Skip to content

Instantly share code, notes, and snippets.

@fsworld009

fsworld009/3.steam.md

Last active April 9, 2025 10:56
Show Gist options
  • Save fsworld009/86f806fcfb7ee2cddaa5df89f4113bf3 to your computer and use it in GitHub Desktop.
Save fsworld009/86f806fcfb7ee2cddaa5df89f4113bf3 to your computer and use it in GitHub Desktop.
Download ZIP
Aurora DX on Framework 16 note
Raw
3.steam.md

disrobox

Used bazzite-arch image

To be able to use gamescope, use the options suggested in 89luca89/distrobox#451 (comment)

distrobox create -n bazzite -i ghcr.io/ublue-os/bazzite-arch:latest --init-hooks "install -o 1000 -g 1000 -d /tmp/.X11-unix-new; mount --bind /tmp/.X11-unix-new /tmp/.X11-unix"

Then we can install gamescope and mangohud inside the container

# Required to upgrade once, otherwise installation would fail
sudo pacman -Syu
# install
sudo pacman -S gamescope, mangohud

Main problem: cannot detect new controllers. So if a controller is disconnected in game, we need to shutdown game and client and restart it. For me, this is a major deal breaker, as it makes Steam Input pretty much useless.

obs-vkcapture plugin is already installed, game can be captured properly in OBS running in another container.

Steam flatpak

gamescope won't work unless using a specific flatpak ProtonGE version

flathub/org.freedesktop.Platform.VulkanLayer.gamescope#6

Other than that seems no particular drawback.

If we want to move flatpak data to another partition, we need to make a symlink at ~/.var/app/com.valvesoftware.Steam and point to an empty disk on target partition, then run flatpak Steam to initialize it.

IME

Got Flatpak fcitx working in Flatpak Steam under KDE Wayland by

  1. Set both XMODIFIERS=@im=fcitx and GTK_IM_MODULE=xim via Flatseal
  2. Enroll in Steam Beta client (2025/02/01: Stable works now)

Screenshot_20241204_095133 Screenshot_20241204_095127 Screenshot_20241204_095248

rpm-ostree layering

considering this option

Protontricks

https://github.com/flathub/com.github.Matoking.protontricks

Add permissions to flatpak steam and any game folder from external drives via flatseal for command line.

Reshade

https://github.com/kevinlekiller/reshade-steam-proton

Use the reshade-linux-flatpak.sh script, that would install Reshade dlls and default effects in Steam flatpak folder so Steam flatpak and other game managers like lutris have permissions to read the dll.

Raw
4. development.md

Install fonts

https://github.com/Homebrew/homebrew-linux-fonts

Fonts for my dev env:

brew install font-hack
brew install --HEAD homebrew/linux-fonts/font-dejavu-sans-mono-for-powerline

Dejavu Sans Mono has been renamed to Dejavu Sans Mono for Powerline

Not needed in Aurora DX

Already included in base image

# Shell tools
brew install starship fish tmux
# Nerd font
brew install font-hack-nerd-font
Raw
auroradx-fw16.md

https://getaurora.dev/

General

  • Based on Fedora Silverblue
  • Identitcal to their another project Bluefin DX (where DE is Gnome)
  • Same maintainers also made Bazztie

So using these projects as keywords when Googling, the information should also be applicable to Aurora DX.

2024/11/05

Silverblue Installation: by default it uses the whole disk, to only install in a partition, need to manually create partition table.

Refer to https://docs.bazzite.gg/General/Installation_Guide/dual_boot_setup_guide/

  1. Set desired btrfs partition size
  2. To encrypt disk, set luks when creating btrfs partition

Post install

  • sudo hostnamectl set-hostname (hostname)
  • disable animation when moving window: system settings -> Window management -> Desktop Effects -> wobby windows
  • Can set Konsole as default terminal app
    • But no Konsole in menu, probably need to manually create like in Bazzite
  • Discover store shows FW16 firmware update, need to reboot to apply it, but it's very convenient that firmware update is reachable this way.
  • Distrobox was empty, but bluefin-cli was added after next boot.

Brave cannot access controller

Flatseal, add access to /run/udev/

2024/11/06

Build portable OBS

Use a Fedora 40 distrobox to build and run

https://github.com/obsproject/obs-studio/wiki/build-instructions-for-linux

Install dependencies for Fedora and do portable build

The previous portable build I made in Arch cannot be run, which suggests the portable build is still platform dependant.

For my version https://github.com/fsworld009/obs-studio/releases/tag/30.2.3-projector-patches I have to use cef_binary_5060_linux_x86_64 for browser build, probably because the branch is based on older version

Using the latest one results in constant

make: *** [Makefile:156: all] Error 2

error, and won't pass obs-browser build.

Plugin path is changed to /home/fsworld009/obs-studio-portable/lib64/obs-plugins/, not sure about the 32bit ones.

Without LD_PRELOAD=/usr/lib/libv4l/v4l1compat.so it can still read video from GC553 properly, not sure why.

Can probably change to non-portable build and export from distrobox? But I suppose that would be replaced by official builds on container upgrade.

Create a script to launch it with one click

distrobox enter fedora -- ~/obs-studio-portable/bin/obs

VeraCrypt

Can install in distrobox, but containers doesn't have permission to access disks

89luca89/distrobox#762 suggests to create a rootful container.

distrobox create --name fedora-veracrypt --image quay.io/fedora/fedora:40 --root --volume /media

distrobox enter --root fedora-veracrypt

sudo dnf install device-mapper fuse gtk3 pcsc-lite 
sudo rpm -i ~/Downloads/veracrypt-1.26.14-Fedora-40-x86_64.rpm

The folder can be mounted successfully, however it is only visible inside the container.

Followed https://www.zihad.com.bd/posts/install-veracrypt-in-fedora-silverblue-kinoite/ to install in home folder instead. Used gtk2 version. No dark theme, but that's fine.

Shadercache process

579: fsworld+  156505 97.6  0.2 670456 182656 ?       SNl  16:12   1:00 /run/media/fsworld009/data/flatpakdata/com.valvesoftware.Steam/.local/share/Steam/ubuntu12_32/../ubuntu12_64/fossilize_replay /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steamapp_pipeline_cache.foz /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache.foz --master-process --quiet-slave --shmem-fd 111 --control-fd 137 --spirv-val --num-threads 4 --on-disk-validation-whitelist /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/

Waydroid

For bazzite: https://universal-blue.discourse.group/t/how-to-go-about-getting-waydroid-on-bluefin-dx/3371/11

However it's not available in Bluefin so maybe https://github.com/89luca89/distrobox/blob/main/docs/useful_tips.md#automated-installation

Might need manually apply stuffs from ujust setup-waydroid configure steps:

https://github.com/ublue-os/bazzite/blob/bdc8d0f6ba73baf64db12b26ea594bac25d8a89b/system_files/desktop/shared/usr/share/ublue-os/just/82-bazzite-waydroid.just#L32-L39

And might also need to manually apply some of the patches in the waydroid systemd services: https://github.com/ublue-os/bazzite/blob/main/system_files/desktop/shared/usr/libexec/waydroid-container-start

2025/03/18:

Automated setup doesn't work as it still requires packages on host https://github.com/Vanilla-OS/waydroid-image

And after initialization I get Unit waydroid-init.service could not be found.

Seems the only way to get it work is to layer it: https://universal-blue.discourse.group/t/unable-to-install-waydroid-through-discover-software-centre-unable-to-repport-issue/5547

Required for running in SE Linux env ublue-os/bazzite#1998 https://github.com/ublue-os/bazzite/blob/master/system_files/desktop/shared/usr/share/ublue-os/just/82-bazzite-waydroid.just

Temp disable SE Linux https://universal-blue.discourse.group/t/waydroid-failing-to-start-fixed/5718

Might be helpful for testing purpose

sudo setenforce 0

flatpak: zombie [bwrap] processes

Sometimes when closing a flatpak chromium app (Brave, Vesktop, Steam, Chromium)... the app might not exit cleanly and leaving zombie processes (mostly [bwrap] processes).

The problem with zombie processes is that we cannot rerun the app again.

Ideally, we could use https://askubuntu.com/a/1121528 to find the parents and kill them. But it doesn't work here because flatpak apps use bwrap and if we keep trace their parent with

ps -f [pid]

It would goes back to the whole systemd process

UID          PID    PPID  C STIME TTY      STAT   TIME CMD
fsworld+    3218       1  0 22:28 ?        Ss     0:00 /usr/lib/systemd/systemd --user

and killing this process would crash the entire login session.

Some explanations on bwrap found here: containers/bubblewrap#620

systemd-cgls shows flatpak apps belong to app.slice:

 └─user-1000.slice
│   ├─[email protected]
│   │ ├─app.slice
│   │ │ ├─app-flatpak-com.brave.Browser-5483.scope
│   │ │ │ ├─5483 /usr/bin/bwrap --args 41 -- /app/bin/zypak-helper child - /app/brave/brave --type=zygote --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54->
│   │ │ │ ├─5492 /usr/bin/bwrap --args 41 -- /app/bin/zypak-helper child - /app/brave/brave --type=zygote --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54->
│   │ │ │ ├─5493 /app/brave/brave --type=zygote --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54-f637-434a-b463-e2d5ab48e6d5,unknown --change-stack-guard-o>
│   │ │ │ ├─5532 /app/brave/brave --type=utility --utility-sub-type=storage.mojom.StorageService --lang=en-US --service-sandbox-type=utility --string-annotations --crashpad-handler-pid=1>
│   │ │ │ ├─5553 /app/brave/brave --type=renderer --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54-f637-434a-b463-e2d5ab48e6d5,unknown --enable-distillabil>
│   │ │ │ ├─5562 /app/brave/brave --type=renderer --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54-f637-434a-b463-e2d5ab48e6d5,unknown --enable-distillabil>
│   │ │ │ ├─5600 /app/brave/brave --type=renderer --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54-f637-434a-b463-e2d5ab48e6d5,unknown --extension-process >
│   │ │ │ ├─5614 /app/brave/brave --type=renderer --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54-f637-434a-b463-e2d5ab48e6d5,unknown --extension-process >
│   │ │ │ ├─5640 /app/brave/brave --type=renderer --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54-f637-434a-b463-e2d5ab48e6d5,unknown --extension-process >
│   │ │ │ ├─5653 /app/brave/brave --type=renderer --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54-f637-434a-b463-e2d5ab48e6d5,unknown --extension-process >
│   │ │ │ ├─5665 /app/brave/brave --type=renderer --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54-f637-434a-b463-e2d5ab48e6d5,unknown --extension-process >
│   │ │ │ └─9368 /app/brave/brave --type=renderer --string-annotations --crashpad-handler-pid=16 --enable-crash-reporter=0c9b6b54-f637-434a-b463-e2d5ab48e6d5,unknown --extension-process >
│   │ │

Not sure if it's possible to restart these

https://bbs.archlinux.org/viewtopic.php?id=227446 systemctl reboot seems to restart all

https://discussion.fedoraproject.org/t/how-to-list-and-kill-flatpak-apps-via-gui-like-gnome-system-monitor/74980/3 there is flatpak ps and flatpak kill commands.

2025/01/13: flatpak kill doesn't clean defunct processes.

Some tips on how to deal with zombie processes link to parent 1: https://serverfault.com/questions/76263/how-to-kill-a-defunct-process-with-parent-1

amdgpu issues

https://community.frame.work/t/arch-amdgpu-screen-freeze-on-zen-kernel-new-warnings-on-vanilla-kernel-6-12-5/62063/4 Screen freeze on 6.11.4+

Useful info:

Reset gpu

cat /sys/kernel/debug/dri/1/amdgpu_gpu_recover

amdgpu.dcdebugmask=0x400 to fix freeze caused by adaptive sync

2025/01/14

journalctl log before screen freeze

Jan 14 22:45:33 fsworld009-fw16 flatpak[26798]: [144:0114/224533.829604:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:33 fsworld009-fw16 flatpak[26697]: [54:0114/224533.862669:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:33 fsworld009-fw16 flatpak[26697]: [54:0114/224533.862776:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:33 fsworld009-fw16 flatpak[26798]: [144:0114/224533.862880:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:33 fsworld009-fw16 flatpak[26697]: [54:0114/224533.895953:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:33 fsworld009-fw16 flatpak[26697]: [54:0114/224533.896064:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:33 fsworld009-fw16 flatpak[26798]: [144:0114/224533.896165:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:33 fsworld009-fw16 flatpak[26697]: [54:0114/224533.929357:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:33 fsworld009-fw16 flatpak[26697]: [54:0114/224533.929478:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:33 fsworld009-fw16 flatpak[26798]: [144:0114/224533.929595:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:33 fsworld009-fw16 flatpak[26697]: [54:0114/224533.962687:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:33 fsworld009-fw16 flatpak[26697]: [54:0114/224533.962814:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:33 fsworld009-fw16 flatpak[26798]: [144:0114/224533.962929:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:33 fsworld009-fw16 flatpak[26697]: [54:0114/224533.996026:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:33 fsworld009-fw16 flatpak[26697]: [54:0114/224533.996173:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:33 fsworld009-fw16 flatpak[26798]: [144:0114/224533.996317:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.029403:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.029582:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26798]: [144:0114/224534.029750:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.062658:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.062761:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26798]: [144:0114/224534.062878:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.095987:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.096117:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26798]: [144:0114/224534.096225:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.129347:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.129482:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26798]: [144:0114/224534.129613:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.162700:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.162890:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26798]: [144:0114/224534.163058:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.196008:ERROR:gbm_pixmap_wayland.cc(82)] Cannot create bo with format= YUV_420_BIPLANAR and usage=SCANOUT_CPU_READ_WRITE
Jan 14 22:45:34 fsworld009-fw16 flatpak[26697]: [54:0114/224534.196152:ERROR:gpu_channel.cc(503)] Buffer Handle is null.
Jan 14 22:45:34 fsworld009-fw16 flatpak[26798]: [144:0114/224534.196273:ERROR:shared_image_interface_proxy.cc(134)] Buffer handle is null. Not creating a mailbox from it.

2025/02/03

Spotify core dump caused the entire laptop to crash

Feb 03 20:54:45 fsworld009-fw16 systemd-coredump[35996]: Process 35990 (spotify) of user 1000 terminated abnormally with signal 6/ABRT, processing...
Feb 03 20:54:45 fsworld009-fw16 audit[35996]: AVC avc:  denied  { sys_admin } for  pid=35996 comm="systemd-coredum" capability=21  scontext=system_u:system_r:systemd_coredump_t:s0 tcontext=system_u:system_r:systemd_coredump_t:s0 tclass=capability permissive=0
Feb 03 20:54:45 fsworld009-fw16 com.spotify.Client.desktop[35997]: *** stack smashing detected ***: terminated
Feb 03 20:54:45 fsworld009-fw16 audit[35997]: ANOM_ABEND auid=1000 uid=1000 gid=1000 ses=3 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=35997 comm="spotify" exe="/app/extra/share/spotify/spotify" sig=6 res=1
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=115 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=116 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=117 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 systemd-coredump[36003]: Process 35997 (spotify) of user 1000 terminated abnormally with signal 6/ABRT, processing...
Feb 03 20:54:45 fsworld009-fw16 audit[36003]: AVC avc:  denied  { sys_admin } for  pid=36003 comm="systemd-coredum" capability=21  scontext=system_u:system_r:systemd_coredump_t:s0 tcontext=system_u:system_r:systemd_coredump_t:s0 tclass=capability permissive=0
Feb 03 20:54:45 fsworld009-fw16 systemd[1]: Started [email protected] - Process Core Dump (PID 35996/UID 0).
Feb 03 20:54:45 fsworld009-fw16 audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-coredump@2-35996-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=118 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=119 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=120 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 systemd[1]: Started [email protected] - Pass systemd-coredump journal entries to relevant user for potential DrKonqi handling.
Feb 03 20:54:45 fsworld009-fw16 audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=drkonqi-coredump-processor@2-35996-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=121 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=122 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=123 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 systemd[1]: Started [email protected] - Process Core Dump (PID 36003/UID 0).
Feb 03 20:54:45 fsworld009-fw16 audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-coredump@3-36003-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=124 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=125 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 audit: BPF prog-id=126 op=LOAD
Feb 03 20:54:45 fsworld009-fw16 systemd[1]: Started [email protected] - Pass systemd-coredump journal entries to relevant user for potential DrKonqi handling.
Feb 03 20:54:45 fsworld009-fw16 audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=drkonqi-coredump-processor@3-36003-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Feb 03 20:54:46 fsworld009-fw16 systemd-coredump[36004]: [🡕] Process 35990 (spotify) of user 1000 dumped core.
                                                         
                                                         Stack trace of thread 241:
                                                         #0  0x00007f7bf99a4db4 n/a (n/a + 0x0)
                                                         #1  0x00007f7bf994c08e n/a (n/a + 0x0)
                                                         #2  0x00007f7bf9933882 n/a (n/a + 0x0)
                                                         #3  0x00007f7bf99346e7 n/a (n/a + 0x0)
                                                         #4  0x00007f7bf9a33c29 n/a (n/a + 0x0)
                                                         #5  0x00007f7bf9a34ed4 n/a (n/a + 0x0)
                                                         #6  0x00005650ae7710e9 n/a (n/a + 0x0)
                                                         ELF object binary architecture: AMD x86-64
Feb 03 20:54:46 fsworld009-fw16 systemd-coredump[36006]: [🡕] Process 35997 (spotify) of user 1000 dumped core.
                                                         
                                                         Stack trace of thread 247:
                                                         #0  0x00007f7bf99a4db4 n/a (n/a + 0x0)
                                                         #1  0x00007f7bf994c08e n/a (n/a + 0x0)
                                                         #2  0x00007f7bf9933882 n/a (n/a + 0x0)
                                                         #3  0x00007f7bf99346e7 n/a (n/a + 0x0)
                                                         #4  0x00007f7bf9a33c29 n/a (n/a + 0x0)
                                                         #5  0x00007f7bf9a34ed4 n/a (n/a + 0x0)
                                                         #6  0x00005650ae7710e9 n/a (n/a + 0x0)
                                                         ELF object binary architecture: AMD x86-64
Feb 03 20:54:46 fsworld009-fw16 systemd[1]: [email protected]: Deactivated successfully.
Feb 03 20:54:46 fsworld009-fw16 audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-coredump@2-35996-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Feb 03 20:54:46 fsworld009-fw16 systemd[1]: [email protected]: Consumed 283ms CPU time, 126.8M memory peak.
Feb 03 20:54:46 fsworld009-fw16 systemd[1]: [email protected]: Deactivated successfully.
Feb 03 20:54:46 fsworld009-fw16 audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-coredump@3-36003-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Feb 03 20:54:46 fsworld009-fw16 systemd[1]: [email protected]: Consumed 281ms CPU time, 126.6M memory peak.
Feb 03 20:54:46 fsworld009-fw16 audit: BPF prog-id=117 op=UNLOAD

May be related? https://bbs.archlinux.org/viewtopic.php?id=299776&p=2

2025/02/01

GUI apps running through distrobox doesn't display CJK fonts

fedora-silverblue/issue-tracker#540 fc-cache -v doesn't work, had to install fonts manually

sudo dnf install google-noto-sans-cjk-fonts

Seems like ideally, distrobox should pick fonts from host, but it didn't happen in my setup.

2025/02/25

Discord flatpak Screen shareing

Screenshareing on Waland wasn't supported for a long time, was using Vesktop as alternative.

However, the support has arrived recently (around Jan 2025) https://www.gamingonlinux.com/2025/01/discord-screen-sharing-with-audio-on-linux-wayland-is-officially-here/

It's not in Flatpak due to a bug related to file chooser flathub/com.discordapp.Discord#483

However according to flathub/com.discordapp.Discord#483 (comment) it's possible to install the test branch manually with it enabled

flatpak install --user https://dl.flathub.org/build-repo/166117/com.discordapp.Discord.flatpakref
flatpak run com.discordapp.Discord//test

file chooser bug and permission

flathub/com.discordapp.Discord#483 Required host system to bump to xdg-desktop-portal 1.20, in the meanwhile, if some files are not accessible, configure file access override in Flatseal as a workaround.

Opt-in Wayland features

Seems required for recording to work properly https://github.com/flathub/com.discordapp.Discord?tab=readme-ov-file#wayland

fcitx5 support

Not enabled by default, similar to Vesktop in the past https://gist.github.com/fsworld009/0a7240c4eab9254aca3f7a334acc1aa9#discord

We could use similar trick to copy the shortcut and override ourselves

cp /var/lib/flatpak/app/com.discordapp.Discord/current/active/export/share/applications/com.discordapp.Discord.desktop ~/.local/share/applications/com.discordapp.Discord.desktop

Modify the Exec line:

Exec=/usr/bin/flatpak run --branch=test --arch=x86_64 --command=com.discordapp.Discord --file-forwarding com.discordapp.Discord @@u %U @@ --enable-wayland-ime

Switch back to stable

(TODO later) Based on flathub/com.discordapp.Discord#483 (comment) Flatpak version shold be bumped with Wayland screenshare support in stable branch. We should be able to use this version once it's released in flathub.

Hardware accelaration

https://www.reddit.com/r/linux_gaming/comments/zqxre4/psa_discord_gpu_acceleration_doesnt_work/

Should launch with --enable-gpu-rasterization

2025/03/14

Hardware video decoding for Chromium based browser

By default software decoding is used by all chromium based browser (verified in brave and vivaldi).

See explanation and how to validate here: https://www.linuxuprising.com/2021/01/how-to-enable-hardware-accelerated.html

But the argument provided in that article doesn't work, need to usee the options from https://community.brave.com/t/request-for-hardware-accelerated-video-decode-vaapi-on-linux/62654/6?u=fsworld009

flatpak run com.vivaldi.Vivaldi --enable-features=VaapiVideoDecoder,VaapiIgnoreDriverChecks,Vulkan,DefaultANGLEVulkan,VulkanFromANGLE

Note: --ozone-platform=wayland seems not needed under Wayland. Also DON'T touch any flags in chrome://flags like Override software rendering list, seems breaking the enable-features flag we pass in.

RyzenAdj

Tools to monitor CPU and set TDP

https://github.com/FlyGoat/RyzenAdj Can build it in Fedora distrobox then use it in Aurora DX host.

Currently blocked FlyGoat/RyzenAdj#278

Due to secure boot, maybe disable it? Not sure if that would break Windows installation.

Raw
docker-windows-mac.md

Windows

https://github.com/dockur/windows

Encountered dockur/windows#943, need to run

sudo modprobe ip_tables iptable_nat

before first boot up, for share folder to work properly.

Create docker-compose.yml with

services:
  windows:
    image: dockurr/windows
    container_name: windows
    environment:
      VERSION: "11"
      DISK_SIZE: "256G"
      RAM_SIZE: "16G"
      CPU_CORES: "4"
      USERNAME: "fsworld009"
    devices:
      - /dev/kvm
      - /dev/net/tun
    cap_add:
      - NET_ADMIN
    ports:
      - 8006:8006
      - 3389:3389/tcp
      - 3389:3389/udp
    stop_grace_period: 2m
    volumes:
      - /run/media/fsworld009/data/windows/storage:/storage
      - /run/media/fsworld009/data:/data

and docker compose up -d

Current Problem: can only have one share folder

Remote desktop

Default browser view is not smooth and has no audio / clipboard support.

Use Remmina from Flatpak

Windows: Group policy editor:

Computer Configuration > Administrative Templates > Windows Components > Remote Desktop Services > Remote Desktop Session Host > Device and Resource Redirection Set the policy Limit audio playback quality to enabled and high and then reboot.

Remmina: New profile

  1. Protocol: RDP
  2. Advanced tab -> Audio output mode: Local

Ref:

Mac OSX

https://github.com/sickcodes/Docker-OSX https://ivonblog.com/posts/linux-docker-osx/

Brave crashes

Seems there's memory leak after playing videos

Dec 16 15:39:17 fsworld009-fw16 systemd[2987]: app-flatpak-com.brave.Browser-32946.scope: Consumed 2h 33min 40.244s CPU time, 14G memory peak.
Dec 16 15:39:19 fsworld009-fw16 systemd[2987]: app-flatpak-com.brave.Browser-32906.scope: Consumed 59min 52.281s CPU time, 2.3G memory peak.

Trying this workaround for now:

https://www.reddit.com/r/brave/comments/1h47jdd/memory_leak_when_watching_youtube/

Update: I have removed Return Youtube Dislike and turned off Dark Reader for Youtube and this has seemed to fix the problem.

Ref:

Raw
Game notes.md

Monster Hunter Rise

Windows 11

Always recompile shader on launch, no solution yet.

Possible cause but the registry change doesn't work for me https://www.reddit.com/r/MonsterHunterRisePC/comments/1e1k2la/mh_rise_pc_version_compiling_shaders_issue_fixed/

Aurora DX

Steam flatpak: most of the time the game crashed when entering the title screen. Have about 10% chance to actually get in.

journalctl

Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu: ring gfx_0.0.0 timeout, but soft recovered
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:7 pasid:32814)
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:  in process MonsterHunterRi pid 84135 thread vkd3d_queue pid 84215)
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x00008000dfc74000 from client 10
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00701431
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          Faulty UTCL2 client ID: SQC (data) (0xa)
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          MORE_FAULTS: 0x1
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          WALKER_ERROR: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          PERMISSION_FAULTS: 0x3
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          MAPPING_ERROR: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          RW: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:7 pasid:32814)
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:  in process MonsterHunterRi pid 84135 thread vkd3d_queue pid 84215)
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x00008000dfc77000 from client 10
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00000000
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          Faulty UTCL2 client ID: CB/DB (0x0)
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          MORE_FAULTS: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          WALKER_ERROR: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          PERMISSION_FAULTS: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          MAPPING_ERROR: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          RW: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:7 pasid:32814)
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:  in process MonsterHunterRi pid 84135 thread vkd3d_queue pid 84215)
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x00008000dfc74000 from client 10
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00000000
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          Faulty UTCL2 client ID: CB/DB (0x0)
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          MORE_FAULTS: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          WALKER_ERROR: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          PERMISSION_FAULTS: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          MAPPING_ERROR: 0x0
Nov 19 12:13:15 fsworld009-fw16 kernel: amdgpu 0000:03:00.0: amdgpu:          RW: 0x0

Some how I fixed it but I don't know which one actually works. Just recording past actions that might have been related

  • Remove Reframework
    • Steam settings -> Download -> Clear cache
  • Steam settings -> Download -> disable "Enable Shader cache"
  • delete steamapps/shadercache
    • trigger redownload and recompiling shader
  • Tried steam tinker launch (but didn't really do any perm change for this)
  • Launched with VKD3D_SHADER_CACHE_PATH=0 %command%
  • delete shader.cache2 and vkd3d-proton.cache
    • trigger recompiling shader
  • ujust update
    • No update
    • Updated to latest kernel on Bazzite and it still works, so unlikely to be a driver issue
  • plugin a monitor into dGPU
    • Probably unrelated
  • Change proton version to GE-Proton7-55
    • Performance is bad with this version
  • Reverted to GE-Proton9-18 after confirming it doesn't crash in 7-55

I suppose the crash is caused by bad shader cache and some of the actions get rid of those bad cache files?

According to https://gitlab.freedesktop.org/mesa/mesa/-/issues/11201#note_2422595 this is a generic error related to shader cache:

Consider that the game works on other Linux machine, currently suspecting the shader precache from Steam for this GPU is problematic.

Trying this solution:

  1. Steam settings -> Downloads -> Clear download cache
  2. Steam settings -> Downloads -> Uncheck Enable shader pre-cache
  3. delete
    1. steamapps/shadercache/1446780
    2. vkd3d-proton.cache, vkd3d-proton.cache.write, shader.cache2 in steamapps/common/MonsterHunterRise/
  4. Launch the game, it should compile shaders without steam's precache data
    1. Confirmed the game works
  5. Close the game, go to steamapps/common/MonsterHunterRise/, vkd3d-proton.cache and shader.cache2 should be presented
  6. Steam settings -> Downloads -> Check Enable shader pre-cache
  7. Launch the game again while keeping previously generated precaches, it should re compile again at first run
    1. The game works on first run
    2. On second run, it won't recompile and it seems still working

Note that I also have VKD3D_VULKAN_DEVICE=1 set to enforce using dGPU, but this is probably not needed, as the game already detected dGPU without it. Still fail after some days.

Enable fcitx, game capture, and Reframework, and be able to type CHT:

XMODIFIERS=@im=fcitx HOST_LC_ALL=zh_TW.UTF-8 OBS_VKCAPTURE=1 WINEDLLOVERRIDES="dinput8.dll=n,b" %command%

Ref: shader cache processes

   534: fsworld+   17237  1.0  0.0  16728 13660 ?        SNs  18:48   0:00 /run/media/fsworld009/data/flatpakdata/com.valvesoftware.Steam/.local/share/Steam/ubuntu12_32/../ubuntu12_64/fossilize_replay /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steamapp_pipeline_cache.foz /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache.foz --master-process --quiet-slave --shmem-fd 13 --control-fd 114 --spirv-val --num-threads 4 --on-disk-validation-whitelist /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache_whitelist --replayer-cache /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/replay_cache --device-index 0 --timeout-seconds 10 --implicit-whitelist 0
   535: fsworld+   17254 93.3  0.2 662276 138688 ?       SNl  18:48   0:11 /run/media/fsworld009/data/flatpakdata/com.valvesoftware.Steam/.local/share/Steam/ubuntu12_32/../ubuntu12_64/fossilize_replay /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steamapp_pipeline_cache.foz /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache.foz --master-process --quiet-slave --shmem-fd 13 --control-fd 114 --spirv-val --num-threads 4 --on-disk-validation-whitelist /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache_whitelist --replayer-cache /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/replay_cache --device-index 0 --timeout-seconds 10 --implicit-whitelist 0
   536: fsworld+   17255  101  0.2 670764 148116 ?       SNl  18:48   0:12 /run/media/fsworld009/data/flatpakdata/com.valvesoftware.Steam/.local/share/Steam/ubuntu12_32/../ubuntu12_64/fossilize_replay /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steamapp_pipeline_cache.foz /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache.foz --master-process --quiet-slave --shmem-fd 13 --control-fd 114 --spirv-val --num-threads 4 --on-disk-validation-whitelist /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache_whitelist --replayer-cache /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/replay_cache --device-index 0 --timeout-seconds 10 --implicit-whitelist 0
   537: fsworld+   17256  101  0.2 670180 149968 ?       SNl  18:48   0:12 /run/media/fsworld009/data/flatpakdata/com.valvesoftware.Steam/.local/share/Steam/ubuntu12_32/../ubuntu12_64/fossilize_replay /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steamapp_pipeline_cache.foz /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache.foz --master-process --quiet-slave --shmem-fd 13 --control-fd 114 --spirv-val --num-threads 4 --on-disk-validation-whitelist /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache_whitelist --replayer-cache /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/replay_cache --device-index 0 --timeout-seconds 10 --implicit-whitelist 0
   538: fsworld+   17257  101  0.2 670364 144384 ?       SNl  18:48   0:12 /run/media/fsworld009/data/flatpakdata/com.valvesoftware.Steam/.local/share/Steam/ubuntu12_32/../ubuntu12_64/fossilize_replay /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steamapp_pipeline_cache.foz /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache.foz --master-process --quiet-slave --shmem-fd 13 --control-fd 114 --spirv-val --num-threads 4 --on-disk-validation-whitelist /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/steam_pipeline_cache_whitelist --replayer-cache /run/media/fsworld009/data/SteamLibrary/steamapps/shadercache/1446780/fozpipelinesv6/steamapprun_pipeline_cache.769baa6fe04cfcb6/replay_cache --device-index 0 --timeout-seconds 10 --implicit-whitelist 0
   551: fsworld+   17740  0.0  0.0  10908  5912 pts/1    S+   18:48   0:00 ug shader

2024/12/15

Pass VKD3D_VULKAN_DEVICE=0 seems to let the game to launch at least for now? Could be related to shadercache command using --device-index 0.

Also, https://bbs.archlinux.org/viewtopic.php?id=298387 indicates the problem might be fixed in newer kernel, so used ujust rebase-helper to switch to latest tag.

2024/12/27: latest tag bumped kernel to 6.12.6 which should contain the related fix according to https://gitlab.freedesktop.org/drm/amd/-/issues/3599#note_2627028 Still observing.

Monster Hunter Wilds

Avoid Proton Experimental Bleeding edge

Most suggestions were proton experimental (bleeding edge). However, with this Proton version, the prefix will be recreated on every launch, and Denuvo would treat every new prefixes as new machine, eventually lock you out of the game for 24HR since you authorized more than 5 machines in short period.

Use proton-GE 9-26 up instead. 9-26 is released on 3/12/2025, and should have most hot fixes from vkd3d-proton since game is launched.

https://www.protondb.com/app/2246340?device=pc https://www.reddit.com/r/linux_gaming/comments/1j088lg/denuvo_takes_every_proton_version_as_its_own/?sort=new

Improve Performance

Config changes: https://www.nexusmods.com/monsterhunterwilds/mods/91

Update Dstorage: https://www.nexusmods.com/monsterhunterwilds/mods/127?tab=posts

Copying the original 1.2.2 files to system32 in prefix (/home/USERNAME/.steam/steam/steamapps/compatdata/2246340/pfx/drive_c/windows/system32)

Reframework and related mods: https://www.nexusmods.com/monsterhunterwilds/mods/93 https://www.nexusmods.com/monsterhunterwilds/mods/221?tab=description https://www.nexusmods.com/monsterhunterwilds/mods/234

Reshade after disabling all effects: https://www.nexusmods.com/monsterhunterwilds/mods/594

Press Home to bring up ReShade window, go to settings, add a new Textures path /path/to/mod/reshade-shaders/Textures/. Make sure this path is on top of all other textures path.

Enable lut effect for this shader to take effect.

Launch options

VKD3D_DISABLE_EXTENSIONS=VK_NV_low_latency2 HOST_LC_ALL=zh_TW.UTF-8 WINEDLLOVERRIDES="d3dcompiler_47=n;dxgi,dinput8, dstorage, dstoragecore=n, b" MANGOHUD_CONFIG=fps_limit=60,no_display mangohud OBS_VKCAPTURE=1 %command%

VKD3D_DISABLE_EXTENSIONS=VK_NV_low_latency2 seems resolve random crash, HOST_LC_ALL is required to input Chinese after enabling reframework, use mangohud to limit fps because the in game fps limiter seems broken (source)

Raw
kvm-qemu-gpu-passthrough.md

GPU: AMD Radeon 7700S

Prerequisite

ujust dx-group

reboot once.

Bind GPU to vfio-pci

Based on https://github.com/FilBot3/fedora-silverblue-gpu-passthrough and my past attempt https://gist.github.com/fsworld009/0a7240c4eab9254aca3f7a334acc1aa9#file-framework-16-gpu-passthrough-md

sudo rpm-ostree kargs \
  --append-if-missing="amd_iommu=on" \
  --append-if-missing="iommu=pt" \
  --append-if-missing="rd.driver.pre=vfio_pci" \
  --append-if-missing="vfio-pci.ids=1002:7480,1002:ab30" \
  --append-if-missing="vfio_iommu_type1.allow_unsafe_interrupts=1"
  
  
sudo rpm-ostree initramfs \
  --enable \
  --arg="--add-drivers" \
  --arg="vfio-pci"

Revert

In case we don't want this anymore...

sudo rpm-ostree kargs \
  --delete-if-present="amd_iommu=on" \
  --delete-if-present="iommu=pt" \
  --delete-if-present="rd.driver.pre=vfio_pci" \
  --delete-if-present="vfio-pci.ids=1002:7480,1002:ab30" \
  --delete-if-present="vfio_iommu_type1.allow_unsafe_interrupts=1"

sudo rpm-ostree initramfs --disable

Create subvolume to store vmdisks

sudo btrfs subvolume create vmdisks
sudo chown -R fsworld009:fsworld009 vmdisks
# apply no cow attribute to avoid bad performance for disk image files
chtttr +C vmdisks
# try to create a file in this folder
lsattr vmdisks # should have 'C' property to files in the output

Permission denied when using disk images on external mounts / disks

According to https://unix.stackexchange.com/a/497605, need to add

sudo chmod a+x /media/${whoami}/

Bazzite guest

There are no free USB channels

Need to add more USB redirector and reboot

Sunshine server (for streaming video back to host)

ujust setup-sunshine

and reboot.

Although the host machine cannot access it, it seems we need to setup bridged network https://ivonblog.com/posts/qemu-kvm-bridged-networking/

Stopped for now as there are too many restrictions for this approach, and clipboard sharing from Spice stop working when not using QXL video

mount host share folders

(Assume data is a vfiofs shared folder)

export DBX_SUDO_PROGRAM="pkexec"
sudo mkdir -p /run/media/fsworld009
sudo mkdir -p /run/media/fsworld009/data
sudo mount -t virtiofs data /run/media/fsworld009/data

Windows guest

spice guest tools https://www.spice-space.org/download/windows/spice-guest-tools/spice-guest-tools-latest.exe for clipboard sharing

vfio guest tools https://fedorapeople.org/groups/virt/virtio-win/direct-downloads/latest-virtio/virtio-win-guest-tools.exe guide https://virtio-fs.gitlab.io/howto-windows.html

Raw
obs-flatpak.md

OBS Flatpak on Wayland

Hardware support

Flatpak version has built-in hardware support for AMD GPU.

  1. Settings -> Output, change to Advanced mode
  2. Video Encoder: FFmpeg VAAPI H.264
  3. Encoder Settings: Set gpu in VAAPI Device, set bitrate in "Bitrate"

Source: https://www.reddit.com/r/linux_gaming/comments/1evpfwz/hardware_accelerated_video_encoding_amd/

Global hotkey

Currently this is not supported, need external script as workaround https://gitlab.com/AndrewShark/obs-scripts/-/tree/main/obs-wayland-shortcuts?ref_type=heads

Pending https://ideas.obsproject.com/posts/2066/implement-globalshortcuts-portal

Enforce projector window to a specific size

Usage: view input capture card source in the correct window size / ratio (1080p)

Method: utilize KDE window rule

image

  1. Window Class (application): Exact match, com.obsproject.Studio
  2. Match whole window class: No
  3. Window types: Normal window
  4. Window title: Exact match, Windowd Projector (Source) - ${input_capture_card_source_name}
  5. Size: Force, 1922x1109 (for 1080p)

Twitch

Video maximum bitrate: 6000kbps Audio max bitrate: 320kbps (160 per channel)

Plugins

obs-vkcapture

For game capture on Linux https://github.com/nowrep/obs-vkcapture

Has native flatpak support, see REAMME.

obs-pipewire-audio-capture

For per-application audio capture https://github.com/dimtpap/obs-pipewire-audio-capture

Need to manually copy files to flatpak installation, see https://github.com/dimtpap/obs-pipewire-audio-capture?tab=readme-ov-file#installation

obs-multi-rtmp

For streaming to multipole platforms https://github.com/sorayuki/obs-multi-rtmp

Download -linux-gnu.deb version, copy obs-multi-rtmp folder in archive to ~/.var/app/com.obsproject.Studio/config/obs-studio/plugins/

obs-move-transition

Support adding filters to move sources based on condition https://github.com/exeldro/obs-move-transition

Has flatpak build, see https://github.com/exeldro/obs-move-transition

obs-branch-output

For recording sources separately https://github.com/OPENSPHERE-Inc/branch-output

Download x86_64-linux-gnu.deb version and extract data.tar

  1. Copy usr/lib/x86_64-linux-gnu/obs-plugins/osi-branch-output.so to ~/.var/app/com.obsproject.Studio/config/obs-studio/plugins/osi-branch-output/bin/64bit/
  2. Copy usr/share/obs/obs-plugins/osi-branch-output/locale to ~/.var/app/com.obsproject.Studio/config/obs-studio/plugins/osi-branch-output/data/locale
Raw
Regressions.md

dGPU not visible

https://community.frame.work/t/graphics-card-not-available/66269/31 https://gitlab.freedesktop.org/drm/amd/-/issues/4083#note_2839425

Last working version: Kernel 6.13.5

VRam calculation error

ValveSoftware/Proton#8206 (comment) https://gitlab.freedesktop.org/mesa/mesa/-/commit/a565f2994fec38e20cce46589698dd5487e3ea61

Last working version: mesa 24.X, fixed in 25.0.2 ValveSoftware/Proton#8206 (comment)

Current working Aurora version: https://github.com/ublue-os/aurora/releases/tag/stable-20250302

To rebase to older version manually:

rpm-ostree rebase ostree-image-signed:docker://ghcr.io/ublue-os/aurora-dx:41-20250302

Might need to check if there is -1 -2 tags..etc at https://github.com/ublue-os/bluefin/pkgs/container/aurora-dx/versions?filters%5Bversion_type%5D=tagged&page=8

ujust rebase-helper would also help, but it only lists about last 15 versions by date.

Ref: https://docs.bazzite.gg/Installing_and_Managing_Software/Updates_Rollbacks_and_Rebasing/rebase_guide/

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment