Replace `getNext` with `nextNonMarker`

Most of the times, when we need to get the next instruction, we actually
want to skip over "marker" function calls (e.g., calls to `newpc` and
`function_call`). `nextNonMarker` does exactly this.

`FunctionCallIdentification::isCall` and `JumpTargetManager::setCFGForm`
have also been extended to correctly handle such situations.

functionboundariesdetection.cpp

@@ -218,7 +218,7 @@ void FBD::collectFunctionCalls() {
     if (auto *Call = dyn_cast<CallInst>(U)) {
       BasicBlock *ReturnBB = getBlock(Call->getOperand(1));
       uint32_t ReturnPC = getLimitedValue(Call->getOperand(2));
-      auto *Terminator = cast<TerminatorInst>(getNext(Call));
+      auto *Terminator = cast<TerminatorInst>(nextNonMarker(Call));
       assert(Terminator != nullptr);
       FunctionCalls[Terminator] = ReturnBB;
       CallPredecessors[ReturnBB].push_back(Call->getParent());

functioncallidentification.h

@@ -44,11 +44,13 @@ public:
   bool isCall(llvm::TerminatorInst *T) const {
     assert(T != nullptr);
     llvm::Instruction *Previous = getPrevious(T);
-    if (Previous == nullptr)
-      return false;
+    while (Previous != nullptr && isMarker(Previous)) {
+      auto *Call = llvm::cast<llvm::CallInst>(Previous);
+      if (Call->getCalledFunction() == FunctionCall)
+        return true;
 
-    if (auto *Call = llvm::dyn_cast<llvm::CallInst>(Previous))
-      return Call->getCalledFunction() == FunctionCall;
+      Previous = getPrevious(Previous);
+    }
 
     return false;
   }

jumptargetmanager.cpp

@@ -1237,7 +1237,7 @@ void JumpTargetManager::setCFGForm(CFGForm NewForm) {
     if (auto *FunctionCall = TheModule.getFunction("function_call")) {
       for (User *U : FunctionCall->users()) {
         auto *Call = cast<CallInst>(U);
-        auto *Terminator = cast<TerminatorInst>(Call->getNextNode());
+        auto *Terminator = cast<TerminatorInst>(nextNonMarker(Call));
         assert(Terminator->getNumSuccessors() == 1);
 
         // Get the correct argument, the first is the callee, the second the

revamb.h

@@ -16,6 +16,8 @@
 #include "llvm/ADT/StringRef.h"
 #include "llvm/ADT/Triple.h"
 
+// Local includes
+#include "ir-helpers.h"
 namespace llvm {
 class GlobalVariable;
 };
@@ -139,4 +141,30 @@ static inline bool contains(T Range, typename T::value_type V) {
   return std::find(std::begin(Range), std::end(Range), V) != std::end(Range);
 }
 
+static const std::array<llvm::StringRef, 3> MarkerFunctionNames = {
+  "newpc",
+  "function_call",
+  "exitTB"
+};
+
+static inline bool isMarker(llvm::Instruction *I) {
+  using namespace std::placeholders;
+  using llvm::any_of;
+  using std::bind;
+
+  return any_of(MarkerFunctionNames, bind(isCallTo, I, _1));
+}
+
+static inline llvm::Instruction *nextNonMarker(llvm::Instruction *I) {
+  auto It = I->getIterator();
+  auto End = I->getParent()->end();
+  do {
+    It++;
+    assert(It != End);
+  } while (isMarker(&*It));
+
+  assert(It != End);
+  return &*It;
+}
+
 #endif // _REVAMB_H