Re: [sv-bc] void casting on void function call

I wouldn't object to allowing any expression in a void cast but
don't feel strongly about that.

I do think that systf task calls are very important to permit and
don't see any reason to disallow void functions.

Gord

Surya Pratik Saha wrote:
> Hi Gordon,
> function_subroutine_call is part of 'primary' in the BNF. Do you want 
> then void function call is an expression? Please do not compare with 
> 'C', as 'C' does not distinguish a statement and expression. In 'C', you 
> can simply write
> 1 + 2;
> which SV does not allow. So as per SV, any statement should not be void 
> casted, hence void function call too should not be allowed like task 
> enable.
> 
> Regards
> Surya
> 
> 
> 
> -------- Original Message  --------
> Subject: Re:[sv-bc] void casting on void function call
> From: Gordon Vreugdenhil <gordonv@model.com>
> To: Surya Pratik Saha <spsaha@cal.interrasystems.com>
> Cc: "sv-bc@eda-stds.org" <sv-bc@eda-stds.org>
> Date: Thursday, August 30, 2007 8:57:46 PM
>>
>> The question is whether the intent of "function_subroutine_call"
>> is restricted to "function's that are expressions".  There
>> is no real reason to make that restriction; C for example
>> allows void casts of void functions as well.
>>
>> If you want to argue this too narrowly, you'll argue yourself
>> into a bigger problem -- grammatically "function_subroutine_call"
>> is any call (including a task).  The semantic intent is in the
>> "function" prefix.  But "function_subroutine_call" isn't a
>> system_tf_call so one could argue that void'($bits(x)) would
>> be illegal.  Clearly that isn't the intent once you consider
>> all the systf file I/O cases where one wants to through away
>> return codes.
>>
>> I don't see any reason to not interpret this in the most
>> permissive way possible.  I'd even support allowing (at least)
>> systf task calls since that would be a great way for real
>> user code to insulate themselves against possible variations
>> in vendor implementations and/or LRM versions.  A great
>> example of that is with "$system" for which there is variation
>> in implementation.  Allowing a systask call in a void cast
>> would allow user code to have:
>>     void'($system("..."));
>> as completely portable code even among different vendors.
>>
>> Gord
>>
>> Surya Pratik Saha wrote:
>>> Hi,
>>> As per LRM-BNF, void casted statment is:
>>>
>>> *Subroutine_call_statement ::= subroutine_call;
>>> | void'(function_subroutine_call);*
>>>
>>> And for void function, LRM mentions following text (page no. 156 of 
>>> IEEE 1800).
>>>
>>> *Function calls are expressions unless of type void, which are 
>>> statements:
>>> a = b + myfunc1(c, d); //call myfunc1 (defined above) as an expression
>>> myprint(a); //call myprint (defined below) as a statement
>>> function void myprint (int a);
>>> ...
>>> endfunction
>>>
>>> *Can **function_subroutine_call **contain void function call? I don't 
>>> think so, as void function call is a statement. But some simulators 
>>> don't think like that and passes the following case:
>>>
>>> module top;
>>>     initial begin
>>>         void'(f());
>>>     end
>>>     function void f;
>>>         return ;
>>>     endfunction
>>> endmodule
>>>
>>> Any expert comment?
>>>
>>> -- 
>>> Regards
>>> Surya
>>>
>>>
>>> -- 
>>> This message has been scanned for viruses and
>>> dangerous content by *MailScanner* <http://www.mailscanner.info/>, 
>>> and is
>>> believed to be clean.
>>
> 
> 
> 
> 

-- 
--------------------------------------------------------------------
Gordon Vreugdenhil                                503-685-0808
Model Technology (Mentor Graphics)                gordonv@model.com


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.