[Unstable Update] October 2024 Edition

[Metam0rph0sis]

I installed the 6.12-rc3 kernel and wanted to check Sched_ext, which should be included in the kernel, but it seems that this is not the case.
When I try to run the systemctl start scx service, I get a message ConditionPathIsDirectory=/sys/kernel/sched_ext was not met and Start scx_scheduler was skipped because of an unmet condition check
This module is not yet included in the kernel build parameters or something else is required ?
Tried to build the package with these fixes, but it did not change the result

[sparky]

It seems the default config for Sched_ext is off
CONFIG_SCHED_CLASS_EXT is not set
if you build again try setting to =y
there may be more needed
sorry wrong ‘reply to’ should have been @ Metam0rph0sis but see no way to change

[freggel.doe]

@philm

I see unstable branch as a stage branch for experiments and builds, which now get widely used by our users too.

Just a side note: I’m looking forward to a possible separation between edge (newest packages from arch with “stable” manjaro packages) and unstable (for developing/building/experimenting with manjaro packages) @romangg mentioned in some talks he gave.

[Metam0rph0sis]

@sparky

It seems the default config for Sched_ext is off

I expected this but I would not like to build my own kernel, hope this setting will be changed for the next versions of the package.

[defunct]

@philm

So as long as we don’t do our proper homework as distro maintainers I won’t ship newer Nvidia drivers on stable nor testing branches!

Totally get this, but for whatever reason, aside from the suspend/resume issues I have (which have always been a problem with Nvidia in some way or another), the performance of 560 drivers are amazing. Gnome in Wayland is super snappy.

[nikgnomic]

@defunct

aside from the suspend/resume issues I have (which have always been a problem with Nvidia in some way or another)

I have been using Manjaro Xfce for many years and not had a major issue with suspend/resume. Minor issues were resolved just by reading this forum

[GeorgeB]

Anyone else seeing a crossed out tray speaker (mute) for their volume control indicator after the last update? It shows it like this regardless of the volume level.

Also, the sddm breeze theme is broken and won’t allow logging in (just greys out the password input when you enter it and stays on screen), so I had to switch tty and change the theme to log in.

edit// the stuck mute volume indicator got fixed for me after setting a new global theme with default config in plasma settings, and rebooting. Just rebooting wouldn’t do it so I guess with 6.2 there were some changes to the tray.

[philm]

@freggel.doe

I’m looking forward to a possible separation between edge (newest packages from arch with “stable” manjaro packages) and unstable (for developing/building/experimenting with manjaro packages)

With the current concept, this would be testing branch. It always depends how long it may take to stablize unstable. Normally you have one branch you pull the packages from Arch, then adopt their changes and check the overlay packages from Manjaro. Put it all nicely together and snap it over to another branch for review. Whenever you sync new from Arch you reset that process. Only when that process is mostly automated you can expect more on pair with Arch stable repos.

[philm]

@GeorgeB

Also, the sddm breeze theme is broken and won’t allow logging in

Which exact package version you see that issue? Cos I adopted upstream changes for 6.2 Plasma just recently. [sddm] add plasma 6.2 support (0266802c) · Commits · Artwork / themes / Breath Theme · GitLab

[GeorgeB]

24.0.0-4

[freggel.doe]

@GeorgeB

Also, the sddm breeze theme is broken and won’t allow logging in (just greys out the password input when you enter it and stays on screen), so I had to switch tty and change the theme to log in.

See further up:

[Unstable Update] October 2024 Edition

sddm login problems/weirdness with multiple monitors on plasma: https://bugs.kde.org/show_bug.cgi?id=494804

and a quote from a comment of mentioned bug report:

FTR, this is caused by Qt 6.8, not Plasma 6.2.1

[cscs]

@cscs

For amdgpu users experiencing hangs or freezes, especially after suspend, beginning with kernel 6.11+;

This boot option ( amdgpu.dcdebugmask=0x200 ) seems to work so far.
The option disables PSR-SU, sometimes called PSR 2.0.

I have not experienced a hang after sleep since applying.

Spoke too soon. Eventually got another hang.

The gore

0 0 0 0 0 i2c i2c-10: SMBus Timeout!
0 0 0 0 0 i2c i2c-10: Failed reset at end of transaction (01)
0 0 0 0 0 i2c i2c-10: Failed! (01)
0 0 0 0 0 kwin_libinput: Libinput: event10 - ELAN06FA:00 04F3:31BE Touchpad: kernel bug: Touch jump detected and discarded.
1 1 1 1 1
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 amdgpu 0000:04:00.0: [drm] *ERROR* [CRTC:73:crtc-0] flip_done timed out
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 amdgpu 0000:04:00.0: [drm] *ERROR* flip_done timed out
0 0 0 0 0 amdgpu 0000:04:00.0: [drm] *ERROR* [CRTC:73:crtc-0] commit wait timed out
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 amdgpu 0000:04:00.0: [drm] *ERROR* flip_done timed out
0 0 0 0 0 amdgpu 0000:04:00.0: [drm] *ERROR* [PLANE:58:plane-3] commit wait timed out
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 amdgpu 0000:04:00.0: [drm] *ERROR* flip_done timed out
0 0 0 0 0 amdgpu 0000:04:00.0: [drm] *ERROR* [PLANE:70:plane-5] commit wait timed out
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug
0 0 0 0 0 kwin_wayland_drm: Pageflip timed out! This is a kernel bug

Will try

amdgpu.dcdebugmask=0x600

Which disables PSR-SU still, as well as Panel Replay.

[philm]

@Metam0rph0sis

I installed the 6.12-rc3 kernel and wanted to check Sched_ext , which should be included in the kernel, but it seems that this is not the case.

We enabled it for 6.12-rc4 build: [config] enable sched_ext (5c00176c) · Commits · Packages / core / linux612 · GitLab

[GeorgeB]

I’ve installed the last kernel updates (6.6.57-1) and my system froze repeating a few seconds of audio playing from a radio, without glitching. Just repeated the same two seconds of the feed from a song. Keyboard unresponsive, no mouse movement, I could only physically shut down the PC. After the restart, it could NOT recover the journalctl and it only had entries from the current boot (last two minutes) and its size is a couple of megabytes. Previous history got deleted (?!)

That’s the first time I’m ever experiencing anything like it. Freezes aren’t that uncommon with kernel updates and specific hardware, but why on earth would it delete the logs?

[freggel.doe]

@GeorgeB

Previous history got deleted (?!)
…
but why on earth would it delete the logs?

Just fyi: those probably weren’t deleted (as in “taken deliberate action to remove”) but not (yet) written to disk (missing sync).

[GeorgeB]

gigabytes of them? If that’s what happened on every power failure, the whole Linux-based server side of the internet wouldn’t know what’s up.

[Aragorn]

@GeorgeB

my system froze repeating a few seconds of audio playing from a radio, without glitching. Just repeated the same two seconds of the feed from a song.

This means that the processor was still trying to do its thing while the kernel was already dead.

@GeorgeB

After the restart, it could NOT recover the journalctl and it only had entries from the current boot (last two minutes) and its size is a couple of megabytes. Previous history got deleted (?!)

That’s the first time I’m ever experiencing anything like it. Freezes aren’t that uncommon with kernel updates and specific hardware, but why on earth would it delete the logs?

Filesystem auto-repair upon boot after the filesystem was in an incoherent sate due to the freeze and the subsequent system reset, depending on which filesystem you use.

[GeorgeB]

Kdump - ArchWiki maybe we can implement this on the unstable branch kernels, so we can diagnose kernel freezes, etc. It will help the testing and stable branches. It’s worth considering anyway.

[defunct]

With the Pamac GUI, I have to manually refresh databases to see all updates with the latest version. It seems to only show aur updates when you go to updates until manually refreshing.

Note I have aur, development versions and flatpak enabled.