SCE-MI: List of uncataloged issues

Hi,

I had as an action item to list all the issues that have been raised
recently (over the last 6 months or so) but that have not made it
into the issues list on the web site (http://www.eda.org/itc/issues).

  1) Questions about input-ready propagation
     http://www.eda.org/itc/hm/0250.html

     I had three questions and got one reply from John, but the
     questions are not fully resolved.

  2) SCE-MI initialization and binding during SystemC constructors
     http://www.eda.org/itc/hm/0256.html

     This issue was resolved with John's response
     http://www.eda.org/itc/hm/0257.html. A followon question given
     John's response is whether it is worth it to update the Routed
     example to do binding in SystemC constructors.

  3) Multiple BindMessage(In|Out)Port() of the same port
     http://www.eda.org/itc/hm/0258.html

  4) SCE-MI error handling questions and observations
     http://www.eda.org/itc/hm/0259.html

     There are several sub-issues here:

     a) Can the implementation call the error handler even if all
        API calls are done with a non-NULL SceMiEC pointer?

     b) Is it OK to mix the error handling styles (NULL and non-NULL
        SceMiEC) on different calls?

     c) The recoverable error thread which split into three threads:

          http://www.eda.org/itc/hm/0272.html
          http://www.eda.org/itc/hm/0285.html
          http://www.eda.org/itc/hm/0288.html

        John started down the road to propose changes to the spec
        to introduce a recoverable error type, but then retracted
        his proposals because he felt that the spec already has
        everything it needs to support the handling of recoverable
        errors. The only outstanding issue is how to report a
        failed lookup of an integer attribute (AttributeIntegerValue()).
        We cannot simply use a 0 return value to indicate the case
        where the attribute did not exist.

  5) John's proposal to eliminate implementation defined attributes
     http://www.eda.org/itc/hm/0291.html about 2/3 to 3/4 down in
     the message.

     This was John's proposal to avoid the issue mentioned in 4c
     above with non-existent integer attributes.

     A consequence of this proposal is that the Override*() methods
     become obsolete http://www.eda.org/itc/hm/0300.html.

  6) Proposal to change the declaration of error and info handler
     registration methods to allow recovery of previous handler
     http://www.eda.org/itc/hm/0294.html.

     I later revisited this in http://www.eda.org/itc/hm/0299.html.

  7) C++ exceptions from SCE-MI error and/or info handler?
     http://www.eda.org/itc/hm/0265.html

  8) Modifying and/or sending message received: is it legal?
     http://www.eda.org/itc/hm/0268.html

     I have received no response on this one yet.

Let me know if I missed anything.

Thanks,
Per

-- 
Per Bojsen                                Email: <bojsen@zaiqtech.com>
Zaiq Technologies, Inc.                   WWW:   http://www.zaiqtech.com
78 Dragon Ct.                             Tel:   781 721 8229
Woburn, MA 01801                          Fax:   781 932 7488