Host crashed after a processor offline

[mzprx]

I'm running VirtualBox 4.3.2 on this host:

• OS: OpenIndiana 151a8
• CPU: Intel Xeon E3-1240 V2

I had one virtual machine guest running.

Once I ran psradm -f 1 on the host to offline one CPU the machine panicked with this info:

> ::status
debugging crash dump vmcore.0 (64-bit) from raven
operating system: 5.11 oi_151a8 (i86pc)
image uuid: bce29343-bd66-4abc-9670-96f69cb1d15c
panic message: BAD TRAP: type=6 (#ud Invalid opcode) rp=ffffff001f36aab0 addr=1
dump content: kernel pages only
> ::stack
0xfffffffff822afa4()
0xfffffffff822aae6()
0xfffffffff823d6bb()
supdrvIOCtlFast+0x8c()
VBoxDrvSolarisIOCtl+0x109()
cdev_ioctl+0x45(10e00000002, 200056c1, 1, 202003, ffffff04e462a468, ffffff001f36ae24)
spec_ioctl+0x5a(ffffff04e65e5700, 200056c1, 1, 202003, ffffff04e462a468, ffffff001f36ae24)
fop_ioctl+0x7b(ffffff04e65e5700, 200056c1, 1, 202003, ffffff04e462a468, ffffff001f36ae24)
ioctl+0x18e(12, 200056c1, 1)
sys_syscall+0x17a()
> ::cpuinfo
 ID ADDR             FLG NRUN BSPL PRI RNRN KRNRN SWITCH THREAD           PROC
  0 fffffffffbc30640  1f    0    0  -1   no    no t-0    ffffff001e005c40 (idle)
  1 ffffff04d865e080  2f    0    0  -1   no    no t-10   ffffff001e364c40 (idle)
  2 ffffff04d8642a80  1f    0    0  59   no    no t-0    ffffff04db148ae0 VBoxHeadless
  3 ffffff04d8ac8540  1f    0    0  -1   no    no t-0    ffffff001e4a2c40 (idle)
  4 ffffff04d8ac7040  1f    0    0  -1   no    no t-0    ffffff001e852c40 (idle)
  5 ffffff04d8ac5080  1f    0    0  -1   no    no t-4    ffffff001e454c40 (idle)
  6 fffffffffbc3ae50  1b    0    0  59   no    no t-0    ffffff04da579bc0 VBoxHeadless
  7 ffffff04d8aba540  1f    1    0  -1   no    no t-9    ffffff001e97bc40 (idle)
>

I'll provide the crash dump file on request.

[Ramshankar Venkataraman]

Could you please attach VBox.log for the VM when the crash happened.

[mzprx]

Last night I had two similar panices:

# ls -al vmdump.*
-rw-r--r-- 1 root root 392953856 2013-11-24 03:21 vmdump.1
-rw-r--r-- 1 root root 320471040 2013-11-24 04:04 vmdump.2
#

Now it was not caused by processor offline on the host, so it is possible that guest somehow caused it (there are some weekly cron jobs running on guest approximately during the time when the host panicked).

Here is the panic stack from vmdump.1 (vmdump.2 have the exactly same panic stack):

> ::status
debugging crash dump vmcore.1 (64-bit) from raven
operating system: 5.11 oi_151a8 (i86pc)
image uuid: a7b255ac-d818-e7f7-9d8c-88b0e1f5ab4a
panic message: 
BAD TRAP: type=e (#pf Page fault) rp=ffffff0024979930 addr=28 occurred in module "<unknown>" due to a NULL pointer dereferenc
e
dump content: kernel pages only
> ::stack
vpanic()
die+0xdd(e, ffffff0024979930, 28, 0)
trap+0x17db(ffffff0024979930, 28, 0)
0xfffffffffb8001d6()
0xfffffffff82938db()
0xfffffffff82c1e14()
0xfffffffff82c18a8()
0xfffffffff8240d53()
0xfffffffff82317aa()
0xfffffffff823124b()
0xfffffffff82c3667()
0xfffffffff823c6f5()
supdrvIOCtlFast+0x8c()
VBoxDrvSolarisIOCtl+0x109()
cdev_ioctl+0x45(10e00000002, 200056c1, 1, 202003, ffffff04e52ba190, ffffff0024979e24)
spec_ioctl+0x5a(ffffff04e78c8140, 200056c1, 1, 202003, ffffff04e52ba190, ffffff0024979e24)
fop_ioctl+0x7b(ffffff04e78c8140, 200056c1, 1, 202003, ffffff04e52ba190, ffffff0024979e24)
ioctl+0x18e(12, 200056c1, 1)
sys_syscall+0x17a()
>

Please find attached all four VBox.log files I found in the Logs directory for the guest machine.

[mzprx]

I just found that I'm not running 4.3.2, but 4.3.0. Sorry for confusion. I'll upgrade to 4.3.2 to see whether it helps.

[aeichner]

Please reopen if still relevant with a recent VirtualBox release.