System randomly won't complete boot

[teo1978]

Since the incomplete/broken "fix" for #1431753 was released, and hence NVidia drivers version 340 were installed AND enabled on my machine (where previously I had 331 installed and disabled) this has started happening:

About 50-75% of the times (very rough estimation), when I boot, the boot gets stuck at the purple screen with the "Ubuntu" word with four red dots below.

Systematically, when this happens, that screen is preceded by a black screen with a lot of errors, of which I attach a picture (yes, I know, it's unreadable, but it's all I can do).
If anybody tells me in which log files those errors can be found, I will be happy to attach them.

Normally, that is, when the issue does not happen and the boot is succesful, the black screen preceding the purple screen would only contain one line with an unrelated error that has to do with USB.

ProblemType: Bug
DistroRelease: Ubuntu 14.04
Package: unity 7.2.5+14.04.20150603-0ubuntu1
ProcVersionSignature: Ubuntu 3.13.0-61.100-generic 3.13.11-ckt22
Uname: Linux 3.13.0-61-generic x86_64
NonfreeKernelModules: nvidia
ApportVersion: 2.14.1-0ubuntu3.11
Architecture: amd64
CompizPlugins: No value set for `/apps/compiz-1/general/screen0/options/active_plugins'
CurrentDesktop: Unity
Date: Mon Aug 10 21:15:32 2015
InstallationDate: Installed on 2013-10-11 (668 days ago)
InstallationMedia: Ubuntu 13.04 "Raring Ringtail" - Release amd64 (20130424)
SourcePackage: unity
UpgradeStatus: Upgraded to trusty on 2014-05-24 (443 days ago)

[teo1978 (teo8976)]

Inability to boot after the update of NVidia drivers to 340 has been confirmed by user Anna at https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-340-updates/+bug/1431753/comments/104

[Alberto Milone (albertomilone)]

please reproduce the problem and attach the following files:

/var/log/Xorg.0.log
/var/log/Xorg.0.log.old
/var/log/kern.log
/var/log/gpu-manager.log

[teo1978 (teo8976)]

/var/log/Xorg.0.log.old is empty
/var/log/gpu-manager.log doesn't exist

[teo1978 (teo8976)]

When this happens, any of these may happen:

A) I force-power-off the computer, try to boot again in the exact same conditions (not plugging/unplugging AC power nor mouse nor external screen), and at the next attempt most of the times it boots normally

B) I do the same as A, but retry for six-seven times and it NEVER completes boot. So, I disconnect the external screen if it was connected, or connect it if it wasn't, and then at the next attempt it works. It may be random, but the odds don't seem high.

[Graham Inggs (ginggs)]

From kern.log in comment #6

Aug 10 13:24:50 xxx1 kernel: [ 14.150733] BUG: unable to handle kernel NULL pointer dereference at (null)
Aug 10 13:24:50 xxx1 kernel: [ 14.150737] IP: [<ffffffff81729f8b>] __down_common+0x4c/0x144
Aug 10 13:24:50 xxx1 kernel: [ 14.150738] PGD 95f32067 PUD 95f31067 PMD 0
Aug 10 13:24:50 xxx1 kernel: [ 14.150739] Oops: 0002 [#1] SMP
Aug 10 13:24:50 xxx1 kernel: [ 14.150753] Modules linked in: acer_wmi(+) sparse_keymap intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel snd_hda_intel(+) aes_x86_64 arc4 lrw gf128mul snd_hda_codec glue_helper ablk_helper cryptd snd_hwdep snd_pcm snd_page_alloc ath3k ath9k btusb snd_seq_midi ath9k_common bluetooth snd_seq_midi_event ath9k_hw snd_rawmidi joydev ath serio_raw mac80211 snd_seq cfg80211 snd_seq_device nvidia(POX+) i915(+) snd_timer lpc_ich snd wmi mei_me video drm_kms_helper drm mei soundcore mac_hid shpchp i2c_algo_bit parport_pc ppdev lp parport hid_generic usbhid hid tg3 psmouse ahci sdhci_pci ptp libahci sdhci pps_core
Aug 10 13:24:50 xxx1 kernel: [ 14.150755] CPU: 2 PID: 533 Comm: nvidia-persiste Tainted: P OX 3.13.0-61-generic #100-Ubuntu
Aug 10 13:24:50 xxx1 kernel: [ 14.150755] Hardware name: Acer Aspire V3-571G/VA50_HC_CR, BIOS V2.07 10/15/2012
Aug 10 13:24:50 xxx1 kernel: [ 14.150756] task: ffff880245d14800 ti: ffff880093fec000 task.ti: ffff880093fec000
Aug 10 13:24:50 xxx1 kernel: [ 14.150757] RIP: 0010:[<ffffffff81729f8b>] [<ffffffff81729f8b>] __down_common+0x4c/0x144
Aug 10 13:24:50 xxx1 kernel: [ 14.150758] RSP: 0018:ffff880093fedb48 EFLAGS: 00010096
Aug 10 13:24:50 xxx1 kernel: [ 14.150759] RAX: 0000000000000000 RBX: ffffffffa0b49430 RCX: 0000000000000000
Aug 10 13:24:50 xxx1 kernel: [ 14.150759] RDX: ffffffffa0b49438 RSI: ffff880093fedb50 RDI: ffffffffa0b49430
Aug 10 13:24:50 xxx1 kernel: [ 14.150760] RBP: ffff880093fedb98 R08: 0000000000000296 R09: ffffffffa07478eb
Aug 10 13:24:50 xxx1 kernel: [ 14.150760] R10: 0000000000000030 R11: 0000000000000030 R12: 7fffffffffffffff
Aug 10 13:24:50 xxx1 kernel: [ 14.150760] R13: ffff880245d14800 R14: 0000000000000002 R15: 0000000000000000
Aug 10 13:24:50 xxx1 kernel: [ 14.150761] FS: 00007fe20489c740(0000) GS:ffff88025f280000(0000) knlGS:0000000000000000
Aug 10 13:24:50 xxx1 kernel: [ 14.150762] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Aug 10 13:24:50 xxx1 kernel: [ 14.150762] CR2: 0000000000000000 CR3: 0000000095f33000 CR4: 00000000001407e0
Aug 10 13:24:50 xxx1 kernel: [ 14.150763] Stack:
Aug 10 13:24:50 xxx1 kernel: [ 14.150764] ffffffff81224ffd ffffffffa0b49438 0000000000000000 ffff880095fb97f8
Aug 10 13:24:50 xxx1 kernel: [ 14.150765] 0000000000000000 ffffffffa0b49430 ffff880090ae8000 ffff880036b1a300
Aug 10 13:24:50 xxx1 kernel: [ 14.150767] 0000000000000002 00000000000000ff ffff880093fedba8 ffffffff8172a0a0
Aug 10 13:24:50 xxx1 kernel: [ 14.150767] Call Trace:
Aug 10 13:24:50 xxx1 kernel: [ 14.150771] [<ffffffff81224ffd>] ? proc_alloc_inode+0x1d/0xb0
Aug 10 13:24:50 xxx1 kernel: [ 14.150773] [<ffffffff8172a0a0>] __down+0x1d/0x1f
Aug 10 13:24:50...

[Teo (teo1978)]

Are you sure it's a duplicate?

This one happened randomly, the other one is described as 100% systematic.
This one results in the "Ubuntu..." screen being displayed forever, the other one results in a black screen.

In my case the issue disappeared after I uninstalled/purged the nvidia drivers and (probably weeks later) reinstalled them.