My waveform does not change:
I am working on my 32-bit comparator project. I already have an 1 bit one. I do not know where is the issue. Anyone can help me find that?
Thanks so much
Code:
1bit:
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
ENTITY comp1 is
port (a : IN std_logic ;
b : IN std_logic ;
g : IN std_logic ;
l : IN std_logic ;
e : IN std_logic ;
great : OUT std_logic ;
less : OUT std_logic ;
equal : OUT std_logic );
END ;
ARCHITECTURE comp1_arch OF comp1 IS
signal s1,s2,s3: std_logic;
begin
s1 <= (a and (not b));
s2 <= (not ((a and (not b)) or (b and (not a))));
s3 <= (b and (not a));
equal <= (e and s2) after 30 ns;
great <= (g or(e and s1)) after 27 ns;
less <= (l or(e and s3)) after 27 ns;
end comp1_arch;
32 bit:
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
ENTITY comp32 is
GENERIC (BW : INTEGER :=32);
PORT ( a_32 : IN STD_LOGIC_VECTOR (BW -1 DOWNTO 0);
b_32 : IN STD_LOGIC_VECTOR (BW -1 DOWNTO 0);
g_32 : OUT STD_LOGIC ;
l_32 : OUT STD_LOGIC ;
e_32 : OUT STD_LOGIC );
END comp32;
ARCHITECTURE comp32_arch OF comp32 IS
COMPONENT comp1
PORT (a,b,g,l,e : IN std_logic ;
great,less,equal : OUT std_logic);
END COMPONENT comp1;
signal gre : std_logic_vector(BW downto 0);
signal les : std_logic_vector(BW downto 0);
signal equ : std_logic_vector(BW downto 0);
begin
gre(0)<='0';les(0)<='0';equ(0)<='0';
gen: for i in 0 to BW-1 generate
biti: comp1 port map( a => a_32(i),b => b_32(i), g => gre(i), l => les(i), e =>equ(i),
great => gre(i+1), less => les(i+1), equal => equ(i+1));
end generate;
g_32 <= gre(BW-1);
l_32 <= les(BW-1);
e_32 <= equ(BW-1);
end comp32_arch;
Test Bench:
LIBRARY ieee;
USE ieee.std_logic_1164.ALL;
ENTITY comp32_TB IS
END comp32_TB;
ARCHITECTURE behavior OF comp32_TB IS
COMPONENT comp32
PORT(
a_32 : IN std_logic_vector(31 downto 0);
b_32 : IN std_logic_vector(31 downto 0);
g_32 : OUT std_logic;
l_32 : OUT std_logic;
e_32 : OUT std_logic
);
END COMPONENT;
signal a_32 : std_logic_vector(31 downto 0) := (others => '0');
signal b_32 : std_logic_vector(31 downto 0) := (others => '0');
signal g_32 : std_logic;
signal l_32 : std_logic;
signal e_32 : std_logic;
BEGIN
uut: comp32 PORT MAP (
a_32 => a_32,
b_32 => b_32,
g_32 => g_32,
l_32 => l_32,
e_32 => e_32
);
stim_proc: process
begin
a_32 <="00000000000000000000000000000000";b_32<="00000000000000000000000000000000";wait for 1500 ns;
a_32 <="00000000000000000000000000000001";b_32<="00000000000000000000000000000000";wait for 1500 ns;
a_32 <="00000000000000000000000000000000";b_32<="10000000000000000000000000000000";wait for 1500 ns;
wait;
end process;
END;
You had your chained signals backward, and the first inputs want to show equal:
architecture comp32_arch of comp32 is
component comp1
port (a,b,g,l,e : in std_logic ;
great,less,equal : out std_logic);
end component comp1;
signal gre : std_logic_vector(BW downto 0);
signal les : std_logic_vector(BW downto 0);
signal equ : std_logic_vector(BW downto 0);
begin
gre(BW) <= '0'; -- gre(0) <= '0';
les(BW) <= '0'; -- les(0) <= '0';
equ(BW) <= '1'; -- equ(0) <= '0';
gen:
for i in 0 to BW-1 generate
biti:
comp1
port map (
a => a_32(i),
b => b_32(i),
g => gre(i+1), -- gre(i),
l => les(i+1), -- les(i),
e => equ(i+1), -- equ(i),
great => gre(i), -- gre(i+1),
less => les(i), -- les(i+1),
equal => equ(i) -- equ(i+1)
);
end generate;
g_32 <= gre(0); -- gre(BW);-- (BW-1);
l_32 <= les(0); -- les(BW); -- (BW-1);
e_32 <= equ(0); -- equ(BW); -- (BW-1);
end architecture comp32_arch;
And that gives:
The most significant bit without an equals defines either less than or greater than. If they're all equal that propagates all the way through.
Related
I have this scheme
I have to write structural VHDL design for it.
So these are my components:
MUX:
library IEEE;
use IEEE.STD_LOGIC_1164.all;
entity mux is
port(
A : in STD_LOGIC_VECTOR(7 downto 0);
B : in STD_LOGIC_VECTOR(7 downto 0);
Sel : in bit;
Z : out STD_LOGIC_VECTOR(7 downto 0)
);
end mux;
architecture Beh of mux is
begin
Z <= A when Sel='1'else
B;
end Beh;
REG:
library IEEE;
use IEEE.STD_LOGIC_1164.all;
entity reg is
port(
C : STD_LOGIC;
LD : in bit;
Reg_in : in STD_LOGIC_VECTOR(7 downto 0);
R_out : out STD_LOGIC_VECTOR(7 downto 0)
);
end reg;
architecture Beh of reg is
begin
process (C)
begin
if (rising_edge (C)) then
if (LD = '1') then
R_out <= Reg_in;
end if;
end if;
end process;
end Beh;
TOP:
library IEEE;
use IEEE.STD_LOGIC_1164.all;
entity top is
port(
LDA, LDB, S1, S0 : in bit;
CLK : in STD_LOGIC;
X, Y : in STD_LOGIC_VECTOR(7 downto 0);
RB : out STD_LOGIC_VECTOR(7 downto 0)
);
end top;
architecture Beh of top is
signal regB_out : STD_LOGIC_VECTOR(7 downto 0);
signal regA_out : STD_LOGIC_VECTOR(7 downto 0);
signal mux1_res : STD_LOGIC_VECTOR(7 downto 0);
signal mux2_res : STD_LOGIC_VECTOR(7 downto 0);
begin
Mux1: entity mux(Beh)
port map
(
A => X,
B => regB_out,
Sel => S1,
Z => mux1_res
);
RegA: entity reg(Beh)
port map
(
LD => LDA,
C => CLK,
Reg_in => mux1_res,
R_out => regA_out
);
Mux2: entity mux(Beh)
port map
(
A => regA_out,
B => Y,
Sel => S0,
Z => mux2_res
);
RB<=regB_out;
RegB: entity reg(Beh)
port map
(
LD => LDB,
C => CLK,
Reg_in => mux2_res,
R_out => regB_out
);
end Beh;
I am not sure I wrote bind between RB, regB_out and B correctly. And in the waveform when S1 and S0 both equal 0, I get nonsence for 1 CLK period. Like on the screenshot at 600ns '01' on RB shouldn't be there. Can some one help me to find mistakes?
TestBench:
library IEEE;
use IEEE.std_logic_1164.all;
use IEEE.std_logic_signed.all;
use IEEE.numeric_std.all;
ENTITY tbt is
END tbt;
ARCHITECTURE behavior OF tbt IS
COMPONENT TOP
PORT (
CLK : in STD_LOGIC;
LDA, LDB, S1, S0 : in bit;
X, Y : in STD_LOGIC_VECTOR(7 downto 0);
RB : out STD_LOGIC_VECTOR(7 downto 0)
);
END COMPONENT;
signal CLK_sig : std_logic;
signal LDA_sig, LDB_sig, S1_sig, S0_sig : bit :='0';
signal X_sig, Y_sig, RB_sig : std_logic_vector(7 downto 0):="00000000";
constant CLK_period : time := 100 ns;
constant s_per : time := 50 ns;
begin
-------------------------------------------------------------
uut: TOP PORT MAP (
CLK => CLK_sig,
LDA => LDA_sig,
LDB => LDB_sig,
S1 => S1_sig,
S0 => S0_sig,
X => X_sig,
Y => Y_sig,
RB=> RB_sig
);
-------------------------------------------------------------
CLK_process :process
begin
CLK_sig <= '0';
wait for CLK_period/2;
CLK_sig <= '1';
wait for CLK_period/2;
end process;
-------------------------------------------------------------
stim_proc: process
variable itertion_backwards : integer := 255;
variable itertion_forward : integer := 0;
begin
wait for CLK_period;
for itertion_forward in 0 to 254 loop
X_sig <= STD_LOGIC_VECTOR(TO_SIGNED(INTEGER(itertion_forward),8));
Y_sig <= STD_LOGIC_VECTOR(TO_SIGNED(INTEGER(itertion_backwards),8));
wait for CLK_period;
S1_sig<= not S1_sig;
wait for CLK_period;
S0_sig<= not S0_sig;
wait for CLK_period;
LDA_sig<= not LDA_sig;
wait for CLK_period;
LDB_sig<= not LDB_sig;
itertion_backwards := itertion_backwards - 1;
end loop;
wait;
end process;
end;
My VHDL-Code is functionaly correct, in simulation it does what it's thought for. I tested in many variations and the code works correct.
But when i program the fpga (Nexyx 4 ddr) everything works well except the preload of the counter.
I don't know if the load enable (load_e) output from the fsm doesn't reach the counter or if the output signal that sais the counter is loaded (counter_loaded) doesn't reach the fsm but when i program the fpga it never pases from state C or D (waiting for counter loaded) to state E or F (where it makes a countdown).
I tested the other parts of the code in the target and it works properly, so the only problema so far is that one and i can't find the error, i'm thinking about timming, but i have no idea of how to solve it.
I leave here the counter and fsm code, as well as the TOP code, i`m new in VHDL and it might be lots of bad practice mistakes.
I'm spanish, that's the reason of my bad English and also the spanish name of some signal, but i add a comment next to them.
--------COUNTER---------------------------------------
entity counter is
Generic (NBITS : positive := 15
);
Port (clk : in STD_LOGIC;
rst : in STD_LOGIC;
ce : in STD_LOGIC;
load : in STD_LOGIC_VECTOR (NBITS-1 downto 0);
load_e : in STD_LOGIC;
unit : out STD_LOGIC_VECTOR(3 downto 0);
dec : out STD_LOGIC_VECTOR(3 downto 0);
zero_n : out STD_LOGIC; --true si cuenta = 0
loaded : out STD_LOGIC);
end counter;
architecture Behavioral of counter is
signal q_i : unsigned (NBITS-1 downto 0) := (others => '1');
begin
process(clk,rst)
begin
if rst = '1' then
q_i <= (OTHERS => '1');
loaded <= '0';
elsif rising_edge(clk) then
if CE = '1' then
if load_e = '1' then --ONE OF MY GUESSES OF THE PROBLEM
q_i <= unsigned(load);
loaded <= '1';
else
q_i <= q_i - 1;
loaded <= '0';
end if;
end if;
end if;
end process;
dec <= std_logic_vector(to_unsigned((to_integer(q_i(14 downto 10)) / 10),dec'length)); --first 5 bits are the tens
unit <= std_logic_vector(to_unsigned((to_integer(q_i(14 downto 10)) rem 10),unit'length)); --fist 5 bits are the unit
zero_n <= '1' WHEN q_i < "000010000000000" ELSE '0'; --cout is zero if the first 5 bits are less tan 1 in binary
end Behavioral;
------FINITE STATE MACHINE--------------------------------
entity maquina_estados is
Port (
clk : in STD_LOGIC;
rst : in STD_LOGIC;
corto : in STD_LOGIC;
largo : in STD_LOGIC;
b_on : in STD_LOGIC;
zero_n : in STD_LOGIC;
counter_loaded : in STD_LOGIC;
load_e : out STD_LOGIC;
load : out STD_LOGIC_VECTOR(14 downto 0);
bomba_led : out STD_LOGIC;
indica_on : out STD_LOGIC);
end maquina_estados;
architecture Behavioral of maquina_estados is
type state_type is (A, B, C, D, E, F); --define state(A = powered off, B = powered on, C = short coffee preload, D = large coffee preload, E = short coffee, F = large coffee)
signal state, next_state : state_type; --type state signal
begin
process(clk,rst)
begin
if rst = '1' then
state <= A;
elsif rising_edge(clk) then
state <= next_state;
end if;
end process;
process(state, b_on, corto, largo, zero_n, counter_loaded)
begin
CASE state IS
WHEN A => if b_on = '1' then
next_state <= B;
else
next_state <= A;
end if;
WHEN B => if b_on = '0' then
next_state <= A;
elsif corto = '1' then
next_state <= C;
elsif largo = '1' then
next_state <= D;
else
next_state <= B;
end if;
WHEN C => if counter_loaded = '1' then
next_state <= E;
else
next_state <= C;
end if;
WHEN D => if counter_loaded = '1' then
next_state <= F;
else
next_state <= D;
end if;
WHEN E => if zero_n = '1' then
next_state <= B;
else
next_state <= E;
end if;
WHEN F => if zero_n = '1' then
next_state <= B;
else
next_state <= F;
end if;
WHEN OTHERS => next_state <= A;
end case;
end process;
process(state)
begin
CASE state IS
WHEN A => load <= "111111111111111"; --default value of the count
load_e <= '0';
bomba_led <= '0';
indica_on <= '0';
WHEN B => load <= "111111111111111";
load_e <= '0';
bomba_led <= '0';
indica_on <= '1';
WHEN C => load <= "010101111111111"; --10 second, this in addition to a 1024 hz clock made posible to use the first 5 bits as the number
load_e <= '1';
bomba_led <= '0';
indica_on <= '1';
WHEN D => load <= "101001111111111"; --20 seconds
load_e <= '1';
bomba_led <= '0';
indica_on <= '1';
WHEN E => load <= "111111111111111";
load_e <= '0';
bomba_led <= '1';
indica_on <= '1';
WHEN F => load <= "111111111111111";
load_e <= '0';
bomba_led <= '1';
indica_on <= '1';
end case;
end process;
end behavioral;
------TOP-----------------------
entity TOP is
Generic(
FIN : positive := 100000000;
FOUT : positive := 1024);
Port ( clk : in STD_LOGIC;
rst : in STD_LOGIC;
corto : in STD_LOGIC;
largo : in STD_LOGIC;
b_on : in STD_LOGIC;
display_number : out STD_LOGIC_VECTOR (6 downto 0);
display_selection : out STD_LOGIC_VECTOR (7 downto 0);
bomba_led : out STD_LOGIC;
indica_on : out STD_LOGIC);
end TOP;
architecture Behavioral of TOP is
--instancies
component clk_divider is
-- Port ( );
generic(
FIN : positive;
FOUT : positive
);
port (
Clk : in STD_LOGIC;
Reset : in STD_LOGIC;
Clk_out : out STD_LOGIC
);
end component;
component maquina_estados is
Port ( clk : in STD_LOGIC;
rst : in STD_LOGIC;
corto : in STD_LOGIC;
largo : in STD_LOGIC;
b_on : in STD_LOGIC;
zero_n : in STD_LOGIC;
counter_loaded : in STD_LOGIC;
load_e : out STD_LOGIC;
load : out STD_LOGIC_VECTOR(14 downto 0);
bomba_led : out STD_LOGIC;
indica_on : out STD_LOGIC);
end component;
component counter is
Generic (NBITS : positive
);
Port (clk : in STD_LOGIC;
rst : in STD_LOGIC;
ce : in STD_LOGIC;
load : in STD_LOGIC_VECTOR (NBITS-1 downto 0);
load_e : in STD_LOGIC;
unit : out STD_LOGIC_VECTOR(3 downto 0);
dec : out STD_LOGIC_VECTOR(3 downto 0);
zero_n : out STD_LOGIC;
loaded : out STD_LOGIC);
end component;
component clk_manager is
generic(
CLK_FREQ : positive
);
Port (
clk : in STD_LOGIC;
rst : in STD_LOGIC;
strobe_1024Hz : out STD_LOGIC;
strobe_128Hz : out STD_LOGIC
);
end component;
component decoder is
Port ( code : in STD_LOGIC_VECTOR(3 downto 0);
led : out STD_LOGIC_vector(6 downto 0)
);
end component;
component display_refresh is
Port ( clk : in STD_LOGIC;
ce : in STD_LOGIC;
segment_unit : in STD_LOGIC_VECTOR (6 downto 0);
segment_dec : in STD_LOGIC_VECTOR (6 downto 0);
display_number : out STD_LOGIC_VECTOR (6 downto 0);
display_selection : out STD_LOGIC_VECTOR (1 downto 0)); --cada elemento del vector corresponde a un 7 seg, true se ve false no
end component;
-- prescaler signals
signal prescaler_clk_out : STD_LOGIC;
--maquina estados signals
signal zero_n_fsm : STD_LOGIC;
signal load_e_fsm : STD_LOGIC;
signal load_fsm : STD_LOGIC_VECTOR(14 downto 0);
signal bomba_led_fsm: STD_LOGIC;
--counter signals
signal unit : STD_LOGIC_VECTOR(3 downto 0);
signal dec : STD_LOGIC_VECTOR(3 downto 0);
signal zero_n_cntr : STD_LOGIC;
signal load_e_cntr : STD_LOGIC;
signal load_cntr : STD_LOGIC_VECTOR(14 downto 0);
signal counter_loaded : STD_LOGIC;
--clk_manager signals
signal strobe_1024Hz : STD_LOGIC;
signal strobe_128Hz : STD_LOGIC;
signal ce_clkm : STD_LOGIC;
signal rst_clkm : STD_LOGIC;
--decoders signals
signal unit_code : STD_LOGIC_VECTOR(6 downto 0);
signal dec_code : STD_LOGIC_VECTOR(6 downto 0);
--display refresh signals
signal display_refresh_number : STD_LOGIC_VECTOR(6 downto 0);
signal display_refresh_selection : STD_LOGIC_VECTOR(1 downto 0);
begin
prescaler: clk_divider
generic map(
FIN => FIN,
FOUT => FOUT
)
port map(
Clk => clk,
Reset => rst,
Clk_out => prescaler_clk_out
);
sm: maquina_estados
Port map( clk => prescaler_clk_out,
rst => rst,
corto => corto,
largo => largo,
b_on => b_on,
zero_n => zero_n_fsm,
counter_loaded => counter_loaded,
load_e => load_e_fsm,
load => load_fsm,
bomba_led => bomba_led_fsm,
indica_on => indica_on);
cntr: counter
Generic map(NBITS => 15
)
Port map(clk => clk,
rst => rst,
ce => strobe_1024Hz,
load => load_cntr,
load_e => load_e_fsm,
unit => unit,
dec => dec,
zero_n => zero_n_cntr,
loaded => counter_loaded);
clk_m: clk_manager
generic map(
CLK_FREQ => FIN
)
Port map(
clk => clk,
rst => rst,
strobe_1024Hz => strobe_1024Hz,
strobe_128Hz => strobe_128Hz
);
unit_dcd: decoder
Port map(
code => unit,
led => unit_code
);
dec_dcd: decoder
Port map(
code => dec,
led => dec_code
);
dr: display_refresh
Port map(
clk => clk,
ce => strobe_128Hz,
segment_unit => unit_code,
segment_dec => dec_code,
display_number => display_refresh_number,
display_selection => display_refresh_selection);
display_number <= display_refresh_number WHEN bomba_led_fsm = '1' ELSE "1111111";
display_selection <= ("111111" & display_refresh_selection) WHEN bomba_led_fsm = '1' ELSE "11111111";
zero_n_fsm <= zero_n_cntr;
bomba_led <= bomba_led_fsm;
load_cntr <= load_fsm;
end Behavioral;
Here are all the reports that the implementation ans sythesis gave me:
Synthesis reports
implementation reports 1/6
implementation reports 2/6
implementation reports 3/6
implementation reports 4/6
implementation reports 5/6
implementation reports 6/6
I hope someone could find the problema and give me a solution or a way of how to debug this problem.
Thanks.
Your FSM is clocked on prescaler_clk_out, and your counter is clocked on clk, which is a red flag. This could easily lead to an implementation failure.
Draw a timing diagram showing all your clocks and resets, and your lower-frequency enables (in particular, strobe_1024Hz)
Try to clock all the logic on the same clock, presumably clk, and make sure that everything is synchronous to this clock (in other words, inputs have sufficient setup and hold times relative to this clock)
Make sure you are actually resetting the chip
Once you've done the timing diagram, write a constraints file that tells the synthesiser what your clocks are. clk_manager and clk_divider may be an issue here, but hopefully everything will be clocked on just 'clk', and the contstraints file will contain only the clock name and frequency. If you still can't get it to work, ask a new question, showing your timing diagram, and your attempt at a constraints file.
I'm supposed to write up a 16-bit ALU. My professor wants us to try and code the adder and sub of the ALU with a
signal tmp : std_logic_vector(16 downto 0); and then in the case for the select input s we put:
tmp <= conv_std_logic_vector(conv_integer(a) + conv_integer(b), 17);
After experimenting with it for a while, my waveform only showed the inputs' values as UUUUUUUUUUUUUUUU. Even after I had commented out the conv_std_logic_vector(...) stuff.
Is there a simple explanation as to why my inputs aren't showing up in the waveform?
Here is my code:
-- 16-Bit ALU
-- By: Logan Jordon
library ieee;
use ieee.std_logic_1164.all;
use ieee.std_logic_unsigned.all;
use IEEE.NUMERIC_STD.ALL;
--use ieee.std_logic_arith.all;
entity alu16 is
port (
a : in std_logic_vector(15 downto 0);
b : in std_logic_vector(15 downto 0);
s : in std_logic_vector(1 downto 0);
r : out std_logic_vector(15 downto 0);
cout : out std_logic;
lt, eq, gt : out std_logic;
overflow : out std_logic
);
end entity alu16;
architecture beh of alu16 is
signal tmp : std_logic_vector(16 downto 0);
signal add_overflow : std_logic;
signal sub_overflow : std_logic;
begin
-- PROCESS
process(a, b, add_overflow, sub_overflow)
begin
case s is
--ADD
when "00" =>
--tmp <= conv_std_logic_vector(conv_integer(a) + conv_integer(b), 17);
tmp <= a + b;
overflow <= add_overflow;
--SUB
when "01" =>
--tmp <= conv_std_logic_vector(conv_integer(a) - conv_integer(b), 17);
tmp <= a - b;
overflow <= sub_overflow;
--AND
when "10" =>
tmp <= '0' & a AND b;
overflow <= '0';
--OR
when "11" =>
tmp <= '0' & a OR b;
overflow <= '0';
when others =>
tmp <= "00000000000000000";
end case;
--One-Bitters
if a > b then
gt <= '1';
lt <= '0';
eq <= '0';
elsif a < b then
lt <= '1';
gt <= '0';
eq <= '0';
elsif a = b then
eq <= '1';
lt <= '0';
gt <= '0';
end if;
end process;
--OUTPUTS
cout <= tmp(16);
r <= tmp(15 downto 0);
add_overflow <= '1' when (a(15) = b(15)) and (a(15) /= tmp(15))
else '0';
sub_overflow <= '1' when (a(15) = NOT b(15)) and (a(15) /= tmp(15))
else '0';
end beh;
EDIT: In the case that it might be my test bench, here's the code for my testbench:
library ieee;
use ieee.std_logic_1164.all;
use ieee.std_logic_unsigned.all;
use IEEE.NUMERIC_STD.ALL;
entity alu16_tb is
end alu16_tb;
architecture behavior of alu16_tb is
component ALU16
port(
a : in std_logic_vector(15 downto 0);
b : in std_logic_vector(15 downto 0);
s : in std_logic_vector(1 downto 0);
r : out std_logic_vector(15 downto 0);
cout : out std_logic;
lt, eq, gt : out std_logic;
overflow : out std_logic
);
end component;
-- Signals to interface with the UUT
-- Set each of the input vectors to unique values to avoid
-- needing a process to drive them below
signal a : std_logic_vector(15 downto 0) := "0000000000000000";
signal b : std_logic_vector(15 downto 0) := "0000000000000000";
signal s : std_logic_vector(1 downto 0) := "00";
signal r : std_logic_vector(15 downto 0):= "0000000000000000";
signal cout : std_logic := '0';
signal lt : std_logic := '0';
signal gt : std_logic := '0';
signal eq : std_logic := '0';
signal overflow : std_logic := '0';
constant tick : time := 10 ns;
begin
-- Instantiate the Unit Under Test (UUT)
uut : ALU16 port map (
a => a,
b => b,
s => s,
r => r,
cout => cout,
lt => lt,
gt => gt,
eq => eq,
overflow => overflow
);
-- Drive selector bits
drive_s : process
begin
a <= "0000000000000001";
b <= "0000000000000010";
wait for (tick*2);
s <= "00";
wait for (tick*2);
s <= "01";
wait for (tick*2);
s <= "10";
wait for (tick*2);
s <= "11";
end process drive_s;
end;
I am implementing a multiplier in which i multiply A (8 bits) and B (8 bits), and store result at S. Number of bit required for output S is 16 bits. S have higher part SH and lower part SL.Every time i shift ,add operation is performed
i am getting following errors in my controller part :-
Attribute event requires a static signal prefix
is not declared.
"**" expects 2 arguments
and my code is:-
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use ieee.std_logic_arith.all;
use ieee.std_logic_unsigned.all;
entity PIPO is
port (reset: in std_logic ;
B:IN STD_LOGIC_VECTOR (7 downto 0 );
LOAD:in std_logic ;
SHIFT:in std_logic ;
ADD:in std_logic ;
Sum:IN STD_LOGIC_VECTOR (7 downto 0 );
C_out:in std_logic ;
CLK:in std_logic ;
result: out STD_LOGIC_VECTOR (15 downto 0) ;
LSB:out std_logic ;
TB:out std_logic_vector (7 downto 0) );
end ;
architecture rtl OF PIPO is
signal temp1 : std_logic_vector(15 downto 0);
----temp2 -add
signal temp2 : std_logic ;
begin
process (CLK, reset)
begin
if reset='0' then
temp1<= (others =>'0');
temp2<= '0';
elsif (CLK'event and CLK='1') then
if LOAD ='1' then
temp1(7 downto 0) <= B;
temp1(15 downto 8) <= (others => '0');
end if ;
if ADD= '1' then
temp2 <='1';
end if;
if SHIFT= '1' then
if ADD= '1' then
------adder result ko add n shift
temp2<= '0';
temp1<=C_out & Sum & temp1( 7 downto 1 );
else
----only shift
temp1<= '0' & temp1( 15 downto 1 );
end if;
end if;
end if;
end process;
LSB <=temp1(0);
result<=temp1( 15 downto 0 );
TB <=temp1(15 downto 8);
end architecture rtl;
-------------------------------------------
-------------------------------------------
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use ieee.std_logic_arith.all;
use ieee.std_logic_unsigned.all;
entity Controller is
Port ( ADD :OUT STD_LOGIC;
SHIFT:OUT STD_LOGIC;
LOAD:OUT STD_LOGIC;
STOP:OUT STD_LOGIC;
STRT:IN STD_LOGIC;
LSB:IN STD_LOGIC;
CLK:IN STD_LOGIC;
reset:IN STD_LOGIC );
end ;
architecture rtl OF Contoller is
---RTL level code is inherently synchronous
signal count : unsigned (2 downto 0);
----differnt states
type state_typ is ( IDLE, INIT, TEST, ADDs, SHIFTs );
signal state : state_typ;
begin
--controller : process (ADD,SHIFT,LOAD,STOP,STRT,LSB,CLK,reset)
process (state)--(CLK, reset,ADD,SHIFT,LOAD,STOP,STRT,LSB)
begin
if reset='0' then
state <= IDLE;
count <= "000";
elsif (CLK'event and CLK='1') then
case state is
when IDLE =>
if STRT = '1' then
--- if STRT = '1' then
state <= INIT;
else
state <= IDLE;
end if;
when INIT =>
state <= TEST;
when TEST =>
if LSB = '0' then
state <= SHIFTs;
else
state <= ADDs;
end if;
when ADDs =>
state <= SHIFTs;
when SHIFTs =>
if count = "111" then
count <= "000";
state <= IDLE;
else
count<= std_logic_vector(unsigned(count) + 1);
state <= TEST;
end if;
end case;
end if;
end process ;
STOP <= '1' when state = IDLE else '0';
ADD <= '1' when state = ADDs else '0';
SHIFT <= '1' when state = SHIFTs else '0';
LOAD <= '1' when state = INIT else '0';
end architecture rtl;
----------------------------------------------
--------------------------------------------
---multiplicand
library ieee;
use ieee.std_logic_1164.all;
entity multiplicand is
port (A : in std_logic(7 downto 0);
reset :in std_logic;
LOAD : in std_logic;
TA : OUT STD_LOGIC(7 downto 0);
CLK : in std_logic );
end entity;
architecture rtl OF multiplicand is
begin
process (CLK, reset)
begin
if reset='0' then
TA <= (others =>'0'); -- initialize
elsif (CLK'event and CLK='1') then
if LOAD_cmd = '1' then
TA(7 downto 0) <= A_in; -- load B_in into register
end if;
end if ;
end process;
end architecture rtl;
------------------------------------------------------
------------------------------------------------------
---Full Adder
library ieee;
use ieee.std_logic_1164.all;
entity Full_Adder is
port (A : in std_logic;
B : in std_logic;
C_in : in std_logic;
Sum : out std_logic ;
C_out : out std_logic);
end;
architecture struc of Full_Adder is
begin
Sum <= A xor B xor C_in;
C_out <= (A and B) or (A and C_in) or (B and C_in);
end struc;
------------------------------------------------------------
-------------------------------------------------------------
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
entity Adder is
Port ( TA : in STD_LOGIC_VECTOR (7 downto 0);
TB : in STD_LOGIC_VECTOR (7 downto 0);
Sum : out STD_LOGIC_VECTOR (7 downto 0);
C_in : in STD_LOGIC;
C_out : out STD_LOGIC);
end Adder;
architecture struc of Adder is
component Full_Adder is
port(A : in std_logic;
B : in std_logic;
C_in : in std_logic;
Sum : out std_logic ;
C_out : out std_logic);
end component;
signal C: std_logic_vector (7 downto 0);
begin
FA0:Full_Adder port map(TA(0), TB(0), C_in, Sum(0), C(0));
FA1: Full_Adder port map(TA(1), TB(1), C(0), Sum(1), C(1));
FA3: Full_Adder port map(TA(2),TB(2), C(1), Sum(2), C(2));
FA4: Full_Adder port map(TA(3), TB(3), C(2), Sum(3), C(3));
FA5: Full_Adder port map(TA(4), TB(4), C(3), Sum(4), C(4));
FA6: Full_Adder port map(TA(5), TB(5), C(4), Sum(5), C(5));
FA7: Full_Adder port map(TA(6), TB(6), C(5), Sum(6), C(6));
FA8: Full_Adder port map(TA(7), TB(7), C(6), Sum(7), C(7));
C_out <= C(7);
end struc;
------------------------------------------------------------
------------------------------------------------------------
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use ieee.std_logic_arith.all;
use ieee.std_logic_unsigned.all;
entity multiplier is
Port ( num1 : in STD_LOGIC_VECTOR (7 downto 0);
num2 : in STD_LOGIC_VECTOR (7 downto 0);
result : out STD_LOGIC_VECTOR (15 downto 0);
CLK:in std_logic ;
reset:IN STD_LOGIC;
STRT:IN STD_LOGIC;
STOP:OUT STD_LOGIC );
end multiplier;
architecture rtl of Multiplier is
signal ADD :STD_LOGIC;
signal SHIFT :STD_LOGIC;
signal LOAD :STD_LOGIC;
signal LSB :STD_LOGIC;
signal A : STD_LOGIC_VECTOR (7 downto 0);
signal B :STD_LOGIC_VECTOR (7 downto 0);
signal Sum:STD_LOGIC_VECTOR (7 downto 0);
signal C_out:STD_LOGIC;
component Controller
port (
ADD :OUT STD_LOGIC;
SHIFT:OUT STD_LOGIC;
LOAD:OUT STD_LOGIC;
STOP:OUT STD_LOGIC;
STRT:IN STD_LOGIC;
LSB:IN STD_LOGIC;
CLK:IN STD_LOGIC;
reset:IN STD_LOGIC );
end component;
component Adder
port (
TA : in STD_LOGIC_VECTOR (7 downto 0);
TB : in STD_LOGIC_VECTOR (7 downto 0);
Sum : out STD_LOGIC_VECTOR (7 downto 0);
C_in : in STD_LOGIC;
C_out : out STD_LOGIC);
end component;
component PIPO
port (reset: in std_logic ;
B:IN STD_LOGIC_VECTOR (7 downto 0 );
LOAD:in std_logic ;
SHIFT:in std_logic ;
ADD:in std_logic ;
Sum:IN STD_LOGIC_VECTOR (7 downto 0 );
C_out:in std_logic ;
CLK:in std_logic ;
result: out STD_LOGIC_VECTOR (15 downto 0) ;
LSB:out std_logic ;
TB:out std_logic );
end component;
component multiplicand
port (A : in std_logic (7 downto 0);
reset :in std_logic;
LOAD : in std_logic;
TA : OUT STD_LOGIC(7 downto 0);
CLK : in std_logic );
end component ;
begin
inst_Controller: Controller
port map (ADD => ADD,
SHIFT =>SHIFT,
LOAD =>LOAD ,
STOP =>STOP,
STRT =>STRT,
LSB =>LSB ,
CLK =>CLK ,
reset =>reset
);
inst_multiplicand :multiplicand
port map (A =>A,
reset=>reset,
LOAD =>LOAD,
TA => TA(7 downto 0),
CLK => CLK
);
inst_PIPO :PIPO
port map ( reset => reset,
B => B ,
LOAD =>LOAD,
SHIFT=>SHIFT,
ADD=>ADD,
Sum=>Sum,
C_out=>C_out,
CLK=>CLK,
result=>result,
LSB=>LSB,
TB=>TB
);
inst_Full_Adder : Full_Adder
port map ( TA => TA,
TB =>TB,
Sum=>Sum ,
C_in=>C_in,
C_out=>C_out
);
end rtl;
Actually the space between CLK and the apostrophe/tick isn't significant
david_koontz#Macbook: token_test
elsif (CLK 'event and CLK ='1') then
KEYWD_ELSIF (151) elsif
DELIM_LEFT_PAREN ( 9) (
IDENTIFIER_TOKEN (128) CLK
DELIM_APOSTROPHE ( 8) '
IDENTIFIER_TOKEN (128) event
KEYWD_AND (134) and
IDENTIFIER_TOKEN (128) CLK
DELIM_EQUAL ( 25) =
CHAR_LIT_TOKEN ( 2) '1'
DELIM_RIGHT_PAREN ( 10) )
KEYWD_THEN (211) then
gives the same answer as:
david_koontz#Macbook: token_test
elsif (CLK'event and CLK ='1') then
KEYWD_ELSIF (151) elsif
DELIM_LEFT_PAREN ( 9) (
IDENTIFIER_TOKEN (128) CLK
DELIM_APOSTROPHE ( 8) '
IDENTIFIER_TOKEN (128) event
KEYWD_AND (134) and
IDENTIFIER_TOKEN (128) CLK
DELIM_EQUAL ( 25) =
CHAR_LIT_TOKEN ( 2) '1'
DELIM_RIGHT_PAREN ( 10) )
KEYWD_THEN (211) then
In vhdl, there is no lexical element parsing requiring a lack of white space. (Sorry Russel).
Correcting the other syntax ambiguities of your code (see below, missing context clause, Controller misspelled in the architecture declaration, count used as both a scalar and array subtype), results in two different VHDL analyzers swallowing the space between CLK and ' just fine.
The problem is in the tool you are using not actually being standard compliant or the code you present as having the problem isn't actually representational of the code generating the error. If a non-compliant tool it's likely a shortcoming you can live with, although there may be more things a bit more irksome.
david_koontz#Macbook: ghdl -a controller.vhdl
david_koontz#Macbook: nvc -a controller.vhdl
david_koontz#Macbook:
(no errors, it also elaborates without a test bench in ghdl, nvc disallows top level ports - which it is permitted to do by the standard)
library ieee;
use ieee.std_logic_1164.all;
use ieee.numeric_std.all;
entity Controller is
Port (
ADD: OUT STD_LOGIC;
SHIFT: OUT STD_LOGIC;
LOAD: OUT STD_LOGIC;
STOP: OUT STD_LOGIC;
STRT: IN STD_LOGIC;
LSB: IN STD_LOGIC;
CLK: IN STD_LOGIC;
reset: IN STD_LOGIC
);
end entity;
architecture rtl OF Controller is
---RTL level code is inherently synchronous
signal count : std_logic_vector (2 downto 0);
----differnt states
type state_typ is ( IDLE, INIT, TEST, ADDs, SHIFTs );
signal state : state_typ;
begin
NOLABEL:
process (CLK, reset)
begin
if reset='0' then
state <= IDLE;
count <= "000";
elsif (CLK 'event and CLK ='1') then
case state is
when IDLE =>
if STRT = '1' then
state <= INIT;
else
state <= IDLE;
end if;
when INIT =>
state <= TEST;
when TEST =>
if LSB = '0' then
state <= SHIFTs;
else
state <= ADDs;
end if;
when ADDs =>
state <= SHIFTs;
when SHIFTs =>
if count = "111" then -- verify if finished
count <= "000"; -- re-initialize counter
state <= IDLE; -- ready for next multiply
else
count <= -- increment counter
std_logic_vector(unsigned(count) + 1);
state <= TEST;
end if;
end case;
end if;
end process;
---end generate; ???
STOP <= '1' when state = IDLE else '0';
ADD <= '1' when state = ADDs else '0';
SHIFT <= '1' when state = SHIFTs else '0';
LOAD <= '1' when state = INIT else '0';
end architecture rtl;
The error message appears to stem from the signal CLK (the prefix for the event attribtute). There is no other use of the event attribute in your code presented with the question. A signal is one of the elements of entity_class that can be decorated with an attribute.
In the VHDL LRM's section on predefined attributes 'EVENT can only decorate a signal, and CLK is a signal (declared in a port). In that section the prefix is required to be denoted by a static signal name.
Is CLK a static signal name? Yes it is. It's a scalar subtype declared in the entity declaration and is locally static (available at analysis time - it's a scalar, a simple name and not involving a generic).
And about now you might get why someone would wonder if the code in the question is representational of the code generating the error or the VHDL tool used is not compliant.
The error message you report is usually associated with trying to use 'EVENT with an indexed signal name, e.g. w(i)'event. (See Signal attributes on a signal vector).
You're going to kick yourself for this one:
elsif (CLK 'event and CLK ='1') then
Should be:
elsif (CLK'event and CLK ='1') then
See the difference?
Even better:
elsif rising_edge(CLK) then
It seems you're missing a clk entry in the process
Change the line reading:
process (state)--(CLK, reset,ADD,SHIFT,LOAD,STOP,STRT,LSB)
to read:
process (clk, reset)
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use IEEE.NUMERIC_STD.ALL;
entity fir_123 is
port( Clk : in std_logic; --clock signal
Xin : in signed(7 downto 0); --input signal
Yout : out signed(15 downto 0) --filter output
);
end fir_123;
architecture Behavioral of fir_123 is
component DFF is
port(
Q : out signed(15 downto 0); --output connected to the adder
Clk :in std_logic; -- Clock input
D :in signed(15 downto 0) -- Data input from the MCM block.
);
end component;
signal H0,H1,H2,H3 : signed(7 downto 0) := (others => '0');
signal MCM0,MCM1,MCM2,MCM3,add_out1,add_out2,add_out3 : signed(15 downto 0) := (others => '0');
signal Q1,Q2,Q3 : signed(15 downto 0) := (others => '0');
begin
--filter coefficient initializations.
--H = [-2 -1 3 4].
H0 <= to_signed(-2,8);
H1 <= to_signed(-1,8);
H2 <= to_signed(3,8);
H3 <= to_signed(4,8);
--Multiple constant multiplications.
MCM3 <= H3*Xin;
MCM2 <= H2*Xin;
MCM1 <= H1*Xin;
MCM0 <= H0*Xin;
--adders
add_out1 <= Q1 + MCM2;
add_out2 <= Q2 + MCM1;
add_out3 <= Q3 + MCM0;
--flipflops(for introducing a delay).
dff1 : DFF port map(Q1,Clk,MCM3);
dff2 : DFF port map(Q2,Clk,add_out1);
dff3 : DFF port map(Q3,Clk,add_out2);
--an output produced at every positive edge of clock cycle.
process(Clk)
begin
if(rising_edge(Clk)) then
Yout <= add_out3;
end if;
end process;
end Behavioral;
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use IEEE.NUMERIC_STD.ALL;
entity dff is
port(`
Q : out signed(15 downto 0); --output connected to the adder
Clk :in std_logic; -- Clock input
D :in signed(15 downto 0) -- Data input from the MCM block.
);
end dff;
architecture Behavioral of dff is
signal qt : signed(15 downto 0) := (others => '0');
begin
Q <= qt;
process(Clk)
begin
if ( rising_edge(Clk) ) then
qt <= D;
end if;
end process;
end Behavioral;
When I run this code it compiles successfully error free syntax but I get several warning and because of that I am not getting desired result. I get Xin, Clkin & Yout undefined in simulation result. I tried in different ways but still I haven't resolved these warnings:
1) WARNING:Xst:1293 - FF/Latch has a constant value of 0 in
block . This FF/Latch will be trimmed during the optimization
process.
2) WARNING:Xst:1293 - FF/Latch has a constant value of
0 in block . This FF/Latch will be trimmed during the
optimization process.
3) WARNING:Xst:1293 - FF/Latch has a
constant value of 0 in block . This FF/Latch will be trimmed
during the optimization process.
4) WARNING:Xst:1896 - Due to other
FF/Latch trimming, FF/Latch has a constant value of 0 in
block . This FF/Latch will be trimmed during
There seems to be no problem with the code. The only thing that I thought could go wrong is the fact that the fir module doesn't have any reset. The code for fir is as follows:
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use IEEE.NUMERIC_STD.ALL;
entity fir_123 is
port( Clk : in std_logic; --clock signal
reset: in std_logic;
Xin : in signed(7 downto 0); --input signal
Yout : out signed(15 downto 0) --filter output
);
end fir_123;
architecture Behavioral of fir_123 is
component DFF is
port(
Q : out signed(15 downto 0); --output connected to the adder
Clk :in std_logic; -- Clock input
reset: in std_logic;
D :in signed(15 downto 0) -- Data input from the MCM block.
);
end component;
signal H0,H1,H2,H3 : signed(7 downto 0) := (others => '0');
signal MCM0,MCM1,MCM2,MCM3,add_out1,add_out2,add_out3 : signed(15 downto 0) := (others => '0');
signal Q1,Q2,Q3 : signed(15 downto 0) := (others => '0');
signal yout_int : signed(15 downto 0);
begin
--filter coefficient initializations.
--H = [-2 -1 3 4].
H0 <= to_signed(-2,8);
H1 <= to_signed(-1,8);
H2 <= to_signed(3,8);
H3 <= to_signed(4,8);
--Multiple constant multiplications.
MCM3 <= H3*Xin;
MCM2 <= H2*Xin;
MCM1 <= H1*Xin;
MCM0 <= H0*Xin;
--adders
add_out1 <= Q1 + MCM2;
add_out2 <= Q2 + MCM1;
add_out3 <= Q3 + MCM0;
--flipflops(for introducing a delay).
dff1 : DFF port map(Q1,Clk,reset,MCM3);
dff2 : DFF port map(Q2,Clk,reset,add_out1);
dff3 : DFF port map(Q3,Clk,reset,add_out2);
--an output produced at every positive edge of clock cycle.
registered_yout: process
begin
wait until rising_edge(clk);
if (reset = '1') then
yout_int <= (others => '0');
else
yout_int <= add_out3;
end if;
end process;
Yout <= yout_int;
end Behavioral;
I also added in reset for dff and the changed file looks like this:
library IEEE;
use IEEE.STD_LOGIC_1164.ALL;
use IEEE.NUMERIC_STD.ALL;
entity dff is
port(
Q : out signed(15 downto 0); --output connected to the adder
Clk :in std_logic; -- Clock input
reset: in std_logic;
D :in signed(15 downto 0) -- Data input from the MCM block.
);
end dff;
architecture Behavioral of dff is
signal qt : signed(15 downto 0) := (others => '0');
begin
Q <= qt;
registered_qt : process
begin
wait until rising_edge(clk);
if (reset = '1') then
qt <= (others => '0');
else
qt <= D;
end if;
end process;
end Behavioral;
The testbench that I used is as follows:
library ieee;
use ieee.numeric_std.all;
use ieee.std_logic_1164.all;
entity tb is
end entity tb;
architecture test_bench of tb is
component fir_123 is
port( Clk : in std_logic;
reset : in std_logic;
Xin : in signed(7 downto 0);
Yout : out signed(15 downto 0)
);
end component fir_123;
constant clk_per : time := 8 ns;
signal clk: std_logic;
signal reset: std_logic;
signal Xin : signed(7 downto 0);
signal Yout : signed(15 downto 0);
begin
dft : component fir_123
port map (
Clk => clk,
reset => reset,
Xin => Xin,
Yout => Yout
);
Clk_generate : process --Process to generate the clk
begin
clk <= '0';
wait for clk_per/2;
clk <= '1';
wait for clk_per/2;
end process;
Rst_generate : process --Process to generate the reset in the beginning
begin
reset <= '1';
wait until rising_edge(clk);
reset <= '0';
wait;
end process;
Test: process
begin
Xin <= (others => '0');
wait until rising_edge(clk);
Xin <= (others => '1');
wait until rising_edge(clk);
Xin <= (others => '0');
wait for clk_per*10;
report "testbench finished" severity failure;
end process test;
end architecture test_bench;
I have checked the waveforms in a simulator and they all seem to be defined after the reset has been deasserted. The fact that Xin and Clk is undefined shows that there is something wrong with the testbench.