From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga03.intel.com (mga03.intel.com [143.182.124.21]) by mail.openembedded.org (Postfix) with ESMTP id 8516D6C6B0 for ; Fri, 20 Sep 2013 14:55:42 +0000 (UTC) Received: from azsmga002.ch.intel.com ([10.2.17.35]) by azsmga101.ch.intel.com with ESMTP; 20 Sep 2013 07:55:44 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.90,944,1371106800"; d="scan'208";a="297320649" Received: from unknown (HELO helios.localnet) ([10.252.123.234]) by AZSMGA002.ch.intel.com with ESMTP; 20 Sep 2013 07:55:43 -0700 From: Paul Eggleton To: Martin Jansa Date: Fri, 20 Sep 2013 15:55:41 +0100 Message-ID: <4410837.umzcAyviZD@helios> Organization: Intel Corporation User-Agent: KMail/4.10.5 (Linux/3.8.0-30-generic; KDE/4.10.5; i686; ; ) In-Reply-To: <20130920144205.GJ18414@jama> References: <1379682086-27877-1-git-send-email-paul.eggleton@linux.intel.com> <1689545.vp0O8sHNEO@helios> <20130920144205.GJ18414@jama> MIME-Version: 1.0 Cc: openembedded-core@lists.openembedded.org Subject: Re: [PATCH] qt4-x11-free: disable QGtkStyle X-BeenThere: openembedded-core@lists.openembedded.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Patches and discussions about the oe-core layer List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 20 Sep 2013 14:55:44 -0000 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="iso-8859-1" On Friday 20 September 2013 16:42:05 Martin Jansa wrote: > On Fri, Sep 20, 2013 at 03:27:32PM +0100, Paul Eggleton wrote: > > On Friday 20 September 2013 15:58:26 Andreas M=FCller wrote: > > > On Fri, Sep 20, 2013 at 3:01 PM, Paul Eggleton > > >=20 > > > wrote: > > > > This currently leads to a floating dependency on GTK+. Since mo= st > > > > users > > > > in an embedded context won't need this, just disable it by defa= ult. > > > >=20 > > > > Fixes [YOCTO #5116]. > > > >=20 > > > > Signed-off-by: Paul Eggleton > > > > --- > > > >=20 > > > > meta/recipes-qt/qt4/qt4-x11-free.inc | 2 +- > > > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > >=20 > > > > diff --git a/meta/recipes-qt/qt4/qt4-x11-free.inc > > > > b/meta/recipes-qt/qt4/qt4-x11-free.inc index 4b29f55..369b3d5 1= 00644 > > > > --- a/meta/recipes-qt/qt4/qt4-x11-free.inc > > > > +++ b/meta/recipes-qt/qt4/qt4-x11-free.inc > > > > @@ -12,7 +12,7 @@ INC_PR =3D "r50" > > > >=20 > > > > QT_GLFLAGS ?=3D "${@base_contains('DISTRO_FEATURES', 'opengl',= > > > > '-opengl', > > > > '-no-opengl', d)} " QT_GLFLAGS_qemux86 =3D "-opengl" > > > > QT_GLFLAGS_qemuppc =3D "-opengl" > > > >=20 > > > > -QT_X11_FLAGS ?=3D "-no-xinerama -no-xkb" > > > > +QT_X11_FLAGS ?=3D "-no-xinerama -no-xkb -no-gtkstyle" > > > >=20 > > > > QT_CONFIG_FLAGS +=3D "${QT_X11_FLAGS}" > > > > QT_BASE_LIB ?=3D "libqt" > > >=20 > > > How about PACKAGECONFIG for this? > >=20 > > Ultimately I decided against introducing PACKAGECONFIG just for thi= s, > > because the Qt recipes currently avoid it (PACKAGECONFIG cannot sup= port a > > number of the additional library options for Qt since they are tri-= state; > > not to mention that PACKAGECONFIG isn't just a drop-in for Qt becau= se it > > doesn't use autotools/cmake and therefore doesn't use > > EXTRA_OECONF/EXTRA_OECMAKE). I figured that the few users who do wa= nt > > QGTKStyle can just set QT_X11_FLAGS as desired. >=20 > FWIW: I'm using PACKAGECONFIG in meta-qt5, you're right that it canno= t > cover all 3 states for each option, but most common scenarios need to= > select only between 2 and automatically added DEPENDS/RDEPENDS is big= > plus. On this wider issue, I have given some thought to how we could expand=20= PACKAGECONFIG to support multi-state options, although I'm hesitant to = make=20 the already slightly opaque PACKAGECONFIG syntax any more complicated.=20= Something to think about further for the next development cycle perhaps= . =20 > To support PACKAGECONFIG without autotools/cmake inherit I just add > generated EXTRA_OECONF to QT_CONFIG_FLAGS. I didn't try it but Peter reported that this caused a failure when he t= ried=20 this as a solution (presumably because there is junk in EXTRA_OECONF al= ready=20 that isn't suitable for passing to Qt4's configure). It ought to be abl= e to=20 work though. Cheers, Paul --=20 Paul Eggleton Intel Open Source Technology Centre