pedrosland
commented
6 years ago Thanks for the report.
It seems that clang can't find its own include files. The documentation says that the default location for the missing file is $(dirname /path/to/tool)/../lib/clang/3.3/include.
My clang is at /usr/lib/llvm-3.9/bin/clang. This would put my stddef.h in /usr/lib/llvm-3.9/lib/clang/3.9.1/include/. There is one there but it seems I'm actually using one from gcc.
I think that there is another problem with your approach. It looks like you're running this on your (non-arm) computer and not a Raspberry Pi. Installing llvm on your computer will generate the bindings as if for your computer unless you use a llvm toolchain that will cross-compile. For this, you will need some variant of apt-get install llvm-3.9-dev libclang-3.9-dev clang-3.9 and to set up your paths so that the build.rs can find it.
Why do you not just use the pre-generated bindings? If there is something wrong with them I can try to correct it.
eberkund
I am getting the following error when running with the "generate-bindings" feature.
I have installed clang according to the bindgen docs
But I get this error