Skip to content

troubleshooting.md

Latest commit

 

History

History
333 lines (243 loc) · 12.1 KB

troubleshooting.md

File metadata and controls

333 lines (243 loc) · 12.1 KB

Linux troublehooting

Wacom tablet not scrolling in QT apps

Wacom tablet scrolls in gtk apps but not in qt apps (kdenlive, freecad, etc). xinput test 21 shows button press 4 and 5 for scroll actions but seems that QT does not naturally recognize them as scroll actions. Firefox. terminals, etc work fine though.

The tablet scroll strip in the tablet works. Here is the output of xinput test 10

button press   4 a[3]=2 a[4]=0 a[5]=0 
button release 4 a[3]=2 a[4]=0 a[5]=0 
motion a[3]=2 a[4]=0 a[5]=0 
button press   4 a[3]=1 a[4]=0 a[5]=0 
button release 4 a[3]=1 a[4]=0 a[5]=0 
motion a[3]=1 a[4]=0 a[5]=0

Compare it with the output of the cursor

button press   5 
button release 5 
motion a[2]=0 a[3]=0 a[4]=0 a[5]=0 
button press   5 
button release 5 
motion a[2]=0 a[3]=0 a[4]=0 a[5]=0 
button press   5 
button release 5 
motion a[2]=0 a[3]=0 a[4]=0 a[5]=0

Fixing a segmentation fault

I am going to show how I fixed the following problem

[unix ~]$ cadence
Using Tray Engine 'Qt'
Segmentation fault (core dumped)

First check the list of coredumps and get info

[unix ~]$ coredumpctl list
...
Sat 2020-11-14 19:29:34 CET  199894  1000   985  11 present   /usr/bin/python3.8
[unix ~]$ coredumpctl info 199894
           PID: 199894 (python3)
           UID: 1000 (unix)
           GID: 985 (users)
        Signal: 11 (SEGV)
     Timestamp: Sat 2020-11-14 19:29:34 CET (8min ago)
  Command Line: /usr/bin/python3 /usr/share/cadence/src/cadence.py
    Executable: /usr/bin/python3.8
 Control Group: /user.slice/user-1000.slice/session-1.scope
          Unit: session-1.scope
         Slice: user-1000.slice
       Session: 1
     Owner UID: 1000 (unix)
       Boot ID: 871e977e8a4c4141a55280436b88b9b0
    Machine ID: f409642b16af41dea88241c5f8a8c126
      Hostname: hal
       Storage: /var/lib/systemd/coredump/core.python3.1000.871e977e8a4c4141a55280436b88b9b0.199894.160537857400000>
       Message: Process 199894 (python3) of user 1000 dumped core.
                
                Stack trace of thread 199894:
                #0  0x00007f878f3e70e2 _ZN10QXcbCursor12queryPointerEP14QXcbConnectionPP18QXcbVirtualDesktopP6QPoin>
                #1  0x00007f878f3f727c _ZN8QXcbDrag4initEv (libQt5XcbQpa.so.5 + 0x6c27c)
                #2  0x00007f878f3f7485 _ZN8QXcbDragC2EP14QXcbConnection (libQt5XcbQpa.so.5 + 0x6c485)
                #3  0x00007f878f3c7c63 _ZN14QXcbConnectionC1EP19QXcbNativeInterfacebjPKc (libQt5XcbQpa.so.5 + 0x3cc>
                #4  0x00007f878f3cb99c _ZN15QXcbIntegrationC2ERK11QStringListRiPPc (libQt5XcbQpa.so.5 + 0x4099c)
                #5  0x00007f878f70561d _ZN21QXcbIntegrationPlugin6createERK7QStringRK11QStringListRiPPc (libqxcb.so>
                #6  0x00007f8790cbb627 _ZN22QGuiApplicationPrivate25createPlatformIntegrationEv (libQt5Gui.so.5 + 0>
                #7  0x00007f8790cbcae1 _ZN22QGuiApplicationPrivate21createEventDispatcherEv (libQt5Gui.so.5 + 0x131>
                #8  0x00007f879488b936 _ZN23QCoreApplicationPrivate4initEv (libQt5Core.so.5 + 0x2ba936)
                #9  0x00007f8790cbfb20 _ZN22QGuiApplicationPrivate4initEv (libQt5Gui.so.5 + 0x134b20)
                #10 0x00007f87913b76aa _ZN19QApplicationPrivate4initEv (libQt5Widgets.so.5 + 0x15f6aa)
                #11 0x00007f8791cc84aa n/a (QtWidgets.abi3.so + 0x37d4aa)
                #12 0x00007f8791cc858a n/a (QtWidgets.abi3.so + 0x37d58a)
                #13 0x00007f8794c87e70 n/a (sip.cpython-38-x86_64-linux-gnu.so + 0x16e70)
                #14 0x00007f8795917570 _PyObject_MakeTpCall (libpython3.8.so.1.0 + 0x127570)
                #15 0x00007f8795912894 _PyEval_EvalFrameDefault (libpython3.8.so.1.0 + 0x122894)
                #16 0x00007f879590c9a4 _PyEval_EvalCodeWithName (libpython3.8.so.1.0 + 0x11c9a4)
                #17 0x00007f87959bcf03 PyEval_EvalCode (libpython3.8.so.1.0 + 0x1ccf03)
                #18 0x00007f87959c8868 n/a (libpython3.8.so.1.0 + 0x1d8868)
                #19 0x00007f87959c2a43 n/a (libpython3.8.so.1.0 + 0x1d2a43)
                #20 0x00007f879588267b PyRun_FileExFlags (libpython3.8.so.1.0 + 0x9267b)
                #21 0x00007f87958820f2 PyRun_SimpleFileExFlags (libpython3.8.so.1.0 + 0x920f2)
                #22 0x00007f87959d582a Py_RunMain (libpython3.8.so.1.0 + 0x1e582a)
                #23 0x00007f87959b17b9 Py_BytesMain (libpython3.8.so.1.0 + 0x1c17b9)
                #24 0x00007f879564f152 __libc_start_main (libc.so.6 + 0x28152)
                #25 0x000055759116904e _start (python3.8 + 0x104e)
                
                Stack trace of thread 199906:
                #0  0x00007f879571c46f __poll (libc.so.6 + 0xf546f)
                #1  0x00007f879069463b n/a (libxcb.so.1 + 0xc63b)
                #2  0x00007f879069637b xcb_wait_for_event (libxcb.so.1 + 0xe37b)
                #3  0x00007f878f3ee1b0 _ZN14QXcbEventQueue3runEv (libQt5XcbQpa.so.5 + 0x631b0)
                #4  0x00007f879469ee8f n/a (libQt5Core.so.5 + 0xcde8f)

Then coredumpctl gdb 199894 will show you some info

Reading symbols from /usr/bin/python3.8...
(No debugging symbols found in /usr/bin/python3.8)
[New LWP 199894]
[New LWP 199906]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
Core was generated by `/usr/bin/python3 /usr/share/cadence/src/cadence.py'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0  0x00007f878f3e70e2 in QXcbCursor::queryPointer(QXcbConnection*, QXcbVirtualDesktop**, QPoint*, int*) ()
   from /usr/lib/python3.8/site-packages/PyQt5/Qt/plugins/platforms/../../lib/libQt5XcbQpa.so.5
[Current thread is 1 (Thread 0x7f87954af740 (LWP 199894))]

From there you can backtrace at the gdb prompt (gdb) bt. Seems that I have multiple versions of libQt5XcbQpa.so.5 I link the version in site-packages to the main /usr/lib but then complains

[unix ~]$ cadence
Using Tray Engine 'Qt'
Cannot mix incompatible Qt library (5.14.1) with this library (5.15.1)
Aborted (core dumped)

SOLVED! Removed all QT5 installed via pip and reinstalled it via pacman. QT and python package managements are nasty. At least I learnt some debugging tools

Canon Lide 120

Create a file in /lib/udev/rules.d/ called 60-canonlide120.rules

Run lsusb:

[irix@x220 ~]$ lsusb
Bus 002 Device 004: ID 04a9:190e Canon, Inc. CanoScan LiDE 120

So content of file is

SUBSYSTEM=="usb", ATTRS{idVendor}=="04a9", ATTRS{idProduct}=="190e", MODE="0666"

And then reload the rules

udevadm control --reload-rules

FOUND HERE http://unix.stackexchange.com/questions/44308/understanding-udev-rules-and-permissions-in-libusb#147777/

In the supported scanners by sane, there's no Lide 120:
cat /etc/sane.d/genesys.conf

Tutorial config in Arch Linux:
https://www.eanderalx.org/linux/scanner_lide110

Old libraries not found

Error while burning bootloader.
/home/irix/opt/arduino-1.8.1/hardware/tools/avr/bin/avrdude: error while loading shared libraries: libtinfo.so.5: cannot open shared object file: No such file or directory

That is because a new library has been installed. Solve it with a symbolic link:

sudo ln -s /usr/lib/libtinfo.so /usr/lib/libtinfo.so.5

[irix@x220 lib]$ ls -l libtin*
lrwxrwxrwx 1 root root 22 May  2 13:37 libtinfo.so -> /usr/lib/libtinfo.so.6
lrwxrwxrwx 1 root root 20 May  2 14:57 libtinfo.so.5 -> /usr/lib/libtinfo.so
lrwxrwxrwx 1 root root 27 May  2 13:37 libtinfo.so.6 -> /usr/lib/libncursesw.so.6.0

Pacman troubleshooting

sudo pacman -S --force pacman-mirrorlist

sudo pacman -Syyu

CUPS

Cannot access http://localhost:631

Connection refused: Make sure CUPS is installed and activated:

systemctl enable cups
systemctl start cups

CUPS web config 403 Forbidden: Add yourself to wheel group

Roland does not appear in CUPS

Roland does not appear

ls /dev/bus/usb/00x/00x where x to be found by lsusb with printer connected and lit up. Are you on the lp user group? Adjust udev rules like this , Ensure that is set 0666.

lpinfo -v

/usr/lib/cups/backend/usb

Create udev rule

[irix@x220 ~]$ lsusb
Bus 002 Device 014: ID 0b75:0405 Roland DG Corp.

custom udev rule in a file /etc/udev/rules.d/10-usbprinter.rules containing

ATTR{idVendor}=="0b75", ATTR{idProduct}=="0405", MODE:="0660", GROUP:="lp"

This is to give permission to the printer

save the file and cat to usb device cat filename > /dev/xxx

If /usr/lib/cups/backend/usb shows Failed to open device, code: -1

Then the power management might be disabling the Device

In /etc/default/tlp add the lines USB_BLACKLIST="0b75:0405" and so on with every device

Send to Roland Vinyl

cat file.camm > /dev/usb/lp0

Not working in Arch, says...

device or resource busy

might it be because of this?

[irix@hal ~]$ cat /etc/modprobe.d/blacklistusblp.conf 
blacklist usblp

Blacklisting usblp should not be required

It works with:

lpr -P vinyl

TODO X220

  • Grub does not show menu
  • Configure coreboot
  • dotfiles
  • Dual Intel / Radeon
  • Hotplug Radeon

Intel NUC freezes with eGPU

I experienced random freezes with intel NUC 11 running an external GPU. At the beginning I thought it was a memory fault but it ended up being some sort of power saving settings in the BIOS. I followed the instructions given by marCH here https://community.intel.com/t5/Intel-NUCs/NUC10i7FNH-Issue-with-Thunderbolt3-eGPU/m-p/1279791/highlight/true but the issue is still happening..

I SOLVED the issue by issuing nvidia-settings -a [gpu:0]/GPUPowerMizerMode=1 at the start of the session. I created an alias for that stayhot which is the equivalent of dontfreeze but sounds better.

Alternatively you can make this change persistent in /etc/X11/xorg.conf

Option         "RegistryDwords" "PowerMizerEnable=0x1; PerfLevelSrc=0x2222; PowerMizerDefault=0x1; PowerMizerDefaultAC=0x1"

Also make sure the thunderbolt cable is not loose. That can also cause freezes in Linux.

X Troubleshooting

If X doesn't start usually is due to an error in xorg.conf. Do the following:

  • CTRL+ALT+F1 to enter a new login screen
  • sudo systemctl stop ly note that ly is my display manager
  • Fix the offending /etc/X11/xorg.conf
  • sudo systemctl start ly

If you are testing features you can restart X with sudo systemctl restart ly. To check for errors or warnings in X check the logs

grep EE ~/.local/share/xorg/Xorg.0.log
grep WW ~/.local/share/xorg/Xorg.0.log

bash: append_path: command not found

Everytime perl is updated this appears 3 times in every terminal window. This is because .bashrc sources /etc/profile which defines a function called appendpath and sources all files in the /etc/profile.d directory. And here's the problem: The file /etc/profile.d/perlbin.sh calls a function called append_path instead of appendpath

# Set path to perl scriptdirs if they exist
# https://wiki.archlinux.org/index.php/Perl_Policy#Binaries_and_scripts
# Added /usr/bin/*_perl dirs for scripts

[ -d /usr/bin/site_perl ] && append_path '/usr/bin/site_perl'

[ -d /usr/bin/vendor_perl ] && append_path '/usr/bin/vendor_perl'

[ -d /usr/bin/core_perl ] && append_path '/usr/bin/core_perl'

export PATH

# If you have modules in non-standard directories you can add them here.
#export PERLLIB=dir1:dir2

At the beginning I was manually fixing the perlbin.sh but I got a better

SOLUTION: Create a new alias append_path='appendpath' and make sure that in .bashrc you load the aliases before sourcing /etc/profile