Bug? - High CPU on Plexamp Flatpak on Linux (Fedora 36 / Gnome / x86)

I’m a huge Plexamp fanboy and I like using Plexamp on my Linux Desktop while I’m working. I’m on Fedora 36 on a plain old x86 laptop, using gnome, wayland, pipewire and I’m using the flatpak installer (the AppImage hasn’t worked for me since I switched to distros using pipewire: Ubuntu 22.04 / Fedora 34).

What I’m seeing is a huge amount of CPU usage when I launch the plexamp client on linux. I’ve seen as high as 12 Cores of a 10th gen i7. Here’s a typical htop screen cap during playback. I once counted over 30 Plexamp processes running, I have 16 cores.

This is streaming over WAN, but my Android app works great and video playback on the Plex webui from this same computer work awesome. I’m not sure what I can do, but this causes the fans to spin up and actually makes it hard to enjoy music.

The only troubleshooting I’ve done is a reinstall (no change), attempting to use the AppImage (doesn’t work at all, possibly pipewire related, I’m not sure), deleting the cache (no change), compiling custom kernels tweaking the scheduler and compiling for optimization on Intel CPUs (no change).

Has anyone encountered this or have any ideas?

My setup:
Fedora 36 (5.19.12-200.fc36.x86_64)
Gnome 42 (Wayland / Pipewire)
Intel i7-10875H (System76 / Clevo Laptop)
Nvidia RX 2060 (Using nvidia binary driver v515.76)
Plexamp 4.4.0 (Flatpak v1.12.7)

Here’s a copy of a 60 seconds of my logs (Plexamp Log 10-10-2022 - Pastebin.com) that covers

  • starting the app
  • the current song attempting to play, but not working
  • my pressing the “previous” button
  • playing about a minute of a song

Huh, that definitely doesn’t sound normal, and nothing crazy in logs.

Do you have any ideas of troubleshooting steps I could take? It looks like debugging is already enabled in the logs, so is there any performance profiling tools you could recommend that might help pinpoint where the hangup might be?

Any help is appreciated.

Thank y’all,
Topher

Honestly, not sure. You could try running strace on a process to see what it’s doing, but that’s likely hard to interpret.

It seems like there must be something environmentally different on your side since I haven’t seen this reported before.

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.