#StackBounty: #drivers #graphics #xorg #google-chrome #intel Screen tearing in Chrome on Ubuntu 19.04 with Intel Graphics

Bounty: 500

I’m using dual monitors, one of which is rotated with

xrandr --output HDMI-1 --rotate left

I get screen tearing on Google Chrome (installed from Chrome’s PPAs, not Chromium) on both monitors, though it’s more noticeable on the rotated one. I don’t seem to get screen tearing in other programs, like gnome-terminal for example.

My GPU is

$ inxi -G
Graphics:  Device-1: Intel UHD Graphics 630 driver: i915 v: kernel 
           Display: x11 server: X.Org 1.20.4 driver: i915 resolution: 1920x1080~60Hz, 1920x1080~60Hz 
           OpenGL: renderer: Mesa DRI Intel UHD Graphics 630 (Coffeelake 3x8 GT2) v: 4.5 Mesa 19.0.8 

I tried adding

Section "Device"
   Identifier  "Intel Graphics"
   Driver      "intel"
   Option      "TripleBuffer" "true"
   Option      "TearFree"     "true"
   Option      "DRI"          "false"
EndSection

to /etc/X11/xorg.conf.d/20-intel-graphics.conf as per this question, but the problem is that command makes my dual monitors act as one monitor (the same picture shows up on both monitors). I tried removing the TripleBuffer and DRI lines, but that changed nothing. This side effect is mentioned in this Arch Linux Forum thread with a possible solution of adding i915.semaphores=1 to the kernel boot parameters. I don’t know if that 7 year old advice is still relevant and if I was in a situation where I could fiddle with kernel boot parameters, I probably wouldn’t be using Ubuntu in the first place.

I also tried setting the chrome://flags/#ignore-gpu-blacklist Chrome flag as per this question, but nothing seems to have changed in chrome://gpu/.

I’m using i3 as my window manager. I’m using X, not Wayland (i3 doesn’t work with Wayland).


Get this bounty!!!

#StackBounty: #18.04 #ssh #xorg #docker #x11-forwarding X11 over remote Docker container stops working after resetting VPN

Bounty: 50

I’m running on ubuntu 18.04, with a “operator” machine VPN’ed into a network that has the “Test” machine.

The test machine runs a Docker container that has a GUI application that I forward over X11. I use SSH on the operator machine to SSH into the test machine. Then, I run the Docker container. Normally, the GUI shows up on my “operator” machine when I run the Docker container.

However, I’ve noticed that as soon as I disconnect my VPN and reconnect,I get a “Could not connect to display $DISPLAY” error.

I would need to create a new container from the image to have the X11 working again.

Here is how I launch it:

SOCK=/tmp/.X11-unix
XAUTH=/tmp/.docker.xauth
xauth nlist $DISPLAY | sed -e 's/^..../ffff/' | xauth -f $XAUTH nmerge -
chmod 777 $XAUTH
sudo docker run  --gpus all -it --privileged --network=host -e DISPLAY=$DISPLAY -v $XSOCK:$XSOCK -v $XAUTH:$XAUTH -e XAUTHORITY=$XAUTH  image

Nothing changed on the “test” machine or docker container in this situation. Even if the VPN reconnect assigned me the same IP as prior, it would still have this issue.


Get this bounty!!!

#StackBounty: #boot #drivers #nvidia #xorg #wayland How to activate Wayland before arriving at login screen?

Bounty: 100

I have a boot video problem Ubuntu 18.04-4 installs in safe graphics mode only, but boots to purple screen where mouse works (NVIDIA) that can be solved by putting the video card in the first slot, arriving at the login screen, changing from xorg to wayland (xorg does not work), then putting the GPU back at slot 2.

Instead of moving my GPU everytime I reinstall my Ubuntu, how can I enable Wayland without arriving at the login screen? I don’t arrive at this screen if I use the GPU in the 2nd slot.


Get this bounty!!!

#StackBounty: #terminal #fonts #xorg #unicode #true-type-fonts Different line spacing depending on font encoding

Bounty: 50

Historically, I’ve been using Monotype Courier New as my XTerm font (it displays well both Cyrillic characters and box-drawing characters). When migrating from legacy KOI8-R to UTF-8 locale, I’ve noticed a change in the inter-line spacing (or leading), which made Courier New no longer suitable for terminals.

Here’s how my XTerm used to look in ru_RU.KOI8-R locale (12pt Courier New):

enter image description here

Here’s how it looks now (ru_RU.UTF-8, 12pt again, character width is the same, but note the increased leading, vertical box-drawing charaters form a double “dashed” line):

enter image description here

Apparently, XTerm is not the only application affected. Basically, all programs using Core X11 Fonts exhibit the same behaviour. Take a look at xfontsel displaying different encodings of the same font:

KOI8-R:

enter image description here

ISO 10646-1 (which stands for Unicode):

enter image description here

Interestingly, Courier New seems to be the only monospaced font affected. Other fonts, like -misc-freemono-*, have the same leading regardless of the encoding.

Now, the question:

Is there any way I can enforce the particular leading for a particular font? Should I look into the FreeType library? Or Xft? Any ideas are appreciated.


Get this bounty!!!

#StackBounty: #linux #keyboard #xorg Remap LeftAlt+J to ArrowDown

Bounty: 100

On a bare bones Linux installation (using xkb / xmodmap) is it possible to remap the LeftAlt+J to send a ArrowDown? I have found all sort of workarounds or slightly different scenarios with CapsLock+J and so on but none using exactly LeftAlt+J.

Ideally this should work without having to change LeftAlt for AltGr and LeftAlt will still keep it’s normal behavior. Eg: LeftAlt+F still opens the File menu in apps that have it.

Also, I mentioned xkb and xmodmap because I would like this to work everywhere instead of just remaps in certain apps.

One of the minimum requirements for this to be an accepted solution is, one should be able to open Firefox, focus the address bar, type something and with LeftAlt+J to navigate down the suggestions.

In my tests tools like xdotool and xte didn’t work well, but I’m happy to be corrected.

System: Linux with BSPWM, so no Gnome or KDE workarounds.

PS: I’m not looking for a spoon-fed solution but if someone can point me in the right direction, I will do the rest myself.

Thank you!


Get this bounty!!!

#StackBounty: #xorg #touchpad Set resistance to Two Finger Scrolling

Bounty: 50

I’m still using the synaptics driver because of the lack of configurability in libinput (the speed of the pointer is bad for small and large distances, there is no coasting, I need to limit the size of the touchpad or my laptop will recognize areas outside, and I’m used to setting the scrolling speed too).

So, the synaptics driver being so configurable, I would like to know if it’s possible to have some sort of resistance before Two Finger Scrolling begins in the synaptics driver. For this I have two reasons:

  1. When I use two finger tap as a right-click, many times it scrolls first and then opens the menu. If I right click a link in the web browser, the link escapes from underneath the cursor first, and I have to try again. I have tried setting the right-click to three fingers, but then I have to move the hand a lot more.
  2. The lack of some resistance makes horizontal scrolling very annoying: it should not begin scrolling horizontally until the user is explicitly moving the fingers sideways. That is the default in Windows, I don’t remember how it works in libinput.


Get this bounty!!!

#StackBounty: #xorg #touchpad #libinput Set resistance to Two Finger Scrolling

Bounty: 50

I’m still using the synaptics driver because of the lack of configurability in libinput (the speed of the pointer is bad for small and large distances, there is no coasting, I need to limit the size of the touchpad or my laptop will recognize areas outside, and I’m used to setting the scrolling speed too).

So, the synaptics driver being so configurable, I would like to know if it’s possible to have some sort of resistance before Two Finger Scrolling begins in the synaptics driver. For this I have two reasons:

  1. When I use two finger tap as a right-click, many times it scrolls first and then opens the menu. If I right click a link in the web browser, the link escapes from underneath the cursor first, and I have to try again. I have tried setting the right-click to three fingers, but then I have to move the hand a lot more.
  2. The lack of some resistance makes horizontal scrolling very annoying: it should not begin scrolling horizontally until the user is explicitly moving the fingers sideways. That is the default in Windows, I don’t remember how it works in libinput.


Get this bounty!!!

#StackBounty: #linux #multiple-monitors #xorg #xrandr xrandr: aligned bottoms of monitors with rotation => mouse / artifact issues

Bounty: 100

I used the solutions in this question to align my two monitors. My left monitor is High DPI 3840×2160 while the right is scaled 2x from 2560×1440 and rotated to portrait mode; they share a bottom edge.

xrandr --output eDP1 --scale 1x1 --mode 3840x2160 --pos 0x2800 --primary --output DP1 --scale 2x2 --mode 2560x1440 --pos 3840x0 --rotate left

This works but has the following undesirable characteristics:

  1. There’s some artifacts when dragging windows — they leave a black trail over the background.
  2. The mouse pointer disappears when it gets within around 100px of the right edge of the right monitor. It still functions as a pointer but I cannot see it.

Any ideas for how I can correct these issues?

Graphics card is reported as “Intel Corporation HD Graphics 630”


Get this bounty!!!

#StackBounty: #xorg #lightdm #chrome-book #arch-arm Lightdm Webkit2 greeter not loading – 'Error running javascript'?

Bounty: 50

I am running Arch Linux ARM on an armv7 Chromebook C201. I am currently experiencing a problem with the lightdm-webkit2-greeter not loading up properly on boot. I have checked /etc/lightdm/lightdm.conf and /etc/lightdm/lightdm-webkit2-greeter.conf and, as far as I can see, everything seems to be in order. The log file /var/log/lightdm/seat0-greeter.log shows the following:

libEGL warning: DRI2: failed to authenticate

** (lightdm-webkit2-greeter:432): WARNING **: 09:35:34.098: Error running javascript: 

I have tried re-installing lightdm and the webkit2 greeter, but it hasn’t helped. Does anyone know what might be causing this error?

It was working before, but something must have changed in the system to have caused this.


Get this bounty!!!

#StackBounty: #xorg #lightdm #chrome-book #arch-arm Lightdm Webkit2 greeter not loading – 'Error running javascript'?

Bounty: 50

I am running Arch Linux ARM on an armv7 Chromebook C201. I am currently experiencing a problem with the lightdm-webkit2-greeter not loading up properly on boot. I have checked /etc/lightdm/lightdm.conf and /etc/lightdm/lightdm-webkit2-greeter.conf and, as far as I can see, everything seems to be in order. The log file /var/log/lightdm/seat0-greeter.log shows the following:

libEGL warning: DRI2: failed to authenticate

** (lightdm-webkit2-greeter:432): WARNING **: 09:35:34.098: Error running javascript: 

I have tried re-installing lightdm and the webkit2 greeter, but it hasn’t helped. Does anyone know what might be causing this error?

It was working before, but something must have changed in the system to have caused this.


Get this bounty!!!