kernel 2.6.3 on JS20

Gazelle Jean-Laurent jean-laurent.gazelle at
Fri Mar 5 03:49:47 EST 2004

Thanks for your help,

I finally applied these 2 patches on 2.6.4-rc1 TAGGED version
(boot sequence of 2.6.3 was stopped because of NVRAM partion read issue)

Now under 2.6.4-rc1 + ide patch + fw patch , I'm almost at the end of boot sequence,
except that there's something wrong about IDE interrupts (interrupt 0x10 received and disabled at boot time).
Is it a still pending problem or is there a fix ?

AMD8111: chipset revision 3
AMD8111: 0000:00:04.1 (rev 03) UDMA133 controller
AMD8111: 100% native mode on irq 32
    ide0: BM-DMA at 0x7c00-0x7c07, BIOS settings: hda:pio, hdb:pio
    ide1: BM-DMA at 0x7c08-0x7c0f, BIOS settings: hdc:pio, hdd:pio
Probing IDE interface ide0...
hda: ST94011A, ATA DISK drive
Interrupt 0x10 (real) is invalid, disabling it.
ide0 at 0x7400-0x7407,0x6c02 on irq 32
Probing IDE interface ide1...
ide1 at 0x7800-0x7807,0x7002 on irq 32
hda: max request size: 1024KiB
hda: lost interrupt
hda: lost interrupt
hda: lost interrupt
hda: 78140160 sectors (40007 MB) w/2048KiB Cache, CHS=16383/255/63, UDMA(33)
hda: lost interrupt
 hda:<4>hda: dma_timer_expiry: dma status == 0x24
hda: DMA interrupt recovery

PS : Yes I've extracted my kernel sources using the 'bk export -rv2.6.3 ../build'
PJ : complete log sequence.

On Wed, 03 Mar 2004 09:36:33 -0600
Jake Moilanen <moilanen at> wrote:

R> Yeah, there are a number of FW regressions lately.  This one is due to
R> FW giving us the SPLPAR hypertas set and the kernel thinking it's on a
R> SPLPAR machine.  I have a patch that has a bunch of hacks to work around
R> the FW issues (there's more past this one).  It's not pretty, but it
R> will get your system booted at least.
R> The second patch is to work around an IDE HW issue where the IO space
R> has to be in the ISA range.  FW is supposed to get a fix for this one
R> soon.
R> Thanks,
R> Jake
R> On Wed, 2004-03-03 at 09:23, Gazelle Jean-Laurent wrote:
R> > Hi all,
R> >
R> > I'm trying to build linux kernel 2.6 on a JS20 platform.
R> >
R> > I extracted the TAGGED version 2.6.3 of ameslab 2.5 repository and rebuilded it
R> > successfully but run into an issue at boot time (complete log is attached):
R> > [boot]0100 MM Init
R> > [boot]0100 MM Init Done
R> > register_vpa: cpu 0x0
R> > Trap instruction interrupt : Invalid Instruction
R> >
R> > Looks like 'register_vpa' hcall fails...
R> >
R> >
R> > I should have missed something. Do I need another kernel version ?
R> > Any kind of information welcome.
R> >
R> >
R> > regards,
R> > --
R> > ------------------------------------------------------------
R> >  Jean-Laurent GAZELLE               150, rue M. Berthelot
R> >  Phone: +33 (0)4 98 16 34 66        Z.I. Toulon Est BP244
R> >  Fax : +33 (0)4 98 16 34 01         83078 TOULON Cedex9
R> >  e-mail : jlg at
R> >  Thales Computers                   FRANCE
R> > _____________ _______________

 Jean-Laurent GAZELLE               150, rue M. Berthelot
 Phone: +33 (0)4 98 16 34 66        Z.I. Toulon Est BP244
 Fax : +33 (0)4 98 16 34 01         83078 TOULON Cedex9
 e-mail : jlg at
 Thales Computers                   FRANCE
_____________ _______________
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Log_264.txt.bz2
Type: application/x-bzip2
Size: 3786 bytes
Desc: not available
Url : 

More information about the Linuxppc64-dev mailing list