Subversion Repositories raytrac

------------------------------------------------

--! @file ema32x2.vhd

--! @brief RayTrac Floating Point Adder  

--! @author Julián Andrés Guarín Reyes

--------------------------------------------------

-- RAYTRAC (FP BRANCH)

-- Author Julian Andres Guarin

-- ema32x2.vhd

-- This file is part of raytrac.

-- 

--     raytrac is free software: you can redistribute it and/or modify

--     it under the terms of the GNU General Public License as published by

--     the Free Software Foundation, either version 3 of the License, or

--     (at your option) any later version.

-- 

--     raytrac is distributed in the hope that it will be useful,

--     but WITHOUT ANY WARRANTY; without even the implied warranty of

--     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the

--     GNU General Public License for more details.

-- 

--     You should have received a copy of the GNU General Public License

--     along with raytrac.  If not, see <http://www.gnu.org/licenses/>

library ieee;

use ieee.std_logic_1164.all;

use ieee.std_logic_unsigned.all;

use ieee.std_logic_arith.all;

--! Esta entidad recibe dos n&uacutemeros en formato punto flotante IEEE 754, de precision simple y devuelve las mantissas signadas y corridas, y el exponente correspondiente al resultado antes de normalizarlo al formato float. 

--!\nLas 2 mantissas y el exponente entran despues a la entidad add2 que suma las mantissas y entrega el resultado en formato IEEE 754.

entity ema32x2 is

        port (

                clk,dpc         : in std_logic;

                a32,b32         : in std_logic_vector (31 downto 0);

                res32           : out std_logic_vector(31 downto 0)

        );

end ema32x2;

architecture ema32x2_arch of ema32x2 is

        component lpm_mult

        generic (

                lpm_hint                        : string;

                lpm_representation      : string;

                lpm_type                        : string;

                lpm_widtha                      : natural;

                lpm_widthb                      : natural;

                lpm_widthp                      : natural

        );

        port (

                dataa   : in std_logic_vector ( lpm_widtha-1 downto 0 );

                datab   : in std_logic_vector ( lpm_widthb-1 downto 0 );

                result  : out std_logic_vector( lpm_widthp-1 downto 0 )

        );

        end component;

        signal s2slr                                                                            : std_logic_vector(1 downto 0);

        signal s3lshift,s4lshift                                                        : std_logic_vector(4 downto 0);

        signal s2exp,s3exp,s4exp                                                        : std_logic_vector(7 downto 0);

        signal s4slab                                                                           : std_logic_vector(15 downto 0);

        signal s2slab                                                                           : std_logic_vector(16 downto 0);

        signal b1s,s4nrmP                                                                       : std_logic_vector(22 downto 0); -- Inversor de la mantissa

        signal s0a,s0b,s1a,s1b                                                          : std_logic_vector(31 downto 0); -- Float 32 bit 

        signal s1sma,s2sma,s2smb,s3sma,s3smb,s3ures,s4ures      : std_logic_vector(24 downto 0); -- Signed mantissas

        signal s3res                                                                            : std_logic_vector(25 downto 0); -- Signed mantissa result

        signal s1pS,s1pH,s1pL,s4nrmL,s4nrmH,s4nrmS                      : std_logic_vector(17 downto 0); -- Shifert Product

        signal s0zeroa,s0zerob,s1z,s4sgr                                : std_logic;

begin

        process (clk)

        begin

                if clk'event and clk='1' then

                        --!Registro de entrada

                        s0a <= a32;

                        s0b(31) <= dpc xor b32(31);     --! Importante: Integrar el signo en el operando B

                        s0b(30 downto 0) <= b32(30 downto 0);

                        --!Etapa 0,Escoger el mayor exponente que sera el resultado desnormalizado, calcula cuanto debe ser el corrimiento de la mantissa con menor exponente y reorganiza los operandos, si el mayor es b, intercambia las posición si el mayor es a las posiciones la mantiene. Zero check.

                        if s0a(30 downto 23) >= s0b (30 downto 23) then

                                --!signo,exponente,mantissa

                                s1b(31) <= s0b(31);

                                s1b(30 downto 23) <= s0a(30 downto 23)-s0b(30 downto 23);

                                s1b(22 downto 0) <= s0b(22 downto 0);

                                --! zero signaling

                                s1z <= s0zerob;

                                --!clasifica a

                                s1a <= s0a;

                        else

                                --!signo,exponente,mantissa

                                s1b(31) <= s0a(31);

                                s1b(30 downto 23) <= s0b(30 downto 23)-s0a(30 downto 23);

                                s1b(22 downto 0) <= s0a(22 downto 0);

                                --! zero signaling

                                s1z <= s0zeroa;

                                --!clasifica b

                                s1a <= s0b;

                        end if;

                        --! Etapa 1: Denormalización de las mantissas.  

                        --! A

                        s2exp <= s1a(30 downto 23);

                        s2sma <= s1sma;

                        --! B

                        for i in 23 downto 15 loop

                                s2smb(i)        <= s1pL(23-i) xor s1b(31);

                        end loop;

                        for i in 14 downto 6 loop

                                s2smb(i)        <= (s1pH(14-i) or s1pL(14-i+9)) xor s1b(31);

                        end loop;

                        for i in 5 downto 0 loop

                                s2smb(i)        <= (s1pS(5-i) or s1pH(5-i+9)) xor s1b(31);

                        end loop;

                        if s1b(30 downto 28)>"000" then

                                s2slr <= "11";

                        else

                                s2slr <= s1b(27 downto 26);

                        end if;

                        s2smb(24) <= s1b(31);

                        --! Etapa2: Finalizar la denormalización de b.

                        --! A

                        s3sma <= s2sma;

                        s3exp <= s2exp;

                        --! B

                        case (s2slr) is

                                when "00" =>

                                        s3smb   <= s2smb(24 downto 0)+s2smb(24);

                                when "01" =>

                                        s3smb   <= ( s2slab(8 downto 0) & s2smb(23 downto 8) ) + s2smb(24);

                                when "10"  =>

                                        s3smb   <= ( s2slab(16 downto 0) & s2smb(23 downto 16)) + s2smb(24);

                                when others =>

                                        s3smb   <= (others => '0');

                        end case;

                        --! Etapa 3: Etapa 3 Realizar la suma, quitar el signo de la mantissa y codificar el corrimiento hacia la izquierda.

                        s4ures  <= s3ures+s3res(25);                            --Resultado no signado

                        s4sgr   <= s3res(25);                                           --Signo

                        s4exp   <= s3exp;                                                       --Exponente 

                        s4lshift <= s3lshift;                                           --Corrimiento hacia la izquierda. 

                        --! Etapa 4: Corrimiento y normalización de la mantissa resultado.

                        res32(31) <= s4sgr;

                        if s4ures(24)='1' then

                                res32(22 downto 0) <= s4ures(23 downto 1);

                                res32(30 downto 23) <= s4exp+1;

                        else

                                case s4lshift(4 downto 3) is

                                        when "00" =>

                                                res32(22 downto 0)       <= s4nrmP(22 downto 0);

                                                res32(30 downto 23) <= s4exp - s4lshift;

                                        when "01" =>

                                                res32(22 downto 0)       <= s4nrmP(14 downto 0)   & s4slab(7 downto 0);

                                                res32(30 downto 23) <= s4exp - s4lshift;

                                        when "10" =>

                                                res32(22 downto 0)       <= s4nrmP(6 downto 0)    & s4slab(15 downto 0);

                                                res32(30 downto 23) <= s4exp - s4lshift;

                                        when others =>

                                                res32(30 downto 0) <= (others => '0');

                                end case;

                        end if;

                end if;

        end process;

        --! Combinatorial gremlin, Etapa 0, Escoger el mayor exponente que sera el resultado desnormalizado,\n 

        --! calcula cuanto debe ser el corrimiento de la mantissa con menor exponente y reorganiza los operandos,\n

        --! si el mayor es b, intercambia las posición si el mayor es a las posiciones la mantiene. Zero check.\n 

        process (s0b(30 downto 23),s0a(30 downto 23))

        begin

                s0zerob <='0';

                s0zeroa <='0';

                for i in 30 downto 23 loop

                        if s0a(i)='1' then

                                s0zeroa <= '1';

                        end if;

                        if s0b(i)='1' then

                                s0zerob <='1';

                        end if;

                end loop;

        end process;

        --! Combinatorial Gremlin, Etapa 1 Denormalización de las mantissas. 

        denormsupershiftermult:lpm_mult

        generic map ("DEDICATED_MULTIPLIER_CIRCUITRY=YES,MAXIMIZE_SPEED=9","UNSIGNED","LPM_MULT",9,9,18)

        port    map ("00"&shl(conv_std_logic_vector(1,7),s1b(25 downto 23)),conv_std_logic_vector(0,3)&b1s(22 downto 17),s1pS);

        denormhighshiftermult:lpm_mult

        generic map ("DEDICATED_MULTIPLIER_CIRCUITRY=YES,MAXIMIZE_SPEED=9","UNSIGNED","LPM_MULT",9,9,18)

        port    map ("00"&shl(conv_std_logic_vector(1,7),s1b(25 downto 23)),b1s(16 downto 8),s1pH);

        denormlowshiftermult:lpm_mult

        generic map ("DEDICATED_MULTIPLIER_CIRCUITRY=YES,MAXIMIZE_SPEED=9","UNSIGNED","LPM_MULT",9,9,18)

        port    map ("00"&shl(conv_std_logic_vector(1,7),s1b(25 downto 23)),b1s(7 downto 0)&s1z,s1pL);

        s1b2b1s:

        for i in 22 downto 0 generate

                b1s(i) <= s1b(22-i);

        end generate s1b2b1s;

        signa:

        for i in 22 downto 0 generate

                s1sma(i) <= s1a(31) xor s1a(i);

        end generate;

        s1sma(23) <= not(s1a(31));

        s1sma(24) <= s1a(31);

        --! Combinatorial Gremlin, Etapa2: Finalizar la denormalización de b.

        s2signslab:

        for i in 16 downto 0 generate

                s2slab(i) <= s2smb(24);

        end generate s2signslab;

        --! Combinatorial Gremlin, Etapa 3 Realizar la suma, quitar el signo de la mantissa y codificar el corrimiento hacia la izquierda. 

        --adder:sadd2

        --port map (s3sma(24)&s3sma,s3smb(24)&s3smb,dpc,s3res);

        process (s3sma,s3smb)

        begin

                --! Magia: La suma ocurre aqui

                s3res <= (s3sma(24)&s3sma)+(s3smb(24)&s3smb);

        end process;

        process(s3res)

                variable lshift : integer range 24 downto 0;

        begin

                lshift:=24;

                for i in 0 to 23 loop

                        s3ures(i) <= s3res(25) xor s3res(i);

                        if (s3res(25) xor s3res(i))='1' then

                                lshift:=23-i;

                        end if;

                end loop;

                s3ures(24) <= s3res(24) xor s3res(25);

                s3lshift <= conv_std_logic_vector(lshift,5);

        end process;

        --! Combinatorial Gremlin, Etapa 4 corrimientos y normalización de la mantissa resultado.

        normsupershiftermult:lpm_mult

        generic map ("DEDICATED_MULTIPLIER_CIRCUITRY=YES,MAXIMIZE_SPEED=9","UNSIGNED","LPM_MULT",9,9,18)

        port    map (shl(conv_std_logic_vector(1,9),s4lshift(2 downto 0)),s4ures(22 downto 14),s4nrmS);

        normhighshiftermult:lpm_mult

        generic map ("DEDICATED_MULTIPLIER_CIRCUITRY=YES,MAXIMIZE_SPEED=9","UNSIGNED","LPM_MULT",9,9,18)

        port    map (shl(conv_std_logic_vector(1,9),s4lshift(2 downto 0)),s4ures(13 downto 5),s4nrmH);

        normlowshiftermult:lpm_mult

        generic map ("DEDICATED_MULTIPLIER_CIRCUITRY=YES,MAXIMIZE_SPEED=9","UNSIGNED","LPM_MULT",9,9,18)

        port    map (shl(conv_std_logic_vector(1,9),s4lshift(2 downto 0)),s4ures(4 downto 0)&conv_std_logic_vector(0,4),s4nrmL);

        process (s4nrmS,s4nrmH,s4nrmL)

        begin

                s4nrmP(22 downto 14) <= s4nrmS(8 downto 0) or s4nrmH(17 downto 9);

                s4nrmP(13 downto 5) <= s4nrmH(8 downto 0) or s4nrmL(17 downto 9);

                s4nrmP(4 downto 0) <= s4nrmL(8 downto 4);

        end process;

        s4signslab:

        for i in 15 downto 0 generate

                s4slab(i) <= '0';

        end generate s4signslab;

end ema32x2_arch;