https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86120
kargl at gcc dot gnu.org changed: What |Removed |Added ---------------------------------------------------------------------------- Summary|SIGSEGV without backtrace |ICE caused by unassociated |when using submodules |pointer in SIZE intrinsic --- Comment #4 from kargl at gcc dot gnu.org --- Just realized the original subject of this PR is nonsense. There isn't a submodule in the attached code. Updating to something a little more reasonable.