Problem report on the SCE-API from Yung-ming Fang <yfang@cadence. com>

Subject: Problem report on the SCE-API from Yung-ming Fang
From: Bailey, Brian (brian_bailey@mentorg.com)
Date: Mon Jul 22 2002 - 10:22:43 PDT

Hi, Dear Sirs,

I'd like to report some possible problems in the document.
In Section 4.2.1, page 59 of 94,

"It shall be assumed that any module that is to be officially considered
a transactor must have
at least one SceMiClockControl instance immediately inside it."

If I read it right,
transactor is defined as any instance (module instance) that has at
least one SceMiClockControl as child.
If this is correct, the example at page 43 is wrong.

" dOriginAnchorage = scemi->BindMessageInPort ("anchorage", "TokenIn"
); "

"anchorage" is NOT a transactor because it has no SceMiClockControl as a
direct descendant.
See Section 3.2.2.2 or Fig 8 at page 24, anchorage is an instance of
module OrigDest, which has no SceMiClockControl as a child.

Another possible problem is the portname.
At page 43 of 94,
" dDestinationMaui = scemi->BindMessageOutPort ("maui", "TokenOut",
&binding ); "

maui is an instance of module Destination (its module is defined in page
27),
According to portname definition in Section 4.2.1, the correct portname
should be "message".

Hope in the next release, the example will be correct.

Regards,

fang

==============================================
Brian Bailey brian_bailey@mentor.com
Chief Technologist Tel: 503 685 1371
System Verification Development Office Fax: 503 685 1652
Mentor Graphics Corp. 8005 S.W. Boeckman Rd. Wilsonville, OR 97070

This archive was generated by hypermail 2b28 : Mon Jul 22 2002 - 10:24:49 PDT