Move the AST from @T to Gc<T>, improve Gc<T> #14250
Conversation
|
cc #14193 |
|
Note that this also purges all remnants of |
Wooo! |
|
@alexcrichton Some tests failed on Travis. |
|
Oops, forgot to push some changes, tests should be fixed now |
|
Reviving #13316 might be nice, although it's likely rebase-hell, given how large it is. |
|
Uh oh, this happened without mentioning me. It's sad how #13316 went stale without getting into any real decision making process. |
|
Travis is still sad |
|
@eddyb, this didn't take too much time, but I meant for this to be as small of a migration as possible. I didn't want to change the actual representation of the AST, I just wanted to remove all I would expect this to make further migrations a little easier because all the autoderef should have been taken care of (in order for this to bootstrap). Sadly, though, it still relies on the implicit copyability of |
|
Thanks @sfackler, should be addressed now |
| @@ -73,8 +54,54 @@ impl<T> Clone for Gc<T> { | |||
| #[cfg(not(test))] | |||
| pub static GC: () = (); | |||
|
|
|||
| #[cfg(test)] | |||
| pub static GC: () = (); | |||
| impl<T: Eq + 'static> Eq for Gc<T> { | |||
There was a problem hiding this comment.
These are buggy because they don't handle cycles, and Gc<T> permits those in valid code.
There was a problem hiding this comment.
I'm not sure I follow. This is not a localized problem for Gc<T>, it also affects Rc<T>. Are you proposing we remove the trait implementations for both pointers, or just Gc<T>?
There was a problem hiding this comment.
With Rc<T>, it's a bug to create a strong reference cycle as it will leak. It's permitted in correct code for Gc<T> though so it seems bad if ==, etc. can stack overflow. I guess it could shove some metadata into a map in TLS to cope with it (not quite sure).
|
I'm okay with this as quick hack to remove |
|
This PR is meant as a step to removing the |
This commit uses the same trick as ~/Box to map Gc<T> to @t internally inside the compiler. This moves a number of implementations of traits to the `gc` module in the standard library. This removes functions such as `Gc::new`, `Gc::borrow`, and `Gc::ptr_eq` in favor of the more modern equivalents, `box(GC)`, `Deref`, and pointer equality. The Gc pointer itself should be much more useful now, and subsequent commits will move the compiler away from @t towards Gc<T> [breaking-change]
This commit removes `@T` from the compiler by moving the AST to using `Gc<T>`. This also starts treating `Gc<T>` as `@T` in the same way that `Box<T>` is the same as `~T` in the compiler. After this hits a snapshot, the `@T` syntax should be able to be removed completely.
This commit removes
@Tfrom the compiler by moving the AST to usingGc<T>. This also starts treatingGc<T>as@Tin the same way thatBox<T>is the same as~Tin the compiler.After this hits a snapshot, the
@Tsyntax should be able to be removed completely.