Endill added inline comments.
================
Comment at: clang/lib/Parse/ParsePragma.cpp:731-736
+ } else if (E.get()->isTypeDependent()) {
+ PP.Diag(StartLoc, diag::warn_pragma_debug_type_dependent_argument)
+ << SourceRange(StartLoc, Tok.getLocation());
+ } else if (E.get()->isValueDependent()) {
+ PP.Diag(StartLoc, diag::warn_pragma_debug_value_dependent_argument)
+ << SourceRange(StartLoc, Tok.getLocation());
----------------
aaron.ballman wrote:
> Endill wrote:
> > aaron.ballman wrote:
> > >
> > I see quite a lot of items inside `ExprDependence`. Are you sure we don't
> > miss any corner cases by using `isValueDependent()` to select between
> > value- and type-dependent diagnostic text?
> I was trying to think of a case where we'd run into problems, but I can't
> come up with one, so I believe we're fine here.
Apparently, we should do it the other way round, because `An id-expression is
value-dependent if: — it is type-dependent` ([[
http://eel.is/c++draft/temp.dep#constexpr-2.2 | temp.dep.constexpr/2.2]]). I
observe that in one of the tests I added, `T{}` is TypeValueInstantiation.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D144115/new/
https://reviews.llvm.org/D144115
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
