From 900c7c61bb6abca5b3324c11ba1b45fa3e31c5fa Mon Sep 17 00:00:00 2001
From: Eugene Zhulenev <ezhulenev@google.com>
Date: Mon, 15 Oct 2018 16:52:33 -0700
Subject: Check if it's allowed to squueze inner dimensions in TensorBlockIO

---
 unsupported/Eigen/CXX11/src/Tensor/TensorBlock.h | 36 +++++++++++++++++++++---
 1 file changed, 32 insertions(+), 4 deletions(-)

(limited to 'unsupported/Eigen/CXX11/src/Tensor/TensorBlock.h')

diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorBlock.h b/unsupported/Eigen/CXX11/src/Tensor/TensorBlock.h
index a59a5d5b2..91c77b05a 100644
--- a/unsupported/Eigen/CXX11/src/Tensor/TensorBlock.h
+++ b/unsupported/Eigen/CXX11/src/Tensor/TensorBlock.h
@@ -246,6 +246,8 @@ class TensorBlockIO {
   typedef TensorBlockCopyOp<Scalar, StorageIndex> BlockCopyOp;
 
  protected:
+  typedef array<StorageIndex, NumDims> Dimensions;
+
   struct BlockIteratorState {
     StorageIndex input_stride;
     StorageIndex output_stride;
@@ -262,22 +264,46 @@ class TensorBlockIO {
           count(0) {}
   };
 
+  // Compute how many inner dimensions it's allowed to squeeze when doing IO
+  // between a tensor and a block. It's safe to squeeze inner dimensions, only
+  // if they are not reordered.
+  static int NumSqueezableInnerDims(const Dimensions& tensor_to_block_dim_map) {
+    int num_squeezable_dims = 0;
+    if (Layout == ColMajor) {
+      for (int i = 0; i < NumDims; ++i) {
+        if (tensor_to_block_dim_map[i] == i) num_squeezable_dims++;
+        else break;
+      }
+    } else {
+      for (int i = NumDims - 1; i >= 0; --i) {
+        if (tensor_to_block_dim_map[i] == i) num_squeezable_dims++;
+        else break;
+      }
+    }
+    return num_squeezable_dims;
+  }
+
   static EIGEN_DEVICE_FUNC EIGEN_STRONG_INLINE void Copy(
       const Block& block, StorageIndex first_coeff_index,
-      const array<StorageIndex, NumDims>& tensor_to_block_dim_map,
-      const array<StorageIndex, NumDims>& tensor_strides, const Scalar* src_data,
+      const Dimensions& tensor_to_block_dim_map,
+      const Dimensions& tensor_strides,
+      const Scalar* src_data,
       Scalar* dst_data) {
+    // Do not squeeze reordered inner dimensions.
+    int num_squeezable_dims = NumSqueezableInnerDims(tensor_to_block_dim_map);
+
     // Find the innermost tensor dimension whose size is not 1. This is the
     // effective inner dim. If all dimensions are of size 1, then fallback to
     // using the actual innermost dim to avoid out-of-bound access.
     StorageIndex num_size_one_inner_dims = 0;
-    for (int i = 0; i < NumDims; ++i) {
+    for (int i = 0; i < num_squeezable_dims; ++i) {
       const int dim = cond<Layout>()(i, NumDims - i - 1);
       if (block.block_sizes()[tensor_to_block_dim_map[dim]] != 1) {
         num_size_one_inner_dims = i;
         break;
       }
     }
+
     // Calculate strides and dimensions.
     const StorageIndex tensor_stride1_dim = cond<Layout>()(
         num_size_one_inner_dims, NumDims - num_size_one_inner_dims - 1);
@@ -286,7 +312,9 @@ class TensorBlockIO {
     StorageIndex block_inner_dim_size =
         NumDims == 0 ? 1
                      : block.block_sizes()[block_dim_for_tensor_stride1_dim];
-    for (Index i = num_size_one_inner_dims + 1; i < NumDims; ++i) {
+
+    // Squeeze multiple inner dims into one for larger inner dim size.
+    for (Index i = num_size_one_inner_dims + 1; i < num_squeezable_dims; ++i) {
       const Index dim = cond<Layout>()(i, NumDims - i - 1);
       const StorageIndex block_stride =
           block.block_strides()[tensor_to_block_dim_map[dim]];
-- 
cgit v1.2.3