zmodem wrote: The linked bug does say "reverting it locally fixed the testing failures" which is a pretty strong indication.
I'm afraid we don't have a small reproducer yet, but https://crbug.com/493894786#comment4 and forward has a call stack and disassembly. It looks like we're failing a control-flow guard check when calling a virtual destructor. I'm guessing that relates to the newly emitted vector deleting dtor. We'll keep digging :-) https://github.com/llvm/llvm-project/pull/185653 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
