<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html xmlns="http://www.w3.org/1999/xhtml"> <head> <meta http-equiv="X-UA-Compatible" content="IE=Edge" /> <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> <title>5. Kaleidoscope: Extending the Language: Control Flow — LLVM 8 documentation</title> <link rel="stylesheet" href="../_static/llvm-theme.css" type="text/css" /> <link rel="stylesheet" href="../_static/pygments.css" type="text/css" /> <script type="text/javascript" id="documentation_options" data-url_root="../" src="../_static/documentation_options.js"></script> <script type="text/javascript" src="../_static/jquery.js"></script> <script type="text/javascript" src="../_static/underscore.js"></script> <script type="text/javascript" src="../_static/doctools.js"></script> <script type="text/javascript" src="../_static/language_data.js"></script> <link rel="index" title="Index" href="../genindex.html" /> <link rel="search" title="Search" href="../search.html" /> <link rel="next" title="6. Kaleidoscope: Extending the Language: User-defined Operators" href="OCamlLangImpl6.html" /> <link rel="prev" title="4. Kaleidoscope: Adding JIT and Optimizer Support" href="OCamlLangImpl4.html" /> <style type="text/css"> table.right { float: right; margin-left: 20px; } table.right td { border: 1px solid #ccc; } </style> </head><body> <div class="logo"> <a href="../index.html"> <img src="../_static/logo.png" alt="LLVM Logo" width="250" height="88"/></a> </div> <div class="related" role="navigation" aria-label="related navigation"> <h3>Navigation</h3> <ul> <li class="right" style="margin-right: 10px"> <a href="../genindex.html" title="General Index" accesskey="I">index</a></li> <li class="right" > <a href="OCamlLangImpl6.html" title="6. Kaleidoscope: Extending the Language: User-defined Operators" accesskey="N">next</a> |</li> <li class="right" > <a href="OCamlLangImpl4.html" title="4. Kaleidoscope: Adding JIT and Optimizer Support" accesskey="P">previous</a> |</li> <li><a href="http://llvm.org/">LLVM Home</a> | </li> <li><a href="../index.html">Documentation</a>»</li> <li class="nav-item nav-item-1"><a href="index.html" accesskey="U">LLVM Tutorial: Table of Contents</a> »</li> </ul> </div> <div class="document"> <div class="documentwrapper"> <div class="body" role="main"> <div class="section" id="kaleidoscope-extending-the-language-control-flow"> <h1>5. Kaleidoscope: Extending the Language: Control Flow<a class="headerlink" href="#kaleidoscope-extending-the-language-control-flow" title="Permalink to this headline">¶</a></h1> <div class="contents local topic" id="contents"> <ul class="simple"> <li><a class="reference internal" href="#chapter-5-introduction" id="id2">Chapter 5 Introduction</a></li> <li><a class="reference internal" href="#if-then-else" id="id3">If/Then/Else</a><ul> <li><a class="reference internal" href="#lexer-extensions-for-if-then-else" id="id4">Lexer Extensions for If/Then/Else</a></li> <li><a class="reference internal" href="#ast-extensions-for-if-then-else" id="id5">AST Extensions for If/Then/Else</a></li> <li><a class="reference internal" href="#parser-extensions-for-if-then-else" id="id6">Parser Extensions for If/Then/Else</a></li> <li><a class="reference internal" href="#llvm-ir-for-if-then-else" id="id7">LLVM IR for If/Then/Else</a></li> <li><a class="reference internal" href="#code-generation-for-if-then-else" id="id8">Code Generation for If/Then/Else</a></li> </ul> </li> <li><a class="reference internal" href="#for-loop-expression" id="id9">‘for’ Loop Expression</a><ul> <li><a class="reference internal" href="#lexer-extensions-for-the-for-loop" id="id10">Lexer Extensions for the ‘for’ Loop</a></li> <li><a class="reference internal" href="#ast-extensions-for-the-for-loop" id="id11">AST Extensions for the ‘for’ Loop</a></li> <li><a class="reference internal" href="#parser-extensions-for-the-for-loop" id="id12">Parser Extensions for the ‘for’ Loop</a></li> <li><a class="reference internal" href="#llvm-ir-for-the-for-loop" id="id13">LLVM IR for the ‘for’ Loop</a></li> <li><a class="reference internal" href="#code-generation-for-the-for-loop" id="id14">Code Generation for the ‘for’ Loop</a></li> </ul> </li> <li><a class="reference internal" href="#full-code-listing" id="id15">Full Code Listing</a></li> </ul> </div> <div class="section" id="chapter-5-introduction"> <h2><a class="toc-backref" href="#id2">5.1. Chapter 5 Introduction</a><a class="headerlink" href="#chapter-5-introduction" title="Permalink to this headline">¶</a></h2> <p>Welcome to Chapter 5 of the “<a class="reference external" href="index.html">Implementing a language with LLVM</a>” tutorial. Parts 1-4 described the implementation of the simple Kaleidoscope language and included support for generating LLVM IR, followed by optimizations and a JIT compiler. Unfortunately, as presented, Kaleidoscope is mostly useless: it has no control flow other than call and return. This means that you can’t have conditional branches in the code, significantly limiting its power. In this episode of “build that compiler”, we’ll extend Kaleidoscope to have an if/then/else expression plus a simple ‘for’ loop.</p> </div> <div class="section" id="if-then-else"> <h2><a class="toc-backref" href="#id3">5.2. If/Then/Else</a><a class="headerlink" href="#if-then-else" title="Permalink to this headline">¶</a></h2> <p>Extending Kaleidoscope to support if/then/else is quite straightforward. It basically requires adding lexer support for this “new” concept to the lexer, parser, AST, and LLVM code emitter. This example is nice, because it shows how easy it is to “grow” a language over time, incrementally extending it as new ideas are discovered.</p> <p>Before we get going on “how” we add this extension, lets talk about “what” we want. The basic idea is that we want to be able to write this sort of thing:</p> <div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">fib</span><span class="p">(</span><span class="n">x</span><span class="p">)</span> <span class="k">if</span> <span class="n">x</span> <span class="o"><</span> <span class="mi">3</span> <span class="n">then</span> <span class="mi">1</span> <span class="k">else</span> <span class="n">fib</span><span class="p">(</span><span class="n">x</span><span class="o">-</span><span class="mi">1</span><span class="p">)</span><span class="o">+</span><span class="n">fib</span><span class="p">(</span><span class="n">x</span><span class="o">-</span><span class="mi">2</span><span class="p">);</span> </pre></div> </div> <p>In Kaleidoscope, every construct is an expression: there are no statements. As such, the if/then/else expression needs to return a value like any other. Since we’re using a mostly functional form, we’ll have it evaluate its conditional, then return the ‘then’ or ‘else’ value based on how the condition was resolved. This is very similar to the C “?:” expression.</p> <p>The semantics of the if/then/else expression is that it evaluates the condition to a boolean equality value: 0.0 is considered to be false and everything else is considered to be true. If the condition is true, the first subexpression is evaluated and returned, if the condition is false, the second subexpression is evaluated and returned. Since Kaleidoscope allows side-effects, this behavior is important to nail down.</p> <p>Now that we know what we “want”, lets break this down into its constituent pieces.</p> <div class="section" id="lexer-extensions-for-if-then-else"> <h3><a class="toc-backref" href="#id4">5.2.1. Lexer Extensions for If/Then/Else</a><a class="headerlink" href="#lexer-extensions-for-if-then-else" title="Permalink to this headline">¶</a></h3> <p>The lexer extensions are straightforward. First we add new variants for the relevant tokens:</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* control *)</span> <span class="o">|</span> <span class="nc">If</span> <span class="o">|</span> <span class="nc">Then</span> <span class="o">|</span> <span class="nc">Else</span> <span class="o">|</span> <span class="nc">For</span> <span class="o">|</span> <span class="nc">In</span> </pre></div> </div> <p>Once we have that, we recognize the new keywords in the lexer. This is pretty simple stuff:</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="o">...</span> <span class="k">match</span> <span class="nn">Buffer</span><span class="p">.</span><span class="n">contents</span> <span class="n">buffer</span> <span class="k">with</span> <span class="o">|</span> <span class="s2">"def"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Def</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"extern"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Extern</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"if"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">If</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"then"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Then</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"else"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Else</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"for"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">For</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"in"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">In</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="n">id</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Ident</span> <span class="n">id</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> </pre></div> </div> </div> <div class="section" id="ast-extensions-for-if-then-else"> <h3><a class="toc-backref" href="#id5">5.2.2. AST Extensions for If/Then/Else</a><a class="headerlink" href="#ast-extensions-for-if-then-else" title="Permalink to this headline">¶</a></h3> <p>To represent the new expression we add a new AST variant for it:</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="k">type</span> <span class="n">expr</span> <span class="o">=</span> <span class="o">...</span> <span class="c">(* variant for if/then/else. *)</span> <span class="o">|</span> <span class="nc">If</span> <span class="k">of</span> <span class="n">expr</span> <span class="o">*</span> <span class="n">expr</span> <span class="o">*</span> <span class="n">expr</span> </pre></div> </div> <p>The AST variant just has pointers to the various subexpressions.</p> </div> <div class="section" id="parser-extensions-for-if-then-else"> <h3><a class="toc-backref" href="#id6">5.2.3. Parser Extensions for If/Then/Else</a><a class="headerlink" href="#parser-extensions-for-if-then-else" title="Permalink to this headline">¶</a></h3> <p>Now that we have the relevant tokens coming from the lexer and we have the AST node to build, our parsing logic is relatively straightforward. Next we add a new case for parsing a if-expression as a primary expression:</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="k">let</span> <span class="k">rec</span> <span class="n">parse_primary</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">...</span> <span class="c">(* ifexpr ::= 'if' expr 'then' expr 'else' expr *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">If</span><span class="o">;</span> <span class="n">c</span><span class="o">=</span><span class="n">parse_expr</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Then</span> <span class="o">??</span> <span class="s2">"expected 'then'"</span><span class="o">;</span> <span class="n">t</span><span class="o">=</span><span class="n">parse_expr</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Else</span> <span class="o">??</span> <span class="s2">"expected 'else'"</span><span class="o">;</span> <span class="n">e</span><span class="o">=</span><span class="n">parse_expr</span> <span class="o">>]</span> <span class="o">-></span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">If</span> <span class="o">(</span><span class="n">c</span><span class="o">,</span> <span class="n">t</span><span class="o">,</span> <span class="n">e</span><span class="o">)</span> </pre></div> </div> </div> <div class="section" id="llvm-ir-for-if-then-else"> <h3><a class="toc-backref" href="#id7">5.2.4. LLVM IR for If/Then/Else</a><a class="headerlink" href="#llvm-ir-for-if-then-else" title="Permalink to this headline">¶</a></h3> <p>Now that we have it parsing and building the AST, the final piece is adding LLVM code generation support. This is the most interesting part of the if/then/else example, because this is where it starts to introduce new concepts. All of the code above has been thoroughly described in previous chapters.</p> <p>To motivate the code we want to produce, lets take a look at a simple example. Consider:</p> <div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">extern</span> <span class="n">foo</span><span class="p">();</span> <span class="n">extern</span> <span class="n">bar</span><span class="p">();</span> <span class="k">def</span> <span class="nf">baz</span><span class="p">(</span><span class="n">x</span><span class="p">)</span> <span class="k">if</span> <span class="n">x</span> <span class="n">then</span> <span class="n">foo</span><span class="p">()</span> <span class="k">else</span> <span class="n">bar</span><span class="p">();</span> </pre></div> </div> <p>If you disable optimizations, the code you’ll (soon) get from Kaleidoscope looks like this:</p> <div class="highlight-llvm notranslate"><div class="highlight"><pre><span></span><span class="k">declare</span> <span class="kt">double</span> <span class="vg">@foo</span><span class="p">()</span> <span class="k">declare</span> <span class="kt">double</span> <span class="vg">@bar</span><span class="p">()</span> <span class="k">define</span> <span class="kt">double</span> <span class="vg">@baz</span><span class="p">(</span><span class="kt">double</span> <span class="nv">%x</span><span class="p">)</span> <span class="p">{</span> <span class="nl">entry:</span> <span class="nv">%ifcond</span> <span class="p">=</span> <span class="k">fcmp</span> <span class="k">one</span> <span class="kt">double</span> <span class="nv">%x</span><span class="p">,</span> <span class="m">0.000000e+00</span> <span class="k">br</span> <span class="k">i1</span> <span class="nv">%ifcond</span><span class="p">,</span> <span class="kt">label</span> <span class="nv">%then</span><span class="p">,</span> <span class="kt">label</span> <span class="nv">%else</span> <span class="nl">then:</span> <span class="c">; preds = %entry</span> <span class="nv">%calltmp</span> <span class="p">=</span> <span class="k">call</span> <span class="kt">double</span> <span class="vg">@foo</span><span class="p">()</span> <span class="k">br</span> <span class="kt">label</span> <span class="nv">%ifcont</span> <span class="nl">else:</span> <span class="c">; preds = %entry</span> <span class="nv">%calltmp1</span> <span class="p">=</span> <span class="k">call</span> <span class="kt">double</span> <span class="vg">@bar</span><span class="p">()</span> <span class="k">br</span> <span class="kt">label</span> <span class="nv">%ifcont</span> <span class="nl">ifcont:</span> <span class="c">; preds = %else, %then</span> <span class="nv">%iftmp</span> <span class="p">=</span> <span class="k">phi</span> <span class="kt">double</span> <span class="p">[</span> <span class="nv">%calltmp</span><span class="p">,</span> <span class="nv">%then</span> <span class="p">],</span> <span class="p">[</span> <span class="nv">%calltmp1</span><span class="p">,</span> <span class="nv">%else</span> <span class="p">]</span> <span class="k">ret</span> <span class="kt">double</span> <span class="nv">%iftmp</span> <span class="p">}</span> </pre></div> </div> <p>To visualize the control flow graph, you can use a nifty feature of the LLVM ‘<a class="reference external" href="http://llvm.org/cmds/opt.html">opt</a>’ tool. If you put this LLVM IR into “t.ll” and run “<code class="docutils literal notranslate"><span class="pre">llvm-as</span> <span class="pre"><</span> <span class="pre">t.ll</span> <span class="pre">|</span> <span class="pre">opt</span> <span class="pre">-analyze</span> <span class="pre">-view-cfg</span></code>”, <a class="reference external" href="../ProgrammersManual.html#viewing-graphs-while-debugging-code">a window will pop up</a> and you’ll see this graph:</p> <div class="figure align-center" id="id1"> <img alt="Example CFG" src="../_images/LangImpl05-cfg.png" /> <p class="caption"><span class="caption-text">Example CFG</span></p> </div> <p>Another way to get this is to call “<code class="docutils literal notranslate"><span class="pre">Llvm_analysis.view_function_cfg</span> <span class="pre">f</span></code>” or “<code class="docutils literal notranslate"><span class="pre">Llvm_analysis.view_function_cfg_only</span> <span class="pre">f</span></code>” (where <code class="docutils literal notranslate"><span class="pre">f</span></code> is a “<code class="docutils literal notranslate"><span class="pre">Function</span></code>”) either by inserting actual calls into the code and recompiling or by calling these in the debugger. LLVM has many nice features for visualizing various graphs.</p> <p>Getting back to the generated code, it is fairly simple: the entry block evaluates the conditional expression (“x” in our case here) and compares the result to 0.0 with the “<code class="docutils literal notranslate"><span class="pre">fcmp</span> <span class="pre">one</span></code>” instruction (‘one’ is “Ordered and Not Equal”). Based on the result of this expression, the code jumps to either the “then” or “else” blocks, which contain the expressions for the true/false cases.</p> <p>Once the then/else blocks are finished executing, they both branch back to the ‘ifcont’ block to execute the code that happens after the if/then/else. In this case the only thing left to do is to return to the caller of the function. The question then becomes: how does the code know which expression to return?</p> <p>The answer to this question involves an important SSA operation: the <a class="reference external" href="http://en.wikipedia.org/wiki/Static_single_assignment_form">Phi operation</a>. If you’re not familiar with SSA, <a class="reference external" href="http://en.wikipedia.org/wiki/Static_single_assignment_form">the wikipedia article</a> is a good introduction and there are various other introductions to it available on your favorite search engine. The short version is that “execution” of the Phi operation requires “remembering” which block control came from. The Phi operation takes on the value corresponding to the input control block. In this case, if control comes in from the “then” block, it gets the value of “calltmp”. If control comes from the “else” block, it gets the value of “calltmp1”.</p> <p>At this point, you are probably starting to think “Oh no! This means my simple and elegant front-end will have to start generating SSA form in order to use LLVM!”. Fortunately, this is not the case, and we strongly advise <em>not</em> implementing an SSA construction algorithm in your front-end unless there is an amazingly good reason to do so. In practice, there are two sorts of values that float around in code written for your average imperative programming language that might need Phi nodes:</p> <ol class="arabic simple"> <li>Code that involves user variables: <code class="docutils literal notranslate"><span class="pre">x</span> <span class="pre">=</span> <span class="pre">1;</span> <span class="pre">x</span> <span class="pre">=</span> <span class="pre">x</span> <span class="pre">+</span> <span class="pre">1;</span></code></li> <li>Values that are implicit in the structure of your AST, such as the Phi node in this case.</li> </ol> <p>In <a class="reference external" href="OCamlLangImpl7.html">Chapter 7</a> of this tutorial (“mutable variables”), we’ll talk about #1 in depth. For now, just believe me that you don’t need SSA construction to handle this case. For #2, you have the choice of using the techniques that we will describe for #1, or you can insert Phi nodes directly, if convenient. In this case, it is really really easy to generate the Phi node, so we choose to do it directly.</p> <p>Okay, enough of the motivation and overview, lets generate code!</p> </div> <div class="section" id="code-generation-for-if-then-else"> <h3><a class="toc-backref" href="#id8">5.2.5. Code Generation for If/Then/Else</a><a class="headerlink" href="#code-generation-for-if-then-else" title="Permalink to this headline">¶</a></h3> <p>In order to generate code for this, we implement the <code class="docutils literal notranslate"><span class="pre">Codegen</span></code> method for <code class="docutils literal notranslate"><span class="pre">IfExprAST</span></code>:</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="k">let</span> <span class="k">rec</span> <span class="n">codegen_expr</span> <span class="o">=</span> <span class="k">function</span> <span class="o">...</span> <span class="o">|</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">If</span> <span class="o">(</span><span class="n">cond</span><span class="o">,</span> <span class="n">then_</span><span class="o">,</span> <span class="n">else_</span><span class="o">)</span> <span class="o">-></span> <span class="k">let</span> <span class="n">cond</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">cond</span> <span class="k">in</span> <span class="c">(* Convert condition to a bool by comparing equal to 0.0 *)</span> <span class="k">let</span> <span class="n">zero</span> <span class="o">=</span> <span class="n">const_float</span> <span class="n">double_type</span> <span class="mi">0</span><span class="o">.</span><span class="mi">0</span> <span class="k">in</span> <span class="k">let</span> <span class="n">cond_val</span> <span class="o">=</span> <span class="n">build_fcmp</span> <span class="nn">Fcmp</span><span class="p">.</span><span class="nc">One</span> <span class="n">cond</span> <span class="n">zero</span> <span class="s2">"ifcond"</span> <span class="n">builder</span> <span class="k">in</span> </pre></div> </div> <p>This code is straightforward and similar to what we saw before. We emit the expression for the condition, then compare that value to zero to get a truth value as a 1-bit (bool) value.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Grab the first block so that we might later add the conditional branch</span> <span class="c"> * to it at the end of the function. *)</span> <span class="k">let</span> <span class="n">start_bb</span> <span class="o">=</span> <span class="n">insertion_block</span> <span class="n">builder</span> <span class="k">in</span> <span class="k">let</span> <span class="n">the_function</span> <span class="o">=</span> <span class="n">block_parent</span> <span class="n">start_bb</span> <span class="k">in</span> <span class="k">let</span> <span class="n">then_bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"then"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="n">position_at_end</span> <span class="n">then_bb</span> <span class="n">builder</span><span class="o">;</span> </pre></div> </div> <p>As opposed to the <a class="reference external" href="LangImpl05.html">C++ tutorial</a>, we have to build our basic blocks bottom up since we can’t have dangling BasicBlocks. We start off by saving a pointer to the first block (which might not be the entry block), which we’ll need to build a conditional branch later. We do this by asking the <code class="docutils literal notranslate"><span class="pre">builder</span></code> for the current BasicBlock. The fourth line gets the current Function object that is being built. It gets this by the <code class="docutils literal notranslate"><span class="pre">start_bb</span></code> for its “parent” (the function it is currently embedded into).</p> <p>Once it has that, it creates one block. It is automatically appended into the function’s list of blocks.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Emit 'then' value. *)</span> <span class="n">position_at_end</span> <span class="n">then_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="k">let</span> <span class="n">then_val</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">then_</span> <span class="k">in</span> <span class="c">(* Codegen of 'then' can change the current block, update then_bb for the</span> <span class="c"> * phi. We create a new name because one is used for the phi node, and the</span> <span class="c"> * other is used for the conditional branch. *)</span> <span class="k">let</span> <span class="n">new_then_bb</span> <span class="o">=</span> <span class="n">insertion_block</span> <span class="n">builder</span> <span class="k">in</span> </pre></div> </div> <p>We move the builder to start inserting into the “then” block. Strictly speaking, this call moves the insertion point to be at the end of the specified block. However, since the “then” block is empty, it also starts out by inserting at the beginning of the block. :)</p> <p>Once the insertion point is set, we recursively codegen the “then” expression from the AST.</p> <p>The final line here is quite subtle, but is very important. The basic issue is that when we create the Phi node in the merge block, we need to set up the block/value pairs that indicate how the Phi will work. Importantly, the Phi node expects to have an entry for each predecessor of the block in the CFG. Why then, are we getting the current block when we just set it to ThenBB 5 lines above? The problem is that the “Then” expression may actually itself change the block that the Builder is emitting into if, for example, it contains a nested “if/then/else” expression. Because calling Codegen recursively could arbitrarily change the notion of the current block, we are required to get an up-to-date value for code that will set up the Phi node.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Emit 'else' value. *)</span> <span class="k">let</span> <span class="n">else_bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"else"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="n">position_at_end</span> <span class="n">else_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="k">let</span> <span class="n">else_val</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">else_</span> <span class="k">in</span> <span class="c">(* Codegen of 'else' can change the current block, update else_bb for the</span> <span class="c"> * phi. *)</span> <span class="k">let</span> <span class="n">new_else_bb</span> <span class="o">=</span> <span class="n">insertion_block</span> <span class="n">builder</span> <span class="k">in</span> </pre></div> </div> <p>Code generation for the ‘else’ block is basically identical to codegen for the ‘then’ block.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Emit merge block. *)</span> <span class="k">let</span> <span class="n">merge_bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"ifcont"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="n">position_at_end</span> <span class="n">merge_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="k">let</span> <span class="n">incoming</span> <span class="o">=</span> <span class="o">[(</span><span class="n">then_val</span><span class="o">,</span> <span class="n">new_then_bb</span><span class="o">);</span> <span class="o">(</span><span class="n">else_val</span><span class="o">,</span> <span class="n">new_else_bb</span><span class="o">)]</span> <span class="k">in</span> <span class="k">let</span> <span class="n">phi</span> <span class="o">=</span> <span class="n">build_phi</span> <span class="n">incoming</span> <span class="s2">"iftmp"</span> <span class="n">builder</span> <span class="k">in</span> </pre></div> </div> <p>The first two lines here are now familiar: the first adds the “merge” block to the Function object. The second changes the insertion point so that newly created code will go into the “merge” block. Once that is done, we need to create the PHI node and set up the block/value pairs for the PHI.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Return to the start block to add the conditional branch. *)</span> <span class="n">position_at_end</span> <span class="n">start_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">build_cond_br</span> <span class="n">cond_val</span> <span class="n">then_bb</span> <span class="n">else_bb</span> <span class="n">builder</span><span class="o">);</span> </pre></div> </div> <p>Once the blocks are created, we can emit the conditional branch that chooses between them. Note that creating new blocks does not implicitly affect the IRBuilder, so it is still inserting into the block that the condition went into. This is why we needed to save the “start” block.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Set a unconditional branch at the end of the 'then' block and the</span> <span class="c"> * 'else' block to the 'merge' block. *)</span> <span class="n">position_at_end</span> <span class="n">new_then_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">build_br</span> <span class="n">merge_bb</span> <span class="n">builder</span><span class="o">);</span> <span class="n">position_at_end</span> <span class="n">new_else_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">build_br</span> <span class="n">merge_bb</span> <span class="n">builder</span><span class="o">);</span> <span class="c">(* Finally, set the builder to the end of the merge block. *)</span> <span class="n">position_at_end</span> <span class="n">merge_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="n">phi</span> </pre></div> </div> <p>To finish off the blocks, we create an unconditional branch to the merge block. One interesting (and very important) aspect of the LLVM IR is that it <a class="reference external" href="../LangRef.html#functionstructure">requires all basic blocks to be “terminated”</a> with a <a class="reference external" href="../LangRef.html#terminators">control flow instruction</a> such as return or branch. This means that all control flow, <em>including fall throughs</em> must be made explicit in the LLVM IR. If you violate this rule, the verifier will emit an error.</p> <p>Finally, the CodeGen function returns the phi node as the value computed by the if/then/else expression. In our example above, this returned value will feed into the code for the top-level function, which will create the return instruction.</p> <p>Overall, we now have the ability to execute conditional code in Kaleidoscope. With this extension, Kaleidoscope is a fairly complete language that can calculate a wide variety of numeric functions. Next up we’ll add another useful expression that is familiar from non-functional languages…</p> </div> </div> <div class="section" id="for-loop-expression"> <h2><a class="toc-backref" href="#id9">5.3. ‘for’ Loop Expression</a><a class="headerlink" href="#for-loop-expression" title="Permalink to this headline">¶</a></h2> <p>Now that we know how to add basic control flow constructs to the language, we have the tools to add more powerful things. Lets add something more aggressive, a ‘for’ expression:</p> <div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">extern</span> <span class="n">putchard</span><span class="p">(</span><span class="n">char</span><span class="p">);</span> <span class="k">def</span> <span class="nf">printstar</span><span class="p">(</span><span class="n">n</span><span class="p">)</span> <span class="k">for</span> <span class="n">i</span> <span class="o">=</span> <span class="mi">1</span><span class="p">,</span> <span class="n">i</span> <span class="o"><</span> <span class="n">n</span><span class="p">,</span> <span class="mf">1.0</span> <span class="ow">in</span> <span class="n">putchard</span><span class="p">(</span><span class="mi">42</span><span class="p">);</span> <span class="c1"># ascii 42 = '*'</span> <span class="c1"># print 100 '*' characters</span> <span class="n">printstar</span><span class="p">(</span><span class="mi">100</span><span class="p">);</span> </pre></div> </div> <p>This expression defines a new variable (“i” in this case) which iterates from a starting value, while the condition (“i < n” in this case) is true, incrementing by an optional step value (“1.0” in this case). If the step value is omitted, it defaults to 1.0. While the loop is true, it executes its body expression. Because we don’t have anything better to return, we’ll just define the loop as always returning 0.0. In the future when we have mutable variables, it will get more useful.</p> <p>As before, lets talk about the changes that we need to Kaleidoscope to support this.</p> <div class="section" id="lexer-extensions-for-the-for-loop"> <h3><a class="toc-backref" href="#id10">5.3.1. Lexer Extensions for the ‘for’ Loop</a><a class="headerlink" href="#lexer-extensions-for-the-for-loop" title="Permalink to this headline">¶</a></h3> <p>The lexer extensions are the same sort of thing as for if/then/else:</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="o">...</span> <span class="k">in</span> <span class="nn">Token</span><span class="p">.</span><span class="n">token</span> <span class="o">...</span> <span class="c">(* control *)</span> <span class="o">|</span> <span class="nc">If</span> <span class="o">|</span> <span class="nc">Then</span> <span class="o">|</span> <span class="nc">Else</span> <span class="o">|</span> <span class="nc">For</span> <span class="o">|</span> <span class="nn">In</span> <span class="p">...</span> <span class="n">in</span> <span class="nn">Lexer</span><span class="p">.</span><span class="n">lex_ident</span><span class="o">...</span> <span class="k">match</span> <span class="nn">Buffer</span><span class="p">.</span><span class="n">contents</span> <span class="n">buffer</span> <span class="k">with</span> <span class="o">|</span> <span class="s2">"def"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Def</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"extern"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Extern</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"if"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">If</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"then"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Then</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"else"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Else</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"for"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">For</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"in"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">In</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="n">id</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Ident</span> <span class="n">id</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> </pre></div> </div> </div> <div class="section" id="ast-extensions-for-the-for-loop"> <h3><a class="toc-backref" href="#id11">5.3.2. AST Extensions for the ‘for’ Loop</a><a class="headerlink" href="#ast-extensions-for-the-for-loop" title="Permalink to this headline">¶</a></h3> <p>The AST variant is just as simple. It basically boils down to capturing the variable name and the constituent expressions in the node.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="k">type</span> <span class="n">expr</span> <span class="o">=</span> <span class="o">...</span> <span class="c">(* variant for for/in. *)</span> <span class="o">|</span> <span class="nc">For</span> <span class="k">of</span> <span class="kt">string</span> <span class="o">*</span> <span class="n">expr</span> <span class="o">*</span> <span class="n">expr</span> <span class="o">*</span> <span class="n">expr</span> <span class="n">option</span> <span class="o">*</span> <span class="n">expr</span> </pre></div> </div> </div> <div class="section" id="parser-extensions-for-the-for-loop"> <h3><a class="toc-backref" href="#id12">5.3.3. Parser Extensions for the ‘for’ Loop</a><a class="headerlink" href="#parser-extensions-for-the-for-loop" title="Permalink to this headline">¶</a></h3> <p>The parser code is also fairly standard. The only interesting thing here is handling of the optional step value. The parser code handles it by checking to see if the second comma is present. If not, it sets the step value to null in the AST node:</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="k">let</span> <span class="k">rec</span> <span class="n">parse_primary</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">...</span> <span class="c">(* forexpr</span> <span class="c"> ::= 'for' identifier '=' expr ',' expr (',' expr)? 'in' expression *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">For</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Ident</span> <span class="n">id</span> <span class="o">??</span> <span class="s2">"expected identifier after for"</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">'='</span> <span class="o">??</span> <span class="s2">"expected '=' after for"</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">begin</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="n">start</span><span class="o">=</span><span class="n">parse_expr</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">','</span> <span class="o">??</span> <span class="s2">"expected ',' after for"</span><span class="o">;</span> <span class="n">end_</span><span class="o">=</span><span class="n">parse_expr</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">let</span> <span class="n">step</span> <span class="o">=</span> <span class="k">begin</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">','</span><span class="o">;</span> <span class="n">step</span><span class="o">=</span><span class="n">parse_expr</span> <span class="o">>]</span> <span class="o">-></span> <span class="nc">Some</span> <span class="n">step</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="nc">None</span> <span class="k">end</span> <span class="n">stream</span> <span class="k">in</span> <span class="k">begin</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">In</span><span class="o">;</span> <span class="n">body</span><span class="o">=</span><span class="n">parse_expr</span> <span class="o">>]</span> <span class="o">-></span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">For</span> <span class="o">(</span><span class="n">id</span><span class="o">,</span> <span class="n">start</span><span class="o">,</span> <span class="n">end_</span><span class="o">,</span> <span class="n">step</span><span class="o">,</span> <span class="n">body</span><span class="o">)</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">raise</span> <span class="o">(</span><span class="nn">Stream</span><span class="p">.</span><span class="nc">Error</span> <span class="s2">"expected 'in' after for"</span><span class="o">)</span> <span class="k">end</span> <span class="n">stream</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">raise</span> <span class="o">(</span><span class="nn">Stream</span><span class="p">.</span><span class="nc">Error</span> <span class="s2">"expected '=' after for"</span><span class="o">)</span> <span class="k">end</span> <span class="n">stream</span> </pre></div> </div> </div> <div class="section" id="llvm-ir-for-the-for-loop"> <h3><a class="toc-backref" href="#id13">5.3.4. LLVM IR for the ‘for’ Loop</a><a class="headerlink" href="#llvm-ir-for-the-for-loop" title="Permalink to this headline">¶</a></h3> <p>Now we get to the good part: the LLVM IR we want to generate for this thing. With the simple example above, we get this LLVM IR (note that this dump is generated with optimizations disabled for clarity):</p> <div class="highlight-llvm notranslate"><div class="highlight"><pre><span></span><span class="k">declare</span> <span class="kt">double</span> <span class="vg">@putchard</span><span class="p">(</span><span class="kt">double</span><span class="p">)</span> <span class="k">define</span> <span class="kt">double</span> <span class="vg">@printstar</span><span class="p">(</span><span class="kt">double</span> <span class="nv">%n</span><span class="p">)</span> <span class="p">{</span> <span class="nl">entry:</span> <span class="c">; initial value = 1.0 (inlined into phi)</span> <span class="k">br</span> <span class="kt">label</span> <span class="nv">%loop</span> <span class="nl">loop:</span> <span class="c">; preds = %loop, %entry</span> <span class="nv">%i</span> <span class="p">=</span> <span class="k">phi</span> <span class="kt">double</span> <span class="p">[</span> <span class="m">1.000000e+00</span><span class="p">,</span> <span class="nv">%entry</span> <span class="p">],</span> <span class="p">[</span> <span class="nv">%nextvar</span><span class="p">,</span> <span class="nv">%loop</span> <span class="p">]</span> <span class="c">; body</span> <span class="nv">%calltmp</span> <span class="p">=</span> <span class="k">call</span> <span class="kt">double</span> <span class="vg">@putchard</span><span class="p">(</span><span class="kt">double</span> <span class="m">4.200000e+01</span><span class="p">)</span> <span class="c">; increment</span> <span class="nv">%nextvar</span> <span class="p">=</span> <span class="k">fadd</span> <span class="kt">double</span> <span class="nv">%i</span><span class="p">,</span> <span class="m">1.000000e+00</span> <span class="c">; termination test</span> <span class="nv">%cmptmp</span> <span class="p">=</span> <span class="k">fcmp</span> <span class="k">ult</span> <span class="kt">double</span> <span class="nv">%i</span><span class="p">,</span> <span class="nv">%n</span> <span class="nv">%booltmp</span> <span class="p">=</span> <span class="k">uitofp</span> <span class="k">i1</span> <span class="nv">%cmptmp</span> <span class="k">to</span> <span class="kt">double</span> <span class="nv">%loopcond</span> <span class="p">=</span> <span class="k">fcmp</span> <span class="k">one</span> <span class="kt">double</span> <span class="nv">%booltmp</span><span class="p">,</span> <span class="m">0.000000e+00</span> <span class="k">br</span> <span class="k">i1</span> <span class="nv">%loopcond</span><span class="p">,</span> <span class="kt">label</span> <span class="nv">%loop</span><span class="p">,</span> <span class="kt">label</span> <span class="nv">%afterloop</span> <span class="nl">afterloop:</span> <span class="c">; preds = %loop</span> <span class="c">; loop always returns 0.0</span> <span class="k">ret</span> <span class="kt">double</span> <span class="m">0.000000e+00</span> <span class="p">}</span> </pre></div> </div> <p>This loop contains all the same constructs we saw before: a phi node, several expressions, and some basic blocks. Lets see how this fits together.</p> </div> <div class="section" id="code-generation-for-the-for-loop"> <h3><a class="toc-backref" href="#id14">5.3.5. Code Generation for the ‘for’ Loop</a><a class="headerlink" href="#code-generation-for-the-for-loop" title="Permalink to this headline">¶</a></h3> <p>The first part of Codegen is very simple: we just output the start expression for the loop value:</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="k">let</span> <span class="k">rec</span> <span class="n">codegen_expr</span> <span class="o">=</span> <span class="k">function</span> <span class="o">...</span> <span class="o">|</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">For</span> <span class="o">(</span><span class="n">var_name</span><span class="o">,</span> <span class="n">start</span><span class="o">,</span> <span class="n">end_</span><span class="o">,</span> <span class="n">step</span><span class="o">,</span> <span class="n">body</span><span class="o">)</span> <span class="o">-></span> <span class="c">(* Emit the start code first, without 'variable' in scope. *)</span> <span class="k">let</span> <span class="n">start_val</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">start</span> <span class="k">in</span> </pre></div> </div> <p>With this out of the way, the next step is to set up the LLVM basic block for the start of the loop body. In the case above, the whole loop body is one block, but remember that the body code itself could consist of multiple blocks (e.g. if it contains an if/then/else or a for/in expression).</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Make the new basic block for the loop header, inserting after current</span> <span class="c"> * block. *)</span> <span class="k">let</span> <span class="n">preheader_bb</span> <span class="o">=</span> <span class="n">insertion_block</span> <span class="n">builder</span> <span class="k">in</span> <span class="k">let</span> <span class="n">the_function</span> <span class="o">=</span> <span class="n">block_parent</span> <span class="n">preheader_bb</span> <span class="k">in</span> <span class="k">let</span> <span class="n">loop_bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"loop"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="c">(* Insert an explicit fall through from the current block to the</span> <span class="c"> * loop_bb. *)</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">build_br</span> <span class="n">loop_bb</span> <span class="n">builder</span><span class="o">);</span> </pre></div> </div> <p>This code is similar to what we saw for if/then/else. Because we will need it to create the Phi node, we remember the block that falls through into the loop. Once we have that, we create the actual block that starts the loop and create an unconditional branch for the fall-through between the two blocks.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Start insertion in loop_bb. *)</span> <span class="n">position_at_end</span> <span class="n">loop_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="c">(* Start the PHI node with an entry for start. *)</span> <span class="k">let</span> <span class="n">variable</span> <span class="o">=</span> <span class="n">build_phi</span> <span class="o">[(</span><span class="n">start_val</span><span class="o">,</span> <span class="n">preheader_bb</span><span class="o">)]</span> <span class="n">var_name</span> <span class="n">builder</span> <span class="k">in</span> </pre></div> </div> <p>Now that the “preheader” for the loop is set up, we switch to emitting code for the loop body. To begin with, we move the insertion point and create the PHI node for the loop induction variable. Since we already know the incoming value for the starting value, we add it to the Phi node. Note that the Phi will eventually get a second value for the backedge, but we can’t set it up yet (because it doesn’t exist!).</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Within the loop, the variable is defined equal to the PHI node. If it</span> <span class="c"> * shadows an existing variable, we have to restore it, so save it</span> <span class="c"> * now. *)</span> <span class="k">let</span> <span class="n">old_val</span> <span class="o">=</span> <span class="k">try</span> <span class="nc">Some</span> <span class="o">(</span><span class="nn">Hashtbl</span><span class="p">.</span><span class="n">find</span> <span class="n">named_values</span> <span class="n">var_name</span><span class="o">)</span> <span class="k">with</span> <span class="nc">Not_found</span> <span class="o">-></span> <span class="nc">None</span> <span class="k">in</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">add</span> <span class="n">named_values</span> <span class="n">var_name</span> <span class="n">variable</span><span class="o">;</span> <span class="c">(* Emit the body of the loop. This, like any other expr, can change the</span> <span class="c"> * current BB. Note that we ignore the value computed by the body, but</span> <span class="c"> * don't allow an error *)</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">codegen_expr</span> <span class="n">body</span><span class="o">);</span> </pre></div> </div> <p>Now the code starts to get more interesting. Our ‘for’ loop introduces a new variable to the symbol table. This means that our symbol table can now contain either function arguments or loop variables. To handle this, before we codegen the body of the loop, we add the loop variable as the current value for its name. Note that it is possible that there is a variable of the same name in the outer scope. It would be easy to make this an error (emit an error and return null if there is already an entry for VarName) but we choose to allow shadowing of variables. In order to handle this correctly, we remember the Value that we are potentially shadowing in <code class="docutils literal notranslate"><span class="pre">old_val</span></code> (which will be None if there is no shadowed variable).</p> <p>Once the loop variable is set into the symbol table, the code recursively codegen’s the body. This allows the body to use the loop variable: any references to it will naturally find it in the symbol table.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Emit the step value. *)</span> <span class="k">let</span> <span class="n">step_val</span> <span class="o">=</span> <span class="k">match</span> <span class="n">step</span> <span class="k">with</span> <span class="o">|</span> <span class="nc">Some</span> <span class="n">step</span> <span class="o">-></span> <span class="n">codegen_expr</span> <span class="n">step</span> <span class="c">(* If not specified, use 1.0. *)</span> <span class="o">|</span> <span class="nc">None</span> <span class="o">-></span> <span class="n">const_float</span> <span class="n">double_type</span> <span class="mi">1</span><span class="o">.</span><span class="mi">0</span> <span class="k">in</span> <span class="k">let</span> <span class="n">next_var</span> <span class="o">=</span> <span class="n">build_add</span> <span class="n">variable</span> <span class="n">step_val</span> <span class="s2">"nextvar"</span> <span class="n">builder</span> <span class="k">in</span> </pre></div> </div> <p>Now that the body is emitted, we compute the next value of the iteration variable by adding the step value, or 1.0 if it isn’t present. ‘<code class="docutils literal notranslate"><span class="pre">next_var</span></code>’ will be the value of the loop variable on the next iteration of the loop.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Compute the end condition. *)</span> <span class="k">let</span> <span class="n">end_cond</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">end_</span> <span class="k">in</span> <span class="c">(* Convert condition to a bool by comparing equal to 0.0. *)</span> <span class="k">let</span> <span class="n">zero</span> <span class="o">=</span> <span class="n">const_float</span> <span class="n">double_type</span> <span class="mi">0</span><span class="o">.</span><span class="mi">0</span> <span class="k">in</span> <span class="k">let</span> <span class="n">end_cond</span> <span class="o">=</span> <span class="n">build_fcmp</span> <span class="nn">Fcmp</span><span class="p">.</span><span class="nc">One</span> <span class="n">end_cond</span> <span class="n">zero</span> <span class="s2">"loopcond"</span> <span class="n">builder</span> <span class="k">in</span> </pre></div> </div> <p>Finally, we evaluate the exit value of the loop, to determine whether the loop should exit. This mirrors the condition evaluation for the if/then/else statement.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Create the "after loop" block and insert it. *)</span> <span class="k">let</span> <span class="n">loop_end_bb</span> <span class="o">=</span> <span class="n">insertion_block</span> <span class="n">builder</span> <span class="k">in</span> <span class="k">let</span> <span class="n">after_bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"afterloop"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="c">(* Insert the conditional branch into the end of loop_end_bb. *)</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">build_cond_br</span> <span class="n">end_cond</span> <span class="n">loop_bb</span> <span class="n">after_bb</span> <span class="n">builder</span><span class="o">);</span> <span class="c">(* Any new code will be inserted in after_bb. *)</span> <span class="n">position_at_end</span> <span class="n">after_bb</span> <span class="n">builder</span><span class="o">;</span> </pre></div> </div> <p>With the code for the body of the loop complete, we just need to finish up the control flow for it. This code remembers the end block (for the phi node), then creates the block for the loop exit (“afterloop”). Based on the value of the exit condition, it creates a conditional branch that chooses between executing the loop again and exiting the loop. Any future code is emitted in the “afterloop” block, so it sets the insertion position to it.</p> <div class="highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(* Add a new entry to the PHI node for the backedge. *)</span> <span class="n">add_incoming</span> <span class="o">(</span><span class="n">next_var</span><span class="o">,</span> <span class="n">loop_end_bb</span><span class="o">)</span> <span class="n">variable</span><span class="o">;</span> <span class="c">(* Restore the unshadowed variable. *)</span> <span class="k">begin</span> <span class="k">match</span> <span class="n">old_val</span> <span class="k">with</span> <span class="o">|</span> <span class="nc">Some</span> <span class="n">old_val</span> <span class="o">-></span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">add</span> <span class="n">named_values</span> <span class="n">var_name</span> <span class="n">old_val</span> <span class="o">|</span> <span class="nc">None</span> <span class="o">-></span> <span class="bp">()</span> <span class="k">end</span><span class="o">;</span> <span class="c">(* for expr always returns 0.0. *)</span> <span class="n">const_null</span> <span class="n">double_type</span> </pre></div> </div> <p>The final code handles various cleanups: now that we have the “<code class="docutils literal notranslate"><span class="pre">next_var</span></code>” value, we can add the incoming value to the loop PHI node. After that, we remove the loop variable from the symbol table, so that it isn’t in scope after the for loop. Finally, code generation of the for loop always returns 0.0, so that is what we return from <code class="docutils literal notranslate"><span class="pre">Codegen.codegen_expr</span></code>.</p> <p>With this, we conclude the “adding control flow to Kaleidoscope” chapter of the tutorial. In this chapter we added two control flow constructs, and used them to motivate a couple of aspects of the LLVM IR that are important for front-end implementors to know. In the next chapter of our saga, we will get a bit crazier and add <a class="reference external" href="OCamlLangImpl6.html">user-defined operators</a> to our poor innocent language.</p> </div> </div> <div class="section" id="full-code-listing"> <h2><a class="toc-backref" href="#id15">5.4. Full Code Listing</a><a class="headerlink" href="#full-code-listing" title="Permalink to this headline">¶</a></h2> <p>Here is the complete code listing for our running example, enhanced with the if/then/else and for expressions.. To build this example, use:</p> <div class="highlight-bash notranslate"><div class="highlight"><pre><span></span><span class="c1"># Compile</span> ocamlbuild toy.byte <span class="c1"># Run</span> ./toy.byte </pre></div> </div> <p>Here is the code:</p> <dl class="docutils"> <dt>_tags:</dt> <dd><div class="first last highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o"><</span><span class="p">{</span><span class="n">lexer</span><span class="p">,</span><span class="n">parser</span><span class="p">}</span><span class="o">.</span><span class="n">ml</span><span class="o">></span><span class="p">:</span> <span class="n">use_camlp4</span><span class="p">,</span> <span class="n">pp</span><span class="p">(</span><span class="n">camlp4of</span><span class="p">)</span> <span class="o"><*.</span><span class="p">{</span><span class="n">byte</span><span class="p">,</span><span class="n">native</span><span class="p">}</span><span class="o">></span><span class="p">:</span> <span class="n">g</span><span class="o">++</span><span class="p">,</span> <span class="n">use_llvm</span><span class="p">,</span> <span class="n">use_llvm_analysis</span> <span class="o"><*.</span><span class="p">{</span><span class="n">byte</span><span class="p">,</span><span class="n">native</span><span class="p">}</span><span class="o">></span><span class="p">:</span> <span class="n">use_llvm_executionengine</span><span class="p">,</span> <span class="n">use_llvm_target</span> <span class="o"><*.</span><span class="p">{</span><span class="n">byte</span><span class="p">,</span><span class="n">native</span><span class="p">}</span><span class="o">></span><span class="p">:</span> <span class="n">use_llvm_scalar_opts</span><span class="p">,</span> <span class="n">use_bindings</span> </pre></div> </div> </dd> <dt>myocamlbuild.ml:</dt> <dd><div class="first last highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="k">open</span> <span class="nc">Ocamlbuild_plugin</span><span class="o">;;</span> <span class="n">ocaml_lib</span> <span class="o">~</span><span class="n">extern</span><span class="o">:</span><span class="bp">true</span> <span class="s2">"llvm"</span><span class="o">;;</span> <span class="n">ocaml_lib</span> <span class="o">~</span><span class="n">extern</span><span class="o">:</span><span class="bp">true</span> <span class="s2">"llvm_analysis"</span><span class="o">;;</span> <span class="n">ocaml_lib</span> <span class="o">~</span><span class="n">extern</span><span class="o">:</span><span class="bp">true</span> <span class="s2">"llvm_executionengine"</span><span class="o">;;</span> <span class="n">ocaml_lib</span> <span class="o">~</span><span class="n">extern</span><span class="o">:</span><span class="bp">true</span> <span class="s2">"llvm_target"</span><span class="o">;;</span> <span class="n">ocaml_lib</span> <span class="o">~</span><span class="n">extern</span><span class="o">:</span><span class="bp">true</span> <span class="s2">"llvm_scalar_opts"</span><span class="o">;;</span> <span class="n">flag</span> <span class="o">[</span><span class="s2">"link"</span><span class="o">;</span> <span class="s2">"ocaml"</span><span class="o">;</span> <span class="s2">"g++"</span><span class="o">]</span> <span class="o">(</span><span class="nc">S</span><span class="o">[</span><span class="nc">A</span><span class="s2">"-cc"</span><span class="o">;</span> <span class="nc">A</span><span class="s2">"g++"</span><span class="o">]);;</span> <span class="n">dep</span> <span class="o">[</span><span class="s2">"link"</span><span class="o">;</span> <span class="s2">"ocaml"</span><span class="o">;</span> <span class="s2">"use_bindings"</span><span class="o">]</span> <span class="o">[</span><span class="s2">"bindings.o"</span><span class="o">];;</span> </pre></div> </div> </dd> <dt>token.ml:</dt> <dd><div class="first last highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(*===----------------------------------------------------------------------===</span> <span class="c"> * Lexer Tokens</span> <span class="c"> *===----------------------------------------------------------------------===*)</span> <span class="c">(* The lexer returns these 'Kwd' if it is an unknown character, otherwise one of</span> <span class="c"> * these others for known things. *)</span> <span class="k">type</span> <span class="n">token</span> <span class="o">=</span> <span class="c">(* commands *)</span> <span class="o">|</span> <span class="nc">Def</span> <span class="o">|</span> <span class="nc">Extern</span> <span class="c">(* primary *)</span> <span class="o">|</span> <span class="nc">Ident</span> <span class="k">of</span> <span class="kt">string</span> <span class="o">|</span> <span class="nc">Number</span> <span class="k">of</span> <span class="kt">float</span> <span class="c">(* unknown *)</span> <span class="o">|</span> <span class="nc">Kwd</span> <span class="k">of</span> <span class="kt">char</span> <span class="c">(* control *)</span> <span class="o">|</span> <span class="nc">If</span> <span class="o">|</span> <span class="nc">Then</span> <span class="o">|</span> <span class="nc">Else</span> <span class="o">|</span> <span class="nc">For</span> <span class="o">|</span> <span class="nc">In</span> </pre></div> </div> </dd> <dt>lexer.ml:</dt> <dd><div class="first last highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(*===----------------------------------------------------------------------===</span> <span class="c"> * Lexer</span> <span class="c"> *===----------------------------------------------------------------------===*)</span> <span class="k">let</span> <span class="k">rec</span> <span class="n">lex</span> <span class="o">=</span> <span class="n">parser</span> <span class="c">(* Skip any whitespace. *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span> <span class="o">(</span><span class="sc">' '</span> <span class="o">|</span> <span class="sc">'\n'</span> <span class="o">|</span> <span class="sc">'\r'</span> <span class="o">|</span> <span class="sc">'\t'</span><span class="o">);</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">lex</span> <span class="n">stream</span> <span class="c">(* identifier: [a-zA-Z][a-zA-Z0-9] *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span> <span class="o">(</span><span class="sc">'A'</span> <span class="o">..</span> <span class="sc">'Z'</span> <span class="o">|</span> <span class="sc">'a'</span> <span class="o">..</span> <span class="sc">'z'</span> <span class="k">as</span> <span class="n">c</span><span class="o">);</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">let</span> <span class="n">buffer</span> <span class="o">=</span> <span class="nn">Buffer</span><span class="p">.</span><span class="n">create</span> <span class="mi">1</span> <span class="k">in</span> <span class="nn">Buffer</span><span class="p">.</span><span class="n">add_char</span> <span class="n">buffer</span> <span class="n">c</span><span class="o">;</span> <span class="n">lex_ident</span> <span class="n">buffer</span> <span class="n">stream</span> <span class="c">(* number: [0-9.]+ *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span> <span class="o">(</span><span class="sc">'0'</span> <span class="o">..</span> <span class="sc">'9'</span> <span class="k">as</span> <span class="n">c</span><span class="o">);</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">let</span> <span class="n">buffer</span> <span class="o">=</span> <span class="nn">Buffer</span><span class="p">.</span><span class="n">create</span> <span class="mi">1</span> <span class="k">in</span> <span class="nn">Buffer</span><span class="p">.</span><span class="n">add_char</span> <span class="n">buffer</span> <span class="n">c</span><span class="o">;</span> <span class="n">lex_number</span> <span class="n">buffer</span> <span class="n">stream</span> <span class="c">(* Comment until end of line. *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span> <span class="o">(</span><span class="sc">'#'</span><span class="o">);</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">lex_comment</span> <span class="n">stream</span> <span class="c">(* Otherwise, just return the character as its ascii value. *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="n">c</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="n">c</span><span class="o">;</span> <span class="n">lex</span> <span class="n">stream</span> <span class="o">>]</span> <span class="c">(* end of stream. *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="o">[<</span> <span class="o">>]</span> <span class="ow">and</span> <span class="n">lex_number</span> <span class="n">buffer</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span> <span class="o">(</span><span class="sc">'0'</span> <span class="o">..</span> <span class="sc">'9'</span> <span class="o">|</span> <span class="sc">'.'</span> <span class="k">as</span> <span class="n">c</span><span class="o">);</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="nn">Buffer</span><span class="p">.</span><span class="n">add_char</span> <span class="n">buffer</span> <span class="n">c</span><span class="o">;</span> <span class="n">lex_number</span> <span class="n">buffer</span> <span class="n">stream</span> <span class="o">|</span> <span class="o">[<</span> <span class="n">stream</span><span class="o">=</span><span class="n">lex</span> <span class="o">>]</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Number</span> <span class="o">(</span><span class="n">float_of_string</span> <span class="o">(</span><span class="nn">Buffer</span><span class="p">.</span><span class="n">contents</span> <span class="n">buffer</span><span class="o">));</span> <span class="n">stream</span> <span class="o">>]</span> <span class="ow">and</span> <span class="n">lex_ident</span> <span class="n">buffer</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span> <span class="o">(</span><span class="sc">'A'</span> <span class="o">..</span> <span class="sc">'Z'</span> <span class="o">|</span> <span class="sc">'a'</span> <span class="o">..</span> <span class="sc">'z'</span> <span class="o">|</span> <span class="sc">'0'</span> <span class="o">..</span> <span class="sc">'9'</span> <span class="k">as</span> <span class="n">c</span><span class="o">);</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="nn">Buffer</span><span class="p">.</span><span class="n">add_char</span> <span class="n">buffer</span> <span class="n">c</span><span class="o">;</span> <span class="n">lex_ident</span> <span class="n">buffer</span> <span class="n">stream</span> <span class="o">|</span> <span class="o">[<</span> <span class="n">stream</span><span class="o">=</span><span class="n">lex</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">match</span> <span class="nn">Buffer</span><span class="p">.</span><span class="n">contents</span> <span class="n">buffer</span> <span class="k">with</span> <span class="o">|</span> <span class="s2">"def"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Def</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"extern"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Extern</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"if"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">If</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"then"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Then</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"else"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Else</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"for"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">For</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="s2">"in"</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">In</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">|</span> <span class="n">id</span> <span class="o">-></span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Ident</span> <span class="n">id</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="ow">and</span> <span class="n">lex_comment</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span> <span class="o">(</span><span class="sc">'\n'</span><span class="o">);</span> <span class="n">stream</span><span class="o">=</span><span class="n">lex</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">stream</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="n">c</span><span class="o">;</span> <span class="n">e</span><span class="o">=</span><span class="n">lex_comment</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">e</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="o">[<</span> <span class="o">>]</span> </pre></div> </div> </dd> <dt>ast.ml:</dt> <dd><div class="first last highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(*===----------------------------------------------------------------------===</span> <span class="c"> * Abstract Syntax Tree (aka Parse Tree)</span> <span class="c"> *===----------------------------------------------------------------------===*)</span> <span class="c">(* expr - Base type for all expression nodes. *)</span> <span class="k">type</span> <span class="n">expr</span> <span class="o">=</span> <span class="c">(* variant for numeric literals like "1.0". *)</span> <span class="o">|</span> <span class="nc">Number</span> <span class="k">of</span> <span class="kt">float</span> <span class="c">(* variant for referencing a variable, like "a". *)</span> <span class="o">|</span> <span class="nc">Variable</span> <span class="k">of</span> <span class="kt">string</span> <span class="c">(* variant for a binary operator. *)</span> <span class="o">|</span> <span class="nc">Binary</span> <span class="k">of</span> <span class="kt">char</span> <span class="o">*</span> <span class="n">expr</span> <span class="o">*</span> <span class="n">expr</span> <span class="c">(* variant for function calls. *)</span> <span class="o">|</span> <span class="nc">Call</span> <span class="k">of</span> <span class="kt">string</span> <span class="o">*</span> <span class="n">expr</span> <span class="kt">array</span> <span class="c">(* variant for if/then/else. *)</span> <span class="o">|</span> <span class="nc">If</span> <span class="k">of</span> <span class="n">expr</span> <span class="o">*</span> <span class="n">expr</span> <span class="o">*</span> <span class="n">expr</span> <span class="c">(* variant for for/in. *)</span> <span class="o">|</span> <span class="nc">For</span> <span class="k">of</span> <span class="kt">string</span> <span class="o">*</span> <span class="n">expr</span> <span class="o">*</span> <span class="n">expr</span> <span class="o">*</span> <span class="n">expr</span> <span class="n">option</span> <span class="o">*</span> <span class="n">expr</span> <span class="c">(* proto - This type represents the "prototype" for a function, which captures</span> <span class="c"> * its name, and its argument names (thus implicitly the number of arguments the</span> <span class="c"> * function takes). *)</span> <span class="k">type</span> <span class="n">proto</span> <span class="o">=</span> <span class="nc">Prototype</span> <span class="k">of</span> <span class="kt">string</span> <span class="o">*</span> <span class="kt">string</span> <span class="kt">array</span> <span class="c">(* func - This type represents a function definition itself. *)</span> <span class="k">type</span> <span class="n">func</span> <span class="o">=</span> <span class="nc">Function</span> <span class="k">of</span> <span class="n">proto</span> <span class="o">*</span> <span class="n">expr</span> </pre></div> </div> </dd> <dt>parser.ml:</dt> <dd><div class="first last highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(*===---------------------------------------------------------------------===</span> <span class="c"> * Parser</span> <span class="c"> *===---------------------------------------------------------------------===*)</span> <span class="c">(* binop_precedence - This holds the precedence for each binary operator that is</span> <span class="c"> * defined *)</span> <span class="k">let</span> <span class="n">binop_precedence</span><span class="o">:(</span><span class="kt">char</span><span class="o">,</span> <span class="kt">int</span><span class="o">)</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">t</span> <span class="o">=</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">create</span> <span class="mi">10</span> <span class="c">(* precedence - Get the precedence of the pending binary operator token. *)</span> <span class="k">let</span> <span class="n">precedence</span> <span class="n">c</span> <span class="o">=</span> <span class="k">try</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">find</span> <span class="n">binop_precedence</span> <span class="n">c</span> <span class="k">with</span> <span class="nc">Not_found</span> <span class="o">-></span> <span class="o">-</span><span class="mi">1</span> <span class="c">(* primary</span> <span class="c"> * ::= identifier</span> <span class="c"> * ::= numberexpr</span> <span class="c"> * ::= parenexpr</span> <span class="c"> * ::= ifexpr</span> <span class="c"> * ::= forexpr *)</span> <span class="k">let</span> <span class="k">rec</span> <span class="n">parse_primary</span> <span class="o">=</span> <span class="n">parser</span> <span class="c">(* numberexpr ::= number *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Number</span> <span class="n">n</span> <span class="o">>]</span> <span class="o">-></span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Number</span> <span class="n">n</span> <span class="c">(* parenexpr ::= '(' expression ')' *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">'('</span><span class="o">;</span> <span class="n">e</span><span class="o">=</span><span class="n">parse_expr</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">')'</span> <span class="o">??</span> <span class="s2">"expected ')'"</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">e</span> <span class="c">(* identifierexpr</span> <span class="c"> * ::= identifier</span> <span class="c"> * ::= identifier '(' argumentexpr ')' *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Ident</span> <span class="n">id</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">let</span> <span class="k">rec</span> <span class="n">parse_args</span> <span class="n">accumulator</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="n">e</span><span class="o">=</span><span class="n">parse_expr</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">begin</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">','</span><span class="o">;</span> <span class="n">e</span><span class="o">=</span><span class="n">parse_args</span> <span class="o">(</span><span class="n">e</span> <span class="o">::</span> <span class="n">accumulator</span><span class="o">)</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">e</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">e</span> <span class="o">::</span> <span class="n">accumulator</span> <span class="k">end</span> <span class="n">stream</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">accumulator</span> <span class="k">in</span> <span class="k">let</span> <span class="k">rec</span> <span class="n">parse_ident</span> <span class="n">id</span> <span class="o">=</span> <span class="n">parser</span> <span class="c">(* Call. *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">'('</span><span class="o">;</span> <span class="n">args</span><span class="o">=</span><span class="n">parse_args</span> <span class="bp">[]</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">')'</span> <span class="o">??</span> <span class="s2">"expected ')'"</span><span class="o">>]</span> <span class="o">-></span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Call</span> <span class="o">(</span><span class="n">id</span><span class="o">,</span> <span class="nn">Array</span><span class="p">.</span><span class="n">of_list</span> <span class="o">(</span><span class="nn">List</span><span class="p">.</span><span class="n">rev</span> <span class="n">args</span><span class="o">))</span> <span class="c">(* Simple variable ref. *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Variable</span> <span class="n">id</span> <span class="k">in</span> <span class="n">parse_ident</span> <span class="n">id</span> <span class="n">stream</span> <span class="c">(* ifexpr ::= 'if' expr 'then' expr 'else' expr *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">If</span><span class="o">;</span> <span class="n">c</span><span class="o">=</span><span class="n">parse_expr</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Then</span> <span class="o">??</span> <span class="s2">"expected 'then'"</span><span class="o">;</span> <span class="n">t</span><span class="o">=</span><span class="n">parse_expr</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Else</span> <span class="o">??</span> <span class="s2">"expected 'else'"</span><span class="o">;</span> <span class="n">e</span><span class="o">=</span><span class="n">parse_expr</span> <span class="o">>]</span> <span class="o">-></span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">If</span> <span class="o">(</span><span class="n">c</span><span class="o">,</span> <span class="n">t</span><span class="o">,</span> <span class="n">e</span><span class="o">)</span> <span class="c">(* forexpr</span> <span class="c"> ::= 'for' identifier '=' expr ',' expr (',' expr)? 'in' expression *)</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">For</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Ident</span> <span class="n">id</span> <span class="o">??</span> <span class="s2">"expected identifier after for"</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">'='</span> <span class="o">??</span> <span class="s2">"expected '=' after for"</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">begin</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="n">start</span><span class="o">=</span><span class="n">parse_expr</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">','</span> <span class="o">??</span> <span class="s2">"expected ',' after for"</span><span class="o">;</span> <span class="n">end_</span><span class="o">=</span><span class="n">parse_expr</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">let</span> <span class="n">step</span> <span class="o">=</span> <span class="k">begin</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">','</span><span class="o">;</span> <span class="n">step</span><span class="o">=</span><span class="n">parse_expr</span> <span class="o">>]</span> <span class="o">-></span> <span class="nc">Some</span> <span class="n">step</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="nc">None</span> <span class="k">end</span> <span class="n">stream</span> <span class="k">in</span> <span class="k">begin</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">In</span><span class="o">;</span> <span class="n">body</span><span class="o">=</span><span class="n">parse_expr</span> <span class="o">>]</span> <span class="o">-></span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">For</span> <span class="o">(</span><span class="n">id</span><span class="o">,</span> <span class="n">start</span><span class="o">,</span> <span class="n">end_</span><span class="o">,</span> <span class="n">step</span><span class="o">,</span> <span class="n">body</span><span class="o">)</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">raise</span> <span class="o">(</span><span class="nn">Stream</span><span class="p">.</span><span class="nc">Error</span> <span class="s2">"expected 'in' after for"</span><span class="o">)</span> <span class="k">end</span> <span class="n">stream</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">raise</span> <span class="o">(</span><span class="nn">Stream</span><span class="p">.</span><span class="nc">Error</span> <span class="s2">"expected '=' after for"</span><span class="o">)</span> <span class="k">end</span> <span class="n">stream</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">raise</span> <span class="o">(</span><span class="nn">Stream</span><span class="p">.</span><span class="nc">Error</span> <span class="s2">"unknown token when expecting an expression."</span><span class="o">)</span> <span class="c">(* binoprhs</span> <span class="c"> * ::= ('+' primary)* *)</span> <span class="ow">and</span> <span class="n">parse_bin_rhs</span> <span class="n">expr_prec</span> <span class="n">lhs</span> <span class="n">stream</span> <span class="o">=</span> <span class="k">match</span> <span class="nn">Stream</span><span class="p">.</span><span class="n">peek</span> <span class="n">stream</span> <span class="k">with</span> <span class="c">(* If this is a binop, find its precedence. *)</span> <span class="o">|</span> <span class="nc">Some</span> <span class="o">(</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="n">c</span><span class="o">)</span> <span class="k">when</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">mem</span> <span class="n">binop_precedence</span> <span class="n">c</span> <span class="o">-></span> <span class="k">let</span> <span class="n">token_prec</span> <span class="o">=</span> <span class="n">precedence</span> <span class="n">c</span> <span class="k">in</span> <span class="c">(* If this is a binop that binds at least as tightly as the current binop,</span> <span class="c"> * consume it, otherwise we are done. *)</span> <span class="k">if</span> <span class="n">token_prec</span> <span class="o"><</span> <span class="n">expr_prec</span> <span class="k">then</span> <span class="n">lhs</span> <span class="k">else</span> <span class="k">begin</span> <span class="c">(* Eat the binop. *)</span> <span class="nn">Stream</span><span class="p">.</span><span class="n">junk</span> <span class="n">stream</span><span class="o">;</span> <span class="c">(* Parse the primary expression after the binary operator. *)</span> <span class="k">let</span> <span class="n">rhs</span> <span class="o">=</span> <span class="n">parse_primary</span> <span class="n">stream</span> <span class="k">in</span> <span class="c">(* Okay, we know this is a binop. *)</span> <span class="k">let</span> <span class="n">rhs</span> <span class="o">=</span> <span class="k">match</span> <span class="nn">Stream</span><span class="p">.</span><span class="n">peek</span> <span class="n">stream</span> <span class="k">with</span> <span class="o">|</span> <span class="nc">Some</span> <span class="o">(</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="n">c2</span><span class="o">)</span> <span class="o">-></span> <span class="c">(* If BinOp binds less tightly with rhs than the operator after</span> <span class="c"> * rhs, let the pending operator take rhs as its lhs. *)</span> <span class="k">let</span> <span class="n">next_prec</span> <span class="o">=</span> <span class="n">precedence</span> <span class="n">c2</span> <span class="k">in</span> <span class="k">if</span> <span class="n">token_prec</span> <span class="o"><</span> <span class="n">next_prec</span> <span class="k">then</span> <span class="n">parse_bin_rhs</span> <span class="o">(</span><span class="n">token_prec</span> <span class="o">+</span> <span class="mi">1</span><span class="o">)</span> <span class="n">rhs</span> <span class="n">stream</span> <span class="k">else</span> <span class="n">rhs</span> <span class="o">|</span> <span class="o">_</span> <span class="o">-></span> <span class="n">rhs</span> <span class="k">in</span> <span class="c">(* Merge lhs/rhs. *)</span> <span class="k">let</span> <span class="n">lhs</span> <span class="o">=</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Binary</span> <span class="o">(</span><span class="n">c</span><span class="o">,</span> <span class="n">lhs</span><span class="o">,</span> <span class="n">rhs</span><span class="o">)</span> <span class="k">in</span> <span class="n">parse_bin_rhs</span> <span class="n">expr_prec</span> <span class="n">lhs</span> <span class="n">stream</span> <span class="k">end</span> <span class="o">|</span> <span class="o">_</span> <span class="o">-></span> <span class="n">lhs</span> <span class="c">(* expression</span> <span class="c"> * ::= primary binoprhs *)</span> <span class="ow">and</span> <span class="n">parse_expr</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="n">lhs</span><span class="o">=</span><span class="n">parse_primary</span><span class="o">;</span> <span class="n">stream</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">parse_bin_rhs</span> <span class="mi">0</span> <span class="n">lhs</span> <span class="n">stream</span> <span class="c">(* prototype</span> <span class="c"> * ::= id '(' id* ')' *)</span> <span class="k">let</span> <span class="n">parse_prototype</span> <span class="o">=</span> <span class="k">let</span> <span class="k">rec</span> <span class="n">parse_args</span> <span class="n">accumulator</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Ident</span> <span class="n">id</span><span class="o">;</span> <span class="n">e</span><span class="o">=</span><span class="n">parse_args</span> <span class="o">(</span><span class="n">id</span><span class="o">::</span><span class="n">accumulator</span><span class="o">)</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">e</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">accumulator</span> <span class="k">in</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Ident</span> <span class="n">id</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">'('</span> <span class="o">??</span> <span class="s2">"expected '(' in prototype"</span><span class="o">;</span> <span class="n">args</span><span class="o">=</span><span class="n">parse_args</span> <span class="bp">[]</span><span class="o">;</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">')'</span> <span class="o">??</span> <span class="s2">"expected ')' in prototype"</span> <span class="o">>]</span> <span class="o">-></span> <span class="c">(* success. *)</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Prototype</span> <span class="o">(</span><span class="n">id</span><span class="o">,</span> <span class="nn">Array</span><span class="p">.</span><span class="n">of_list</span> <span class="o">(</span><span class="nn">List</span><span class="p">.</span><span class="n">rev</span> <span class="n">args</span><span class="o">))</span> <span class="o">|</span> <span class="o">[<</span> <span class="o">>]</span> <span class="o">-></span> <span class="k">raise</span> <span class="o">(</span><span class="nn">Stream</span><span class="p">.</span><span class="nc">Error</span> <span class="s2">"expected function name in prototype"</span><span class="o">)</span> <span class="c">(* definition ::= 'def' prototype expression *)</span> <span class="k">let</span> <span class="n">parse_definition</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Def</span><span class="o">;</span> <span class="n">p</span><span class="o">=</span><span class="n">parse_prototype</span><span class="o">;</span> <span class="n">e</span><span class="o">=</span><span class="n">parse_expr</span> <span class="o">>]</span> <span class="o">-></span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Function</span> <span class="o">(</span><span class="n">p</span><span class="o">,</span> <span class="n">e</span><span class="o">)</span> <span class="c">(* toplevelexpr ::= expression *)</span> <span class="k">let</span> <span class="n">parse_toplevel</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="n">e</span><span class="o">=</span><span class="n">parse_expr</span> <span class="o">>]</span> <span class="o">-></span> <span class="c">(* Make an anonymous proto. *)</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Function</span> <span class="o">(</span><span class="nn">Ast</span><span class="p">.</span><span class="nc">Prototype</span> <span class="o">(</span><span class="s2">""</span><span class="o">,</span> <span class="o">[||]),</span> <span class="n">e</span><span class="o">)</span> <span class="c">(* external ::= 'extern' prototype *)</span> <span class="k">let</span> <span class="n">parse_extern</span> <span class="o">=</span> <span class="n">parser</span> <span class="o">|</span> <span class="o">[<</span> <span class="k">'</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Extern</span><span class="o">;</span> <span class="n">e</span><span class="o">=</span><span class="n">parse_prototype</span> <span class="o">>]</span> <span class="o">-></span> <span class="n">e</span> </pre></div> </div> </dd> <dt>codegen.ml:</dt> <dd><div class="first last highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(*===----------------------------------------------------------------------===</span> <span class="c"> * Code Generation</span> <span class="c"> *===----------------------------------------------------------------------===*)</span> <span class="k">open</span> <span class="nc">Llvm</span> <span class="k">exception</span> <span class="nc">Error</span> <span class="k">of</span> <span class="kt">string</span> <span class="k">let</span> <span class="n">context</span> <span class="o">=</span> <span class="n">global_context</span> <span class="bp">()</span> <span class="k">let</span> <span class="n">the_module</span> <span class="o">=</span> <span class="n">create_module</span> <span class="n">context</span> <span class="s2">"my cool jit"</span> <span class="k">let</span> <span class="n">builder</span> <span class="o">=</span> <span class="n">builder</span> <span class="n">context</span> <span class="k">let</span> <span class="n">named_values</span><span class="o">:(</span><span class="kt">string</span><span class="o">,</span> <span class="n">llvalue</span><span class="o">)</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">t</span> <span class="o">=</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">create</span> <span class="mi">10</span> <span class="k">let</span> <span class="n">double_type</span> <span class="o">=</span> <span class="n">double_type</span> <span class="n">context</span> <span class="k">let</span> <span class="k">rec</span> <span class="n">codegen_expr</span> <span class="o">=</span> <span class="k">function</span> <span class="o">|</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Number</span> <span class="n">n</span> <span class="o">-></span> <span class="n">const_float</span> <span class="n">double_type</span> <span class="n">n</span> <span class="o">|</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Variable</span> <span class="n">name</span> <span class="o">-></span> <span class="o">(</span><span class="k">try</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">find</span> <span class="n">named_values</span> <span class="n">name</span> <span class="k">with</span> <span class="o">|</span> <span class="nc">Not_found</span> <span class="o">-></span> <span class="k">raise</span> <span class="o">(</span><span class="nc">Error</span> <span class="s2">"unknown variable name"</span><span class="o">))</span> <span class="o">|</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Binary</span> <span class="o">(</span><span class="n">op</span><span class="o">,</span> <span class="n">lhs</span><span class="o">,</span> <span class="n">rhs</span><span class="o">)</span> <span class="o">-></span> <span class="k">let</span> <span class="n">lhs_val</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">lhs</span> <span class="k">in</span> <span class="k">let</span> <span class="n">rhs_val</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">rhs</span> <span class="k">in</span> <span class="k">begin</span> <span class="k">match</span> <span class="n">op</span> <span class="k">with</span> <span class="o">|</span> <span class="sc">'+'</span> <span class="o">-></span> <span class="n">build_add</span> <span class="n">lhs_val</span> <span class="n">rhs_val</span> <span class="s2">"addtmp"</span> <span class="n">builder</span> <span class="o">|</span> <span class="sc">'-'</span> <span class="o">-></span> <span class="n">build_sub</span> <span class="n">lhs_val</span> <span class="n">rhs_val</span> <span class="s2">"subtmp"</span> <span class="n">builder</span> <span class="o">|</span> <span class="sc">'*'</span> <span class="o">-></span> <span class="n">build_mul</span> <span class="n">lhs_val</span> <span class="n">rhs_val</span> <span class="s2">"multmp"</span> <span class="n">builder</span> <span class="o">|</span> <span class="sc">'<'</span> <span class="o">-></span> <span class="c">(* Convert bool 0/1 to double 0.0 or 1.0 *)</span> <span class="k">let</span> <span class="n">i</span> <span class="o">=</span> <span class="n">build_fcmp</span> <span class="nn">Fcmp</span><span class="p">.</span><span class="nc">Ult</span> <span class="n">lhs_val</span> <span class="n">rhs_val</span> <span class="s2">"cmptmp"</span> <span class="n">builder</span> <span class="k">in</span> <span class="n">build_uitofp</span> <span class="n">i</span> <span class="n">double_type</span> <span class="s2">"booltmp"</span> <span class="n">builder</span> <span class="o">|</span> <span class="o">_</span> <span class="o">-></span> <span class="k">raise</span> <span class="o">(</span><span class="nc">Error</span> <span class="s2">"invalid binary operator"</span><span class="o">)</span> <span class="k">end</span> <span class="o">|</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Call</span> <span class="o">(</span><span class="n">callee</span><span class="o">,</span> <span class="n">args</span><span class="o">)</span> <span class="o">-></span> <span class="c">(* Look up the name in the module table. *)</span> <span class="k">let</span> <span class="n">callee</span> <span class="o">=</span> <span class="k">match</span> <span class="n">lookup_function</span> <span class="n">callee</span> <span class="n">the_module</span> <span class="k">with</span> <span class="o">|</span> <span class="nc">Some</span> <span class="n">callee</span> <span class="o">-></span> <span class="n">callee</span> <span class="o">|</span> <span class="nc">None</span> <span class="o">-></span> <span class="k">raise</span> <span class="o">(</span><span class="nc">Error</span> <span class="s2">"unknown function referenced"</span><span class="o">)</span> <span class="k">in</span> <span class="k">let</span> <span class="n">params</span> <span class="o">=</span> <span class="n">params</span> <span class="n">callee</span> <span class="k">in</span> <span class="c">(* If argument mismatch error. *)</span> <span class="k">if</span> <span class="nn">Array</span><span class="p">.</span><span class="n">length</span> <span class="n">params</span> <span class="o">==</span> <span class="nn">Array</span><span class="p">.</span><span class="n">length</span> <span class="n">args</span> <span class="k">then</span> <span class="bp">()</span> <span class="k">else</span> <span class="k">raise</span> <span class="o">(</span><span class="nc">Error</span> <span class="s2">"incorrect # arguments passed"</span><span class="o">);</span> <span class="k">let</span> <span class="n">args</span> <span class="o">=</span> <span class="nn">Array</span><span class="p">.</span><span class="n">map</span> <span class="n">codegen_expr</span> <span class="n">args</span> <span class="k">in</span> <span class="n">build_call</span> <span class="n">callee</span> <span class="n">args</span> <span class="s2">"calltmp"</span> <span class="n">builder</span> <span class="o">|</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">If</span> <span class="o">(</span><span class="n">cond</span><span class="o">,</span> <span class="n">then_</span><span class="o">,</span> <span class="n">else_</span><span class="o">)</span> <span class="o">-></span> <span class="k">let</span> <span class="n">cond</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">cond</span> <span class="k">in</span> <span class="c">(* Convert condition to a bool by comparing equal to 0.0 *)</span> <span class="k">let</span> <span class="n">zero</span> <span class="o">=</span> <span class="n">const_float</span> <span class="n">double_type</span> <span class="mi">0</span><span class="o">.</span><span class="mi">0</span> <span class="k">in</span> <span class="k">let</span> <span class="n">cond_val</span> <span class="o">=</span> <span class="n">build_fcmp</span> <span class="nn">Fcmp</span><span class="p">.</span><span class="nc">One</span> <span class="n">cond</span> <span class="n">zero</span> <span class="s2">"ifcond"</span> <span class="n">builder</span> <span class="k">in</span> <span class="c">(* Grab the first block so that we might later add the conditional branch</span> <span class="c"> * to it at the end of the function. *)</span> <span class="k">let</span> <span class="n">start_bb</span> <span class="o">=</span> <span class="n">insertion_block</span> <span class="n">builder</span> <span class="k">in</span> <span class="k">let</span> <span class="n">the_function</span> <span class="o">=</span> <span class="n">block_parent</span> <span class="n">start_bb</span> <span class="k">in</span> <span class="k">let</span> <span class="n">then_bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"then"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="c">(* Emit 'then' value. *)</span> <span class="n">position_at_end</span> <span class="n">then_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="k">let</span> <span class="n">then_val</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">then_</span> <span class="k">in</span> <span class="c">(* Codegen of 'then' can change the current block, update then_bb for the</span> <span class="c"> * phi. We create a new name because one is used for the phi node, and the</span> <span class="c"> * other is used for the conditional branch. *)</span> <span class="k">let</span> <span class="n">new_then_bb</span> <span class="o">=</span> <span class="n">insertion_block</span> <span class="n">builder</span> <span class="k">in</span> <span class="c">(* Emit 'else' value. *)</span> <span class="k">let</span> <span class="n">else_bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"else"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="n">position_at_end</span> <span class="n">else_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="k">let</span> <span class="n">else_val</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">else_</span> <span class="k">in</span> <span class="c">(* Codegen of 'else' can change the current block, update else_bb for the</span> <span class="c"> * phi. *)</span> <span class="k">let</span> <span class="n">new_else_bb</span> <span class="o">=</span> <span class="n">insertion_block</span> <span class="n">builder</span> <span class="k">in</span> <span class="c">(* Emit merge block. *)</span> <span class="k">let</span> <span class="n">merge_bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"ifcont"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="n">position_at_end</span> <span class="n">merge_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="k">let</span> <span class="n">incoming</span> <span class="o">=</span> <span class="o">[(</span><span class="n">then_val</span><span class="o">,</span> <span class="n">new_then_bb</span><span class="o">);</span> <span class="o">(</span><span class="n">else_val</span><span class="o">,</span> <span class="n">new_else_bb</span><span class="o">)]</span> <span class="k">in</span> <span class="k">let</span> <span class="n">phi</span> <span class="o">=</span> <span class="n">build_phi</span> <span class="n">incoming</span> <span class="s2">"iftmp"</span> <span class="n">builder</span> <span class="k">in</span> <span class="c">(* Return to the start block to add the conditional branch. *)</span> <span class="n">position_at_end</span> <span class="n">start_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">build_cond_br</span> <span class="n">cond_val</span> <span class="n">then_bb</span> <span class="n">else_bb</span> <span class="n">builder</span><span class="o">);</span> <span class="c">(* Set a unconditional branch at the end of the 'then' block and the</span> <span class="c"> * 'else' block to the 'merge' block. *)</span> <span class="n">position_at_end</span> <span class="n">new_then_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">build_br</span> <span class="n">merge_bb</span> <span class="n">builder</span><span class="o">);</span> <span class="n">position_at_end</span> <span class="n">new_else_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">build_br</span> <span class="n">merge_bb</span> <span class="n">builder</span><span class="o">);</span> <span class="c">(* Finally, set the builder to the end of the merge block. *)</span> <span class="n">position_at_end</span> <span class="n">merge_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="n">phi</span> <span class="o">|</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">For</span> <span class="o">(</span><span class="n">var_name</span><span class="o">,</span> <span class="n">start</span><span class="o">,</span> <span class="n">end_</span><span class="o">,</span> <span class="n">step</span><span class="o">,</span> <span class="n">body</span><span class="o">)</span> <span class="o">-></span> <span class="c">(* Emit the start code first, without 'variable' in scope. *)</span> <span class="k">let</span> <span class="n">start_val</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">start</span> <span class="k">in</span> <span class="c">(* Make the new basic block for the loop header, inserting after current</span> <span class="c"> * block. *)</span> <span class="k">let</span> <span class="n">preheader_bb</span> <span class="o">=</span> <span class="n">insertion_block</span> <span class="n">builder</span> <span class="k">in</span> <span class="k">let</span> <span class="n">the_function</span> <span class="o">=</span> <span class="n">block_parent</span> <span class="n">preheader_bb</span> <span class="k">in</span> <span class="k">let</span> <span class="n">loop_bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"loop"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="c">(* Insert an explicit fall through from the current block to the</span> <span class="c"> * loop_bb. *)</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">build_br</span> <span class="n">loop_bb</span> <span class="n">builder</span><span class="o">);</span> <span class="c">(* Start insertion in loop_bb. *)</span> <span class="n">position_at_end</span> <span class="n">loop_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="c">(* Start the PHI node with an entry for start. *)</span> <span class="k">let</span> <span class="n">variable</span> <span class="o">=</span> <span class="n">build_phi</span> <span class="o">[(</span><span class="n">start_val</span><span class="o">,</span> <span class="n">preheader_bb</span><span class="o">)]</span> <span class="n">var_name</span> <span class="n">builder</span> <span class="k">in</span> <span class="c">(* Within the loop, the variable is defined equal to the PHI node. If it</span> <span class="c"> * shadows an existing variable, we have to restore it, so save it</span> <span class="c"> * now. *)</span> <span class="k">let</span> <span class="n">old_val</span> <span class="o">=</span> <span class="k">try</span> <span class="nc">Some</span> <span class="o">(</span><span class="nn">Hashtbl</span><span class="p">.</span><span class="n">find</span> <span class="n">named_values</span> <span class="n">var_name</span><span class="o">)</span> <span class="k">with</span> <span class="nc">Not_found</span> <span class="o">-></span> <span class="nc">None</span> <span class="k">in</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">add</span> <span class="n">named_values</span> <span class="n">var_name</span> <span class="n">variable</span><span class="o">;</span> <span class="c">(* Emit the body of the loop. This, like any other expr, can change the</span> <span class="c"> * current BB. Note that we ignore the value computed by the body, but</span> <span class="c"> * don't allow an error *)</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">codegen_expr</span> <span class="n">body</span><span class="o">);</span> <span class="c">(* Emit the step value. *)</span> <span class="k">let</span> <span class="n">step_val</span> <span class="o">=</span> <span class="k">match</span> <span class="n">step</span> <span class="k">with</span> <span class="o">|</span> <span class="nc">Some</span> <span class="n">step</span> <span class="o">-></span> <span class="n">codegen_expr</span> <span class="n">step</span> <span class="c">(* If not specified, use 1.0. *)</span> <span class="o">|</span> <span class="nc">None</span> <span class="o">-></span> <span class="n">const_float</span> <span class="n">double_type</span> <span class="mi">1</span><span class="o">.</span><span class="mi">0</span> <span class="k">in</span> <span class="k">let</span> <span class="n">next_var</span> <span class="o">=</span> <span class="n">build_add</span> <span class="n">variable</span> <span class="n">step_val</span> <span class="s2">"nextvar"</span> <span class="n">builder</span> <span class="k">in</span> <span class="c">(* Compute the end condition. *)</span> <span class="k">let</span> <span class="n">end_cond</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">end_</span> <span class="k">in</span> <span class="c">(* Convert condition to a bool by comparing equal to 0.0. *)</span> <span class="k">let</span> <span class="n">zero</span> <span class="o">=</span> <span class="n">const_float</span> <span class="n">double_type</span> <span class="mi">0</span><span class="o">.</span><span class="mi">0</span> <span class="k">in</span> <span class="k">let</span> <span class="n">end_cond</span> <span class="o">=</span> <span class="n">build_fcmp</span> <span class="nn">Fcmp</span><span class="p">.</span><span class="nc">One</span> <span class="n">end_cond</span> <span class="n">zero</span> <span class="s2">"loopcond"</span> <span class="n">builder</span> <span class="k">in</span> <span class="c">(* Create the "after loop" block and insert it. *)</span> <span class="k">let</span> <span class="n">loop_end_bb</span> <span class="o">=</span> <span class="n">insertion_block</span> <span class="n">builder</span> <span class="k">in</span> <span class="k">let</span> <span class="n">after_bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"afterloop"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="c">(* Insert the conditional branch into the end of loop_end_bb. *)</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">build_cond_br</span> <span class="n">end_cond</span> <span class="n">loop_bb</span> <span class="n">after_bb</span> <span class="n">builder</span><span class="o">);</span> <span class="c">(* Any new code will be inserted in after_bb. *)</span> <span class="n">position_at_end</span> <span class="n">after_bb</span> <span class="n">builder</span><span class="o">;</span> <span class="c">(* Add a new entry to the PHI node for the backedge. *)</span> <span class="n">add_incoming</span> <span class="o">(</span><span class="n">next_var</span><span class="o">,</span> <span class="n">loop_end_bb</span><span class="o">)</span> <span class="n">variable</span><span class="o">;</span> <span class="c">(* Restore the unshadowed variable. *)</span> <span class="k">begin</span> <span class="k">match</span> <span class="n">old_val</span> <span class="k">with</span> <span class="o">|</span> <span class="nc">Some</span> <span class="n">old_val</span> <span class="o">-></span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">add</span> <span class="n">named_values</span> <span class="n">var_name</span> <span class="n">old_val</span> <span class="o">|</span> <span class="nc">None</span> <span class="o">-></span> <span class="bp">()</span> <span class="k">end</span><span class="o">;</span> <span class="c">(* for expr always returns 0.0. *)</span> <span class="n">const_null</span> <span class="n">double_type</span> <span class="k">let</span> <span class="n">codegen_proto</span> <span class="o">=</span> <span class="k">function</span> <span class="o">|</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Prototype</span> <span class="o">(</span><span class="n">name</span><span class="o">,</span> <span class="n">args</span><span class="o">)</span> <span class="o">-></span> <span class="c">(* Make the function type: double(double,double) etc. *)</span> <span class="k">let</span> <span class="n">doubles</span> <span class="o">=</span> <span class="nn">Array</span><span class="p">.</span><span class="n">make</span> <span class="o">(</span><span class="nn">Array</span><span class="p">.</span><span class="n">length</span> <span class="n">args</span><span class="o">)</span> <span class="n">double_type</span> <span class="k">in</span> <span class="k">let</span> <span class="n">ft</span> <span class="o">=</span> <span class="n">function_type</span> <span class="n">double_type</span> <span class="n">doubles</span> <span class="k">in</span> <span class="k">let</span> <span class="n">f</span> <span class="o">=</span> <span class="k">match</span> <span class="n">lookup_function</span> <span class="n">name</span> <span class="n">the_module</span> <span class="k">with</span> <span class="o">|</span> <span class="nc">None</span> <span class="o">-></span> <span class="n">declare_function</span> <span class="n">name</span> <span class="n">ft</span> <span class="n">the_module</span> <span class="c">(* If 'f' conflicted, there was already something named 'name'. If it</span> <span class="c"> * has a body, don't allow redefinition or reextern. *)</span> <span class="o">|</span> <span class="nc">Some</span> <span class="n">f</span> <span class="o">-></span> <span class="c">(* If 'f' already has a body, reject this. *)</span> <span class="k">if</span> <span class="n">block_begin</span> <span class="n">f</span> <span class="o"><></span> <span class="nc">At_end</span> <span class="n">f</span> <span class="k">then</span> <span class="k">raise</span> <span class="o">(</span><span class="nc">Error</span> <span class="s2">"redefinition of function"</span><span class="o">);</span> <span class="c">(* If 'f' took a different number of arguments, reject. *)</span> <span class="k">if</span> <span class="n">element_type</span> <span class="o">(</span><span class="n">type_of</span> <span class="n">f</span><span class="o">)</span> <span class="o"><></span> <span class="n">ft</span> <span class="k">then</span> <span class="k">raise</span> <span class="o">(</span><span class="nc">Error</span> <span class="s2">"redefinition of function with different # args"</span><span class="o">);</span> <span class="n">f</span> <span class="k">in</span> <span class="c">(* Set names for all arguments. *)</span> <span class="nn">Array</span><span class="p">.</span><span class="n">iteri</span> <span class="o">(</span><span class="k">fun</span> <span class="n">i</span> <span class="n">a</span> <span class="o">-></span> <span class="k">let</span> <span class="n">n</span> <span class="o">=</span> <span class="n">args</span><span class="o">.(</span><span class="n">i</span><span class="o">)</span> <span class="k">in</span> <span class="n">set_value_name</span> <span class="n">n</span> <span class="n">a</span><span class="o">;</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">add</span> <span class="n">named_values</span> <span class="n">n</span> <span class="n">a</span><span class="o">;</span> <span class="o">)</span> <span class="o">(</span><span class="n">params</span> <span class="n">f</span><span class="o">);</span> <span class="n">f</span> <span class="k">let</span> <span class="n">codegen_func</span> <span class="n">the_fpm</span> <span class="o">=</span> <span class="k">function</span> <span class="o">|</span> <span class="nn">Ast</span><span class="p">.</span><span class="nc">Function</span> <span class="o">(</span><span class="n">proto</span><span class="o">,</span> <span class="n">body</span><span class="o">)</span> <span class="o">-></span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">clear</span> <span class="n">named_values</span><span class="o">;</span> <span class="k">let</span> <span class="n">the_function</span> <span class="o">=</span> <span class="n">codegen_proto</span> <span class="n">proto</span> <span class="k">in</span> <span class="c">(* Create a new basic block to start insertion into. *)</span> <span class="k">let</span> <span class="n">bb</span> <span class="o">=</span> <span class="n">append_block</span> <span class="n">context</span> <span class="s2">"entry"</span> <span class="n">the_function</span> <span class="k">in</span> <span class="n">position_at_end</span> <span class="n">bb</span> <span class="n">builder</span><span class="o">;</span> <span class="k">try</span> <span class="k">let</span> <span class="n">ret_val</span> <span class="o">=</span> <span class="n">codegen_expr</span> <span class="n">body</span> <span class="k">in</span> <span class="c">(* Finish off the function. *)</span> <span class="k">let</span> <span class="o">_</span> <span class="o">=</span> <span class="n">build_ret</span> <span class="n">ret_val</span> <span class="n">builder</span> <span class="k">in</span> <span class="c">(* Validate the generated code, checking for consistency. *)</span> <span class="nn">Llvm_analysis</span><span class="p">.</span><span class="n">assert_valid_function</span> <span class="n">the_function</span><span class="o">;</span> <span class="c">(* Optimize the function. *)</span> <span class="k">let</span> <span class="o">_</span> <span class="o">=</span> <span class="nn">PassManager</span><span class="p">.</span><span class="n">run_function</span> <span class="n">the_function</span> <span class="n">the_fpm</span> <span class="k">in</span> <span class="n">the_function</span> <span class="k">with</span> <span class="n">e</span> <span class="o">-></span> <span class="n">delete_function</span> <span class="n">the_function</span><span class="o">;</span> <span class="k">raise</span> <span class="n">e</span> </pre></div> </div> </dd> <dt>toplevel.ml:</dt> <dd><div class="first last highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(*===----------------------------------------------------------------------===</span> <span class="c"> * Top-Level parsing and JIT Driver</span> <span class="c"> *===----------------------------------------------------------------------===*)</span> <span class="k">open</span> <span class="nc">Llvm</span> <span class="k">open</span> <span class="nc">Llvm_executionengine</span> <span class="c">(* top ::= definition | external | expression | ';' *)</span> <span class="k">let</span> <span class="k">rec</span> <span class="n">main_loop</span> <span class="n">the_fpm</span> <span class="n">the_execution_engine</span> <span class="n">stream</span> <span class="o">=</span> <span class="k">match</span> <span class="nn">Stream</span><span class="p">.</span><span class="n">peek</span> <span class="n">stream</span> <span class="k">with</span> <span class="o">|</span> <span class="nc">None</span> <span class="o">-></span> <span class="bp">()</span> <span class="c">(* ignore top-level semicolons. *)</span> <span class="o">|</span> <span class="nc">Some</span> <span class="o">(</span><span class="nn">Token</span><span class="p">.</span><span class="nc">Kwd</span> <span class="sc">';'</span><span class="o">)</span> <span class="o">-></span> <span class="nn">Stream</span><span class="p">.</span><span class="n">junk</span> <span class="n">stream</span><span class="o">;</span> <span class="n">main_loop</span> <span class="n">the_fpm</span> <span class="n">the_execution_engine</span> <span class="n">stream</span> <span class="o">|</span> <span class="nc">Some</span> <span class="n">token</span> <span class="o">-></span> <span class="k">begin</span> <span class="k">try</span> <span class="k">match</span> <span class="n">token</span> <span class="k">with</span> <span class="o">|</span> <span class="nn">Token</span><span class="p">.</span><span class="nc">Def</span> <span class="o">-></span> <span class="k">let</span> <span class="n">e</span> <span class="o">=</span> <span class="nn">Parser</span><span class="p">.</span><span class="n">parse_definition</span> <span class="n">stream</span> <span class="k">in</span> <span class="n">print_endline</span> <span class="s2">"parsed a function definition."</span><span class="o">;</span> <span class="n">dump_value</span> <span class="o">(</span><span class="nn">Codegen</span><span class="p">.</span><span class="n">codegen_func</span> <span class="n">the_fpm</span> <span class="n">e</span><span class="o">);</span> <span class="o">|</span> <span class="nn">Token</span><span class="p">.</span><span class="nc">Extern</span> <span class="o">-></span> <span class="k">let</span> <span class="n">e</span> <span class="o">=</span> <span class="nn">Parser</span><span class="p">.</span><span class="n">parse_extern</span> <span class="n">stream</span> <span class="k">in</span> <span class="n">print_endline</span> <span class="s2">"parsed an extern."</span><span class="o">;</span> <span class="n">dump_value</span> <span class="o">(</span><span class="nn">Codegen</span><span class="p">.</span><span class="n">codegen_proto</span> <span class="n">e</span><span class="o">);</span> <span class="o">|</span> <span class="o">_</span> <span class="o">-></span> <span class="c">(* Evaluate a top-level expression into an anonymous function. *)</span> <span class="k">let</span> <span class="n">e</span> <span class="o">=</span> <span class="nn">Parser</span><span class="p">.</span><span class="n">parse_toplevel</span> <span class="n">stream</span> <span class="k">in</span> <span class="n">print_endline</span> <span class="s2">"parsed a top-level expr"</span><span class="o">;</span> <span class="k">let</span> <span class="n">the_function</span> <span class="o">=</span> <span class="nn">Codegen</span><span class="p">.</span><span class="n">codegen_func</span> <span class="n">the_fpm</span> <span class="n">e</span> <span class="k">in</span> <span class="n">dump_value</span> <span class="n">the_function</span><span class="o">;</span> <span class="c">(* JIT the function, returning a function pointer. *)</span> <span class="k">let</span> <span class="n">result</span> <span class="o">=</span> <span class="nn">ExecutionEngine</span><span class="p">.</span><span class="n">run_function</span> <span class="n">the_function</span> <span class="o">[||]</span> <span class="n">the_execution_engine</span> <span class="k">in</span> <span class="n">print_string</span> <span class="s2">"Evaluated to "</span><span class="o">;</span> <span class="n">print_float</span> <span class="o">(</span><span class="nn">GenericValue</span><span class="p">.</span><span class="n">as_float</span> <span class="nn">Codegen</span><span class="p">.</span><span class="n">double_type</span> <span class="n">result</span><span class="o">);</span> <span class="n">print_newline</span> <span class="bp">()</span><span class="o">;</span> <span class="k">with</span> <span class="nn">Stream</span><span class="p">.</span><span class="nc">Error</span> <span class="n">s</span> <span class="o">|</span> <span class="nn">Codegen</span><span class="p">.</span><span class="nc">Error</span> <span class="n">s</span> <span class="o">-></span> <span class="c">(* Skip token for error recovery. *)</span> <span class="nn">Stream</span><span class="p">.</span><span class="n">junk</span> <span class="n">stream</span><span class="o">;</span> <span class="n">print_endline</span> <span class="n">s</span><span class="o">;</span> <span class="k">end</span><span class="o">;</span> <span class="n">print_string</span> <span class="s2">"ready> "</span><span class="o">;</span> <span class="n">flush</span> <span class="n">stdout</span><span class="o">;</span> <span class="n">main_loop</span> <span class="n">the_fpm</span> <span class="n">the_execution_engine</span> <span class="n">stream</span> </pre></div> </div> </dd> <dt>toy.ml:</dt> <dd><div class="first last highlight-ocaml notranslate"><div class="highlight"><pre><span></span><span class="c">(*===----------------------------------------------------------------------===</span> <span class="c"> * Main driver code.</span> <span class="c"> *===----------------------------------------------------------------------===*)</span> <span class="k">open</span> <span class="nc">Llvm</span> <span class="k">open</span> <span class="nc">Llvm_executionengine</span> <span class="k">open</span> <span class="nc">Llvm_target</span> <span class="k">open</span> <span class="nc">Llvm_scalar_opts</span> <span class="k">let</span> <span class="n">main</span> <span class="bp">()</span> <span class="o">=</span> <span class="n">ignore</span> <span class="o">(</span><span class="n">initialize_native_target</span> <span class="bp">()</span><span class="o">);</span> <span class="c">(* Install standard binary operators.</span> <span class="c"> * 1 is the lowest precedence. *)</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">add</span> <span class="nn">Parser</span><span class="p">.</span><span class="n">binop_precedence</span> <span class="sc">'<'</span> <span class="mi">10</span><span class="o">;</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">add</span> <span class="nn">Parser</span><span class="p">.</span><span class="n">binop_precedence</span> <span class="sc">'+'</span> <span class="mi">20</span><span class="o">;</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">add</span> <span class="nn">Parser</span><span class="p">.</span><span class="n">binop_precedence</span> <span class="sc">'-'</span> <span class="mi">20</span><span class="o">;</span> <span class="nn">Hashtbl</span><span class="p">.</span><span class="n">add</span> <span class="nn">Parser</span><span class="p">.</span><span class="n">binop_precedence</span> <span class="sc">'*'</span> <span class="mi">40</span><span class="o">;</span> <span class="c">(* highest. *)</span> <span class="c">(* Prime the first token. *)</span> <span class="n">print_string</span> <span class="s2">"ready> "</span><span class="o">;</span> <span class="n">flush</span> <span class="n">stdout</span><span class="o">;</span> <span class="k">let</span> <span class="n">stream</span> <span class="o">=</span> <span class="nn">Lexer</span><span class="p">.</span><span class="n">lex</span> <span class="o">(</span><span class="nn">Stream</span><span class="p">.</span><span class="n">of_channel</span> <span class="n">stdin</span><span class="o">)</span> <span class="k">in</span> <span class="c">(* Create the JIT. *)</span> <span class="k">let</span> <span class="n">the_execution_engine</span> <span class="o">=</span> <span class="nn">ExecutionEngine</span><span class="p">.</span><span class="n">create</span> <span class="nn">Codegen</span><span class="p">.</span><span class="n">the_module</span> <span class="k">in</span> <span class="k">let</span> <span class="n">the_fpm</span> <span class="o">=</span> <span class="nn">PassManager</span><span class="p">.</span><span class="n">create_function</span> <span class="nn">Codegen</span><span class="p">.</span><span class="n">the_module</span> <span class="k">in</span> <span class="c">(* Set up the optimizer pipeline. Start with registering info about how the</span> <span class="c"> * target lays out data structures. *)</span> <span class="nn">DataLayout</span><span class="p">.</span><span class="n">add</span> <span class="o">(</span><span class="nn">ExecutionEngine</span><span class="p">.</span><span class="n">target_data</span> <span class="n">the_execution_engine</span><span class="o">)</span> <span class="n">the_fpm</span><span class="o">;</span> <span class="c">(* Do simple "peephole" optimizations and bit-twiddling optzn. *)</span> <span class="n">add_instruction_combination</span> <span class="n">the_fpm</span><span class="o">;</span> <span class="c">(* reassociate expressions. *)</span> <span class="n">add_reassociation</span> <span class="n">the_fpm</span><span class="o">;</span> <span class="c">(* Eliminate Common SubExpressions. *)</span> <span class="n">add_gvn</span> <span class="n">the_fpm</span><span class="o">;</span> <span class="c">(* Simplify the control flow graph (deleting unreachable blocks, etc). *)</span> <span class="n">add_cfg_simplification</span> <span class="n">the_fpm</span><span class="o">;</span> <span class="n">ignore</span> <span class="o">(</span><span class="nn">PassManager</span><span class="p">.</span><span class="n">initialize</span> <span class="n">the_fpm</span><span class="o">);</span> <span class="c">(* Run the main "interpreter loop" now. *)</span> <span class="nn">Toplevel</span><span class="p">.</span><span class="n">main_loop</span> <span class="n">the_fpm</span> <span class="n">the_execution_engine</span> <span class="n">stream</span><span class="o">;</span> <span class="c">(* Print out all the generated code. *)</span> <span class="n">dump_module</span> <span class="nn">Codegen</span><span class="p">.</span><span class="n">the_module</span> <span class="o">;;</span> <span class="n">main</span> <span class="bp">()</span> </pre></div> </div> </dd> <dt>bindings.c</dt> <dd><div class="first last highlight-c notranslate"><div class="highlight"><pre><span></span><span class="cp">#include</span> <span class="cpf"><stdio.h></span><span class="cp"></span> <span class="cm">/* putchard - putchar that takes a double and returns 0. */</span> <span class="k">extern</span> <span class="kt">double</span> <span class="nf">putchard</span><span class="p">(</span><span class="kt">double</span> <span class="n">X</span><span class="p">)</span> <span class="p">{</span> <span class="n">putchar</span><span class="p">((</span><span class="kt">char</span><span class="p">)</span><span class="n">X</span><span class="p">);</span> <span class="k">return</span> <span class="mi">0</span><span class="p">;</span> <span class="p">}</span> </pre></div> </div> </dd> </dl> <p><a class="reference external" href="OCamlLangImpl6.html">Next: Extending the language: user-defined operators</a></p> </div> </div> </div> </div> <div class="clearer"></div> </div> <div class="related" role="navigation" aria-label="related navigation"> <h3>Navigation</h3> <ul> <li class="right" style="margin-right: 10px"> <a href="../genindex.html" title="General Index" >index</a></li> <li class="right" > <a href="OCamlLangImpl6.html" title="6. Kaleidoscope: Extending the Language: User-defined Operators" >next</a> |</li> <li class="right" > <a href="OCamlLangImpl4.html" title="4. Kaleidoscope: Adding JIT and Optimizer Support" >previous</a> |</li> <li><a href="http://llvm.org/">LLVM Home</a> | </li> <li><a href="../index.html">Documentation</a>»</li> <li class="nav-item nav-item-1"><a href="index.html" >LLVM Tutorial: Table of Contents</a> »</li> </ul> </div> <div class="footer" role="contentinfo"> © Copyright 2003-2020, LLVM Project. Last updated on 2020-09-07. Created using <a href="http://sphinx-doc.org/">Sphinx</a> 1.8.4. </div> </body> </html>