Speed up Parser::expected_token_types.

The parser pushes a `TokenType` to `Parser::expected_token_types` on
every call to the various `check`/`eat` methods, and clears it on every
call to `bump`. Some of those `TokenType` values are full tokens that
require cloning and dropping. This is a *lot* of work for something
that is only used in error messages and it accounts for a significant
fraction of parsing execution time.

This commit overhauls `TokenType` so that `Parser::expected_token_types`
can be implemented as a bitset. This requires changing `TokenType` to a
C-style parameterless enum, and adding `TokenTypeSet` which uses a
`u128` for the bits. (The new `TokenType` has 105 variants.)

The new types `ExpTokenPair` and `ExpKeywordPair` are now arguments to
the `check`/`eat` methods. This is for maximum speed. The elements in
the pairs are always statically known; e.g. a
`token::BinOp(token::Star)` is always paired with a `TokenType::Star`.
So we now compute `TokenType`s in advance and pass them in to
`check`/`eat` rather than the current approach of constructing them on
insertion into `expected_token_types`.

Values of these pair types can be produced by the new `exp!` macro,
which is used at every `check`/`eat` call site. The macro is for
convenience, allowing any pair to be generated from a single identifier.

The ident/keyword filtering in `expected_one_of_not_found` is no longer
necessary. It was there to account for some sloppiness in
`TokenKind`/`TokenType` comparisons.

The existing `TokenType` is moved to a new file `token_type.rs`, and all
its new infrastructure is added to that file. There is more boilerplate
code than I would like, but I can't see how to make it shorter.

Author: nnethercote

compiler/rustc_builtin_macros/src/asm.rs

@@ -1,16 +1,16 @@
 use ast::token::IdentIsRaw;
 use lint::BuiltinLintDiag;
-use rustc_ast::AsmMacro;
 use rustc_ast::ptr::P;
-use rustc_ast::token::{self, Delimiter};
 use rustc_ast::tokenstream::TokenStream;
+use rustc_ast::{AsmMacro, token};
 use rustc_data_structures::fx::{FxHashMap, FxIndexMap};
 use rustc_errors::PResult;
 use rustc_expand::base::*;
 use rustc_index::bit_set::GrowableBitSet;
-use rustc_parse::parser::Parser;
+use rustc_parse::exp;
+use rustc_parse::parser::{ExpKeywordPair, Parser};
 use rustc_session::lint;
-use rustc_span::symbol::{Ident, Symbol, kw, sym};
+use rustc_span::symbol::{Ident, Symbol, kw};
 use rustc_span::{ErrorGuaranteed, InnerSpan, Span};
 use rustc_target::asm::InlineAsmArch;
 use smallvec::smallvec;
@@ -39,16 +39,16 @@ pub struct AsmArgs {
 /// - `Err(_)` if the current token matches the keyword, but was not expected
 fn eat_operand_keyword<'a>(
     p: &mut Parser<'a>,
-    symbol: Symbol,
+    exp: ExpKeywordPair,
     asm_macro: AsmMacro,
 ) -> PResult<'a, bool> {
     if matches!(asm_macro, AsmMacro::Asm) {
-        Ok(p.eat_keyword(symbol))
+        Ok(p.eat_keyword(exp))
     } else {
         let span = p.token.span;
-        if p.eat_keyword_noexpect(symbol) {
+        if p.eat_keyword_noexpect(exp.kw) {
             // in gets printed as `r#in` otherwise
-            let symbol = if symbol == kw::In { "in" } else { symbol.as_str() };
+            let symbol = if exp.kw == kw::In { "in" } else { exp.kw.as_str() };
             Err(p.dcx().create_err(errors::AsmUnsupportedOperand {
                 span,
                 symbol,
@@ -96,28 +96,28 @@ pub fn parse_asm_args<'a>(
 
     let mut allow_templates = true;
     while p.token != token::Eof {
-        if !p.eat(&token::Comma) {
+        if !p.eat(exp!(Comma)) {
             if allow_templates {
                 // After a template string, we always expect *only* a comma...
                 return Err(dcx.create_err(errors::AsmExpectedComma { span: p.token.span }));
             } else {
                 // ...after that delegate to `expect` to also include the other expected tokens.
-                return Err(p.expect(&token::Comma).err().unwrap());
+                return Err(p.expect(exp!(Comma)).err().unwrap());
             }
         }
         if p.token == token::Eof {
             break;
         } // accept trailing commas
 
         // Parse clobber_abi
-        if p.eat_keyword(sym::clobber_abi) {
+        if p.eat_keyword(exp!(ClobberAbi)) {
             parse_clobber_abi(p, &mut args)?;
             allow_templates = false;
             continue;
         }
 
         // Parse options
-        if p.eat_keyword(sym::options) {
+        if p.eat_keyword(exp!(Options)) {
             parse_options(p, &mut args, asm_macro)?;
             allow_templates = false;
             continue;
@@ -129,65 +129,65 @@ pub fn parse_asm_args<'a>(
         let name = if p.token.is_ident() && p.look_ahead(1, |t| *t == token::Eq) {
             let (ident, _) = p.token.ident().unwrap();
             p.bump();
-            p.expect(&token::Eq)?;
+            p.expect(exp!(Eq))?;
             allow_templates = false;
             Some(ident.name)
         } else {
             None
         };
 
         let mut explicit_reg = false;
-        let op = if eat_operand_keyword(p, kw::In, asm_macro)? {
+        let op = if eat_operand_keyword(p, exp!(In), asm_macro)? {
             let reg = parse_reg(p, &mut explicit_reg)?;
-            if p.eat_keyword(kw::Underscore) {
+            if p.eat_keyword(exp!(Underscore)) {
                 let err = dcx.create_err(errors::AsmUnderscoreInput { span: p.token.span });
                 return Err(err);
             }
             let expr = p.parse_expr()?;
             ast::InlineAsmOperand::In { reg, expr }
-        } else if eat_operand_keyword(p, sym::out, asm_macro)? {
+        } else if eat_operand_keyword(p, exp!(Out), asm_macro)? {
             let reg = parse_reg(p, &mut explicit_reg)?;
-            let expr = if p.eat_keyword(kw::Underscore) { None } else { Some(p.parse_expr()?) };
+            let expr = if p.eat_keyword(exp!(Underscore)) { None } else { Some(p.parse_expr()?) };
             ast::InlineAsmOperand::Out { reg, expr, late: false }
-        } else if eat_operand_keyword(p, sym::lateout, asm_macro)? {
+        } else if eat_operand_keyword(p, exp!(Lateout), asm_macro)? {
             let reg = parse_reg(p, &mut explicit_reg)?;
-            let expr = if p.eat_keyword(kw::Underscore) { None } else { Some(p.parse_expr()?) };
+            let expr = if p.eat_keyword(exp!(Underscore)) { None } else { Some(p.parse_expr()?) };
             ast::InlineAsmOperand::Out { reg, expr, late: true }
-        } else if eat_operand_keyword(p, sym::inout, asm_macro)? {
+        } else if eat_operand_keyword(p, exp!(Inout), asm_macro)? {
             let reg = parse_reg(p, &mut explicit_reg)?;
-            if p.eat_keyword(kw::Underscore) {
+            if p.eat_keyword(exp!(Underscore)) {
                 let err = dcx.create_err(errors::AsmUnderscoreInput { span: p.token.span });
                 return Err(err);
             }
             let expr = p.parse_expr()?;
-            if p.eat(&token::FatArrow) {
+            if p.eat(exp!(FatArrow)) {
                 let out_expr =
-                    if p.eat_keyword(kw::Underscore) { None } else { Some(p.parse_expr()?) };
+                    if p.eat_keyword(exp!(Underscore)) { None } else { Some(p.parse_expr()?) };
                 ast::InlineAsmOperand::SplitInOut { reg, in_expr: expr, out_expr, late: false }
             } else {
                 ast::InlineAsmOperand::InOut { reg, expr, late: false }
             }
-        } else if eat_operand_keyword(p, sym::inlateout, asm_macro)? {
+        } else if eat_operand_keyword(p, exp!(Inlateout), asm_macro)? {
             let reg = parse_reg(p, &mut explicit_reg)?;
-            if p.eat_keyword(kw::Underscore) {
+            if p.eat_keyword(exp!(Underscore)) {
                 let err = dcx.create_err(errors::AsmUnderscoreInput { span: p.token.span });
                 return Err(err);
             }
             let expr = p.parse_expr()?;
-            if p.eat(&token::FatArrow) {
+            if p.eat(exp!(FatArrow)) {
                 let out_expr =
-                    if p.eat_keyword(kw::Underscore) { None } else { Some(p.parse_expr()?) };
+                    if p.eat_keyword(exp!(Underscore)) { None } else { Some(p.parse_expr()?) };
                 ast::InlineAsmOperand::SplitInOut { reg, in_expr: expr, out_expr, late: true }
             } else {
                 ast::InlineAsmOperand::InOut { reg, expr, late: true }
             }
-        } else if eat_operand_keyword(p, sym::label, asm_macro)? {
+        } else if eat_operand_keyword(p, exp!(Label), asm_macro)? {
             let block = p.parse_block()?;
             ast::InlineAsmOperand::Label { block }
-        } else if p.eat_keyword(kw::Const) {
+        } else if p.eat_keyword(exp!(Const)) {
             let anon_const = p.parse_expr_anon_const()?;
             ast::InlineAsmOperand::Const { anon_const }
-        } else if p.eat_keyword(sym::sym) {
+        } else if p.eat_keyword(exp!(Sym)) {
             let expr = p.parse_expr()?;
             let ast::ExprKind::Path(qself, path) = &expr.kind else {
                 let err = dcx.create_err(errors::AsmSymNoPath { span: expr.span });
@@ -390,31 +390,31 @@ fn parse_options<'a>(
 ) -> PResult<'a, ()> {
     let span_start = p.prev_token.span;
 
-    p.expect(&token::OpenDelim(Delimiter::Parenthesis))?;
-
-    while !p.eat(&token::CloseDelim(Delimiter::Parenthesis)) {
-        const OPTIONS: [(Symbol, ast::InlineAsmOptions); ast::InlineAsmOptions::COUNT] = [
-            (sym::pure, ast::InlineAsmOptions::PURE),
-            (sym::nomem, ast::InlineAsmOptions::NOMEM),
-            (sym::readonly, ast::InlineAsmOptions::READONLY),
-            (sym::preserves_flags, ast::InlineAsmOptions::PRESERVES_FLAGS),
-            (sym::noreturn, ast::InlineAsmOptions::NORETURN),
-            (sym::nostack, ast::InlineAsmOptions::NOSTACK),
-            (sym::may_unwind, ast::InlineAsmOptions::MAY_UNWIND),
-            (sym::att_syntax, ast::InlineAsmOptions::ATT_SYNTAX),
-            (kw::Raw, ast::InlineAsmOptions::RAW),
+    p.expect(exp!(OpenParen))?;
+
+    while !p.eat(exp!(CloseParen)) {
+        const OPTIONS: [(ExpKeywordPair, ast::InlineAsmOptions); ast::InlineAsmOptions::COUNT] = [
+            (exp!(Pure), ast::InlineAsmOptions::PURE),
+            (exp!(Nomem), ast::InlineAsmOptions::NOMEM),
+            (exp!(Readonly), ast::InlineAsmOptions::READONLY),
+            (exp!(PreservesFlags), ast::InlineAsmOptions::PRESERVES_FLAGS),
+            (exp!(Noreturn), ast::InlineAsmOptions::NORETURN),
+            (exp!(Nostack), ast::InlineAsmOptions::NOSTACK),
+            (exp!(MayUnwind), ast::InlineAsmOptions::MAY_UNWIND),
+            (exp!(AttSyntax), ast::InlineAsmOptions::ATT_SYNTAX),
+            (exp!(Raw), ast::InlineAsmOptions::RAW),
         ];
 
         'blk: {
-            for (symbol, option) in OPTIONS {
+            for (exp, option) in OPTIONS {
                 let kw_matched = if asm_macro.is_supported_option(option) {
-                    p.eat_keyword(symbol)
+                    p.eat_keyword(exp)
                 } else {
-                    p.eat_keyword_noexpect(symbol)
+                    p.eat_keyword_noexpect(exp.kw)
                 };
 
                 if kw_matched {
-                    try_set_option(p, args, asm_macro, symbol, option);
+                    try_set_option(p, args, asm_macro, exp.kw, option);
                     break 'blk;
                 }
             }
@@ -423,10 +423,10 @@ fn parse_options<'a>(
         }
 
         // Allow trailing commas
-        if p.eat(&token::CloseDelim(Delimiter::Parenthesis)) {
+        if p.eat(exp!(CloseParen)) {
             break;
         }
-        p.expect(&token::Comma)?;
+        p.expect(exp!(Comma))?;
     }
 
     let new_span = span_start.to(p.prev_token.span);
@@ -438,14 +438,14 @@ fn parse_options<'a>(
 fn parse_clobber_abi<'a>(p: &mut Parser<'a>, args: &mut AsmArgs) -> PResult<'a, ()> {
     let span_start = p.prev_token.span;
 
-    p.expect(&token::OpenDelim(Delimiter::Parenthesis))?;
+    p.expect(exp!(OpenParen))?;
 
-    if p.eat(&token::CloseDelim(Delimiter::Parenthesis)) {
+    if p.eat(exp!(CloseParen)) {
         return Err(p.dcx().create_err(errors::NonABI { span: p.token.span }));
     }
 
     let mut new_abis = Vec::new();
-    while !p.eat(&token::CloseDelim(Delimiter::Parenthesis)) {
+    while !p.eat(exp!(CloseParen)) {
         match p.parse_str_lit() {
             Ok(str_lit) => {
                 new_abis.push((str_lit.symbol_unescaped, str_lit.span));
@@ -457,10 +457,10 @@ fn parse_clobber_abi<'a>(p: &mut Parser<'a>, args: &mut AsmArgs) -> PResult<'a,
         };
 
         // Allow trailing commas
-        if p.eat(&token::CloseDelim(Delimiter::Parenthesis)) {
+        if p.eat(exp!(CloseParen)) {
             break;
         }
-        p.expect(&token::Comma)?;
+        p.expect(exp!(Comma))?;
     }
 
     let full_span = span_start.to(p.prev_token.span);
@@ -483,7 +483,7 @@ fn parse_reg<'a>(
     p: &mut Parser<'a>,
     explicit_reg: &mut bool,
 ) -> PResult<'a, ast::InlineAsmRegOrRegClass> {
-    p.expect(&token::OpenDelim(Delimiter::Parenthesis))?;
+    p.expect(exp!(OpenParen))?;
     let result = match p.token.uninterpolate().kind {
         token::Ident(name, IdentIsRaw::No) => ast::InlineAsmRegOrRegClass::RegClass(name),
         token::Literal(token::Lit { kind: token::LitKind::Str, symbol, suffix: _ }) => {
@@ -497,7 +497,7 @@ fn parse_reg<'a>(
         }
     };
     p.bump();
-    p.expect(&token::CloseDelim(Delimiter::Parenthesis))?;
+    p.expect(exp!(CloseParen))?;
     Ok(result)
 }
 

compiler/rustc_builtin_macros/src/assert.rs

@@ -7,6 +7,7 @@ use rustc_ast::{DelimArgs, Expr, ExprKind, MacCall, Path, PathSegment, UnOp, tok
 use rustc_ast_pretty::pprust;
 use rustc_errors::PResult;
 use rustc_expand::base::{DummyResult, ExpandResult, ExtCtxt, MacEager, MacroExpanderResult};
+use rustc_parse::exp;
 use rustc_parse::parser::Parser;
 use rustc_span::symbol::{Ident, Symbol, sym};
 use rustc_span::{DUMMY_SP, Span};
@@ -144,7 +145,7 @@ fn parse_assert<'a>(cx: &ExtCtxt<'a>, sp: Span, stream: TokenStream) -> PResult<
             cx.dcx().emit_err(errors::AssertMissingComma { span: parser.token.span, comma });
 
             parse_custom_message(&mut parser)
-        } else if parser.eat(&token::Comma) {
+        } else if parser.eat(exp!(Comma)) {
             parse_custom_message(&mut parser)
         } else {
             None

compiler/rustc_builtin_macros/src/cfg.rs

@@ -6,6 +6,7 @@ use rustc_ast::token;
 use rustc_ast::tokenstream::TokenStream;
 use rustc_errors::PResult;
 use rustc_expand::base::{DummyResult, ExpandResult, ExtCtxt, MacEager, MacroExpanderResult};
+use rustc_parse::exp;
 use rustc_span::Span;
 use {rustc_ast as ast, rustc_attr as attr};
 
@@ -48,9 +49,9 @@ fn parse_cfg<'a>(
 
     let cfg = p.parse_meta_item_inner()?;
 
-    let _ = p.eat(&token::Comma);
+    let _ = p.eat(exp!(Comma));
 
-    if !p.eat(&token::Eof) {
+    if !p.eat(exp!(Eof)) {
         return Err(cx.dcx().create_err(errors::OneCfgPattern { span }));
     }
 

compiler/rustc_builtin_macros/src/format.rs

@@ -12,6 +12,7 @@ use rustc_errors::{Applicability, Diag, MultiSpan, PResult, SingleLabelManySpans
 use rustc_expand::base::*;
 use rustc_lint_defs::builtin::NAMED_ARGUMENTS_USED_POSITIONALLY;
 use rustc_lint_defs::{BufferedEarlyLint, BuiltinLintDiag, LintId};
+use rustc_parse::exp;
 use rustc_parse_format as parse;
 use rustc_span::symbol::{Ident, Symbol};
 use rustc_span::{BytePos, ErrorGuaranteed, InnerSpan, Span};
@@ -94,12 +95,12 @@ fn parse_args<'a>(ecx: &ExtCtxt<'a>, sp: Span, tts: TokenStream) -> PResult<'a,
     let mut first = true;
 
     while p.token != token::Eof {
-        if !p.eat(&token::Comma) {
+        if !p.eat(exp!(Comma)) {
             if first {
                 p.clear_expected_token_types();
             }
 
-            match p.expect(&token::Comma) {
+            match p.expect(exp!(Comma)) {
                 Err(err) => {
                     match token::TokenKind::Comma.similar_tokens() {
                         Some(tks) if tks.contains(&p.token.kind) => {
@@ -123,7 +124,7 @@ fn parse_args<'a>(ecx: &ExtCtxt<'a>, sp: Span, tts: TokenStream) -> PResult<'a,
         match p.token.ident() {
             Some((ident, _)) if p.look_ahead(1, |t| *t == token::Eq) => {
                 p.bump();
-                p.expect(&token::Eq)?;
+                p.expect(exp!(Eq))?;
                 let expr = p.parse_expr()?;
                 if let Some((_, prev)) = args.by_name(ident.name) {
                     ecx.dcx().emit_err(errors::FormatDuplicateArg {

compiler/rustc_builtin_macros/src/pattern_type.rs

@@ -3,7 +3,8 @@ use rustc_ast::tokenstream::TokenStream;
 use rustc_ast::{Pat, Ty, ast};
 use rustc_errors::PResult;
 use rustc_expand::base::{self, DummyResult, ExpandResult, ExtCtxt, MacroExpanderResult};
-use rustc_span::{Span, sym};
+use rustc_parse::exp;
+use rustc_span::Span;
 
 pub(crate) fn expand<'cx>(
     cx: &'cx mut ExtCtxt<'_>,
@@ -24,7 +25,7 @@ fn parse_pat_ty<'a>(cx: &mut ExtCtxt<'a>, stream: TokenStream) -> PResult<'a, (P
     let mut parser = cx.new_parser_from_tts(stream);
 
     let ty = parser.parse_ty()?;
-    parser.expect_keyword(sym::is)?;
+    parser.expect_keyword(exp!(Is))?;
     let pat = parser.parse_pat_no_top_alt(None, None)?;
 
     Ok((ty, pat))

compiler/rustc_builtin_macros/src/util.rs

@@ -7,7 +7,7 @@ use rustc_expand::expand::AstFragment;
 use rustc_feature::AttributeTemplate;
 use rustc_lint_defs::BuiltinLintDiag;
 use rustc_lint_defs::builtin::DUPLICATE_MACRO_ATTRIBUTES;
-use rustc_parse::{parser, validate_attr};
+use rustc_parse::{exp, parser, validate_attr};
 use rustc_session::errors::report_lit_error;
 use rustc_span::{BytePos, Span, Symbol};
 
@@ -204,7 +204,7 @@ pub(crate) fn get_single_expr_from_tts(
         Ok(ret) => ret,
         Err(guar) => return ExpandResult::Ready(Err(guar)),
     };
-    let _ = p.eat(&token::Comma);
+    let _ = p.eat(exp!(Comma));
 
     if p.token != token::Eof {
         cx.dcx().emit_err(errors::OnlyOneArgument { span, name });
@@ -237,7 +237,7 @@ pub(crate) fn get_exprs_from_tts(
         let expr = cx.expander().fully_expand_fragment(AstFragment::Expr(expr)).make_expr();
 
         es.push(expr);
-        if p.eat(&token::Comma) {
+        if p.eat(exp!(Comma)) {
             continue;
         }
         if p.token != token::Eof {