https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91599
Martin Sebor <msebor at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Keywords| |diagnostic
Status|UNCONFIRMED |NEW
Last reconfirmed| |2019-08-29
CC| |msebor at gcc dot gnu.org
Blocks| |88443
Ever confirmed|0 |1
Severity|enhancement |normal
--- Comment #1 from Martin Sebor <msebor at gcc dot gnu.org> ---
Confirmed. gimple_location(stmt) returns zero for the statement but
EXPR_LOCATION(lhs) has the right location so it can be used as a fallback.
Unfortunately, that alone isn't enough. The %G directive in the
warning_at(loc, "%G...", stmt) call seems to insist on using the stmt location
in preference to loc even when it's invalid/unknown, and so the warning still
doesn't point in the right place. I could hack around it at the call site but
a better fix is in the warning machinery. (Ideally, of course, every statement
would have location and this wouldn't be a problem at all. But what fun would
that be?)
Referenced Bugs:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88443
[Bug 88443] [meta-bug] bogus/missing -Wstringop-overflow warnings
