After http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=3b9b8d571da6bb3652427e8ccc7948cbcec0e517, busybox-hwclock is no longer created because it only contains /etc/init.d/hwclock.sh which is removed when DISTRO_FEATURES contains systemd but not sysvinit. The busybox-hwclock RDEPEND in packagegroup-core-boot should be excluded in this case.
Is there an equivlant for the hwclock setting with systemd? Can you provide the correct patch, (I will assign it to you, if you can).
For systemd, you generally use timedatectl to set the date/time (http://www.freedesktop.org/software/systemd/man/timedatectl.html). I don't think there is a service for hwclock in systemd. It is generally recommended for systemd that the RTC is kept as UTC. Setting the date/time with timedatectl will update both the local and RTC time. "timedatectl set-local-rtc 0" is used to set systemd to maintain RTC in universal time. "timedatectl set-local-rtc 1" is used to set systemd to maintain RTC in local time. Note that "timedatectl set-local-rtc 1" gives the following warning: Warning: The RTC is configured to maintain time in the local time zone. This mode is not fully supported and will create various problems with time zone changes and daylight saving adjustments. If at all possible use RTC in UTC, by calling 'timedatectl set-local-rtc 0'. The set-local-rtc option is persisted across reboots. I think busybox-hwclock is just not necessary for systemd. I am wondering what the most concise way would be to test that DISTRO_FEATURES contains systemd but not sysvinit when setting RDEPENDS.
What about modutils-initscripts? I think the same logic applies there. P.S. I just noticed there's a new field 'Document Change' here. I set it as 'Yes (doc changes required)' because I think we may need some documents on systemd and sysv issues. //Chen Qi
Possibly also init-ifupdown needs to be added to that list.
http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=be13994d00ebb9973306fd2045dd142a859a2a62
Hi, What are the impacts to the user for this fix? Is there something specific the user needs to do under certain conditions? Is it simply to be sure that busybox-hwclock is not an IMAGE_FEATURE for a Systemd-based image? I need someone to clearly note the impact(s) to the user here to help me figure out where to update the docs for this fix. Thanks, Scott