qt-x11-free-3.0.5 compile error
declan.moriarty at ntlworld.ie
Wed Jul 30 05:20:00 PDT 2003
On Fri, Jul 25, 2003 at 09:16:09PM +0100, Matthew Burgess enlightened us thusly
> On Fri, 25 Jul 2003 20:49:11 +0100
> Matthew Burgess <ca9mbu at hermes.sunderland.ac.uk> wrote:
> > On Fri, 25 Jul 2003 18:51:56 +0000 (UTC)
> > declan.moriarty at ntlworld.ie ("Declan. Moriarty") wrote:
> > > /usr/src/qt-x11-free-3.0.5/include/qvaluevector.h: In method
> > > `QValueVectorPrivate<Element>::QValueVectorPrivate(const
> > > QValueVectorPrivate<Element> &)':
> > > /usr/src/qt-x11-free-3.0.5/include/qvaluevector.h:476:
> > > instantiated from `QValueVector<Element>::detachInternal()'
> > > /usr/src/qt-x11-free-3.0.5/include/qvaluevector.h:459:
> > > instantiated from here
> > > /usr/src/qt-x11-free-3.0.5/include/qvaluevector.h:125: Internal
> > > compiler error.
> > > /usr/src/qt-x11-free-3.0.5/include/qvaluevector.h:125: Please submit
> > > a full bug
> > > report.
> > You're trying to compile a C++ library with gcc-2.95.3? Good luck...
> > :)
> > As you are probably no doubt aware it's ability to handle anything
> > resembling standards compliant C++ code is fairly minimal. Anyway,
> > onto the point in hand, I assumed you have tried to re-run make after
> > receiving the error. I believe that Internal Compiler Errors (ICEs)
> > can occur for a number of reasons, although admittedly certain code
> > constructs (in particular C++ templates) can be the cause so it may
> > well fail at the same point next time around. If you have gcc-3.x I'd
> > strongly suggest compiling QT with that instead.
> > A search of google turns up nothing (as I'm sure you're aware) other
> > than that some other poor sods have had exactly the same failure
> > (please forgive the wrapping in the google url):
> > http://www.google.com/search?q=qvaluevector.h+%22Internal+compiler+error%22&hl=en&lr=&ie=UTF-8&oe=UTF-8
> There's a response to the Rosegarden message hosted on Geocrawler.com
> (linked to by the above google URL) by the OP which suggest that
> removing the -O2 from the compiler flags may fix it.
I'm getting somewhere, but I'm not sure where :-/
On this box, qt is about a 6 hour compile. I haven't been able to let it
continue through, and anyhow it barfs halfway with the above error.
This compile requires the $QTDIR variable to be source top dir.
When I go straight back in, I get an error because qmake won't execute -
some permission denied thing even though I'm running as root. If I set
$QTDIR to the installed version, I'm away, but I'm not happy that this
is going to work as I'm using different options.
Does that compile set other variables that need to be set up before I
can resume a make on it?
With best Regards,
More information about the blfs-support