Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

elewis · 2016-04-15 00:52:22

I upgraded my Arch in Virtual Box from a successful upgrade made a few days via 'sudo pacman -Syu' with the result that Arch will not start at all.
After launching, and passing through GRUB, a screen appears with three messages (one of which I have seen for a long time)
[0.854749] Failed to Find cpu0 device mode. (This has appeared for several months; it has caused no issues.)
[4.690631] piix4_smbus 0000:00:7.0 SMBus base uninitialized - upgrade BIOS or use force_addr=0xaddr (I never noticed this before)
[5.116758] intel_rapl: no valid rapl domains found in package 0 (I never noticed this before)
At this point, the boot process stops. I have attempted the following:
Boot the VM with gparted. Mount the guest Arch VM and edit the line /etc/modprobe.d/fsub-blacklist.conf by adding the lines: blacklist i2c_piix4; blacklist piix4_smbus; blacklist intel_rapl.
This activity does nothing. I have also tried enabling (from the VirtualBox GUI) pae/nx (it was not enabled in the previous working installation. I have also enabled and disabled 'nested paging'.
None of the above measures has had any impact on the startup sequence for Arch. Startup fails, with the black screen and the three messages noted above. it is not possible to start Arch at all at this time.
I have a backup version created on February 1 (immediately after upgrade to Plasma 5 Desktops) which mounts and starts fine. After the upgrade to Linux 4.5 (4/18/16) startup fails.
Note that a machine that boots natively with Arch works fine, even after the upgrade to Linux 4.5. This issue appears related to VirtualBox hosting.
Any help would be appreciated.

fredbezies · 2016-04-15 06:08:21

Related to https://bugs.archlinux.org/task/48598 ?

elewis · 2016-04-15 21:55:42

Looks similar. In my case, it does not appear that any drive or operating system is loading at all. Thanks.

kartiman · 2016-04-17 15:15:00

I have experienced the same halt on boot process due to the same reasons and under the same conditions.
I have tried disabling (from the VirtualBox GUI) pae/nx (it was enabled in the previous working installation; I have also enabled and disabled 'nested paging'; both, unfortunately, have had no success.
I have not attempted any other strategy to circumvent the problem.
It is not possible to start Arch at all at this time.

torkjel · 2016-04-22 07:35:23

I'm having the same problem. Worked around it by booting the install iso and downgrading the kernel and virtualbox guest stuff as described on the wiki: https://wiki.archlinux.org/index.php/do … the_kernel

pacman -U linux-4.4.5-1-x86_64.pkg.tar.xz linux-headers-4.4.5-1-x86_64.pkg.tar.xz virtualbox-guest-dkms-5.0.16-3-x86_64.pkg.tar.xz virtualbox-guest-utils-5.0.16-3-x86_64.pkg.tar.xz

The log from the last boot that did not work shows that the problem is sddm crashing:

sudo journalctl --boot=-1

Apr 22 08:55:44 arch.my.vm sddm[410]: Running display setup script "/usr/share/sddm/scripts/Xsetup"
Apr 22 08:55:44 arch.my.vm sddm[410]: Display server started.
Apr 22 08:55:44 arch.my.vm sddm[410]: Socket server starting...
Apr 22 08:55:44 arch.my.vm sddm[410]: Socket server started.
Apr 22 08:55:44 arch.my.vm sddm[410]: Greeter starting...
Apr 22 08:55:44 arch.my.vm sddm[410]: Adding cookie to "/var/run/sddm/{74c276bd-89a3-4087-b540-74744107494f}"
Apr 22 08:55:44 arch.my.vm sddm[410]: /usr/bin/xauth: (stdin):1: bad "remove" command line
Apr 22 08:55:44 arch.my.vm sddm[410]: /usr/bin/xauth: (stdin):2: bad "add" command line
Apr 22 08:55:44 arch.my.vm sddm[410]: QProcess: Destroyed while process ("/usr/lib/sddm/sddm-helper") is still running.
Apr 22 08:55:44 arch.my.vm kernel: sddm[410]: segfault at 0 ip 00007fb6a1ec4f23 sp 00007ffe081257e0 error 4 in libQt5Core.so.5.6.0[7fb6a1c14000+4ba000]
Apr 22 08:55:44 arch.my.vm sddm-helper[442]: [PAM] Starting...
Apr 22 08:55:44 arch.my.vm sddm-helper[442]: [PAM] Authenticating...
Apr 22 08:55:44 arch.my.vm sddm-helper[442]: [PAM] returning.
Apr 22 08:55:44 arch.my.vm systemd[1]: Started Process Core Dump (PID 444/UID 0).
Apr 22 08:55:44 arch.my.vm systemd-coredump[445]: Removed old coredump core.sddm.0.1c1414a35fad49d8b61ba8fa4dbea169.288.1461308143000000000000.lz4.
Apr 22 08:55:44 arch.my.vm sddm-helper[442]: Received a wrong opcode instead of AUTHENTICATED: 0
Apr 22 08:55:44 arch.my.vm sddm-helper[442]: pam_unix(sddm-greeter:session): session opened for user sddm by (uid=0)
Apr 22 08:55:44 arch.my.vm systemd[1]: sddm.service: Main process exited, code=dumped, status=11/SEGV
Apr 22 08:55:44 arch.my.vm systemd[1]: sddm.service: Unit entered failed state.
Apr 22 08:55:44 arch.my.vm systemd[1]: sddm.service: Failed with result 'core-dump'.
Apr 22 08:55:44 arch.my.vm systemd-coredump[445]: Process 410 (sddm) of user 0 dumped core.

Stack trace of thread 410:
#0 0x00007fb6a1ec4f23 _ZN7QObject7connectEPKS_PKcS1_S3_N2Qt14ConnectionTypeE (libQt5Core.so.5)
#1 0x000000000042fed5 _ZN4SDDM4Auth7Private9setSocketEP12QLocalSocket (sddm)
#2 0x0000000000431c24 _ZN4SDDM4Auth12SocketServer19handleNewConnectionEv (sddm)
#3 0x00007fb6a1ebe870 _ZN11QMetaObject8activateEP7QObjectiiPPv (libQt5Core.so.5)
#4 0x00007fb6a23c2bbf n/a (libQt5Network.so.5)
#5 0x00007fb6a1ebe870 _ZN11QMetaObject8activateEP7QObjectiiPPv (libQt5Core.so.5)
#6 0x00007fb6a1f3a93e _ZN15QSocketNotifier9activatedEiNS_14QPrivateSignalE (libQt5Core.so.5)
#7 0x00007fb6a1ecb44b _ZN15QSocketNotifier5eventEP6QEvent (libQt5Core.so.5)
#8 0x00007fb6a1e923ba _ZN16QCoreApplication6notifyEP7QObjectP6QEvent (libQt5Core.so.5)
#9 0x00007fb6a1e92518 _ZN16QCoreApplication15notifyInternal2EP7QObjectP6QEvent (libQt5Core.so.5)
#10 0x00007fb6a1ee814d n/a (libQt5Core.so.5)
#11 0x00007fb69e39af07 g_main_context_dispatch (libglib-2.0.so.0)
#12 0x00007fb69e39b160 n/a (libglib-2.0.so.0)
#13 0x00007fb69e39b20c g_main_context_iteration (libglib-2.0.so.0)
#14 0x00007fb6a1ee7c8f _ZN20QEventDispatcherGlib13processEventsE6QFlagsIN10QEventLoop17ProcessEventsFlagEE (libQt5Core.s
#15 0x00007fb6a1e906fa _ZN10QEventLoop4execE6QFlagsINS_17ProcessEventsFlagEE (libQt5Core.so.5)
#16 0x00007fb6a1e98b7c _ZN16QCoreApplication4execEv (libQt5Core.so.5)
#17 0x000000000041a786 main (sddm)
#18 0x00007fb6a12fb710 __libc_start_main (libc.so.6)
#19 0x000000000041aa69 _start (sddm)

Stack trace of thread 414:
#0 0x00007fb6a13b9c3d poll (libc.so.6)
#1 0x00007fb69e39b0fc n/a (libglib-2.0.so.0)
#2 0x00007fb69e39b20c g_main_context_iteration (libglib-2.0.so.0)
#3 0x00007fb6a1ee7c8f _ZN20QEventDispatcherGlib13processEventsE6QFlagsIN10QEventLoop17ProcessEventsFlagEE (libQt5Core.s
#4 0x00007fb6a1e906fa _ZN10QEventLoop4execE6QFlagsINS_17ProcessEventsFlagEE (libQt5Core.so.5)
#5 0x00007fb6a1cb9fe4 _ZN7QThread4execEv (libQt5Core.so.5)
#6 0x00007fb6a2c50a55 n/a (libQt5DBus.so.5)
#7 0x00007fb6a1cbede8 n/a (libQt5Core.so.5)
#8 0x00007fb6a10c5424 start_thread (libpthread.so.0)
#9 0x00007fb6a13c2cbd __clone (libc.so.6)
Apr 22 08:55:44 arch.my.vm systemd[1]: sddm.service: Service hold-off time over, scheduling restart.
Apr 22 08:55:44 arch.my.vm systemd[1]: Stopped Simple Desktop Display Manager.
Apr 22 08:55:44 arch.my.vm systemd[1]: Started Simple Desktop Display Manager.

fredbezies · 2016-04-22 07:42:53

The only true fix is to get VirtualBox 5.0.18 on repositories. You can apply these workarounds, but it will be short term answer

torkjel · 2016-04-22 07:48:42

Luckily, if past experience is anything to go by, we can trust the arch devs to get 5.0.18 into the repos, in the short term

fredbezies · 2016-04-22 09:18:02

torkjel wrote:

Luckily, if past experience is anything to go by, we can trust the arch devs to get 5.0.18 into the repos, in the short term

It will be great indeed. Maybe with kernel 4.5.2-1 which lands in testing yesterday... Please, developpers

bhrgunatha · 2016-05-11 10:27:59

I recently upgraded to VirtualBox 5.0.20 (r106931) on a Windows 7 host (5.0.20 is in the repos now).
I just tried rebooting my Arch VM and I get the same problem.
No settings for Setting pae/nx (on or off) work. Same with 3d acceleration (on or off). Same with nested paging.

Sometimes it hangs directly after the message
intel_rapl: no valid rapl domains found in package 0
Sometimes it hangs directy after the message
piix4_smbus 0000:00:7.0 SMBus base uninitialized - upgrade BIOS or use force_addr=0xaddr

It's been a while since I last booted the Arch VM - I don't even remeber which kernel version is currently installed.

Is booting the install ISO and downgrading the kernel and virtualbox guest additions the only solution?

Last edited by bhrgunatha (2016-05-11 10:29:05)

entodoays · 2016-05-26 15:04:56

I tried installing Arch in VirtualBox and am having this same error message (intel_rapl: no valid rapl domains found in package 0 ) on boot.
Blacklisting the module and recreating initramfs removes the error but the VM still doesn't boot.

Arch Linux

#1 2016-04-15 00:52:22

Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

#2 2016-04-15 06:08:21

Re: Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

#3 2016-04-15 21:55:42

Re: Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

#4 2016-04-17 15:15:00

Re: Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

#5 2016-04-22 07:35:23

Re: Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

#6 2016-04-22 07:42:53

Re: Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

#7 2016-04-22 07:48:42

Re: Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

#8 2016-04-22 09:18:02

Re: Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

#9 2016-05-11 10:27:59

Re: Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

#10 2016-05-26 15:04:56

Re: Upgrading to Linux 4.5 (pacman -Syu) causes boot failure in VirtualBox

Board footer