[llvm] r298688 - [ValueTracking] Use uint64_t for CarryIn in computeKnownBitsAddSub instead of a creating a temporary APInt. NFC

[Craig Topper via llvm-commits]

Author: ctopper
Date: Fri Mar 24 00:38:09 2017
New Revision: 298688

URL: http://llvm.org/viewvc/llvm-project?rev=298688&view=rev
Log:
[ValueTracking] Use uint64_t for CarryIn in computeKnownBitsAddSub instead of a creating a temporary APInt. NFC

Modified:
    llvm/trunk/lib/Analysis/ValueTracking.cpp

Modified: llvm/trunk/lib/Analysis/ValueTracking.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Analysis/ValueTracking.cpp?rev=298688&r1=298687&r2=298688&view=diff
==============================================================================
--- llvm/trunk/lib/Analysis/ValueTracking.cpp (original)
+++ llvm/trunk/lib/Analysis/ValueTracking.cpp Fri Mar 24 00:38:09 2017
@@ -264,11 +264,11 @@ static void computeKnownBitsAddSub(bool
   computeKnownBits(Op1, KnownZero2, KnownOne2, Depth + 1, Q);
 
   // Carry in a 1 for a subtract, rather than a 0.
-  APInt CarryIn(BitWidth, 0);
+  uint64_t CarryIn = 0;
   if (!Add) {
     // Sum = LHS + ~RHS + 1
     std::swap(KnownZero2, KnownOne2);
-    CarryIn.setBit(0);
+    CarryIn = 1;
   }
 
   APInt PossibleSumZero = ~LHSKnownZero + ~KnownZero2 + CarryIn;