On 11/9/2021 at 7:17 PM, Blobbonator said:

Just fyi - on my 12600k i can see using /dmesg

I will try it out with the suggestions in the message in the coming days but maybe alder lake needs an updated linux kernel we'll see

 

edit: tried to use "i915.force_probe=4680" but didn't work for me

Any chance you found a solution for this? FWIW, I'm running 6.10-rc2, but still the same issue.

 

[   23.505832] i915 0000:00:02.0: Your graphics device 4680 is not properly supported by the driver in this
               kernel version. To force driver probe anyway, use i915.force_probe=4680
               module parameter or CONFIG_DRM_I915_FORCE_PROBE=4680 configuration option,
               or (recommended) check for kernel updates.

As you are on rc2.

 

either add to syslinux 

 

or add to i915.conf in modprobed.

options i915 force_probe=4680

 

or blacklist i915 and load intel gpu top as per this post.

 

 

2 hours ago, SimonF said:

As you are on rc2.

 

either add to syslinux 

 

or add to i915.conf in modprobed.

options i915 force_probe=4680

 

or blacklist i915 and load intel gpu top as per this post.

 

 

Got it. Thanks!

 

I will inevitably forget about making these changes. Will this effect anything in the future?

8 hours ago, flyize said:

Got it. Thanks!

 

I will inevitably forget about making these changes. Will this effect anything in the future?

Shouldn't do. Native support is due to be included in Kernel 5.16 from what I have read.

18 hours ago, flyize said:

Will this effect anything in the future?

Depends what you've done, if you've installed the Intel GPU TOP plugin it will handle everything for you if you've added the file manually it should also not harm your system.

So transcoding seems to be 'soft' crashing my server. It responds to pings but nothing else. syslog shows this:

 

Jan 16 16:39:12 Truffle kernel: i2c /dev entries driver
Jan 16 17:10:06 Truffle kernel: i915 0000:00:02.0: [drm] Resetting vcs0 for preemption time out
Jan 16 17:10:06 Truffle kernel: i915 0000:00:02.0: [drm] GPU HANG: ecode 12:4:28fffffd, in Plex Transcoder [8854]

 

Any ideas?

On 1/17/2022 at 9:17 AM, flyize said:

So transcoding seems to be 'soft' crashing my server. It responds to pings but nothing else. syslog shows this:

 

Jan 16 16:39:12 Truffle kernel: i2c /dev entries driver
Jan 16 17:10:06 Truffle kernel: i915 0000:00:02.0: [drm] Resetting vcs0 for preemption time out
Jan 16 17:10:06 Truffle kernel: i915 0000:00:02.0: [drm] GPU HANG: ecode 12:4:28fffffd, in Plex Transcoder [8854]

 

Any ideas?

Anyone even got a guess as to what this could be?

 

I've recently upgrading the server hardware and am running 6.10.-rc2. Server works perfectly, as long as I don't turn on hardware transcoding in Plex. Also, I can't imagine that it matters, but I'm using binhex-plex.

Edited January 20, 20224 yr by flyize

I got hw transcoding working with my i5-11600 on H264 materials, but not on h265. Is there some additional step that has to be done for it to work with plex?

 

CPU is i5-11600

Edited January 29, 20224 yr by ProphetSe7en

Exactly what I was after! Thank you!

On 1/28/2022 at 7:51 PM, ProphetSe7en said:

I got hw transcoding working with my i5-11600 on H264 materials, but not on h265. Is there some additional step that has to be done for it to work with plex?

 

CPU is i5-11600

 

I seem to be having the same issue with an 11500T, h264 starts up immediately and shows hw encoding.  When I try to do the same on an h265 file it shows it starts and shows hw encoding but the client just sits there spinning.  It looks like the cpu has one thread that goes to 100% while the gpu shows no load.

 

edit-Just figured it out, uncheck Use hardware accelerated video encoding but leave Use hardware acceleration when available checked.  Immediately worked after doing this and cpu load is <20% (4 threads being used for VM so not necessarily just from transcoding) while gpu video load is ~8%.

Edited February 28, 20224 yr by Dural

So what it appears to be doing is using the QS hardware for decoding but the CPU for encoding.  If I turn the hardware encoding on it will work for anything h264 but h265 just spins, plex shows it's transcoding but nothing happens and 1 CPU thread is maxed out.  Issue with the current drivers for 11th gen hardware?

Is there a guide to get HW transcoding working with a 10th series iGPU on 6.9.2?

 

I have Intel Top and gtv-g installed and the igpu passe through to plex, but i am still unable to get it to hardware transcode

 

 

I'm using the hotio container, if that matters

hello,

i am using unraid 6.9.2

with intel core i5-10400 processor

I am using the HOTIO plex container.

Hardware transcoding only works on 720 or 1080p videos, all 4K files are transcoded by the CPU. I don't know what to do, can someone help me?

 

 

 

and it's not the HEVC the problem:

 

Edited March 11, 20224 yr by totoleouf

On 3/11/2022 at 12:53 PM, totoleouf said:

hello,

i am using unraid 6.9.2

with intel core i5-10400 processor

I am using the HOTIO plex container.

Hardware transcoding only works on 720 or 1080p videos, all 4K files are transcoded by the CPU. I don't know what to do, can someone help me?

 

 

 

and it's not the HEVC the problem:

 

 

You never transcode 4k, regardless if it's by CPU or iGPU. It just uses way too much resources. Always Direct Play. If you need to transcode 4k to play it, maybe you shouldn't be collecting it in the first place.

On 3/30/2022 at 11:12 PM, kalfun said:

 

You never transcode 4k, regardless if it's by CPU or iGPU. It just uses way too much resources. Always Direct Play. If you need to transcode 4k to play it, maybe you shouldn't be collecting it in the first place.

 

It worked before, I even did a tutorial on unraid (lien), I already did 6 simultaneous 4K>SD transcoding with 100gb files.
So no, it's not a resource problem!
And I think, the problem comes from PLEX because by disabling HDR MAPPING it works.

I am again transcoding several 4K sources with igpu and the processor is not very busy!

 

 

 

Just wanted to say thanks to OP for a very simple guide that worked flawlessly first time. Gone from 100% CPU usage for one transcode to less than 20%, on an i3-8100T. Transcode video quality is great compared to i5 2500k!

On 4/1/2022 at 6:28 PM, totoleouf said:

 

It worked before, I even did a tutorial on unraid (lien), I already did 6 simultaneous 4K>SD transcoding with 100gb files.
So no, it's not a resource problem!
And I think, the problem comes from PLEX because by disabling HDR MAPPING it works.

I am again transcoding several 4K sources with igpu and the processor is not very busy!

 

 

 

any solution? having the same problem and similar setup - active tone mapping will use the cpu and not the igpu

Anyone has tried with an intel 11700k. ?


Thx

Envoyé de mon Pixel 5 en utilisant Tapatalk

you have to wait for an unraid update I think, on a discord someone with an intel 12th processor had problems, passing on the beta update of unraid it works, but the version is not stable

 

On 1/17/2022 at 3:17 PM, flyize said:

So transcoding seems to be 'soft' crashing my server. It responds to pings but nothing else. syslog shows this:

 

Jan 16 16:39:12 Truffle kernel: i2c /dev entries driver
Jan 16 17:10:06 Truffle kernel: i915 0000:00:02.0: [drm] Resetting vcs0 for preemption time out
Jan 16 17:10:06 Truffle kernel: i915 0000:00:02.0: [drm] GPU HANG: ecode 12:4:28fffffd, in Plex Transcoder [8854]

 

Any ideas?

I am having the very same issue. Did you find a solution for this yet?

4 hours ago, jsrk said:

I am having the very same issue. Did you find a solution for this yet?

Nope. Its a known issue right now.

On 4/30/2022 at 6:04 PM, shiz said:

any solution? having the same problem and similar setup - active tone mapping will use the cpu and not the igpu

 

I'm having the same issue, 4K HW tone mapping on my i3 10300 worked before, but now its broke

Edited May 14, 20224 yr by macieksoft

23 hours ago, flyize said:

Nope. Its a known issue right now.

 

20 hours ago, jsrk said:

 

I'm using Alder Lake.