Subject: Re: [sv-ac] Updated Assertion chapter
From: Surrendra Dudani (Surrendra.Dudani@synopsys.com)
Date: Fri Apr 18 2003 - 10:59:59 PDT
At 12:46 PM 4/14/2003 -0500, you wrote:
Hi Adam,
Please see comments.
Thanks
Surrendra
>HI Surrendra;
>
>More comments on the document.
>
>
>Fig 1-1 I thought we were going to add a sampling indicator to the figure.
> Also, I thought the first dashed line should move forward one tick to the
> one just before clock tick 3.
Fixed in draft5.
>pg 2, bottom - Would it be possible to reference the name of the declared
>property
> if a property instance is asserted? E.g.
>
> property afail; ... endproperty
> assert property (afail);
>
> The error message reported through $error() could include the fact
> that we used
> the property named 'afail'.
Yes.
>pg 4, last paragraph ,,, "An expression such as (clk && gating_signal) or
>(clk iff gating_signal) may be used to represent a gated clock."
>
>Suggest ... "represent gated clocks."
>
>The former suggests that they both are equivalent methods.
will be fixed in the next draft
>Syntax 1-2.
>
> Must note the following:
>
> '*='
> '*->'
> '##'
>
> These are all multi-character operators - that is they can not have
> embedded space,
> like " * - > ".
They seem correct in the draft5 document.
>pg 6. last example shows "##1`true" - This will produce, after the
>preprocessor,
> "##11". Insert a space between 1 and `true.
Will be fixed in next draft
>Add to the note that `true is used only for visual effectiveness, not for
>technical reasons.
Will be fixed in next draft
>pg. 11 Same problem with missing spaces between ##1 and `true. [As
>experienced, this
> will be a source of syntax errors by users too...]
Will be fixed in next draft
> End of page, 'nonconsecutive exact repetition' should be goto
> repetition or other
> explanatory phrase.
fixed in draft5
>pg 19. We should have a figure showing first_match functionality like we
>do for the
> other operators.
>
>pg 22. We should have a figure showing within functionality like we do for
>the other
> operators.
>
>
>pg 23. We should have a figure showing ended method functionality like we
>have for
> the operators.
Good idea. However, there is a strict time constraint to finish the LRM.
Examples with figures are taking too much time to get it right.
>pg 24. The points made about implication dances around this fact:
> a |-> b is equivalent to a |-> $first_match(b)
Will be fixed in next draft
> Thus we should include this here - instead of requiring the user to
> intuitively figure
> this out.
>
>pg 31, Multi-clock sequences needs a figure for a few examples.
>
>pg 33, needs a figure for matched functionality.
Same as for other figures.
>pg 41. Why only allow binding of module|interface|program to modules? Why not
> interfaces to be symmetric? Program blocks ?
SV-AC must approve to extend this.
>pg 42. Add example of program block being bound to module/instance, etc.
There is an example of binding a program instance to a module without
showing the contents of the program. There is another example of showing
the assertion contents of an interface and binding the interface to a
module. These two examples should be adequate.
**********************************************
Surrendra A. Dudani
Synopsys, Inc.
377 Simarano Drive, Suite 300
Marlboro, MA 01752
Tel: 508-263-8072
Fax: 508-263-8123
email: Surrendra.Dudani@synopsys.com
**********************************************
This archive was generated by hypermail 2b28 : Fri Apr 18 2003 - 11:01:43 PDT