search

ValveSoftware / steam-for-linux

Issue tracking for the Steam for Linux beta client
4.25k stars 175 forks source link

steamwebhelper crashing on NVIDIA with GPU-accelerated web rendering (workaround in beta) #9780

Open onegentig opened 1 year ago

onegentig commented 1 year ago

Since yesterday (1. July 2023), Steam hangs on start, unable to create a window for itself (new GUI problem?). It appears that start.sh is repeatedly trying to create a steamwebhelper that just crashes soon after spawning.

Update: Identified as a CEF (libcef.so) issue, it’s currently tracked by NVIDIA and looks like a workaround is out in Steam Beta (see changelog). Should be fixed by disabling "GPU-accelerated web rendering" and wiping GLCache, see #9780 (comment).

System Information

$ dnf list installed "*steam*" ``` Installed Packages steam.i686 1.0.0.78-1.fc38 @rpmfusion-nonfree-updates steam-devices.i686 1.0.0.78-1.fc38 @rpmfusion-nonfree-steam ```
$ dnf list installed "*nvidia*" ``` Installed Packages akmod-nvidia.x86_64 3:535.54.03-1.fc38 @rpmfusion-nonfree-nvidia-driver kmod-nvidia-6.3.8-200.fc38.x86_64.x86_64 3:535.54.03-1.fc38 @@commandline nvidia-persistenced.x86_64 3:535.54.03-1.fc38 @rpmfusion-nonfree-nvidia-driver nvidia-settings.x86_64 3:535.54.03-1.fc38 @rpmfusion-nonfree-nvidia-driver nvidia-vaapi-driver.x86_64 0.0.10-1.fc38 @rpmfusion-nonfree-nvidia-driver xorg-x11-drv-nvidia.x86_64 3:535.54.03-2.fc38 @rpmfusion-nonfree-nvidia-driver xorg-x11-drv-nvidia-cuda.x86_64 3:535.54.03-2.fc38 @rpmfusion-nonfree-nvidia-driver xorg-x11-drv-nvidia-cuda-libs.x86_64 3:535.54.03-2.fc38 @rpmfusion-nonfree-nvidia-driver xorg-x11-drv-nvidia-cuda-libs.i686 3:535.54.03-2.fc38 @rpmfusion-nonfree-nvidia-driver xorg-x11-drv-nvidia-kmodsrc.x86_64 3:535.54.03-2.fc38 @rpmfusion-nonfree-nvidia-driver xorg-x11-drv-nvidia-libs.i686 3:535.54.03-2.fc38 @rpmfusion-nonfree-nvidia-driver xorg-x11-drv-nvidia-libs.x86_64 3:535.54.03-2.fc38 @rpmfusion-nonfree-nvidia-driver xorg-x11-drv-nvidia-power.x86_64 3:535.54.03-2.fc38 @rpmfusion-nonfree-nvidia-driver ```
$ inxi -F ``` System: Host: ongn-zetaxi270 Kernel: 6.3.8-200.fc38.x86_64 arch: x86_64 bits: 64 Desktop: GNOME v: 44.2 Distro: Fedora release 38 (Thirty Eight) Machine: Type: Desktop System: Gigabyte product: Z270X-Ultra Gaming v: N/A serial: Mobo: Gigabyte model: Z270X-Ultra Gaming-CF v: x.x serial: UEFI: American Megatrends v: F8 date: 10/27/2017 CPU: Info: quad core model: Intel Core i7-6700 bits: 64 type: MT MCP cache: L2: 1024 KiB Speed (MHz): avg: 3724 min/max: 800/4000 cores: 1: 3870 2: 3400 3: 3400 4: 3885 5: 3820 6: 3801 7: 3737 8: 3884 Graphics: Device-1: NVIDIA GP107 [GeForce GTX 1050] driver: nvidia v: 535.54.03 Display: x11 server: X.Org v: 1.20.14 with: Xwayland v: 22.1.9 driver: X: loaded: nvidia unloaded: fbdev,modesetting,nouveau,vesa gpu: nvidia,nvidia-nvswitch resolution: 1920x1080 API: OpenGL v: 4.6.0 NVIDIA 535.54.03 renderer: NVIDIA GeForce GTX 1050/PCIe/SSE2 Audio: Device-1: Intel 200 Series PCH HD Audio driver: snd_hda_intel Device-2: NVIDIA GP107GL High Definition Audio driver: snd_hda_intel Device-3: Trust GXT 258 Microphone driver: hid-generic,snd-usb-audio,usbhid type: USB API: ALSA v: k6.3.8-200.fc38.x86_64 status: kernel-api Server-1: PipeWire v: 0.3.72 status: active Network: Device-1: Intel Ethernet I219-V driver: e1000e IF: enp0s31f6 state: up speed: 1000 Mbps duplex: full mac: 1c:1b:0d:97:f8:f7 Bluetooth: Device-1: Cambridge Silicon Radio Bluetooth Dongle (HCI mode) driver: btusb type: USB Report: rfkill ID: hci0 state: up address: see --recommends Drives: Local Storage: total: 2.69 TiB used: 1.61 TiB (59.8%) ID-1: /dev/sda vendor: Samsung model: SSD 870 QVO 1TB size: 931.51 GiB ID-2: /dev/sdb vendor: Western Digital model: WD1003FZEX-00K3CA0 size: 931.51 GiB ID-3: /dev/sdc vendor: SanDisk model: EMTEC X150 960GB size: 894.25 GiB Partition: ID-1: / size: 145.01 GiB used: 72.64 GiB (50.1%) fs: ext4 dev: /dev/sda6 ID-2: /boot/efi size: 96 MiB used: 46.6 MiB (48.5%) fs: vfat dev: /dev/sda2 ID-3: /home size: 191.45 GiB used: 122.63 GiB (64.1%) fs: ext4 dev: /dev/sda8 Swap: ID-1: swap-1 type: zram size: 8 GiB used: 0 KiB (0.0%) dev: /dev/zram0 ID-2: swap-2 type: partition size: 4.66 GiB used: 0 KiB (0.0%) dev: /dev/sda7 Sensors: System Temperatures: cpu: 45.0 C mobo: N/A gpu: nvidia temp: 36 C Fan Speeds (RPM): N/A gpu: nvidia fan: 35% Info: Processes: 367 Uptime: 13m Memory: available: 15.56 GiB used: 4.28 GiB (27.5%) Shell: Zsh inxi: 3.3.27 ```

Description

When I try to start Steam either from the terminal or from the application launcher, the process just hangs and doesn't open the Steam client (has to be killed using kill or just killall steam).

The issue started appeared one day ago–Steam was working perfectly fine before that. I made no changes to the system, settings or anything in that time (auto-update might have happened though).

Executing steam in the terminal didn't print any errors or any lines that would differ from a successful launch. After "RegisterForAppOverview", it appears that steamwebhelper was repeatedly executed:

steamwebhelper.sh[330664]: Runtime for steamwebhelper: defaulting to /home/onegen/.steam/debian-installation/ubuntu12_64/steam-runtime-heavy
steamwebhelper.sh[330664]: glibc >= 2.34, partially disabling sandbox until CEF supports clone3()

The longer I let it run, the more steamwebhelpers were spawned (seen in htop and ps aux | grep 'steam', no window ever appeared. At this point, the program didn't respond to even SIGINTs and had to be forcibly killed.

\ By some miracle, I was able to start Steam in offline mode (steam -offline) once after running steam --reset. Games, incl. online ones, ran as expected. But sadly, after I restarted the computer again, not even the offline option helped (I tried resetting again, no change).

Addendum: Running steam -bigpicture fails in a…peculiar way (video) [Screencast from 2023-07-03 00-08-47.webm](https://github.com/ValveSoftware/steam-for-linux/assets/84882649/84987064-f1ae-499d-8bd2-37af1d181f91)

Steps for reproducing this issue:

  1. Run Steam

What I tried so far:

LuNeder commented 1 year ago

This problem also happens on OpenSUSE Tumbleweed.

Steam opens and the tray icon appears and all, but no UI. This started to happen the day a mesa downgrade happened on Tumbleweed, 7 days ago. Back then running zypper dup with --allow-vendor-change used to fix it, but now it's broken again and that does not fix it anymore. Systemd journal also seems to briefly mention a problem with libcef.so:

  jul 02 19:22:19│script[3096]          │-> checking /home/luana/.local/share/Steam/ubuntu12_32/steam-runtime                                                                                                                            │
 │jul 02 19:22:22│dbus-daemon[2514]     │[session uid=1000 pid=2514] Activating service name='org.gnome.ScreenSaver' requested by ':1.135' (uid=1000 pid=4073 comm="/home/luana/.local/share/Steam/ubuntu12_32/steam -")                 │
 │jul 02 19:22:22│script[3096]          │waiting for steam to start...done                                                                                                                                                               │
 │jul 02 19:22:24│dbus-daemon[2514]     │[session uid=1000 pid=2514] Activating service name='org.gnome.ScreenSaver' requested by ':1.140' (uid=1000 pid=4252 comm="/home/luana/.local/share/Steam/ubuntu12_32/../ubun")                 │
 │jul 02 19:22:24│dbus-daemon[2514]     │[session uid=1000 pid=2514] Activating service name='org.gnome.ScreenSaver' requested by ':1.140' (uid=1000 pid=4252 comm="/home/luana/.local/share/Steam/ubuntu12_32/../ubun")                 ┬
 │jul 02 19:22:25│dbus-daemon[2514]     │[session uid=1000 pid=2514] Activating service name='org.gnome.ScreenSaver' requested by ':1.144' (uid=1000 pid=4341 comm="/home/luana/.local/share/Steam/ubuntu12_32/../ubun")                 │
 │jul 02 19:22:25│dbus-daemon[2514]     │[session uid=1000 pid=2514] Activating service name='org.gnome.ScreenSaver' requested by ':1.144' (uid=1000 pid=4341 comm="/home/luana/.local/share/Steam/ubuntu12_32/../ubun")                 │
 │jul 02 19:22:28│kernel                │traps: steamwebhelper[4138] trap invalid opcode ip:7f86113db794 sp:7fff9f5686a0 error:0 in libcef.so[7f860eaef000+7770000]                                                                      ┴
 │jul 02 19:22:29│systemd-coredum[4435] │Process 4138 (steamwebhelper) of user 1000 dumped core.                                                                                                                                         │
 │jul 02 19:22:29│drkonqi-coredum[4436] │"/home/luana/.local/share/Steam/ubuntu12_64/steamwebhelper" 4138 "/var/lib/systemd/coredump/core.steamwebhelper.1000.73defe08011544b8a54fa9222e25b4df.4138.1688336548000000.zst"                │
 │jul 02 19:22:39│systemd-coredum[4807] │Process 4482 (steamwebhelper) of user 1000 dumped core.                                                                                                                                         │
 │jul 02 19:22:39│drkonqi-coredum[4808] │"/home/luana/.local/share/Steam/ubuntu12_64/steamwebhelper" 4482 "/var/lib/systemd/coredump/core.steamwebhelper.1000.73defe08011544b8a54fa9222e25b4df.4482.1688336559000000.zst"                │
 │jul 02 19:22:49│systemd-coredum[4955] │Process 4820 (steamwebhelper) of user 1000 dumped core.                                                                                                                                         │
 │jul 02 19:22:50│drkonqi-coredum[4957] │"/home/luana/.local/share/Steam/ubuntu12_64/steamwebhelper" 4820 "/var/lib/systemd/coredump/core.steamwebhelper.1000.73defe08011544b8a54fa9222e25b4df.4820.1688336569000000.zst"

Trying to run the steamwebhelper also errors on libcef.so: /home/luana/.local/share/Steam/ubuntu12_64/steamwebhelper: error while loading shared libraries: libcef.so: cannot open shared object file: No such file or directory

Running with -bigpicture I get the same result as you do too.

However, one thing seems to work: if I open Steam's old UI instead of the new one with steam -vgui, it launches fine and the UI works! Can you try that? (Tho I'm sure the old UI won't stay forever so we do need a fix lol)

onegentig commented 1 year ago

(( More possibly related issues: #9383 (diff: Flatpak, opt) and #9706 (diff: KDE) ))

However, one thing seems to work: if I open Steam's old UI instead of the new one with steam -vgui, it launches fine and the UI works! Can you try that? (Tho I'm sure the old UI won't stay forever so we do need a fix lol) – @LuNeder

Thanks, running steam -vgui got at least the old GUI to open!

However, it felt slow and attempting to change "Steam Library Folders" setting crashed the entire GUI (reloading it). As my games are installed in a non-default location and my settings were reset, I cannot run anything. I was able to get Steam to recognise my games folder by editing ~/.steam/root/steamapps/libraryfolders.vdf by hand. It recognised most of the apps there – Steam wanted to re-download a few games like "Warframe", but by clicking "Cancel" it [slowly] verified its presence.

Additionally, the "Friends Network" (chat) didn't work, and opening "Big Picture Mode" with -vgui had the same outcome as without it (the glitching screen from original post). Odd…

Playing with libraryfolders.vdf
 > I am leaving this here for anyone who has a similar problem. > Or myself if I need it again. :) > > Before anything, make sure Steam is turned off and remember to make a backup: > > ```sh > killall steam > cp ~/.steam/root/steamapps/libraryfolders.vdf ~/.steam/root/steamapps/libraryfolders.vdf.bak > ``` > > Open `libraryfolders.vdf` in your editor of choice (gedit, nano, vim,…). > It should look something like this: > > ```json > "libraryfolders" > { > "0" > { > "path" "/home/username/.local/share/Steam" > "label" "" > "contentid" "999999999999999" > "totalsize" "0" > "update_clean_bytes_tally" "0" > "time_last_update_corruption" "0" > "apps" > { > } > } > } > ``` > > Steam luckily can fill most information by itself on startup. All we really have to do it give it the "path" and a random "contentid" integer. > > - `path` is usually called "SteamLibrary". You want the folder that *contains* `steamapps`, but not `steamapps` itself. > - `contentid` is just a random unique integer. Make up something, as long as it is different from the "0" (just touch the numpad a bit and you're done). > > Then insert it like so: > > ```json > "libraryfolders" > { > "0" > { > "path" "/home/username/.local/share/Steam" > "label" "" > "contentid" "999999999999999" > "totalsize" "0" > "update_clean_bytes_tally" "0" > "time_last_update_corruption" "0" > "apps" > { > } > } > "1" > { > "path" "/run/media/username/GAMEHDD/SteamLibrary" > "label" "" > "contentid" "1234567890000" > "totalsize" "0" > "update_clean_bytes_tally" "0" > "time_last_update_corruption" "0" > "apps" > { > } > } > } > ``` > > On next startup, Steam should automatically detect "apps" and everything, even the ID was regenerated. > > Of course, no guarantees this workaround works for everyone. If you try this and it doesn't work, feel free to reply here. > > ------------------

Tho I'm sure the old UI won't stay forever so we do need a fix lol

Yeah, here's hoping that Valve fixes this issue soon. 🙏🏽

For now I'm just glad I'm not an isolated incident. 😄

Systemd journal also seems to briefly mention a problem with libcef.so

So I ventured to the endless pits of journalctl and I was able to get a similar core dump note:

systemd-coredump[66477]: [🡕] Process 66315 (steamwebhelper) of user 1000 dumped core.

Indeed there was the same line about libcef:

kernel: traps: steamwebhelper[66315] trap invalid opcode ip:7fe8823db794 sp:7fff18da8e00 error:0 in libcef.so[7fe87>

Here is the entire relevant part of the journal:

$ journalctl ``` Jul 03 00:01:41 ongn-zetaxi270 abrt-dump-journal-oops[1874]: abrt-dump-journal-oops: Found oopses: 1 Jul 03 00:01:41 ongn-zetaxi270 abrt-dump-journal-oops[1874]: abrt-dump-journal-oops: Creating problem directories Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66298]: Executable '/home/onegen/.steam/debian-installation/ubuntu12_64/steamwebhelper' doesn't belong > Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66298]: 'post-create' on '/var/spool/abrt/ccpp-2023-07-03-00:01:41.284470-66131' exited with 1 Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66298]: Deleting problem directory '/var/spool/abrt/ccpp-2023-07-03-00:01:41.284470-66131' Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66300]: Can't find a meaningful backtrace for hashing in '.' Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66300]: Deleting non-reportable oops '.' because DropNotReportableOopses is set to 'yes' Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66300]: 'post-create' on '/var/spool/abrt/oops-2023-07-03-00:01:41-1874-0' exited with 1 Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66298]: Deleting problem directory '/var/spool/abrt/ccpp-2023-07-03-00:01:41.284470-66131' Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66300]: Can't find a meaningful backtrace for hashing in '.' Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66300]: Deleting non-reportable oops '.' because DropNotReportableOopses is set to 'yes' Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66300]: 'post-create' on '/var/spool/abrt/oops-2023-07-03-00:01:41-1874-0' exited with 1 Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66300]: Deleting problem directory '/var/spool/abrt/oops-2023-07-03-00:01:41-1874-0' Jul 03 00:01:41 ongn-zetaxi270 abrt-server[66300]: Lock file '.lock' was locked by process 66313, but it crashed? Jul 03 00:01:42 ongn-zetaxi270 abrt-dump-journal-oops[1874]: Reported 1 kernel oopses to Abrt Jul 03 00:01:43 ongn-zetaxi270 guake.desktop[3561]: Showing the terminal Jul 03 00:01:47 ongn-zetaxi270 guake.desktop[3561]: Hiding the terminal Jul 03 00:01:50 ongn-zetaxi270 audit[66315]: ANOM_ABEND auid=1000 uid=1000 gid=1000 ses=3 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c10> Jul 03 00:01:50 ongn-zetaxi270 kernel: traps: steamwebhelper[66315] trap invalid opcode ip:7fe8823db794 sp:7fff18da8e00 error:0 in libcef.so[7fe87> Jul 03 00:01:50 ongn-zetaxi270 audit: BPF prog-id=292 op=LOAD Jul 03 00:01:50 ongn-zetaxi270 audit: BPF prog-id=293 op=LOAD Jul 03 00:01:50 ongn-zetaxi270 audit: BPF prog-id=294 op=LOAD Jul 03 00:01:50 ongn-zetaxi270 systemd[1]: Started systemd-coredump@65-66476-0.service - Process Core Dump (PID 66476/UID 0). Jul 03 00:01:50 ongn-zetaxi270 audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=syste> Jul 03 00:01:51 ongn-zetaxi270 systemd-coredump[66477]: [🡕] Process 66315 (steamwebhelper) of user 1000 dumped core. Module /home/onegen/.steam/debian-installation/ubuntu12_64/steamwebhelper from rpm libxcb-> Module /home/onegen/.steam/debian-installation/ubuntu12_64/libminigbm.so from rpm libxcb-1> Module /home/onegen/.steam/debian-installation/ubuntu12_64/libSDL3.so.0 from rpm libxcb-1.> Module /home/onegen/.steam/debian-installation/ubuntu12_64/libcef.so from rpm libxcb-1.13.> Module libxcb-glx.so.0 from rpm libxcb-1.13.1-11.fc38.x86_64 Module libXss.so.1 from rpm libXScrnSaver-1.2.3-12.fc38.x86_64 Module libudev.so.1 from rpm systemd-253.5-1.fc38.x86_64 Module libdconfsettings.so from rpm dconf-0.40.0-8.fc38.x86_64 Module libbrotlicommon.so.1 from rpm brotli-1.0.9-11.fc38.x86_64 Module libbrotlidec.so.1 from rpm brotli-1.0.9-11.fc38.x86_64 Module libbz2.so.1 from rpm bzip2-1.0.8-13.fc38.x86_64 Module libgmp.so.10 from rpm gmp-6.2.1-4.fc38.x86_64 Module libhogweed.so.6 from rpm nettle-3.8-3.fc38.x86_64 Module libnettle.so.8 from rpm nettle-3.8-3.fc38.x86_64 Module libtasn1.so.6 from rpm libtasn1-4.19.0-2.fc38.x86_64 Module libunistring.so.5 from rpm libunistring-1.1-3.fc38.x86_64 Module libidn2.so.0 from rpm libidn2-2.3.4-2.fc38.x86_64 Module libp11-kit.so.0 from rpm p11-kit-0.24.1-6.fc38.x86_64 Module libcrypto.so.3 from rpm openssl-3.0.9-1.fc38.x86_64 Module libkeyutils.so.1 from rpm keyutils-1.6.1-6.fc38.x86_64 Module libkrb5support.so.0 from rpm krb5-1.20.1-8.fc38.x86_64 Module libcom_err.so.2 from rpm e2fsprogs-1.46.5-4.fc38.x86_64 Module libk5crypto.so.3 from rpm krb5-1.20.1-8.fc38.x86_64 Module libkrb5.so.3 from rpm krb5-1.20.1-8.fc38.x86_64 Module libdatrie.so.1 from rpm libdatrie-0.2.13-5.fc38.x86_64 Module libpixman-1.so.0 from rpm pixman-0.42.2-1.fc38.x86_64 Module libxcb-shm.so.0 from rpm libxcb-1.13.1-11.fc38.x86_64 Module libxcb-render.so.0 from rpm libxcb-1.13.1-11.fc38.x86_64 Module libpng16.so.16 from rpm libpng-1.6.37-14.fc38.x86_64 Module libxml2.so.2 from rpm libxml2-2.10.4-1.fc38.x86_64 Module libgraphite2.so.3 from rpm graphite2-1.3.14-11.fc38.x86_64 Module libfreetype.so.6 from rpm freetype-2.13.0-2.fc38.x86_64 Module liblz4.so.1 from rpm lz4-1.9.4-2.fc38.x86_64 Module libzstd.so.1 from rpm zstd-1.5.5-1.fc38.x86_64 Module liblzma.so.5 from rpm xz-5.4.1-1.fc38.x86_64 Module libcap.so.2 from rpm libcap-2.48-6.fc38.x86_64 Module libblkid.so.1 from rpm util-linux-2.38.1-4.fc38.x86_64 Module libgnutls.so.30 from rpm gnutls-3.8.0-2.fc38.x86_64 Module libavahi-client.so.3 from rpm avahi-0.8-22.fc38.x86_64 Module libavahi-common.so.3 from rpm avahi-0.8-22.fc38.x86_64 Module libgssapi_krb5.so.2 from rpm krb5-1.20.1-8.fc38.x86_64 Module libthai.so.0 from rpm libthai-0.1.29-4.fc38.x86_64 Module libfribidi.so.0 from rpm fribidi-1.0.12-3.fc38.x86_64 Module libcairo.so.2 from rpm cairo-1.17.8-4.fc38.x86_64 Module libfontconfig.so.1 from rpm fontconfig-2.14.2-1.fc38.x86_64 Module libharfbuzz.so.0 from rpm harfbuzz-7.1.0-1.fc38.x86_64 Module libpangoft2-1.0.so.0 from rpm pango-1.50.14-1.fc38.x86_64 Module libsystemd.so.0 from rpm systemd-253.5-1.fc38.x86_64 Module libXau.so.6 from rpm libXau-1.0.11-2.fc38.x86_64 Module libplds4.so from rpm nss-3.90.0-1.fc38.x86_64 Module libplc4.so from rpm nss-3.90.0-1.fc38.x86_64 Module libdrm.so.2 from rpm libdrm-2.4.114-2.fc38.x86_64 Module libGLdispatch.so.0 from rpm libglvnd-1.6.0-2.fc38.x86_64 Module libGLX.so.0 from rpm libglvnd-1.6.0-2.fc38.x86_64 Module libselinux.so.1 from rpm libselinux-3.5-1.fc38.x86_64 Module libmount.so.1 from rpm util-linux-2.38.1-4.fc38.x86_64 Module libz.so.1 from rpm zlib-1.2.13-3.fc38.x86_64 Module libgmodule-2.0.so.0 from rpm glib2-2.76.3-1.fc38.x86_64 Module libpcre2-8.so.0 from rpm pcre2-10.42-1.fc38.1.x86_64 Module libffi.so.8 from rpm libffi-3.4.4-2.fc38.x86_64 Module libatspi.so.0 from rpm at-spi2-core-2.48.3-1.fc38.x86_64 Module libcups.so.2 from rpm cups-2.4.6-1.fc38.x86_64 Module libpango-1.0.so.0 from rpm pango-1.50.14-1.fc38.x86_64 Module libpangocairo-1.0.so.0 from rpm pango-1.50.14-1.fc38.x86_64 Module libasound.so.2 from rpm alsa-lib-1.2.9-1.fc38.x86_64 Module libdbus-1.so.3 from rpm dbus-1.14.8-1.fc38.x86_64 Module libexpat.so.1 from rpm expat-2.5.0-2.fc38.x86_64 Module libXfixes.so.3 from rpm libXfixes-6.0.0-5.fc38.x86_64 Module libXcursor.so.1 from rpm libXcursor-1.2.1-3.fc38.x86_64 Module libxcb-dri3.so.0 from rpm libxcb-1.13.1-11.fc38.x86_64 Module libxcb.so.1 from rpm libxcb-1.13.1-11.fc38.x86_64 Module libX11-xcb.so.1 from rpm libX11-1.8.6-1.fc38.x86_64 Module libatk-bridge-2.0.so.0 from rpm at-spi2-core-2.48.3-1.fc38.x86_64 Module libatk-1.0.so.0 from rpm at-spi2-core-2.48.3-1.fc38.x86_64 Module libnspr4.so from rpm nss-3.90.0-1.fc38.x86_64 Module libibus-1.0.so.5 from rpm ibus-1.5.28-6.fc38.x86_64 Module libGL.so.1 from rpm libglvnd-1.6.0-2.fc38.x86_64 Module libXdamage.so.1 from rpm libXdamage-1.1.5-9.fc38.x86_64 Module libXcomposite.so.1 from rpm libXcomposite-0.4.5-9.fc38.x86_64 Module libXrandr.so.2 from rpm libXrandr-1.5.2-10.fc38.x86_64 Module libXtst.so.6 from rpm libXtst-1.2.4-2.fc38.x86_64 Module libXrender.so.1 from rpm libXrender-0.9.11-2.fc38.x86_64 Module libXext.so.6 from rpm libXext-1.3.5-2.fc38.x86_64 Module libXi.so.6 from rpm libXi-1.8.1-1.fc38.x86_64 Module libX11.so.6 from rpm libX11-1.8.6-1.fc38.x86_64 Module libgio-2.0.so.0 from rpm glib2-2.76.3-1.fc38.x86_64 Module libglib-2.0.so.0 from rpm glib2-2.76.3-1.fc38.x86_64 Module libgobject-2.0.so.0 from rpm glib2-2.76.3-1.fc38.x86_64 Stack trace of thread 66315: #0 0x00007fe8823db794 n/a (/home/onegen/.steam/debian-installation/ubuntu12_64/libcef.so > #1 0x00007fe882467522 n/a (/home/onegen/.steam/debian-installation/ubuntu12_64/libcef.so > #2 0x00007fe88246a764 n/a (/home/onegen/.steam/debian-installation/ubuntu12_64/libcef.so > #3 0x00007fe85a8f77f2 n/a (libnvidia-glcore.so.535.54.03 + 0x14f77f2) #4 0x00007fe85a3fcf45 n/a (libnvidia-glcore.so.535.54.03 + 0xffcf45) #5 0x00007fe85a40844a n/a (libnvidia-glcore.so.535.54.03 + 0x100844a) #6 0x00007fe85a746182 n/a (libnvidia-glcore.so.535.54.03 + 0x1346182) #7 0x00007fe85a749cb5 n/a (libnvidia-glcore.so.535.54.03 + 0x1349cb5) #8 0x00007fe87deb9695 n/a (/home/onegen/.steam/debian-installation/ubuntu12_64/libSDL3.so> #9 0x00007fe87deb9a1e n/a (/home/onegen/.steam/debian-installation/ubuntu12_64/libSDL3.so> ELF object binary architecture: AMD x86-64 Jul 03 00:01:51 ongn-zetaxi270 systemd[1]: systemd-coredump@65-66476-0.service: Deactivated successfully. Jul 03 00:01:51 ongn-zetaxi270 audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=system> Jul 03 00:01:51 ongn-zetaxi270 audit: BPF prog-id=294 op=UNLOAD Jul 03 00:01:51 ongn-zetaxi270 audit: BPF prog-id=293 op=UNLOAD Jul 03 00:01:51 ongn-zetaxi270 audit: BPF prog-id=292 op=UNLOAD ```
LuNeder commented 1 year ago

Okay yeah we surely have the same problem then!

For me, the friends window on -vgui tells me I'm offline (I'm clearly not, even updating games in Steam itself work witgh -vgui). Tho I did not try to actually launch any games yet lol.

Leaving my logs here too since we do have the same problem: logs.tar.gz

I'm on OpenSUSE Tumbleweed, XFCE and NVidia 1070Ti

grillo-delmal commented 1 year ago

I'm having the same issues as @nickonegen , I think we only differ in our hardware

inxi -F ``` System: Host: rivaia Kernel: 6.3.8-200.fc38.x86_64 arch: x86_64 bits: 64 Desktop: GNOME v: 44.2 Distro: Fedora release 38 (Thirty Eight) Machine: Type: Desktop Mobo: MSI model: Z97 GAMING 7 (MS-7916) v: 1.0 serial: UEFI-[Legacy]: American Megatrends v: 1.12 date: 02/16/2016 CPU: Info: quad core model: Intel Core i7-4790K bits: 64 type: MT MCP cache: L2: 1024 KiB Speed (MHz): avg: 1666 min/max: 800/4400 cores: 1: 4399 2: 800 3: 800 4: 861 5: 800 6: 800 7: 800 8: 4068 Graphics: Device-1: NVIDIA TU106 [GeForce RTX 2060 Rev. A] driver: nvidia v: 535.54.03 Display: x11 server: X.Org v: 1.20.14 with: Xwayland v: 22.1.9 driver: X: loaded: nvidia gpu: nvidia,nvidia-nvswitch resolution: 1: 1920x1080~60Hz 2: 1920x1080~60Hz API: OpenGL v: 4.6.0 NVIDIA 535.54.03 renderer: NVIDIA GeForce RTX 2060/PCIe/SSE2 Audio: Device-1: Intel 9 Series Family HD Audio driver: snd_hda_intel Device-2: NVIDIA TU106 High Definition Audio driver: snd_hda_intel Device-3: Jieli USBAudio2.0 driver: hid-generic,snd-usb-audio,usb-storage,usbhid type: USB Device-4: Razer USA Seiren Mini driver: hid-generic,snd-usb-audio,usbhid type: USB API: ALSA v: k6.3.8-200.fc38.x86_64 status: kernel-api Server-1: PipeWire v: 0.3.72 status: active Network: Device-1: Qualcomm Atheros Killer E220x Gigabit Ethernet driver: alx IF: enp5s0 state: up speed: 1000 Mbps duplex: full mac: 44:8a:5b:d5:f1:a2 Drives: Local Storage: total: 3.17 TiB used: 203.74 GiB (6.3%) ID-1: /dev/sda vendor: Western Digital model: WD Green 2.5 480GB size: 447.13 GiB ID-2: /dev/sdb vendor: Western Digital model: WD10EZEX-08M2NA0 size: 931.51 GiB ID-3: /dev/sdc vendor: Western Digital model: WD20EZAZ-00L9GB0 size: 1.82 TiB Partition: ID-1: / size: 446.11 GiB used: 12.52 GiB (2.8%) fs: btrfs dev: /dev/dm-0 ID-2: /boot size: 973.4 MiB used: 668 MiB (68.6%) fs: ext4 dev: /dev/sda2 ID-3: /home size: 1.82 TiB used: 190.57 GiB (10.2%) fs: btrfs dev: /dev/dm-1 ID-4: /opt size: 1.82 TiB used: 190.57 GiB (10.2%) fs: btrfs dev: /dev/dm-1 ID-5: /var size: 1.82 TiB used: 190.57 GiB (10.2%) fs: btrfs dev: /dev/dm-1 Swap: ID-1: swap-1 type: zram size: 8 GiB used: 0 KiB (0.0%) dev: /dev/zram0 Sensors: System Temperatures: cpu: 43.0 C mobo: N/A gpu: nvidia temp: 54 C Fan Speeds (RPM): N/A gpu: nvidia fan: 0% Info: Processes: 362 Uptime: 48m Memory: available: 31.27 GiB used: 4.07 GiB (13.0%) Shell: Bash inxi: 3.3.27 ```
fr33zing commented 1 year ago

I was able to get Steam to launch normally without -vgui by deleting ~/.local/share/Steam

onegentig commented 1 year ago

@grillo-delmal: I'm having the same issues as nickonegen, I think we only differ in our hardware

Right, so far this (or similar) issue was see on various DEs (GNOME1, KDE4, XFCE2) and distros, (Fedora1 4, Ubuntu5, openSUSE2, Arch7), even on the Flatpak8 release – but all affected systems have a NVIDIA GPU in the GeForce 900 generation or younger. I haven't seen a report of someone with older GPU or a non-NVIDIA GPU have this problem.

Can't speak for all distros (GPU driver version isn't shown in most issues) but at least for Fedora, I can safely say that all these GPUs use the same akmod-nvidia driver from RPMFusion (v535.54.03). I looked for an alike Fedora issue that would have a different version, but found none. Maybe I'll meddle with that as well, but I'd very much prefer not doing that… 😓

Perhaps the label:NVIDIA drivers could be applied to this issue?

Now, I am by no means an expert in this field, but if they [can] use the same driver, their interface can't be all that different, and might be possibly the same in eyes of apps like Steam?

onegentig commented 1 year ago

@fr33zing: I was able to get Steam to launch normally without -vgui by deleting ~/.local/share/Steam

While I tried this, I decided to give it another shot, and I found something.

Removing ~/.local/share/Steam (further referred to as the Steam directory) worked sometimes to get it running. Removed it, tried running steam via terminal, and it worked without the performance and crashing issues as -vgui. Then I tried opening it using the icon (.desktop) and everything was as normal – no window, live-die loop of steamwebhelper. Wiping the Steam directory again allowed steam to be ran from the terminal once more.

However after a few attempts, not even rm -Rf ~/.local/share/Steam got it running again. It re-downloaded everything but hung at the same spot once more. Can't say I have made any other changes, I only tried experimenting with removing individual subfolders. Not even a full reinstall with sudo dnf remove steam helped.

Given how many symlinks are in that directory, I can't say I am surprised. 😅 I saved copies of the Steam directory while it worked and after it stopped working and ran it through diff -sr – aside from minor differences, cache and logs, there were no changes. Attempting to restore the functioning copy didn't get it running either.

I will continue experimenting with it, but so far, only steam -vgui works.

Addendum: I've seen some people report that they at least get to the login screen – I wasn't so lucky. DRI_PRIME=0 has no impact, nor did PrefersNonDefaultGPU=true or X-KDE-RunOnDiscreteGpu=true in the steam.desktop, regardless of -vgui or new GUI. Changing any settings in old GUI like "GPU-accelerated rendering" or "Hardware video decoding" had no effect on anything either.

cyrasil commented 1 year ago

@grillo-delmal: I'm having the same issues as nickonegen, I think we only differ in our hardware

Right, so far this (or similar) issue was see on various DEs (GNOME1 3, KDE4 6, XFCE2) and distros, (Fedora1 3 4, Ubuntu5, openSUSE2, Arch7), even on the Flatpak8 release – but all affected systems have a NVIDIA GPU in the GeForce 900 generation or younger. I haven't seen a report of someone with older GPU or a non-NVIDIA GPU have this problem.

I have similar (Don't want to say "same" yet, because I have not read through all the posts carefully), issue with Arch+i3wm+amd RX6600 GPU. Ussually some variation of the following sequence gets the steam running, but it requires several tries: steam-native, steam-runtime, steam -vgui -bigpicture.

Edit: I've removed cinnamon and openbox DE that i had installed from the past + the dependencies installed by those packages and steam now starts normally even without the -vgui option.

Edit2: From the pacman log xdg-desktop-portal, xdg-desktop-portal-xapp were one of the dependencies removed. But i still have the -gnome and -gtk versions installed. Mentioning this, because these were mentioned in some of the other related issues.

LuNeder commented 1 year ago

Quoting someone from NVidia in some discord group: “from what I could see it looks like it might be a race condition in CEF”

So it makes sense that -vgui launches, since it didn’t use CEF while the new UI is apparently made in HTML and uses CEF for everything.

grillo-delmal commented 1 year ago

I opened my steam client again with the -vgui flag and was able to download an update (from the beta branch) that fixed the problem for me.

LuNeder commented 1 year ago

I opened my steam client again with the -vgui flag and was able to download an update (from the beta branch) that fixed the problem for me.

I’ve opened with -vgui and joined the beta, but the problem persists when not using -vgui for me.

grillo-delmal commented 1 year ago

disregard what I said, the problem persists. It's just that if I leave the big picture looping like how @nickonegen showed at some point at random it starts normally if you leave it looping, but when you start a game it starts looping again and the game ends up being unplayable... at least -vgui works and now I can open the game options panel there, which was actually a problem until now.

LuNeder commented 1 year ago

After joining the beta yesterday night, I booted my PC today and Steam is indeed working without -vgui lol (kinda laggy tho)

mabenson00 commented 1 year ago

not working even with the beta for me.

Kizaing commented 1 year ago

Did a dnf update today and I'm hit with the exact same issue :( running with flatpak steam, I was able to get in with the -vgui flag but it seems to have no network connectivity. The new UI or big picture mode won't launch at all, just stuck in a loop

drunknsorry commented 1 year ago

With an RTX 3080, this was what I had to do to get steam working properly

1) Open terminal and steam -vgui 2) Steam > Settings > Interface, then disable hardware accelerated web views 3) In terminal do steam --reset 4) Once steam launches, close it fully 5) In terminal sudo rm -r ~/.cache/nvidia/GLCache 5) Steam should now keep launching normally

Edit: On x11 with 535 drivers

iddm commented 1 year ago

How come you have this directory there and I don't?

RustyReich commented 1 year ago

Replying to https://github.com/ValveSoftware/steam-for-linux/issues/9780#issuecomment-1622073358

Can confirm, this appears to have fixed the issue for me, at least for the time being, on a GTX 1070.

Kizaing commented 1 year ago

Replying to https://github.com/ValveSoftware/steam-for-linux/issues/9780#issuecomment-1622073358

No dice for me :( RTX 3060 on driver 535, Fedora 38

grillo-delmal commented 1 year ago

Replying to https://github.com/ValveSoftware/steam-for-linux/issues/9780#issuecomment-1622073358

It fixed the problem for me too, (I skipped step 3 though)

onegentig commented 1 year ago

from #9780 (comment):  5. In terminal sudo rm -r ~/.cache/nvidia/GLCache

Apologies for the late reply, but I wanted to test things out before replying.

The command sudo rm -r ~/.cache/nvidia/GLCache(why sudo?) didn’t do anything – like @vityafx, I didn't have the folder there at all. With locate, I found the GLCache folder in ~/.nv.

And sure enough, after wiping this folder, it looks like it worked!

  1. Launched steam -vgui via terminal
  2. Opted into Steam Beta Update(not sure if necessary?)
  3. Disabled GPU-accelerated web rendering
  4. Closed Steam, wait until the Steam process fully exits
  5. rm -r ~/.nv/GLCache
  6. Steam worked properly now!

Steam now seems to work correctly via terminal and .desktop link. I restarted it several times to make sure, and for now, it looks like it… just works! steam --reset was not necessary.

I wonder why the folder difference – according to NVIDIA Dev-Guide, ~/.nv should be the default folder for NVIDIA shader cache (but it can be changed by setting the $__GL_SHADER_DISK_CACHE_PATH variable). Not sure if it’s distro repack difference or we just installed things differently…? (I doubt that, but for info, I just followed the RPMFusion guide to the letter.)

Either way, most of y’all should have it at ~/.nv/GLCache or ~/.cache/nvidia/GLCache or maybe ~/.cache/.nv/GLCache. If neither, you’ll have to look for the folder with locate or something (it's not ~/.local/share/Steam/steamapps/**, that is Vulkan shader cache for games).

Quick sidenote: I noticed that I was able to run all games and even the GUI through Lutris, and I couldn’t for a long time figure out what was different. Well, if the GLCache was at fault, Lutris had to do something with or around that – and it looks like it just straight up used its own cache in ~/.cache/lutris/shaders/steam/GLCache.

On the latest Beta Update (steam-runtime_0.20230509.49499), I re-enabled "GPU-accelerated web rendering", and so far it looks like it all works without problems (incl. Big Picture). According to the changelog, the issue should be patched – though it’s likely still worked on (they call it a workaround), and I kinda won’t be surprised if it broke in a few restarts (it’s beta afterall) – I’ll update this post if that happens. :)

Kizaing commented 1 year ago

Alright I was able to get it working again by uninstalling the Flatpak version and installing the rpm-fusion package, seems to be working as normal now

Drogobo commented 1 year ago

Mine is still broken after trying everyone's advice. I am an arch user on latest everything with a 2070 super.

Buccaneersdan commented 1 year ago

@Drogobo I tend to believe the solution posted by @nickonegen works for me on kernel 6.4.1-arch2-1 and nvidia-dkms 535.54.03. I've done nothing before that but a reinstall of the drivers after a failed downgrade to 530.41.03 (the downgrade can't compile the dkms for kernel 6.4.1 anymore).

Have you tried steam-native or steam -vgui from console? There are a few red herrings in the output but I had an "terminate called after throwing an instance of 'dxvk::DxvkError" at some point which I fixed by putting VK_ICD_FILENAMES=/usr/share/vulkan/icd.d/nvidia_icd.json in /etc/environment. If you think you found an error, maybe create a gist of it and post the link here.

IBBoard commented 1 year ago

I seem to be hitting this on openSUSE Tumbleweed. With updates up to Friday last week, it worked if I just disabled hardware acceleration. With today's updates it only works with -vgui.

I've tried https://github.com/ValveSoftware/steam-for-linux/issues/9780#issuecomment-1623489003 and it didn't work for me. I've tried a couple of CEF switches, which don't fix the crash but sometimes change the sandboxing message. I've tried a steam --reset and still getting the crash loop after logging in again.

NVidia 535.54.03, kernel 6.3.9-1, glibc-2.37-4.4 and Steam 1.0.0.78 by the RPM, 1687386907 by the UI. GeForce GTX 970. Steam Beta isn't any different.

Drogobo commented 1 year ago

I don't know about what everyone else's type of glitch is, but mine is really screwed. steam -vgui crashes xorg whenever I open a game. The regular one is even worse; I will just run steam, and xorg will kill itself. I am on the currently-available newest version of the kernel and drivers for Arch Linux.

xi-pinmping commented 1 year ago

I'm on arch as well with the newest nvidia drivers and can't start steam through the command line, using -vgui, native or anything else mentioned in this thread. Unfortunately I just scrolled all the way through.

iddm commented 1 year ago

To everyone who posted the comments after @nickonegen : I had the same issue with 3080 and 4090, until I did what @nickonegen suggested. Mane people find the GLCache directory by another path, I found it exactly where @nickonegen said it would be. I deleted it as per suggestion, restarted steam without any flags, and it has been all working since then, and working great - not a single issue with whatever there might be (except for the non-related HiDPI problem, but this is off-topic).

This isn't going to be fixed with new kernel or with new driver from NVIDIA, this can only be fixed (if it is ever going to be) by the maintainers of the NVIDIA package, who would delete/erase this GLCache directory with the new package release of the same driver.

And a few more things: initially, yes, the "vgui" argument was doing the job for me (even though with some caveats), but then I also couldn't see the window even with this flag. I also couldn't launch any single game (apparently, they refused to load the libsteam_api.so, or some parts of it, at least the GameServer API for sure), and steam couldn't even place the tray icon. So yes, for me it got worse for some reason. But, again, I got it all fixed thanks to @nickonegen. @kisak-valve you suggested that I had a relevant bug where the only option I had was downgrading the driver. As you may see, this isn't the case. Just an FYI. Thanks for trying to help!

iddm commented 1 year ago

I seem to be hitting this on openSUSE Tumbleweed. With updates up to Friday last week, it worked if I just disabled hardware acceleration. With today's updates it only works with -vgui.

I've tried https://github.com/ValveSoftware/steam-for-linux/issues/9780#issuecomment-1623489003 and it didn't work for me. I've tried a couple of CEF switches, which don't fix the crash but sometimes change the sandboxing message. I've tried a steam --reset and still getting the crash loop after logging in again.

NVidia 535.54.03, kernel 6.3.9-1, glibc-2.37-4.4 and Steam 1.0.0.78 by the RPM, 1687386907 by the UI. GeForce GTX 970. Steam Beta isn't any different.

For the sake of sharing knowledge and helping others - did you find the GLCache directory there, deleted it and it didn't work? Or you didn't find the directory there?

IBBoard commented 1 year ago

It was there (not in .cache), and deleting didn't change anything. Still won't open unless I use -vgui.

I'll try checking other locations later on case there's another cache that Steam is using.

iddm commented 1 year ago

It was there (not in .cache), and deleting didn't change anything. Still won't open unless I use -vgui.

I'll try checking other locations later on case there's another cache that Steam is using.

That might make sense, as the distro you are using is different. Try "sudo find -iname GLCache /" Or something like that. Or you may ask the maintainers of the package of your distro, or look at how it is packaged, perhaps, they change this setting there.

IBBoard commented 1 year ago

Ahah! ~/.cache/.nv/GLCache on openSUSE Tumbleweed using the G06 drivers from https://download.nvidia.com/opensuse/tumbleweed. Steam now opens with the new UI and works with GPU accelerated web views. The .nv directory existed, but doesn't seem to have anything later than 2018.

As a user, I like the idea of the package cleaning up the cache if it's going to cause bugs.

But as an occasional developer and packager then I know that's a PITA and don't think it's a great idea. It would be better if the code automatically invalidated the cache to prevent whatever problem this was, rather than having to forcefully delete it on package update etc.

Thanks everyone!

Drogobo commented 1 year ago

I don't know why, but clearing my nvidia cache doesn't do anything to help my issue.

Veronicap98 commented 1 year ago

Wiping GLCache worked for me.

Garuda Linux, 6.4.1-zen2-1-zen, nvidia version 535.54.03-1

Simply erased all the contents of ~/.cache/nvidia/GLCache, did nothing else, Steam works perfectly now.

spoolcron commented 1 year ago

Arch Linux / 2070S user.

Found my solution here https://bugs.archlinux.org/task/79006

"Installing our lib32-libnm to avoid the old one from the runtime also avoids the crash."

ViNi-Arco commented 1 year ago

@spoolcron

Arch Linux / 2070S user.

Found my solution here https://bugs.archlinux.org/task/79006

"Installing our lib32-libnm to avoid the old one from the runtime also avoids the crash."

That was it, thank you very much! the lib32-libgudev 237-2 -> 238-1 update that broke Steam starting on Archlinux. if you don't want to downgrade lib32-libgudev, installing lib32-libnm will work fine too.

lib32-gst-plugins-good pulls lib32-libgudev as dependency if you don't know how it got on your system.

Edit: installing lib32-libudev0-shim is also an option, you decide, see.

xi-pinmping commented 1 year ago

I have searched my whole system for the GLCache and removed all to no avail. Installing lib32-libudev0-shim, lib32-libnm, and downgrading lib32-libgudev has not worked. I cannot even get to the menu to disable hardware acceleration using any of the listed options on both runtime and native.

Co3co commented 1 year ago

Removing GLCache fixed it for me. The only thing is, I have to do it every time I try to launch Steam, which isn't really a problem as I can use an alias to do both.

gotama commented 1 year ago

Removing GLCache fixed it for me. The only thing is, I have to do it every time I try to launch Steam, which isn't really a problem as I can use an alias to do both.

Did you switch off hard ware acceleration for WebView in steam settings?

steffenWi commented 1 year ago

Thank you. Executing

cd /var/cache/pacman/pkg/
pacman -U ./libgudev-237-2-x86_64.pkg.tar.zst ./lib32-libgudev-237-2-x86_64.pkg.tar.zst

fixed it. Was able to launch Steam immediately after executing that command. This is on Arch Linux with an AMD GPU in case it matters.

That was it, thank you very much! the lib32-libgudev 237-2 -> 238-1 update that broke Steam starting on Archlinux. if you don't want to downgrade lib32-libgudev, installing lib32-libnm will work fine too.

Co3co commented 1 year ago

Did you switch off hard ware acceleration for WebView in steam settings?

Yes, I did. The only thing I didn't do was steam --reset.

ipaqmaster commented 1 year ago

steam -vgui seemed to do the trick for me (3900x, 2080Ti, nvidia-535.54.03). Followed up with killing it and running rm -r ~/.nv/GLCache. After that I was able to launch Steam via its start menu shortcut again as normal.

maxpain commented 1 year ago

The same problem in Docker on Ubuntu 22.04 and 23.04 with 535 Nvidia drivers.

carlos-aurelio commented 1 year ago

And sure enough, after wiping this folder, it looks like it worked!

1. Launched `steam -vgui` via terminal

2. Opted into Steam Beta Update(not sure if necessary?)

3. Disabled GPU-accelerated web rendering

4. Closed Steam, wait until the Steam process fully exits

5. `rm -r ~/.nv/GLCache`

6. Steam worked properly now!

Works for me but I need to remove the GLCache everytime before I start Steam

Reger95 commented 1 year ago

For AMD GPU users you can have a try as below: If amdgpu stack has been installed already do what as below: $ sudo amdgpu-uninstall $ sudo amdgpu-install --usecase=graphics --vulkan=pro If amdgpu stack has not been installed, Install amdgpu stack according to https://amdgpu-install.readthedocs.io/en/latest/preamble.html Attention: --usecase don't use workstation option. After above, reboot your machine. Though something maybe changed but it works for me.

ragardner commented 1 year ago

I am also having this issue, amd gpu pop os. The only way I can get steam to run normally is by installing the flatpak version and then clicking open from within pop shop

Edit: I found another way to get it to start normally, right clicking on the steam icon and clicking "Library" to start steam in the library works for me

garrett commented 1 year ago

It seems like I'm having this issue too, on an AMD 7900 XTX on Fedora Silverblue 38. I can no longer run the Steam from Flatpak from GNOME. It works from the terminal, oddly, if I do flatpak run com.valvesoftware.Steam or if I rebase to Kinoite and run it from KDE.

The workarounds listed above don't seem to work for me.

I even went as far as rebuilding my Steam configs in ~/.var/app/com.valvesoftware.Steam/ completely from scratch (including re-adding games... sigh). It's still broken.

kisak-valve commented 1 year ago

Hello @garrett, that set of symptoms makes your issue #9383.

garrett commented 1 year ago

@kisak-valve Thanks! Yeah, that's exactly my issue. I just got a new computer and this one has integrated graphics as well as the 7900. (My last only had a PCI GPU.)

EntityinArray commented 1 year ago
  1. Shut down Steam entirely
  2. Go to ~/.cache/nvidia/GLCache
  3. Delete everything there (you will see a bunch of files with hex gibberish)
  4. Launch Steam, it worked in my case after that

I think this is NVIDIA driver bug, not Steam's fault. I recently bought my first NVIDIA card, spent hours trying to set everything up on Arch Linux and stumbled upon this and now i'm dissapointed. All other GPUs work fine on Linux, what's wrong NVIDIA?