================
@@ -670,14 +670,39 @@ class LifetimeSafetySemaHelperImpl : public
LifetimeSafetySemaHelper {
const Expr *LastExpr = OriginExprChain.back();
std::string IssueStr = getDiagSubjectDescription(LastExpr);
+ std::optional<LifetimeBoundParamInfo> HiddenLifetimeBound;
----------------
iitianpushkar wrote:
The reason I added this was because some call/constructor expressions are part
of the origin flow chain but are not printed in the alias chain because they
have the same source range as another expression.
The below test failed because of that :
```cpp
template <typename T> struct [[gsl::Pointer]] Pointer {
Pointer(const T &bar [[clang::lifetimebound]]);
};
void nested_local_pointer() {
Pointer<Pointer<Pointer<Bar>>> ppp;
Pointer<Pointer<Bar>> pp;
Pointer<Bar> p;
{
Bar v;
p = Pointer(v); // expected-warning {{local variable 'v' does not live
long enough}}
pp = Pointer(p); // expected-note {{local variable 'p' aliases the
storage of local variable 'v' because parameter 'bar' is marked
'lifetimebound'}}
ppp = Pointer(pp); // expected-note {{local variable 'pp' aliases the
storage of local variable 'v'}}
} // expected-note {{local variable 'v' is destroyed
here}}
use(***ppp); // expected-note {{later used here}}
}
```
For pp = Pointer(p), the lifetimebound info comes from the implicit constructor
call Pointer(p) but that step can be skipped by shouldShowInAliasChain because
it shares the same source range as the visible alias step. Without carrying the
info over, the diagnostic becomes only:
```
local variable 'p' aliases the storage of local variable 'v'
```
The value is reset immediately after emitting the next visible alias note, so
it cannot leak further down the chain.
I see your concern that this may attach the explanation to a visible edge that
is not exactly the hidden edge. If you think that is too confusing, I can
switch to the more conservative behaviour and only print the lifetimebound
explanation when the currently visible edge itself re-derives it.
What are your thoughts on this ?
Cc @Xazax-hun
https://github.com/llvm/llvm-project/pull/206337
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
