https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111669
Xi Ruoyao <xry111 at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Severity|normal |enhancement
--- Comment #7 from Xi Ruoyao <xry111 at gcc dot gnu.org> ---
(In reply to Zeb Figura from comment #6)
> It is my impression that gcc is interested in avoiding false positives for
> its warnings.
Correct, but we are also interested in avoiding false negatives. Without extra
information provided by something like __builtin_unreachable, any change
decreasing false positives will increase false negatives (unless the false
positive is completely stupid: for the simplified test case I think the false
positive not completely stupid, but maybe it is completely stupid for your
original program).
> It is also my impression that -Wnonnull is not *supposed* to emit warnings
> for cases where, from the compiler's point of view, NULL might be passed,
> but some high-level invariant prevents this. Compare -Wmaybe-uninitialized,
> where the documentation clearly specifies otherwise.
Maybe we can separate -Wnonnull into -Wmaybe-nonnull and -Wnonnull, or just
make -Wnonnull not to emit warnings for conditional paths and tell users
expecting a nonnull warning in conditional paths to use the analyzer (it's very
supposed to warn even in conditional paths) instead.
> If both of these impressions are incorrect, this bug report can be closed as
> WONTFIX.
I'll keep it open but make it an enhancement.
