Minutes, IBIS Quality Committee

22 March 2016

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
Signal Integrity Software         * Mike LaBonte
Teraspeed Labs:                   * Bob Ross

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: We have a new bug from Walter.

ARs:

- Mike update BUG174
  - Done

- Mike upload DLL_checking_draft2.docx
  - The intended file was actually DLL_Checking_draft3.docx.
  - This was not uploaded because it is identical to DLL_Checking_draft2.docx,
    already uploaded.

- Mike add comments to CMPNT, DLY, MSPEC sections.
  - No new progress.

- Lance add comments to EBD section.
  - No new progress.

- Bob add comments to CIRCUIT section.
  - No new progress.

BUG from Walter Katz:

- Bob showed fail.ibs from Walter.
- Bob: A [Model Selector] has Series and Input models.
  - There is a [Series Pin Mapping] so the Input model is an error.
- Lance: How can a Series model be assigned to a single pin?
- Bob: The [Series Pin Mapping] assigns it to both.
  - but it probably should point to Selector, not Series.
- Mike: [Model Selector] should not point to 1 pin and 2 pin classes of Model_type.
  - The [Series Pin Mapping] is not required to point to the Selector.
- Bob: Correct.
- Mike: The model names in the testcase should not be confused with Model_type names.
  - IBISCHK find no errors or warnings?
- Bob: Correct.

- Bob changed the file to IBIS_Ver 6.1 and tested.
  - The result was the same.
  - IBISCHK would flag an error it the Series model had been used directly.

- Mike: The model_name in [Pin] is not supposed to be a Series type.
  - The use of [Model Selector] is obscuring the type
  - We should add two rules to IBIS:
    - If [Pin] calls a [Model Selector] all models under that must be one-pin.
    - If [Series Pin Mapping] calls a [Model Selector] all models under that must be two-pin.

- Bob: We can file a bug and note that the spec needs to change.
- Mike: We should change the spec first to justify making it an error.
- Bob: If [Pin] calls a [Model Selector] that has only Series types no error is flagged.
- Mike: That should be an error.

- Mike: We should not have the one-pin vs. two-pin issue fixed until the spec is in order.
  - There may be a change or two from the IBIS 6.2 work.
- Bob: Agree.
- Lance: There is no rush, this can wait for another contract.
- Bob: We should file the bug and discuss in Open Forum.

- Bob showed that IBISCHK flags errors for [Pin] that calls Series models
  through a [Model Selector] that has only Series models.


AMI Executable checking:

- Mike showed 2 docx files and a spreadsheet.
- Lance: This is too complicated.
- Mike: We should consider simple sanity checks.
  - AMI_Init() may exist but it might crash.
- Lance: Successfully loading the DLL should be enough of a test.
- Mike: Without calling it we would not know if there is a library dependency.

Bob motioned to adjourn. Lance seconded.

Meeting ended: 12:30 ET


Next meeting Mar 29, 2016