Le 27/11/2010 08:43, Dirk Behme a écrit : > Doing a './MAKEALL ARMV7' on the recent master (d4752d5d2...) doesn't > look nice [2].
Globally, this is expected and does not hit only ARMV7 but any ARM board -- I think on the whole MAKEALL arm, only 35 out of 168 boards build without errors or warnings (more do build, but with warnings though). ELF relocation was introduced during this cycle, on all ARM cpus but only on a handful of boards. Others need to be fixed by their maintainers. Typically, those which miss a CONFIG_SYS_SDRAM_BASE are hit by this issue. Note that not fixing all boards is voluntary: those boards whose maintainer will not have fixed during the coming cycle will be removed--as a janitorial matter, one could say. For others -- the break may or may not be a consequence of ELF relocation. This will have to be fixed on an individual basis. > Are there any other pending patches to fix the compilation of the > failing boards? These shall be fixed by their maintainers. Wolfgang, should I send out a (single) mail to each ARM board maintainer right after this cycle asking them to either fix their board or say that they are not going to? This could help weed out during next cycle. > Thanks > > Dirk Amicalement, -- Albert. _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
