Subversion Repositories pci

//////////////////////////////////////////////////////////////////////

////                                                              ////

////  File name "pci_user_constants.v"                            ////

////                                                              ////

////  This file is part of the "PCI bridge" project               ////

////  http://www.opencores.org/cores/pci/                         ////

////                                                              ////

////  Author(s):                                                  ////

////      - Miha Dolenc (mihad@opencores.org)                     ////

////      - Tadej Markovic (tadej@opencores.org)                  ////

////                                                              ////

//////////////////////////////////////////////////////////////////////

////                                                              ////

//// Copyright (C) 2000 Miha Dolenc, mihad@opencores.org          ////

////                                                              ////

//// This source file may be used and distributed without         ////

//// restriction provided that this copyright statement is not    ////

//// removed from the file and that any derivative work contains  ////

//// the original copyright notice and the associated disclaimer. ////

////                                                              ////

//// This source file is free software; you can redistribute it   ////

//// and/or modify it under the terms of the GNU Lesser General   ////

//// Public License as published by the Free Software Foundation; ////

//// either version 2.1 of the License, or (at your option) any   ////

//// later version.                                               ////

////                                                              ////

//// This source 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 Lesser General Public License for more ////

//// details.                                                     ////

////                                                              ////

//// You should have received a copy of the GNU Lesser General    ////

//// Public License along with this source; if not, download it   ////

//// from http://www.opencores.org/lgpl.shtml                     ////

////                                                              ////

//////////////////////////////////////////////////////////////////////

//

// CVS Revision History

//

// $Log: not supported by cvs2svn $

// Revision 1.3  2002/08/13 11:03:53  mihad

// Added a few testcases. Repaired wrong reset value for PCI_AM5 register. Repaired Parity Error Detected bit setting. Changed PCI_AM0 to always enabled(regardles of PCI_AM0 define), if image 0 is used as configuration image

//

// Revision 1.2  2002/03/05 11:53:47  mihad

// Added some testcases, removed un-needed fifo signals

//

// Revision 1.1  2002/02/01 14:43:31  mihad

// *** empty log message ***

//

//

// Fifo implementation defines:

// If FPGA and XILINX are defined, Xilinx's BlockSelectRAM+ is instantiated for Fifo storage.

// 16 bit width is used, so 8 bits of address ( 256 ) locations are available. If RAM_DONT_SHARE is not defined (commented out),

// then one block RAM is shared between two FIFOs. That means each Fifo can have a maximum address length of 7 - depth of 128 and only 6 block rams are used

// If RAM_DONT_SHARE is defined ( not commented out ), then 12 block RAMs are used and each Fifo can have a maximum address length of 8 ( 256 locations )

// If FPGA is not defined, then ASIC RAMs are used. Currently there is only one version of ARTISAN RAM supported. User should generate synchronous RAM with

// width of 40 and instantiate it in pci_tpram.v. If RAM_DONT_SHARE is defined, then these can be dual port rams ( write port

// in one clock domain, read in other ), otherwise it must be two port RAM ( read and write ports in both clock domains ).

// If RAM_DONT_SHARE is defined, then all RAM address lengths must be specified accordingly, otherwise there are two relevant lengths - PCI_FIFO_RAM_ADDR_LENGTH and

// WB_FIFO_RAM_ADDR_LENGTH.

`define WBW_ADDR_LENGTH 6

`define WBR_ADDR_LENGTH 4

`define PCIW_ADDR_LENGTH 3

`define PCIR_ADDR_LENGTH 3

//`define FPGA

//`define XILINX

//`define WB_RAM_DONT_SHARE

//`define PCI_RAM_DONT_SHARE

`ifdef FPGA

    `ifdef XILINX

        `define PCI_FIFO_RAM_ADDR_LENGTH 8      // PCI target unit fifo storage definition

        `define WB_FIFO_RAM_ADDR_LENGTH 8       // WB slave unit fifo storage definition

        `define PCI_XILINX_RAMB4

        `define WB_XILINX_RAMB4

        //`define PCI_XILINX_DIST_RAM

        //`define WB_XILINX_DIST_RAM

    `endif

`else

    `define PCI_FIFO_RAM_ADDR_LENGTH 4      // PCI target unit fifo storage definition when RAM sharing is used ( both pcir and pciw fifo use same instance of RAM )

    `define WB_FIFO_RAM_ADDR_LENGTH 7       // WB slave unit fifo storage definition when RAM sharing is used ( both wbr and wbw fifo use same instance of RAM )

//    `define WB_ARTISAN_SDP

//    `define PCI_ARTISAN_SDP

//    `define PCI_VS_STP

//    `define WB_VS_STP

`endif

// these two defines allow user to select active high or low output enables on PCI bus signals, depending on

// output buffers instantiated. Xilinx FPGAs use active low output enables.

`define ACTIVE_LOW_OE

//`define ACTIVE_HIGH_OE

// HOST/GUEST implementation selection - see design document and specification for description of each implementation

// only one can be defined at same time

`define GUEST

//`define HOST

// if NO_CNF_IMAGE is commented out, then READ-ONLY access to configuration space is ENABLED:

// - ENABLED Read-Only access from WISHBONE for GUEST bridges

// - ENABLED Read-Only access from PCI for HOST bridges

// with defining NO_CNF_IMAGE, one decoder and one multiplexer are saved

`define NO_CNF_IMAGE

// number defined here specifies how many MS bits in PCI address are compared with base address, to decode

// accesses. Maximum number allows for minimum image size ( number = 20, image size = 4KB ), minimum number

// allows for maximum image size ( number = 1, image size = 2GB ). If you intend on using different sizes of PCI images,

// you have to define a number of minimum sized image and enlarge others by specifying different address mask.

// smaller the number here, faster the decoder operation

`define PCI_NUM_OF_DEC_ADDR_LINES 20

// no. of PCI Target IMAGES

// - PCI provides 6 base address registers for image implementation.

// PCI_IMAGE1 definition is not required and has no effect, since PCI image 1 is always implemented

// If GUEST is defined, PCI Image 0 is also always implemented and is used for configuration space

// access.

// If HOST is defined and NO_CNF_IMAGE is not, then PCI Image 0 is used for Read Only access to configuration

// space. If HOST is defined and NO_CNF_IMAGE is defined, then user can define PCI_IMAGE0 as normal image, and there

// is no access to Configuration space possible from PCI bus.

// Implementation of all other PCI images is selected by defining PCI_IMAGE2 through PCI_IMAGE5 regardles of HOST

// or GUEST implementation.

`ifdef HOST

    `ifdef NO_CNF_IMAGE

        `define PCI_IMAGE0

    `endif

`endif

`define PCI_IMAGE2

`define PCI_IMAGE3

`define PCI_IMAGE4

`define PCI_IMAGE5

// initial value for PCI image address masks. Address masks can be defined in enabled state,

// to allow device independent software to detect size of image and map base addresses to

// memory space. If initial mask for an image is defined as 0, then device independent software

// won't detect base address implemented and device dependent software will have to configure

// address masks as well as base addresses!

`define PCI_AM0 20'hffff_e

`define PCI_AM1 20'hffff_c

`define PCI_AM2 20'hffff_8

`define PCI_AM3 20'hffff_0

`define PCI_AM4 20'hfffe_0

`define PCI_AM5 20'h0000_0

// initial value for PCI image maping to MEMORY or IO spaces.  If initial define is set to 0,

// then IMAGE with that base address points to MEMORY space, othervise it points ti IO space. D

// Device independent software sets the base addresses acording to MEMORY or IO maping!

`define PCI_BA0_MEM_IO 1'b0 // considered only when PCI_IMAGE0 is used as general PCI-WB image!

`define PCI_BA1_MEM_IO 1'b1

`define PCI_BA2_MEM_IO 1'b0

`define PCI_BA3_MEM_IO 1'b1

`define PCI_BA4_MEM_IO 1'b0

`define PCI_BA5_MEM_IO 1'b1

// number defined here specifies how many MS bits in WB address are compared with base address, to decode

// accesses. Maximum number allows for minimum image size ( number = 20, image size = 4KB ), minimum number

// allows for maximum image size ( number = 1, image size = 2GB ). If you intend on using different sizes of WB images,

// you have to define a number of minimum sized image and enlarge others by specifying different address mask.

// smaller the number here, faster the decoder operation

`define WB_NUM_OF_DEC_ADDR_LINES 20

// no. of WISHBONE Slave IMAGES

// WB image 0 is always used for access to configuration space. In case configuration space access is not implemented,

// ( both GUEST and NO_CNF_IMAGE defined ), then WB image 0 is not implemented. User doesn't need to define image 0.

// WB Image 1 is always implemented and user doesnt need to specify its definition

// WB images' 2 through 5 implementation by defining each one.

//`define WB_IMAGE2

`define WB_IMAGE3

`define WB_IMAGE4

//`define WB_IMAGE5

// If this define is commented out, then address translation will not be implemented.

// addresses will pass through bridge unchanged, regardles of address translation enable bits.

// Address translation also slows down the decoding

//`define ADDR_TRAN_IMPL

// decode speed for WISHBONE definition - initial cycle on WISHBONE bus will take 1 WS for FAST, 2 WSs for MEDIUM and 3 WSs for slow.

// slower decode speed can be used, to provide enough time for address to be decoded.

//`define WB_DECODE_FAST

`define WB_DECODE_MEDIUM

//`define WB_DECODE_SLOW

// Base address for Configuration space access from WB bus. This value cannot be changed during runtime

`define WB_CONFIGURATION_BASE 20'hF300_0

// Turn registered WISHBONE slave outputs on or off

// all outputs from WB Slave state machine are registered, if this is defined - WB bus outputs as well as

// outputs to internals of the core.

//`define REGISTER_WBS_OUTPUTS

/*-----------------------------------------------------------------------------------------------------------

Core speed definition - used for simulation and 66MHz Capable bit value in status register indicating 66MHz

capable device

-----------------------------------------------------------------------------------------------------------*/

`define PCI33

//`define PCI66

/*-----------------------------------------------------------------------------------------------------------

[000h-00Ch] First 4 DWORDs (32-bit) of PCI configuration header - the same regardless of the HEADER type !

        Vendor_ID is an ID for a specific vendor defined by PCI_SIG - 2321h does not belong to anyone (e.g.

        Xilinx's Vendor_ID is 10EEh and Altera's Vendor_ID is 1172h). Device_ID and Revision_ID should be used

        together by application.

-----------------------------------------------------------------------------------------------------------*/

`define HEADER_VENDOR_ID    16'h2321

`define HEADER_DEVICE_ID    16'h0001

`define HEADER_REVISION_ID  8'h01

// Turn registered WISHBONE master outputs on or off

// all outputs from WB Master state machine are registered, if this is defined - WB bus outputs as well as

// outputs to internals of the core.

`define REGISTER_WBM_OUTPUTS

// MAX Retry counter value for WISHBONE Master state-machine

//      This value is 8-bit because of 8-bit retry counter !!!

`define WB_RTY_CNT_MAX                  8'hff