Initial support for dumping DWARF corresponding to Swift code

The DWARF data for Swift code has a top-level DW_TAG_module DIE as the
child of the DW_TAG_compile_unit DIE and the parent of the
DW_TAG_subprogram DIEs that dump_syms uses to locate functions.
dump_syms needs to process DW_TAG_module DIEs as introducing nested
scopes to make it work with Swift.

This also reworks demangling to be language-specific, so that the C++
demangler isn't invoked when processing Swift code. The DWARF data for
Swift code presents its mangled names in the same form as used for C++
(DW_AT_MIPS_linkage_name or DW_AT_linkage_name) but the mangling is
Swift-specific (beginning with _T instead of _Z). There is no
programmatic interface to a Swift name demangler as an analogue to C++'s
__cxa_demangle(), so mangled Swift names are exposed as-is. Xcode's
"xcrun swift-demangle" can be used to post-process these mangled Swift
names on macOS.

Support for mangled names presented in a DW_AT_linkage_name attribute,
as used by DWARF 4, is added. This supersedes the earlier use of
DW_AT_MIPS_linkage_name.

BUG=google-breakpad:702,google-breakpad:715
R=ted.mielczarek@gmail.com

Review URL: https://codereview.chromium.org/2147523005 .

6 files changed, 130 insertions, 33 deletions

diff --git a/src/common/dwarf/dwarf2enums.h b/src/common/dwarf/dwarf2enums.h
index 6b8a7245..09567555 100644
--- a/src/common/dwarf/dwarf2enums.h
+++ b/src/common/dwarf/dwarf2enums.h
@@ -232,6 +232,8 @@ enum DwarfAttribute {
   DW_AT_call_column   = 0x57,
   DW_AT_call_file     = 0x58,
   DW_AT_call_line     = 0x59,
+  // DWARF 4
+  DW_AT_linkage_name  = 0x6e,
   // SGI/MIPS extensions.
   DW_AT_MIPS_fde = 0x2001,
   DW_AT_MIPS_loop_begin = 0x2002,
@@ -499,7 +501,7 @@ enum DwarfOpcode {
   DW_OP_call_frame_cfa               =0x9c,
   DW_OP_bit_piece                    =0x9d,
   DW_OP_lo_user                      =0xe0,
-  DW_OP_hi_user                      =0xff,  
+  DW_OP_hi_user                      =0xff,
   // GNU extensions
   DW_OP_GNU_push_tls_address         =0xe0,
   // Extensions for Fission.  See http://gcc.gnu.org/wiki/DebugFission.
@@ -542,6 +544,7 @@ enum DwarfLanguage
     DW_LANG_ObjC_plus_plus           =0x0011,
     DW_LANG_UPC                      =0x0012,
     DW_LANG_D                        =0x0013,
+    DW_LANG_Swift                    =0x001e,
     // Implementation-defined language code range.
     DW_LANG_lo_user = 0x8000,
     DW_LANG_hi_user = 0xffff,
@@ -668,7 +671,7 @@ enum DwarfPointerEncoding
     // encoding (except DW_EH_PE_aligned), and indicates that the
     // encoded value represents the address at which the true address
     // is stored, not the true address itself.
-    DW_EH_PE_indirect	= 0x80  
+    DW_EH_PE_indirect	= 0x80
   };
 
 }  // namespace dwarf2reader
diff --git a/src/common/dwarf_cu_to_module.cc b/src/common/dwarf_cu_to_module.cc
index 479e39b2..100f3962 100644
--- a/src/common/dwarf_cu_to_module.cc
+++ b/src/common/dwarf_cu_to_module.cc
@@ -39,9 +39,6 @@
 #include "common/dwarf_cu_to_module.h"
 
 #include <assert.h>
-#if !defined(__ANDROID__)
-#include <cxxabi.h>
-#endif
 #include <inttypes.h>
 #include <stdint.h>
 #include <stdio.h>
@@ -350,20 +347,22 @@ void DwarfCUToModule::GenericDIEHandler::ProcessAttributeString(
     case dwarf2reader::DW_AT_name:
       name_attribute_ = AddStringToPool(data);
       break;
-    case dwarf2reader::DW_AT_MIPS_linkage_name: {
-      char* demangled = NULL;
-      int status = -1;
-#if !defined(__ANDROID__)  // Android NDK doesn't provide abi::__cxa_demangle.
-      demangled = abi::__cxa_demangle(data.c_str(), NULL, NULL, &status);
-#endif
-      if (status != 0) {
-        cu_context_->reporter->DemangleError(data, status);
-        demangled_name_ = "";
-        break;
-      }
-      if (demangled) {
-        demangled_name_ = AddStringToPool(demangled);
-        free(reinterpret_cast<void*>(demangled));
+    case dwarf2reader::DW_AT_MIPS_linkage_name:
+    case dwarf2reader::DW_AT_linkage_name: {
+      string demangled;
+      Language::DemangleResult result =
+          cu_context_->language->DemangleName(data, &demangled);
+      switch (result) {
+        case Language::kDemangleSuccess:
+          demangled_name_ = AddStringToPool(demangled);
+          break;
+
+        case Language::kDemangleFailure:
+          cu_context_->reporter->DemangleError(data);
+          // fallthrough
+        case Language::kDontDemangle:
+          demangled_name_.clear();
+          break;
       }
       break;
     }
@@ -676,11 +675,10 @@ void DwarfCUToModule::WarningReporter::UnnamedFunction(uint64 offset) {
           filename_.c_str(), offset);
 }
 
-void DwarfCUToModule::WarningReporter::DemangleError(
-    const string &input, int error) {
+void DwarfCUToModule::WarningReporter::DemangleError(const string &input) {
   CUHeading();
-  fprintf(stderr, "%s: warning: failed to demangle %s with error %d\n",
-          filename_.c_str(), input.c_str(), error);
+  fprintf(stderr, "%s: warning: failed to demangle %s\n",
+          filename_.c_str(), input.c_str());
 }
 
 void DwarfCUToModule::WarningReporter::UnhandledInterCUReference(
@@ -761,6 +759,7 @@ dwarf2reader::DIEHandler *DwarfCUToModule::FindChildHandler(
     case dwarf2reader::DW_TAG_class_type:
     case dwarf2reader::DW_TAG_structure_type:
     case dwarf2reader::DW_TAG_union_type:
+    case dwarf2reader::DW_TAG_module:
       return new NamedScopeHandler(cu_context_.get(), child_context_.get(),
                                    offset);
     default:
@@ -774,6 +773,10 @@ void DwarfCUToModule::SetLanguage(DwarfLanguage language) {
       cu_context_->language = Language::Java;
       break;
 
+    case dwarf2reader::DW_LANG_Swift:
+      cu_context_->language = Language::Swift;
+      break;
+
     // DWARF has no generic language code for assembly language; this is
     // what the GNU toolchain uses.
     case dwarf2reader::DW_LANG_Mips_Assembler:
diff --git a/src/common/dwarf_cu_to_module.h b/src/common/dwarf_cu_to_module.h
index fca92710..a36b82b0 100644
--- a/src/common/dwarf_cu_to_module.h
+++ b/src/common/dwarf_cu_to_module.h
@@ -202,7 +202,7 @@ class DwarfCUToModule: public dwarf2reader::RootDIEHandler {
     virtual void UnnamedFunction(uint64 offset);
 
     // __cxa_demangle() failed to demangle INPUT.
-    virtual void DemangleError(const string &input, int error);
+    virtual void DemangleError(const string &input);
 
     // The DW_FORM_ref_addr at OFFSET to TARGET was not handled because
     // FilePrivate did not retain the inter-CU specification data.
diff --git a/src/common/dwarf_cu_to_module_unittest.cc b/src/common/dwarf_cu_to_module_unittest.cc
index 619e90a2..e789f9da 100644
--- a/src/common/dwarf_cu_to_module_unittest.cc
+++ b/src/common/dwarf_cu_to_module_unittest.cc
@@ -83,7 +83,7 @@ class MockWarningReporter: public DwarfCUToModule::WarningReporter {
   MOCK_METHOD1(UncoveredFunction, void(const Module::Function &function));
   MOCK_METHOD1(UncoveredLine, void(const Module::Line &line));
   MOCK_METHOD1(UnnamedFunction, void(uint64 offset));
-  MOCK_METHOD2(DemangleError, void(const string &input, int error));
+  MOCK_METHOD1(DemangleError, void(const string &input));
   MOCK_METHOD2(UnhandledInterCUReference, void(uint64 offset, uint64 target));
 };
 
diff --git a/src/common/language.cc b/src/common/language.cc
index c2fd81f6..3a598f69 100644
--- a/src/common/language.cc
+++ b/src/common/language.cc
@@ -34,18 +34,66 @@
 
 #include "common/language.h"
 
+#include <stdlib.h>
+
+#if !defined(__ANDROID__)
+#include <cxxabi.h>
+#endif
+
+#include <limits>
+
+namespace {
+
+string MakeQualifiedNameWithSeparator(const string& parent_name,
+                                      const char* separator,
+                                      const string& name) {
+  if (parent_name.empty()) {
+    return name;
+  }
+
+  return parent_name + separator + name;
+}
+
+}  // namespace
+
 namespace google_breakpad {
 
 // C++ language-specific operations.
 class CPPLanguage: public Language {
  public:
   CPPLanguage() {}
+
   string MakeQualifiedName(const string &parent_name,
                            const string &name) const {
-    if (parent_name.empty())
-      return name;
-    else
-      return parent_name + "::" + name;
+    return MakeQualifiedNameWithSeparator(parent_name, "::", name);
+  }
+
+  virtual DemangleResult DemangleName(const string& mangled,
+                                      std::string* demangled) const {
+#if defined(__ANDROID__)
+    // Android NDK doesn't provide abi::__cxa_demangle.
+    demangled->clear();
+    return kDontDemangle;
+#else
+    int status;
+    char* demangled_c =
+        abi::__cxa_demangle(mangled.c_str(), NULL, NULL, &status);
+
+    DemangleResult result;
+    if (status == 0) {
+      result = kDemangleSuccess;
+      demangled->clear();
+    } else {
+      result = kDemangleFailure;
+      demangled->assign(demangled_c);
+    }
+
+    if (demangled_c) {
+      free(reinterpret_cast<void*>(demangled_c));
+    }
+
+    return result;
+#endif
   }
 };
 
@@ -54,19 +102,45 @@ CPPLanguage CPPLanguageSingleton;
 // Java language-specific operations.
 class JavaLanguage: public Language {
  public:
+  JavaLanguage() {}
+
   string MakeQualifiedName(const string &parent_name,
                            const string &name) const {
-    if (parent_name.empty())
-      return name;
-    else
-      return parent_name + "." + name;
+    return MakeQualifiedNameWithSeparator(parent_name, ".", name);
   }
 };
 
 JavaLanguage JavaLanguageSingleton;
 
+// Swift language-specific operations.
+class SwiftLanguage: public Language {
+ public:
+  SwiftLanguage() {}
+
+  string MakeQualifiedName(const string &parent_name,
+                           const string &name) const {
+    return MakeQualifiedNameWithSeparator(parent_name, ".", name);
+  }
+
+  virtual DemangleResult DemangleName(const string& mangled,
+                                      std::string* demangled) const {
+    // There is no programmatic interface to a Swift demangler. Pass through the
+    // mangled form because it encodes more information than the qualified name
+    // that would have been built by MakeQualifiedName(). The output can be
+    // post-processed by xcrun swift-demangle to transform mangled Swift names
+    // into something more readable.
+    demangled->assign(mangled);
+    return kDemangleSuccess;
+  }
+};
+
+SwiftLanguage SwiftLanguageSingleton;
+
 // Assembler language-specific operations.
 class AssemblerLanguage: public Language {
+ public:
+  AssemblerLanguage() {}
+
   bool HasFunctions() const { return false; }
   string MakeQualifiedName(const string &parent_name,
                            const string &name) const {
@@ -78,6 +152,7 @@ AssemblerLanguage AssemblerLanguageSingleton;
 
 const Language * const Language::CPlusPlus = &CPPLanguageSingleton;
 const Language * const Language::Java = &JavaLanguageSingleton;
+const Language * const Language::Swift = &SwiftLanguageSingleton;
 const Language * const Language::Assembler = &AssemblerLanguageSingleton;
 
 } // namespace google_breakpad
diff --git a/src/common/language.h b/src/common/language.h
index bbe30334..05786199 100644
--- a/src/common/language.h
+++ b/src/common/language.h
@@ -77,9 +77,25 @@ class Language {
   virtual string MakeQualifiedName (const string &parent_name,
                                     const string &name) const = 0;
 
+  enum DemangleResult {
+    // Demangling was not performed because it’s not appropriate to attempt.
+    kDontDemangle = -1,
+
+    kDemangleSuccess,
+    kDemangleFailure,
+  };
+
+  // Wraps abi::__cxa_demangle() or similar for languages where appropriate.
+  virtual DemangleResult DemangleName(const string& mangled,
+                                      std::string* demangled) const {
+    demangled->clear();
+    return kDontDemangle;
+  }
+
   // Instances for specific languages.
   static const Language * const CPlusPlus,
                         * const Java,
+                        * const Swift,
                         * const Assembler;
 };