Minutes, IBIS Quality Task Group

20 October 2020

11:00-12:00 EST (08:00-09:00 PST)

ROLL CALL

Achronix Semiconductor:             Hansel Dsilva
Intel Corp.                         Michael Mirmak
Mentor, a Siemens Business          Arpad Muranyi
Micron Technology                   Randy Wolff
Signal Integrity Software:        * Mike LaBonte
SPISim                              Wei-hsing Huang
Teraspeed Labs:                   * Bob Ross
Zuken USA:                        * Lance Wang
SAE ITC                             José Godoy

Everyone in attendance marked by *

NOTE: "AR" = Action Required.

-----------------------MINUTES ---------------------------
Mike LaBonte conducted the meeting.

Call for IBIS related patent disclosures:

- None

Call for opens:

- None

Review of previous meeting minutes:
Minutes from the October 6 meeting were reviewed. Bob Ross moved to approve.
Lance Wang seconded.  Without objection the minutes were approved.


ARs:
- Mike LaBonte to update and email IBIS-ISS parsing bug report
  - Done but not sent
  - Bob Ross also had sent a second draft.


New parser bugs:
Bob Ross reported there were no new bug reports.


IBIS-ISS parser:
Bob Ross showed his draft 2 of BUG214. He renamed it draft 3 for new edits.  He said
the parts written by Mike LaBonte had not been altered.  The new text added checking
of references to IBIS-ISS from .ims files and from multi-lingual.

Lance Wang noted that parameters could be specified in multiple places.  He felt that
could be a lot of work to handle correctly.  Mike LaBonte said checking parameters
fully would require a full parser, that would read everything into memory.  Bob said
multi-lingual models starting with IBIS 4.0 would be checked.  Mike said IBIS-ISS
came later than 4.0. It was introduced by BIRD116.2, incorporated into IBIS 6.0.
Bob changed 4.0 to 6.0 in his draft.

We discussed aspects of preference order for various means of setting parameters.
Bob felt that number/string type checking should be performed.  Mike felt that intrinsic
element parameters could be checked rigidly, throwing errors, because their usage
is documented.  Parameters created in subckt definitions could only be checked for
consistency because it was not known which type was actually intended. Also we could
warn when a parameter appeared in an X element call but was not used in the definition.
Bob added checks related to parameter passing to the document.

Mike noted that we had previously decided not to specify in-depth checking.  Bob said
that was a decision yet to be made.  Mike agreed that the cost might not be too much,
and we were not in a hurry to complete the ibischk update.

Bob said there was an intrinsic parameter that specified the name of a file to read.
Mike said that would increase the complexity of the parser.  Lance felt that extensive
checking was not necessary, and that parameters in particular could get complicated.

Bob showed examples of expressions that could be hard to check.  Mike suggested that
expressions needed to be checked, but a full syntax check would not be necessary.
The presence of arithmetic operators in an expression would indicate that it was
of numeric type, and that might be sufficient for checking.  Bob said we should have
interface checks, and everything about the interface between the call and the definition
should be checked.  Mike felt that a useful level of checking would be to check for
syntax that is legal for HSPICE or some other simulator, but not legal for IBIS-ISS.
Bob said many of these checks would apply to EMD and EMS files.

Mike said we should resolve the overlap between his text and the new text from Bob.
Bob said Mike's text was very detailed, more appropriate for a code change specification.
Mike suggested keeping his text in draft 3, to be considered for deletion in the
next meeting.

Bob suggested include statements might be handled by producing a flattened file.
Mike said that should be up to the developer.  Bob said it also didn't matter much if
the files were parsed into memory or read repeatedly. Mike agreed, with the exception
that certain checks could only be performed after all IBIS-ISS contents had been read,
which could require retaining some information.

Bob said we would need to decide if include statements should be allowed to reference
files residing above the IBIS file tree.  Mike suggested that would be a good Open
Forum discussion topic.

AR: Bob Ross to update BUG214 draft to describe parameters that name files


Tabled topics (no discussion without motion):
  - BIRD181.2


Bob Ross moved to adjourn. Lance Wang seconded. Without objection the meeting ended.

Meeting ended: 12:17 ET

Next meeting October 27, 2020