* add bench/benchVecAdd.cpp by Gael, fix crash (ei_pload on non-aligned)

* introduce packet(int), make use of it in linear vectorized paths
  --> completely fixes the slowdown noticed in benchVecAdd.
* generalize coeff(int) to linear-access xprs
* clarify the access flag bits
* rework api dox in Coeffs.h and util/Constants.h
* improve certain expressions's flags, allowing more vectorization
* fix bug in Block: start(int) and end(int) returned dyn*dyn size
* fix bug in Block: just because the Eval type has packet access
  doesn't imply the block xpr should have it too.

1 files changed, 12 insertions, 1 deletions

diff --git a/Eigen/src/Core/CwiseBinaryOp.h b/Eigen/src/Core/CwiseBinaryOp.h
index ec4619781..6672edcbe 100644
--- a/Eigen/src/Core/CwiseBinaryOp.h
+++ b/Eigen/src/Core/CwiseBinaryOp.h
@@ -69,7 +69,7 @@ struct ei_traits<CwiseBinaryOp<BinaryOp, Lhs, Rhs> >
         HereditaryBits
       | (int(LhsFlags) & int(RhsFlags) & LinearAccessBit)
       | (ei_functor_traits<BinaryOp>::PacketAccess && ((int(LhsFlags) & RowMajorBit)==(int(RhsFlags) & RowMajorBit))
-        ? int(LhsFlags) & int(RhsFlags) & PacketAccessBit : 0)),
+        ? (int(LhsFlags) & int(RhsFlags) & PacketAccessBit) : 0)),
     CoeffReadCost = LhsCoeffReadCost + RhsCoeffReadCost + ei_functor_traits<BinaryOp>::Cost
   };
 };
@@ -108,6 +108,17 @@ class CwiseBinaryOp : ei_no_assignment_operator,
       return m_functor.packetOp(m_lhs.template packet<LoadMode>(row, col), m_rhs.template packet<LoadMode>(row, col));
     }
 
+    inline const Scalar _coeff(int index) const
+    {
+      return m_functor(m_lhs.coeff(index), m_rhs.coeff(index));
+    }
+
+    template<int LoadMode>
+    inline PacketScalar _packet(int index) const
+    {
+      return m_functor.packetOp(m_lhs.template packet<LoadMode>(index), m_rhs.template packet<LoadMode>(index));
+    }
+
   protected:
     const LhsNested m_lhs;
     const RhsNested m_rhs;