| 1 |
On Friday 24 April 2009 10:01:29 Duncan wrote: |
| 2 |
> Markos Chandras <hwoarang@g.o> posted |
| 3 |
> 200904240940.01333.hwoarang@g.o, excerpted below, on Fri, 24 Apr |
| 4 |
> |
| 5 |
> 2009 09:39:52 +0100: |
| 6 |
> > qmake will read the QMAKESPEC variable. This variable is 'usually' set |
| 7 |
> > as /usr/share/qt4/mkspecs/default which is a symlink to the actual |
| 8 |
> > mkspec. So my symlink is right |
| 9 |
> > |
| 10 |
> > That is |
| 11 |
> > |
| 12 |
> > Eternity ~ # ls -la /usr/share/qt4/mkspecs/|grep default lrwxrwxrwx 1 |
| 13 |
> > root root 12 2009-04-23 19:33 default -> linux-g++-64 |
| 14 |
> |
| 15 |
> But could it be failing to dereference the symlink? IOW, is it trying to |
| 16 |
> match "default" rather than "linux-g++-64"? I've seen that error a |
| 17 |
> number of times and it's a hard one to catch in the logic flow if one's |
| 18 |
> dereference assumptions are incorrect. |
| 19 |
> |
| 20 |
According to qt-creator upstream , qmake is always looking for default symlink |
| 21 |
in order to identify the correct mkspec. So , internally, qmake will run |
| 22 |
something like this |
| 23 |
|
| 24 |
qmake QMAKESPEC=/usr/share/qt4/mkspec/default qt-creator.pro |
| 25 |
|
| 26 |
According to generated makefiles, QMAKESPEC is set to /usr/share/qt4/linux- |
| 27 |
g++-64 , but that path is wrong. It should be /usr/share/qt4/mkspecs/linux- |
| 28 |
g++-64. |
| 29 |
But that shouldnt be a problem as long as the mkspec host is identified |
| 30 |
corrently. |
| 31 |
|
| 32 |
Gosh, what a mess |
| 33 |
> > Of course it needs further investigation but it is rather complicated |
| 34 |
> > to do this right now |
| 35 |
> |
| 36 |
> Understood. |
| 37 |
> |
| 38 |
> The problem just interested me enough to see what I could make of it, and |
| 39 |
> I posted what I came up with in the hope that it might be of help to you |
| 40 |
> or anyone else deciding to take it further... which decision would depend |
| 41 |
> on having the time, of course. If you/we are lucky, some other user with |
| 42 |
> a bit of time and technical skills is reading, and will find the |
| 43 |
> discussion intriguing enough to followup on and file a bug with a patch. |
| 44 |
> =:^) |
| 45 |
> |
| 46 |
> Meanwhile, as you said, "the workaround is quite safe and doesn't cause |
| 47 |
> any further problems :)" |
| 48 |
|
| 49 |
-- |
| 50 |
Markos Chandras (hwoarang) |
| 51 |
Gentoo Linux Developer |
| 52 |
Qt/KDE/Sunrise/Sound |
| 53 |
Web: http://hwoarang.silverarrow.gr |
Archives