On Mon, May 19, 2003 at 09:30:52AM +0200, Andreas Tille wrote: > On Sat, 17 May 2003, Junichi Uekawa wrote: > > > > It's strange, because on my i386 system > > > > objdump -T /usr/lib/libgdk-X11-2.0.so gives > > > > > > > > 0002e71c g DF .text 00000049 Base _gdk_display_x11_get_type > > > > 00045ebc g DF .text 0000014c Base _gdk_windowing_window_init > > > > 0005017c g DF .text 00000106 Base XineramaIsActive > > > > 00010a20 g DF .init 00000000 Base _init > > > > 0002b8b0 g DF .text 00000041 Base gdk_image_type_get_type > > > > 00050284 g DF .text 000001ba Base XineramaQueryScreens > > > > 00000000 DF *UND* 000000c3 g_get_charset > > > > 00026e3c g DF .text 00000097 Base _gdk_screen_close > > > > > > > > and that probably means the symbol should be defined within that library. > > > > > > On crest (m68k) this results in: > > > > > > 00000000 D *UND* 00000000 XineramaIsActive > > > 00016970 g DF .init 00000000 Base _init > > > 0002ffd4 g DF .text 0000004c Base gdk_image_type_get_type > > > 00000000 D *UND* 00000000 XineramaQueryScreens > > > 00000000 DF *UND* 000000c6 g_get_charset > > > 0002b8fa g DF .text 0000009a Base _gdk_screen_close > > > 00000000 DF *UND* 00000148 XkbSetDetectableAutoRepeat > > > > > > Any idea why it seems to be undefined (*UND*) on m68k? > > > > Noone seems to have replied to this mail yet, but > > for this kind of thing to happen, a possible cause would be > > a XineramaQueryScreens function that is expected to be > > defined in a static library (and linked into the > > shared library) was once provided as a shared library. > > > > libXinerama.a is only provided as a static library, > > but apparently, on m68k, I suspect some symbols were provided > > through some shared library, at some time. > Thanks for having a look at problem. I have to admit that I have no idea > what to do to fix the problem. Sorry, should've replied to this thread a bit earlier. The exact problem you're quoting (the Xinerama thing), being a bug in libgtk2.0 on m68k only, was of course known to us m68k porters. It turned out to be a problem with logtool, which has now been fixed. The new libgtk2.0 package is linked correctly again. However, that's not why the bug was filed. There are a lot of undefined references to other functions in the bugreport, that have nothing to do with the (now fixed) Xinerama-problem. You may want to check them out. -- Debian GNU/Linux -- http://www.debian.org Nederlandstalige Linux-documentatie -- http://nl.linux.org "An expert can usually spot the difference between a fake charge and a full one, but there are plenty of dead experts." -- National Geographic Channel, in a documentary about large African beasts.
Description: PGP signature