[llvm-commits] CallbackVH implementation

[Dan Gohman]

This is really useful. In a local tree I've converted the  
ScalarEvolution
analysis to use CallbackVH and it works quite well.  It allows makes
some really tricky code that handled some involved
ReplaceAllUsesWith situations simply unnecessary :-).

Can CallbackVH's copy constructor be made protected instead of
public? That ought to limit the range of possible trouble at least.  The
destructor can probably also be protected.

Thanks,

Dan

On Apr 29, 2009, at 4:24 PM, Jeffrey Yasskin wrote:

> Let me know if you had different names or an entirely different
> interface in mind.
>
> I'm a little skeptical of the public copy constructor on CallbackVH,
> since it'll slice off any subclass-specific data, but that may not be
> a big deal in actual use.
>
> Jeffrey
> <CallbackVH.patch>_______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits