ojhunt wrote: > I'm testing changes as you suggested. One thing I have noticed is that I need > to duplicate much of the code from getMangledNameImpl(). Some of it like the > CUDA & multiversion doesn't apply to z/OS other parts like should you mangle > or not and (possibly) the unique internal linkage do apply. Should CGM expose > a function like getMangledNameImpl() but under a different name?
What/where is the duplication? And yes - the intention would be for CGM to expose a function that does a "get the mangled name, ignoring asm labels" that then gets the getMangledNameImpl behind the scenes https://github.com/llvm/llvm-project/pull/209215 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
