Open Firmawre nvramrc problem
  • mat
  • Caterpillar
    Caterpillar
    mat
    Posts: 23 from 2003/10/22
    From: Torun, Poland
    Hi,
    I need help (and hope to get it quick ;)) to resolve the following problem. I've been
    experimenting with Open Firmware scripting and in the process I wrote some instructions to nvramrc
    that made my Pegasos to stop booting. I've connected the serial console and I see that the OF
    starts fine. It reaches the memory tests, passes it displays CPU info and then I assume it tries to
    start nvramrc. I then get an error (I've tried to find out something and in the test process made it
    to setenv something which gives the error). Then it shows:

    no/bad nvramrc - performing default startup script
    nvramrc error
    nvramrc script error: undefined word
    no console:
    error: cannot open console for I/O
    no keyboard!
    error: cannot open keyboard for console input
    entering main read/eval loop...

    And then I just get OF prompt. It seems it works almost fine. However it does not display command
    results (just error messages). Of course I can't start nvedit to correct the error I made. It seems
    also that other devices are not initialized too - I can't force it to boot or do any other job. So I
    need someone with better experience (maybe someone from bPlan) to help me resolve the problem. Is
    there any way to hand-init the console? Or maybe there is some simple way to reset nvramrc
    settings?

    Thanx in advance for any hints,

    M.
  • »19.06.04 - 15:05
    Profile
  • mat
  • Caterpillar
    Caterpillar
    mat
    Posts: 23 from 2003/10/22
    From: Torun, Poland
    I was just told that when you press ESC just after the first keyboar led blink on power up it should
    bypass the nvramrc processing. Unfortunately it doesn't work for me.
  • »19.06.04 - 15:23
    Profile
  • mat
  • Caterpillar
    Caterpillar
    mat
    Posts: 23 from 2003/10/22
    From: Torun, Poland
    Ok. I did it - what I did was to use the emergency procedure published after problems with the first batch of P2 motherboards in december.
  • »19.06.04 - 16:08
    Profile
  • Priest of the Order of the Butterfly
    Priest of the Order of the Butterfly
    bbrv
    Posts: 750 from 2003/2/14
    From: Earth
    Good Mat!

    Glad you overcame that problem. Happy Pegasos!

    R&B :-)
  • »19.06.04 - 16:56
    Profile Visit Website
  • mat
  • Caterpillar
    Caterpillar
    mat
    Posts: 23 from 2003/10/22
    From: Torun, Poland
    Yes I did. Thanks. And you know what? The problem taught me nothing - I am still willing to mess with nvramrc to make my Pegasos recognize my new disks right after I boot it without additional reset.

    What I found out using the serial console is that nvramrc is executed before harddisks detection (during disks spin-up). I'll try to add some delay loop to the nvramrc (as I couldn't find OF Forth equivalent of sleep/wait/pause command) to make it wait a few seconds before disks detection - I hope this will be enough for the disks to spin-up.

    M.
  • »21.06.04 - 07:29
    Profile
  • Butterfly
    Butterfly
    ultraspec
    Posts: 94 from 2004/1/29
    You're a brave man mat. Thank you for sharing this experience with everyone, we can all learn from it :)
  • »21.06.04 - 15:28
    Profile Visit Website
  • mat
  • Caterpillar
    Caterpillar
    mat
    Posts: 23 from 2003/10/22
    From: Torun, Poland
    Just wait a moment - I'm writing something that might be interesting in another thread :-D
  • »21.06.04 - 15:30
    Profile