Are operands of short-circuit ops self-determined?

D

David Jones

Guest
In Table 29 on p. 60 of IEEE Std 1364-2001, it is clearly indicated that
the operand lengths for the "&&" and "||" operators, as well as the comparison
operators are sized to max(L(i),L(j)).

OTOH, Table 4-21 on p. 44 of IEEE Std 1364-1995, it is clearly indicated that
the operand lengths for these operators are self-determined.

I would have assumed that the operands of "&&" and "||" should be self-
determined, and the operands of the comparison operators should be sized
to max(L(i),L(j)). Is this actually true?

Somewhat related, are Verilog errata freely available (assuming that one has
already paid for both IEEE Std 1364-1995 and IEEE Std 1364-2001)?
 
dej@coup.inode.org (David Jones) wrote in message news:<F8lgb.2639$oO4.332959@news20.bellglobal.com>...
I would have assumed that the operands of "&&" and "||" should be self-
determined, and the operands of the comparison operators should be sized
to max(L(i),L(j)). Is this actually true?
You are correct. These operators were on the same line in the 1995 standard.
They fixed the table to say that comparison operands got sized to match each
other (though they are not affected by the size of the rest of the expression
like a context-determined expression would be). Unfortunately, the logical
operators got left in the same line of the table, even though the changed
line doesn't apply to them.

You can see the official erratum on this at

http://www.boyd.com/1364_btf/report/full_pr/70.html


Somewhat related, are Verilog errata freely available (assuming that one has
already paid for both IEEE Std 1364-1995 and IEEE Std 1364-2001)?
There has been discussion of how to distribute errata, but I don't think
anything has been done about it. I believe that there is going to be a
new printing to replace the 2001 standard, because the IEEE actually printed
an earlier draft instead of the final version that got approved. This will
also include some obvious typographical fixes, but not anything that changes
the actual meaning (even situations like this where the existing meaning is
known to be wrong).
 

Welcome to EDABoard.com

Sponsor

Back
Top