[Bug target/94934] Failure to inline addv

jakub at gcc dot gnu.org gcc-bugzilla@gcc.gnu.org
Mon May 18 09:32:31 GMT 2020


https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94934

--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
(In reply to Gabriel Ravier from comment #3)
> In that case, it looks really easy to reimplemnet `-ftrapv` as literally
> just enabling `-fsanitize=signed-integer-overflow
> -fsanitize-undefined-trap-on-error`.

It shouldn't be literally that, because -ftrapv should be able to coexist with
other UBSan sanitizers (just not signed-integer-overflow) and e.g. let the
other UBSan sanitizers have different decision on what to do on error (the
trap. vs. fatal error vs. just diagnostics and continue).
And for -ftrapv we should consider using the library functions rather than
emitting code inline if it would be too large to match past behavior.


More information about the Gcc-bugs mailing list