Bug 33374

Assigning to kernel - you?!

Assignee: bugsquad => kernel

Source RPMS
kernel-linus-6.6.37-1.mga9

Binaries RPMS in x86_64

Repository: 9-x86_64-core-updates_testing
kernel-linus-6.6.37-1.mga9
kernel-linus-devel-6.6.37-1.mga9
kernel-linus-devel-latest-6.6.37-1.mga9
kernel-linus-doc-6.6.37-1.mga9
kernel-linus-latest-6.6.37-1.mga9
kernel-linus-source-6.6.37-1.mga9

Binaries RPMS in i586

Repository: 9-i586-core-updates_testing
kernel-linus-6.6.37-1.mga9
kernel-linus-devel-6.6.37-1.mga9
kernel-linus-devel-latest-6.6.37-1.mga9
kernel-linus-doc-6.6.37-1.mga9
kernel-linus-latest-6.6.37-1.mga9
kernel-linus-source-6.6.37-1.mga9

RH mageia 8 x86_64

LC_ALL=C urpmi kernel-linus kernel-linus-devel

installing kernel-linus-6.6.37-1.mga9.x86_64.rpm kernel-linus-devel-6.6.37-1.mga9.x86_64.rpm from //home/katnatek/qa-testing/x86_64
Preparing...                     ##################################################################################################
      1/2: kernel-linus-devel    ##################################################################################################
      2/2: kernel-linus          ##################################################################################################
remove-boot-splash: Format of /boot/initrd-6.6.37-1.mga9.img not recognized

vhba (20211218-2.mga9): Installing module.
............
........
You should restart your computer for kernel-desktop, kernel-linus, kernel-server

Tomorrow the test

RH mageia 8 x86_64, uname -r
6.6.37-1.mga9

Video OK
Webcam OK
Ethernet OK
Mount isos with cdemu OK
Not issues detected

As previous iterations, this flavor makes my car use radeon driver, not regression

I get a kernel panic in my RH mageia 9 i586

https://www.imagebam.com/view/MEULEZO
https://www.imagebam.com/view/MEULEZN

Mageia9, x64
Intel model: NUC12WSBi7
12-core Intel Core i7-1260P - 16 with some HT
Intel Alder Lake-P Integrated Graphics - i915
Intel Alder Lake-P PCH CNVi WiFi driver: iwlwifi

Kernel updated to linus.
Rebooted smoothly.  Browsers working fine.  Networking OK including NFS shares.
bluez working fine - Bluetooth speaker connected as soon as switched on.
$ stress -c 5 -m 3 -i 2 -d 2 -t 25
stress: info: [111693] dispatching hogs: 5 cpu, 2 io, 3 vm, 2 hdd
stress: info: [111693] successful run completed in 26s

Checked libreoffice writer, VirtualBox, vlc for videos, created 8GB ramdisk and ran tar on it to generate a backup of a user directory then extracted the files onto the ramdisk.  All OK.  Leaving this to run.  No issues apparent.

CC: (none) => tarazed25

(In reply to katnatek from comment #5)
> I get a kernel panic in my RH mageia 9 i586
> 
> https://www.imagebam.com/view/MEULEZO
> https://www.imagebam.com/view/MEULEZN

Try to boot with one of clocksource=hpet|acpi_pm|jiffies in kernel boot cmdline.

(In reply to Giuseppe Ghibò from comment #7)
> (In reply to katnatek from comment #5)
> > I get a kernel panic in my RH mageia 9 i586
> > 
> > https://www.imagebam.com/view/MEULEZO
> > https://www.imagebam.com/view/MEULEZN
> 
> Try to boot with one of clocksource=hpet|acpi_pm|jiffies in kernel boot
> cmdline.

cat /proc/cmdline 
BOOT_IMAGE=/boot/vmlinuz-6.6.37-1.mga9 root=UUID=d8ee3489-3e11-46a1-8aa1-e8e3637f404a ro ro splash quiet noiswmd root=UUID=d8ee3489-3e11-46a1-8aa1-e8e3637f404a audit=0 resume=UUID=b772d414-4192-401b-bdd4-43174b560b41 vga=791 clocksource=hpet

Tje first did the trick, later I test with others

uname -a
Linux cefiro 6.6.37-1.mga9 #1 SMP PREEMPT_DYNAMIC Sat Jul  6 02:22:43 UTC 2024 i686 GNU/Linux

(In reply to Giuseppe Ghibò from comment #7)
> (In reply to katnatek from comment #5)
> > I get a kernel panic in my RH mageia 9 i586
> > 
> > https://www.imagebam.com/view/MEULEZO
> > https://www.imagebam.com/view/MEULEZN
> 
> Try to boot with one of clocksource=hpet|acpi_pm|jiffies in kernel boot
> cmdline.

Just clocksource=hpet works I guess for this kernel and system will be a permanent addition

mga9-64 on my workstation svarten
Plasma X11, Intel i7-870, Intel P55, nvidia GTX750
Filesystem: LVM on LUKS, on SSD

Improvement over previously released update:
The 6.6.37-1 kernels (linus, desktop, and sever) solve the hang from suspend-resume that nvidia-current 550.90-07 (but not the former version) show using kernels 6.6.28.
(But still nvidia-newfeature-555.58.02-1 hangs on resume with any kernel https://bugs.mageia.org/show_bug.cgi?id=33375#c2 )

Various desktop apps, flatpak and apps, BOINC, syncthing, java app ...

vt switching: OK

Suspend-resume: OK

Hibernate-resume: OK

Bad, regression:
VirtualBox, using locally dkms built kmod, running with MSW7 guest: using USB 2 chip programmer Galep5... connection failed and when I in VB menu chose to disconnect that USB device, guest just stopped, using no CPU!  I then let VB send cpi off, and then guest started using one core 100% but for minutes nothing else happens, no repaint of resized guest window.    :-(

$ dkms status|grep 6.6.37
virtualbox, 7.0.18-1.mga9, 6.6.37-1.mga9, x86_64: installed 
virtualbox, 7.0.18-1.mga9, 6.6.37-server-1.mga9, x86_64: installed 
virtualbox, 7.0.18-1.mga9, 6.6.37-desktop-1.mga9, x86_64: installed 
nvidia-current, 550.90.07-1.mga9.nonfree, 6.6.37-1.mga9, x86_64: installed 
nvidia-current, 550.90.07-1.mga9.nonfree, 6.6.37-desktop-1.mga9, x86_64: installed

mga9-64 OK on Laptop Thinkpad T510
- except it fail to shut off power when hibernating (no regression)
Details see https://bugs.mageia.org/show_bug.cgi?id=33373#c22

Not was validated with the other kernel update why?

CC: (none) => andrewsfarm

I think it's ok to validate this too.

I have not had time to test and report as I usually do...

For the regression in VirtualBox USB (my previous comment above) I see the same on desktop kernel as on this linus.  I see I reported it did work using kernels 6.6.28.  I should try 6.6.28 again and see if it still works there or it is some other problem now maybe even a guest system security update (windows7) is at fault.  Flash drive works, but not the Galep5 chip programmer.

Regardless, I think 6.6.37 should be released as it at least on my worksattion works better with latest nvidia-current which hang on resume from suspend with 6.6.28.

As normal, really we should be many more testing such important thing as the kernel, but we are few testing so I too think we should release it.

Yes. Consider also that kernel-linus is a vanilla kernel, without (apart a couple of exceptions) any extra patch or fixes beyond those included upstream.

I wouldn't want to disclaim that the usb problem could be from Win7. My Win7 guest has a problem identifying my usb ports as anything but 1.0 when they are 3.0, but Mageia guests don't have that problem. 

Validating.

CC: (none) => sysadmin-bugs
Whiteboard: (none) => MGA9-32-OK MGA9-64-OK
Keywords: (none) => validated_update

An update for this issue has been pushed to the Mageia Updates repository.

https://advisories.mageia.org/MGASA-2024-0266.html

Status: NEW => RESOLVED
Resolution: (none) => FIXED

ADDENDUM

The problem with my USB connected chip programmer
(bad communication and then the guest hangs)
is tied to VirtualBox 7.0.18! - not kernel.
With that VB version I see the problem also on kernels 6.6.28.

I did not QA check 7.0.18 with *that* USB device.
I did only try USB flash stick was OK.
Live and learn...

After downgrading VB to 7.0.14, that device works with both 6.6.28 and 6.6.37 kernels.

(In reply to Morgan Leijström from comment #18)
> ADDENDUM
> 
> The problem with my USB connected chip programmer
> (bad communication and then the guest hangs)
> is tied to VirtualBox 7.0.18! - not kernel.
> With that VB version I see the problem also on kernels 6.6.28.
> 
> I did not QA check 7.0.18 with *that* USB device.
> I did only try USB flash stick was OK.
> Live and learn...
> 
> After downgrading VB to 7.0.14, that device works with both 6.6.28 and
> 6.6.37 kernels.

Please open a new bug against virtualbox, thank you for your research

-> Bug 33400