For maximum portability, the shebang for bash
scripts should be:
#!/usr/bin/env bash
If, however, your script is written in proper POSIX shell, use:
#!/bin/sh
In any case, though, this must be the first line in a script.
For maximum portability, the shebang for bash
scripts should be:
#!/usr/bin/env bash
If, however, your script is written in proper POSIX shell, use:
#!/bin/sh
In any case, though, this must be the first line in a script.
Sadly, I'm still not seeing the Hardware option show up.
I added the shebang to the first line of the "go" file.
Rebooted unRaid and restarted the Channels Docker, but still nothing.
And the command: ls -alh /dev/dri
still returns:
/bin/ls: cannot access '/dev/dri': No such file or directory
Frustrating, but not the end of the world. Thanks for all the help.
Yes it must be the very first line. That might be why its not working.
I have this as the line after the transcode activation, and it’s fine for me....should I change it?
Yes the shebang must always be the very first line of a shell script.
I put the shebang first, followed by this:
modprobe i915
lsmod | grep i915
Different syntax from what I put in the GO file before, but it also seems to work in activating the Intel Quick Sync driver upon bootup. Should there be a problem in this, or no?
No problem, but the lsmod
command isn't really necessary; that command just lists the installed and activated kernel modules, and the pipe into grep
just filters the output.
so just listing modprobe i915 is all that is truly necessary?!
so just listing modprobe i915 is all that is truly necessary?!
Yes, that should be all that is necessary.
To understand what is actually happening: modprobe
checks if a kernel module is present and enabled, and if not (but it is available on the system) then it enables it and inserts it into the kernel.
Basically, the command modprobe i915
ensures that the Intel graphics driver is installed and active. That should be all that is needed on your end.
Oh snap!
Okay cool. Once I am finished on my test setup, I will create a guide, SPECIFICALLY for Linux N00B's. I think that when deployed properly, this is a very good setup for Channels DVR
So I have narrowed down the syntax of my go file, to look exactly like this:
#!/bin/bash
/usr/local/sbin/emhttp &
modprobe i915
Everything works, including hardware transcoding, thanks to modprobe i915 enabling the Intel Graphics driver.
Is NVIDIA based hardware still only supported on Windows? I have a Xeon CPU w/ a Quadro P400 running Unraid. I'm currently using the Plex Linuxserver.io container with full Nvidia support (without requiring Quicksync support).
Are there any plans to support Linux in the future? Thanks.
NVENC is already supported on Linux.
tmm1, is there a diag command I can run to find out why it's not enabled in my setup?
Also, does NVENC work without the need for /dev/dri/render* as you state earlier in this thread. That's for the i915 but I don't have Quicksync since I have a Xeon CPU.
As far as I'm aware, /dev/dri
access is needed regardless of vendor in order to access the card's hardware. It may not be necessary, but I've heard anecdotally that access to /dev/shm
is also needed (at least in reference to Docker), but I can't personally confirm this.
Looks like I was mistaken about nvenc on Linux. We do not currently have it enabled. It's a little tricky to do in a way that will work across multiple Linux distros
I've updated the instructions on https://getchannels.com/dvr-server/#unraid to include hardware transcoding setup.
any chance you'll be able to do the same for Nvidia transcoding? Thanks.
Is it possible to add support to hardware acceleration on Nvidia cards in Docker? I know that Nvidia does support Docker containers (see the nvidia-docker runtime) and it is also supported in Plex, but I'm not sure if the Channels DVR docker container supports it?
I'd love to see this as well. I don't have quicksync, but transcoding on my nvidia gpu is working really well in docker.
Please email us support@getchannels.com as we have an nvidia test build available but no machine to try it.
Anyway I could test that build for Nvidia transcoding on Linux?