Posts by sky42

e123enitan update 1st to a sky42 release with a device that official LE knows
e.g. https://sky42.libreelec.tv/release/11.0/G…6-%23240204.tar and then to official LE

update https://gitlab.freedesktop.org/drm/intel/-/issues/10199
i was ask to reproduce with drm-tip and i did

ah my Intel issue has labels now.

jernej as short as i can

playback screen resolution 3840x2160 @ 23.976 Hz Pixel Depth 12 bit results in HD audio dropout on at least i7-1270p and N100
- Pixel Depth reported by my Denon AVR
- i use only 2160p in all refresh rates
why is long and stated here https://forum.kodi.tv/showthread.php?tid=344912

problem goes away
- playback 2160p @ 24 Hz
- playback 2160p @ 23.976 Hz with max bpc 10 (Denon reports pixel depth 10 bit)

other infos
- max bpc set to 10 it still uses 8 bit pixel depth with e.g. DVD remux
- the GUI comes up at boot in 4K60 with 8 bit

All i did was try the matrix of settings and find out when it is OK. Hints to do what are from smp and from you (removing the need of proptest).
I have no clue whatsoever about GPU kernel drivers. Just stating facts about things i tested and try to create a easy to use workaround.

Quote from chewitt

Pursuing the fix upstream then submitting a backported kernel or intel driver patch will probably be faster. I think you'll find that script is a little too hacky for peoples tastes and i'd be surprised if it gets merged at this stage in the release cycle.

Ofc that is a hacky script because it is a workaround and not a fix. My focus was to make it usable with as many hints as possible.
For more then 5 years i deliver contributions to this project. Do you have the impression i would not follow up and remove the script when there is a fix upstream?
I can not pursue anything on that intel issue. It was days to prepare all the information and create clean infos for that issue. Now it sits there and nobody responsible has acted in any way at all. What would be the way to pursue that? Being pushy and asking when somebody is doing something? That is not very nice and i will not do that. If i could fix it i would have done it. Instead i try the next best thing and try to help users with the same problem.
In your opinion the user of LE should be denied the easier workaround of including this hacky script and instead waiting that somebody will fix it upstream so the it comes to LE. Until then the users are screwed and poking in the dark if they hit the same issue. Buying cables or do other desperate things. From what i saw by researching it i found hints that it is a years old problem. Now i am doing something and being punished for it. That is how i feel.

Quote from chewitt

In context the number of users rocking up here with latest generation Intel hardware is small and in continuous slow in-decline for the last decade. So audience is not large, there is a workaround (albeit forum sourced) available, and 'Yes' adding hacky stuff to the core image matters and is something to be avoided; largely because once workarounds are merged history shows people end up being happy with the workaround and the root cause is never pursued. If you refuse the hack up-front there's a significantly higher probability of the real issue being resolved (tried and tested). Backporting patches merged/submitted upstream is a no-brainer.

So if it is just a bunch of users they can be ignored for the good of the code base? .oO(i am being sarcastic)
Again do you have the impression i will not follow up when Intel fixes that problem?
And yes there is be a easier way for the users: to include my hacky script until Intel fixed it.
The scope is unclear so i did not restrict when it will act, i already had some PCI id filtering in there, but i can only guess which PCI id are problematic.
Your way leads to user download the script and with a new image with the fix it will not be removed and still be there and i have no chance in updating that script, because it is not part of the image. I could not put in the parameters when it is fixed, because i do not know them. And that is better in your opinion?

chewitt What exactly do i need to do that it gets included now? So people with nice and expensive hardware can just use there hardware.

There will be no more ADL10 images and i will delete the existing ones.
Use the solution from post #46.

After some more help and this time from jernej . Thank you.

I have now a script to fix the audio problem for any LE11/12 without extra binary. The same script works on LE11 or 12.
wget https://raw.githubusercontent.com/sky42src/LibreELEC.tv/11.80.5-240204/projects/Generic/filesystem/usr/bin/intel-gpu-max-10-bit.sh && chmod +x intel-gpu-max-10-bit.sh && ./intel-gpu-max-10-bit.sh

I opened a pull request to make the helper script part of official LE.

Generic: added intel-gpu-max-10-bit.sh helper by sky42src · Pull Request #8588 · LibreELEC/LibreELEC.tv

This uses modetest to set max 10 bit to solve…

github.com

added Zotac CI343

MiniX Z100 wakeup from S5 by USB is fixed. I did get a new BIOS (compiled today) after asking 2 days ago in the evening with 1st response yesterday.

openden the issue
https://gitlab.freedesktop.org/drm/intel/-/issues/10199

Put in the kernel command line ipv6.disable=1
for a Pi that goes into /flash/cmdline.txt

Opened a issue with MiniX support for the S5 USB wakeup problem.

MatteN thx Coil whine sound better

other boxes details worth to know

HUNSN BM34
- USB port only on the front
- pre assembled DDR4 not 3200 only 2666 (the box i tested)
- pre assembled SSD M2 SATA (the box i tested)

Zotac CI337
- DDR5 4800

Zotac CI343
- DDR5 4800

KingnovyPC
- no WiFi or BT

MiniX Z100-0db details

note worthy:
- very tiny full aluminium body and they say it will not throttle ever (not yet tested that, but i will)
- 2x nice external WiFi/BT antennas
- with Windows 11 Pro license embedded in BIOS and preinstalled (tested with reinstall W11/W10)
- working wakeup from S5 by USB/WoL
- working wakeup from S3 with USB/WoL
- less then 0.18 Wh per day used 4-8h, no sleep or poweroff
- HDMI / USB ports are upside down

LE11 dmesg, lspci, lsusb, kodi.log
https://paste.libreelec.tv/legal-raven.log

LE12 dmesg, lspci, lsusb, kodi.log
https://paste.libreelec.tv/suitable-sawfish.log

For now that will be my daily driver.

ASUS PN42 details

note worthy:
- with Consumer Infrared RC6: works out of the box with LE and a WMC/MCE remote
- with HDMI-CEC but ONLY Power Control no using TV remote for Kodi
- working wakeup from S5 with IR remote (reacts on MCE power toggle and power on)
- working wakeup from S5 by USB
- working wakeup from S5 by WoL
- wakeup from S3 works too for USB/WoL
- less then 0.18 Wh per day used 4-8h, no sleep or poweroff
- power button is very bright

LE11 dmesg, lspci, lsusb, kodi.log
https://paste.libreelec.tv/complete-tetra.log

LE12 dmesg, lspci, lsusb, kodi.log
https://paste.libreelec.tv/awaited-rhino.log

Sadly my exemplar makes coil whine voltage whistles (hope that is the right word for Spannungspfeifen). I have very good hearing and that is annoying me.

I want to share my tests about Intel N100 as a replacement for my Streacom FC8 Evo i3-8100

tested by me
- ASUS PN42-BBN100MV https://www.amazon.de/dp/B0C53D43JF
- MiniX NEO Z100-0dB (512GB) https://www.amazon.de/dp/B0CPLQWN73

other boxes i did not test, but know of
- Zotac CI337 https://www.zotac.com/us/product/min…7-nano-barebone
- Zotac CI343 https://www.zotac.com/us/product/min…-ci343-barebone
- HUNSN BM34 https://www.amazon.de/dp/B0CHS5RSX8/
- KingnovyPC https://www.amazon.de/dp/B0CNH27KCW/


The PN42 is a very nice box and i used it for about 2 month now. The Z100 i did get 2 days ago and it is very nice too.
Details i will write down later.

I am preparing to write all down that i learned in making working perfectly for my needs.
It was very interesting chasing down this problem Intel Alder Lake 2160p @ 23.976 Hz passthrough HD Audio dropouts (i7-1270p/N100) and i am not yet finished (need to open a bug @ Intel), but have found a workaround.

New workaround script for official LE 11 and 12, written and tested on official LE 11.0.6.

SSH to your box an run
curl https://sky42.libreelec.tv/testing/proptest/install.sh | bash /dev/stdin

now you can run
/storage/.config/activate-adl-10-bit-fix.sh

or be sure it is in
/storage/.config/autostart.sh
and reboot


Use the solution from post #46.

11.0.6 was just taged. If that tag does not change my 240124 is a 11.0.6.

I will still build a clean 11.0.6 and delete then 240124 and 240116 to safe some space on my image server.