PAE dom0 is broken with the default config

PAE dom0 is broken with the default config

[Kamble, Nitin A]

[-- Attachment #1.1: Type: text/plain, Size: 1939 bytes --]

Hi Keir, Ian,

  I built PAE Xen with the default config, and it is not booting. Log is
bellow. If I replace the linux-defconfig_xen_x86_32 file with the
linux-defconfig_xen0_x86_32 before building then dom0 is able to boot
and see more than 4Gig memory. This looks like a dom0 kernel config
issue. Do you care PAE dom0 for 3.0.2 with the default config? Is
anybody working on this?

 

IP route cache hash table entries: 262144 (order: 8, 1048576 bytes)

TCP established hash table entries: 524288 (order: 10, 4194304 bytes)

TCP bind hash table entries: 65536 (order: 7, 524288 bytes)

TCP: Hash tables configured (established 524288 bind 65536)

TCP reno registered

NET: Registered protocol family 8

NET: Registered protocol family 20

Using IPI No-Shortcut mode

Freeing unused kernel memory: 172k freed

scsi_mod: no version for "struct_module" found: kernel tainted.

SCSI subsystem initialized

ACPI: PCI Interrupt 0000:00:1f.2[A] -> GSI 18 (level, low) -> IRQ 17

ata1: SATA max UDMA/133 cmd 0x0 ctl 0xC402 bmdma 0xBC00 irq 17

ata2: SATA max UDMA/133 cmd 0x0 ctl 0xC002 bmdma 0xBC08 irq 17

ata1: dev 0 ATA-6, max UDMA/133, 312581808 sectors: LBA48

ata1: dev 0 configured for UDMA/133

scsi0 : ata_piix

ATA: abnormal status 0x7F on port 0xC087

scsi1 : ata_piix

  Vendor: ATA       Model: ST3160023AS       Rev: 3.00

  Type:   Direct-Access                      ANSI SCSI revision: 05

SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)

sda: Write Protect is off

SCSI device sda: drive cache: write back

SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)

sda: Write Protect is off

SCSI device sda: drive cache: write back

 sda: unknown partition table

sd 0:0:0:0: Attached scsi disk sda

Kernel panic - not syncing: Attempted to kill init!

 (XEN) Domain 0 crashed: rebooting machine in 5 seconds.

 

 

 

 

 


[-- Attachment #1.2: Type: text/html, Size: 8518 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

RE: PAE dom0 is broken with the default config

[Ian Pratt]

>   I built PAE Xen with the default config, and it is not 
> booting. Log is bellow. If I replace the 
> linux-defconfig_xen_x86_32 file with the 
> linux-defconfig_xen0_x86_32 before building then dom0 is able 
> to boot and see more than 4Gig memory. This looks like a dom0 
> kernel config issue. Do you care PAE dom0 for 3.0.2 with the 
> default config? Is anybody working on this?

This seems pretty unlikely to be a PAE issue -- we do quite extensive
testing.

Are you sure you have the right initrd? Is the same scsi driver being
used as in the -xen0 case?

Thanks,
Ian
 
>  
> 
> IP route cache hash table entries: 262144 (order: 8, 1048576 bytes)
> 
> TCP established hash table entries: 524288 (order: 10, 4194304 bytes)
> 
> TCP bind hash table entries: 65536 (order: 7, 524288 bytes)
> 
> TCP: Hash tables configured (established 524288 bind 65536)
> 
> TCP reno registered
> 
> NET: Registered protocol family 8
> 
> NET: Registered protocol family 20
> 
> Using IPI No-Shortcut mode
> 
> Freeing unused kernel memory: 172k freed
> 
> scsi_mod: no version for "struct_module" found: kernel tainted.
> 
> SCSI subsystem initialized
> 
> ACPI: PCI Interrupt 0000:00:1f.2[A] -> GSI 18 (level, low) -> IRQ 17
> 
> ata1: SATA max UDMA/133 cmd 0x0 ctl 0xC402 bmdma 0xBC00 irq 17
> 
> ata2: SATA max UDMA/133 cmd 0x0 ctl 0xC002 bmdma 0xBC08 irq 17
> 
> ata1: dev 0 ATA-6, max UDMA/133, 312581808 sectors: LBA48
> 
> ata1: dev 0 configured for UDMA/133
> 
> scsi0 : ata_piix
> 
> ATA: abnormal status 0x7F on port 0xC087
> 
> scsi1 : ata_piix
> 
>   Vendor: ATA       Model: ST3160023AS       Rev: 3.00
> 
>   Type:   Direct-Access                      ANSI SCSI revision: 05
> 
> SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
> 
> sda: Write Protect is off
> 
> SCSI device sda: drive cache: write back
> 
> SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
> 
> sda: Write Protect is off
> 
> SCSI device sda: drive cache: write back
> 
>  sda: unknown partition table
> 
> sd 0:0:0:0: Attached scsi disk sda
> 
> Kernel panic - not syncing: Attempted to kill init!
> 
>  (XEN) Domain 0 crashed: rebooting machine in 5 seconds.
> 
>  
> 
>  
> 
>  
> 
>  
> 
>  
> 
> 

RE: PAE dom0 is broken with the default config

[Kamble, Nitin A]

   I did rebuild the initrd for the new kernel/modules. The system has
sata disk. scsi driver is the same ata_piix, libata. The same config is
working fine for the non-PAE build. 
   The VMX status report mail says PAE is fine, but they are not using
the default (linux-defconfig_xen_x86_32) config. They are using the xen0
config for PAE testing. What config do you use for your PAE testing?

In summary here is my observation
PAE build with linux-defconfig_xen_x86_32 config:	dom0 does not
boot
PAE build with linux-defconfig_xen0_x86_32 config:	boots
Non-PAE build with either config:				boots

All of the above kernels need corresponding initrd to boot.

Thanks & Regards,
Nitin
------------------------------------------------------------------------
-----------
Open Source Technology Center, Intel Corp

>-----Original Message-----
>From: Ian Pratt [mailto:m+Ian.Pratt@cl.cam.ac.uk]
>Sent: Wednesday, April 05, 2006 4:34 PM
>To: Kamble, Nitin A; Keir.Fraser@cl.cam.ac.uk
>Cc: xen-devel@lists.xensource.com; ian.pratt@cl.cam.ac.uk
>Subject: RE: PAE dom0 is broken with the default config
>
>>   I built PAE Xen with the default config, and it is not
>> booting. Log is bellow. If I replace the
>> linux-defconfig_xen_x86_32 file with the
>> linux-defconfig_xen0_x86_32 before building then dom0 is able
>> to boot and see more than 4Gig memory. This looks like a dom0
>> kernel config issue. Do you care PAE dom0 for 3.0.2 with the
>> default config? Is anybody working on this?
>
>This seems pretty unlikely to be a PAE issue -- we do quite extensive
>testing.
>
>Are you sure you have the right initrd? Is the same scsi driver being
>used as in the -xen0 case?
>
>Thanks,
>Ian
>
>>
>>
>> IP route cache hash table entries: 262144 (order: 8, 1048576 bytes)
>>
>> TCP established hash table entries: 524288 (order: 10, 4194304 bytes)
>>
>> TCP bind hash table entries: 65536 (order: 7, 524288 bytes)
>>
>> TCP: Hash tables configured (established 524288 bind 65536)
>>
>> TCP reno registered
>>
>> NET: Registered protocol family 8
>>
>> NET: Registered protocol family 20
>>
>> Using IPI No-Shortcut mode
>>
>> Freeing unused kernel memory: 172k freed
>>
>> scsi_mod: no version for "struct_module" found: kernel tainted.
>>
>> SCSI subsystem initialized
>>
>> ACPI: PCI Interrupt 0000:00:1f.2[A] -> GSI 18 (level, low) -> IRQ 17
>>
>> ata1: SATA max UDMA/133 cmd 0x0 ctl 0xC402 bmdma 0xBC00 irq 17
>>
>> ata2: SATA max UDMA/133 cmd 0x0 ctl 0xC002 bmdma 0xBC08 irq 17
>>
>> ata1: dev 0 ATA-6, max UDMA/133, 312581808 sectors: LBA48
>>
>> ata1: dev 0 configured for UDMA/133
>>
>> scsi0 : ata_piix
>>
>> ATA: abnormal status 0x7F on port 0xC087
>>
>> scsi1 : ata_piix
>>
>>   Vendor: ATA       Model: ST3160023AS       Rev: 3.00
>>
>>   Type:   Direct-Access                      ANSI SCSI revision: 05
>>
>> SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
>>
>> sda: Write Protect is off
>>
>> SCSI device sda: drive cache: write back
>>
>> SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
>>
>> sda: Write Protect is off
>>
>> SCSI device sda: drive cache: write back
>>
>>  sda: unknown partition table
>>
>> sd 0:0:0:0: Attached scsi disk sda
>>
>> Kernel panic - not syncing: Attempted to kill init!
>>
>>  (XEN) Domain 0 crashed: rebooting machine in 5 seconds.
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>

RE: PAE dom0 is broken with the default config

[Boris Ostrovsky]

>> I built PAE Xen with the default config, and it is not 
>> booting. Log is bellow. If I replace the 
>> linux-defconfig_xen_x86_32 file with the 
>> linux-defconfig_xen0_x86_32 before building then dom0 is able 
>> to boot and see more than 4Gig memory. This looks like a dom0 
>> kernel config issue. Do you care PAE dom0 for 3.0.2 with the 
>> default config? Is anybody working on this?
>
>
> This seems pretty unlikely to be a PAE issue -- we do quite extensive
> testing.
>
> Are you sure you have the right initrd? Is the same scsi driver being
> used as in the -xen0 case?

I suspect there may indeed be some sort of problem with PAE. I have domU RHEL4-U2
SMP 32-bit kernel that dies when it tries to set PAE bit in CR4 --- it
can't find translation for the next instruction (and for anything else, for
that matter: it keeps then getting GPFs on ignore_int's first instruction).

UP version of the same kernel and SMP versions of other kernels, such as 2.6.16,
don't have this problem.

-boris

RE: PAE dom0 is broken with the default config

[Ian Pratt]

>    I did rebuild the initrd for the new kernel/modules. The 
> system has sata disk. scsi driver is the same ata_piix, 
> libata. The same config is working fine for the non-PAE build. 
>    The VMX status report mail says PAE is fine, but they are 
> not using the default (linux-defconfig_xen_x86_32) config. 
> They are using the xen0 config for PAE testing. What config 
> do you use for your PAE testing?

The majority testing is done with the -xen kernel, which seems to boot
and run fine on all of our machines, including those with root fs on
SATA.

It might be worth comparing the scsi/acpi/pci etc options closely
between the two different configs, and try a 'binary' chop to see which
are important.

Thanks,
Ian
 
> In summary here is my observation
> PAE build with linux-defconfig_xen_x86_32 config:	dom0 does not
> boot
> PAE build with linux-defconfig_xen0_x86_32 config:	boots
> Non-PAE build with either config:				boots
> 
> All of the above kernels need corresponding initrd to boot.
> 
> Thanks & Regards,
> Nitin
> --------------------------------------------------------------
> ----------
> -----------
> Open Source Technology Center, Intel Corp
> 
> >-----Original Message-----
> >From: Ian Pratt [mailto:m+Ian.Pratt@cl.cam.ac.uk]
> >Sent: Wednesday, April 05, 2006 4:34 PM
> >To: Kamble, Nitin A; Keir.Fraser@cl.cam.ac.uk
> >Cc: xen-devel@lists.xensource.com; ian.pratt@cl.cam.ac.uk
> >Subject: RE: PAE dom0 is broken with the default config
> >
> >>   I built PAE Xen with the default config, and it is not 
> booting. Log 
> >> is bellow. If I replace the
> >> linux-defconfig_xen_x86_32 file with the
> >> linux-defconfig_xen0_x86_32 before building then dom0 is 
> able to boot 
> >> and see more than 4Gig memory. This looks like a dom0 
> kernel config 
> >> issue. Do you care PAE dom0 for 3.0.2 with the default config? Is 
> >> anybody working on this?
> >
> >This seems pretty unlikely to be a PAE issue -- we do quite 
> extensive 
> >testing.
> >
> >Are you sure you have the right initrd? Is the same scsi 
> driver being 
> >used as in the -xen0 case?
> >
> >Thanks,
> >Ian
> >
> >>
> >>
> >> IP route cache hash table entries: 262144 (order: 8, 1048576 bytes)
> >>
> >> TCP established hash table entries: 524288 (order: 10, 
> 4194304 bytes)
> >>
> >> TCP bind hash table entries: 65536 (order: 7, 524288 bytes)
> >>
> >> TCP: Hash tables configured (established 524288 bind 65536)
> >>
> >> TCP reno registered
> >>
> >> NET: Registered protocol family 8
> >>
> >> NET: Registered protocol family 20
> >>
> >> Using IPI No-Shortcut mode
> >>
> >> Freeing unused kernel memory: 172k freed
> >>
> >> scsi_mod: no version for "struct_module" found: kernel tainted.
> >>
> >> SCSI subsystem initialized
> >>
> >> ACPI: PCI Interrupt 0000:00:1f.2[A] -> GSI 18 (level, low) 
> -> IRQ 17
> >>
> >> ata1: SATA max UDMA/133 cmd 0x0 ctl 0xC402 bmdma 0xBC00 irq 17
> >>
> >> ata2: SATA max UDMA/133 cmd 0x0 ctl 0xC002 bmdma 0xBC08 irq 17
> >>
> >> ata1: dev 0 ATA-6, max UDMA/133, 312581808 sectors: LBA48
> >>
> >> ata1: dev 0 configured for UDMA/133
> >>
> >> scsi0 : ata_piix
> >>
> >> ATA: abnormal status 0x7F on port 0xC087
> >>
> >> scsi1 : ata_piix
> >>
> >>   Vendor: ATA       Model: ST3160023AS       Rev: 3.00
> >>
> >>   Type:   Direct-Access                      ANSI SCSI revision: 05
> >>
> >> SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
> >>
> >> sda: Write Protect is off
> >>
> >> SCSI device sda: drive cache: write back
> >>
> >> SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
> >>
> >> sda: Write Protect is off
> >>
> >> SCSI device sda: drive cache: write back
> >>
> >>  sda: unknown partition table
> >>
> >> sd 0:0:0:0: Attached scsi disk sda
> >>
> >> Kernel panic - not syncing: Attempted to kill init!
> >>
> >>  (XEN) Domain 0 crashed: rebooting machine in 5 seconds.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> 

Re: PAE dom0 is broken with the default config

[Keir Fraser]

On 6 Apr 2006, at 11:50, Ian Pratt wrote:

>>    I did rebuild the initrd for the new kernel/modules. The
>> system has sata disk. scsi driver is the same ata_piix,
>> libata. The same config is working fine for the non-PAE build.
>>    The VMX status report mail says PAE is fine, but they are
>> not using the default (linux-defconfig_xen_x86_32) config.
>> They are using the xen0 config for PAE testing. What config
>> do you use for your PAE testing?
>
> The majority testing is done with the -xen kernel, which seems to boot
> and run fine on all of our machines, including those with root fs on
> SATA.
>
> It might be worth comparing the scsi/acpi/pci etc options closely
> between the two different configs, and try a 'binary' chop to see which
> are important.

Yes. There's obviously an underlying bug, but it'll probably be easier 
to track down if we know what config option triggers it.

  -- Keir

RE: PAE dom0 is broken with the default config

[Kamble, Nitin A]

As Ian said earlier it is an initrd issue. The initrd file passed to the
dom0 kernel was not right, after I fixed it PAE dom0 is able to boot
with the default config.

Thanks & Regards,
Nitin
------------------------------------------------------------------------
-----------
Open Source Technology Center, Intel Corp

>-----Original Message-----
>From: Ian Pratt [mailto:m+Ian.Pratt@cl.cam.ac.uk]
>Sent: Thursday, April 06, 2006 3:50 AM
>To: Kamble, Nitin A; Keir.Fraser@cl.cam.ac.uk
>Cc: xen-devel@lists.xensource.com; ian.pratt@cl.cam.ac.uk;
>ian.pratt@cl.cam.ac.uk
>Subject: RE: PAE dom0 is broken with the default config
>
>>    I did rebuild the initrd for the new kernel/modules. The
>> system has sata disk. scsi driver is the same ata_piix,
>> libata. The same config is working fine for the non-PAE build.
>>    The VMX status report mail says PAE is fine, but they are
>> not using the default (linux-defconfig_xen_x86_32) config.
>> They are using the xen0 config for PAE testing. What config
>> do you use for your PAE testing?
>
>The majority testing is done with the -xen kernel, which seems to boot
>and run fine on all of our machines, including those with root fs on
>SATA.
>
>It might be worth comparing the scsi/acpi/pci etc options closely
>between the two different configs, and try a 'binary' chop to see which
>are important.
>
>Thanks,
>Ian
>
>> In summary here is my observation
>> PAE build with linux-defconfig_xen_x86_32 config:	dom0 does not
>> boot
>> PAE build with linux-defconfig_xen0_x86_32 config:	boots
>> Non-PAE build with either config:				boots
>>
>> All of the above kernels need corresponding initrd to boot.
>>
>> Thanks & Regards,
>> Nitin
>> --------------------------------------------------------------
>> ----------
>> -----------
>> Open Source Technology Center, Intel Corp
>>
>> >-----Original Message-----
>> >From: Ian Pratt [mailto:m+Ian.Pratt@cl.cam.ac.uk]
>> >Sent: Wednesday, April 05, 2006 4:34 PM
>> >To: Kamble, Nitin A; Keir.Fraser@cl.cam.ac.uk
>> >Cc: xen-devel@lists.xensource.com; ian.pratt@cl.cam.ac.uk
>> >Subject: RE: PAE dom0 is broken with the default config
>> >
>> >>   I built PAE Xen with the default config, and it is not
>> booting. Log
>> >> is bellow. If I replace the
>> >> linux-defconfig_xen_x86_32 file with the
>> >> linux-defconfig_xen0_x86_32 before building then dom0 is
>> able to boot
>> >> and see more than 4Gig memory. This looks like a dom0
>> kernel config
>> >> issue. Do you care PAE dom0 for 3.0.2 with the default config? Is
>> >> anybody working on this?
>> >
>> >This seems pretty unlikely to be a PAE issue -- we do quite
>> extensive
>> >testing.
>> >
>> >Are you sure you have the right initrd? Is the same scsi
>> driver being
>> >used as in the -xen0 case?
>> >
>> >Thanks,
>> >Ian
>> >
>> >>
>> >>
>> >> IP route cache hash table entries: 262144 (order: 8, 1048576
bytes)
>> >>
>> >> TCP established hash table entries: 524288 (order: 10,
>> 4194304 bytes)
>> >>
>> >> TCP bind hash table entries: 65536 (order: 7, 524288 bytes)
>> >>
>> >> TCP: Hash tables configured (established 524288 bind 65536)
>> >>
>> >> TCP reno registered
>> >>
>> >> NET: Registered protocol family 8
>> >>
>> >> NET: Registered protocol family 20
>> >>
>> >> Using IPI No-Shortcut mode
>> >>
>> >> Freeing unused kernel memory: 172k freed
>> >>
>> >> scsi_mod: no version for "struct_module" found: kernel tainted.
>> >>
>> >> SCSI subsystem initialized
>> >>
>> >> ACPI: PCI Interrupt 0000:00:1f.2[A] -> GSI 18 (level, low)
>> -> IRQ 17
>> >>
>> >> ata1: SATA max UDMA/133 cmd 0x0 ctl 0xC402 bmdma 0xBC00 irq 17
>> >>
>> >> ata2: SATA max UDMA/133 cmd 0x0 ctl 0xC002 bmdma 0xBC08 irq 17
>> >>
>> >> ata1: dev 0 ATA-6, max UDMA/133, 312581808 sectors: LBA48
>> >>
>> >> ata1: dev 0 configured for UDMA/133
>> >>
>> >> scsi0 : ata_piix
>> >>
>> >> ATA: abnormal status 0x7F on port 0xC087
>> >>
>> >> scsi1 : ata_piix
>> >>
>> >>   Vendor: ATA       Model: ST3160023AS       Rev: 3.00
>> >>
>> >>   Type:   Direct-Access                      ANSI SCSI revision:
05
>> >>
>> >> SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
>> >>
>> >> sda: Write Protect is off
>> >>
>> >> SCSI device sda: drive cache: write back
>> >>
>> >> SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
>> >>
>> >> sda: Write Protect is off
>> >>
>> >> SCSI device sda: drive cache: write back
>> >>
>> >>  sda: unknown partition table
>> >>
>> >> sd 0:0:0:0: Attached scsi disk sda
>> >>
>> >> Kernel panic - not syncing: Attempted to kill init!
>> >>
>> >>  (XEN) Domain 0 crashed: rebooting machine in 5 seconds.
>> >>
>> >>
>> >>
>> >>
>> >>
>> >>
>> >>
>> >>
>> >>
>> >>
>> >>
>> >>
>>