On 5/8/19 4:39 AM, Joakim Zhang wrote: >> > Another question is that CAN driver doesn’t ensure the total number of >> > time quanta in a bit time programabled at least from 8 to 25 via CAN >> > bittiming calculation(CAN_CALC_BITTIMING). > >> Where does this 8 to 25 limitation come from? > > [Joakim] Both ISO11898-1:2003(E) and ISO11898-1:2015(E) all mention it: > ISO11898-1:2003(E): > ISO11898-1:2015(E): I still read the spec as I've written in the previous mail: >> > 1)ISO 11898-1:2003(E) show that : >> > >> > 12.4.1.2 Programming of bit time >> > >> > *The total number of time quanta in a bit time shall be >> > programmable at least from 8 to 25*. >> > >> > 2)CAN2.0 spec show that: >> > >> > 8 BIT TIMING REQUIREMENTS >> > >> > *The total number of TIME QUANTA in a bit time has to >> > be programmable at least from 8 to 25.* I read the above quoted paragraph as: If you implement a spec conforming CAN core, the time quanta has to programmable in a certain range (according to CAN2.0, while SO 11898-1:2003(E) says it "shall be"). That range must be at least 8...25. To my understanding you might implement a wider range. > And the total number in a nominal bit time and a data bit time also > have a least limitation when can fd mode enable. I understand the spec that it defines a minimum of configuration range. You might implement a wider range. Is here a native speaker listening, who can read the spec and tell what's the correct interpretation? Marc -- Pengutronix e.K. | Marc Kleine-Budde | Industrial Linux Solutions | Phone: +49-231-2826-924 | Vertretung West/Dortmund | Fax: +49-5121-206917-5555 | Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de |
Attachment:
signature.asc
Description: OpenPGP digital signature
