Merge changes from github.

PiperOrigin-RevId: 160344052
STCYC · Jun 27, 2017 · 50b999a · 50b999a
1 parent d6d58a3
commit 50b999a
Show file tree

Hide file tree

Showing 303 changed files with 6,499 additions and 1,404 deletions.
diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
@@ -159,7 +159,12 @@ There are two ways to run TensorFlow unit tests.
    bazel test ${flags} //tensorflow/python/...
    ```
 
-2. Using Docker and TensorFlow's CI scripts.
+2. Using [Docker](www.docker.com) and TensorFlow's CI scripts.
+
+   ```bash
+   # Install Docker first, then this will build and run cpu tests
+   tensorflow/tools/ci_build/ci_build.sh CPU bazel test //tensorflow/...
+   ```
 
    See
    [TensorFlow Builds](https://github.com/tensorflow/tensorflow/tree/master/tensorflow/tools/ci_build) for details.

diff --git a/ISSUE_TEMPLATE.md b/ISSUE_TEMPLATE.md
@@ -6,6 +6,7 @@ If you open a GitHub issue, here is our policy:
 
 1. It must be a bug or a feature request.
 2. The form below must be filled out.
+3. It shouldn't be a TensorBoard issue. Those go [here](https://github.com/tensorflow/tensorboard/issues).
 
 **Here's why we have that policy**: TensorFlow developers respond to issues. We want to focus on work that benefits the whole community, e.g., fixing bugs and adding features. Support only helps individuals. GitHub also notifies thousands of people when issues are filed. We want them to see you communicating an interesting problem, rather than being redirected to Stack Overflow.
 

diff --git a/README.md b/README.md
@@ -34,13 +34,13 @@ and discussion.**
 
 People who are a little more adventurous can also try our nightly binaries:
 
-* Linux CPU-only: [Python 2](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=cpu-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.0rc2-cp27-none-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=cpu-slave)) / [Python 3.4](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=cpu-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.0rc2-cp34-cp34m-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=cpu-slave/)) / [Python 3.5](https://ci.tensorflow.org/view/Nightly/job/nightly-python35-linux-cpu/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.0rc2-cp35-cp35m-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-python35-linux-cpu/))
-* Linux GPU: [Python 2](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-linux/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.2.0rc2-cp27-none-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-linux/)) / [Python 3.4](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=gpu-linux/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.2.0rc2-cp34-cp34m-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=gpu-linux/)) / [Python 3.5](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3.5,label=gpu-linux/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.2.0rc2-cp35-cp35m-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3.5,label=gpu-linux/))
-* Mac CPU-only: [Python 2](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=mac-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.0rc2-py2-none-any.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=mac-slave/)) / [Python 3](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=mac-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.0rc2-py3-none-any.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=mac-slave/))
-* Mac GPU: [Python 2](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-mac-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-mac/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.2.0rc2-py2-none-any.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-mac-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-mac/)) / [Python 3](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-mac-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=gpu-mac/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.2.0rc2-py3-none-any.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-mac-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=gpu-mac/))
-* Windows CPU-only: [Python 3.5 64-bit](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows,PY=35/lastSuccessfulBuild/artifact/cmake_build/tf_python/dist/tensorflow-1.2.0rc2-cp35-cp35m-win_amd64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows,PY=35/)) / [Python 3.6 64-bit](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows,PY=36/lastSuccessfulBuild/artifact/cmake_build/tf_python/dist/tensorflow-1.2.0rc2-cp36-cp36m-win_amd64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows,PY=36/))
-* Windows GPU: [Python 3.5 64-bit](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows-gpu,PY=35/lastSuccessfulBuild/artifact/cmake_build/tf_python/dist/tensorflow_gpu-1.2.0rc2-cp35-cp35m-win_amd64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows-gpu,PY=35/)) / [Python 3.6 64-bit](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows-gpu,PY=36/lastSuccessfulBuild/artifact/cmake_build/tf_python/dist/tensorflow_gpu-1.2.0rc2-cp36-cp36m-win_amd64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows-gpu,PY=36/))
-* Android: [demo APK](https://ci.tensorflow.org/view/Nightly/job/nightly-android/lastSuccessfulBuild/artifact/out/tensorflow_demo.apk), [native libs](http://ci.tensorflow.org/view/Nightly/job/nightly-android/lastSuccessfulBuild/artifact/out/native/)
+* Linux CPU-only: [Python 2](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=cpu-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.0-cp27-none-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=cpu-slave)) / [Python 3.4](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=cpu-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.0-cp34-cp34m-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=cpu-slave/)) / [Python 3.5](https://ci.tensorflow.org/view/Nightly/job/nightly-python35-linux-cpu/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.0-cp35-cp35m-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-python35-linux-cpu/))
+* Linux GPU: [Python 2](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-linux/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.2.0-cp27-none-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-linux/)) / [Python 3.4](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=gpu-linux/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.2.0-cp34-cp34m-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=gpu-linux/)) / [Python 3.5](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3.5,label=gpu-linux/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.2.0-cp35-cp35m-linux_x86_64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3.5,label=gpu-linux/))
+* Mac CPU-only: [Python 2](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=mac-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.0-py2-none-any.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=mac-slave/)) / [Python 3](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=mac-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.2.0-py3-none-any.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=mac-slave/))
+* Mac GPU: [Python 2](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-mac-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-mac/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.2.0-py2-none-any.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-mac-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-mac/)) / [Python 3](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-mac-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=gpu-mac/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.2.0-py3-none-any.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-mac-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=gpu-mac/))
+* Windows CPU-only: [Python 3.5 64-bit](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows,PY=35/lastSuccessfulBuild/artifact/cmake_build/tf_python/dist/tensorflow-1.2.0-cp35-cp35m-win_amd64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows,PY=35/)) / [Python 3.6 64-bit](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows,PY=36/lastSuccessfulBuild/artifact/cmake_build/tf_python/dist/tensorflow-1.2.0-cp36-cp36m-win_amd64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows,PY=36/))
+* Windows GPU: [Python 3.5 64-bit](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows-gpu,PY=35/lastSuccessfulBuild/artifact/cmake_build/tf_python/dist/tensorflow_gpu-1.2.0-cp35-cp35m-win_amd64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows-gpu,PY=35/)) / [Python 3.6 64-bit](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows-gpu,PY=36/lastSuccessfulBuild/artifact/cmake_build/tf_python/dist/tensorflow_gpu-1.2.0-cp36-cp36m-win_amd64.whl) ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-win/M=windows-gpu,PY=36/))
+* Android: [demo APK](https://ci.tensorflow.org/view/Nightly/job/nightly-android/lastSuccessfulBuild/artifact/out/tensorflow_demo.apk), [native libs](https://ci.tensorflow.org/view/Nightly/job/nightly-android/lastSuccessfulBuild/artifact/out/native/)
 ([build history](https://ci.tensorflow.org/view/Nightly/job/nightly-android/))
 
 #### *Try your first TensorFlow program*

diff --git a/RELEASE.md b/RELEASE.md
@@ -113,6 +113,8 @@
   checkpoints containing such RNN cells, in which case you can use the
   [checkpoint_convert script](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/contrib/rnn/python/tools/checkpoint_convert.py)
   to convert the variable names in your old checkpoints.
+* Added `tf.contrib.kernel_methods` module with Ops and estimators for primal
+  (explicit) kernel methods in TensorFlow.
 
 ## Bug Fixes and Other Changes
 * In python, `Operation.get_attr` on type attributes returns the Python DType

diff --git a/configure b/configure
@@ -162,8 +162,12 @@ bazel version > bazel.version
 curr_bazel_version=$(head -n 1 bazel.version | cut -d ' ' -f3)
 rm -f bazel.version
 
+
 echo "You have bazel $curr_bazel_version installed."
-if [ "$(version "$MIN_BAZEL_VERSION")" -gt "$(version "$curr_bazel_version")" ]; then
+if [ -z "$curr_bazel_version" ]; then
+  echo "WARNING: current bazel installation is not a release version."
+  echo "Make sure you are running at least bazel $MIN_BAZEL_VERSION."
+elif [ "$(version "$MIN_BAZEL_VERSION")" -gt "$(version "$curr_bazel_version")" ]; then
   echo "Please upgrade your bazel installation to version $MIN_BAZEL_VERSION or higher to build TensorFlow!"
   echo "Exiting..."
   exit 1
@@ -535,9 +539,9 @@ done
 # Set default CUDA version if not set
 if [ -z "$TF_CUDA_VERSION" ]; then
   TF_CUDA_VERSION="8.0"
-  export TF_CUDA_VERSION 
+  export TF_CUDA_VERSION
 fi
-write_action_env_to_bazelrc "TF_CUDA_VERSION" "$TF_CUDA_VERSION" 
+write_action_env_to_bazelrc "TF_CUDA_VERSION" "$TF_CUDA_VERSION"
 
 # Set up which gcc nvcc should use as the host compiler
 # No need to set this on Windows
@@ -586,6 +590,9 @@ while true; do
     # Result returned from "read" will be used unexpanded. That make "~" unusable.
     # Going through one more level of expansion to handle that.
     CUDNN_INSTALL_PATH=`"${PYTHON_BIN_PATH}" -c "import os; print(os.path.realpath(os.path.expanduser('${CUDNN_INSTALL_PATH}')))"`
+    if is_windows; then
+      CUDNN_INSTALL_PATH="$(cygpath -m "$CUDNN_INSTALL_PATH")"
+    fi
   fi
 
   if [[ -z "$TF_CUDNN_VERSION" ]]; then
@@ -652,16 +659,22 @@ write_action_env_to_bazelrc "TF_CUDNN_VERSION" "$TF_CUDNN_VERSION"
 
 # Configure the compute capabilities that TensorFlow builds for.
 # Since Cuda toolkit is not backward-compatible, this is not guaranteed to work.
+function get_native_cuda_compute_capabilities {
+  device_query_bin="$CUDA_TOOLKIT_PATH/extras/demo_suite/deviceQuery" # Also works on Windows without .exe
+  "$device_query_bin" | grep 'Capability' | grep -o '[0-9]*\.[0-9]*' | sed ':a;{N;s/\n/,/};ba'
+  exit 0 # ensure that this function always exit success even if device detection fails, to prevent the whole configure from aborting
+}
 while true; do
   fromuser=""
-  default_cuda_compute_capabilities="3.5,5.2"
+  native_cuda_compute_capabilities=$(get_native_cuda_compute_capabilities)
+  default_cuda_compute_capabilities=${native_cuda_compute_capabilities:-"3.5,5.2"}
   if [ -z "$TF_CUDA_COMPUTE_CAPABILITIES" ]; then
 cat << EOF
 Please specify a list of comma-separated Cuda compute capabilities you want to build with.
 You can find the compute capability of your device at: https://developer.nvidia.com/cuda-gpus.
 Please note that each additional compute capability significantly increases your build time and binary size.
 EOF
-    read -p "[Default is: \"3.5,5.2\"]: " TF_CUDA_COMPUTE_CAPABILITIES
+    read -p "[Default is: \"$default_cuda_compute_capabilities\"]: " TF_CUDA_COMPUTE_CAPABILITIES
     fromuser=1
   fi
   if [ -z "$TF_CUDA_COMPUTE_CAPABILITIES" ]; then
@@ -832,29 +845,29 @@ while true; do
     if [ -e "$MPI_HOME/include" ] && [ -e "$MPI_HOME/lib" ]; then
         break
     fi
- 
+
     echo "Invalid path to the MPI Toolkit. ${MPI_HOME}/include or ${MPI_HOME}/lib cannot be found."
     if [ -z "$fromuser" ]; then
         exit 1
     fi
 
     # Retry
-    MPI_HOME="" 
+    MPI_HOME=""
 done
-    
-    
+
+
 if [ "$TF_NEED_MPI" == "1" ]; then
   write_to_bazelrc 'build --define with_mpi_support=true'
 
   #Link the MPI header files
   ln -sf "${MPI_HOME}/include/mpi.h" third_party/mpi/mpi.h
 
 
-  #Determine if we use OpenMPI or MVAPICH, these require different header files 
+  #Determine if we use OpenMPI or MVAPICH, these require different header files
   #to be included here to make bazel dependency checker happy
 
   if [ -e "${MPI_HOME}/include/mpi_portable_platform.h" ]; then
-        #OpenMPI 
+        #OpenMPI
         ln -sf "${MPI_HOME}/include/mpi_portable_platform.h" third_party/mpi/
         sed -i -e "s/MPI_LIB_IS_OPENMPI=False/MPI_LIB_IS_OPENMPI=True/" third_party/mpi/mpi.bzl
  else
@@ -864,7 +877,7 @@ if [ "$TF_NEED_MPI" == "1" ]; then
         sed -i -e "s/MPI_LIB_IS_OPENMPI=True/MPI_LIB_IS_OPENMPI=False/" third_party/mpi/mpi.bzl
  fi
 
-  
+
   if [ -e "${MPI_HOME}/lib/libmpi.so" ]; then
     ln -sf "${MPI_HOME}/lib/libmpi.so" third_party/mpi/
   else

diff --git a/tensorflow/BUILD b/tensorflow/BUILD
@@ -208,6 +208,7 @@ filegroup(
         "//tensorflow/compiler/jit/kernels:all_files",
         "//tensorflow/compiler/jit/legacy_flags:all_files",
         "//tensorflow/compiler/jit/ops:all_files",
+        "//tensorflow/compiler/plugin/executor:all_files",
         "//tensorflow/compiler/tests:all_files",
         "//tensorflow/compiler/tf2xla:all_files",
         "//tensorflow/compiler/tf2xla/cc:all_files",

diff --git a/tensorflow/c/generate-pc.sh b/tensorflow/c/generate-pc.sh
@@ -26,7 +26,7 @@ usage() {
 [ $# == 0 ] && usage && exit 0
 
 # read the options
-ARGS=`getopt -o p:v:h --long prefix:,version:,help -n $0 -- "$@"`
+ARGS=$(getopt -o p:v:h --long prefix:,version:,help -n $0 -- "$@")
 eval set -- "$ARGS"
 
 # extract options and their arguments into variables.

diff --git a/tensorflow/cc/BUILD b/tensorflow/cc/BUILD
@@ -472,10 +472,23 @@ cc_binary(
     name = "tutorials_example_trainer",
     srcs = ["tutorials/example_trainer.cc"],
     copts = tf_copts(),
-    linkopts = [
-        "-lpthread",
-        "-lm",
-    ],
+    linkopts = select({
+        "//tensorflow:windows": [],
+        "//tensorflow:windows_msvc": [],
+        "//tensorflow:darwin": [
+            "-lm",
+            "-lpthread",
+        ],
+        "//tensorflow:ios": [
+            "-lm",
+            "-lpthread",
+        ],
+        "//conditions:default": [
+            "-lm",
+            "-lpthread",
+            "-lrt",
+        ],
+    }),
     deps = [
         ":cc_ops",
         "//tensorflow/core:core_cpu",

diff --git a/tensorflow/cc/gradients/math_grad.cc b/tensorflow/cc/gradients/math_grad.cc
@@ -162,6 +162,32 @@ Status Log1pGrad(const Scope& scope, const Operation& op,
 }
 REGISTER_GRADIENT_OP("Log1p", Log1pGrad);
 
+Status SinhGrad(const Scope& scope, const Operation& op,
+                const std::vector<Output>& grad_inputs,
+                std::vector<Output>* grad_outputs) {
+  // y = sinh(x)
+  // dy/dx = cosh(x)
+  auto dydx = Cosh(scope, op.input(0));
+  // grad(x) = grad(y) * conj(dy/dx)
+  grad_outputs->push_back(
+      Mul(scope, grad_inputs[0], ConjugateHelper(scope, dydx)));
+  return scope.status();
+}
+REGISTER_GRADIENT_OP("Sinh", SinhGrad);
+
+Status CoshGrad(const Scope& scope, const Operation& op,
+                const std::vector<Output>& grad_inputs,
+                std::vector<Output>* grad_outputs) {
+  // y = cosh(x)
+  // dy/dx = sinh(x)
+  auto dydx = Sinh(scope, op.input(0));
+  // grad(x) = grad(y) * conj(dy/dx)
+  grad_outputs->push_back(
+      Mul(scope, grad_inputs[0], ConjugateHelper(scope, dydx)));
+  return scope.status();
+}
+REGISTER_GRADIENT_OP("Cosh", CoshGrad);
+
 Status TanhGrad(const Scope& scope, const Operation& op,
                 const std::vector<Output>& grad_inputs,
                 std::vector<Output>* grad_outputs) {