[llvm-cxxfilt] Improve strip-underscore behavior

[steven_wu]

Summary

For platform that uses macho format, c++filt should be stripping the
leading underscore by default. Introduce the binutil compatible "-n"
option to control strip-undercore behaivor together with the existing
"-_" option and fallback to system default if none of them are set.

rdar://problem/57173514

[steven_wu]

ping

[erik.pilkington]

LGTM

llvm/tools/llvm-cxxfilt/llvm-cxxfilt.cpp

77

Side comment: Why does this function take an ostream then only write "import thunk for" into it? Looks like that will produce broken output for import thunks in the Split case in demangleLine. Can't we just return a string with "import thunk for" prefixed in that case?

[compnerd]

Sorry about the delay on the review!

llvm/tools/llvm-cxxfilt/llvm-cxxfilt.cpp

69–76

Please rewrite this as:

if (StripUnderscore)
  return true;
if (NoStripUnderscore)
  return false;
return Triple(sys::getProcessTriple()).isOSBinFormatMachO();

[steven_wu]

Address review feedback

[steven_wu]

llvm/tools/llvm-cxxfilt/llvm-cxxfilt.cpp

77

Agree with that. That will be a followup patch.

[steven_wu]

Fix comments

llvm/test/tools/llvm-cxxfilt/abitag.test

-RUN: llvm-cxxfilt _Z14returns_stringB5cxx11v _Z6globalB5cxx11 _Z6globalB12a_longer_tag | FileCheck %s
+RUN: llvm-cxxfilt -n _Z14returns_stringB5cxx11v _Z6globalB5cxx11 _Z6globalB12a_longer_tag | FileCheck %s
 
 CHECK: returns_string[abi:cxx11]()
 CHECK-NEXT: global[abi:cxx11]
 CHECK-NEXT: global[abi:a_longer_tag]
 

llvm/test/tools/llvm-cxxfilt/coff-import.test

 RUN: llvm-cxxfilt -_ ___imp__ZSt6futureIvE | FileCheck %s
-RUN: llvm-cxxfilt __imp__ZSt6futureIvE | FileCheck %s
+RUN: llvm-cxxfilt -n __imp__ZSt6futureIvE | FileCheck %s
 
 CHECK: import thunk for std::future<void>
 

llvm/test/tools/llvm-cxxfilt/darwin.test

+REQUIRES: system-darwin
+
+RUN: llvm-cxxfilt __Z1fv | FileCheck %s
+CHECK: f()

llvm/test/tools/llvm-cxxfilt/delimiters.test

 RUN:      '_Z3Foo]'   \
 RUN:      '_Z3Foo^'   \
 RUN:      '_Z3Foo`'   \
 RUN:      "\"_Z3Foo{\"" \
 RUN:      '_Z3Foo|'   \
 RUN:      '_Z3Foo}'   \
 RUN:      '_Z3Foo~,,' \
 RUN:      '_Z3Foo,,_Z3Bar::_Z3Baz  _Z3Foo,_Z3Bar:_Z3Baz' \
-RUN:      '_Z3Foo$ ._Z3Foo' | llvm-cxxfilt | FileCheck %s
+RUN:      '_Z3Foo$ ._Z3Foo' | llvm-cxxfilt -n | FileCheck %s
 
 CHECK: ,,Foo!
 CHECK: Foo"
 CHECK: Foo#
 CHECK: Foo%
 CHECK: Foo&
 CHECK: Foo'
 CHECK: Foo(

llvm/test/tools/llvm-cxxfilt/empty-line.test

 # Show that llvm-cxxfilt can handle empty lines in the input.
 RUN: echo "" > %t
 RUN: echo "_Z3foov" >> %t
 RUN: echo "" >> %t
-RUN: llvm-cxxfilt < %t | FileCheck %s
+RUN: llvm-cxxfilt -n < %t | FileCheck %s
 
 CHECK:       {{^$}}
 CHECK-NEXT:  foo()
 CHECK-EMPTY:

llvm/test/tools/llvm-cxxfilt/invalid.test

-RUN: llvm-cxxfilt _Z1fi __Z1fi f ___ZSt1ff_block_invoke | FileCheck %s
+RUN: llvm-cxxfilt -n _Z1fi __Z1fi f ___ZSt1ff_block_invoke | FileCheck %s
 
 CHECK: f(int)
 CHECK-NEXT: __Z1fi
 CHECK-NEXT: f
 CHECK-NEXT: invocation function for block in std::f(float)

llvm/test/tools/llvm-cxxfilt/noargs.test

-RUN: sed -n 's/^STDIN: //p' %s | llvm-cxxfilt | FileCheck %s
+RUN: sed -n 's/^STDIN: //p' %s | llvm-cxxfilt -n | FileCheck %s
 
 STDIN: _Znw
 STDIN: _Znwj
 STDIN: _Znwm
 
 CHECK: operator new
 CHECK: operator new(unsigned int)
 CHECK: operator new(unsigned long)
 

llvm/test/tools/llvm-cxxfilt/simple.test

-RUN: llvm-cxxfilt _Z1fi abc | FileCheck %s
-RUN: echo "Mangled _Z1fi and _Z3foov in string." | llvm-cxxfilt \
+RUN: llvm-cxxfilt -n _Z1fi abc | FileCheck %s
+RUN: echo "Mangled _Z1fi and _Z3foov in string." | llvm-cxxfilt -n \
 RUN:   | FileCheck %s --check-prefix=CHECK-STRING
-RUN: llvm-cxxfilt "CLI remains mangled _Z1fi" \
+RUN: llvm-cxxfilt -n "CLI remains mangled _Z1fi" \
 RUN:   | FileCheck %s --check-prefix=CHECK-MANGLED
 
 CHECK: f(int)
 CHECK-NEXT: abc
 CHECK-STRING: Mangled f(int) and foo() in string.
 CHECK-MANGLED: CLI remains mangled _Z1fi

llvm/test/tools/llvm-cxxfilt/types.test

-RUN: llvm-cxxfilt -t f i | FileCheck %s
+RUN: llvm-cxxfilt -n -t f i | FileCheck %s
 RUN: echo "f i" | llvm-cxxfilt -t | FileCheck %s --check-prefix="CHECK-STRING"
 
 CHECK: float
 CHECK-NEXT: int
 CHECK-STRING: float int

llvm/test/tools/llvm-cxxfilt/underscore.test

 RUN: llvm-cxxfilt -_ __ZN2ns1fE _ZSt1f _f | FileCheck %s -check-prefix CHECK-STRIPPED
-RUN: llvm-cxxfilt __ZN2ns1fE _ZSt1f _f | FileCheck %s -check-prefix CHECK-UNSTRIPPED
+RUN: llvm-cxxfilt -n __ZN2ns1fE _ZSt1f _f | FileCheck %s -check-prefix CHECK-UNSTRIPPED
 
 CHECK-STRIPPED: ns::f
 CHECK-STRIPPED: _ZSt1f
 CHECK-STRIPPED: _f
 
 CHECK-UNSTRIPPED: __ZN2ns1fE
 CHECK-UNSTRIPPED: std::f
 CHECK-UNSTRIPPED: _f
 

llvm/tools/llvm-cxxfilt/llvm-cxxfilt.cpp

 //===-- llvm-c++filt.cpp --------------------------------------------------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 
 #include "llvm/ADT/StringExtras.h"
+#include "llvm/ADT/Triple.h"
 #include "llvm/Demangle/Demangle.h"
 #include "llvm/Support/CommandLine.h"
+#include "llvm/Support/Host.h"
 #include "llvm/Support/InitLLVM.h"
 #include "llvm/Support/raw_ostream.h"
 #include <cstdlib>
 #include <iostream>
 
 using namespace llvm;
 
 enum Style {
static cl::alias FormatShort("s", cl::desc("alias for --format"),
                              cl::aliasopt(Format));
 
 static cl::opt<bool> StripUnderscore("strip-underscore",
                                      cl::desc("strip the leading underscore"),
                                      cl::init(false));
 static cl::alias StripUnderscoreShort("_",
                                       cl::desc("alias for --strip-underscore"),
                                       cl::aliasopt(StripUnderscore));
+static cl::opt<bool>
+    NoStripUnderscore("no-strip-underscore",
+                      cl::desc("do not strip the leading underscore"),
+                      cl::init(false));
+static cl::alias
+    NoStripUnderscoreShort("n", cl::desc("alias for --no-strip-underscore"),
+                           cl::aliasopt(NoStripUnderscore));
 
 static cl::opt<bool>
     Types("types",
           cl::desc("attempt to demangle types as well as function names"),
           cl::init(false));
 static cl::alias TypesShort("t", cl::desc("alias for --types"),
                             cl::aliasopt(Types));
 
 static cl::list<std::string>
 Decorated(cl::Positional, cl::desc("<mangled>"), cl::ZeroOrMore);
 
 static cl::extrahelp
     HelpResponse("\nPass @FILE as argument to read options from FILE.\n");
 
+static bool shouldStripUnderscore() {
+  if (StripUnderscore)
+    return true;
+  if (NoStripUnderscore)
+    return false;
+  // If none of them are set, use the default value for platform.
+  // macho has symbols prefix with "_" so strip by default.
+  return Triple(sys::getProcessTriple()).isOSBinFormatMachO();
+}
+
 static std::string demangle(llvm::raw_ostream &OS, const std::string &Mangled) {
   int Status;
 
   const char *DecoratedStr = Mangled.c_str();
-  if (StripUnderscore)
+  if (shouldStripUnderscore())
     if (DecoratedStr[0] == '_')
       ++DecoratedStr;
   size_t DecoratedLength = strlen(DecoratedStr);
 
   char *Undecorated = nullptr;
 
   if (Types ||
       ((DecoratedLength >= 2 && strncmp(DecoratedStr, "_Z", 2) == 0) ||