[LTO][Legacy] Add API for passing LLVM options separately

[thegameg]

Summary

In order to correctly pass options to LLVM, including options containing spaces which are used as delimiters for multiple options in lto_codegen_debug_options, add a new API: lto_codegen_debug_options_array.

Unfortunately, tools/lto has no testing infrastructure yet, so there are no tests associated with this patch.

[steven_wu]

LGTM with one nitpick.

llvm/lib/LTO/LTOCodeGenerator.cpp

630

Maybe it would be cleaner if remove the old setCodeGenDebugOptions implementation and make the old C API to build an array and call the new interface?

[thegameg]

• Move string splitting to lto_codegen_debug_options.
• Remove null checks since none of the other lto_* functions do it.

llvm/include/llvm-c/lto.h

 
 /**
  * @defgroup LLVMCLTO LTO
  * @ingroup LLVMC
  *
  * @{
  */
 
-#define LTO_API_VERSION 25
+#define LTO_API_VERSION 26
 
 /**
  * \since prior to LTO_API_VERSION=3
  */
 typedef enum {
     LTO_SYMBOL_ALIGNMENT_MASK              = 0x0000001F, /* log2 of alignment */
     LTO_SYMBOL_PERMISSIONS_MASK            = 0x000000E0,
     LTO_SYMBOL_PERMISSIONS_CODE            = 0x000000A0,

  * \since LTO_API_VERSION=12
  */
 extern unsigned int
 lto_api_version(void);
 
 /**
  * Sets options to help debug codegen bugs.
  *
+ * This function takes one or more options separated by spaces.
+ * Warning: passing file paths through this function may confuse the argument
+ * parser if the paths contain spaces.
+ *
  * \since prior to LTO_API_VERSION=3
  */
 extern void
 lto_codegen_debug_options(lto_code_gen_t cg, const char *);
 
 /**
+ * Same as the previous function, but takes every option separately through an
+ * array.
+ *
+ * \since prior to LTO_API_VERSION=26
+ */
+extern void lto_codegen_debug_options_array(lto_code_gen_t cg,
+                                            const char *const *, int number);
+
+/**
  * Initializes LLVM disassemblers.
  * FIXME: This doesn't really belong here.
  *
  * \since LTO_API_VERSION=5
  */
 extern void
 lto_initialize_disassembler(void);
 

llvm/include/llvm/LTO/legacy/LTOCodeGenerator.h

 // with the machine specific code generator.
 //
 //===----------------------------------------------------------------------===//
 
 #ifndef LLVM_LTO_LTOCODEGENERATOR_H
 #define LLVM_LTO_LTOCODEGENERATOR_H
 
 #include "llvm-c/lto.h"
+#include "llvm/ADT/ArrayRef.h"
 #include "llvm/ADT/SmallPtrSet.h"
 #include "llvm/ADT/StringMap.h"
 #include "llvm/ADT/StringSet.h"
 #include "llvm/IR/GlobalValue.h"
 #include "llvm/IR/Module.h"
 #include "llvm/Support/CommandLine.h"
 #include "llvm/Support/Error.h"
 #include "llvm/Support/ToolOutputFile.h"
struct LTOCodeGenerator {
   void addMustPreserveSymbol(StringRef Sym) { MustPreserveSymbols.insert(Sym); }
 
   /// Pass options to the driver and optimization passes.
   ///
   /// These options are not necessarily for debugging purpose (the function
   /// name is misleading).  This function should be called before
   /// LTOCodeGenerator::compilexxx(), and
   /// LTOCodeGenerator::writeMergedModules().
-  void setCodeGenDebugOptions(StringRef Opts);
+  void setCodeGenDebugOptions(ArrayRef<const char *> Opts);
 
   /// Parse the options set in setCodeGenDebugOptions.
   ///
   /// Like \a setCodeGenDebugOptions(), this must be called before
   /// LTOCodeGenerator::compilexxx() and
   /// LTOCodeGenerator::writeMergedModules().
   void parseCodeGenDebugOptions();
 

llvm/lib/LTO/LTOCodeGenerator.cpp

bool LTOCodeGenerator::compileOptimized(ArrayRef<raw_pwrite_stream *> Out) {
 
   reportAndResetTimings();
 
   finishOptimizationRemarks();
 
   return true;
 }
 
-/// setCodeGenDebugOptions - Set codegen debugging options to aid in debugging
-/// LTO problems.
-void LTOCodeGenerator::setCodeGenDebugOptions(StringRef Options) {
-  for (std::pair<StringRef, StringRef> o = getToken(Options); !o.first.empty();
-       o = getToken(o.second))
-    CodegenOptions.push_back(o.first);
+void LTOCodeGenerator::setCodeGenDebugOptions(ArrayRef<const char *> Options) {
+  for (StringRef Option : Options)
+    CodegenOptions.push_back(Option);
 }
 
 void LTOCodeGenerator::parseCodeGenDebugOptions() {
   // if options were requested, set them
   if (!CodegenOptions.empty()) {
     // ParseCommandLineOptions() expects argv[0] to be program name.
     std::vector<const char *> CodegenArgv(1, "libLLVMLTO");
     for (std::string &Arg : CodegenOptions)
       CodegenArgv.push_back(Arg.c_str());
     cl::ParseCommandLineOptions(CodegenArgv.size(), CodegenArgv.data());
   }

llvm/tools/lto/lto.cpp

 //===-lto.cpp - LLVM Link Time Optimizer ----------------------------------===//
 //
 // 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
 //
 //===----------------------------------------------------------------------===//
 //
 // This file implements the Link Time Optimization library. This library is
 // intended to be used by linker to optimize code at link time.
 //
 //===----------------------------------------------------------------------===//
 
 #include "llvm-c/lto.h"
 #include "llvm/ADT/STLExtras.h"
+#include "llvm/ADT/StringExtras.h"
 #include "llvm/Bitcode/BitcodeReader.h"
 #include "llvm/CodeGen/CommandFlags.inc"
 #include "llvm/IR/DiagnosticInfo.h"
 #include "llvm/IR/DiagnosticPrinter.h"
 #include "llvm/IR/LLVMContext.h"
 #include "llvm/LTO/LTO.h"
 #include "llvm/LTO/legacy/LTOCodeGenerator.h"
 #include "llvm/LTO/legacy/LTOModule.h"

 bool lto_codegen_compile_to_file(lto_code_gen_t cg, const char **name) {
   maybeParseOptions(cg);
   return !unwrap(cg)->compile_to_file(
       name, DisableVerify, DisableInline, DisableGVNLoadPRE,
       DisableLTOVectorization);
 }
 
 void lto_codegen_debug_options(lto_code_gen_t cg, const char *opt) {
-  unwrap(cg)->setCodeGenDebugOptions(opt);
+  std::vector<const char *> Options;
+  for (std::pair<StringRef, StringRef> o = getToken(opt); !o.first.empty();
+       o = getToken(o.second))
+    Options.push_back(o.first.data());
+
+  unwrap(cg)->setCodeGenDebugOptions(Options);
+}
+
+void lto_codegen_debug_options_array(lto_code_gen_t cg,
+                                     const char *const *options, int number) {
+  unwrap(cg)->setCodeGenDebugOptions(makeArrayRef(options, number));
 }
 
 unsigned int lto_api_version() { return LTO_API_VERSION; }
 
 void lto_codegen_set_should_internalize(lto_code_gen_t cg,
                                         bool ShouldInternalize) {
   unwrap(cg)->setShouldInternalize(ShouldInternalize);
 }

llvm/tools/lto/lto.exports

 lto_codegen_compile
 lto_codegen_create
 lto_codegen_create_in_local_context
 lto_codegen_dispose
 lto_codegen_set_debug_model
 lto_codegen_set_pic_model
 lto_codegen_write_merged_modules
 lto_codegen_debug_options
+lto_codegen_debug_options_array
 lto_codegen_set_assembler_args
 lto_codegen_set_assembler_path
 lto_codegen_set_cpu
 lto_codegen_compile_to_file
 lto_codegen_optimize
 lto_codegen_compile_optimized
 lto_codegen_set_should_internalize
 lto_codegen_set_should_embed_uselists