I need to compile Boost for an embedded target running GNU/Linux on an ARM CPU. The toolchain for this is proprietary, compiler executable names are like "arm-none-linux-gnueabi-g++". I'm having trouble compiling with this toolchain though - I can't figure out (even with the documentation) what I'm doing wrong, but AFAICT Boost.Build keeps using my default gcc toolchain. In case it's relevant I'm on Cygwin using a self-compiled bjam executable, Boost is v1.34.1 My user-config.jam looks like this: <<<<<<<<<< # Compiler configuration # using gcc ; using gcc : : arm-none-linux-gnueabi-gcc ; using g++ : : arm-none-linux-gnueabi-g++ ; # Python configuration using python : 2.5 : /usr ;
>>>>
A dry run of the compilation with command: ./bjam --with-iostreams --with-regex --with-serialization --with-test --with-thread --toolset=gcc stage -n Produces a lot of output along these lines: <<<<<<<<<< Building Boost.Regex with the optional Unicode/ICU support disabled. Please refer to the Boost.Regex documentation for more information (don't panic: this is a strictly optional feature). ...patience... ...found 2215 targets... ...updating 977 targets... [--- snipped a load of 'MkDir1' lines ---] gcc.compile.c++ bin.v2/libs/serialization/build/gcc-3.4.4/release/threading-multi/basic_archive.o "g++" -ftemplate-depth-128 -O3 -finline-functions -Wno-inline -Wall -mthreads -DBOOST_ALL_NO_LIB=1 -DBOOST_SERIALIZATION_DYN_LINK=1 -DNDEBUG -I"." -c -o "bin.v2\libs\serialization\build\gcc-3.4.4\release\threading-multi\basic_archive.o" "libs\serialization\src\basic_archive.cpp"
>>>>
Why is it continuing to use "g++" when i've specified a different g++ executable to be used? Can anyone help me please? TIA, --rob