Skip to contentSkip to conversation

CAN documentation

DISCUSSION Nikita Edward Baruzdin 2 participants 2 messages Jul 25, 2014 CAN network drivers

lore
Share & tools
lore
#0 lore raw

Hello, guys,

I'm looking at linux-can/can-modules tree:

2011-06-26 12:47 Oliver Hartkopp
minor doc changes ... tbc

2011-06-10 19:08 Oliver Hartkopp
documentation: rework chapter numbering & references

...

2010-11-22 19:41 Oliver Hartkopp
Added new documentation layout contributed by Daniele Venzano.

Why are all these changes still not in the kernel? I'm not sure if this is a good idea to split the documentation, but the patches introduce some new helpful information.

Hence a little follow-up question:

Quoted 2 lines

The timestamp on Linux has a resolution of one microsecond and it is set automatically at the reception of a CAN frame.

It seems to me this is not entirely true as the kernel has a support for SO_TIMESTAMPNS socket option for quite a long time. Could you really use that option with CAN? Does it make sense to use nanosecond timestamps? I'm also a bit confused whether these are only the software timestamps or you can somehow get them from a controller too.

#1 lore raw

Hi Nikita,

indeed the BCM patches have been added to the tree as well as CAN FD and filter documentation updates, see commit log at:

http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/log/Documentation/networking/can.txt

Don't know if it's really useful to split up the file.

Would you like to provide a patch for the timestamping stuff?

Regards,
Oliver

Quoted 30 lines · Nikita Edward Baruzdin

Hello, guys,

I'm looking at linux-can/can-modules tree:

2011-06-26 12:47 Oliver Hartkopp
minor doc changes ... tbc

2011-06-10 19:08 Oliver Hartkopp
documentation: rework chapter numbering & references

...

2010-11-22 19:41 Oliver Hartkopp
Added new documentation layout contributed by Daniele Venzano.

Why are all these changes still not in the kernel? I'm not sure if this is a good idea to split the documentation, but the patches introduce some new helpful information.

Hence a little follow-up question:

Quoted 2 lines

The timestamp on Linux has a resolution of one microsecond and it is set automatically at the reception of a CAN frame.

It seems to me this is not entirely true as the kernel has a support for SO_TIMESTAMPNS socket option for quite a long time. Could you really use that option with CAN? Does it make sense to use nanosecond timestamps? I'm also a bit confused whether these are only the software timestamps or you can somehow get them from a controller too.