277 lines
8.5 KiB
Verilog
277 lines
8.5 KiB
Verilog
// ***************************************************************************
|
|
// ***************************************************************************
|
|
// Copyright 2014 - 2017 (c) Analog Devices, Inc. All rights reserved.
|
|
//
|
|
// In this HDL repository, there are many different and unique modules, consisting
|
|
// of various HDL (Verilog or VHDL) components. The individual modules are
|
|
// developed independently, and may be accompanied by separate and unique license
|
|
// terms.
|
|
//
|
|
// The user should read each of these license terms, and understand the
|
|
// freedoms and responsibilities that he or she has by using this source/core.
|
|
//
|
|
// This core is distributed in the hope that it will be useful, but WITHOUT ANY
|
|
// WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
|
|
// A PARTICULAR PURPOSE.
|
|
//
|
|
// Redistribution and use of source or resulting binaries, with or without modification
|
|
// of this file, are permitted under one of the following two license terms:
|
|
//
|
|
// 1. The GNU General Public License version 2 as published by the
|
|
// Free Software Foundation, which can be found in the top level directory
|
|
// of this repository (LICENSE_GPL2), and also online at:
|
|
// <https://www.gnu.org/licenses/old-licenses/gpl-2.0.html>
|
|
//
|
|
// OR
|
|
//
|
|
// 2. An ADI specific BSD license, which can be found in the top level directory
|
|
// of this repository (LICENSE_ADIBSD), and also on-line at:
|
|
// https://github.com/analogdevicesinc/hdl/blob/master/LICENSE_ADIBSD
|
|
// This will allow to generate bit files and not release the source code,
|
|
// as long as it attaches to an ADI device.
|
|
//
|
|
// ***************************************************************************
|
|
// ***************************************************************************
|
|
`timescale 1ns/1ps
|
|
|
|
module util_axis_fifo #(
|
|
parameter DATA_WIDTH = 64,
|
|
parameter ADDRESS_WIDTH = 5,
|
|
parameter ASYNC_CLK = 1,
|
|
parameter M_AXIS_REGISTERED = 1,
|
|
parameter [ADDRESS_WIDTH-1:0] ALMOST_EMPTY_THRESHOLD = 16,
|
|
parameter [ADDRESS_WIDTH-1:0] ALMOST_FULL_THRESHOLD = 16
|
|
) (
|
|
input m_axis_aclk,
|
|
input m_axis_aresetn,
|
|
input m_axis_ready,
|
|
output m_axis_valid,
|
|
output [DATA_WIDTH-1:0] m_axis_data,
|
|
output m_axis_tlast,
|
|
output [ADDRESS_WIDTH-1:0] m_axis_level,
|
|
output m_axis_empty,
|
|
output m_axis_almost_empty,
|
|
|
|
input s_axis_aclk,
|
|
input s_axis_aresetn,
|
|
output s_axis_ready,
|
|
input s_axis_valid,
|
|
input [DATA_WIDTH-1:0] s_axis_data,
|
|
input s_axis_tlast,
|
|
output [ADDRESS_WIDTH-1:0] s_axis_room,
|
|
output s_axis_full,
|
|
output s_axis_almost_full
|
|
);
|
|
|
|
generate if (ADDRESS_WIDTH == 0) begin : zerodeep /* it's not a real FIFO, just a 1 stage pipeline */
|
|
|
|
if (ASYNC_CLK) begin
|
|
|
|
(* KEEP = "yes" *) reg [DATA_WIDTH-1:0] cdc_sync_fifo_ram;
|
|
reg axis_tlast_d;
|
|
reg s_axis_waddr = 1'b0;
|
|
reg m_axis_raddr = 1'b0;
|
|
|
|
wire m_axis_waddr;
|
|
wire s_axis_raddr;
|
|
|
|
sync_bits #(
|
|
.NUM_OF_BITS(1),
|
|
.ASYNC_CLK(ASYNC_CLK)
|
|
) i_waddr_sync (
|
|
.out_clk(m_axis_aclk),
|
|
.out_resetn(m_axis_aresetn),
|
|
.in_bits(s_axis_waddr),
|
|
.out_bits(m_axis_waddr)
|
|
);
|
|
|
|
sync_bits #(
|
|
.NUM_OF_BITS(1),
|
|
.ASYNC_CLK(ASYNC_CLK)
|
|
) i_raddr_sync (
|
|
.out_clk(s_axis_aclk),
|
|
.out_resetn(s_axis_aresetn),
|
|
.in_bits(m_axis_raddr),
|
|
.out_bits(s_axis_raddr)
|
|
);
|
|
|
|
assign m_axis_valid = m_axis_raddr != m_axis_waddr;
|
|
assign m_axis_level = ~m_axis_ready;
|
|
assign m_axis_empty = 0;
|
|
assign m_axis_almost_empty = 0;
|
|
assign s_axis_ready = s_axis_raddr == s_axis_waddr;
|
|
assign s_axis_full = 0;
|
|
assign s_axis_almost_full = 0;
|
|
assign s_axis_room = s_axis_ready;
|
|
|
|
always @(posedge s_axis_aclk) begin
|
|
if (s_axis_ready == 1'b1 && s_axis_valid == 1'b1)
|
|
cdc_sync_fifo_ram <= s_axis_data;
|
|
axis_tlast_d <= s_axis_tlast;
|
|
end
|
|
|
|
always @(posedge s_axis_aclk) begin
|
|
if (s_axis_aresetn == 1'b0) begin
|
|
s_axis_waddr <= 1'b0;
|
|
end else if (s_axis_ready & s_axis_valid) begin
|
|
s_axis_waddr <= s_axis_waddr + 1'b1;
|
|
end
|
|
end
|
|
|
|
always @(posedge m_axis_aclk) begin
|
|
if (m_axis_aresetn == 1'b0) begin
|
|
m_axis_raddr <= 1'b0;
|
|
end else begin
|
|
if (m_axis_valid & m_axis_ready)
|
|
m_axis_raddr <= m_axis_raddr + 1'b1;
|
|
end
|
|
end
|
|
|
|
assign m_axis_data = cdc_sync_fifo_ram;
|
|
assign m_axis_tlast = axis_tlast_d;
|
|
|
|
end else begin /* !ASYNC_CLK */
|
|
|
|
// Note: In this mode, the write and read interface must have a symmetric
|
|
// aspect ratio
|
|
reg [DATA_WIDTH-1:0] axis_data_d;
|
|
reg axis_valid_d;
|
|
reg axis_tlast_d;
|
|
|
|
always @(posedge s_axis_aclk) begin
|
|
if (!s_axis_aresetn) begin
|
|
axis_data_d <= {DATA_WIDTH{1'b0}};
|
|
axis_valid_d <= 1'b0;
|
|
axis_tlast_d <= 1'b0;
|
|
end else if (s_axis_ready) begin
|
|
axis_data_d <= s_axis_data;
|
|
axis_valid_d <= s_axis_valid;
|
|
axis_tlast_d <= s_axis_tlast;
|
|
end
|
|
end
|
|
|
|
assign m_axis_data = axis_data_d;
|
|
assign m_axis_valid = axis_valid_d;
|
|
assign m_axis_tlast = axis_tlast_d;
|
|
assign s_axis_ready = m_axis_ready | ~m_axis_valid;
|
|
assign m_axis_empty = 1'b0;
|
|
assign m_axis_almost_empty = 1'b0;
|
|
assign m_axis_level = 1'b0;
|
|
assign s_axis_full = 1'b0;
|
|
assign s_axis_almost_full = 1'b0;
|
|
assign s_axis_room = 1'b0;
|
|
|
|
end
|
|
|
|
end else begin : fifo /* ADDRESS_WIDTH != 0 - this is a real FIFO implementation */
|
|
|
|
wire [ADDRESS_WIDTH-1:0] s_axis_waddr;
|
|
wire [ADDRESS_WIDTH-1:0] m_axis_raddr;
|
|
wire _m_axis_ready;
|
|
wire _m_axis_valid;
|
|
|
|
wire s_mem_write;
|
|
wire m_mem_read;
|
|
|
|
reg valid = 1'b0;
|
|
|
|
/* Control for first falls through */
|
|
always @(posedge m_axis_aclk) begin
|
|
if (m_axis_aresetn == 1'b0) begin
|
|
valid <= 1'b0;
|
|
end else begin
|
|
if (_m_axis_valid)
|
|
valid <= 1'b1;
|
|
else if (m_axis_ready)
|
|
valid <= 1'b0;
|
|
end
|
|
end
|
|
|
|
assign s_mem_write = s_axis_ready & s_axis_valid;
|
|
assign m_mem_read = (~valid || m_axis_ready) && _m_axis_valid;
|
|
|
|
util_axis_fifo_address_generator #(
|
|
.ASYNC_CLK(ASYNC_CLK),
|
|
.ADDRESS_WIDTH(ADDRESS_WIDTH),
|
|
.ALMOST_EMPTY_THRESHOLD (ALMOST_EMPTY_THRESHOLD),
|
|
.ALMOST_FULL_THRESHOLD (ALMOST_FULL_THRESHOLD))
|
|
i_address_gray (
|
|
.m_axis_aclk(m_axis_aclk),
|
|
.m_axis_aresetn(m_axis_aresetn),
|
|
.m_axis_ready(_m_axis_ready),
|
|
.m_axis_valid(_m_axis_valid),
|
|
.m_axis_raddr(m_axis_raddr),
|
|
.m_axis_level(m_axis_level),
|
|
.m_axis_empty(m_axis_empty),
|
|
.m_axis_almost_empty(m_axis_almost_empty),
|
|
.s_axis_aclk(s_axis_aclk),
|
|
.s_axis_aresetn(s_axis_aresetn),
|
|
.s_axis_ready(s_axis_ready),
|
|
.s_axis_valid(s_axis_valid),
|
|
.s_axis_full(s_axis_full),
|
|
.s_axis_almost_full(s_axis_almost_full),
|
|
.s_axis_waddr(s_axis_waddr),
|
|
.s_axis_room(s_axis_room)
|
|
);
|
|
|
|
if (ASYNC_CLK == 1) begin : async_clocks /* Asynchronous WRITE/READ clocks */
|
|
|
|
// The assumption is that in this mode the M_AXIS_REGISTERED is 1
|
|
// When the clocks are asynchronous instantiate a block RAM
|
|
// regardless of the requested size to make sure we threat the
|
|
// clock crossing correctly
|
|
ad_mem #(
|
|
.DATA_WIDTH (DATA_WIDTH+1),
|
|
.ADDRESS_WIDTH (ADDRESS_WIDTH))
|
|
i_mem (
|
|
.clka(s_axis_aclk),
|
|
.wea(s_mem_write),
|
|
.addra(s_axis_waddr),
|
|
.dina({s_axis_data, s_axis_tlast}),
|
|
.clkb(m_axis_aclk),
|
|
.reb(m_mem_read),
|
|
.addrb(m_axis_raddr),
|
|
.doutb({m_axis_data, m_axis_tlast})
|
|
);
|
|
|
|
assign _m_axis_ready = ~valid || m_axis_ready;
|
|
assign m_axis_valid = valid;
|
|
|
|
end else begin : sync_clocks /* Synchronous WRITE/READ clocks */
|
|
|
|
reg [DATA_WIDTH:0] ram[0:2**ADDRESS_WIDTH-1];
|
|
|
|
// When the clocks are synchronous use behavioral modeling for the SDP RAM
|
|
// Let the synthesizer decide what to infer (distributed or block RAM)
|
|
always @(posedge s_axis_aclk) begin
|
|
if (s_mem_write)
|
|
ram[s_axis_waddr] <= {s_axis_data, s_axis_tlast};
|
|
end
|
|
|
|
if (M_AXIS_REGISTERED == 1) begin
|
|
|
|
reg [DATA_WIDTH:0] data;
|
|
|
|
always @(posedge m_axis_aclk) begin
|
|
if (m_mem_read)
|
|
data <= ram[m_axis_raddr];
|
|
end
|
|
|
|
assign _m_axis_ready = ~valid || m_axis_ready;
|
|
assign m_axis_data = data[DATA_WIDTH:1];
|
|
assign m_axis_tlast = data[0];
|
|
assign m_axis_valid = valid;
|
|
|
|
end else begin
|
|
|
|
assign _m_axis_ready = m_axis_ready;
|
|
assign m_axis_valid = _m_axis_valid;
|
|
assign {m_axis_data, m_axis_tlast} = ram[m_axis_raddr];
|
|
|
|
end
|
|
end
|
|
end /* fifo */
|
|
endgenerate
|
|
|
|
endmodule
|