dblaikie added a comment. In D144181#4133056 <https://reviews.llvm.org/D144181#4133056>, @Michael137 wrote:
> In D144181#4133025 <https://reviews.llvm.org/D144181#4133025>, @dblaikie > wrote: > >> Ah, accidentally posted to the lldb part of this stack... instead: >> >> Any chance we can make these work more like member functions (could the >> ctors include their mangled names, for instance)? Or is it the innate nature >> of ctors having the various C1 <https://reviews.llvm.org/C1>/C2/etc versions? > > Initially we tried that in https://reviews.llvm.org/D143652. The existence of > multiple constructor definitions which aren't linked to the `DISubprogram` > declaration makes it tough. We need to start with a pretty expensive search > through the index for all the possible definitions. But then we need to > somehow choose the right one to take the linkage name from. And that context > isn't available at the point where LLDB parses DWARF. > > I'll post some numbers of how much space this would take with Adrian's > suggestion and go from there Hmm, I'd sort of still be inclined to look further at the linkage name option - but maybe the existence of a situation where the usage is built with debug info, but the out of line ctor definitions are all in another library without debug info is a sufficiently motivating use case to justify the addition of these attributes to the ctor declaration. *shrug* Carry on then. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D144181/new/ https://reviews.llvm.org/D144181 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
