aaron.ballman added inline comments.
================
Comment at: clang/lib/Sema/SemaConcept.cpp:1065-1068
assert(TC &&
"TPL must have a template type parameter with a type constraint");
auto *Constraint =
+ cast<ConceptSpecializationExpr>(TC->getImmediatelyDeclaredConstraint());
----------------
mlychkov wrote:
> aaron.ballman wrote:
> > If we're going to be touching this code, there's more suspect code here
> > that needs to be cleaned up a bit. Directly above this is:
> > ```
> > const TypeConstraint *TC =
> > cast<TemplateTypeParmDecl>(TPL->getParam(0))->getTypeConstraint();
> > assert(TC &&
> > "TPL must have a template type parameter with a type constraint");
> > ```
> > That assertion can be removed entirely -- if the `cast<>` fails, it doesn't
> > return null, it asserts.
> Actually, cast return nonnull value, but
> TC = "cast_result"->getTypeConstraint();
> and TC may become nullptr, I suppose.
Good catch! I think we should add the assertion on `TC` being nonnull back.
I'll take care of that.
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D108481/new/
https://reviews.llvm.org/D108481
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
