Yegor Yefremov wrote: > Are there any plans to mainline slcan? Or what issues are still open > (kernel/2.6/README.known-issues doesn't have any)? >
Hi Yegor, so far i have three items on my personal mainlining list: - netlink gateway net/can/gw.c (needs an API rework to handle non-CAN networks, working on that topic here) - ISO15765-2 transport protocol net/can/isotp.c (needs some more feedback from users - currently got only three peoples using it) - SLCAN driver drivers/net/can/slcan.c I don't have any general objections on mainlining the slcan.c - mostly i do have the gut feeling that we should take a closer look into the latest changes of the tty-layer and the changes that have been made to slip.c (where slcan.c is derived from mostly). If you are interested in reviewing the slcan driver you're invited to do so. I also wonder if the slcan driver should support the IFF_ECHO functionality as i'm quite sure that checking the "\n" on the receiver side is not really a proper indicator for a successfully transmission of a single CAN frame. "\n"'s are also in the RX ASCII stream ... the slcan driver is a just a driver for low cost interfaces using a simple unreliable ASCII communication protocol. IMO the slcan driver can never act better than the underlying protocol. So finally only my gut feeling seems to slow down the mainlining schedule ;-) Best regards, Oliver _______________________________________________ Socketcan-core mailing list [email protected] https://lists.berlios.de/mailman/listinfo/socketcan-core
