https://gcc.gnu.org/bugzilla/show_bug.cgi?id=93952
--- Comment #1 from Frédéric Recoules <frederic.recou...@univ-grenoble-alpes.fr> --- (In reply to Frédéric Recoules from comment #0) > Now I wonder if the code where we replace the constraint by "rm" is valid > because the returned value depends of the constraint the compiler have > chosen. Ok, it seems that if the compiler have the choice between "r" and "m", it returns the same result whatever it chooses. I do shake my head a bit because I don't understand why solo "m" is treated in a different way.
