syntax: ABI-oblivious grammar #65884

Centril · 2019-10-28T03:31:28Z

This PR has the following effects:

extern $lit is now legal where $lit:literal and $lit is substituted for a string literal.
extern "abi_that_does_not_exist" is now syntactically legal whereas before, the set of ABI strings was hard-coded into the grammar of the language. With this PR, the set of ABIs are instead validated and translated during lowering. That seems more appropriate.
ast::FloatTy is now distinct from rustc_target::abi::FloatTy. The former is used substantially more and the translation between them is only necessary in a single place.
As a result of 2-3, libsyntax no longer depends on librustc_target, which should improve pipe-lining somewhat.

cc @rust-lang/lang -- the points 1-2 slightly change the definition of the language but in a way which seems consistent with our general principles (in particular wrt. the discussions of turning things into semantic errors). I expect this to be uncontroversial but it's worth letting y'all know. :)

r? @varkor

Centril · 2019-10-28T03:36:05Z

src/libsyntax/feature_gate/check.rs

-            Abi::C |
-            Abi::System => {}
+            abi => {
+                self.parse_sess.span_diagnostic.delay_span_bug(


(The purpose of doing this is so that if we add a new ABI but forget to feature gate it, an ICE will occur and it won't pass CI.)

petrochenkov · 2019-10-28T04:59:44Z

Assigning myself as well, I had some plans in this area.
cc #60493

src/libsyntax/parse/parser.rs

eddyb · 2019-10-28T06:15:05Z

I think @oli-obk moved FloatTy around, perhaps now we can have Primitive::{F32, F64} again instead of rustc_target::abi::FloatTy existing.

src/libsyntax/ast.rs

nikomatsakis · 2019-10-28T13:52:01Z

the points 1-2 slightly change the definition of the language but in a way which seems consistent with our general principles (in particular wrt. the discussions of turning things into semantic errors). I expect this to be uncontroversial but it's worth letting y'all know. :)

👍 from me

Centril · 2019-10-28T18:41:15Z

I think @oli-obk moved FloatTy around, perhaps now we can have Primitive::{F32, F64} again instead of rustc_target::abi::FloatTy existing.

@eddyb I don't have an opinion on this other than to say that it could be done in a follow-up. :)

eddyb · 2019-10-28T18:53:28Z

it could be done in a follow-up. :)

My argument would be that it's confusing to have two copies of the same type.
But I was also hoping for @oli-obk's opinion on this.

src/libsyntax/ast.rs

petrochenkov · 2019-10-28T19:47:08Z

src/libsyntax_pos/symbol.rs

@@ -561,6 +561,7 @@ symbols! {
        rust_2018_preview,
        rust_begin_unwind,
        rustc,
+        Rust,


src/libsyntax/parse/parser.rs

petrochenkov · 2019-10-29T20:07:36Z

r=me after addressing the remaining comments, unless varkor wants to review as well.

Centril · 2019-11-06T10:07:53Z

@bors r=petrochenkov

bors · 2019-11-06T10:07:54Z

📌 Commit 09979759ba46574d39d8ab0c4dc14566b918e949 has been approved by petrochenkov

We also sever syntax's dependency on rustc_target as a result. This should slightly improve pipe-lining. Moreover, some cleanup is done in related code.

Centril · 2019-11-07T04:27:22Z

@bors r=petrochenkov

bors · 2019-11-07T04:27:23Z

📌 Commit 55f76cd has been approved by petrochenkov

@varkor

…chenkov syntax: ABI-oblivious grammar This PR has the following effects: 1. `extern $lit` is now legal where `$lit:literal` and `$lit` is substituted for a string literal. 2. `extern "abi_that_does_not_exist"` is now *syntactically* legal whereas before, the set of ABI strings was hard-coded into the grammar of the language. With this PR, the set of ABIs are instead validated and translated during lowering. That seems more appropriate. 3. `ast::FloatTy` is now distinct from `rustc_target::abi::FloatTy`. The former is used substantially more and the translation between them is only necessary in a single place. 4. As a result of 2-3, libsyntax no longer depends on librustc_target, which should improve pipe-lining somewhat. cc @rust-lang/lang -- the points 1-2 slightly change the definition of the language but in a way which seems consistent with our general principles (in particular wrt. the discussions of turning things into semantic errors). I expect this to be uncontroversial but it's worth letting y'all know. :) r? @varkor

@ghost

Rollup of 5 pull requests Successful merges: - #59789 (Revert two unapproved changes to rustc_typeck.) - #65752 (Use structured suggestions for missing associated items) - #65884 (syntax: ABI-oblivious grammar) - #65974 (A scheme for more macro-matcher friendly pre-expansion gating) - #66017 (Add future incompatibility lint for `array.into_iter()`) Failed merges: - #66056 (rustc_metadata: Some reorganization of the module structure) r? @ghost

@oli-obk

rustc_target: inline abi::FloatTy into abi::Primitive. This effectively undoes a small part of @oli-obk's rust-lang#50967, now that the rest of the compiler doesn't use the `FloatTy` definition from `rustc_target`, post-rust-lang#65884.

Centril added T-lang Relevant to the language team, which will review and decide on the PR/issue. relnotes Marks issues that should be documented in the release notes of the next release. labels Oct 28, 2019

Centril added this to the 1.40 milestone Oct 28, 2019

rust-highfive assigned varkor Oct 28, 2019