Meeting: P1450.1 (STIL) Working Group
Date: Monday 10/2/00 @ ITC-Atlantic City
Time: 8:30am to 12:30pm

General:
This was a short meeting due to other ITC activities that many wanted to
participate in (P1500, CTL, IBM-LBIST to name a few). Even so, progress was
made on some key issues. The working group is still planning to complete the
1450.1 document by the end of 2000. The 1450.2 is in good shape for
completion by year's end, but was not discussed at this meeting due to lack
of time and lack of key participants. Follow on work will be done with email
and conference calls.

Attendees:
Tony Taylor - Synopsys
Peter Wohl - Synopsys
Doug Sprague - IBM.
Klaus-Dieter Hilliges - Agilent.
Ernie Wahl - Lucent. keeping tabs on STIL; interest in Auto Program Gen
Bernard Heilbler - SZ testsysteme
Gordon Robinson - Credence
Hira Ranga - 3MTS
Rohit Kapur - Synopsys (part time)

1. XML (Gordon):
The use of XML has been suggested by Gordon several times and was discussed
at this meeting. It was felt that: a) we don't want to make any changes that
are disruptive to the STIL standard at this time; and b) the STIL, STIL
extensions, and CTL standards all need to be based on compatible syntax.
That said, there still may be a way to utilize XML if we can identify a way
to move to/from an XML based format.

The main benefits identified for an XML version are: a) it is a standard
that is getting major acceptance for inter operability; b) there are many
tools available for XML (browsers, navigators, editors).
The decision was made to research XML in parallel, and without affecting the
work currently in progress. Gordon agreed to do some initial investigation.
[AI-Gordon]

Gordon: "There is no intent at all to alter course on anything that is
defined or under development. Long term, some of us see great promise
in using XML tools and technologies on some STIL and extension
datasets. Therefore I agreed to produce a white paper describing how
we can get to such a point, while preserving everything that is
currently defined and in development. While I'm sure of the importance
of XML long-term, I put preservation of current progress and momentum
higher on my priority list. To repeat: There is no intent at all to
alter course on anything that is defined or under development."

Hira made the suggestion that an XML version of STDF (STandard Datalog
Format) should be created.

2. Document Structure:
The numbering of the syntax lines needs to be re-sequenced [AI-tony]
Section 11.3 "ScanStructure statement" needs to be moved to the annex
[AI-peter]

Still need input from Greg Maston on the 1450.0 supplement and his sections
of 1450.1 [AI-greg]

3. Pattern Variables, Enums, TestResult (Tony)
Tony has provided a more complete definition of pattern variable syntax,
including examples. In addition, a new function named TestResult has been
added to address the need initially proposed by Kim Smith to have an implied
Pass/Fail variable on every V statement.

The discussion of this syntax prompted the question of what is the purpose
of putting such capabilities in the STIL language. The discussion produced
the following conclusions:

A) Is STIL.1 to be a "test bench" tool? NO! Although there may be some
limited applications where STIL may be used as a design "test bench", this
is not the intended application.

B) Is STIL.1 to be used as a simulation stimulus? YES! Simulation of STIL is
needed for test verification and fault coverage computation. The language
needs to have constructs to make this efficient (like parallel application
of scan vector data).

C) Is STIL.1 to specify ATE patterns with micro code? NO! This is not the
purpose of the language at this time. The 1450.3 work is to address this
need. A limited sub-set of what STIL.1 is defining may be used.

With the above conclusions in mind, it was decided to remove the TestResult
function and put it in the "recommended usage" document. We also decided
that section 7.0 needs to state that user defined functions are allowed in
pattern expressions.

4. WFCMap, \j, \m (Peter)
Peter has worked with Greg Maston to make some simplifying definitions as to
how the map and join functions work. Also, the WCFMap definition was moved
to the PatternBurst block. These changes were discussed and approved by the
group.

5. X-statement (Doug)
Several minor changes were made (Loop becomes Iteration) as well as some
corrections to the examples. These were discussed and approved by the group.

6. Wait, Extend (Tony)
Minor changes to the Wait and Extend have been made in the document. The
discussion of these changes opened up a fuller discussion as to the purpose
of these statements. It was questioned wether a signal-wait form of syntax
would work better. The meeting ran out of time with no conclusion on this
issue. This issues is to be continued via discussion and email.

7. BIST Structures
A discussion the previous day in the CTL meeting had identified a proposed
solution the BISTStructure syntax. The meeting disbanded before this could
be reviewed. It is to be continued via phone and email discussion.

8. 1450.2-Levels
This document has received considerable attention by Gregg Wilder and his
sub-task-force. It appears to be in good shape, but was not reviewed in this
meeting due to lack of time and lack of needed participants.

9. Programmable buffers (Ernie)
Ernie wanted to bring up a new issue. It was not even presented due to lack
of time. [AI-Ernie to explain the issue]

Action Items:
[AI-1] - Gordon, Tony, and Rohit to meet to discuss XML and potential for
incorporation into the STIL and CTL standards. Gordon to do initial
investigation and proof of concept.
[AI-2] - Tony - re sequence syntax numbering
[AI-3] - Peter - move section 11.3 to the annex
[AI-4] - Greg - complete 1450.0 supplement and 1450.1 assigned parts
[AI-5] - Ernie - explain your "programmable buffer" issue