Electron 8: Linux unit tests are not starting

[bpasero]

It seems evident that we have issues running our tests with Electron 8.

E.g. https://monacotools.visualstudio.com/DefaultCollection/Monaco/_build/results?buildId=71995&view=results

[bpasero]

Pushed changes that seem to help macOS, but Linux is still entirely broken and not even starting the tests:

.build/electron/code-exploration: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by .build/electron/code-exploration)
/__w/1/s/.build/electron/code-exploration: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /__w/1/s/.build/electron/code-exploration)
[1444:0420/074319.126225:WARNING:discardable_shared_memory_manager.cc(199)] Less than 64MB of free space in temporary directory for shared memory files: 63
Fontconfig warning: "/etc/fonts/fonts.conf", line 86: unknown element "blank"
Xlib:  extension "RANDR" missing on display ":10".
[1444:0420/074319.323355:ERROR:bus.cc(393)] Failed to connect to the bus: Failed to connect to socket /var/run/dbus/system_bus_socket: No such file or directory
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
(electron) The default value of app.allowRendererProcessReuse is deprecated, it is currently "false".  It will change to be "true" in Electron 9.  For more information please check https://github.com/electron/electron/issues/18397
[1444:0420/074319.348232:ERROR:bus.cc(393)] Failed to connect to the bus: Could not parse server address: Unknown address type (examples of valid types are "tcp" and on UNIX "unix")
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[1467:0420/074319.385360:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x0, 0x0, 0x0.  Please report this to https://crbug.com/1025266
[1444:0420/074319.725571:WARNING:gpu_process_host.cc(1166)] The GPU process has crashed 1 time(s)
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[1444:0420/074319.772832:WARNING:gpu_process_host.cc(935)] Reinitialized the GPU process after a crash. The reported initialization time was 27 ms
[1495:0420/074319.778036:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x0, 0x0, 0x0.  Please report this to https://crbug.com/1025266
[1444:0420/074319.914018:WARNING:gpu_process_host.cc(1166)] The GPU process has crashed 2 time(s)
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[1444:0420/074319.949666:WARNING:gpu_process_host.cc(935)] Reinitialized the GPU process after a crash. The reported initialization time was 16 ms
[1505:0420/074319.954354:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x0, 0x0, 0x0.  Please report this to https://crbug.com/1025266
[1444:0420/074320.070486:WARNING:gpu_process_host.cc(1166)] The GPU process has crashed 3 time(s)
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[1512:0420/074320.092145:ERROR:sandbox_linux.cc(374)] InitializeSandbox() called with multiple threads in process gpu-process.
[1444:0420/074320.096272:WARNING:gpu_process_host.cc(935)] Reinitialized the GPU process after a crash. The reported initialization time was 0 ms
[1512:0420/074320.097028:ERROR:gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.
[1512:0420/074320.106618:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x0, 0x0, 0x0.  Please report this to https://crbug.com/1025266
[1444:0420/074320.225318:WARNING:gpu_process_host.cc(1166)] The GPU process has crashed 4 time(s)
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[1519:0420/074320.247360:ERROR:sandbox_linux.cc(374)] InitializeSandbox() called with multiple threads in process gpu-process.

[deepak1556]

@bpasero this is the reason why the gpu process is crashing

Less than 64MB of free space in temporary directory for shared memory files: 63
gpu_channel_manager.cc(459)] ContextResult::kFatalFailure: Failed to create shared context for virtualization.

Can we free up some space in the vm ?

[bpasero]

@deepak1556 interesting, to be clear, is this disk space to free or more memory?

[deepak1556]

On linux /dev/shm needs to have atleast 64MB of space for shared memory to be used by chromium. We can either try increasing that memory in CI if possible (I would say 4GB/8GB is a good amount) or we can use --disable-dev-shm-usage as outlined here puppeteer/puppeteer#1834

[bpasero]

I don't think that command line argument helps, I can reproduce now running the docker image:

root@b2d7abbdd0f0:~# DISPLAY=:10 ./electron --no-sandbox --disable-dev-shm-usage --disable-gpu
./electron: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by ./electron)
/root/electron: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /root/electron)
Fontconfig warning: "/etc/fonts/fonts.conf", line 86: unknown element "blank"

Electron 8.2.3 - Build cross platform desktop apps with JavaScript, HTML, and CSS
Usage: electron [options] [path]

A path to an Electron app may be specified. It must be one of the following:
  - index.js file.
  - Folder containing a package.json file.
  - Folder containing an index.js file.
  - .html/.htm file.
  - http://, https://, or file:// URL.

Options:
  -i, --interactive     Open a REPL to the main process.
  -r, --require         Module to preload (option can be repeated).
  -v, --version         Print the version.
  -a, --abi             Print the Node ABI version.
Xlib:  extension "RANDR" missing on display ":10".
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[2984:0420/120330.937759:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x0, 0x0, 0x0.  Please report this to https://crbug.com/1025266
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[3007:0420/120330.981019:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x0, 0x0, 0x0.  Please report this to https://crbug.com/1025266
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[3025:0420/120331.023597:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x0, 0x0, 0x0.  Please report this to https://crbug.com/1025266
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[3034:0420/120331.059460:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x0, 0x0, 0x0.  Please report this to https://crbug.com/1025266
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[3041:0420/120331.094133:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x0, 0x0, 0x0.  Please report this to https://crbug.com/1025266
/proc/self/exe: /lib/x86_64-linux-gnu/libdbus-1.so.3: no version information available (required by /proc/self/exe)
[3048:0420/120331.127820:FATAL:x11_util.cc(1312)] Unsupported visual with rgb mask 0x0, 0x0, 0x0.  Please report this to https://crbug.com/1025266
[2963:0420/120331.131368:FATAL:gpu_data_manager_impl_private.cc(1034)] The display compositor is frequently crashing. Goodbye.
Trace/breakpoint trap
root@b2d7abbdd0f0:~# 

Any clues how to make changes to the shared memory thingy to try?

[deepak1556]

The command line did work, gpu process no longer crashes. The fatal error that is remaining in the output is wrt initializing headless display. Some dependency changed in 8 that our docker container is missing, will look into it tomorrow.

[deepak1556]

Okay its an upstream bug fixed by https://bugs.chromium.org/p/chromium/issues/detail?id=1025266#c19 , will backport the fix in electron.

[bpasero]

Oh ok great, let me know when I can try again.