Closed Yarn closed 1 year ago
Could you try again with new images?
Could you try again with new images?
That worked for me, had same issue.
I'm still getting a black screen with no obvious error message. I assume these images have NEKO_HWENC: VAAPI
by default since it happens with or without that setting now.
It has been changed to use by default, you can turn it off using NEKO_HWENC: none
.
You could set gst debugging to more verbose to see why its failing, e.g. GST_DEBUG=*:4
adding that I get some output about Xorg crashing
The error at the end of the log file /home/neko/.local/share/xorg/Xorg.99.0.log
inside the container
[761391.567] (II) LoadModule: "void"
[761391.567] (WW) Warning, couldn't open module void
[761391.567] (EE) Failed to load module "void" (module does not exist, 0)
[761391.567] (EE) No input driver matching `void'
[761391.567] (II) Falling back to input driver `libinput'
[761391.567] (II) LoadModule: "libinput"
[761391.567] (WW) Warning, couldn't open module libinput
[761391.567] (EE) Failed to load module "libinput" (module does not exist, 0)
[761391.567] (II) LoadModule: "void"
[761391.567] (WW) Warning, couldn't open module void
[761391.567] (EE) Failed to load module "void" (module does not exist, 0)
[761391.567] (EE) No input driver matching `void'
[761391.567] (II) Falling back to input driver `libinput'
[761391.567] (II) LoadModule: "libinput"
[761391.567] (WW) Warning, couldn't open module libinput
[761391.567] (EE) Failed to load module "libinput" (module does not exist, 0)
On my intel n100, the VAAPI method doesn't work in any way. I'm using VAAPI fine on the media server, but on the n.eko I just get a black screen.
neko:
image: ghcr.io/m1k1o/neko/intel-chromium:latest
restart: "unless-stopped"
shm_size: "2gb"
ports:
- "8090:8080"
- "52000-52100:52000-52100/udp"
cap_add:
- SYS_ADMIN
privileged: true
environment:
NEKO_SCREEN: '1920x1080@30'
NEKO_PASSWORD: xxxx
NEKO_PASSWORD_ADMIN: xxxx
NEKO_EPR: 52000-52100
NEKO_NAT1TO1: 192.168.31.xx
NEKO_HWENC: VAAPI
devices:
- /dev/dri:/dev/dri
I'm also still facing this issue with intel-firefox.
version: "3.4"
services:
neko:
image: "ghcr.io/m1k1o/neko/intel-firefox:latest"
restart: "unless-stopped"
shm_size: "2gb"
environment:
NEKO_SCREEN: 1920x1080@30
NEKO_PASSWORD: neko
NEKO_PASSWORD_ADMIN: <redacted>
NEKO_EPR: 52000-52100
NEKO_ICELITE: 1
NEKO_NAT1TO1: <redacted>
NEKO_HWENC: VAAPI
LIBVA_DRIVER_NAME: i915
devices:
- /dev/dri:/dev/dri
- /dev/fb0:/dev/fb0
volumes:
- '${PWD}/data:/home/neko/.mozilla/firefox/'
- '${PWD}/policies.json:/usr/lib/firefox/distribution/policies.json'
container_name: neko
network_mode: "service:gluetun"
Running vainfo
within the container returns:
libva info: VA-API version 1.10.0
libva info: User environment variable requested driver 'i915'
libva info: Trying to open /usr/lib/x86_64-linux-gnu/dri/i915_drv_video.so
libva info: va_openDriver() returns -1
vaInitialize failed with error code -1 (unknown libva error),exit
I'm running docker on an LXD container but I think I have the gpu fully passed through. The /dev/dri folder is visible within the docker container
I tried VLC and firefox with the same result, I don't see anything helpful in the logs for either. These are the logs for VLC. neko.log
I installed some packages to try to get things to work, one of the pci ones got rid of an error (not sure which) but none of them changed the result.