Minutes from the 18 September ibis-atm meeting are attached.
IBIS Macromodel Task Group
Meeting date: 18 September 2018
Members (asterisk for those attending):
ANSYS: Dan Dvorscak
* Curtis Clark
Cadence Design Systems: * Ambrish Varma
Brad Brim
Kumar Keshavan
Ken Willis
eASIC: David Banas
GlobalFoundries: Steve Parker
IBM Luis Armenta
Trevor Timpane
Intel: * Michael Mirmak
Keysight Technologies: Fangyi Rao
* Radek Biernacki
Ming Yan
Stephen Slater
Mentor, A Siemens Business: John Angulo
* Arpad Muranyi
Micron Technology: * Randy Wolff
* Justin Butterfield
SiSoft: * Walter Katz
* Mike LaBonte
SPISim: * Wei-hsing Huang
Synopsys: Rita Horner
Kevin Li
Teraspeed Consulting Group: Scott McMorrow
Teraspeed Labs: * Bob Ross
The meeting was led by Arpad Muranyi. Curtis Clark took the minutes.
--------------------------------------------------------------------------------
Opens:
- None.
-------------
Review of ARs:
- Randy to investigate if/why/how a clock waveform input might be used.
- In progress.
- Michael M. to investigate if/why/how a clock waveform input might be used.
- In progress.
--------------------------
Call for patent disclosure:
- None.
-------------------------
Review of Meeting Minutes:
Arpad asked for any comments or corrections to the minutes of the September 11
meeting. Walter moved to approve the minutes. Michael M. seconded the motion.
There were no objections.
-------------
New Discussion:
IBIS v7.0 Section 3.2 issue:
Arpad noted that the Editorial task group had come upon an issue at their Friday
meeting and thought it might be more than a pure editorial issue. They wanted
to discuss it in ATM.
Michael M. shared the latest draft of v7.0. Section 3.2, item 3., discusses the
character set allowed in filenames, stems, extensions, and the rules surrounding
them. BIRD186 added a large number of definitions to IBIS (file name, stem,
path, etc. - section 3.1). All are meant to clarify what we mean by file name,
full path, etc., because we now have so many types of files that may have
slightly different naming conventions. Extension is defined as the portion
after the last period. One concern is whether the syntax rules should allow
file names that have no extension but have a period at the end. The rules are
generally pretty well defined, but for parameter definition files in particular
no specific extension was defined.
Arpad noted one particular conflict in the definitions. Section 3.2, item 3.,
defines a character set that includes the period, but in 3.1 the definition of
extension says it is the portion after the last period. Bob and Radek thought
the intent was clear. Walter agreed, but noted we could fix it if we wanted to
bother.
Bob asked if it's possible to have a file name with a null stem, i.e., no stem
and then ".txt" for example. Radek said such files are legitimate in the OSs.
Radek noted that IBIS does not allow them because the stem must be at least one
character long. Michael M. noted that the Editorial Task group had explicitly
added the requirement that the stem not be null. That language had not existed
previously.
Walter reviewed an email he'd sent in reply to the ATM agenda email. It
proposed a change to address Arpad's concern about the conflicting definitions
of extension and the character set it can contain. Instead of Walter's change,
Arpad proposed adding "(Note that "." is not allowed in extension)" at the end
of the second sentence in 3.2, item 3. Bob said he was okay with this
parenthetical. Walter agreed and said he felt this change was purely an
editorial change (clarification).
Walter's email also proposed adding ".." along with the existing "./" as
character sequences that are not allowed (section 3.2, item 3). Walter noted
that this would help resolve pathological cases.
In addition, in section 3.1, Walter proposed adding the sentence, "The stem
should contain at least one character." to the definition of stem.
Arpad noted an additional correction for section 3.2, item 3. In the second
sentence of that section, the use of "file name" should probably be replaced
with "stem". Walter then noted an additional issue, that the character set for
path information had not really been addressed.
Walter said the question is whether we wanted to make the proposed changes as
editorial, or whether they needed a BIRD. Radek and Walter noted that they
felt this was purely editorial. Bob noted that it could be editorial, but he
would like it added to the list of notes on clarifications of BIRDs for 7.0.
Arpad asked Michael M. if he was okay with these changes. Michael said he was
okay with them, and if the list of changes were sent out to the ATM and
Editorial groups, Editorial could tackle the changes in their next meeting.
Michael noted Walter's question about whether these require a BIRD, and said
that if they are handled as purely editorial then the changes will only be
reviewed by everyone once the final IBIS 7.0 is reviewed. Michael said his real
question was whether EDA vendors, and to a lesser extent model makers, were
going to have any problems with the rules being proposed. Are there any
significant differences between EDA tool assumptions and these IBIS definitions?
Arpad noted that he did not like any rule that allowed "." as the only character
in a stem (e.g. "..txt"). He said OSs might allow it, but he anticipated tools
having trouble with it. Radek said he agreed with Walter's proposal to state
that ".." is not permitted. This would address the issue. Radek noted that it
is okay for IBIS to be more restrictive than the OSs. Arpad proposed explicitly
stating that more than two periods in a row is also prohibited. Walter/Radek
noted that if ".." is disallowed, then "..." is automatically disallowed.
Bob noted concern about overloading ibischk with rules to test, but Walter noted
that the checks required here would be trivial for the parser. Michael M. said
he was not concerned about people violating these rules when creating files by
hand, but he was concerned that some automated processes might create files that
violated these rules. Radek noted that we are only defining the rules for files
distributed by the model makers, and the EDA tools are free to do what they
want.
Walter noted that he thought file names should not be allowed to end with a
period (e.g. xyz.). Radek said that this is explicitly allowed in the parameter
definition files section. Walter said he thought this was an error in that
section. Bob noted that because we never specified a particular extension for
parameter definition files, we had left it open. Bob agreed that is was a
mistake to have left this undefined.
Arpad asked why we explicitly allowed "." at the end of a file name in the
parameter definition file sections. Bob noted that such language already
existed in v6.1. Arpad agreed but suggested we should remove that if there
was no good reason for having it. Walter agreed and suggested that we continue
to allow a file name that only has a stem (e.g. xyz), but remove the language
that allows "xyz.". Radek said removing this language wouldn't solve the
problem entirely. Arpad noted it could help us simplify the rules in 3.2.
Walter proposed that we also add the following to section 3.2, item 3.:
File name should not begin with a ".".
File name should not end with a ".".
Walter said he thought these changes would not affect any IBIS file that has
ever been created to date. Michael M. said we just want to be careful about
making changes in the Editorial task group that no one else will see until the
final v7.0 comes up for review. Mike L. noted that we have written BIRDs before
for purely editorial clarifications, so the bar is low for what could require
a BIRD. Walter suggested that it should be a trivial BIRD to write, and the
Editorial task group could incorporate the changes immediately and assume that
the BIRD documenting them would be approved. Bob said we would talk about it
at the next Editorial meeting. Michael M. said we need to decide who will
write the BIRD. Walter said he would send out a revised email containing the
changes discussed in the meeting.
- Mike L.: Motion to adjourn.
- Radek: Second.
- Arpad: Thank you all for joining.
AR: Walter to send a follow up email containing the discussed revisions to the
file name conventions.
-------------
Next meeting: 25 September 2018 12:00pm PT
-------------
IBIS Interconnect SPICE Wish List:
1) Simulator directives
