[cfe-commits] [PATCH] [VCPP] Add __ptr64 qualifier

[Aaron Ballman]

That's my understanding as well -- they're pretty goofy to have as
type attributes, truth be told.  But we should still be consistent
with the way they're mangled (eg, if you use __ptr64 in a 32-bit
build, does it get mangled as such even though you have a 32-bit
pointer).

~Aaron

On Mon, Nov 5, 2012 at 7:06 PM, Charles Davis <cdavis5x at gmail.com> wrote:
>
>   He means you should add a Clang test case (like you did in {D99}) showing that names get mangled properly in 64-bit mode. (Speaking of which, I have something to say about that over there...)
>
>   @AaronBallman The mangling for `__ptr32` is... nothing. This mangling scheme predates support for 64-bit pointers (and `__ptr64`). In fact, `__ptr64` was something that was tacked on later--hence why 64-bit pointer types have a special flag '`E`' set in their mangled names. This was carried over into the Win64 environment. Why? I don't know; probably because it makes the 64-bit mangler simpler (i.e. it's essentially identical to the 32-bit one).
>
> http://llvm-reviews.chandlerc.com/D101
> _______________________________________________
> cfe-commits mailing list
> cfe-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits