Common Issues: Difference between revisions

From miki
Jump to navigation Jump to search
Line 1,544: Line 1,544:
:* The following [https://www.virtualbox.org/ticket/20180 virtualbox issue #20180] contained detailed information, also related to '''split lock detection''' feature.
:* The following [https://www.virtualbox.org/ticket/20180 virtualbox issue #20180] contained detailed information, also related to '''split lock detection''' feature.
:* This [https://lwn.net/ml/linux-kernel/1556134382-58814-1-git-send-email-fenghua.yu%40intel.com/ lwn post] describes the split lock detection feature with many details, and how the hypervisor should manage that.
:* This [https://lwn.net/ml/linux-kernel/1556134382-58814-1-git-send-email-fenghua.yu%40intel.com/ lwn post] describes the split lock detection feature with many details, and how the hypervisor should manage that.
:* We see the same messages in {{file|~/vmware/.../vmware.log}}.
:* https://communities.vmware.com/t5/VMware-Workstation-Player/x86-Split-Lock-Detection-AC-Alignment-Check/td-p/2894457


* '''{{red|Issue}}''' VMWare crashed (Win10 guest). Log has no information on the crash.
* '''{{red|Issue}}''' VMWare crashed (Win10 guest). Log has no information on the crash.

Revision as of 20:42, 26 April 2022

Issues on ALL (ALL)

  • Issue — Cannot paste text in Youtube comment
  • Fix Go to about:config, and set dom.event.clipboardevents.enabled to false, [1]
  • Issue — Proxy does not work in Thunderbird 38 with IMAP when connecting to localhost (see Thunderbird).
  • Workaround is to disable the proxy. Proxy works for add-on though.
  • Issue — (reported as bug 1625739) Midnight Commander (mc) starts slowly when current directory is on a samba share. Also when pressing C-O.
We debug the root cause with strace:
strace -r -tt -o /tmp/mc.strace mc
grep -v 0.00 /tmp/mc.strace
# 0.099773 read(5, "\33]0;peetersm@zavcxl0005: ~\7\r\r\n\33["..., 128) = 98
# 0.048970 lstat("/smb/lacie-cloudbox/family/backup/nintendo_3ds/nintendo/cia/SaveDataFiler.cia.txt", {st_mode=S_IFREG|0755, st_size=1701186, ...}) = 0
# 3.310479 stat("/smb/lacie-cloudbox/family", {st_mode=S_IFDIR|0755, st_size=0, ...}) = 0
# 0.010677 getuid()                  = 1000
stat -f /smb/lacie-cloudbox/family
# ... slow ...
So it is the stat call that takes several seconds.
  • Issue — Dropbox will stop syncing
  • Fix Make sure Dropbox is on ext4 with user_xattr flag (see Dropbox).
  • Issue — Some monospace fonts not showing in Firefox (typically GitHub) / Thunderbird.
Fix Using Firefly, extract the CSS style of code element. Narrow down to one font style, consolas. Running fc-match shows that found is font though:
fc-match consolas
# consolas.ttf: "Consolas" "Regular"
But we have multiple occurences:
locate -i consolas.ttf
# /home/data/work/_git/homelight-new/.fonts/ms/consolas.ttf
# /home/peetersm/.fonts/ms/consolas.ttf
# /usr/local/share/fonts/ms/consolas.ttf

Turns out the permissions were incorrect: 

ll /home/peetersm/.fonts/ms/consolas.ttf /usr/local/share/fonts/ms/consolas.ttf
# -rw-rw---- 1 peetersm peetersm 358256 Jun 10  2009 /home/peetersm/.fonts/ms/consolas.ttf
# -rw-r----- 1 root     staff    358256 May  9 18:31 /usr/local/share/fonts/ms/consolas.ttf

We fix that: 

sudo find /usr/local/share/fonts -type f -print0 | sudo xargs -0 chmod 644
find ~/.fonts -type f -print0 | xargs -0 chmod 644

Issues on Bionic (ALL)

  • Issue — Owncloud client complains with error Reading from keychain failed with error: 'No keychain service available', and request password at each login.
Fixed: Install package libgnome-keyring0 (see bug 1761936).
  • Issue — Console tty not showing up with keyboard sequence Ctrl+Alt+Fx.
Fixed: Make sure to type Ctrl and Alt separately. First press Ctrl, then Alt, then a function key Fx.
  • Issue — System monitor gnome extensions conflicts with Ubuntu AppIndicator.
  • Issue — Login password from GDM is shown in plain text on the VT1 console bug 1767918. This occured to me when being at the login screen, and switching back and forth with Ctrl-Alt-F1 and Ctrl-Alt-F2.

Issues on Trusty (ALL)

  • Issue — Garbage tilde characters when copy-pasting in Terminator. This happens due to bracked paste mode, triggered by Midnight Commander. See my bug report bug 1350334.
    • Work-around — Run printf "\e[?2004l" in the terminal [2]
    • Fixed — See patch in referenced bug report. Waiting official update. Build locally, and update packages (need reboot, or force reload vte lib, dunno how to do)
sudo dpkg -i libvte-common_0.28.2-5ubuntu1foo2_all.deb libvte9_0.28.2-5ubuntu1foo2_amd64.deb python-vte_0.28.2-5ubuntu1foo2_amd64.deb
  • Issue — GTK warning when launching i386 applications (like acroread):
Gtk-Message: Failed to load module "overlay-scrollbar"
Gtk-Message: Failed to load module "unity-gtk-module"
(acroread:5353): Gtk-WARNING **: Unable to locate theme engine in module_path: "murrine",
(acroread:5353): Gtk-WARNING **: Unable to locate theme engine in module_path: "pixmap",
Gtk-Message: Failed to load module "canberra-gtk-module"
  • Fixed — Install the following GTK libraries for i386 arch:
sudo apt-get install libcanberra-gtk-module:i386 unity-gtk2-module:i386 gtk2-engines-murrine:i386 gtk2-engines-pixbuf:i386
apt-get download overlay-scrollbar-gtk2:i386
sudo dpkg --ignore-depends overlay-scrollbar -i overlay-scrollbar-gtk*.deb
sudo vi /var/lib/dpkg/status
# Search 'Package: overlay-scrollbar-gtk2', then for the i386 package, delete 'overlay-scrollbar, ' on line 'Depends:'.
sudo apt-get install -f
  • Issue — Fonts like Calibri not rendered correctly at small size. In particular normal letter appears aliased, but ligatures are anti-aliased.
  • Fixed — The solution is to disable embedded bitmaps. See Fonts.
  • Issue — Meta-key (Alt) not working and Midnight Commander displays Press any key... after each command when in tmux session:
  • Fixed — Add the following to ~/.bash_profile and ~/.profile [3]:
# This fix meta-key support and stop mc displaying "Press any key..."
if [ $TERM = "screen" ]; then
    export TERM=xterm
fi
if [ -n "$TMUX" ]; then
    export COLORTERM=rxvt
fi

Issues on Precise (nxl67170ux)

  • Issue — Git — Cannot install tileqt (tcl theme) to beautify gitk.
  • Issue — TeX Live still old version (2009) — This may be the cause of errors LaTeX Font Warning: Some font shapes were not available, defaults substituted. (see [4])
    • fixed — install TeX Live 2012 from backports repository.
  • Issue — Annoying file C:\nppdf32Log\debuglog.txt. Caused by acroread plugin (see [5], [6], [7]). A solution would be to edit nppdf.so with hex editor.
  • Issue — Acrobat reader leaves ghost process SynchronizerApp-binary. Disable it!
  • Issue — Opera fonts too small, not as good as windows.
  • Issue — Touchpad scroll not working, detected as ps/2 mouse.
    • partial fix — Install Kevin Cernekee's DKMS package (see below and bug 606238). See also [8] for more solutions.
sudo -s
add-apt-repository ppa:cernekee/ppa
apt-get update
apt-get install psmouse-dkms
reboot
  • Issue — Annoying popup failure to download extra data files flashplugin-installer. See bug 977178 and bug 982684.
    Pop-up is from update-notified. Might be related to sudo ignoring some environment variables / profile.
    • tentative fix — sudo apt-get --reinstall install flashplugin-installer
  • Issue — Synergy, still heavy keyboard corruption issue (v.1.4.10), in particular ctrl stuck key (must reboot to fix), and custom layout not honored (my custom belgian on us layout).
  • Issue — Virtualbox module not recompiling (broken DKMS)
  • Issue — Can't reconnect to Bluetooth Microsoft Sculpt Mouse.
    • Solved — Install Bluetooth manager (package blueman). See [9], bug 1094744.
  • Issue — animated gifs do not work.
    • Solved — This is a feature of privoxy, to ban ads using such animation. It can be disabled (via setting deanimate_gifs), or better yet, temporarily disable privoxy, and clear cache/reload the page.
  • Issue — Acrobat reader, errant tooltips
    • Solved — Click on the document area. Tooltips appears when the focus is on the document tab.
    • Other solution, see Banishing errant tooltips (using Xlib and Python)

Issues on Trusty (nxl67170ux)

  • Issue — Controlling laptop brightness via brightness keys is very slow (small delta, long latency). See bug 847001 (and possibly related kernel bug. See also X page for work-arounds.
    • Solved — Enable vendor acpi in grub and use the 20-intel.conf script as explained on X page.
  • Issue — Fonts in Opera look really ugly. Note that Opera did not upgrade (still v12.16). Other applications do not seem affected.
  • Issue — xdotool can no longer move (unmaximized) windows to top left corner (see bug 1312412)
  • Issue — lots of failure due to pcscd in /var/log/syslog (see below). apt-get remove pcscd as temporary fix.
Apr 29 14:28:22 nxl67170ux pcscd: ifdwrapper.c:348:IFDStatusICC() Card not transacted: 612
Apr 29 14:28:22 nxl67170ux pcscd: eventhandler.c:313:EHStatusHandlerThread() Error communicating to: O2 Micro Oz776 00 00
Apr 29 14:28:22 nxl67170ux kernel: [  754.501492] usb 1-1.3.2: usbfs: process 1883 (pcscd) did not claim interface 0 before use
Apr 29 14:28:22 nxl67170ux kernel: [  754.501546] usb 1-1.3.2: usbfs: process 1883 (pcscd) did not claim interface 0 before use
Apr 29 14:28:22 nxl67170ux pcscd: pcscdaemon.c:684:at_exit() write() failed: Connection refused
  • Issue — XBMC Frodo 12.3 cannot read DVD (from disc media or from hdd). Filed as bug 1315873.
    • Work around — Install XBMC Gotham 13.0 from xbmc ppa (see XBMC).
  • Issue — GPU hangs stuck on render ring + dmesg error - see bug 76956 and bug 77104, launchpad bug #1274779
  • Issue — keyboard shortcut for Indicator-synapse does not work.
  • Issue — auto-completion works badly with command starting with sudo (lots of false positive).
    • Partial fix — This is due to doing complete -cF sudo in bash profile, which was necessary since u12.04 or so. However in u14.04 is no longer necessary. Removing it fixes the problem.
    • Doing partial fix above however the auto-completion is mostly inactive when using sudo.
  • Issue — Unity hot corners not working anymore after reboot. See bugs +1305438, +1128041 and +858845.
    • Partial fix — Open up Unity Tweak app, then off/on the hot corners.
Wireless
  • Wireless deauthentication code from [12]:
  • About syslog wlan0: deauthenticated from 00:1d:7e:0c:24:5e (Reason: 14), see [13]:
    That means that the WPA encryption code is failing and the packet is corrupt. I would worry about that corrupt beacon from the AP.
  • Troubleshooting WiFi (opensuse post)
  • Getting info on WiFi (opensuse post)
3 Deauthenticated because sending station is leaving (or has left) IBSS or ESS.
14 MIC failure.
  • Issue — Unstable Wi-Fi connection. Frequent disconnect. See syslog below.
  • Fix? (Done, no obvious effect, maybe less spam in syslog) Edit wireless connection, click the down arrow next to BSSID which is blank, select the mac address that was hidden before. [14]
  • Fix? Done, no obvious effect Disable IPv6? Edit wireless connection, click IPv6 settings, Click on drop-down list for method, and select Ignore.
  • Fix? Several fixes from [15]
  • (Done, no obvious effect, but less spam in syslog) Set regulatory agency for wireless
iw reg get            # Get agency. '00' means unset
sudo iw reg set US    # Or pick country code (eg. 'BE' for Belgium)
  • (Reverted. See if manual IP is sufficient)Disable 802.11n protocol. Edit /etc/modprobe.d/iwlwifi.conf
options iwlwifi 11n_disable=1
  • (Not tested yet) Disable ipv6 protocol. Edit /etc/sysctl.conf:
net.ipv6.conf.all.disable_ipv6 = 1
net.ipv6.conf.default.disable_ipv6 = 1 
net.ipv6.conf.lo.disable_ipv6 = 1
  • (Done, more stable, but reverted because link downgraded to 1Mb/s) Make sure the router is not set not to use QOS/WWN
  • Fix? (Done, more stable?) Deal with message wlan0: deauthenticating from 00:1d:7e:0c:24:5e by local choice (reason=3) [16], bug 548992.
  • Message might be due to DHCP (timeout...). See [17], [18]. As a fix, use manual IP in wireless config (don't use DHCP). See dhclient.leases to know which value to use for gateway, domain...
  • Kill wpa_supplicant [19] with sudo killall wpa_supplicant. NetworkManager will respawn it automatically.

Issues on Trusty (griffin)

  • Issue — v-sync tearing during DVD playback in XBMC 13.0 Gotham (see XBMC)
    • Add to /etc/profile: 
    # Fix vsync tearing - http://askubuntu.com/questions/456355/have-tearing-no-vsync-in-movies-nvidia-proprietary-driver
export __GL_YIELD="USLEEP"
export __GL_THREADED_OPTIMIZATIONS=1
  • Issue — Evolution continuously reporting error TCP connetion reset by peer or similar
    • Possible fix — Change gandi account to use STARTTLS on port 587 for sending mail (see [20], [21])

Issues on Precise (griffin)

  • Issue — Can't map Super-Space for Gnome-Do
    • solved? — open ccsm → Ubuntu Plugin → temporarily change the shortcut for the launcher. Change gnome-do shortcut again, and then restore the launcher shortcut to super.
    • solved — Use Synapse instead with Alt-Space as mapping.
  • Issue — Current openjdk-6-jdk does not print file full path in error message, but only basename, which breaks vim "goto bug" feature.
    • partially solved — This is solved in openjdk-7-sdk, but can't install 32_bit version though. So for 32-bit, we need to refer to openjdk-6 libraries.
  • Issue — Super-Space does not work. Also cannot bind to Alt-Space either (get error (Do:6011): libdo-WARNING **: Binding '<Alt>space' failed!). See bug 988699, [22] (Bugs: Do)
    • temporary fix — killing and restarting gnome-do in a shell (gnome-do&) to re-enable Super-Space
    • solved — Use Synapse instead with Alt-Space as mapping.
  • Issue — Compiz Ring Switch requires pressing Enter to select a window.
    • solved — Change Key to show the launcher to Shift-Super in Ubuntu Unity Plugin. Also, changed Key to show the HUD to Ctrl-Super
  • Issue — Compiz Scale plugin Initiate Window Picker for All Windows does not work (show windows for current workspace only)
    • See bug 933776
    • A patch can be found in this ppa (sudo add-apt-repository ppa:mc3man/fixing-scale)

Issues on Maverick (griffin)

  • Issue — No Configuration Editor link in the menu. Must be added by hand?
    • Solved — Enable it by editing Gnome menu, right click on menu → Edit Menu → System Tools → Enable icon for Configuration Editor.
  • Issue — Known issues documented in Maverick release manifest & Release Notes:
    • Connected DisplayPort monitors may prevent booting. (655795)
    • 802.11n support for the iwlagn driver has been temporarily disabled. Intel is actively working to get this properly fixed up in the firmware. This workaround should be reverted once updated firmware is available. (630748, 654152). See also Ubuntu Forums thread, Bug 16691 kernel.org, Intel WiFi driver for linux and official Linux Wireless wiki.
      • cfr file /etc/modprobe.d/intel-5300-iwlagn-disable11n.conf 
      dmesg | grep -i iwl
sudo modprobe -r iwlagn
sudo modprobe iwlagn 11n_disable=1
iwconfig                                           # Check that 11n is disabled
# ...
# wlan0     IEEE 802.11abg  ESSID:off/any
# 
# ...
    • The logo will not display during boot when using the Nvidia proprietary driver. (653274)
  • Issue — [arrandale] Integrated intel graphics GMA HD doesn't work (590504). See also [23]
    • This relates to kernel message at boot: 
    dmesg | grep i915
# ... intel ips 0000:00:1f.6: failed to get i915 symbols, graphics turbo disabled
  • Issue — Suspend does not work; pc goes suspend but then restart immediately it seems in login screen. Related bug: 570604.
    • Solved — Problem is linked with xhci_hcd that prevents suspend when loaded (this is because sleep/suspend is not implemented yet in that module). Unloading it will allow suspend (workaround). Bug is fixed in Ubuntu 11.04. Launchpad bug 522998
    echo 'SUSPEND_MODULES="xhci_hcd"' >> /etc/pm/config.d/unload_modules   # As root
  • Issue — Window Manager does not display window decorators after Macbuntu install. Temporary fix:
    • compiz --replace&
  • To Do — Setup BEH (backend error handler) for CUPS

Hardware issues (nxl67002ux)

  • Issue — Heavy HW interrupts related to intel graphic card, power saving (sleep), and use of DVI/HDMI interface (module i915, drm, drm_kms_helper).
    • 2.6.32 kernel — udevd eats CPU resource! Launchpad bug 440411, talked on forum 1356215. There was a partial workaround, but it does not work anymore on 2.6.32.28:
    sudo service udevd stop
sudo service udevd start     # After this unfortunately there is no battery state icon anymore (unless reboot, or maybe suspend)
udevadm monitor              # Check that there is no events triggered
    • 2.6.35 kernel (Maverick) — Install a newer version of the kernel, like 2.6.35. But then we have issues with the processes [kslowd001] ... [kslowd004] that hog the cpu even more, and make mouse movements jerky. Launchpad bug 662946, 661012 (dup). See fix in forum 1594239 (Install latest kernel 2.6.37 with partial fix with module drm_kms_helper / i915..., or install kernel 2.6.34). Other links [24]. Workaround:
    echo N > /sys/module/drm_kms_helper/parameters/poll
# OR...
echo "options drm_kms_helper poll=N">/etc/modprobe.d/local.conf   # As root.
    • Some issues seem to be fixed in the latest kernel 2.6.33 for Lucid. See [25] (drm/i915: Disable FBC on 915GM and 945GM...)
    • Workarounds to investigate
      • Install kernel 2.6.34 as suggested at [26]
      • Install latest kernel 2.6.33 for Lucid from Ubuntu Kernel ([27])
  • Issue — Frequent wireless disconnect.
    • Reference HW info on this machine
    • My other bugs in launchpad.
    • Related to intel 5300 chipset (module iwlagn) and 802.11n networks.
    • Done — Installed a linux-backports-modules-wireless package (linux-backports-modules-wireless-2.6.37-lucid-generic-pae).
      • Brought some improvements, in particular with kernel 2.6.32-32. Not disconnection, max ssh connect 1.2s, but still some drop to 1Mb/s. Still old firmware version 8.24.2.12 (file iwlwifi-5000-2.ucode)
    • Done — Installed a compat-wireless package from compat-wireless project (linux-backports-modules-wireless-lucid-generic-pae).
      • Not sure it does anything better. Still stuck with the old firmware version 8.24.2.12 (file iwlwifi-5000-2.ucode)
    • SOLVED — Force upgrade the iwlagn driver as indicated on this forum (see also bug 794291, and post on tech drive-in where I first learned about this solution):
    #Remove any linux backports modules
sudo apt-get autoremove linux-backports-modules-wireless*

#Download and install the compat-wireless iwlagn driver
wget http://www.orbit-lab.org/kernel/compat-wireless-2.6-stable/v2.6.38/compat-wireless-2.6.38.2-2-ns.tar.bz2
sudo apt-get install build-essential linux-headers-$(uname -r)
sudo apt-get remove lsh-server
tar jxvf compat-wireless-2.6.38.2-2-ns.tar.bz2 
cd compat-wireless-2.6.38.2-2-ns/
make
sudo make install

#Now reboot... or you can try reload the module:

sudo make unload
sudo modprobe iwlagn
sudo modprobe bnep
sudo modprobe btusb
sudo modprobe rfcomm
sudo modprobe sco
sudo /etc/init.d/networking restart

#Check driver:
lshw -C network
# ...
# configuration: broadcast=yes driver=iwlagn driverversion=2.6.32-32-generic-pae firmware=8.83.5.1 build 33692 ip=134.27.182.162 latency=0 multicast=yes wireless=IEEE 802.11abgn
modinfo iwlagn
# filename:       /lib/modules/2.6.32-32-generic-pae/updates/drivers/net/wireless/iwlwifi/iwlagn.ko
# ...
# firmware:       iwlwifi-5000-5.ucode

    Running my ssh tests over WiFi show a very steady connection, without drop of connection speed (however it's capped at 54 Mb/s)

Hardware issues (griffin)

  • Issue — Smartcard reader (Ricoh Co Ltd RL5c476 II) does not work. No available drivers [28].
sudo lspci | grep CardBus
# 46:06.4 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev bb)

VirtualBox issues (nxl67002ux)

  • Issue — VirtualBox - UE slower gui refresh since update to VirtualBox 3.2.4 (previous version was 3.2.0)
  • Issue — VirtualBox - ksoftirqd/0 taking 10-20% CPU. Possible work-around is to add nohz=off to kernel command line, i.e. in /boot/grub/menu.list) (see [29] and [30])
  • Issue — VirtualBox - BT dongle not reconnecting after suspend. It used to work correctly in the past. Is it due to an update?
  • Issue — VirtualBox - Cygwin very slow in VirtualBox (while (true); do date --utc; done | uniq -c gives me ~4 fork / second. As an alternative, try MSYS (MGWin), UWIN or Interix. Some other causes: [31]
    • SOLVED? — The speed issue seems to be solved in the last version (1.7.9). Fork is still slow though, but rxvt is blazing fast. rxvt still much slower.
  • Issue — VirtualBox - Can't go to synergy client when VB guest is in fullscreen. See X#Synergy.
    • SOLVED? — Press host key (ctrl) before/after going to synergy client.

Other issues (nxl67002ux)

  • Issue — ssh disconnects after some time, autossh can't reconnect. Even doing a simple ssh does not work anymore. Might be due to broken ssh-agent
    • Disable auto-run of Gnome ssh-agent, enabled back my ssh-agent script
  • Issue — synergy unstable. Sometimes CTRL key does not work anymore in the client...
    • Look for alternatives: Synergy+ (SynergyPlus), or x2x (see Synergy page on wikipedia).
  • Issue — Midnight Commander does not colorize different file types. Check here.
    • SOLVED — Use mc official repository instead. See MC page
  • Issues — Serious interferences from Privoxy:
    • Reject URL like .../advies/... and .../ADV_AUTH.CAB (work-around: disable Privoxy AdBlocking via http://p.p).
    • Interference with Altran Timesheet. ActiveX / ASP.NET controls do not work in (Wine) Internet Explorer 6.
  • Issue — Opera 11 - v11 does not support my autoproxy .pac file anymore.
  • Issue — Evince, error in console EggSMClient-WARNING **: Failed to connect to the session manager.... It seems related to apparmor.
    • SOLVED — To solve it [32]:
    cd /etc/apparmor.d/disable
sudo ln -s /etc/apparmor.d/usr.bin.evince usr.bin.evince
sudo /etc/init.d/apparmor restart
  • Issue — ufw user[6].rules files should be stored in /etc/ufw instead of /lib/ufw so that they can be tracked by etckeeper (filed a bug in launchpad, 728128)
  • Issue — Kernel panic when updating to kernel 2.6.32-29, due to trailing dots in /boot/initrd file (bug submitted on launchpad 728430)
    • SOLVED — No problem with 2.6.32-30, but again same problem with 2.6.32-32. Problem gone after formatting /boot in Ext3.
  • 'Issue — Opera 11 → shaky (at best) clipboard support! Copy & paste to/from utility xsel, or to/from virtualbox guest/host is broken. The best is to pass through an intermediate editor, like gedit.

Issues on Precise (nxl67063ux)

  • Issue — When connecting to Host's Cygwin via SSH, there network drive shares are not visible.

opengl / hdmi / optimus Issues (Reddragon)

Issues on Precise/Trusty (Reddragon)

  • Issue — External headphone (jack) does not work
    • (see bug 978880)
    • Broken solution — See [37]. Edit /usr/share/pulseaudio/alsa-mixer/paths/analog-output-headphones.conf (and similar ...-2.conf files if any) (more links [38], [39], [40]). Change the lines:
     [Element Speaker]
-switch = off
+switch = on
-volume = off
+volume = ignore
    Update — this does not work. Settings are reverted after pulseaudio update. Moreover the bass speaker is not correctly handled.
    • Solved The solution is to install the latest Realtek codec from here, v5.18 (more information [41], [42], [43], [44]). To install, just un-tar and run given ./install script:
    tar -jxvzf ...
cd ...
vi install                   # install script has a bug, fetch exec at wrong place after untar
sudo ./install
  • To Do — Setup BEH (backend error handler) for CUPS
  • Issue — Cannot read movies from DVD-RW
    • Solved — Require patch on udisks (see Udisks).
    • Ubuntu 14.04 LTS — Require patch on udisks2 (see Udisks).
  • Issue — Sometimes keyboard in US layout
    • Temp fix: Add bin/setkbmap.sh and run it at startup:
    #! /bin/bash setxkbmap be
  • Issue — Flash instability
    • Temp fix: Uninstall all nvidia drivers
    • Temp fix: Uninstall flashplugin-installer
    • Temp fix: Install dansguardian & privoxy to prevent access to underground/unstable sites
      • Setup local network proxy: localhost:8080
      • Idem in firefox
      • ! for thunderbird, select no proxy (if proxy configured in system parameters → network... incredible but true [45])
    Other issues after upgrade to Trusty Tahr
    • System still thinks the DVD is present after pressing the eject button. One must select eject DVD in the launcher.
    • DVD-RW still un-readable (but now package udisks2)

Issues on Trusty/Xenial (Zyratastic)

  • Issue — Fn-F2 (WLAN / Bluetooth switch) and Fn-F5/Fn-F6 (screen brightness) do not work.
    • (NOK) Add acpi_osi='!Windows 2012' to /etc/default/grub (see [46], and also [47], [48], [49]) — Fn-F5/F6 still do not work, but Fn-F2 does work now!!!
    • (NOK) Add acpi= to /etc/default/grub (as per [50]) — Does not work either (even got black screen → Ctrl-Alt-BS to get out)
    • (Solved) Add acpi_osi= to /etc/default/grub (see guide N550JV)
  • Issue — NVidia GT750M card not detected.
    • sudo apt-add-repository ppa:xorg-edgers/ppa sudo apt-get update sudo apt-get install nvidia-331 nvidia-settings-331 nvidia-prime
    • Reboot. Multiple issues from apport (→ disabled now)
    • Nvidia card works with accelerated OpenGL, but keyboard is stuck to US layout, and Bluetooth is lost. Fonts are not rendered as usual it seems.
    • Purge xorg-edgers to see if we can revert back to previous situation:
    sudo ppa-purge xorg-edgers sudo apt-get autoremove nvidia-prime
    • (Solved) Install bumblebee and nvidia drivers v319 (see guide N550JV)
  • Issue — Black screen after package update
    • Cause is either update to a new kernel, and/or updating nvidia-331 to nvidia-340.
    • (Solved) See Osgyan's guide to fix the problem. This requires updating to nvidia-349 drivers and whitelisting/blacklisting a few modules.
  • Issue — Minecraft — Graphics glitches, missing limbs on intel card.
    • (NOK) Changing settings, to no avail: Graphics: Fast / Fancy, View Bobbing: ON / OFF: Advanced OpenGL: OFF/Fast/Fancy
    • (NOK) Tried these VM, and they have glitches: OpenJDK 7 64, Oracle JDK 7 64.
    • (NOK) From [52].
      • Run MC with -Xmx512M
      • Run in fullscreen
      • OpenGL / Multicore settings: Smooth FPS: On, Load Far: Off, Chunk UpF: 2, Lazy CL: On, Smooth World: On, Preload Chunks: Off, Dynamic Updates: On
      • Setting priority of proSeems to workcess java to high (aka sudo -k nice -n -5 su -c "..." -USER, see [53])
    • (NOK) Disable compiz before launching minecraft (see [54])
    sudo apt-get install metacity metacity --replace && ./launch-minecraft-now && unity --replace &
    • Check more solution from google query "optifine missing limbs" — ...
    • (Alt. solution) Disable OptiFine
    • (Solved) — use NVIDIA OpenGL (via bumblebee)
  • Issue — Image really dark
    • Use xrandr (see [55]):
    xrandr --output eDP1 --gamma 1.4:1.4:1.4
xrandr --verbose
    • Q? — How to make that settings permanent (setting is lost after quit / restart)?
      • I could make a custom script per seat, that is run right before launching minecraft.
  • Issue — Minecraft — FPS stuck at 60 FPS.
    • Solved — This is due to vsync. Disable it with env. var vblank_mode ([56]):
    vblank_mode=0 java ... # Only for one app echo "0/SyncToVBlank=0" >> ~/.nvidia-settings-rc # ... or for all apps
    • (or uncheck the nvidia-settings option at X Screen 0 / OpenGL Settings / Sync to VBlank)
  • Issue — Bluetooth — Bluetooth sometimes not active, even with nvidia disabled
    • [ 13.663135] Bluetooth: Error in firmware loading err = -110,len = 448, size = 4096 [ 13.663173] Bluetooth: Loading patch file failed [ 13.663198] ath3k: probe of 3-5:1.0 failed with error -110 sudo modprobe -r aht3k; sleep 1; sudo modprobe ath3k # work-around - NOT WORKING
    • Might be due to interaction with libmtp [57]. But uninstalling libmtp-common as no effect.
  • Issue — Keyboard layout — BE layout not applied on login
    • Adding .xinitrc to setxkbmap be (to fix keepassx issue) — this might fix the problem, or more precisely hide the problem.
  • Issue — Boot — Boot stuck right after grub. empty screen. no disk activity.
    • We can trigger disk activity 2x by pressing return key.
  • Done — Check BIOS version and upgrade if necessary (see [58])
    • We have version 204. Updated to version 207
  • To Do — Many errors in dmesg.
    • [ 12.493360] ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver [ 12.493374] lpc_ich: Resource conflict(s) found affecting gpio_ich ...
  • To Do — Minecraft — Do we get more FPS if we disable compiz (or any compositor, like metacity)?
  • To Do — Check install guides above.
  • To Do — Try latest nvidia drivers from xorg-edgers.
  • To Do — Backup
    • Backup Windows Rescue / System Image DVD (to mnemosyne)
    • Backup Restore / Recovery partition (to mnemosyne)
    • Backup current windows partition (locally)
    • Backup partition table, EFI partition, etc.
  • To Do — Install better temperature management from Intel, see [59]
  • Issue — The battery led is blinking orange. This usually indicates an issue with the Power Management unit in the laptop. See this post [60]. One solution apparently is to press the power button for at least 5-7 seconds until the PM unit is reset (led stops blinking). Another solution is to remove the battery, wait 10-20 second, and reinsert the battery; but this requires opening the laptop case.
  • Issue — XBMC touch interface does not work in v13 Gotham RC1. Filed as bug 15184.
  • Issue — Fan always on.
    • Fix &mdahs; Install fancontrol and lm-sensors. Must run sensors-detect (see SensorInstallHowto). Reboot the PC.
    • To try: fwts (from FirmwareTestSuite), and install module sudo modprobe asus-laptop [61]
    • Other references: [62]

Issues on Trusty (Graphicdaemon)

  • Issue — Fn-F2 (WLAN / Bluetooth switch) and Fn-F5/Fn-F6 (screen brightness) do not work.
    • (NOK) Add acpi_osi='!Windows 2012' to /etc/default/grub (see [63], and also [64], [65], [66]) — Fn-F5/F6 still do not work, but Fn-F2 does work now!!!
    • (NOK) Add acpi= to /etc/default/grub (as per [67]) — Does not work either (even got black screen → Ctrl-Alt-BS to get out)
    • (Solved) Add acpi_osi= to /etc/default/grub (see guide N550JV)
  • Issue — NVidia GT750M card not detected.
    • sudo apt-add-repository ppa:xorg-edgers/ppa sudo apt-get update sudo apt-get install nvidia-337 bumblebee wine
    • Reboot. Multiple issues from apport (→ disabled now)
    • Nvidia card works with accelerated OpenGL, but keyboard is stuck to US layout, and Bluetooth is lost. Fonts are not rendered as usual it seems.
    • Purge xorg-edgers to see if we can revert back to previous situation:
    sudo ppa-purge xorg-edgers sudo apt-get autoremove nvidia-prime
    • (Solved) Install bumblebee and nvidia drivers v319 (see guide N550JV)
  • Issue — Minecraft — Graphics glitches, missing limbs on intel card.
    • (NOK) Changing settings, to no avail: Graphics: Fast / Fancy, View Bobbing: ON / OFF: Advanced OpenGL: OFF/Fast/Fancy
    • (NOK) Tried these VM, and they have glitches: OpenJDK 7 64, Oracle JDK 7 64.
    • (NOK) From [69].
      • Run MC with -Xmx512M
      • Run in fullscreen
      • OpenGL / Multicore settings: Smooth FPS: On, Load Far: Off, Chunk UpF: 2, Lazy CL: On, Smooth World: On, Preload Chunks: Off, Dynamic Updates: On
      • Setting priority of proSeems to workcess java to high (aka sudo -k nice -n -5 su -c "..." -USER, see [70])
    • (NOK) Disable compiz before launching minecraft (see [71])
    sudo apt-get install metacity metacity --replace && ./launch-minecraft-now && unity --replace &
    • Check more solution from google query "optifine missing limbs" — ...
    • (Alt. solution) Disable OptiFine
    • (Solved) — use NVIDIA OpenGL (via bumblebee)
  • Issue — Image really dark
    • Use xrandr (see [72]):
    xrandr --output eDP1 --gamma 1.4:1.4:1.4
xrandr --verbose
    • Q? — How to make that settings permanent (setting is lost after quit / restart)?
      • I could make a custom script per seat, that is run right before launching minecraft.
  • Issue — Minecraft — FPS stuck at 60 FPS.
    • Solved — This is due to vsync. Disable it with env. var vblank_mode ([73]):
    vblank_mode=0 java ... # Only for one app echo "0/SyncToVBlank=0" >> ~/.nvidia-settings-rc # ... or for all apps
    • (or uncheck the nvidia-settings option at X Screen 0 / OpenGL Settings / Sync to VBlank)
  • Issue — Bluetooth — Bluetooth sometimes not active, even with nvidia disabled
    • [ 13.663135] Bluetooth: Error in firmware loading err = -110,len = 448, size = 4096 [ 13.663173] Bluetooth: Loading patch file failed [ 13.663198] ath3k: probe of 3-5:1.0 failed with error -110 sudo modprobe -r aht3k; sleep 1; sudo modprobe ath3k # work-around - NOT WORKING
  • Issue — Keyboard layout — BE layout not applied on login
    • Adding .xinitrc to setxkbmap be (to fix keepassx issue) — this might fix the problem, or more precisely hide the problem.
  • Issue — Boot — Boot stuck right after grub. empty screen. no disk activity.
    • We can trigger disk activity 2x by pressing return key.
  • Done — Check BIOS version and upgrade if necessary (see [74])
    • We have version 204. Updated to version 207
  • To Do — Many errors in dmesg.
    • [ 12.493360] ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver [ 12.493374] lpc_ich: Resource conflict(s) found affecting gpio_ich ...
  • To Do — Minecraft — Do we get more FPS if we disable compiz (or any compositor, like metacity)?
  • To Do — Check install guides above.
  • To Do — Try latest nvidia drivers from xorg-edgers.
    • First make sure that bumblebee and nvidia is uninstalled
    sudo apt-get purge "nvidia*" "bumblebee*" sudo apt-get install nvidia-346 nvidia-libopencl1-346
    • The following packages should be installed:
    nvidia-346 nvidia-346-uvm nvidia-libopencl1-346 nvidia-opencl-icd-346 nvidia-prime nvidia-settings
    • Probem, blank login screen. Some links: [75] [76]
    • To recover login screen, switching to intel works. From a console:
    sudo service lightdm stop sudo prime-select intel sudo service lightdm start
    • Now, let's see if we can use bumblebee again (along with nvidia prime):
    sudo apt-get install optirun-nvidia sudo shutdown -r now
    • Login screen visible. Check that it works:
    $ optirun /opt/VirtualGL/bin/glxspheres64 Polygons in scene: 62464 Visual ID of window: 0x20 Context is Direct OpenGL Renderer: GeForce GTX 850M/PCIe/SSE2 281.837052 frames/sec - 314.530150 Mpixels/sec
    • Why does bumblebee works, and not nvidia-prime? Could it be that the device id in /etc/X11/xorg.conf was incorrect? We have
    $ sudo lspci | grep -i nvidia
01:00.0 3D controller: NVIDIA Corporation GM107M [GeForce GTX 850M] (rev ff)
    Content of /etc/bumblebee/xorg.conf.nvidia:
    Section "ServerLayout"
    Identifier  "Layout0"
    Option      "AutoAddDevices" "false"
    Option      "AutoAddGPU" "false"
EndSection

Section "Device"
    Identifier  "DiscreteNvidia"
    Driver      "nvidia"
    VendorName  "NVIDIA Corporation"

#   If the X server does not automatically detect your VGA device,
#   you can manually set it here.
#   To get the BusID prop, run `lspci | egrep 'VGA|3D'` and input the data
#   as you see in the commented example.
#   This Setting may be needed in some platforms with more than one
#   nvidia card, which may confuse the proprietary driver (e.g.,
#   trying to take ownership of the wrong device). Also needed on Ubuntu 13.04.
    BusID "PCI:01:00:0"

#   Setting ProbeAllGpus to false prevents the new proprietary driver
#   instance spawned to try to control the integrated graphics card,
#   which is already being managed outside bumblebee.
#   This option doesn't hurt and it is required on platforms running
#   more than one nvidia graphics card with the proprietary driver.
#   (E.g. Macbook Pro pre-2010 with nVidia 9400M + 9600M GT).
#   If this option is not set, the new Xorg may blacken the screen and
#   render it unusable (unless you have some way to run killall Xorg).
    Option "ProbeAllGpus" "false"

    Option "NoLogo" "true"
    Option "UseEDID" "false"
    Option "UseDisplayDevice" "none"
EndSection
    Content of /etc/X11/xorg.conf:
    Section "ServerLayout"
    Identifier "layout"
    Screen 0 "nvidia"
    Inactive "intel"
EndSection

Section "Device"
    Identifier "intel"
    Driver "intel"
    BusID "PCI:0@0:2:0"
    Option "AccelMethod" "SNA"
EndSection

Section "Screen"
    Identifier "intel"
    Device "intel"
EndSection

Section "Device"
    Identifier "nvidia"
    Driver "nvidia"
    BusID "PCI:1@0:0:0"
    Option "ConstrainCursor" "off"
EndSection

Section "Screen"
    Identifier "nvidia"
    Device "nvidia"
    Option "AllowEmptyInitialConfiguration" "on"
    Option "IgnoreDisplayDevices" "CRT"
EndSection
  • Done — Backup, done on 20140701
    • Backup Windows Rescue / System Image DVD (to mnemosyne)
    • Backup Restore / Recovery partition (to mnemosyne)
    • Backup current windows partition (locally)
    • Backup partition table, EFI partition, etc.
  • To Do — Install better temperature management from Intel, see [77]
  • Issue — The battery led is blinking orange. This usually indicates an issue with the Power Management unit in the laptop. See this post [78]. One solution apparently is to press the power button for at least 5-7 seconds until the PM unit is reset (led stops blinking). Another solution is to remove the battery, wait 10-20 second, and reinsert the battery; but this requires opening the laptop case.
  • Issue — XBMC touch interface does not work in v13 Gotham RC1. Filed as bug 15184.
  • Issue — Minecraft FPS not going higher than 45fps for 1.7.2, and not higher than 55fps for 1.5.2 (with nvidia-331-updates). Is nvidia driver really running?
    • Todo: start minecraft with optirun

Issues on Trusty (ZAVCWL0005)

  • Issue — pcscd crash at boot — pcscd crashed with SIGSEGV in __efl_set__libc_thread_subfreeres_element___rpc_thread_destroy__(), package pcscd 1.8.10-1ubuntu1.
To restart pcscd daemon (in case of Card not found / Cannot connect to PC/SC daemon!):
sudo pcscd
  • Issue — regular kernel panic after several days (with standby/resume cycles, though not sure it is related).
  • No traces in /var/log. Enable netconsole. Collected this trace:
[36421.360837] ata1.00: exception Emask 0x0 SAct 0x7fffffff SErr 0x0 action 0x6 frozen
[36421.360853] ata1.00: failed command: WRITE FPDMA QUEUED
[36421.360858] ata1.00: cmd 61/08:00:f0:56:0d/00:00:00:00:00/40 tag 0 ncq 4096 out
[36421.360858]          res 40/00:ff:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
[36421.360861] ata1.00: status: { DRDY }
[36421.361022] ata1.00: failed command: WRITE FPDMA QUEUED
...
[36426.726787] ata1: link is slow to respond, please be patient (ready=0)
[36431.387945] ata1: COMRESET failed (errno=-16)
[36431.387956] ata1: hard resetting link
...
[36481.495600] ata1: reset failed, giving up
[36481.495604] ata1.00: disabled
[36481.495609] ata1.00: device reported invalid CHS sector 0
...
[36481.495688] ata1: EH complete
[36481.495711] sd 0:0:0:0: [sda] FAILED Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
[36481.495714] sd 0:0:0:0: [sda] CDB: 
[36481.495716] Write(10): 2a 00 00 0d 56 d8 00 00 08 00
[36481.495728] blk_update_request: I/O error, dev sda, sector 874200
[36481.495736] Buffer I/O error on dev dm-1, logical block 45787, lost async page write
[36481.495748] sd 0:0:0:0: [sda] FAILED Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
[36481.495750] sd 0:0:0:0: [sda] CDB: 
[36481.495752] Write(10): 2a 00 00 0d 56 d0 00 00 08[36481.497838] EXT4-fs warning (device dm-3): ext4_end_bio:317: I/O error -5 writing to inode 5505588 (offset 0 size 4096 starting block 22100526)
[36481.497849] Buffer I/O error on device dm-3, logical block 22100526
[36481.497864] Write-error on swap-device (252:2:6137680)
...
[36481.513686] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000007
[36481.513686] 
[36481.513751] CPU: 1 PID: 1 Comm: init Tainted: G        W  OE  3.19.0-37-generic #42~14.04.1-Ubuntu
[36481.513798] Hardware name: Hewlett-Packard HP EliteBook 840 G1/198F, BIOS L71 Ver. 01.31 03/24/2015
[36481.513845]  ffff880234a66c80 ffff880235223c88 ffffffff817af92b[36481.514034]  [<ffffffff817af92b>] dump_stack+0x45/0x57
[36481.514077]  [<ffffffff817a976b>] panic+0xc1/0x1f5
[36481.514114]  [<ffffffff81077b61>] do_exit+0xa11/0xb00
[36481.514144]  [<ffffffff81077cdf>] do_group_exit+0x3f/0xa0
[36481.514182]  [<ffffffff81083bc0>] get_signal+0x1e0/0x710
[36481.514220]  [<ffffffff81014e70>] do_signal+0x20/0x120
[36481.514255]  [<ffffffff811a0000>] ? generic_file_remap_pages+0x250/0x2c0
[36481.514295]  [<ffffffff81690652>] ? __sys_sendmsg+0x42/0x80
[36481.514329]  [<ffffffff81014fd9>] do_notify_resume+0x69/0xb0
[36481.514371]  [<ffffffff817b82a2>] retint_signal+0x48/0x86
[36481.514412] Kernel Offset: 0x0 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)
[36481.514446] drm_kms_helper: panic occurred, switching back to text console
  • Check for badblock
sudo badblocks -v /dev/sda          # Or anything else?
  • comment #32 - disable AHCI in BIOS
  • libata.force=noncq to grub line
cat /sys/block/sda/device/queue_depth
# if returns 31 means NCQ enabled, 1 means disabled
  • Kernel options like "libata.noacpi=1", "libata.force=noncq" or "sata_nv.adma=0"
  • "pci=noacpi" or "irqpoll"
  • Add 'ifconfig eth0 down' to /etc/rc.local.shutdown
  • or disable Wake-On-LAN in BIOS
Procedure to debug hard disk issues. No link to my problem there.
  • Not read yet:
Looking for "WRITE FPDMA QUEUED":
  • [79] and [80] - Report a possible ext4 bug --> WRONG. Very unlikely.
  • [81] — Relate to "5184 hours" bug on micron SSDs. Propose to do a firmware update.
  • [82] — Refers to a program that can reproduce the error
  • Issue — Very high CPU usage on screen dimming triggered by the screen saver, with high probability of crash.
Happens with kernel 3.19.0-58-generic.
Reverting to 3.19.0-56-generic: no difference.
Enabling vendor acpi in grub. Edit /etc/default/grub
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash acpi_backlight=vendor"
Then update grub and reboot:
sudo update-grub
  • Issue — X.org vivid backport leaks like hell when using some applications (for instance when Thunderbird is busy waiting, or with the Prisma Puzzle Timer).
workaround — Revert back to trusty x.org:
sudo apt-get install xserver-xorg

Issues on Xenial (ZAVCXL0005)

  • Issue — On-screen text disappears after suspend bug 1573959.
  • Fix Didn't have this issue anymore, so probably fixed in recent kernel as documented. Install kernel 4.8+ with sudo apt install linux-generic-hwe-16.04.
  • Issue — High CPU usage by kworker/ksoftirqd.
ps faux reports 100% usage by processes named kworker and ksoftirqd.
  • Workaround: problem was gone after the following sequence: undock laptop, plug in usb devices in all usb port, remove such devices, close the lid and wait for suspend, wake up from suspend, dock back the laptop while powered up. When docking the laptop, the problem went immediately away, so maybe is it enough to undock/dock the laptop when powered.
  • This may be correlated to ethernet failure and BIOS update (original BIOS version was 1.31. Version 1.37 and 1.39 at least exhibit the bug). A solution might be then to downgrade BIOS back to v1.31;
  • Other option is to debug the kernel. See [83], [84], [85].
  • To collect perf trace:
sudo apt-get install linux-tools-common
perf
sudo apt install linux-tools-generic linux-cloud-tools-generic
sudo perf record -g -a sleep 10
sudo perf report
I get the following traces:
  Children      Self  Command          Shared Object                          Symbol
+   79.76%     0.00%  kworker/0:1      [kernel.kallsyms]                      [k] kthread
+   79.76%     0.00%  kworker/0:1      [kernel.kallsyms]                      [k] ret_from_fork
+   79.75%     0.06%  kworker/0:1      [kernel.kallsyms]                      [k] worker_thread
+   79.67%     0.31%  kworker/0:1      [kernel.kallsyms]                      [k] process_one_work
+   78.28%     0.14%  kworker/0:1      [kernel.kallsyms]                      [k] rpm_idle
+   77.78%     0.06%  kworker/0:1      [kernel.kallsyms]                      [k] pm_runtime_work
+   77.72%     0.18%  kworker/0:1      [kernel.kallsyms]                      [k] rpm_suspend
+   77.48%     0.14%  kworker/0:1      [kernel.kallsyms]                      [k] __rpm_callback
+   77.39%     0.03%  kworker/0:1      [kernel.kallsyms]                      [k] usb_runtime_idle
+   77.35%     0.09%  kworker/0:1      [kernel.kallsyms]                      [k] __pm_runtime_suspend
+   76.93%     0.04%  kworker/0:1      [kernel.kallsyms]                      [k] rpm_callback
+   76.77%     0.06%  kworker/0:1      [kernel.kallsyms]                      [k] usb_runtime_suspend
+   76.68%     0.19%  kworker/0:1      [kernel.kallsyms]                      [k] usb_suspend_both
+   72.68%     0.25%  kworker/0:1      [kernel.kallsyms]                      [k] usb_control_msg
+   70.67%     0.27%  kworker/0:1      [kernel.kallsyms]                      [k] usb_start_wait_urb
+   65.29%     0.13%  kworker/0:1      [kernel.kallsyms]                      [k] usb_submit_urb
+   65.17%     0.35%  kworker/0:1      [kernel.kallsyms]                      [k] usb_submit_urb.part.6
+   64.42%     1.50%  kworker/0:1      [kernel.kallsyms]                      [k] usb_hcd_submit_urb
+   55.81%     0.16%  kworker/0:1      [kernel.kallsyms]                      [k] hub_suspend
+   54.43%    53.56%  kworker/0:1      [kernel.kallsyms]                      [k] xhci_hub_control
+   53.84%     0.03%  kworker/0:1      [kernel.kallsyms]                      [k] set_port_feature
+   20.44%     0.05%  kworker/0:1      [kernel.kallsyms]                      [k] usb_resume_interface.isra.6
+   20.38%     0.02%  kworker/0:1      [kernel.kallsyms]                      [k] hub_resume
+   20.34%     0.28%  kworker/0:1      [kernel.kallsyms]                      [k] hub_activate
  • Last resort is to update the kernel, from current v4.4. There is kernel 4.8 available by installing package linux-image-generic-hwe-16.04-edge (see RollingLTSEnablementStack).
  • Update 2017/01/10 — Install HWE 4.10 kernel made the problem worse (kworker kicking in anytime, even before sleep, and docking/undocking not permanent fix). So trying HWE Edge kernel 4.13.
  • Issue — High CPU usage by compiz
  • Some reports that it might be related to the GVFS metadata bug [86] (from [87]). Same report relates the bug to NFS / Samba shares.
  • The GVFS metadata bug may cause various other high CPU use, in particular with remote FS. The fix is:
rm -rf ~/.local/share/gvfs-metadata
# ... or more drastic, prevent gvfs metadata daemon to run:
sudo chmod -x /usr/lib/gvfs/gvfsd-metadata
  • work-around Problem went away either by using the wall plugin (unlikely) and/or closing VirtualBox (note however that problem was happening with VM being suspended).
  • Issue — WiFi not working after suspend (does not see SSID anymore)
  • We see that wifi is managed by driver iwlwifi:
lshw|grep -i wifi
#                configuration: broadcast=yes driver=iwlwifi driverversion=4.4.0-64-generic firmware=17.352738.0 
# latency=0 link=no multicast=yes wireless=IEEE 802.11abgn
  • Workaround: We unload / reload it:
# Reset the wifi driver
sudo rmmod iwlmvm iwlwifi
sudo modprobe iwlwifi iwlmvm
  • Issue — Permission denied error when running apt update
sudo apt update
#Err:32 http://security.ubuntu.com/ubuntu xenial-security/main DEP-11 64x64 Icons                              
#  Could not open file /var/lib/apt/lists/partial/security.ubuntu.com_ubuntu_dists_xenial-security_main_dep11_icons-64x64.tar.gz - open (13: Permission denied) [IP: 127.0.0.1 8118]
#...
#E: Failed to fetch http://security.ubuntu.com/ubuntu/dists/xenial-security/main/dep11/icons-64x64.tar  Could not open file /var/lib/apt/lists/partial/security.ubuntu.com_ubuntu_dists_xenial-security_main_dep11_icons-64x64.tar.gz - open (13: Permission denied) [IP: 127.0.0.1 8118]
Let's delete the apt cache and rebuild it. Maybe this is due to download corruption, so for safety I must disable proxy as well.
# Disable proxy
sudo rm /etc/apt/apt.conf
unset $(env|grep proxy|cut -f 1 -d=)

# Delete the apt cache and rebuild it
rm -rf /var/lib/apt/lists
mkdir -p /var/lib/apt/lists/partial
chmod 755 /var/lib/apt/lists
chmod 700 /var/lib/apt/lists/partial
chown _apt:root /var/lib/apt/lists/partial
touch /var/lib/apt/lists/lock
chmod 640 /var/lib/apt/lists/lock
Note that the chown above is necessary or we get the following error [88]:
W: Can't drop privileges for downloading as file '/var/lib/apt/lists/partial/be.archive.ubuntu.com_ubuntu_dists_xenial_InRelease' couldn't be accessed by user '_apt'. - pkgAcquire::Run (13: Permission denied)
Update — Got the problem again:
# Err:16 http://archive.canonical.com/ubuntu xenial/partner Sources
#   Could not open file /var/lib/apt/lists/partial/archive.canonical.com_ubuntu_dists_xenial_partner_source_Sources.xz - open (13: Permission denied) [IP: 127.0.0.1 8118]
# ...
# E: Failed to fetch http://archive.canonical.com/ubuntu/dists/xenial/partner/source/Sources  Could not open file /var/lib/apt/lists/partial/archive.canonical.com_ubuntu_dists_xenial_partner_source_Sources.xz - open (13: Permission denied) [IP: 127.0.0.1 8118]
# E: Some index files failed to download. They have been ignored, or old ones used instead.
As a fix, disable the partner repository from /etc/apt/soure.list, then enable it again:
sudo sed -ri '/partner/s/^/##/' /etc/apt/sources.list
sudo apt update
sudo sed -ri '/partner/s/^##//' /etc/apt/sources.list
  • Issue — dwww gives Internal Server Error errors when accessing /usr/share/doc pages:
  • Fix — Install dwww=1.13.3 package [89].
  • Issue — dwww gives Access denied errors when accessing cppreferences manual.
  • Fix Update /etc/dwww/dwww.conf [90]
-# DWWW_DOCPATH='/usr/share/doc:/usr/share/info:/usr/share/man:/usr/local/share/doc:/usr/local/share/info:/usr/local/share/man:/usr/share/common-licenses'
+DWWW_DOCPATH='/usr/share/doc:/usr/share/info:/usr/share/man:/usr/local/share/doc:/usr/local/share/info:/usr/local/share/man:/usr/share/common-licenses:/usr/share/cppreference
  • Issue — Interference between setting PROMPT_COMMAND and Midnight Commander.
  • Issue — Stuck mouse button issue in VirtualBox / Windows Guest.
  • Happened since switch to xserver-xorg-hwe-16.04 and linux-generic-hwe-16.04-edge. Might got even worse when using virtualbox-5.2.6, but now reverted back to virtualbox-5.1.30.
  • Reverting to the previous kernel and xorg did not help however. So the dependencies are not obvious.
  • Symptoms:
  • Can't alt-tab between applications Linux hosts. Ctrl-Alt-arrows to shift between desktops works though.
  • In Windows guest, left click behave like a right click. Also, mouse is apparently captured by one Windows application, for which the mouse behaves correctly. Clicking on another application gives strange result.
  • Workarounds:
  • In Windows guest, select the application we want to use, then ctrl-home, disable and enable back the mouse input. Doing this seems to fix the mouse for the new application.
  • In Windows guest, right-ctrl-delete to lock the screen then log in back. This seems to solve all issues in guest, but leaves the host in stuck state (can't alt-tab).
  • In Linux host, ctrl-home to bring VB menu, then alt-tab to force application switch. Locking the screen then login back. Ctrl-P in VirtualBox to pause the VM.
  • In Linux host, kill compiz. If the desktop is completely stuck, Ctrl-Alt-Left still works, then use Ctrl-Alt-T to open a new terminator windows. This works best if the desktop is completely empty (and does not work if current desktop already contains a terminator windows because then the new windows appears in the background and is out of reach). Then compiz --replace&, or better metacity --replace&. Metacity looks more stable.
  • Things to try (read from some forums):
  • Reinstall the Windows guest additions but enable full 3D support.
  • This post describes a problem very similar to mine, but windows host, linux guest. They suggest to Kill the VboxClient process that has a --draganddrop command line flag.
  • Workaround — Just noticed the bug is correlated to my Jabra Pro headset, connected to host via USB. Everytime I take the headset from the pod, the problem appears. Hanging up the set solves the problem. I noticed the USB device was no longer attached to the virtual machine, so maybe long term solution is to add the USB device to the USB device filter. Could the problem be linked to the KWorker-taking-100%-cpu issue?
  • Fix — Restore the USB filter to attach Jabra Pro automatically to VM (existing filters were old, not matching anymore).
  • Better fix — Disconnect the Jabra Pro USB cable.
  • Issue — indicator-multiload displays CPU usage in percent next to graphs.
  • Fix Open Preferences, then Indicator Items, and add an empty indicator item before the first line [91].
  • Issue — Stuck mouse button issue sporadically.
  • Fix — Was due to my Bluetooth mouse, still powered on in my PC bag! Make sure the damn thing is off.
  • Issue — update-initramfs returns the error below:
    • Fix — Remove some leftover file: sudo rm /var/lib/initramfs-tools/4.13.0-21-generic.
update-initramfs: Generating /boot/initrd.img-4.13.0-21-generic
WARNING: missing /lib/modules/4.13.0-21-generic
Ensure all necessary drivers are built into the linux image!

Issues on Zesty (ZAVCWL0005)

No indicator icon for Skype, Dropbox, Discord

Source: [92] 

cat /usr/share/applications/discord.desktop|grep Exec
# Exec=env XDG_CURRENT_DESKTOP=Unity /usr/share/discord/Discord
cat /usr/share/applications/skypeforlinux.desktop|grep Exec
# Exec=env XDG_CURRENT_DESKTOP=Unity /usr/bin/skypeforlinux %U

Issues on XBMCBuntu

  • No audio in some applications with HDMI output — Some applications (like Firefox or flash player) are silent.
Install and launch pavucontrol to select the output device. Also, select HDMI as the fallback device.

Issues on VirtualBox Debian 9

  • File changes done in the VM not immediately visible when served via Samba — This may be due to SMB2 protocol [93].
  • Trying to set server max protocol = SMB2_10 (servermaxprotocol). No effect.
  • Trying to disable SMB2.0 in Windows [94]. No effect.
Rem to disable SMB 2.0:
sc config lanmanworkstation depend= bowser/mrxsmb10/nsi
sc config mrxsmb20 start= disabled
Rem to enable back:
sc config lanmanworkstation depend= bowser/mrxsmb10/mrxsmb20/nsi
sc config mrxsmb20 start= auto
  • Trying to disable opportunistic lock on Windows [95], [96], [97]. No effect.
Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\MRXSmb\Parameters\]
OplocksDisabled=dword:00000001
  • Trying to disable oplocks in Samba config [98]. We add to /etc/samba/smb.conf:
[Global]
oplocks = False
level2 oplocks = False

kernel oplocks = yes

# Do this for every share
[shareXYZ]
oplocks = False
level2 oplocks = False
kernel oplocks = yes
... This works! Is it enough to disable on server-side or we must also keep the Windows client-side configuration.

Issues on Kraken - Xubuntu 18.04

  • Fixed — Screen keeps locking itself.
  • In Settings -> Power -> Display -> Set everything to 0 (for Never).
  • In Settings -> Application Startup -> uncheck screensaver and light-locker. Reboot or killall light-locker.
  • In a terminal, do sudo xset s 0 0. Make sure variable DISPLAY is defined [99].

Issues on Debian Buster (zavcxl0006)

  • Fixed — see some.pdf opens Gimp, not Evince
Permanent fix — Must add an entry in /etc/mailcap.order, but this requires creating a MIME entry [100]:
echo 'application/pdf; evince %s; test=test -n "$DISPLAY"' | sudo tee /usr/lib/mime/packages/evince
echo 'evince:application/pdf'| sudo tee -a /etc/mailcap.order
sudo update-mime
Temporary fix — Edit file /etc/mailcap, and remove the line application/pdf; gimp-2.10 %s; test=test -n "$DISPLAY"
  • Issue (with workaround) — spam from tracker-extract in /var/log/messages
Mar 20 09:25:58 zavcxl0006 tracker-extract[20935]: Could not insert metadata for item "file:///home/peetersm/Documents/news/iter/Unofficial%20ITER%20fan%20club%20-%20Sandia%E2%80%99s%20Z%20machine%20exceeds%20two%20billion%20degrees%20Kelvin.htm": 15.48: invalid UTF-8 character
Mar 20 09:25:58 zavcxl0006 tracker-extract[20935]: If the error above is recurrent for the same item/ID, consider running "tracker-extract" in the terminal with the TRACKER_VERBOSITY=3 environment variable, and filing a bug with the additional information
  • there were only 2 occurences (found with sudo cat /var/log/messages|grep tracker-extract|sed -r 's/^.*]://'|sort -u). These were 2 HTML files we converted to PDF.
  • Question: Should we disable tracker-extract?
  • Recently, tracker-extract eating lot of CPU. So let's disable it!
  • Workaround: First, disable Gnome settings → Search. Turn off everything, and turn off main switch (in window bar)source lang="bash". Then disable tracker service as well:
systemctl --user mask tracker-store.service tracker-miner-fs.service tracker-miner-rss.service tracker-extract.service tracker-miner-apps.service tracker-writeback.service
  • ISSUE — spam from gnome-shell:
Mar 20 09:30:22 zavcxl0006 gnome-shell[7751]: JS ERROR: TypeError: Shell.StageInputMode is undefined#012_onFocusChange@/home/peetersm/.local/share/gnome-shell/extensions/no-title-bar@franglais125.gmail.com/app_menu.js:125:1#012wrapper@resource:///org/gnome/gjs/modules/_legacy.js:82:22#012activateWindow@resource:///org/gnome/shell/ui/main.js:573:9#012_finish@resource:///org/gnome/shell/ui/altTab.js:615:9#012wrapper@resource:///org/gnome/gjs/modules/_legacy.js:82:22#012_keyReleaseEvent@resource:///org/gnome/shell/ui/switcherPopup.js:209:17#012wrapper@resource:///org/gnome/gjs/modules/_legacy.js:82:22
  • ISSUE — Paste mode not closed correctly when pasting in neovim/terminator. Instead we see a final tag <Paste> being padded.
  • Fixed — Screen blacking out after 15sec of inactivity, even though power saving is disabled in Gnome settings (no screen dimming when inactive, never blank screen).
Killing one of these applications fixed the issue
pkill -9 -fu $USERNAME "gsd-(power|print-notifications|rfkill|screensaver-proxy|sharing|smartcard)"
Killing gsd-power is enough apparently (or maybe with gsd-screensaver-proxy):
pkill -fu $USERNAME "gsd-power"
Permanent fix Someone suggested to add consoleblank=0 [101]. Edit the file /etc/default/grub, then run update-grub:
GRUB_CMDLINE_LINUX="consoleblank=0"
  • Issue (with workaround) — Hickups (micro-freeze) when using Virtualbox with Windows guest, scrolling with mouse pad very slow and jerky. VirtualBox quite sluggish.
  • Workaround — Switch to Oracle version, 6.0.5-129923 (testing version).
  • Uninstall debian virtualbox package, and install Oracle own package (here even testing version): VirtualBox-6.0.5-129923-Linux_amd64.run, and extension pack Oracle_VM_VirtualBox_Extension_Pack-6.0.5-129921.vbox-extpack.
  • Update: debian VirtualBox 6.0.6 package, still same issue. Removing VNC extension package, same. Reverting to 6.0.5
  • Workaround — Upgraded to virtualbox-6.0=6.0.14-133895~Debian~buster (from oracle debian repo [102]).
  • Decrease monitor count from 2 to 1.
  • Decrease video memory from 256MB to 128MB.
  • Hold the package: sudo apt-mark hold virtualbox-6.0.
  • → Best results so far!
  • Issue (with workaround) — Gnome extension User Themes crashing after suspend resume (symptom: top bar becomes black)
Workaround — Restart Gnome-Shell with Alt-F2, then r.
  • ISSUE — Zoom with CTRL + touchpad mousewheel not working on Chromium on Linux.
Known issue [103] but upstream won't fix. Instead use extension ctrlZoom.
  • ISSUE — Win7 in VirtualBox, Skype for Business graphical glitch when using driver VBoxSVGA (with 3D accel).
  • Workaround: Revert to VBoxVGA, but deprecation warning regarding 3D acceleration.
  • See ticket 18982.
  • Fixed — cmus not playing through BT headset when started before BT headset powered on.
fix: we add these lines to /etc/pulse/defaut.pa [104]
### Issue: cmus not connecting to BT headset when started before BT connected
### Trying this from https://wiki.archlinux.org/index.php/Bluetooth_headset
### automatically switch to newly-connected devices
load-module module-switch-on-connect
Then we restart the daemon with pulseaudio -k (NOTE: maybe restarting PA was sufficient in fact...)
  • Done — Upgrade bluez to connect to new BT headset WH-1000XM3. We follow advice on askubuntu.com to upgrade to bluez 5.52.
git clone https://salsa.debian.org/bluetooth-team/bluez.git # URL got when doing 'apt-get source bluez'
cd bluez
sudo apt build-dep bluez
sudo apt install devscripts
debchange --local xeyownt --preserve
dpkg-buildpackage -rfakeroot -b
sudo dpkg -i bluetooth_5.52-1xeyownt1_all.deb bluez_5.52-1xeyownt1_amd64.deb bluez-obexd_5.52-1xeyownt1_amd64.deb libbluetooth3_5.52-1xeyownt1_amd64.deb 
# Restart bluetooth service
sudo service bluetooth restart
  • Issue — Some trouble connecting back to new BT headset WH-1000XM3.
  • Current working recipe:
  • First check if a2dp_sink is reported as active profile with
pactl list | grep -C2 A2DP
# ...
#     Active Profile: a2dp_sink
  • If not, run the one liner until it is fixed:
sudo systemctl restart bluetooth.service; sleep 1; echo "connect 38:18:4C:4B:6A:3A" | bluetoothctl; sleep 1; while ! (pactl list | grep -C2 A2DP); do sleep 1; done
  • Note: to get PA current profile: pactl list, or to make sure PA uses the correct profile, pactl list | grep -C2 A2DP:
    Profiles:
        headset_head_unit: Headset Head Unit (HSP/HFP) (sinks: 1, sources: 1, priority: 30, available: no)
        a2dp_sink: High Fidelity Playback (A2DP Sink) (sinks: 1, sources: 0, priority: 40, available: yes)
        off: Off (sinks: 0, sources: 0, priority: 0, available: yes)
    Active Profile: a2dp_sink
  • Reference page for bluetooth: see archlinux wiki.
  • Troubleshoot and workaround:
# Switch on headset and wait BT connection. Then check if active profile is set to a2dp_sink:
pactl list | grep -C2 A2DP
#      Profiles:
#          headset_head_unit: Headset Head Unit (HSP/HFP) (sinks: 1, sources: 1, priority: 30, available: no)
#          a2dp_sink: High Fidelity Playback (A2DP Sink) (sinks: 1, sources: 0, priority: 40, available: yes)
#          off: Off (sinks: 0, sources: 0, priority: 0, available: yes)
#      Active Profile: a2dp_sink

# If not, we will restart BT daemon, then manually connect using bluetoothctl:
sudo systemctl start bluetooth.service
sleep 3                                    # TODO: try a simple restart
sudo systemctl stop bluetooth.service
bluetoothctl

# In bluetoothctl:
connect 38:18:4C:4B:6A:3A
quit

# Check profile again
pactl list | grep -C2 A2DP
  • ISSUE — Lost ethernet device, or connection going really slow.
  • FIXED - Doing lshw, we notice our device is configure for 10Mbit/s
      *-network
            description: Ethernet interface
            logical name: enp0s31f6
            size: 10Mbit/s
            capacity: 1Gbit/s
  • We can use ethtools (from package ethtool) to reconfigure the device [105], [106]:
sudo ethtool -s enp0s31f6 speed 1000 duplex full
sudo ethtool enp0s31f6
# Settings for enp0s31f6:
# ...
#     Speed: 1000Mb/s
#     Duplex: Full
  • ISSUE gdm3 showing black screen after suspend resume. Occuring once every 10 days or so.
  • Other problem: GDM refusing to lock the screen because VirtualBox is in full-screen. WTF?
  • Trying: Use lightdm and gnome-screensaver instead of gdm3.
  • Install lightdm and gnome-screensaver, and choose lightdm as default window manager.
  • Edit /etc/xdg/autostart/gnome-screensaver.desktop and comment out line OnlyShowIn (see Freedesktop). Alternatively copy that file into ~/.config/autostart, and edit accordingly.
  • Update: black screen with lightdm on March 2nd, around 9:30am. Reboot around 10:00am. This redhat bug report is relevant (searching for Timeout waiting for DDI BUF C idle bit).
  • Trying: create file /usr/share/X11/xorg.conf.d/20-intel.conf as suggested redhat bug report.
  • Still issues. Installing buster-backports repository, and upgrading all packages, including kernel 5.4.13.
  • As result of kernel upgrade, we can't compile virtualbox 6.0.14 module anymore (sudo /sbin/vboxconfig).
  • *NEW* Now, getting heavy GPU load in VirtualBox (Win10 task manager showing heavy GPU load related to OUTLOOK.EXE).
  • *NEW* GPU freezes in Linux. System still alive but mostly unresponsive. VT switching possible but slow. In kernel log, spam of Resetting rcs0 for hang on rcs0. This relates to some kernel bug in 5.4.x series (see Google). Trying to upgrade to latest sid kernel. If not, will have to roll-back to 4.x kernel.
  • Since kernel update, cannot use vmware player anymore. To fix, use vmmon and vmnet from mkubecek/vmware-host-modules. Using tag (p15.5.1-k5.4). See INSTALL:
make tarballs
sudo cp -v vmmon.tar vmnet.tar /usr/lib/vmware/modules/source/
sudo vmware-modconfig --console --install-all
  • ISSUE We have the following error message at boot:
Mar 02 10:24:48 zavcxl0006 kernel: i915 0000:00:02.0: firmware: failed to load i915/kbl_dmc_ver1_04.bin (-2)
Mar 02 10:24:48 zavcxl0006 kernel: firmware_class: See https://wiki.debian.org/Firmware for information about missing firmware
  • Fix: install package firmware-misc-nonfree [108], [109].
  • ISSUE We have the following ACPI error message at boot:
Mar 02 10:24:48 zavcxl0006 kernel: ACPI Error: Field [CAP1] at bit offset/length 64/32 exceeds size of target Buffer (64 bits) (2018081
Mar 02 10:24:48 zavcxl0006 kernel: ACPI Error: Method parse/execution failed \_SB._OSC, AE_AML_BUFFER_LIMIT (20180810/psparse-516)
  • Relevant bug report: [110] (more [111], [112], [113], [114]).
  • However this is apparently only a minor issue (like some hotkeys not working).
  • ISSUE On kernel 5.4.19-1, suspend not always working.
  • We get messages like INFO: task systemd-sleep:723565 blocked for more than 1087 seconds.
  • Workaround?: Turning off the Wifi (and turning back on ethernet, which apparently was turned off) triggered the suspend... (going suspend finally on Mar 09 17:59:45, see journalctl-mar_09-175945-going_suspend_after_stale.log)
  • ISSUE Intermittent freeze in VMWare guest. When it happens CPU usage skyrockets, and we see 100% usage for processes vmware-vmx and khugepaged.
  • See this vmware report. Recommendation is to disable transparent huge pages:
echo never > /sys/kernel/mm/transparent_hugepage/enabled
echo never > /sys/kernel/mm/transparent_hugepage/defrag
echo 0 > /sys/kernel/mm/transparent_hugepage/khugepaged/defrag
  • See this post for more information on huge pages (and advice to disable transparent huge pages). Also here, we don't miss anything by turning them off. More troubleshooting here as well.
  • See here to disable the transparent huge page permanently using systemd, that is create the file /etc/systemd/system/disable-transparent-huge-pages.service:
[Unit]
Description=Disable Transparent Huge Pages (THP)
DefaultDependencies=no
After=sysinit.target local-fs.target
Before=mongod.service

[Service]
Type=oneshot
ExecStart=/bin/sh -c 'echo never | tee /sys/kernel/mm/transparent_hugepage/enabled > /dev/null'
ExecStart=/bin/sh -c 'echo never | tee /sys/kernel/mm/transparent_hugepage/defrag > /dev/null'
ExecStart=/bin/sh -c 'echo 0 | tee /sys/kernel/mm/transparent_hugepage/khugepaged/defrag > /dev/null'

[Install]
WantedBy=basic.target
  • Then reload and start the services:
sudo systemctl daemon-reload
sudo systemctl start disable-transparent-huge-pages
  • Then make sure the service is started at boot:
sudo systemctl enable disable-transparent-huge-pages
  • ISSUE Freeze forcing hard reboot (after 8 days uptime):
Mar 14 15:58:06 zavcxl0006 kernel: [331200.493098] Asynchronous wait on fence i915:gnome-shell[2277]:2e3bce timed out (hint:intel_atomic_commit_ready+0x0/0x50 [i915])
Mar 14 15:58:09 zavcxl0006 kernel: [331204.333102] list_del corruption. prev->next should be ffff9b0e1f17f078, but was 0000000000000000
Mar 14 15:58:09 zavcxl0006 kernel: [331204.333120] ------------[ cut here ]------------
Mar 14 15:58:09 zavcxl0006 kernel: [331204.333123] kernel BUG at lib/list_debug.c:51!
Mar 14 15:58:09 zavcxl0006 kernel: [331204.333133] invalid opcode: 0000 [#1] SMP PTI
Mar 14 15:58:09 zavcxl0006 kernel: [331204.333139] CPU: 6 PID: 1167171 Comm: kworker/u16:1 Tainted: G           OE     5.4.0-4-amd64 #1 Debian 5.4.19-1
Mar 14 15:58:09 zavcxl0006 kernel: [331204.333142] Hardware name: HP HP EliteBook 840 G5/83B2, BIOS Q78 Ver. 01.04.00 09/12/2018
Mar 14 15:58:09 zavcxl0006 kernel: [331204.333222] Workqueue: i915 retire_work_handler [i915]
  • Fixed Missing firmware when creating initramfs:
update-initramfs: Generating /boot/initrd.img-5.4.0-4-amd64
W: Possible missing firmware /lib/firmware/i915/icl_dmc_ver1_07.bin for module i915
W: Possible missing firmware /lib/firmware/i915/tgl_dmc_ver2_04.bin for module i915
W: Possible missing firmware /lib/firmware/i915/bxt_huc_ver01_8_2893.bin for module i915
  • Relevant AskUbuntu post
  • FIXDownload the relevant drivers from intel github, and copy them into /lib/firmware/i915. Then sudo update-initramfs -u.
  • ISSUE Samba shared (in particular those mounted with autofs) are staling when network/server no longer available.
  • ISSUE — Suspend does not work anymore on HP 8540w (prob since switch Ubuntu -> Debian)
  • Might be related to wifi not working [116]. See errors in sudo journalctl -b. Blacklisting drivers might help [117]:
for m in iwlwifi iwldvm; do echo blacklist $m > $m.conf; done
depmod -ae
udpate-initramfs -u
  • No difference. Let's remove the blacklist and install package firmware-iwlwifi.
  • TODO — See Debian On HP 8540w for more stuff to fix. In particular it says that suspend will not work with nouveau, and that nvidia driver must be installed.
  • ISSUE Promiscuous-mode not enabled for VMWare. This is necessary when starting the Cisco AnyConnect VPN on Windows.
  • Create a new systemd unit file /etc/systemd/system/vmware-promiscuous.service:
[Unit]
Description=Enable vmware promiscuous mode
After=vmware.service

[Service]
Type=oneshot
ExecStart=/bin/bash -c "chgrp vmnet /dev/vmnet*; chmod g+rw /dev/vmnet*"
ExecStop=/bin/bash -c "chgrp root /dev/vmnet*; chmod g-rw /dev/vmnet*"
RemainAfterExit=yes

[Install]
WantedBy=multi-user.target
  • Then create the group and enable the service:
# Create a group vmnet
sudo addgroup --system vmnet
# Add user 'vbox' to that group
sudo gpasswd -a vbox vmnet
# Enable systemd service:
sudo systemctl daemon-reload
sudo systemctl enable vmware-promiscuous.service 
sudo systemctl start vmware-promiscuous.service
  • FIXED We lose bluetooth. In gnome control-center, it says "no bluetooth found, insert usb dongle...". At boot, very early (during crypt password), we get usb 1-5: device not accepting address 21, error -32.
  • Fix: Issue maybe related to USB over current protection. Shutdown the PC, disconnect all USB device, wait a couple minutes [118].
  • Listed workarounds: Upgrade to more recent Courier-IMAP, which apparently reduces freq. of occurence; or Switch to Dovecot.
  • ISSUE Poor microphone quality on my BT headset Sony WH-1000XM3. See bug +1838151, in progress on Ubuntu.
  • ISSUE Poor quality of internal microphone on HP elitebook 640 G4. Is this a Linux issue? Maybe related. In fact, microphone is not bad, but we have to boost volume a lot, and lot of ambient + white noise is captured.
  • ISSUE On Kernel 5.9.6-1, we have huge freeze in vmware win10. We see process vmware-vmw going to 400% and kcompactd0 going to 100% CPU. This does not happen on Kernel 5.8.10-1.
  • Maybe linked to new proactive memory compaction commit Kernel 5.9 features.
  • This is apparently again related to the transparent huge page issues (see issue above, related to khugepaged).
  • Seems hard to disable that thing.
  • FIX attemptThe recommended way [122] is to add transparent_hugepage=never to kernel command line (edit /etc/default/grub, then sudo update-grub).
  • Also, we must set the relevant files in /proc (see issue above). Some user reports it does not work [123]. RedHat may have fix for that but we don't have access [124], [125], [126]
  • TO TRY: Set 0 into /sys/kernel/mm/transparent_hugepage/khugepaged/defrag, echo 0 > /sys/kernel/mm/transparent_hugepage/khugepaged/defrag [127]. Use systemd trick (see issue above) to make it permanent.
  • Background info on transparent huge pages: [128], [129], [130]
  • Filed as a bug in debian: #995470
  • ISSUE On Vim with latest YCM + UltiSnips plugin, snippets with auto-expand or with braces inside are incorrectly expanded (see [131], [132])
  • FIXED Since libcurl3-gnutls:amd64 upgrade from Debian backports (got version 7.64.0-4+deb10u1 over 7.74.0-1.2~bpo10+1), the git clone fails:
$ git clone https://github.com/pyexcel/pyexcel-xls.git
Cloning into 'pyexcel-xls'...
fatal: unable to access 'https://github.com/pyexcel/pyexcel-xls.git/': Failed sending HTTP2 data
Fix is to reinstall from stable (or oldstable now). See [133], [134].
sudo apt reinstall libcurl3-gnutls/oldstable
sudo apt-mark hold libcurl3-gnutls
  • ISSUE Severe out-of-memory leading to complete freeze since upgrade recoll:amd64 from 1.27.12-1~ppa1~buster to 1.31.0-1~ppa1~buster. The cause is the processing of a file named fat-loop.xls in Mediawiki website source:
python3 /usr/share/recoll/filters/xls-dump.py --dump-mode=canonical-xml --utf-8 --catch /home/data/www/html/mw1.35.1/tests/phpunit/data/MSCompoundFileReader/fat-loop.xls
See bug #8.
Workaround tried: Add fat-loop.xls to excluded file. Rebuild the index.
More workaround: change the oom_score when recollindex starts. On my system the highest value is from Chromium that has a oom_score_adj of 300. So let's use same value, change user crontab (crontab -e):
15 12  * * *     RCLCRON_RCLINDEX= RECOLL_CONFDIR="/home/peetersm/.recoll" choom -n 300 recollindex
Workaround submitted as Issue 117. There is a fix available in Recoll 1.31.1 already.
  • ISSUE File copy freezes in Nautilus (aka files) when trying to copy some XML file from Samba share (eg. '/smb/mnemosyne/movie/children/Les Mysterieuses Cites d Or - DVD 1/VIDEO_TS/VIDEO_TS.nfo').

Issue on Debian Buster (Griffin)

  • Issue Losing SSH connect to zavcxl0006 regularly, like once a day.
Using ssh zavcxl0006; date when connecting to get the exact time, and cross-ref with /var/log/messages.
It seems to be related to DHCP lease time. Checking the log I notice the client is regularly changing IP address.
Workaround: Setting static addresses for griffin & zavcxl0006 devices.

Issue on Debian Bullseye (Wonderzyra)

  • Issue — Losing BT mouse connection at every boot.
  • Workaround - open settings, remove device. Force pairing on mouse. Connect device again.

Issue on Debian Buster in VMWare + Windows host

  • Issue — Slow nvim startup due to slow clipboard
  • Host: Windows 7, XMing 7.7.0.52 — Client: Debian 10 Buster 64, nvim 0.4.4, xsel.
  • nvim startup analysis (rm -f vim.log && time vi --startuptime vim.log -c 'qa!' && grep -C 3 clipboard vim.log) reveal that clipboard.vim has a 1s timeout:
   091.664  008.686  000.372: sourcing /usr/local/share/nvim/runtime/syntax/syntax.vim
   1132.111  1039.313  1039.313: sourcing /usr/local/share/nvim/runtime/autoload/provider/clipboard.vim
  • Checking in that file, we can reproduce the same delay with
xsel -o -b
# real	0m1.008s
# user	0m0.001s
# sys	0m0.006s
  • Disabling VMWare clipboard integration has no effect. Installing open-vm-tools-desktop has no effect. Finally the fix was to install xclip [135]:
# Install xclip to fix nvim startup issue due to xsel timeout
sudo apt install xclip

Issue on Debian BookWorm (zavcxl0007,ZAVCWL0136)

  • Issue — Issue reported by sudo journalctl -b (expand for details)
Apr 11 14:40:41 zavcxl0007 kernel: x86/tme: enabled by BIOS
Apr 11 14:40:41 zavcxl0007 kernel: x86/mktme: No known encryption algorithm is supported: 0x0
Apr 11 14:40:41 zavcxl0007 kernel: x86/mktme: disabled by BIOS
...
Apr 11 14:40:41 zavcxl0007 kernel: ENERGY_PERF_BIAS: Set to 'normal', was 'performance'
...
Apr 11 14:40:41 zavcxl0007 kernel: ACPI: [Firmware Bug]: BIOS _OSI(Linux) query ignored
...
Apr 11 14:40:41 zavcxl0007 kernel: pci 0000:00:07.0: DPC: RP PIO log size 0 is invalid
Apr 11 14:40:41 zavcxl0007 kernel: pci 0000:00:07.1: [8086:9a25] type 01 class 0x060400
Apr 11 14:40:41 zavcxl0007 kernel: pci 0000:00:07.1: PME# supported from D0 D3hot D3cold
Apr 11 14:40:41 zavcxl0007 kernel: pci 0000:00:07.1: DPC: RP PIO log size 0 is invalid
...
Apr 11 14:40:41 zavcxl0007 kernel: tpm tpm0: [Firmware Bug]: TPM interrupt not working, polling instead
Apr 11 14:40:41 zavcxl0007 kernel: AMD-Vi: AMD IOMMUv2 functionality not available on this system - This is not a bug.
Apr 11 14:40:41 zavcxl0007 kernel: i8042: PNP: PS/2 Controller [PNP0303:PS2K] at 0x60,0x64 irq 1
Apr 11 14:40:41 zavcxl0007 kernel: i8042: PNP: PS/2 appears to have AUX port disabled, if this is incorrect please boot with i8042.nopnp
...
Apr 11 14:40:41 zavcxl0007 kernel: i915 0000:00:02.0: firmware: failed to load i915/tgl_dmc_ver2_12.bin (-2)
Apr 11 14:40:41 zavcxl0007 kernel: firmware_class: See https://wiki.debian.org/Firmware for information about missing firmware
Apr 11 14:40:41 zavcxl0007 kernel: i915 0000:00:02.0: Direct firmware load for i915/tgl_dmc_ver2_12.bin failed with error -2
Apr 11 14:40:41 zavcxl0007 kernel: i915 0000:00:02.0: [drm] Failed to load DMC firmware i915/tgl_dmc_ver2_12.bin. Disabling runtime power management.
Apr 11 14:40:41 zavcxl0007 kernel: i915 0000:00:02.0: [drm] DMC firmware homepage: https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/i915
Fixed i915 fixed now with recent non-free firmware upgrade
Apr 11 16:28:30 zavcxl0007 kernel: i915 0000:00:02.0: firmware: direct-loading firmware i915/tgl_dmc_ver2_12.bin
Apr 11 16:28:30 zavcxl0007 kernel: i915 0000:00:02.0: [drm] Finished loading DMC firmware i915/tgl_dmc_ver2_12.bin (v2.12)

...
Apr 11 14:40:41 zavcxl0007 kernel: Lockdown: resume: hibernation is restricted; see man kernel_lockdown.7
Apr 11 14:40:41 zavcxl0007 kernel: Lockdown: systemd: /dev/mem,kmem,port is restricted; see man kernel_lockdown.7
...
Apr 11 14:40:41 zavcxl0007 kernel: iwlwifi 0000:00:14.3: firmware: failed to load iwlwifi-QuZ-a0-hr-b0-67.ucode (-2)
Apr 11 14:40:41 zavcxl0007 kernel: iwlwifi 0000:00:14.3: Direct firmware load for iwlwifi-QuZ-a0-hr-b0-67.ucode failed with error -2
Apr 11 14:40:41 zavcxl0007 kernel: iwlwifi 0000:00:14.3: firmware: failed to load iwlwifi-QuZ-a0-hr-b0-66.ucode (-2)
Apr 11 14:40:41 zavcxl0007 kernel: iwlwifi 0000:00:14.3: Direct firmware load for iwlwifi-QuZ-a0-hr-b0-66.ucode failed with error -2
Apr 11 14:40:41 zavcxl0007 kernel: iwlwifi 0000:00:14.3: firmware: failed to load iwlwifi-QuZ-a0-hr-b0-65.ucode (-2)
Apr 11 14:40:41 zavcxl0007 kernel: iwlwifi 0000:00:14.3: Direct firmware load for iwlwifi-QuZ-a0-hr-b0-65.ucode failed with error -2
Apr 11 14:40:41 zavcxl0007 kernel: iwlwifi 0000:00:14.3: firmware: failed to load iwlwifi-QuZ-a0-hr-b0-64.ucode (-2)
Apr 11 14:40:41 zavcxl0007 kernel: iwlwifi 0000:00:14.3: Direct firmware load for iwlwifi-QuZ-a0-hr-b0-64.ucode failed with error -2
Apr 11 14:40:41 zavcxl0007 kernel: 'iwlwifi 0000:00:14.3: firmware: failed to load iwl-debug-yoyo.bin (-2)
Fixed iwlwifi fixed with non-free firmware upgrade now (otherwise see also intel wireless firmware.
Apr 11 16:28:30 zavcxl0007 kernel: iwlwifi 0000:00:14.3: firmware: direct-loading firmware iwlwifi-QuZ-a0-hr-b0-63.ucode

...
Apr 11 14:40:43 zavcxl0007 kernel: Bluetooth: hci0: Failed to read codec capabilities (-56)
Apr 11 14:40:43 zavcxl0007 kernel: NET: Registered PF_ALG protocol family
Apr 11 14:40:43 zavcxl0007 bluetoothd[740]: profiles/sap/server.c:sap_server_register() Sap driver initialization failed.
Apr 11 14:40:43 zavcxl0007 bluetoothd[740]: sap-server: Operation not permitted (1)
...
Apr 11 14:40:44 zavcxl0007 pipewire[974]: spa.alsa: '_ucm0001.hw:sofhdadsp,5': playback open failed: Device or resource busy
Apr 11 14:40:44 zavcxl0007 pipewire[974]: mod.adapter: 0x55ba38aff7a0: can't get format: Device or resource busy
...
Apr 11 14:40:44 zavcxl0007 pipewire[974]: spa.alsa: '_ucm0001.hw:sofhdadsp': capture open failed: Device or resource busy
Apr 11 14:40:44 zavcxl0007 pipewire[974]: spa.alsa: '_ucm0001.hw:sofhdadsp,6': capture open failed: Device or resource busy
  • Issue Synapse — keyboard shortcut doesn't work under Wayland. Also, receive a message on launch saying Synapse wants to inhibit shortcuts... You can restore shortcuts by pressing Super+Escape, with no obvious effect.
  • Issue kcompactd0 popping up regularly in the top. Not taking as much CPU as it used too. Stability?
  • Applying this fix into /etc/sysctl.conf source
# Fix problem where vmware battles with kcompactd0.
vm.compaction_proactiveness=0
... However, still seeing kcompactd0 kicking in and slowing down the VM (but no huge dead-locked freeze)
  • Alternative is to add similar lines in VM vmx file [136]
# Fix problem where vmware battles with kcompactd0
vm.compaction_proactiveness=0
... Same... still kicking in. Also doing:
echo never > /sys/kernel/mm/transparent_hugepage/defrag        # was 'madvise'
echo 0 > /sys/kernel/mm/transparent_hugepage/khugepaged/defrag # was '1'
echo never > /sys/kernel/mm/transparent_hugepage/enabled       # was 'always'
  • ... with changes above, kcompactd0 kicks when the VM starts, but apparently not anymore afterwards... to be confirmed.
  • See also [137], [138]. See also latest patch. Also this nice lwn.net write up.
  • [2022-04-26 11:48] 2min freeze in VM
  • Issue Spam message in kernel log. See [139].
Apr 15 10:27:16 zavcxl0007 kernel: x86/split lock detection: #AC: main-svga/8996 took a split_lock trap at address: 0x55bd812a2e3d
Apr 15 10:27:17 zavcxl0007 kernel: x86/split lock detection: #AC: main-svga/8996 took a split_lock trap at address: 0x55bd812a2e3d

Apr 15 10:41:22 zavcxl0007 kernel: x86/split lock detection: #AC: main-svga/8996 took a split_lock trap at address: 0x55bd812a2e3d
Apr 15 10:41:24 zavcxl0007 kernel: x86/split lock detection: #AC: main-svga/8996 took a split_lock trap at address: 0x55bd812a2e3d
Apr 15 10:41:34 zavcxl0007 kernel: x86/split lock detection: #AC: main-svga/8996 took a split_lock trap at address: 0x55bd812a2e3d
  • Issue VMWare crashed (Win10 guest). Log has no information on the crash.
Apr 15 10:41:45 zavcxl0007 synapse.desktop[8816]: VMware Player Error:
Apr 15 10:41:45 zavcxl0007 synapse.desktop[8816]: VMware Player unrecoverable error: (vmplayer)
Apr 15 10:41:45 zavcxl0007 synapse.desktop[8816]: Unexpected signal: 11.
Apr 15 10:41:45 zavcxl0007 synapse.desktop[8816]: A log file is available in "/tmp/vmware-peetersm/vmware-vmplayer-8816.log".
Apr 15 10:41:45 zavcxl0007 synapse.desktop[8816]: You can request support.
Apr 15 10:41:45 zavcxl0007 synapse.desktop[8816]: To collect data to submit to VMware technical support, run "vm-support".
Apr 15 10:41:45 zavcxl0007 synapse.desktop[8816]: We will respond on the basis of your support entitlement.
  • Fixed Do not see the Bluetooth Logitech Mouse M555b.
We connect manually, after collecting address on previous PC
bluetoothctl
> connect 00:1F:20:97:ED:C8
  • Issue — Terminator is stuck when awaking from stand-by. Must kill it and start afresh.
  • Issue — ZAVCWV0009 (VMWare) - Choppy sound when using high definition audio device in Webex, and less choppy when using jabra headset, but issue with mic reported as not working. However the VM was not using lot of CPU power in host.
  • We generate a new VM *.vmx file from scratch, adding back the hard disk later. This requires keeping the disk / system UID, and also restoring EFI boot in BIOS (press F2).
  • Issue — ZAVCWV0009 (VMWare) - Corrupted graphics in Skype (textbox background alternating between white and light gray). Looks like a 3D accel glitch. Was not present on zavcxl0006 (Debian Buster).
Content of /var/log/unattended-upgrades/unattended-upgrades.log, we see 20 minutes gap:
2022-04-21 11:18:26,923 INFO Starting unattended upgrades script
2022-04-21 11:18:26,923 INFO Allowed origins are: origin=Debian,codename=bookworm,label=Debian, origin=Debian,codename=bookworm,label=Debian-Security, origin=Debian,codename=bookworm-security,label=Debian-Security
2022-04-21 11:18:26,924 INFO Initial blacklist: 
2022-04-21 11:18:26,924 INFO Initial whitelist (not strict): 
2022-04-21 11:37:14,655 WARNING SIGTERM received, will stop
workaround:
sudo apt purge unattended-upgrades
  • Issue — following bluetooth message since April 24th: (see also [140]).
Can this be due to selecting mSBC profile on the WH-1000XM3 headset? Or due to the fact that we should disable one of the audio gateway role in the headset-roles? (see bluetooth.lua.d/50-bluez-config.lua, also [141])
Apr 19 15:48:14 zavcxl0007 bluetoothd[14403]: Starting SDP server
Apr 19 15:48:15 zavcxl0007 bluetoothd[14403]: src/profile.c:record_cb() Unable to get Hands-Free Voice gateway SDP record: Host is down
Apr 19 15:55:28 zavcxl0007 bluetoothd[14403]: src/profile.c:record_cb() Unable to get Hands-Free Voice gateway SDP record: Host is down
Apr 19 15:55:35 zavcxl0007 bluetoothd[14403]: src/profile.c:record_cb() Unable to get Hands-Free Voice gateway SDP record: Host is down
Apr 19 15:55:45 zavcxl0007 bluetoothd[14403]: src/profile.c:record_cb() Unable to get Hands-Free Voice gateway SDP record: Host is down
Apr 19 15:55:59 zavcxl0007 bluetoothd[14403]: src/profile.c:record_cb() Unable to get Hands-Free Voice gateway SDP record: Host is down
Apr 19 15:56:21 zavcxl0007 bluetoothd[14403]: src/profile.c:record_cb() Unable to get Hands-Free Voice gateway SDP record: Host is down
Apr 19 15:56:59 zavcxl0007 bluetoothd[14403]: src/profile.c:record_cb() Unable to get Hands-Free Voice gateway SDP record: Host is down
Apr 19 15:58:09 zavcxl0007 bluetoothd[14403]: src/profile.c:record_cb() Unable to get Hands-Free Voice gateway SDP record: Host is down
Also, many other bluetooth errors (related to spa_server<:code>, wireplumber[2206]: sbc_decode failed: -3, ...
  • Issue — On headset WH-1000XM3, when headset profile is used, we have audio loopback from the microphone, ie. we hear back everything the mic is recording, which is quite annoying. Also microphone is very sensitive, it would be very helpful if we could mute / unmute by double-tapping the headset.
  • Fixed — Missing firmware
update-initramfs: Generating /boot/initrd.img-5.17.0-1-amd64
W: Possible missing firmware /lib/firmware/i915/adlp_dmc_ver2_14.bin for module i915
We download the latest firmware Linux admin
cd ~/build
wget https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/plain/i915/adlp_dmc_ver2_14.bin
wget https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/plain/i915/adlp_dmc_ver2_12.bin
sudo mkdir /usr/local/lib/firmware
sudo cp adlp_dmc_ver2_12.bin /lib/firmware/i915
sudo cp adlp_dmc_ver2_14.bin /lib/firmware/i915
Note however that the message was probably irrelevant since the firmware is not loaded at boot.
  • Issue — Background color of gnome tray menu is black, making menu items hard to read. Was it like that in the original theme?

Issue on Windows 10 (ZAVCWL0136)

  • Issue Google Chrome, software_reporter_tool.exe consuming lot of CPU [142]
  • Go to three dots, More tools, Clear browsing data, and clear all browsing history.
  • Got to three dots, Settings, three dots, advanced, reset and clean up, clean up computer, and uncheck Report details to Google about harmful software...
Retrieved from "https://miki.immie.org/mediawiki/index.php?title=Common_Issues&oldid=15097"