[Lldb-commits] [PATCH] D75626: Add support for owning module information to TypeSystemClang.

Pavel Labath via Phabricator via lldb-commits lldb-commits at lists.llvm.org
Thu Mar 5 01:09:02 PST 2020


labath added a comment.

Thanks for splitting this up. This seems fine to me. I'll leave the approval to the type system folks.

I don't know how much we care about the size of the Type class. If we did, there are ways this could be optimized. But I guess the fact that we stored `m_is_complete_objc_class` as a lone bool flag  (and so changing that to a uint32_t doesn't cost anything) means that we don't really care that much about it.



================
Comment at: lldb/source/Plugins/TypeSystem/Clang/TypeSystemClang.cpp:1445-1446
       ast, (TagDecl::TagKind)kind,
-      decl_ctx, // What decl context do we use here? TU? The actual decl
-                // context?
+      decl_ctx, // What decl context do we use here? TU? The actual
+                            // decl context?
       SourceLocation(), SourceLocation(), &identifier_info);
----------------
bad formatting


================
Comment at: lldb/source/Plugins/TypeSystem/Clang/TypeSystemClang.cpp:7621
         isDefined, impControl, HasRelatedResultType);
+        SetMemberOwningModule(setter, class_interface_decl);
 
----------------
here too


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D75626/new/

https://reviews.llvm.org/D75626





More information about the lldb-commits mailing list