[sv-ac] RE: Review of Mantis 2093: Checker construct should permit output arguments

Hi Jacob,

Please, see my comments below.

Thanks,
Dmitry

From: Katz, Jacob
Sent: Monday, August 29, 2011 16:14
To: Korchemny, Dmitry
Subject: RE: Review of Mantis 2093: Checker construct should permit output arguments

17.2
Acc. to the rules of type and direction inference where it is omitted, it turns out that in the following:
                checker c(input sequence s, output o)
the type "sequence" is inferred for "o", which is illegal, while the declaration seems intuitive. I think the inference of type should stop not just on the next type, but also on next direction specification
[Korchemny, Dmitry] I agree, will modify the definition.

Example 3: comma missing after "clock"

checker my_check3 (logic a, b, event clock output bit failure, undef);

[Korchemny, Dmitry] Fixed.

17.3
The sentence:
                The mechanism for passing arguments to a checker is similar to the mechanism for passing arguments to a property
should probably talk only about input arguments. Output arguments are described separately and the mechanism there is different.
[Korchemny, Dmitry] Properties can also have output arguments, if the actual arguments are local variables in another property. I tried to follow up these definitions. We can discuss this further at today's meeting.

Why is the difference between the cases when the actual is variable_lvalue and formal is typed vs. untyped?
[Korchemny, Dmitry] If the formal is untyped we can use substitution semantics. If the formal is typed then the substitution semantics is not always possible. This is consistent with the definitions for sequences and properties. See 16.8.1 Typed formal arguments in sequence declarations:
If the actual argument is a variable_lvalue, references to the formal shall be considered as having the formal's type with any assignment to the formal being treated as though there was a subsequent assignment from the formal to the actual argument.

--------------------------------
Jacob M. Katz | jacob.katz@intel.com<mailto:jacob.katz@intel.com> | Work: +972-4-865-5726 | iNet: (8)-465-5726

From: Korchemny, Dmitry
Sent: Monday, August 29, 2011 15:12
To: Katz, Jacob
Subject: Review of Mantis 2093: Checker construct should permit output arguments

Hi Jacob,

Could you send your review?

Thanks,
Dmitry
---------------------------------------------------------------------
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.