[PATCH] D98268: [dfsan] Tracking origins at phi nodes

[Matt Morehouse via Phabricator via llvm-commits]

morehouse added a comment.

Can we add an end-to-end test as well?



================
Comment at: llvm/lib/Transforms/Instrumentation/DataFlowSanitizer.cpp:1572-1575
+    for (std::vector<DFSanFunction::PHIFixupElement>::iterator
+             I = DFSF.PHIFixups.begin(),
+             E = DFSF.PHIFixups.end();
+         I != E; ++I) {
----------------
Nit:  range for loop is cleaner if we don't need to advance iterators ourselves.


================
Comment at: llvm/lib/Transforms/Instrumentation/DataFlowSanitizer.cpp:2716
+  }
+
   Value *RawDestShadow = DFSF.DFS.getShadowAddress(I.getDest(), &I);
----------------
This is unrelated and shouldn't be in this diff...


================
Comment at: llvm/lib/Transforms/Instrumentation/DataFlowSanitizer.cpp:3144-3145
+    Value *UndefOrigin = UndefValue::get(DFSF.DFS.OriginTy);
+    for (PHINode::block_iterator i = PN.block_begin(), e = PN.block_end();
+         i != e; ++i) {
+      OriginPN->addIncoming(UndefOrigin, *i);
----------------
Nit:  range for loop is cleaner


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D98268/new/

https://reviews.llvm.org/D98268