cmtice wrote: > I'm not intimately familiar with this code, but I am somewhat suspicious of > implementation. > > Adding the number of base classes to the result makes sense to me. What > surprises me is that this should be done only for fields with anonymous > types. Can you explain why should the index of a field depend on whether its > type is anonymous or not?
I knew that I needed to add it for anonymous structs; I wasn't sure about other cases. I've tested it more now, and it seems good to add it in all cases, so I will do that. > > I am also surprised by the hardcoding of `omit_empty_base_classes` to true in > the `GetNumBaseClasses` call. All of the other calls in this function pass > the value from the argument. Why should this one be special? I knew that in the case of anonymous structs we did not want to count empty base classes; since this line was for anonymous structs, I hard-coded it to make sure. Now that I've updated the code to always add the number of base classes, I'm removing that hard-coding. https://github.com/llvm/llvm-project/pull/138487 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
