Quoting Julien Cristau <[EMAIL PROTECTED]>: > On Sun, Mar 23, 2008 at 10:44:25 +0100, Brice Goglin wrote: > >> * symbols conflicts, probably won't ever be fixed. > > uh? I don't think that should be the case. It shouldn't be too hard to > mark internal symbols for each driver with _X_INTERNAL so they aren't > exported, or build with -Bsymbolic or similar.
I think you mean't _X_HIDDEN. Using internal or private visibility is most likely to cause obscure bugs. Talking about it, I have posted patches and updated the related bugzilla entries several times, to allow compiling everything with -fvisibility=hidden; this includes fixing several symbol name clashes. But I don't think it would be the proper solution for this problem, because the usage of hidden symbols should be an option, and if not enabling it, it should just make all symbols visible. I think the proper solution should be to have a single AtomBios module that would be shared by radeon and radeonhd. And probably compile it in the server tree, somehow following the pattern of the different int10 backends. > Cheers, > Julien Paulo _______________________________________________ xorg-driver-ati mailing list [email protected] http://lists.x.org/mailman/listinfo/xorg-driver-ati
