[cfe-commits] r72481 - in /cfe/trunk: lib/Parse/ParseDeclCXX.cpp lib/Sema/SemaLookup.cpp lib/Sema/SemaTemplateInstantiateDecl.cpp lib/Sema/SemaTemplateInstantiateExpr.cpp test/SemaTemplate/example-dynarray.cpp test/SemaTemplate/instantiate-declref.cpp

[Douglas Gregor]

On May 29, 2009, at 10:32 AM, Sebastian Redl wrote:

> Douglas Gregor wrote:
>>
>> On May 29, 2009, at 9:50 AM, Sebastian Redl wrote:
>>
>>> Sebastian Redl wrote:
>>>> Douglas Gregor wrote:
>>>>
>>>>> Did you see the FIXMEs? Something's not working with placement  
>>>>> new,
>>>>> but I haven't looked into it yet to see what's going on.
>>>>>
>>>> I'll take a look at it. Good break from going crazy over C++'s  
>>>> biggest
>>>> misfeature.
>>>>
>>> Got it. See:
>>> http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20090525/017732.html
>>>
>>
>> Ah, okay. Any chance that we could improve the error message, to ask
>> that the user include the <new> header when that is the error?
> Yes, by recognizing that the overload asked for is precisely one of
> those listed in 18.6p1 (i.e. one the <new> header provides) but not  
> one
> of those that are implicitly defined.

Ah, that would be wonderful. I do need to add support for "include  
this header" Fix-It hints, since there are a few other places we need  
them, too.

> I'll do that when I need the next break from exception specifications.


Thanks!

	- Doug