Demo Confidential Task for Apache TVM Machine Learning Framework

You should follow the guide to let it work in the enclave.

Install TVM

$ git clone https://github.com/apache/tvm.git $HOME/tvm
$ cd $HOME/tvm

Compile TVM and modify some settings.
```
$ mkdir build && cp ./cmake/config.cmake ./build
$ cd build
```
Change the following keys to ON. This is very important.
- USE_MICRO
- USE_MICRO_STANDALONE_RUNTIME
These keys allow the TVM to build a self-contained ML model; that is, the library build by TVM is a self-runnable bundle with the required interpreter runtime modules such as runtime::GraphExecutor, the graph JSON, and the params, which does not need to link against TVM runtime libraries anymore.

Then compile it.
```
$ cmake .. && make -j$(nproc)
```
By default, you can set the TVM_HOME to $HOME/tvm or the parent directory from where the TVM libraries are built.
Then you may also need to configure the path python looks for (assume you have set TVM_HOME).

$ export PYTHONPATH=$TVM_HOME/python:$TVM_HOME/topi/python:$TVM_HOME/nnvm/python:${PYTHONPATH}

Build the standalone library for enclave, which will automatically copies the libmodel_entry.a to ./platform_sgx/lib.

$ cd ./model_deploy
$ make

Provide feedback