https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64010
--- Comment #7 from Peter A. Bigot <pab at pabigot dot com> --- I don't know if https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64160 is related; it appears to be a similar problem affecting a value returned in multiple registers. It's present both before and after the alternative patch was applied (now that the code that reset the target has been fixed).