[SandboxIR] Add callbacks for instruction insert/remove/move ops

llvm/include/llvm/SandboxIR/Context.h

@@ -9,18 +9,31 @@
 #ifndef LLVM_SANDBOXIR_CONTEXT_H
 #define LLVM_SANDBOXIR_CONTEXT_H
 
+#include "llvm/ADT/DenseMap.h"
+#include "llvm/ADT/SmallVector.h"
 #include "llvm/IR/LLVMContext.h"
 #include "llvm/SandboxIR/Tracker.h"
 #include "llvm/SandboxIR/Type.h"
 
 namespace llvm::sandboxir {
 
-class Module;
-class Value;
 class Argument;
+class BBIterator;
 class Constant;
+class Module;
+class Value;
 
 class Context {
+public:
+  // A RemoveInstrCallback receives the instruction about to be removed.
+  using RemoveInstrCallback = std::function<void(Instruction *)>;
+  // A InsertInstrCallback receives the instruction about to be created.
+  using InsertInstrCallback = std::function<void(Instruction *)>;
+  // A MoveInstrCallback receives the instruction about to be moved, the
+  // destination BB and an iterator pointing to the insertion position.
+  using MoveInstrCallback =
+      std::function<void(Instruction *, const BBIterator &)>;
+
 protected:
   LLVMContext &LLVMCtx;
   friend class Type;        // For LLVMCtx.
@@ -48,6 +61,21 @@ class Context {
   /// Type objects.
   DenseMap<llvm::Type *, std::unique_ptr<Type, TypeDeleter>> LLVMTypeToTypeMap;
 
+  /// Callbacks called when an IR instruction is about to get removed. Keys are
+  /// used as IDs for deregistration.
+  DenseMap<int, RemoveInstrCallback> RemoveInstrCallbacks;
+  /// Callbacks called when an IR instruction is about to get inserted. Keys are
+  /// used as IDs for deregistration.
+  DenseMap<int, InsertInstrCallback> InsertInstrCallbacks;
+  /// Callbacks called when an IR instruction is about to get moved. Keys are
+  /// used as IDs for deregistration.
+  DenseMap<int, MoveInstrCallback> MoveInstrCallbacks;
+
+  /// A counter used for assigning callback IDs during registration. The same
+  /// counter is used for all kinds of callbacks so we can detect mismatched
+  /// registration/deregistration.
+  static int NextCallbackId;
+
   /// Remove \p V from the maps and returns the unique_ptr.
   std::unique_ptr<Value> detachLLVMValue(llvm::Value *V);
   /// Remove \p SBV from all SandboxIR maps and stop owning it. This effectively
@@ -70,6 +98,10 @@ class Context {
   Constant *getOrCreateConstant(llvm::Constant *LLVMC);
   friend class Utils; // For getMemoryBase
 
+  void runRemoveInstrCallbacks(Instruction *I);
+  void runInsertInstrCallbacks(Instruction *I);
+  void runMoveInstrCallbacks(Instruction *I, const BBIterator &Where);
+
   // Friends for getOrCreateConstant().
 #define DEF_CONST(ID, CLASS) friend class CLASS;
 #include "llvm/SandboxIR/Values.def"
@@ -198,6 +230,26 @@ class Context {
 
   /// \Returns the number of values registered with Context.
   size_t getNumValues() const { return LLVMValueToValueMap.size(); }
+
+  /// Register a callback that gets called when a SandboxIR instruction is about
+  /// to be removed from its parent. Note that this will also be called when
+  /// reverting the creation of an instruction.
+  /// \Returns a callback ID for later deregistration.
+  int registerRemoveInstrCallback(RemoveInstrCallback CB);
+  void unregisterRemoveInstrCallback(int CallbackId);
+
+  /// Register a callback that gets called right after a SandboxIR instruction
+  /// is created. Note that this will also be called when reverting the removal
+  /// of an instruction.
+  /// \Returns a callback ID for later deregistration.
+  int registerInsertInstrCallback(InsertInstrCallback CB);
+  void unregisterInsertInstrCallback(int CallbackId);
+
+  /// Register a callback that gets called when a SandboxIR instruction is about
+  /// to be moved. Note that this will also be called when reverting a move.
+  /// \Returns a callback ID for later deregistration.
+  int registerMoveInstrCallback(MoveInstrCallback CB);
+  void unregisterMoveInstrCallback(int CallbackId);
 };
 
 } // namespace llvm::sandboxir

llvm/lib/SandboxIR/Context.cpp

@@ -35,17 +35,20 @@ Value *Context::registerValue(std::unique_ptr<Value> &&VPtr) {
   assert(VPtr->getSubclassID() != Value::ClassID::User &&
          "Can't register a user!");
 
+  Value *V = VPtr.get();
+  [[maybe_unused]] auto Pair =
+      LLVMValueToValueMap.insert({VPtr->Val, std::move(VPtr)});
+  assert(Pair.second && "Already exists!");
+
   // Track creation of instructions.
   // Please note that we don't allow the creation of detached instructions,
   // meaning that the instructions need to be inserted into a block upon
   // creation. This is why the tracker class combines creation and insertion.
-  if (auto *I = dyn_cast<Instruction>(VPtr.get()))
+  if (auto *I = dyn_cast<Instruction>(V)) {
     getTracker().emplaceIfTracking<CreateAndInsertInst>(I);
+    runInsertInstrCallbacks(I);
+  }
 
-  Value *V = VPtr.get();
-  [[maybe_unused]] auto Pair =
-      LLVMValueToValueMap.insert({VPtr->Val, std::move(VPtr)});
-  assert(Pair.second && "Already exists!");
   return V;
 }
 
@@ -660,4 +663,57 @@ Module *Context::createModule(llvm::Module *LLVMM) {
   return M;
 }
 
+void Context::runRemoveInstrCallbacks(Instruction *I) {
+  for (const auto &CBEntry : RemoveInstrCallbacks) {
+    CBEntry.second(I);
+  }
+}
+
+void Context::runInsertInstrCallbacks(Instruction *I) {
+  for (auto &CBEntry : InsertInstrCallbacks) {
+    CBEntry.second(I);
+  }
+}
+
+void Context::runMoveInstrCallbacks(Instruction *I, const BBIterator &WhereIt) {
+  for (auto &CBEntry : MoveInstrCallbacks) {
+    CBEntry.second(I, WhereIt);
+  }
+}
+
+int Context::NextCallbackId = 0;
+
+int Context::registerRemoveInstrCallback(RemoveInstrCallback CB) {
+  int Id = NextCallbackId++;
+  RemoveInstrCallbacks[Id] = CB;
+  return Id;
+}
+void Context::unregisterRemoveInstrCallback(int CallbackId) {
+  [[maybe_unused]] bool erased = RemoveInstrCallbacks.erase(CallbackId);
+  assert(erased &&
+         "Callback id not found in RemoveInstrCallbacks during deregistration");
+}
+
+int Context::registerInsertInstrCallback(InsertInstrCallback CB) {
+  int Id = NextCallbackId++;
+  InsertInstrCallbacks[Id] = CB;
+  return Id;
+}
+void Context::unregisterInsertInstrCallback(int CallbackId) {
+  [[maybe_unused]] bool erased = InsertInstrCallbacks.erase(CallbackId);
+  assert(erased &&
+         "Callback id not found in InsertInstrCallbacks during deregistration");
+}
+
+int Context::registerMoveInstrCallback(MoveInstrCallback CB) {
+  int Id = NextCallbackId++;
+  MoveInstrCallbacks[Id] = CB;
+  return Id;
+}
+void Context::unregisterMoveInstrCallback(int CallbackId) {
+  [[maybe_unused]] bool erased = MoveInstrCallbacks.erase(CallbackId);
+  assert(erased &&
+         "Callback id not found in MoveInstrCallbacks during deregistration");
+}
+
 } // namespace llvm::sandboxir

llvm/lib/SandboxIR/Instruction.cpp

@@ -64,6 +64,8 @@ Instruction *Instruction::getPrevNode() const {
 }
 
 void Instruction::removeFromParent() {
+  Ctx.runRemoveInstrCallbacks(this);
+
   Ctx.getTracker().emplaceIfTracking<RemoveFromParent>(this);
 
   // Detach all the LLVM IR instructions from their parent BB.
@@ -73,6 +75,8 @@ void Instruction::removeFromParent() {
 
 void Instruction::eraseFromParent() {
   assert(users().empty() && "Still connected to users, can't erase!");
+
+  Ctx.runRemoveInstrCallbacks(this);
   std::unique_ptr<Value> Detached = Ctx.detach(this);
   auto LLVMInstrs = getLLVMInstrs();
 
@@ -100,6 +104,7 @@ void Instruction::moveBefore(BasicBlock &BB, const BBIterator &WhereIt) {
     // Destination is same as origin, nothing to do.
     return;
 
+  Ctx.runMoveInstrCallbacks(this, WhereIt);
   Ctx.getTracker().emplaceIfTracking<MoveInstr>(this);
 
   auto *LLVMBB = cast<llvm::BasicBlock>(BB.Val);

llvm/unittests/SandboxIR/SandboxIRTest.cpp

@@ -22,6 +22,7 @@
 #include "llvm/SandboxIR/Value.h"
 #include "llvm/Support/SourceMgr.h"
 #include "gmock/gmock-matchers.h"
+#include "gmock/gmock-more-matchers.h"
 #include "gtest/gtest.h"
 
 using namespace llvm;
@@ -5962,3 +5963,87 @@ TEST_F(SandboxIRTest, CheckClassof) {
   EXPECT_NE(&sandboxir::CLASS::classof, &sandboxir::Instruction::classof);
 #include "llvm/SandboxIR/Values.def"
 }
+
+TEST_F(SandboxIRTest, InstructionCallbacks) {
+  parseIR(C, R"IR(
+    define void @foo(ptr %ptr, i8 %val) {
+      ret void
+    }
+  )IR");
+  Function &LLVMF = *M->getFunction("foo");
+  sandboxir::Context Ctx(C);
+
+  auto &F = *Ctx.createFunction(&LLVMF);
+  auto &BB = *F.begin();
+  sandboxir::Argument *Ptr = F.getArg(0);
+  sandboxir::Argument *Val = F.getArg(1);
+  sandboxir::Instruction *Ret = &BB.front();
+
+  SmallVector<sandboxir::Instruction *> Inserted;
+  int InsertCbId = Ctx.registerInsertInstrCallback(
+      [&Inserted](sandboxir::Instruction *I) { Inserted.push_back(I); });
+
+  SmallVector<sandboxir::Instruction *> Removed;
+  int RemoveCbId = Ctx.registerRemoveInstrCallback(
+      [&Removed](sandboxir::Instruction *I) { Removed.push_back(I); });
+
+  // Keep the moved instruction and the instruction pointed by the Where
+  // iterator so we can check both callback arguments work as expected.
+  SmallVector<std::pair<sandboxir::Instruction *, sandboxir::Instruction *>>
+      Moved;
+  int MoveCbId = Ctx.registerMoveInstrCallback(
+      [&Moved](sandboxir::Instruction *I, const sandboxir::BBIterator &Where) {
+        // Use a nullptr to signal "move to end" to keep it single. We only
+        // have a basic block in this test case anyway.
+        if (Where == Where.getNodeParent()->end())
+          Moved.push_back(std::make_pair(I, nullptr));
+        else
+          Moved.push_back(std::make_pair(I, &*Where));
+      });
+
+  Ctx.save();
+  auto *NewI = sandboxir::StoreInst::create(Val, Ptr, /*Align=*/std::nullopt,
+                                            Ret->getIterator(), Ctx);
+  EXPECT_THAT(Inserted, testing::ElementsAre(NewI));
+  EXPECT_THAT(Removed, testing::IsEmpty());
+  EXPECT_THAT(Moved, testing::IsEmpty());
+
+  Ret->moveBefore(NewI);
+  EXPECT_THAT(Inserted, testing::ElementsAre(NewI));
+  EXPECT_THAT(Removed, testing::IsEmpty());
+  EXPECT_THAT(Moved, testing::ElementsAre(std::make_pair(Ret, NewI)));
+
+  Ret->eraseFromParent();
+  EXPECT_THAT(Inserted, testing::ElementsAre(NewI));
+  EXPECT_THAT(Removed, testing::ElementsAre(Ret));
+  EXPECT_THAT(Moved, testing::ElementsAre(std::make_pair(Ret, NewI)));
+
+  NewI->eraseFromParent();
+  EXPECT_THAT(Inserted, testing::ElementsAre(NewI));
+  EXPECT_THAT(Removed, testing::ElementsAre(Ret, NewI));
+  EXPECT_THAT(Moved, testing::ElementsAre(std::make_pair(Ret, NewI)));
+
+  // Check that after revert the callbacks have been called for the inverse
+  // operations of the changes made so far.
+  Ctx.revert();
+  EXPECT_THAT(Inserted, testing::ElementsAre(NewI, NewI, Ret));
+  EXPECT_THAT(Removed, testing::ElementsAre(Ret, NewI, NewI));
+  EXPECT_THAT(Moved, testing::ElementsAre(std::make_pair(Ret, NewI),
+                                          std::make_pair(Ret, nullptr)));
+
+  // Check that deregistration works. Do an operation of each type after
+  // deregistering callbacks and check.
+  Inserted.clear();
+  Removed.clear();
+  Moved.clear();
+  Ctx.unregisterInsertInstrCallback(InsertCbId);
+  Ctx.unregisterRemoveInstrCallback(RemoveCbId);
+  Ctx.unregisterMoveInstrCallback(MoveCbId);
+  auto *NewI2 = sandboxir::StoreInst::create(Val, Ptr, /*Align=*/std::nullopt,
+                                            Ret->getIterator(), Ctx);
+  Ret->moveBefore(NewI2);
+  Ret->eraseFromParent();
+  EXPECT_THAT(Inserted, testing::IsEmpty());
+  EXPECT_THAT(Removed, testing::IsEmpty());
+  EXPECT_THAT(Moved, testing::IsEmpty());
+}