[okl4-developer] help getting started

[Laborator SRCT]

Nelson Tam wrote:
So my overall advice would be to stick to gumstix, and simulate using the  commands I've suggested.


I'll do that from now on. I'm gonna try to modify gumstix for my ARM11 target  once gumstix works.

Nelson Tam wrote:
We use various distros here, such as Ubuntu and CentOS, but OpenSuse should work fine also.  It really doesn't make a difference as long as you have the correct tool versions.


Than you for the reply Nelson. Yes. that is what I'm trying to verify. That I use the correct tool versions. 

I'm also interested to know if somebody has actually ran the gumstix target up to a succesfull simulation stage starting from the archive found in the wiki.ok-labs.com. If gumstix is buggy then I'm trying to search for better tool versions when there are none available.


So far I went with:

   openSUSE 10.2


   for OKl4 build:
      arm-linux-gcc-3.4.4 from NICTA site's link
      skyeye 121n (from NICTA site's link)
      skyeye 124Rel (from skyeye site, compiled with gcc-4.1.2)
      skyeye 122Rel (from skyeye site, compiled into executable with gcc-3.3.1)
      python2.4.4 (because there is an error with building with python2.5.1 too)

I don't get errors in the build any more,  
build.py machine=gumstix project=iguana wombat=yes 

runs without errors.

However I get errors in the simulation that seem to indicate an error in the build scripts.

If I use skyeye121n (binary from NICTA) I get:
fpga-lin:/download/okl4_release_1.5.2 # skyeye  -c tools/sim_config/gumstix.skyeye -e build/images/image.sim
arch: arm
cpu info: xscale,  pxa25x, 69052100, fffffff0, 2 
mach info: name pxa_lubbock, mach_init addr  0x805f300
SKYEYE: use xscale mmu ops

OKL4 - (provider: Open Kernel Labs) built on Feb  5  2008 13:37:48 using gcc version 3.4.4.
roottask execute pagefault at  7ffc7cd0, ip=7ffc7cd0 - deny
--- KD# roottask pagefault   ---



and if I use skyeye built from sources on my openSuse box:

   for skyeye 122

load section vmlinux.initcall.init: lma =  0xa066d850 (vma = 0x8136d850)  size = 0x000000dc.
load section vmlinux.con_initcall.init: lma =  0xa066d92c (vma = 0x8136d92c)  size = 0x00000004.
not load section  vmlinux.security_initcall.init: addr = 0x8136d930  size = 0x00000000 .
load section  vmlinux.init.ramfs: lma = 0xa066e000 (vma = 0x8136e000)  size =  0x00000086.
load section vmlinux.exitcall: lma = 0xa066f000 (vma  = 0x8136f000)  size = 0x0000004c.
not load section vmlinux.bss: addr = 0x8136f060  size  = 0x000204e8 .
load section rootfs: lma = 0xa0700000 (vma =  0x80600000)  size = 0x007d0000.
load section bootinfo: lma = 0xa0074000 (vma =  0x80014000)  size = 0x00004000.
start addr is set to 0xa0000000 by exec file. <and here the simulation freezes>


   for skyeye 124
fpga-lin:/download/okl4_release_1.5.2 # skyeye -c  tools/sim_config/gumstix.skyeye -e build/images/image.sim
big_endian is  false.
arch: arm
cpu info: xscale, pxa25x, 69052100, fffffff0, 2 
mach  info: name pxa_lubbock, mach_init addr 0x806d880
uart_mod:0, desc_in:,  desc_out:, converter:
SKYEYE: use xscale mmu ops
exec file  "build/images/image.sim"'s format is elf32-littlearm.
load section  kernel.text: lma = 0xa0000000 (vma = 0xf0000000)  size =  0x00010814.
SKYEYE:NumInstrs 0, mem_write_byte addr = f0000000 no bank


Thank you for your support

Gabi

       
---------------------------------
Be a better friend, newshound, and know-it-all with Yahoo! Mobile.  Try it now.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.okl4.org/pipermail/developer/attachments/20080205/ecf8e332/attachment.htm