Anastasia added inline comments.
================ Comment at: test/SemaOpenCL/extension-begin.cl:26 + #ifndef IMPLICIT_INCLUDE #include "extension-begin.h" ---------------- sidorovd wrote: > Anastasia wrote: > > sidorovd wrote: > > > Anastasia wrote: > > > > Can we also test that macro `my_ext` is not defined here but defined > > > > above? > > > > > > > > It seems we are not testing anything like this... > > > #pragma OPENCL EXTENSION my_ext : begin doesn't define an appropriate > > > macro. And so cl-ext=+my_ext. > > But don't you need to expose the definition of it? > Certainly I need, but now the only proper way to do so is by adding an > extension via adding it in OpenCLExtensions.def. Previously we decided to > avoid adding an extension directly into clang, so with a new approach I'd > prefer not to add a definition of the macro in the header but define it > somewhere else, otherwise the macro becomes defined where it's not supposed > to be (even for ARM and AMD =) ). However, my understanding is that you should define the macro when you define the extension itself. ``` #pragma OPENCL EXTENSION my_ext : begin #define my_ext ... #pragma OPENCL EXTENSION my_ext : end ``` does it not work for you? CHANGES SINCE LAST ACTION https://reviews.llvm.org/D58666/new/ https://reviews.llvm.org/D58666 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
