 1.5 million cycles after the rework (factor 3.55) for the
logging of the exact same CAN content.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
The creation of the logfile format was using fprint_canframe() and two other
fprintf() calls to produce a single logfile format line. Instead of using
fprint_canframe() use sprint_canframe() directly and reduce the printf() calls
from three to one.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Large file support (LFS) in recent C libraries allows 32 bit applications to
create large files with file length values > 0x7FFFFFFF when the filesystem
supports it.
This patch adds large file support to the Makefile, if you don't want LFS use
the autotools version and call configure like this:
./configure --disable-largefile
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
When printing the current time on the first line the bitstuffing mode is
displayed depending on the command line settings.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
The dummyFile pointer was only needed to redirect stdxxx streams but it's
content was never used. Re-use the pFile pointer for this reason which is
unused at that point to silent the compiler warning about the set and unused
dummyFile variable.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Fixed coding style inspired by Linux checkpatch.pl
Additionally removed a comment in the write() result check in the 'if (speed)'
statement to make sure the write() is not optimized away as the executed code
was empty.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
slcand takes settings known from slcan_attach to configure CAN
interface, like open/close channel, set bitrate etc (Oliver Hartkopp).
In addition -S option sets UART's baudrate (Yegor Yefremov).
This way slcand combines three autilities in one: old slcand,
slcan_attach and stty.
Example: slcand -o -s8 -S 3000000 ttyUSB0
Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Reviewed-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
/dev/null returns EOF therefore select() finishes immediately.
Now EOF is probed on start.
Other /dev/null as stdin workarounds:
- lscanpty ... < /dev/ptmx (dirty but works)
- cat </dev/null | slcanpty ... (doesn't work for me and even the patch
is "disabled" then)
Signed-off-by: Janusz Uzycki <j.uzycki@elproma.com.pl>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Make sure that we have minimum one complete SLCAN messages from pty in the
receive buffer before we start processing the received SLCAN message.
Fragments of a received incomplete message are stored to be appended by the
next read() syscall in pty2can().
This patch is a rework of an initial patch from Ulrich Escher. Tnx!
Reported-by: Ulrich Escher <git@myvdr.de>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Terminate slcanpty when the pty file descriptor is closed instead of performing
an endless loop in the select() statement.
Reported-by: Janusz Uzycki <j.uzycki@elproma.com.pl>
Signed-off-by: Janusz Uzycki <j.uzycki@elproma.com.pl>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
As Hakan Engblom pointed out a CAN 2.0B RTR frame can contain a non-zero DLC
value which was not addressed by parse_canframe() and sprint_canframe().
This patch adds support for non-zero DLC values in a way that legacy logfiles
are still usable as the DLC can optionally be added.
Reported-by: Hakan Engblom <H.Engblom@tele-radio.com>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
A configured inverse filter e.g.
cangw -A -s can0 -d can1 -e -f 123~7FF
has been displayed with 'cangw -L' like this:
cangw -A -s can0 -d can1 -e -f 20000123:7FF # 0 handled 0 dropped 0 deleted
While 0x20000000 was the value of CAN_INV_FILTER.
This patch fixes the output to the expected display:
cangw -A -s can0 -d can1 -e -f 123~7FF # 0 handled 0 dropped 0 deleted
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Added new option '-i' for can-gw rules to allow to route CAN frames back to the
originating/incoming CAN interface (which is disabled by default).
CAN frames that are deleted due to the violation of the max_hops limit are now
printed when displaying the configured rules with 'cangw -L'.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
nbytes is used after writing the CAN frame to check for an additional command
in the string buffer. Therefore using nbytes while writing the CAN frame is
wrong. Replace it with 'tmp' which can be used here without problems.
Signed-off-by: ulrich escher<git@myvdr.de>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
The options to for adding checksums (-c or -x) can only be used in conjunction
with modifications (-m) applied to the routed CAN frames.
Signed-off-by: Andre Naujoks <nautsch2@googlemail.com>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Added -x option to print extra message infos per frame:
RX/TX : Indicates whether the frame was sent or received from the local host
BRS : bit rate setting enabled (CAN FD only)
ESI : error state indicator (CAN FD only)
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Added new view CANLIB_VIEW_INDENT_SFF flags to fix the sloppy output of
fprint_long_canframe() when mixing EFF & SFF CAN identifiers.
candump: Once an EFF frame is detected the indention is enabled.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
This is a major upgrade of the basic tools to handle CAN FD frames.
The library to parse and print CAN frames and logfiles has been extended.
In detail:
asc2log.c | 5 +
candump.c | 24 ++++---
cangen.c | 172 +++++++++++++++++++++++++++++++++------------------
canlogserver.c | 28 +++++---
canplayer.c | 25 ++++---
cansend.c | 55 ++++++++++++----
lib.c | 189 ++++++++++++++++++++++++++++++++++++++-------------------
lib.h | 109 ++++++++++++++++++++++++--------
log2asc.c | 8 +-
log2long.c | 26 ++++++-
10 files changed, 440 insertions(+), 201 deletions(-)
asc2log.c / log2asc.c
- updates for new lib functions
- still can only handle CAN2.0 frames (no new info about ASC file layout)
log2long.c / canlogserver.c / canplayer.c
- updates for new lib functions to handle CAN FD
lib.h / lib.c
- reworked lib functions to handle CAN FD
- parse_canframe() now returns CAN_MTU and CANFD_MTU on success, 0 at failure
- added can_dlc2len() and can_len2dlc() helpers
- moved hexstring2candata to hexstring2data to support simple byte buffers
- in the long CAN frame representation use %03X/%08X instead of %3X/%8X
- introduced unified buffer size definitions for ASCII CAN frames
- updated documentation
cangen.c
- support CAN FD frames (added -f option to create CAN FD frames)
- added -m option ('mix') to create random extended / RTR / CAN FD frames
- fixed the 'fixed data' option which was zero'ing the payload by the time
- updated help text
candump.c
- support CAN FD frames (print, bridge, log)
- distinguish frame types by length info: [0] = CAN2.0 [00] = CAN FD frame
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
There has been a change with __kernel_sa_family_t in Linux 3.1 which was not
adopted in this update so far to be backward compatible with old environments.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
This way a user can specify a gap below 1ms. This is somehwat similar
to setting the gap to 0, but the ENOBUFS are greatly reduced, so a given
amount of CAN messages by -n is still usefull and useable.
Signed-off-by: Alexander Stein <alexander.stein@systec-electronic.com>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
As pointed out by Andre Naujoks the pseudo random generator should be
initialized with a seed to prevend this ugly behaviour:
$ ./cangen vcan2 -v & ./cangen vcan1 -v
[1] 5995
vcan2 567#69.98.3C.64.73.48
vcan1 567#69.98.3C.64.73.48
vcan2 451#4A.94.E8.2A.EC.58.55.62
vcan1 451#4A.94.E8.2A.EC.58.55.62
vcan2 729#BA.58.1B.3D.AB.D7.7E.50
vcan1 729#BA.58.1B.3D.AB.D7.7E.50
(..)
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
In the case of pending bus errors on some slcan adapters the read of the
status flags resets the adapter and clears the error LED.
This patch adds the commandline option to send the status read command to the
adapter as suggested by Thorsten Godau.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
As investigated by Thorsten Godau the bitrate can only be set after the
slcan adapter is in 'close' state.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Tested-by: Thorsten Godau <dl9sec@gmx.net>
The buffers to build the pathnames had been malloc'ed and never been free'd.
This patch removes the malloc stuff entirely.
Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
misbehave for regression tests.
CAN_ISOTP_TX_STMIN:
Set a value in nano secs that's used as minimum CF gap by the sender.
This value is used instead of the value provided by the receiver inside the FC.
CAN_ISOTP_RX_STMIN:
Ignores received CF frames which timestamps differ less than this value in nano
secs. This is used to test the receivers misbehaviour when the receiver
provides a lower stmin value that he's able to cope with. Received CF frames
are silently dropped when they come faster than specified by this value.
On 64bit systems a 'long' is a 64bit value but the target values of scanf()
are always 32bit in the affected code - which means just an 'unsigned int'.
Unsigned int is fine on 32 and 64 bit systems.
Thanks to Andre Naujoks for reporting this issue.
*worst* case calculation for stuff-bits.
(see "Guaranteeing Message Latencies on Controller Area Network" 1st ICC'94)
Due to the bitstuffing estimation the calculated busload may exceed 100%.
Major changes:
- rework internal structures to prepare routings and modifications of CAN traffic also to non-CAN interfaces
- fix reading of netlink messages in cangw.c (added RTCAN_RTA / RTCAN_PAYLOAD macros)
- rework reading of gw-job lists in cgw_dump_jobs()
- rename of functions and API definitions to have a common namespace cgw_
- added infrastructure to perform crc8 and xor checksums in CAN frame data[]
TODO:
- add and test functionality for crc8 and xor checksums in CAN frame data[]
- add help text for crc8 and xor checksums in CAN frame data[]
repository. Building with exported headers from an unpatched kernel will
fail due to missing symbols or headers.
This patch adds two make variables to optionally disable building such
binaries, like this:
make PROGRAMS_ISOTP= PROGRAMS_CANGW= PROGRAMS_SLCAN= all
Signed-off-by: Olaf Hering <olaf@aepfle.de>
Signed-off-by: Oliver Hartkopp <oliver@hartkopp.net>
Oliver Hartkopp
