Minutes, IBIS Quality Task Group

07 August 2018

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

ROLL CALL

eASIC:                              David Banas
Ericsson:                           Anders Ekholm
Intel:                              Michael Mirmak
                                    Eugene Lim
IO Methodology:                   * Lance Wang
Mentor, a Siemens Business          Arpad Muranyi
Signal Integrity Software:        * Mike LaBonte
Teraspeed Labs:                   * Bob Ross
Visteon:                            Preetesh Rathod

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:

- Bob: I sent new bugs.

Review of previous meeting minutes:

- Mike: Minutes of the previous meeting have date issues, and may
  not been sent yet. Need to investigate and get caught up.

AR: Mike LaBonte to check previous meeting minutes and send.

ARs:

- AR: Bob Ross to ask Jose Godoy for his title.
  - No reply.
- AR: Bob Ross to send IBISCHK6.1.5 contract draft to Atul for quote.
  - Atul has said he is available. The contract will be sent after
    new bugs have been added and the proposed cost quoted.
- AR: Mike LaBonte to send IBISCHK700_PARSER_CONTRACT_draft11.docx 
  - Done.

New IBISCHK bugs:
- Bob Ross showed bug198.txt
- Bob: This originated from a testcase file sent by Arpad Muranyi.
  He got 14 errors, mostly reported as other bugs. This bug is one that
  was not previously reported. The 5.1 parser is correct, it gives 4 errors.
  The 6.0 parser reported zero errors. The test file is 4.2.
- Mike: Why is the severity MODERATE?
- Lance: FATAL should be used for a crash. SEVERE should be used for a false error.
- Mike: A false negative should be MODERATE.
- Bob: If it misses a check, it could be MODERATE or SEVERE.
- Mike: In this case the data read into memory by the parser might also be
  incorrect, which would impact source code users.

- Bob Ross showed bug199.txt
- Bob: This is co-authored by Arpad Muranyi. It needs to have the actual
  false error message text added to the bug. The problem appeared in 6.1.3.
- Mike: This should be SEVERE severity.
- Bob showed the testcase results.

- Bob Ross showed bug200.txt
- Bob: This is also co-authored by Arpad. This gives a BUG message.
  In the testcase, [Model] sub-parameters are placed after a secondary keyword.
- Mike: Does the IBIS spec have this requirement? In that case the B message should
  be an E.
- Bob: That is the recommended remedy.
- Mike: The parser might not easily determine whether the text is a
  subparameter at that point.
- Bob added "or text" to the recommended message.

- Bob Ross showed bug201.txt
- Bob: This is from Lance Wang. There should be a warning that [Pullup] and
  [Pulldown] should not exist for Terminators.
- Mike: Does IBIS forbid that?
- Bob: No, but there is no way for a Terminator to use them.
  It is a warning in case they accidentally mark a driver as Terminator.
- Mike: This is LOW priority, but we no doubt will fix it.
  - Priority doesn't matter much, we always fix all open bugs.
  - That said, HIGH priority might trigger a new parser update.
- Bob: Sometimes bug fixes create new bugs, which become HIGH priority.
  - BUG200 occurred in 6.1.3.
  - BUG201 should be checked for all IBIS versions.
- Mike: In the past, people were not happy when we made existing IBIS files no longer
  pass IBISCHK. Maybe this new check should affect only IBIS 6.1+ models.
- Bob: This happened in an existing model. It probably was a cut and paste
  error, resulting in an Output model being called Terminator.

- Bob: Another bug from Arpad is a current check error given when polarity
  is reversed. The error message is correct, so that bug will not be filed.

- Bob: Another bug from Freescale gives non-monotonic note messages, but the
  non-monotonicity originally couldn't be detected by Mike LaBonte.
  A testcase can be created but there is no point in pursuing.
  IBISCHK is working as expected.

IBISCHK6.1.5:
- Bob: I might ask for a quote before the new bugs are classified.
  The contract will come after, and will state a contract price.

IBISCHK7 parser development:
- Bob: I have some new content to add to the contract document.
  - There is an issue with BCI_State if it is Failed or Error.
- Mike: That would be strange, but the EDA tool and models set the actual values.
  - That might be an ATM discussion.

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

Meeting ended: 12:23 ET

Next meeting August 14, 2018