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

Re: booting L2000 from netinstall image and netboot fails with panic



On Fri, Jun 01, 2007 at 02:55:27PM +0200, Christian Hilgers wrote:
> Hi,
> 
> I tried to install one (of two) L2000-36 (both running fine with
> HPUX 11.11, ioscan below) with Linux on a seperate disk.
> 
> I tried to install using debian-40r0-hppa-netinst.iso
> (CD was verified, md5sum taken from CD matches the md5sum
> on the webpage)
> 
> Both systems fail after (log below)
...

> logfile from cdboot
> 
> Firmware Version  41.39
...
> Linux version 2.6.18-4-parisc64 (Debian 2.6.18.dfsg.1-11) 
> (waldi@debian.org) (gcc version 4.1.2 20061115 (prerelease) (Debian 
> 4.1.1-21)) #1 Wed Feb 21 10:20:02 MST 2007

This kernel should boot fine on the L2000.
I'm using -parisc64-smp version on my a500's.

...
> Searching for devices...
> Found devices:
> 1. Rhapsody 360 at 0xfffffffffffa0000 [160] { 0, 0x0, 0x5c5, 0x00004 }
> 2. Rhapsody 360 at 0xfffffffffffa2000 [162] { 0, 0x0, 0x5c5, 0x00004 }
> 3. Rhapsody 360 at 0xfffffffffffa4000 [164] { 0, 0x0, 0x5c5, 0x00004 }
> 4. Rhapsody 360 at 0xfffffffffffa6000 [166] { 0, 0x0, 0x5c5, 0x00004 }
> 5. Memory at 0xfffffffffed08000 [8] { 1, 0x0, 0x095, 0x00009 }
> 6. Astro BC Runway Port at 0xfffffffffed00000 [0] { 12, 0x0, 0x582, 0x0000b 
> }
> 7. Elroy PCI Bridge at 0xfffffffffed30000 [0/0] { 13, 0x0, 0x782, 0x0000a }
> 8. Elroy PCI Bridge at 0xfffffffffed32000 [0/1] { 13, 0x0, 0x782, 0x0000a }
> 9. Elroy PCI Bridge at 0xfffffffffed34000 [0/2] { 13, 0x0, 0x782, 0x0000a }
> 10. Elroy PCI Bridge at 0xfffffffffed36000 [0/3] { 13, 0x0, 0x782, 0x0000a }
> 11. Elroy PCI Bridge at 0xfffffffffed38000 [0/4] { 13, 0x0, 0x782, 0x0000a }
> 12. Elroy PCI Bridge at 0xfffffffffed3a000 [0/5] { 13, 0x0, 0x782, 0x0000a }
> 13. Elroy PCI Bridge at 0xfffffffffed3c000 [0/6] { 13, 0x0, 0x782, 0x0000a }
> 14. Elroy PCI Bridge at 0xfffffffffed3e000 [0/7] { 13, 0x0, 0x782, 0x0000a }
> Enabling PDC_PAT chassis codes support v0.05
> 
> ********** VIRTUAL FRONT PANEL **********
> System Boot detected
> *****************************************
> LEDs:  RUN      ATTENTION     FAULT     REMOTE     POWER
>        FLASH    FLASH         OFF       ON         ON
> LED State: Running non-OS code.  Non-critical error detected.
> Check Chassis and Console Logs for error messages.
> 
> processor                 system initialization      1C00
> 
> *****************************************
> 
> ************ EARLY BOOT VFP *************
> End of early boot detected
> *****************************************

ugh. I really hate how HP VFP hides/prempts console output.
In order to see the missing console bits, one has to type ^B and
then "cl" (console log) to find the missing parts.

> allocation failed: out of vmalloc space - use vmalloc=<size> to increase 
> size.

all of these messages are known problems. But they shouldn't cause
a boot failure.

> Serial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing enabled
> 0000:00:04.0: ttyS0 at MMIO 0xfffffffff8000000 (irq = 69) is a 16550A
> 0000:00:04.0: ttyS1 at MMIO 0xfffffffff8000008 (irq = 69) is a 16550A
> 0000:00:04.0: ttyS2 at MMIO 0xfffffffff8000010 (irq = 69) is a 16550A
> 0000:00:05.0: ttyS3 at MMIO 0xfffffffff8006000 (irq = 70) is a 16550A
> Couldn't register serial port 0000:00:05.0: -28

I've never noticed this before. 28 == ENOSPC so my guess is this is
related to the vmalloc errors previous.

> Freeing unused kernel memory:
> ********** VIRTUAL FRONT PANEL **********
...
> processor                 system initialization      1C00

OS tells Firmware that it's now booted.

...
> processor                 system panic               1B00

OS tells firmware it panic'd.

...
> processor                 system panic               1B00

Twice?
Anyway, the actual tombstone that we want to see can be acquired
in two ways:
1) ^B and then "cl" to review the entire console output. My guess is
the boot history is long gone and you'll need to try this again.

2) answer "y" to "Interact with Boot loader" question.
   entire the number of the last parameter (as if you wanted
   to edit that paramter. _append_ "pdcchassis=0" to disable
   the OS from sending chassis codes to the firmware.
   This avoids the VFP from preempting the normal console output.

...
> here an ioscan from running hpux:

Nothing unusual here. I was looking for PCI-PCI bridges which do cause
problems with a number of kernel releases. Seems to be fixed in the
2.6.21 and 2.6.22 developement (pre-release) kernels.

> fc          0  0/7/0/0      td        CLAIMED     INTERFACE    HP Tachyon 
> TL/TS Fibre Channel Mass Storage Adapter

This is the last I've heard about a driver for Tachlite:
http://www.ussg.iu.edu/hypermail/linux/kernel/0508.2/0113.html

cpq driver is no longer in 2.6.22 tree and I don't see a replacement.

> Any ideas welcome. The systems are rack mounted so removing HW
> is not that easy

You shouldn't have to remove anything. Should be fine.
Please retry to collect the Tombstone from the panic.

thanks,
grant



Reply to: