Re: [trinity-devel] kdewebdev - kafkacommon.cpp:745:56: error: using temporary as lvalue [-fpermissive]

Le Mon, 16 May 2011 14:51:11 -0500, "David C. Rankin"<drankinatty(a)suddenlinkmail.com> a écrit : > On 05/16/2011 01:51 AM, /dev/ammo42 wrote: >> Le Sun, 15 May 2011 18:01:31 -0500, >> "David C. Rankin"<drankinatty(a)suddenlinkmail.com> a écrit : >> >>> On 05/15/2011 06:13 AM, /dev/ammo42 wrote: >>>> Le Sat, 14 May 2011 22:11:41 -0500, >>>> "David C. Rankin"<drankinatty(a)suddenlinkmail.com> a écrit : >>>> >>>>> Tim, all, >>>>> >>>>> kdevelop build fine, but I ran into a lvalue problem in >>>>> kdewebdev. I'm building with: >>>>> >>>>> msg "Starting cmake..." >>>>> cmake ${srcdir}/${_svnmod} \ >>>>> -DCMAKE_INSTALL_PREFIX=${trinity_prefix} \ >>>>> -DCMAKE_VERBOSE_MAKEFILE=ON \ >>>>> -DQT_VERSION=3 \ >>>>> -DCMAKE_CXX_FLAGS="-fpermissive" \ >>>>> -DWITH_T1LIB=ON \ >>>>> -DWITH_LIBPAPER=ON \ >>>>> -DWITH_TIFF=ON \ >>>>> -DWITH_OPENEXR=ON \ >>>>> -DWITH_PAM=ON \ >>>>> -DBUILD_ALL=ON >>>>> make >>>>> >>>>> The error is: >>>>> >>>>> [ 77%] Building CXX object >>>>> quanta/parts/kafka/CMakeFiles/kafkalibrary-static.dir/kafkacommon.cpp.o >>>>> cd /build/src/build/quanta/parts/kafka&& /usr/bin/c++ >>>>> -DHAVE_CONFIG_H -DQT_NO_ASCII_CAST -DQT_CLEAN_NAMESPACE >>>>> -DQT_NO_STL -DQT_NO_COMPAT -DQT_NO_TRANSLATION >>>>> -DQT_THREAD_SUPPORT -D_REENTRANT -march=x86-64 -mtune=generic >>>>> -O2 -pipe -include tqt.h -I/build/src/build/quanta/parts/kafka >>>>> -I/build/src/kdewebdev/lib -I/build/src/kdewebdev/quanta/src >>>>> -I/build/src/kdewebdev/quanta/project >>>>> -I/build/src/kdewebdev/quanta/parsers >>>>> -I/build/src/kdewebdev/quanta/utility >>>>> -I/build/src/kdewebdev/quanta/treeviews -I/opt/trinity/include >>>>> -I/opt/qt/include -I/opt/qt/include/tqt -include tqt.h -fPIC >>>>> -o CMakeFiles/kafkalibrary-static.dir/kafkacommon.cpp.o >>>>> -c /build/src/kdewebdev/quanta/parts/kafka/kafkacommon.cpp /build/src/kdewebdev/quanta/parts/kafka/kafkacommon.cpp: >>>>> In static member function 'static void >>>>> kafkaCommon::fitsNodesPosition(Node*, int, int, int, >>>>> int)': /build/src/kdewebdev/quanta/parts/kafka/kafkacommon.cpp:745:56: >>>>> error: using temporary as lvalue >>>>> [-fpermissive] /build/src/kdewebdev/quanta/parts/kafka/kafkacommon.cpp:746:55: >>>>> error: using temporary as lvalue >>>>> [-fpermissive] /build/src/kdewebdev/quanta/parts/kafka/kafkacommon.cpp:747:57: >>>>> error: using temporary as lvalue >>>>> [-fpermissive] /build/src/kdewebdev/quanta/parts/kafka/kafkacommon.cpp:748:56: >>>>> error: using temporary as lvalue >>>>> [-fpermissive] /build/src/kdewebdev/quanta/parts/kafka/kafkacommon.cpp:752:56: >>>>> error: using temporary as lvalue >>>>> [-fpermissive] /build/src/kdewebdev/quanta/parts/kafka/kafkacommon.cpp:753:57: >>>>> error: using temporary as lvalue [-fpermissive] make[2]: *** >>>>> [quanta/parts/kafka/CMakeFiles/kafkalibrary-static.dir/kafkacommon.cpp.o] >>>>> Error 1 make[2]: Leaving directory `/build/src/build' make[1]: >>>>> *** [quanta/parts/kafka/CMakeFiles/kafkalibrary-static.dir/all] >>>>> Error 2 make[1]: Leaving directory `/build/src/build' make: *** >>>>> [all] Error 2 ==> ERROR: A failure occurred in build(). >>>>> Aborting... >>>>> >>>>> I'm not that swift at debugging, but I know enough to >>>>> know an lvalue problem -- ain't in my build script :p >>>> The code is obviously wrong, I even wonder why older gcc allowed >>>> such code. >>>> Does this kludge^Wpatch help (the patch is attached)? >>>> >>> >>> Patch applied, but it looks like it got worse :) >> Oops, bad programming on my side, my C-formatted mind told me that >> attrs >> + index ==&attrs[index] :D >> Does this patch help better [still an attached one] ? >>> >>> [ 10%] Building CXX object >>> quanta/utility/CMakeFiles/utility-static.dir/quantacommon.cpp.o

> > Hmm.. Nope... didn't work either :) > > [ 10%] Building CXX object > quanta/utility/CMakeFiles/utility-static.dir/quantacommon.cpp.o > cd /build/src/build/quanta/utility&& /usr/bin/c++ > -DHAVE_CONFIG_H

It stops at 10% because I had to edit a file other than the one to add a kludgey method ­— and the method I wrote ain't right. Does the (again) corrected and attached version of the patch work better ? >