summaryrefslogtreecommitdiffstats
path: root/core
Commit message (Collapse)AuthorAgeFilesLines
* [slx-brightness] Add new module to control backlit displaysSimon Rettberg2024-03-117-0/+418
| | | | | | | | | We do this raw by reading /dev/input/eventX to make sure no running program (e.g. VMware Player) can grab the keyboard and prevent any other running daemon (e.g. xfce-power-manager) from seeing the keystrokes. Another advantage is that this will work everywhere (tty, lightdm, VM, desktop session) since the daemon starts on system bootup.
* [hardware-stats] Cleanup python scriptSimon Rettberg2024-03-041-54/+67
| | | | | Make it look less like it was written by a sysadmin using python for the first time.
* [hardware-stats] Not needed with ./dataSimon Rettberg2024-03-041-2/+0Star
|
* [hardware-stats] Move collect script to this repoSimon Rettberg2024-03-043-8/+588
|
* [etherwake] udev: Mask exit code 76 from ethtoolSimon Rettberg2024-03-012-2/+9
|
* [vmware-version-check/nvidia-libs@] Silence non-critical errors in ↵Simon Rettberg2024-03-013-6/+15
| | | | addon-required
* [vmware-common] Silence warning if SHARE_REMAP_MODE is emptySimon Rettberg2024-02-291-1/+1
|
* [run-virt] Don't have writelog at this pointSimon Rettberg2024-02-291-1/+2
|
* [xorg] Implement has_option in XsessionSimon Rettberg2024-02-291-1/+21
|
* [dnbd3-proxy-mode] Make local-switch-service Type=simpleSimon Rettberg2024-02-161-1/+3
| | | | | | | This will not wait for the script to finish, like with oneshot, to consider the job done and thus the system being booted up fully. Because this really is a background job that isn't vital to the system's operation.
* [vmware.*] Try to avoid chroot in addon-required check scriptSimon Rettberg2024-02-165-50/+114
|
* [remote-access] Disable xfce's compositingSimon Rettberg2024-02-141-0/+7
| | | | This avoids the screen only refreshing at 1fps with no active outputs.
* [slxlog] Use is_debug for checking SLX_DEBUG modeSimon Rettberg2024-02-081-1/+2
|
* [run-virt] Use SLX_DEBUG, like slxlog does, by switching to is_debugSimon Rettberg2024-02-081-2/+2
|
* [vmware-common] Fix keys getting stuck on VMware 17.5.xSimon Rettberg2024-02-061-0/+2
|
* [dnbd3-proxy-mode] local proxy: Skip last iteration of forced switchSimon Rettberg2024-02-021-1/+1
| | | | | | | This looks ugly in systemd-analyze plot - the service takes forever. The remaining aproaches should be enough to cover the bootup process - if we need something more permanent, dnbd3 should be extended to forbid switching unless the server is actually unreachable.
* [vbox-src] Add progress output, skip udev reload with MaxiLinuxSimon Rettberg2024-02-021-3/+11
|
* Make useradd and groupadd return 0Simon Rettberg2024-02-012-5/+6
| | | | | These functions will perror on error, so make sure on success we return 0, even though checking the return value is useless...
* [vbox-src] Need bashSimon Rettberg2024-02-011-1/+1
|
* [libvirt-src/libvirt] Add demo to groups at build time, not run timeSimon Rettberg2024-02-013-13/+6Star
|
* [vbox-src] Add vboxusers at build time, add demo to groupSimon Rettberg2024-02-013-2/+16
|
* [disk-partitions] Refactor gettmpSimon Rettberg2024-02-011-20/+16Star
|
* [disk-partitions] Run everything in parallelSimon Rettberg2024-02-011-47/+50
|
* [dnbd3-proxy-mode] Add user dnbd3 to fuse group at build timeSimon Rettberg2024-02-013-3/+4
|
* [redsocks] config is downloaded in stage 3, no need to waitSimon Rettberg2024-02-011-7/+0Star
|
* [run-virt] Use tcp-reset for TCP firewallingSimon Rettberg2024-01-301-3/+6
|
* [vmware-common] SimplifySimon Rettberg2024-01-301-3/+1Star
|
* [disk-partitions] Don't wait for udev if we have disks from stage 3Simon Rettberg2024-01-302-1/+2
| | | | | | | | When running the new dracut-init approach, we already loaded all the disk controller modules and scanned the buses. No need to wait for udev settle etc. in stage 4 again. For the old (unused) stage3.1 + stage3.2 approach, add a service droplet that keeps the old order.
* [run-virt] Fix corrupted vmx with additional NICsSteffen Ritter2024-01-301-3/+4
| | | | | | | Don't accidentally add an empty line to TMPCONFIG. Also make awk ignore empty lines when looking for duplicates so we don't end up with an ' = ' line which will crash vmplayer. Bonuspoint: sort the output for better readability
* [disk-partitions] Scan for ID83 in background; skip ro-partitions for ↵Simon Rettberg2024-01-291-4/+8
| | | | rw-type use cases
* [disk-partitions] Don't wait for udev if we scanned for HDDs in stage 3Simon Rettberg2024-01-291-5/+8
|
* [bwlp-stage4-tweaks] +xSimon Rettberg2024-01-291-0/+0
|
* [bwlp-stage4-tweaks] Add service to grow rootfs to blockdev sizeSimon Rettberg2024-01-293-0/+63
| | | | | | | | This was previously done in initrd, but it turns out this was happening rather late and blocked the switchroot for a couple seconds. Let's do it in stage 4 instead as early as possible, and order it before gather-hw-info.service, as that service will determine the rootfs free space and needs to see the properly resized rootfs.
* [run-virt] Split mounting of /tmp/virt into own serviceSimon Rettberg2024-01-295-32/+51
|
* Rename hardware-stats service to gather-hw-infoSimon Rettberg2024-01-298-8/+7Star
|
* [dhcpc-busybox] Remove check for non-existent varSimon Rettberg2024-01-291-5/+0Star
| | | | | | | | | Nobody else ever checks or sets SLX_DHCP_SECONDARY. We already have SLX_DHCP_OTHER_NICS to control whether to run a DHCP client on other interfaces. This happens in systemd-bridge_additional_nics where we conditionally spawn udhcpc instances for appropriate interfaces if the above variable is set. No need for another check in the dhcpc client wrapper script.
* [system-check/hardware-stats] Loosen dependenciesSimon Rettberg2024-01-298-47/+36Star
| | | | | | | | | | | | | For some reason, system-check is ordered before hardware-report, but I can't figure out why. system-check merely generates the warnings to display on the login screen, and hardware-report sends hw data to the boot server. I don't see anything in hardware-report actually depending on anything system-check does. There are some other minor changes, like moving the generation of the json report from hardware-report to hardware-stats, and putting the according python script into a directory named hardware-stats instead of system-check, which might imply it belongs to that module which is wrong.
* UTF-8Simon Rettberg2024-01-293-5/+5
|
* [kdm/kdm-theme-bwlp] RemoveSimon Rettberg2024-01-2949-1026/+0Star
|
* Enable bwlp-desktop-tweaks in bwlp targetSimon Rettberg2024-01-231-0/+1
|
* [debug-report-bwlp] Include special bwlp log directorySteffen Ritter2024-01-222-1/+2
| | | | | | | | | | Sometimes we need more debug output, so we extend our scripts with additional logging information. But where to store such logs without always overwriting 'debug_report' binary or hacking log file names to somethin we already include? So we include a special '/tmp/bwlp_debug_logs' directory where we could store and report e.g. 'set -x' output from our scripts.
* [vmware-common] Improve handling of syntax errors in vmxSimon Rettberg2024-01-191-4/+6
|
* [run-virt] Make copy of downloaded vmx before passing it on to pluginsSimon Rettberg2024-01-191-9/+17
| | | | | This makes sure the debug_report contains the unmodified version as well as the modified version of the configuration for the VM session.
* [dhcpc-busybox] Don't start via udev for additional nicsSimon Rettberg2024-01-181-2/+3
| | | | This depends on having SLX_DHCP_OTHER_NICS enabled.
* [idleaction] Delete rm of unused flag-fileSimon Rettberg2024-01-171-6/+0Star
|
* [vmchooser2] Properly implement the re-autologin delaySimon Rettberg2024-01-171-7/+3Star
|
* [kiosk-common] Remove unused flag fileSimon Rettberg2024-01-171-4/+0Star
|
* [vmware17] Also always use patched modules if availableSimon Rettberg2024-01-171-1/+1
|
* [hardware-stats] Change low memory warning conditionsSimon Rettberg2024-01-171-2/+5
|
* [nvidia-common] Remove unused varSimon Rettberg2024-01-161-3/+1Star
|