Chromium

OS should be linux. I don't know if windows is affected too?

Thanks for pointing out. I missed the OS thing but the title says Linux.

I don't know how to edit the initial submission. Perhaps someone with more rights could do it.

[Empty comment from Monorail migration]

Tried testing the issue on reported chrome version #77.0.3865.120 using Linux 16.04 as per https://crbug.com/chromium/1016383#c0 added xorg.conf file and launched chrome but didn't observe distorted UI.

Adding gayane@chromium.org to cc list for further inputs.

Thanks..

[Monorail components: -UI Internals>GPU UI>Browser]

"but didn't observe distorted UI"
Can you confirm from the xorg.log that 30 bit color mode was used?

I really don't believe the outcome described  in https://crbug.com/chromium/1016383#c4 . You either have an incompatible monitor or an incompatible video card.

In addition to xorg.log  you could run   "xdpyinfo". and look at what is printed at depth. You should see 30 planes.

you should see something like:
====================================================================
depths (8):    30, 1, 4, 8, 15, 16, 24, 32
  root window id:    0x253
  depth of root window:    30 planes
  number of colormaps:    minimum 1, maximum 1
  default colormap:    0x20
  default number of colormap cells:    1024
  preallocated pixels:    black 0, white 1073741823
  options:    backing-store WHEN MAPPED, save-unders NO
  largest cursor:    256x256
  current input event mask:    0xfa8033
    KeyPressMask             KeyReleaseMask           EnterWindowMask          
    LeaveWindowMask          ExposureMask             StructureNotifyMask      
    SubstructureNotifyMask   SubstructureRedirectMask FocusChangeMask          
    PropertyChangeMask       ColormapChangeMask      
  number of visuals:    180
  default visual id:  0x21
  visual:
    visual id:    0x21
    class:    TrueColor
    depth:    30 planes
    available colormap entries:    1024 per subfield
    red, green, blue masks:    0x3ff, 0xffc00, 0x3ff00000
    significant bits in color specification:    11 bits
============================================================

[Empty comment from Monorail migration]

Observed depth of root window as 24 planes, this issue probably requires specific monitor and video card hence adding Hardware dependency label and requesting respective team for help in further triaging.

Adding thomasanderson@chromium.org to cc list for further help.

Thanks..

@ santhoshkumar@chromium.org  "this issue probably requires specific monitor and video card"  

Not "probably", it is "certainly". You need proper hardware monitor, video card and driver. You will get 30 bit colors only on a Display port. It was all documented in the filing of the bug, but it was probably overlooked.  I should have put it in bold and itemized. Thanks for acknowledging the bug. Can you set it as confirmed now?

To add to the urgency of this request, there are a number of projects that depend on Chromium:

 -Chrome browser,
 -Brave browser,
 -Opera browser,
 -Electron  
 just to name a few.

I don't know what priority 2 means, but I, and many others have been using only Firefox for the better part of the year, which works perfectly in the 30 bit color depth setting.  

There must be a way to get the right hardware and test?  If not is there a way to try to help?

Here a nice and useful read bout 30 bit color depth https://www.nvidia.com/docs/IO/40049/TB-04701-001_v02_new.pdf

->thomasanderson to triage

Able to repro.  Not sure when we can get to this though.

Thanks.

[Empty comment from Monorail migration]

[Empty comment from Monorail migration]

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c5c97769c8d7e6a4ce3edae9263b327709bad5b0

commit c5c97769c8d7e6a4ce3edae9263b327709bad5b0
Author: Tom Anderson <thomasanderson@chromium.org>
Date: Tue Nov 05 23:50:10 2019

Fix 30bpp rendering on Linux/X11

This CL adds support for some additional X11 color types: RGB565, ARGB4444,
RGBA1010102. Also avoid unnecessarily rendering into an RGBA8888 buffer and
converting to the window format; just render into the correct format directly.
Also simplify/refactor some code.

BUG=1016383
R=rjkroege

Change-Id: I08051526b81d364a236c37654fda1be6bad1bc20
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1899324
Reviewed-by: Robert Kroeger <rjkroege@chromium.org>
Commit-Queue: Thomas Anderson <thomasanderson@chromium.org>
Cr-Commit-Position: refs/heads/master@{#712812}

[modify] https://crrev.com/c5c97769c8d7e6a4ce3edae9263b327709bad5b0/components/viz/service/display_embedder/software_output_device_x11.cc
[modify] https://crrev.com/c5c97769c8d7e6a4ce3edae9263b327709bad5b0/ui/base/x/x11_shm_image_pool_base.cc
[modify] https://crrev.com/c5c97769c8d7e6a4ce3edae9263b327709bad5b0/ui/base/x/x11_software_bitmap_presenter.cc
[modify] https://crrev.com/c5c97769c8d7e6a4ce3edae9263b327709bad5b0/ui/base/x/x11_software_bitmap_presenter.h
[modify] https://crrev.com/c5c97769c8d7e6a4ce3edae9263b327709bad5b0/ui/base/x/x11_util.cc
[modify] https://crrev.com/c5c97769c8d7e6a4ce3edae9263b327709bad5b0/ui/base/x/x11_util.h
[modify] https://crrev.com/c5c97769c8d7e6a4ce3edae9263b327709bad5b0/ui/ozone/platform/x11/x11_canvas_surface.cc
[modify] https://crrev.com/c5c97769c8d7e6a4ce3edae9263b327709bad5b0/ui/ozone/platform/x11/x11_canvas_surface.h

The fix should appear in Chrome 80 and later.

Thank you!

I hardly can wait to test it. I hope that the other projects depending on chromium such as Electron and other browsers will follow suit.

Is there any build I can test or I have to check out  from GIT and build it myself?

Thanks.

the change should roll out to the dev channel within a week: https://www.google.com/chrome/dev/

I can confirm that the colours are good in chrome-unstable 80.xx on 30 bit color monitor. Thank you very much.

[Empty comment from Monorail migration]

Just got chrome version  80.0.3987.87 and colors are sill wrong.

@xhejt ...

What OS  and if it is 32 or 64 bits and the  window manager you are usin?  
Can you attach here a print screen with the offending app with the About page and superimposed a window that show the color depth, similar to what I am attaching below?  
Also the make of the monitor and what type of connector and video card you have and the video driver used. Thanks
 
As you can ssee in my picture I have 30  

Linux, ubuntu 19.10, xfce 4.14.1, 64bits. Actually, it does not start:

$ google-chrome
[672:672:0207/202531.383966:ERROR:gl_surface_glx.cc(129)] Failed to get GLXConfig
[672:672:0207/202531.384111:ERROR:gl_surface_glx.cc(478)] CreateDummyWindow(gfx::GetXDisplay()) failed
[672:672:0207/202531.384162:ERROR:gl_initializer_x11.cc(148)] GLSurfaceGLX::InitializeOneOff failed.
[672:672:0207/202531.387365:ERROR:viz_main_impl.cc(161)] Exiting GPU process due to errors during initialization
[704:735:0207/202531.544362:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[743:1:0207/202531.636928:ERROR:command_buffer_proxy_impl.cc(125)] ContextResult::kTransientFailure: Failed to send GpuChannelMsg_CreateCommandBuffer.
[784:797:0207/202531.710333:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[706:1:0207/202531.780901:ERROR:child_process_sandbox_support_impl_linux.cc(79)] FontService unique font name matching request did not receive a response.
[706:1:0207/202531.781460:ERROR:child_process_sandbox_support_impl_linux.cc(79)] FontService unique font name matching request did not receive a response.
[801:808:0207/202531.887480:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[811:811:0207/202531.995885:ERROR:gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.
[811:817:0207/202532.080661:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[820:820:0207/202532.221712:ERROR:gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.
[820:827:0207/202532.246179:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[831:831:0207/202532.346455:ERROR:gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.
[831:839:0207/202532.430226:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[636:653:0207/202532.536021:FATAL:gpu_data_manager_impl_private.cc(1034)] The display compositor is frequently crashing. Goodbye.
Trace/breakpoint trap (core dumped)

screen #0:
  dimensions:    3840x2160 pixels (1016x571 millimeters)
  resolution:    96x96 dots per inch
  depths (8):    30, 1, 4, 8, 15, 16, 24, 32
  root window id:    0x11d
  depth of root window:    30 planes
  number of colormaps:    minimum 1, maximum 1
  default colormap:    0x22
  default number of colormap cells:    1024
  preallocated pixels:    black 0, white 1073741823
  options:    backing-store WHEN MAPPED, save-unders NO
  largest cursor:    256x256
  current input event mask:    0x7a003f
    KeyPressMask             KeyReleaseMask           ButtonPressMask
    ButtonReleaseMask        EnterWindowMask          LeaveWindowMask
    StructureNotifyMask      SubstructureNotifyMask   SubstructureRedirectMask
    FocusChangeMask          PropertyChangeMask
  number of visuals:    12
  default visual id:  0x20
  visual:
    visual id:    0x20
    class:    TrueColor
    depth:    30 planes
    available colormap entries:    1024 per subfield
    red, green, blue masks:    0x3ff00000, 0xffc00, 0x3ff
    significant bits in color specification:    10 bits
  visual:
    visual id:    0x21
    class:    DirectColor
    depth:    30 planes
    available colormap entries:    1024 per subfield
    red, green, blue masks:    0x3ff00000, 0xffc00, 0x3ff
    significant bits in color specification:    10 bits


however, there is no such glxfbconfig:
126 GLXFBConfigs:
    visual  x   bf lv rg d st  colorbuffer  sr ax dp st accumbuffer  ms  cav
  id dep cl sp  sz l  ci b ro  r  g  b  a F gb bf th cl  r  g  b  a ns b eat
----------------------------------------------------------------------------
0x095  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0  0 0 None
0x096  0 tc  0  16  0 r  . .   5  6  5  0 .  .  0  0  0  0  0  0  0  0 0 None
0x097  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  0 0 None
0x098  0 tc  0  16  0 r  . .   5  6  5  0 .  .  0 16  0  0  0  0  0  0 0 None
0x099  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0 24  8  0  0  0  0  0 0 None
0x09a  0 tc  0  16  0 r  . .   5  6  5  0 .  .  0 24  8  0  0  0  0  0 0 None
0x09b  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0  0 0 None
0x09c  0 tc  0  32  0 r  . .   8  8  8  8 .  .  0  0  0  0  0  0  0  0 0 None
0x09d  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x09e  0 tc  0  32  0 r  . .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x09f  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0  0 0 None
0x0a0  0 tc  0  24  0 r  . .   8  8  8  0 .  .  0  0  0  0  0  0  0  0 0 None
0x0a1  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0a2  0 tc  0  24  0 r  . .   8  8  8  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0a3  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0  0 0 None
0x0a4  0 tc  0  32  0 r  . .   8  8  8  8 .  s  0  0  0  0  0  0  0  0 0 None
0x0a5  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None
0x0a6  0 tc  0  32  0 r  . .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None
0x0a7  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  0 0 None
0x0a8  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0 16 16 16  0  0 0 Slow
0x0a9  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x0aa  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8 16 16 16 16  0 0 Slow
0x0ab  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0ac  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8 16 16 16  0  0 0 Slow
0x0ad  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None
0x0ae  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8 16 16 16 16  0 0 Slow
0x0af  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0  2 1 None
0x0b0  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0  4 1 None
0x0b1  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0  8 1 None
0x0b2  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0 16 1 None
0x0b3  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  2 1 None
0x0b4  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  4 1 None
0x0b5  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  8 1 None
0x0b6  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0 16 1 None
0x0b7  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0  2 1 None
0x0b8  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0  4 1 None
0x0b9  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0  8 1 None
0x0ba  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0 16 1 None
0x0bb  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  2 1 None
0x0bc  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  4 1 None
0x0bd  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  8 1 None
0x0be  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0 16 1 None
0x0bf  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0  2 1 None
0x0c0  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0  4 1 None
0x0c1  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0  8 1 None
0x0c2  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0 16 1 None
0x0c3  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  2 1 None
0x0c4  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  4 1 None
0x0c5  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  8 1 None
0x0c6  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0 16 1 None
0x0c7  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0  2 1 None
0x0c8  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0  4 1 None
0x0c9  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0  8 1 None
0x0ca  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0 16 1 None
0x0cb  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  2 1 None
0x0cc  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  4 1 None
0x0cd  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  8 1 None
0x0ce  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0 16 1 None
0x0cf  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0  0 0 None
0x0d0  0 dc  0  16  0 r  . .   5  6  5  0 .  .  0  0  0  0  0  0  0  0 0 None
0x0d1  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  0 0 None
0x0d2  0 dc  0  16  0 r  . .   5  6  5  0 .  .  0 16  0  0  0  0  0  0 0 None
0x0d3  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0d4  0 dc  0  16  0 r  . .   5  6  5  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0d5  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0  0 0 None
0x0d6  0 dc  0  32  0 r  . .   8  8  8  8 .  .  0  0  0  0  0  0  0  0 0 None
0x0d7  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x0d8  0 dc  0  32  0 r  . .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x0d9  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0  0 0 None
0x0da  0 dc  0  24  0 r  . .   8  8  8  0 .  .  0  0  0  0  0  0  0  0 0 None
0x0db  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0dc  0 dc  0  24  0 r  . .   8  8  8  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0dd  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0  0 0 None
0x0de  0 dc  0  32  0 r  . .   8  8  8  8 .  s  0  0  0  0  0  0  0  0 0 None
0x0df  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None
0x0e0  0 dc  0  32  0 r  . .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None
0x0e1  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  0 0 None
0x0e2  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0 16 16 16  0  0 0 Slow
0x0e3  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x0e4  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8 16 16 16 16  0 0 Slow
0x0e5  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0e6  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8 16 16 16  0  0 0 Slow
0x0e7  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None
0x0e8  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8 16 16 16 16  0 0 Slow
0x0e9  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0  2 1 None
0x0ea  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0  4 1 None
0x0eb  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0  8 1 None
0x0ec  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0 16 1 None
0x0ed  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  2 1 None
0x0ee  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  4 1 None
0x0ef  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  8 1 None
0x0f0  0 dc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0 16 1 None
0x0f1  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0  2 1 None
0x0f2  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0  4 1 None
0x0f3  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0  8 1 None
0x0f4  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0 16 1 None
0x0f5  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  2 1 None
0x0f6  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  4 1 None
0x0f7  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  8 1 None
0x0f8  0 dc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0 16 1 None
0x0f9  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0  2 1 None
0x0fa  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0  4 1 None
0x0fb  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0  8 1 None
0x0fc  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0 16 1 None
0x0fd  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  2 1 None
0x0fe  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  4 1 None
0x0ff  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  8 1 None
0x100  0 dc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0 16 1 None
0x101  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0  2 1 None
0x102  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0  4 1 None
0x103  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0  8 1 None
0x104  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0 16 1 None
0x105  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  2 1 None
0x106  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  4 1 None
0x107  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  8 1 None
0x108  0 dc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0 16 1 None
0x109 32 tc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0  0 0 None
0x10a 32 tc  0  32  0 r  . .   8  8  8  8 .  .  0  0  0  0  0  0  0  0 0 None
0x10b 32 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x10c 32 tc  0  32  0 r  . .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x10d 32 tc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0  0 0 None
0x10e 32 tc  0  32  0 r  . .   8  8  8  8 .  s  0  0  0  0  0  0  0  0 0 None
0x10f 32 tc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None
0x110 32 tc  0  32  0 r  . .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None
0x111 32 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x112 32 tc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None



it is lenovo carbon x1, gen 7, local panel, 4k version offers hdr/10bit display, intel driver (intel_drv, not modeset)

@xhejt ... Try a supported window manager such as Plasma5/kde or Gnome. I didn't try xfce and I don't know what kind of support it offers for wide gamut.

From your dump it appears that you don't have the OpenGL running properly on your platform. Fix that first (test with glxinfo & glxgears). Your problem is not a chromium problem. We all run on 30 bit color depth and is fine as shown in the attachment.

You did not do what you were asked for: provide a proof of color depth and a screen of the browser.    "actually does not start" is not what you said before "Just got chrome version  80.0.3987.87 and colors are sill wrong."  This is misleading just to use an euphemism and actually not nice.

Ok, I posted too soon, but still your problem is your configuration and this is not related to the browser.
NVIDIA drivers are ok, I don't know much about the Intel drivers. Until some time ago only proprietary NVIDIA drivers were providing proper wide gamut on linux.

glxgears and glxinfo work just fine even on 30bits. xfce desktop works ok as well.

does glxinfo on nvidia return 10bit glxfbconfig? This could be the issue the browser does not handle well - 30bit display with 24bit only glx visuals..

btw, the browser starts for a second before it crashes and it displays a *pink* window anyway, therefore I reported colors are still wrong. Not sure if it is usefull to capture it.

I am not at my computer to test but you could try to disable hardware acceleration from your browser and see what happens.
The "one second" pink window probably happens because it cannot render on your system.  Not useful to capture. I can assure you and you saw proof in my post that on NVIDIA platform works fine.

Try to disable hardware acceleration or the opposite to figure out what is wrong on your system.   Does Firefox work?  With hardware acceleration on?

/etc/drirc:

 <option name="allow_rgb10_configs" value="true" />

it exposed 10bit configs and chrome runs ok now and colors are ok. However, that pink screen flashes anyway while chrome is starting..

So it was the hardware acceleration issue after all. I did not
experience the issue because I believe that the proprietary NVIDIA driver
does not use the freedesktop.org DRI configuration.

The pink blip is most likely caused by a reverse order in the rendering
code that enables rendering before setting the 30 bit configuration in
place.
ᐧ

However, it crashes from time to time:

[8874:8874:0207/214734.761089:ERROR:gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.
[8874:8880:0207/214734.794008:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[8884:8884:0207/214734.905588:ERROR:gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.
[8884:8890:0207/214734.934464:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[8893:8893:0207/214735.053487:ERROR:gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.
[8893:8899:0207/214735.086722:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[7827:7844:0207/214735.168979:FATAL:gpu_data_manager_impl_private.cc(1034)] The display compositor is frequently crashing. Goodbye.
Trace/breakpoint trap (core dumped)


which is a bit strange as now, 10bit configs are supported:
66 GLX Visuals
    visual  x   bf lv rg d st  colorbuffer  sr ax dp st accumbuffer  ms  cav
  id dep cl sp  sz l  ci b ro  r  g  b  a F gb bf th cl  r  g  b  a ns b eat
----------------------------------------------------------------------------
0x020 30 tc  0  32  0 r  y .  10 10 10  2 .  .  0 24  8  0  0  0  0  0 0 None
0x021 30 dc  0  32  0 r  y .  10 10 10  2 .  .  0 24  8  0  0  0  0  0 0 None
0x14b 30 tc  0  32  0 r  y .  10 10 10  2 .  .  0  0  0  0  0  0  0  0 0 None
0x14c 30 tc  0  32  0 r  . .  10 10 10  2 .  .  0  0  0  0  0  0  0  0 0 None
0x14d 30 tc  0  32  0 r  . .  10 10 10  2 .  .  0 24  8  0  0  0  0  0 0 None
0x14e 30 tc  0  30  0 r  y .  10 10 10  0 .  .  0  0  0  0  0  0  0  0 0 None
0x14f 30 tc  0  30  0 r  . .  10 10 10  0 .  .  0  0  0  0  0  0  0  0 0 None
0x150 30 tc  0  30  0 r  y .  10 10 10  0 .  .  0 24  8  0  0  0  0  0 0 None
0x151 30 tc  0  30  0 r  . .  10 10 10  0 .  .  0 24  8  0  0  0  0  0 0 None

182 GLXFBConfigs:
    visual  x   bf lv rg d st  colorbuffer  sr ax dp st accumbuffer  ms  cav
  id dep cl sp  sz l  ci b ro  r  g  b  a F gb bf th cl  r  g  b  a ns b eat
----------------------------------------------------------------------------
0x095  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0  0  0  0  0  0  0  0 0 None
0x096  0 tc  0  16  0 r  . .   5  6  5  0 .  .  0  0  0  0  0  0  0  0 0 None
0x097  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0 16  0  0  0  0  0  0 0 None
0x098  0 tc  0  16  0 r  . .   5  6  5  0 .  .  0 16  0  0  0  0  0  0 0 None
0x099  0 tc  0  16  0 r  y .   5  6  5  0 .  .  0 24  8  0  0  0  0  0 0 None
0x09a  0 tc  0  16  0 r  . .   5  6  5  0 .  .  0 24  8  0  0  0  0  0 0 None
0x09b  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0  0  0  0  0  0  0  0 0 None
0x09c  0 tc  0  32  0 r  . .   8  8  8  8 .  .  0  0  0  0  0  0  0  0 0 None
0x09d  0 tc  0  32  0 r  y .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x09e  0 tc  0  32  0 r  . .   8  8  8  8 .  .  0 24  8  0  0  0  0  0 0 None
0x09f  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0  0  0  0  0  0  0  0 0 None
0x0a0  0 tc  0  24  0 r  . .   8  8  8  0 .  .  0  0  0  0  0  0  0  0 0 None
0x0a1  0 tc  0  24  0 r  y .   8  8  8  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0a2  0 tc  0  24  0 r  . .   8  8  8  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0a3  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0  0  0  0  0  0  0  0 0 None
0x0a4  0 tc  0  32  0 r  . .   8  8  8  8 .  s  0  0  0  0  0  0  0  0 0 None
0x0a5  0 tc  0  32  0 r  y .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None
0x0a6  0 tc  0  32  0 r  . .   8  8  8  8 .  s  0 24  8  0  0  0  0  0 0 None
0x0a7 30 tc  0  32  0 r  y .  10 10 10  2 .  .  0  0  0  0  0  0  0  0 0 None
0x0a8 30 tc  0  32  0 r  . .  10 10 10  2 .  .  0  0  0  0  0  0  0  0 0 None
0x0a9 30 tc  0  32  0 r  y .  10 10 10  2 .  .  0 24  8  0  0  0  0  0 0 None
0x0aa 30 tc  0  32  0 r  . .  10 10 10  2 .  .  0 24  8  0  0  0  0  0 0 None
0x0ab 30 tc  0  30  0 r  y .  10 10 10  0 .  .  0  0  0  0  0  0  0  0 0 None
0x0ac 30 tc  0  30  0 r  . .  10 10 10  0 .  .  0  0  0  0  0  0  0  0 0 None
0x0ad 30 tc  0  30  0 r  y .  10 10 10  0 .  .  0 24  8  0  0  0  0  0 0 None
0x0ae 30 tc  0  30  0 r  . .  10 10 10  0 .  .  0 24  8  0  0  0  0  0 0 None


and also, it seems that message is not so deadly as I can see it from time to time. The last time, the browser crashed after about 20 minutes.

It never crashes on NVIDIA driver. I am almost convinced that it has to do with the Intel driver. Try to disable HW acceleration and see your mileage.
Another option is to install Plasma5, their compositor works well.

disabled HW accel, see attached screenshot and it won't start at all:
google-chrome
[12119:12134:0207/220850.904945:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[12164:1:0207/220850.996277:ERROR:command_buffer_proxy_impl.cc(125)] ContextResult::kTransientFailure: Failed to send GpuChannelMsg_CreateCommandBuffer.
[12145:1:0207/220851.076207:ERROR:child_process_sandbox_support_impl_linux.cc(79)] FontService unique font name matching request did not receive a response.
[12145:1:0207/220851.078672:ERROR:child_process_sandbox_support_impl_linux.cc(79)] FontService unique font name matching request did not receive a response.
[12200:12210:0207/220851.098785:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[12214:12221:0207/220851.259583:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[12224:12224:0207/220851.363225:ERROR:gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.
[12224:12230:0207/220851.448850:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[12233:12233:0207/220851.576910:ERROR:gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.
[12233:12239:0207/220851.601109:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[12242:12242:0207/220851.701193:ERROR:gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.
[12242:12248:0207/220851.782111:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
[12089:12106:0207/220851.889746:FATAL:gpu_data_manager_impl_private.cc(1034)] The display compositor is frequently crashing. Goodbye.
Trace/breakpoint trap (core dumped)


tried several times to start, no luck.

disabled hw accel is ok if I turn off 30bits.

I assume that the video driver has rendering issues with 30bits color depth.  The other thing you can try is to enable back  HW and try another compositor. Plasma 5's kwin is quite good.  

Have you tried outer browsers such as Firefox, which are not chromium based?

Firefox is just ok. It seems that chrome is the only app that has serious problem (crash). openbox has color issues but except that, I do not see any other crash or something.

It is hard to tell, as they may be configured differently.  Try another compositor and then one can dig further if it is a problem on the chromium end. Also, you have an opensource driver that has not been tested with wide gamut.  On NVIDIA driver there are no issues whatsoever so we cannot blame it on Chrome.

Just tried xfwm4 with compositor, the same result. Previous attempts were without any compositor.

Xfwm4 is not plasma5/kwin

I do not have KDE, only xfce, sorry.

Perhaps chromium people are more willing to fix bugs.

"do not have KDE, only xfce, sorry.
Perhaps chromium people are more willing to fix bugs.

Sorry, Perhaps you are using unsupported software (intel driver/xfce) and are willing to try what is supported ...

As I do not use compositor, xfce is completely unrelated. And unsupported intel driver? This is just ridiculous as it is actually the most used HW.

And honestly, I do not care anymore, tried to help to improve chrome and got only *rude* answers.

"And unsupported intel driver? This is just ridiculous as it is actually the most used HW."  

* Not for performance video cards so far!
* Did you read anywhere that the intel video driver supports  30 bits color depth?  
* Or that xfce supports 10 bits/channel color depth?

If you have that information please forward it to the chromium developers as I did it for fixing the colors in Chromium.
Please stick to the facts, posting messages that say half truths without willing to really help with checking the trusted platforms.  
The problems are complex and can be in more than one place, while it appears that you are dead sure that the problem is in chromium. That the problem is in chromium could be true, but without testing the browser a with known and validated environments for wide gamut we cannot. know.

Regarding rude answers, please read the above and judge for yourself what stands out.:
"I do not have KDE, only xfce, sorry.

Perhaps chromium people are more willing to fix bugs.
"

The only acid answer was the answer to your from above where you push the load on others without willing to help further. That's rude
Have a nice day and hope for a better attitude.

As of intel driver:
https://www.phoronix.com/scan.php?page=news_item&px=Mesa-Lands-10-bit-Color

and I say again. Everything is working on wide gamut, only chrome crashes.

This is a good step, but it is not a proof. It is just an announcement that it will be implemented. Open source drivers are known for their  slow implementation and many bugs. There are just not enough people with linux and 30 it colour depth monitors.

It is entirely possible that chromium uses different API than other apps, therefore the problem has not been exposed by those.

My suggestion was to try KDE so that we eliminate one other possible weak link xfce. With that done only the driver or chromium can be at fault.

Given the fact I have absolutely no issues with Chrome/KDE//proprietary NVIDIA driver, I am inclined to think that the driver or xfce are at fault, but until we don't eliminate them one by one it is hard to tell.

The difference is in the mask layouts for your systems.  The one that works has rgb masks:
0x3ff, 0xffc00, 0x3ff00000
and the one that doesn't has:
0x3ff00000, 0xffc00, 0x3ff

So the issue is just RGB vs BGR

Any idea what could switch the masks? Any partilar reason why chrome should request BGR sometimes? Or the driver should support both RGB and BGR?

What is the full output of xdpyinfo?

Attached. There is only one RGB oder.

Yes it appears only BGR is supported on your system.  I've opened this bug with skia (Chromium's graphics library) to add support for this new colortype: https://bugs.chromium.org/p/skia/issues/detail?id=9893
We revisit this issue once that bug is fixed.

"Attached. There is only one RGB oder."

It appears that the file contains the order BGR, the one that is not supported. I'll test my output when I get  home, but if you look at https://crbug.com/chromium/1016383#c6 you can see that I have the order 0x3ff, 0xffc00, 0x3ff00000, the one that works.

So what is causing the order of the mask layouts on the system to be different?

Not sure why color order is different. Maybe more interesting question is, why does it sometimes works and sometimes does not? As in this case, the browser does not fail instantly, but only after some time of running. It looks like there are some cases where BGR is OK and some where it does not.

Anyway, thanks a lot for requesting support for this colortype.

btw, if I switch back to 24bit/8bit, I have the same pixel order:

screen #0:
  dimensions:    3840x2160 pixels (336x189 millimeters)
  resolution:    290x290 dots per inch
  depths (7):    24, 1, 4, 8, 15, 16, 32
  root window id:    0x1a1
  depth of root window:    24 planes
  number of colormaps:    minimum 1, maximum 1
  default colormap:    0x22
  default number of colormap cells:    256
  preallocated pixels:    black 0, white 16777215
  options:    backing-store WHEN MAPPED, save-unders NO
  largest cursor:    256x256
  current input event mask:    0xfa803f
    KeyPressMask             KeyReleaseMask           ButtonPressMask          
    ButtonReleaseMask        EnterWindowMask          LeaveWindowMask          
    ExposureMask             StructureNotifyMask      SubstructureNotifyMask  
    SubstructureRedirectMask FocusChangeMask          PropertyChangeMask      
    ColormapChangeMask      
  number of visuals:    94
  default visual id:  0x20
  visual:
    visual id:    0x20
    class:    TrueColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits

and this time, it is ok.

I was curios if the color order is given by the video driver or by the window manager. if it is the window manager there is a simple temporary fix it by changing it until the skia issue is fixed.

If I run: X :1.0 & export DISPLAY=:1.0; xterm and there xdpyinfo, I got the same BGR mask order. It has nothing to do with xfce or something.

The issue is likely caused by differences in drivers between Intel and Nvidia

@xhejt ...

"btw, if I switch back to 24bit/8bit, I have the same pixel order:
...
red, green, blue masks:    0xff0000, 0xff00, 0xff"

Your conclusion is not correct . If you pay some closer attention you'll see that the mask is different for 8 it/color. there is no "3" in the mask
for 10 bit /color you have either "0x3ff, 0xffc00, 0x3ff00000" which works now  or " 0x3ff00000, 0xffc00, 0x3ff" which does not work yet.

On my NVIDIA card I am also getting  0xff0000, 0xff00, 0xff mask for 8 bit/color.

Also your conclusion that "It has nothing to do with xfce or something" maybe is not correct because the only difference between our working setup and yours is the card/driver (which Thomas thinks is the issue) and the window manager. Changing cards is not easy but installing another proven window manager may be easier to test than you think ... or ...

No, the point is, that for both 24 and 30 bits, R component occupies most significant bits. It completely does not matter, whether the mask is actually with 3ff or ff, that's not the point.

On nvidia, on the other hand, 24bits uses BGR order while 30bits uses RGB order. On intel, both 24bits and 30bits use BGR order.

Moreover, the point with plain xorg + terminal (*WITHOUT* window manager) showed that noone is messing with masks, it is the native mask Xorg uses, BRG one, not any other.

I will wait for fix in skia.

And btw, the browser itself asks please report this issue: Unsupported visual with rgb mask 0x3ff00000, 0xffc00, 0x3ff.  Please report this to https://crbug.com/1025266
so I clearly do not see why I should not.

As you wish, it was just a suggestion for a possible temporary fix.  Definitely, there is a need for a fix.

3ff and ff are different masks because they have different color representation sizes.  However you missed the point, on my system the order is one for 30 bits and a different one 24 bits. What I wanted to say that your discovery that the order is the same for 8 bits/color has no relevance for the good color rendering in 24 bits color depth as it appears that other drivers have the same order.  

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/fd5829812cbaa1f7031ef4e0cf7257b86d631fe6

commit fd5829812cbaa1f7031ef4e0cf7257b86d631fe6
Author: Tom Anderson <thomasanderson@chromium.org>
Date: Wed Feb 12 20:47:25 2020

Support BGRA 30bpp visuals for X11 software rendering

At least with Intel drivers, deepcolor visuals may be of BGR format
isntead of RGB like on Nvidia.  Skia recently added support for
30bpp BGR formats [1], so use this new format to support this
configuration.

[1] https://skia.googlesource.com/skia/+/f7eb0544a8853eb087e20a0e5e7bc3597de02074

R=mtklein

Bug: 1016383
Change-Id: I22ffd649f7cfd7c6ed132250dee35c7960210fa6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2052206
Auto-Submit: Thomas Anderson <thomasanderson@chromium.org>
Commit-Queue: Scott Violet <sky@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#740804}

[modify] https://crrev.com/fd5829812cbaa1f7031ef4e0cf7257b86d631fe6/ui/base/x/x11_util.cc

[Empty comment from Monorail migration]

What version will have the fix?

Versions later than 82.0.4057.0

I tried version 83.0.4103.61, it does not crash. It has some issues however.

1) If I enable hardware acceleration, chrome is dark (it seems to be missing shift colors by 2 bits up), see attached screenshot.
2) when chrome creates new window, displays menu, it is always pink at first, kinda annoying as almost everything flashes pink for a bit.

And this is still not fixed.

Version 90.0.4430.72 (Official Build) (64-bit)

Still not fixed.

Version 101.0.4951.41 (Official Build) (64-bit)


can we expect any progress here?

It’s been 2-3 years… this issue still exists to this date.

Chromium, we urgently ask for this to get fixed.

[Comment Deleted]

I switched from NVIDIA to AMDGPU and the problem with grayed out/dark chromium colors on  10 bit/color display is still present as shown in https://crbug.com/chromium/1016383#c66

The annoying thing  is that all the Chromium based applications have the problem : Electron based apps such as Signal Desktop and Skype, Chrome browser, Brave Browser, etc.

No other application behaves like that in Linux 30  bit color displays. Firefox is showing good colors

What can we do have this fixed ?  Thanks

For 24 planes this are the color masks
default visual id:  0x45
  visual:
    visual id:    0x45
    class:    TrueColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits

These are different from what I was getting on NVIDIA and probably explains the issue

on the 30 bit color setting the colour masks are as follows:

visual:
    visual id:    0x21
    class:    TrueColor
    depth:    30 planes
    available colormap entries:    1024 per subfield
    red, green, blue masks:    0x3ff00000, 0xffc00, 0x3ff
    significant bits in color specification:    10 bits

The masks are in reverse order from what I had on the NVIDIA drivers

FYI, I reopened the issue under 1364850

[Comment Deleted]

10 bit is working on Chromium for me but not Chrome. I have Ubuntu 22.04 LTS and created an xorg.conf to get it working. For regular Chrome, I'm getting the same distorted UI.

This issue was migrated from crbug.com/chromium/1016383?no_tracker_redirect=1

[Multiple monorail components: Internals>GPU, UI>Browser]
[Monorail blocked-on: crbug.com/skia/9893]
[Monorail blocking: crbug.com/chromium/1021659]
[Monorail mergedwith: crbug.com/chromium/1001800, crbug.com/chromium/1038680]
[Monorail components added to Component Tags custom field.]