[GH-ISSUE #498] After installing a new AMD GPU and setting render node to it, moonlight crashes on connection #303

New issue

Closed

opened 2026-10-15 12:59:28 +00:00 by bsakaguchi · 3 comments

bsakaguchi commented

2026-10-15 12:59:28 +00:00

Owner

Originally created by @one-moment-please on GitHub (Sep 5, 2026).
Original GitHub issue: https://github.com/games-on-whales/wolf/issues/498

Basically I had everything working when i just has a Intel Arc A380 gpu, however that gpu is not sufficient for gaming. I installed an AMD RX 5700 XT and installed the driver package and updated the render node in the compose file to it but now i can't seem to connect to the server. Switching back to the intel gpu as the render node everything works, but performance is hampered.

servarr@archie
--------------
OS: Debian GNU/Linux 13 (trixie) x86_64
Host: X99-D3
Kernel: Linux 6.12.107+deb13-amd64
Uptime: 8 hours, 19 mins
Packages: 863 (dpkg)
Shell: bash 5.2.37
Cursor: Adwaita
Terminal: /dev/pts/1
CPU: Intel(R) Xeon(R) E5-2696 v3 (36) @ 3.80 GHz
GPU 1: AMD Radeon RX 5700 XT [Discrete]
GPU 2: Intel Arc A380 @ 2.45 GHz [Discrete]
Memory: 6.97 GiB / 125.70 GiB (6%)
Swap: 3.09 GiB / 24.02 GiB (13%)
Disk (/): 204.92 GiB / 432.80 GiB (47%) - ext4
Disk (/mnt/backup): 172.30 GiB / 915.82 GiB (19%) - ext4
Disk (/mnt/dsk/1): 6.62 TiB / 12.73 TiB (52%) - xfs
Disk (/mnt/dsk/2): 6.72 TiB / 12.73 TiB (53%) - xfs
Disk (/mnt/dsk/3): 6.83 TiB / 12.73 TiB (54%) - xfs
Disk (/mnt/dsk/4): 6.68 TiB / 12.73 TiB (53%) - xfs
Disk (/mnt/dsk/5): 8.71 TiB / 14.55 TiB (60%) - xfs
Disk (/mnt/dsk/6): 9.07 TiB / 14.55 TiB (62%) - xfs
Disk (/mnt/ext): 1.03 TiB / 1.82 TiB (56%) - xfs
Disk (/mnt/games): 39.47 GiB / 219.00 GiB (18%) - ext4
Disk (/mnt/wrk): 456.02 GiB / 953.40 GiB (48%) - xfs
Local IP (bridge0): 192.168.0.2/24
Locale: en_CA.UTF-8

services:
  wolf:
    image: ghcr.io/games-on-whales/wolf:stable
    container_name: wolf
    volumes:
      - /mnt/games/wolf:/etc/wolf
      - /var/run/docker.sock:/var/run/docker.sock:rw
      - /dev/:/dev/:rw
      - /run/udev:/run/udev:rw
    device_cgroup_rules:
      - 'c 13:* rmw'
    environment:
      - WOLF_RENDER_NODE=/dev/dri/renderD129
    devices:
      - /dev/dri
      - /dev/uinput
      - /dev/uhid
    network_mode: host
    restart: unless-stopped

18:24:36.187489937 ERROR | [GSTREAMER] Pipeline error: Internal data stream error.
18:24:36.192001376 WARN  | [DOCKER] Failed to get major numbers for hidraw and input
8:09:12.576050206   220 0x7fefc0007160 WARN               interpipe gstinterpipe.c:236:gst_inter_pipe_leave_node_priv: Node 1796920474626130326_video not found. Could not leave node.
18:24:36.545292953 INFO  | [DOCKER] Starting container: /WolfXFCE_1796920474626130326
18:24:40.337690294 INFO  | Stopped container: /WolfXFCE_1796920474626130326
18:24:40.382086431 INFO  | [GSTREAMER] Pipeline reached End Of Stream
18:24:42.341139265 INFO  | [GSTREAMER] Pipeline reached End Of Stream
8:09:18.721687950   220 0x7fefe801b290 WARN                audiosrc gstaudiosrc.c:227:audioringbuffer_thread_func:<pulsesrc2> error reading data -1 (reason: Success), skipping segment

Originally created by @one-moment-please on GitHub (Sep 5, 2026). Original GitHub issue: https://github.com/games-on-whales/wolf/issues/498 Basically I had everything working when i just has a Intel Arc A380 gpu, however that gpu is not sufficient for gaming. I installed an AMD RX 5700 XT and installed the driver package and updated the render node in the compose file to it but now i can't seem to connect to the server. Switching back to the intel gpu as the render node everything works, but performance is hampered. <img width="621" height="359" alt="Image" src="https://github.com/user-attachments/assets/bcd7d943-339d-42f6-8699-87fb352da2d6" /> ``` servarr@archie -------------- OS: Debian GNU/Linux 13 (trixie) x86_64 Host: X99-D3 Kernel: Linux 6.12.107+deb13-amd64 Uptime: 8 hours, 19 mins Packages: 863 (dpkg) Shell: bash 5.2.37 Cursor: Adwaita Terminal: /dev/pts/1 CPU: Intel(R) Xeon(R) E5-2696 v3 (36) @ 3.80 GHz GPU 1: AMD Radeon RX 5700 XT [Discrete] GPU 2: Intel Arc A380 @ 2.45 GHz [Discrete] Memory: 6.97 GiB / 125.70 GiB (6%) Swap: 3.09 GiB / 24.02 GiB (13%) Disk (/): 204.92 GiB / 432.80 GiB (47%) - ext4 Disk (/mnt/backup): 172.30 GiB / 915.82 GiB (19%) - ext4 Disk (/mnt/dsk/1): 6.62 TiB / 12.73 TiB (52%) - xfs Disk (/mnt/dsk/2): 6.72 TiB / 12.73 TiB (53%) - xfs Disk (/mnt/dsk/3): 6.83 TiB / 12.73 TiB (54%) - xfs Disk (/mnt/dsk/4): 6.68 TiB / 12.73 TiB (53%) - xfs Disk (/mnt/dsk/5): 8.71 TiB / 14.55 TiB (60%) - xfs Disk (/mnt/dsk/6): 9.07 TiB / 14.55 TiB (62%) - xfs Disk (/mnt/ext): 1.03 TiB / 1.82 TiB (56%) - xfs Disk (/mnt/games): 39.47 GiB / 219.00 GiB (18%) - ext4 Disk (/mnt/wrk): 456.02 GiB / 953.40 GiB (48%) - xfs Local IP (bridge0): 192.168.0.2/24 Locale: en_CA.UTF-8 ``` ``` services: wolf: image: ghcr.io/games-on-whales/wolf:stable container_name: wolf volumes: - /mnt/games/wolf:/etc/wolf - /var/run/docker.sock:/var/run/docker.sock:rw - /dev/:/dev/:rw - /run/udev:/run/udev:rw device_cgroup_rules: - 'c 13:* rmw' environment: - WOLF_RENDER_NODE=/dev/dri/renderD129 devices: - /dev/dri - /dev/uinput - /dev/uhid network_mode: host restart: unless-stopped ``` ``` 18:24:36.187489937 ERROR | [GSTREAMER] Pipeline error: Internal data stream error. 18:24:36.192001376 WARN | [DOCKER] Failed to get major numbers for hidraw and input 8:09:12.576050206 220 0x7fefc0007160 WARN interpipe gstinterpipe.c:236:gst_inter_pipe_leave_node_priv: Node 1796920474626130326_video not found. Could not leave node. 18:24:36.545292953 INFO | [DOCKER] Starting container: /WolfXFCE_1796920474626130326 18:24:40.337690294 INFO | Stopped container: /WolfXFCE_1796920474626130326 18:24:40.382086431 INFO | [GSTREAMER] Pipeline reached End Of Stream 18:24:42.341139265 INFO | [GSTREAMER] Pipeline reached End Of Stream 8:09:18.721687950 220 0x7fefe801b290 WARN audiosrc gstaudiosrc.c:227:audioringbuffer_thread_func:<pulsesrc2> error reading data -1 (reason: Success), skipping segment ```

bsakaguchi closed this issue

2026-10-15 12:59:28 +00:00

bsakaguchi commented

2026-10-15 12:59:29 +00:00

Author

Owner

@one-moment-please commented on GitHub (Sep 6, 2026):

Edit: I managed to get it to start by commenting out the lines with "vah264lpenc" as the clankers had divined that wolf was not correctly detecting that low power encoding is not available on the 5700 XT

#    [[gstreamer.video.h264_encoders]]
#    check_elements = [ 'vah264lpenc', 'vapostproc' ]
#    encoder_pipeline = '''vah264lpenc aud=false b-frames=0 ref-frames=1 num-slices={slices_per_frame} bitrate={bitrate} cpb-size={vbv_buffer_size} min-qp=20 key-int-max=1024 rate-c>
#h264parse !
#video/x-h264, profile=main, stream-format=byte-stream'''
#    plugin_name = 'va'

So now it starts and runs, but it is a really, really choppy framerate and I'm trying to troubleshoot that as both the server and my laptop are completely hardwired and I can't see the slowdowns. I tried to set it to use the Intel GPU for transcoding but it just keeps crashing when I set the encoder node:

services:
  wolf:
    image: ghcr.io/games-on-whales/wolf:stable
    container_name: wolf
    volumes:
      - /mnt/games/wolf:/etc/wolf
      - /var/run/docker.sock:/var/run/docker.sock:rw
      - /dev/:/dev/:rw
      - /run/udev:/run/udev:rw
    device_cgroup_rules:
      - 'c 13:* rmw'
    environment:
      - WOLF_RENDER_NODE=/dev/dri/renderD129
#      - WOLF_ENCODER_NODE=/dev/dri/renderD128
      - WOLF_LOG_LEVEL=DEBUG
    devices:
      - /dev/dri:/dev/dri
      - /dev/uinput:/dev/uinput
      - /dev/kfd:/dev/kfd
      - /dev/uhid:/dev/uhid
    network_mode: host
    restart: unless-stopped

@one-moment-please commented on GitHub (Sep 6, 2026): Edit: I managed to get it to start by commenting out the lines with "vah264lpenc" as the clankers had divined that wolf was not correctly detecting that low power encoding is not available on the 5700 XT ``` # [[gstreamer.video.h264_encoders]] # check_elements = [ 'vah264lpenc', 'vapostproc' ] # encoder_pipeline = '''vah264lpenc aud=false b-frames=0 ref-frames=1 num-slices={slices_per_frame} bitrate={bitrate} cpb-size={vbv_buffer_size} min-qp=20 key-int-max=1024 rate-c> #h264parse ! #video/x-h264, profile=main, stream-format=byte-stream''' # plugin_name = 'va' ``` So now it starts and runs, but it is a really, really choppy framerate and I'm trying to troubleshoot that as both the server and my laptop are completely hardwired and I can't see the slowdowns. I tried to set it to use the Intel GPU for transcoding but it just keeps crashing when I set the encoder node: ``` services: wolf: image: ghcr.io/games-on-whales/wolf:stable container_name: wolf volumes: - /mnt/games/wolf:/etc/wolf - /var/run/docker.sock:/var/run/docker.sock:rw - /dev/:/dev/:rw - /run/udev:/run/udev:rw device_cgroup_rules: - 'c 13:* rmw' environment: - WOLF_RENDER_NODE=/dev/dri/renderD129 # - WOLF_ENCODER_NODE=/dev/dri/renderD128 - WOLF_LOG_LEVEL=DEBUG devices: - /dev/dri:/dev/dri - /dev/uinput:/dev/uinput - /dev/kfd:/dev/kfd - /dev/uhid:/dev/uhid network_mode: host restart: unless-stopped ```

bsakaguchi commented

2026-10-15 12:59:29 +00:00

Author

Owner

@ABeltramo commented on GitHub (Sep 6, 2026):

This is a known issue with multiple AMD/Intel GPUs; see if this helps: https://github.com/games-on-whales/wolf/issues/225#issuecomment-3585056921

@ABeltramo commented on GitHub (Sep 6, 2026): This is a known issue with multiple AMD/Intel GPUs; see if this helps: https://github.com/games-on-whales/wolf/issues/225#issuecomment-3585056921

bsakaguchi commented

2026-10-15 12:59:29 +00:00

Author

Owner

@one-moment-please commented on GitHub (Sep 7, 2026):

I think that fixed it. I was't able to test right away as I had tried to test the gpu by passing it to a windows vm guest and that ended up crashing the libvirtd service and rendering the AMD gpu inaccessible until i rebooted - which I had to schedule for when the server wasn't being used by anything else.

It's not perfect but it's playable for the types of games i enjoy which is enough, I suppose if I wanted more I would've jst made a proper seperate gaming computer

@one-moment-please commented on GitHub (Sep 7, 2026): I think that fixed it. I was't able to test right away as I had tried to test the gpu by passing it to a windows vm guest and that ended up crashing the libvirtd service and rendering the AMD gpu inaccessible until i rebooted - which I had to schedule for when the server wasn't being used by anything else. It's not perfect but it's playable for the types of games i enjoy which is enough, I suppose if I wanted more I would've jst made a proper seperate gaming computer