disabled TensorView for PlusNode::BackpropTo() and SigmoidNode, now gives the same for Image/QuickE2E

Source/Common/Include/DataTensor.h

@@ -452,8 +452,8 @@ namespace Microsoft { namespace MSR { namespace CNTK {
     // Eventually this can go away once we switch completely to cudnn layout.
     enum ImageLayoutKind
     {
-        CHW,    // cudnn; default for BrainScript
-        HWC     // legacy; default for NDL
+        HWC,    // legacy; default for NDL
+        CHW     // cudnn; default for BrainScript
     };
     static inline ImageLayoutKind ImageLayoutKindFrom(const wstring & s)
     {

Source/ComputationNetworkLib/ConvolutionalNodes.h

@@ -177,7 +177,7 @@ namespace Microsoft { namespace MSR { namespace CNTK {
 #endif
         }
 
-        // BUGBUG: Should not be here. Use PlusNode and m_sampleLayout.
+        // BUGBUG: Should not be here. Use PlusNode and m_sampleLayout.  TODO: Bad naming:'output' is actually an 'input'
         void AddBias(const Matrix<ElemType>& output, const Matrix<ElemType>& bias, Matrix<ElemType>& dst)
         {
             assert(m_convEng != nullptr);

Source/ComputationNetworkLib/LinearAlgebraNodes.h

@@ -43,7 +43,7 @@ namespace Microsoft { namespace MSR { namespace CNTK {
 
         virtual void /*ComputationNode::*/BackpropTo(const size_t inputIndex, const FrameRange & fr) override
         {
-#ifdef ENABLE_TENSORVIEW
+#if 0//def ENABLE_TENSORVIEW
             // BUGBUG: This gives us a huge perf hit for Image/QuickE2E.
             size_t rank = DetermineElementwiseTensorRank();
             auto gradient      =                    GradientTensorFor(rank, fr);

Source/ComputationNetworkLib/NonlinearityNodes.h

@@ -155,7 +155,7 @@ namespace Microsoft { namespace MSR { namespace CNTK {
     // SigmoidNode (input) -- sigmoid non-linearity
     // -----------------------------------------------------------------------
 
-#ifdef ENABLE_TENSORVIEW
+#if 0//def ENABLE_TENSORVIEW
     template<class ElemType>
     class SigmoidNode : public UnaryElementWiseNode<ElemType>
     {

Source/Math/ConvolutionEngine.cpp

@@ -239,7 +239,7 @@ namespace Microsoft { namespace MSR { namespace CNTK {
             assert(outT.w() * outT.h() * outT.c() == out.GetNumRows());
             assert(outT.n() == out.GetNumCols());
 
-            Mat o = out.ColumnSlice(0, out.GetNumCols());
+            Mat o = out.ColumnSlice(0, out.GetNumCols());   // same as .AsReference()
             Mat d = dst.Reshaped(biasT.c(), outT.w() * outT.h() * outT.n());
             d.AssignSumOf(o.Reshaped(biasT.c(), outT.w() * outT.h() * outT.n()), bias);
         }