W dniu 16.01.2019 o 02:25, Philip Prindeville pisze: > Inline > >> On Jan 15, 2019, at 12:40 AM, Tomasz Maciej Nowak <tome...@o2.pl> wrote: >> >> Hi Philip >> >> W dniu 14.01.2019 o 21:10, Philip Prindeville pisze: >> >>> Why is the hwclock being dropped everywhere? >> >> It's provided by busybox and installed by default, also greping for its >> usage in target/linux/x86/base-files returned nothing. Am I missing >> something, is the full fledged one used somewhere or by something? > > > It’s used at least two places. > > When you first get the board and bring it up and set the clock, since it’s > not usually set correctly by the factory, and it’s used by the ntpd script to > save the synchronized time when the system is taken down.
But it uses options available in busybox hwclock so I still don't see a reason to add it. > >> >>> And who says the geos2, net5501, or alix2 are obsolete? >> >> They are not obsolete by any means, only the way to assign packages for >> specific device changes. Please take a look at patch nr 2 in this RFC and >> section adding target/linux/x86/image/geode.mk. There You have almost all >> packages replicated with additional geos device profile. Now You'll get two >> images, one generic, and one for geos. This distinction is here because with >> old way of adding packages, the generic image would end up with lot of >> useless, geos/dsl specific packages. > > > So why is there a Device/geos but none of the other SBC’s? None other SBC is so specific (I followed what packages were specified in profiles), and for other most drivers are enabled in kernel config. > And why doesn’t the geos DEVICE_PACKAGES include -kmod-via-rhine, since it > uses the kmod-8139cp instead? I can remove the call for generic profile and add all relevant packages if that is preferred. Anyway both of these drivers are enabled in kernel config as built-in and those kmod packages are empty. In next version of this series I'll remove kmod packages which are enabled in kernel. > > -Philip Please give this series a test. Regards, Tomasz. > > > >> >>> I have some running right here… >> >> I have WRAP which runs perfectly fine with these changes. >> -- TMN _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel
