Auto merge of #34652 - jseyfried:fix_expansion_perf, r=nrc

Fix expansion performance regression

**syntax-[breaking-change] cc #31645**

This fixes #34630 by reverting commit 5bf7970 of PR #33943, which landed in #34424.

By removing the `Rc<_>` wrapping around `Delimited` and `SequenceRepetition` in `TokenTree`, 5bf7970 made cloning `TokenTree`s more expensive. While this had no measurable performance impact on the compiler's crates, it caused an order of magnitude performance regression on some macro-heavy code in the wild. I believe this is due to clones of `TokenTree`s in `macro_parser.rs` and/or `macro_rules.rs`.

r? @nrc

diff --git a/src/libsyntax/ext/expand.rs b/src/libsyntax/ext/expand.rs
index 877293bac54f77547aa64bcee932e0893f5037bd..b2b63d0dbb4bd76b4b8fe8ac3121adf460438d67 100644 (file)
--- a/src/libsyntax/ext/expand.rs
+++ b/src/libsyntax/ext/expand.rs
@@ -237,7 +237,7 @@ fn mac_result<'a>(path: &ast::Path, ident: Option<Ident>, tts: Vec<TokenTree>, m
                     },
                 });
 
-                let marked_tts = mark_tts(tts, mark);
+                let marked_tts = mark_tts(&tts, mark);
                 Some(expandfun.expand(fld.cx, call_site, &marked_tts))
             }
 
@@ -257,7 +257,7 @@ fn mac_result<'a>(path: &ast::Path, ident: Option<Ident>, tts: Vec<TokenTree>, m
                     }
                 });
 
-                let marked_tts = mark_tts(tts, mark);
+                let marked_tts = mark_tts(&tts, mark);
                 Some(expander.expand(fld.cx, call_site, ident, marked_tts))
             }
 
@@ -1130,7 +1130,7 @@ fn fold_mac(&mut self, Spanned {node, span}: ast::Mac) -> ast::Mac {
         Spanned {
             node: Mac_ {
                 path: self.fold_path(node.path),
-                tts: self.fold_tts(node.tts),
+                tts: self.fold_tts(&node.tts),
             },
             span: self.new_span(span),
         }
@@ -1145,7 +1145,7 @@ fn new_span(&mut self, mut span: Span) -> Span {
 }
 
 // apply a given mark to the given token trees. Used prior to expansion of a macro.
-fn mark_tts(tts: Vec<TokenTree>, m: Mrk) -> Vec<TokenTree> {
+fn mark_tts(tts: &[TokenTree], m: Mrk) -> Vec<TokenTree> {
     noop_fold_tts(tts, &mut Marker{mark:m, expn_id: None})
 }
 

diff --git a/src/libsyntax/ext/quote.rs b/src/libsyntax/ext/quote.rs
index 68527b0797d5b6031da5ca6dcc4a2d9eb2c24341..ffc950d76dd27f43047bb06e54def08e5f265266 100644 (file)
--- a/src/libsyntax/ext/quote.rs
+++ b/src/libsyntax/ext/quote.rs
@@ -32,6 +32,7 @@ pub mod rt {
     use ext::base::ExtCtxt;
     use parse::{self, token, classify};
     use ptr::P;
+    use std::rc::Rc;
 
     use tokenstream::{self, TokenTree};
 
@@ -215,12 +216,12 @@ fn to_tokens(&self, cx: &ExtCtxt) -> Vec<TokenTree> {
             if self.node.style == ast::AttrStyle::Inner {
                 r.push(TokenTree::Token(self.span, token::Not));
             }
-            r.push(TokenTree::Delimited(self.span, tokenstream::Delimited {
+            r.push(TokenTree::Delimited(self.span, Rc::new(tokenstream::Delimited {
                 delim: token::Bracket,
                 open_span: self.span,
                 tts: self.node.value.to_tokens(cx),
                 close_span: self.span,
-            }));
+            })));
             r
         }
     }
@@ -235,12 +236,12 @@ fn to_tokens(&self, cx: &ExtCtxt) -> Vec<TokenTree> {
 
     impl ToTokens for () {
         fn to_tokens(&self, _cx: &ExtCtxt) -> Vec<TokenTree> {
-            vec![TokenTree::Delimited(DUMMY_SP, tokenstream::Delimited {
+            vec![TokenTree::Delimited(DUMMY_SP, Rc::new(tokenstream::Delimited {
                 delim: token::Paren,
                 open_span: DUMMY_SP,
                 tts: vec![],
                 close_span: DUMMY_SP,
-            })]
+            }))]
         }
     }
 
@@ -791,9 +792,14 @@ fn statements_mk_tt(cx: &ExtCtxt, tt: &TokenTree, matcher: bool) -> Vec<ast::Stm
                                 id_ext("tokenstream"),
                                 id_ext("SequenceRepetition")];
             let e_seq_struct = cx.expr_struct(sp, cx.path_global(sp, seq_path), fields);
+            let e_rc_new = cx.expr_call_global(sp, vec![id_ext("std"),
+                                                        id_ext("rc"),
+                                                        id_ext("Rc"),
+                                                        id_ext("new")],
+                                                   vec![e_seq_struct]);
             let e_tok = cx.expr_call(sp,
                                      mk_tt_path(cx, sp, "Sequence"),
-                                     vec!(e_sp, e_seq_struct));
+                                     vec!(e_sp, e_rc_new));
             let e_push =
                 cx.expr_method_call(sp,
                                     cx.expr_ident(sp, id_ext("tt")),

diff --git a/src/libsyntax/ext/tt/macro_rules.rs b/src/libsyntax/ext/tt/macro_rules.rs
index 23f0b1fff0ae72b79c28763d4bfab6befbfc8706..84572b84963f3812511a2efdb4c8622d875a4297 100644 (file)
--- a/src/libsyntax/ext/tt/macro_rules.rs
+++ b/src/libsyntax/ext/tt/macro_rules.rs
@@ -28,6 +28,7 @@
 use std::cell::RefCell;
 use std::collections::{HashMap};
 use std::collections::hash_map::{Entry};
+use std::rc::Rc;
 
 struct ParserAnyMacro<'a> {
     parser: RefCell<Parser<'a>>,
@@ -262,7 +263,7 @@ pub fn compile<'cx>(cx: &'cx mut ExtCtxt,
     let match_lhs_tok = MatchNt(lhs_nm, token::str_to_ident("tt"));
     let match_rhs_tok = MatchNt(rhs_nm, token::str_to_ident("tt"));
     let argument_gram = vec![
-        TokenTree::Sequence(DUMMY_SP, tokenstream::SequenceRepetition {
+        TokenTree::Sequence(DUMMY_SP, Rc::new(tokenstream::SequenceRepetition {
             tts: vec![
                 TokenTree::Token(DUMMY_SP, match_lhs_tok),
                 TokenTree::Token(DUMMY_SP, token::FatArrow),
@@ -271,14 +272,14 @@ pub fn compile<'cx>(cx: &'cx mut ExtCtxt,
             separator: Some(token::Semi),
             op: tokenstream::KleeneOp::OneOrMore,
             num_captures: 2,
-        }),
+        })),
         // to phase into semicolon-termination instead of semicolon-separation
-        TokenTree::Sequence(DUMMY_SP, tokenstream::SequenceRepetition {
+        TokenTree::Sequence(DUMMY_SP, Rc::new(tokenstream::SequenceRepetition {
             tts: vec![TokenTree::Token(DUMMY_SP, token::Semi)],
             separator: None,
             op: tokenstream::KleeneOp::ZeroOrMore,
             num_captures: 0
-        }),
+        })),
     ];
 
     // Parse the macro_rules! invocation (`none` is for no interpolations):

diff --git a/src/libsyntax/ext/tt/transcribe.rs b/src/libsyntax/ext/tt/transcribe.rs
index 40944a9a1c2d360bf13c78a9b1a45d742f2fb998..7c0d10669f30e108e6e173510ebbd2a9e8fc601c 100644 (file)
--- a/src/libsyntax/ext/tt/transcribe.rs
+++ b/src/libsyntax/ext/tt/transcribe.rs
@@ -79,11 +79,11 @@ pub fn new_tt_reader_with_doc_flag(sp_diag: &Handler,
     let mut r = TtReader {
         sp_diag: sp_diag,
         stack: vec!(TtFrame {
-            forest: TokenTree::Sequence(DUMMY_SP, tokenstream::SequenceRepetition {
+            forest: TokenTree::Sequence(DUMMY_SP, Rc::new(tokenstream::SequenceRepetition {
                 tts: src,
                 // doesn't matter. This merely holds the root unzipping.
                 separator: None, op: tokenstream::KleeneOp::ZeroOrMore, num_captures: 0
-            }),
+            })),
             idx: 0,
             dotdotdoted: false,
             sep: None,

diff --git a/src/libsyntax/fold.rs b/src/libsyntax/fold.rs
index ed6f09eed645f7741951542aaeea87665667266c..ac3d643b185cac725fd7139d14ae4541970a3be6 100644 (file)
--- a/src/libsyntax/fold.rs
+++ b/src/libsyntax/fold.rs
@@ -28,6 +28,8 @@
 use util::small_vector::SmallVector;
 use util::move_map::MoveMap;
 
+use std::rc::Rc;
+
 pub trait Folder : Sized {
     // Any additions to this trait should happen in form
     // of a call to a public `noop_*` function that only calls
@@ -222,11 +224,11 @@ fn fold_ty_params(&mut self, tps: P<[TyParam]>) -> P<[TyParam]> {
         noop_fold_ty_params(tps, self)
     }
 
-    fn fold_tt(&mut self, tt: TokenTree) -> TokenTree {
+    fn fold_tt(&mut self, tt: &TokenTree) -> TokenTree {
         noop_fold_tt(tt, self)
     }
 
-    fn fold_tts(&mut self, tts: Vec<TokenTree>) -> Vec<TokenTree> {
+    fn fold_tts(&mut self, tts: &[TokenTree]) -> Vec<TokenTree> {
         noop_fold_tts(tts, self)
     }
 
@@ -501,7 +503,7 @@ pub fn noop_fold_mac<T: Folder>(Spanned {node, span}: Mac, fld: &mut T) -> Mac {
     Spanned {
         node: Mac_ {
             path: fld.fold_path(node.path),
-            tts: fld.fold_tts(node.tts),
+            tts: fld.fold_tts(&node.tts),
         },
         span: fld.new_span(span)
     }
@@ -528,26 +530,32 @@ pub fn noop_fold_arg<T: Folder>(Arg {id, pat, ty}: Arg, fld: &mut T) -> Arg {
     }
 }
 
-pub fn noop_fold_tt<T: Folder>(tt: TokenTree, fld: &mut T) -> TokenTree {
-    match tt {
+pub fn noop_fold_tt<T: Folder>(tt: &TokenTree, fld: &mut T) -> TokenTree {
+    match *tt {
         TokenTree::Token(span, ref tok) =>
             TokenTree::Token(span, fld.fold_token(tok.clone())),
-        TokenTree::Delimited(span, delimed) => TokenTree::Delimited(span, Delimited {
-            delim: delimed.delim,
-            open_span: delimed.open_span,
-            tts: fld.fold_tts(delimed.tts),
-            close_span: delimed.close_span,
-        }),
-        TokenTree::Sequence(span, seq) => TokenTree::Sequence(span, SequenceRepetition {
-            tts: fld.fold_tts(seq.tts),
-            separator: seq.separator.clone().map(|tok| fld.fold_token(tok)),
-            ..seq
-        }),
+        TokenTree::Delimited(span, ref delimed) => {
+            TokenTree::Delimited(span, Rc::new(
+                            Delimited {
+                                delim: delimed.delim,
+                                open_span: delimed.open_span,
+                                tts: fld.fold_tts(&delimed.tts),
+                                close_span: delimed.close_span,
+                            }
+                        ))
+        },
+        TokenTree::Sequence(span, ref seq) =>
+            TokenTree::Sequence(span,
+                       Rc::new(SequenceRepetition {
+                           tts: fld.fold_tts(&seq.tts),
+                           separator: seq.separator.clone().map(|tok| fld.fold_token(tok)),
+                           ..**seq
+                       })),
     }
 }
 
-pub fn noop_fold_tts<T: Folder>(tts: Vec<TokenTree>, fld: &mut T) -> Vec<TokenTree> {
-    tts.move_map(|tt| fld.fold_tt(tt))
+pub fn noop_fold_tts<T: Folder>(tts: &[TokenTree], fld: &mut T) -> Vec<TokenTree> {
+    tts.iter().map(|tt| fld.fold_tt(tt)).collect()
 }
 
 // apply ident folder if it's an ident, apply other folds to interpolated nodes
@@ -605,7 +613,7 @@ pub fn noop_fold_interpolated<T: Folder>(nt: token::Nonterminal, fld: &mut T)
             token::NtIdent(Box::new(Spanned::<Ident>{node: fld.fold_ident(id.node), ..*id})),
         token::NtMeta(meta_item) => token::NtMeta(fld.fold_meta_item(meta_item)),
         token::NtPath(path) => token::NtPath(Box::new(fld.fold_path(*path))),
-        token::NtTT(tt) => token::NtTT(tt.map(|tt| fld.fold_tt(tt))),
+        token::NtTT(tt) => token::NtTT(P(fld.fold_tt(&tt))),
         token::NtArm(arm) => token::NtArm(fld.fold_arm(arm)),
         token::NtImplItem(arm) =>
             token::NtImplItem(arm.map(|arm| fld.fold_impl_item(arm)

diff --git a/src/libsyntax/parse/mod.rs b/src/libsyntax/parse/mod.rs
index bbcc044d43c6b74b8701924847df75867f0a8ec6..9502bc48a3e110f84e0ddf107a3a3181ec92c06d 100644 (file)
--- a/src/libsyntax/parse/mod.rs
+++ b/src/libsyntax/parse/mod.rs
@@ -662,6 +662,7 @@ pub fn integer_lit(s: &str,
 #[cfg(test)]
 mod tests {
     use super::*;
+    use std::rc::Rc;
     use syntax_pos::{Span, BytePos, Pos, NO_EXPANSION};
     use codemap::Spanned;
     use ast::{self, PatKind};
@@ -763,7 +764,7 @@ fn string_to_tts_macro () {
                             )
                             if first_delimed.delim == token::Paren
                             && ident.name.as_str() == "a" => {},
-                            _ => panic!("value 3: {:?}", *first_delimed),
+                            _ => panic!("value 3: {:?}", **first_delimed),
                         }
                         let tts = &second_delimed.tts[..];
                         match (tts.len(), tts.get(0), tts.get(1)) {
@@ -774,10 +775,10 @@ fn string_to_tts_macro () {
                             )
                             if second_delimed.delim == token::Paren
                             && ident.name.as_str() == "a" => {},
-                            _ => panic!("value 4: {:?}", *second_delimed),
+                            _ => panic!("value 4: {:?}", **second_delimed),
                         }
                     },
-                    _ => panic!("value 2: {:?}", *macro_delimed),
+                    _ => panic!("value 2: {:?}", **macro_delimed),
                 }
             },
             _ => panic!("value: {:?}",tts),
@@ -793,7 +794,7 @@ fn string_to_tts_1() {
             TokenTree::Token(sp(3, 4), token::Ident(str_to_ident("a"))),
             TokenTree::Delimited(
                 sp(5, 14),
-                tokenstream::Delimited {
+                Rc::new(tokenstream::Delimited {
                     delim: token::DelimToken::Paren,
                     open_span: sp(5, 6),
                     tts: vec![
@@ -802,10 +803,10 @@ fn string_to_tts_1() {
                         TokenTree::Token(sp(10, 13), token::Ident(str_to_ident("i32"))),
                     ],
                     close_span: sp(13, 14),
-                }),
+                })),
             TokenTree::Delimited(
                 sp(15, 21),
-                tokenstream::Delimited {
+                Rc::new(tokenstream::Delimited {
                     delim: token::DelimToken::Brace,
                     open_span: sp(15, 16),
                     tts: vec![
@@ -813,7 +814,7 @@ fn string_to_tts_1() {
                         TokenTree::Token(sp(18, 19), token::Semi),
                     ],
                     close_span: sp(20, 21),
-                })
+                }))
         ];
 
         assert_eq!(tts, expected);

diff --git a/src/libsyntax/parse/parser.rs b/src/libsyntax/parse/parser.rs
index a06270bb7727a4ac3c5927c0afdbbe25f55750e8..4cf14e62299547153364166c4a0542b35f300a17 100644 (file)
--- a/src/libsyntax/parse/parser.rs
+++ b/src/libsyntax/parse/parser.rs
@@ -2688,12 +2688,13 @@ fn parse_unquoted(&mut self) -> PResult<'a, TokenTree> {
                     )?;
                     let (sep, repeat) = self.parse_sep_and_kleene_op()?;
                     let name_num = macro_parser::count_names(&seq);
-                    return Ok(TokenTree::Sequence(mk_sp(sp.lo, seq_span.hi), SequenceRepetition {
-                        tts: seq,
-                        separator: sep,
-                        op: repeat,
-                        num_captures: name_num
-                    }));
+                    return Ok(TokenTree::Sequence(mk_sp(sp.lo, seq_span.hi),
+                                      Rc::new(SequenceRepetition {
+                                          tts: seq,
+                                          separator: sep,
+                                          op: repeat,
+                                          num_captures: name_num
+                                      })));
                 } else if self.token.is_keyword(keywords::Crate) {
                     self.bump();
                     return Ok(TokenTree::Token(sp, SpecialVarNt(SpecialMacroVar::CrateMacroVar)));
@@ -2848,12 +2849,12 @@ pub fn parse_token_tree(&mut self) -> PResult<'a, TokenTree> {
                     _ => {}
                 }
 
-                Ok(TokenTree::Delimited(span, Delimited {
+                Ok(TokenTree::Delimited(span, Rc::new(Delimited {
                     delim: delim,
                     open_span: open_span,
                     tts: tts,
                     close_span: close_span,
-                }))
+                })))
             },
             _ => {
                 // invariants: the current token is not a left-delimiter,

diff --git a/src/libsyntax/tokenstream.rs b/src/libsyntax/tokenstream.rs
index 35377d14bab7cc0fe4618fb7be541311b1a987c9..f0f0a7bc580d3b6883f8fabc91d00cf40810aeff 100644 (file)
--- a/src/libsyntax/tokenstream.rs
+++ b/src/libsyntax/tokenstream.rs
@@ -21,6 +21,8 @@
 use parse::lexer;
 use parse::token;
 
+use std::rc::Rc;
+
 /// A delimited sequence of token trees
 #[derive(Clone, PartialEq, Eq, RustcEncodable, RustcDecodable, Hash, Debug)]
 pub struct Delimited {
@@ -94,13 +96,13 @@ pub enum TokenTree {
     /// A single token
     Token(Span, token::Token),
     /// A delimited sequence of token trees
-    Delimited(Span, Delimited),
+    Delimited(Span, Rc<Delimited>),
 
     // This only makes sense in MBE macros.
 
     /// A kleene-style repetition sequence with a span
     // FIXME(eddyb) #12938 Use DST.
-    Sequence(Span, SequenceRepetition),
+    Sequence(Span, Rc<SequenceRepetition>),
 }
 
 impl TokenTree {
@@ -149,7 +151,7 @@ pub fn get_tt(&self, index: usize) -> TokenTree {
                     Some(*cnt)
                 }).max().unwrap_or(0);
 
-                TokenTree::Delimited(sp, Delimited {
+                TokenTree::Delimited(sp, Rc::new(Delimited {
                     delim: token::Bracket,
                     open_span: sp,
                     tts: vec![TokenTree::Token(sp, token::Ident(token::str_to_ident("doc"))),
@@ -157,7 +159,7 @@ pub fn get_tt(&self, index: usize) -> TokenTree {
                               TokenTree::Token(sp, token::Literal(
                                   token::StrRaw(token::intern(&stripped), num_of_hashes), None))],
                     close_span: sp,
-                })
+                }))
             }
             (&TokenTree::Delimited(_, ref delimed), _) => {
                 if index == 0 {