*** C++ compiler and preprocessor checking for i686-pc-linux-gnu-g++... i686-pc-linux-gnu-g++ checking whether we are using the GNU C++ compiler... yes checking whether i686-pc-linux-gnu-g++ accepts -g... yes checking dependency style of i686-pc-linux-gnu-g++... gcc3 checking how to run the C++ preprocessor... i686-pc-linux-gnu-g++ -E checking if want C++ exception handling... no checking whether i686-pc-linux-gnu-g++ has new(std::nothrow)... no checking for C++ template repository directory... templates not supported? not used checking for C++ compiler template parameters... none needed checking for C++ optimization flags... -O2 -march=i686 -mtune=athlon-xp -msse -mmmx -pipe -ggdb3 checking map usability... yes checking map presence... yes checking for map... yes checking for type bool... no checking for true/false... no configure: WARNING: *** Your C++ compiler does not support the bool data type. configure: WARNING: *** LAM requires a C++ compiler with support for the bool configure: WARNING: *** data type. configure: error: cannot continue
Created attachment 151287 [details, diff] 7.1.4-gcc43.patch Fixes the configure system.
Created attachment 151289 [details] config.log The specific error messages. I'll mail upstream and point them at this bug. Debian are just patching $S/configure, but that's not the Correct and Authorized way to do it. Especially when we're doing eautoreconf.
The patch is valid and work. Only, I seem to have stepped in some libtool foo.
Hrmmph... This is the problem I ran into. *** [Gentoo] sanity check failed! *** *** libtool.m4 and ltmain.sh have a version mismatch! *** *** (libtool.m4 = 1.5.26, ltmain.sh = 1.5.22) *** I solved it by adding these lines to the ebuild, right after eautoreconf: for oldltmain in $(grep -l 1.5.22 `find ./ -name 'ltmain.sh'`) ;do cp config/ltmain.sh ${oldltmain} done Not pretty, but it works.
On the lam mailing list I was referred by this post: http://www.lam-mpi.org/MailArchives/lam/2008/04/13673.php to this patch: http://www.lam-mpi.org/MailArchives/lam/att-13641/m4-sq-brackets.patch Which does the same, only better.
Just added the fixes to cvs for this, should hit the mirrors soon. I don't have gcc-4.3.0 on any machines to test against so please reopen if you hit any problems. Thanks for the patch and report. Extra thanks for emailing upstream :)