https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91388
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Assignee|unassigned at gcc dot gnu.org |pinskia at gcc dot
gnu.org
Status|NEW |ASSIGNED
--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
So I think the fix is to gate pass_warn_function_return on error_count. Because
we don't know 100% the state of the IR coming in. As errors are already removed
by the time we get there and that means things like an error on static_assert
or and error on COMPOUND_EXPR (like in PR 119149), there are no markers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60440#c9 definitely makes it sound
like it is the same issue too.
Which makes me thing we should not warn if there has been an error already.
