Page 144 of 147
Re: New kernels
Posted: Tue Oct 14, 2025 9:01 pm
by xeno74
Hi All,
The merge window is over. Linus Torvalds has released the
RC1 of kernel
6.18 for testing.
I have compiled it for X1000 and e5500 machines (X5000/20, X5000/40, Mirari, and QEMU VMs).
The kernel developers have activated the power management for all devices connected to PCI and PCI Express.
Unfortunately not all devices are working with these modifications. For example AMD Radeons HD6870 and HD5870. There are some kernels available with different patches for testing.
Download and further information:
github.com
Kernel 6.18.0-rc1 testing:
Please test the kernels.
Thanks,
Christian
Re: New kernels
Posted: Wed Oct 15, 2025 6:56 pm
by Hypex
xeno74 wrote: Tue Oct 14, 2025 9:01 pm
Hi All,
The merge window is over. Linus Torvalds has released the
RC1 of kernel
6.18 for testing.
Hi. Just testing the RC1 now. Seen no issues on my X1000 system. Kernel experience stable so far as I type this.
Re: New kernels
Posted: Thu Oct 16, 2025 5:07 am
by xeno74
Hypex wrote: Wed Oct 15, 2025 6:56 pm
Hi. Just testing the RC1 now. Seen no issues on my X1000 system. Kernel experience stable so far as I type this.
This sounds good. Thanks for testing.
Re: New kernels
Posted: Fri Oct 17, 2025 5:37 am
by xeno74
Hi All,
I have compiled the
RC1-2 of kernel
6.18 for X1000 and e5500 machines (X5000/20, X5000/40, Mirari, and QEMU VMs) without the power management for PCI and PCI Express (CONFIG_PCIEASPM and CONFIG_PCIEASPM_DEFAULT).
Disabling both kernel options resolved the boot and performance issues.
Download and further information:
github.com
Please test the kernels.
Thanks,
Christian
Re: New kernels
Posted: Mon Oct 20, 2025 6:45 am
by xeno74
New for the
RC2:
We have continued to disable power management for PCI and PCI Express because of boot issues and performance issues
#17. (
CONFIG_PCIEASPM and
CONFIG_PCIEASPM_DEFAULT)
Re: New kernels
Posted: Tue Oct 21, 2025 2:36 pm
by Hypex
xeno74 wrote: Fri Oct 17, 2025 5:37 am
Hi All,
I have compiled the
RC1-2 of kernel
6.18 for X1000 and e5500 machines (X5000/20, X5000/40, Mirari, and QEMU VMs) without the power management for PCI and PCI Express (CONFIG_PCIEASPM and CONFIG_PCIEASPM_DEFAULT).
Hi. I tested the RC1-2 kernel on my X1000. It booted fine but the system was sluggish in use. Similar to Fienix or Debian but a lot slower. I ran top and Xorg was using almost 99% of my CPU. That was on the MATE Remix. So using that much CPU was very unusual. Usually that system is fairly snappy and the previous alpha was fine.
Re: New kernels
Posted: Tue Oct 21, 2025 3:02 pm
by xeno74
Hypex wrote: Tue Oct 21, 2025 2:36 pm
Hi. I tested the RC1-2 kernel on my X1000. It booted fine but the system was sluggish in use. Similar to Fienix or Debian but a lot slower. I ran top and Xorg was using almost 99% of my CPU. That was on the MATE Remix. So using that much CPU was very unusual. Usually that system is fairly snappy and the previous alpha was fine.
Thanks for testing the RC1-2. I also tested the RC1-2 last week, but I don’t have 99% CPU usage. I have about 20%-30% CPU usage when not in use.
Maybe the fbdev driver? I use the Radeon driver. I tested with Void PPC. Have you seen any error messages in the dmesg?
Re: New kernels
Posted: Tue Oct 21, 2025 5:37 pm
by xeno74
Hi All,
Linus Torvalds has released the
RC2 of kernel
6.18 for testing.
I have compiled it for X1000 and e5500 machines (X5000/20, X5000/40, Mirari, and QEMU VMs).
Download and further information:
github.com
Please test the kernels.
Thanks,
Christian
Re: New kernels
Posted: Wed Oct 22, 2025 5:57 am
by xeno74
I had only about 20% CPU usage with the RC2 if the X1000 not in use.
Re: New kernels
Posted: Wed Oct 22, 2025 4:17 pm
by Hypex
xeno74 wrote: Tue Oct 21, 2025 3:02 pm
Thanks for testing the RC1-2. I also tested the RC1-2 last week, but I don’t have 99% CPU usage. I have about 20%-30% CPU usage when not in use.
I checked with Mint as well tonight and all is well there. Xorg didn't go much over 7% and system was responsive. Modules loaded.
Maybe the fbdev driver? I use the Radeon driver. I tested with Void PPC. Have you seen any error messages in the dmesg?
It could be. It is bound to be slower. But sometimes I do notice a glitch. For example I once booted into my Xenial and it was sluggish. It was a fresh boot so didn't make sense. I couldn't work with it so I logged out. I logged back in and it was fine! Didn’t even need to reboot.
I've now booted MATE again. This time it's fine. Checked dmesg and no obvious errors aside from what I usually find. Xorg topping out at about 12%. Well now I have installed the rc2 kernel now and should be able to test tomorrow night.
Code: Select all
amigaone@ubuntu:~$ lsmod
Module Size Used by
i2c_piix4 20480 0
snd_hda_codec_atihdmi 16384 1
snd_hda_codec_idt 53248 1
snd_hda_codec_hdmi 45056 1 snd_hda_codec_atihdmi
snd_hda_codec_generic 86016 1 snd_hda_codec_idt
snd_hda_intel 36864 6
snd_intel_dspcfg 12288 1 snd_hda_intel
snd_hda_codec 135168 5 snd_hda_codec_generic,snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec_atihdmi,snd_hda_codec_idt
snd_hda_core 98304 6 snd_hda_codec_generic,snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec,snd_hda_codec_atihdmi,snd_hda_codec_idt
amigaone@ubuntu:~$ dmesg | grep -i error
[ 2.271968] [drm:uvd_v1_0_ib_test] *ERROR* radeon: fence wait timed out.
[ 2.271998] [drm:radeon_ib_ring_tests] *ERROR* radeon: failed testing IB on ring 5 (-110).
[ 5.708885] faux_driver regulatory: Direct firmware load for regulatory.db failed with error -2
amigaone@ubuntu:~$ dmesg | grep -i fail
[ 2.271998] [drm:radeon_ib_ring_tests] *ERROR* radeon: failed testing IB on ring 5 (-110).
[ 5.708885] faux_driver regulatory: Direct firmware load for regulatory.db failed with error -2
[ 5.725737] cfg80211: failed to load regulatory.db
[ 22.187083] systemd[1]: systemd-journald.socket: SO_PASSSEC failed: Operation not supported
[ 22.201579] systemd[1]: systemd-journald.socket: SO_PASSSEC failed: Operation not supported
[ 22.270985] systemd[1]: systemd-journald-dev-log.socket: SO_PASSSEC failed: Operation not supported
