[PATCH] D139028: [RFC][clang] Add attribute-like keywords

[Richard Sandiford via Phabricator via cfe-commits]

rsandifo-arm added a comment.

Thanks @aaron.ballman and @erichkeane for the comments.  I'll split it into stages and use tablegen more, like you say.

In D139028#4150761 <https://reviews.llvm.org/D139028#4150761>, @aaron.ballman wrote:

> One thing that might also help is to split this into a few stages. 1) Add the new general functionality, 2) Replacing the existing implementation of keyword attributes with the new functionality where possible, 3) Add new attributes using the new functionality. It seems to me that we should be able to replace a bunch of our existing keyword attributes with the newer generalized approach (I'm thinking about ones like the calling convention attributes specifically, but we've got others as well), and that will help us prove the design as well as simplify the compiler implementation. WDYT?

I've tried to go through the existing keyword attributes, but unfortunately, I don't think any of them match the new scheme exactly:

__declspec
----------

I don't think anyone was suggesting we try to convert this :-) but for completeness:

  int __declspec(align(16)) a1; // OK, but standard attribute rules would say this appertains to the type
  int a2 __declspec();          // Not valid, unlike for standard decl attributes



__forceinline
-------------

  int __forceinline a1() { return 1; } // OK, but standard attribute rules would say this appertains to the type
  int a2 __forceinline() { return 1; } // Not valid, but would be for standard decl attributes



Calling convention keywords
---------------------------

These are generally DeclOrType attributes, with attributes sliding from the decl to the type where necessary.

  __stdcall int a1();   // OK, but appertains to the decl and relies on sliding behaviour
  int a2 __stdcall();   // Not valid, but is another place to put standard decl attributes
  int a3() __stdcall;   // Not valid, but is another place to put standard type attributes
  int (__stdcall a4)(); // OK, but standard attributes aren't allowed in this position
  extern int (*const __stdcall volatile a5) (); // OK, but standard attributes wouldn't be allowed here



Nullability keywords
--------------------

Like the calling-convention keywords, the nullability keywords appertain to types but are allowed in some (but not all) positions that would normally appertain to decls:

  using ptr = int *;
  _Nullable ptr a1; // OK, but appertains to the decl and relies on sliding behaviour
  ptr a2 _Nullable; // Not valid, but is another place to put standard decl attributes
  extern int *const _Nullable volatile a3; // OK, but a standard attribute wouldn't be allowed here

The same distinction applies to Microsoft pointer attributes.

Would it be OK just to use the new scheme for new attributes?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D139028/new/

https://reviews.llvm.org/D139028