[PATCH] D20143: New pass: guard widening

[David Majnemer via llvm-commits]

majnemer added a subscriber: majnemer.

================
Comment at: lib/Transforms/Scalar/GuardWidening.cpp:81-95
@@ +80,17 @@
+  /// Used to keep track of which widening potential is more effective.
+  enum WideningScore {
+    /// Don't widen.
+    WS_IllegalOrNegative,
+
+    /// Widening is performance neutral as far as the cycles spent in check
+    /// conditions goes (but can still help, e.g., code layout, having less
+    /// deopt state).
+    WS_Neutral,
+
+    /// Widening is profitable.
+    WS_Positive,
+
+    /// Widening is very profitable.  Not significantly different from \c
+    /// WS_Positive, except by the order.
+    WS_VeryPositive
+  };
----------------
Your enum is defined in a class, I think you can forgo the WS prefix if you'd like.

================
Comment at: lib/Transforms/Scalar/GuardWidening.cpp:185-186
@@ +184,4 @@
+    for (auto &I : *BB)
+      if (auto *II = dyn_cast<IntrinsicInst>(&I))
+        if (II->getIntrinsicID() == Intrinsic::experimental_guard)
+          CurrentList.push_back(II);
----------------
`match(&I, m_Intrinsic<Intrinsic::experimental_guard>())` ?

================
Comment at: lib/Transforms/Scalar/GuardWidening.cpp:357-358
@@ +356,4 @@
+
+    IRBuilder<> B(InsertPt);
+    Result = B.CreateAnd(Cond0, Cond1, "wide.chk");
+  }
----------------
Is this saving anything over `BinaryOperator::CreateAnd(Cond0, Cond1, "wide.chk", InsertPt);`?


http://reviews.llvm.org/D20143