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

Re: sarge & sid installer failure on r4k indy



Meelis Roos wrote:
> I have a R4600PC Indy (32M RAM) and I'm trying the new installer on it,
> booting from network via bootp/tftp using netboot-boot.img.
> 
> The installer from sid (20040403) complains about umount but seems
> continue somewhat before hanging:
> 
> RAMDISK: Compressed image found at block 0
> Freeing initrd memory: 1804k freed
> VFS: Mounted root (ext2 filesystem) readonly.
> Freeing unused kernel memory: 88k freed
> Setting up filesystem, please wait ...
> umount: /initrd: Invalid argument

Which is normal. Beyond that point it copies the ramdisk and chroots
into the copy.

> Then it waits for quite some seconds. Sometimes it doesn't proceed
> anywhere, sometimes it proceeds with the install and hangs in different
> places (before language chooser, 71% of hardware detection, after
> setting the hostname).

Probably you are just a bit impatient, my test machine is similiar
to yours, and quite slow with the install.

> The installer from sarge (20040308 - an older one so this is probably
> uninportant) complains about missing console:
> 
> RAMDISK: Compressed image found at block 0
> Freeing initrd memory: 1748k freed
> VFS: Mounted root (ext2 filesystem) readonly.
> Freeing unused kernel memory: 88k freed
> Warning: unable to open an initial console.
> Kernel panic: Attempted to kill init!
> 
> I switched it to the serial console via setenv console d, now it
> complained the same it to the serial console, with or without
> console=ttyS0,9600 parameter.

For older images, you'll have to use
bootp():<filename> append="devfs=mount"

[snip]
> About possible harware failures: IRIX 6.2 and Debian Woody run fine. But
> when I have keyboard unplugged and using serial console, I get these 2
> lines on power-on:
> 
> Data path test                             *FAILED*
>         RTC path test                      *FAILED*
> 
> It probably means that the clock battery is dead (the clock is mostly
> right, difference 9 minutes for a long time with no NTP) but maybe they
> are somewhat important.

I see this also, after longer downtimes the clock starts at 1970-01-01,
and the file dates are then broken as a consequence. It doesn't matter
for the installer.


Thiemo



Reply to: