Auto merge of #85020 - lrh2000:named-upvars, r=tmandry

Name the captured upvars for closures/generators in debuginfo

Previously, debuggers print closures as something like
```
y::main::closure-0 (0x7fffffffdd34)
```
The pointer actually references to an upvar. It is not very obvious, especially for beginners.

It's because upvars don't have names before, as they are packed into a tuple. This PR names the upvars, so we can expect to see something like
```
y::main::closure-0 {_captured_ref__b: 0x[...]}
```

r? `@tmandry`
Discussed at #84752 (comment) .

compiler/rustc_codegen_llvm/src/debuginfo/metadata.rs

@@ -1280,6 +1280,31 @@ fn prepare_struct_metadata(
 // Tuples
 //=-----------------------------------------------------------------------------
 
+/// Returns names of captured upvars for closures and generators.
+///
+/// Here are some examples:
+///  - `name__field1__field2` when the upvar is captured by value.
+///  - `_ref__name__field` when the upvar is captured by reference.
+fn closure_saved_names_of_captured_variables(tcx: TyCtxt<'tcx>, def_id: DefId) -> Vec<String> {
+    let body = tcx.optimized_mir(def_id);
+
+    body.var_debug_info
+        .iter()
+        .filter_map(|var| {
+            let is_ref = match var.value {
+                mir::VarDebugInfoContents::Place(place) if place.local == mir::Local::new(1) => {
+                    // The projection is either `[.., Field, Deref]` or `[.., Field]`. It
+                    // implies whether the variable is captured by value or by reference.
+                    matches!(place.projection.last().unwrap(), mir::ProjectionElem::Deref)
+                }
+                _ => return None,
+            };
+            let prefix = if is_ref { "_ref__" } else { "" };
+            Some(prefix.to_owned() + &var.name.as_str())
+        })
+        .collect::<Vec<_>>()
+}
+
 /// Creates `MemberDescription`s for the fields of a tuple.
 struct TupleMemberDescriptionFactory<'tcx> {
     ty: Ty<'tcx>,
@@ -1289,14 +1314,25 @@ struct TupleMemberDescriptionFactory<'tcx> {
 
 impl<'tcx> TupleMemberDescriptionFactory<'tcx> {
     fn create_member_descriptions(&self, cx: &CodegenCx<'ll, 'tcx>) -> Vec<MemberDescription<'ll>> {
+        let mut capture_names = match *self.ty.kind() {
+            ty::Generator(def_id, ..) | ty::Closure(def_id, ..) => {
+                Some(closure_saved_names_of_captured_variables(cx.tcx, def_id).into_iter())
+            }
+            _ => None,
+        };
         let layout = cx.layout_of(self.ty);
         self.component_types
             .iter()
             .enumerate()
             .map(|(i, &component_type)| {
                 let (size, align) = cx.size_and_align_of(component_type);
+                let name = if let Some(names) = capture_names.as_mut() {
+                    names.next().unwrap()
+                } else {
+                    format!("__{}", i)
+                };
                 MemberDescription {
-                    name: format!("__{}", i),
+                    name,
                     type_metadata: type_metadata(cx, component_type, self.span),
                     offset: layout.fields.offset(i),
                     size,

compiler/rustc_middle/src/query/mod.rs

@@ -342,6 +342,16 @@ rustc_queries! {
         }
     }
 
+    query symbols_for_closure_captures(
+        key: (LocalDefId, DefId)
+    ) -> Vec<rustc_span::Symbol> {
+        desc {
+            |tcx| "symbols for captures of closure `{}` in `{}`",
+            tcx.def_path_str(key.1),
+            tcx.def_path_str(key.0.to_def_id())
+        }
+    }
+
     /// MIR after our optimization passes have run. This is MIR that is ready
     /// for codegen. This is also the only query that can fetch non-local MIR, at present.
     query optimized_mir(key: DefId) -> &'tcx mir::Body<'tcx> {

compiler/rustc_middle/src/ty/closure.rs

@@ -3,10 +3,12 @@ use crate::hir::place::{
 };
 use crate::{mir, ty};
 
+use std::fmt::Write;
+
 use rustc_data_structures::fx::{FxHashMap, FxIndexMap};
 use rustc_hir as hir;
 use rustc_hir::def_id::{DefId, LocalDefId};
-use rustc_span::Span;
+use rustc_span::{Span, Symbol};
 
 use super::{Ty, TyCtxt};
 
@@ -159,6 +161,43 @@ impl CapturedPlace<'tcx> {
         place_to_string_for_capture(tcx, &self.place)
     }
 
+    /// Returns a symbol of the captured upvar, which looks like `name__field1__field2`.
+    fn to_symbol(&self, tcx: TyCtxt<'tcx>) -> Symbol {
+        let hir_id = match self.place.base {
+            HirPlaceBase::Upvar(upvar_id) => upvar_id.var_path.hir_id,
+            base => bug!("Expected an upvar, found {:?}", base),
+        };
+        let mut symbol = tcx.hir().name(hir_id).as_str().to_string();
+
+        let mut ty = self.place.base_ty;
+        for proj in self.place.projections.iter() {
+            match proj.kind {
+                HirProjectionKind::Field(idx, variant) => match ty.kind() {
+                    ty::Tuple(_) => write!(&mut symbol, "__{}", idx).unwrap(),
+                    ty::Adt(def, ..) => {
+                        write!(
+                            &mut symbol,
+                            "__{}",
+                            def.variants[variant].fields[idx as usize].ident.name.as_str(),
+                        )
+                        .unwrap();
+                    }
+                    ty => {
+                        bug!("Unexpected type {:?} for `Field` projection", ty)
+                    }
+                },
+
+                // Ignore derefs for now, as they are likely caused by
+                // autoderefs that don't appear in the original code.
+                HirProjectionKind::Deref => {}
+                proj => bug!("Unexpected projection {:?} in captured place", proj),
+            }
+            ty = proj.ty;
+        }
+
+        Symbol::intern(&symbol)
+    }
+
     /// Returns the hir-id of the root variable for the captured place.
     /// e.g., if `a.b.c` was captured, would return the hir-id for `a`.
     pub fn get_root_variable(&self) -> hir::HirId {
@@ -209,6 +248,15 @@ impl CapturedPlace<'tcx> {
     }
 }
 
+fn symbols_for_closure_captures<'tcx>(
+    tcx: TyCtxt<'tcx>,
+    def_id: (LocalDefId, DefId),
+) -> Vec<Symbol> {
+    let typeck_results = tcx.typeck(def_id.0);
+    let captures = typeck_results.closure_min_captures_flattened(def_id.1);
+    captures.into_iter().map(|captured_place| captured_place.to_symbol(tcx)).collect()
+}
+
 /// Return true if the `proj_possible_ancestor` represents an ancestor path
 /// to `proj_capture` or `proj_possible_ancestor` is same as `proj_capture`,
 /// assuming they both start off of the same root variable.
@@ -392,3 +440,7 @@ impl BorrowKind {
         }
     }
 }
+
+pub fn provide(providers: &mut ty::query::Providers) {
+    *providers = ty::query::Providers { symbols_for_closure_captures, ..*providers }
+}

compiler/rustc_middle/src/ty/mod.rs

@@ -16,7 +16,6 @@ pub use self::IntVarValue::*;
 pub use self::Variance::*;
 pub use adt::*;
 pub use assoc::*;
-pub use closure::*;
 pub use generics::*;
 pub use vtable::*;
 
@@ -55,6 +54,12 @@ pub use rustc_type_ir::*;
 
 pub use self::binding::BindingMode;
 pub use self::binding::BindingMode::*;
+pub use self::closure::{
+    is_ancestor_or_same_capture, place_to_string_for_capture, BorrowKind, CaptureInfo,
+    CapturedPlace, ClosureKind, MinCaptureInformationMap, MinCaptureList,
+    RootVariableMinCaptureList, UpvarBorrow, UpvarCapture, UpvarCaptureMap, UpvarId, UpvarListMap,
+    UpvarPath, CAPTURE_STRUCT_LOCAL,
+};
 pub use self::consts::{Const, ConstInt, ConstKind, InferConst, ScalarInt, Unevaluated, ValTree};
 pub use self::context::{
     tls, CanonicalUserType, CanonicalUserTypeAnnotation, CanonicalUserTypeAnnotations,
@@ -1980,6 +1985,7 @@ pub fn ast_uint_ty(uty: UintTy) -> ast::UintTy {
 }
 
 pub fn provide(providers: &mut ty::query::Providers) {
+    closure::provide(providers);
     context::provide(providers);
     erase_regions::provide(providers);
     layout::provide(providers);

compiler/rustc_mir_build/src/build/mod.rs

@@ -15,7 +15,7 @@ use rustc_middle::mir::*;
 use rustc_middle::thir::{BindingMode, Expr, ExprId, LintLevel, PatKind, Thir};
 use rustc_middle::ty::subst::Subst;
 use rustc_middle::ty::{self, Ty, TyCtxt, TypeFoldable, TypeckResults};
-use rustc_span::symbol::{kw, sym};
+use rustc_span::symbol::sym;
 use rustc_span::Span;
 use rustc_target::spec::abi::Abi;
 
@@ -902,13 +902,16 @@ impl<'a, 'tcx> Builder<'a, 'tcx> {
                 ty::Generator(_, substs, _) => ty::UpvarSubsts::Generator(substs),
                 _ => span_bug!(self.fn_span, "upvars with non-closure env ty {:?}", closure_ty),
             };
+            let def_id = self.def_id.as_local().unwrap();
+            let capture_syms = tcx.symbols_for_closure_captures((def_id, fn_def_id));
             let capture_tys = upvar_substs.upvar_tys();
-            let captures_with_tys =
-                hir_typeck_results.closure_min_captures_flattened(fn_def_id).zip(capture_tys);
+            let captures_with_tys = hir_typeck_results
+                .closure_min_captures_flattened(fn_def_id)
+                .zip(capture_tys.zip(capture_syms));
 
             self.upvar_mutbls = captures_with_tys
                 .enumerate()
-                .map(|(i, (captured_place, ty))| {
+                .map(|(i, (captured_place, (ty, sym)))| {
                     let capture = captured_place.info.capture_kind;
                     let var_id = match captured_place.place.base {
                         HirPlaceBase::Upvar(upvar_id) => upvar_id.var_path.hir_id,
@@ -917,14 +920,6 @@ impl<'a, 'tcx> Builder<'a, 'tcx> {
 
                     let mutability = captured_place.mutability;
 
-                    // FIXME(project-rfc-2229#8): Store more precise information
-                    let mut name = kw::Empty;
-                    if let Some(Node::Binding(pat)) = tcx_hir.find(var_id) {
-                        if let hir::PatKind::Binding(_, _, ident, _) = pat.kind {
-                            name = ident.name;
-                        }
-                    }
-
                     let mut projs = closure_env_projs.clone();
                     projs.push(ProjectionElem::Field(Field::new(i), ty));
                     match capture {
@@ -935,7 +930,7 @@ impl<'a, 'tcx> Builder<'a, 'tcx> {
                     };
 
                     self.var_debug_info.push(VarDebugInfo {
-                        name,
+                        name: sym,
                         source_info: SourceInfo::outermost(tcx_hir.span(var_id)),
                         value: VarDebugInfoContents::Place(Place {
                             local: ty::CAPTURE_STRUCT_LOCAL,

src/test/debuginfo/captured-fields-1.rs

@@ -0,0 +1,99 @@
+// compile-flags:-g
+
+// === GDB TESTS ===================================================================================
+
+// gdb-command:run
+// gdb-command:print test
+// gdbr-check:$1 = captured_fields_1::main::{closure#0} {_ref__my_ref__my_field1: 0x[...]}
+// gdb-command:continue
+// gdb-command:print test
+// gdbr-check:$2 = captured_fields_1::main::{closure#1} {_ref__my_ref__my_field2: 0x[...]}
+// gdb-command:continue
+// gdb-command:print test
+// gdbr-check:$3 = captured_fields_1::main::{closure#2} {_ref__my_ref: 0x[...]}
+// gdb-command:continue
+// gdb-command:print test
+// gdbr-check:$4 = captured_fields_1::main::{closure#3} {my_ref: 0x[...]}
+// gdb-command:continue
+// gdb-command:print test
+// gdbr-check:$5 = captured_fields_1::main::{closure#4} {my_var__my_field2: 22}
+// gdb-command:continue
+// gdb-command:print test
+// gdbr-check:$6 = captured_fields_1::main::{closure#5} {my_var: captured_fields_1::MyStruct {my_field1: 11, my_field2: 22}}
+// gdb-command:continue
+
+// === LLDB TESTS ==================================================================================
+
+// lldb-command:run
+// lldb-command:print test
+// lldbg-check:(captured_fields_1::main::{closure#0}) $0 = { _ref__my_ref__my_field1 = 0x[...] }
+// lldb-command:continue
+// lldb-command:print test
+// lldbg-check:(captured_fields_1::main::{closure#1}) $1 = { _ref__my_ref__my_field2 = 0x[...] }
+// lldb-command:continue
+// lldb-command:print test
+// lldbg-check:(captured_fields_1::main::{closure#2}) $2 = { _ref__my_ref = 0x[...] }
+// lldb-command:continue
+// lldb-command:print test
+// lldbg-check:(captured_fields_1::main::{closure#3}) $3 = { my_ref = 0x[...] }
+// lldb-command:continue
+// lldb-command:print test
+// lldbg-check:(captured_fields_1::main::{closure#4}) $4 = { my_var__my_field2 = 22 }
+// lldb-command:continue
+// lldb-command:print test
+// lldbg-check:(captured_fields_1::main::{closure#5}) $5 = { my_var = { my_field1 = 11 my_field2 = 22 } }
+// lldb-command:continue
+
+#![feature(capture_disjoint_fields)]
+#![allow(unused)]
+
+struct MyStruct {
+    my_field1: u32,
+    my_field2: u32,
+}
+
+fn main() {
+    let mut my_var = MyStruct {
+        my_field1: 11,
+        my_field2: 22,
+    };
+    let my_ref = &mut my_var;
+
+    let test = || {
+        let a = &mut my_ref.my_field1;
+    };
+
+    _zzz(); // #break
+
+    let test = || {
+        let a = &my_ref.my_field2;
+    };
+
+    _zzz(); // #break
+
+    let test = || {
+        let a = &my_ref;
+    };
+
+    _zzz(); // #break
+
+    let test = || {
+        let a = my_ref;
+    };
+
+    _zzz(); // #break
+
+    let test = move || {
+        let a = my_var.my_field2;
+    };
+
+    _zzz(); // #break
+
+    let test = || {
+        let a = my_var;
+    };
+
+    _zzz(); // #break
+}
+
+fn _zzz() {}

src/test/debuginfo/captured-fields-2.rs

@@ -0,0 +1,55 @@
+// compile-flags:-g
+
+// === GDB TESTS ===================================================================================
+
+// gdb-command:run
+// gdb-command:print my_ref__my_field1
+// gdbr-check:$1 = 11
+// gdb-command:continue
+// gdb-command:print my_var__my_field2
+// gdbr-check:$2 = 22
+// gdb-command:continue
+
+// === LLDB TESTS ==================================================================================
+
+// lldb-command:run
+// lldb-command:print my_ref__my_field1
+// lldbg-check:(unsigned int) $0 = 11
+// lldb-command:continue
+// lldb-command:print my_var__my_field2
+// lldbg-check:(unsigned int) $1 = 22
+// lldb-command:continue
+
+#![feature(capture_disjoint_fields)]
+#![allow(unused)]
+
+struct MyStruct {
+    my_field1: u32,
+    my_field2: u32,
+}
+
+fn main() {
+    let mut my_var = MyStruct {
+        my_field1: 11,
+        my_field2: 22,
+    };
+    let my_ref = &mut my_var;
+
+    let test = || {
+        let a = my_ref.my_field1;
+
+        _zzz(); // #break
+    };
+
+    test();
+
+    let test = move || {
+        let a = my_var.my_field2;
+
+        _zzz(); // #break
+    };
+
+    test();
+}
+
+fn _zzz() {}