FC5 & IBM xSeries 100

[Jouni Viikari]

Hi all,

I have installed FC5 to an brand new IBM xSeries 100 type 8486 server.
On short trials It seems to work almost normally except:

1)  On boot I always get following messages on console:

PCI: Cannot allocate resource region 7 of bridge 0000:00:01.0
PCI: Cannot allocate resource region 8 of bridge 0000:00:01.0
PCI: Cannot allocate resource region 9 of bridge 0000:00:01.0
PCI: Cannot allocate resource region 7 of bridge 0000:00:1c.1
PCI: Cannot allocate resource region 8 of bridge 0000:00:1c.1
PCI: Cannot allocate resource region 9 of bridge 0000:00:1c.1

2)  On short trial the server seems to work quite normally except when I
try to start X.  startx (or gdm at runlevel 5) freezes most of the times
(not always) and leaves screen blank and in unusable state.  I still can
login to the box through ssh (Server works slowly in this state).
Following gets logged during X crash:

kernel: [drm] Initialized drm 1.0.1 20051102
kernel: ACPI: PCI Interrupt 0000:0a:04.0[A] -> GSI 16 (level, low) ->
IRQ 16
kernel: [drm] Initialized radeon 1.22.0 20051229 on minor 0
kernel: irq 20: nobody cared (try booting with the "irqpoll" option)
kernel:  [<c0145ac7>] __report_bad_irq+0x2b/0x69     [<c0145c88>]
note_interrupt+0x183/0x1af
kernel:  [<c01455f2>] handle_IRQ_event+0x23/0x4c     [<c01456c8>]
__do_IRQ+0xad/0xdd
kernel:  [<c0105eee>] do_IRQ+0x60/0x7b     =======================
kernel:  [<c0104786>] common_interrupt+0x1a/0x20
kernel:  [<c0102f65>] mwait_idle+0x1f/0x33     [<c0102f2d>] cpu_idle
+0x8f/0xa8
kernel:  [<c03c8715>] start_kernel+0x2fe/0x304    <3>handlers:
kernel: [<c026e461>] (usb_hcd_irq+0x0/0x50)
kernel: Disabling IRQ #20

This has happened with every kernel available to FC5 so far.

Since this is a box I promised one person to setup as a linux server and
it has been already sitting here one month unused, I wonder if anybody
knows if it is safe to put it under 'production-use' (If no X is used
and just acting as a pure mail/web server)?

I have tried some boot options (e.g. acpi=off pci=noacpi acpi=noirq, but
on the kernels I tried they did not help).  

Added comments on error #186557 

BR,

Jouni

[Karl Larsen]

Jouni Viikari wrote:
>  Hi all,
>
>  I have installed FC5 to an brand new IBM xSeries 100 type 8486 server.
>  On short trials It seems to work almost normally except:
>
>  1)  On boot I always get following messages on console:
>
>  PCI: Cannot allocate resource region 7 of bridge 0000:00:01.0
>  PCI: Cannot allocate resource region 8 of bridge 0000:00:01.0
>  PCI: Cannot allocate resource region 9 of bridge 0000:00:01.0
>  PCI: Cannot allocate resource region 7 of bridge 0000:00:1c.1
>  PCI: Cannot allocate resource region 8 of bridge 0000:00:1c.1
>  PCI: Cannot allocate resource region 9 of bridge 0000:00:1c.1
>
>  2)  On short trial the server seems to work quite normally except when I
>  try to start X.  startx (or gdm at runlevel 5) freezes most of the times
>  (not always) and leaves screen blank and in unusable state.  I still can
>  login to the box through ssh (Server works slowly in this state).
>  Following gets logged during X crash:
>
>  kernel: [drm] Initialized drm 1.0.1 20051102
>  kernel: ACPI: PCI Interrupt 0000:0a:04.0[A] -> GSI 16 (level, low) ->
>  IRQ 16
>  kernel: [drm] Initialized radeon 1.22.0 20051229 on minor 0
>  kernel: irq 20: nobody cared (try booting with the "irqpoll" option)
>  kernel:  [<c0145ac7>] __report_bad_irq+0x2b/0x69     [<c0145c88>]
>  note_interrupt+0x183/0x1af
>  kernel:  [<c01455f2>] handle_IRQ_event+0x23/0x4c     [<c01456c8>]
>  __do_IRQ+0xad/0xdd
>  kernel:  [<c0105eee>] do_IRQ+0x60/0x7b     =======================
>  kernel:  [<c0104786>] common_interrupt+0x1a/0x20
>  kernel:  [<c0102f65>] mwait_idle+0x1f/0x33     [<c0102f2d>] cpu_idle
>  +0x8f/0xa8
>  kernel:  [<c03c8715>] start_kernel+0x2fe/0x304    <3>handlers:
>  kernel: [<c026e461>] (usb_hcd_irq+0x0/0x50)
>  kernel: Disabling IRQ #20
>
>  This has happened with every kernel available to FC5 so far.
>
>  Since this is a box I promised one person to setup as a linux server and
>  it has been already sitting here one month unused, I wonder if anybody
>  knows if it is safe to put it under 'production-use' (If no X is used
>  and just acting as a pure mail/web server)?
>     
    If the deal is that you will service this server from afar using ssh 
then I would NEVER turn on X Windows. With ssh you can do all things to 
this computer so it should be safe.

    If the user looks at the computer you might need X windows to please 
him.

Karl


>  I have tried some boot options (e.g. acpi=off pci=noacpi acpi=noirq,
> but
>  on the kernels I tried they did not help).  
>
>  Added comments on error #186557 
>
>  BR,
>
>  Jouni
>
>

[Nicholas Winn]

I don't know the hardware specs you have with that xseries server. I 
have a batch of X 360 IBM servers. To get everything to work properly I 
had to go to a 2.4.9 kernel (only way to see all 16 gigs of mem.).

Karl Larsen wrote:

>  Jouni Viikari wrote:
>
> > Hi all,
> >
> > I have installed FC5 to an brand new IBM xSeries 100 type 8486 server.
> > On short trials It seems to work almost normally except:
> >
> > 1)  On boot I always get following messages on console:
> >
> > PCI: Cannot allocate resource region 7 of bridge 0000:00:01.0
> > PCI: Cannot allocate resource region 8 of bridge 0000:00:01.0
> > PCI: Cannot allocate resource region 9 of bridge 0000:00:01.0
> > PCI: Cannot allocate resource region 7 of bridge 0000:00:1c.1
> > PCI: Cannot allocate resource region 8 of bridge 0000:00:1c.1
> > PCI: Cannot allocate resource region 9 of bridge 0000:00:1c.1
> >
> > 2)  On short trial the server seems to work quite normally except when I
> > try to start X.  startx (or gdm at runlevel 5) freezes most of the times
> > (not always) and leaves screen blank and in unusable state.  I still can
> > login to the box through ssh (Server works slowly in this state).
> > Following gets logged during X crash:
> >
> > kernel: [drm] Initialized drm 1.0.1 20051102
> > kernel: ACPI: PCI Interrupt 0000:0a:04.0[A] -> GSI 16 (level, low) ->
> > IRQ 16
> > kernel: [drm] Initialized radeon 1.22.0 20051229 on minor 0
> > kernel: irq 20: nobody cared (try booting with the "irqpoll" option)
> > kernel:  [<c0145ac7>] __report_bad_irq+0x2b/0x69     [<c0145c88>]
> > note_interrupt+0x183/0x1af
> > kernel:  [<c01455f2>] handle_IRQ_event+0x23/0x4c     [<c01456c8>]
> > __do_IRQ+0xad/0xdd
> > kernel:  [<c0105eee>] do_IRQ+0x60/0x7b     =======================
> > kernel:  [<c0104786>] common_interrupt+0x1a/0x20
> > kernel:  [<c0102f65>] mwait_idle+0x1f/0x33     [<c0102f2d>] cpu_idle
> > +0x8f/0xa8
> > kernel:  [<c03c8715>] start_kernel+0x2fe/0x304    <3>handlers:
> > kernel: [<c026e461>] (usb_hcd_irq+0x0/0x50)
> > kernel: Disabling IRQ #20
> >
> > This has happened with every kernel available to FC5 so far.
> >
> > Since this is a box I promised one person to setup as a linux server and
> > it has been already sitting here one month unused, I wonder if anybody
> > knows if it is safe to put it under 'production-use' (If no X is used
> > and just acting as a pure mail/web server)?
> >   
>
>     If the deal is that you will service this server from afar using 
>  ssh then I would NEVER turn on X Windows. With ssh you can do all 
>  things to this computer so it should be safe.
>
>     If the user looks at the computer you might need X windows to 
>  please him.
>
>  Karl
>
>
> > I have tried some boot options (e.g. acpi=off pci=noacpi acpi=noirq, but
> > on the kernels I tried they did not help). 
> > Added comments on error #186557
> > BR,
> >
> > Jouni
> >
> >   
>
>  

-- 


Best Regards,

Nicholas Winn
IT Project Manager
Vacation Express

"If you make it idiot proof, someone will make a better idiot" - Taz

"Trial By Fire..." - Myself

"Quite frankly, I found it mostly interesting in a Jerry Springer kind of way. White
trash battling it out in public, throwing chairs at each other. SCO crying about IBM's
other women....Fairly entertaining," - Linus Torvalds

"Some people have told me they don't think a fat penguin really embodies the
grace of Linux, which just tells me they have never seen a angry penguin charging at them
in excess of 100 mph. They'd be a lot more careful about what they say if they
had." - Linux Torvalds