Problem with SMP and NCCH-DL motherboard

Problem with SMP and NCCH-DL motherboard

[Pedro Pla]

Hello,

I'm running a system with an Asus NCCH-DL motherboard with dual Nocona
Xeon 3.2 GHZ cpu's, when I boot the system with a single cpu kernel with
acpi not compiled in it works fine, however when I try to boot a kernel
with smp it gives a timeout detecting the cpus and then the machine
reboots after I think trying to work out the irq tables, I say I think
because it happens so fast that I can barely see what is on the screen
before it reboots.

I've tried recompiling the kernel with many different options both in
and out and the only one that lets me boot is a single cpu without acpi,
apic and hpet, I tried googling for similar problems but haven't been
able to find anything that applies. I've also tried using an EM64T
kernel in case it had to do with the Nocona being incompatible with smp
in 32bit mode but that didn't seem to work and gave the same error.

Is it a kernel problem or a hardware issue? I tried swapping cpu's in
case one was broken but that didn't help, I also tried flashing the bios
to the latest one from Asus in case there was some issue with that but
no luck either.

Also in case this is helpful in locating the problem, when I get it to
boot a single cpu 2.6.12.1 kernel without acpi or apic I get the
following errors with the PCI:

NET: Registered protocol family 16
PCI: PCI BIOS revision 2.10 entry at 0xf1d30, last bus=4
PCI: Using configuration type 1
SCSI subsystem initialized
PCI: Probing PCI hardware
PCI: Probing PCI hardware (bus 00)
PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.1
Boot video device is 0000:01:00.0
PCI: Transparent bridge - 0000:00:1e.0
PCI: Using IRQ router PIIX/ICH [8086/25a1] at 0000:00:1f.0
PCI: IRQ 0 for device 0000:00:1f.1 doesn't match PIRQ mask - try
pci=usepirqmask
PCI: Found IRQ 11 for device 0000:00:1f.1
PCI: Cannot allocate resource region 0 of device 0000:01:00.0

Any ideas or advice would be greatly appreciated.

Best regards,
Pedro

Re: Problem with SMP and NCCH-DL motherboard

[Pedro Pla]

Hi Again,

I've done some further tests disabling hyperthreading which then lets me
disable apic on the motherboard and when I disable these I can boot a
kernel compiled with smp, however it boots detecting only 1 cpu rather
than the 2 on the motherboard. If I then reenable APIC in the bios with
hyperthreading still disabled it fails in the same way, so the problem
seems to be with APIC. Any ideas what more I can try?

Thanks in advance!!
Pedro

Pedro Pla wrote:

>Hello,
>
>I'm running a system with an Asus NCCH-DL motherboard with dual Nocona
>Xeon 3.2 GHZ cpu's, when I boot the system with a single cpu kernel with
>acpi not compiled in it works fine, however when I try to boot a kernel
>with smp it gives a timeout detecting the cpus and then the machine
>reboots after I think trying to work out the irq tables, I say I think
>because it happens so fast that I can barely see what is on the screen
>before it reboots.
>
>I've tried recompiling the kernel with many different options both in
>and out and the only one that lets me boot is a single cpu without acpi,
>apic and hpet, I tried googling for similar problems but haven't been
>able to find anything that applies. I've also tried using an EM64T
>kernel in case it had to do with the Nocona being incompatible with smp
>in 32bit mode but that didn't seem to work and gave the same error.
>
>Is it a kernel problem or a hardware issue? I tried swapping cpu's in
>case one was broken but that didn't help, I also tried flashing the bios
>to the latest one from Asus in case there was some issue with that but
>no luck either.
>
>Also in case this is helpful in locating the problem, when I get it to
>boot a single cpu 2.6.12.1 kernel without acpi or apic I get the
>following errors with the PCI:
>
>NET: Registered protocol family 16
>PCI: PCI BIOS revision 2.10 entry at 0xf1d30, last bus=4
>PCI: Using configuration type 1
>SCSI subsystem initialized
>PCI: Probing PCI hardware
>PCI: Probing PCI hardware (bus 00)
>PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.1
>Boot video device is 0000:01:00.0
>PCI: Transparent bridge - 0000:00:1e.0
>PCI: Using IRQ router PIIX/ICH [8086/25a1] at 0000:00:1f.0
>PCI: IRQ 0 for device 0000:00:1f.1 doesn't match PIRQ mask - try
>pci=usepirqmask
>PCI: Found IRQ 11 for device 0000:00:1f.1
>PCI: Cannot allocate resource region 0 of device 0000:01:00.0
>
>Any ideas or advice would be greatly appreciated.
>
>Best regards,
>Pedro
>
>
>-
>To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>Please read the FAQ at  http://www.tux.org/lkml/
>
>  
>

GOT IT WORKING - Problem with SMP and NCCH-DL motherboard

[Pedro Pla]

Hi,

In case anyone has the same problem I got it solved. I basically
disabled everything that was on the motherboard such as the promise
sata, sound, usb, etc. It then seemed to work except it choked on
loading the AACRAID driver so I set the pci=usepirqmask and it then got
it. I then tried reenabling USB and it gave the same problem again so I
disabled the USB legacy support and left USB 2.0 on and tried again,
this time it worked. Now why would USB legacy support prevent linux from
detecting the other cpus and make it reboot? Not that I'm too worried
about it but it might be something a kernel hacker might want to look at...

Pedro

Pedro Pla wrote:

>Hi Again,
>
>I've done some further tests disabling hyperthreading which then lets me
>disable apic on the motherboard and when I disable these I can boot a
>kernel compiled with smp, however it boots detecting only 1 cpu rather
>than the 2 on the motherboard. If I then reenable APIC in the bios with
>hyperthreading still disabled it fails in the same way, so the problem
>seems to be with APIC. Any ideas what more I can try?
>
>Thanks in advance!!
>Pedro
>
>Pedro Pla wrote:
>
>  
>
>>Hello,
>>
>>I'm running a system with an Asus NCCH-DL motherboard with dual Nocona
>>Xeon 3.2 GHZ cpu's, when I boot the system with a single cpu kernel with
>>acpi not compiled in it works fine, however when I try to boot a kernel
>>with smp it gives a timeout detecting the cpus and then the machine
>>reboots after I think trying to work out the irq tables, I say I think
>>because it happens so fast that I can barely see what is on the screen
>>before it reboots.
>>
>>I've tried recompiling the kernel with many different options both in
>>and out and the only one that lets me boot is a single cpu without acpi,
>>apic and hpet, I tried googling for similar problems but haven't been
>>able to find anything that applies. I've also tried using an EM64T
>>kernel in case it had to do with the Nocona being incompatible with smp
>>in 32bit mode but that didn't seem to work and gave the same error.
>>
>>Is it a kernel problem or a hardware issue? I tried swapping cpu's in
>>case one was broken but that didn't help, I also tried flashing the bios
>>to the latest one from Asus in case there was some issue with that but
>>no luck either.
>>
>>Also in case this is helpful in locating the problem, when I get it to
>>boot a single cpu 2.6.12.1 kernel without acpi or apic I get the
>>following errors with the PCI:
>>
>>NET: Registered protocol family 16
>>PCI: PCI BIOS revision 2.10 entry at 0xf1d30, last bus=4
>>PCI: Using configuration type 1
>>SCSI subsystem initialized
>>PCI: Probing PCI hardware
>>PCI: Probing PCI hardware (bus 00)
>>PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.1
>>Boot video device is 0000:01:00.0
>>PCI: Transparent bridge - 0000:00:1e.0
>>PCI: Using IRQ router PIIX/ICH [8086/25a1] at 0000:00:1f.0
>>PCI: IRQ 0 for device 0000:00:1f.1 doesn't match PIRQ mask - try
>>pci=usepirqmask
>>PCI: Found IRQ 11 for device 0000:00:1f.1
>>PCI: Cannot allocate resource region 0 of device 0000:01:00.0
>>
>>Any ideas or advice would be greatly appreciated.
>>
>>Best regards,
>>Pedro
>>
>>
>>-
>>To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>>the body of a message to majordomo@vger.kernel.org
>>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>Please read the FAQ at  http://www.tux.org/lkml/
>>
>> 
>>
>>    
>>
>
>-
>To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>Please read the FAQ at  http://www.tux.org/lkml/
>
>  
>