Hangs at boot on AMD64

Where does it hang?

A machine has 4 cpus and it hangs after having initialized them and sometimes on the following step, after having initialized pci.
I've tried to boot with acpi=off and noapic but it didn't solve.

Sorry, the system has 2 Dual core opterons, not 4 cpus.
Anyway i've tried it on a third machine with dual core opterons and it hanghed there too.

The latest messages I see are:

Initializing CPU#1
Initializing CPU#2
Brought up 4 CPUs
Initializing CPU#3
migration_cost=522
Grant table initialized
NET: registered protocol 16
ACPI: bus type pci registered
PCI: Using configuration type 1

And nothing more.
The same happened on these three dual-cpu system while 2.6.16 works well instead (tried it loading almost 8 virtual machines).
Any idea?

I'm going to try with the i386 distribution right now, to see if it behaves differently.

Well, the same kernel works fine using the i386 distro.

I have same issue on i386 machine. It gets to the same point and then fans go on like mad.

Processor is Intel Dual Core with VT enabled. With VT disabled, it works ok.

Hangs on my laptop.

nafallo@darkelf:~ $ cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 4
model name : Mobile AMD Athlon(tm) 64 Processor 2800+
stepping : 10
cpu MHz : 800.000
cache size : 1024 KB
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext lm 3dnowext 3dnow up
bogomips : 1605.21
TLB size : 1024 4K pages
clflush size : 64
cache_alignment : 64
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp

Chuck, what info do you want from us? :-)

Have you tried booting with noapic?

Yes. Didn't work. And as reported above 2.6.16 works. So what bits have changed in 2.6.17? :-)

MSI-K9N AMD- 64x2 4600

Xen kernel starts to boot,
runs scripts/local-premount
then scripts/local-bottom and
after scripts/init-bottom it hangs.

xen 2.6.16.29 binary works fine.

reports of similar problems can be found at link below;

http://ubuntuforums.org/showthread.php?t=288993

BTW

when I tried to make initrd with yaird if gave an error.
mkinitrd did not give an error.

perhaps this has some relevance?

I am also having similar issues with xen-source-2.6.17 on my amd64 system. The system panics shortly after starting the xen kernel. unfortunatly, the system panics and reboots to quickly for me to gather any meaningful information. I will attempt to rememdy this and post further information.
System mainboard is an MSI K8N Neo2 Platinum nForce3, cpu is an AMD Athlon64 3500+

I am having the same problem on an HP DL385 system. The 2.6.16 xen kernel package works (although it is misnamed, and should probably have a convenience package to fix the naming conventions).

xen-image-xen0-2.6.16-11.2-generic works

xen-image-xen0-2.6.17-6-generic-xen0 hangs as described above (though always after the four CPUs have been reported, the progress after seems inconsistent with each reboot. It never appears to make it to init.)

I should clarify that xen-image-xen0-2.6.16-11.2-generic works for the host kernel, but seems unable to start any guests. I'll file a separate bug about this later.

There seem to be enough cases of this bug to confirm it. The previous xen-image (2.6.16) is reported to work fine. It's the newer one (2.6.17) that hangs at boot. "noapic" does not solve the issue. Users are willing to provide more information as necessary.

Can you try booting with initcall_debug=1? Hopefully this will narrow down where it is hanging without having to build custom kernels for you to try booting.

Seem to have the same problem. Kernel xen0-linux-2.6.17-6-generic-xen0 + initrd. Enclosed is the complete console output, might contain a hint? Changing memory, maxcpus or selinux parameters gves not result

Can you try booting after adding "noreplacement" to the kernel command line?

I should have an update for the xen kernel soonish. I will need help testing.

chuck

Any status on this? I have three systems (one AMD Turion 64X2 and two Intel C2D) that exhibit the behavior, and would be happy to help testing in anyway.

Any updates? I would be interested in assisting with testing as well.

Is there any further news on this xen problem? Willing to help with testing.

Is Xen able to support LVM2? All my partitions except for /boot are lvm2.

My system hangs when I boot Xen, sometimes as described in this bug, sometimes when it is examing the disks. Thanks.

Hi,
I have a Core 2 Duo T7200 (it's VT). I run Edgy amd64 on it.
I have the same Kernel panic as described, followed by a reboot.
What's up now? Has someone found the cause?

Same issue as ddave01 on my new turionX2. It just doesn't do anything after scripts/init-bottom

After adding the 'noreplacement' option like Kyle suggested, i ran into the kernel panic error everyone else is talking about

Similar to ddabe and RJ.

Running Sun SunFire V20Z.

Xen kernel starts to boot,
runs scripts/local-premount
then scripts/local-bottom and
after scripts/init-bottom it hangs.

CTRL-ALT-DELETE is picked up for a reboot, "Sync SCSI Data" or something like that is shown before the reboot. Everything running off LVM2 on SCSI LSI RAID card.

Hi,
I saw Feisty has this:
http://packages.ubuntu.com/feisty/devel/xen-headers-2.6.19-1-generic-amd64

Has someone tested?
I think this bug will desperately be unsolved...

I meant:
http://packages.ubuntu.com/feisty/source/xen-source

Please test in feisty.

I am now running Feisty Faun. KVM worked like a dream until I
upgraded. I believe KVM is the way to go (once Feisty is released )
if you have a HVM dual core processor.

ddave

On 1/20/07, Mark Petersen <mpetersen@1sync.org> wrote:
> Similar to ddabe and RJ.
>
> Running Sun SunFire V20Z.
>
> Xen kernel starts to boot,
> runs scripts/local-premount
> then scripts/local-bottom and
> after scripts/init-bottom it hangs.
>
> CTRL-ALT-DELETE is picked up for a reboot, "Sync SCSI Data" or something
> like that is shown before the reboot. Everything running off LVM2 on
> SCSI LSI RAID card.
>
> --
> Hangs at boot on AMD64
> https://launchpad.net/bugs/65788
>

Anyone hacking panics might want to rebuild their initrds.

for me this method (will blow away your current initrd so be careful) works for me:

# update-initramfs -k 2.6.22-11-xen -c
# update-grub

I may be having the same problem on gutsy x86 on a Core 2 Duo T7200 Toshiba laptop.

In my case the linux-image-xen kernel boots fine as long as the virtualization support is turned off in the BIOS, but when I turn it on in the BIOS, boot on dom0 hangs. The last console output is:

  Setting up standard PCI resources

I note that on a successful boot, this line comes right before some ACPI stuff:

  ACPI: EC: Look up EC in DSDT
  ACPI: Interpreter enabled
  ACPI: (supports S0 S3 S4 S5)
  ACPI: Using IOAPIC for interrupt routing

I'll be trying to boot with ACPI off next.

Booting with virtualization support enabled in the BIOS and "acpi=off" passed to xen works. It works without any ACPI support, which on a laptop is pretty much unacceptable. I also tried "acpi=noirq" but that boot hanged as well.

Some more details:

noacpi -> Hangs
acpi=off -> Boots, no ACPI
acpi=force -> Hangs
acpi=ht -> Boots, no ACPI
acpi=noirq -> Hangs

Still no combination that boots and gives me ACPI. Trying acpi=strict now.

Few more:

acpi -> Hangs
acpi=strict -> Hangs
acpi=1 -> Hangs

This bug is old and probably no longer applies.

Regards
chuck