https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66461
--- Comment #4 from Jerry DeLisle <jvdelisle at gcc dot gnu.org> --- Yes, I was looking at this in gdb lst night and I am really suspicious of the comment about guaranteed to match. I tried with what you did only I did not return at that point. The fact that it has to do with fixed form continuation had me digging in scanner.c. I also noticed that the expr returned after what looks like a good free_expr (good meaning no ICE) leaves the contents of the expr as garbage. I wonder if the free_expression should set all the sub component types in the structure to valid values to avoid these issues.
