wiki/ news/ 2003-11-18

2003-11-18

Attendees: Andrew, Ian, Tyler, Peter & a couple of bright high-schoolers

CanBusLinuxDriver

Ian inserted a Phase 0 to the driver rewrite plan: a command line tool to exercise the i82527 chip interactively. This is made possible by the Linux hackHHHH system call iopl, which allows a user-level program (running as root) to gain kernel-level IO port access rights. Ian used this tool, sucan (pronounced "yes you can" ;-), on the FlightComputer to send the first ever RTR message from a MOPS board!

The uncanny driver will be updated with the newly gained knowledge, and should be able to query status from the LTC by the next meeting.

The bad news: Andrew and Ian pored over i82527 spec sheet, but could find no way for the chip to tell the CPU whether an incoming message is an RTR message or not. The flag we thought would show this is only active during a message object's automatic response to a particular ID. We may have to redo the CAN ID scheme and all the node protocols to eliminate use of the RTR bit!

Cantalope

Tyler and Peter outdid themselves, demonstrating reading messages from and writing messages to the CantalopeHardware... in hex, decimal, or binary! Way to go, guys! By the end of the meeting, we were using Cantalope to turn the GPS node on and off, verify Cantalope itself, and show the operation of the sucan test program. Andrew has the latest source, soon to be commited to CVS.

Can King is dead! Long live Cantalope!

Open Issues

-- IanOsgood - 19 Nov 2003