Use GetArgumentVector to retrieve the utf-8 encoded arguments on all platforms

[asmith]

Summary

After running the lld unit tests on Windows, it was discovered that various llvm tools on Windows are not handling unicode command lines. An example of a test that does not work on Windows because of this issue is lld/test/ELF/format-binary-non-ascii.s.

This problem was originally fixed in r329468 by modifying Windows/Path.inc to try and use the current codepage. Eli Friedman recommended instead modifying the llvm tools to properly handle unicode command lines by calling sys::Process::GetArgumentVector.

This reverts r329468 and updates llvm-mc and llvm-objdump to handle unicode command lines. With this change and the change in D45550, lld/test/ELF/format-binary-non-ascii.s passes.

[efriedma]

LGTM

Support/Windows/Path.inc

 namespace sys  {
 namespace path {
 
-// Convert a (likely) UTF-8 path to UTF-16.  Also, if the absolute equivalent of the
+// Convert a UTF-8 path to UTF-16.  Also, if the absolute equivalent of the
 // path is longer than CreateDirectory can tolerate, make it absolute and
 // prefixed by '\\?\'.
 std::error_code widenPath(const Twine &Path8,

 
   // Several operations would convert Path8 to SmallString; more efficient to
   // do it once up front.
-  SmallString<2*MAX_PATH> Path8Str;
+  SmallString<128> Path8Str;
   Path8.toVector(Path8Str);
 
   // If we made this path absolute, how much longer would it get?

       else
         llvm::sys::path::append(FullPath, *I);
     }
-    Path8Str = FullPath;
+    return UTF8ToUTF16(FullPath, Path16);
   }
 
-  // Path8Str now contains the full path or the original path
-  // If the conversion from UTF8 to UTF16 fails because of ERROR_NO_UNICODE_TRANSLATION,
-  // we also try using the current code page before giving up
-  auto ec = UTF8ToUTF16(Path8Str, Path16);
-  if (ec == mapWindowsError(ERROR_NO_UNICODE_TRANSLATION)) {
-    ec = CurCPToUTF16(Path8Str, Path16);
-  }
-  return ec;
+  // Just use the caller's original path.
+  return UTF8ToUTF16(Path8Str, Path16);
 }
 } // end namespace path
 

llvm-mc/llvm-mc.cpp

 #include "llvm/Support/ManagedStatic.h"
 #include "llvm/Support/MemoryBuffer.h"
 #include "llvm/Support/PrettyStackTrace.h"
+#include "llvm/Support/Process.h"
 #include "llvm/Support/Signals.h"
 #include "llvm/Support/SourceMgr.h"
 #include "llvm/Support/TargetRegistry.h"

 }
 
 static std::string DwarfDebugFlags;
-static void setDwarfDebugFlags(int argc, char **argv) {
+static void setDwarfDebugFlags(int argc, const char **argv) {
   if (!getenv("RC_DEBUG_OPTIONS"))
     return;
   for (int i = 0; i < argc; i++) {

   return Res;
 }
 
-int main(int argc, char **argv) {
+static ExitOnError ExitOnErr;
+
+int main(int argc_, const char **argv_) {
   // Print a stack trace if we signal out.
-  sys::PrintStackTraceOnErrorSignal(argv[0]);
-  PrettyStackTraceProgram X(argc, argv);
-  llvm_shutdown_obj Y;  // Call llvm_shutdown() on exit.
+  sys::PrintStackTraceOnErrorSignal(argv_[0]);
+  PrettyStackTraceProgram X(argc_, argv_);
+
+  ExitOnErr.setBanner("llvm-mc: ");
+
+  SmallVector<const char *, 256> argv;
+  SpecificBumpPtrAllocator<char> ArgAllocator;
+  ExitOnErr(errorCodeToError(sys::Process::GetArgumentVector(
+      argv, makeArrayRef(argv_, argc_), ArgAllocator)));
+
+  llvm_shutdown_obj Y; // Call llvm_shutdown() on exit.
 
   // Initialize targets and assembly printers/parsers.
   llvm::InitializeAllTargetInfos();

   // Register the target printer for --version.
   cl::AddExtraVersionPrinter(TargetRegistry::printRegisteredTargetsForVersion);
 
-  cl::ParseCommandLineOptions(argc, argv, "llvm machine code playground\n");
+  cl::ParseCommandLineOptions(argv.size(), argv.data(), "llvm machine code playground\n");
   MCTargetOptions MCOptions = InitMCTargetOptionsFromFlags();
   TripleName = Triple::normalize(TripleName);
-  setDwarfDebugFlags(argc, argv);
+  setDwarfDebugFlags(argv.size(), argv.data());
 
   setDwarfDebugProducer();
 

llvm-objdump/llvm-objdump.cpp

 #include "llvm/Support/ManagedStatic.h"
 #include "llvm/Support/MemoryBuffer.h"
 #include "llvm/Support/PrettyStackTrace.h"
+#include "llvm/Support/Process.h"
 #include "llvm/Support/Signals.h"
 #include "llvm/Support/SourceMgr.h"
 #include "llvm/Support/TargetRegistry.h"

     report_error(file, object_error::invalid_file_type);
 }
 
-int main(int argc, char **argv) {
+static ExitOnError ExitOnErr;
+
+int main(int argc_, const char **argv_) {
   // Print a stack trace if we signal out.
-  sys::PrintStackTraceOnErrorSignal(argv[0]);
-  PrettyStackTraceProgram X(argc, argv);
-  llvm_shutdown_obj Y;  // Call llvm_shutdown() on exit.
+  sys::PrintStackTraceOnErrorSignal(argv_[0]);
+  PrettyStackTraceProgram X(argc_, argv_);
+
+  ExitOnErr.setBanner("llvm-objdump: ");
+
+  SmallVector<const char *, 256> argv;
+  SpecificBumpPtrAllocator<char> ArgAllocator;
+  ExitOnErr(errorCodeToError(sys::Process::GetArgumentVector(
+      argv, makeArrayRef(argv_, argc_), ArgAllocator)));
+
+  llvm_shutdown_obj Y; // Call llvm_shutdown() on exit.
 
   // Initialize targets and assembly printers/parsers.
   llvm::InitializeAllTargetInfos();

   // Register the target printer for --version.
   cl::AddExtraVersionPrinter(TargetRegistry::printRegisteredTargetsForVersion);
 
-  cl::ParseCommandLineOptions(argc, argv, "llvm object file dumper\n");
+  cl::ParseCommandLineOptions(argv.size(), argv.data(), "llvm object file dumper\n");
   TripleName = Triple::normalize(TripleName);
 
   ToolName = argv[0];