Re: KVM guest crashes

[Alexander Graf <agraf@suse.de>]

Avi Kivity wrote:
> Avi Kivity wrote:
>>
>> I suggest checking if you have the latest BIOS update applied.  I've
>> had bad experiences with un-updated processors.
>>
>
> FWIW, I have an 8-way F9 guest (2.6.27.5-blah) running on an 2x4
> Barcelona host, happily make -j16ing an allmodconfig kernel.

Strange. I started the tests again with an updated BIOS now, installing
an Intel machine to test on in parallel.

old:

# ./rdmsr /dev/cpu/0/msr $(( 0x0000008b ))
0x1000065

new:

# ./rdmsr /dev/cpu/0/msr $(( 0x0000008b ))
0x1000083


But I already got one guest crashing:

int3: 0000 [1] SMP
last sysfs file: /sys/kernel/uevent_seqnum
CPU 2
Modules linked in: nls_utf8 cifs(X) af_packet virtio_net virtio_pci
virtio_ring virtio edd ext3 mbcache jbd fan ide_pci_generic ide_core
ata_generic sata_nv libata scsi_mod dock thermal processor thermal_sys
 hwmon
Supported: Yes, External
Pid: 0, comm: swapper Tainted: G S        2.6.27.7-9-default #1
RIP: 0010:[<ffffffff80a500f1>]  [<ffffffff80a500f1>]
per_cpu__cpu_state+0x1/0x4
RSP: 0018:ffff88007a493fa8  EFLAGS: 00000083
RAX: ffffffff806f5fa0 RBX: ffffffff80a500f0 RCX: 0000000000000000
RDX: ffff880001033200 RSI: 0000000000000000 RDI: ffffffffff5fc0b0
RBP: ffff88007a48beb0 R08: 0000000000000000 R09: ffff880001039638
R10: 00000000ffffffff R11: ffffffff8021c5d9 R12: 0000000000000000
R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000
FS:  00007fe3252e4950(0000) GS:ffff88007a461f40(0000) knlGS:0000000000000000
CS:  0010 DS: 0018 ES: 0018 CR0: 000000008005003b
CR2: 000000000062d000 CR3: 000000007c10a000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process swapper (pid: 0, threadinfo ffff88007a48a000, task ffff88007a488280)
Stack:  ffff88007a48beb0 ffffffff8020ca2e ffff88007a48beb0 <EOI> 
0000007dd83ce327
 0000000000000086 ffff8800010396d0 0000000002625a00 0000000000000002
 000000010000eadc 0000007dd83ce327 0000000000000292 0000000000000292
Call Trace:
Inexact backtrace:

 <IRQ>  [<ffffffff8020ca2e>] ? ret_from_intr+0x0/0x29
 <EOI>  [<ffffffff804a6992>] ? notifier_call_chain+0x29/0x4c
 [<ffffffff80213465>] ? default_idle+0x38/0x54
 [<ffffffff8020b34a>] ? cpu_idle+0xa9/0xf1


Code: cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc
cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc <cc> cc
cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc cc
RIP  [<ffffffff80a500f1>] per_cpu__cpu_state+0x1/0x4
 RSP <ffff88007a493fa8>
---[ end trace 17313f34f216af07 ]---
Kernel panic - not syncing: Attempted to kill the idle task!
------------[ cut here ]------------
WARNING: at kernel/smp.c:331 smp_call_function_mask+0x38/0x1f2()
Modules linked in: nls_utf8 cifs(X) af_packet virtio_net virtio_pci
virtio_ring virtio edd ext3 mbcache jbd fan ide_pci_generic ide_core
ata_generic sata_nv libata scsi_mod dock thermal processor thermal_sys
 hwmon
Supported: Yes, External
Pid: 0, comm: swapper Tainted: G S    D   2.6.27.7-9-default #1

Call Trace:
 [<ffffffff8020e42e>] show_trace_log_lvl+0x41/0x58
 [<ffffffff804a1e97>] dump_stack+0x69/0x6f
 [<ffffffff80240eb2>] warn_on_slowpath+0x51/0x77
 [<ffffffff80261fef>] smp_call_function_mask+0x38/0x1f2
 [<ffffffff802621d2>] smp_call_function+0x29/0x2e
 [<ffffffff8021ba16>] native_smp_send_stop+0x1a/0x3f
 [<ffffffff804a1f59>] panic+0xbc/0x170
 [<ffffffff802449e2>] do_exit+0x6b/0x334
 [<ffffffff804a4b9b>] oops_begin+0x0/0x9e
 [<ffffffff804a524a>] do_int3+0x7d/0xa1
 [<ffffffff804a46e6>] int3+0xb6/0xf0
 [<ffffffff80a500f1>] per_cpu__cpu_state+0x1/0x4
DWARF2 unwinder stuck at per_cpu__cpu_state+0x1/0x4

Leftover inexact backtrace:

 <IRQ>  [<ffffffff8020ca2e>] ret_from_intr+0x0/0x29
 <EOI>  [<ffffffff804a6992>] notifier_call_chain+0x29/0x4c
 [<ffffffff80213465>] default_idle+0x38/0x54
 [<ffffffff8020b34a>] cpu_idle+0xa9/0xf1

---[ end trace 17313f34f216af07 ]---


The UP guests seemed to work fine - will start them again now.

Alex