It will not build. I am thinking arch-arm needs to rebuild linux-tools against the latest 6.12 kernel tree. I tried here and it also failed due to format error which usually means detecting the wrong architecture.
There was a bpf package upgrade yesterday so I did rebuild scx-scheds and pushed it to unstable but I believe it will not help your situation with the freezing. I am thinking it has more to do with something in your sway as others have not complained with their other DE’s.
test fail, maybe is wayland issue?
do you upgrade to xfce4 wayland yet?
xfce v4.20 has not hit arch-arm repo’s yet. Still waiting on it and the new python and llvm.
ok, just ask.
start practice labwc, prepare back to PiOS.
I have had this in my cmdline.txt for a long time (mainly for something else). You might want to try adding this to your cmdline.txt and see if it makes a difference. It seems that the order of things that gets loaded supposed to make a difference. I read that it can cause a conflict. Notice bpf is last.
lsm=landlock,lockdown,yama,integrity,apparmor,bpf
Additional:
v6.13-rc7 kernel packages pushed to unstable when the mirrors sync.
lsm line did not help.
edit,
Pi4/sway, long time sleep can not wake up, does it kernel matter?
edit2, @Darksky
I can’t read this in english, does idle=freeze?
Not a word about freeze.
Updated 2025-01-17:
v6.12.10-1 kernels (linux-rpi4/5) and raspberrypi-bootloader-20250114-1 packages pushed to unstable when the mirrors sync.
6.12.10 sched_ext freeze 10 seconds, worse.
No problems here.
ok, test Lxqt/labwc hours, sched_ext did not freeze anymore, so not wayland issue, its wlroots/sway 
v6.13.0 is out v6.12.10 has an update as well as v6.6.72 even I can remember that it is not on our repo anymore
I have been monitoring the v6.13.0 kernel to see if they are going to do anything with it. It failed their vc4 test. Sometimes it means nothing with code changes being made.
Some checks were not successful
10 successful and 1 failing checks
KUnit Tests / VC4 Unit Tests on ARM64 (push) Failing after 5m
I usually do not rebuild with everyday changes unless it is something major. I do rebuild with kernel every kernel upstream upgrade which can sometimes be 2 times a week.
6.13
*drm/v3d: Ensure job pointer is set to NULL after job completion
I went ahead and compiled v6.13.0 and it freezes every time with youtube with the audio in a short loop. I see this error in the terminal:
Crash Annotation GraphicsCriticalError: |[C0][GFX1-]: Couldn't sanitize GL_RENDERER
"V3D 7.1.7.0" (t=34.0883) [GFX1-]: Couldn't sanitize GL_RENDERER "V3D 7.1.7.0"
can we have a google drive link to test ?
I have only tested with the pi5 so far. Click on the Download button and unpack artifacts.zip then install the kernel packages.
linux-rpi4-mainline:
https://gitlab.manjaro.org/manjaro-arm/packages/core/linux-rpi4-mainline/-/jobs/15341
linux-rpi5-mainline:
https://gitlab.manjaro.org/manjaro-arm/packages/core/linux-rpi5-mainline/-/jobs/15343
Confirmed.
I have that ?
Jan 21 11:02:19 bloodmoon-pc kernel: vc4-drm axi:gpu: [drm] *ERROR* Failed to get ddc i2c adapter by node
Jan 21 11:02:19 bloodmoon-pc kernel: vc4-drm axi:gpu: bound 107c580000.hvs (ops vc4_hvs_ops [vc4])
Jan 21 11:02:19 bloodmoon-pc kernel: rpi-gpiomem 107d510700.gpiomem: window base 0x107d510700 size 0x00000020
Jan 21 11:02:19 bloodmoon-pc kernel: rpi-gpiomem 107d504100.gpiomem: initialised 1 regions as /dev/gpiomem3
Jan 21 11:02:19 bloodmoon-pc kernel: rpi-gpiomem 107d504100.gpiomem: window base 0x107d504100 size 0x00000020
I opened an issue on the RPi git. At first they were going to tell me it was working ok for them then had to come back with a kernel panic. I feel confident they will get it fixed now soon.