IEEE 1532 Meeting - December 4-5, 2001

• Attendees
• Minutes
• Attachments

Back to Minutes page
Back to IEEE 1532 home page

Attendees

The following individuals attended the meeting:

    Neil Jacobson – Xilinx
    Ken Parker - Agilent
    Dennis Lia - GenRad
    Dave Bonnett – ASSET
    Brad Ishihara - Altera
    Mark Moyer – Lattice
    Jack Griffin – Cypress
    Howard Tang – Lattice
    Jason Baumbach – Cypress
    Sanjeev Sharma – Altera
    Alan Hermann – Altera
    Ray Dellecker –JTAG
    Dan Gravitz - ASSET

Minutes

    Meeting: IEEE P1532
    Date: Dec/4 & 5 / 2000
    Location: San Jose, Ca
    Purpose: Latest Draft Clean up

Agenda:

• Welcome and Introductions
• Standards Document Update
• Discussion
  • Review of latest draft
  • Review of latest draft

Date: 12-4-00 Monday
Time: 9:00 AM

Welcome and Introductions:
Standards Document Update: Neil

• IEEE copy of the draft received by Ken and Neil and being edited.
• PDF version will be available from IEEE. About $50.
• Neil is getting numerous inquires requesting information on the standard.
• Request to write a book on 1532. IEEE and Kluwer Academic Press pursuing Neil.
• Xilinxs building a software tool – Will do individual devices(not multiple devices, or concurrence), written in C, Parser in Flex and Bison, Xilinx Parallel I/O port, Source and executable code, due this month, being tested now, Should this be industry wide open sourced( Would there be an academic resource interested). There was considerable discussion on the use of the Xilinx.

Actions: Neil

• Make available on IEEE website.
• Make code modular
• Pursue open source administration by academic resource.

Discussion: Latest Draft - Page 100: 8.5.5.1 (r)

• USERCODE as 32 1 bit quantities.
• Is there sufficient guidance in the notes and recommendations?
• Is the 32 1 bit quantities needed?

Motion: Remove the recommendation (r) and it’s Note
Made: Mark
Second: Ray
Vote: 8:0:0
Actions:

Remove the recommedation – Ken
Add further description on use of USECODE – Add in 8.5.5.2 somewhere – Ken
What about Security Code - Ken

Time: 10:45 AM
Break

Time: 11AM
Discussion: Latest Draft - Page 110: Example 11

• Not an issue.

Action: Add Array_B and Array_C to example - Ken

Discussion: Latest Draft - Page 110: Example 12

• Need to design the example to illustrate a “round trip” read-out.
• Working Group drew diagrams to understand the “round trip” mechanism.
• Linear Array, Unprocessed Flag, Vendor Post Process data reformat tool.
• Need ability to read back a device to obtain data for the input flow.
• Output Flow -> Data stream -> Reformat tool -> Data file -> Input Flow.
• Need a Tag in BSDL to indicate to tool vendors that the flow produces an unprocessed output.
• ?! used together on the capture and update side of flow instruction? (? is read, ! is output)

Action: – Neil / Ken

• allow ?! on the update side.
• Add unprocess Tag to output data file

Time: 12:10 PM
Lunch

Discussion: New Issues - Codification of PROPRIETARY keyword

• Does it mean apply absolutely no optimization?
• Action: Redo this section. Define PROPRIETARY to mean Non-optilmizable and Non-concurrentizable. Add verbage. Add optimize to the glossary.
• 1532 Optimization allowed: IRshift
  • IR
    Always Load Instructiion
  • DR
    Update/Capture Overlay – assumes IR optimization

Discussion: New Issues - Add VCC attribute

• Request was to specify the programming voltage.
• Working Group considered it a non-ISP request.
• Considered to be a need of socket programmers.

Action: Invite socket programmer tool vendors to attend next meeting and express their requirements.

Discussion: New Issues - Incomplete data in ISC file should be an error

• Data file and BSDL file must match for data length with in a shift. The data file may have less bits for a read operation than the width specified in the flow in which case the tool will pad with zeros. Data can not run out in the middle of a repeat loop.

Discussion: New Issues - Binary, Hex, Decimal Data formats

• Are we satisfied with the current situation?
• Yes.
• Action: Leave as is. Scan data is Hex. Ever thing else is integer.

Time: 3PM
Break

Time: 3:15 PM
Discussion: New Issues - ISC_PIN_BEHAVIOR

• Leave as documented

Discussion: New Issues - ISC_STATUS

• Leave as documented.

Discussion: New Issues - Abstract Data Tag Keyword

• Array
• Security
• Idcode
• Usercode
• Are these enough? Answer is Yes. Too many? Answer is No.

Discussion: New Issues - PROC_ERROR_EXIT

• Mandatory or Optional? Manditory
• Single point of exit. Called for data mismatch errors and for tool aborts situations.
• User tool will provide “Continue/Stop on Error” options (complete action or not). If Continue on Error then execution stops before proc_program_done is not called. Abort would call proc_error_exit immediately.
• PROC_ERROR_EXIT is called before proc_program_done.
• Program security after proc_program_done.

Time: 5:00PM – Adjourn Meeting

Time: 8:30 AM
Date: Tuesday 12/5/00

Discussion: New Issues – “!” to a variable

• There is currently no way to save capture data in a variable – should this change?
• No.

Discussion: New Issues – Security Operation

• Pre-1532 Security is set by designer and encoded in JEDEC file.
• 1532: Security is set at configuration time and decided by operator of the configuration tool.
• Add new record(s) to ISC data file
• define Off/On/optional/recommended
• Objective to specify or override the vendor specified BSSL Action procedure

Motion: Add “define off / on / optional / recommended” , after the header in the data file.

Made: Mark
Second: Alan
Vote: 8:0:0

Next Steps:

• ISC data file generation
• Integration with configuration tools from multiple vendors.
• Concurrentization.
• Demo board with multiple vendor devices.
       1532, 1149.1, 1532 proprietary

Tasks:

1. BSDL Draft update – Ken 2 weeks
2. BSDL files to Neil/Reflector – IC Vendors 2 weeks
3. Check BSDL files and redistribute– Neil 2 weeks
4. Configuration Tests, data files – IC Vendors 4 weeks(Single device) --- Late Feb
5. Develop tools – tools vendors -
6. Demo board development – Altera – 8 weeks
7. Concurrent Algorithm development – 10 weeks
8. Verification of Concurrentization – 4 weeks --- Mid April

Note: Ken/Neil captured an MS Project file for the above tasks.

Discussion: New Issues – Choose default padding value?

• No. Leave default as zero.

Discussion: New Issues – New Rules?

• Reviewed ISC Extension Structure section 8.4
• Reviewed ISC BSDL Attribute Definitions section 8.5
• Need to rework 8.5.4 – Attribute ISC_Security - ken

Time: 12:00 PM
Lunch

Time: 12:45 PM

• ISC_Security 8.5.4 (continued): Change ISC_PROTECTION to ISC_SECURITY. Used, if Security is implemented and addressing the ISC_PDATA register. Make it optional. Must exist when ISC_PROGRAM_SECURITY exists. Must use “security” tag for associated data.
• Reviewed Section 8.5.5 Flow Attribute for rules.

  Add rule that Time units are in seconds
  Delete rule (p)
  Add shift rules? Yes, describe over flow conditions on shift operations.

• Reviewed Attribute IAC_Procedure section 8.5.6

  Need rule that proc_error_exit and proc_program_done are to be manditory
  Brad respectfully declined to assume editorial responsibilities.
  

• Reviewed Attribute ISC_Action section 8.5.7
• Reviewed Attribute ISC_Illegal_Exit section 8.5.8
• Reviewed Attribute ISC_Design_Warning section 8.5.9
• Reviewed Data File Format section 9.

Thanks to Lattice for hosting the meeting.

Note: Neil has captured agenda items in a Power Point file describing the issues and resolutions in bullet form.

Time: 3:15 PM
Adjourn

Next Meetings:
Date: February 19/20/2001Mon-Tue,
Location: Altera, San Jose
Purpose: Single Device Configuration Review

Secretary: Dennis Lia

Back to Minutes page
Back to IEEE P1532 home page