RE: [vhdl-200x] Modular types

OK, I think I see now... let me see if I'm right :)

One way to do it would be to allow the type-conversion "M (Integer'(-1))" to wrap-around in the same way as all other operations on a modular type do.  The argument I see in the appendix to AI-00340 is that type conversion is sometimes "lossy" in other areas (eg real->integer), so why not here too.

The other option (which is what was implemented in Ada as I understand it) was to add an attribute to the modular type to allow a "different-sort-of-type-conversion".

Does that covert it?

Thanks,
Martin

-----Original Message-----
From: owner-vhdl-200x@eda.org [mailto:owner-vhdl-200x@eda.org] On Behalf Of tgingold@free.fr
Sent: 26 June 2014 18:06
To: vhdl-200x@eda.org
Subject: Re: [vhdl-200x] Modular types

> Thanks Tristan - I noticed this behaviour but forgot to comment on it.  
> For other interested parties, the relevant bit of the LRM is:

[...]

> Now, I can see the reasoning - without an explicit type-conversion 
> (using 'mod), it's more in the VHDL-like "fail as early as possible"
> mindset.  But perhaps without an explicit type-conversion you 
> shouldn't be allowed to assign an integer to a modular type, so it 
> should fail at compile-time?

Ada is strongly typed as VHDL, so you cannot assign a signed integer to a modular integer without an explicit type-conversion.  So indeed, such direct assignment is rejected at compile time.

> I imagine you've written rather more Ada code than I have, so could 
> you elucidate a bit more on the problems the current Ada Behaviour 
> causes please?  You mention it is "error-prone".  What would your 
> preferred behaviour be - to fail at compile-time, or treat all 
> assignments as asking for a "wraparound", or something else entirely?

The issue is the behaviour of the type-conversion.  When you use modular type (in Ada), you think that all operations wraps-around.  That's true except for type conversion.  If M is a modular type, M (Integer'(-1)) raise an error instead of returning M'Modulus - 1.  I think this was a mistake (and I am not alone, see AI-00340 for opinions).

Regards,
Tristan.

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


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