Created attachment 355360 [details, diff] ebuild diff needs gpm, has missing blockers on emul-libs: will enter the tree masked and get proper blockers upon unmasking please review carefuly as this one is non trivial
Created attachment 355362 [details] the ebuild itself might be easier to read that way
Maybe patches stuff from src_unpack should be moved to src_prepare ;)
Created attachment 355412 [details, diff] ebuild diff right, fixed also moved the tc-export & friends to the right place (src_configure vs multilib_src_configure) and reordered the function definitions in an imho more natural way
Created attachment 355414 [details] the ebuild itself
maybe we could move the gpm[multilib_usedep] dep to PDEPEND to try to avoid circular deps with gpm since its dlopened. opinions?
(In reply to Alexis Ballier from comment #5) > maybe we could move the gpm[multilib_usedep] dep to PDEPEND to try to avoid > circular deps with gpm since its dlopened. opinions? We can't. Even with --with-dlsym, the code still does some checks on the headers and library.
ping @maintainers: I will commit this in a few days unless you state here you want a "delay" for reviewing.
in cvs gpm moved to pdepend and --with-gpm=libgpm.so.1 works without multilib gpm for non native abis.