Re: 2.1.80 no-go

Michael Marxmeier (mike@msede.com)
Wed, 21 Jan 1998 21:51:46 +0100

Michael Marxmeier wrote:
> Kernel 2.1.80 hangs during boot on an SMP box.
>
> The following unusual message is displayed during boot
> process (before the usual FDC message).
>
> "more enable_irq(6) than disable_irq(6)"
>
> Scanning the SCSI bus (AIC7XXX) never finishes.
> Scrolling the display and ALT+SCROLL still work.

Ingo Molnar suggested:
> if you have a default Adaptec setup, then the following lilo.conf line
> should get you going:
>
> append="pirq=11,11,11,11"
>
>let us know if there are other problems.

It works. Stupid me!
I tried the pirq= argument but had a typo which went unnoticed.

scanpci returns a rather strange result on my system:
# echo -n pirq=; echo `scanpci | grep T_L | cut -c56-` | sed 's/ /,/g'
pirq=0x00,0xff,0x0f,0x09,0x0a,0x0b

This is an SMP box using the ASUS P2L97-DS motherboard (440LX chipset).
According to scanpci, the bogus 0x00 and 0xff entries are related to

pci bus 0x0 cardnum 0x01 function 0x0000: vendor 0x8086 device 0x7181
Intel Device unknown
pci bus 0x0 cardnum 0x04 function 0x0002: vendor 0x8086 device 0x7112
Intel Device unknown

I had to specify the pirq= argument on the kernel commandline (omitting
the bogus 0x0,0xff entries).
I am using 2.1.80 + devfs patch + tcp patch as suggested by Andy Kleen.

Thanks for your help
Michael

-- 
Michael Marxmeier           Marxmeier Software GmbH
E-Mail: mike@msede.com      Suedstrasse 58
Voice : +49 202 2431440     42119 Wuppertal, Germany
Fax   : +49 202 2431420     http://www.msede.com/