URL https://opencores.org/ocsvn/raytrac/raytrac/trunk

Subversion Repositories raytrac

[/] [raytrac/] [branches/] [fp_sgdma/] [ap_n_dpc.vhd] - Blame information for rev 256

Go to most recent revision | Details | Compare with Previous | View Log


--! @file ap_n_dpc.vhd
--! @brief Decodificador de operaci&eacute;n. Sistema de decodificación de los \kdatapaths, cuyo objetivo es a partir del par&acute;ametro de entrada DCS.\nSon 4 las posibles configuraciones de \kdatapaths que existen. Los valores de los bits DC son los que determinan y decodifican la interconexi&oacute;n entre los componentes aritm&eacute;ticos. El componente S determina el signo de la operaci&oacute;n cuando es una suma la que operaci&oacute;n se es&eacutea; ejecutando en el momento.  
--! @author Juli&aacute;n Andr&eacute;s Guar&iacute;n Reyes
--------------------------------------------------------------
-- RAYTRAC
-- Author Julian Andres Guarin
-- ap_n_dpc.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 work.arithpack.all;
 
library altera_mf;
use altera_mf.altera_mf_components.all;
 
 
entity ap_n_dpc is
 
        port (
 
                p0,p1,p2,p3,p4,p5,p6,p7,p8: out std_logic_vector(31 downto 0);
 
 
                clk                                             : in    std_logic;
                rst                                             : in    std_logic;
 
                ax                                              : in    std_logic_vector(31 downto 0);
                ay                                              : in    std_logic_vector(31 downto 0);
                az                                              : in    std_logic_vector(31 downto 0);
                bx                                              : in    std_logic_vector(31 downto 0);
                by                                              : in    std_logic_vector(31 downto 0);
                bz                                              : in    std_logic_vector(31 downto 0);
                vx                                              : out   std_logic_vector(31 downto 0);
                vy                                              : out   std_logic_vector(31 downto 0);
                vz                                              : out   std_logic_vector(31 downto 0);
                sc                                              : out   std_logic_vector(31 downto 0);
                ack                                             : in    std_logic;
                empty                                   : out   std_logic;
                sign_switcheroo                 : in            std_logic;
 
                 --paraminput                           : in    vectorblock06;  --! Vectores A,B
 
                dcs                                             : in    std_logic_vector(2 downto 0);            --! Bit con el identificador del bloque AB vs CD e identificador del sub bloque (A/B) o (C/D). 
 
                sync_chain_1                    : in    std_logic;              --! Se&ntilde;al de dato valido que se va por toda la cadena de sincronizacion.
                pipeline_pending                : out   std_logic               --! Se&ntilde;al para indicar si hay datos en el pipeline aritm&eacute;tico.    
 
 
 
                --qresult_d                             : out   vectorblock04   --! 4 salidas de resultados, pues lo m&aacute;ximo que podr&aacute; calcularse por cada clock son 2 vectores. 
 
 
 
        );
end entity;
 
architecture ap_n_dpc_arch of ap_n_dpc is
        --!Constantes de apoyo
        constant ssync_chain_max : integer :=32;
        constant ssync_chain_min : integer :=2;
 
        --! Tunnning delay
        constant adder2_delay: integer := 1;
        constant adder1_delay : integer := 1;
 
        --!TBXSTART:FACTORS_N_ADDENDS
        signal sfactor0 : std_logic_vector(31 downto 0);
        signal sfactor1 : std_logic_vector(31 downto 0);
        signal sfactor2 : std_logic_vector(31 downto 0);
        signal sfactor3 : std_logic_vector(31 downto 0);
        signal sfactor4 : std_logic_vector(31 downto 0);
        signal sfactor5 : std_logic_vector(31 downto 0);
        signal sfactor6 : std_logic_vector(31 downto 0);
        signal sfactor7 : std_logic_vector(31 downto 0);
        signal sfactor8 : std_logic_vector(31 downto 0);
        signal sfactor9 : std_logic_vector(31 downto 0);
        signal sfactor10        : std_logic_vector(31 downto 0);
        signal sfactor11        : std_logic_vector(31 downto 0);
        --signal sfactor                : vectorblock12;
 
        signal ssumando0        : std_logic_vector(31 downto 0);
        signal ssumando1        : std_logic_vector(31 downto 0);
        signal ssumando2        : std_logic_vector(31 downto 0);
        signal ssumando3        : std_logic_vector(31 downto 0);
        signal ssumando4        : std_logic_vector(31 downto 0);
        signal ssumando5        : std_logic_vector(31 downto 0);
        --signal ssumando               : vectorblock06;
 
        signal sq0_q            : std_logic_vector(31 downto 0);
        --!TBXEND
 
 
        --!TBXSTART:ARITHMETIC_RESULTS
 
        signal sp0                      : std_logic_vector(31 downto 0);
        signal sp1                      : std_logic_vector(31 downto 0);
        signal sp2                      : std_logic_vector(31 downto 0);
        signal sp3                      : std_logic_vector(31 downto 0);
        signal sp4                      : std_logic_vector(31 downto 0);
        signal sp5                      : std_logic_vector(31 downto 0);
        --signal sprd32blk      : vectorblock06;
 
        signal sa0                      : std_logic_vector(31 downto 0);
        signal sa1                      : std_logic_vector(31 downto 0);
        signal sa2                      : std_logic_vector(31 downto 0);
 
        --signal sadd32blk      : vectorblock03;
 
        signal ssq32    : std_logic_vector(31 downto 0);
        signal sinv32   : std_logic_vector(31 downto 0);
 
        signal sqx_q            : std_logic_vector(31 downto 0);
        signal sqy_q            : std_logic_vector(31 downto 0);
        signal sqz_q            : std_logic_vector(31 downto 0);
        --signal sqxyz_q                : vectorblock03;
 
        signal sq1_e            : std_logic;
        --!TBXEND
 
 
        --!TBXSTART:SYNC_CHAIN
        signal ssync_chain      : std_logic_vector(ssync_chain_max downto ssync_chain_min);
        --!TBXEND
 
        --signal qxyzd          : std_logic_vector(95 downto 0);
 
        --signal qxyzq          : std_logic_vector(95 downto 0);
 
        signal sq2_d            : std_logic_vector(31 downto 0);
        signal sq2_q            : std_logic_vector(31 downto 0);
        signal sq2_w            : std_logic;
        signal sq2_e            : std_logic;
 
        signal sqr_e            : std_logic;
        signal sqr_w            : std_logic;            --! Salidas de escritura y lectura en las colas de resultados.
        signal sqr_dx           : std_logic_vector(31 downto 0);
        signal sqr_dy           : std_logic_vector(31 downto 0);
        signal sqr_dz           : std_logic_vector(31 downto 0);
        signal sqr_dsc          : std_logic_vector(31 downto 0);
 
 
 
        signal sa0o                     : std_logic_vector(31 downto 0);
        signal sa1o                     : std_logic_vector(31 downto 0);
        signal sa2o                     : std_logic_vector(31 downto 0);
        --signal sadd32blko     : vectorblock03;        --! Salidas de los 3 sumadores.
 
        signal sp0o                     : std_logic_vector(31 downto 0);
        signal sp1o                     : std_logic_vector(31 downto 0);
        signal sp2o                     : std_logic_vector(31 downto 0);
        signal sp3o                     : std_logic_vector(31 downto 0);
        signal sp4o                     : std_logic_vector(31 downto 0);
        signal sp5o                     : std_logic_vector(31 downto 0);
        --signal sprd32blko     : vectorblock06;        --! Salidas de los 6 multiplicadores.
 
        signal sinv32o  : std_logic_vector(31 downto 0);         --! Salidas de la raiz cuadradas y el inversor.
        signal ssq32o   : std_logic_vector(31 downto 0);         --! Salidas de la raiz cuadradas y el inversor.
 
        --! Bloque Aritmetico de Sumadores y Multiplicadores (madd)
        component arithblock
        port (
 
                clk     : in std_logic;
                rst : in std_logic;
 
                sign            : in std_logic;
                sign_switch     : in std_logic;
 
                factor0         : in std_logic_vector(31 downto 0);
                factor1         : in std_logic_vector(31 downto 0);
                factor2         : in std_logic_vector(31 downto 0);
                factor3         : in std_logic_vector(31 downto 0);
                factor4         : in std_logic_vector(31 downto 0);
                factor5         : in std_logic_vector(31 downto 0);
                factor6         : in std_logic_vector(31 downto 0);
                factor7         : in std_logic_vector(31 downto 0);
                factor8         : in std_logic_vector(31 downto 0);
                factor9         : in std_logic_vector(31 downto 0);
                factor10        : in std_logic_vector(31 downto 0);
                factor11        : in std_logic_vector(31 downto 0);
                --prd32blki     : in vectorblock06;
 
                sumando0        : in std_logic_vector(31 downto 0);
                sumando1        : in std_logic_vector(31 downto 0);
                sumando2        : in std_logic_vector(31 downto 0);
                sumando3        : in std_logic_vector(31 downto 0);
                sumando4        : in std_logic_vector(31 downto 0);
                sumando5        : in std_logic_vector(31 downto 0);
                --add32blki     : in vectorblock06;
 
                a0                      : out std_logic_vector(31 downto 0);
                a1                      : out std_logic_vector(31 downto 0);
                a2                      : out std_logic_vector(31 downto 0);
                --add32blko     : out vectorblock03;
 
                p0                      : out std_logic_vector(31 downto 0);
                p1                      : out std_logic_vector(31 downto 0);
                p2                      : out std_logic_vector(31 downto 0);
                p3                      : out std_logic_vector(31 downto 0);
                p4                      : out std_logic_vector(31 downto 0);
                p5                      : out std_logic_vector(31 downto 0);
                --prd32blko     : out vectorblock06;
 
                sq32o           : out std_logic_vector(31 downto 0);
                inv32o          : out std_logic_vector(31 downto 0)
 
        );
        end component;
 
begin
 
        --! Bloque Aritm&eacute;tico
        ap : arithblock
        port map (
                clk             => clk,
                rst                     => rst,
 
                sign            => dcs(0),
 
                factor0 =>sfactor0,
                factor1 =>sfactor1,
                factor2 =>sfactor2,
                factor3 =>sfactor3,
                factor4 =>sfactor4,
                factor5 =>sfactor5,
                factor6 =>sfactor6,
                factor7 =>sfactor7,
                factor8 =>sfactor8,
                factor9 =>sfactor9,
                factor10=>sfactor10,
                factor11=>sfactor11,
                --prd32blki     => sfactor,
 
                sumando0=>ssumando0,
                sumando1=>ssumando1,
                sumando2=>ssumando2,
                sumando3=>ssumando3,
                sumando4=>ssumando4,
                sumando5=>ssumando5,
                --add32blki     => ssumando,
 
                a0=>sa0o,
                a1=>sa1o,
                a2=>sa2o,
                --add32blko     => sadd32blko, 
 
                p0=>sp0o,
                p1=>sp1o,
                p2=>sp2o,
                p3=>sp3o,
                p4=>sp4o,
                p5=>sp5o,
                --prd32blko     => sprd32blko,
 
                sq32o=> ssq32o,
                inv32o=> sinv32o
        );
 
        --! Cadena de sincronizaci&oacute;n: 29 posiciones.
        pipeline_pending <= sync_chain_1 or not(sq2_e) or not(sq1_e) or not(sqr_e);
        empty <= sqr_e;
        sync_chain_proc:
        process(clk,rst,sync_chain_1)
        begin
                if rst=rstMasterValue then
                        ssync_chain(ssync_chain_max downto ssync_chain_min) <= (others => '0');
 
                        p0 <= (others => '0');
                        p1 <= (others => '0');
                        p2 <= (others => '0');
 
                elsif clk'event and clk='1' then
                        for i in ssync_chain_max downto ssync_chain_min+1 loop
                                ssync_chain(i) <= ssync_chain(i-1);
                        end loop;
                        ssync_chain(ssync_chain_min) <= sync_chain_1;
 
                        --! Salida de los multiplicadores p0 p1 p2 
                        if ssync_chain(23)='1' then
                                p0 <= ssq32; -- El resultado quedara consignado en VZ1=BASE+1
                        elsif ssync_chain(28)='1' then
                                p1 <= sq2_q; -- El resultado quedara consignado en VX1=BASE+3
                        elsif ssync_chain(24)='1' then
                                p2 <= sinv32; -- El resutlado quedara consignado en VY1=BASE+2
                                p3 <= sqx_q;
                                p4 <= sqy_q;
                                p5 <= sqz_q;
                        elsif ssync_chain(28)='1' then
                                p6 <= sp3o;
                                p7 <= sp4o;
                                p8 <= sp5o;
                        end if;
 
                end if;
        end process sync_chain_proc;
 
 
 
 
 
        --! El siguiente c&oacute;digo sirve para conectar arreglos a se&ntilde;ales std_logic_1164, son abstracciones de c&oacute;digo tambi&eacute;n, sin embargo se realizan a trav&eacute;s de registros. 
        register_products_outputs:
        process (clk)
        begin
                if clk'event and clk='1' then
                        sp0 <= sp0o;
                        sp1 <= sp1o;
                        sp2 <= sp2o;
                        sp3 <= sp3o;
                        sp4 <= sp4o;
                        sp5 <= sp5o;
                        sa0 <= sa0o;
                        sa1 <= sa1o;
                        sa2 <= sa2o;
                        sinv32 <= sinv32o;
                        ssq32 <= ssq32o;
                end if;
        end process;
 
        --! Decodificaci&oacute;n del Datapath.
        datapathproc:process(dcs,ax,bx,ay,by,az,bz,sinv32,sp0,sp1,sp2,sp3,sp4,sp5,sa0,sa1,sa2,sq0_q,sqx_q,sqy_q,sqz_q,ssync_chain,ssq32,sq2_q)
        begin
 
                case dcs is
                        when "011"  =>
 
                                sq2_w <= '0';
                                sq2_d <= ssq32;
 
                                sfactor0 <= ay;
                                sfactor1 <= bz;
                                sfactor2 <= az;
                                sfactor3 <= by;
                                sfactor4 <= az;
                                sfactor5 <= bx;
                                sfactor6 <= ax;
                                sfactor7 <= bz;
                                sfactor8 <= ax;
                                sfactor9 <= by;
                                sfactor10 <= ay;
                                sfactor11 <= bx;
 
                                ssumando0 <= sp0;
                                ssumando1 <= sp1;
                                ssumando2 <= sp2;
                                ssumando3 <= sp3;
                                ssumando4 <= sp4;
                                ssumando5 <= sp5;
 
                                sqr_dx <= sa0;
                                sqr_dy <= sa1;
                                sqr_dz <= sa2;
 
                                sqr_w <= ssync_chain(13+adder2_delay);
 
                        when"000"|"001" =>
 
                                sq2_w <= '0';
                                sq2_d <= ssq32;
 
                                sfactor0 <= ay;
                                sfactor1 <= bz;
                                sfactor2 <= az;
                                sfactor3 <= by;
                                sfactor4 <= az;
                                sfactor5 <= bx;
                                sfactor6 <= ax;
                                sfactor7 <= bz;
                                sfactor8 <= ax;
                                sfactor9 <= by;
                                sfactor10 <= ay;
                                sfactor11 <= bx;
 
 
                                ssumando0 <= ax;
                                ssumando1 <= bx;
                                ssumando2 <= ay;
                                ssumando3 <= by;
                                ssumando4 <= az;
                                ssumando5 <= bz;
 
                                sqr_dx <= sa0;
                                sqr_dy <= sa1;
                                sqr_dz <= sa2;
 
                                sqr_w <= ssync_chain(9+adder2_delay);
 
                        when"110" |"100" =>
 
 
 
                                sfactor0 <= ax;
                                sfactor1 <= bx;
                                sfactor2 <= ay;
                                sfactor3 <= by;
                                sfactor4 <= az;
                                sfactor5 <= bz;
 
                                sfactor6 <= sinv32;
                                sfactor7 <= sqx_q;
                                sfactor8 <= sinv32;
                                sfactor9 <= sqy_q;
                                sfactor10 <= sinv32;
                                sfactor11 <= sqz_q;
 
 
                                ssumando0 <= sp0;
                                ssumando1 <= sp1;
                                ssumando2 <= sa0;
                                ssumando3 <= sq0_q;
                                ssumando4 <= az;
                                ssumando5 <= bz;
 
                                if dcs(1)='1' then
                                        sq2_d <= ssq32;
                                        sq2_w <= ssync_chain(22+adder1_delay);
                                else
                                        sq2_d <= sa1;
                                        sq2_w <= ssync_chain(21+adder1_delay);
                                end if;
 
                                sqr_dx <= sp3;
                                sqr_dy <= sp4;
                                sqr_dz <= sp5;
 
                                sqr_w <= ssync_chain(27+adder1_delay);
 
                        when others =>
 
                                sq2_w <= '0';
                                sq2_d <= ssq32;
 
                                sfactor0 <= ax;
                                sfactor1 <= bx;
                                sfactor2 <= ay;
                                sfactor3 <= by;
                                sfactor4 <= az;
                                sfactor5 <= bz;
 
                                sfactor6 <= ax;
                                sfactor7 <= bx;
                                sfactor8 <= ay;
                                sfactor9 <= by;
                                sfactor10 <= az;
                                sfactor11 <= bz;
 
                                ssumando0 <= sp0;
                                ssumando1 <= sp1;
                                ssumando2 <= sa0;
                                ssumando3 <= sq0_q;
                                ssumando4 <= az;
                                ssumando5 <= bz;
 
                                sqr_dx <= sp3;
                                sqr_dy <= sp4;
                                sqr_dz <= sp5;
 
                                sqr_w <= ssync_chain(5);
 
                end case;
 
 
 
 
        end process;
 
        --! Colas internas de producto punto, ubicada en el pipe line aritm&eacute;co. Paralelo a los sumadores a0 y a2.  
        q0 : scfifo --! Debe ir registrada la salida.
        generic map (
                allow_rwcycle_when_full => "ON",
                lpm_widthu                              => 4,
                lpm_numwords                    => 16,
                lpm_showahead                   => "ON",
                lpm_width                               => 32,
                overflow_checking               => "ON",
                underflow_checking              => "ON",
                use_eab                                 => "ON"
        )
        port    map (
                sclr            => '0',
                clock           => clk,
                rdreq           => ssync_chain(13),
                wrreq           => ssync_chain(5),
                data            => sp2,
                q                       => sq0_q
        );
        --! Colas internas de producto punto, ubicada en el pipe line aritm&eacute;co. Paralelo a los sumadores a0 y a2.  
        q2 : scfifo --! Debe ir registrada la salida.
        generic map (
                allow_rwcycle_when_full => "ON",
                lpm_widthu                              => 4,
                lpm_numwords                    => 16,
                lpm_showahead                   => "ON",
                lpm_type                                => "SCIFIFO",
                lpm_width                               => 32,
                overflow_checking               => "ON",
                underflow_checking              => "ON",
                use_eab                                 => "ON"
        )
        port map (
                rdreq           => ssync_chain(28),
                sclr            => '0',
                clock           => clk,
                empty           => sq2_e,
                q                       => sqr_dsc,
                wrreq           => sq2_w,
                data            => sq2_d
        );
 
        --! Cola interna de normalizaci&oacute;n de vectores, ubicada entre el pipeline aritm&eacute;tico
        qx : scfifo
        generic map (
                allow_rwcycle_when_full => "ON",
                lpm_widthu                              => 5,
                lpm_numwords                    => 32,
                lpm_showahead                   => "ON",
                lpm_width                               => 32,
                overflow_checking               => "ON",
                underflow_checking              => "ON",
                use_eab                                 => "ON"
        )
        port    map (
                aclr            => '0',
                clock           => clk,
                empty           => sq1_e,
                rdreq           => ssync_chain(23+adder1_delay),
                wrreq           => sync_chain_1,
                data            => ax,
                q                       => sqx_q
        );
        qy : scfifo
        generic map (
                allow_rwcycle_when_full => "ON",
                lpm_widthu                              => 5,
                lpm_numwords                    => 32,
                lpm_showahead                   => "ON",
                lpm_width                               => 32,
                overflow_checking               => "ON",
                underflow_checking              => "ON",
                use_eab                                 => "ON"
        )
        port    map (
                aclr            => '0',
                clock           => clk,
                rdreq           => ssync_chain(23+adder1_delay),
                wrreq           => sync_chain_1,
                data            => ay,
                q                       => sqy_q
        );
        qz : scfifo
        generic map (
                allow_rwcycle_when_full => "ON",
                lpm_widthu                              => 5,
                lpm_numwords                    => 32,
                lpm_showahead                   => "ON",
                lpm_width                               => 32,
                overflow_checking               => "ON",
                underflow_checking              => "ON",
                use_eab                                 => "ON"
        )
        port    map (
                aclr            => '0',
                clock           => clk,
                rdreq           => ssync_chain(23+adder1_delay),
                wrreq           => sync_chain_1,
                data            => az,
                q                       => sqz_q
        );
--!***********************************************************************************************************
--!Q RESULT
--!***********************************************************************************************************
 
        --Colas de resultados
        rx : scfifo
        generic map (
                allow_rwcycle_when_full => "ON",
                lpm_widthu                              => 5,
                lpm_numwords                    => 32,
                lpm_showahead                   => "ON",
                lpm_width                               => 32,
                overflow_checking               => "ON",
                underflow_checking              => "ON",
                use_eab                                 => "ON"
        )
        port    map (
                aclr            => '0',
                clock           => clk,
                empty           => sqr_e,
                rdreq           => ack,
                wrreq           => sqr_w,
                data            => sqr_dx,
                q                       => vx
        );
        ry : scfifo
        generic map (
                allow_rwcycle_when_full => "ON",
                lpm_widthu                              => 5,
                lpm_numwords                    => 32,
                lpm_showahead                   => "ON",
                lpm_width                               => 32,
                overflow_checking               => "ON",
                underflow_checking              => "ON",
                use_eab                                 => "ON"
        )
        port    map (
                aclr            => '0',
                clock           => clk,
                rdreq           => ack,
                wrreq           => sqr_w,
                data            => sqr_dy,
                q                       => vy
        );
        rz : scfifo
        generic map (
                allow_rwcycle_when_full => "ON",
                lpm_widthu                              => 5,
                lpm_numwords                    => 32,
                lpm_showahead                   => "ON",
                lpm_width                               => 32,
                overflow_checking               => "ON",
                underflow_checking              => "ON",
                use_eab                                 => "ON"
        )
        port    map (
                aclr            => '0',
                clock           => clk,
                rdreq           => ack,
                wrreq           => sqr_w,
                data            => sqr_dz,
                q                       => vz
        );
        rsc : scfifo
        generic map (
                allow_rwcycle_when_full => "ON",
                lpm_widthu                              => 5,
                lpm_numwords                    => 32,
                lpm_showahead                   => "ON",
                lpm_width                               => 32,
                overflow_checking               => "ON",
                underflow_checking              => "ON",
                use_eab                                 => "ON"
        )
        port    map (
                aclr            => '0',
                clock           => clk,
                rdreq           => ack,
                wrreq           => sqr_w,
                data            => sqr_dsc,
                q                       => sc
        );
 
 
end architecture;

Line No.	Rev	Author	Line
1	219	jguarin200	`--! @file ap_n_dpc.vhd`
2	196	jguarin200	`--! @brief Decodificador de operacién. Sistema de decodificación de los \kdatapaths, cuyo objetivo es a partir del par´ametro de entrada DCS.\nSon 4 las posibles configuraciones de \kdatapaths que existen. Los valores de los bits DC son los que determinan y decodifican la interconexión entre los componentes aritméticos. El componente S determina el signo de la operación cuando es una suma la que operación se es&eacutea; ejecutando en el momento.`
3	128	jguarin200	`--! @author Julián Andrés Guarín Reyes`
4	122	jguarin200	`--------------------------------------------------------------`
5			`-- RAYTRAC`
6			`-- Author Julian Andres Guarin`
7	219	jguarin200	`-- ap_n_dpc.vhd`
8	122	jguarin200	`-- This file is part of raytrac.`
9			`--`
10			`-- raytrac is free software: you can redistribute it and/or modify`
11			`-- it under the terms of the GNU General Public License as published by`
12			`-- the Free Software Foundation, either version 3 of the License, or`
13			`-- (at your option) any later version.`
14			`--`
15			`-- raytrac is distributed in the hope that it will be useful,`
16			`-- but WITHOUT ANY WARRANTY; without even the implied warranty of`
17			`-- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the`
18			`-- GNU General Public License for more details.`
19			`--`
20			`-- You should have received a copy of the GNU General Public License`
21			`-- along with raytrac. If not, see <http://www.gnu.org/licenses/>.`
22
23			`library ieee;`
24			`use ieee.std_logic_1164.all;`
25	212	jguarin200	`use ieee.std_logic_unsigned.all;`
26	151	jguarin200	`use work.arithpack.all;`
27	134	jguarin200
28	212	jguarin200	`library altera_mf;`
29			`use altera_mf.altera_mf_components.all;`
30	158	jguarin200
31	212	jguarin200
32	219	jguarin200	`entity ap_n_dpc is`
33	152	jguarin200
34	122	jguarin200	`port (`
35	229	jguarin200
36	248	jguarin200	`p0,p1,p2,p3,p4,p5,p6,p7,p8: out std_logic_vector(31 downto 0);`
37	242	jguarin200
38
39	204	jguarin200	`clk : in std_logic;`
40			`rst : in std_logic;`
41
42	229	jguarin200	`ax : in std_logic_vector(31 downto 0);`
43			`ay : in std_logic_vector(31 downto 0);`
44			`az : in std_logic_vector(31 downto 0);`
45			`bx : in std_logic_vector(31 downto 0);`
46			`by : in std_logic_vector(31 downto 0);`
47			`bz : in std_logic_vector(31 downto 0);`
48			`vx : out std_logic_vector(31 downto 0);`
49			`vy : out std_logic_vector(31 downto 0);`
50			`vz : out std_logic_vector(31 downto 0);`
51			`sc : out std_logic_vector(31 downto 0);`
52			`ack : in std_logic;`
53			`empty : out std_logic;`
54	256	jguarin200	`sign_switcheroo : in std_logic;`
55	204	jguarin200
56	229	jguarin200	`--paraminput : in vectorblock06; --! Vectores A,B`
57	204	jguarin200
58	229	jguarin200	`dcs : in std_logic_vector(2 downto 0); --! Bit con el identificador del bloque AB vs CD e identificador del sub bloque (A/B) o (C/D).`
59
60	212	jguarin200	`sync_chain_1 : in std_logic; --! Señal de dato valido que se va por toda la cadena de sincronizacion.`
61	229	jguarin200	`pipeline_pending : out std_logic --! Señal para indicar si hay datos en el pipeline aritmético.`
62	204	jguarin200
63	229	jguarin200
64
65			`--qresult_d : out vectorblock04 --! 4 salidas de resultados, pues lo máximo que podrá calcularse por cada clock son 2 vectores.`
66
67	204	jguarin200
68
69	122	jguarin200	`);`
70	153	jguarin200	`end entity;`
71	122	jguarin200
72	219	jguarin200	`architecture ap_n_dpc_arch of ap_n_dpc is`
73	228	jguarin200	`--!Constantes de apoyo`
74	248	jguarin200	`constant ssync_chain_max : integer :=32;`
75	228	jguarin200	`constant ssync_chain_min : integer :=2;`
76	125	jguarin200
77	235	jguarin200	`--! Tunnning delay`
78			`constant adder2_delay: integer := 1;`
79	242	jguarin200	`constant adder1_delay : integer := 1;`
80	228	jguarin200
81	161	jguarin200	`--!TBXSTART:FACTORS_N_ADDENDS`
82	242	jguarin200	`signal sfactor0 : std_logic_vector(31 downto 0);`
83			`signal sfactor1 : std_logic_vector(31 downto 0);`
84			`signal sfactor2 : std_logic_vector(31 downto 0);`
85			`signal sfactor3 : std_logic_vector(31 downto 0);`
86			`signal sfactor4 : std_logic_vector(31 downto 0);`
87			`signal sfactor5 : std_logic_vector(31 downto 0);`
88			`signal sfactor6 : std_logic_vector(31 downto 0);`
89			`signal sfactor7 : std_logic_vector(31 downto 0);`
90			`signal sfactor8 : std_logic_vector(31 downto 0);`
91			`signal sfactor9 : std_logic_vector(31 downto 0);`
92	229	jguarin200	`signal sfactor10 : std_logic_vector(31 downto 0);`
93			`signal sfactor11 : std_logic_vector(31 downto 0);`
94			`--signal sfactor : vectorblock12;`
95
96			`signal ssumando0 : std_logic_vector(31 downto 0);`
97			`signal ssumando1 : std_logic_vector(31 downto 0);`
98			`signal ssumando2 : std_logic_vector(31 downto 0);`
99			`signal ssumando3 : std_logic_vector(31 downto 0);`
100			`signal ssumando4 : std_logic_vector(31 downto 0);`
101			`signal ssumando5 : std_logic_vector(31 downto 0);`
102			`--signal ssumando : vectorblock06;`
103
104			`signal sq0_q : std_logic_vector(31 downto 0);`
105	161	jguarin200	`--!TBXEND`
106	163	jguarin200
107
108			`--!TBXSTART:ARITHMETIC_RESULTS`
109	229	jguarin200
110			`signal sp0 : std_logic_vector(31 downto 0);`
111			`signal sp1 : std_logic_vector(31 downto 0);`
112			`signal sp2 : std_logic_vector(31 downto 0);`
113			`signal sp3 : std_logic_vector(31 downto 0);`
114			`signal sp4 : std_logic_vector(31 downto 0);`
115			`signal sp5 : std_logic_vector(31 downto 0);`
116			`--signal sprd32blk : vectorblock06;`
117
118			`signal sa0 : std_logic_vector(31 downto 0);`
119			`signal sa1 : std_logic_vector(31 downto 0);`
120			`signal sa2 : std_logic_vector(31 downto 0);`
121
122			`--signal sadd32blk : vectorblock03;`
123
124			`signal ssq32 : std_logic_vector(31 downto 0);`
125			`signal sinv32 : std_logic_vector(31 downto 0);`
126
127			`signal sqx_q : std_logic_vector(31 downto 0);`
128			`signal sqy_q : std_logic_vector(31 downto 0);`
129			`signal sqz_q : std_logic_vector(31 downto 0);`
130			`--signal sqxyz_q : vectorblock03;`
131
132			`signal sq1_e : std_logic;`
133	163	jguarin200	`--!TBXEND`
134
135	160	jguarin200
136			`--!TBXSTART:SYNC_CHAIN`
137	228	jguarin200	`signal ssync_chain : std_logic_vector(ssync_chain_max downto ssync_chain_min);`
138	171	jguarin200	`--!TBXEND`
139	212	jguarin200
140	229	jguarin200	`--signal qxyzd : std_logic_vector(95 downto 0);`
141
142			`--signal qxyzq : std_logic_vector(95 downto 0);`
143
144	228	jguarin200	`signal sq2_d : std_logic_vector(31 downto 0);`
145			`signal sq2_q : std_logic_vector(31 downto 0);`
146			`signal sq2_w : std_logic;`
147			`signal sq2_e : std_logic;`
148	219	jguarin200
149	229	jguarin200	`signal sqr_e : std_logic;`
150			`signal sqr_w : std_logic; --! Salidas de escritura y lectura en las colas de resultados.`
151			`signal sqr_dx : std_logic_vector(31 downto 0);`
152			`signal sqr_dy : std_logic_vector(31 downto 0);`
153			`signal sqr_dz : std_logic_vector(31 downto 0);`
154			`signal sqr_dsc : std_logic_vector(31 downto 0);`
155
156	219	jguarin200
157
158	229	jguarin200	`signal sa0o : std_logic_vector(31 downto 0);`
159			`signal sa1o : std_logic_vector(31 downto 0);`
160			`signal sa2o : std_logic_vector(31 downto 0);`
161			`--signal sadd32blko : vectorblock03; --! Salidas de los 3 sumadores.`
162	163	jguarin200
163	229	jguarin200	`signal sp0o : std_logic_vector(31 downto 0);`
164			`signal sp1o : std_logic_vector(31 downto 0);`
165			`signal sp2o : std_logic_vector(31 downto 0);`
166			`signal sp3o : std_logic_vector(31 downto 0);`
167			`signal sp4o : std_logic_vector(31 downto 0);`
168			`signal sp5o : std_logic_vector(31 downto 0);`
169			`--signal sprd32blko : vectorblock06; --! Salidas de los 6 multiplicadores.`
170
171			`signal sinv32o : std_logic_vector(31 downto 0); --! Salidas de la raiz cuadradas y el inversor.`
172			`signal ssq32o : std_logic_vector(31 downto 0); --! Salidas de la raiz cuadradas y el inversor.`
173
174	219	jguarin200	`--! Bloque Aritmetico de Sumadores y Multiplicadores (madd)`
175			`component arithblock`
176			`port (`
177
178			`clk : in std_logic;`
179			`rst : in std_logic;`
180
181			`sign : in std_logic;`
182	256	jguarin200	`sign_switch : in std_logic;`
183	229	jguarin200
184			`factor0 : in std_logic_vector(31 downto 0);`
185			`factor1 : in std_logic_vector(31 downto 0);`
186			`factor2 : in std_logic_vector(31 downto 0);`
187			`factor3 : in std_logic_vector(31 downto 0);`
188			`factor4 : in std_logic_vector(31 downto 0);`
189			`factor5 : in std_logic_vector(31 downto 0);`
190			`factor6 : in std_logic_vector(31 downto 0);`
191			`factor7 : in std_logic_vector(31 downto 0);`
192			`factor8 : in std_logic_vector(31 downto 0);`
193			`factor9 : in std_logic_vector(31 downto 0);`
194			`factor10 : in std_logic_vector(31 downto 0);`
195			`factor11 : in std_logic_vector(31 downto 0);`
196			`--prd32blki : in vectorblock06;`
197	219	jguarin200
198	229	jguarin200	`sumando0 : in std_logic_vector(31 downto 0);`
199			`sumando1 : in std_logic_vector(31 downto 0);`
200			`sumando2 : in std_logic_vector(31 downto 0);`
201			`sumando3 : in std_logic_vector(31 downto 0);`
202			`sumando4 : in std_logic_vector(31 downto 0);`
203			`sumando5 : in std_logic_vector(31 downto 0);`
204			`--add32blki : in vectorblock06;`
205	219	jguarin200
206	229	jguarin200	`a0 : out std_logic_vector(31 downto 0);`
207			`a1 : out std_logic_vector(31 downto 0);`
208			`a2 : out std_logic_vector(31 downto 0);`
209			`--add32blko : out vectorblock03;`
210	219	jguarin200
211	229	jguarin200	`p0 : out std_logic_vector(31 downto 0);`
212			`p1 : out std_logic_vector(31 downto 0);`
213			`p2 : out std_logic_vector(31 downto 0);`
214			`p3 : out std_logic_vector(31 downto 0);`
215			`p4 : out std_logic_vector(31 downto 0);`
216			`p5 : out std_logic_vector(31 downto 0);`
217			`--prd32blko : out vectorblock06;`
218
219			`sq32o : out std_logic_vector(31 downto 0);`
220			`inv32o : out std_logic_vector(31 downto 0)`
221	219	jguarin200
222			`);`
223			`end component;`
224
225	123	jguarin200	`begin`
226	204	jguarin200
227	219	jguarin200	`--! Bloque Aritmético`
228			`ap : arithblock`
229			`port map (`
230			`clk => clk,`
231			`rst => rst,`
232	204	jguarin200
233	229	jguarin200	`sign => dcs(0),`
234	219	jguarin200
235	229	jguarin200	`factor0 =>sfactor0,`
236			`factor1 =>sfactor1,`
237			`factor2 =>sfactor2,`
238			`factor3 =>sfactor3,`
239			`factor4 =>sfactor4,`
240			`factor5 =>sfactor5,`
241			`factor6 =>sfactor6,`
242			`factor7 =>sfactor7,`
243			`factor8 =>sfactor8,`
244			`factor9 =>sfactor9,`
245			`factor10=>sfactor10,`
246			`factor11=>sfactor11,`
247			`--prd32blki => sfactor,`
248
249			`sumando0=>ssumando0,`
250			`sumando1=>ssumando1,`
251			`sumando2=>ssumando2,`
252			`sumando3=>ssumando3,`
253			`sumando4=>ssumando4,`
254			`sumando5=>ssumando5,`
255			`--add32blki => ssumando,`
256	219	jguarin200
257	229	jguarin200	`a0=>sa0o,`
258			`a1=>sa1o,`
259			`a2=>sa2o,`
260			`--add32blko => sadd32blko,`
261	219	jguarin200
262	229	jguarin200	`p0=>sp0o,`
263			`p1=>sp1o,`
264			`p2=>sp2o,`
265			`p3=>sp3o,`
266			`p4=>sp4o,`
267			`p5=>sp5o,`
268			`--prd32blko => sprd32blko,`
269
270			`sq32o=> ssq32o,`
271			`inv32o=> sinv32o`
272	219	jguarin200	`);`
273	122	jguarin200
274	142	jguarin200	`--! Cadena de sincronización: 29 posiciones.`
275	229	jguarin200	`pipeline_pending <= sync_chain_1 or not(sq2_e) or not(sq1_e) or not(sqr_e);`
276			`empty <= sqr_e;`
277	140	jguarin200	`sync_chain_proc:`
278	212	jguarin200	`process(clk,rst,sync_chain_1)`
279	140	jguarin200	`begin`
280			`if rst=rstMasterValue then`
281	228	jguarin200	`ssync_chain(ssync_chain_max downto ssync_chain_min) <= (others => '0');`
282	230	jguarin200
283	242	jguarin200	`p0 <= (others => '0');`
284			`p1 <= (others => '0');`
285			`p2 <= (others => '0');`
286
287	229	jguarin200	`elsif clk'event and clk='1' then`
288	228	jguarin200	`for i in ssync_chain_max downto ssync_chain_min+1 loop`
289	142	jguarin200	`ssync_chain(i) <= ssync_chain(i-1);`
290	140	jguarin200	`end loop;`
291	228	jguarin200	`ssync_chain(ssync_chain_min) <= sync_chain_1;`
292	242	jguarin200
293			`--! Salida de los multiplicadores p0 p1 p2`
294	248	jguarin200	`if ssync_chain(23)='1' then`
295			`p0 <= ssq32; -- El resultado quedara consignado en VZ1=BASE+1`
296			`elsif ssync_chain(28)='1' then`
297			`p1 <= sq2_q; -- El resultado quedara consignado en VX1=BASE+3`
298			`elsif ssync_chain(24)='1' then`
299			`p2 <= sinv32; -- El resutlado quedara consignado en VY1=BASE+2`
300			`p3 <= sqx_q;`
301			`p4 <= sqy_q;`
302			`p5 <= sqz_q;`
303			`elsif ssync_chain(28)='1' then`
304			`p6 <= sp3o;`
305			`p7 <= sp4o;`
306			`p8 <= sp5o;`
307	242	jguarin200	`end if;`
308
309	140	jguarin200	`end if;`
310			`end process sync_chain_proc;`
311	144	jguarin200
312	163	jguarin200
313	158	jguarin200
314
315	124	jguarin200
316	140	jguarin200	`--! El siguiente código sirve para conectar arreglos a señales std_logic_1164, son abstracciones de código también, sin embargo se realizan a través de registros.`
317			`register_products_outputs:`
318			`process (clk)`
319			`begin`
320			`if clk'event and clk='1' then`
321	229	jguarin200	`sp0 <= sp0o;`
322			`sp1 <= sp1o;`
323			`sp2 <= sp2o;`
324			`sp3 <= sp3o;`
325			`sp4 <= sp4o;`
326			`sp5 <= sp5o;`
327			`sa0 <= sa0o;`
328			`sa1 <= sa1o;`
329			`sa2 <= sa2o;`
330			`sinv32 <= sinv32o;`
331			`ssq32 <= ssq32o;`
332	140	jguarin200	`end if;`
333			`end process;`
334	148	jguarin200
335	196	jguarin200	`--! Decodificación del Datapath.`
336	229	jguarin200	`datapathproc:process(dcs,ax,bx,ay,by,az,bz,sinv32,sp0,sp1,sp2,sp3,sp4,sp5,sa0,sa1,sa2,sq0_q,sqx_q,sqy_q,sqz_q,ssync_chain,ssq32,sq2_q)`
337	196	jguarin200	`begin`
338	229	jguarin200
339			`case dcs is`
340			`when "011" =>`
341
342			`sq2_w <= '0';`
343			`sq2_d <= ssq32;`
344
345			`sfactor0 <= ay;`
346			`sfactor1 <= bz;`
347			`sfactor2 <= az;`
348			`sfactor3 <= by;`
349			`sfactor4 <= az;`
350			`sfactor5 <= bx;`
351			`sfactor6 <= ax;`
352			`sfactor7 <= bz;`
353			`sfactor8 <= ax;`
354			`sfactor9 <= by;`
355			`sfactor10 <= ay;`
356			`sfactor11 <= bx;`
357
358			`ssumando0 <= sp0;`
359			`ssumando1 <= sp1;`
360			`ssumando2 <= sp2;`
361			`ssumando3 <= sp3;`
362			`ssumando4 <= sp4;`
363			`ssumando5 <= sp5;`
364
365			`sqr_dx <= sa0;`
366			`sqr_dy <= sa1;`
367			`sqr_dz <= sa2;`
368
369	230	jguarin200	`sqr_w <= ssync_chain(13+adder2_delay);`
370	229	jguarin200
371			`when"000"\|"001" =>`
372
373			`sq2_w <= '0';`
374			`sq2_d <= ssq32;`
375
376			`sfactor0 <= ay;`
377			`sfactor1 <= bz;`
378			`sfactor2 <= az;`
379			`sfactor3 <= by;`
380			`sfactor4 <= az;`
381			`sfactor5 <= bx;`
382			`sfactor6 <= ax;`
383			`sfactor7 <= bz;`
384			`sfactor8 <= ax;`
385			`sfactor9 <= by;`
386			`sfactor10 <= ay;`
387			`sfactor11 <= bx;`
388
389
390			`ssumando0 <= ax;`
391			`ssumando1 <= bx;`
392			`ssumando2 <= ay;`
393			`ssumando3 <= by;`
394			`ssumando4 <= az;`
395			`ssumando5 <= bz;`
396
397			`sqr_dx <= sa0;`
398			`sqr_dy <= sa1;`
399			`sqr_dz <= sa2;`
400
401	230	jguarin200	`sqr_w <= ssync_chain(9+adder2_delay);`
402	229	jguarin200
403			`when"110" \|"100" =>`
404
405
406
407			`sfactor0 <= ax;`
408			`sfactor1 <= bx;`
409			`sfactor2 <= ay;`
410			`sfactor3 <= by;`
411			`sfactor4 <= az;`
412			`sfactor5 <= bz;`
413
414			`sfactor6 <= sinv32;`
415			`sfactor7 <= sqx_q;`
416			`sfactor8 <= sinv32;`
417			`sfactor9 <= sqy_q;`
418			`sfactor10 <= sinv32;`
419			`sfactor11 <= sqz_q;`
420
421
422			`ssumando0 <= sp0;`
423			`ssumando1 <= sp1;`
424			`ssumando2 <= sa0;`
425			`ssumando3 <= sq0_q;`
426			`ssumando4 <= az;`
427			`ssumando5 <= bz;`
428
429			`if dcs(1)='1' then`
430			`sq2_d <= ssq32;`
431	242	jguarin200	`sq2_w <= ssync_chain(22+adder1_delay);`
432	229	jguarin200	`else`
433			`sq2_d <= sa1;`
434	242	jguarin200	`sq2_w <= ssync_chain(21+adder1_delay);`
435	229	jguarin200	`end if;`
436
437			`sqr_dx <= sp3;`
438			`sqr_dy <= sp4;`
439			`sqr_dz <= sp5;`
440
441	248	jguarin200	`sqr_w <= ssync_chain(27+adder1_delay);`
442	229	jguarin200
443			`when others =>`
444
445			`sq2_w <= '0';`
446			`sq2_d <= ssq32;`
447
448			`sfactor0 <= ax;`
449			`sfactor1 <= bx;`
450			`sfactor2 <= ay;`
451			`sfactor3 <= by;`
452			`sfactor4 <= az;`
453			`sfactor5 <= bz;`
454
455			`sfactor6 <= ax;`
456			`sfactor7 <= bx;`
457			`sfactor8 <= ay;`
458			`sfactor9 <= by;`
459			`sfactor10 <= az;`
460			`sfactor11 <= bz;`
461
462			`ssumando0 <= sp0;`
463			`ssumando1 <= sp1;`
464			`ssumando2 <= sa0;`
465			`ssumando3 <= sq0_q;`
466			`ssumando4 <= az;`
467			`ssumando5 <= bz;`
468
469			`sqr_dx <= sp3;`
470			`sqr_dy <= sp4;`
471			`sqr_dz <= sp5;`
472
473			`sqr_w <= ssync_chain(5);`
474
475			`end case;`
476
477
478
479
480	123	jguarin200	`end process;`
481
482	204	jguarin200	`--! Colas internas de producto punto, ubicada en el pipe line aritméco. Paralelo a los sumadores a0 y a2.`
483			`q0 : scfifo --! Debe ir registrada la salida.`
484			`generic map (`
485	212	jguarin200	`allow_rwcycle_when_full => "ON",`
486	229	jguarin200	`lpm_widthu => 4,`
487			`lpm_numwords => 16,`
488	204	jguarin200	`lpm_showahead => "ON",`
489			`lpm_width => 32,`
490			`overflow_checking => "ON",`
491			`underflow_checking => "ON",`
492	228	jguarin200	`use_eab => "ON"`
493	204	jguarin200	`)`
494			`port map (`
495	212	jguarin200	`sclr => '0',`
496			`clock => clk,`
497	228	jguarin200	`rdreq => ssync_chain(13),`
498	242	jguarin200	`wrreq => ssync_chain(5),`
499	229	jguarin200	`data => sp2,`
500			`q => sq0_q`

Browse

Tools

Subversion Repositories raytrac

[/] [raytrac/] [branches/] [fp_sgdma/] [ap_n_dpc.vhd] - Blame information for rev 256