Add a library for disassembling machine instructions at a given address

CMake/LLVM/LLVM.cmake

@@ -62,4 +62,16 @@ endif()
 
 # Add the external llvm project to the build.
 add_subdirectory(${propeller_llvm_src_dir}/llvm ${propeller_llvm_build_dir}/llvm EXCLUDE_FROM_ALL)
-include_directories(${propeller_llvm_src_dir}/llvm/include ${propeller_llvm_build_dir}/llvm/include)
+include_directories(
+  ${propeller_llvm_src_dir}
+  ${propeller_llvm_src_dir}/llvm/include
+  ${propeller_llvm_build_dir}
+  ${propeller_llvm_build_dir}/llvm/include
+)
+# Add generated target-specific library directories.
+foreach (tgt ${LLVM_TARGETS_TO_BUILD})
+  include_directories(
+    ${propeller_llvm_src_dir}/llvm/lib/Target/${tgt}
+    ${propeller_llvm_build_dir}/llvm/lib/Target/${tgt}
+  )
+endforeach()

propeller/BUILD

@@ -649,6 +649,29 @@ cc_library(
     ],
 )
 
+cc_library(
+    name = "mini_disassembler",
+    srcs = ["mini_disassembler.cc"],
+    hdrs = ["mini_disassembler.h"],
+    deps = [
+        "@abseil-cpp//absl/base:nullability",
+        "@abseil-cpp//absl/log",
+        "@abseil-cpp//absl/log:check",
+        "@abseil-cpp//absl/memory",
+        "@abseil-cpp//absl/status",
+        "@abseil-cpp//absl/status:statusor",
+        "@abseil-cpp//absl/strings",
+        "@abseil-cpp//absl/strings:str_format",
+        "@llvm-project//llvm:AllTargetsAsmParsers",  # buildcleaner: keep
+        "@llvm-project//llvm:AllTargetsDisassemblers",  # buildcleaner: keep
+        "@llvm-project//llvm:MC",
+        "@llvm-project//llvm:MCDisassembler",
+        "@llvm-project//llvm:Object",
+        "@llvm-project//llvm:Support",
+        "@llvm-project//llvm:TargetParser",
+    ],
+)
+
 ########################
 #  Tests & Test Utils  #
 ########################
@@ -1125,3 +1148,21 @@ cc_test(
         "@com_google_googletest//:gtest_main",
     ],
 )
+
+cc_test(
+    name = "mini_disassembler_test",
+    srcs = ["mini_disassembler_test.cc"],
+    data = [
+        "//propeller/testdata:llvm_function_samples.binary",
+    ],
+    deps = [
+        ":binary_content",
+        ":mini_disassembler",
+        ":status_testing_macros",
+        "@abseil-cpp//absl/status:status_matchers",
+        "@abseil-cpp//absl/strings",
+        "@com_google_googletest//:gtest_main",
+        "@llvm-project//llvm:MC",
+        "@llvm-project//llvm:X86UtilsAndDesc",
+    ],
+)

propeller/CMakeLists.txt

@@ -42,6 +42,7 @@ add_library(propeller_lib OBJECT
   file_perf_data_provider.cc
   frequencies_branch_aggregator.cc
   lbr_branch_aggregator.cc
+  mini_disassembler.cc
   node_chain.cc
   node_chain_assembly.cc
   node_chain_builder.cc
@@ -66,6 +67,11 @@ target_link_libraries(propeller_lib
   quipper_protos
   # keep-sorted end
 )
+foreach (tgt ${LLVM_TARGETS_TO_BUILD})
+  foreach (tool AsmParser Desc Disassembler Info)
+    target_link_libraries(propeller_lib LLVM${tgt}${tool})
+  endforeach()
+endforeach()
 
 # Build all CXX test utilities into a unified library.
 add_library(propeller_test_lib OBJECT

propeller/mini_disassembler.cc

@@ -0,0 +1,139 @@
+#include "propeller/mini_disassembler.h"
+
+#include <cstdint>
+#include <memory>
+#include <string>
+
+#include "absl/base/nullability.h"
+#include "absl/log/check.h"
+#include "absl/log/log.h"
+#include "absl/memory/memory.h"
+#include "absl/status/status.h"
+#include "absl/status/statusor.h"
+#include "absl/strings/str_format.h"
+#include "absl/strings/string_view.h"
+#include "llvm/ADT/ArrayRef.h"
+#include "llvm/ADT/StringRef.h"
+#include "llvm/MC/MCInst.h"
+#include "llvm/MC/MCInstrDesc.h"
+#include "llvm/MC/MCInstrInfo.h"
+#include "llvm/MC/TargetRegistry.h"
+#include "llvm/Object/ObjectFile.h"
+#include "llvm/Support/Error.h"
+#include "llvm/Support/TargetSelect.h"
+#include "llvm/Support/raw_ostream.h"
+#include "llvm/TargetParser/Triple.h"
+
+namespace propeller {
+absl::StatusOr<absl::Nonnull<std::unique_ptr<MiniDisassembler>>>
+MiniDisassembler::Create(const llvm::object::ObjectFile *object_file) {
+  auto disassembler =
+      absl::WrapUnique<MiniDisassembler>(new MiniDisassembler(object_file));
+
+  std::string err;
+  llvm::InitializeAllTargetInfos();
+  llvm::InitializeAllTargetMCs();
+  llvm::InitializeAllAsmParsers();
+  llvm::InitializeAllDisassemblers();
+
+  llvm::Triple triple;
+  triple.setArch(
+      llvm::Triple::ArchType(object_file->getArch()));
+  const llvm::Target *target =
+      llvm::TargetRegistry::lookupTarget(triple.normalize(), err);
+  if (target == nullptr) {
+    return absl::FailedPreconditionError(absl::StrFormat(
+        "no target for triple '%s': %s", triple.getArchName(), err));
+  }
+  disassembler->mri_ =
+      absl::WrapUnique(target->createMCRegInfo(triple.getTriple()));
+  if (disassembler->mri_ == nullptr) {
+    return absl::FailedPreconditionError(absl::StrFormat(
+        "createMCRegInfo failed for triple '%s'", triple.getArchName()));
+  }
+  disassembler->asm_info_ = absl::WrapUnique(target->createMCAsmInfo(
+      *disassembler->mri_, triple.getTriple(), llvm::MCTargetOptions()));
+  if (disassembler->asm_info_ == nullptr) {
+    return absl::FailedPreconditionError(absl::StrFormat(
+        "createMCAsmInfo failed for triple '%s'", triple.getArchName()));
+  }
+
+  disassembler->sti_ = absl::WrapUnique(target->createMCSubtargetInfo(
+      triple.getTriple(), /*CPU=*/"", /*Features=*/""));
+  if (disassembler->sti_ == nullptr) {
+    return absl::FailedPreconditionError(absl::StrFormat(
+        "createMCSubtargetInfo failed for triple '%s'", triple.getArchName()));
+  }
+
+  disassembler->mii_ = absl::WrapUnique(target->createMCInstrInfo());
+  if (disassembler->mii_ == nullptr) {
+    return absl::FailedPreconditionError(absl::StrFormat(
+        "createMCInstrInfo failed for triple '%s'", triple.getArchName()));
+  }
+
+  disassembler->mia_ =
+      absl::WrapUnique(target->createMCInstrAnalysis(disassembler->mii_.get()));
+  if (disassembler->mia_ == nullptr) {
+    return absl::FailedPreconditionError(absl::StrFormat(
+        "createMCInstrAnalysis failed for triple '%s'", triple.getArchName()));
+  }
+
+  disassembler->ctx_ = std::make_unique<llvm::MCContext>(
+      triple, disassembler->asm_info_.get(), disassembler->mri_.get(),
+      disassembler->sti_.get());
+  disassembler->disasm_ = absl::WrapUnique(
+      target->createMCDisassembler(*disassembler->sti_, *disassembler->ctx_));
+  if (disassembler->disasm_ == nullptr)
+    return absl::FailedPreconditionError(
+        absl::StrFormat("createMCDisassembler failed"));
+
+  return disassembler;
+}
+
+absl::StatusOr<llvm::MCInst> MiniDisassembler::DisassembleOne(
+    uint64_t binary_address) {
+  for (const auto &section : object_file_->sections()) {
+    if (!section.isText() || section.isVirtual()) {
+      continue;
+    }
+    if (binary_address < section.getAddress() ||
+        binary_address >= section.getAddress() + section.getSize()) {
+      continue;
+    }
+    llvm::Expected<llvm::StringRef> content = section.getContents();
+    if (!content) {
+      return absl::FailedPreconditionError("section has no content");
+    }
+    llvm::ArrayRef<uint8_t> content_bytes(
+        reinterpret_cast<const uint8_t *>(content->data()), content->size());
+    uint64_t section_offset = binary_address - section.getAddress();
+    llvm::MCInst inst;
+    uint64_t size;
+    if (!disasm_->getInstruction(inst, size,
+                                 content_bytes.slice(section_offset),
+                                 binary_address, llvm::nulls())) {
+      return absl::FailedPreconditionError(absl::StrFormat(
+          "getInstruction failed at binary address 0x%lx", binary_address));
+    }
+    return inst;
+  }
+  return absl::FailedPreconditionError(absl::StrFormat(
+      "no section containing address 0x%lx found", binary_address));
+}
+
+bool MiniDisassembler::MayAffectControlFlow(const llvm::MCInst &inst) {
+  return mii_->get(inst.getOpcode()).mayAffectControlFlow(inst, *mri_);
+}
+
+llvm::StringRef MiniDisassembler::GetInstructionName(
+    const llvm::MCInst &inst) const {
+  return mii_->getName(inst.getOpcode());
+}
+
+absl::StatusOr<bool> MiniDisassembler::MayAffectControlFlow(
+    uint64_t binary_address) {
+  auto inst = DisassembleOne(binary_address);
+  if (!inst.ok()) return inst.status();
+  return MayAffectControlFlow(inst.value());
+}
+}  // namespace propeller

propeller/mini_disassembler.h

@@ -0,0 +1,55 @@
+#ifndef PROPELLER_MINI_DISASSEMBLER_H_
+#define PROPELLER_MINI_DISASSEMBLER_H_
+
+#include <cstdint>
+#include <memory>
+
+#include "absl/base/nullability.h"
+#include "absl/status/statusor.h"
+#include "llvm/ADT/StringRef.h"
+#include "llvm/MC/MCAsmInfo.h"
+#include "llvm/MC/MCContext.h"
+#include "llvm/MC/MCDisassembler/MCDisassembler.h"
+#include "llvm/MC/MCInst.h"
+#include "llvm/MC/MCInstrAnalysis.h"
+#include "llvm/MC/MCInstrInfo.h"
+#include "llvm/MC/MCRegisterInfo.h"
+#include "llvm/MC/MCSubtargetInfo.h"
+#include "llvm/Object/ObjectFile.h"
+
+namespace propeller {
+class MiniDisassembler {
+ public:
+  // Creates a MiniDisassembler for `object_file`. Does not take ownership of
+  // `object_file`, which must point to a valid object that outlives the
+  // `MiniDisassembler`.
+  static absl::StatusOr<absl::Nonnull<std::unique_ptr<MiniDisassembler>>>
+  Create(const llvm::object::ObjectFile *object_file);
+
+  MiniDisassembler(const MiniDisassembler &) = delete;
+  MiniDisassembler(MiniDisassembler &&) = delete;
+
+  MiniDisassembler &operator=(const MiniDisassembler &) = delete;
+  MiniDisassembler &operator=(MiniDisassembler &&) = delete;
+
+  absl::StatusOr<llvm::MCInst> DisassembleOne(uint64_t binary_address);
+  bool MayAffectControlFlow(const llvm::MCInst &inst);
+  llvm::StringRef GetInstructionName(const llvm::MCInst &inst) const;
+  absl::StatusOr<bool> MayAffectControlFlow(uint64_t binary_address);
+
+ private:
+  explicit MiniDisassembler(const llvm::object::ObjectFile *object_file)
+      : object_file_(object_file) {}
+
+  const llvm::object::ObjectFile *object_file_;
+  std::unique_ptr<const llvm::MCRegisterInfo> mri_;
+  std::unique_ptr<const llvm::MCAsmInfo> asm_info_;
+  std::unique_ptr<const llvm::MCSubtargetInfo> sti_;
+  std::unique_ptr<const llvm::MCInstrInfo> mii_;
+  std::unique_ptr<llvm::MCContext> ctx_;
+  std::unique_ptr<const llvm::MCInstrAnalysis> mia_;
+  std::unique_ptr<const llvm::MCDisassembler> disasm_;
+};
+}  // namespace propeller
+
+#endif  // PROPELLER_MINI_DISASSEMBLER_H_

propeller/mini_disassembler_test.cc

@@ -0,0 +1,104 @@
+#include "propeller/mini_disassembler.h"
+
+#include <memory>
+#include <string>
+
+#include "propeller/status_testing_macros.h"
+#include "gmock/gmock.h"
+#include "gtest/gtest.h"
+#include "absl/status/status_matchers.h"
+#include "absl/strings/str_cat.h"
+#include "llvm/MC/MCInst.h"
+#include "llvm/lib/Target/X86/MCTargetDesc/X86MCTargetDesc.h"
+#include "propeller/binary_content.h"
+
+namespace propeller {
+namespace {
+using ::absl_testing::IsOk;
+using ::absl_testing::IsOkAndHolds;
+using ::testing::Not;
+
+TEST(MiniDisassemblerTest, DisassembleOne) {
+  const std::string binary =
+      absl::StrCat(::testing::SrcDir(),
+                   "_main/propeller/testdata/"
+                   "llvm_function_samples.binary");
+  ASSERT_OK_AND_ASSIGN(std::unique_ptr<BinaryContent> binary_content,
+                       GetBinaryContent(binary));
+  ASSERT_OK_AND_ASSIGN(
+      std::unique_ptr<MiniDisassembler> md,
+      MiniDisassembler::Create(binary_content->object_file.get()));
+  ASSERT_OK_AND_ASSIGN(llvm::MCInst inst, md->DisassembleOne(0x4008e4));
+  EXPECT_EQ(inst.getOpcode(), llvm::X86::RET64);
+}
+
+TEST(MiniDisassemblerTest, DisassembleOneFailure) {
+  const std::string binary =
+      absl::StrCat(::testing::SrcDir(),
+                   "_main/propeller/testdata/"
+                   "llvm_function_samples.binary");
+  ASSERT_OK_AND_ASSIGN(std::unique_ptr<BinaryContent> binary_content,
+                       GetBinaryContent(binary));
+  ASSERT_OK_AND_ASSIGN(
+      std::unique_ptr<MiniDisassembler> md,
+      MiniDisassembler::Create(binary_content->object_file.get()));
+  EXPECT_THAT(md->DisassembleOne(0x999999999), Not(IsOk()));
+}
+
+TEST(MiniDisassemblerTest, RetMayAffectControlFlow) {
+  const std::string binary =
+      absl::StrCat(::testing::SrcDir(),
+                   "_main/propeller/testdata/"
+                   "llvm_function_samples.binary");
+  ASSERT_OK_AND_ASSIGN(std::unique_ptr<BinaryContent> binary_content,
+                       GetBinaryContent(binary));
+  ASSERT_OK_AND_ASSIGN(
+      std::unique_ptr<MiniDisassembler> md,
+      MiniDisassembler::Create(binary_content->object_file.get()));
+  ASSERT_OK_AND_ASSIGN(llvm::MCInst ret_inst, md->DisassembleOne(0x4008e4));
+  EXPECT_TRUE(md->MayAffectControlFlow(ret_inst));
+}
+
+TEST(MiniDisassemblerTest, CallMayAffectControlFlow) {
+  const std::string binary =
+      absl::StrCat(::testing::SrcDir(),
+                   "_main/propeller/testdata/"
+                   "llvm_function_samples.binary");
+  ASSERT_OK_AND_ASSIGN(std::unique_ptr<BinaryContent> binary_content,
+                       GetBinaryContent(binary));
+  ASSERT_OK_AND_ASSIGN(
+      std::unique_ptr<MiniDisassembler> md,
+      MiniDisassembler::Create(binary_content->object_file.get()));
+  ASSERT_OK_AND_ASSIGN(llvm::MCInst call_inst, md->DisassembleOne(0x4008c9));
+  EXPECT_TRUE(md->MayAffectControlFlow(call_inst));
+}
+
+TEST(MiniDisassemblerTest, BranchMayAffectControlFlow) {
+  const std::string binary =
+      absl::StrCat(::testing::SrcDir(),
+                   "_main/propeller/testdata/"
+                   "llvm_function_samples.binary");
+  ASSERT_OK_AND_ASSIGN(std::unique_ptr<BinaryContent> binary_content,
+                       GetBinaryContent(binary));
+  ASSERT_OK_AND_ASSIGN(
+      std::unique_ptr<MiniDisassembler> md,
+      MiniDisassembler::Create(binary_content->object_file.get()));
+  EXPECT_THAT(md->MayAffectControlFlow(0x4008b6), IsOkAndHolds(true));
+}
+
+TEST(MiniDisassemblerTest, PushMayNotAffectControlFlow) {
+  const std::string binary =
+      absl::StrCat(::testing::SrcDir(),
+                   "_main/propeller/testdata/"
+                   "llvm_function_samples.binary");
+  ASSERT_OK_AND_ASSIGN(std::unique_ptr<BinaryContent> binary_content,
+                       GetBinaryContent(binary));
+  ASSERT_OK_AND_ASSIGN(
+      std::unique_ptr<MiniDisassembler> md,
+      MiniDisassembler::Create(binary_content->object_file.get()));
+  ASSERT_OK_AND_ASSIGN(llvm::MCInst push_inst, md->DisassembleOne(0x400590));
+  EXPECT_FALSE(md->MayAffectControlFlow(push_inst));
+}
+
+}  // namespace
+}  // namespace propeller

propeller/testdata/BUILD

@@ -122,7 +122,7 @@ genrule(
         "propeller_sample_1.perfdata2.gen",
     ],
     cmd = "$(CC) $(CC_FLAGS) -O0 -gsplit-dwarf=split -gmlt -Wl,-build-id -pie -fbasic-block-sections=labels $< -o $(RULEDIR)/propeller_sample_1.bin.gen && mv propeller_sample_1.dwo $(RULEDIR)/ && " +
-          "$(location //third_party/llvm/llvm-project/llvm:llvm-dwp) -e $(RULEDIR)/propeller_sample_1.bin.gen -o $(RULEDIR)/propeller_sample_1.dwp.gen $(RULEDIR)/propeller_sample_1.dwo && " +
+          "$(location //llvm:llvm-dwp) -e $(RULEDIR)/propeller_sample_1.bin.gen -o $(RULEDIR)/propeller_sample_1.dwp.gen $(RULEDIR)/propeller_sample_1.dwo && " +
           "cd $(RULEDIR) ; /usr/bin/perf record -o propeller_sample_1.perfdata1.gen -e cycles -b -- ./propeller_sample_1.bin.gen ;" +
           "/usr/bin/perf record -o propeller_sample_1.perfdata2.gen -e cycles -b -- ./propeller_sample_1.bin.gen 1 2 3 4",
     tags = [