[cfe-commits] r68894 - in /cfe/trunk: lib/Parse/ParseObjc.cpp lib/Sema/SemaDeclObjC.cpp test/Analysis/region-1.m test/SemaObjC/method-conflict.m test/SemaObjC/protocol-forward-circular.m test/SemaObjC/protocol-test-2.m test/SemaObjC/protocol-unde

[Fariborz Jahanian]

On Apr 13, 2009, at 1:50 PM, Daniel Dunbar wrote:

> On Sun, Apr 12, 2009 at 1:43 AM, Chris Lattner <sabre at nondot.org>  
> wrote:
> Author: lattner
> Date: Sun Apr 12 03:43:13 2009
> New Revision: 68894
>
> URL: http://llvm.org/viewvc/llvm-project?rev=68894&view=rev
> Log:
> Fix rdar://6771034: don't warn on use of forward declared protocol  
> in protocol
> list of another protocol definition.  This warning is very noisy and  
> GCC doesn't
> produce it so existing code doesn't expect it.
>
> Ok, seems like a necessary concession. However, this construct is  
> also somewhat questionable and at some point we should move back  
> towards having the warning enabled. Can we retain a -W.... for this?

I think the best place for the warning is when protocol is actually  
used in a class (or in a qualified type maybe):

@protocol P1

@protocol P2<P1> @end

@interface I <P2> // warn here
@end

I am not sure though, how noisy this one will be.

- Fariborz

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20090413/3f05b158/attachment.html>