[cxx-abi-dev] Function call mangling again

[David Vandevoorde]

On Mar 27, 2009, at 2:18 PM, Jason Merrill wrote:

> Doug Gregor wrote:
>> Mangling just the source name seems reasonable for a dependent  
>> name. Would we then mangle (f) as a parentheses expression whose  
>> subexpression is the source name?
>
> Well, we don't currently have a mangling for parentheses, so we  
> would need to invent something, either to represent the parens or a  
> modifier for the call to indicate that we don't do arg-dependent  
> lookup.


I think I prefer encoding the parentheses, since there may be other  
contexts where they matter (decltype?).

	Daveed