[DFSan][NewPM] Port DataFlowSanitizer to NewPM

[ychen]

LGTM with two nits. Please wait one day or two in case other reviewers want to have a look.

llvm/include/llvm/Transforms/Instrumentation/DataFlowSanitizer.h

9

LLVM_TRANSFORMS_INSTRUMENTATION_DATAFLOWSANITIZER_H

12

Is Function.h needed?

[aeubanks]

Address review comments

[aeubanks]

llvm/include/llvm/Transforms/Instrumentation/DataFlowSanitizer.h

12

Nope, removed.

[morehouse]

llvm/lib/Passes/PassRegistry.def

92

Nit: maybe "dfsan-module" for consistency?

llvm/lib/Transforms/Instrumentation/DataFlowSanitizer.cpp

778

Do we need a bool to avoid calling init more than once?

[aeubanks]

llvm/lib/Passes/PassRegistry.def

92

Almost all of the others have both a module and function pass, so the "-module" is to distinguish between those. I think sancov-module is the only one that doesn't fit that, and I actually did want to rename it to sancov.

I'd like to keep the name the same as the legacy pass for migration reasons (basically so I don't have to touch every dfsan test).

llvm/lib/Transforms/Instrumentation/DataFlowSanitizer.cpp

778

A new DataFlowSanitizer is constructed for each run/runOnModule so that shouldn't be necessary.

[morehouse]

LGTM

clang/lib/CodeGen/BackendUtil.cpp

static void addThreadSanitizerPass(const PassManagerBuilder &Builder,
   PM.add(createThreadSanitizerLegacyPassPass());
 }
 
 static void addDataFlowSanitizerPass(const PassManagerBuilder &Builder,
                                      legacy::PassManagerBase &PM) {
   const PassManagerBuilderWrapper &BuilderWrapper =
       static_cast<const PassManagerBuilderWrapper&>(Builder);
   const LangOptions &LangOpts = BuilderWrapper.getLangOpts();
-  PM.add(createDataFlowSanitizerPass(LangOpts.SanitizerBlacklistFiles));
+  PM.add(
+      createDataFlowSanitizerLegacyPassPass(LangOpts.SanitizerBlacklistFiles));
 }
 
 static TargetLibraryInfoImpl *createTLII(llvm::Triple &TargetTriple,
                                          const CodeGenOptions &CodeGenOpts) {
   TargetLibraryInfoImpl *TLII = new TargetLibraryInfoImpl(TargetTriple);
 
   switch (CodeGenOpts.getVecLib()) {
   case CodeGenOptions::Accelerate:

llvm/bindings/go/llvm/InstrumentationBindings.cpp

 
 void LLVMAddDataFlowSanitizerPass(LLVMPassManagerRef PM,
                                   int ABIListFilesNum,
                                   const char **ABIListFiles) {
   std::vector<std::string> ABIListFilesVec;
   for (int i = 0; i != ABIListFilesNum; ++i) {
     ABIListFilesVec.push_back(ABIListFiles[i]);
   }
-  unwrap(PM)->add(createDataFlowSanitizerPass(ABIListFilesVec));
+  unwrap(PM)->add(createDataFlowSanitizerLegacyPassPass(ABIListFilesVec));
 }

llvm/include/llvm/InitializePasses.h

 void initializeControlHeightReductionLegacyPassPass(PassRegistry&);
 void initializeCorrelatedValuePropagationPass(PassRegistry&);
 void initializeCostModelAnalysisPass(PassRegistry&);
 void initializeCrossDSOCFIPass(PassRegistry&);
 void initializeDAEPass(PassRegistry&);
 void initializeDAHPass(PassRegistry&);
 void initializeDCELegacyPassPass(PassRegistry&);
 void initializeDSELegacyPassPass(PassRegistry&);
-void initializeDataFlowSanitizerPass(PassRegistry&);
+void initializeDataFlowSanitizerLegacyPassPass(PassRegistry &);
 void initializeDeadInstEliminationPass(PassRegistry&);
 void initializeDeadMachineInstructionElimPass(PassRegistry&);
 void initializeDebugifyMachineModulePass(PassRegistry &);
 void initializeDelinearizationPass(PassRegistry&);
 void initializeDemandedBitsWrapperPassPass(PassRegistry&);
 void initializeDependenceAnalysisPass(PassRegistry&);
 void initializeDependenceAnalysisWrapperPassPass(PassRegistry&);
 void initializeDetectDeadLanesPass(PassRegistry&);

llvm/include/llvm/Transforms/Instrumentation.h

 /// Insert frontend instrumentation based profiling. Parameter IsCS indicates if
 // this is the context senstive instrumentation.
 ModulePass *createInstrProfilingLegacyPass(
     const InstrProfOptions &Options = InstrProfOptions(), bool IsCS = false);
 
 ModulePass *createInstrOrderFilePass();
 
 // Insert DataFlowSanitizer (dynamic data flow analysis) instrumentation
-ModulePass *createDataFlowSanitizerPass(
+ModulePass *createDataFlowSanitizerLegacyPassPass(
     const std::vector<std::string> &ABIListFiles = std::vector<std::string>());
 
 // Options for sanitizer coverage instrumentation.
 struct SanitizerCoverageOptions {
   enum Type {
     SCK_None = 0,
     SCK_Function,
     SCK_BB,

llvm/include/llvm/Transforms/Instrumentation/DataFlowSanitizer.h

+//===- DataFlowSanitizer.h - dynamic data flow analysis -------------------===//
+//
+// 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
+//
+//===----------------------------------------------------------------------===//
+#ifndef LLVM_TRANSFORMS_INSTRUMENTATION_DATAFLOWSANITIZER_H
+#define LLVM_TRANSFORMS_INSTRUMENTATION_DATAFLOWSANITIZER_H
+
+#include "llvm/IR/Module.h"
+#include "llvm/IR/PassManager.h"
+#include <string>
+#include <vector>
+
+namespace llvm {
+
+class DataFlowSanitizerPass : public PassInfoMixin<DataFlowSanitizerPass> {
+private:
+  std::vector<std::string> ABIListFiles;
+
+public:
+  DataFlowSanitizerPass(
+      const std::vector<std::string> &ABIListFiles = std::vector<std::string>())
+      : ABIListFiles(ABIListFiles) {}
+  PreservedAnalyses run(Module &M, ModuleAnalysisManager &AM);
+};
+
+} // namespace llvm
+
+#endif

llvm/lib/Passes/PassBuilder.cpp

 #include "llvm/Transforms/IPO/SyntheticCountsPropagation.h"
 #include "llvm/Transforms/IPO/WholeProgramDevirt.h"
 #include "llvm/Transforms/InstCombine/InstCombine.h"
 #include "llvm/Transforms/Instrumentation.h"
 #include "llvm/Transforms/Instrumentation/AddressSanitizer.h"
 #include "llvm/Transforms/Instrumentation/BoundsChecking.h"
 #include "llvm/Transforms/Instrumentation/CGProfile.h"
 #include "llvm/Transforms/Instrumentation/ControlHeightReduction.h"
+#include "llvm/Transforms/Instrumentation/DataFlowSanitizer.h"
 #include "llvm/Transforms/Instrumentation/GCOVProfiler.h"
 #include "llvm/Transforms/Instrumentation/HWAddressSanitizer.h"
 #include "llvm/Transforms/Instrumentation/InstrOrderFile.h"
 #include "llvm/Transforms/Instrumentation/InstrProfiling.h"
 #include "llvm/Transforms/Instrumentation/MemorySanitizer.h"
 #include "llvm/Transforms/Instrumentation/PGOInstrumentation.h"
 #include "llvm/Transforms/Instrumentation/PoisonChecking.h"
 #include "llvm/Transforms/Instrumentation/SanitizerCoverage.h"

llvm/lib/Passes/PassRegistry.def

 MODULE_PASS("rpo-function-attrs", ReversePostOrderFunctionAttrsPass())
 MODULE_PASS("sample-profile", SampleProfileLoaderPass())
 MODULE_PASS("scc-oz-module-inliner",
   buildInlinerPipeline(OptimizationLevel::Oz, ThinLTOPhase::None, DebugLogging))
 MODULE_PASS("strip-dead-prototypes", StripDeadPrototypesPass())
 MODULE_PASS("synthetic-counts-propagation", SyntheticCountsPropagation())
 MODULE_PASS("wholeprogramdevirt", WholeProgramDevirtPass(nullptr, nullptr))
 MODULE_PASS("verify", VerifierPass())
+MODULE_PASS("dfsan", DataFlowSanitizerPass())
 MODULE_PASS("asan-module", ModuleAddressSanitizerPass(/*CompileKernel=*/false, false, true, false))
 MODULE_PASS("msan-module", MemorySanitizerPass({}))
 MODULE_PASS("tsan-module", ThreadSanitizerPass())
 MODULE_PASS("kasan-module", ModuleAddressSanitizerPass(/*CompileKernel=*/true, false, true, false))
 MODULE_PASS("sancov-module", ModuleSanitizerCoveragePass())
 MODULE_PASS("poison-checking", PoisonCheckingPass())
 #undef MODULE_PASS
 

llvm/lib/Transforms/Instrumentation/DataFlowSanitizer.cpp

 /// address into the shadow memory range.  See the function
 /// DataFlowSanitizer::getShadowAddress below.
 ///
 /// For more information, please refer to the design document:
 /// http://clang.llvm.org/docs/DataFlowSanitizerDesign.html
 //
 //===----------------------------------------------------------------------===//
 
+#include "llvm/Transforms/Instrumentation/DataFlowSanitizer.h"
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/DepthFirstIterator.h"
 #include "llvm/ADT/None.h"
 #include "llvm/ADT/SmallPtrSet.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringExtras.h"
 #include "llvm/ADT/StringRef.h"

 #include "llvm/IR/InstVisitor.h"
 #include "llvm/IR/InstrTypes.h"
 #include "llvm/IR/Instruction.h"
 #include "llvm/IR/Instructions.h"
 #include "llvm/IR/IntrinsicInst.h"
 #include "llvm/IR/LLVMContext.h"
 #include "llvm/IR/MDBuilder.h"
 #include "llvm/IR/Module.h"
+#include "llvm/IR/PassManager.h"
 #include "llvm/IR/Type.h"
 #include "llvm/IR/User.h"
 #include "llvm/IR/Value.h"
 #include "llvm/InitializePasses.h"
 #include "llvm/Pass.h"
 #include "llvm/Support/Casting.h"
 #include "llvm/Support/CommandLine.h"
 #include "llvm/Support/ErrorHandling.h"
AttributeList TransformFunctionAttributes(
 
   return AttributeList::get(
       Ctx,
       CallSiteAttrs.getFnAttributes(),
       CallSiteAttrs.getRetAttributes(),
       llvm::makeArrayRef(ArgumentAttributes));
 }
 
-class DataFlowSanitizer : public ModulePass {
+class DataFlowSanitizer {
   friend struct DFSanFunction;
   friend class DFSanVisitor;
 
   enum { ShadowWidthBits = 16, ShadowWidthBytes = ShadowWidthBits / 8 };
 
   /// Which ABI should be used for instrumented functions?
   enum InstrumentedABI {
     /// Argument and return value labels are passed through additional
class DataFlowSanitizer {
   void addGlobalNamePrefix(GlobalValue *GV);
   Function *buildWrapperFunction(Function *F, StringRef NewFName,
                                  GlobalValue::LinkageTypes NewFLink,
                                  FunctionType *NewFT);
   Constant *getOrBuildTrampolineFunction(FunctionType *FT, StringRef FName);
   void initializeCallbackFunctions(Module &M);
   void initializeRuntimeFunctions(Module &M);
 
-public:
-  static char ID;
+  bool init(Module &M);
 
-  DataFlowSanitizer(const std::vector<std::string> &ABIListFiles =
-                        std::vector<std::string>());
+public:
+  DataFlowSanitizer(const std::vector<std::string> &ABIListFiles);
 
-  bool doInitialization(Module &M) override;
-  bool runOnModule(Module &M) override;
+  bool runImpl(Module &M);
 };
 
 struct DFSanFunction {
   DataFlowSanitizer &DFS;
   Function *F;
   DominatorTree DT;
   DataFlowSanitizer::InstrumentedABI IA;
   bool IsNativeABI;
public:
   void visitAllocaInst(AllocaInst &I);
   void visitSelectInst(SelectInst &I);
   void visitMemSetInst(MemSetInst &I);
   void visitMemTransferInst(MemTransferInst &I);
 };
 
 } // end anonymous namespace
 
-char DataFlowSanitizer::ID;
-
-INITIALIZE_PASS(DataFlowSanitizer, "dfsan",
-                "DataFlowSanitizer: dynamic data flow analysis.", false, false)
-
-ModulePass *llvm::createDataFlowSanitizerPass(
-    const std::vector<std::string> &ABIListFiles) {
-  return new DataFlowSanitizer(ABIListFiles);
-}
-
 DataFlowSanitizer::DataFlowSanitizer(
-    const std::vector<std::string> &ABIListFiles)
-    : ModulePass(ID) {
+    const std::vector<std::string> &ABIListFiles) {
   std::vector<std::string> AllABIListFiles(std::move(ABIListFiles));
   AllABIListFiles.insert(AllABIListFiles.end(), ClABIListFiles.begin(),
                          ClABIListFiles.end());
   // FIXME: should we propagate vfs::FileSystem to this constructor?
   ABIList.set(
       SpecialCaseList::createOrDie(AllABIListFiles, *vfs::getRealFileSystem()));
 }
 
TransformedFunction DataFlowSanitizer::getCustomFunctionType(FunctionType *T) {
   Type *RetType = T->getReturnType();
   if (!RetType->isVoidTy())
     ArgTypes.push_back(ShadowPtrTy);
   return TransformedFunction(
       T, FunctionType::get(T->getReturnType(), ArgTypes, T->isVarArg()),
       ArgumentIndexMapping);
 }
 
-bool DataFlowSanitizer::doInitialization(Module &M) {
+bool DataFlowSanitizer::init(Module &M) {
   Triple TargetTriple(M.getTargetTriple());
   bool IsX86_64 = TargetTriple.getArch() == Triple::x86_64;
   bool IsMIPS64 = TargetTriple.isMIPS64();
   bool IsAArch64 = TargetTriple.getArch() == Triple::aarch64 ||
                    TargetTriple.getArch() == Triple::aarch64_be;
 
   const DataLayout &DL = M.getDataLayout();
 
void DataFlowSanitizer::initializeCallbackFunctions(Module &M) {
   DFSanStoreCallbackFn = Mod->getOrInsertFunction(
       "__dfsan_store_callback", DFSanLoadStoreCmpCallbackFnTy);
   DFSanMemTransferCallbackFn = Mod->getOrInsertFunction(
       "__dfsan_mem_transfer_callback", DFSanMemTransferCallbackFnTy);
   DFSanCmpCallbackFn = Mod->getOrInsertFunction("__dfsan_cmp_callback",
                                                 DFSanLoadStoreCmpCallbackFnTy);
 }
 
-bool DataFlowSanitizer::runOnModule(Module &M) {
+bool DataFlowSanitizer::runImpl(Module &M) {
+  init(M);
+
   if (ABIList.isIn(M, "skip"))
     return false;
 
   const unsigned InitialGlobalSize = M.global_size();
   const unsigned InitialModuleSize = M.size();
 
   bool Changed = false;
 
void DFSanVisitor::visitPHINode(PHINode &PN) {
   for (PHINode::block_iterator i = PN.block_begin(), e = PN.block_end(); i != e;
        ++i) {
     ShadowPN->addIncoming(UndefShadow, *i);
   }
 
   DFSF.PHIFixups.push_back(std::make_pair(&PN, ShadowPN));
   DFSF.setShadow(&PN, ShadowPN);
 }
+
+class DataFlowSanitizerLegacyPass : public ModulePass {
+private:
+  std::vector<std::string> ABIListFiles;
+
+public:
+  static char ID;
+
+  DataFlowSanitizerLegacyPass(
+      const std::vector<std::string> &ABIListFiles = std::vector<std::string>())
+      : ModulePass(ID), ABIListFiles(ABIListFiles) {}
+
+  bool runOnModule(Module &M) override {
+    return DataFlowSanitizer(ABIListFiles).runImpl(M);
+  }
+};
+
+char DataFlowSanitizerLegacyPass::ID;
+
+INITIALIZE_PASS(DataFlowSanitizerLegacyPass, "dfsan",
+                "DataFlowSanitizer: dynamic data flow analysis.", false, false)
+
+ModulePass *llvm::createDataFlowSanitizerLegacyPassPass(
+    const std::vector<std::string> &ABIListFiles) {
+  return new DataFlowSanitizerLegacyPass(ABIListFiles);
+}
+
+PreservedAnalyses DataFlowSanitizerPass::run(Module &M,
+                                             ModuleAnalysisManager &AM) {
+  if (DataFlowSanitizer(ABIListFiles).runImpl(M)) {
+    return PreservedAnalyses::none();
+  }
+  return PreservedAnalyses::all();
+}

llvm/lib/Transforms/Instrumentation/Instrumentation.cpp

void llvm::initializeInstrumentation(PassRegistry &Registry) {
   initializePGOMemOPSizeOptLegacyPassPass(Registry);
   initializeCGProfileLegacyPassPass(Registry);
   initializeInstrOrderFileLegacyPassPass(Registry);
   initializeInstrProfilingLegacyPassPass(Registry);
   initializeMemorySanitizerLegacyPassPass(Registry);
   initializeHWAddressSanitizerLegacyPassPass(Registry);
   initializeThreadSanitizerLegacyPassPass(Registry);
   initializeModuleSanitizerCoverageLegacyPassPass(Registry);
-  initializeDataFlowSanitizerPass(Registry);
+  initializeDataFlowSanitizerLegacyPassPass(Registry);
 }
 
 /// LLVMInitializeInstrumentation - C binding for
 /// initializeInstrumentation.
 void LLVMInitializeInstrumentation(LLVMPassRegistryRef R) {
   initializeInstrumentation(*unwrap(R));
 }

llvm/test/Instrumentation/DataFlowSanitizer/call.ll

 ; RUN: opt < %s -dfsan -S | FileCheck %s
+; RUN: opt < %s -passes=dfsan -S | FileCheck %s
 target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 ; CHECK-LABEL: @__dfsan_arg_tls
 ; CHECK: = external thread_local(initialexec) global [64 x i16]
 
 ; CHECK-LABEL: @__dfsan_retval_tls
 ; CHECK: = external thread_local(initialexec) global i16