mgorny added a comment. In https://reviews.llvm.org/D26249#585869, @zturner wrote:
> Is this necessary because of a standalone build, or for some other reason? > If this is to get Standalone build working, then it seems like this isn't > really a complete solution, because standalone build still wouldn't work on > Windows since it would still be trying to include the private header. Yes, it is. However, since I don't have a Windows environment, I can't really test any Windows solution. With the very limited support for stand-alone builds right now, I see this as a feasible improvement to make tests work on Gentoo ;-). > What about making a separate unittest file called `SymbolFilePDBDIATests.cpp` > and excluding it at the CMake level? I don't mind doing that; however, it won't fix the issue in question. CMake doesn't know the state of HAVE_DIA either since it is available only during LLVM build. If we want to have Windows support there, we probably need to copy the check result to the public LLVMConfig.h. https://reviews.llvm.org/D26249 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
