no X11 after reboot :-(

[Scott Bennett]

Henry Hu <henry.hu.sh at gmail.com> wrote:
> On Wed, May 13, 2015 at 1:35 PM Scott Bennett <bennett at sdf.org> wrote:
>
> >      On Tue, 12 May 2015 13:36:05 +0200 Jan Bramkamp <crest at rlwinm.de>
> > wrote:
> > > On 12/05/15 08:53, Scott Bennett wrote:
> > > >       A week ago, after my system had been up for a few weeks, I shut
> > it
> > > > down and then rebooted it, after which X11 can't seem to get going.
> > The
> > > > Xorg.0.log ends with
> > > > [ log file removed ]
> > > > Initializing built-in extension DRI2
> > > > Loading extension GLX
> > > > Loading extension NV-GLX
> > > > Loading extension NV-CONTROL
> > > > Loading extension XINERAMA
> > > > /usr/lib/libgomp.so.1: version OMP_3.0 required by
> > /usr/local/lib/libMagickCore-6.so.2 not found
> > > > xinit: connection to X server lost
> > > >
> > > > waiting for X server to shut down (EE) Server terminated successfully
> > (0). Closing log file.
> > > >
> > > > [hellas] 185 %
> > > >
> > > >       What is OMP?  And why does the X server care about what
> > libMagickCore*
> > > > wants in order to start up successfully?  And how do I *fix* this?
> > Being
> > > > limited to a console and virtual consoles (and window(1), thank
> > goodness!)
> > > > for the past week has been quite aggravating.
> > > >       Many thanks in advance to anyone who can show me the way around
> > this
> > > > situation!
> > >
> > > This looks like a broken dependency on a shared lib. The proper fix is
> >
> >      But why is a base system library (/usr/lib/libgomp.so.1) issuing a
> > message about what something in /usr/local/lib wants?  And, I repeat, why
> > does the X server care about libMagickCore* when it is starting up?  Why
> > cannot the server run without it?
> >
> You read it wrong,  the message means that /usr/lib/libgomp.so.1 does not
> satisfy the requirements of /usr/local/lib/libMagickCore-6.so.2.

     Okay.  I understood that the file name before the first colon was the
name of the module issuing the message.  Sigh.  I wonder why libMagickCore*
issues a message in a reversed format.

> I've just checked that mine does not have that tag also, but libgomp.so in
> /usr/local/lib/gcc48 has that tag. Maybe you need to install some version

     Henry, thank you very much for that explanation.  Unfortunately, that
leaves me with no solution at present because, although gcc48 is already
built as lang/gcc, due to a corrupted pkg data base record in
/var/db/pkg/local.sqlite, I cannot install it.  Until I find a way to fix
or recreate the local.sqlite file, there will be no gcc48 on my system, as
I wrote before. :-(  But at least now, thanks to your info, I see the
problem:  I use WindowMaker as my window manager, which lists
libMagickWand-6.so in its LIB_DEPENDS, and that is apparently one of the
libraries from graphics/ImageMagick.

> of gcc to get the runtime library.

     So lang/gcc49 might do it?

> Interestingly, on my machine libMagickCore does not depend on gcc's
> libraries.
>  Do you have option "OPENMP" enabled for ImageMagick port?

     I don't remember offhand, but quite probably so.
>
> It's possible that X server started, but the session fails to start. X
> terminates after the session terminates.
>
> >
> > > to install compatible versions of all executables and shared libs.
> > > Upgrade your system for packages or ports like usual. Run pkg check -B
> >
> >      Well, that is a bit of a problem for me at the moment because the
> > /var/db/pkg/local.sqlite file on my system has somehow become corrupted,
> > and I have not been able so far to find a way to fix it or to rebuild it
> > without the corruption.  (See my posts in freebsd-ports@ in the last 30
> > hours or so.)  Running pkg check -B against the existing local.sqlite
> > didn't turn up a problem with libMagickCore* stuff, nor would it likely
> > find anything regarding a base system library.
> >
> > > afterward (or install sysutils/bsdadminscripts and use pkg_libchk for
> > > more details). If the upgrade did not fix your problem query pkgng for
> > > for affected files (the executable and shared lib) e.g. pkg which $path.
> > >
> >      Thanks for your suggestions.  If I can find a way to fix the package
> > data base corruption, I will attempt to run "portmaster -a" again, but I
> > don't see how that will fix a mismatch between the base system and ports.
> > Port distfiles are supposed to match the major releases of FreeBSD, right?


                                  Scott Bennett, Comm. ASMELG, CFIAG
**********************************************************************
* Internet:   bennett at sdf.org   *xor*   bennett at freeshell.org  *
*--------------------------------------------------------------------*
* "A well regulated and disciplined militia, is at all times a good  *
* objection to the introduction of that bane of all free governments *
* -- a standing army."                                               *
*    -- Gov. John Hancock, New York Journal, 28 January 1790         *
**********************************************************************