Merge branch 'main' into lwg-4140

Author: frederick-vs-ja

bolt/include/bolt/Core/BinaryContext.h

@@ -537,6 +537,17 @@ class BinaryContext {
   BinaryFunction *createInjectedBinaryFunction(const std::string &Name,
                                                bool IsSimple = true);
 
+  /// Patch the original binary contents at address \p Address with a sequence
+  /// of instructions from the \p Instructions list. The callee is responsible
+  /// for checking that the sequence doesn't cross any function or section
+  /// boundaries.
+  ///
+  /// Optional \p Name can be assigned to the patch. The name will be emitted to
+  /// the symbol table at \p Address.
+  BinaryFunction *createInstructionPatch(uint64_t Address,
+                                         InstructionListType &Instructions,
+                                         const Twine &Name = "");
+
   std::vector<BinaryFunction *> &getInjectedBinaryFunctions() {
     return InjectedBinaryFunctions;
   }

bolt/include/bolt/Core/MCPlusBuilder.h

@@ -637,10 +637,6 @@ class MCPlusBuilder {
     return false;
   }
 
-  virtual void getADRReg(const MCInst &Inst, MCPhysReg &RegName) const {
-    llvm_unreachable("not implemented");
-  }
-
   virtual bool isMoveMem2Reg(const MCInst &Inst) const { return false; }
 
   virtual bool mayLoad(const MCInst &Inst) const {
@@ -1538,6 +1534,13 @@ class MCPlusBuilder {
     llvm_unreachable("not implemented");
   }
 
+  /// Undo the linker's ADRP+ADD to ADR relaxation. Take \p ADRInst and return
+  /// ADRP+ADD instruction sequence.
+  virtual InstructionListType undoAdrpAddRelaxation(const MCInst &ADRInst,
+                                                    MCContext *Ctx) const {
+    llvm_unreachable("not implemented");
+  }
+
   /// Return not 0 if the instruction CurInst, in combination with the recent
   /// history of disassembled instructions supplied by [Begin, End), is a linker
   /// generated veneer/stub that needs patching. This happens in AArch64 when

bolt/include/bolt/Passes/PatchEntries.h

@@ -26,8 +26,6 @@ class PatchEntries : public BinaryFunctionPass {
   struct Patch {
     const MCSymbol *Symbol;
     uint64_t Address;
-    uint64_t FileOffset;
-    BinarySection *Section;
   };
 
 public:

bolt/lib/Core/BinaryContext.cpp

@@ -2400,6 +2400,32 @@ BinaryContext::createInjectedBinaryFunction(const std::string &Name,
   return BF;
 }
 
+BinaryFunction *BinaryContext::createInstructionPatch(
+    uint64_t Address, InstructionListType &Instructions, const Twine &Name) {
+  ErrorOr<BinarySection &> Section = getSectionForAddress(Address);
+  assert(Section && "cannot get section for patching");
+  assert(Section->hasSectionRef() && Section->isText() &&
+         "can only patch input file code sections");
+
+  const uint64_t FileOffset =
+      Section->getInputFileOffset() + Address - Section->getAddress();
+
+  std::string PatchName = Name.str();
+  if (PatchName.empty()) {
+    // Assign unique name to the patch.
+    static uint64_t N = 0;
+    PatchName = "__BP_" + std::to_string(N++);
+  }
+
+  BinaryFunction *PBF = createInjectedBinaryFunction(PatchName);
+  PBF->setOutputAddress(Address);
+  PBF->setFileOffset(FileOffset);
+  PBF->setOriginSection(&Section.get());
+  PBF->addBasicBlock()->addInstructions(Instructions);
+
+  return PBF;
+}
+
 std::pair<size_t, size_t>
 BinaryContext::calculateEmittedSize(BinaryFunction &BF, bool FixBranches) {
   // Adjust branch instruction to match the current layout.

bolt/lib/Passes/ADRRelaxationPass.cpp

@@ -71,14 +71,10 @@ void ADRRelaxationPass::runOnFunction(BinaryFunction &BF) {
             continue;
       }
 
-      MCPhysReg Reg;
-      BC.MIB->getADRReg(Inst, Reg);
-      int64_t Addend = BC.MIB->getTargetAddend(Inst);
-      InstructionListType Addr;
-
+      InstructionListType AdrpAdd;
       {
         auto L = BC.scopeLock();
-        Addr = BC.MIB->materializeAddress(Symbol, BC.Ctx.get(), Reg, Addend);
+        AdrpAdd = BC.MIB->undoAdrpAddRelaxation(Inst, BC.Ctx.get());
       }
 
       if (It != BB.begin() && BC.MIB->isNoop(*std::prev(It))) {
@@ -99,7 +95,7 @@ void ADRRelaxationPass::runOnFunction(BinaryFunction &BF) {
         PassFailed = true;
         return;
       }
-      It = BB.replaceInstruction(It, Addr);
+      It = BB.replaceInstruction(It, AdrpAdd);
     }
   }
 }

bolt/lib/Passes/PatchEntries.cpp

@@ -83,9 +83,8 @@ Error PatchEntries::runOnFunctions(BinaryContext &BC) {
         return false;
       }
 
-      PendingPatches.emplace_back(Patch{Symbol, Function.getAddress() + Offset,
-                                        Function.getFileOffset() + Offset,
-                                        Function.getOriginSection()});
+      PendingPatches.emplace_back(
+          Patch{Symbol, Function.getAddress() + Offset});
       NextValidByte = Offset + PatchSize;
       if (NextValidByte > Function.getMaxSize()) {
         if (opts::Verbosity >= 1)
@@ -118,16 +117,12 @@ Error PatchEntries::runOnFunctions(BinaryContext &BC) {
     }
 
     for (Patch &Patch : PendingPatches) {
-      BinaryFunction *PatchFunction = BC.createInjectedBinaryFunction(
+      // Add instruction patch to the binary.
+      InstructionListType Instructions;
+      BC.MIB->createLongTailCall(Instructions, Patch.Symbol, BC.Ctx.get());
+      BinaryFunction *PatchFunction = BC.createInstructionPatch(
+          Patch.Address, Instructions,
           NameResolver::append(Patch.Symbol->getName(), ".org.0"));
-      // Force the function to be emitted at the given address.
-      PatchFunction->setOutputAddress(Patch.Address);
-      PatchFunction->setFileOffset(Patch.FileOffset);
-      PatchFunction->setOriginSection(Patch.Section);
-
-      InstructionListType Seq;
-      BC.MIB->createLongTailCall(Seq, Patch.Symbol, BC.Ctx.get());
-      PatchFunction->addBasicBlock()->addInstructions(Seq);
 
       // Verify the size requirements.
       uint64_t HotSize, ColdSize;

bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp

@@ -278,13 +278,23 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
     return Inst.getOpcode() == AArch64::ADDXri;
   }
 
-  void getADRReg(const MCInst &Inst, MCPhysReg &RegName) const override {
+  MCPhysReg getADRReg(const MCInst &Inst) const {
     assert((isADR(Inst) || isADRP(Inst)) && "Not an ADR instruction");
     assert(MCPlus::getNumPrimeOperands(Inst) != 0 &&
            "No operands for ADR instruction");
     assert(Inst.getOperand(0).isReg() &&
            "Unexpected operand in ADR instruction");
-    RegName = Inst.getOperand(0).getReg();
+    return Inst.getOperand(0).getReg();
+  }
+
+  InstructionListType undoAdrpAddRelaxation(const MCInst &ADRInst,
+                                            MCContext *Ctx) const override {
+    assert(isADR(ADRInst) && "ADR instruction expected");
+
+    const MCPhysReg Reg = getADRReg(ADRInst);
+    const MCSymbol *Target = getTargetSymbol(ADRInst);
+    const uint64_t Addend = getTargetAddend(ADRInst);
+    return materializeAddress(Target, Ctx, Reg, Addend);
   }
 
   bool isTB(const MCInst &Inst) const {

bolt/test/X86/Inputs/define_bar.s

@@ -1,5 +1,5 @@
 # Mocks a vtable object weak def in the C++ stdlib.
-  .data.rel.ro
+  .section .data.rel.ro,"aw"
   .weak bar
   .type bar, %object
 bar:

clang/bindings/python/clang/cindex.py

@@ -2713,6 +2713,21 @@ def visitor(base, children):
         conf.lib.clang_visitCXXBaseClasses(self, fields_visit_callback(visitor), bases)
         return iter(bases)
 
+    def get_methods(self):
+        """Return an iterator for accessing the methods of this type."""
+
+        def visitor(method, children):
+            assert method != conf.lib.clang_getNullCursor()
+
+            # Create reference to TU so it isn't GC'd before Cursor.
+            method._tu = self._tu
+            methods.append(method)
+            return 1  # continue
+
+        methods: list[Cursor] = []
+        conf.lib.clang_visitCXXMethods(self, fields_visit_callback(visitor), methods)
+        return iter(methods)
+
     def get_exception_specification_kind(self):
         """
         Return the kind of the exception specification; a value from
@@ -4020,6 +4035,7 @@ def set_property(self, property, value):
     ),
     ("clang_visitChildren", [Cursor, cursor_visit_callback, py_object], c_uint),
     ("clang_visitCXXBaseClasses", [Type, fields_visit_callback, py_object], c_uint),
+    ("clang_visitCXXMethods", [Type, fields_visit_callback, py_object], c_uint),
     ("clang_Cursor_getNumArguments", [Cursor], c_int),
     ("clang_Cursor_getArgument", [Cursor, c_uint], Cursor),
     ("clang_Cursor_getNumTemplateArguments", [Cursor], c_int),

clang/bindings/python/tests/cindex/test_type.py

@@ -559,3 +559,21 @@ class Template : public A, public B, virtual C {
         self.assertEqual(bases[1].get_base_offsetof(cursor_type_decl), 96)
         self.assertTrue(bases[2].is_virtual_base())
         self.assertEqual(bases[2].get_base_offsetof(cursor_type_decl), 128)
+
+    def test_class_methods(self):
+        source = """
+        template <typename T>
+        class Template { void Foo(); };
+        typedef Template<int> instance;
+        instance bar;
+        """
+        tu = get_tu(source, lang="cpp", flags=["--target=x86_64-linux-gnu"])
+        cursor = get_cursor(tu, "instance")
+        cursor_type = cursor.underlying_typedef_type
+        self.assertEqual(cursor.kind, CursorKind.TYPEDEF_DECL)
+        methods = list(cursor_type.get_methods())
+        self.assertEqual(len(methods), 4)
+        self.assertEqual(methods[0].kind, CursorKind.CXX_METHOD)
+        self.assertEqual(methods[1].kind, CursorKind.CONSTRUCTOR)
+        self.assertEqual(methods[2].kind, CursorKind.CONSTRUCTOR)
+        self.assertEqual(methods[3].kind, CursorKind.CONSTRUCTOR)