Line 38... |
Line 38... |
---- You should have received a copy of the GNU Lesser General ----
|
---- You should have received a copy of the GNU Lesser General ----
|
---- Public License along with this source; if not, download it ----
|
---- Public License along with this source; if not, download it ----
|
---- from http://www.opencores.org/lgpl.shtml ----
|
---- from http://www.opencores.org/lgpl.shtml ----
|
---- ----
|
---- ----
|
----------------------------------------------------------------------------
|
----------------------------------------------------------------------------
|
-- ----
|
---- ----
|
-- CVS Revision History ----
|
-- CVS Revision History ----
|
-- ----
|
---- ----
|
-- $Log: not supported by cvs2svn $ ----
|
-- $Log: not supported by cvs2svn $ ----
|
-- ----
|
---- ----
|
--
|
---- ----
|
-- quick description
|
-- quick description
|
--
|
--
|
-- Based upon the using a shift register as a fifo which has been
|
-- Based upon the using a shift register as a fifo which has been
|
-- around for years ( decades ), but really came of use to VHDL
|
-- around for years ( decades ), but really came of use to VHDL
|
-- when the Xilinx FPGA's started having SRL's.
|
-- when the Xilinx FPGA's started having SRL's.
|
Line 55... |
Line 55... |
-- In my view, the definitive article on shift register logic fifo's
|
-- In my view, the definitive article on shift register logic fifo's
|
-- comes from Mr Chapman at Xilinx, in the form of his BBFIFO
|
-- comes from Mr Chapman at Xilinx, in the form of his BBFIFO
|
-- tecXeclusive article, which as at early 2008, Xilinx have
|
-- tecXeclusive article, which as at early 2008, Xilinx have
|
-- removed.
|
-- removed.
|
--
|
--
|
-- This version is for 'earlier' devices that have an inherent shift
|
--
|
-- register of 16 bits.
|
-- using Xilinx ISE 10.1 and later, the tools are getting real clever.
|
|
-- In previous version of ISE, SRL inferance was not this clever.
|
|
-- now if one infers a 32 bit srl in a device that has inherantly 16
|
|
-- bit srls, then an srl and a series of registers was created.
|
|
-- In 10,1 and later, if you infer a 32 bit srl for a device with
|
|
-- 16 bit srls in, then you end up with the cascaded srls as expected.
|
|
--
|
|
-- Well done Xilinx..
|
--
|
--
|
|
|
library IEEE;
|
library IEEE;
|
use IEEE.STD_LOGIC_1164.ALL;
|
use IEEE.STD_LOGIC_1164.ALL;
|
use ieee.NUMERIC_STD.all;
|
use ieee.NUMERIC_STD.all;
|