[Xen-devel] Xen 4.0.4, kernel 3.5.0 HVM crash and kernel BUG

[Valtteri Kiviniemi <kiviniemi.valtteri@xxxxxxxxx>]

Hi,

I'm trying to get Windows Server 2008 R2 installation booting on Xen 4.0.4. Using the following config:

kernel = "/usr/lib/xen/boot/hvmloader"
builder = "hvm"
shadow_memory = "8"
memory = "4096"
name = "ts"
vcpus = "8"
cpus = ["0", "1", "2", "3", "4", "5", "6", "7"]
pae = "1"
acpi = "1"
apic = "1"
vfb = [ 'type=vnc, vnclisten=10.100.100.1, vncpasswd=xxxx' ]
xen_extended_power_mgmt = "0"
vif = [ "type=ioemu, mac=00:16:3e:d7:d7:5d, bridge=xenbr0" ]
disk = [ "phy:/dev/virtuals/ts,hda,w", "file:/media/iso/windows_server_2008_r2_sp1.iso,hdc:cdrom,r" ]
>>>
viridian = "1"
device_model = "/usr/lib/xen/bin/qemu-dm"
boot = "dc"
snapshot = "0"
vncconsole = "1"
sdl = "0"
opengl = "0"
vnc = "1"
nographic = "0"
stdvga = "0"
tsc_mode = "1"
monitor = "0"
localtime = "1"
usb = "0"
keymap = "fi"
xen_platform_pci = "1"
pci_msitranslate = "1"
pci_power_mgmt = "0"

The domU will start booting just fine, but after a few minutes the VNC screen goes black. After that when typing "xm destroy ts" it will trigger a kernel BUG:

BUG: unable to handle kernel NULL pointer dereference at 0000000000000030
IP: [<ffffffff810c50c4>] iput+0x3e/0x195
PGD 0
Oops: 0000 [#1] SMP
CPU 6
Pid: 3571, comm: qemu-dm Not tainted 3.5.0-dom0 #1                  /DQ77MK
RIP: e030:[<ffffffff810c50c4>]  [<ffffffff810c50c4>] iput+0x3e/0x195
RSP: e02b:ffff8800389ffbf8  EFLAGS: 00010246
RAX: 0000000000000001 RBX: ffff8800377b0720 RCX: ffff8800501c0000
RDX: ffff8800501c0000 RSI: ffff8800377b0790 RDI: ffff8800377b0790
RBP: 0000000000000000 R08: ffffffff815cdd00 R09: 0000000000000016
R10: fefefefefefefeff R11: ffff8800377b0400 R12: 00000001000a3e0c
R13: 0000000000000000 R14: 00000001000a3e0c R15: ffff8800389ffc28
FS:  00007f1af70a8700(0000) GS:ffff880050180000(0000) knlGS:0000000000000000
CS:  e033 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000000030 CR3: 000000000156d000 CR4: 0000000000002660
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process qemu-dm (pid: 3571, threadinfo ffff8800389fe000, task ffff88003a721260)
Stack:
 ffff88003a6d6400 ffff8800377b0000 00000001000a3e0c ffffffff8133ce8f
 ffff8800377b0400 ffffffff8134b6cd ffff8800389ffc28 ffff8800389ffc28
 ffff8800377b00f8 ffff8800377b0680 ffff880038cdcd60 ffff8800377b0000
Call Trace:
 [<ffffffff8133ce8f>] ? sk_release_kernel+0x23/0x39
 [<ffffffff8134b6cd>] ? netdev_run_todo+0x1e9/0x206
 [<ffffffff8129798f>] ? tun_chr_close+0x4c/0x7b
 [<ffffffff810b39d3>] ? fput+0xe4/0x1c5
 [<ffffffff810b202c>] ? filp_close+0x61/0x68
 [<ffffffff81035e62>] ? put_files_struct+0x62/0xb9
 [<ffffffff81036374>] ? do_exit+0x24a/0x74c
 [<ffffffff81036906>] ? do_group_exit+0x6b/0x9d
 [<ffffffff8103ea0b>] ? get_signal_to_deliver+0x449/0x46e
 [<ffffffff81009fa5>] ? do_signal+0x28/0x4c4
 [<ffffffff81027079>] ? pvclock_clocksource_read+0x48/0xbf
 [<ffffffff8105b745>] ? ktime_get_ts+0x66/0xa8
 [<ffffffff810bfb18>] ? poll_select_copy_remaining+0xe0/0xf5
 [<ffffffff8100a48d>] ? do_notify_resume+0x3b/0x74
 [<ffffffff81411a70>] ? int_signal+0x12/0x17
Code: 00 00 00 40 74 02 0f 0b 48 8d 77 70 48 8d bf 08 01 00 00 e8 8b 71 10 00 85 c0 0f 84 5d 01 00 00 48 8b 6b 18 f6 83 80 00 00 00 08 <4c> 8b 65 30 74 11 be 68 05 00 00 48 c7 c7 8e df 4f 81 e8 bb d0
RIP  [<ffffffff810c50c4>] iput+0x3e/0x195
 RSP <ffff8800389ffbf8>
CR2: 0000000000000030
---[ end trace 67cc1654658fedcc ]---
Fixing recursive fault but reboot is needed!

I also tested Xen 4.2.0 and problem is the same. So is this a Xen bug or a kernel bug? I am running vanilla kernel.org kernel 3.5.0 and my hardware is Intel Core i7-3770 CPU and Intel DQ77MK motherboard with latest bios.

Best regards,
Valtteri Kiviniemi
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel