Update LLVM lib/Demangle to 87039c048c0cbc3d8cbba86187269b006bf2f373

This adds support for demangling Rust symbols.

Added two simple test cases for demangling Rust.

Ran:
      cp ~/src/llvm-project/llvm/include/llvm/Demangle/*.h third_party/llvm/include/llvm/Demangle/
      cp ~/src/llvm-project/llvm/lib/Demangle/*.cpp third_party/llvm/lib/Demangle/
      cp ~/src/llvm-project/llvm/LICENSE.TXT third_party/llvm/LICENSE.txt

Author: tengyifei

CMakeLists.txt

@@ -33,9 +33,11 @@ endif()
 include_directories(third_party/llvm/include)
 add_executable(demumble
                demumble.cc
+               third_party/llvm/lib/Demangle/Demangle.cpp
                third_party/llvm/lib/Demangle/ItaniumDemangle.cpp
                third_party/llvm/lib/Demangle/MicrosoftDemangle.cpp
                third_party/llvm/lib/Demangle/MicrosoftDemangleNodes.cpp
+               third_party/llvm/lib/Demangle/RustDemangle.cpp
 )
 set_target_properties(demumble PROPERTIES CXX_STANDARD 14
                                           CXX_STANDARD_REQUIRED ON)

README.md

@@ -1,13 +1,13 @@
 # demumble
 
-`demumble` demangles both Itanium and Visual Studio symbols. It runs on both
-POSIX and Windows.
+`demumble` demangles both Itanium, Rust, and Visual Studio symbols. It runs on
+both POSIX and Windows.
 
     $ demumble _Z4funcPci
     func(char*, int)
     $ demumble '?Fx_i@@YAHP6AHH@Z@Z'
     int __cdecl Fx_i(int (__cdecl *)(int))
-    
+
 ## Download
 
 There are prebuilt x64 binaries for Linux, Mac (10.9+), and Windows on the

demumble.cc

@@ -26,6 +26,9 @@ static void print_demangled(const char* format, const char* s, size_t* n_used) {
   if (char* itanium = llvm::itaniumDemangle(s, NULL, NULL, NULL)) {
     printf(format, itanium, s);
     free(itanium);
+  } else if (char* rust = llvm::rustDemangle(s, NULL, NULL, NULL)) {
+    printf(format, rust, s);
+    free(rust);
   } else if (char* ms = llvm::microsoftDemangle(s, n_used, NULL, NULL, NULL)) {
     printf(format, ms, s);
     free(ms);
@@ -39,6 +42,12 @@ static bool is_mangle_char_itanium(char c) {
          (c >= '0' && c <= '9') || c == '_' || c == '$';
 }
 
+static bool is_mangle_char_rust(char c) {
+  // See https://rust-lang.github.io/rfcs/2603-rust-symbol-name-mangling-v0.html.
+  return (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') ||
+         (c >= '0' && c <= '9') || c == '_';
+}
+
 static bool is_mangle_char_win(char c) {
   return (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') ||
          (c >= '0' && c <= '9') || strchr("?_@$", c);
@@ -53,6 +62,11 @@ static bool is_plausible_itanium_prefix(char* s) {
   return strstr(prefix, "_Z");
 }
 
+static bool is_plausible_rust_prefix(char* s) {
+  // Rust symbols start with "_R".
+  return s[0] == '_' && s[1] == 'R';
+}
+
 static char buf[8192];
 int main(int argc, char* argv[]) {
   enum { kPrintAll, kPrintMatching } print_mode = kPrintAll;
@@ -121,6 +135,9 @@ int main(int argc, char* argv[]) {
         else if (is_plausible_itanium_prefix(cur))
           while (cur + n_sym != end && is_mangle_char_itanium(cur[n_sym]))
             ++n_sym;
+        else if (is_plausible_rust_prefix(cur))
+          while (cur + n_sym != end && is_mangle_char_rust(cur[n_sym]))
+            ++n_sym;
         else {
           if (print_mode == kPrintAll)
             printf("_");

demumble_test.py

@@ -6,6 +6,10 @@
     ('demumble hello', 'hello\n'),
     ('demumble _Z4funcPci _Z1fv', 'func(char*, int)\nf()\n'),
     ('demumble < _Z4funcPci _Z1fv', 'func(char*, int)\nf()\n'),
+    ('demumble _RINvNtC3std3mem8align_ofdE _RNvNvC5mylib3foo3bar',
+     'std::mem::align_of::<f64>\nmylib::foo::bar\n'),
+    ('demumble < _RINvNtC3std3mem8align_ofdE _RNvNvC5mylib3foo3bar',
+     'std::mem::align_of::<f64>\nmylib::foo::bar\n'),
     ('demumble ?Fxi@@YAHP6AHH@Z@Z', 'int __cdecl Fxi(int (__cdecl *)(int))\n'),
     ('demumble ??0S@@QEAA@$$QEAU0@@Z', 'public: __cdecl S::S(struct S &&)\n'),
     ('demumble ??_C@_02PCEFGMJL@hi?$AA@', '"hi"\n'),