Is it better to install driver on the server then forward /dev/dri to the container or install drivers straight into the container ?
Because all I’ve done is install drivers on server than forward /dev/dri and this
echo “options i915 enable_guc=2” >> /etc/modprobe.d/i915.conf
modprobe i915
and it seems to work very well
Somethings wrong here, pulling the latest linuxserver’ image broke HW HDR tone mapping
It appears to be some sort of a regression in the intel drivers. The linuxserver image installs the latest version of them (released 3 days ago). 21.50.21939 breaks hardware tone mapping. I manually installed the prior version (21.49.21786) in the image and it restored hardware tone mapping.
Indeed, thanks
Downgrading drivers seems to work
Anyone know how we can report this? Obviously an issue can be opened here but I don’t have enough technical knowledge about these drivers to be useful. Can the Plex devs help translate what went wrong in the latest drivers?
Finally got everything working on my NUC 11 TNKi7. I was on Ubuntu Desktop (21.10) and tried various versions of Intel compute runtime. Tone Mapping broken on all instances. Nuke the NUC and install a fresh copy of Ubuntu 20.04 with version 21.49.21786 and voila! everything now works. HW acceleration with tone mapping.
This is definitely still a problem.
I opened an issue here Hardware Accelerated Tonemapping for Tiger Lake broken after 21.49.21786 · Issue #488 · intel/compute-runtime · GitHub
Thanks, indeed still an issue. Is someone from Plex checking this issue at the moment?
Just checking the commits from version to version… seems that between working en non-working version some memory optimalizations have been made. Could it possible be that the memory optimalization broke the drivers?
Do you mind sharing how you managed to downgrade your intel driver?
I don’t have experience doing this and I’m not finding useful information
No problem, first off how do you have it installed?
Do you use docker?
Finding transcoding really unstable after setting up based upon all of the information above.
Ubuntu 20.04 LTS VM (+pci passthrough) - 5.10.0-1057-oem
Docker image - linuxserver/plex:1.25.4
Within container, installed by linuxserver:
intel-igc-cm/now 1.0.97+i643~u20.04 amd64 [installed,local]
intel-opencl-icd/now 21.49.21786+i643~u20.04 amd64 [installed,local]
Plex - 1.25.4.5487
What have I missed 
[Mon Feb 7 19:53:05 2022] i915 0000:00:10.0: [drm] GPU HANG: ecode 12:10:2fffefff, in Plex Transcoder [49480]
[Mon Feb 7 19:53:17 2022] i915 0000:00:10.0: [drm] GPU HANG: ecode 12:10:2fffefff, in Plex Transcoder [49480]
[Mon Feb 7 19:53:17 2022] i915 0000:00:10.0: [drm] Resetting vcs1 for stopped heartbeat on vcs1
[Mon Feb 7 19:53:17 2022] i915 0000:00:10.0: [drm] ERROR vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[Mon Feb 7 19:53:17 2022] i915 0000:00:10.0: [drm] Resetting chip for stopped heartbeat on vcs1
[Mon Feb 7 19:53:17 2022] i915 0000:00:10.0: [drm] ERROR vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[Mon Feb 7 19:53:17 2022] i915 0000:00:10.0: [drm] ERROR vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[Mon Feb 7 19:53:17 2022] i915 0000:00:10.0: [drm] Plex Transcoder[49480] context reset due to GPU hang
[Mon Feb 7 19:53:20 2022] i915 0000:00:10.0: [drm] Resetting vcs1 for preemption time out
[Mon Feb 7 19:53:20 2022] i915 0000:00:10.0: [drm] ERROR vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[Mon Feb 7 19:53:20 2022] i915 0000:00:10.0: [drm] GPU HANG: ecode 12:10:2fffefff, in Plex Transcoder [49480]
If memory serves,
They fixed that in the 5.11 kernel.
The Plex-side changes were implemented back in 1.24.x
Thanks, I’ll drop 5.13 (unless you suggest otherwise?) on tomorrow and see if it makes any difference
Running on Ubuntu VM with iGPU passthrough and seeing no issues:
dpkg -l | grep plexmediaserver
ii plexmediaserver 1.25.1.5286-34f965be8 amd64 Plex organizes all of your personal media so you can easily access and enjoy it.
uname -a
Linux internalplex 5.13.0-27-generic #29-Ubuntu SMP Wed Jan 12 17:36:47 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux
ii intel-igc-core 1.0.8744
ii intel-opencl-icd 21.44.21506
Upgraded to 5.13 (linux-oem-20.04c) and seems to be running smoothly
Thanks guys.
Are you using the desktop or server version of Ubuntu?
Server 21.10
Server 20.04 LTS
Still seeing quite a few regular GPU hangs - am I alone? Installed on a clean VM for testing:
Proxmox VM + PCI GPU passthrough
OS - Ubuntu Server 21.10
Kernel - 5.13.0-28-generic
Plex - PlexMediaServer-1.25.5.5492
lspci output:
00:00.0 Host bridge: Intel Corporation 440FX - 82441FX PMC [Natoma] (rev 02)
00:01.0 ISA bridge: Intel Corporation 82371SB PIIX3 ISA [Natoma/Triton II]
00:01.1 IDE interface: Intel Corporation 82371SB PIIX3 IDE [Natoma/Triton II]
00:01.2 USB controller: Intel Corporation 82371SB PIIX3 USB [Natoma/Triton II] (rev 01)
00:01.3 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ACPI (rev 03)
00:02.0 VGA compatible controller: Device 1234:1111 (rev 02)
00:03.0 Unclassified device [00ff]: Red Hat, Inc. Virtio memory balloon
00:05.0 SCSI storage controller: Red Hat, Inc. Virtio SCSI
00:10.0 VGA compatible controller: Intel Corporation TigerLake-LP GT2 [Iris Xe Graphics] (rev 01)
00:12.0 Ethernet controller: Red Hat, Inc. Virtio network device
00:1e.0 PCI bridge: Red Hat, Inc. QEMU PCI-PCI bridge
00:1f.0 PCI bridge: Red Hat, Inc. QEMU PCI-PCI bridge
dmesg error:
[Sun Feb 13 11:20:51 2022] i915 0000:00:10.0: [drm] Resetting vcs1 for CS error
[Sun Feb 13 11:20:51 2022] i915 0000:00:10.0: [drm] GPU HANG: ecode 12:4:9da89fa7, in Plex Transcoder [1135]
[Sun Feb 13 11:20:51 2022] i915 0000:00:10.0: [drm] Resetting vcs1 for CS error
[Sun Feb 13 11:20:51 2022] i915 0000:00:10.0: [drm] GPU HANG: ecode 12:4:f6fa86e6, in Plex Transcoder [1135]
[Sun Feb 13 11:20:57 2022] i915 0000:00:10.0: [drm] Resetting vcs1 for preemption time out
[Sun Feb 13 11:20:57 2022] i915 0000:00:10.0: [drm] GPU HANG: ecode 12:4:44b279e6, in Plex Transcoder [1135]
Happy to provide further logs, remote access etc. if needed.
Nathan
