aaron.ballman added a comment.
In D99861#2697449 <https://reviews.llvm.org/D99861#2697449>, @Qix- wrote:
> I'm not sure exactly how to continue after the last few comments - what
> should the approach be for this patch? Or are these things we can shoot for
> in later patches?
I don't think they're things we should shoot for in a later patch; the token
replay approach doesn't seem like it would work for more complicated attribute
arguments. As a concrete example of what would be super difficult to support
would be:
struct S {
int member;
void func(int i, int j) [[plugin::attr(i + j + member)]];
};
because it would be very difficult to recognize that `member` is looked up in
the context of the declaration of `S::func()` after having left the parsing
context for the structure.
I think the one way to continue is along the lines of what's described in
https://bugs.llvm.org/show_bug.cgi?id=46446#c11 so that the plugin handles
parsing the arguments by invoking calls on the parser.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D99861/new/
https://reviews.llvm.org/D99861
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
