how to write VHDL for shifting?

[walala]

For example,

T2:=CONV_INTEGER(X)*(2**5)-CONV_INTEGER(X)*(2**3)+CONV_INTEGER(X)

how to use shifting to represent it?

thanks a lot,

-Walala

 

[Egbert Molenkamp]

"walala" <mizhael@yahoo.com> schreef in bericht
news:bphc5c$om4$1@mozo.cc.purdue.edu...

For example,

T2:=CONV_INTEGER(X)*(2**5)-CONV_INTEGER(X)*(2**3)+CONV_INTEGER(X)

how to use shifting to represent it?

thanks a lot,

-Walala

Maybe your synthesis tool can handle this! You want to multiply X with the
constant 41.
Try this

LIBRARY ieee;
USE ieee.std_logic_1164.ALL;
USE ieee.numeric_std.ALL;
ENTITY mul IS
PORT (x : IN unsigned(7 downto 0);
y : OUT unsigned(15 downto 0));
END mul;

ARCHITECTURE behaviour OF mul IS
SIGNAL qi : std_logic;
BEGIN
y <= 41 * x;
END behaviour;

Egbert Molenkamp

 

[MM]

"walala" <mizhael@yahoo.com> wrote in message
news:bphc5c$om4$1@mozo.cc.purdue.edu...

For example,

T2:=CONV_INTEGER(X)*(2**5)-CONV_INTEGER(X)*(2**3)+CONV_INTEGER(X)

how to use shifting to represent it?

signal foo: std_logic_vector(7 downto 0);
signal shifted_by_2bits: std_logic_vector(7 downto 0);

shifted_by_2bits <= "00" & foo(7 downto 2);

Remember about the sign extension if you do signed arithmetic.

There are standard shift functions as well...


/Mikhail

 

[Brent Hayhoe]

MM wrote:

"walala" <mizhael@yahoo.com> wrote in message
news:bphc5c$om4$1@mozo.cc.purdue.edu...

For example,

T2:=CONV_INTEGER(X)*(2**5)-CONV_INTEGER(X)*(2**3)+CONV_INTEGER(X)

how to use shifting to represent it?

signal foo: std_logic_vector(7 downto 0);
signal shifted_by_2bits: std_logic_vector(7 downto 0);

shifted_by_2bits <= "00" & foo(7 downto 2);

Remember about the sign extension if you do signed arithmetic.

There are standard shift functions as well...

/Mikhail

Consider using the IEEE Numeric_Std package and defining your signals as either
Signed or Unsigned types dependant on the arithmetic that you are performing.
All the operator overloading is then predefined for you and you don't need to
concern yourself with sign extension when using Signed... it's handled by the
package functions.

N.B. integer conversion then becomes the 'To_Integer' function.

In general this should lead to more comprehensible code, and by inference, more
maintainable code.

Believe me, I have suffered the nightmare of converting a huge DSP design to
Numeric_Std from the Synopsys Std_Logic_Arith and it's associated packages.
Trying to work out when the designer switches from Unsigned to Signed numbers,
when all the signals/variables are defined as Std_Logic_Vectors is not as easy
as you might think!

I hope these packages are not standardized into the IEEE library!

--

Regards,

Brent Hayhoe.

Aftonroy Limited
Email: <A
HREF="mailto:&#066;&#114;&#101;&#110;&#116;&#046;&#072;&#097;&#121;&#104;&#111;&#101;&#064;&#065;&#102;&#116;&#111;&#110;&#114;&#111;&#121;&#046;&#099;&#111;&#109;">