llvm/tools/gccld/GenerateCode.cpp

//===- GenerateCode.cpp - Functions for generating executable files  ------===//
//
//                     The LLVM Compiler Infrastructure
//
// This file was developed by the LLVM research group and is distributed under
// the University of Illinois Open Source License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file contains functions for generating executable files once linking
// has finished.  This includes generating a shell script to run the JIT or
// a native executable derived from the bytecode.
//
//===----------------------------------------------------------------------===//

#include "gccld.h"
#include "llvm/System/Program.h"
#include "llvm/Module.h"
#include "llvm/PassManager.h"
#include "llvm/Analysis/LoadValueNumbering.h"
#include "llvm/Analysis/Passes.h"
#include "llvm/Analysis/Verifier.h"
#include "llvm/Bytecode/Archive.h"
#include "llvm/Bytecode/WriteBytecodePass.h"
#include "llvm/Target/TargetData.h"
#include "llvm/Transforms/IPO.h"
#include "llvm/Transforms/Scalar.h"
#include "llvm/Support/SystemUtils.h"
#include "llvm/Support/CommandLine.h"

using namespace llvm;

namespace {
  cl::opt<bool>
  DisableInline("disable-inlining", cl::desc("Do not run the inliner pass"));

  cl::opt<bool>
  Verify("verify", cl::desc("Verify intermediate results of all passes"));

  cl::opt<bool>
  DisableOptimizations("disable-opt",
                       cl::desc("Do not run any optimization passes"));
}

/// CopyEnv - This function takes an array of environment variables and makes a
/// copy of it.  This copy can then be manipulated any way the caller likes
/// without affecting the process's real environment.
///
/// Inputs:
///  envp - An array of C strings containing an environment.
///
/// Return value:
///  NULL - An error occurred.
///
///  Otherwise, a pointer to a new array of C strings is returned.  Every string
///  in the array is a duplicate of the one in the original array (i.e. we do
///  not copy the char *'s from one array to another).
///
static char ** CopyEnv(char ** const envp) {
  // Count the number of entries in the old list;
  unsigned entries;   // The number of entries in the old environment list
  for (entries = 0; envp[entries] != NULL; entries++)
    /*empty*/;

  // Add one more entry for the NULL pointer that ends the list.
  ++entries;

  // If there are no entries at all, just return NULL.
  if (entries == 0)
    return NULL;

  // Allocate a new environment list.
  char **newenv = new char* [entries];
  if ((newenv = new char* [entries]) == NULL)
    return NULL;

  // Make a copy of the list.  Don't forget the NULL that ends the list.
  entries = 0;
  while (envp[entries] != NULL) {
    newenv[entries] = new char[strlen (envp[entries]) + 1];
    strcpy (newenv[entries], envp[entries]);
    ++entries;
  }
  newenv[entries] = NULL;

  return newenv;
}


/// RemoveEnv - Remove the specified environment variable from the environment
/// array.
///
/// Inputs:
///  name - The name of the variable to remove.  It cannot be NULL.
///  envp - The array of environment variables.  It cannot be NULL.
///
/// Notes:
///  This is mainly done because functions to remove items from the environment
///  are not available across all platforms.  In particular, Solaris does not
///  seem to have an unsetenv() function or a setenv() function (or they are
///  undocumented if they do exist).
///
static void RemoveEnv(const char * name, char ** const envp) {
  for (unsigned index=0; envp[index] != NULL; index++) {
    // Find the first equals sign in the array and make it an EOS character.
    char *p = strchr (envp[index], '=');
    if (p == NULL)
      continue;
    else
      *p = '\0';

    // Compare the two strings.  If they are equal, zap this string.
    // Otherwise, restore it.
    if (!strcmp(name, envp[index]))
      *envp[index] = '\0';
    else
      *p = '=';
  }
}

static void dumpArgs(const char **args) {
  std::cerr << *args++;
  while (*args)
    std::cerr << ' ' << *args++;
  std::cerr << '\n' << std::flush;
}

static inline void addPass(PassManager &PM, Pass *P) {
  // Add the pass to the pass manager...
  PM.add(P);

  // If we are verifying all of the intermediate steps, add the verifier...
  if (Verify) PM.add(createVerifierPass());
}

static bool isBytecodeLibrary(const sys::Path &FullPath) {
  // Check for a bytecode file
  if (FullPath.isBytecodeFile()) return true;
  // Check for a dynamic library file
  if (FullPath.isDynamicLibrary()) return false;
  // Check for a true bytecode archive file
  if (FullPath.isArchive() ) {
    std::string ErrorMessage;
    Archive* ar = Archive::OpenAndLoadSymbols( FullPath, &ErrorMessage );
    return ar->isBytecodeArchive();
  }
  return false;
}

static bool isBytecodeLPath(const std::string &LibPath) {
  sys::Path LPath(LibPath);

  // Make sure it exists and is a directory
  try {
    if (!LPath.exists() || !LPath.isDirectory())
      return false;
  } catch (std::string& xcptn) {
    return false;
  }
  
  // Grab the contents of the -L path
  std::set<sys::Path> Files;
  LPath.getDirectoryContents(Files);
  
  // Iterate over the contents one by one to determine
  // if this -L path has any bytecode shared libraries
  // or archives
  std::set<sys::Path>::iterator File = Files.begin();
  std::string dllsuffix = sys::Path::GetDLLSuffix();
  for (; File != Files.end(); ++File) {

    if ( File->isDirectory() )
      continue;

    std::string path = File->toString();

    // Check for an ending '.dll,.so' or '.a' suffix as all
    // other files are not of interest to us here
    if (path.find(dllsuffix, path.size()-dllsuffix.size()) == std::string::npos
        && path.find(".a", path.size()-2) == std::string::npos)
      continue;

    // Finally, check to see if the file is a true bytecode file
    if (isBytecodeLibrary(*File))
      return true;
  }
  return false;
}

/// GenerateBytecode - generates a bytecode file from the specified module.
///
/// Inputs:
///  M           - The module for which bytecode should be generated.
///  StripLevel  - 2 if we should strip all symbols, 1 if we should strip
///                debug info.
///  Internalize - Flags whether all symbols should be marked internal.
///  Out         - Pointer to file stream to which to write the output.
///
/// Returns non-zero value on error.
///
int llvm::GenerateBytecode(Module *M, int StripLevel, bool Internalize,
                           std::ostream *Out) {
  // In addition to just linking the input from GCC, we also want to spiff it up
  // a little bit.  Do this now.
  PassManager Passes;

  if (Verify) Passes.add(createVerifierPass());

  // Add an appropriate TargetData instance for this module...
  addPass(Passes, new TargetData(M));

  // Often if the programmer does not specify proper prototypes for the
  // functions they are calling, they end up calling a vararg version of the
  // function that does not get a body filled in (the real function has typed
  // arguments).  This pass merges the two functions.
  addPass(Passes, createFunctionResolvingPass());

  if (!DisableOptimizations) {
    // Now that composite has been compiled, scan through the module, looking
    // for a main function.  If main is defined, mark all other functions
    // internal.
    addPass(Passes, createInternalizePass(Internalize));

    // Now that we internalized some globals, see if we can hack on them!
    addPass(Passes, createGlobalOptimizerPass());

    // Linking modules together can lead to duplicated global constants, only
    // keep one copy of each constant...
    addPass(Passes, createConstantMergePass());

    // Propagate constants at call sites into the functions they call.
    addPass(Passes, createIPSCCPPass());

    // Remove unused arguments from functions...
    addPass(Passes, createDeadArgEliminationPass());

    if (!DisableInline)
      addPass(Passes, createFunctionInliningPass()); // Inline small functions

    addPass(Passes, createPruneEHPass());            // Remove dead EH info
    addPass(Passes, createGlobalOptimizerPass());    // Optimize globals again.
    addPass(Passes, createGlobalDCEPass());          // Remove dead functions

    // If we didn't decide to inline a function, check to see if we can
    // transform it to pass arguments by value instead of by reference.
    addPass(Passes, createArgumentPromotionPass());

    // The IPO passes may leave cruft around.  Clean up after them.
    addPass(Passes, createInstructionCombiningPass());

    addPass(Passes, createScalarReplAggregatesPass()); // Break up allocas

    // Run a few AA driven optimizations here and now, to cleanup the code.
    addPass(Passes, createGlobalsModRefPass());      // IP alias analysis

    addPass(Passes, createLICMPass());               // Hoist loop invariants
    addPass(Passes, createLoadValueNumberingPass()); // GVN for load instrs
    addPass(Passes, createGCSEPass());               // Remove common subexprs
    addPass(Passes, createDeadStoreEliminationPass()); // Nuke dead stores

    // Cleanup and simplify the code after the scalar optimizations.
    addPass(Passes, createInstructionCombiningPass());

    // Delete basic blocks, which optimization passes may have killed...
    addPass(Passes, createCFGSimplificationPass());

    // Now that we have optimized the program, discard unreachable functions...
    addPass(Passes, createGlobalDCEPass());
  }

  // If the -s or -S command line options were specified, strip the symbols out
  // of the resulting program to make it smaller.  -s and -S are GLD options
  // that we are supporting.
  if (StripLevel)
    addPass(Passes, createStripSymbolsPass(StripLevel == 1));

  // Make sure everything is still good.
  Passes.add(createVerifierPass());

  // Add the pass that writes bytecode to the output file...
  addPass(Passes, new WriteBytecodePass(Out));

  // Run our queue of passes all at once now, efficiently.
  Passes.run(*M);

  return 0;
}

/// GenerateAssembly - generates a native assembly language source file from the
/// specified bytecode file.
///
/// Inputs:
///  InputFilename  - The name of the output bytecode file.
///  OutputFilename - The name of the file to generate.
///  llc            - The pathname to use for LLC.
///
/// Return non-zero value on error.
///
int llvm::GenerateAssembly(const std::string &OutputFilename,
                           const std::string &InputFilename,
                           const sys::Path &llc,
                           bool Verbose) {
  // Run LLC to convert the bytecode file into assembly code.
  std::vector<const char*> args;
  args.push_back(llc.c_str());
  args.push_back("-f");
  args.push_back("-o");
  args.push_back(OutputFilename.c_str());
  args.push_back(InputFilename.c_str());
  args.push_back(0);
  if (Verbose) dumpArgs(&args[0]);
  return sys::Program::ExecuteAndWait(llc, &args[0]);
}

/// GenerateCFile - generates a C source file from the specified bytecode file.
int llvm::GenerateCFile(const std::string &OutputFile,
                        const std::string &InputFile,
                        const sys::Path &llc,
                        bool Verbose) {
  // Run LLC to convert the bytecode file into C.
  std::vector<const char*> args;
  args.push_back(llc.c_str());
  args.push_back("-march=c");
  args.push_back("-f");
  args.push_back("-o");
  args.push_back(OutputFile.c_str());
  args.push_back(InputFile.c_str());
  args.push_back(0);
  if (Verbose) dumpArgs(&args[0]);
  return sys::Program::ExecuteAndWait(llc, &args[0]);
}

/// GenerateNative - generates a native executable file from the specified
/// assembly source file.
///
/// Inputs:
///  InputFilename  - The name of the output bytecode file.
///  OutputFilename - The name of the file to generate.
///  Libraries      - The list of libraries with which to link.
///  gcc            - The pathname to use for GGC.
///  envp           - A copy of the process's current environment.
///
/// Outputs:
///  None.
///
/// Returns non-zero value on error.
///
int llvm::GenerateNative(const std::string &OutputFilename,
                         const std::string &InputFilename,
                         const std::vector<std::string> &LibPaths,
                         const std::vector<std::string> &Libraries,
                         const sys::Path &gcc, char ** const envp,
                         bool Shared,
                         bool ExportAllAsDynamic,
                         const std::vector<std::string> &RPaths,
                         const std::string &SOName,
                         bool Verbose) {
  // Remove these environment variables from the environment of the
  // programs that we will execute.  It appears that GCC sets these
  // environment variables so that the programs it uses can configure
  // themselves identically.
  //
  // However, when we invoke GCC below, we want it to use its normal
  // configuration.  Hence, we must sanitize its environment.
  char ** clean_env = CopyEnv(envp);
  if (clean_env == NULL)
    return 1;
  RemoveEnv("LIBRARY_PATH", clean_env);
  RemoveEnv("COLLECT_GCC_OPTIONS", clean_env);
  RemoveEnv("GCC_EXEC_PREFIX", clean_env);
  RemoveEnv("COMPILER_PATH", clean_env);
  RemoveEnv("COLLECT_GCC", clean_env);


  // Run GCC to assemble and link the program into native code.
  //
  // Note:
  //  We can't just assemble and link the file with the system assembler
  //  and linker because we don't know where to put the _start symbol.
  //  GCC mysteriously knows how to do it.
  std::vector<const char*> args;
  args.push_back(gcc.c_str());
  args.push_back("-fno-strict-aliasing");
  args.push_back("-O3");
  args.push_back("-o");
  args.push_back(OutputFilename.c_str());
  args.push_back(InputFilename.c_str());

  // StringsToDelete - We don't want to call c_str() on temporary strings.
  // If we need a temporary string, copy it here so that the memory is not
  // reclaimed until after the exec call.  All of these strings are allocated
  // with strdup.
  std::vector<char*> StringsToDelete;

  if (Shared) args.push_back("-shared");
  if (ExportAllAsDynamic) args.push_back("-export-dynamic");
  if (!RPaths.empty()) {
    for (std::vector<std::string>::const_iterator I = RPaths.begin(),
        E = RPaths.end(); I != E; I++) {
      std::string rp = "-Wl,-rpath," + *I;
      StringsToDelete.push_back(strdup(rp.c_str()));
      args.push_back(StringsToDelete.back());
    }
  }
  if (!SOName.empty()) {
    std::string so = "-Wl,-soname," + SOName;
    StringsToDelete.push_back(strdup(so.c_str()));
    args.push_back(StringsToDelete.back());
  }

  // Add in the libpaths to find the libraries.
  //
  // Note:
  //  When gccld is called from the llvm-gxx frontends, the -L paths for
  //  the LLVM cfrontend install paths are appended.  We don't want the
  //  native linker to use these -L paths as they contain bytecode files.
  //  Further, we don't want any -L paths that contain bytecode shared
  //  libraries or true bytecode archive files.  We omit them in all such
  //  cases.
  for (unsigned index = 0; index < LibPaths.size(); index++)
    if (!isBytecodeLPath(LibPaths[index])) {
      std::string Tmp = "-L"+LibPaths[index];
      StringsToDelete.push_back(strdup(Tmp.c_str()));
      args.push_back(StringsToDelete.back());
    }

  // Add in the libraries to link.
  for (unsigned index = 0; index < Libraries.size(); index++)
    // HACK: If this is libg, discard it.  This gets added by the compiler
    // driver when doing: 'llvm-gcc main.c -Wl,-native -o a.out -g'. Note that
    // this should really be fixed by changing the llvm-gcc compiler driver.
    if (Libraries[index] != "crtend" && Libraries[index] != "g") {
      std::string Tmp = "-l"+Libraries[index];
      StringsToDelete.push_back(strdup(Tmp.c_str()));
      args.push_back(StringsToDelete.back());
    }
  args.push_back(0); // Null terminate.

  // Run the compiler to assembly and link together the program.
  if (Verbose) dumpArgs(&args[0]);
  int Res = sys::Program::ExecuteAndWait(gcc, &args[0],(const char**)clean_env);

  delete [] clean_env;

  while (!StringsToDelete.empty()) {
    free(StringsToDelete.back());
    StringsToDelete.pop_back();
  }
  return Res;
}