RE: [sv-cc] review of 0226 implementation in draft8

As far as I can see, the last P1800 meeting preceded the Annex N discussions.

Here is an excerpt from the January 14 SV-CC meeting minutes:


* Item 2226

Chuck has tried to do some review of Item 2226. John also looked at it some too, but is concerned that he may miss details. Chuck asked if we should roll some of the include file issues into comments back to the editor. Abi pointed out that they were missing from the proposal, so we cannot blame Stu. John pointed out that common sense wise, we should ask him to put it in, regardless of where the error came in. Jim suggested that we pass a motion to ask the P1800 committee to allow a new proposal to handle the header file issue. Chas thinks there is no harm in creating such a proposal. Abi and Chuck will work on such a proposal. John will review and mark as closed or send back to the editor. Chas will also review. Chuck will look to see if there is a record of his last review of it.

* Issues in the sv_vpi_user.h file

There was a discussion on how to go about verifying the include file. It is a fairly complicated process. Discussion on what to do with the range from 800-899. The reader API in 1800-2005 used up to 874. John suggested that we just excise the term 'Data Read API' from the include file only.

Shalom
________________________________
From: owner-sv-cc@server.eda.org [mailto:owner-sv-cc@server.eda.org] On Behalf Of Stuart Sutherland
Sent: Tuesday, January 27, 2009 10:31 AM
To: 'Shields, John'; 'SV-CC'
Subject: RE: [sv-cc] review of 0226 implementation in draft8

All of John's notes below can be made as editorial corrections, and will be made in the final version of draft 8.

Chas., I forget whether it was decided at the last working group meeting to correct the numbering in Annex N as editorial changes, or to flag the errata during the ballot process.  Do you remember?

Stu
~~~~~~~~~~~~~~
Stuart Sutherland
stuart@sutherland-hdl.com
(503) 692-0898

From: owner-sv-cc@eda.org [mailto:owner-sv-cc@eda.org] On Behalf Of Shields, John
Sent: Monday, January 26, 2009 11:22 PM
To: SV-CC
Subject: [sv-cc] review of 0226 implementation in draft8

Hi,

Gosh this was fun!  Clause 36, 38, and Annex C and L were implemented correctly.  Clause 37 has some minor cross reference issues that were not what was documented and one issue with diagram 37.39 Frames.  Annex N has been discussed and concur with the concerns raised about numbering issues with vpiAllocScheme and cbEndOfObject.  These are editorial issues which should be corrected. There is also an omission of vpiObjId, as Abi noted a couple of weeks ago.  It was the first implementation of 0226 changes but not in any subsequent revisions of our Annex N inputs to the editor. Clearly, I dropped it and we all missed it across at least 2 reviews.  It has always been on the information model diagram, of course.  I don't care how we handle it procedurally, but it would be just sad to drop the ball on it again if we should defer it.

Regards, John
---------------------
The details are:

Page 879, 37.3.7 is currently:

Other transient objects include:
1) threads (see 37.39).
2) outdated and out of scope references made within a thread
3) iterators (objects of type vpiIterator), which are created by calls to vpi_iterate() (see 37.23).
4) a vpiSchedEvent created by vpi_put_value() (see 37.34)
5) callbacks (see 37.36).

and should be:

Other transient objects include:
1) threads (see 37.40).
2) outdated and out of scope references made within a thread
3) iterators (objects of type vpiIterator), which are created by calls to vpi_iterate() (see 38.23).
4) a vpiSchedEvent created by vpi_put_value() (see 38.34)
5) callbacks (see 38.36).

On page 915, in note 10) of diagram  37.29 Class variables and class objects, it should be:

10) For details on class object specific callbacks, see 38.36.1

In diagram 37.39 Frames, there remains a property:
-> validity
int: vpiValid

which no longer exists in the information model.  I had removed this from the framemaker version showing "changes only" for this clause.  In the full version of clause 37, this property was still there.  As you all know, it easier for everyone to review a succinct changes only version and much easier for the editor to incorporate my integrated version of those changes. This was my editorial mistake and cleverly hidden, I might add.

I will not repeat the Annex N changes to editor assigned values, as Charlie documented them adequately in his Mantis note to 0226.  I will emphasize that we must have property vpiObjId in Annex N with an appropriately assigned value:

#define vpiObjId                 <editor value> /* Mantis 2226 */




--
This message has been scanned for viruses and
dangerous content by MailScanner<http://www.mailscanner.info/>, and is
believed to be clean.

--
This message has been scanned for viruses and
dangerous content by MailScanner<http://www.mailscanner.info/>, and is
believed to be clean.
---------------------------------------------------------------------
Intel Israel (74) Limited

This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.