Add development mode dockerfile and info on use (#78)

Author: boomanaiden154

experimental/docker/README.md

@@ -0,0 +1,23 @@
+# MLGO Development Dockerfiles
+
+This folder contains dockerfiles with all the dependencies necessary to do
+development work with MLGO. Note that these dockerfiles do not contain a
+clone of LLVM, only system dependencies, python dependencies, and the current
+repository. You are expected to clone LLVM and configure the build yourself.
+To see how to do this, the [inliner demo](../docs/demo/demo.md) should have
+some useful guidance.
+
+### Building the image
+
+To build the image, make sure that you are in the root of the ml-compiler-opt
+repository and then run:
+```bash
+docker build -t <your tag here> -f ./docker/development.Dockerfile .
+```
+
+### Disclaimers
+
+**WARNING**: These development dockerfiles are not guaranteed to work. We will
+do a best effort to make sure these dockerfiles are working, but they might stop
+working at any point in time and should absolutely not be used in production
+environments.

experimental/docker/development.Dockerfile

@@ -0,0 +1,53 @@
+# TODO(boomanaiden154): Refactor to 22.04 and remove custom PPAs once we aren't
+# dependent upon a specific tensorflow version and thus a specific python version
+# (once TFLite patch lands)
+FROM ubuntu:20.04
+RUN apt-get update
+RUN DEBIAN_FRONTEND=noninteractive apt-get install -y \
+    python3-distutils \
+    python-is-python3 \
+    python3 \
+    python3-pip \
+    tmux \
+    g++ \
+    ccache \
+    binutils-gold \
+    binutils-dev \
+    ninja-build \
+    pkg-config \
+    gcc-multilib \
+    g++-multilib \
+    gawk \
+    dos2unix \
+    libxml2-dev \
+    rsync \
+    git \
+    libtool \
+    m4 \
+    automake \
+    libgcrypt-dev \
+    liblzma-dev \
+    libssl-dev \
+    libgss-dev \
+    python3-dev \
+    wget \
+    zlib1g-dev \
+    tcl-dev \
+    libpfm4-dev \
+    software-properties-common
+RUN wget --quiet https://storage.googleapis.com/tensorflow/libtensorflow/libtensorflow-cpu-linux-x86_64-1.15.0.tar.gz && \
+    mkdir /tmp/tensorflow && \
+    tar xfz libtensorflow-cpu-linux-x86_64-1.15.0.tar.gz -C /tmp/tensorflow
+# install latest cmake rather than the one in the 20.04 repos as that version
+# doesn't support some cmake options used by MLGO
+RUN wget -O - https://apt.kitware.com/keys/kitware-archive-latest.asc 2>/dev/null | gpg --dearmor - | tee /usr/share/keyrings/kitware-archive-keyring.gpg >/dev/null && \
+    echo 'deb [signed-by=/usr/share/keyrings/kitware-archive-keyring.gpg] https://apt.kitware.com/ubuntu/ focal main' | tee /etc/apt/sources.list.d/kitware.list >/dev/null && \
+    apt-get update && \
+    apt-get install -y cmake
+# install latest git rather than 20.04 default due to that git version
+# having some issues with projects commonly used by MLGO such as chromium
+RUN apt-add-repository ppa:git-core/ppa && \
+    apt-get update && \
+    apt-get install -y git vim
+COPY . /ml-compiler-opt
+RUN python3 -m pip install -r /ml-compiler-opt/requirements-dev.txt