Re: [sv-bc] Is '1 allowed in a concatenation?

Hi All.

It could be argued that "unsized" refers to a user operation,
rather than to a characteristic of the literal. "unsized"
means that the user has not added a size. If so, then
failing to "size" a literal might permit it, nevertheless,
to be concatenated. This seems to be Cliff's argument.

I more or less agree with Shalom's argument: If a literal
falls into an unsized category, it should not be legal
in a concatenation, regardless of whether its
width is implied or specified in the Std. So, it would
be a SYNTAX issue, not a semantic one: A literal lacking
an explicit size can't be concatenated.

The Std definitely is amphibolic on this and should be
changed, no matter what the decision.

On 03/22/2010 10:24 AM, Clifford E. Cummings wrote:
> Hi, Shalom -
>
> It is a bit of a stretch, but here is why I think we are covered (this
> could be better):
>
> At 05:46 AM 3/22/2010, Bresticker, Shalom wrote:
>> Is a literal like '1 allowed in a concatenation?
>>
>> On the one hand, 11.4.12 says, "Unsized constant numbers shall not be
>> allowed in concatenations. This is because the size of each operand in
>> the concatenation is needed to calculate the complete size of the
>> concatenation."
>>
>> On the other hand, 5.7.1 says, "An unsized single-bit value can be
>> specified by preceding the single-bit value with an apostrophe ( ' ),
>> but without the base specifier. All bits of the unsized value shall be
>> set to the value of the specified bit. In a self-determined context,
>> an unsized single-bit value shall have a width of 1 bit, and the value
>> shall be treated as unsigned."
>
> (1) unsized single-bit value can be preceded with an apostrophe ( ' )
> (2) All bits of the unsized value shall be set to the value of the
> specified bit
> (3) In a self-determined context, an unsized single-bit value shall have
> a width of 1 bit (now it has a size)
>
> Just my reading of a somewhat obscure passage. I don't think anyone
> would object to better wording, if you can propose it.
>
> Regards - Cliff
>
> At 07:47 AM 3/22/2010, Bresticker, Shalom wrote:
>> Hi, Cliff.
>>
>> It is clear that if '1 is legal in a concatenation, then it is a
>> single bit.
>> The question is whether it is legal.
>>
>> At best, the LRM is ambiguous about it.
>>
>> Since 11.4.12 says that unsized constants are not allowed in
>> concatenations and 5.7.1 calls '1 unsized, it is reasonable to
>> conclude that it is not legal. As pointed out, a simple unsized
>> constant is also not allowed even though its size is well-defined (32
>> bits).
>>
>> Regards,
>> Shalom
>>
>> > -----Original Message-----
>> > From: owner-sv-bc@eda.org [mailto:owner-sv-bc@eda.org] On
>> > Behalf Of Clifford E. Cummings
>> > Sent: Monday, March 22, 2010 4:40 PM
>> > To: sv-bc@eda.org
>> > Subject: Re: [sv-bc] Is '1 allowed in a concatenation?
>> >
>> > Hi, Shalom -
>> >
>> > There may be a wording problem in the standard, but adding '1 or '0
>> > to a concatenation is the same as adding 1'b1 or 1'b0.
>> >
>> > I point this out as an exception to '0 and '1 expansion in my
>> > training classes.
>> >
>> > Rationale:
>> > What would be expected if you had declared:
>> > logic [15:0] data1 = {'1,8'hAA};
>> > logic [15:0] data2 = {8'hAA, '1};
>> > logic [15:0] data3 = {'0, 8'hAA,'0, '1};
>> >
>> > There would have to be some rather interesting assumptions on the
>> > part of the compiler to determine the intended size of the '0 or '1
>> > expansion, including a number of special cases. The assumptions might
>> > not be obvious to an end user and debugging these declarations might
>> > be rather time-consuming.
>> >
>> > It was just easier to say that in these self-determined contexts,
>> > each '1 or '0 is just a single bit.
>> >
>> > Regards - Cliff
>
> ----------------------------------------------------
> Cliff Cummings - Sunburst Design, Inc.
> 14355 SW Allen Blvd., Suite #100, Beaverton, OR 97005
> Phone: 503-641-8446 / FAX: 503-641-8486
> cliffc@sunburst-design.com / www.sunburst-design.com
> World Class Verilog & SystemVerilog Training
>

-- 
      John Michael Williams
      Senior Adjunct Faculty
Silicon Valley Technical Institute
-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.