Posts by chewitt

Some things to try (not in any specific sequence):

a) Remove the SD card and power on the RPi board. Check that the firmware screen shows up and HDMI and EDID are 'OK'

b) Check that you've updated to the lasted RPi-eeprom firmware? If not, update and try again

c) See if the image here https://chewitt.libreelec.tv/testing/LibreE…-12.90.1.img.gz behaves any different?

d) Using a spare SD card write the nightly image to the card (so clean install not update) .. booting?

I'm using an RPi5 with my own nightly image for aeons without issues so there's no general problem. I have a hunch it might be eeprom firmware related (b).

Quote from Da Flex

Try the latest LE 13 nightly. It has some fixes for the WiFi driver.

How do you imagine the WiFi driver is linked to this issue? (or did you reply to the wrong thread)

Ensure the wireless regulatory domain used by the RPi is aligned to the one used by the router; else the router may decide to use radio configurations that the RPi cannot handle (or vice versa).

Read the explanation given in [RPi5] H.265 TS Severe Artefacts when Skipping Backwards and continue there if needed.

You can delete the Addons DB file and it will be regenerated on next boot from the add-ons found in /storage/.kodi/addons though you might need to manually (re)enable some of the add-ons again. You can also delete the EPG database as this contains transient data and will be repopulated from Tvheadend.

The online repo the .tar.xz file is downloaded from has (for unknown reasons) regenerated the download and this has caused the hash of the downloaded file to change. We see the same on GitHub sometimes.

The correct fix is updating PKG_SHA256 in package.mk to use the 'got' hash value instead of the current 'wanted' hash. The alternate workaround is setting PKG_SHA256="" (null) which disables the hash check. The one-line command vpeter posted does this.

https://git.zx2c4.com/wireguard-tools/snapshot/wireguard-tools-v1.0.20250521.tar.xz

This URL ^ (correctly formatted, unlike the one you posted) works fine for me.

Or bump bump the wireguard-tools package to the current release, e.g. https://github.com/LibreELEC/Libr…ools/package.mk

Or download the .tar.xz file from somewhere else (and storing it with the correct filename) to sources/wireguard-tools/ and then manually generate the required .url and .sha256 files and content (look at existing files to crib the simple format).

Quote from Da Flex

Doesn't the kernel detect devices at boot time?

It should, but the user has provided zero evidence that it did, and they've also been fiddling with files they shouldn't. So the normal protocol is to ask for evidence (the LE13 pastekodi output, which will show the full EDID block, full log, and system log) so that blind guessing isn't involved.

cp /usr/lib/udev/rules.d/95-udevil-mount.rules /storage/.config/udev.rules.d

Clone the udev rules file ^

# check for blockdevices, /dev/sd*, /dev/sr*, /dev/mmc*, and /dev/nvme*
SUBSYSTEM!="block", KERNEL!="sd*|sr*|mmc*|nvme*", GOTO="exit"       <= edit this

# check for blockdevices, /dev/sd*, /dev/sr*, /dev/mmc*, and /dev/nvme*
SUBSYSTEM!="block", KERNEL!="sd*|sr*|nvme*", GOTO="exit"            <= to this, i.e. drop mmc*

Then edit the file using nano and change this block towards the top of the file ^ then save/reboot to effect the change.

The change means for any mmc* blockdevice it skips straight to exit and doesn't process partition mounting. NB: This will also prevent SD cards from ever mounting as those are also mmc* devices. I doubt you need that though.

Quote from Da Flex

There must be an audio routing error on LE side.

Are you using tea leaves or a crystal ball to derive this knowledge?

Using an LE13 nightly image (as 12.2 will not contain the info I'm looking for) do a clean install, put Kodi in debug mode, reboot, then run "pastekodi" and share the URL so we can see the log.

The Kodi release process has been a little challenging over the last year. This is partly due to different people picking up a complex an unpopular task, but also because there's a desire to release for all platforms at the same time and we've repeatedly seen issues with one platform which takes days-to-weeks to resolve. In the meantime development continues to roll forwards and more things get merged into the master branch: hence the staggered timing between LE which simply tracks the master branch in nightlies instead of following Alpha releases (these days we start formal releases at Beta) and official Kodi binaries finally being released once the build issues are resolved.

Quote from wyup

Is there a way to pass kernel options or modify filesystem node parameters for the KMS to customise these settings?

I have fuzzy recall RPi5 uses 10-bit internally padded to 12-bit for output in some circumstances as the SoC doesn't natively support the required 10-bit output. There is some upstream work being done to improve output from Kodi which might indirectly influence things, but the direct and short answer to the question above is "No"

I'm not aware of any reason the update should not work, but Kodi doesn't support downgrades so systemctl stop kodi && cp -R /storage/.kodi /storage/.kodi-12.2 before doing a manual update will clone the Kodi folder and give you a simple way to revert back to 12.2 if you need to.

PEBKAC

It looks like CONFIG_KASAN=y snuck into a kernel defconfig and this increases the KERNEL size to a point where standard vendor u-boot configurations (as used with Android STBs) will fail to load the kernel and fault. This is nothing to do with RAM size.

I have a large list of changes to work on and current progress is rather slow, so I will update images at some point soon, but to avoid breaking promises I'm not going to give a timeline on when.

Collabora are working on broader colour support, but there are background kernel colour UAPI changes at the same time so it's likely to continue for a while to allow for things being merged in a sensible sequence. There is also development on multi-core support which is a requirement for 8K (an 8K image is essentially 4x 4K frames being parallel processed). I also haven't seen any signs of people working on AFBC yet, but that will also be needed to reduce memory bandwidth requirements.

Configure the OneForAll to behave like an IR6/MCE remote. Configure the flirc reciever to use the Kodi profile and go through each button of the profile to map what the remote sends for that button to what the receiver sees. Kodi should then work using the MCE remote profile without needing any XML fiddling. It's a PEBKAC issue not a kernel issue.