[Qt] Make it possible to build without QtTest/QtPrintSuport
Created attachment 157241 [details] Patch
Comment on attachment 157241 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=157241&action=review > Tools/QtTestBrowser/launcherwindow.cpp:727 > +#if !defined(QT_NO_PRINTER) && defined(HAVE_QTPRINTSUPPORT) && HAVE_QTPRINTSUPPORT Why not just HAVE(QPRINTSUPPORT)? > Tools/Tools.pro:12 > + contains(DEFINES, HAVE_QTTESTLIB=1): SUBDIRS += DumpRenderTree/qt/DumpRenderTree.pro I'd like some sort of warning here, like in default_pre, that we're disabling DRT due to missing testlib.
(In reply to comment #2) > (From update of attachment 157241 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=157241&action=review > > > Tools/QtTestBrowser/launcherwindow.cpp:727 > > +#if !defined(QT_NO_PRINTER) && defined(HAVE_QTPRINTSUPPORT) && HAVE_QTPRINTSUPPORT > > Why not just HAVE(QPRINTSUPPORT)? It was a copy-paste of what I did in launcherwindow.h, where HAVE() macro is not defined. But in fact, the code in the header can be moved to the cpp. > > > Tools/Tools.pro:12 > > + contains(DEFINES, HAVE_QTTESTLIB=1): SUBDIRS += DumpRenderTree/qt/DumpRenderTree.pro > > I'd like some sort of warning here, like in default_pre, that we're disabling DRT due to missing testlib. Only for DRT, or WTR and tests too ?
Comment on attachment 157241 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=157241&action=review > Source/WebKit/qt/Api/qwebframe.cpp:1468 > +#if HAVE(QTPRINTSUPPORT) I think here and in the other function it might be good to have an #else UNUSED_PARAM(printer) #endif or the like, to avoid a compiler warning.
(In reply to comment #4) > (From update of attachment 157241 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=157241&action=review > > > Source/WebKit/qt/Api/qwebframe.cpp:1468 > > +#if HAVE(QTPRINTSUPPORT) > > I think here and in the other function it might be good to have an > > #else > UNUSED_PARAM(printer) > #endif > > or the like, to avoid a compiler warning. There is no warning since -Wno-unused-parameter is added to QMAKE_CXXFLAGS in default_post.prf.
Comment on attachment 157241 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=157241&action=review >>> Source/WebKit/qt/Api/qwebframe.cpp:1468 >>> +#if HAVE(QTPRINTSUPPORT) >> >> I think here and in the other function it might be good to have an >> >> #else >> UNUSED_PARAM(printer) >> #endif >> >> or the like, to avoid a compiler warning. > > There is no warning since -Wno-unused-parameter is added to QMAKE_CXXFLAGS in default_post.prf. Ok >>> Tools/Tools.pro:12 >>> + contains(DEFINES, HAVE_QTTESTLIB=1): SUBDIRS += DumpRenderTree/qt/DumpRenderTree.pro >> >> I'd like some sort of warning here, like in default_pre, that we're disabling DRT due to missing testlib. > > Only for DRT, or WTR and tests too ? All three of them I'd say, or more generally - if done in default_pre.prf like suggested - visible early when WebKit.pro is parsed.
(In reply to comment #6) > >>> Tools/Tools.pro:12 > >>> + contains(DEFINES, HAVE_QTTESTLIB=1): SUBDIRS += DumpRenderTree/qt/DumpRenderTree.pro > >> > >> I'd like some sort of warning here, like in default_pre, that we're disabling DRT due to missing testlib. > > > > Only for DRT, or WTR and tests too ? > > All three of them I'd say, or more generally - if done in default_pre.prf like suggested - visible early when WebKit.pro is parsed. I added it in features.prf, to have the message and test at the same place. It's printed in configure step, just after pkg-config tests. Tell me if you prefer to have it in default_pre.prf instead.
Created attachment 157900 [details] Patch
Comment on attachment 157900 [details] Patch Clearing flags on attachment: 157900 Committed r125377: <http://trac.webkit.org/changeset/125377>
All reviewed patches have been landed. Closing bug.