r219029 - Suppress defined-but-unused warnings

Jingyue Wu jingyue at google.com
Sat Oct 4 11:40:20 PDT 2014


Hi Markus,

I am not familiar with this piece of code, but I think the assert is worth
keeping. It verifies function resolveKind does not make ResolveKind
ambiguous if it wasn't ambiguous before. Am I correct?

Jingyue

On Sat Oct 04 2014 at 3:42:59 AM Markus Trippelsdorf <markus at trippelsdorf.de>
wrote:

>
>
> On 2014.10.03 at 22:16 -0000, Jingyue Wu wrote:
> > Author: jingyue
> > Date: Fri Oct  3 17:16:40 2014
> > New Revision: 219029
> >
> > URL: http://llvm.org/viewvc/llvm-project?rev=219029&view=rev
> > Log:
> > Suppress defined-but-unused warnings
> >
> > by adding a fake use
>
> Why not get rid of WasAmbiguous altogether? It does nothing AFAICS.
>
> diff --git a/clang/include/clang/Sema/Lookup.h b/clang/include/clang/Sema/
> Lookup.h
> index 067e42210fde..b88b9682788e 100644
> --- a/clang/include/clang/Sema/Lookup.h
> +++ b/clang/include/clang/Sema/Lookup.h
> @@ -425,10 +425,8 @@ public:
>        }
>      } else {
>        AmbiguityKind SavedAK;
> -      bool WasAmbiguous = false;
>        if (ResultKind == Ambiguous) {
>          SavedAK = Ambiguity;
> -        WasAmbiguous = true;
>        }
>        ResultKind = Found;
>        resolveKind();
> @@ -436,8 +434,6 @@ public:
>        // If we didn't make the lookup unambiguous, restore the old
>        // ambiguity kind.
>        if (ResultKind == Ambiguous) {
> -        (void)WasAmbiguous;
> -        assert(WasAmbiguous);
>          Ambiguity = SavedAK;
>        } else if (Paths) {
>          deletePaths(Paths);
>
> --
> Markus
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20141004/628ebefc/attachment.html>


More information about the cfe-commits mailing list