On Fri, Jul 11, 2014 at 12:24:15PM -0400, Cliff Brake wrote: > On Fri, Jul 11, 2014 at 12:08 PM, Dmytriyenko, Denys <[email protected]> wrote: > > Unless you also have meta-yocto-bsp layer (before daisy) defining > > beagleboard machine... > > or unless you have meta-gumstix-community included ... > > DEBUG: LOAD > /scratch/oe/oe-build-daisy/sources/meta-ti/conf/machine/beagleboard.conf > DEBUG: CONF > /scratch/oe/oe-build-daisy/sources/meta-ti/conf/machine/beagleboard.conf:5: > including conf/machine/include/omap3.inc > DEBUG: BB > /scratch/oe/oe-build-daisy/sources/meta-gumstix-community/conf/machine/include/omap3.inc: > handle(data, include) > DEBUG: LOAD > /scratch/oe/oe-build-daisy/sources/meta-gumstix-community/conf/machine/include/omap3.inc > > Supporting multiple machines in one build present some interesting > challenges -- what are the best practices in situations like this? > > I guess the good news is I got to try 3.14 on the bb ...
That is interesting indeed. If they felt the need to overwrite the entire include file for the SOC, then they should care to keep it updated or set all necessary defaults. Otherwise they should have used a more unique name, like omap3-gumstix.inc and made necessary customizations in there, keeping the original upstream omap3.inc intact and using or not using it from their machine configs accordingly. For example, I've dealt with u-boot.inc in a similar fashion in meta-ti: http://git.yoctoproject.org/cgit/cgit.cgi/meta-ti/tree/recipes-bsp/u-boot/u-boot-ti.inc -- Denys -- _______________________________________________ meta-ti mailing list [email protected] https://lists.yoctoproject.org/listinfo/meta-ti
