tools/llvmc/doc/LLVMC-Reference.rst

===================================
Customizing LLVMC: Reference Manual
===================================
:Author: Mikhail Glushenkov <foldr@codedegers.com>

LLVMC is a generic compiler driver, designed to be customizable and
extensible. It plays the same role for LLVM as the ``gcc`` program
does for GCC - LLVMC's job is essentially to transform a set of input
files into a set of targets depending on configuration rules and user
options. What makes LLVMC different is that these transformation rules
are completely customizable - in fact, LLVMC knows nothing about the
specifics of transformation (even the command-line options are mostly
not hard-coded) and regards the transformation structure as an
abstract graph. The structure of this graph is completely determined
by plugins, which can be either statically or dynamically linked. This
makes it possible to easily adapt LLVMC for other purposes - for
example, as a build tool for game resources.

Because LLVMC employs TableGen [1]_ as its configuration language, you
need to be familiar with it to customize LLVMC.


.. contents::


Compiling with LLVMC
====================

LLVMC tries hard to be as compatible with ``gcc`` as possible,
although there are some small differences. Most of the time, however,
you shouldn't be able to notice them::

     $ # This works as expected:
     $ llvmc -O3 -Wall hello.cpp
     $ ./a.out
     hello

One nice feature of LLVMC is that one doesn't have to distinguish
between different compilers for different languages (think ``g++`` and
``gcc``) - the right toolchain is chosen automatically based on input
language names (which are, in turn, determined from file
extensions). If you want to force files ending with ".c" to compile as
C++, use the ``-x`` option, just like you would do it with ``gcc``::

      $ # hello.c is really a C++ file
      $ llvmc -x c++ hello.c
      $ ./a.out
      hello

On the other hand, when using LLVMC as a linker to combine several C++
object files you should provide the ``--linker`` option since it's
impossible for LLVMC to choose the right linker in that case::

    $ llvmc -c hello.cpp
    $ llvmc hello.o
    [A lot of link-time errors skipped]
    $ llvmc --linker=c++ hello.o
    $ ./a.out
    hello


Predefined options
==================

LLVMC has some built-in options that can't be overridden in the
configuration libraries:

* ``-o FILE`` - Output file name.

* ``-x LANGUAGE`` - Specify the language of the following input files
  until the next -x option.

* ``-load PLUGIN_NAME`` - Load the specified plugin DLL. Example:
  ``-load $LLVM_DIR/Release/lib/LLVMCSimple.so``.

* ``-v`` - Enable verbose mode, i.e. print out all executed commands.

* ``--view-graph`` - Show a graphical representation of the compilation
  graph. Requires that you have ``dot`` and ``gv`` programs
  installed. Hidden option, useful for debugging.

* ``--write-graph`` - Write a ``compilation-graph.dot`` file in the
  current directory with the compilation graph description in the
  Graphviz format. Hidden option, useful for debugging.

* ``--save-temps`` - Write temporary files to the current directory
  and do not delete them on exit. Hidden option, useful for debugging.

* ``--help``, ``--help-hidden``, ``--version`` - These options have
  their standard meaning.


Compiling LLVMC plugins
=======================

It's easiest to start working on your own LLVMC plugin by copying the
skeleton project which lives under ``$LLVMC_DIR/plugins/Simple``::

   $ cd $LLVMC_DIR/plugins
   $ cp -r Simple MyPlugin
   $ cd MyPlugin
   $ ls
   Makefile PluginMain.cpp Simple.td

As you can see, our basic plugin consists of only two files (not
counting the build script). ``Simple.td`` contains TableGen
description of the compilation graph; its format is documented in the
following sections. ``PluginMain.cpp`` is just a helper file used to
compile the auto-generated C++ code produced from TableGen source. It
can also contain hook definitions (see `below`__).

__ hooks_

The first thing that you should do is to change the ``LLVMC_PLUGIN``
variable in the ``Makefile`` to avoid conflicts (since this variable
is used to name the resulting library)::

   LLVMC_PLUGIN=MyPlugin

It is also a good idea to rename ``Simple.td`` to something less
generic::

   $ mv Simple.td MyPlugin.td

Note that the plugin source directory must be placed under
``$LLVMC_DIR/plugins`` to make use of the existing build
infrastructure. To build a version of the LLVMC executable called
``mydriver`` with your plugin compiled in, use the following command::

   $ cd $LLVMC_DIR
   $ make BUILTIN_PLUGINS=MyPlugin DRIVER_NAME=mydriver

To build your plugin as a dynamic library, just ``cd`` to its source
directory and run ``make``. The resulting file will be called
``LLVMC$(LLVMC_PLUGIN).$(DLL_EXTENSION)`` (in our case,
``LLVMCMyPlugin.so``). This library can be then loaded in with the
``-load`` option. Example::

    $ cd $LLVMC_DIR/plugins/Simple
    $ make
    $ llvmc -load $LLVM_DIR/Release/lib/LLVMCSimple.so

Sometimes, you will want a 'bare-bones' version of LLVMC that has no
built-in plugins. It can be compiled with the following command::

    $ cd $LLVMC_DIR
    $ make BUILTIN_PLUGINS=""

How plugins are loaded
======================

It is possible for LLVMC plugins to depend on each other. For example,
one can create edges between nodes defined in some other plugin. To
make this work, however, that plugin should be loaded first. To
achieve this, the concept of plugin priority was introduced. By
default, every plugin has priority zero; to specify the priority
explicitly, put the following line in your plugin's TableGen file::

    def Priority : PluginPriority<$PRIORITY_VALUE>;
    # Where PRIORITY_VALUE is some integer > 0

Plugins are loaded in order of their (increasing) priority, starting
with 0. Therefore, the plugin with the highest priority value will be
loaded last.


Customizing LLVMC: the compilation graph
========================================

Each TableGen configuration file should include the common
definitions::

   include "llvm/CompilerDriver/Common.td"
   // And optionally:
   // include "llvm/CompilerDriver/Tools.td"
   // which contains some useful tool definitions.

Internally, LLVMC stores information about possible source
transformations in form of a graph. Nodes in this graph represent
tools, and edges between two nodes represent a transformation path. A
special "root" node is used to mark entry points for the
transformations. LLVMC also assigns a weight to each edge (more on
this later) to choose between several alternative edges.

The definition of the compilation graph (see file
``plugins/Base/Base.td`` for an example) is just a list of edges::

    def CompilationGraph : CompilationGraph<[
        Edge<"root", "llvm_gcc_c">,
        Edge<"root", "llvm_gcc_assembler">,
        ...

        Edge<"llvm_gcc_c", "llc">,
        Edge<"llvm_gcc_cpp", "llc">,
        ...

        OptionalEdge<"llvm_gcc_c", "opt", (case (switch_on "opt"),
                                          (inc_weight))>,
        OptionalEdge<"llvm_gcc_cpp", "opt", (case (switch_on "opt"),
                                                  (inc_weight))>,
        ...

        OptionalEdge<"llvm_gcc_assembler", "llvm_gcc_cpp_linker",
            (case (input_languages_contain "c++"), (inc_weight),
                  (or (parameter_equals "linker", "g++"),
                      (parameter_equals "linker", "c++")), (inc_weight))>,
        ...

        ]>;

As you can see, the edges can be either default or optional, where
optional edges are differentiated by an additional ``case`` expression
used to calculate the weight of this edge. Notice also that we refer
to tools via their names (as strings). This makes it possible to add
edges to an existing compilation graph in plugins without having to
know about all tool definitions used in the graph.

The default edges are assigned a weight of 1, and optional edges get a
weight of 0 + 2*N where N is the number of tests that evaluated to
true in the ``case`` expression. It is also possible to provide an
integer parameter to ``inc_weight`` and ``dec_weight`` - in this case,
the weight is increased (or decreased) by the provided value instead
of the default 2. It is also possible to change the default weight of
an optional edge by using the ``default`` clause of the ``case``
construct.

When passing an input file through the graph, LLVMC picks the edge
with the maximum weight. To avoid ambiguity, there should be only one
default edge between two nodes (with the exception of the root node,
which gets a special treatment - there you are allowed to specify one
default edge *per language*).

When multiple plugins are loaded, their compilation graphs are merged
together. Since multiple edges that have the same end nodes are not
allowed (i.e. the graph is not a multigraph), an edge defined in
several plugins will be replaced by the definition from the plugin
that was loaded last. Plugin load order can be controlled by using the
plugin priority feature described above.

To get a visual representation of the compilation graph (useful for
debugging), run ``llvmc --view-graph``. You will need ``dot`` and
``gsview`` installed for this to work properly.


Writing a tool description
==========================

As was said earlier, nodes in the compilation graph represent tools,
which are described separately. A tool definition looks like this
(taken from the ``include/llvm/CompilerDriver/Tools.td`` file)::

  def llvm_gcc_cpp : Tool<[
      (in_language "c++"),
      (out_language "llvm-assembler"),
      (output_suffix "bc"),
      (cmd_line "llvm-g++ -c $INFILE -o $OUTFILE -emit-llvm"),
      (sink)
      ]>;

This defines a new tool called ``llvm_gcc_cpp``, which is an alias for
``llvm-g++``. As you can see, a tool definition is just a list of
properties; most of them should be self-explanatory. The ``sink``
property means that this tool should be passed all command-line
options that lack explicit descriptions.

The complete list of the currently implemented tool properties follows:

* Possible tool properties:

  - ``in_language`` - input language name. Can be either a string or a
    list, in case the tool supports multiple input languages.

  - ``out_language`` - output language name.

  - ``output_suffix`` - output file suffix.

  - ``cmd_line`` - the actual command used to run the tool. You can
    use ``$INFILE`` and ``$OUTFILE`` variables, output redirection
    with ``>``, hook invocations (``$CALL``), environment variables
    (via ``$ENV``) and the ``case`` construct (more on this below).

  - ``join`` - this tool is a "join node" in the graph, i.e. it gets a
    list of input files and joins them together. Used for linkers.

  - ``sink`` - all command-line options that are not handled by other
    tools are passed to this tool.

The next tool definition is slightly more complex::

  def llvm_gcc_linker : Tool<[
      (in_language "object-code"),
      (out_language "executable"),
      (output_suffix "out"),
      (cmd_line "llvm-gcc $INFILE -o $OUTFILE"),
      (join),
      (prefix_list_option "L", (forward),
                          (help "add a directory to link path")),
      (prefix_list_option "l", (forward),
                          (help "search a library when linking")),
      (prefix_list_option "Wl", (unpack_values),
                          (help "pass options to linker"))
      ]>;

This tool has a "join" property, which means that it behaves like a
linker. This tool also defines several command-line options: ``-l``,
``-L`` and ``-Wl`` which have their usual meaning. An option has two
attributes: a name and a (possibly empty) list of properties. All
currently implemented option types and properties are described below:

* Possible option types:

   - ``switch_option`` - a simple boolean switch, for example ``-time``.

   - ``parameter_option`` - option that takes an argument, for example
     ``-std=c99``;

   - ``parameter_list_option`` - same as the above, but more than one
     occurence of the option is allowed.

   - ``prefix_option`` - same as the parameter_option, but the option name
     and parameter value are not separated.

   - ``prefix_list_option`` - same as the above, but more than one
     occurence of the option is allowed; example: ``-lm -lpthread``.

   - ``alias_option`` - a special option type for creating
     aliases. Unlike other option types, aliases are not allowed to
     have any properties besides the aliased option name. Usage
     example: ``(alias_option "preprocess", "E")``


* Possible option properties:

   - ``append_cmd`` - append a string to the tool invocation command.

   - ``forward`` - forward this option unchanged.

   - ``forward_as`` - Change the name of this option, but forward the
     argument unchanged. Example: ``(forward_as "--disable-optimize")``.

   - ``output_suffix`` - modify the output suffix of this
     tool. Example: ``(switch "E", (output_suffix "i")``.

   - ``stop_compilation`` - stop compilation after this phase.

   - ``unpack_values`` - used for for splitting and forwarding
     comma-separated lists of options, e.g. ``-Wa,-foo=bar,-baz`` is
     converted to ``-foo=bar -baz`` and appended to the tool invocation
     command.

   - ``help`` - help string associated with this option. Used for
     ``--help`` output.

   - ``required`` - this option is obligatory.


Option list - specifying all options in a single place
======================================================

It can be handy to have all information about options gathered in a
single place to provide an overview. This can be achieved by using a
so-called ``OptionList``::

    def Options : OptionList<[
    (switch_option "E", (help "Help string")),
    (alias_option "quiet", "q")
    ...
    ]>;

``OptionList`` is also a good place to specify option aliases.

Tool-specific option properties like ``append_cmd`` have (obviously)
no meaning in the context of ``OptionList``, so the only properties
allowed there are ``help`` and ``required``.

Option lists are used at file scope. See the file
``plugins/Clang/Clang.td`` for an example of ``OptionList`` usage.

.. _hooks:

Using hooks and environment variables in the ``cmd_line`` property
==================================================================

Normally, LLVMC executes programs from the system ``PATH``. Sometimes,
this is not sufficient: for example, we may want to specify tool names
in the configuration file. This can be achieved via the mechanism of
hooks - to write your own hooks, just add their definitions to the
``PluginMain.cpp`` or drop a ``.cpp`` file into the
``$LLVMC_DIR/driver`` directory. Hooks should live in the ``hooks``
namespace and have the signature ``std::string hooks::MyHookName
(void)``. They can be used from the ``cmd_line`` tool property::

    (cmd_line "$CALL(MyHook)/path/to/file -o $CALL(AnotherHook)")

It is also possible to use environment variables in the same manner::

   (cmd_line "$ENV(VAR1)/path/to/file -o $ENV(VAR2)")

To change the command line string based on user-provided options use
the ``case`` expression (documented below)::

    (cmd_line
      (case
        (switch_on "E"),
           "llvm-g++ -E -x c $INFILE -o $OUTFILE",
        (default),
           "llvm-g++ -c -x c $INFILE -o $OUTFILE -emit-llvm"))

Conditional evaluation: the ``case`` expression
===============================================

The 'case' construct can be used to calculate weights of the optional
edges and to choose between several alternative command line strings
in the ``cmd_line`` tool property. It is designed after the
similarly-named construct in functional languages and takes the form
``(case (test_1), statement_1, (test_2), statement_2, ... (test_N),
statement_N)``. The statements are evaluated only if the corresponding
tests evaluate to true.

Examples::

    // Increases edge weight by 5 if "-A" is provided on the
    // command-line, and by 5 more if "-B" is also provided.
    (case
        (switch_on "A"), (inc_weight 5),
        (switch_on "B"), (inc_weight 5))

    // Evaluates to "cmdline1" if option "-A" is provided on the
    // command line, otherwise to "cmdline2"
    (case
        (switch_on "A"), "cmdline1",
        (switch_on "B"), "cmdline2",
        (default), "cmdline3")

Note the slight difference in 'case' expression handling in contexts
of edge weights and command line specification - in the second example
the value of the ``"B"`` switch is never checked when switch ``"A"`` is
enabled, and the whole expression always evaluates to ``"cmdline1"`` in
that case.

Case expressions can also be nested, i.e. the following is legal::

    (case (switch_on "E"), (case (switch_on "o"), ..., (default), ...)
          (default), ...)

You should, however, try to avoid doing that because it hurts
readability. It is usually better to split tool descriptions and/or
use TableGen inheritance instead.

* Possible tests are:

  - ``switch_on`` - Returns true if a given command-line switch is
    provided by the user. Example: ``(switch_on "opt")``. Note that
    you have to define all possible command-line options separately in
    the tool descriptions. See the next section for the discussion of
    different kinds of command-line options.

  - ``parameter_equals`` - Returns true if a command-line parameter equals
    a given value. Example: ``(parameter_equals "W", "all")``.

  - ``element_in_list`` - Returns true if a command-line parameter list
    includes a given value. Example: ``(parameter_in_list "l", "pthread")``.

  - ``input_languages_contain`` - Returns true if a given language
    belongs to the current input language set. Example:
    ``(input_languages_contain "c++")``.

  - ``in_language`` - Evaluates to true if the language of the input
    file equals to the argument. At the moment works only with
    ``cmd_line`` property on non-join nodes. Example: ``(in_language
    "c++")``.

  - ``not_empty`` - Returns true if a given option (which should be
    either a parameter or a parameter list) is set by the
    user. Example: ``(not_empty "o")``.

  - ``default`` - Always evaluates to true. Should always be the last
    test in the ``case`` expression.

  - ``and`` - A standard logical combinator that returns true iff all
    of its arguments return true. Used like this: ``(and (test1),
    (test2), ... (testN))``. Nesting of ``and`` and ``or`` is allowed,
    but not encouraged.

  - ``or`` - Another logical combinator that returns true only if any
    one of its arguments returns true. Example: ``(or (test1),
    (test2), ... (testN))``.


Language map
============

One last thing that you will need to modify when adding support for a
new language to LLVMC is the language map, which defines mappings from
file extensions to language names. It is used to choose the proper
toolchain(s) for a given input file set. Language map definition looks
like this::

    def LanguageMap : LanguageMap<
        [LangToSuffixes<"c++", ["cc", "cp", "cxx", "cpp", "CPP", "c++", "C"]>,
         LangToSuffixes<"c", ["c"]>,
         ...
        ]>;

Debugging
=========

When writing LLVMC plugins, it can be useful to get a visual view of
the resulting compilation graph. This can be achieved via the command
line option ``--view-graph``. This command assumes that Graphviz [2]_ and
Ghostview [3]_ are installed. There is also a ``--dump-graph`` option that
creates a Graphviz source file(``compilation-graph.dot``) in the
current directory.


References
==========

.. [1] TableGen Fundamentals
       http://llvm.cs.uiuc.edu/docs/TableGenFundamentals.html

.. [2] Graphviz
       http://www.graphviz.org/

.. [3] Ghostview
       http://pages.cs.wisc.edu/~ghost/