default is syntactically allowed before items in trait definitions.
Items in impl s (i.e. const s, type s, and fn s) may syntactically leave out their bodies in favor of ; .
Bounds on associated types in impl s are now syntactically allowed (e.g. type Foo: Ord; ).
... (the C-variadic type) may occur syntactically directly as the type of any function parameter.
This means that this parses now, but not much else. I think it means attribute-like macros can consume function definitions with ..., as long as the output is transformed so that it's removed. It's a stabilization in the syntax, but not in the functionality.
Like @daboross said, it's only implemented on the syntactic level. Trying to use it for anything more than that (eg implementing a function) therefore has no chance of working, even in nightly rust.
@sarvi In your particular example, I believe the problem is probably that you've written it as a method, not a function? Like, it seems to only be supported for freestanding functions, not for methods declared in impl blocks.
If I change your function definition to a freestanding function, then it compiles:
Thx daboross.
It does work as a function an not as a method.
Should I report it as issue in the issue?
Because non-c_varadic ones work both as standalone functions and as method.
The crate redhook for example uses that method behaviour to provide a very nice library to build ld_preload programs very conveniently.
And these use this approach and it looks very simple and clean.