afrog
afrog copied to clipboard
zan8in
Atom crash everytime on new Fedora Worktation 35
Prerequisites
- [x] Put an X between the brackets on this line if you have done all of the following:
- Reproduced the problem in Safe Mode: https://flight-manual.atom.io/hacking-atom/sections/debugging/#using-safe-mode
- Followed all applicable steps in the debugging guide: https://flight-manual.atom.io/hacking-atom/sections/debugging/
- Checked the FAQs on the message board for common solutions: https://discuss.atom.io/c/faq
- Checked that your issue isn't already filed: https://github.com/issues?utf8=✓&q=is%3Aissue+user%3Aatom
- Checked that there is not already an Atom package that provides the described functionality: https://atom.io/packages
Description
atom crash on first launch and everytime
Steps to Reproduce
- On new install Fedora Workstation 35
- Localinstall of atom rpm
- atom .
- Crash
atom openning
but Crash
Reproduces how often:
everytime
Versions
atom --version
Atom : 1.58.0 Electron: 9.4.4 Chrome : 83.0.4103.122 Node : 12.14.1
apm --version
apm 2.6.2 npm 6.14.13 node 12.14.1 x64 atom 1.58.0 python 3.10.0 git 2.33.1
OS version
Fedora Workstation 35
Additional Information
- /usr/bin/atom: line 190: 19480 Illegal instruction (core dumped) nohup "$ATOM_PATH" --executed-from="$(pwd)" --pid=$$ "$@" > "$ATOM_HOME/nohup.out" 2>&1
- [19480:1113/224938.915770:FATAL:gpu_data_manager_impl_private.cc(439)] GPU process isn't usable. Goodbye.
Same for me on Fedora 35. I think that the issue is related to https://github.com/atom/atom/issues/23036 and https://github.com/atom/atom/issues/23167.
See in particular this comment.
As a temporary workaround, atom can be started with the --no-sandbox flag.
Edit: Atom seems to run fine also with the --in-process-gpu flag, without disabling the sandbox.
Same for me on Fedora 35. Same version, same problem, same error... Starting with --in-process-gpu looks good at first sight... Thanks @mattbit , at least I can still work this way until a fix has shipped...
Same for me on Fedora 35. Same version, same problem, same error... Starting with --in-process-gpu looks good at first sight... Thanks @mattbit , at least I can still work this way until a fix has shipped...
I was too quickly happy. Unfortunately Atom crashes after some times again with many electron deprecated errors.
Same happens to me on Fedora 35.
atom --in-process-gpu launches it fine.
UPDATE: not really:
[----@localhost ~]$atom ----in-process-gpu --enable-electron-logging [----@localhost ~]$ /usr/bin/atom: line 190: 734441 Illegal instruction (core dumped) nohup "$ATOM_PATH" --executed-from="$(pwd)" --pid=$$ "$@" > "$ATOM_HOME/nohup.out" 2>&1 Fontconfig warning: "/usr/share/fontconfig/conf.avail/05-reset-dirs-sample.conf", line 6: unknown element "reset-dirs" [734441:0104/120635.480704:INFO:CONSOLE(357)] "document.registerElement is deprecated and will be removed in M73, around March 2019. Please use window.customElements.define instead. See https://www.chromestatus.com/features/4642138092470272 for more details.", source: /usr/share/atom/resources/app.asar/src/text-editor-element.js (357) [734441:0104/120635.503447:WARNING:quota_manager.cc(1689)] No storage quota provided in QuotaSettings. [734441:0104/120635.742282:INFO:CONSOLE(1451)] "WebSocket connection to 'wss://ws-mt1.pusher.com/app/f119821248b7429bece3?protocol=7&client=js&version=4.4.0&flash=false' failed: WebSocket is closed before the connection is established.", source: /home/----/.atom/packages/teletype/node_modules/pusher-js/dist/web/pusher.js (1451) [734441:0104/120635.744700:INFO:CONSOLE(51)] "teletype: Using pusher key:", source: /home/----/.atom/packages/teletype/lib/teletype-package.js (51) [734441:0104/120635.744736:INFO:CONSOLE(52)] "teletype: Using base URL:", source: /home/----/.atom/packages/teletype/lib/teletype-package.js (52) [734441:0104/120635.746330:ERROR:simple_index_file_posix.cc(61)] Could not get file info for /home/----/.config/Atom/Cache/b0b780dc8652849f_0 [734441:0104/120635.881949:FATAL:gpu_data_manager_impl_private.cc(892)] The display compositor is frequently crashing. Goodbye. Failed to generate minidump.
[----@localhost ~]$ atom ----in-process-gpu [----@localhost ~]$ /usr/bin/atom: line 190: 734813 Illegal instruction (core dumped) nohup "$ATOM_PATH" --executed-from="$(pwd)" --pid=$$ "$@" > "$ATOM_HOME/nohup.out" 2>&1 Fontconfig warning: "/usr/share/fontconfig/conf.avail/05-reset-dirs-sample.conf", line 6: unknown element "reset-dirs" [734813:0104/121006.785414:FATAL:gpu_data_manager_impl_private.cc(892)] The display compositor is frequently crashing. Goodbye. Failed to generate minidump.
Also none for me: https://github.com/atom/atom/issues/23167#issuecomment-999113262
Fails when launching from the Menu. From syslog:
Mar 8 07:58:45 fedora-35-vm abrt-server[19195]: Package 'atom' isn't signed with proper key
To make Atom startable by Menu you can alter the corresponding desktop-file:
/usr/share/applications/atom.desktop
Exec=env ATOM_DISABLE_SHELLING_OUT_FOR_ENVIRONMENT=false /usr/bin/atom --in-process-gpu
Thank you. That's what I ended up doing.
To make Atom startable by Menu you can alter the corresponding desktop-file:
/usr/share/applications/atom.desktop
Exec=env ATOM_DISABLE_SHELLING_OUT_FOR_ENVIRONMENT=false /usr/bin/atom --in-process-gpu
Hi,
for me, "--in-process-gpu" didn't solve the issue.
I had to add --no-sandbox, and it worked.
Regards,
I don't know if others are getting this, so cross posting here, but suddenly on Fedora 35 latest Atom 1.60 if you do anything in Atom which triggers a file open popup then it freezes GNOME. I believe some very recent Fedora 35 update in the last week or so, maybe Linux kernel 5.17 not sure, is now not playing well with Atom. This is a big deal because it makes Atom much harder to use.
https://github.com/atom/atom/issues/25437
Fedora Linux 35, Atom 1.60.0 x64, kernel 5.17.5-200, GNOME 41.6, X11.
Using the GPU switch workaround. The application opens, works, and can open files.
Be sure to use sudo update-desktop-database if making a change does not fix the application.
Fedora Linux 35, Atom 1.60.0 x64, kernel 5.17.5-200, GNOME 41.6, X11.
Using the GPU switch workaround. The application opens, works, and can open files.
Be sure to use
sudo update-desktop-databaseif making a change does not fix the application.
Most people using modern Fedora don't use X11 anymore. I've used Wayland since probably Fedora 30 something. At least 2-3 years ago. This could mean why yours works differently.
I use KDE, which is hardly usable with wayland, especially with NVIDIA drivers.I would like to use wayland, but KDE is n'ont ready for it.Le 5 mai 2022 21:05, Leandro Hermida @.***> a écrit :
Fedora Linux 35, Atom 1.60.0 x64, kernel 5.17.5-200, GNOME 41.6, X11. Using the GPU switch workaround. The application opens, works, and can open files. Be sure to use sudo update-desktop-database if making a change does not fix the application.
Most people using modern Fedora don't use X11 anymore. I've used Wayland since probably Fedora 30 something. At least 2-3 years ago. This could mean why yours works differently.
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you commented.Message ID: @.***>
That's for some reason the way I had it. Probably like the article says, with proprietary NVIDIA graphics drivers (470xx):
> rpm -q -a 'name=*nvidia*' | sort
akmod-nvidia-470xx-470.103.01-2.fc36.x86_64
kmod-nvidia-470xx-470.103.01-2.fc36.x86_64
kmod-nvidia-470xx-5.17.5-200.fc35.x86_64-470.103.01-1.fc35.x86_64
kmod-nvidia-470xx-5.17.6-200.fc35.x86_64-470.103.01-1.fc35.x86_64
kmod-nvidia-470xx-5.17.6-300.fc36.x86_64-470.103.01-2.fc36.x86_64
nvidia-persistenced-510.68.02-1.fc36.x86_64
nvidia-settings-470xx-470.103.01-3.fc36.x86_64
xorg-x11-drv-nvidia-470xx-470.103.01-4.fc36.x86_64
xorg-x11-drv-nvidia-470xx-cuda-470.103.01-4.fc36.x86_64
xorg-x11-drv-nvidia-470xx-cuda-libs-470.103.01-4.fc36.i686
xorg-x11-drv-nvidia-470xx-cuda-libs-470.103.01-4.fc36.x86_64
xorg-x11-drv-nvidia-470xx-devel-470.103.01-4.fc36.i686
xorg-x11-drv-nvidia-470xx-devel-470.103.01-4.fc36.x86_64
xorg-x11-drv-nvidia-470xx-kmodsrc-470.103.01-4.fc36.x86_64
xorg-x11-drv-nvidia-470xx-libs-470.103.01-4.fc36.i686
xorg-x11-drv-nvidia-470xx-libs-470.103.01-4.fc36.x86_64
xorg-x11-drv-nvidia-470xx-power-470.103.01-4.fc36.x86_64
I made sure not to use the latest/rolling packages again!
Just did a system upgrade, ran through the 2 tests again, and all still seems fine here. Seems Wayland is the most important change here, where at least my release and GNOME versions changed as well.
Fedora 35 v Fedora 36: What's the Difference? - It's FOSS News
⋊> ~ grep -iP '^Exec' '/usr/share/applications/atom.desktop'
Exec=env ATOM_DISABLE_SHELLING_OUT_FOR_ENVIRONMENT=false /usr/bin/atom --in-process-gpu %F
Graphics: NV124 OS: Fedora Linux 36 (Workstation Edition) OS Type: 64-bit GNOME Version: 42.1 Windowing System: Wayland
Atom: 1.60.0
In Fedora 35 --no-sandbox or --in-process-gpu both worked and solved the editor start crash problem. But now with Fedora 36 using --in-process-gpu the app starts but it crashes when editing settings and clicking on a setting with a drop-down menu... the menu won't show any items and then a second later the entire editor will crash. It doesn't crash every time you click on a setting drop-down but after a couple tries it will crash. When I changed startup to --no-sandbox it doesn't crash when displaying settings drop-down menus.
Workaround:
atom --no-sandbox
Be advised: using --no-sandbox disables the electron sandbox, and could leave you vulnerable to attacks when installing outdated/unverified packages.
https://github.com/electron/electron/issues/15760
Read more above to find info about the electron sandbox.
https://github.com/atom/atom/issues/23036#issuecomment-1139239233
This is a duplicate of #23036 and the issue should be closed. Please follow the link above for more information on this issue.
