OpenWrt Forum Archive

Topic: SparkLAN WAPD-170AG / IXP425 problem, need help with kernel

The content of this topic has been archived on 3 Apr 2018. There are no obvious gaps in this topic, but there may still be some posts missing at the end.

Hi,

I got a SparkLAN WAPD-170AG routerboard that was running some older version of OpenWRT on it...

Got this screen shot...

BusyBox v1.4.2 (2008-07-02 19:09:29 CST) Built-in shell (ash)
Enter 'help' for a list of built-in commands.

  _______                  _    _   ___ __ _
 |    ___|.-----.----.----| |_.| | | . |  \ |
 \___    \|  _  | -  |   _|   /| |_| _ | .  |
 |_______||   __|___||__| |_\_\|___|_|_|_|\_|
          |__| W I R E L E S S   F R E E D O M
  SparkLAN LiteSDK V1.4 ----------------------------
  *
  * Built at 2008,07,03 Taiwan (WAPD-170AG)
  *
 ---------------------------------------------------
===========WAPD-170AG ath0 Access Point of scanning==================

Tried to compile and copy a newer version of OpenWRT, but I can't boot anything anymore, and the kernel that was shipped with the board is, of course, overwritten.

Can anyone help me with a running kernel for this board? Or give me some hints how to make it work?

Thx.

Hi dajhard,
Sorry I never responded before, been on a long overseas trip. I got back to playing with the Sparklan boards this weekend. Right now, the following procedure works to get OpenWrt on the board up and running with the latest SVN.

1. make kernel_menuconfig -> system type -> Intel IXP4xx Implementation Options  --->   [*]Coyote

2. edit trunk/scripts/arm-magic.sh. add a new board with a hexid of 290.

(i.e.
Index: arm-magic.sh
===================================================================
--- arm-magic.sh        (revision 16213)
+++ arm-magic.sh        (working copy)
@@ -22,7 +22,7 @@
# NOTE: for now it's for only IXP4xx in big endian mode

# list of supported boards, in "boardname machtypeid" format
-for board in "avila 526" "gateway7001 731" "nslu2 597" "nas100d 865" "wg302v1 889" "wg302v2 890" "pronghorn 928" "pronghornmetro 1040" "compex 1273" "wrt300nv2 1077" "loft 849" "dsmg600 964" "fsg3 1091" "ap1000 1543" "tw2662 1658" "tw5334 1664" "ixdpg425 604" "cambria 1468" "sidewinder 1041"
+for board in "avila 526" "gateway7001 731" "nslu2 597" "nas100d 865" "wg302v1 889" "wg302v2 890" "pronghorn 928" "pronghornmetro 1040" "compex 1273" "wrt300nv2 1077" "loft 849" "dsmg600 964" "fsg3 1091" "ap1000 1543" "tw2662 1658" "tw5334 1664" "ixdpg425 604" "cambria 1468" "sidewinder 1041" "sparklan 290"
do
   set -- $board
   hexid=$(printf %x\\n $2)
)


You should now be able to load the openwrt-sparklan-zImage to boot the board. (Note, You must set the image options in the kernel to use ramdisk)


Will continue my previous thread with any progress. I think I should be able to add it as a proper target, using the coyote and the sparklan SDK as a basis.



Regards,

Fins

Thx for reply, I got a question...

I got a 2.6.26.6 kernel from SparkLAN a month ago. It works stable.
I am using it as a router with 2 x SuperA links (MadWifi).
Got ntpclient and quagga-unstable (0.99.11) running on it, nothing more.
With 1 wireless link onboard the board was working perfectly. Good signal levels, good throughput, not overkilling the CPU.

With 2 wireless links onboard, the first link is working good, but the second link doesn't work. Whatever I do, i get very low signal levels, at both sides of the link (remote router, Debian router, reports -20dBm weaker signals, Sparklan routerboard shows cca the same).

I tought it was a SparkLAN issue, or IXP4xx, but now I have the same simptoms on my RB433AH (AR7100). It's running kernel 2.6.26.5.

So, now I will try to compile the latest version of the Kamikaze. And I need to make quagga-unstable package.

The question...
Can u please tell me how to do this... "(Note, You must set the image options in the kernel to use ramdisk)".

Thx.

make menuconfig

Target Images ---> [*]ramdisk


That's all smile

I cant remeber what version I got from sparklan, the atheros binary driver they sent me for the WMIA-199N cards wouldn't work, their build system pales in comparrison to openWRT's.

edit... have you tried the 2 cards operating on distant channels? Just an idea, if they were on the same channel it may cause issues.

(Last edited by fins on 1 Jun 2009, 16:42)

ath0      IEEE 802.11Ta  ESSID:"2156H"  Nickname:""
          Mode:Master  Frequency:5.8 GHz  Access Point: 00:10:C6:5D:2B:63
          Bit Rate:0 kb/s   Tx-Power=14 dBm   Sensitivity=1/1
          Retry:off   RTS thr:off   Fragment thr:off
          Encryption key:off
          Power Management:off
          Link Quality=0/70  Signal level=-93 dBm  Noise level=-93 dBm
          Rx invalid nwid:128319  Rx invalid crypt:0  Rx invalid frag:0
          Tx excessive retries:0  Invalid misc:0   Missed beacon:0

ath1      IEEE 802.11Ta  ESSID:"2127H"  Nickname:""
          Mode:Master  Frequency:5.25 GHz  Access Point: 00:10:C6:5D:28:47
          Bit Rate:0 kb/s   Tx-Power=14 dBm   Sensitivity=1/1
          Retry:off   RTS thr:off   Fragment thr:off
          Encryption key:off
          Power Management:off
          Link Quality=37/70  Signal level=-56 dBm  Noise level=-93 dBm
          Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
          Tx excessive retries:0  Invalid misc:0   Missed beacon:0

Can u help me out with this...
I did ...

svn co svn://svn.openwrt.org/openwrt/branches/8.09
cd 8.09
make kernel_menuconfig
-> Target System (Intel IXP4xx [2.6])  ---> (X) Intel IXP4xx [2.6] 
EXIT
-> System Type  ---> Intel IXP4xx Implementation Options  ---> [*] Coyote

Got this error...

*
* Library routines
*
CRC-CCITT functions (CRC_CCITT) [M/y/?] m
CRC16 functions (CRC16) [Y/n/m/?] y
CRC ITU-T V.41 functions (CRC_ITU_T) [N/m/y/?] n
CRC32 functions (CRC32) [Y/?] y
CRC7 functions (CRC7) [N/m/y/?] n
CRC32c (Castagnoli, et al) Cyclic Redundancy-Check (LIBCRC32C) [M/n/y/?] m
#
# configuration written to .config
#
make[4]: armeb-linux-uclibc-gcc: Command not found
make[4]: armeb-linux-uclibc-gcc: Command not found
scripts/kconfig/conf -s arch/arm/Kconfig
make[4]: armeb-linux-uclibc-gcc: Command not found
make[4]: armeb-linux-uclibc-gcc: Command not found
  CHK     include/linux/version.h
  UPD     include/linux/version.h
make[4]: armeb-linux-uclibc-gcc: Command not found
  SYMLINK include/asm-arm/arch -> include/asm-arm/arch-ixp4xx
make[4]: armeb-linux-uclibc-gcc: Command not found
  Generating include/asm-arm/mach-types.h
make[4]: armeb-linux-uclibc-gcc: Command not found
make[4]: armeb-linux-uclibc-gcc: Command not found
  CHK     include/linux/utsrelease.h
  UPD     include/linux/utsrelease.h
make[4]: armeb-linux-uclibc-gcc: Command not found
  SYMLINK include/asm -> include/asm-arm
make[4]: armeb-linux-uclibc-gcc: Command not found
make[4]: armeb-linux-uclibc-gcc: Command not found
make[4]: armeb-linux-uclibc-gcc: Command not found
  CC      kernel/bounds.s
/bin/sh: armeb-linux-uclibc-gcc: not found
make[5]: *** [kernel/bounds.s] Error 127
make[4]: *** [prepare0] Error 2
make[3]: *** [prepare] Error 2
xake[3]: Leaving directory `/home/dajhard/openwrt/svn/8.09/build_dir/linux-ixp4x _generic/linux-2.6.26.8'
make[2]: [menuconfig] Error 2 (ignored)
rm -rf /home/dajhard/openwrt/svn/8.09/build_dir/linux-ixp4xx_generic/modules
*home/dajhard/openwrt/svn/8.09/staging_dir/host/bin/sed -i -e 's,.*CONFIG_AEABI.4,# CONFIG_AEABI is not set,' /home/dajhard/openwrt/svn/8.09/build_dir/linux-ixp xx_generic/linux-2.6.26.8/.config
e-: /home/dajhard/openwrt/svn/8.09/staging_dir/host/bin/sed: No such file or dir ctory
make[2]: *** [menuconfig] Error 127
make[2]: Leaving directory `/home/dajhard/openwrt/svn/8.09/target/linux/ixp4xx'
make[1]: *** [menuconfig] Error 2
make[1]: Leaving directory `/home/dajhard/openwrt/svn/8.09/target/linux'
make: *** [kernel_menuconfig] Error 2
dajhard@sizif:~/openwrt/svn/8.09$

The
"make[4]: armeb-linux-uclibc-gcc: Command not found"
Lines suggest that the toolchain hasn't been built. What I'd do is do a
"make clean"
followed by
"make"
followed by
"make menu_kernelconfig"
and do the steps above,
finally another
"make"

If that's not workin g, then post the output of the final make, or wherever you get errors that halt compilation.

The discussion might have continued from here.