[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [openrisc] Problem about configure the gcc with "--target=or32-uclinux"



Hi,

Checkout these modules from opencores CVS:
# cvs -d :pserver:cvs@cvs.opencores.org:/home/oc/cvs -z9 co -r tn_m001 or1k/binutils
# cvs -d :pserver:cvs@cvs.opencores.org:/home/oc/cvs -z9 co -r tn_m001 or1k/gcc
# cvs -d :pserver:cvs@cvs.opencores.org:/home/oc/cvs -z9 co or1k/newlib-1.10.0
# cvs -d :pserver:cvs@cvs.opencores.org:/home/oc/cvs -z9 co or1k/rtems-20020807

prepare build directory, let's say in or1k directory:
# cd or1k
# mkdir bb

Build binutils:
# mkdir bb/binutils
# cd bb/binutils
# ../../binutils/configure --target=or32-rtems --prefix=/opt/or32-rtems
# make all install

export path:
# export PATH="/opt/or32-rtems/bin:$PATH"

Build gcc:
# mkdir bb/gcc

# cd or1k/gcc
# ln -fs ../newlib-1.10.0/newlib/ .
# ./contrib/egcs-update --touch
# cd ../bb/gcc
# ../../gcc/configure --target=or32-rtems --with-gnu-as --with-gnu-ld --with-newlib --verbose --enable-threads --prefix=/opt/or32-rtems  --enable-languages="c"
# make all install

Build rtems:
# mkdir bb/rtems
# cd bb/rtems
# ../../rtems-20020807/configure --target=or32-rtems
# make all

This should successfully build rtems toolchain and binaries.

Sample programs are then located in bb/rtems/or32-rtems/orp/samples

I attached configuration file for or1ksim that is configured properly to run build samples. Also note, that you have to create uart0.rx and uart0.tx files in a directory you are running or1ksim from (please look inside sim.cfg for details). These are console input/output.
# touch uart0.rx uart0.tx

You have to use or1k/binutils and or1k/gcc with tag tn_m001, and or1k/newlib-1.10.0.
We are switching rtems binary format from coff to elf32-or32 and transition is not complete yet.

regards,
Ivan

xu hu wrote:
20020913143819.75455.qmail@web14601.mail.yahoo.com">
Hi, come on, give me some advice. 

when i set the target to or32-rtems, i have the
toolchain compiled successful, but when i configure
the
rtems, there is another error, or32-rtems-gcc can't
compile correctly, so i write a simple c code, then
compile it with or32-rtems-gcc, ld says " target
elf32-or32 not found"

why? Help me please.

giro.
--- xu hu <giro_cn@yahoo.com> wrote:
Today, i get the source of gcc from the cvs, but
when
i try to compile it, i meet some problem,
"#> ../gcc/configure --target=or32-uclinux ....."
run error.
it says " Invalid configuration 'or32- \
uclinux':system:'uclinux' not recognized.
Unrecognized target system name or32-uclinux"
why?
Any help will be appreciated!

Thanks.

giro.


__________________________________________________
Do you Yahoo!?
Yahoo! News - Today's headlines
http://news.yahoo.com
--
To unsubscribe from openrisc mailing list please
visit http://www.opencores.org/mailinglists.shtml


__________________________________________________
Do you Yahoo!?
Yahoo! News - Today's headlines
http://news.yahoo.com
--
To unsubscribe from openrisc mailing list please visit http://www.opencores.org/mailinglists.shtml


/* sim.cfg -- Simulator configuration script file
   Copyright (C) 2001, Marko Mlinar, markom@opencores.org

This file includes a lot of help about configurations and default one

This file is part of OpenRISC 1000 Architectural Simulator.

This program 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 2 of the License, or
(at your option) any later version.

This program 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 this program; if not, write to the Free Software
Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */


/* INTRODUCTION

   The or1ksim have various parameters, which can be set in configuration
   files.  Multiple configurations may be used and switched between at
   or1ksim startup.
   By default, or1ksim loads condfiguration file from './sim.cfg' and if not
   found it checks '~/.or1k/sim.cfg'. If even this file is not found or
   all parameters are not defined, default configuration is used.
   Users should not rely on default configuration, but rather redefine all
   critical settings, since default configuration may differ in newer
   versions of the or1ksim.
   If multiple configurations are used, user can switch between them by
   supplying -f <filename.cfg> option when starting simulator.

   This file may contain (standard C) only comments - no // support.
   
   Configure files can also be included, using:
   
   include "file_name_to_include"

   Like normal configuration file, this file is divided in sections,
   where each section is described in detail also.

   Some section also have subsections. One example of such subsection is
   block:

   device <index>
     instance specific parameters...
   enddevice

   which creates a device instance.
*/

/* MEMORY SECTION

   This section specifies how is initial memory generated and which blocks
   it consist of.

   type = random/unknown/pattern
      specifies the initial memory values. 'random' parameter generate
      random memory using seed 'random_seed' parameter. 'pattern' parameter
      fills memory with 'pattern' parameter and 'unknown' does not specify
      how memory should be generated - the fastest option.

   random_seed = <value>
      random seed for randomizer, used if type = random

   pattern = <value>
      pattern to fill memory, used if type = pattern

   nmemories = <value>
      number of memory instances connected

   instance specific:
     baseaddr = <hex_value>
        memory start address

     size = <hex_value>
        memory size

     name = "<string>"
        memory block name

     ce = <value>
        chip enable index of the memory instance

     delayr = <value>
        cycles, required for read access, -1 if instance does not support reading

     delayw = <value>
        cycles, required for write access, -1 if instance does not support writing

     16550 = 0/1
        0, if this device is uart 16450 and 1, if it is 16550

     log = "<filename>"
        filename, where to log memory accesses to, no log, if log command is not specified
*/

section memory
  /*random_seed = 12345
  type = random*/
  pattern = 0x00
  type = unknown /* Fastest */

  nmemories = 2
  device 0
    name = "FLASH"
    ce = 0
    baseaddr = 0x00000000
    size = 0x00200000
    delayr = 10
    delayw = -1
  enddevice

  device 1
    name = "RAM"
    ce = 1
    baseaddr = 0x10000000
    size = 0x02000000
    delayr = 2
    delayw = 4
  enddevice
end


/* IMMU SECTION

    This section configures Instruction Memory Menangement Unit

    enabled = 0/1
       whether IMMU is enabled
       (NOTE: UPR bit is set)

    nsets = <value>
       number of ITLB sets; must be power of two

    nways = <value>
       number of ITLB ways

    pagesize = <value>
       instruction page size; must be power of two

    entrysize = <value>
       instruction entry size in bytes

    ustates = <value>
       number of ITLB usage states (2, 3, 4 etc., max is 4)
       
    hitdelay = <value>
       number of cycles immu hit costs
    
    missdelay = <value>
       number of cycles immu miss costs
*/

section immu
  enabled = 0
  nsets = 32
  nways = 1
  pagesize = 8192
  hitdelay = 0
  missdelay = 0
end


/* DMMU SECTION

    This section configures Data Memory Menangement Unit

    enabled = 0/1
       whether DMMU is enabled
       (NOTE: UPR bit is set)

    nsets = <value>
       number of DTLB sets; must be power of two

    nways = <value>
       number of DTLB ways

    pagesize = <value>
       data page size; must be power of two

    entrysize = <value>
       data entry size in bytes

    ustates = <value>
       number of DTLB usage states (2, 3, 4 etc., max is 4)

    hitdelay = <value>
       number of cycles immu hit costs
    
    missdelay = <value>
       number of cycles immu miss costs
*/

section dmmu
  enabled = 0
  nsets = 32
  nways = 1
  pagesize = 8192
  hitdelay = 0
  missdelay = 0
end


/* IC SECTION

   This section configures Instruction Cache

   enabled = 0/1
      whether IC is enabled
      (NOTE: UPR bit is set)

   nsets = <value>
      number of IC sets; must be power of two

   nways = <value>
      number of IC ways

   blocksize = <value>
      IC block size in bytes; must be power of two

   ustates = <value>
      number of IC usage states (2, 3, 4 etc., max is 4)

   hitdelay = <value>
      number of cycles ic hit costs
    
    missdelay = <value>
      number of cycles ic miss costs
*/

section ic
  enabled = 0
  nsets = 512
  nways = 1
  blocksize = 16
  hitdelay = 0
  missdelay = 0
end


/* DC SECTION

   This section configures Data Cache

   enabled = 0/1
      whether DC is enabled
      (NOTE: UPR bit is set)

   nsets = <value>
      number of DC sets; must be power of two

   nways = <value>
      number of DC ways

   blocksize = <value>
      DC block size in bytes; must be power of two

   ustates = <value>
      number of DC usage states (2, 3, 4 etc., max is 4)

   load_hitdelay = <value>
      number of cycles dc load hit costs
   
   load_missdelay = <value>
      number of cycles dc load miss costs
       
   store_hitdelay = <value>
      number of cycles dc load hit costs
    
   store_missdelay = <value>
      number of cycles dc load miss costs
*/

section dc
  enabled = 0
  nsets = 512
  nways = 1
  blocksize = 16
  load_hitdelay = 0
  load_missdelay = 0
  store_hitdelay = 0
  store_missdelay = 0
end

/* SIM SECTION

  This section specifies how should sim behave.

  verbose = 0/1
      whether to print out extra messages

  debug = 0-9
      = 0 disabled debug messages
      1-9 level of sim debug information, greater the number more verbose is
          the output

  profile = 0/1
      whether to generate profiling file 'sim.profile'

  prof_fn = "<filename>"
      filename, where to generate profiling info, used
      only if 'profile' is set
      
  mprofile = 0/1
      whether to generate memory profiling file 'sim.mprofile'

  mprof_fn = "<filename>"
      filename, where to generate memory profiling info, used
      only if 'mprofile' is set

  history = 0/1
      whether instruction execution flow is tracked for
      display by simulator hist command. Useful for
      back-trace debugging.

  iprompt = 0/1
      whether we strart in interactive prompt

  exe_log = 0/1
      whether execution log should be generated
      
  exe_log = default/hardware/simple/software
      type of executed log, default is used if not specified
      
  exe_log_start = <value>
      index of first instruction to start log with, default = 0
        
  exe_log_end = <value>
      index of last instruction to end log with; not limited, if omitted
  
  exe_log_marker = <value>
      <value> specifies number of instructions before horizontal marker is
      printed; if zero, markers are disabled (default)

  exe_log_fn = "<filename>"
      where to put execution log in, used only if 'exe_log'
      is set

  spr_log = 0/1
      whether log of writes/reads to/from sprs should be generated

  spr_log_fn = "<filename>"
      where to put sprs writes/reads in log, used only if 'spr_log'
      is set
      
  clkcycle = <value>[ps|ns|us|ms]
      specifies time measurement for one cycle
*/

section sim
  /* verbose = 1 */
  debug = 4
  profile = 0
  prof_fn = "sim.profile"
  mprofile = 0
  mprof_fn = "sim.mprofile"

  history = 1
  /* iprompt = 0 */
  exe_log = 0
  exe_log_type = hardware
  exe_log_start = 0
  exe_log_end = 0
  exe_log_marker = 0
  exe_log_fn = "executed.log"
  spr_log = 0 
  spr_log_fn = "spr.log"
  clkcycle = 100ns
end


/* SECTION VAPI

    This section configures Verification API, used for Advanced
    Core Verification.

    enabled = 0/1
        whether to start VAPI server

    server_port = <value>
        TCP/IP port to start VAPI server on

    log_enabled = 0/1
       whether logging of VAPI requests is enabled

    hide_device_id = 0/1
       whether to disable logging of device id (for compatability with old version)

    vapi_fn = <filename>
       specifies filename where to log into, if log_enabled is selected
*/

section VAPI
  enabled = 0
  server_port = 9998
  log_enabled = 0
  vapi_log_fn = "vapi.log"
end


/* CPU SECTION

   This section specifies various CPU parameters.

   ver = <value>
   rev = <value>
      specifies version and revision of the CPU used

   upr = <value>
      changes the upr register
      
   sr = <value>
      sets the initial Supervision Register value

   superscalar = 0/1
      whether CPU is scalar or superscalar
      (modify cpu/or32/execute.c to tune superscalar model)

   hazards = 0/1
      whether data hazards are tracked in superscalar CPU
      and displayed by the simulator r command

   dependstats = 0/1
      whether inter-instruction dependencies are calculated
      and displayed by simulator stats command.

   raw_range = <value>
      range in cycles for raw register over time analysis, 0 = disabled
      
   sbuf_len = <value>
      length of store buffer (<= 256), 0 = disabled
*/

section cpu
  ver = 0x1200
  rev = 0x0001
  /* upr = */
  sr = 0x00008003
  superscalar = 0
  hazards = 0
  dependstats = 0
  raw_range = 0
  sbuf_len = 0
end


/* PM SECTION

   This section specifies Power Menagement paramaters
   
   enabled = 0/1
      whether power menagement is enabled
*/

section pm
  enabled = 0
end

/* BPB SECTION

   This section specifies how branch prediction should behave.
   
   enabled = 0/1
      whether bpb is enabled
      
   btic = 0/1
      enable branch target instruction cache model

   sbp_bf_fwd = 0/1
      whether static branch prediction for l.bf uses forward prediction
   
   sbp_bnf_fwd = 0/1
      whether static branch prediction for l.bnf uses forward prediction

   hitdelay = <value>
       number of cycles bpb hit costs
    
   missdelay = <value>
       number of cycles bpb miss costs
*/

section bpb
  enabled = 0
  btic = 0
  sbp_bf_fwd = 0
  sbp_bnf_fwd = 0
  hitdelay = 0
  missdelay = 0
end


/* DEBUG SECTION

   This sections specifies how debug unit should behave.

   enabled = 0/1
      whether debug unit is enabled

   gdb_enabled = 0/1
      whether to start gdb server at 'server_port' port

   server_port = <value>
      TCP/IP port to start gdb server on, used only if gdb_enabled
      is set

   vapi_id = <hex_value>
      Used to create "fake" vapi log file containing the JTAG proxy messages.
*/

section debug
  enabled = 1 
  gdb_enabled = 0
  server_port = 9999
  vapi_id = 0xFFFF
end


/* MC SECTION

   This section configures the memory controller

   enabled = 0/1
      whether memory controller is enabled

   baseaddr = <hex_value>
      address of first MC register

   POC = <hex_value>
      Power On Configuration register
*/

section mc
  enabled = 0
  baseaddr = 0x60000000
  POC = 0x00000008                 /* Power on configuration register */
end


/* UART SECTION

   This section configures UARTs

   nuarts = <value>
      make specified number of instances, configure each
      instance within device - enddevice construct.

   instance specific:
     baseaddr = <hex_value>
        address of first UART register for this device

     rxfile = "<filename>"
        filename, where to read data from

     txfile = "<filename>"
        filename, where to write data to

     irq = <value>
        irq number for this device

     16550 = 0/1
        0, if this device is uart 16450 and 1, if it is 16550

     jitter = <value>
        in msecs... time to block, -1 to disable it

     vapi_id = <hex_value>
        VAPI id of this instance
*/

section uart
  nuarts = 1

  device 0
    baseaddr = 0x80000000
    irq = 2
    rxfile = "uart0.rx"
    txfile = "uart0.tx"
    jitter = -1                     /* async behaviour */
  enddevice
end


/* DMA SECTION

   This section configures DMAs

   ndmas = <value>
      make specified number of instances, configure each
      instance within device - enddevice construct.

   instance specific:
     baseaddr = <hex_value>
        address of first DMA register for this device

     irq = <value>
        irq number for this device

     vapi_id = <hex_value>
        VAPI id of this instance
*/

section dma
  ndmas = 0
/*
  device 0
    baseaddr = 0x90000000
    irq = 4
  enddevice
*/
end


/* ETHERNET SECTION

   This section configures ethernets

   nethernets = <value>
      make specified number of instances, configure each
      instance within device - enddevice construct.

   instance specific:
     baseaddr = <hex_value>
        address of first ethernet register for this device

     dma = <value>
        which controller is this ethernet "connected" to

     irq = <value>
        ethernet mac IRQ level

     rtx_type = <value>
        use 0 - file interface, 1 - socket interface

     rx_channel = <value>
        DMA channel used for RX

     tx_channel = <value>
        DMA channel used for TX

     rxfile = "<filename>"
        filename, where to read data from

     txfile = "<filename>"
        filename, where to write data to

     sockif = "<ifacename>"
        interface name of ethernet socket

     vapi_id = <hex_value>
        VAPI id of this instance
*/

section ethernet
  nethernets = 0

/*
  device 0
    baseaddr = 0x88000000
    dma = 0
    irq = 15
    rtx_type = 0
    tx_channel = 0
    rx_channel = 1
    rxfile = "/tmp/eth0.rx"
    txfile = "/tmp/eth0.tx"
    sockif = "lo"
  enddevice
*/
end

/* GPIO SECTION

   This section configure GPIOs

   ngpios = <value>
      make specified number of instances, configure each
      instance within device - enddevice construct.

   instance specific:
     baseaddr = <hex_value>
        address of first GPIO register for this device

     irq = <value>
        irq number for this device

     base_vapi_id = <hex_value>
        first VAPI id of this instance
	GPIO uses 8 consecutive VAPI IDs
*/

section gpio
  ngpios = 0
/*
  device 0
    baseaddr = 0xA0000000
    irq = 23
    base_vapi_id = 0x0200
  enddevice
*/
end

/* VGA SECTION
    
    This section configures VGA/LCD controller
    
    nvgas = <value>
       number of VGA devices connected
       
    instance specific:
      baseaddr = <hex_value>
        address of first VGA register
        
      irq = <value>
        irq number for this device
        
      refresh_rate = <value>
        number of cycles between screen dumps
        
      filename = "<filename>"
        template name for generated names (e.g. "primary" produces "primary0023.bmp")
*/

section vga
  nvgas = 0
  
/*
  device 0
    baseaddr = 0xb0000000
    irq = 20
    refresh_rate = 100000
    filename = "primary"
  enddevice
*/
end

/* FB SECTION
    
    This section configures frame buffer
    
    enabled = 0/1
      whether frame buffer is enabled
    
    baseaddr = <hex_value>
      base address of frame buffer
        
    paladdr = <hex_value>
      base address of first palette entry
        
    refresh_rate = <value>
      number of cycles between screen dumps
        
    filename = "<filename>"
      template name for generated names (e.g. "primary" produces "primary0023.bmp")
*/

section fb
  enabled = 0
  baseaddr = 0xb8000000
  refresh_rate = 100000
  filename = "primary"
end

/* KBD SECTION

    This section configures PS/2 compatible keyboard
    
    enabled = 0/1
      whether keyboard is enabled
      
    baseaddr = <hex_value>
      base address of the keyboard device
      
    rxfile = "<filename>"
      filename, where to read data from
*/

section kbd
  enabled = 0
  irq = 21
  baseaddr = 0xb1000000
  rxfile = "/tmp/kbd.rx"
end