http://bugzilla.novell.com/show_bug.cgi?id=584484
http://bugzilla.novell.com/show_bug.cgi?id=584484#c14
Jiří Suchomel changed:
What |Removed |Added
----------------------------------------------------------------------------
Priority|P5 - None |P2 - High
Version|Factory |Milestone 4
--- Comment #14 from Jiří Suchomel 2010-03-31 08:55:53 UTC ---
I've reproduced it finally. It is not related to NTP.
Every call of "/sbin/hwclock --systz --localtime --noadjfile && touch
/dev/shmwarpclock" moves the time, and the calls are repeated on each enter of
the timezone dialog.
When I have local time in BIOS hwclock --localtime reports it correctly, and
with the original way used in YaST also the system time was set correctly (when
UTC/localtime button is appropriately checked). But now, hwclock --systz moves
the system time for 2 hours every time it is called.
Original way was to use /sbin/hwclock --hctosys --localtime and this works
well.
So, what's the problem with "hwclock --systz..." call? Does it really have to
be called only after mkinitrd? We are talking about showing the time during
first stage of the installation, and here mkinitrd is not available. If
mkinitrd must be called before, than we have to move "hwclock --systz" call to
the very end of installation.
Does it need to be called at all?
Recently we've introduced it to SLE11SP1 as well, seems like the problem will
be also there, so we'd have to find solution rather fast.
--
Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.