[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#195920: missing linker flags for texk/xdvik/ on GNU/FreeBSD



On 07.07.03 Robert Millan (zeratul2@wanadoo.es) wrote:
> On Mon, Jul 07, 2003 at 04:45:06PM +0200, Hilmar Preusse wrote:

Hi,

http://bugs.debian.org/195920

> > > > > on GNU/FreeBSD, libXpm has a library dependency on libXt,
> > > > > so you need a -lXt flag for linking the texk/xdvik stuff
> > > > > with libXpm, otherwise the linking fails:
> > > > 
> > > > No, that's the wrong solution.  If libXpm needs libXt, it
> > > > should be linked to it.  This patch is entirely bogus.
> > > 
> > > I see... then we probably hit the same linker bug we have on
> > > GNU/Hurd.
> > > 
> > > will have a closer look.
> > > 
> > Any progress here?
> 
> No, but it's probably not the same bug. I guess when Xfree86
> stabilises for GNU/FreeBSD we might have a clearer view on the
> problem. I'll look into this when I hack on Xfree86 again.
> 
I've read the latest changelog of xfree:

  * Fix default library build defines to not build the (X)Render,
    Xft, Xft1, or Xcursor libraries if BuildLibraries is not "YES". 
    This prevents gratuitous library builds when, e.g.,
    BuildServersOnly is set to YES.  Clarify and improve related
    patch annotation.  Also, fix the Imake configuration for Debian
    GNU/Linux, the Hurd, and GNU/FreeBSD to not force
    BuildXft1Library to YES unconditionally.  These changes resolve a
    FTBFS in environments where libext-dev is not installed, since
    the Xft1 library was getting gratuitously built during the
    debugging server build, and couldn't find the Xext library to
    link against.  As a minor bonus, this should speed the build up a
    little, since a library that isn't needed is no longer being
    built.

Does that fix your bug?

> If you or anyone wants to help, i could give an account in my
> public GNU/FreeBSD box. See
> http://people.debian.org/~rmh/gnu-freebsd/pub/ACCESS
> 
Well, I don't understand too much about the whole story. It wouldn't
be very helpful.

Thanks anyway,
  Hilmar 
-- 
sigmentation fault



Reply to: