C++ PATCH for c++/50508 (ICE on constexpr &&)

Gabriel Dos Reis gdr@integrable-solutions.net
Sun Oct 2 16:10:00 GMT 2011


On Sun, Oct 2, 2011 at 7:37 AM, Jason Merrill <jason@redhat.com> wrote:
> On 10/01/2011 08:05 PM, Gabriel Dos Reis wrote:
>>
>> It is weird though that GCC does not maintain a properly typed
>> internal representation.
>
> Huh?  Different typedefs need to be compared with same_type_p rather than
> ==.  I don't see how that makes the representation not properly typed.
>

The comment wasn't about comparison of typedefs -- the patch did not compare
typedefs.

*Value* representations should not change just because a type name was
introduced
via a typedef.  In particular, in my opinion comparing for "true" or
"false" should just be
an equality test to boolean_true_node or boolean_false_nore, not a
comparison of their
 integer representation.

-- Gaby



More information about the Gcc-patches mailing list