Date: Tue, 22 Feb 2005 16:18:51 -0500 From: Mikhail Teterin <Mikhail.Teterin@murex.com> To: Felix =?iso-8859-1?q?K=FChling?= <fxkuehl@gmx.de> Cc: acardenas@bsdperu.org Subject: Re: ports/76257: nvidia_driver breaks xorg-clients build Message-ID: <200502222109.j1ML995v026956@harik.murex.com> In-Reply-To: <1109106485.6323.13.camel@trabant> References: <200502220508.j1M58MAN094098@corbulon.video-collage.com> <200502221710.j1MH9a5v009787@harik.murex.com> <1109106485.6323.13.camel@trabant>
next in thread | previous in thread | raw e-mail | index | archive | help
> The attached patch (undo my other patch first) implements something like > this without actually loading libGL dynamically. It uses > dlopen(NULL, ...) to get a handle for the main program (including > libGL). Yes, this is what I had in mind. > I hope this is portable. Does this work for you? I've seen `configure' scripts checking "whether dynamically linked program can dlopen itself". It can on Linux and FreeBSD... > Note that the patch also changes the Imakefile. In xc/programs/xdriinfo > run this to update the Makefile: > > make Makefile > make depend > > Then recompile xdriinfo as usual. Thanks! I'll try it, unless someone beats me to it. One thing, however (for now) -- when reporting failures of dlopen/dlsym, why not use dlerror() function to provide actual error message? Yours, -mi
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200502222109.j1ML995v026956>