Drivers for Hardware Architecture - Brainstorm

Hardware Architectures Document
Please debate the content of Volume 3 here!

Drivers for Hardware Architecture - Brainstorm

Postby Ian McIntosh » Mon Nov 23, 2009 7:05 pm

The bullet points we raised at the November 23 meeting:
  • Need to deal with multiple assemblies in a system
  • When externally controlled, generally a single TAP is desired
    • Multiple TAPs, e.g. for concurrency, like Intellitech cJTAG, shouldn't be discounted
  • Means to isolate assembly under test from other signals on the backplane
  • Co-existence of embedded control with option for external control
  • Signals going off-board must not be floating (DFT rules)
  • Fan-out may be large, especially across large backplanes
  • Individual control of tests in a FRU
  • Signal integrity of 5-wire TAP when taken external to a system
  • Uniformity of interface - don't mix multi-drop and star
  • Means to address assemblies:
    • "Slot code"
    • Hardwired by board type
    • Mixture of the two
Please reply with additions to the above, or with comments and observations.
Ian McIntosh
Testability Lead
SELEX Galileo
User avatar
Ian McIntosh
SJTAG Chair
 
Posts: 167
Joined: Mon Nov 05, 2007 11:49 pm
Location: SELEX GALILEO, UK

Re: Drivers for Hardware Architecture - Brainstorm

Postby Ian McIntosh » Wed Dec 02, 2009 7:21 pm

Additional point raised at the November 30th meeting:
  • Security
    • Protect against unauthorised retrieval of firmware, software or other device register contents
    • Protect against substitution of unauthorised firmware or software or other device register contents
    • Permit access for board test
    • Restrict retrieval of circuit design information

There is a thread dedicated to this topic: viewtopic.php?f=29&t=111
Ian McIntosh
Testability Lead
SELEX Galileo
User avatar
Ian McIntosh
SJTAG Chair
 
Posts: 167
Joined: Mon Nov 05, 2007 11:49 pm
Location: SELEX GALILEO, UK

Re: Drivers for Hardware Architecture - Brainstorm

Postby Bradford Van Treuren » Fri Dec 04, 2009 6:32 pm

To try to stimulate more discussion on the subject of management requirements SJTAG needs to support, I offer the following diagram and description to help gain more insight into the problem space to assist our group.
SM_AM.jpg
management hierarchy model

System Manager:
  • Tracks and records the configuration and state of the system
  • Regulates the operational state of each assembly in the system
  • Communicates configuration and state information to the outside world through an external interface
  • Applies operations/functions to the system as requested by the external inteface
Assembly Manger:
  • Tracks and records the configuration and state of any sub-assemblies that make up this assembly
  • Regulates the operational state of each sub-assembly that makes up this assembly
  • Applies operations/functions to the assembly as requested by the System/Parent Assembly Manager

For POST or self initiated testing, the assembly manager would post its state information in a publically accessible memory resources (e.g., dual ported RAM, FLASH file system) where the system manager could extract the inforation in a pull control model or the assembly manager could asynchronously send a message to the system manager with any configuration or state change in a push control model.

Ideally, the information is maintained by the system manager on a softare bulletin board/data base resource.
Bradford Van Treuren
Distinguished Member of Technical Staff
Alcatel-Lucent
User avatar
Bradford Van Treuren
SJTAG Chair Emeritus
 
Posts: 36
Joined: Fri Nov 16, 2007 2:06 pm
Location: Alcatel-Lucent / USA

Re: Drivers for Hardware Architecture - Brainstorm

Postby Ian McIntosh » Mon Dec 07, 2009 8:37 pm

The slides I used during today's meeting, work from the opposite end that Brad's do. I was trying work things from a viewpoint of how a system typically appears and what the problems are that are presented for JTAG, since that's fundamentally what SJTAG is trying to solve.

http://files.sjtag.org/IanMc/Back2Basics.pdf

What I'd hoped to do was build up layers or overlays on the basic physical architecture that showed how the management and routing entities we'd identified could be placed, but I wasn't finding it easy :?
Ian McIntosh
Testability Lead
SELEX Galileo
User avatar
Ian McIntosh
SJTAG Chair
 
Posts: 167
Joined: Mon Nov 05, 2007 11:49 pm
Location: SELEX GALILEO, UK


Return to Volume 3

Who is online

Users browsing this forum: No registered users and 1 guest

cron