Description of problem: Until initscripts 9.25, clock was configure at start by udev (/lib/udev/rules.d/88-clock.rules). 88-clock.rules was dropped recently upstream http://git.fedorahosted.org/git/?p=initscripts.git;a=commit;h=73d7cf7234f4c903e13de28d71a839afff9b93ca because it causes issues in conjunction with systemd (e.g. mga bug 2521), and is not installed anymore with initscripts 9.34. But these udev rules are still needed when using sysVinit in order to restore correctly the clock. As we will support both systemd and sysVinit in Mga 2, I propose to reintroduce 88-clock.rules, but adding in it a test to not apply its rules when systemd is used (TEST=="/sys/fs/cgroup/systemd" ..., syntax from 70-udev-acl.rules). Version-Release number of selected component (if applicable): initscripts-9.34-2.mga2, sysvinit-2.87-10.mga1 How reproducible: Steps to Reproduce: 1. configure HW clock in localtime 2. boot using sysVinit 3.
CC: (none) => dmorganec, mageia, tmb
Created attachment 1146 [details] 88-clock.rules with test on systemd
Keywords: (none) => PATCHCC: (none) => thierry.vignaudBlocks: (none) => 2120Assignee: bugsquad => mageia
Seems like a valid approach to me. I'll add this back in with your additional safety check.
Status: NEW => ASSIGNED
*** Bug 2521 has been marked as a duplicate of this bug. ***
CC: (none) => davidwhodgins
I am running Mageia 2 from the KDE Live CD, as a hard disk install. I had the same symptoms as in Bug 2521: clock was off by the amount of my timezone difference. I added the rule in the attachment to my rules.d. It was a complete fix.
CC: (none) => laidlaws
CC: (none) => guillomovitchSummary: clock is not set properly when using sysVinit and hw clock in localtime => clock setting issue with sysVinit
I'll try and get this merged today...
Merged now. Thanks... only a week late which is pretty good for me :D
Resolution: (none) => FIXEDStatus: ASSIGNED => RESOLVED
See Also: (none) => https://bugs.mageia.org/show_bug.cgi?id=26990
This is a very old bug. Mageia no longer uses SysV. Should the bug be closed?