Assign a value to a position in a std_logic_vector

[ionutca...@gmail.com]

Hi , I'm new to VHDL and I can't figure out how to do this :

I have a vector OP3 : std_logic_vector(31 downto 0) and another one S : std_logic_vector(4 downto 0) ...
My question is , is it possible , and if yes , how can I do OP3(S) <= '1' ?

(For example , if S = 00111 then OP3(7) <= '1').

[Thomas Stanka]

Convert S to natural (integer) type.

signal S_Nat : natural;
S_nat <= to_integer(unsigned(S));

than you can use following code.
OP3(S_nat) <= '1';

ofc it is possible to have conversion "Inline" if you don't like to
have an additional signal.

best regards
Thomas

[Tricky]

Or if OP3(n) is a register, you probably want to do the type conversion via a variable to avoid the extra register that would be implied using a signal inside the process.

[ionutca...@gmail.com]

Thank you for the replies , it was really helpful , but now I have another problem.
I have OP3 <= RdData2 (if RdData2 = ABABABAB OP3 <= ABABABAB) , and then
OP3(to_integer(unsigned(Sa))) <= '1' , and here the problem occurs , that one bit is set correctly to 1 , but all the others become U or X-es . How can i solve this?

[Andy]

On Tuesday, February 26, 2013 5:56:27 AM UTC-6, ionutca...@gmail.com wrote:
> Thank you for the replies , it was really helpful , but now I have another problem. I have OP3 <= RdData2 (if RdData2 = ABABABAB OP3 <= ABABABAB) , and then OP3(to_integer(unsigned(Sa))) <= '1' , and here the problem occurs , that one bit is set correctly to 1 , but all the others become U or X-es . How can i solve this?

You should set the entire vector to zeroes before setting the single bit to '1'.

opp3 <= (others => '0'); -- prior to assigning single bit

Andy

[GaborSzakacs]

ionutca...@gmail.com wrote:
> Thank you for your fast replies , it is helpful , though I have another question.
> I have : OP3 <= x"0000_0000" and Sa < = "10101" . Then , if I do
> OP3(4 downto 0) <= Sa , the result is 00...X0X0X rather than 00..10101 (.. means many zeros :) ) ... What am I doing wrong here ?
>
If the assignmet of all zeroes is not within a process, then
what you're seeing is the result of conflicting drivers.
i.e. All bits that are driven to the same value by
both assignments will show that value. All bits that
are driven to different values by the two assignments
will show as 'X'. In your case one value is all zero,
so any ones in the other value will be in conflict and
show up as 'X'.

-- Gabor

[ionutca...@gmail.com]

I've tried to use processes , but still nothing (I'm pretty sure something is still wrong , as I said I'm new to VHDL). Here is my code :

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 instantiating
---- any Xilinx primitives in this code.
--library UNISIM;
--use UNISIM.VComponents.all;

entity ALU is
Port (
RdData1 : in STD_LOGIC_VECTOR (31 downto 0);
RdData2 : in STD_LOGIC_VECTOR (31 downto 0);
FAddr : in STD_LOGIC_VECTOR (15 downto 0);
ALUSrc : in STD_LOGIC;
ALUOP : in STD_LOGIC_VECTOR (2 downto 0);
Sa : in STD_LOGIC_VECTOR (4 downto 0);
Y : out STD_LOGIC_VECTOR (31 downto 0)
);
end ALU;

architecture Behavioral of ALU is
signal SEAddr : STD_LOGIC_VECTOR(31 downto 0);
signal OP1 : STD_LOGIC_VECTOR(31 downto 0);
signal OP2 : STD_LOGIC_VECTOR(31 downto 0);
signal OP3 : STD_LOGIC_VECTOR(31 downto 0);
signal S_nat : natural;
begin
OP2 <= RdData2 when AluSrc='0' else SEAddr;
SEAddr(15 downto 0) <= FAddr;
SEAddr(31 downto 16) <= x"0000" when FAddr(15)='0' else x"FFFF";

OP1 <= RdData1;

process (RdData2)
begin
OP3 <= RdData2;
end process;

with Sa select
S_nat <= 0 when "00000",1 when "00001",2 when "00010",3 when "00011",4 when "00100",5 when "00101",
6 when "00110",7 when "00111",8 when "01000",9 when "01001",10 when "01010",11 when "01011",
12 when "01100",13 when "01101",14 when "01110",15 when "01111",16 when "10000",17 when "10001",
18 when "10010",19 when "10011",20 when "10100",21 when "10101",22 when "10110",23 when "10111",
24 when "11000",25 when "11001",26 when "11010",27 when "11011",28 when "11100",29 when "11101",
30 when "11110",31 when others;

process (S_nat)
begin
OP3(S_nat) <= '1';
end process;

with ALUOP select
y <= OP1 + OP2 when "000", OP1 - OP2 when "001", OP1 AND OP2 when "010", OP1 OR OP2 when "011", OP3 when others;

end Behavioral;

What I want is that OP3 to be exactly as RdData2 (and if I just do OP3 <= RdData2 ,it is) but also that 1 bit which is determined by Sa to become 1.

For example , RdData 2 is 5555 5555 , and if I do only Op3 <= RdData2 then OP3 becomes 5555 5555 , but if i also do OP3(S_nat) <= '1' , then the result is XXXX XXX (to be more precise is xxxx xxxx xxxx xxxx xxxx xx1x xxxx xxx1 , so that 1 specific bit (10 in this case) becomes 1 , but the others are X-es).

Any help would be highly appreciated , I'm working on this problem for several days and it still bugs me :(.

[GaborSzakacs]

The problem is that you are making the two assignments in two different
processes, so they are "fighting" eachother rather than one having
higher priority. For what you're doing, it may be easier to have
a separate vector with a single 1 in a position determined by S_nat
and then OR that vector with OP3 in the same process where you
currently assign it to RdData2.

-- Gabor

[ionutca...@gmail.com]

Thank You , Thank You ALL , finally it is working . Again , Thank You very much :)

[Thomas Stanka]

On 26 Feb., 15:54, ionutcazacu2...@gmail.com wrote:
> use IEEE.STD_LOGIC_ARITH.ALL;
> use IEEE.STD_LOGIC_UNSIGNED.ALL;

Not your question but usage of these libraries is bad style. You will
see them a lot in old books and unfortunately too often in even not so
old books).

Use ieee.numeric_std.all instead. The _arith and _unsigned itself are
vendor dependent, while numeric_std is real ieee standard and
therefore vendor independent.

>         process (RdData2)
>           begin
>                 OP3 <= RdData2;
>         end process;
>         with Sa select
>                 S_nat <= 0 when "00000",1 when "00001",2 when "00010",3 when "00011",4 when "00100",5 when "00101",
>                                         6 when "00110",7 when "00111",8 when "01000",9 when "01001",10 when "01010",11 when "01011",
>                                         12 when "01100",13 when "01101",14 when "01110",15 when "01111",16 when "10000",17 when "10001",
>                                         18 when "10010",19 when "10011",20 when "10100",21 when "10101",22 when "10110",23 when "10111",
>                                         24 when "11000",25 when "11001",26 when "11010",27 when "11011",28 when "11100",29 when "11101",
>                                         30 when "11110",31 when others;
>
>         process (S_nat)
>                 begin
>         OP3(S_nat) <= '1';
>         end process;

Better:

process (Rddata2, Sa)
begin
OP3 <= RdData2; -- initailise all bits of
OP3 with RdData2
OP3(to_integer(unsigned(Sa))) <= '1'; -- modify bit selected
with Sa
end process;

This code eliminates the fact that two process drive the same signal,
is by far more readable and less errorprone than the code above.
It is important that if you use some branches, each branch defines all
bits of signal even if you would use OP3 <= OP3; OP3(Sa) <= '1'.

bye Thomas