https://gcc.gnu.org/bugzilla/show_bug.cgi?id=93934
--- Comment #11 from Uroš Bizjak <ubizjak at gmail dot com> --- (In reply to Zoltan Vajda from comment #9) > As I understand it, it is acknowledged, that this is a bug. However, the > issue is in state NEW for a quite long time. The issue is still present in > GCC 11.2. Do you see any chances for some progress in this case? The Comment #5 summarises all the troubles with trapping FLDL/FLDS. I'm afraid that trapping FLDL/FLDS renders -fsignaling-nans on x87 practically impossible to implement. Even the documentation advises against usage of -fsignaling-nans: This option is experimental and does not currently guarantee to disable all GCC optimizations that affect signaling NaN behavior. You should really use -mfpmath=sse here.