Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 22 Nov 2012 13:23:55 +0100
From:      Florian Smeets <flo@smeets.im>
To:        Stefan Esser <se@freebsd.org>
Cc:        freebsd-ports@freebsd.org
Subject:   Re: Firefox build failure on -CURRENT (with CLANG and with GCC)
Message-ID:  <50AE195B.9030409@smeets.im>
In-Reply-To: <50AE1445.2090900@freebsd.org>
References:  <50AE1445.2090900@freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On 11/22/2012 13:02, Stefan Esser wrote:
> Hi Florian,
> 
> seems that Firefox cannot be built with CLANG. The build fails due to a
> visibility default set in mozilla-release/config/gcc_hidden.h, which is
> in conflict with Qt declarations:
> 
> In file included from
> work/mozilla-release/intl/locale/src/nsLocaleService.cpp:7:
> In file included from /usr/local/include/qt4/QtCore/QString:1:
> In file included from /usr/local/include/qt4/QtCore/qstring.h:46:
> /usr/local/include/qt4/QtCore/qbytearray.h:143:7: error: visibility does
> not match previous declaration
> class Q_CORE_EXPORT QByteArray
>       ^
> /usr/local/include/qt4/QtCore/qglobal.h:1412:27: note: expanded from
> macro 'Q_CORE_EXPORT'
> #    define Q_CORE_EXPORT Q_DECL_EXPORT
>                           ^
> /usr/local/include/qt4/QtCore/qglobal.h:1269:42: note: expanded from
> macro 'Q_DECL_EXPORT'
> #    define Q_DECL_EXPORT __attribute__((visibility("default")))
>                                          ^
> ../../../config/gcc_hidden.h:6:13: note: previous attribute is here
> #pragma GCC visibility push(hidden)
> 
> Seems that this GCC pragma has different semantics in GLANG ...
> 
> If the contents of gcc_hidden.h is commented out, CLANG completes the
> build, but the installation fails in the phase where hyphenation tables
> are built:
> 
> 
> Any ideas how to get Firefox compiled and installed on -CURRENT?
> 

Yes, don't use the experimental QT4 option. :) Firefox builds just fine
on head with clang and the default options, that's what I'm using here
right now.

We'll have to look into the QT4 failure.

Thanks for the report.
Florian



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?50AE195B.9030409>