Minutes of SJTAG Fringe Meeting at ITC'07, 2007-10-25
Call to order at 10:35am
Bradford Van Treuren
- Brad provided some history of SJTAG as well as an overview of and its role in MicroTCA and ATCA;
- Brad showed Use Case presentation; including goals; will be available on SJTAG website;
- Brad requested feedback on matrix "use cases vs. goals"
- A. Sparks mentioned that dot6 would not be supported for backplane testing if one board would be the driver and other boards would be the receiver and both boards would NOT go through the Update-Capture sequence concurrently (i.e. the driver goes through Update and then stays in RTI, while the receiver goes through the Capture state); Brad noted that this is a true restrictions that designers would have to be considering, but that the desire to allow driver and receiver not both needing to be in Update-Capture sequence in lockstep should not make it into SJTAG in order to avoid such restrictions as dot6 not being supported;
- Bernard Sutton noted that TAP security is always coming up;
- some "firewall" circuitry on the system TAP interface or in individual devices could provide a solution;
- should security issues be handled by a working group of its own (another "dot") or is it already covered sufficiently by IJTAG?
- Erik Larsson: how would you implement test data for POST or for other embedded test? How to avoid configuration conflicts and manage the test data?
- Andrew Firth: a board containing test data for POST, etc., should also record test results (fault data, signature, diagnostic results), so that they are available on the board when it is shipped back for repair;
- furthermore: the card could record whether it really failed and that the board has not been declared as having failed by error (intentionally or not); maybe a history of the board tests, etc.
- prompted by Ken Filiter, Brad mentioned that the system JTAG implementation for ATCA would probably drafted by an ATCA sub committee (apparently lead by Brad) and the SJTAG group will also likely continue working past that event to standardize a more general system level JTAG infrastructure;
Adjourn at 12:25pm
Many thanks to Heiko for taking excellent notes during the meeting!