Finally after around 8 months of crossdeving I realized that this won't workout (especially with QT packages). There are just too many build failures. Regardless I've published my findings if anyone would like to do something like this.

The final nail in the coffin for crossdev was kwin which came ...

So I had to migrate my crossdev setup which I was using to maintain my laptop to a chroot-maintained setup on my non-avx512 capable workstation. Taking advantage of the situation I did some with vs without avx512 benchmark on the same machine.

http://delogics.blogspot.com/2025/08/gentoo-benchmarks ...

Rebuild pam from the running Gentoo machine. You build it in chroot which causes issues for some reason.

@devs may take note, this issue I'm also facing.

Your user is in the render and video group?

Ok, so once it starts taking 30%, it always takes 30%. There is no 'spike' as such.

If you've screencast USE enabled in kwin just try to disable it temporarily to see if it helps.

Start KDE directly by executing /usr/bin/startplasma-wayland

Also put yourself to the render and video groups.

'randomly' implies that if the system is idling you see the GPU spike to 30% for extended periods of time or is it something else?

During this period you may like to see if there is any increase in CPU utilization of any other process which may give you pointers on which process might be triggering ...

Can you report your CPU utilization while playing music?

Are you using pulse? You may try to play a wav file using aplay and selecting the correct alsa device (to ensure it's not using pulse) and see if stuttering is there. If not, it's a pulseaudio problem.

You may also like to change your cpu ...

WIREPLUMBER WAS NOT RUNNING!! THIS WAS THE ISSUE
systemctl --user start wireplumber.service
I wonder was it disabled...

Apparently Apache is not fixed. I tried to apply the patches to the apache 2 eclass manually but it didn't fix the issue (same pkg-config problem). And emerge --sync is not syncing with the fresh git tree as of now, so I'll wait and report.

Ok, Apache is also fixed. I'll try it out. Thanks! This was much needed.

Yes, it compiles now. Thanks so much for the fix. Apache itself also has a similar problem, but because it seems my toolchain is fine, I'll go ahead and file a bug.

Hello everyone.

I've always been having problems cross compiling app-admin/apache-tools -- which is pretty surprising because this is a VERY common package for embedded systems and no one's complaining -> something's wrong with my toolchain.

config.log --
https://paste.ubuntu.com/p/JYHPMP84V8 ...

This is becoming a serious headache. I can't share screen on browsers (webRTC screen sharing not working). And I've all reasons to believe that the root cause is the same.

Therefore I did some benchmarks (taking advantage of the current situation).

http://delogics.blogspot.com/2025/04/de ... hmark.html

In the mean time this cross boss script works for many packages.

Yup, rust is the future. It even ended up in the kernel.

sccache is like distcc for rust.

dE_logics,

The SDE must be available inside the chroot, or the chroot will leave the SDE outside the chroot and therefore inaccessible.
Putting SDE in your path, when you want to chroot is tricky as it will either be wrong inside or outside the chroot.

I'm surprised that you only pass ./SDE, a ...

I'll try it out. But still a better approach will be to make the kernel execute the emulator when it detects that the executable lies in a certain matching path (like /tmp/portage). That way cross compiling bugs can be overcome and it'll still be reasonably fast.

I found update-binfmts --detector ...

So what I can do is use distcc when a package cannot be cross compiled. It'll be a backup option.

Thanks for suggesting this.

For for the story of by burnt out laptop, it was an Athlon x2. The CPU didn't have problems, but the mobo did.

I'm just trying to avoid frying the laptop like it happened last time. I ran gentoo on it since 2009 to I think at 2010 I died. Although I believe other CPU intensive tasks where also to blame.

So I rather drop avx512 instructions.

What happens with distcc is that that compiling is done remotely ...

But AMD is supposed to have extra instructions which Intel never incorporated. So that shouldn't be a problem. Also sde binary must be built with x86-64 baseline instruction set so it must run on even the 2003 era Athlon processors, but this zen. So it shouldn't be a problem.

So what I'll do is ...

but this zen. So it shouldn't be a problem. Zen has some "missing" instructions compared to Bulldozer, I'was bitten by it first 8 or so years ago during a "hardware upgrade", and avoided it recently again: Doing an emerge -e world on new hardware is always kinda cool, but having to do it before the ...

In brief, I'm trying to cross compile x86-64-v4 binaries on x86-64-v3.

stage3 must be in baseline.

But AMD is supposed to have extra instructions which Intel never incorporated. So that shouldn't be a problem. Also sde binary must be built with x86-64 baseline instruction set so it must run on even the 2003 era Athlon processors, but this zen. So it shouldn't be a problem.

So what I'll do is put ...