- openSUSE Kernel

Considering disabling WAN in SUSE kernels
by Jiri Slaby 19 Apr '24

19 Apr '24

Hi all, I've filed: https://bugzilla.suse.com/show_bug.cgi?id=1223131 to consider disabling WAN in the kernel. SLE kernels do not provide that functionality and I assume there is noone actually using this in Leap or Tumbleweed. If there is anyone in the wild, raise up your voice in the bug with your use case and exact driver you are using. We might end up disabling only a subset (or nothing at all). thanks, -- js suse labs

1 0

Linux 6.8.4-rc1 in Kernel:stable
by Frank Krüger 05 Apr '24

05 Apr '24

Hi there, Is there any deeper reason that RC1 of kernel 6.4.1 has already been published? Reference to bsc#1012628 in the changelog is not really helpful. Thanks. Regards, Frank

2 1

[Slowroll] Building and signing KMPs
by Manfred Hollstein 03 Apr '24

03 Apr '24

Hi there, I build the old broadcom-wl kmp for a variety of OS/kernel combinations. With Slowroll however it fails when the built rpms will be signed using a generated pesign.spec. This is from https://build.opensuse.org/public/build/home:manfred-h:KMPs/openSUSE_Slowro…: [ 33s] /etc/profile.d/suse-buildsystem.sh: line 53: test: : integer expression expected [ 33s] + exec rpmbuild -ba --define '_srcdefattr (-,root,root)' --nosignature --undefine _enable_debug_packages --define 'disturl obs://build.opensuse.org/home:manfred-h:KMPs/openSUSE_Slowroll/9aa3d5cf0236…' /home/abuild/rpmbuild/SOURCES/pesign-repackage.spec [ 33s] error: source_date_epoch_from_changelog set but %changelog is missing [ 33s] [ 33s] RPM build errors: [ 33s] source_date_epoch_from_changelog set but %changelog is missing [ 33s] ### VM INTERACTION START ### [ 34s] [ 28.473717][ T1] sysrq: Power Off [ 34s] [ 28.474748][ T8] reboot: Power down [ 34s] ### VM INTERACTION END ### [ 34s] [ 34s] i02-ch1d failed "build pesign-repackage.spec" at Wed Mar 27 08:11:09 UTC 2024. Perhaps Bernhard is reading this, is there anything special building a kmp for Slowroll? TIA, cheers. l8er manfred

5 9

Wrong dates in shared folder mounted in VirtualBox
by Larry Finger 07 Mar '24

07 Mar '24

Hi, As noted in boo#1220524 - Files in virtualbox shared folder (on the guest) show wrong timestamp, the dates are not passed from host to guest if the guest is TW. This problem began sometime in kernel 6.7.X. To try to debug this issue, I have generated a v6.7.7-1-default kernel locally. The standard openSUSE configuration was used with one change. As I do not have access to the .pem file specified in the standard configuration, I changed to the kernel source signing key with 'CONFIG_MODULE_SIG_KEY="certs/signing_key.pem"'. All other configuration variables are unchanged. The kernel generated in this manner works correctly; however, what should be the equivalent installed with zypper from the TW repos fails!!! The local kernel was generated with gcc-13.2.1 20240206. I will be happy to provide further information that will help fix this bug. Larry

2 1

Building the distribution kernel
by Larry Finger 05 Mar '24

05 Mar '24

Hi, I am trying to debug the problem shown in boo#1220524 - Files in virtualbox shared folder (on the guest) show wrong timestamp. Using the standard 6.6.5 default kernel in a TW guest (Xfce desktop) shows the problem. When I generate a 6.6.5 kernel starting with the mainline 6.5.0 and applying the patches in the kernel-source repo in the patches.kernel.org and patches.suse directories, the generated kernel works correctly! What am I missing? I am installing the kernel using 'make modules_install install', not from an rpm, but I do not expect that to cause a problem. Are there patches applied to the standard kernel that I am missing? Does installing from an rpm install other packages? Thanks, Larry

2 1

Errors returned from dracut only with Kernel 6.8
by Gene Snider 30 Jan '24

30 Jan '24

I use the Kernel_HEAD repository. Two new errors have appeared ever since the first version of kernel 6.8, 6.8.0-rc1-1.gc2014a1-default, was installed and for all subsequent version. The errors occur when dracut is called from zypper and tries to install the i18n module. Here is the error from /var/log/zypp/history: dracut[I]: *** Including module: i18n *** /usr/bin/zgrep: line 271: /usr/bin/gzip: Permission denied /usr/bin/zgrep: line 286: /usr/bin/grep: Permission denied /usr/bin/zgrep: line 271: /usr/bin/gzip: Permission denied /usr/bin/zgrep: line 286: /usr/bin/grep: Permission denied dracut[I]: *** Including module: btrfs *** I've checked the permissions on /usr/bin/grep and /usr/bin/gzip and they appear correct. Also, both programs work correctly from the command line, and for all other modules. Both errors are not present in the last version of 6.7, 6.7.0-12.g020caa6-default, and all earlier versions. Is this a regression in kernel 6.8.0-rc1-1.gc2014a1-default? Regards, Gene

1 1

Kernel 6.7
by Frank Krüger 12 Jan '24

12 Jan '24

Hi there, Is there any deeper reason that kernel 6.7 hasn't reached Kernel:stable yet? Thx. Regards, Frank

2 1

Re: Disabling /dev/fb0 on TW
by Michal Suchánek 14 Nov '23

14 Nov '23

On Tue, Oct 10, 2023 at 10:35:53AM +0200, Jan Engelhardt wrote: > > On Tuesday 2023-10-10 10:26, Thomas Zimmermann via openSUSE Factory wrote: > > fbiterm has been unmaintained for >15 years > > 8 years give or take; https://github.com/jengelh/fbterm/commits/master > > > But there's no good drop-in replacement. > > kmscon. > > $ cat /etc/kmscon/kmscon.conf > font-name=Consoleet VGA 9x16 > font-size=32 > > (systemctl enable kmsconvt@tty2) > > It even does CJK (which is what fb(i)term was previously used for). It does not, however, use the kernel keymap, and does not support changing the keymap on the fly which is something the installer abuses. There is an laternative that does support this: mlterm-sdl. It is a hassle to set up, though. And nobody cared to create and test the setup for the installer to use mlterm instead of fbiterm. Thanks Michal

1 0

CONFIG_NFS_V4_2_READ_PLUS in master
by Petr Tesařík 31 Oct '23

31 Oct '23

Hi Michal, I can see that CONFIG_NFS_V4_2_READ_PLUS config option was unselected by commit ff9060bd2d1965e27f2da795a27a15d93d4cf67d. This is understandable, given mainline commit 21e31401fc4595aeefa224cd36ab8175ec867b87. However, it seems that all issues have been worked out, see mainline commit 9cf2744d249144fc0fe17667b56da78216678378. Could we re-enable CONFIG_NFS_V4_2_READ_PLUS in master now? Cheers, Petr T

3 5

cifsd-cfid-laundromat in D state
by Wolfgang Rosenauer 20 Oct '23

20 Oct '23

Hi, just tried to analyze why my idle system shows a constant load of around 7 since a while. This is 6.5.6-1-default #1 SMP PREEMPT_DYNAMIC Fri Oct 6 11:20:48 UTC 2023 (c97c2df) x86_64 x86_64 x86_64 GNU/Linux and I found it's most likely 7 cifsd-cfid-laundromat threads in D state which again pointed me at https://lore.kernel.org/lkml/457a5483c3c4ca5bb6c7ec6f4231074c.pc@manguebit.… Is the fix already in the pipeline for Tumbleweed? Thanks, Wolfgang

2 2