Re: Recovery from a Blown Serial Port

davygravy — Thu, 15 Mar 2012 17:04:44 -0400

It will only work if it is just a simple blown port. If there is something within the processor that is shot/shorted/failed, then it wouldn't work.

http://archlinuxarm.org/forum/viewtopic.php?f=18&t=1120#p13778

shows my efforts ...

After a successful jtag flash, power down the kirkwood device, and disconnect the JTAG adapter from it. Restart & cross fingers... may work, may not.

You may have to leave off the netconsole part... first, just get it to reboot correctly.

Re: Recovery from a Blown Serial Port

nielsek — Thu, 15 Mar 2012 02:46:47 -0400

I was quite excited reading about this, hoping I could revive my dockstar with broken serial. Unfortunately it didn't work for me.
What did you do after flashing? Just removed JTAG and resetted the dockstar or are some other commands needed? Could you also share which dockstar.cfg you used?

Thanks in advance!

Recovery from a Blown Serial Port

davygravy — Sun, 11 Mar 2012 18:41:07 -0400

Just for posterity/anyone who needs an idea ...

I had an expendable Pogoplug V2/E02 that I'd picked up as bricked... recovered it via serial cable and used it as a test subject. Due to some unforeseen circumstances, the serial port became fried... and I was unable to boot it any longer, plus the NAND contents had become trashed.

It seems the fix for this was pretty predictable... using OpenOCD...
1. Flash u-boot.mtd0.kwb as normally done
2. Flash uboot.environment to the correct location
3. Attached a USB drive that already has Debian set up on it.

> dockstar_init
dockstar_init
target state: halted
target halted in ARM state due to debug-request, current mode: Supervisor
cpsr: 0x000000d3 pc: 0xffff0000
MMU: disabled, D-Cache: disabled, I-Cache: disabled
> nand probe 0
nand probe 0
NAND flash device 'NAND 128MiB 3.3V 8-bit (Hynix)' found
> nand erase 0 0x0 0xa0000
nand erase 0 0x0 0xa0000
erased blocks 0 to 5 on NAND flash device #0 'NAND 128MiB 3.3V 8-bit'
> nand erase 0 0xa0000 0xe0000
nand erase 0 0xa0000 0xe0000
erased blocks 5 to 12 on NAND flash device #0 'NAND 128MiB 3.3V 8-bit'
> nand write 0 uboot.mtd0.kwb-pogo_e02-DGL2CDisabled 0 oob_softecc_kw
nand write 0 uboot.mtd0.kwb-pogo_e02-DGL2CDisabled 0 oob_softecc_kw
> nand write 0 uboot_E02.environment 0x0c0000 oob_softecc_kw
nand write 0 uboot_E02.environment 0x0c0000 oob_softecc_kw

Now, to get some kind of control over it... go back in while booted in Debian and turn on the netconsole:

fw_setenv serverip 192.168.11.149
fw_setenv ipaddr 192.168.11.150
fw_setenv ethaddr 00:25:31:02:CD:67
fw_setenv if_netconsole 'ping $serverip'
fw_setenv start_netconsole 'setenv ncip $serverip; setenv bootdelay 10; setenv stdin nc; setenv stdout nc; setenv stderr nc; version;'
fw_setenv preboot 'run if_netconsole start_netconsole'

...my preferred invocation of nc/netcat to chatter back and forth via netconsole...

nc -klu 6666

Alive again.

It is my guess that a capacitor, resistor or fuse has blown on the board, as booting it now requires a momentary shorting of pins 3 & 4 on the JTAG port.