Minutes of Weekly Meeting, 2015-03-23

Meeting called to order: 11:07 AM EDT

1. Roll Call

Heiko Ehrenberg
Ian McIntosh
Adam Ley
Brad Van Treuren
Carl Walker
Tim Pender (joined at 11:15 AM)
Peter Horwood (joined at 11:38 AM)
Eric Cormack (joined at 11:48 AM)

Michele Portolan
Brian Erickson

2. Review and approve previous minutes:

3/16/2015 minutes (draft circulated 3/16/2015):

  • No corrections noted;
  • moved to approve by Eric, second by Tim;
    • No objections or abstentions -> minutes approved.

3. Review old action items

  • All: do we feel SJTAG is requiring a new test language to obtain the information needed for diagnostics or is STAPL/SVF sufficient? See also Gunnar's presentation, in particular the new information he'd be looking for in a test language (http://files.sjtag.org/Ericsson-Nov2006/STAPL-Ideas.pdf)
  • Ian: Add the previously discussed lists to the 'master' template. Ongoing.
    • Some sections need further expansion that may take time to develop.
  • Add "Power manager" to "Illustrative SJTAG infrastructure" diagram and circulate it to the group. - DONE

4. Reminders

  • Consider Adam's three points (from the action from the first weekly meeting) and suggest what is preventing us from answering those questions:
    • Establish consensus on goals and constraints
    • What are we trying to achieve?
    • What restrictions are we faced with?
  • Forum thread for discussion: http://forums.sjtag.org/viewtopic.php?f=3&t=172

5. Discussion Topics

a. Revisit draft PAR statements - continuation, inferences from ITEA Journal item.

  • Infrastructure Diagram:
  • {Ian shared "Illustrative SJTAG infrastructure" diagram, now with Power Manager and PM Chain added}
  • Ian mentions that it is becoming a hassle to add stuff to the drawing as Power Point is not really the right tool for such diagrams, but there may be other ways to connect the PM Chain as Peter had noted last week.
  • Brad and Adam concur that we have this now added to the diagram as a circuit element and if necessary we can refine it later as needed.
  • {Tim joined}
  • Ian asked if we need to add anything else; Brad suggests that the only thing he can think of right now is a I2C sensor that's wired to the micro processor, as the handling and usage would be different when compared to the I2C sensor connected to the FPGA.  May be hard to fit in, but perhaps it can be added in the Power Manager "neighbourhood".  Ian agrees that that should be possible and accepts adding another I2C sensor as an ACTION item.
  • ITEA Article:
  • {Ian shared the "Embedded Instrumentation Systems Architecture" (EISA) paper from ITA Journal 2009, issue 30}
  • There was some email discussion on the reflector about the paper, and the general consensus seems to be that it is interesting for us.
  • Ian had a concern that the vision in the paper was moving too far away from what might be seen as the traditional market space for JTAG tool vendors.
  • Brad: Push (EISA) vs. pull (JTAG) system status reports. EISA assumes intelligent sensors. JTAG has no such mechanism.
  • Ian: Can use the Push model because they have relatively smart sub-system controllers throughout. A lot of the sensors are probably fairly "dumb" and status reports could be pushed at a higher level by a sub-system configuration that connects to those sensors. You could get something similar in ATCA if the Board Manager presented virtual instruments to the Shelf Controller.
  • Brad: Implementation of tests for system blocks with multi-drop addressing is very different compared to direct access to system blocks (e.g via IPMI).  This also impacts diagnostics (how to identify and communicate the exact fault location, unambiguously).  The different schemes can't be modelled identically.
  • EISA has everything cut and dried in the hierarchy.  In many cases we're trying to deal with a system made up of new designed assemblies alongside legacy designs and COTS items that we may have limited data on.  It's not simple.  Ian wonders if we should we consider only the parts we have total control over for the core standard and leave things like addressing COTS boards for extensions?  Not ideal, but may be a simplification.
  • Brad: the point of me showing this EISA was to make us aware that there are other ways of supporting system test / configuration etc., than what we may be used to or what we may see in the tools that we use in our industry today.  Things like using a Configuration Manager and the use of meta languages which can give vendor independence.
  • Ian: How do we identify key features in this work/paper that we could learn from for the purpose of SJTAG?
  • {Peter joined}
  • Brad drew a comparison with C/Atlas, raised in an early discussion with Adam. Agreed that C/Atlas had proved costly to implement.  Ian thought an lesson may be to avoid something that required expensive tooling support and significant engineering effort to implement.
  • Brad: Different user expectations are being pushed back to the tool vendors in respect of 1149.1-2013 and 1687.
  • Ian: Those have given the user additional control, but the user now needs to think more about how the UUT operates to make use of that.  It's not as "push button" as the traditional JTAG tools have been able to offer.
  • To avoid distraction from the main topics of the meetings, it was agreed that we all read the paper and continue discussions on the SJTAG email reflector.
  • {Eric joined}

6. Key Takeaway for today's meeting

  • None.

7. Glossary terms from this meeting

  • None.

8. Schedule next meeting

  • Next meeting March 30. Heiko will be absent. Michele likely to be absent.
  • April schedule: 6, 13, 20, 27.  April 6 is Easter Monday - Eric and Peter will be absent.

9. Any other business

  • Next newsletter is due! - Update on Green Paper:
    • Brad sent a draft of the Green Paper.
    • {Ian shared the draft Green Paper titled "A Tale of Two Domains: Preprocessed Tests vs. Interactive Configuration"}
    • Brad asked the group to look for readability, typos, etc., and other suggestions. In particular, check for changes in tense or anything confusing.
    • Initial comments: Enlarge text in Figure 3; Figure 6 may need a label for the Debugger window.
    • Feedback comments via the email reflector or at least to Brad {ACTION}.
  • ITC: Nothing to add.

10. Review new action items

  • Ian: Add an I2C sensor connected to the uP on the "Illustrative SJTAG infrastructure" diagram.
  • All: Review the draft Green Paper titled "A Tale of Two Domains: Preprocessed Tests vs. Interactive Configuration" and submit corrections / suggestions via the email reflector or directly to Brad.

11. Adjourn

Moved to adjourn by Peter, second by Eric. Meeting adjourned at 12:03 PM EDT

Thanks to Heiko for providing the majority of these notes.

Respectfully submitted,
Ian McIntosh