sammccall added a comment. It should be possible to test this by adapting a test case from the original ed8dadb <https://reviews.llvm.org/rGed8dadb37c7e1a7f4889d868ac9b19bfe7762237>, but I'm not certain...
================ Comment at: clang-tools-extra/clang-tidy/ClangTidyDiagnosticConsumer.cpp:525 + case clang::DiagnosticsEngine::ak_addrspace: + Builder << Qualifiers::getAddrSpaceAsString( + *reinterpret_cast<LangAS *>(Info.getRawArg(Index))); ---------------- Not very familiar with this code, but it looks like this is feeding the arg back into a builder, and ed8dadb added support for sending LangAS itself to the builder. So I don't think we should stringify it, but just send the LangAS ================ Comment at: clang-tools-extra/clang-tidy/ClangTidyDiagnosticConsumer.cpp:526 + Builder << Qualifiers::getAddrSpaceAsString( + *reinterpret_cast<LangAS *>(Info.getRawArg(Index))); + break; ---------------- I think the stored thing is a LangAS (an enum), rather than a pointer to one. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D71486/new/ https://reviews.llvm.org/D71486 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
