RE: [sv-bc] Mantis 2102

Hi,

I agree that there is benefit to a statement about atomicity.
Consider a statement like

A[2:1] = A[1:0];

You could do this even in Verilog-1995 with vectors, as in a shift
register.

The statement probably belongs more properly in 7.6 "Array assignments".

Thanks,
Shalom

> This is the kind of sentence that needs a good context to 
> bring out its meaning. I agree that the prose here has become 
> confused.
> 
> There are at least two possible topics for the sentence which 
> was dropped:
>    - The atomicity of the assignment operation (vs references 
> to the assigned
>      elements in the LHS of that assignment) - we don't want 
> a data dependence
>      between the LHS and the RHS of a single assignment operation.
>    - The granularity of checking for multiply-driven nets.
> 
> In context, it is this latter topic that is being discussed.
> The dropped sentence does not add anything essential to that 
> discussion.
> Perhaps it was added because the first sentence - about the 
> elements being "examined" individually suggests that they 
> might "evaluate" individually and therefore there /is/ data 
> dependence between LHS and RHS of a continuous assignment.
> 
> There is a real semantic issue in play in these sentences.  
> Until a better wording can be developed, I would leave this 
> text alone.  The concept of continuous assignment of arrays 
> with feedback loops is strange and too delicate for me to 
> tackle while eating lunch :-)
---------------------------------------------------------------------
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.