Timothy Miller wrote:
And I wonder if there isn't a better way to do a rotate in verilog
that the synthesizer will make sense of. You know, every time I try
to so something clever with shifters, like manually code a barrel
shifter, I always get a much worse result than if I just use the >> or
<< operator in verilog.
I thought that was a fairly slick way of implementing the rotate. I
know how to do the clocked rotate in discrete logic, but I've had no
idea what the verilog implementation would look like. I don't know that
it is worth the gates to do it as a true barrel shift/rotate rather than
the clocked version. I'll leave that up to you. My take on it would be
to implement the clocked version, then later when/if we find out we have
the spare gates and/or the performance is too lousy, we code it up as a
true barrel shift.
Something in the back of my mind is telling me that there are <<< and
operators in either VHDL or Verilog, but I cannot remember what
they do... and google chokes on that.
The few code examples for barrel shifters that I was able find used the
same code you did.
Patrick M
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
