[DOCS][APP] Add Example for C++ deployment (#398)

* [DOCS][APP] Add Example for C++ deployment

* fix lint

apps/README.md

-# TVM Application Extensions
+# TVM Application Extensions and Examples
 This folder contains various extension projects using TVM,
 they also serve as examples on how to use TVM in your own project.
 
@@ -6,3 +6,7 @@ If you are interested in writing optimized kernels with TVM, checkout [TOPI: TVM
 
 - [extension](extension) How to extend TVM C++ api along with python API.
 - [graph_executor](graph_executor) Build nnvm graph executor with TVM.
+- [ios_rpc](ios_rpc) iOS RPC server.
+- [android_rpc](android_rpc) Android RPC server.
+- [cpp_deploy](cpp_deploy) Example to deploy with C++ runtime.
+

apps/cpp_deploy/Makefile

+# Minimum Makefile for the extension package
+TVM_ROOT=$(shell cd ../..; pwd)
+NNVM_PATH=nnvm
+DMLC_CORE=${TVM_ROOT}/dmlc-core
+
+PKG_CFLAGS = -std=c++11 -O2 -fPIC\
+	-I${TVM_ROOT}/include\
+	-I${DMLC_CORE}/include\
+	-I${TVM_ROOT}/dlpack/include\
+
+PKG_LDFLAGS = -L${TVM_ROOT}/lib -ltvm_runtime
+
+
+lib/cpp_deploy: cpp_deploy.cc lib/test_addone_sys.o
+	@mkdir -p $(@D)
+	$(CXX) $(PKG_CFLAGS) -o $@  $^ $(PKG_LDFLAGS)
+
+lib/test_addone_sys.o: prepare_test_libs.py
+	python prepare_test_libs.py

apps/cpp_deploy/README.md

+C++ Deployment API Example
+==========================
+This folder contains an example to demonstrate how to use TVM C++ Runtime
+API to load and run modules built by TVM.
+
+Type the following command to run the example under current folder(need to build TVM first).
+```bash
+./run_example.sh
+```
+

apps/cpp_deploy/cpp_deploy.cc

+/*!
+ *  Copyright (c) 2017 by Contributors
+ * \brief Example code on load and run TVM module.s
+ * \file cpp_deploy_example.cc
+ */
+#include <cstdio>
+#include <dlpack/dlpack.h>
+#include <tvm/runtime/module.h>
+#include <tvm/runtime/registry.h>
+#include <tvm/runtime/packed_func.h>
+
+void Verify(tvm::runtime::Module mod, std::string fname) {
+  // Get the function from the module.
+  tvm::runtime::PackedFunc f = mod.GetFunction(fname);
+  CHECK(f != nullptr);
+  // Allocate the DLPack data structures.
+  //
+  // Note that we use TVM runtime API to allocate the DLTensor in this example.
+  // TVM accept DLPack compatible DLTensors, so function can be invoked
+  // as long as we pass correct pointer to DLTensor array.
+  //
+  // For more information please refer to dlpack.
+  // One thing to notice is that DLPack contains alignment requirement for
+  // the data pointer and TVM takes advantage of that.
+  // If you plan to use your customized data container, please
+  // make sure the DLTensor you pass in meet the alignment requirement.
+  //
+  DLTensor* x;
+  DLTensor* y;
+  int ndim = 1;
+  int dtype_code = kFloat;
+  int dtype_bits = 32;
+  int dtype_lanes = 1;
+  int device_type = kCPU;
+  int device_id = 0;
+  int64_t shape[1] = {10};
+  TVMArrayAlloc(shape, ndim, dtype_code, dtype_bits, dtype_lanes,
+                device_type, device_id, &x);
+  TVMArrayAlloc(shape, ndim, dtype_code, dtype_bits, dtype_lanes,
+                device_type, device_id, &y);
+  for (int i = 0; i < shape[0]; ++i) {
+    static_cast<float*>(x->data)[i] = i;
+  }
+  // Invoke the function
+  // PackedFunc is a function that can be invoked via positional argument.
+  // The signature of the function is specified in tvm.build
+  f(x, y);
+  // Print out the output
+  for (int i = 0; i < shape[0]; ++i) {
+    CHECK_EQ(static_cast<float*>(y->data)[i], i + 1.0f);
+  }
+  LOG(INFO) << "Finish verification...";
+}
+
+int main(void) {
+  // Normally we can directly
+  tvm::runtime::Module mod_dylib =
+      tvm::runtime::Module::LoadFromFile("lib/test_addone_dll.so");
+  LOG(INFO) << "Verify dynamic loading from test_addone_dll.so";
+  Verify(mod_dylib, "addone");
+  // For libraries that are directly packed as system lib and linked together with the app
+  // We can directly use GetSystemLib to get the system wide library.
+  LOG(INFO) << "Verify load function from system lib";
+  tvm::runtime::Module mod_syslib = (*tvm::runtime::Registry::Get("module._GetSystemLib"))();
+  Verify(mod_syslib, "addonesys");
+  return 0;
+}

apps/cpp_deploy/prepare_test_libs.py

+"""Script to prepare test_addone.so"""
+import tvm
+import os
+
+def prepare_test_libs(base_path):
+    n = tvm.var("n")
+    A = tvm.placeholder((n,), name='A')
+    B = tvm.compute(A.shape, lambda *i: A(*i) + 1.0, name='B')
+    s = tvm.create_schedule(B.op)
+    fadd_dylib = tvm.build(s, [A, B], "llvm", name="addone")
+    fadd_syslib = tvm.build(s, [A, B], "llvm --system-lib", name="addonesys")
+    dylib_path = os.path.join(base_path, "test_addone_dll.so")
+    syslib_path = os.path.join(base_path, "test_addone_sys.o")
+    fadd_dylib.export_library(dylib_path)
+    fadd_syslib.save(syslib_path)
+
+if __name__ == "__main__":
+    curr_path = os.path.dirname(os.path.abspath(os.path.expanduser(__file__)))
+    prepare_test_libs(os.path.join(curr_path, "./lib"))

apps/cpp_deploy/run_example.sh

+#!/bin/bash
+echo "Build the libraries.."
+make
+echo "Run the example"
+export LD_LIBRARY_PATH=../../lib
+export DYLD_LIBRARY_PATH=../../lib
+lib/cpp_deploy
\ No newline at end of file

docs/api/python/topi.rst

@@ -20,7 +20,6 @@ Index
    topi.min
    topi.nn.relu
    topi.nn.dilate
-   topi.nn.scale_shift
    topi.nn.conv2d_nchw
    topi.nn.conv2d_hwcn
    topi.nn.depthwise_conv2d_nchw
@@ -55,7 +54,6 @@ topi.nn
 ~~~~~~~
 .. autofunction:: topi.nn.relu
 .. autofunction:: topi.nn.dilate
-.. autofunction:: topi.nn.scale_shift
 .. autofunction:: topi.nn.conv2d_nchw
 .. autofunction:: topi.nn.conv2d_hwcn
 .. autofunction:: topi.nn.depthwise_conv2d_nchw

include/tvm/runtime/module.h

@@ -39,7 +39,7 @@ class Module {
    * \return The result function.
    *  This function will return PackedFunc(nullptr) if function do not exist.
    */
-  PackedFunc GetFunction(const std::string& name, bool query_imports);
+  PackedFunc GetFunction(const std::string& name, bool query_imports = false);
   /*!
    * \brief Import another module into this module.
    * \param other The module to be imported.
@@ -56,7 +56,7 @@ class Module {
    *  Re-create import relationship by calling Import.
    */
   static Module LoadFromFile(const std::string& file_name,
-                             const std::string& format);
+                             const std::string& format = "");
   /*! \return internal container */
   inline ModuleNode* operator->();
   /*! \return internal container */

topi/python/topi/reduction.py

@@ -44,6 +44,7 @@ def comm_reduce(data, axis=None, keepdims=False, func=tvm.sum):
     Parameters
     ----------
     data : tvm.Tensor
+        The input data
 
     axis : None or int or tuple of int
         Axis or axes along which a sum is performed.
@@ -120,7 +121,7 @@ def sum(data, axis=None, keepdims=False):
 
     keepdims : bool
         If this is set to True, the axes which are reduced are left in the result as dimensions
-         with size one.
+        with size one.
         With this option, the result will broadcast correctly against the input array.
 
     Returns
@@ -145,7 +146,7 @@ def max(data, axis=None, keepdims=False):
 
     keepdims : bool
         If this is set to True, the axes which are reduced are left in the result as dimensions
-         with size one.
+        with size one.
         With this option, the result will broadcast correctly against the input array.
 
     Returns
@@ -170,7 +171,7 @@ def min(data, axis=None, keepdims=False):
 
     keepdims : bool
         If this is set to True, the axes which are reduced are left in the result as dimensions
-         with size one.
+        with size one.
         With this option, the result will broadcast correctly against the input array.
 
     Returns