Rollup merge of #138682 - Alexendoo:extra-symbols, r=fee1-dead

Allow drivers to supply a list of extra symbols to intern

Allows adding new symbols as `const`s in external drivers, desirable in Clippy so we can use them in patterns to replace code like 75530e9f72/src/tools/clippy/clippy_lints/src/casts/cast_ptr_alignment.rs (L66)

The Clippy change adds a couple symbols as a demo, the exact `clippy_utils` API and replacing other usages can be done on the Clippy side to minimise sync conflicts

---

try-job: aarch64-gnu

compiler/rustc_driver_impl/src/lib.rs

@@ -264,6 +264,7 @@ pub fn run_compiler(at_args: &[String], callbacks: &mut (dyn Callbacks + Send))
         hash_untracked_state: None,
         register_lints: None,
         override_queries: None,
+        extra_symbols: Vec::new(),
         make_codegen_backend: None,
         registry: diagnostics_registry(),
         using_internal_features: &USING_INTERNAL_FEATURES,

compiler/rustc_hir/src/tests.rs

@@ -17,7 +17,7 @@ fn def_path_hash_depends_on_crate_id() {
     // the crate by changing the crate disambiguator (e.g. via bumping the
     // crate's version number).
 
-    create_session_globals_then(Edition::Edition2024, None, || {
+    create_session_globals_then(Edition::Edition2024, &[], None, || {
         let id0 = StableCrateId::new(Symbol::intern("foo"), false, vec!["1".to_string()], "");
         let id1 = StableCrateId::new(Symbol::intern("foo"), false, vec!["2".to_string()], "");
 

compiler/rustc_interface/src/interface.rs

@@ -348,6 +348,10 @@ pub struct Config {
     /// the list of queries.
     pub override_queries: Option<fn(&Session, &mut Providers)>,
 
+    /// An extra set of symbols to add to the symbol interner, the symbol indices
+    /// will start at [`PREDEFINED_SYMBOLS_COUNT`](rustc_span::symbol::PREDEFINED_SYMBOLS_COUNT)
+    pub extra_symbols: Vec<&'static str>,
+
     /// This is a callback from the driver that is called to create a codegen backend.
     ///
     /// Has no uses within this repository, but is used by bjorn3 for "the
@@ -409,6 +413,7 @@ pub fn run_compiler<R: Send>(config: Config, f: impl FnOnce(&Compiler) -> R + Se
         &early_dcx,
         config.opts.edition,
         config.opts.unstable_opts.threads,
+        &config.extra_symbols,
         SourceMapInputs { file_loader, path_mapping, hash_kind, checksum_hash_kind },
         |current_gcx| {
             // The previous `early_dcx` can't be reused here because it doesn't

compiler/rustc_interface/src/tests.rs

@@ -53,7 +53,7 @@ where
         checksum_hash_kind,
     });
 
-    rustc_span::create_session_globals_then(DEFAULT_EDITION, sm_inputs, || {
+    rustc_span::create_session_globals_then(DEFAULT_EDITION, &[], sm_inputs, || {
         let temps_dir = sessopts.unstable_opts.temps_dir.as_deref().map(PathBuf::from);
         let io = CompilerIO {
             input: Input::Str { name: FileName::Custom(String::new()), input: String::new() },

compiler/rustc_interface/src/util.rs

@@ -117,6 +117,7 @@ fn run_in_thread_with_globals<F: FnOnce(CurrentGcx) -> R + Send, R: Send>(
     thread_stack_size: usize,
     edition: Edition,
     sm_inputs: SourceMapInputs,
+    extra_symbols: &[&'static str],
     f: F,
 ) -> R {
     // The "thread pool" is a single spawned thread in the non-parallel
@@ -134,9 +135,12 @@ fn run_in_thread_with_globals<F: FnOnce(CurrentGcx) -> R + Send, R: Send>(
         // name contains null bytes.
         let r = builder
             .spawn_scoped(s, move || {
-                rustc_span::create_session_globals_then(edition, Some(sm_inputs), || {
-                    f(CurrentGcx::new())
-                })
+                rustc_span::create_session_globals_then(
+                    edition,
+                    extra_symbols,
+                    Some(sm_inputs),
+                    || f(CurrentGcx::new()),
+                )
             })
             .unwrap()
             .join();
@@ -152,6 +156,7 @@ pub(crate) fn run_in_thread_pool_with_globals<F: FnOnce(CurrentGcx) -> R + Send,
     thread_builder_diag: &EarlyDiagCtxt,
     edition: Edition,
     threads: usize,
+    extra_symbols: &[&'static str],
     sm_inputs: SourceMapInputs,
     f: F,
 ) -> R {
@@ -168,12 +173,18 @@ pub(crate) fn run_in_thread_pool_with_globals<F: FnOnce(CurrentGcx) -> R + Send,
     let registry = sync::Registry::new(std::num::NonZero::new(threads).unwrap());
 
     if !sync::is_dyn_thread_safe() {
-        return run_in_thread_with_globals(thread_stack_size, edition, sm_inputs, |current_gcx| {
-            // Register the thread for use with the `WorkerLocal` type.
-            registry.register();
+        return run_in_thread_with_globals(
+            thread_stack_size,
+            edition,
+            sm_inputs,
+            extra_symbols,
+            |current_gcx| {
+                // Register the thread for use with the `WorkerLocal` type.
+                registry.register();
 
-            f(current_gcx)
-        });
+                f(current_gcx)
+            },
+        );
     }
 
     let current_gcx = FromDyn::from(CurrentGcx::new());
@@ -230,7 +241,7 @@ pub(crate) fn run_in_thread_pool_with_globals<F: FnOnce(CurrentGcx) -> R + Send,
     // pool. Upon creation, each worker thread created gets a copy of the
     // session globals in TLS. This is possible because `SessionGlobals` impls
     // `Send` in the parallel compiler.
-    rustc_span::create_session_globals_then(edition, Some(sm_inputs), || {
+    rustc_span::create_session_globals_then(edition, extra_symbols, Some(sm_inputs), || {
         rustc_span::with_session_globals(|session_globals| {
             let session_globals = FromDyn::from(session_globals);
             builder

compiler/rustc_macros/src/symbols.rs

@@ -142,13 +142,13 @@ pub(super) fn symbols(input: TokenStream) -> TokenStream {
     output
 }
 
-struct Preinterned {
+struct Predefined {
     idx: u32,
     span_of_name: Span,
 }
 
 struct Entries {
-    map: HashMap<String, Preinterned>,
+    map: HashMap<String, Predefined>,
 }
 
 impl Entries {
@@ -163,7 +163,7 @@ impl Entries {
             prev.idx
         } else {
             let idx = self.len();
-            self.map.insert(s.to_string(), Preinterned { idx, span_of_name: span });
+            self.map.insert(s.to_string(), Predefined { idx, span_of_name: span });
             idx
         }
     }
@@ -295,10 +295,14 @@ fn symbols_with_errors(input: TokenStream) -> (TokenStream, Vec<syn::Error>) {
     }
 
     let symbol_digits_base = entries.map["0"].idx;
-    let preinterned_symbols_count = entries.len();
+    let predefined_symbols_count = entries.len();
     let output = quote! {
         const SYMBOL_DIGITS_BASE: u32 = #symbol_digits_base;
-        const PREINTERNED_SYMBOLS_COUNT: u32 = #preinterned_symbols_count;
+
+        /// The number of predefined symbols; this is the the first index for
+        /// extra pre-interned symbols in an Interner created via
+        /// [`Interner::with_extra_symbols`].
+        pub const PREDEFINED_SYMBOLS_COUNT: u32 = #predefined_symbols_count;
 
         #[doc(hidden)]
         #[allow(non_upper_case_globals)]
@@ -315,10 +319,13 @@ fn symbols_with_errors(input: TokenStream) -> (TokenStream, Vec<syn::Error>) {
         }
 
         impl Interner {
-            pub(crate) fn fresh() -> Self {
-                Interner::prefill(&[
-                    #prefill_stream
-                ])
+            /// Creates an `Interner` with the predefined symbols from the `symbols!` macro and
+            /// any extra symbols provided by external drivers such as Clippy
+            pub(crate) fn with_extra_symbols(extra_symbols: &[&'static str]) -> Self {
+                Interner::prefill(
+                    &[#prefill_stream],
+                    extra_symbols,
+                )
             }
         }
     };

compiler/rustc_metadata/src/rmeta/decoder.rs

@@ -562,9 +562,9 @@ impl<'a, 'tcx> SpanDecoder for DecodeContext<'a, 'tcx> {
                     Symbol::intern(s)
                 })
             }
-            SYMBOL_PREINTERNED => {
+            SYMBOL_PREDEFINED => {
                 let symbol_index = self.read_u32();
-                Symbol::new_from_decoded(symbol_index)
+                Symbol::new(symbol_index)
             }
             _ => unreachable!(),
         }

compiler/rustc_metadata/src/rmeta/encoder.rs

@@ -201,9 +201,9 @@ impl<'a, 'tcx> SpanEncoder for EncodeContext<'a, 'tcx> {
     }
 
     fn encode_symbol(&mut self, symbol: Symbol) {
-        // if symbol preinterned, emit tag and symbol index
-        if symbol.is_preinterned() {
-            self.opaque.emit_u8(SYMBOL_PREINTERNED);
+        // if symbol predefined, emit tag and symbol index
+        if symbol.is_predefined() {
+            self.opaque.emit_u8(SYMBOL_PREDEFINED);
             self.opaque.emit_u32(symbol.as_u32());
         } else {
             // otherwise write it as string or as offset to it

compiler/rustc_metadata/src/rmeta/mod.rs

@@ -586,7 +586,7 @@ impl SpanTag {
 // Tags for encoding Symbol's
 const SYMBOL_STR: u8 = 0;
 const SYMBOL_OFFSET: u8 = 1;
-const SYMBOL_PREINTERNED: u8 = 2;
+const SYMBOL_PREDEFINED: u8 = 2;
 
 pub fn provide(providers: &mut Providers) {
     encoder::provide(providers);

compiler/rustc_middle/src/query/on_disk_cache.rs

@@ -46,7 +46,7 @@ const TAG_EXPN_DATA: u8 = 1;
 // Tags for encoding Symbol's
 const SYMBOL_STR: u8 = 0;
 const SYMBOL_OFFSET: u8 = 1;
-const SYMBOL_PREINTERNED: u8 = 2;
+const SYMBOL_PREDEFINED: u8 = 2;
 
 /// Provides an interface to incremental compilation data cached from the
 /// previous compilation session. This data will eventually include the results
@@ -674,9 +674,9 @@ impl<'a, 'tcx> SpanDecoder for CacheDecoder<'a, 'tcx> {
                     Symbol::intern(s)
                 })
             }
-            SYMBOL_PREINTERNED => {
+            SYMBOL_PREDEFINED => {
                 let symbol_index = self.read_u32();
-                Symbol::new_from_decoded(symbol_index)
+                Symbol::new(symbol_index)
             }
             _ => unreachable!(),
         }
@@ -892,9 +892,9 @@ impl<'a, 'tcx> SpanEncoder for CacheEncoder<'a, 'tcx> {
 
     // copy&paste impl from rustc_metadata
     fn encode_symbol(&mut self, symbol: Symbol) {
-        // if symbol preinterned, emit tag and symbol index
-        if symbol.is_preinterned() {
-            self.encoder.emit_u8(SYMBOL_PREINTERNED);
+        // if symbol predefined, emit tag and symbol index
+        if symbol.is_predefined() {
+            self.encoder.emit_u8(SYMBOL_PREDEFINED);
             self.encoder.emit_u32(symbol.as_u32());
         } else {
             // otherwise write it as string or as offset to it

compiler/rustc_span/src/lib.rs

@@ -116,9 +116,13 @@ pub struct SessionGlobals {
 }
 
 impl SessionGlobals {
-    pub fn new(edition: Edition, sm_inputs: Option<SourceMapInputs>) -> SessionGlobals {
+    pub fn new(
+        edition: Edition,
+        extra_symbols: &[&'static str],
+        sm_inputs: Option<SourceMapInputs>,
+    ) -> SessionGlobals {
         SessionGlobals {
-            symbol_interner: symbol::Interner::fresh(),
+            symbol_interner: symbol::Interner::with_extra_symbols(extra_symbols),
             span_interner: Lock::new(span_encoding::SpanInterner::default()),
             metavar_spans: Default::default(),
             hygiene_data: Lock::new(hygiene::HygieneData::new(edition)),
@@ -129,6 +133,7 @@ impl SessionGlobals {
 
 pub fn create_session_globals_then<R>(
     edition: Edition,
+    extra_symbols: &[&'static str],
     sm_inputs: Option<SourceMapInputs>,
     f: impl FnOnce() -> R,
 ) -> R {
@@ -137,7 +142,7 @@ pub fn create_session_globals_then<R>(
         "SESSION_GLOBALS should never be overwritten! \
          Use another thread if you need another SessionGlobals"
     );
-    let session_globals = SessionGlobals::new(edition, sm_inputs);
+    let session_globals = SessionGlobals::new(edition, extra_symbols, sm_inputs);
     SESSION_GLOBALS.set(&session_globals, f)
 }
 
@@ -156,7 +161,7 @@ where
     F: FnOnce(&SessionGlobals) -> R,
 {
     if !SESSION_GLOBALS.is_set() {
-        let session_globals = SessionGlobals::new(edition, None);
+        let session_globals = SessionGlobals::new(edition, &[], None);
         SESSION_GLOBALS.set(&session_globals, || SESSION_GLOBALS.with(f))
     } else {
         SESSION_GLOBALS.with(f)
@@ -172,7 +177,7 @@ where
 
 /// Default edition, no source map.
 pub fn create_default_session_globals_then<R>(f: impl FnOnce() -> R) -> R {
-    create_session_globals_then(edition::DEFAULT_EDITION, None, f)
+    create_session_globals_then(edition::DEFAULT_EDITION, &[], None, f)
 }
 
 // If this ever becomes non thread-local, `decode_syntax_context`

compiler/rustc_span/src/symbol.rs

@@ -2540,15 +2540,10 @@ rustc_index::newtype_index! {
 }
 
 impl Symbol {
-    const fn new(n: u32) -> Self {
+    pub const fn new(n: u32) -> Self {
         Symbol(SymbolIndex::from_u32(n))
     }
 
-    /// for use in Decoder only
-    pub fn new_from_decoded(n: u32) -> Self {
-        Self::new(n)
-    }
-
     /// Maps a string to its interned representation.
     #[rustc_diagnostic_item = "SymbolIntern"]
     pub fn intern(string: &str) -> Self {
@@ -2634,11 +2629,14 @@ struct InternerInner {
 }
 
 impl Interner {
-    fn prefill(init: &[&'static str]) -> Self {
-        Interner(Lock::new(InternerInner {
-            arena: Default::default(),
-            strings: init.iter().copied().collect(),
-        }))
+    fn prefill(init: &[&'static str], extra: &[&'static str]) -> Self {
+        let strings = FxIndexSet::from_iter(init.iter().copied().chain(extra.iter().copied()));
+        assert_eq!(
+            strings.len(),
+            init.len() + extra.len(),
+            "`init` or `extra` contain duplicate symbols",
+        );
+        Interner(Lock::new(InternerInner { arena: Default::default(), strings }))
     }
 
     #[inline]
@@ -2762,9 +2760,9 @@ impl Symbol {
         self != kw::Empty && self != kw::Underscore && !self.is_path_segment_keyword()
     }
 
-    /// Is this symbol was interned in compiler's `symbols!` macro
-    pub fn is_preinterned(self) -> bool {
-        self.as_u32() < PREINTERNED_SYMBOLS_COUNT
+    /// Was this symbol predefined in the compiler's `symbols!` macro
+    pub fn is_predefined(self) -> bool {
+        self.as_u32() < PREDEFINED_SYMBOLS_COUNT
     }
 }
 

compiler/rustc_span/src/symbol/tests.rs

@@ -3,7 +3,7 @@ use crate::create_default_session_globals_then;
 
 #[test]
 fn interner_tests() {
-    let i = Interner::prefill(&[]);
+    let i = Interner::prefill(&[], &[]);
     // first one is zero:
     assert_eq!(i.intern("dog"), Symbol::new(0));
     // re-use gets the same entry: