Implicit declaration of type_info

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Implicit declaration of type_info

Steven Watanabe
AMDG

MSVC implicitly declares type_info in the global scope.
and at least one header (eh.h) depends on this.  I'm not
having any luck figuring out how to mirror this behavior
in clang (obviously only in Microsoft compatibility mode).
I figured that the best place to put it would be with the
declarations of operator new and delete, but I could not
find where that is.  Help?

In Christ,
Steven Watanabe

_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev
Reply | Threaded
Open this post in threaded view
|

Re: Implicit declaration of type_info

Sebastian Redl
On 07/09/2010 08:41 PM, Steven Watanabe wrote:

> AMDG
>
> MSVC implicitly declares type_info in the global scope.
> and at least one header (eh.h) depends on this.  I'm not
> having any luck figuring out how to mirror this behavior
> in clang (obviously only in Microsoft compatibility mode).
> I figured that the best place to put it would be with the
> declarations of operator new and delete, but I could not
> find where that is.  Help?
>  
Sema::DeclareGlobalNewDelete, in SemaExprCXX.cpp, around line 1200.
Beware, though, this function is lazily called only when we have to look
up a global operator new/delete for the first time.

Sebastian
_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev