Description:
I am trying to generate a test bench for a 5 state sequential state machine that detects 110 or any combination of (2) 1's and (1) 0. I already have written the code. see below. I am having trouble with the test bench which is wrong. I want to test for all possible sequences as well as input combinations that are off sequence.
Please give me examples of a good test bench to achieve what I need for a mealy machine.
vhdl code:
library IEEE;
use IEEE.STD_LOGIC_1164.all;
entity state is
port( clk, x : in std_logic;
z : out std_logic
);
end entity;
architecture behavioral of state is
type state_type is (s0,s1,s2,s3,s4);
signal state,next_s: state_type;
------------------------------------------------------------------------------
begin
process (state,x)
begin
if clk='1' and clk'event then
case state is
when s0 =>
if(x ='0') then
z <= '0';
next_s <= s4;
else
z <= '0';
next_s <= s1;
end if;
when s1 => --when current state is "s1"
if(x ='0') then
z <= '0';
next_s <= s3;
else
z <= '0';
next_s <= s2;
end if;
when s2 => --when current state is "s2"
if(x ='0') then
z <= '1';
next_s <= s0;
else
z <= '0';
next_s <= s0;
end if;
when s3 => --when current state is "s3"
if(x ='0') then
z <= '0';
next_s <= s0;
else
z <= '1';
next_s <= s0;
end if;
when s4 => --when current state is s4
if (x = '0') then
z <= '0';
next_s <= s0;
else
z <= '0';
next_s <= s3;
end if;
end case;
end if;
end process;
end behavioral;
Test Bench code:
library ieee;
use ieee.std_logic_1164.all;
-- Add your library and packages declaration here ...
entity state_tb is
end state_tb;
architecture TB_ARCHITECTURE of state_tb is
-- Component declaration of the tested unit
component state
port(
clk : in STD_LOGIC;
x : in STD_LOGIC;
z : out STD_LOGIC );
end component;
-- Stimulus signals - signals mapped to the input and inout ports of tested entity
signal clk : STD_LOGIC;
signal x : STD_LOGIC;
-- Observed signals - signals mapped to the output ports of tested entity
signal z : STD_LOGIC;
-- Add your code here ...
begin
-- Unit Under Test port map
UUT : state
port map (
clk => clk,
x => x,
z => z
);
-- CLOCK STIMULI
CLOCK: process
begin
CLK <= not clk after 20 ns;
wait for 40 ns;
end process;
-- X input STIMULI
X_Stimuli: process
begin
X <= not x after 40 ns;
wait for 80 ns;
end process;
end TB_ARCHITECTURE;
configuration TESTBENCH_FOR_state of state_tb is
for TB_ARCHITECTURE
for UUT : state
use entity work.state(behavioral);
end for;
end for;
end TESTBENCH_FOR_state;
These are some problems with both the FSM code and the testbench code in your example, but the main issue is that to test an FSM you need t apply a sequence of input values and check the outputs. You can't just toggle your input signal between 1 and 0. So, here's some advice:
First, you have to decide whether you want a generic FSM that detects any input sequence, or a
FSM that detects only a single sequence (your code shows the second option)
You need to consider the time dimension in your test. Yours is a clocked circuit, meaning that each test will take several clock cycles.
To test every possible input sequence, I suggest that you create a procedure that takes as arguments:
A sequence of 4 input values to the FSM (could be a std_logic_vector)
A sequence of 4 output values that you expect to see
(optionally) a sequence of the 4 states you expect the FSM will go through
Your procedure could look like:
procedure test_sequence(
input_sequence: std_logic_vector;
expected_output_sequence: std_logic_vector
) is begin
for i in input_sequence'range loop
x <= input_sequence(i);
wait until rising_edge(clk);
assert z = expected_output_sequence(i);
end loop;
end;
Then, in your main tests process, you can test one sequence with:
test_sequence(
input_sequence => "110",
expected_output_sequence => "001"
);
Some other suggestions:
You should add a reset signal to make testing easier and to prevent mismatchs between simulation and synthesis
There is no need for a configuration in your case, you can remove it from your code
Your FSM code is incomplete because you never update your current state
In a testbench like the one you are using, you need to initialize the signals used as input to the DUT (x and clk)
Note that the procedure above needs to be inside a process' declarative region. Something like:
main_test_process: process is
procedure test_sequence(
input_sequence: std_logic_vector;
expected_output_sequence: std_logic_vector
) is begin
for i in input_sequence'range loop
x <= input_sequence(i);
wait until rising_edge(clk);
assert z = expected_output_sequence(i);
end loop;
end;
begin
test_sequence( input_sequence => "000", expected_output_sequence => "000");
test_sequence( input_sequence => "001", expected_output_sequence => "000");
-- (add any other input sequences here...)
test_sequence( input_sequence => "110", expected_output_sequence => "001");
std.env.finish;
end process;
should work.
Your statemachine has the following possible cycles with 2 or 3 steps before going back to s0 and correctly detects sequencies of two 1s.
Case (x1,x2,x3) States (z1,z2,z3)
0 0,0,0 4,0,... 0,0,... (starts again at s0)
1 0,0,1 4,0,... 0,0,... (starts again at s0)
2 0,1,0 4,3,0 0,0,0 (covered by your TB)
3 0,1,1 4,3,0 0,0,1
4 1,0,0 1,3,0 0,0,0
5 1,0,1 1,3,0 0,0,1 (covered by your TB)
6 1,1,0 1,2,0 0,0,1
7 1,1,1 1,2,0 0,0,0
As I see it your creating you're stimuli as follows
__ __ __ __ __
clk __| |__| |__| |__| |__| |__...
_____ _____ _____
x _____| |_____| |_____| |...
I.e. because in each section with x=1 you have exactly one rising clock and therefore you're only testing sequencies with a 0101010... pattern, where your statemachine would go on one of the two paths marked in the table above. This means that the other 6 possible pathes are never executed in your testbench.
Since this statemachine has a small and finite number of paths I would recommend an exhaustive test where you would essentially cycle through the 8 possible cases listed above; this could be easily implemented with a 3bit counter. So you would create a sequence in the form
reset
test-case 0 (sequence 0,0,0)
reset
test-case 1 (and so on)
That would require you to add a reset to the entity state. Alternatively you could modify your statemachine to remain in s0 with a zero-input; then you could reset with a sequence of 0,0,0 at any time.
Related
I have to generate the vhdl code for the signal generator above as a finite state machine. What I cant manage to implement is this. How to generate
the clock values to output. To be more specific, I want for half period
output 0 and the other half 1, when the state machine is in state clock. My implementation, as you can see on the
code below, is output <= clock but this obviously does not work. I made a finite machine with 4 states (s1, s2, s3, s4)
My testbench and my vhdl code are these:
testbench
library ieee;
use ieee.std_logic_1164.all;
entity signal_generator_tb is
end entity;
architecture signal_generator_tb_arch of signal_generator_tb is
signal clock_tb, reset_tb: std_logic;
signal output_tb: std_logic;
component signal_generator
port(clock, reset: in std_logic;
output: out std_logic);
end component;
begin
dut : signal_generator port map(clock_tb, reset_tb, output_tb);
stim_reset : process
begin
reset_tb <= '0'; wait for 10 ns;
reset_tb <= '1'; wait;
end process;
stim_clock : process
begin
clock_tb <= '1'; wait for 10 ns;
clock_tb <= '0'; wait for 10 ns;
end process;
end architecture;
and my vhdl code is that:
library ieee;
use ieee.std_logic_1164.all;
entity signal_generator is
port (clock, reset: in std_logic;
output: out std_logic);
end entity;
architecture signal_generator_arch of signal_generator is
type state_type is (s0, s1, s2, s3); --This is the states of the
finite state machine and we can create signlas with this type
signal current_state, next_state: state_type; --We can only assign
w_open and w_closed because there are type of state_type
begin
STATE_MEMORY : process(clock, reset)
begin
if(reset = '0') then
current_state <= s0;
elsif(falling_edge(clock)) then
current_state <= next_state;
end if;
end process;
NEXT_STATE_LOGIC : process(current_state)
begin
case(current_state) is
when s0 => next_state <= s1;
when s1 => next_state <= s2;
when s2 => next_state <= s3;
when s3 => next_state <= s0;
when others => next_state <= s0;
end case;
end process;
OUTPUT_LOGIC : process(current_state)
begin
case(current_state) is
when s0 => output <= '1';
when s1 => if(rising_edge(clock)) then
output <= '1';
else
output <= '0';
end if;
when s2 => output <= '1';
when s3 => output <= '0';
when others => output <= '0';
end case;
end process;
end architecture;
wave form
Your state machine does not work because of your OUTPUT_LOGIC process. It should also be sensitive to clock and when in state s1 it should output clock itself:
OUTPUT_LOGIC : process(current_state, clock)
begin
case(current_state) is
when s0 => output <= '1';
when s1 => output <= clock;
when s2 => output <= '1';
when s3 => output <= '0';
when others => output <= '0';
end case;
end process;
Note that you could simplify all this by simply passing your clock through a AND gate (to force it to 0) and a OR gate (to force it to 1) and output the result. A two bits counter that you would initialize to "11" and decrement would provide the current state, encoded in a very convenient way to force the output to 0 and 1:
library ieee;
use ieee.std_logic_1164.all;
use ieee.numeric_std_unsigned.all;
entity signal_generator is
port(clock, reset: in std_ulogic;
output: out std_ulogic);
end entity signal_generator;
architecture arc of signal_generator is
signal cnt: std_ulogic_vector(1 downto 0); -- the state
begin
process(clock, reset)
begin
if reset = '0' then
cnt <= "11";
elsif falling_edge(clock) then
cnt <= cnt - 1;
end if;
end process;
output <= (clock and cnt(1)) or cnt(0);
end architecture arc;
Note: always prefer std_ulogic (unresolved) to std_logic (resolved) when you do not plan to have multiple drive situations on a signal. This way, if you accidentally create a multiple drive situation you will get a meaningful error message instead of spending hours trying to understand where all these X values in your waveforms come from.
I have the VHDL implementation that works on board, it detects the sequence 01110 and will raise a flag for 2 clock counts. It detects overlapping sequences as well where 011101110 would raise the flag twice.
I've checked my implementation with a logic analyzer on the board and am fairly confident that it works. I am feeding in a repetition sequence of 0111 at 10 kHz, on the board, it has a clock at 100 MHz where I scale it to 10 kHz with a prescaler.
My problem is, when trying to recreate a similar scenario using a simulation, I do not get any outputs as expected
Image from logic analyzer from board
Image from Test Bench
Test Bench Code
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use IEEE.STD_LOGIC_ARITH.ALL;
use IEEE.STD_LOGIC_UNSIGNED.ALL;
entity test_FSM_prac4 is
-- Port ( );
end test_FSM_prac4;
architecture Behavioral of test_FSM_prac4 is
component FSM_prac4 is
port (
inputSignal : in STD_LOGIC;
pushButton : in STD_LOGIC;
clk100mhz : in STD_LOGIC;
logic_analyzer : out STD_LOGIC_VECTOR (7 downto 0);
LEDs: out STD_LOGIC
); end component;
signal inputSignal : std_logic := '0';
signal pushButton: std_logic := '0';
signal clk100mhz: std_logic := '0';
signal logic_analyzer: std_logic_vector(7 downto 0);
signal LEDs : std_logic;
begin
uut : FSM_prac4 port map(
inputSignal => inputSignal,
pushButton => pushButton,
clk100mhz => clk100mhz,
logic_analyzer => logic_analyzer,
LEDs => LEDs
);
--generate clock 100mhz
clock_tic: process begin
loop
clk100mhz <= '0';
wait for 5ns;
clk100mhz <= '1';
wait for 5ns;
end loop;
end process;
input_changes: process begin
loop
inputSignal <= '0';
wait for 100us;
inputSignal <= '1';
wait for 100us;
inputSignal <= '1';
wait for 100us;
inputSignal <= '1';
wait for 100us;
end loop;
end process;
end Behavioral;
To show the mapping for logic Analyzer
logic_analyzer(0) <= masterReset;
logic_analyzer(1) <= newClock -- 10Khz Clock;
logic_analyzer(2) <= outputZ;
--FSM States
logic_analyzer(3) <= '1' when y = A ELSE '0';
logic_analyzer(4) <= '1' when y = B ELSE '0';
logic_analyzer(5) <= '1' when y = C ELSE '0';
logic_analyzer(6) <= '1' when y = D ELSE '0';
logic_analyzer(7) <= '1' when y = E ELSE '0';
If anyone could direct to what I am doing wrong on the test bench and how to replicate to get similar results as the first image as it shows that in simulation, it always stays at state A and the new clock is not toggling meaning that clk100mhz is somehow not connected but I can't figure out why.
Any help is greatly appreciated, thanks guys
edit:
I wrote a simple program to test my scalar clock
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use IEEE.STD_LOGIC_ARITH.ALL;
use IEEE.STD_LOGIC_UNSIGNED.ALL;
entity scaler_clk is
Port (
pushButton : in std_logic;
indicator : out std_logic;
clk100mhz : in STD_LOGIC;
clk10khz: out STD_LOGIC
);
end scaler_clk;
architecture Behavioral of scaler_clk is
signal clockScalers : std_logic_vector (12 downto 0):= (others => '0') ;
signal prescaler: std_logic_vector(12 downto 0) := "1001110001000";
signal newClock: std_logic := '0';
signal masterReset : std_logic;
begin
clk10khz <= newClock;
masterReset <= pushButton;
process (clk100mhz,masterReset) begin
if(masterReset <= '1') then <--- error occurs here
clockScalers <= "0000000000000";
newClock <= '0';
indicator <= '1';
elsif (clk100mhz'event and clk100mhz = '1')then
indicator <= '0';
clockScalers <= clockScalers + 1;
if(clockScalers > prescaler) then
newClock <= not newClock;
clockScalers <= (others => '0');
end if;
end if;
end process;
end Behavioral;
test bench code
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
entity test_scaler_clk is
-- Port ( );
end test_scaler_clk;
architecture Behavioral of test_scaler_clk is
component scaler_clk Port (
pushButton : in std_logic;
indicator : out std_logic;
--input clock
clk100mhz : in STD_LOGIC;
clk10khz: out STD_LOGIC
);end component;
signal clk100mhz: std_logic := '0';
signal clk10khz : std_logic;
signal pushButton: std_logic;
signal indicator : std_logic;
begin
uut: scaler_clk port map(
pushButton => pushButton,
indicator => indicator,
clk100mhz => clk100mhz,
clk10khz => clk10khz
);
pushButton <= '0';
clock_tic: process begin
loop
clk100mhz <= '0';
wait for 5ns;
clk100mhz <= '1';
wait for 5ns;
end loop;
end process;
end Behavioral;
Even though I set pushButton to '0', it is still triggering masterReset, anyone knows why, that's why the 10 kHz clock isn't working
There are several things that you could (should) improve in your code. As Brian already explained, in your Behavioral architecture of scaler_clk, you should have:
if(masterReset = '1') then
instead of:
if(masterReset <= '1') then
Now, let's start with the most likely cause of your initial problem: unbound components. Your test benches instantiate the design to validate as components. VHDL components are kind of prototypes of actual entities. Prototypes are enough to compile because the compiler can perform all necessary syntax and type checking. But they are not enough to simulate because the simulator also needs the implementation behind the prototype. Some tools have a default binding strategy for unbound components: if they find an entity with the same name and if it has only one architecture, they use that. Your simulator apparently does not use such strategy (at least not by default, there is maybe an option for that but it is disabled). Note that most simulators I know issue warnings when they find unbound components. You probably missed these warnings.
Anyway, your component instances are unbound (they have no associated entity/architecture) and the simulator considers them as black boxes. Their outputs are not driven, except by the initial values you declared (1).
How to fix this? Two options:
Use a configuration to specify which entity/architecture pair shall be used for each component instance:
for all: scaler_clk use entity work.scaler_clk(Behavioral);
Use entity instantiations instead of components:
uut: entity work.scaler_clk(Behavioral) port map...
Now, let's go through some other aspects of your code that could be improved:
You are using non-standard packages, that are frequently not even compatible: IEEE.STD_LOGIC_ARITH and IEEE.STD_LOGIC_UNSIGNED. As they are not standard they should not even be in the standard IEEE library. You should use IEEE.NUMERIC_STD instead, and only that one. It declares the SIGNED and UNSIGNED types (with the same declaration as STD_LOGIC_VECTOR) and overloads the arithmetic operators on them.
Your test benches generate the 100MHz clock with:
clock_tic: process begin
loop
clk100mhz <= '0';
wait for 5ns;
clk100mhz <= '1';
wait for 5ns;
end loop;
end process;
The infinite loop is useless: a process is already an infinite loop:
clock_tic: process
begin
clk100mhz <= '0';
wait for 5ns;
clk100mhz <= '1';
wait for 5ns;
end process clock_tic;
would do the same. Same remark for your input_changes process.
Your input_changes process uses wait for <duration> statements. This is not a good idea because you do not know when the inputSignal signal toggles, compared to the clock. Is it just before, just after or exactly at the same time as the rising edge of clk100mhz? And if it is exactly at the same time, what will happen? Of course, you can carefully chose the <durations> to avoid such ambiguities but it is error prone. You should use the wait for <duration> only in the clock generating process. Everywhere else, it is better to synchronize with the clock:
input_changes: process
begin
inputSignal <= '0';
for i in 1 to 10000 loop
wait until rising_edge(clk100mhz);
end loop;
inputSignal <= '1';
for i in 1 to 10000 loop
wait until rising_edge(clk100mhz);
end loop;
inputSignal <= '1';
for i in 1 to 10000 loop
wait until rising_edge(clk100mhz);
end loop;
inputSignal <= '1';
for i in 1 to 10000 loop
wait until rising_edge(clk100mhz);
end loop;
end process input_changes;
This guarantees that inputSignal changes just after the rising edge of the clock. And you could rewrite it in a bit more elegant way (and probably a bit easier to maintain):
input_changes: process
constant values: std_logic_vector(0 to 3) := "0111";
begin
for i in values'range loop
inputSignal <= values(i);
for i in 1 to 10000 loop
wait until rising_edge(clk100mhz);
end loop;
end loop;
end process input_changes;
You are using resolved types (STD_LOGIC and STD_LOGIC_VECTOR). These types allow multiple drive, that is, having a hardware wire (VHDL signal) that is driven by several devices (VHDL processes). Usually you do not want this. Usually you even want to avoid this like the plague because it can cause short-circuits. In most cases it is wiser to use non-resolved types (STD_ULOGIC and STD_ULOGIC_VECTOR) because the compiler and/or the simulator will raise errors if you accidentally create a short circuit in your design.
One last thing: if, as its name suggests, you intend to use the clk10khz signal as a real clock, you should reconsider this decision. It is a signal that you generate with your custom logic. Clocks have very specific electrical and timing constraints that cannot really be fulfilled by regular signals. Before using clk10khz as a clock you must deal with clock skew, clock buffering... Not impossible but tricky. If you did use it as a clock your synthesizer probably issued critical warnings that you also missed (have a look maybe at the timing report). Moreover, this is probably useless in your case: an enable signal generated from clk100mhz could probably be used instead, avoiding all these problems. Instead of:
process (clk100mhz,masterReset) begin
if(masterReset = '1') then
clockScalers <= "0000000000000";
newClock <= '0';
indicator <= '1';
elsif (clk100mhz'event and clk100mhz = '1')then
indicator <= '0';
clockScalers <= clockScalers + 1;
if(clockScalers > prescaler) then
newClock <= not newClock;
clockScalers <= (others => '0');
end if;
end if;
end process;
use:
signal tick10khz: std_ulogic;
...
process(clk100mhz, masterReset) begin
if masterReset = '1') then
clockScalers <= "0000000000000";
tick10khz <= '0';
elsif rising_edge(clk100mhz) then
clockScalers <= clockScalers + 1;
tick10khz <= '0'
if(clockScalers > prescaler) then
tick10khz <= '1';
clockScalers <= (others => '0');
end if;
end if;
end process;
And then, instead of:
process(clk10khz)
begin
if rising_edge(clk10khz) then
register <= register_input;
end if;
end process;
use:
process(clk100mhz)
begin
if rising_edge(clk100mhz) then
if tick10khz = '1' then
register <= register_input;
end if;
end if;
end process;
The result will be the same but with only one single 100MHz clock, which avoids clock skew, clock buffering and clock domain crossing problems.
(1) This illustrates why declaring variables and signals with initial values is usually not a good idea: it hides potential problems. Without this your signals would have been stuck at 'U' (uninitialized) and it would maybe have helped understanding where the problem comes from.
Have JUST started learning how to use this tool so if my question seems silly i apologize in advance. I have searched the error in numerous forums (already answered posts , not mine) and couldn't understand what i was doing wrong so here is my question:
My Behavioral Code:
----------------------------------------------------------------------------- -----
-- Company:
-- Engineer:
--
-- Create Date: 01:47:22 07/07/2015
-- Design Name:
-- Module Name: Module_1 - Behavioral
-- Project Name:
-- Target Devices:
-- Tool versions:
-- Description:
--
-- Dependencies:
--
-- Revision:
-- Revision 0.01 - File Created
-- Additional Comments:
--
----------------------------------------------------------------------------- -----
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use IEEE.STD_LOGIC_ARITH.ALL;
use IEEE.STD_LOGIC_UNSIGNED.ALL;
-- Uncomment the following library declaration if using
-- arithmetic functions with Signed or Unsigned valuessss
--use IEEE.NUMERIC_STD.ALL;
-- Uncomment the following library declaration if instantiating
-- any Xilinx primitives in this code.
--library UNISIM;
--use UNISIM.VComponents.all;
entity Module_1 is
port (A,B,WE,reset : in std_logic;
clk : in std_logic;
DIN : in signed(3 downto 0);
FULL,EMPTY,ERROR : out std_logic:= '0';
PARKFREE : out signed(3 downto 0)
);
end Module_1;
architecture Behavioral of Module_1 is
signal current_state,next_state:std_ulogic_vector(1 downto 0);
signal empty_bf, full_bf :std_ulogic;
signal enter, reset_b : std_ulogic := '0' ;
constant s0: std_ulogic_vector (1 downto 0):="00";
constant s1: std_ulogic_vector (1 downto 0):="10";
constant s2: std_ulogic_vector (1 downto 0):="11";
constant s3: std_ulogic_vector (1 downto 0):="01";
signal park_counter,buffr: signed(3 downto 0):="0000";
signal PARKTOTAL,free_park_counter: signed(3 downto 0):= "1111";
begin
p1: process (clk,reset,reset_b)
begin
if (reset = '1') then
current_state <= s0;
elsif clk'event and clk = '1' then
current_state <= next_state;
end if;
end process p1;
p2: process (current_state,A,B)
begin
next_state <= current_state;
case current_state is
when s0 =>
if A = '1' then
enter <= '1';
next_state <= s1;
elsif B = '1' then
next_state <= s3;
end if;
when s1 =>
if A = '0' then
enter <= '0';
next_state <= s0;
elsif B = '1' then
next_state <= s2;
end if;
when s2 =>
if A = '0' then
next_state <= s3;
elsif B = '0' then
next_state <= s1;
end if;
when s3 =>
if B = '0' then
enter <= '0';
next_state <= s0;
elsif A = '1' then
next_state <= s2;
end if;
when others =>
end case;
end process p2;
p3: process(current_state,A,B)
begin
case current_state is
when s1 =>
if enter = '0' and A = '0' and empty_bf = '0' then
park_counter <= park_counter - 1;
free_park_counter <= free_park_counter + 1;
ERROR <= '0';
end if;
when s3 =>
if enter = '1' and B = '0' and full_bf = '0' then
park_counter <= park_counter + 1;
free_park_counter <= free_park_counter - 1;
ERROR <= '0';
end if;
when others =>
end case;
end process p3;
max: process(WE)
begin
if clk'event and clk = '1' and WE = '1' then
PARKTOTAL <= DIN ;
buffr <= DIN ;
if (free_park_counter < buffr - park_counter) then
ERROR <= '1';
reset_b <= '1';
else free_park_counter <= buffr - park_counter;
end if;
end if;
end process max;
incr: process(free_park_counter,DIN)
begin
PARKFREE <= free_park_counter;
if (free_park_counter = 15) then
EMPTY <= '1';
empty_bf <= '1';
else EMPTY <= '0';
empty_bf <= '0';
end if;
if (free_park_counter = 0) then
FULL <= '1';
full_bf <= '1';
else FULL <= '0';
full_bf <= '0';
end if;
end process incr;
end Behavioral;
My Testbench
----------------------------------------------------------------------------- ---
-- Company:
-- Engineer:
--
-- Create Date: 02:17:07 07/11/2015
-- Design Name:
-- Module Name: D:/Users/ErgasiaFPGA/Testbench.vhd
-- Project Name: ErgasiaFPGA
-- Target Device:
-- Tool versions:
-- Description:
--
-- VHDL Test Bench Created by ISE for module: Module_1
--
-- Dependencies:
--
-- Revision:
-- Revision 0.01 - File Created
-- Additional Comments:
--
-- Notes:
-- This testbench has been automatically generated using types std_logic and
-- std_logic_vector for the ports of the unit under test. Xilinx recommends
-- that these types always be used for the top-level I/O of a design in order
-- to guarantee that the testbench will bind correctly to the post-implementation
-- simulation model.
--------------------------------------------------------------------------------
LIBRARY ieee;
USE ieee.std_logic_1164.ALL;
use IEEE.STD_LOGIC_ARITH.ALL;
use IEEE.STD_LOGIC_UNSIGNED.ALL;
-- Uncomment the following library declaration if using
-- arithmetic functions with Signed or Unsigned values
--USE ieee.numeric_std.ALL;
ENTITY Testbench IS
END Testbench;
ARCHITECTURE behavior OF Testbench IS
-- Component Declaration for the Unit Under Test (UUT)
COMPONENT Module_1
PORT(
A : IN std_logic;
B : IN std_logic;
WE : IN std_logic;
reset : IN std_logic;
clk : IN std_logic;
DIN : IN signed(3 downto 0);
FULL : OUT std_logic;
EMPTY : OUT std_logic;
ERROR : OUT std_logic;
PARKFREE : OUT signed(3 downto 0)
);
END COMPONENT;
--Inputs
signal A : std_logic := '0';
signal B : std_logic := '0';
signal WE : std_logic := '0';
signal reset : std_logic := '0';
signal clk : std_logic := '0';
signal DIN : signed(3 downto 0) := (others => '0');
--Outputs
signal FULL : std_logic;
signal EMPTY : std_logic;
signal ERROR : std_logic;
signal PARKFREE : signed(3 downto 0);
-- Clock period definitions
constant clk_period : time := 10 ns;
BEGIN
-- Instantiate the Unit Under Test (UUT)
uut: Module_1 PORT MAP (
A => A,
B => B,
WE => WE,
reset => reset,
clk => clk,
DIN => DIN,
FULL => FULL,
EMPTY => EMPTY,
ERROR => ERROR,
PARKFREE => PARKFREE
);
-- Clock process definitions
clk_process :process
begin
clk <= '0';
wait for clk_period/2;
clk <= '1';
wait for clk_period/2;
end process;
-- Stimulus process
stim_proc: process
begin
-- hold reset state for 100 ns.
reset <= '1' ;
wait for 100 ns;
reset <= '0' ;
wait for clk_period*10;
-- insert stimulus here
A <= '1' ;
wait for clk_period*5;
B <= '1' ;
wait for clk_period*5;
A <= '0' ;
wait for clk_period*5;
B <= '0' ;
wait for clk_period*5;
B <= '1' ;
wait for clk_period*5;
A <= '1' ;
wait for clk_period*5;
B <= '0' ;
wait for clk_period*5;
A <= '0' ;
wait;
end process;
END;
I posted the whole code just in case i'm missing something in some part of it that i wouldn't think about. So , when i ISim it , with any "succesful" trigger of p3...
Referencing it again here:
p3: process(current_state,A,B)
begin
case current_state is
when s1 =>
if enter = '0' and A = '0' and empty_bf = '0' then
park_counter <= park_counter - 1;
free_park_counter <= free_park_counter + 1;
ERROR <= '0';
end if;
when s3 =>
if enter = '1' and B = '0' and full_bf = '0' then
park_counter <= park_counter + 1;
free_park_counter <= free_park_counter - 1;
ERROR <= '0';
end if;
when others =>
end case;
end process p3;
...the ISim says that in this part
"There is an 'U'|'X'|'W'|'Z'|'-' in an arithmetic operand, the result will be 'X'(es)."
and proceeds to make Xs of some of the values after that part , although all of the signals have been initialized (at least the ones in this part)
The "park_counter <= park_counter + 1;" part works correctly in the simulation but the "free_park_counter <= free_park_counter -1;" doesn't. This completely baffles me as they are declared as the same type and are both initialized the same way , even with different values.
So what am i missing or even doing blatantly wrong? Any help will be incredibly appreciated. Only looking for the error , if you could please contain optimizations since i'm looking to learn through trial and error and thought and would like to struggle to make it better myself
In addition , please be patient with my responses since i log on 2 to 3 times per day. Thanks in advance
Your design is non-workable per Brian's answer. Your testbench causes the messages when going from s3 or s1 to s0 before the clock edge. free_park_counter goes to 'U's. (Once it gets U's it won't loop further, no events occur without a signal value change).
Your counters should be clocked to prevent combinatorial looping, plus they likely won't synthesize a clock usefully due to uneven combinatorial delays. Sensitivity lists should likewise be complete, if for no other reason than the intent is to make simulation match the synthesized result.
Looking at the result of your testbench:
(clickable)
We can compare that with the messages from the arithmetic operators found in the Synopsys package std_logic_arith:
../../../src/synopsys/std_logic_arith.vhdl:315:20:#350ns:(assertion warning): There is an 'U'|'X'|'W'|'Z'|'-' in an arithmetic operand, the result will be 'X'(es).
../../../src/synopsys/std_logic_arith.vhdl:315:20:#350ns:(assertion warning): There is an 'U'|'X'|'W'|'Z'|'-' in an arithmetic operand, the result will be 'X'(es).
../../../src/synopsys/std_logic_arith.vhdl:315:20:#550ns:(assertion warning): There is an 'U'|'X'|'W'|'Z'|'-' in an arithmetic operand, the result will be 'X'(es).
The signals displayed in the waveform are selected in order of importance and appearance a first pass selection and we immediately see we also get 'U's on free_park_counter as well as ERROR.
ERROR catches attention because you hadn't mentioned it previously. When asking 'where to the 'U' come from ?' it becomes apparent the issue is there are drivers on ERROR and free_park_counter in both processes p3 and max. The messages are a side effect.
Each process assigning a signal provides a driver. Signals with multiple drivers are either resolved or result in an error for non-resolved types.
The resolved value of free_park_counter with one or more elements having a metavalue will cause the diagnostic messages produced by package std_logic_arith. The 'U's in the waveform are caused by the resolution of the two drivers.
The difficulty your audience had in noticing the two drivers may be in part due to your strong insistence on focusing on process p3, which is not well specified. The title and focus of your question also seems a bit unclear. Without a Minimal Complete and Verifiable example there was also bound to be less scrutiny.
You might expect as a minimum to consolidate all the assignments to ERROR and free_park_counter into a single process. ERROR should likely be registered, and I'd expect something named park_counter would likely want to be registered, too.
There is some confusion in the question title : declaring a signal and setting its value are entirely separate.
Initialising a signal (in the declaration) will influence its value, but not fully determine it. If the initialisation and another driving value are different, the result probably will be 'X'. Likewise if the signal is driven from different processes which disagree on its value.
Now, you are using a multiple-process form of state machine, where the operations are split between clocked and combinational processes. These are recommended by more than one textbook. This is unfortunate because they are notoriously difficult to get right, and for example, a moment's inspection will show that the sensitivity list on process P3 is wrong.
Fixing P3's sensitivity list may not affect the problem, because P3 also drives its own inputs in what is known as a combinational loop. Consider that, if the process wakes up several times because of glitches on the combinational inputs in its sensitivity list, the additions will take place several times...
Rewriting these three processes in the form of a single clocked process P1, (which is, unfortunately, not well taught in several textbooks) will avoid all of these difficulties.
In ISim, if you browse the tree menu on the left you are able to add to then signals window any internal signal you want. Add all of them, rerun the simulation and look for the signals that have 'U'|'X'|'W'|'Z'|'-' values. This should give us a lead to track down the problem.
If you are really new to VHDL, this answer of mine should help you undersand some of the basic concepts of this description language :)
VHDL - iSIM output uninitialised, doesn't change states
Another advice that I learned the hard way, but you can think about it after we solved this problem: textbooks and even Xilinx describe how to implement finite state machines with two or even three distinct processes. This comes from an educational approach where FSM are splitted in synchronous logic and asynchronous logic. In practice, this is doing more harm than good: most of the FSM can be described with a single synchronous process. Google it (or if you are interested we can talk about it) and try it, you will get the hang of it really quickly and it will really simplify the code (you won't even need two separate signals for the states anymore!).
i am a beginner in VHDL. i want ot know why there is a delay of one cycle in the following code.and how to avoid it..at the same time in verilog the statement always #(posedge clk) dont have any delay.. how to do the same in VHDL
library IEEE;
use IEEE.std_logic_1164.all;
-- entity
entity t_ff_s is
port ( T,S,CLK : in std_logic;
Q : out std_logic);
end t_ff_s;
-- entity
architecture my_t_ff_s of t_ff_s is
signal t_tmp : std_logic; -- intermediate signal declaration
begin
tff: process (S,rising_edge(clk))
begin
if (S = '0') then
t_tmp <= '1';
--elsif (rising_edge(CLK)) then
else
t_tmp <= T XOR t_tmp; -- temp output assignment
end if;
end process tff;
Q <= t_tmp; -- final output assignment
end my_t_ff_s;
Sensitivity lists in VHDL don't take an edge specification like in Verilog. VHDL is more flexible in that you can freely use the 'event signal attribute anywhere within a process to implement edge triggered behavior. You can mix level and edge sensitive logic without resorting to split blocks/processes or hacks like negedge for resets. Function calls like rising_edge(clk) (which implements a test for clk'event) are not permitted in a sensitivity list. It only contains signal names. Your code won't compile as is.
If some other syntactically correct version of your code compiles cleanly, the delays you see are artifacts of the simulation model or having a broken sensitivity list. If you want a synchronous clock driven process then you only need the clock signal and possibly an asynchronous reset in the sensitivity list.
Consider the following process:
tff: process(S, clk)
begin
if S = '0' then -- Asynchronous reset (level sensitive)
t_tmp <= '1';
elsif rising_edge(clk) then -- Synchronous logic (edge sensitive)
t_tmp <= T xor t_tmp;
end if;
end process;
Q <= t_tmp;
This process executes when an event occurs on S or clk. If S is '0' then the reset condition is executed with priority over the elsif clause (clk is a don't-care). The assignment to t_tmp takes effect on the next delta cycle which is still the same as the current simulation time. Otherwise, if rising_edge(clk) evaluates to true then an event occurred on clk and it's state changed from '0' (or 'L') to '1' (or 'H') indicating that the event was a rising edge. The synchronous assignment takes place and the new xored t_tmp takes effect on the next delta cycle. Changes in T don't cause the process to execute since it isn't (and shouldn't be) in the sensitivity list.
Because there is no unconditional else clause the t_tmp signal retains its last assigned value if both of the two if conditions are false. It will change the next time there is an event on S or clk that causes a new assignment to t_tmp. This will either be the next clock edge or a re-application of asynchronous reset.
The assignment to Q is continuous and is effectively the same as a process with t_tmp in its sensitivity list. As a consequence, the assignment to Q takes place a delta cycle after events on t_tmp which is two delta cycles after the rising edge. If Q is feeding into logic that updates earlier than the second delta cycle of an edge, it will appear to take an extra clock cycle for it to propagate.
The behavior surrounding delta cycles can sometimes create confusing results when inspecting waveforms. You may have a rising edge that should capture a data input that appears to transition simultaneously on the same time step when, in fact, the data is transitioning on a later delta cycle and will only be captured on the next clock edge.
Similarly, if you construct a simple gated clock without any time delay, its edges will occur at the same time but on later delta cycles than the ungated version of the clock. Data driven from the "earlier" ungated clock will be captured by the gated logic a clock cycle earlier than expected as a result. Data driven the other direction will appear to have an unexpected delay by a clock cycle.
It isn't clear what is causing the problem you see without more information on how you're driving the S, T, and clk signals but it is likely connected to the delta cycle behavior of the simulation engine in some way.
The problem
A little more succinctly than Kevin, rising_edge is an expression and not a signal, a sensitivity list requires a named signal, a transaction on which you resume execution of a suspended process. Put the elsif back in and have only S and clk in the sensitivity list.
Note that because t_tmp isn't in the sensitivity list, you won't see Q follow t_tmp until the next clock event causing the delay you noted.
The fixed syntax process:
tff: process (S,clk) -- was (S, risingedge(CLK)), a syntax error)
begin
if (S = '0') then
t_tmp <= '1';
elsif (rising_edge(CLK)) then -- put back
-- else
t_tmp <= T XOR t_tmp; -- temp output assignment
end if;
Q <= t_tmp; -- final output assignment
end process tff;
Which shows the delay between t_tmp and Q:
(clickable)
Fix it by making Q a concurrent signal assignment
To cure the half clock delay you could make the assignment to Q a concurrent signal assignment statement (move it outside of the process).
tff:
process (S, clk)
begin
if S = '0' then
t_tmp <= '1';
elsif rising_edge(clk) then
t_tmp <= T xor t_tmp;
end if;
end process;
Q <= t_tmp; -- concurrent signal assignment
Which gives:
(clickable)
And you can see above that t_tmp and Q are now in phase.
Fix it by making t_tmp a variable
You could also declare t_tmp as a variable in process dff instead of a signal and switching assignments to it as variable assignments will also cure the one clock delay between t_tmp and Q.
tff:
process (S, clk)
variable t_tmp: std_logic;
begin
if S = '0' then
t_tmp := '1';
elsif rising_edge(clk) then
t_tmp := T xor t_tmp;
end if;
Q <= t_tmp;
end process;
Which shows:
(clickable)
And ghdl using gtkwave doesn't output variables or show delta cycles. You can see Q occurs on the rising edge of the clock.
Making t_tmp a variable also has the effect of eliminating a delta cycle between a transaction on t_tmp and a transaction on Q.
Eliminating delta cycles makes your model execute faster (while occurring at the current simulation time). Signal assignments don't take effect while any process is executing and variable assignments take effect immediately.
Fix it by adding t_tmp to the sensitivity list
And alternatively you could just add t_tmp to the sensitivity list (along with S and clk).
tff:
process (S, clk, t_tmp)
begin
if S = '0' then
t_tmp <= '1';
elsif rising_edge(clk) then
t_tmp <= T xor t_tmp;
end if;
Q <= t_tmp;
end process;
(clickable)
And this is slower than all the rest of the fixes because the if statement is executed each time t_tmp has an event as well as S or CLK. rising_edge is a function call which dynamically elaborates it's interface list, a significant simulator performance penalty particularly if you use a lot of these primitives.
These were done with a test bench:
library IEEE;
use IEEE.std_logic_1164.all;
-- entity
entity t_ff_s is
port ( T,S,CLK : in std_logic;
Q : out std_logic);
end entity t_ff_s;
architecture my_t_ff_s of t_ff_s is
signal t_tmp : std_logic; -- intermediate signal declaration
begin
tff: process (S,clk) -- was (S, risingedge(CLK)), a syntax error)
begin
if (S = '0') then
t_tmp <= '1';
elsif (rising_edge(CLK)) then -- put back
-- else
t_tmp <= T XOR t_tmp; -- temp output assignment
end if;
Q <= t_tmp; -- final output assignment
end process tff;
end my_t_ff_s;
architecture foe of t_ff_s is
signal t_tmp: std_logic;
begin
tff:
process (S, clk)
begin
if S = '0' then
t_tmp <= '1';
elsif rising_edge(clk) then
t_tmp <= T xor t_tmp;
end if;
end process;
Q <= t_tmp; -- concurrent signal assignment
end architecture;
architecture fie of t_ff_s is
begin
tff:
process (S, clk)
variable t_tmp: std_logic;
begin
if S = '0' then
t_tmp := '1';
elsif rising_edge(clk) then
t_tmp := T xor t_tmp;
end if;
Q <= t_tmp;
end process;
end architecture;
architecture fee of t_ff_s is
signal t_tmp: std_logic;
begin
tff:
process (S, clk, t_tmp)
begin
if S = '0' then
t_tmp <= '1';
elsif rising_edge(clk) then
t_tmp <= T xor t_tmp;
end if;
Q <= t_tmp;
end process;
end architecture;
library ieee;
use ieee.std_logic_1164.all;
entity test_tff is
end entity;
architecture foo of test_tff is
signal CLK: std_logic := '0';
signal T: std_logic := '0';
signal S: std_logic := '0';
signal Q: std_logic;
component t_ff_s is
port (
signal CLK: in std_logic;
signal T: in std_logic;
signal S: in std_logic;
signal Q: out std_logic
);
end component;
begin
DUT:
t_ff_s
port map (
T => T,
S => S,
CLK => CLK,
Q => Q
);
CLOCK:
process
begin
wait for 10 ns;
CLK <= not CLK;
if Now > 250 ns then
wait;
end if;
end process;
SET:
process
begin
S <= '0';
wait for 20 ns;
S <= '1';
wait;
end process;
TOGGLE:
process
begin
wait for 20 ns;
T <= '1';
wait for 60 ns;
T <= '0';
wait for 40 ns;
T <= '1';
wait;
end process;
end architecture;
configuration my_t_ff_s_config of test_tff is
for foo
for DUT: t_ff_s
use entity work.t_ff_s(my_t_ff_s);
end for;
end for;
end configuration;
configuration concurrent_config of test_tff is
for foo
for DUT: t_ff_s
use entity work.t_ff_s(foe);
end for;
end for;
end configuration;
configuration variable_config of test_tff is
for foo
for DUT: t_ff_s
use entity work.t_ff_s(fie);
end for;
end for;
end configuration;
configuration sensitivity_config of test_tff is
for foo
for DUT: t_ff_s
use entity work.t_ff_s(fee);
end for;
end for;
end configuration;
note the use of configuration
Using VHDL's configuration declarations to allow the use of multiple architectures. (my_t_ff_s - the original, foe - with concurrent assignment to Q, fie - with t_tmp as a variable and fee - with t_tmp in the sensitivity list).
And amazingly enough ghdl's analyzer was quite helpful getting the configuration declarations syntax right. Once you get the first one, the others are easy.
We tend to get rusty using configuration, it wasn't generally supported historically by synthesis tools. But then again, this is simulation for verification.
And for those with ghdl and gtkwave this is how it was done:
ghdl -a t_ff.vhdl
ghdl -e my_t_ff_s_config
ghdl -e concurrent_config
ghdl -e concurrent_config
ghdl -e sensitivity_config
ghdl -r my_t_ff_s_config --wave=test_tff_my_t_ff_s.ghw
ghdl -r concurrent_config --wave=test_tff_foe.ghw
ghdl -r variable_config --wave=test_tff_fie.ghw
ghdl -r sensitivity_config --wave=test_tff_fee.ghw
GHW is ghdl's native waveform dump file format, understood by gtkwave.
In gtkwave:
open t_ff_s.gtkw (reads in test_tff_my_t_ff_s.ghw)
(otherwise read in test_tff_my_t_ff_s.ghw and add signals to
waveform display, format the window, save save file to t_ff_s.gtkw)
new tab open test_tff_foe.ghw
read save file open t_ff_s.gtkw
new tab open test_tff_fie.ghw
read save file open t_ff_s.gtkw
new tab open test_tff_fee.ghw
read save file open t_ff_s.gtkw
Note ghdl doesn't save variable state or delta cycles, t_tmp won't show up in the waveform for test_ff_fie.ghw.
I'm trying to implement a finite state machine identifier with xilinix 10.1
I've seen those errors in previous questions but the answers didn't include my question.. I'm not searching for an answer but rather a meaning for the FFd1 part
The following error is generated
WARNING:Xst:1293 - FF/Latch <machine1/current_state_FFd1> has a constant value of 0 in block <Main>. This FF/Latch will be trimmed during the optimization process.
WARNING:Xst:1293 - FF/Latch <machine1/current_state_FFd2> has a constant value of 0 in block <Main>. This FF/Latch will be trimmed during the optimization process.
this is my code
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use IEEE.STD_LOGIC_ARITH.ALL;
use IEEE.STD_LOGIC_UNSIGNED.ALL;
entity M_1 is
Port ( x : in STD_LOGIC;
clk : in STD_LOGIC;
state : out integer range 0 to 5 := 0;
z : out STD_LOGIC );
end M_1;
architecture Behavioral of M_1 is
type state_type is (A, B, C, D);
signal next_state, current_state: state_type := A;
begin
process(clk) is
begin
if (clk = '1' and clk'event) then
current_state <= next_state;
end if;
end process;
process(x,current_state)
begin
case current_state is
when A =>
if(x='0') then
next_state <= B;
z <='0';
elsif(x='1') then
next_state <= C;
z <='1';
end if;
when B =>
if(x='0') then
next_state <= C;
z <='1';
elsif(x='1') then
next_state <= D;
z <='0';
end if;
when C =>
if(x='0') then
next_state <= A;
z <='0';
elsif(x='1') then
next_state <= D;
z <='1';
end if;
when D =>
if(x='0') then
next_state <= B;
z <='0';
elsif(x='1') then
next_state <= C;
z <='0';
end if;
end case;
end process;
process (current_State) is
begin
case current_state is
when A =>
state <=0;
when B =>
state <=1;
when C =>
state <=2;
when D =>
state <=3;
end case;
end process;
end Behavioral;
can anyone tell me
What does current_state_FFd1 and what's the difference between it and current_State_1 ?
how can i solve this error ?
Thanks inadvance
The "current_state" signal is mapped onto a 2-bit flip-flop primitive by the CAD tools. The flip-flop will look something like the FD16CE primitive, shown here.
The flip-flop will take 2 data inputs (current_state_FFd1 and current_state_FFd2) and a clock, and produce two data outputs (current_state_FFq1 and current_state_FFq2). The inputs determine the value of the current_state signal sampled at the next clock edge, and the outputs reflect the current state.
The message you're seeing suggests that the CAD tools can prove that "current_state" never changes from the "00" encoding ("A" in your enumerated type), and so the flip-flop can be optimized away with a hard-wired output of "00".
The VHDL you posted looks reasonable -- changes on the 'x' input should cause a change in current_state. I'll bet the 'x' input is somehow hard-wired to 0 in higher-level VHDL (or in your testbench.)