Server Version: 1.21.4.4079-1b7748a7b
Player Version: any
Hey guys.
I bought an Intel NUC 11 Performance kit recently, with the “Tiger Lake” 11th gen Core i7-1165G7 and Iris Xe iGPU (successor of UHD Graphics). This iGPU has really wide Hardware Accelerated video transcoding support and I was expecting this NUC to replace the embedded Plex Media Server of my Nvidia Shield.
Unfortunately, I can’t make it run under Ubuntu Server 20.10 “Groovy”: I ran into strange issues, like fan stopping, overheating, reported to Intel.
I tuned a bit the BIOS to be sure the fan would have enough speed (I have more confidence in my settings than in Intel BIOS legendary reliability) then I decide to return to a Ubuntu LTS distro instead.
I tried Ubuntu Server 20.04 LTS “Focal Fossa”, from the 5.8 generic kernel, to the 5.8 HWE kernel and the 5.10 OEM kernel, with Intel non-free packages from here.
I had bad crashes, like the following one:
[ 2077.553653] i915 0000:00:02.0: [drm] Resetting vcs1 for preemption time out
[ 2077.554371] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2077.555710] i915 0000:00:02.0: [drm] GPU HANG: ecode 12:10:4ab6fff5, in Plex Transcoder [6235]
[ 2077.555712] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[ 2077.555713] Please file a _new_ bug report at https://gitlab.freedesktop.org/drm/intel/issues/new.
[ 2077.555714] Please see https://gitlab.freedesktop.org/drm/intel/-/wikis/How-to-file-i915-bugs for details.
[ 2077.555715] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[ 2077.555715] The GPU crash dump is required to analyze GPU hangs, so please always attach it.
[ 2077.555716] GPU crash dump saved to /sys/class/drm/card0/error
[ 2088.914150] i915 0000:00:02.0: [drm] GPU HANG: ecode 12:10:4ab6fff5, in Plex Transcoder [6235]
[ 2089.015206] i915 0000:00:02.0: [drm] Resetting vcs1 for stopped heartbeat on vcs1
[ 2089.015913] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.016065] i915 0000:00:02.0: [drm] Resetting chip for stopped heartbeat on vcs1
[ 2089.119928] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.120701] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.121996] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.141356] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.142140] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.143438] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.169411] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.170179] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.171445] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.209398] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.210168] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.211440] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.211994] i915 0000:00:02.0: [drm] *ERROR* Failed to reset chip
[ 2089.212098] i915 0000:00:02.0: [drm:add_taint_for_CI [i915]] CI tainted:0x9 by intel_gt_reset+0x221/0x290 [i915]
[ 2089.315925] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.316689] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.317957] i915 0000:00:02.0: [drm] *ERROR* vcs1 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 2089.318539] i915 0000:00:02.0: [drm] Plex Transcoder[6235] context reset due to GPU hang
[ 2090.242782] show_signal_msg: 1 callbacks suppressed
[ 2090.242785] Plex Media Serv[6171]: segfault at 0 ip 0000000000000000 sp 00007f3e90ff0378 error 14 in Plex Media Server[200000+80c000]
I followed all material I found on Intel Website or Plex website to have proper HW support of this iGPU, but I’m running out of ideas.
I put the same packages on a Live Ubuntu Server installation, booted up my Dell XPS 7590 (which has Intel Graphics UHD 630), and Plex HW transcoding works flawless on my laptop.
If someone managed to get the NUC 11 work with Plex, I would be please to know how.
Thanks !