And it requires a trivial patch that would be affected by Linux systems and a faily major patch that can only be applid to FreeBSD systems as it adds FreeBSD libs to the Makefile. As such, two patches will follow - they are based on the patches in FreeBSD ports (and one from steev) and they enable system-monitor to run correctly.
Created attachment 103948 [details, diff] trivial patch
Created attachment 103949 [details, diff] FreeBSD only patch Should only be applied with use kernel_FreeBSD
Created attachment 104287 [details, diff] FreeBSD patch New patch should work fine with other OS's too
Created attachment 104289 [details, diff] trivial patch Same as before, just with the configure patch removed as it's been included in the new FreeBSD patch above
upstream bug http://bugzilla.gnome.org/show_bug.cgi?id=387200
Created attachment 107376 [details, diff] FreeBSD patch Patch for 2.14.6
The upstream bug has a comment, that I happen to agree with. It really appears that result is declared, then checked without being set...
Upstream have merged my patch here. http://bugzilla.gnome.org/attachment.cgi?id=81514&action=view Can we get this in portage pretty please :)
Done. Can you test it to make sure I got my use check correct? I'd prefer to not have to run eautoreconf everywhere, since it adds ~1/4 again to the merge time... I did test the the patch and eautoreconf works correctly with the if statement commented out.
(In reply to comment #9) > Done. > > Can you test it to make sure I got my use check correct? I'd prefer to not > have to run eautoreconf everywhere, since it adds ~1/4 again to the merge > time... WFM Blessed with the ~x86-fbsd keyword :)
