Am 03/09/13 17:02, schrieb Dimitry Andric: > On Mar 9, 2013, at 16:36 , "Hartmann, O." <ohartman_at_zedat.fu-berlin.de> wrote: >> I have one specific FreeBSD 10.0-CURRENT box (FreeBSD 10.0-CURRENT #0 >> r248061: Fri Mar 8 19:44:30 CET 2013 amd64) which rejects to build >> either lang/gcc or lang/gcc46 with the very same error shown below. > … >> checking whether cc supports -pedantic -Wno-long-long... yes >> checking dependency style of cc... none >> configure: error: no usable dependency style found >> gmake[2]: *** [configure-stage1-libcpp] Error 1 >> gmake[2]: Leaving directory `/usr/ports/lang/gcc/work/build' >> gmake[1]: *** [stage1-bubble] Error 2 >> gmake[1]: Leaving directory `/usr/ports/lang/gcc/work/build' >> gmake: *** [all] Error 2 >> *** [do-build] Error code 1 > > What is the actual error in the resulting config.log file? Unfortunately autoconf error message are virtually information-free… > > My first guess would be that you have non-default CFLAGS in your build environment, which confuse gcc's build stages. Hello. CFLAGS settings are either root_at_thor:/usr/ports/lang/gcc # make -VCFLAGS -O2 -pipe -march=native -I/usr/local/include -fno-strict-aliasing or root_at_thor:/usr/ports/lang/gcc # make -VCFLAGS -O2 -pipe -O3 -march=native -I/usr/local/include -fno-strict-aliasing There are several config.log files: ./work/build/config.log ./work/build/lto-plugin/config.log ./work/build/gcc/config.log ./work/build/intl/config.log ./work/build/libcpp/config.log ./work/build/build-x86_64-portbld-freebsd10.0/libiberty/config.log ./work/build/build-x86_64-portbld-freebsd10.0/fixincludes/config.log ./work/build/libiberty/config.log Most recent touched is ./work/build/libcpp/config.log and since libcpp occurs in the error, it is the requested log file. I can not see anything useful in that file - I'm sorry :-( It is attached to this post. If you need further material, feel free to ask. By the way, I copied my /etc/make.conf and /etc/src.conf to another CURRENT machine which has the very same revision status (as well as the /usr/src AND /usr/ports, both boxes are up to date even with "portmaster" run this morning). I do not see the problem on this other machine.Received on Sat Mar 09 2013 - 15:57:35 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:35 UTC