- openSUSE Users

Did solver.dupAllowVendorChange change? Dupe last night listed "Problems" for newer packages in alternate repositories?
by David C. Rankin 30 Nov '24

30 Nov '24

All, Devs, zypper dupe last night was really odd from a "problems" and chose a solution standpoint. It seemed like zypper dup was ignoring my setting in zypp.conf of solver.dupAllowVendorChange = false It was suggesting overwriting OSS packages with packages from the science and mingw repositories. There were multiple packages where I had to tell the chooser to keep the OSS package and not pull the package from some other repo. I thought that setting prevented that. It also wanted suggested switching Nvidia drivers from my current repo to the openSUSE repo with lower version (newer package but lower release count). I was fine with the change, but was surprised that it too seemed to ignore the dupAllowVendorChange setting. Did something change how that setting works? -- David C. Rankin, J.D.,P.E.

2 6

Why do we have system-user-wwwrun? It conflict with nginx user:group and is NOT a real dependency of php.
by David C. Rankin 30 Nov '24

30 Nov '24

All, Devs, Why do we have the package system-user-wwwrun? Yes it creates the user/group for the webserver to run as, but it totally conflicts with the nginx webserver default nginx:nginx, and there is no requirement that a webserver run as www:wwwrun -- none at all. The downside to this type "nanny" package is it creates confusion on updates suggesting that php must be completely uninstalled if the package system-user-wwwrun or requiring the user to choose "break" php by ignoring some of its dependencies. Running php-fpm under user/group www:wwwrun is NOT a valid php dependency - period. After installing nginx, I configured the web server to run as nginx:nginx based on the default installed nginx user/group. It already creates the nginx user/group, so there is no reason for the system-user-wwwrun dependency on php trying to change that. Now every time I try and update Tumbleweed it tries to uninstall php and I'm forced to choose the false option to "break" php every update. These nanny packages should be optional and NOT a dependency. They can be "Recommended", but dependencies are supposed to be real dependencies not some made up user-preference masquerading as a dependency. Php-fpm does not care what user it runs as. That is a simple 2 line config issue: user = nginx group = nginx So why do we have system-user-wwwrun as a hard php dependency, and why is the user not capable of following the wealth of examples and tutorials out there to set up the server? It just seems this is a totally unnecessary package that causes more problems than it purportedly solves. It creates a circular-dependency hell that shouldn't exist in the first place and screws up updates. -- David C. Rankin, J.D.,P.E.

2 2

Recent update
by Marc Chamberlin 29 Nov '24

29 Nov '24

Hello - I am running OpenSuSE 15.6 with the KDE/X11/Plasma desktop and the latest set of updates (looks like mostly nVidia stuff) gave me the following error message - Package nvidia-compute-G06-32bit;550.135-lp156.28.1;x86_64;NVIDIA:repo-non-free is not installed What does this mean and how do I resolve it? Marc...

2 2

plocate updatedb failing to index files in $HOME ? (bug or idiot error?)
by David C. Rankin 29 Nov '24

29 Nov '24

All, I don't have to search for squirreled away files too often, but when I do, I rely on plocate (mlocate successor). However, tonight I was surprised it had not indexed by Readme-doxygen.txt and I can't figure out why. The file I wanted to find has been there since Jan 15: $ l /home/david/doc/linux/doxygen/Readme-doxygen.txt -rw-r--r-- 1 david david 5067 Jan 15 2024 /home/david/doc/linux/doxygen/Readme-doxygen.txt My install has 3 partitons /boot, swap and / and /home is part of / and is mounted: /dev/sda3 / ext4 rw,relatime 0 0 My /etc/updatedb.conf does not exclude /home or ext4 filesystems, e.g. $ noc /etc/updatedb.conf PRUNEFS="9p afs anon_inodefs auto autofs bdev binfmt binfmt_misc ceph fuse.ceph cgroup cifs coda configfs cramfs cpuset debugfs devfs devpts devtmps ecryptfs eventpollfs exofs futexfs ftpfs fuse fusectl gfs gfs2 gpfs hostfs hugetlbfs inotifyfs iso9660 jffs2 lustre misc mqueue ncpfs nfs NFS nfs4 nfsd nnpfs ocfs ocfs2 pipefs proc ramfs rpc_pipefs securityfs selinuxfs sfs shfs smbfs sockfs spufs sshfs subfs supermount sysfs tmpfs ubifs udf usbfs vboxsf vperfctrfs" PRUNEPATHS="/tmp /var/tmp /var/cache /var/lock /var/run /var/spool /mnt /cdrom /usr/tmp /proc /media /sys /.snapshots /var/run/media" PRUNENAMES = ".git .hg .svn .bzr .arch-ids {arch} CVS" PRUNE_BIND_MOUNTS="no" My plocate-updatedb.timer just ran less than an hour ago: $ scs plocate-updatedb ○ plocate-updatedb.service - Update the locate database Loaded: loaded (/usr/lib/systemd/system/plocate-updatedb.service; static5:185m) Active: inactive (dead) since Sun 2024-11-24 00:33:56 CST; 14min ago Invocation: 633dc34e995a4a658e4d66f1f61a0622 TriggeredBy: ● plocate-updatedb.timer Docs: man:updatedb Process: 158324 ExecStart=/bin/sh -c chown -R ${RUN_UPDATEDB_AS}:root /var/lib/plocate && su --shell=/bin/sh ${RUN_UPDATEDB_AS} -c 'umask 0022; /usr/sbin/updatedb' (code=exited, status=0/SUCCESS) Main PID: 158324 (code=exited, status=0/SUCCESS) CPU: 53ms Nov 24 00:33:54 wizard systemd[1]: Starting Update the locate database... Nov 24 00:33:54 wizard su[158324]: (to nobody) root on none Nov 24 00:33:54 wizard su[158324]: pam_systemd(su:session): New sd-bus connection (system-bus-pam-systemd-158324) opened. Nov 24 00:33:54 wizard su[158324]: pam_unix(su:session): session opened for user nobody(uid=65534) by (uid=0) Nov 24 00:33:56 wizard systemd[1]: plocate-updatedb.service: Deactivated successfully. Nov 24 00:33:56 wizard systemd[1]: Finished Update the locate database. 'locate' is properly linked to 'plocate': $ l /usr/bin/locate lrwxrwxrwx 1 root root 7 Jun 12 10:04 /usr/bin/locate -> plocate Yet running 'locate Readme' or 'locate doxygen' does not turn up my file. Why is this? What magic am I missing? -- David C. Rankin, J.D.,P.E.

5 9

Can't install Virtual Box - 15.6 KDE
by James Knott 28 Nov '24

28 Nov '24

I'm trying to Virtual Box but I get this error: /usr/sbin/vboxconfig Building kernel modules... Kernel modules built correctly. They will now be installed. insmod /lib/modules/6.4.0-150600.23.25-default/extra/vboxdrv.ko modprobe: ERROR: could not insert 'vboxnetflt': Key was rejected by service insmod /lib/modules/6.4.0-150600.23.25-default/extra/vboxdrv.ko modprobe: ERROR: could not insert 'vboxnetadp': Key was rejected by service Kernel modules are installed and loaded. When I searched on the first error, it mentioned virtualbox-dkms package, but that does not seem to be available as something I can install. Suggestions? tnx jk

5 30

Re: Leap 15.6 nvidia troubles again - cc instead of gcc?
by Paul Neuwirth 27 Nov '24

27 Nov '24

On Wednesday 2024-11-27 15:40, Patrick Shanahan wrote: >Date: Wed, 27 Nov 2024 15:40:10 >From: Patrick Shanahan <ptilopteri(a)gmail.com> >To: Paul Neuwirth <mail(a)paul-neuwirth.nl> >Cc: users(a)lists.opensuse.org >Subject: Re: Leap 15.6 nvidia troubles again - cc instead of gcc? > > * Paul Neuwirth via openSUSE Users <users(a)lists.opensuse.org> [11-27-24 09:36]: >> Hello together, >> >> after an update today (zypper patch && zypper up) parts of nvidia fail to >> build. >> noticable at start it mentions another compiler is being used than used for >> building the kernel. cc instead of gcc (both 7.5 thou). >> >> # warning: the compiler differs from the one used to build the kernel >> # The kernel was built by: gcc (SUSE Linux) 7.5.0 >> # You are using: cc (SUSE Linux) 7.5.0 >> >> subsequently, everything looks normal until: >> >> # make: command: Command not found >> # Makefile:63: WARNING: Unable to locate the compiler gcc from CONFIG_CC_VERSION_TEXT in the kernel configuration. >> # rm -f -r conftest >> # make[1]: Entering directory '/usr/src/linux-6.4.0-150600.23.17' >> # make[2]: Entering directory '/usr/src/linux-6.4.0-150600.23.17-obj/x86_64/default' >> # CLEAN /usr/src/kernel-modules/nvidia-550.135-default >> # CLEAN /usr/src/kernel-modules/nvidia-550.135-default/Module.symvers >> # make[2]: Leaving directory '/usr/src/linux-6.4.0-150600.23.17-obj/x86_64/default' >> # make[1]: Leaving directory '/usr/src/linux-6.4.0-150600.23.17' >> # / >> # EFI variables are not supported on this system >> # # Modprobe blacklist files have been created at /etc/modprobe.d to prevent >> Nouveau from loading. This can be reverted by deleting >> /etc/modprobe.d/nvidia-*.conf. >> # # *** Reboot your computer and verify that the NVIDIA graphics driver can >> be loaded. *** >> # # sort: write failed: 'standard output': Broken pipe >> # sort: write error >> # /usr/bin/lsinitrd: line 203: echo: write error: Broken pipe >> >> may it be caused by the other compiler? how to fix/change compiler? >> >> if you need full log, I can try. not an expert in vim (yet) - no X available >> on any computer now. >> >> Thank you, >> Regards >> >> Paul > > boot a prior kernel until you have an updated nvidia package > > -- > (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri > facebook/ptilopteri > Photos: http://wahoo.no-ip.org/piwigo paka @ IRCnet oftc > yea, somehow it did work despite the errors!? don't know anymore - only trouble with the proprietary nvidia. after sending the e-mail I tried one more thing: rmmod all loaded nvidia- modules modprobe nvidia and it worked. Display froze at booting (forget to mention that) around mounting the root filesystem. Are there other, better supported (including computation and 3d acceleration, etc.) video cards, I could consider in the future? Thanks again, Paul. PS: # rpm -qa | grep "nvidia-" libva-nvidia-driver-0.0.12-lp155.2.1.x86_64 nvidia-video-G06-550.135-lp156.28.1.x86_64 pcp-pmda-nvidia-gpu-6.2.0-150600.3.9.1.x86_64 nvidia-driver-G06-kmp-default-550.135_k6.4.0_150600.21-lp156.28.1.x86_64 nvidia-gl-G06-550.135-lp156.28.1.x86_64 nvidia-compute-utils-G06-550.135-lp156.28.1.x86_64 nvidia-video-G06-32bit-550.135-lp156.28.1.x86_64 kernel-firmware-nvidia-20240728-150600.3.6.1.noarch nvidia-compute-G06-32bit-550.135-lp156.28.1.x86_64 nvidia-compute-G06-550.135-lp156.28.1.x86_64 nvidia-gl-G06-32bit-550.135-lp156.28.1.x86_64

2 2

Leap 15.6 nvidia troubles again - cc instead of gcc?
by Paul Neuwirth 27 Nov '24

27 Nov '24

Hello together, after an update today (zypper patch && zypper up) parts of nvidia fail to build. noticable at start it mentions another compiler is being used than used for building the kernel. cc instead of gcc (both 7.5 thou). # warning: the compiler differs from the one used to build the kernel # The kernel was built by: gcc (SUSE Linux) 7.5.0 # You are using: cc (SUSE Linux) 7.5.0 subsequently, everything looks normal until: # make: command: Command not found # Makefile:63: WARNING: Unable to locate the compiler gcc from CONFIG_CC_VERSION_TEXT in the kernel configuration. # rm -f -r conftest # make[1]: Entering directory '/usr/src/linux-6.4.0-150600.23.17' # make[2]: Entering directory '/usr/src/linux-6.4.0-150600.23.17-obj/x86_64/default' # CLEAN /usr/src/kernel-modules/nvidia-550.135-default # CLEAN /usr/src/kernel-modules/nvidia-550.135-default/Module.symvers # make[2]: Leaving directory '/usr/src/linux-6.4.0-150600.23.17-obj/x86_64/default' # make[1]: Leaving directory '/usr/src/linux-6.4.0-150600.23.17' # / # EFI variables are not supported on this system # # Modprobe blacklist files have been created at /etc/modprobe.d to prevent Nouveau from loading. This can be reverted by deleting /etc/modprobe.d/nvidia-*.conf. # # *** Reboot your computer and verify that the NVIDIA graphics driver can be loaded. *** # # sort: write failed: 'standard output': Broken pipe # sort: write error # /usr/bin/lsinitrd: line 203: echo: write error: Broken pipe may it be caused by the other compiler? how to fix/change compiler? if you need full log, I can try. not an expert in vim (yet) - no X available on any computer now. Thank you, Regards Paul

1 0

openSUSE-SLE-15.6-2024-4060 conflicts with NVidia G06
by Darryl Gregorash 27 Nov '24

27 Nov '24

So what are we supposed to do with this thing? The patch description is rather cryptic: > This update for ocl-icd fixes the following issues: > > get rid of update-alternatives; it's no longer needed because with upcoming G06 production drivers, the nvidia packages will rely on this implementation (JIRA#PED-11444) > added conflicts to packages for legacy nvidia drivers (G04,G05) Does this mean that G06 is/will be obsolete?

2 1

US Mirrors Down??
by David C. Rankin 25 Nov '24

25 Nov '24

All, Are all US mirrors down? I just tried to list updates for Tumbleweed and got: Retrieving repository 'Main Repository (OSS)' metadata .....................................................[error] Repository 'Main Repository (OSS)' is invalid. [download.opensuse.org-oss|http://download.opensuse.org/tumbleweed/repo/oss/] Failed to retrieve new repository metadata. History: - [|] Error trying to read from 'http://download.opensuse.org/tumbleweed/repo/oss/' - Location 'http://download.opensuse.org/tumbleweed/repo/oss/content' is temporarily unaccessible. Please check if the URIs defined for this repository are pointing to a valid repository. Warning: Skipping repository 'Main Repository (OSS)' because of the above error. Retrieving repository 'Main Update Repository' metadata ....................................................[error] Repository 'Main Update Repository' is invalid. [download.opensuse.org-tumbleweed|http://download.opensuse.org/update/tumble…] Failed to retrieve new repository metadata. History: - [|] Error trying to read from 'http://download.opensuse.org/update/tumbleweed/' - Location 'http://download.opensuse.org/update/tumbleweed/content' is temporarily unaccessible. Please check if the URIs defined for this repository are pointing to a valid repository. Warning: Skipping repository 'Main Update Repository' because of the above error. Retrieving repository 'kde3' metadata ......................................................................[error] Repository 'kde3' is invalid. [kde3|http://download.opensuse.org/repositories/KDE:/KDE3/openSUSE_Tumblewee…] Failed to retrieve new repository metadata. History: - [|] Error trying to read from 'http://download.opensuse.org/repositories/KDE:/KDE3/openSUSE_Tumbleweed/' - Location 'http://download.opensuse.org/repositories/KDE:/KDE3/openSUSE_Tumbleweed/con…' is temporarily unaccessible. Please check if the URIs defined for this repository are pointing to a valid repository. Warning: Skipping repository 'kde3' because of the above error. I'm hoping "temporarily unaccessible" means what it says despite the misspelling. -- David C. Rankin, J.D.,P.E.

5 7

Live USB and partitioner
by Marc Chamberlin 23 Nov '24

23 Nov '24

I am using an OpenSuSE 15.6 live usb stick to help me debug a system that will not boot up. To gain access to the various drives and partitions, of the broken OpenSuSE system, I tried to use the YaST partitioner on the live version of OpenSuSE and was unsuccessful. The partitioner does allow me to edit/define mount points for all of the partitions that are part of the system I am trying to debug. The problem lies with the fact that there is some sort of magic going on with the definition for the root file system '/'. Somehow the live OpenSuSE system is able to mount something??? as the root system, but the YaST partitioner does not show any partition being mounted at the root '/' mount point. And the YaST partitioner will not allow me to exit or set up the mount points until the '/' mount point is defined. So I feel like I am caught in a Catch 22, somehow the '/' mount point is being defined and data from the live USB partitions is being mounted there, but the YaST partitioner doesn't see it and won't allow me to set any mount points until the '/' mount point is defined. The partitioner does show that 2 of the partitions on the live USB stick are defined and mounted. The first has a mount point for the disk label openSUSE_Leap_15.6_KDE_Live defined at /run/overlay/live. The second has a mount point for disk labeled cow defined at /run/overlay/overlay/overlayfs. Interesting also is that /etc/fstab is an empty file. So I am totally confused and don't grok Live USB sticks. How do I mount my own systems partitions on a live USB stick? Thanks, Marc.

3 2