This is a resend. The eda.org reflector rejected the previous message because the attachment was too large. The example files are now attached instead to Mantis item 3087.
Jim
From: Jim Vellenga
Sent: Wednesday, 5 Oct 2011 9:58 AM
To: Arturo Salz (Arturo.Salz@synopsys.com); SV-CC (sv-cc@eda.org)
Subject: Arturo's comments -- esp. 35.5
Arturo,
Well, that was certainly a detailed and comprehensive review. Thank you for doing that and for getting it out to us in such a short time. There are a number of questions that the document will be much better for when we insert explicit answers. The formatting comments are certainly helpful too.
I'll have to admit that I had many of the same concerns about 35.5 and the way that inheritance relations are and are not passed across the intermediate layer. Can this possibly work, and if so, how? Accordingly, I spent quite a bit of time last spring prototyping the operations. I am attaching three of the resulting test cases to this e-mail. These specifically show that it is possible to do what is proposed in terms of import classes in SystemVerilog inheriting from a nonimported SystemVerilog class, or recovering the identity of a handle as a derived class object when the handle is passed back and forth as a base class object pointer.
I notice also that a number of your objections are against having an intermediate layer developed within the DPI_OO:: namespace. You have already told us that what your customers want is really a direct programming interface to C++. That is certainly a reasonable concern, and indeed, an intermediate layer is not necessary for that. We are proposing the intermediate layer, however, for the following two reasons:
-- This gives SystemVerilog a base for extending DPI to other object-oriented languages in the future, and
-- The C++ intermediate layer is explicitly structured to handle the kind of objections you made in connection with section 35.5.
While we may need to be more explicit in section 35.5 and other places, I hope an inspection of the attached examples will give you some hope that the details can be worked out.
Regards,
Jim
Jim Vellenga | Member of Consulting Staff | Cadence
P: 978.262.6015 F: 978.262.6636 www.cadence.com<http://www.cadence.com>
-- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.
This archive was generated by hypermail 2.1.8 : Wed Oct 05 2011 - 07:05:04 PDT